The Bloomsbury Companion to Bertrand Russell 9781474278058, 9781474278089, 9781474278072

Table of contents :
Title Page
Copyright Page
Contents
List of Contributors
Acknowledgments
Bertrand Russell—A Timeline
Abbreviations for Works Written by Russell
Introduction
Notes
Part 1 Russell in Context
Chapter 1 Russell and Idealism
1 Idealism: Berkeley, Kant, and Hegel
2 Three British idealists: McTaggart, Bradley, and Ward
3 Russell as idealist
4 Russell’s Moorean philosophy21
5 Russell’s post-Peano philosophy and idealism
Notes
References
Chapter 2 Russell and the Pragmatists
1 The straightforward story: Russell the anti-pragmatist
2 A spanner in the works of the straightforward story
3 Russell at Harvard: 1914
4 Russell’s pragmatism: Perception, belief, and behavior
Notes
References
Chapter 3 Russell’s Objections to Frege and Meinong in Context
1 The origins of Russell’s Paradox and his disagreements with Frege
2 What Russell learned from Frege
3 Where Russell agreed with Meinong
4 The origins of “The present King of France” objection
Notes
References
Works by Russell
Chapter 4 Russell’s Initial Encounter with Wittgenstein 1911–14
1 Introduction
2 Wittgenstein and Russell 1911–12: The matter of matter
3 1913: Theory of knowledge
4 Russell’s reaction to Wittgenstein’s criticisms
Notes
References
Chapter 5 From Russell to the Vienna Circle and Back
1 Introduction
2 Carnap, the Principia, and logicism
3 OKEW, Aufbau: Logical construction and type theory
4 Russell and the circle
Notes
References
Part 2 Philosophical Issues
Chapter 6 Russell’s Logicism
1 Introduction
2 The regressive method and its prior successes
3 The paradoxes and the development of Russell’s logic
4 Logicism in Principia Mathematica
5 Gödel’s results and the scope of Logicism
6 The controversial “axioms” and if-then-ism
Notes
References
Chapter 7 Russell on Denoting and Language
Introduction
1 Russellian propositions
2 Russell’s theory of descriptions
3 Logical form, metaphysics, and semantics
4 Going forward: Two problems with the theory of descriptions
5 Conclusion
Notes
References
Chapter 8 Russell’s Logic as the Essence of Philosophy
1 Introduction
2 Logic as the essence of philosophy: Problems
3 Logic as the essence of philosophy: Our Knowledge
4 Reconstruction in philosophy
Notes
References
Chapter 9 Sense-Data, Sensibilia, and Percepts
1 Introduction
2 Sense-data introduced
3 Logical constructions and sensibilia
4 Sense-data rejected
5 Russell’s epistemology after the rejection of sense-data
Notes
References
Chapter 10 Russell on Introspection and Self-Knowledge
1 Introduction
2 Knowing things and knowing truths
3 Knowing the mind by acquaintance
4 Inner perception and introspective judgments
5 Is there acquaintance with the self?
6 Constructing the self
7 Toward neutral monism
Notes
References
Chapter 11 Russell’s Bridge
1 Building Russell’s Bridge
2 Russell’s Bridge and the structure of epistemic justification
3 Russell’s regressive method
4 The relevance of Russell’s method
Notes
References
Chapter 12 Neutral Monism
1 Introduction
2 Three arguments against neutral monism
3 Sensation
4 Judgment
5 Images and emphatic particulars
6 Neutral monism and physics
Notes
References
Chapter 13 Bertrand Russell’s Moral Philosophy
1 A decade with the Moorean good
2 Three decades with ethical subjectivism 1913–44
3 Russell’s ethics in practice: The nuclear peril
4 Final thought: A call to duty
Notes
References
Chapter 14 Russell’s Literary Approach to History
1 Introduction
2 The scientific and the literary approaches to history
3 Russell on the literary approach
4 Trevelyan and Russell on truth
5 Beyond the literary approach
Notes
References
Bibliography
1 Books by Russell in chronological order
2 Collections of essays and stories
3 Collections of letters
4 Papers by Russell
5 Secondary Sources
Index

Citation preview

The Bloomsbury Companion to Bertrand Russell

Bloomsbury Companions The Bloomsbury Companions series is a major series of single volume companions to key research fields in the humanities aimed at postgraduate students, scholars and libraries. Each companion offers a comprehensive reference resource giving an overview of key topics, research areas, new directions and a manageable guide to beginning or developing research in the field. A distinctive feature of the series is that each companion provides practical guidance on advanced study and research in the field, including research methods and subject-specific resources.

Titles currently available in the series: Aesthetics, edited by Anna Christina Ribeiro Continental Philosophy, edited by John Mullarkey and Beth Lord Epistemology, edited by Andrew Cullison Ethics, edited by Christian Miller Existentialism, edited by Jack Reynolds, Felicity Joseph and Ashley Woodward Hobbes, edited by S.A. Lloyd Hume, edited by Alan Bailey and Dan O’Brien Kant, edited by Gary Banham, Dennis Schulting and Nigel Hems Leibniz, edited by Brandon C. Look Locke, edited by S.-J. Savonious-Wroth, Paul Schuurman and Jonathan Walmsley Metaphysics, edited by Robert W. Barnard and Neil A. Manson Philosophical Logic, edited by Leon Horston and Richard Pettigrew Philosophy of Language, edited by Manuel Garcia-Carpintero and Max Kolbel Philosophy of Mind, edited by James Garvey Philosophy of Science, edited by Steven French and Juha Saatsi Plato, edited by Gerald A. Press Pragmatism, edited by Sami Pihlström Spinoza, edited by Wiep van Bunge

The Bloomsbury Companion to Bertrand Russell Edited by Russell Wahl

BLOOMSBURY ACADEMIC Bloomsbury Publishing Plc 50 Bedford Square, London, WC1B 3DP, UK 1385 Broadway, New York, NY 10018, USA 29 Earlsfort Terrace, Dublin 2, Ireland BLOOMSBURY, BLOOMSBURY ACADEMIC and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2019 This edition published in 2022 Copyright © Russell Wahl, 2019 Russell Wahl has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as Editor of this work. For legal purposes the Acknowledgments on p. ix constitute an extension of this copyright page. Cover image © HIROKAZU EBIHARA / EyeEm / Getty Images All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party websites referred to or in this book. All internet addresses given in this book were correct at the time of going to press. The author and publisher regret any inconvenience caused if addresses have changed or sites have ceased to exist, but can accept no responsibility for any such changes. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Wahl, Russell, 1952– editor. Title: The Bloomsbury companion to Bertrand Russell / edited by Russell Wahl. Description: New York : Bloomsbury, 2018. | Includes bibliographical references and index. Identifiers: LCCN 2018003675 (print) | LCCN 2018022868 (ebook) | ISBN 9781474278065 (ePub) | ISBN 9781474278072 (ePDF) | ISBN 9781474278058 (hardback) Subjects: LCSH: Russell, Bertrand, 1872–1970. Classification: LCC B1649.R94 (ebook) | LCC B1649.R94 B59 2018 (print) | DDC 192–dc23 LC record available at https://lccn.loc.gov/2018003675 ISBN: HB: 978-1-4742-7805-8 PB: 978-1-3503-5603-0 ePDF: 978-1-4742-7807-2 eBook: 978-1-4742-7806-5 Typeset by Deanta Global Publishing Services, Chennai, India

To find out more about our authors and books visit www.bloomsbury.com and sign up for our newsletters.

Contents List of Contributors Acknowledgments Bertrand Russell—A Timeline Abbreviations for Works Written by Russell

vii ix x xv

Introduction 1 Part 1  Russell in Context 1

Russell and Idealism  James Levine 17

2

Russell and the Pragmatists  Cheryl Misak 59

3

Russell’s Objections to Frege and Meinong in Context  Bernard Linsky 75

4

Russell’s Initial Encounter with Wittgenstein 1911–14  Russell Wahl 99

5

From Russell to the Vienna Circle and Back  François Schmitz 122

Part 2  Philosophical Issues 6

Russell’s Logicism  Kevin C. Klement 151

7

Russell on Denoting and Language  Graham Stevens 179

8

Russell’s Logic as the Essence of Philosophy  Gregory Landini 205

9

Sense-Data, Sensibilia, and Percepts  Russell Wahl 237

10 Russell on Introspection and Self-Knowledge  Donovan Wishon 256 11 Russell’s Bridge  Dustin Olson and Nicholas Griffin 286 12 Neutral Monism  Christopher Pincock 312

vi

Contents

13 Bertrand Russell’s Moral Philosophy  Ray Perkins, Jr. 334 14 Russell’s Literary Approach to History  Peter Stone 361 Bibliography Index

387 417

List of Contributors Nicholas Griffin Department of Philosophy McMaster University Hamilton, ON, Canada

Dustin Olson Department of Philosophy University of Regina, Luther College, Regina, SK, Canada

Kevin C. Klement Department of Philosophy University of Massachusetts Amherst, MA, USA

Ray Perkins, Jr. Department of Philosophy Plymouth State University Plymouth, NH, USA

Gregory Landini Department of Philosophy University of Iowa Iowa City, IA, USA

Christopher Pincock Department of Philosophy Ohio State University Columbus, OH, USA

James Levine Department of Philosophy Trinity College Dublin, Ireland

François Schmitz Département Philosophie Université de Nantes Nantes, France

Bernard Linsky Department of Philosophy University of Alberta Edmonton, AL, Canada

Graham Stevens Department of Philosophy University of Manchester Manchester, UK

Cheryl Misak Department of Philosophy University of Toronto Toronto, ON, Canada

Peter Stone Department of Political Science Trinity College Dublin, Ireland

viii

List of Contributors

Russell Wahl Department of English and Philosophy Idaho State University Pocatello, ID, USA

Donovan Wishon Department of Philosophy University of Mississippi Oxford, MS, USA

Acknowledgments I would like to acknowledge especially the work of the contributors to the book. I benefited a lot from reading their work and from discussions with them. My reviewing of two of the chapters was helped very much by James Skidmore and Pamela Park, who also helped with the bibliography. Chris Pincock kindly reviewed one of my own chapters. Some of the contributors adjusted their topics, some were added late, and some dropped out of the project. All of the contributors have shown patience during this time. I regret that after some pulled out there was only one woman philosopher contributing to the book. The work on my own chapters was facilitated by McMaster University during my time as the Bertrand Russell Visiting Professor and by a sabbatical leave granted by Idaho State University during spring semester, 2013. I would also like to thank Kenneth Blackwell for the help he has provided to me and the other contributors over the years.

Bertrand Russell—A Timeline 1872 (May 18) Bertrand Russell born in Ravenscroft, the country home of his parents, near Trellech, Wales. 1874

Mother and sister die of diphtheria.

1876 Father dies. Russell and his brother move to Pembroke Lodge to live with his grandparents. 1878

Grandfather, Lord John Russell, dies.

1890  Enters Trinity College Cambridge to study for Part I of the Mathematical Tripos. 1893 Receives a first in the Mathematical Tripos. Begins one year of philosophy. 1894

Marries Alys Pearsall Smith.

1895 With Alys, attends economics lectures in Berlin; works on German social democracy. Fellowship dissertation on geometry finished. 1896 Lectures at London School of Economics. First visit to America. Lectures at Bryn Mawr and Johns Hopkins. German Social Democracy published. 1897  An Essay on the Foundations of Geometry, a revision of Russell’s fellowship thesis, published. 1899  McTaggart on leave; Russell teaches the course on Leibniz at Cambridge; and begins his turn away from idealism. Moore’s “On the Nature of Judgement” published. 1900  “The most important year of my intellectual life.” Attends the International Congress of Philosophy in Paris and becomes acquainted with Peano’s work. A Critical Exposition of the Philosophy of Leibniz published. 1901

Discovery of Russell’s paradox.

Bertrand Russell—A Timeline

1903

xi

Principles of Mathematics published.

1903–1910 Working with Whitehead on Principia Mathematica. 1905  “On Denoting” published in Mind. Sees the new theory of descriptions as a way out of the paradoxes. 1906 At work on the substitution theory of classes and relations as the solution to the paradoxes. 1907 Stands as a candidate for Parliament supporting women’s suffrage, but is defeated. 1908 “Mathematical Logic as Based on the Theory of Types” published. 1910 Vol. I of Principia Mathematica published. Philosophical Essays published. 1911 Becomes involved with Lady Ottoline Morrell, works on Problems of Philosophy, meets Wittgenstein. 1912  Problems of Philosophy and Vol. II of Principia Mathematica published. Russell works on “the problem of matter.” 1913 Vol. III of Principia Mathematica published. Russell works on the unfinished Theory of Knowledge. Wittgenstein dictates the Notes on Logic. 1914  Lectures at Harvard on theory of knowledge and logic. Gives Lowell Lectures, published later in the year as Our Knowledge of the External World. The First World War begins. Joins the antiwar Union of Democratic Control. First six chapters of Theory of Knowledge published in the Monist. 1915 “Ultimate Constituents of Matter” delivered and published. Mostly involved in anti-war activity. 1916  Active in No-Conscription Fellowship. Principles of Social Reconstruction published. Denied passport to take up teaching position in Harvard. Dismissed from Trinity College. 1917 Lectures on mathematical logic (later published as Introduction to Mathematical Philosophy). Political Ideals published. 1918 Lectures on the philosophy of logical atomism (January–March). In Brixton prison (May–September). Roads to Freedom published.

xii

Bertrand Russell—A Timeline

1919 “On Propositions,” Introduction to Mathematical Philosophy published. Meets Wittgenstein in The Hague. 1920 Visits Russia with British labour delegation. Goes to China with Dora Black. Practice and Theory of Bolshevism published. Begins lectures in China. 1921 Continues lectures in China. Marries Dora Black. Analysis of Mind published. Son John Conrad born. 1922 Visits Wittgenstein in Innsbruck. The Problem of China published. Stands for parliament from Chelsea and is defeated. 1923  Prospects of Industrial Civilization (with Dora Russell) and The ABC of Atoms published. Daughter, Katherine Jane, born. Again stands for parliament from Chelsea and is defeated. 1924  Icarus or the Future of Science published. Lectures in American (April–June). 1925  ABC of Relativity and the second edition of Principia Mathematica I published. 1926  On Education Especially in Early Childhood published. 1927 “Why I am Not a Christian” read and first published. The Analysis of Matter and An Outline of Philosophy published. Beacon Hill School opens. Lectures in America. 1928

Sceptical Essays published.

1929

Marriage and Morals published. Lectures in America.

1930  The Conquest of Happiness published. Marriage to Dora Russell begins to disintegrate. 1931 Becomes the Third Earl Russell. Moore and Russell are examiners for Wittgenstein’s PhD. Discusses parts of Philosophical Remarks with Wittgenstein. Lectures in America. The Scientific Outlook published. 1932  Education and the Social Order published. Separation from Dora Russell and the end of Russell’s work with Beacon Hill School. 1934

Freedom and Organization 1814-1914 published.

1935

Religion and Science published.

Bertrand Russell—A Timeline

xiii

1936  Russell marries Patricia (Peter) Spence. Which Way to Peace? published. “The Limits of Empiricism” published. 1937  The Amberley Papers (with Patricia Russell) published. Son Conrad born. Education and Democracy published. 1938  Power: A new Social Analysis and the second edition of Principles of Mathematics published. Russell goes to America to lecture. 1939

Lectures at the University of Chicago, then at UCLA.

1940 Accepts a position at College of the City of New York. Public campaign against Russell. Russell’s appointment revoked by a New York court. Delivers William James lectures at Harvard. An Inquiry into Meaning and Truth published. 1941–42 In Philadelphia working for the Barnes Foundation. Work on A History of Western Philosophy. 1944

Returns to Great Britain.

1945

A History of Western Philosophy published.

1946

Physics and Experience published.

1948  Human Knowledge: Its Scope and Limits published. Lectures in Sweden. 1949  Authority and the Individual published. Receives the Order of Merit. 1950 Visits Australia and America. Unpopular Essays published. Receives the Nobel Prize for literature. 1951 Russell on a long lecture tour of America. The Impact of Science on Society and New Hopes for a Changing World published. 1952

Marries Edith Finch.

1954

Human Society in Ethics and Politics published.

1955

Russell-Einstein manifesto. Moves to Plas Penrhyn in Wales.

1957

Why I am Not a Christian and Other Essays published.

1958

Russell president of Campaign for Nuclear Disarmament (CND).

1959  Common Sense and Nuclear Warfare and My Philosophical Development published.

xiv

Bertrand Russell—A Timeline

1960  Bertrand Russell Speaks his Mind published. Committee of 100 formed. 1961 After a march on Hiroshima Day with the Committee of 100, Russell sentenced to a week in prison. Fact and Fiction published. 1963

Unarmed Victory published.

1967 International War Crimes Tribunal, Stockholm and Copenhagen. War Crimes in Vietnam published. 1967–69

Autobiography (3 volumes) published.

1970 (February 2) Russell dies at Plas Penrhyn in Penrhyndeudraeth, Wales.

Abbreviations for Works Written by Russell Many of Russell’s articles and papers now appear in the Collected Papers of Bertrand Russell. These are abbreviated as Papers in this book. The full bibliographic reference to them is as follows: Papers 1: Cambridge Essays: 1888-99. Edited by Kenneth Blackwell, Andrew Brink, Nicholas Griffin, Richard Rempel, and John G. Slater. London: Allen and Unwin, 1983. Papers 2: Philosophical Papers: 1896-99. Edited by Nicholas Griffin and Albert C. Lewis. London: Unwin Hyman, 1990. Papers 3: Towards the “Principles of Mathematics,” 1900-2. Edited by Gregory H. Moore. London: Routledge, 1993. Papers 4: Foundations of Logic, 1903-5. Edited by Alasdair Urquart. London: Routledge, 1992. Papers 5: Towards “Principia Mathematica,” 1905-08. Edited by Gregory H. Moore. London: Routledge 2014. Papers 6: Logical and Philosophical Papers, 1909-13. Edited by John G. Slater, London: Routledge, 1992. Papers 7: Theory of Knowledge: The 1913 Manuscript. Edited by Elizabeth Ramsden Eames and Kenneth Blackwell. London: Allen and Unwin, 1984. Papers 8: The Philosophy of Logical Atomism and Other Essays: 1914-19. Edited by John G. Slater. London: Allen and Unwin, 1986. Papers 9: Essays on Language, Mind, and Matter, 1919-26. Edited by John G. Slater. London: Unwin Hyman, 1988. Papers 10: A Fresh Look at Empiricism: 1927-42. Edited by John G. Slater. London: Routledge, 1996. Papers 11: Last Philosophical Testament: 1943-68. Edited by John G. Slater. London: Routledge, 1997.

Abbreviations for Works Written by Russell

xvi

Papers 12: Contemplation and Action: 1902-14. Edited by Richard A. Rempel, Andrew Brink, and Margaret Moran. London: George Allen and Unwin, 1985. Papers 13: Prophecy and Dissent, 1914-16. Edited by Richard A. Rempel, London: Unwin Hyman, 1988. Papers 14: Pacificsm and Revolution: 1916-18. Edited by Richard A. Rempel, Louis Greenspan, Beryl Haslam, Albert C. Lewis and Mark Lippincourt. London: Routledge, 1995. Papers 15: Uncertain Paths to Freedom: Russia and China, 1919-22. Edited by Richard A. Rempel and Beryl Haslam. London: Routledge, 2000. Papers 21: How to Keep the Peace: The Pacifists Dilemma, 1935-38. Edited by Andrew G. Bone and Michael D. Stevenson. London: Routledge, 2008. Papers 28: Man’s Peril, 1954-55. Edited by Andrew G. Bone. London: Routledge, 2003.

Abbreviations of books by Russell used in this edition AUTO The Autobiography of Bertrand Russell, which is in three volumes. London: Allen and Unwin, 1967-9. AMa

The Analysis of Matter. London, Kegan Paul, 1927. 

AMi

The Analysis of Mind. London, Allen and Unwin, 1921. 

CH

The Conquest of Happiness. London, Allen and Unwin, 1930.

GSD

German Social Democracy. London, Longmans, Green, 1896.

HK

Human Knowledge. London, Allen and Unwin 1948.

HWP

History of Western Philosophy. London: Allen and Unwin, 1946.

HSEP

Human Society in Ethics and Politics. London: Allen and Unwin, 1954.

IMP

Introduction to Mathematical Philosophy. London: Allen and Unwin, 1919. 

IMT

Inquiry into Meaning and Truth. London: Allen and Unwin, 1940.

MM

Marriage and Morals. London, Allen and Unwin, 1929. 

Abbreviations for Works Written by Russell

xvii

MPD

My Philosophical Development. London, Allen and Unwin, 1959.

OKEW

Our Knowledge of the External World. Chicago and London: Open Court, 1914.

OKEW2 Our Knowledge of the External World. London: Allen and Unwin, 1926. OOP

Outline of Philosophy. London, Allen and Unwin 1927. 

PL

A Critical Exposition of the Philosophy of Leibniz. Cambridge: Cambridge University Press, 1900. Second edition London, Allen and Unwin 1937.

PM

Principia Mathematica 3 volumes. Second edition Cambridge: Cambridge University Press, 1925–27.

POM

Principles of Mathematics London, Allen and Unwin, 1938, first edition, 1903.

POP

Problems of Philosophy London: Willias and Norgate, 1912. 2nd edition (cited in this book), Oxford: Oxford University Press, 1959.

Power

Power: A New Social Analysis. London: Allen and Unwin, 1938.

PSR

Principles of Social Reconstruction. London, Allen and Unwin, 1916.

RS

Religion and Science. London, Thorton Butterworth, 1935.

TK

Theory of Knowledge (Papers 7 above).

Collections of essays BW

The Basic Writings of Bertrand Russell. Edited by Robert Egner and Lester Denonn. London: Allen and Unwin, 1961. Revised edition Oxford: Routledge, 2009.

EA

Essays in Analysis. Edited by Douglas Lackey. London: Allen and Unwin, 1974. 

LK

Logic and Knowledge. Edited by Robert C. Marsh. London: Allen and Unwin, 1956.

ML

Mysticism and Logic and other Essays. London: Longmans Green, 1918.

xviii

Abbreviations for Works Written by Russell

PE

Philosophical Essays. London: Longmans Green, 1910. Revised edition, London: Allen and Unwin, 1966.

RoE

Russell on Ethics. Edited by Charles Pigden. London, Routledge, 1999.

UE

Unpopular Essays. New York: Simon Schuster, 1950.

WNC

Why I am Not a Christian and other essays on religion and related subjects. Edited by Paul Edwards. London: Allen and Unwin, 1957.

Selected articles Most references to articles cite the source in the Collected Papers or one of the collections of essays. Nevertheless, a few articles are also referred to by these abbreviations. KAKD “Knowledge by Acquaintance and Knowledge by Description.” First published in PAS, 1911. In Papers 6 and ML. LA “Logical Atomism.” First published in Contemporary British Philosophy, edited by J. H. Muirhead (1924). In Papers 9 and LK. MMD “My Mental Development.” In Schilpp (ed.), The Philosophy of Bertrand Russell (1944). In Papers 11. OD “On Denoting.” First published in Mind, 1905. In Papers 4, LK, and EA. OP “On Propositions: What They Are and How They Mean.” First published in PAS Supplementary Volume, 1919. In Papers 8 and LK. PLA “The Philosophy of Logical Atomism.” First published in The Monist in four installments, Oct. 1918, Jan. 1919, Apr. 1919, and July 1919. In Papers 8 and LK. RSDP “The Relation of Sense-Data to Physics.” First published in Scientia 1914. In Papers 8 and ML. UCM “The Ultimate Constituents of Matter.” First published in The Monist 1915. In Papers 8 and ML.

Introduction

Bertrand Russell’s life spans from 1872 to 1970. He was born in the middle of the Victorian era, when Gladstone was prime minister, and lived through the changes of the first two-thirds of the twentieth century, with its two World Wars and the Cold War and nuclear standoff of the 1950s and 1960s. He was born into an aristocratic family: his grandfather was twice prime minister of Great Britain, and was made an earl by Queen Victoria. Unlike his grandparents, Russell’s parents were fairly radical free thinkers and chose John Stuart Mill to be his (secular) godfather. However, they died when he was quite young, and he and his brother were left to the charge of their grandmother who did not share their parents’ radical views and annulled the request that they be brought up agnostic. In later life, Russell came on his own to views more like his parents’ despite the efforts of his grandmother. His grandmother was domineering and hoped that he would, like his grandfather, become a politician. She strongly disapproved of his first marriage to Alys Pearsall Smith, an American Quaker. In his autobiography Russell related the extreme efforts, which included many discussions of madness in the family and even an assignment to the British embassy in Paris, of those he called “my people,” foremost his grandmother, to prevent the marriage. After coming of age and dutifully going to Paris, Russell married Alys anyway.1 Prior to his marriage with Alys, Russell had been very interested in philosophical issues, but when he went to Cambridge as an undergraduate, he went first into mathematics. There he studied with A. N. Whitehead. It was not until his fourth year that he studied philosophy, when he was exposed mostly to the idealism of McTaggart and others, which was derived from Hegel. Russell’s fellowship dissertation, which was later published with some modifications as An Essay on the Foundations of Geometry, was written in that style.2 The search for philosophical foundations of mathematics was already present, although his later logicism, the view that all mathematical concepts are upon analysis, logical concepts and that all mathematical truths are logical truths, was not yet developed. From his studies at Cambridge, he became acquainted with Spinoza’s work, and while teaching in Cambridge in 1899 he was asked to teach

2

The Bloomsbury Companion to Bertrand Russell

McTaggart’s course on Leibniz. For the rest of his philosophical life Spinoza and Leibniz remained his two philosophical heroes, despite his disagreements with their metaphysics. During the years 1898 to 1900 Russell had a major change of heart. First, along with G.E. Moore, he rejected the main theses of his idealist teachers. Second, he attended the International Congress of Philosophy in Paris in 1900, a year he later characterized as “the most important year of my intellectual life.”3 There he learned of the work of Giussepi Peano and a symbolic language within which to express his ideas. What followed was the advent of what is broadly known as “analytic philosophy.” Along with Moore, who had published an important paper, “The Nature of Judgement,” in January 1899, Russell was responsible for the development of analytic philosophy, especially in Great Britain. Frege was an important precursor, but Frege’s work became known in the English-speaking world primarily through Russell. During the first part of the twentieth century, Russell was responsible for two extremely important works on the foundations of mathematics, the Principles of Mathematics and the three-volume Principia Mathematica written with his former mathematics teacher, A. N. Whitehead. These works, along with Frege’s Basic Laws of Arithmetic, are the major works in logicism. His works at the beginning of the century on topics in epistemology, philosophy of language, and philosophy of science, especially those written around 1914 and his slightly later papers on logical atomism, had an outsized influence on philosophy in the twentieth century. The movement known as logical positivism owed a great deal to Russell, even as they differed with him on key points. The later critics of positivism also owed a debt to Russell. W. V. Quine, whose developed position differs in many ways from Russell’s, also was influenced by Russell. The ordinary language philosophers of the 1950s and 1960s, while harshly critical of Russell, were working in an area made possible by Russell. Wittgenstein had come to England in 1911 to study engineering, but then turned to philosophy, inspired initially by Russell. He was both influenced by Russell and had a strong influence on him. Russell made possible the publication of Wittgenstein’s Tractatus. Russell’s last major philosophical work, Human Knowledge, did not have the impact that Russell had hoped. Yet more than anything else of its era, it anticipates the analytic philosophy of the last quarter of the twentieth century and the beginning of the twenty-first. Russell was also a public intellectual. He campaigned for women’s suffrage in 1907, running for parliament on a pro-suffrage ticket. He became wellknown as a pacifist during the First World War, going to jail at one point.

Introduction

3

He was an outspoken advocate of new ideas in education and marriage, and in later life became involved in the movement for nuclear disarmament. As a public intellectual he was often vilified. The anti-war position he took in 1914 was hardly popular. He was viewed as an extremist, banned by the government from going to various places, fired from Trinity College for his position on conscription and even jailed for remarks critical of United States. After he and his second wife Dora had children, they became very interested in education and opened a school called the Beacon Hill School. They and their school were often criticized as radical. Two of Russell’s more popular books, Marriage and Morals and The Conquest of Happiness, were published during this time and again Russell was seen as a radical in the public eye, especially as these books advocated sex before marriage and open marriages. Outrage at this radical line was behind the campaign by religious leaders to force CCNY to withdraw an offer of a professorship to Russell in 1940. In a case where a woman filed a suit ordering the Board of Education to rescind the appointment, a judge ruled in her favor. The suit claimed her daughter (who would not have been able to attend the classes anyway) would be at risk; her lawyer characterized Russell as “libidinous, lustful, venerous, erotomaniac, aphrodisiac, irreverent, narrowminded, untruthful, and bereft of moral fibre.” That Russell would have been teaching foundations of mathematics and logic to an all-male student body did not deter the judge from ordering the New York Board of Higher Education not to hire Russell.4 During the relatively short time after the end of the Second World War and before the McCarthy era, on the other hand, Russell was a more popular figure in the public mind. He appeared to be more a part of the establishment: he was awarded the Order of Merit, Great Britain’s highest civilian award, in 1949, and the Nobel Prize for Literature in 1950. His book A History of Western Philosophy became a best seller during this time. Soon again, though, he was at the center of controversy, first as an advocate for nuclear disarmament and then as a critic of American policy in Vietnam. At the end of his life his Bertrand Russell Peace Foundation, especially under the rule of Ralph Schoeneman, came out with some fairly radical positions, supporting Che Guevara’s abortive uprising in Bolivia as well as having a tribunal for war crimes against the United States. In these activities Russell’s group was seen by the political right as largely an anti-American front aiding global communism. The political right seemed to have forgotten how critical Russell had been of the Soviet Union throughout his life. His Practice and Theory of Bolshevism was seen by those on the left as a betrayal of the left, but Russell had been thoroughly

4

The Bloomsbury Companion to Bertrand Russell

shaken by what he saw firsthand on a trip to Russia in 1920. When Germany invaded the Soviet Union, Russell, who had abandoned his former pacifism because of his disgust of Hitler, wrote to his friend Gilbert Murray, “I have no doubt that the Soviet Government is even worse than Hitler’s, and it will be a misfortune if it survives.” Within professional philosophical circles Russell’s star also rose and fell. He was an extremely important figure in the first decades of the twentieth century, known not only in Great Britain, but also in the United States and France. He was invited to give lectures in the United States as early as 1896, and taught a course in Harvard in 1914. He was asked to join the Harvard philosophy department in 1916 but was denied a passport by the War Office. He gave lectures in France several times during this early period and published several articles in French journals. In the first decade of the twentieth century, he carried on a lively debate with Henri Poincaré on the relation of mathematics to logic. Louis Couturat was an ally of his in the debate with Poincaré and earlier had been a sympathetic reader of Russell’s work on Leibniz. During the 1920s Russell’s work was very influential on rising British philosophers and the logical positivists. His work from that time, though, did not have the impact of his earlier work, and there is no doubt his reputation in philosophy suffered as he devoted more time to his popular works. When he returned to working in mainstream philosophy in the late 1930s he was considered dated in Britain, but still drew large numbers of students in the United States where he taught at Chicago, UCLA, and Harvard before the CCNY fiasco. While one can find in his later Human Knowledge many precursors to examples and topics prominent in contemporary analytic philosophy, the work did not get much attention, and while A History of Western Philosophy was popular with nonacademic audiences, professional philosophers were largely critical of it. While some of the logical positivists took their inspiration from some of Russell’s enthusiasm for science and skepticism of ordinary philosophy, their adoption of the verification theory of meaning put them at odds with much of Russell’s work. In particular, the philosophical problems that he took seriously were seen by them to be pseudo-problems. With the advent of ordinary language philosophy and the rise in popularity of Wittgenstein’s later work, Russell’s work went further out of favor in professional philosophical circles. While these philosophers rejected the verification theory of meaning of the positivists, they still held that philosophical problems were pseudo-problems to be dissolved by an understanding of language. Those who held that the primary purpose of

Introduction

5

philosophy was therapeutic found little sympathy with Russell’s work even though one can see an important precursor to this view in some of Russell’s own earlier works, where the theory of descriptions was used to eliminate metaphysical difficulties. The neglect of Human Knowledge may in part be explained by this lack of favor in the profession. After Russell’s death, Anglo-American philosophy became less hostile to metaphysics and the epistemological enterprise Russell had been engaged with. As analytic philosophy has revived interest in traditional problems of epistemology and metaphysics there has been a renewal of interest in Russell’s attempts to solve some of these problems. With the establishment of the Russell Archives and the publication of the Collected Papers, many of Russell’s previously obscure or unpublished papers are now available. There have been many studies of Russell’s philosophy and many collections of essays and issues of journals devoted to Russell’s work. This Companion strives to add something new to this body of literature, and so does not attempt to provide a comprehensive treatment to all of the issues involved with Russell’s philosophy. The chapters here are divided into two sections, the first treating Russell’s relations with other philosophers, and the second focusing on philosophical issues with which Russell grappled. The section on Russell in context looks at Russell’s interaction with philosophers from the British idealists to the positivists. The section on philosophical issues looks at a wide range of issues that concerned Russell, including his philosophy of mathematics, his metaphysics and epistemology, his philosophy of language, and his moral philosophy. Given the breadth of his work and the number of philosophers he worked with or reacted to during his long life, there are many topics that are not covered in this study. In the section “Russell in Context,” we have studies on Russell’s relationships with several philosophers. James Levine looks at the idealist background to Russell’s earlier philosophy, helpfully distinguishing between the different strains of Hegelianism found in Bradley and McTaggart, and the Kantianism favored by Ward, who was Russell’s supervisor. Levine then looks at Russell’s own idealism, which was influenced by all of these people. The importance of putting Russell’s views in this context becomes clear, for example, in understanding Russell’s claim to having accepted the ontological argument. The argument in question, Levine makes clear, was not the argument from Anselm and the early moderns, but rather Bradley’s version. Levine also looks at Russell’s break with idealism and his embrace of what Levine calls “Moorean realism.” Levine sees Russell as retreating from extreme Moorean realism and argues that while Russell

6

The Bloomsbury Companion to Bertrand Russell

continued in his strident criticism of the method of the idealists, he modified his views on some of the very doctrines that had initially pushed him to reject idealism. In the end Levine sees hints of Russell’s earlier views surfacing in his later philosophy. Cheryl Misak discusses Russell’s attitudes toward and interactions with the pragmatists. Russell is known for his criticisms of the pragmatic theory of truth, criticisms which were reprinted in his 1910 Philosophical Essays. While Russell maintained his critique of the pragmatic theory of truth, he took seriously many of the positions advocated by the pragmatists, in the end coming around to James’s rejection of the ego and embracing his neutral monism. Russell’s flirtation with behaviorism was also a consequence of his engagement with the pragmatists. Misak points out that Russell was not seen as simply misunderstanding pragmatism, but as someone they saw sympathetic enough that, in 1916, the Harvard philosophers invited him to teach a course on Peirce. Because of his anti-war activities the British government denied him a passport so he never did that. Bernard Linsky’s chapter on Russell’s relationship with Meinong and Frege first looks at issues surrounding Russell’s discovery of the paradox of the set of all sets that are not members of themselves. He remarks on an anticipation of the paradox in Zermelo’s discussion of an argument given by Schröder that had also been criticized by Frege. Linsky notes that Russell’s discovery of the paradox was independent of these earlier ruminations. In a careful examination of Russell’s notes on Frege’s writings he points out what Russell did and also did not accept from Frege’s system. He then looks at Russell’s interaction with Meinong focusing on both Russell’s correspondence with Meinong and the important article “Meinong’s Theory of Complexes and Assumptions.” He ends with a discussion of Russell’s famous example of the present king of France. My own chapter on Russell and Wittgenstein does not attempt to cover all that has been said about their relationship, as that has been covered in detail by many works, but instead focuses on those particular details of their work together, which had a lasting impact on Russell’s thought. Russell worked on a paper on what he called the problem of matter through much of the year 1912 and on a book he never published, Theory of Knowledge, in 1913. In both of these we have an insight into the interaction between Russell and Wittgenstein and have a sense of what Russell took away from his encounter with Wittgenstein and which he developed in his later lectures on logical atomism. Wittgenstein most affected Russell’s views on logic, including logical grammar. I argue that

Introduction

7

Russell’s reflections on logic led him to be suspicious of the kinds of inferences he allowed in his 1912 Problems of Philosophy, and the reflections on logical grammar had many consequences which we can see as early as Russell’s lectures at Harvard in the spring of 1914. In the last chapter in this section, François Schmitz takes a fresh look at the relation between Russell and the logical positivists. He examines the influence Russell had on the logical positivists, specifically Carnap, and looks at two of Carnap’s works, the Abriss der Logistik and the Logische Aufbau der Welt, pointing out what Carnap took from Russell, particularly his theory of relations and his idea of a scientific method for philosophy. Schmitz also looks at those positivists who were less influenced by Russell’s work and more influenced by Wittgenstein, such as Schlick and Waismann. Returning to Russell, Schmitz discusses Russell’s own views on the positivists, particularly the criticism Russell leveled in An Inquiry into Meaning and Truth toward the positivists’ theories of protocol sentences, meaning, and truth. The next section of the Companion revolves around philosophical issues that were important to Russell at least one time or other in his development. All of the discussions seek to shed new light on Russell’s work and his ideas. Some of the papers defend Russell’s view or developments based on Russellian ideas. As there have recently been collections of essays on various aspects of Russell’s philosophy, certain topics are not covered here.5 In the first chapter in this section, Kevin Klement takes a close look at the doctrine of logicism, “the thesis that mathematics is logic.” This is a thesis that was central to Frege (at least with respect to arithmetic) and to Russell. Although working within a different framework, Carnap later adopted a form of logicism. As Klement points out at the beginning of his paper, it is a position that is not looked on favorably these days, because among other things, if one has a conception of logic confined to first-order logic, the thesis will be false. It is clear that Russell, or anyone who espouses a form of logicism, needs a more robust sense of logic as a starting-point. Russell, of course, had a stronger logic. Klement points out that a genuine criticism of the thesis would need to show an example of a truth of mathematics which is not a truth of logic. Many have noted that as a consequence of Gödel’s incompleteness proof there is no deductive system which is consistent and has a recursive set of axioms such that it will have as theorems all mathematical truths. Klement recognizes that most logicists thought their systems would have this feature, but he does not see it as an essential feature for logicisms, pointing out that the theorem applies to

8

The Bloomsbury Companion to Bertrand Russell

second-order logic. A different criticism of logicism involves axioms, such as the axiom of infinity and the axiom of choice, needed for certain important results, but which even Russell did not think were logical axioms. Here Klement takes a new approach, questioning whether the theorems which result from these axioms should be thought of as being themselves part of pure mathematics. Klement concludes that Russell’s logicism may have been more successful than even Russell himself thought. In his chapter on denoting and language in Russell’s philosophy, Graham Stevens seeks to combat the view that Russell only had a minimal interest in philosophy of language when he formulated his theory of descriptions. While it is true that the development of the theory was connected to many concerns Russell had with, for example, the theory of types and issues in his epistemology, Stevens argues that since his early work on Leibniz, Russell was interested in issues of propositional content and so those issues are central. Stevens argues that Russell’s insight specifically into the quantificational structure of descriptions as a thesis in philosophy of language can be defended even if doctrines associated with it are less defensible. Russell himself did not often distinguish his basic thesis concerning descriptions from other doctrines, such as the principle of acquaintance and the descriptive theory of names. Stevens also questions Russell’s characterization of descriptive phrases as incomplete symbols, but again thinks this issue can be separated from Russell’s basic insight. Gregory Landini looks at the thesis, from Russell’s 1914 Our Knowledge of the External World, that logic is the essence of philosophy. As Russell understood the thesis it is that every purported philosophical problem is seen after analysis to be either not a philosophical problem or is logical. Landini holds that Russell had a robust logic, which he calls “cp-Logic,” where cp refers to a comprehension principle. He points out that originally Russell thought that logic (and mathematics) was a synthetic a priori science grounded in our awareness of universals. As Russell had argued that mathematical truths ultimately are logical truths, so too he says that philosophical claims were best understood as claims about logical relations. Using the logical techniques of Principia, together with empirical advances in physics, philosophy can reconstruct physical objects and attack the philosophical problem of the relation of our minds to the physical world. Landini argues that Russell can be understood as holding that the only necessities are logical necessities, even though some of Russell’s own remarks about necessity obscure this. Landini says that in both Problems of Philosophy and Our Knowledge of the External World Russell sees his scientific philosophy

Introduction

9

as freeing people from the tyranny of dogmas which arise from supposed nonlogical necessities. At the end of his chapter he offers a contrast between Russell’s scientific method in philosophy and that of Dewey and Quine. In my chapter on sense-data, sensibilia and percepts, I begin with Russell’s reasons for positing sense-data, which are distinct from the reasons of those who see sense-data as items about which we can have guaranteed certain knowledge. Thus, Russell did not hold that sense-data were mental entities, nor did he hold that our judgments about them were infallible. In fact, they were very much part of a realist philosophy rather than a phenomenalist one. Russell’s position appears to be more phenomenalist when he moved from the epistemological position of inferring enduring physical objects from features of sense-data to seeing physical objects as constructions from sense-data and sensibilia. I discuss Russell’s reasons for this move and argue that Russell’s inclusion of sensibilia is not incoherent given his actual project. I then discuss Russell’s abandonment of sense-data and argue that the later views are not as radical a break from the earlier ones as they are often portrayed. In particular, Russell did not completely abandon the constructions he had articulated in “The Relation of Sense-Data to Physics” and Our Knowledge of the External World. Russell’s philosophy of matter after Problems of Philosophy began in his unpublished “On Matter,” but was articulated in “The Relation of Sense-data to Physics,” where Russell gave as a maxim that logical constructions are to be preferred over inferred entities. Russell’s constructions have often been taken to be of a piece with a quest for Cartesian certainty, but Russell instead is involved with a dialectic between on the one hand what he takes physics to tell us, and on the other what he takes our evidence to be. The goal of the constructions was to develop an account of the relation of sense-data to physics. Unlike the Cartesian claim that we should withhold assent to a belief unless our evidence is beyond doubt, Russell developed the position that we should strive to accommodate both our epistemology and our physics, and even went further in later years to say that the results of physics were more likely to be true than any philosophical claim. As a result, he modified his view, including sensibilia at first and later rejecting sense-data as items with which a pinpoint mind had acquaintance. But his percept-events of his later philosophy played many of the roles sense-data had and so his position is best seen as evolving rather than radically changing over time. Donovan Wishon argues for a similar conclusion this time on the matter of introspection. He also thinks that Russell’s views on introspection have been

10

The Bloomsbury Companion to Bertrand Russell

somewhat distorted by linking Russell’s overall project to a Cartesian quest for certainty. Knowledge by acquaintance even for psychological matters is independent of knowledge of truths. Introspection covers not only our psychological states, but at least in Problems of Philosophy also the self. Wishon points out that while Russell initially held that introspection involved some kind of direct acquaintance with the self, he gradually came to see our knowledge of the self as indirect, first, and then to see that the self was best understood as a logical construction not unlike what he had said about physical objects, except that here the data were not sense-data but the objects of “inner perception.” Wishon discusses Russell’s view of perception of complexes or facts with regard  to introspective data and also the move from these perceptions to judgments. As knowledge about the physical world was augmented by knowledge by description, so too knowledge about the mental world is also often augmented by knowledge by description. Initially, in that part of Theory  of  Knowledge published as the “Nature of Acquaintance,” Russell had held our knowledge of the self was descriptive in this way. As in the case with his views about sense-data, Russell’s views about introspection changed with the onset of his neutral monism. In the case of sense-data, Russell had to reject the distinction he had made between sensations and their objects, ultimately arguing that there was no genuine distinction between them. Wishon points out the Russell made analogous changes with respect to introspective data. Prior to adopting neutral monism, Russell thought our understanding of other people’s minds were best seen on the model of construction, as he had thought of our knowledge of ordinary objects. He soon thought this of our understanding of our own minds. At this time, Russell was not denying that minds existed as substances, but rather that we should substitute our constructions for those inferred entities. With the advent of his neutral monism, the major change was that the basic elements that went into the construction of minds and matter were no longer taken to be ontologically distinct. In their chapter, “Russell’s Bridge,” Duston Olson and Nicholas Griffin look at issues of epistemic justification in Russell’s later philosophy. They take their title from a passage in Human Knowledge that likens the edifice of knowledge to a bridge, which rests on many piers, but whose piers’ own stability is increased by the interconnecting girders (HK, 413). Olson and Griffin see in this metaphor a precursor of recent attempts to blend coherence theories with foundational ones. In developing the analogy, Russell likens the piers to “propositions having

Introduction

11

some intrinsic credibility” while the superstructure is likened to propositions which are only inferred. The first point may suggest that what we have here is a foundationalism, but Olson and Griffin argue that it is neither a straightforward foundationalism nor a straightforward coherentism. They see Russell’s view as sharing features of such contemporary philosophers as Earl Conee, Richard Feldman, Susan Haack, Catherine Elgin, and Kevin McCain. All of these views reject an infallibilist conception of knowledge and embrace an epistemic holism, but do not accept simple coherentists positions either, as they think there is a class of beliefs which, while not infallible, have an intrinsic credibility. After characterizing Russell’s epistemology in Human Knowledge in this way, Olson and Griffin then go on to argue that this view of epistemic justification is not such a radical break with Russell’s previous philosophy as it may seem. Here they do not focus as much on Russell’s discussion of empirical knowledge, but rather on the methodology of Principia Mathematica, which has often been seen as a paradigm of the quest for epistemic certainty. They focus on Russell’s very important 1907 paper “The Regressive Method of Discovering the Premisses of Mathematics,” which Russell presented, but didn’t publish. Examining several others of Russell’s works Olson and Griffin see that the regressive method was something Russell was fairly consistent in holding. The continuation of the story of sense-data on the one hand and introspection on the other is taken up in Christopher Pincock’s paper on neutral monism. Neutral monism comes in moderate and more extreme flavors. A moderate neutral monist might hold that there is no intrinsic ontological difference between those items which enter into mental constructions and those which enter into material constructions. Particulars on this view are neither mental nor material in themselves, but properly only constructs are. An even weaker neutral monism might hold that there is no known intrinsic difference between particulars which enter into mental states and those that enter into physical ones. Pincock discusses Russell’s earlier, dualist views and his 1914 engagement with neutral monism and the objections he raised there. After a careful examination of the objections Russell raised, Pincock argues that Russell’s adoption of neutral monism is very gradual. Russell rejected, as early as 1913, the view that we were acquainted with our own minds as mental substances, but still had difficulties with neutral monism, especially with regard to a proper analysis of beliefs and what he called “emphatic particulars.” Russell appears to accept neutral monism in his 1919 “On Propositions” and 1921 Analysis of Mind, but Pincock is reluctant to see these as genuinely neutral monist works as in both of them Russell thinks

12

The Bloomsbury Companion to Bertrand Russell

that the particulars which are images are such that they do not enter into any complexes which are material. In particular they are not located in any physical space. Pincock sees Russell adopting a more robust neutral monism only after 1925, when he adopts the view that percepts, which include images, are events which occur in a perceiver’s brain. Russell was well-known for his popular books on ethical topics, but much of his work on ethical theory is buried in chapters of his more socially oriented books, reviews, papers read to audiences but not published until recently or in other fairly obscure places. Consequently, Russell’s actual work on ethical theory is little known. Raymond Perkins argues that Russell did serious work in ethical theory and divides Russell’s writings into an initial realist period, much under the influence of G. E. Moore, followed by what he calls “three decades of subjectivism.” These decades see the development of an evolving emotivist theory, punctuated by a brief flirtation with error theory. Perkins argues that Russell anticipated Mackie in his development of error theory and did pioneering work in emotivism. Perkins finds the most sophisticated statement of Russell’s emotivism in his Religion and Science. Perkins then points out some of the advantages Russell’s mature theory has over for example, Ayer and Stevenson’s emotivist ethics, but also brings up several criticisms which he thinks had some effect on Russell and led him to modify his view even more in Human Society in Ethics and Politics. This last work also dealt with issues of peace and war in the nuclear age, and Perkins closes his paper by tackling some of Russell’s more controversial statements during this period, and how Russell’s arguments about the immorality of the doctrine of nuclear deterrence fit with his ethical theory. In the closing chapter, Peter Stone looks at Russell’s view on the study of history. Stone begins with the debate from the turn of the century concerning whether history should be considered a science, or whether a literary approach was a more fruitful way of doing history. While Russell was later to champion a scientific method in philosophy and often said his goal was to make philosophy more scientific, he came down in this debate on the side of the literary approach, rejecting the view that history was best seen as of a piece with the sciences. Russell here sided with his friend George Trevelyan, in his essay, “Clio, a Muse.” Stone notes that Russell had in fact embraced this view prior to Trevelyan’s essay, but the two influenced each other for a long period. He looks at several of Russell’s essays on history and the arguments Russell gives there against the scientific or economic approach to history. Russell was especially skeptical of discovering the equivalent of scientific laws in the study of history and thought

Introduction

13

that the emphasis on such laws overlooked the actual value of history, which he sees as developing certain attitudes toward society in the general public. One of the criticisms of the “literary” approach to history has been the tenuous connection between the study of history on this model and the actual truth, as it may be the case that such attitudes are even better inculcated by fictional stories. In a later part of his chapter, Stone takes on this question and argues that neither Trevelyan nor Russell fully addressed the difficulties. In some passages, Russell talked about Greek history and Gibbon’s Decline and Fall of the Roman Empire where Russell appears to suggest that an interesting lively account is more valuable than a strictly accurate one. Stone argues, though, that other features of Russell’s reflections on the importance of history would bring back the importance of accuracy and that Russell is in a better position in this regard than Trevelyan. The reader will notice that in all the essays there is a revisiting of topics which have been part of a standard account of Russell since the second half of the twentieth century. In many cases much of the standard view is retained, but with interesting qualifications. Thus the papers by Levine, Misak, and Schmitz, for example, accept at one level the overall story of Russell’s relationship with the idealists, pragmatists, and positivists, but also point out important points which modify that story. Some of the papers in the second section, such as those by Klement, Landini, Stevens, seek to defend Russellian positions, although often with modifications that Russell might not have accepted. Others, such as those by myself, Wishon, Pincock, Griffin, and Olson, build on work that has called into question Russell as holding a traditional foundationalist epistemology centered on our certain knowledge of our sense-data and mental states.

Notes 1 See Auto I, 122–27. As Russell lived a long and very interesting life, I will not be able to give a biography in this short introduction. I will confine myself to moments associated with his philosophical development. Besides Russell’s own three-volume autobiography, there are biographies written by Clark (1975), by Moorehead (1992) and, by far the most detailed, the two volumes by Ray Monk (1996 and 2000). There are several other works which discuss Russell’s life, including those by his daughter and his second wife. Monk picks up on the passages about madness in the Autobiography and sees it as a theme in Russell’s life. 2 For more on Russell’s Idealist phase, see Chapter 1.

14

The Bloomsbury Companion to Bertrand Russell

3 See “My Mental Development” in Schilpp, 1944. 12. Also Auto I, 232. 4 John Dewey and Horace Kallen collected essays on this topic in defense of Russell in their The Bertrand Russell Case (New York: Viking Press, 1941). For more on this case, see Weidlich, Appointment Denied: The Inquisition of Bertrand Russell (Amherst, NY: Prometheus Books, 2000). 5 For example, acquaintance is dealt with in detail in a recent collection edited by Wishon and Linsky, Acquaintance, Knowledge and Logic. Detailed issues concerning the logic of Principia Mathematica are dealt with in two collections edited by Griffin and Linsky, Principia Mathematica at 100 and The Palgrave Centenary Companion to Principia Mathematica. Two other collections deal specifically with “On Denoting,” one edited by Jacquette and Griffin, Russell vs Meinong: the Legacy of “On Denoting,” and another edited by Jacquette, Griffin and Blackwell, After “On Denoting”: Themes from Russell and Meinong.



Part One

Russell in Context

16

1

Russell and Idealism James Levine

As Russell presents it, there are three central turning points in his early philosophical development: his acceptance of a form of idealism in 1894; his break with idealism, influenced by G. E. Moore, toward the end of 1898; and his attending the International Congress of Philosophy in August 1900, where he encountered the mathematician Guiseppe Peano and his students. In discussing (in 1944) the first of these turning points, Russell “remember[s] the precise moment” in 1894 when he “bec[a]me a Hegelian” (MMD, 10). However, elsewhere he writes that in 1894 he “went over completely to a semiKantian, semi-Hegelian metaphysic” (MPD, 38), and that while his dissertation on geometry “was mainly Kantian” (MPD, 40), he subsequently became “a fullfledged Hegelian,” so that “[w]herever Kant and Hegel were in conflict, I sided with Hegel” (MPD, 42). In introducing the second of these turning points, Russell writes: It was towards the end of 1898 that Moore and I rebelled against both Kant  and  Hegel. Moore led the way, but I followed closely in his footsteps, (MPD, 54)

while with regard to the third, he writes: The most important year in my intellectual life was the year 1900, and the most important event in this year was my visit to the International Congress of Philosophy in Paris. . . . In Paris in 1900, I was impressed by the fact that, in all discussions, Peano and his pupils had a precision which was not possessed by others. (MMD, 12)

At the outset of MPD, Russell presents these two turning points together as constituting the one “revolution” in his thinking—a revolution that both

18

The Bloomsbury Companion to Bertrand Russell

supersedes his earlier views and provides the framework for later changes in his view: There is one major division in my philosophical work: in the years 1899–1900 I adopted the philosophy of logical atomism and the technique of Peano in mathematical logic. This was so great a revolution as to make my previous work, except such as was purely mathematical, irrelevant to everything that I did later. The change in these years was a revolution; subsequent changes have been of the nature of an evolution. (MPD, 11)

In accord with characterizing his idealist philosophy as almost entirely “irrelevant to everything that I did later,” Russell disparages his idealist writings. Thus, he writes of his work on geometry that it “seems to me now somewhat foolish” (MPD, 39); characterizes a paper he wrote in 1897 as both “unadulterated Hegel” (MPD, 40) and “unmitigated rubbish” (MPD, 41); and writes of his idealist work on the philosophy of physics that “it seems to me complete nonsense and I find it hard to imagine how I can ever have thought otherwise” (MPD, 43). However, regardless of Russell’s retrospective assessment of his work as an idealist, his engagement with, and reaction to, idealism is central to his philosophical development and helps illuminate fundamental aspects of his post-idealist philosophy. The philosophy he accepts, following Moore, in breaking with idealism is a kind of “pre-critical realism,” devised in conscious opposition to various forms of idealism—not only Hegelian, but also Kantian as well as Berkeleyan. The changes in his philosophy initiated by his attending the Paris Congress of 1900 provide him with further objections to idealism, but they also undermine central features of the Moorean philosophy he accepted in breaking with idealism—rendering his philosophy more “critical” and less straightforwardly “realist”—and thus modify the character of his opposition to idealism, which comes to focus less on specific doctrines and more on methodology and intellectual temperament. I proceed in five parts. I discuss some aspects, first, of Berkeleyan, Kantian, and Hegelian idealism; second, of the views of British idealists who influenced Russell’s early philosophical development; third, of Russell’s idealist philosophy of 1894–1898; fourth, of the philosophy Russell accepts immediately following his break with idealism; and fifth, of the philosophy he develops following the Paris Congress and how it affects his engagement with idealism.

Russell and Idealism

19

1  Idealism: Berkeley, Kant, and Hegel Russell characterizes idealism as “the doctrine that whatever exists, or at any rate whatever can be known to exist, must be in some sense mental” and claims that there is “no doubt” that the doctrine is opposed to “common sense,” which “regards tables and chairs and the sun and moon and material objects generally as something radically different from minds and the contents of minds, and as having an existence which might continue if minds ceased” (POP, 37).1 For Russell, while idealism “is very widely held among philosophers,” it “has several forms, and is advocated on several different grounds” (ibid.). I outline here some features of the different forms that idealism takes in Berkeley, Kant, and Hegel.

1.1  Berkeley’s idealism On Berkeley’s idealism, there is nothing besides minds and mind-dependent “ideas.” Berkeley sometimes presents his idealism as following from two assumptions regarding “the things we immediately perceive” or “the objects of sense.” First, he holds, with “the vulgar,” that what we “immediately perceive” are “real things,” such as houses, mountains, and rivers; second, he holds, with “the philosophers,” that “the things immediately perceived are ideas, which exist only in the mind.” By “put[ting] together” these two views, he thereby accepts the idealist view that “real things,” such as houses, mountains, and rivers, are mind-dependent ideas.2

1.2  Kant’s idealism Unlike Berkeley, Kant holds that the “objects of the senses” are “things existing . . . outside of us” (1783, 33). Thus, Kant’s idealism, unlike Berkeley’s, is not generated by the view that the “objects of the senses” are “ideas” that “exist in the mind” and does not incorporate the view that there is nothing besides minds and mind-dependent “ideas.” Instead, Kant’s idealism is generated by the view that there are conditions “in us” that structure how “things” appear to us. In Russell’s terminology, Kant accepts a version of that view that “whatever can be known to exist”—as opposed to whatever can exist simpliciter—“must be in some sense mental.”

20

The Bloomsbury Companion to Bertrand Russell

In particular, Kant holds that space and time are “in us” as “pure forms of sensibility.” Since Kant holds both that “intuition” is the means by which a thinking subject relates immediately to objects that are given to it and that humans “intuit” objects through the faculty of sensibility, he calls space and time “forms of intuition.” Given that space and time are forms “in us,” Kant distinguishes between our knowledge of things “as they appear to us” and “as they are in themselves.” Since space and time condition how things appear to us, we can know a priori that whatever “appears to us” is in space and time; but, because we cannot remove these forms, we cannot know things “as they are in themselves”—that is, cannot know how “things” would appear to a being who had no “forms of intuition” conditioning the appearances of those things. Similarly, for Kant, the pure concepts of the understanding, or categories— including, for example, cause and effect—condition “judgments of experience,” which apply only to things “as they appear to us.” Hence, while we can know a priori regarding things as they “appear to us”—including ordinary material objects and all the objects of physics—that “every event is determined by a cause, according to constant laws,” we cannot know whether things “as they are in themselves” are likewise causally determined. Kant argues in a number of ways that space and time are “in us.” One is by considering the faculty of sensibility (as distinct from the understanding) and arguing that space and time are “pure forms” of that faculty.3 Another is by considering how we have a priori knowledge of geometry and arithmetic.4 For Kant, since knowledge of geometry requires the intuition of space, while knowledge of pure arithmetic requires the intuition of time, we can have a priori knowledge of arithmetic and geometry only if we have a priori intuitions of space and time. Moreover, for Kant, we can have such a priori intuitions only if space and time are forms “in us” that precede all sensations. Third, Kant argues that assuming that space and time characterize “things as they are in themselves” leads to “antinomies” that can be avoided only by rejecting that assumption. Thus, for example, he argues that if space and time characterize “things as they are in themselves,” then the world, understood as “the sum of all appearances,” would be a “whole existing in itself,” which is either infinite or finite. If so, the opposition between “The world is infinite in magnitude” and “The world is finite in magnitude” would be contradictory: one of these statements would be true and the other false. Kant argues, however, that neither can be true, so that if they were contradictories, we would face an antinomy. For Kant, avoiding this antinomy requires rejecting the assumption

Russell and Idealism

21

that the world, as it appears in space and time, is a “thing in itself.” By holding instead that the world “as it appears to us” is not the way it is “in itself,” we can hold that “the sum of all appearances” is not a completed “totality” and hence is neither infinite nor finite. For Kant, the opposition between those statements has now become merely “dialectical,” and both statements may be false without contradiction.5

1.3  Hegel’s idealism In presenting his “Absolute Idealism,” Hegel distinguishes aspects of Kant’s views he accepts from those he rejects. In particular, for Hegel, while Kant’s discussion of the antinomies is “one of the most important and profound advances of the philosophy of modern times” (1830, 92), Kant “did not penetrate to the cognition of [their] true and positive significance” (1830, 93). For Hegel, Kant rightly used the antinomies to demonstrate that our categories, as they stand, are not adequate to reality “in itself ” and thus give us only “appearance.” However, like many other philosophers, Hegel holds that there is a “supreme inconsistency” (1830, 105) in Kant’s view that we cannot know things “as they are in themselves”; for in claiming this, Kant appears to have some knowledge of that about which he claims we can know nothing.6 Hence, for Hegel, the proper response to the antinomies is not to rest content with our categories as they stand—to leave them “uncontested” (1830, 106) so that reality “in itself ” “remains an inaccessible beyond” (1830, 108)—but rather to develop them “dialectically” until we attain the “higher land” (ibid.) of absolute idealism in which the real is identical with fully developed or “sublated” thought. Hence, Hegel’s dictum that “the real is the rational and the rational is the real” (1830, §6): fully developed rational thought—thought that is the outcome of completed dialectic—is identical with the real. Here, the Kantian view that there is an insurmountable barrier separating cognition, on the one side, from reality “in itself,” on the other, is overcome; here, our categories are not left “uncontested,” but are rather “sublated,” thereby enabling us to reach “the Absolute.”

2  Three British idealists: McTaggart, Bradley, and Ward As an undergraduate at Cambridge from 1890 to 1894, Russell studied mathematics in his first three years, and philosophy in the fourth. However,

22

The Bloomsbury Companion to Bertrand Russell

shortly after arriving there, Russell befriended J. E. M. McTaggart, who had entered Cambridge in 1885, was elected a Prize Fellow in 1891, and had, according to Russell, already established a “metaphysical reputation” (Auto I, 63). Hence, McTaggart influenced Russell before his formal study of philosophy began. Russell writes of McTaggart that he “dominated the philosophical outlook of my generation at Cambridge” (Papers 10, 24), that his “wit recommended his Hegelian philosophy,” and that “[h]e taught me to consider British empiricism ‘crude’,” so that “I was willing to believe that Hegel (and in a lesser degree Kant) had a profundity not to be found in Locke, Berkeley, and Hume, or in my former pope, Mill7” (MMD, 10). In his fourth year, Russell’s teachers in philosophy were Henry Sidgwick, James Ward, and G. F. Stout. Given McTaggart’s influence, Russell regarded Sidgwick as “represent[ing] the British point of view, which I believed myself to have seen through” (ibid.), so that Ward, a “Kantian,” and Stout, a “Hegelian,” played a more central role in his education (MPD, 38). Ward, who functioned as Russell’s de facto supervisor, led Russell to read Kant, but was not himself a doctrinaire Kantian. Stout was “Hegelian” only in virtue of his praise for F. H. Bradley, so that Russell’s understanding of what it is to be “Hegelian” was derived primarily from “reading Bradley and talking with McTaggart” (Papers 5, 434), not from any detailed study of Hegel. In fact, he writes that one of the “various things [that] caused me to abandon” (MMD, 11) idealism was that “[i]n a rash moment, . . . I turned from the disciples to the Master and found in Hegel himself a farrago of confusions” (Papers 11, 99). I outline here some features of the views of McTaggart, Bradley, and Ward— three British idealists who played a formative role in Russell’s early philosophical development.

2.1  McTaggart on dialectic and the absolute Among McTaggart’s writings are three books of Hegel exegesis in which he both interprets Hegel and defends Hegel, so interpreted. The first of these—Studies in the Hegelian Dialectic (1896)—based on his 1891 dissertation, is the most relevant here. According to McTaggart, “dialectic” is generated by contradictions that arise from “the instability of all finite categories, due to their imperfect nature” (1896, 4). We begin with a category but find that predicating it of a subject also requires (“by consistency”) predicating the contrary of that category of the

Russell and Idealism

23

subject, thereby “violat[ing] the law of contradiction”; however, we also find that the given category and its contrary are “capable of reconciliation in a higher category,” in which they are “absorbed . . . as moments or aspects of which they can exist without contradiction” (1896, 1). However, that “higher category” will, in general, be vulnerable to a similar contradiction, thus requiring us to ascend to a yet higher category. For McTaggart, we cannot “rest anywhere in this alternate production and removal of contradictions until we reach the end of the ladder of categories” (1896, 2). Following Hegel, McTaggart holds that “the real is always the completely rational” (1896, 3); that the reason “finite categories” are “imperfect” is that they are abstractions and hence “one-sided” (ibid.); and that dialectic begins with the most abstract of categories—namely, “pure Being” (1896, 91)—and proceeds until it reaches the highest category—the Absolute Idea, which is “completely concrete” (1896, 3). Moreover, for McTaggart (and Hegel as McTaggart interprets him), the Absolute is “Spirit” which is “made up of spirits.” Thus, for McTaggart, an atheist, the Absolute is not an infinite Spirit above and beyond finite spirits, but is rather a “whole” whose parts are finite “spirits.” In particular, for McTaggart, each finite spirit is itself “a perfect individual” that “expresses the whole nature of Spirit” (1893, 215), so that “the Absolute . . . consist[s] of individuals bound together by a unity closer than any which we know at present” (1893, 216).

2.2  Bradley on experience, relations, thought, and the absolute Unlike McTaggart, Bradley presents himself as rejecting fundamental Hegelian doctrines, including Hegel’s “main principle” (1883a, x; 1883b, 590–91), that reality is the rational. For Bradley, since thought is inherently defective, overcoming that defect, thereby becoming one with the Absolute, requires transcending thought (and hence reason). Likewise, since the “dialectical method” is a process of reason, it too is “defective” (1883b, 601–02) and fails to attain the Absolute. Further, as against McTaggart’s (and, argues McTaggart, Hegel’s) view that the Absolute is a whole whose parts are “perfect individuals,” Bradley holds that “only the Absolute” itself “is individual or perfect” (1893, 217). In Russell’s terminology, while McTaggart’s view of the Absolute is “monadistic,” Bradley’s is “monistic.” Despite Bradley’s departures from Hegel, Russell (in 1924) presents him as accepting the characteristic view of “Hegelianism” that “logic alone can tell us a

24

The Bloomsbury Companion to Bertrand Russell

great deal about the real world” and as arguing, more specifically, that since “the real world . . . must be logically self-consistent” while the “the world as it seems to be is self-contradictory,” then “the world as it seems to be”—including being in time and space, and containing various interrelated entities, including distinct selves and material things—is “illusory” and radically different from “the real world” (Papers 9, 452). Further, while rejecting the “dialectical method,” Bradley accepts a developmental account of “experience,” according to which it has three stages: the first—“feeling” or “immediate experience”—leads to the second— “relational experience,” which includes thought and judgment—which, in turn, aims at the third—“supra-relational experience” or “the Absolute.” For Bradley, “immediate experience” is a “non-relational” unity and thus does not consist of distinct entities—including “subject” and “object,” or “self ” and “not-self ”—standing in relation to one another. For Bradley, “if .  .  . we go to experience, what we certainly do not find is a subject or an object, or indeed any other thing whatever, standing separate and on its own bottom” (1893, 128; here, as elsewhere emphasis is in the original), so that “in feeling . . . self and notself are not yet present and opposed” (1914, 247). For Bradley, while “feeling” is not itself “self-contradictory” (1999a, 358), once we become “reflective,” we go beyond it to the “relational” stage of experience, in which the “discursive understanding” supposes that distinct, independent entities stand in various relations to one another (1999a, 362). And while we are “fully justified” in moving from “immediate” to “relational” experience, the “relational stage” is “in the end a self-contradictory makeshift” (1999a, 356). In particular, employing what is now called “Bradley’s regress,” he argues that there can be no genuinely “relational unities”—unities composed of distinct, independent entities.8 For Bradley, if a unity were composed of distinct, independent entities, A and B, there would have to be a further entity—a relation R—to bind those entities into a unified whole. However, if R itself is simply one further distinct, independent entity, there would need to be a further relation R* to bind R to A and B, and so on to infinity. On the other hand, if we hold that when R connects A and B into a unified whole, R, A, and B are interdependent, then we have, as Bradley writes, “fall[en] back upon the unity which is not relational” (1999a, 271) and so “haven’t analyzed at all” (1999a, 268). For Bradley, that is, where a relation is independent of the terms it is to relate, there is no genuine “unity” composed of those terms and that relation; and where there is a genuine “unity,” it does not consist of distinct, independent parts.

Russell and Idealism

25

Bradley’s anti-Hegelian view that thought is aiming at a “mode of apprehension” that is no longer thought depends on his theory of judgment. For Bradley, “in judgment an idea is predicated of a reality” that “appears in presentation” (1893, 144; 1883a, 50). The “reality” of which the “idea” is predicated is the judgment’s “real subject” (1883a, 22, 50); it is “individual,” “substantial,” and “self-existent”; it is the entity that the judgment is about. The “idea” predicated of that “reality” is “not a mental image,” but rather an “ideal content” or “universal meaning”; it is “universal,” “adjectival,” and “symbolical” (1883a, 43–44). In Bradley’s terminology in Appearance and Reality, the subject of a given judgment is the “the ‘that’” while the “idea” predicated of that subject is “the ‘what’” (1893, 148–49). Further, for Bradley, to make a judgment is to venture a claim on reality; and where a genuine claim has been ventured, there is always the possibility of error. Hence, for Bradley, no judgment can be tautologous—say, of the form “A is A”—for then there is “no judgment at all . . . for, while professing to say something, it really says nothing” (1883a, 141). More generally, for Bradley, no judgment is “in principle infallible” or has “warrant” that is “immediate.” In determining what we should believe, “the test which we do apply, and which we must apply, is that of system (1914, 202). The view that knowledge requires infallible “foundations” is, he argues, “a false doctrine due to a misleading metaphor” (1914, 209). The way Bradley underpins his view that no judgment is infallible is by holding that judgment requires a “discrepancy” between the “that” and the “what” (1893, 150). In particular, for Bradley (like Leibniz), if we succeeded in fully identifying the individual that a given judgment is about, we would thereby know of any given “what” whether or not it is true of that individual; but, in that case, there would no longer be the possibility of error essential to judgment. Thus, for Bradley, essential to judgment is a defect—that in each judgment, we fail to determine one individual uniquely as the “reality” that judgment is about. Given this defect, no judgment is unrevisable; for, with a fuller specification of the subject of a given judgment, we may come to retract that judgment. Similarly, for Bradley, no judgment is “absolutely true” or “absolutely false”: since it fails to specify uniquely the individual it is about, each judgment is, in Bradley’s terminology, “ambiguous,” and hence may be true of one of the possible individuals that it may be regarded as being about and false of another (1914, 252, 261–63).9 Moreover, given this defect, thought seeks its “happy suicide”; for while we seek, in judgment, to attain “uniqueness” in specifying the subject of the

26

The Bloomsbury Companion to Bertrand Russell

judgment, were we to do so, “predicate and subject” would be “merged” (1893, 152), and we would have attained a “mode of apprehension” that is no longer “relational” or “discursive” (1893, 150) but is instead “immediate like feeling,” albeit at a level above rather than “below distinction and relation” (1893, 213). For Bradley, the “mode of apprehension” we thus seek is “identical with reality” (1893, 151–52); it is the Absolute. It is the stage of experience at which “spirit” has been aiming from the start, so that while Bradley denies that the real is the rational, he holds that “the more that anything is spiritual, so much the more it is veritably real” (1893, 489).

2.3  Ward on method, experience, and reality Just as Bradley is not an orthodox Hegelian, neither is Ward an orthodox Kantian. Thus, for example, like Hegelians, he rejects Kant’s distinctions between “phenomena and things per se,” sensibility and understanding, and matter and form (1922, 57); however, with regard to philosophical method, Ward regards himself as following Kant, not Hegel or Hegelians—in particular, not Bradley.10 According to Ward, Hegelians begin “from above” (1919, 15)11 by accepting first principles, which serve as “a royal road [to] demonstrate a priori what is the reality and what the meaning that underlie appearances” (1904, 4); however, by doing so, they find themselves in the position that “Kant abhorred and stigmatized as philosophical rhapsodizing” (1919, 23)—namely, of accepting a view of ultimate reality without any coherent account as to how it relates to the world of “appearances” in which we live. In contrast, for Ward, we should “begin from where we are, i.e., in mediis rebus” (1919, 3; 1882, 370–71; 1911, 442)12 and then “advance continuously,” so that our knowledge “grow[s] from within”; the “whole procedure will be tentative” and we will never attain a final or “complete” system (1919, 23–24). For Ward, we should follow Kant in regarding the problems of philosophy as “inverse problems” (1919, 2–6), in which the “golden rule” (1919, 23) is to make “partial advance from the more known” to “secure[ ] partial knowledge of the less known” (1919, 3).13 Thus, Ward claims that Bradley’s view of immediate experience as a “prerelational” unity in which “self and not-self are not yet present and opposed” is a “travesty of the facts” (1925, 15), derived not from “any indubitable fact actually found in immediate experience, but [from] a speculative conviction” (1925, 29) regarding the character of ultimate reality. For by regarding ultimate reality as a

Russell and Idealism

27

unity that transcends relations, Bradley has a motive for holding that “immediate experience” intimates at a level “below relations . . . the idea of a superior unity above them” (1925, 14–15, quoting from Bradley, 1893, 462). For Ward, once we set aside “ultimate conceptions to which metaphysical considerations may lead us” (1893, 64) and “start working continuously from our own self-consciousness,” we will “regard the duality of subject and object as the essential characteristic of experience” (1925, 31). In particular, for Ward, “presentation” is “the name for that relation between the object and its subject which is the indispensable condition of attention, or, as we usually say, of consciousness” (1882, 375), so that all consciousness is awareness by a “subject” of an “object” presented to it. In accord with his philosophical method, Ward’s 1899 argument for idealism proceeds not from “logical” first principles, but by discussing contemporary scientific theory, which he argues does not justify either the naturalism or agnosticism with which it is often associated. Citing Kirchhoff, Mach, and Boltzmann, among others, he argues that mechanics has become an abstract “mathematical science,” in which such notions as “matter,” “substance,” and “cause” play no role (1899a, Lecture II). Hence, he distinguishes the science of mechanics from the “mechanical theory of nature”—the view that given any state of the universe and the laws of nature, all subsequent states of the universe are thereby determined—and argues that far from supporting the “mechanical theory of nature,” the science of mechanics actually undermines it. For Ward, although causality plays no role in physics, causality, understood as teleologically directed, is needed to explain evolution so that “it is only in terms of mind that we can understand that unity, activity, and regularity that nature presents” (1899a, xii). Thus, for Ward, “[p]hilosophy cannot dispense with science nor can science, however complete, render philosophy superfluous” (1904, 4): while he begins his argument for idealism from current scientific theory, he concludes that science by itself is incapable of revealing the ultimate character of reality. Ward’s view of experience is central to his argument for idealism. For he holds that each experience not only involves a relation between subject and object but also is an “organic unity” (see, for example, 1899b, 152–53, 178)—in which case, “subject” and “object” are of the same metaphysical nature.14 Since Ward takes his arguments against the “mechanical theory of nature” as establishing the role of teleology—and hence mind or “spirit”—in nature, he concludes that “object” and “subject” are both “spiritual” in nature. However, he does not thereby resolve the issue dividing Bradley and McTaggart as to whether reality consists of a single “spiritual” individual or different “spiritual” individuals,

28

The Bloomsbury Companion to Bertrand Russell

thereby acknowledging his inability to reach a final metaphysical conclusion. As he writes to William James in 1899: “I have long felt that there will always remain something hopelessly insoluble about the relation of the one and the many. . . . My sympathy is so far with the many. . . . But again it seems absurd to regard God as merely one of ourselves, the Supreme One among the many” (Perry, 1935, 647). By 1911, while still “disavow[ing] any attempt completely to solve the so-called riddle of the universe” (1911, 430), Ward’s sympathies have moved toward monism. He argues that while, at the outset—that is, in mediis rebus (1911, 442)—we find “many minds in mutual interaction” (1911, v), we are led to conceive a more fundamental monistic and theistic standpoint, as least as “an ideal,” that following Kant, Ward claims “may claim our faith though it transcend our knowledge” (1911, vi).

3  Russell as idealist Russell’s philosophy during his idealist period reflects, in different ways, the influences of Bradley, McTaggart, and Ward. From Bradley (through Stout), Russell learned the argument that was decisive in his “becom[ing] a Hegelian.” Like McTaggart, he regards “dialectic” as the fundamental method of philosophy; however, the way he carries out “dialectic” involves an engagement, akin to Ward’s, with the details of ongoing science. I outline here three aspects of Russell’s idealist philosophy: how Bradley’s theory of judgment figures in Russell’s “becom[ing] a Hegelian”; Russell’s view of the “dialectic of the sciences”; and “the contradiction of relativity” that he comes to emphasize.

3.1  Russell’s Hegelianism and Bradley’s ontological argument15 In recalling “the precise moment” when he “bec[a]me a Hegelian,” Russell writes: [O]ne day in 1894, . . . I saw in a flash (or thought I saw) that the ontological argument is valid. I had gone out to buy a tin of tobacco; on my way back, I suddenly threw it up in the air, and exclaimed as I caught it: “Great Scott, the ontological argument is sound.” I read Bradley at this time with avidity, and admired him more than any other recent philosopher. (MMD, 10)

The “ontological argument” to which Russell here refers is not the argument, in its classical form used to prove God’s existence, but rather an argument derived from Bradley’s theory of judgment. Put in terms of Bradley’s distinction between the

Russell and Idealism

29

“that” and the “what,” classical versions of the argument attempt to show that there is a “what”—namely, the “universal meaning” expressed by the word “God”—from which it can be proved that there is an existent thing that it is true of. For Bradley, since no “what” is ever absolutely true of reality, there is a “real failure” in the ontological proof as traditionally presented (1893, 350). However, by holding that all judgment (and thought) involves both a “that”—an existing individual that the judgment (thought) is about—and a “what”—a “universal meaning” ascribed to that individual—Bradley recognizes an “irrefragable” principle “underlying” both the ontological argument and the cosmological argument (ibid.). For Bradley, in each thought, we “refer” a “what” (however “vaguely”) to an existent “that”; hence, as long as we think at all, we will be thinking about an existent “that” and will be ascribing some “what” to that “that.” In that case, it is incoherent so suppose either that a given “what” is “wholly un-referred” to an existent or that a given “that” is wholly unqualified by any “what.” And for Bradley, recognizing that each thought has two “sides”—a “that” and a “what”— underlies both the “ontological argument”—which moves from a “what” to an existent that it characterizes—and the “cosmological argument”—which moves from an existent to its “character.” It is to this Bradleyan version of the “ontological argument” that Russell alludes in describing “the precise moment” when he became a “Hegelian.” Thus, in a (May 1894) term paper for Stout, in which he criticizes Descartes’s ontological argument, Russell writes: But those who wish to maintain the ontological argument may reply that whatever we think we cannot get away from reality; if we judge at all, we must affirm some predicate of reality. . . . We must think the Absolute . . . ; hence the ontological argument can, it would seem, only be met by complete skepticism, by abstaining from judgment altogether; which is a negligible alternative. (Papers 1, 179)

For Russell, that is, by holding, with Bradley, that “if we judge at all, we must affirm some predicate of reality,” it will be incoherent to deny “reality as a whole,” in which case the only way to avoid committing ourselves to “reality as a whole”—that is, the Absolute—is to avoid judging (or thinking) at all. After rejecting idealism, Russell characterizes Bradley as accepting a version of “the existential theory of judgment”—according to which for every judgment there is an existent which that judgment is about16—and as developing an argument that “partakes of both the Ontological and Cosmological arguments” that is based on that view of judgment (PL, 177).

30

The Bloomsbury Companion to Bertrand Russell

3.2  The idealist Russell and dialectic Although Russell became a “Hegelian” by accepting an “ontological argument” derived from Bradley, his idealist writings focus primarily not on such Bradleyan themes as the nature of experience, judgment, or the Absolute, but rather on what he calls the “dialectic of the sciences.” In regarding “dialectic” as his overarching project, Russell is more like McTaggart than Bradley; but by engaging in the details of “the sciences” in carrying out his “dialectic”, he is more like Ward than either McTaggart or Bradley. For Russell, since each science “works with” fewer “fundamental ideas” than all the fundamental ideas, it contains contradictions that “unavoidably result” from its “incompleteness” and is thus “condemnable” if it is regarded “as independent and self-subsistent knowledge” (Papers 2, 5). Hence, we are led to “pass outside” the given science to a new science, which will likewise have unavoidable contradictions if it is incomplete. For Russell, the “transition” that results from resolving an “unavoidable contradiction” in a more “abstract science” by “pass[ing]” to a “new”—less abstract—science “is dialectical in the true Hegelian sense, and shews that the notion of the science in question is fundamentally selfcontradictory, and must be throughout replaced by another, in any metaphysical construction of the real” (ibid.). Thus, Russell’s dialectic, like McTaggart’s, proceeds by identifying contradictions that arise from the “incompleteness” and “abstractness” of a given body of thought, so that reaching “the real” requires attaining a state in which our “ideas” are no longer contradictory because they are no longer “incomplete” or “abstract.” However, while Russell’s view of “dialectic” is broadly in accord with McTaggart’s, the way he implements dialectic reflects Ward’s methodology (even though, as an anti-Hegelian, Ward does not regard himself as engaging in “dialectic”). Thus, he writes: It seems possible . . . to obtain a Dialectic having closer relation to Appearance than that of pure Logic. . . . We might thus find a method of turning Appearance into Reality, instead of first constructing Reality and then being confronted by a hopeless dualism, (Papers 2, 24)

thus objecting, like Ward, to proceeding, in the style of Hegelians, from “pure Logic” to a view of “Reality” that is radically divorced from “Appearance.” Accordingly, Russell’s dialectic proceeds, not as McTaggart’s does, in “logical” steps from the most abstract category—pure Being—to the Absolute, but rather by focusing on the details of specific sciences (so that his idealist writings are no

Russell and Idealism

31

less technical than his later writings), identifying “unavoidable contradictions” in them, and resolving those contradictions dialectically. Thus, although Russell’s work on geometry—a topic suggested by Ward, who “instructed” Russell “in the Bibliography of the subject” and recommended that he consult with Whitehead regarding technical issues (Papers 1, 246)—was “mainly Kantian,” it also has a “Hegelian” aspect. On the “Kantian” side, Russell argues that by addressing what, in Ward’s terminology, is an “inverse problem” as to what is “necessary to experience,” he can derive “a priori properties of space”(EFG, 2-5), although not the same a priori properties Kant derives.17 However, he also argues, in a “Hegelian” manner, that those a priori properties of space involve “inevitable” contradictions that can be avoided only by introducing matter in the form of unextended atoms, and hence by a dialectical transition from geometry to physics. Because the matter thus introduced involves no causal properties, the transition is to kinematics, not dynamics; but in the final paragraph of EFG, Russell indicates that “new contradictions” in this kinematic notion of matter “would demand a fresh treatise, leading us . . . into the domain [ ] of Dynamics” (EFG, 201). Further, in some writings, Russell indicates that an “antinomy” facing dynamics (whose fundamental notion is force) requires a dialectical transition to psychology (whose fundamental notion, Russell holds, with Ward, is “conation” or volition) (Papers 2, 16, 18) and, with it, the idealist view that ultimate reality is mental in nature. However, the idealist Russell does not appear to have a settled view either of the proper dialectical order or—what is related—on the issue of monism versus monadism. Thus, in some writings, he indicates there is a “true Hegelian” dialectical transition from “continua” to “discreta” (Papers 2, 5, 18), and hence to a version of monadism; but in others, he indicates that there is “a dialectical transition from punctual matter to the plenum” (Papers 2, 22), thereby supporting a version of monism. In 1897, in discussing “the choice between Monadism and Monism” (Papers 2, 95) as it arises in the philosophy of space and time, Russell concludes without reaching a verdict: “To decide between these alternatives would be to solve the most fundamental question of metaphysics; but to state the alternatives may surely be not unimportant” (Papers 2, 97).18

3.3  The contradiction of relativity By 1898, Russell focuses on what he calls the “contradiction of relativity,” which helps “to define the realm of Mathematics” (Papers 2, 166). The contradiction

32

The Bloomsbury Companion to Bertrand Russell

arises because Russell holds that “the foremost class of judgments, from every point of view, is the class in which a predicate is asserted of a subject” (Papers 2, 167), so that a relation between two entities, A and B, should be grounded in intrinsic (and hence non-relational) properties of A and B. However, he also holds that asymmetric relations (such as greater than, earlier than, and to the right of)—which, he writes, “pervade almost the whole of Mathematics, since they are involved in number, in order, in quantity, and in space and time” (Papers 2, 226)—cannot be so grounded. For Russell, because those relations both should be, but cannot be, grounded in intrinsic properties of the entities related, an “unavoidable contradiction” arises that “infects all of Mathematics” (Papers 2, 328, fn. 5).19 More specifically, for the idealist Russell, the only relations that can be grounded in intrinsic properties of the entities related are symmetric relations— either symmetric transitive relations of “identity of content” or symmetric nontransitive relations of “diversity of content.” Thus, if A and B are each red, then their each having the intrinsic property of being red grounds the symmetric transitive relation of having the same color as obtaining between them. And if A is red while B is green, then their having, respectively, the intrinsic properties of being red and being green, grounds the symmetric non-transitive relation of having different colors as obtaining between them (see Papers 2, 224). However, for Russell, an asymmetric relation between entities—say, A’s being taller than B—cannot be similarly grounded in intrinsic properties of the related entities. It might seem that A and B have intrinsic height-properties in virtue of which one is taller than the other, so that, for example, A’s being 6’ tall and B’s being is 5’ tall would ground A’s being taller than B. However, for Russell, that A and B have these height-properties does not, by itself, ground A’s being taller than B; only together with 6’ being greater than 5’—that is, only with the asymmetric relation of being greater than relating 6’ to 5’—will they do so. In that case, we are still left with an ungrounded asymmetrical relation between two entities and have thus failed to ground A’s being taller than B merely from intrinsic properties of A and B.20 Hence, for the idealist Russell, while any asymmetric relation relating two entities should be grounded in intrinsic properties of the related entities, it cannot be so grounded. Russell concludes that in such cases, we have the “contradiction of relativity,” in which we have “a conception of difference without a difference of conception” (Papers 2, 132). We suppose that there should be “intrinsic adjectives” of A and B in virtue of which A is taller than B—that is, we have a

Russell and Idealism

33

“conception of difference” between them grounding that relation. However, there can be no such “intrinsic adjectives”—that is, no “difference of conception,” no qualitative difference between them—that grounds it. Since he denies that there are “intrinsic adjectives” of height (or, more generally, “intrinsic” magnitudes) grounding comparative relations of height (or magnitude), the idealist Russell holds that the true judgment of magnitude is not: “this has such and such a magnitude,” but “this is equal to, or greater than, or less than that,” (Papers 2, 328)

and regards magnitudes (such as heights) as “invented, as intrinsic conceptions, to rationalize” comparative relations between quantities (Papers 2, 133). However, he also holds that even as “invented,” heights (or magnitudes) do not succeed in “rationalizing” the comparative relations, since there are no “intrinsic conceptions”—invented or otherwise—that can ground asymmetric comparative relations, and hence “rationalize,” the relations. Hence, the “contradiction of relativity” remains.

4  Russell’s Moorean philosophy21 I outline here some central elements—in metaphysics and epistemology, as well as regarding language and philosophical methodology—of the “Moorean” philosophy Russell accepts in rejecting idealism and then discuss how accepting the Moorean philosophy enables Russell to avoid the “contradiction of relativity” and defend what he calls “absolute” theories of order, including absolute theories of time, magnitude, and number.

4.1  Russell’s post-idealist metaphysics: Simples, propositions, and relations Like Hegelians, but unlike Ward, the Moorean Russell confidently characterizes the ultimate nature of reality; however, he disagrees with Hegelians as to what it is. In particular, as against Bradley’s monism, the Moorean Russell holds that reality consists of distinct entities in various relations to one another. Further, Russell rejects the view—common to monistic and monadistic idealists—that wholes are “organic unities” in which there is a mutual dependence between a

34

The Bloomsbury Companion to Bertrand Russell

whole and its parts. Instead, for Russell, the being of a whole depends on the being of its parts but not vice versa, and each whole consists, ultimately, of simples—entities that have no parts—the “ultimate constituents of the universe” (Papers 3, 160–61, see also 51–52). Moreover, as against idealism, according to which everything (or at least everything known) is “mental” in nature, Russell holds (and claims to know) that while some of the “ultimate constituents of the universe,” namely minds, are mental, some are physical, and others, including “[n]umbers, the Homeric gods, relations, chimeras, and four-dimensional spaces” (Papers 3, 276), have “being” but not existence and are thus neither mental nor physical. Besides countenancing simples, the Moorean Russell recognizes two sorts of complex wholes: “aggregates” (or classes), whose “parts have no direct connection inter se,” and “unities” (or propositions), whose parts have such a “direct connection” (Papers 3, 36–37). Thus, for Russell, the relation preceding functions differently in the aggregate whose constituents are A, B, and preceding and in the proposition “A precedes B”: in the aggregate, it functions merely as a relation “in itself ”; in the proposition, it functions “as relating”—specifically, relating from A to B. A Russellian proposition is just the sort of complex unity that Bradley argues is a “self-contradictory makeshift.” Bradley poses a dilemma: either preceding is just one more independent entity, in which case it cannot actually relate A and B (so that it stands in need of a further relation relating it to A and B, launching the regress); or it connects A and B into a unified whole, in which case they are interdependent, so that that we have “fallen back upon the unity which is not relational.” However, by holding that relations have a “twofold use,” Russell rejects the dilemma. For Russell, preceding is exactly the same entity when it occurs “in itself ” and when it occurs “as relating” in different propositions: when it occurs “in itself,” it does not relate; when it occurs “as relating,” it not only relates (thereby cutting off the regress) but also does so while remaining independent of the entities that it relates, so that “A precedes B” is analyzable into three distinct and independent entities.22 More generally, Russell’s view of propositions is opposed, in a number of ways, to Bradley’s account of judgment. As Russell writes to Bradley (in 1900): “The effort of Mr. Moore’s philosophy, which I followed, was to endeavor wholly to deny the dualism of that and what” (Bradley, 1999b, 180). First, on Bradley’s “dualism,” since the “that” a given judgment is about is “individual” and “substantial,” while its “what” is “universal” and “adjectival,” no

Russell and Idealism

35

“what” can be the subject of a judgment. In contrast, for Russell, since “verbs” have a “twofold use,” every constituent of a proposition is capable of functioning as “logical subject.” Although relations occurring “as relating” do not function as “logical subjects,” the same relations occurring “in themselves” do so function; hence, while “A precedes B” is not about preceding, “Preceding is a relation” is. And for Russell, since each constituent of any proposition can function as “subject,” then no constituent of any proposition is any less “individual” or “substantial” than any other constituent. Hence, as against Bradley: “[T]he theory that there are adjectives or attributes or ideal things, or whatever they may be called, which are in some way less substantial, less self-subsistent, less self-identical, than true substantives, appears to be wholly erroneous” (POM, 46).23 Further, given that each “what” is just as “substantial” as a “that,” Russell has “an easy and simple refutation”(Papers 3, 232) of the view, which he takes Bradley to exemplify, that “every proposition, true or false, consists in the ascription of a predicate to a subject and that this ascription is not a relation”: The objection is that the predicate is either something or nothing. If it is nothing, it cannot be ascribed, and the pretended proposition vanishes. If it is something, predication expresses a relation. (Papers 3, 252)

For Russell, that is, since predicates are just as “substantial” as the “subjects” to which they are attributed, then a proposition attributing a predicate to a subject is ultimately relational in form, where predication is the “relating relation.”24 Thus, the Moorean Russell does not merely hold that some propositions are ultimately relational in form; he holds that all are, so that “a proposition must contain two terms at least” along with “a relation between them” (Papers 3, 94).25 Moreover, for the Moorean Russell, since each entity can function as a logical subject and some entities have “being” but not existence, then—as against the “existential theory of judgment” that led him to become a “Hegelian”—some propositions are about entities that have “being” but not existence. In particular, for Russell, since numbers have being but not existence, philosophers— including idealists and empiricists—committed to the “existential theory of judgment” err in attempting to find “existents”—either mental or physical—for arithmetical propositions to be about.26 Accordingly, for the Moorean Russell, “perhaps the most important classification of which propositions are capable” is that between “existential” and “non-existential” propositions—between proposition about existents and those about entities that have being but not existence (see PL, 25, 29).

36

The Bloomsbury Companion to Bertrand Russell

Finally, whereas Bradley holds that in each judgment the “what” is “symbolical,” Russell denies that each proposition contains at least one representational or “symbolical” entity, such as an “idea” or word, among its constituents. For Russell, a unified complex composed of any entities, representational or not, just is a proposition; and since unified complexes do not, in general, contain representational entities, neither do propositions. Thus, for Russell, although each proposition is either true or false, it is not true or false in virtue of representing correctly or incorrectly how things are; instead, truth and falsity are primitive properties of unified complexes. A true proposition does not “correspond” to a fact; it just is a fact (Papers 4, 492). A false proposition is no less real—no less a unified complex of genuine entities—than a true proposition; it is what Russell comes to call a “false objective” (Papers 6, 119). Since “symbolical” entities are not essential to propositions, issues regarding representation—including issues as to whether representations are intrinsically “vague” or “indeterminate”—are irrelevant to the nature of truth and thus do not create an obstacle for Russell’s anti-Bradleyan view that each proposition (and, derivatively, each judgment) is either absolutely true or absolutely false.

4.2 Epistemology The central element of Russell’s Moorean epistemology is the relation he comes to call “acquaintance,” which he characterizes as “the simplest and most pervading aspect of experience” (TK, 5).27 As against Bradley’s view that in immediate experience, there is “as yet” no distinction between “object or subject” or “self and not-self,” Russell holds, like Ward, that each experience involves a relation—for Russell, the relation of acquaintance, whose converse, in accord with Ward’s terminology, is presentation—between a “subject” and “object.” However, whereas Ward holds that each “experience” is an “organic unity,” for Russell, each experience is a complex consisting of distinct and independent entities. Hence, whereas Ward regards “subject” and “object” in an experience as having the same nature, for Russell acquaintance is a “relation between a subject and an object which need not have any community of nature” (TK, 5). Thus, while a “subject” can be acquainted with its own mental acts, it can also be acquainted with physical existents28 and with entities that have being but not existence. For Russell, acquaintance enables us to have the kind of knowledge of mind-independent entities that idealists deny is possible. Whereas Berkeley

Russell and Idealism

37

holds that objects of the senses are mind-dependent entities whose esse est percipi, for Russell, the being of an object of acquaintance does not, in general, depend on its being experienced.29 Whereas Kant holds that while the “objects of sense” are mind-independent “things,” there are “forms of intuition” in us that structure how things are given to us, so that we cannot know things as they are “in themselves,” for Russell, acquaintance is a direct and unmediated cognitive relation between “subject” and “object.” Not only are there no mental representations—such as an “ideas” or images—between “subject” and “object”;30 neither is there any Kantian “form of intuition” structuring how we apprehend “objects” with which we are acquainted. Hence, for Russell, in being acquainted with an entity, we thereby know that entity as it is in itself. Moreover, whereas Kant accounts for (synthetic) a priori knowledge by forms of intuition and categories “in us” that structure and apply to things as they appear to us, for Russell, since we can be acquainted with entities that have being but not existence, we can have (synthetic) a priori knowledge of an objective mindindependent realm.31 Further, whereas Hegelians hold that identifying an object requires knowing every truth about it, Russell comes to hold that we can know an object by acquaintance “completely,” without, theoretically, having to know any truth about it.32 Further, in holding that we can be acquainted with complexes as well as simples, Russell accepts the sort of foundationalist epistemology that Bradley calls “a false doctrine due to a misleading metaphor.” For the Moorean Russell, all proof or justification depends ultimately on our “intuition” or “immediate perception” of—that is, our acquaintance with—“self-evident” propositions. In Russell’s terminology, these “self-evident” propositions are “indemonstrables”: they are accepted as true, not by “demonstrating” their truth from other propositions we accept as true, but rather by an act of “immediate perception.” Thus, an “indemonstrable” propositions have “no evidence except selfevidence” (PL, 166) and “no reason whatever can be given for their truth” (Papers 2, 163).33 The Moorean Russell accepts parallel foundationalist views of empirical and a priori knowledge: while the ultimate justification of empirical knowledge depends on immediate (sensory) perception of self-evident “existential propositions,” the justification of a priori knowledge depends ultimately only on immediate (non-sensory) perception of self-evident “non-existential propositions.” Moreover, for Russell, since “no existential proposition can be deduced from one which is not existential” (Papers 3, 504), then we can have no

38

The Bloomsbury Companion to Bertrand Russell

a priori knowledge regarding what exists, so that idealists are wrong to “attempt, from logical or a priori considerations, to deduce the nature of what exists” (Papers 3, 503).

4.3  Meaning, understanding, and analysis Besides holding that representations play no fundamental role in cognition, the Moorean Russell holds that representations—specifically, linguistic representations—create no in-principle barrier to our accessing reality as it is in itself. In particular, he accepts views of meaning and understanding, according to which in analysis we not only identify the ultimate constituents of the proposition expressed by a given sentence but can do so, moreover, merely by considering what we were already aware of in understanding that sentence prior to analysis. For the Moorean Russell, each word has a definite meaning; its meaning is the entity—simple or complex—that it stands for; a sentence has as its meaning the proposition it expresses; the constituents of the proposition expressed by a sentence are the entities that are the meanings of the words in that sentence; and understanding a sentence requires being acquainted with each constituent of the proposition it expresses.34 Given we can be acquainted with extramental entities without any intervening mental representations, Russell can hold that the proposition expressed by a sentence we understand includes among its constituents the entities in the world that sentence is about, so that unless a sentence is about words or mental entities, the proposition it expresses will not contain words or mental entities among its constituents.35 For Russell, where a word stands for a complex entity, “the term is philosophically defined by enumerating its simple constituents”; and where a word stands for a simple entity, “no philosophical definition is possible” and that entity is philosophically “indefinable” (Papers 2, 410). Hence, for Russell, analyzing proposition P expressed by sentence S1 involves identifying the simple constituents of P. In particular, it requires replacing each word in S1 that stands for a complex entity by words that stand for the simple constituents of that entity. Where analysis is complete, S1 will be transformed into S2, each of whose words stands for a simple (that is, ultimate) constituent of P. Thus, while S1 and S2 both express P, S2 is a fully transparent representation of the proposition they both express: there will be exactly as many words in S2 as there are ultimate constituents in P, so that S2, as opposed to S1, mirrors the ultimate constitution of P.

Russell and Idealism

39

Further, for the Moorean Russell, since understanding a sentence requires being acquainted with each constituent of the proposition expressed by that sentence, analysis makes explicit what we are already aware of, prior to analysis, when we understand the sentence initiating the analysis. That is, for Russell, where S1 expresses P, understanding S1 requires being acquainted with each constituent of P, so that even where analysis involves transforming S1 into a sentence S2 that transparently represents P, the analysis does no more than make explicit the simple entities with which one is already acquainted in understanding S1 prior to analysis.

4.4  The business of philosophy For the Moorean Russell, reality consists of simple entities related to one another in various ways, and our apprehension of reality derives, ultimately, from our acquaintance with simple entities, or indefinables, and with “self-evident” (and hence true) propositions, or “indemonstrables.” Accordingly, for the Moorean Russell, the primary philosophical activity—the “business of philosophy”—is simply to become acquainted with (or to “intuit” or “immediately perceive”) indefinables and indemonstrables. Thus, he writes: [T]he business of philosophy is just the discovery of those simple notions, and those primitive axioms, upon which any calculus or science must be based. . . . [A]n emphasis on results rather than premisses … is radically opposed to the true philosophic method. . . . An idea which can be defined, or a proposition which can be proved, is of only subordinate philosophical interest. The emphasis should be laid on the indefinable and indemonstrable, and here no method is available save intuition. (PL, 170–71)

Likewise, in POM, he writes: A certain body of indefinable entities and indemonstrable propositions must form the starting-point for any mathematical reasoning; and it is this startingpoint that concerns the philosopher. . . . All depends, in the end, upon immediate perception; and philosophical argument, strictly speaking, consists mainly of an endeavour to cause the reader to perceive what has been perceived by the author. (POM, 129–30)

Like Hegelians, the Moorean Russell holds that the central task of philosophy is to facilitate the apprehension of reality as it is in itself. However, for Hegelians, to do so—and hence to attain the kind of “immediacy” that Bradley holds

40

The Bloomsbury Companion to Bertrand Russell

constitutes the Absolute—we must first recognize by means of argument or “dialectic” that our present stage of consciousness—including both ordinary and scientific belief—contains unavoidable contradictions. For the Moorean Russell, in contrast, there is nothing inherently defective in our ordinary and scientific beliefs, so that the “immediate apprehension” the philosopher seeks is not something transcending ordinary and scientific belief but is rather the absolute “starting-point” of our ordinary and scientific cognition (the sort of “starting-point” or ultimate foundations that Ward denies we can ever reach). Here, no argument is needed before we are in a position to apprehend reality as it is in itself: it is already available to us, if we simply attend to what is “selfevident” and to what our words, as ordinarily used, mean.

4.5  The Moorean Russell on time, magnitude, and number A—and plausibly the—decisive event in Russell’s rejection of idealism occurs when he brings the Moorean philosophy to bear on the issues that had previously led him to recognize the “contradiction of relativity.”36 As an idealist, Russell held that asymmetric relations should be, but cannot be, grounded in intrinsic properties of the entities they relate, and concluded that such relations, which pervade the mathematical sciences, are infected with “the contradiction of relativity.” Given his view that any limited science— including the mathematical sciences—contains “unavoidable contradictions,” this conclusion is not problematic. In breaking with idealism, Russell continues to hold that asymmetric relations cannot be grounded in intrinsic properties of the entities they relate;37 however, he now holds that “contradictions should cease to be regarded as commending a theory” (Papers 3, 783) and holds that all propositions are ultimately relational in form. Accordingly, he no longer regards the irreducibility of asymmetric relations to intrinsic properties of entities they relate as establishing an unavoidable contradiction infecting the mathematical sciences; instead, he regards it as providing a reductio of the assumption—which he labels “the axiom of internal relations”38—that all relations should be grounded in intrinsic properties of the entities they relate.39 However, for the Moorean Russell, once it is recognized that A’s being taller than B, for example, cannot be grounded in intrinsic properties of A and B, the question remains as to whether it is grounded at all. For by holding that all propositions are relational, the Moorean Russell holds not only that some

Russell and Idealism

41

relations are not (and should not be) grounded but also that the only way to ground a relation is (ultimately) in terms of other (ungrounded) relations. In particular, for the Moorean Russell, there are two competing theories of height (which is one sort of magnitude). On the relative theory, there are physical quantities along with three indefinable—that is, ungrounded— comparative relations of height that can obtain between them: the asymmetric transitive relations taller and shorter, and the symmetric transitive relation equal in height. This is the theory that the idealist Russell was driven to accept, when he writes: [T]he true judgment of magnitude is not: “this has such and such a magnitude,” but “this is equal to, or greater than, or less than that,” (Papers 2, 328)

thereby regarding magnitudes (such as heights) as “invented, as intrinsic conceptions, to rationalize” comparative relations between quantities (Papers 2, 133). On the absolute theory, there are indefinable heights in addition to quantities. That is, heights are not merely “invented,” but are rather among the ultimate constituents of the universe. Nor, on this theory, are heights “intrinsic properties” of quantities; rather, they constitute an “independent series” through which quantities are ordered “by correlation” (see Papers 3, 291). More specifically, on this theory, there are three indefinable (ungrounded) relations: a (many-one) relation of having that relates a quantity to its height; and two (asymmetric, transitive) relations of greater and less that can obtain between heights (not quantities). Moreover, on this theory, these three relations help define, and thus ground, the comparative relations of height between quantities. Thus, to say that A is taller than B is, when fully analyzed, really to say that the height that A has is greater than the height that B has; and to say that A is equal in height to B is really to say that A and B have the same height. On each theory, there are indefinable relations, including indefinable asymmetric relations. Hence, for the idealist Russell, both theories face the “contradiction of relativity.” In contrast, for the Moorean Russell, who no longer holds that all relations should be grounded, each theory is coherent; but the question remains as to which is correct. For the Moorean Russell, deciding between the two sorts of theories is a paradigm example of the sort of issue it is “the business of philosophy” to address;40 for it amounts to deciding what are the indefinables (and indemonstrables41) with regard to height. Further, in accord with his methodology, according to which fundamental philosophical issues are

42

The Bloomsbury Companion to Bertrand Russell

to be decided by “insight” and “perception,” not argument, the Moorean Russell holds that the issue can be decided by “mere inspection.” One way Russell focuses the issue is by considering propositions expressed by sentences of the form “A is equal in height to B.” Do they include among their constituents an indefinable relation equal in height? Or is that relation definable in terms of the relation of having, so that what is really said is that A and B bear the relation of having to the same height? For the Moorean Russell, since understanding such a sentence requires being acquainted with the constituents of the proposition it expresses, mere reflection on what we are acquainted with in understanding it suffices for determining which theory is correct. Hence, he writes that if we “consider what we mean when we say that two quantities are equal” in a given respect (say, height), we will recognize that “what we mean” is that those two quantities have a “common property” (Papers 3, 58), where that “property” is itself “a third term”—in this case, an indefinable height—“to which both have the same relation”—in this case, the relation of having (Papers 3, 94). For the Moorean Russell, that is, “mere inspection” of what we mean when we say that two quantities are equal in height enables us to recognize that the absolute theory of height is correct. Likewise, for the Moorean Russell, other absolute theories of order—including absolute theories of time and number—are also obviously correct. Thus, instead of the relative theory of time—that there are only events and indefinable relations before, after, and simultaneous between them—he accepts the absolute theory— that indefinable moments constitute an “independent series” by “correlation” to which events are temporally ordered. And instead of the “relative” theory of number—that there are only classes and indefinable relations greater in number, lesser in number, and equal in number between them—he accepts the “absolute” theory—according to which indefinable numbers constitute an “independent series” by “correlation” to which classes are numerically compared. In each case, Russell indicates that “inspection” alone suffices for recognizing that the absolute theory is correct. Thus, he writes that “a direct consideration of the question” reveals that simultaneous events have something “in common beyond the common qualities of all events” (Papers 3, 227)—namely, the moment at which they occur; and that “equality [in number] plainly consists in possession of the same number,” so that “the absolute theory [of number] is plainly correct” (Papers 3, 225). As an idealist Russell, took it as obvious that A’s being taller than B should be grounded; but, by accepting the “axiom of internal relations,” he was forced

Russell and Idealism

43

to deny that it can be grounded and thereby accepted a “relative” theory of magnitude along with the “contradiction of relativity.” Hence, once he rejects that “axiom”—and, more generally, rejects a “dogmatically assumed scholastic logic” that privileges subject-predicate propositions (construed as non-relational)— he can avoid that contradiction and accept “common sense” absolute theories of order (see Papers 3, 232–33). Accordingly, in retrospectively discussing his Moorean philosophy, Russell writes that whereas “Bradley argued that everything common sense believes in is mere appearance,” Moore and he “reverted to the opposite extreme, and thought that everything is real that common sense, uninfluenced by philosophy or theology, supposes real” (MMD,  12), so that whereas Hegelians “convicted” “number, space, time, matter . . . of being selfcontradictory,” he embraced “a very full universe,” including “all the numbers sitting in a row in a Platonic heaven” as well as the “points of space and instants of time” (MPD, 62).

5  Russell’s post-Peano philosophy and idealism Although Russell characterizes his attending the Paris Congress in August 1900 as “the most important event” in “the most important year in [his] intellectual life”—and hence, by implication, as more important than his break with idealism almost two years earlier—following the Congress, he does not announce wholesale changes in fundamental philosophy on a par with those he presents in rejecting idealism. Instead, the new developments in his view—embracing the theories of Weierstrass, Dedekind, and, especially, Cantor regarding infinity and continuity; becoming a logicist; discovering Russell’s paradox—may seem, at least, to be more narrow and technical in nature. But Russell’s post-Peano engagement with mathematical logic is not a merely technical development in his philosophy. I discuss here some ways in which it undermines central elements of his Moorean philosophy and how these changes in Russell’s postPeano philosophy affect his critique of idealism.

5.1  Set theory and justification In accepting Cantor’s theory of the transfinite, the post-Peano Russell rejects “common sense” and apparently “self-evident” principles—such as that adding one to a number always yields a greater number and that “the part

44

The Bloomsbury Companion to Bertrand Russell

is less than the whole” (understood so that a proper subclass of a given class has fewer members than that class)—that he previously accepted and used to argue against Cantor’s theory.42 Accordingly, in defending Cantor’s theory, the post-Peano Russell does not claim, as his Moorean epistemology requires, that it is based on “self-evident” principles; instead, he questions the epistemic value of “self-evidence”—writing, for example, “Self-evidence is often a mere will-o’-the-wisp, which is sure to lead us astray if we take it as our guide” (Papers 3, 368)—and admits that Cantor’s theory is “shocking to common sense” (POM, 358). Further, once Russell discovers his set-theoretic “contradiction”43 and spends years exploring different ways to resolve it, he explicitly rejects the foundationalist view of justification he previously accepted. Thus, in 1906, in dispelling “a misconception” regarding “the nature of the evidence on which [logistic] relies” that “I shared . . . until I came upon the contradictions” (Papers 5, 280), Russell writes: The method of logistic is fundamentally the same as that of every other science. There is the same fallibility, the same uncertainty, the same mixture of induction and deduction, and the same necessity of appealing, in confirmation of principles, to the diffused agreement of calculated results with observation. . . . The “primitive propositions” with which the deductions of logistic begin should, if possible, be evident to intuition; but that is not indispensable, nor is it, in any case, the whole reason for their acceptance. This reason is inductive, namely that, among their known consequences (including themselves), many appear to intuition to be true, none appear to intuition to be false, and those that appear to intuition to be true are not, so far as can be seen, deducible from any system of indemonstrable propositions inconsistent with the system in question. (Papers 5, 280–81)

Thus, instead of holding that axioms or “indemonstrables” have “no evidence except self-evidence,” Russell now holds that no proposition is considered in isolation and accepted simply on its “evidence to intuition.” We may accept a “primitive proposition” that fails to be “self-evident,” so long as it appears to be the best way of deriving consequences we take to be true (and avoiding those we take to be false); and we may reject a “primitive proposition” that is “evident to intuition,” if (in conjunction with other primitive propositions we accept) it has consequences that we take to be false. For Russell, we (do and should) accept the axioms of logic, as we accept the general laws of science, by considerations as to how well they cohere with and systematize our other

Russell and Idealism

45

beliefs. In Bradley’s terminology, Russell now holds—and continues to hold in subsequent writings44—that in determining what to believe “the test which we do apply, and which we must apply, is that of system.”

5.2  Number and analysis In presenting his logicism in POM, Russell defines the cardinal number of a given class α as the class of classes “similar to” α (where two classes are “similar,” or equal in number, if and only if the members of those classes can be put into a one-to-one correspondence with each other). By doing so, Russell abandons the “absolute” theory of number he took to be “plainly correct” during his Moorean period and accepts a “relative” theory. No longer are cardinal numbers indefinables—ultimate constituents of the universe—metaphysically distinct from classes that are used to define the relation of similarity (or equality in number) between classes; instead, the relation of similarity is used to define cardinal numbers.45 Further, Russell recognizes that the procedure he uses to define the cardinal numbers can be used to define other entities that, in his Moorean period, he took to be indefinable.46 Just as he defines cardinal numbers as classes of similar classes, he defines heights as classes of quantities equal in height, and moments as classes of simultaneous events,47 claiming that doing so “clears away incredible accumulations of metaphysical lumber” (OKEW, 42). Thus, he rejects absolute theories of order that had been central to his Moorean philosophy in favor of relative theories of the sort he accepted as an idealist. However, even more fundamentally than that, he rejects his Moorean conception of analysis. No longer does he present analysis as capturing the meaning the analyzed expression already had prior to analysis. Instead, he indicates that in analysis we make “precise” what was previously “vague” (see POM, xix; PM, 12; OKEW, 205; Papers 8, 161–62). In particular, we begin with “vague” or “ambiguous” sentences that we take to be obviously true and then assign precise meanings to those sentences in such a way as to render them true.48 Here, “you cannot for a moment suppose” that the precise meaning assigned to a given sentence “is what we really mean[t]” in uttering it originally (Papers 8, 162). Nor, is there only one correct analysis of the original sentences; for there are a variety of precise meanings we could assign to render the sentences in question true, and Russell presents himself as opting for analyses that “avoid needless assumptions” (OKEW, 126). Thus, in introducing his

46

The Bloomsbury Companion to Bertrand Russell

definitions of the cardinal numbers, Russell does not positively deny that there are indefinables of the sort he previously took the cardinal numbers to be; he only claims they are not needed for “the deduction of all the usual properties of numbers” (POM, 116). Hence, for the post-Peano—unlike the Moorean—Russell, vagueness plays a key role in analysis. However, by continuing to accept (at least until 1918) his Moorean view of acquaintance, Russell holds that vagueness can be overcome. For if we are acquainted with an entity, we can label it and thereby secure a precise meaning for the word we have used to stand for that entity. Hence, although Russell presents analysis as beginning with vague language, he does not focus on vague language as such—on how, or in what sense, it has meaning and can be understood; for his concern is to attain precise language.

5.3  Logical construction and acquaintance Following the publication of PM, Russell applies his post-Peano method of analysis, “so fruitful in the philosophy of mathematics” (Papers 8, 12), to other areas of philosophy. In doing so, he formulates “the supreme maxim in scientific philosophizing”: Wherever possible, logical constructions are to be substituted for inferred entities. (Papers 8, 11)

Thus, by 1914, instead of treating material objects (as he had in POP) as entities with which we are not acquainted that we “infer” from sense-data, he attempts to “construct” them out of sense-data (and sensibilia). By 1919, he applies the program of “logical construction” to the philosophy of mind, thereby introducing the “last substantial change” (MPD, 13) to his philosophy. No longer are thinking “subjects” and the relation of acquaintance indefinable; instead, “logical analogies . . . have led me to [the] conclusion” that, like “points and instants, . . . numbers and particles, and the rest of the apparatus of mathematics,” (thinking) “subjects” and (mental) “acts” (including acquaintance) are “to be constructed, not postulated” (Papers 8, 294; see also AMi, 141–42). The primitive relation of acquaintance between “subject” and “object” was to make it unproblematic how minds can gain knowledge of nonmental entities, thereby refuting idealism. By ceasing to countenance such a relation, Russell presents himself as renouncing a “false simplicity” (AMi, 16) and embarking on a project that “render[s]” the “analysis of knowledge .  .  .  more difficult”

Russell and Idealism

47

(AMi, 15). Moreover, he no longer presents his philosophy as straightforwardly “realist” (see, for example, Papers 9, 162, 410–11, 459–60). For, instead of holding that in sensation, a “subject” is acquainted with a nonmental “object,” he now identifies the sensation with the entity (a particular) that he previously took to be its “object” and holds that, so construed, the sensation is not, in itself, either mental or physical, but rather may be regarded, in a given context, as mental or physical depending on its “causal correlations” with other particulars.49 Hence, like Bradley, Russell now holds that, at the most fundamental level, experience is prior to any distinction between “self ” and “not-self.” Further, by rejecting his former view of acquaintance, Russell lacks the mechanism that he previously held enables us to give precise meanings to words. Accordingly, after 1918 he holds that “all language is vague” (Papers 9, 147; see also AMi 197–98) and holds that analysis proceeds—not from the vague to the precise—but merely from the vague to the “less vague” (IMT, 320). Thus, he writes, echoing Ward, that “philosophy . . . is a is a continuing activity, not something in which we can achieve final perfection once and for all” (OOP, 3). Moreover, by denying that analysis has a final end-point, Russell grants that we never reach “simples” and concedes that “by greater logical skill, the need for assuming them could be avoided” (Papers 9, 173). Furthermore, by 1919, Russell holds that all truth-bearers contain representations; and since he holds that all representations are vague, he holds also that no truth-bearer is absolutely true or absolutely false. Thus, he writes in 1924 that since “[b]eliefs are vague . . . pointing not to one precise fact, but to several vague regions of fact,” they “are not sharply opposed as true or false, but are a blur of truth and falsehood; they are of varying shades of grey, never black or white” (Papers 9, 456). No longer does Russell regard each belief as absolutely true or absolutely false; instead, like Bradley, he holds that each has degrees of both truth and falsity.

5.4  Russell’s post-Peano critique of idealism While the post-Peano Russell abandons central elements of his Moorean philosophy—including its foundational view of justification, its view of analysis as making explicit what we were already aware of in understanding language prior to analysis, its commitment to absolute theories of order, to acquaintance as an indefinable capable of relating thinking subject and extramental object, to simples as the end-points of analysis, and even to absolute truth and falsity— thereby moving his philosophy closer, in some respects, to forms of idealism

48

The Bloomsbury Companion to Bertrand Russell

against which he rebelled,50 he is no less critical of idealism and idealist philosophers than he had been during his Moorean period. However, the nature of his critique changes: instead of specific idealist doctrines he rejects, he focuses on philosophical method, including the relation between philosophy and science. In many respects, Russell’s post-Peano critique of idealism emerges out of his view that mathematicians, especially Cantor, have “completely solved” traditional problems of infinity and continuity that Zeno first raised (Papers 3, 370).51 Soon after the Paris Congress and in numerous writings thereafter,52 Russell claims that by solving those problems, mathematicians have undermined idealist arguments for the view that space and time belong to “appearance,” not “reality.” For Russell, just as Zeno used his paradoxes to support to Parmenides’ monism, so too Kant’s antinomies support his view that space and time are merely “ideal,” and Hegel’s dialectic, which proceeds by generating contradictions, supports his “Absolute Idealism.” But for the postPeano Russell, since these paths to idealism and/or monism depend on holding that the notions of infinity and continuity cannot be made coherent, they are all invalidated by Cantor’s work. Initially, Russell uses Cantor to provide indirect support for aspects of his Moorean philosophy. Thus, he holds that Cantor’s theory of the transfinite undermines idealist arguments against absolute space and time and that Cantor’s view of the continuum as composed of elements supports his metaphysical atomism.53 However, taking a mathematical theory—including its definitions and theorems—as relevant to philosophy is opposed to the Moorean view that philosophy is concerned solely with “the starting-point”—the indefinable and the indemonstrable—not the “results” of mathematical reasoning. Hence, in arguing that philosophers should pay attention to Cantor’s theory, Russell appeals to anti-Moorean views of analysis and justification. He claims that, as opposed to philosophers, Cantor has given precise meanings to the words “infinity” and “continuity” and has developed a theory that, however lacking “self-evidence” and opposed to “common sense” it may be, is adequate for the needs of physics and free from contradiction. More generally, Russell comes to hold that the way that mathematicians have “solved” the problems of infinity and continuity and that he himself addressed the paradoxes provides the model for how philosophy should be conducted. Thus, in 1911, after citing Weierstrass and Cantor as making “all the past difficulties in the notions of the infinite and the continuum disappear” and mentioning his own work in addressing the set-theoretic

Russell and Idealism

49

paradoxes, Russell introduces what he later calls “the scientific method in philosophy”: Heroic solutions have been abused in philosophy, detailed work has been neglected, and there has been too little patience. As with physics in the past, a hypothesis is invented, and on this hypothesis a bizarre world is built which is never compared to the real world. The true method, in philosophy as in science, should be inductive, meticulous, respectful of detail, and should reject the belief that it is the duty of each philosopher to solve all problems by himself. (Papers 6, 138–39)

The “scientific method in philosophy” that emerges out of Russell’s postPeano engagement with mathematical logic is opposed to the Moorean view of philosophy as well as the “heroic” system-building Russell associates with idealism, specifically Hegelian idealism. As against the Moorean view that philosophy is concerned with matters that are so fundamental that “intuition” or “immediate perception,” rather that rational debate, is the only method for dealing with them, Russell now regards philosophy as a collective enterprise capable of “tentative, partial, and progressive advance” (Papers 8, 73). As against Hegelians, who Russell holds are driven by “love of system” (OKEW, 237) to construct, from antecedently accepted principles, “a bizarre world . . . which is never compared to the real world,” Russell now holds that philosophers should follow scientists in adopting a method characterized by “patience” and “detailed work,” by a willingness to work on relatively limited problems even when solutions “are not immediately forthcoming,” and by “inductive[ly]” weighing up and comparing the costs and benefits of alternative hypotheses.54 Furthermore, Russell increasingly regards idealists as intellectually complacent and inexcusably ignorant of philosophically relevant developments in mathematics and science. Thus, after delivering his 1914 lecture “On the Scientific Method in Philosophy” at Oxford, which was then dominated by idealists, Russell writes to Ottoline Morrell: “[I]t was designed to infuriate the Oxford pundits. What I hate is their attempt to make an impressive manner a cloak for complete intellectual atrophy” (Papers 8, 55). In a 1920 review of H. H. Joachim, an Oxford idealist, Russell mentions “advances [that] have all been made by men whose training was predominantly scientific or mathematical, and [that] have been opposed or ignored by orthodox philosophers” (Papers 9, 405), then writes, “[O]fficial academic philosophy, now as at the time of the Renaissance, is engaged in the endeavour to keep alive an antiquated technique, and to ignore the new knowledge which is rendering old problems trivial”

50

The Bloomsbury Companion to Bertrand Russell

(Papers 9, 406). In 1924, he writes that while adherents of German idealism still control academic philosophy in Germany, France, and the United Kingdom, “[t] hey have the strength and weakness of those who stand for the status quo: the strength that comes of tradition, and the weakness that comes of lack of fresh thought” (Papers 9, 452). And for the post-Peano Russell, the “fresh thought” driving his philosophy that is lacking in idealists derives from openness to developments in mathematics and science. Finally, although Russell’s post-Peano engagement with mathematics and science leads him (eventually) to become less “realist” than he had been in his Moorean period, it strengthens his opposition to the “intellectual temper” he associates with idealism. For Russell, idealism (like pragmatism) is one of those “philosophies which assimilate the universe to Man” (POP, 159) and which thereby only “mak[e] Man himself smaller by depriving the universe which he contemplates of all its splendour” (Papers 6, 284). Thus, for Russell, the more “alien” the universe is to us, the more gaining knowledge of it requires shedding “prejudices derived from common sense” (POP, 157), and the more we are thereby “enlarged” by gaining that knowledge (POP, 158–61). Accordingly, for Russell, by demonstrating that the supposed “impossibility of infinite collections was a mistake” whose correction required overcoming “certain rather obstinate mental prejudices” (POP, 147), Cantor illustrates the liberating effect of mathematical logic and more generally of a philosophy that is open to scientific developments. Thus, in both POP and OOP, after mentioning Cantor’s defense of transfinite mathematics, Russell compares logic in the idealist tradition with the mathematical logic embraced by Russell: Formerly, a priori logic was used to prove that various hypotheses which looked possible were impossible, leaving only one possibility, which philosophy therefore pronounced true. Now a priori logic is used to prove the exact contrary, namely, that hypotheses which looked impossible are possible. . . . Instead of being shut in within narrow walls, of which every nook and cranny could be explored, we find ourselves in an open world of free possibilities, where much remains unknown because there is so much to know. . . . [L]ogic, instead of being, as formerly, a bar to possibilities, has become the great liberator of the imagination, presenting innumerable alternatives which are closed to unreflective common sense. (OOP, 296–97; see POP, 148)

While Russell’s concern here is to contrast the liberating effects of mathematical logic with the constraining effects of idealist logic, he is also, implicitly,

Russell and Idealism

51

portraying his post-Peano mathematical logic as liberating him from constraints of the Moorean philosophy. For on the Moorean philosophy, while we can gain knowledge of extramental reality, the knowledge we gain is in accord with what Russell here calls “unreflective common sense” and is available to us by means of “immediate perception,” so that gaining knowledge of the extramental world does not require upsetting our “mental prejudices.” Thus, while the Moorean Russell criticizes idealist logic for preventing us from accepting “common-sense” theories of space and time, the post-Peano Russell regards logic as “loosen[ing] the prejudices of common sense” (POP, 147) and (eventually) regards physics as making “visible the dependence of traditional philosophy upon many indefensible common-sense prejudices” (Papers 9, 405–06), “prejudices” of both sorts that the Moorean Russell had accepted.55 Hence, it is only in his post-Peano period that Russell can present his antiidealism as having the virtue of “enlarg[ing] our conception of what is possible, of enrich[ing] our intellectual imagination and diminish[ing] the dogmatic assurance which closes the mind against speculation” (POP, 161). Like idealists, the post-Peano Russell regards philosophy as undermining common sense; however, unlike idealists, he takes it to be undermined, not by “a priori principles” that “prescribe to the universe” what it must be like, but rather by the “honest toil” (see IMP, 71)—meticulous, tentative, and detailed—that characterizes the scientific method in philosophy. For Russell, while the method is more modest than that pursued by idealists, it ultimately places greater demands on us and also offers more “solid and durable progress” than “heroic solutions” offered by idealists.56

Notes 1 See similar remarks at the outset of Moore (1903b). 2 See Berkeley (1713, 224–5); for “objects of sense,” see, for example (1713, 158). 3 See Kant (1781/1787), Transcendental Aesthetic, Metaphysical Exposition, B edition. 4 See Kant (1781/1787), Transcendental Aesthetic, Transcendental Exposition, B edition (for geometry), and also (1783, Part I). 5 See Kant (1781/1787, A504-507/B532-535). 6 See similar objections to Kant in McTaggart (1896, 27, 107–8) and Bradley (1893, Chapter XII). 7 Whom Russell read avidly as a teenager (see MMD, 8). 8 See, for example, Bradley (1883a, 96; 1893, Chapters 2 and 3).

52

The Bloomsbury Companion to Bertrand Russell

9 For more on this argument see my (1998a, 43–64). 10 While I follow Griffin (1991, 35–45) in highlighting Ward’s influence on Russell, I think he underplays the sense in which Ward regards himself as “Kantian” (see 1991, 40). 11 In citing Ward’s (1919) in this paragraph, and likewise his (1925) in the following paragraph, I take him to be expressing views he held when he was teaching Russell. 12 Dreben (1994) emphasizes that Quine begins “in mediis rebus” (although, unlike Ward, Dreben indicates that Hegel does as well). 13 Ward’s methodology and his use of it to critique “Hegelianism” bears certain similarities to pragmatism. See, in this connection, Dunham (2014). However, while Ward had a substantial and friendly correspondence with William James, he writes to him in 1909: “All the worth I see in pragmatism is to be found—don’t kick—in Kant, in his ‘primacy of practical reason’, and in his showing that there is ‘room for faith’” (Perry 1935, 655). In 1911 (Papers 6, 362), Russell comments on the similarity between Ward’s justification of faith and pragmatism. 14 Given that “subject” and “object” are of the same nature, Ward rejects a metaphysical dualism in favor of some form of “monism” (understood as the view that the universe is made of one kind of stuff). He considers three kinds of “monism”—materialism, a “neutral” or “agnostic” monism, and “spiritual” monism, arguing for the latter (see 1899b, Part V). 15 For further detail on the issues discusses in this section, see Spadoni (1976) and Griffin (1991, 71–78). 16 For Bradley, ultimately, every judgment is about the same (existing) individual—the Absolute. In the passage from PL, Russell characterizes the “existential theory of judgment” as requiring only that for each judgment there is an existent that that judgment is about. 17 Russell argues only that the axioms of projective geometry, which are common to Euclidean and non-Euclidean geometries, are a priori. Hence, Russell regards his arguments regarding space as akin to Kant’s arguments in the Metaphysical Exposition, not the Transcendental Exposition (see EFG, 55–6 and notes 3 and 4 above). Moreover, unlike Kant (as Russell interprets him), Russell refrains from inferring that what is a priori is thereby “subjective” (EFG, 3–4). For more on Russell’s idealist views of geometry see Griffin (1991, Chapter 4) and Gandon (2012, 36ff.). 18 For more on Russell’s idealist views of physics and how they figure in his “dialectic of the sciences,” see Griffin (1991, Chapter 5). 19 As Russell presents it, the “contradiction of relativity” arises in attempting to ground an asymmetric relation between distinct entities; that is, he presents it as an issue arising for “monadism,” suggesting that he may, at this point have regarded the “contradiction of relativity” as mandating a “dialectical transition” to monism (see in this connection Griffin (2013, 389)). Thus, in one of his last writings in which he

Russell and Idealism

20

21

22 23 24 25

26

27

28

53

presents the “contradiction of relativity” as an “unavoidable contradiction” infecting mathematics, Russell writes: “I wonder if a world in which a sole logical subject existed—a Spinozistic world in short—would not be more in conformity with Logic than a world with a plurality of subjects” (Papers 2, 337). It may seem that by the same reasoning, in the example discussed in the previous paragraph, A’s differing from depends on Blue’s differing from green and hence on an ungrounded relation between two entities; and after he breaks with idealism, Russell indicates that it does (see Papers 5, 446). However, assuming Bradley’s view that, being “tautologous,” there is no genuine judgment of the form, or no genuine relation affirmed by a statement of the form, “A is A,” the same reasoning does not apply to show that A’s having the same color as B depends on Red’s being identical with red. See in this connection, note 25 of this chapter. For more on the issues discussed in this Part, see Hylton (1990, Part II) and my (2009, 11–44; 2016, 168–77). While I refer to Russell’s “Moorean philosophy,” I do not discuss here the extent to which Russell’s views immediately after he breaks with idealism are similar to Moore’s (although see references to Moore in some notes below). For further discussion and assessment of Russell’s response to Bradley, see Candlish (2007) and my (2014). Compare the last five sentences of Moore (1899). As late as 1913 (TK, 80–1), Russell questions whether any complexes consist of only two constituents—a “subject” and a predicate functioning “as predicating.” See similarly Moore (1900, 295): “Every proposition, it would seem, must contain at least two different terms and their relation.” By accepting this view, neither Russell nor Moore allow a proposition to consist of a single term in relation to itself and so do not regard identity as a genuine relation. See Russell (Papers 2, 140); Moore (1900, 295). See, for example, Papers 3 (254) and POM (450–51) and for a later version of the same argument (Papers 6, 137–38). Moore (1903a, 123–26) similarly argues that accepting the existential theory of judgment has a pernicious effect in discussions of what is good “in itself.” Although in discussing here Russell’s Moorean views of acquaintance, I cite passages from well after he broke with idealism, I argue elsewhere (2009, 17–28; 2016, 168–72) that Russell accepted the views I attribute to him here (except one— see note 32) when he originally broke with idealism. Initially, Moore and Russell seem to hold that we can be acquainted in sensation with ordinary material objects; however, by POP (Chapter I), if not sooner, Russell holds that in sensation we are acquainted only with “sense-data.” But in accepting this view, Russell holds that the objects of sensation—sense-data—are extramental, physical entities (see Papers 8, 8–9, 76–77).

54

The Bloomsbury Companion to Bertrand Russell

29 See Moore (1903b), Russell (POP, Chapter IV). 30 See Moore (1903b, 24f), Russell (TK, 41-4). 31 See Moore (1904), Russell (POP, Chapters VIII–X). It is only after the influence of Wittgenstein that Russell comes to deny that logical truths are synthetic. 32 See POP (143–44); see Proops (2014, Sections 4–6) for a discussion of when and how Russell came to use acquaintance in this way. 33 See similarly Moore (1903a, 144): “[W]hen any proposition is self-evident, … there are no reasons which prove its truth.” 34 For passages supporting this interpretation of the Moorean Russell, see my (2016, 168–72). 35 In POM (Chapter V), in which he introduces his theory of “denoting concepts,” Russell recognizes an exception to the principle that a proposition contains among its constituents the entities it is about; however, two year later, in OD, he rejects the theory of denoting concepts and is able to restore the principle (albeit with the somewhat Bradleyan consequence that the “logical subject” of a proposition may be quite different from what it initially appears to be). 36 See Griffin 2013, 391–92. 37 See, for example, POM, 223–24 Papers 5, 448; OKEW, 48–49; Papers 8, 183–84; MPD, 54–55. 38 See, for example, Papers 5, 445. 39 For more detailed discussion of the issues involved here, see Griffin (1991, Chapter 8, Section 6; 2013, 387–92) and my work (1998b, 100–07). 40 See (Papers 3, 57–58). 41 See (Papers 3, 57–59; POM, 163–68), where Russell compares the indefinables and indemonstrables of the two theories. 42 See, for example, Papers 2 (265) and PL (109) along with Leibniz’s argument quoted on 244. For discussion of these issues, see Moore (1995) and my work (2009, §2.1). 43 As Moore (1995) discusses, Russell’s idealist background, with its emphasis on “unavoidable contradictions” and “antinomies,” plays a role in his discovering his paradox. 44 See, for example, Papers 5 (576–78), PM (59–60), and Papers 9 (163–64). 45 Although here, unlike the relative theory of number he considered during his Moorean period, similarity is not indefinable but is rather defined in terms of oneto-one correspondence. However, the central point is that Russell has reversed the order of definitional (and metaphysical) priority. 46 See POM (116), OKEW (42, 124–26) and Papers 9 (164–65). 47 The technical difficulty in defining moments and points is that since the “events” in terms of which they are to be defined “have a finite extent,” events can be “overlapping” without being entirely simultaneous, so that moments (and points) will have to be “constructed” out of “overlapping,” rather than “simultaneous”

Russell and Idealism

55

events. Russell credits Whitehead with the solution of this problem (see, for example, OKEW, 114ff; Papers 9, 166). 48 Given his fallibilist epistemology, the post-Peano Russell allows that in the course of an analysis, we may come to reject at least some of the sentences we initially took to be obviously true. See, for example, OKEW (211–12). 49 Thus, with regard to the analysis of sensation, at least, Russell’s post-1918 view is in accord with “neutral monism.” See Chapter 12 of this book. 50 And also closer to pragmatism—see my (2018). 51 While in his published writings Russell consistently presents Cantor as having definitively resolved all the traditional problems of infinity and continuity, there is at least one manuscript, apparently written in the late 1930s, in which Russell indicates that the set-theoretic paradoxes show that the problems remain unsolved (see Vianelli 2001). 52 See, for example, Papers 3, 372; POM, 192, 347, 458–61; POP, 146–47; Papers 6, 138–39; Papers 9, 163, 461–62; HWP, 783–84. 53 See POM, 287–88, 347. 54 In some respects—its modesty, tentativeness, opposition to “heroic” Hegelian “solutions” that have little connection to “the real world”—the method Russell advocates is similar to that presented by Ward. However, in introducing his scientific method in philosophy, Russell emphasizes that “ethical and religious motives . . . ought now to be consciously thrust aside by those who wish to discover philosophical truth” (Papers 8, 57). And it is on these grounds that Russell criticizes Ward, who, unlike McTaggart and Bradley, defends a specifically Christian version of idealism. In particular, Russell criticizes Ward for his “moral arguments” intended to induce faith (Papers 6, 362) and also for appealing to a notion of causality (in arguing for idealism) that he acknowledges plays no role in science (Papers 6, 193). 55 See in this connection Russell’s review of Moore (1922), where he writes that rather than attempting to “vindicate the common-sense beliefs of everyday life,” Moore should attempt to vindicate “common sense” as developed by physicists and that “Dr. Moore’s method tends to become somewhat verbal and dry for lack of the material which science would supply” (Papers 9, 408). 56 Thanks to the editor for his helpful comments.

References Berkeley, George [1713] 1965. Three Dialogues between Hylas and Philonous. Reprinted in Berkeley’s Philosophical Writings. Edited by David M. Armstrong, 129–225. New York: Collier Books.

56

The Bloomsbury Companion to Bertrand Russell

Bradley, Francis H. 1883a. The Principles of Logic, Vol. I. Oxford: University Press. (References to the 1950 reprint of the 1922 second edition.) Bradley, Francis H. 1883b. The Principles of Logic, Vol. II. Oxford: University Press. (References to the 1950 reprint of the 1922 second edition.) Bradley, Francis H. 1893. Appearance and Reality. Oxford: Clarendon Press (References to the 1930, ninth impression.) Bradley, Francis H. 1914. Essays on Truth and Reality. Oxford: Clarendon Press. Bradley, Francis H. 1999a. Refinement and Revision 1903–1924, Collected Works of F. H. Bradley, Volume 3. Edited by Carol A. Keene. Bristol: Thoemmes Press. Bradley, Francis H. 1999b. Selected Correspondence June 1872–December 1904, Collected Works of F. H. Bradley, Volume 4. Edited by Carol A. Keene. Bristol: Thoemmes Press. Candlish, Stewart 2007. The Russell/Bradley Dispute and its Significance for the Twentieth Century Philosophy. Basingstoke: Palgrave Macmillan. Dreben, Burton 1994. “In Mediis Rebus.” Inquiry 37: 441–47. Dunham, Jeremy 2014. “Was Ward a Cambridge Pragmatist?” British Journal for the History of Philosophy 22: 557–81. Gandon, Sébastien 2012. Russell’s Unknown Logicism. Basingstoke: Palgrave Macmillan. Griffin, Nicholas 1991. Russell’s Idealist Apprenticeship. Oxford: Clarendon Press. Griffin, Nicholas 2013. “Russell and Moore’s Revolt Against British Idealism.” In The Oxford Handbook of the History of Analytic Philosophy, edited by Michael Beaney, 383–406. Oxford: University Press. Hegel, Georg W. F. 1830. The Encyclopedia Logic. (References to translation by T. F. Geraets, W. A. Suchting, and H. S. Harris, Indianapolis: Hacket Publishing Company, 1991.) Hylton, Peter 1990. Russell, Idealism and the Emergence of Analytic Philosophy. Oxford: Clarendon Press. Kant, Immanuel 1781/1787. Critique of Pure Reason. (References to translation by Paul Guyer and Allen Wood, Cambridge: Cambridge University Press, 1998.) Kant, Immanuel 1783. Prolegomena to Any Future Metaphysics. (References to translation by James Q. Ellington, revised from Paul Carus, Indianapolis: Hacket Publishing Company, 1977.) Levine, James 1998a. “The What and the That: Theories of Singular Thought in Bradley, Russell, and the early Wittgenstein.” In Appearance versus Reality: New Essays on Bradley’s Metaphysics, edited by G. Stock, 19–72. Oxford: Clarendon Press. Levine, James 1998b. “From Absolute Idealism to the Principles of Mathematics.” International Journal of Philosophical Studies 6: 87–127. Levine, James 2009. “From Moore to Peano to Watson: The Mathematical Roots of Russell’s Naturalism and Behaviorism.” The Baltic International Yearbook of Cognition, Logic and Communication 4: 1–126. (http://newprairiepress.org/cgi/ viewcontent.cgi?article=1024&context=biyclc)

Russell and Idealism

57

Levine, James 2014. “Russell, Particularized Relations, and Bradley’s Dilemma.” dialectica 68 (2): 231–61. Levine, James 2016. “The Place of Vagueness in Russell’s Philosophical Development.” In Early Analytic Philosophy—New Perspectives on the Tradition, edited by Sorin Costreie, 161–212. Cham: Switzerland, Springer. Levine, James 2018. “Russell, Pragmatism, and the Priority of Use over Meaning.” In Pragmatism and the European Traditions: Encounters with Analytic Philosophy and Phenomenology Before the Great Divide, edited by M. Baghramian and S. Marchetti. New York, NY: Routledge. McTaggart, John M. E. 1893. The Further Determination of the Absolute. Reprinted in his Philosophical Studies. Edited by S. V. Keeling, 210–72. London: Edward Arnold & Co. McTaggart, John M. E. 1896. Studies in the Hegelian Dialectic. Cambridge: Cambridge University Press. Moore, George E. 1899. “The Nature of Judgment.” Mind 8 (new series): 176–93. Moore, George E. 1900. “Necessity.” Mind 9 (new series): 289–304. Moore, George E. 1903a. Principia Ethica. Cambridge: Cambridge University Press. Moore, George E. 1903b. “The Refutation of Idealism.” Reprinted in his Philosophical Studies, 1–30. London: Routledge and Kegan Paul. Moore, George E. 1904. “Kant’s Idealism.” Proceedings of the Aristotelian Society 4 (new series): 127–40. Moore, George E. 1922. Philosophical Studies. London: Routledge and Kegan Paul. Moore, Gregory H. 1995. “The Origins of Russell’s Paradox: Russell, Couturat, and the Antinomy of Infinite Number.” In Essays on the Development of the Foundations of Mathematics, edited by J. Hintikka, 215–39. Dordrecht: Klewer Academic Publishers. Perry, Ralph B. 1935. The Thought and Character of William James, Volume II. Boston: Little, Brown, and Company. Proops, Ian 2014. “Russellian Acquaintance Revisited.” Journal of the History of Philosophy 52: 779–811. Spadoni, Carl 1976. “‘Great God in Boots!—The Ontological Argument is Sound!” Russell: The Journal of the Bertrand Russell Archives 23–24 (old series): 37–41. Vianelli, Giovanni 2001. “A Newly Discovered Text by Russell on Pythagoras and the History of Mathematics.” Russell: The Journal of the Bertrand Russell Archives 21 (new series): 5–30.Ward, James 1882. “A General Analysis of Mind.” Journal of Speculative Philosophy 16: 366–85. Ward, James 1893. “‘Modern’ Psychology: A Reflexion.” Mind 2 (5): 54–82. Ward, James 1899a. Naturalism and Agnosticism, Volume I. London: Adam and Charles Black. Ward, James 1899b. Naturalism and Agnosticism, Volume II. London: Adam and Charles Black. Ward, James 1904. Philosophical Orientation and Scientific Standpoint. Berkeley: University Press.

58

The Bloomsbury Companion to Bertrand Russell

Ward, James 1911. The Realm of Ends. Cambridge: Cambridge University Press. Ward, James 1919. “In the Beginning . . . ,” Proceedings of the Aristotelian Society 20 (new series): 1–24. Ward, James 1922. A Study of Kant. Cambridge: Cambridge University Press. Ward, James 1925. “Bradley’s Doctrine of Experience.” Mind 34: 13–38.

2

Russell and the Pragmatists1 Cheryl Misak

1  The straightforward story: Russell the anti-pragmatist It might be thought that the relationship between Russell and the pragmatists is straightforward: he was resolutely antagonistic to pragmatism. The straightforward story seems compelling when we reflect on the fact that Russell wrote something critical on pragmatism every year from 1908 to 1912. It is further bolstered by the fact that he never relented in taking pragmatism’s “cardinal” and “genuinely new” point—its theory of truth—to be silly at best, dangerous at worst (Papers 6, 260–61). As far as Russell was concerned, the pragmatist theory of truth was the position expressed at times by one of the founders of pragmatism, William James. On his most condensed and least careful expressions of that theory, truth is what works or pays. Russell’s review of James’s 1907 Pragmatism: A New Name for Some Old Ways of Thinking, in which he thought he detected that simplistic view of truth, pulls no punches. He says the book “is like a bath with hot water running in so slowly that you don’t know when to scream” (Papers 5, 472). James’s collection The Will to Believe did not improve Russell’s opinion. James says in the preface to that volume: If religious hypotheses about the universe be in order at all, then the active faiths of individuals in them, freely expressing themselves in life, are the experimental tests by which they are verified, and the only means by which their truth or falsehood can be wrought out. The truest scientific hypothesis is that which, as we say, “works” best; and it can be no otherwise with religious hypotheses. (WB, 8)

While James engaged in sustained backtracking from and nuancing of such statements, they stuck with Russell. He mounted several strong arguments

60

The Bloomsbury Companion to Bertrand Russell

against them, and pragmatism’s reputation has never fully recovered from the onslaught. One of Russell’s arguments is that it is near impossible to determine whether a belief pays or has useful consequences: We must suppose that this means that the consequences of entertaining the belief are better than those of rejecting it. In order to know this, we must know what are the consequences of entertaining it, and what are the consequences of rejecting it; we must know also what consequences are good, what bad, what consequences are better, and what worse. (Papers 5, 476–77)

This is a very tall order, which he illustrates with two examples. First, the consequences of believing the doctrine of the Catholic faith might be that the belief makes a person happy “at the expense of a certain amount of stupidity and priestly domination” (Papers 5, 477). It is unclear how we are to weigh these benefits and burdens against each other. Second, the effects of Rousseau’s doctrines were far-reaching—Europe is a different place from what it would have been without them. But how can we isolate from the whole history of Europe after Rousseau all and only those events that are effects of his views? And even if we could do that, whether we take these effects to be good or bad will itself depend on our political views. The question of whether the consequences of believing something are good or bad is an extraordinarily difficult one: often much more difficult to settle, Russell thinks, than the simple question of whether the belief is true. In a related objection, Russell challenges James’s use of the concepts “works” or “pays” and objects to James’s idea, expressed in “The Will to Believe” that when the evidence is underdetermined and we are forced to make a decision, we can believe what makes our lives go better. Russell rightly takes pragmatism to be an extension of the method of the inductive sciences, and he commends it for being such. But pragmatism, he says, misapplies this method in its understanding of what “works”: When science says that a hypothesis works, it means that from this hypothesis we can deduce a number of propositions which are verifiable..  .  . But when pragmatism says that a hypothesis works, it means that the effects of believing it are good, including among the effects . . . the emotions entailed by it or its perceived consequences, and the actions to which we are prompted by it or its perceived consequences. This is a totally different conception of “working,” and one for which the authority of scientific procedure cannot be invoked. (Papers 5 , 484)

Russell and the Pragmatists

61

Finally, Russell accuses the pragmatist of conflating the criterion of truth with the meaning of truth. He illustrates the distinction as follows. If you wish to consult a certain book in the library, you check the catalogue—but a book’s being in the catalogue is not the same thing as its being in the library. In a similar way, utility is a mark or a criterion of truth, but it is not identical to truth. Just as what we want is for the book to be in the library and not merely in the catalogue, what we want is for beliefs to be actually true, not merely to work for us. Russell considers, for example, the belief in the existence of other minds: If I am troubled by solipsism, the discovery that a belief in the existence of others is “true” in the pragmatist’s sense is not enough to allay my sense of loneliness: the perception that I should profit by rejecting solipsism is not alone sufficient to make me reject it. For what I desire is not that the belief in solipsism should be false in the pragmatic sense, but that other people should in fact exist. (Papers 5, 479)

We can have the experience of actually hearing Russell, in a 1959 interview, repeat his antipathy to pragmatism. In answer to the question of whether he thinks there might be a “practical reason” for some people to believe in God, Russell replies, gesturing at the Jamesian account of truth: “It seems to me a fundamental dishonesty and a fundamental treachery to intellectual integrity to hold a belief because you think it’s useful and not because you think it’s true.”2 Of course, Jamesian pragmatism, even at its best, rather than its worst Russellian interpretation of it, is not the only manifestation of the tradition. The straightforward story, though, seems to survive this observation, as Russell seems to have been just as harsh about John Dewey. He wrote to Ottoline Morrell: “In 1914, I liked Dewey better than any other academic American: now I can’t stand him” (Papers 6 , 258; 1921). As for F. C. S. Schiller, pragmatism’s representative in Oxford, Russell thought he was “a pathetic fool” (Auto, 195). The only pragmatist he didn’t pour scorn on was C. S. Peirce, the unsung, personally difficult, and technically sophisticated pragmatist. Perhaps Russell simply hadn’t read much Peirce when he was engaged in battle with pragmatism. Decades later, after the war had cooled, he said of Peirce: “Beyond doubt .  .  . he was one of the most original minds of the later nineteenth century, and certainly the greatest American thinker ever” (WW, 276). But that, it would seem, does not change the apparent fact that for the first half of the twentieth century, when Russell was at the forefront of philosophy and wielding his considerable influence, he was

62

The Bloomsbury Companion to Bertrand Russell

as anti-pragmatist as one could be. Hence the straightforward story that Russell was the enemy of pragmatism has remained firmly entrenched. Indeed, I have, in the past, articulated it myself.3 But things turn out, as so often, to be far more complex than they first appear.

2  A spanner in the works of the straightforward story Frank Ramsey, the brilliant Cambridge philosopher, mathematician, and economist who, when he was still an undergraduate, helped Russell with Principia Mathematica and translated Wittgenstein’s Tractatus, had become a card-carrying pragmatist by the mid-1920s. In 1927, he read “Facts and Propositions” to the Joint Session of the Aristotelian Society and the Mind Association, with G. E. Moore commenting at great length. In that paper, which is largely concerned with Russell’s account of facts and propositions, Ramsey took his first public step toward pragmatism. He also says something that goes very much against the grain of the straightforward story often related about Russell. He asserted, without explanation or hesitation: “My pragmatism is derived from Mr. Russell” (FP, 51). This assertion went unremarked upon in the commentary. What could possibly have been going on? Was Ramsey out of touch with Russell and his antipathy toward pragmatism? The answer to that question is: most definitely not. As his friend and colleague Richard Braithwaite put it, the young philosophers in Cambridge were word perfect on Russell.4 And after Ramsey’s death, just shy of his twenty-seventh birthday in 1930, the summations of his work casually repeated the connection between Russell and Ramsey’s pragmatism. Here is Richard Braithwaite in his obituary of Ramsey: “Recently (in company with Bertrand Russell) he had been descending the slippery path to a sort of pragmatism” (Braithwaite 1930, 216). Here is Keynes in his review of The Foundations of Mathematics, Ramsey’s posthumously published papers, after he says that Ramsey and Wittgenstein were helping Russell to perfect the formal matters first set out in Principia Mathematica: The effect was gradually to empty it of content and to reduce it more and more to mere dry bones, until finally it seemed to exclude not only all experience, but most of the principles, usually reckoned logical, of reasonable thought. Wittgenstein’s solution was to regard everything else as a sort of inspired nonsense, having great value indeed for the individual, but incapable of being exactly discussed. Ramsey’s reaction was towards what he himself described as a

Russell and the Pragmatists

63

sort of pragmatism, not unsympathetic to Russell, but repugnant to Wittgenstein. . . . Thus he was led to consider “human logic” as distinguished from “formal logic.” (Keynes 1972 [1931], 338)

Russell, in his own review of The Foundations of Mathematics, noted without demurral that Ramsey attributes his pragmatism “to the present reviewer,” although later in the review says that Ramsey was “perhaps too pragmatist in his attitude” (1931, 479, 482). In order to see what was going on here, we need to take a careful look at Russell’s exposure to pragmatism and to the themes in pragmatism that cannot be summed up in the snappy slogan “truth is what works.” For Ramsey was just as wary of those slogans, yet he found something worth expanding and defending in the pragmatist tradition. Pragmatism, he says, “is not simply to be mocked at. . . . What is ludicrous, is not the general idea” of pragmatism, but “the way in which William James confused it especially in its application to religious beliefs” (OT, 91). What Ramsey was interested in was the pragmatist idea that a belief is a habit or disposition to behave, and can be evaluated as such. Russell, we shall see, shared that interest.

3  Russell at Harvard: 1914 While Russell knew James and Schiller well, both in print and in person, he also knew a lot more about pragmatism than what James and Schiller said about truth and the will to believe. In the spring of 1914 Russell delivered the Lowell Lectures in Boston. They were on scientific method in philosophy, and were published as Our Knowledge of the External World. He also gave two courses. James had died in 1910 and Peirce would die in April 1914, in the middle of Russell’s visit. Russell knew some of Peirce’s work prior to his arrival at Harvard. In 1904, Peirce was hoping to make a visit to England, and Russell wrote to Peirce’s correspondent Lady Victoria Welby that it would be “the greatest interest and pleasure” to meet Peirce, as “he is a man for whose work I have a great respect.”5 He went on in the letter to talk about Peirce’s categories—his complex division of the real into quality, the brute reaction of experience, and interpretation. During his visit Russell had more exposure to Peirce’s thought. The physicist Victor Lenzen, then a Ph.D. student in Philosophy at Harvard, described the visit thus: The spring semester of 1914 was made notable at Harvard by the residence of Bertrand Russell, who was scheduled to give courses on Logic and on

64

The Bloomsbury Companion to Bertrand Russell Theory of Knowledge. Russell was then at the height of his fame as a logician, and Professor Royce declared that he had received more attention than any logician since Aristotle. Russell’s arrival was delayed, and for several weeks Costello and Royce lectured for him in the courses on Logic and Theory of Knowledge, respectively. The latter . . . contrasted the theories of Russell and Peirce. Professor Royce explained that Russell’s theory, in which mind and object were related by a dyadic relation, was a dyadic theory of knowledge, whereas Peirce’s theory, which involved the concept of interpretation, was a triadic theory. Royce agreed with Peirce that the process of cognition is a social one. (Lenzen 1965, 4)

On Russell’s dyadic theory of knowledge, perception is a two-term relation between subject and object. On Peirce’s triadic account, perception always involves a third term: an interpretation. Harry Costello was a young instructor in the Harvard Department, who served as an assistant lecturer to Russell during his advanced logic course, and was the pinch hitter in that course prior to Russell’s arrival in mid-March. Royce did a similar duty for Russell’s Theory of Knowledge course, with Costello and Lenzen each taking notes on his lectures, partly so that Russell could review what was said during his absence in the first part of his course. T.S. Eliot took notes on Russell’s Advanced Logic class. In addition, Royce’s long-standing comparative methodology seminar ran for the whole of the academic year, with Costello and Lenzen again taking careful notes.6 Russell arrived midway through it as well, and it is unsurprising that his logic and theory of knowledge were frequently discussed in the seminar, along with Moore’s definition of the good. All in all, it was a term in which Cambridge, England figured prominently in Cambridge, Massachusetts. But the knowledge-sharing went in the other direction as well. During Royce’s seminar and the during the first parts of the Theory of Knowledge and Advanced Logic courses, the views of Peirce, James, Dewey, and Schiller enjoyed frequent and careful discussion, with pragmatism distinguished from, as Royce put it, the “wider” humanism of Schiller and James. The pragmatist account of truth, as well as “its misunderstanders” (as James had termed them), figured heavily in some lectures.7 Royce was a very good understander of pragmatism generally and Peirce in particular. Peirce died in Pennsylvania toward the end of this term, and thus it is also no surprise that Royce would have focused on him. He spent time on Peirce’s abductive inference, categories, and account of perception, and on Peirce’s account of meaning or understanding, on

Russell and the Pragmatists

65

which meaning is not just a dictionary definition, nor a pointing to objects in the world, but also the habits of behavior or the expectations for action that are sparked or entailed. His first suggestion for that last, pragmatist, grade of clarity is that the meaning of a belief is the “habits it produces” (W 3, 265; 1878). Perhaps most significantly, as far as Russell’s own development went, Royce talked about Peirce’s theory of truth, belief, and inquiry. A belief, Peirce argued, is an expectation or a habit of behavior. We act on our beliefs until and unless some experience upsets our expectation, in which case inquiry is sparked. Inquiry proceeds from settled belief, to the doubt caused by recalcitrant experience, to a new and better belief. A belief that would forever be permanently settled, or would always serve us well in our interactions with the world, would be true. These views would be published in a volume of posthumously collected papers in 1923: Chance, Love and Logic. It was published in Britain simultaneous with the American edition and Russell read it shortly after. But we shall see that by 1923, Russell was already steeped in pragmatism. Russell’s own lectures were well attended by students and faculty. Royce, in particular, had sustained critical interactions with him. Costello says that both “Royce and Russell were big enough to take it and like it” (Costello 1963, 194). Royce announced in his seminar that Russell’s first Lowell Lecture had a “harte Schönheit” (hard beauty), and at the end of the lectures, his view of Russell had not dimmed (Costello 1963, 151, 186). Others with pragmatist sympathies engaged Russell as well. Lenzen tells us that “on one occasion when Mr. Russell was expounding his dualism between mind and matter,” Ralph Barton Perry was “drawn into defense of his behavioristic theory of mind” (Lenzen 1971, 5). Through these and other interactions during his Harvard semester, Russell learned a lot about pragmatism. So much so that on September 23, 1916, the chair of the Harvard Philosophy and Psychology Department, James Woods, invited Russell to return to Harvard for a year, to teach a seminar on Peirce, and with the help of some graduate students and the Harvard logician Sheffer, to edit two or three volumes of Peirce’s writings.8 Russell was keen. On March 14, 1916, in a letter to Woods, he said: I am looking forward to your answers as to my lectures next year. I am very anxious to give the sort of course that will be acceptable. (Eliot and Haughton 2011, 145)

66

The Bloomsbury Companion to Bertrand Russell

Russell was at that time caught up in political activities, which were to result in his dismissal from Trinity College and land him in jail. The Foreign Office refused to give him a passport, and as a result he never did give his seminar on Peirce.

4  Russell’s pragmatism: Perception, belief, and behavior Royce was right to see that one pivotal difference between the 1914 Russell and the pragmatists concerned perception and knowledge. Russell was then looking to set out the relationship between the mind and the world as a correspondence between simple things in the mind and simple things in the world. For the pragmatists, the mind’s relation to the world is not straightforward or unmediated. Russell was eventually to change his view on this matter, partly under the influence of James. The logical atomist view was under all sorts of strain, but at least one source involved the theory of perception that seemed to go hand in hand with it. For instance, Russell says in My Philosophical Development that by 1918 he had become convinced “that William James had been right in denying the relational character of sensations” (MPD, 134). But well before 1918, indeed, before even his 1914 visit, Russell had started to put forward a view of perception he found in James. When James died, Russell wrote an intellectual eulogy in a short piece titled “The Philosophy of William James.” He spoke highly of James’s psychology; of his empiricist aptitude for seeing “facts first, instead of first seeing theories and then searching out facts to confirm or refute them”; and of his refusal to begin, as psychologists are apt to do, with sensations, because a mere sensation is an abstraction which never really occurs. He begins instead with “the stream of thought,” taking thought as we find it, with all kinds of thinking mixed in a vague continuum. This leads him to a long and very able discussion of “the consciousness of self,” in which he reduces the Self to the passing thought, with its memory of other thoughts and its consciousness of the body. (Papers 6, 286; 1910)

In 1912 Russell wrote his final review of a book of James’s—the posthumously published Essays in Radical Empiricism—for Mind, and we find in it a change in tone from his previous reviews. The volume’s editor was Ralph Barton Perry, who tells us in the preface that James initially intended this volume to contain much else, about meaning and truth, for instance, but he had already reprinted

Russell and the Pragmatists

67

that material while he was still alive. Perhaps this is why Russell was so positive about Essays in Radical Empiricism. It was relatively shorn of the things that he did not like in pragmatism, leaving only the rich conception of experience and its relationship to the world. That was a conception he liked, and he was heartened by James’s claim that one could be a radical empiricist without being a pragmatist about truth. For despite not quite believing “that empiricism, however radical, requires that we should deny the difference between mind and matter” (Papers 6, 304; 1912), Russell was now attracted to James’s view that we directly experience relations, as well as particular things. Russell felt the very pressures that are felt by all empiricists. In the 1911 “Knowledge by Acquaintance and Knowledge by Description,” he had distinguished between two kinds of knowledge, a distinction he acquired from James’s Principles of Psychology and then built upon in a very non-Jamesian way.9 Knowledge by acquaintance is knowledge in which I have “a direct cognitive relation” with the object of the proposition known. I “am directly aware” of it (Papers 6, 148; 1911). We can know a simple object by being directly acquainted with it. What I am directly acquainted with is not, as the idealist would have it, “part of the subject”—part of me. Russell says that only “the most desperate contortions” could prevent that idealist position from leading to solipsism— to the conclusion that only I and my states exist. If all knowledge is rooted in perception, but the objects of perception are mental, then it seems we have no epistemic access to the external world. Russell wants to keep the external world and “to preserve the dualism of subject and object” (CP 6, 148; 1911). Knowledge by acquaintance is a two-term relation between a subject and an object. We are directly acquainted with “sense-data”—noises, colors, etc.,— that are capable of existing unobserved. But then it is hard to make sense of perceiving those independently existing objects without infecting perception with human subjectivity (RSDP, 5, 7). Russell was thus intrigued by James’s idea that, as Russell put it: “There is no stuff out of which thoughts, as opposed to matter, are made; pure experience is the only stuff of the world; what distinguishes consciousness is a certain function, namely the function of knowing” (Papers 6, 300–01; 1912). While Russell posed a number of questions for James’s position and wished James had offered more elaboration, he finds interesting the idea that “the knower and the known . . . may be the same piece of experience in different contexts” (Papers 6, 301; 1912). Russell’s attraction to James’s account of perception which Russell called “neutral monism” is well-known. But what is less well-known, or better, less

68

The Bloomsbury Companion to Bertrand Russell

well-remembered, is how Russell was influenced by the behaviorist aspect of pragmatism.10 Russell came away from his 1914 Harvard visit with a new, positive thought about what is good in pragmatism. That thought culminated in his 1921 The Analysis of Mind, in which Russell tries to analyze the relationship between thought and the world by appealing to a rich, pragmatist account of experience, and to a pragmatist account of the relationship between belief and behavior. He appeals, that is to the idea, first held by Peirce and taken further by Dewey, that a belief is a disposition or a habit of action. In 1919 Russell wrote a long review of Dewey’s 1916 Essays in Experimental Logic, in which Dewey was mighty rude about Russell. He sets out some detailed criticisms of Dewey’s “misunderstanding” of the “analytic realists” and their conception of data. He also criticizes the “will-to-believers.” But he sees one important thing in Dewey’s book. Dewey takes logic to be the study of the nature, origin, and justification of belief—a set of topics that Russell thinks is important, but is more properly thought of as a mix of psychology and logic. He does not like Dewey’s calling them problems of logic. But, once properly labeled, Russell is not altogether hostile to the core of Dewey’s position, which he characterizes as follows: The essence of a belief is the behaviour which exemplifies it (which is it, one is tempted to say); this behaviour is such as is intended to achieve a certain end, and the belief is shown in the behaviour adopted for that purpose. The belief is called true when the behaviour which exemplifies it achieves its end, and false when it does not—omitting refinements due to cooperation of different beliefs. Knowledge is like a railway journey: it is a humanly constructed means of moving from place to place, and its matter, like the rails, is as much a human product as the rest of it, though dependent upon a crude ore which, in its unmanufactured state, would be as useless to intellectual locomotion as iron ore to locomotion by train. (Papers 8, 143; 1919)

Russell argues that there is something good in this view, although he cannot buy it in its totality. He argues that not all knowledge will serve as a means to our ends; that Dewey’s position seems to leave no room for contemplation; that it nonchalantly helps itself to our common-sense view of causality, since “we must know what effects are caused by our beliefs” if we are to “test their value as instruments” (Papers 8, 145; 1919); that it ignores the skeptical challenge; and that it allows us to know only the “world” of our own construction—“we know how the trains will move, since we laid down the rails for them” (Papers 8, 148; 1919). That’s a lot to not like. But the part of Dewey’s view that Russell does like

Russell and the Pragmatists

69

is the behaviorist account of belief, “since it would enable me to define a belief as a certain series of acts” (Papers 8, 147; 1919). Russell, that is, was coming round to the pragmatist idea that belief and action are linked.11 In The Analysis of Mind, he is even more pragmatist, in both of the ways I have just canvassed. In the preface to the book, he says that his motivation in writing the book stems from the idea of combining two tendencies in recent thought. Psychologists had become more interested in the material realm—in physiology and external observation of behavior. And physicists, working under the paradigm set by Einstein’s theory of relativity, had been “making ‘matter’ less and less material.” Russell states his intention: The view that seems to me to reconcile the materialistic tendency of psychology with the anti-materialistic tendency of physics is the view of William James and the American new realists, according to which the “stuff ” of the world is neither mental nor material, but a “neutral stuff,” out of which both are constructed. I have endeavoured in this work to develop this view in some detail as regards the phenomena with which psychology is concerned. (Ami, iii)

There is no thing called “consciousness” that is the substance within which (token) mental phenomena—for instance, beliefs and desires—occur. With James (he quotes extensively from The Principles of Psychology, “A World of Pure Experience,” and “Does Consciousness Exist?”), Russell now rejects the dualisms between subject and object, on the one hand, and mind and body, on the other.12 But he retains his resistance to the idea that there is no distinction at all to be drawn between the respective entities to be found in the mental and the physical worlds. Some entities, he thinks, will be subject only to mental laws, and some will be subject only to physical laws. The images produced in dreams or by imagination, for instance, are not objects within the physical world (Ami, 82). And there may be physical objects that no subject ever experiences, in which case they would not be part of the mental world (AMi, 11). Nonetheless, “the distinction between mind and matter is not so fundamental as is commonly supposed” (Ami, 63). The Analysis of Mind also adopts a pragmatist account of meaning, with its focus on habits of action. Those who are familiar with Peirce’s account of understanding as being not merely about definition and connotation, but also a matter of habits or knowing what to expect, will be struck by the similarities in the following passage: It is not necessary, in order that a man should “understand” a word, that he should “know what it means,” in the sense of being able to say “this word means

70

The Bloomsbury Companion to Bertrand Russell so-and-so.” Understanding words does not consist in knowing their dictionary definitions, or in being able to specify the objects to which they are appropriate. . . . Understanding language is more like understanding cricket: it is a matter of habits, acquired in oneself and rightly presumed in others. To say that a word has a meaning is not to say that those who use the word correctly have ever thought out what the meaning is: the use of the word comes first, and the meaning is to be distilled out of it by observation and analysis. (Ami, 117–18)

Russell does not want to adopt an extreme or pure behaviorist view, on which “nothing can be known except by external observation” (Ami, 11). But he is suddenly struck by a case of charitability and he sees that even J.B. Watson might not hold such a “crude” view. While the behaviorists “somewhat overstate their case,” still he thinks that “there is an important element of truth in their contention,” even when it comes to introspection of our own thoughts and desires (Ami, 13). We can observe what people desire from what they do, and one thing we can observe is that people—including oneself, viewed retrospectively— can be mistaken about what they desire. At bottom, human beings are animals, and in analyzing the nature of the mind “it is of the greatest importance to remember that from the protozoa to man there is nowhere a very wide gap either in structure or behaviour” (Ami, 21). Braithwaite put Russell’s position thus: “He accepts the analysis of will and desire as merely characteristics of certain cycles of actions, but he will not agree with Professor J. B. Watson that thinking consists entirely of muttering words under the breath” (1921, 605). What he accepts is the pragmatist linking of belief and behavior, which was more moderate than that of the behaviorists. Our whole intellectual life, Russell says, consists of belief “and of the passage from one belief to another” by, for instance, our processes of reasoning. Peirce too had argued that inquiry is the passage from belief to doubt to belief again. A 1905 article in The Monist titled “What Pragmatism Is,” whose companion piece, the 1906 “Prolegomena to an Apology for Pragmatism,” was certainly making the rounds in Cambridge England,13 says the following: Belief is not a momentary mode of consciousness; it is a habit of mind essentially enduring for some time, and mostly (at least) unconscious; and like other habits, it is (until it meets with some surprise that begins its dissolution) perfectly selfsatisfied. Doubt is of an altogether contrary genus. It is not a habit, but the privation of a habit. Now a privation of a habit, in order to be anything at all, must be a condition of erratic activity that in some way must get superseded by a habit. (Papers 5, 417; 1905)

Russell and the Pragmatists

71

Here is Russell in The Analysis of Mind. Most of our beliefs are not present to our mind—they “display themselves when the expectations that they arouse fail in any way” (Ami, 146). The “shock of surprise” we feel, say, on drinking a glass of what seemed like beer but turns out to be tea “makes us aware of the expectations that habitually enter into our perceptions; and such expectations must be classed as beliefs, in spite of the fact that we do not normally take note of them or put them into words” (Ami, 146). Russell has become a pragmatist about many kinds of belief: they are expectations that guide our behavior and can be upset by the shock of recalcitrant experience. He not only agrees with Peirce about the connection between belief and habit, but also with his thought that beliefs, rather than propositions, “are the vehicles of truth and falsehood” (Ami, 139). This is the precisely the pragmatist insight that started Ramsey off. Ramsey’s assertion that his pragmatism was derived from Russell now seems makes good sense. While Ramsey got his pragmatism in greater intensity and in greater detail from reading Peirce directly, it was perfectly reasonable in the 1920s to think that Russell was generally on board as well. I say “generally,” because while Russell was never attracted to a pragmatist about truth. And, unlike the pragmatist, he still hankers after the reductionist project: It is quite likely that, if we knew more about animal bodies, we could deduce all their movements from the laws of chemistry and physics. It is already fairly easy to see how chemistry reduces to physics. . . . We only know in part how to reduce physiology to chemistry, but we know enough to make it likely that the reduction is possible. (Ami, 24–25)

The reduction of human behavior to physical laws “can only be effected by entering into great minuteness”—like a spark to dynamite, a postcard saying “All is discovered; fly!” will set off an explosive chain of actions (Ami, 25). While it is highly unlikely that we will be able to predict how the postcard will determine subsequent behavior in detail, Russell thinks that reductionist projects—for instance, the “analysis” of the “material unit” into “constituents analogous to sensations”—remains of “the utmost importance” for philosophy (Ami, 185). This would have left Peirce and James cold, while Russell was still hot for analysis. But one thing is clear. Despite these differences, the relationship between Russell and the pragmatists was a close one in the 1920s, so close that Russell seemed to many to be a pragmatist, rather than the arch anti-pragmatist that he is usually supposed to be. The straightforward story misses out on an important

72

The Bloomsbury Companion to Bertrand Russell

and vital slice of the history of analytic philosophy. While Russell may have encouraged it later in life, by retaining his intense and critical gaze on the worst expression of the pragmatist account of truth, he was partly responsible for planting a seed of pragmatism into the ground of Cambridge, England. That seed had unexpected flowerings. Most significantly, but beyond the scope of this chapter, Ramsey criticized Wittgenstein’s Tractatus on pragmatist grounds and was responsible for Wittgenstein’s turn away from logical analysis and to thinking of meaning in terms of use.14

Notes 1 This chapter draws on my much more expansive Cambridge Pragmatism: From Peirce and James to Ramsey and Wittgenstein. 2 http://www.youtube.com/watch?v=tP4FDLegX9s 3 See Misak (2008). 4 TFL/MS/COLL/375: 13/42A, 43A. 5 York University Clara Thomas Archives 1970-010/13(22). 6 We owe Bernard Linsky a huge debt for compiling and editing the notes for these courses. 7 Royce also discussed his own student C. I. Lewis’s anti-Russellian, pragmatist views about implication. 8 The letter can be found in the Russell Archives, William Ready Division of Archives and Research Collections, Mills Memorial Library, McMaster University. 9 Moore in some undated lecture notes on metaphysics, notes that both Russell and James make use of this distinction, Cambridge University Library Add 8875 13/39/33/1. See Proops (2014) for an excellent account of the historical connection between Russell’s distinction and James’s, as well as the similarities and the differences. 10 But see Acero (2005) and Levine (forthcoming). 11 See James Levine (2009) for an excellent and more full account of Russell’s post1918 behaviourism. See also Baldwin (2003). 12 He later noted that this wasn’t an easy move: “This change in my opinions greatly increased the difficulty of problems involved in connecting experience with the outer world” (MPD, 13). 13 C. K. Ogden had made extensive notes on it, which ended up in Ramsey’s hands. 14 This argument is made in my Cambridge Pragmatism.

Russell and the Pragmatists

73

References Acero, Juan Jose. 2005. “Mind, Intentionality, and Language: The Impact of Russell’s Pragmatism on Ramsey.” In F.P. Ramsey: Critical Reassessments, edited by M. J. Frápolli, 7–40. New York: Continuum. Baldwin, Thomas. 2003. “From Knowledge by Acquaintance to Knowledge by Causation.” In The Cambridge Companion to Bertrand Russell, edited by N. Griffin, 420–48. Cambridge: Cambridge University Press. Braithwaite, Richard. 1930. “Frank Plumpton Ramsey.” Cambridge Review 31 January: 216. Costello, Harry T. 1963. Josiah Royce’s Seminar, 1913-14: As Recorded in the Notebooks of Harry T. Costello. Edited by G. Smith. New York: Rutgers University Press. Eliot, Valerie, and Hugh Haughton (eds.) 2011. The Letters of T.S. Eliot, volume 1: 18981922. Rev. ed. New Haven: Yale University Press. James, William. [1907] 1919. Pragmatism: A New Name for Some Old Ways of Thinking. Reprint. New York: Longmans, Green and Co. James, William. [1912] 1976. Essays in Radical Empiricism. In James (1975–88), vol. iii. The Works of William James. 18 vol, edited by F. H. Burkhard, F. Bowers, and I. K. Skrupskelis. Cambridge, MA: Harvard University Press. James, William. [1897] 1979. The Will to Believe and Other Essays in Popular Philosophy. In James (1975–88), vol. vi. The Works of William James. 18 vol, edited by F. H. Burkhard, F. Bowers, and I. K. Skrupskelis. Cambridge, MA: Harvard University Press. James, William. [1890 ] 1981. The Principles of Psychology. In James (1975–88), vols. viii–x. The Works of William James. 18 vol, edited by F. H. Burkhard, F. Bowers, and I. K. Skrupskelis. Cambridge, MA: Harvard University Press. Keynes, John Maynard. [1931] 1972. Review of The Foundations of Mathematics, by F. P. Ramsey. Originally appeared in The New Statesman and Nation, October 3. Reprinted in The Collected Writings of John Maynard Keynes, vol. X, edited by D. E. Moggridge. Reprint. London: Macmillan. Lenzen, Victor F. 1965. “Reminiscences of a Mission to Milford, Pennsylvania.” Transactions of the Charles S. Peirce Society 1 (1): 3–11. Lenzen, Victor F. 1971. “Bertrand Russell at Harvard.” Russell: The Journal of Bertrand Russell Studies, o.s., 3: 4–6. Levine, James. 2009. “From Moore to Peano to Russell: The Mathematical Roots of Russell’s Naturalism and Behaviourism.” The Baltic International Yearbook of Cognition, Logic and Communication 4: 1–126. Levine, James. forthcoming. “Russell, Pragmatism, and Meaning as Use.” In Philosophical Revolutions: Encounters of Pragmatism, Phenomenology and Analytic Philosophy in the Early Twentieth Century, edited by M. Baghramian and S. Marchetti. Oxford: Oxford University Press.

74

The Bloomsbury Companion to Bertrand Russell

Misak, Cheryl. 2008. “The Reception of Early American Pragmatism.” In The Oxford Handbook of American Philosophy, edited by C. Misak, 197–223. Oxford: Oxford University Press. Misak, Cheryl. 2016. Cambridge Pragmatism: From Peirce and James to Ramsey and Wittgenstein. Oxford: Oxford University Press. Peirce, Charles Sanders. 1923. Chance, Love and Logic: Philosophical Essays. Edited by M. R. Cohen. London: Kegan Paul, Trench, Trubner & Co, Ltd. Peirce, Charles Sanders. (1931–58). (CP) Collected Papers of Charles Sanders Peirce. 8 vols., edited by C. Hartshorne and P. Weiss (vols. i–vi), A. Burks (vols. vii and viii). Cambridge, MA: Belknap Press. Proops, Ian. 2014. “Russellian Acquaintance Revisited.” Journal of the History of Philosophy 52 (4): 779–811. Ramsey, Frank (FP). [1927] 1990. “Facts and Propositions.” Reprinted in F.P. Ramsey: Philosophical Papers, edited by D. H. Mellor, 34–51. Cambridge: Cambridge University Press. Ramsey, Frank (FP). [1930] 1991. On Truth. Edited by N. Rescher and U. Majer. Dordrecht: Kluwer. Russell. 1931. “Critical Notice of The Foundations of Mathematics and Other Logical Essays, Frank Plumpton Ramsey.” Mind 40 (160): 476–82. TFL: Thomas Fisher Rare Book Library, University of Toronto.

3

Russell’s Objections to Frege and Meinong in Context Bernard Linsky

Bertrand Russell’s writings on Gottlob Frege and Alexius Meinong are best known for laying out “Russell’s Paradox” and for, in “On Denoting,” the example of “The present King of France.” Both arguments came to Russell as he was preparing comments on the works of both Frege and Meinong and they each reveal isolated, if devastating, points of disagreement within a wider range of issues. This chapter will assess the scope of the agreement, and in particular, the nature of what Russell learned from Frege. In June 1902 Russell wrote a letter to Gottlob Frege describing his “paradox” of the class of all classes that are not members of themselves. In 1905 he published his theory of descriptions in “On Denoting” in Mind, with its criticism of Alexius Meinong’s theory of nonexistent objects based on the example sentence “The present King of France is bald.” Recent scholarship has shown that we should not see these as merely clever counterexamples that Russell invented to demolish the theories of his rivals Frege and Meinong. Both objections are best understood as the outcome of developments within Russell’s own thinking over a longer period of time. Russell had come across the paradox while writing the body of The Principles of Mathematics well before he realized that it applied to Frege’s system as well as his own. In the same survey of Frege’s writings that lead to the letter, Russell also learned the details of Frege’s reduction of arithmetic to logic. While he and Whitehead did not follow the details of Frege’s analysis in their Principia Mathematica, they learned from his example. Similarly, the target of “On Denoting” is now is now seen to be primarily Russell’s own prior theory of “denoting concepts,” rather a Meinongian theory employing “non-existing” objects to serve as the referents of empty descriptions. Russell’s objections to Meinong’s nonexistent objects in fact

76

The Bloomsbury Companion to Bertrand Russell

emerged from considerations about true and false propositions, and he only turned to nonexistent objects such as “the present King of France” late in his thinking. Russell studied Meinong’s work in preparation for a long, three-part essay “Meinong’s Theory of Assumptions and Complexes,” in Mind 1904, and his review in 1905 of Investigations in Object Theory and Psychology, written by Meinong and his students. The review is published in the same October 1905 issue of Mind in which “On Denoting” appears. An examination of these works, and some correspondence that Russell had with Meinong and two of his students before writing the review, show both the range of agreement of Russell with Meinong’s doctrines and the origin of the famous example of “the present King of France” he used in “On Denoting.”

1  The origins of Russell’s Paradox and his disagreements with Frege Russell’s objections to Frege emerged from a range of agreement on the fundamental nature of the distinction between concepts and objects. They disagreed, however, over the notion of mathematical functions, which Frege used to base his logic, while Russell proposed to analyze the notion of mathematical functions in terms of his notion of propositional function. Frege and Russell also disagreed over the nature of classes (and the extensions of concepts). Frege analyzed the extensions of concepts as objects, and it was that which Russell first saw as the consequence of his paradox. Russell’s Paradox was specifically aimed at Frege’s treatment of the extensions of concepts. Concepts were just one sort of function for Frege, and so had, like other functions, a “course of values” which were counted as objects in his strictly bifurcated ontology of concept and objects. Russell did not share Frege’s insistence on this sharp dichotomy, and the associated unsaturated nature of concepts, which resulted in his startling denial that concepts can even be named by singular terms. Russell held that the mathematical notion of function was to be analyzed first in terms of denoting concepts and then, after the theory of descriptions, as functional relations. In Principia Mathematica “*30 Descriptive Functions,” Russell explains how a relation Rxy, such as “x is father of y,” can be turned into a function “the father of y” with the use of the theory of descriptions, as “the x such that x is father of y,” where that is expanded as with other descriptions, to be expressed as a sentence “There is a z such that for all w, w is father of y if and only if w = z and z is

Russell’s Objections to Frege and Meinong in Context

77

father of y” (cf. Df *30.01). Logicians call a relation R such that there is a unique y to which any x is related a “functional relation,” and it is an exercise in elementary logic to show that functional expressions, called “terms” and, symbolized by “fx,” “gxy,” and so on, can be eliminated in terms of functional relations.1 One standard example of the use of this reduction of functions to relations is in the formal number theory where the two-place function of addition, “x + y” is eliminated in favor of a three-place relational constant “Sum(x,y,z)” which is true when z is the sum of x and y. This seemingly small technical point becomes very significant if one sees a difference between mathematical functions and predicates, or propositional functions, as Russell did. Despite this disagreement, Russell agreed on the fundamental distinction between objects and the other notions needed in logic, whether that might be concepts and functions, as with Frege, or propositional functions, in his own logic. There are traces of Frege’s views about the difference in logical category of concepts and objects in Russell’s immediate reaction to his paradox that he interpreted as proving that there are “more” classes or propositional functions than objects, and also in the later distinctions of the theory of types.2 As will be shown below, the technical structure of Russell’s Paradox was familiar to Ernst Zermelo. Yet, while it is correct to say that others, especially Zermelo, may have “anticipated” Russell’s Paradox, it is clear that Russell came to it on his own as a difficulty for his own system, and then observed that it would apply to Frege’s system when he studied Frege’s work carefully in May 1902, just before he reported the paradox to Frege. Ernst Schröder’s three-volume “Lectures on the Algebra of Logic” (Vorlesungen ueber die Algebra der Logik, 1890), was designed to introduce the then new algebraic logic of George Boole and C. S. Peirce to a German audience. It remained the dominant alternative to the later symbolic logic of Peano, Frege, and Russell even after the appearance of Principia Mathematica in 1910.3 The new logic of Peano, Frege, and Russell paid careful attention to the distinction singular terms and predicates, which Schröder appeared to have ignored by using a single symbol “€” for “is,” whether that was a case of the membership or subset, marked by “ε” and “⊂” in Peano’s notation. Logicians in the “algebraic” tradition of Peirce and Schröder, including Giuseppe Peano who had started out in this school, held that many of the formulas of their calculi could be read as valid on two interpretations, as about propositions or about classes although Peano explicitly noted those sentences that could be interpreted in both ways, and those that could not (Peano, 90).

78

The Bloomsbury Companion to Bertrand Russell

That Schröder, however, did not observe the distinction between the subset and membership relations is demonstrated in his proof that there is a universal class. His argument led to Zermelo’s anticipation of Russell’s Paradox. In the fourth lecture of the Algebra der Logik on the theory of classes, Schröder presents the algebraic account of classes using one primitive notion of “subsumption” (Subsumtion). The assertion that a class a is subsumed by a class b, symbolized as “a € b,” (which is read indifferently as “a is b” or “all a are b”), means what we would express as “a is a subset of b,” symbolized as “a ⊆ b,” and the predication “a is b,” relating an individual to a class to which it belongs. One part of Schröder’s account of sets is a rejection of Boole’s notion of the universe of discourse, which in this framework will be the element 1 in the algebra of classes. The empty set will be the object 0. The argument below is from page 245 of Volume I (Schröder 1890) as quoted by Frege, who also objected to the argument in his discussion of Schröder’s logic: As we have laid down, 0 would have to be contained in every class that can be got out of the manifold 1; . . . 0 would have to be the subject of every predicate. Now suppose we took a to be the class of those classes of the manifold that are equal to 1 (which would certainly be permissible if we could bring everything thinkable into the manifold 1), then this class of its very nature contains just one class, viz. the symbol 1 itself, or alternatively the whole of the manifold, which constitutes the reference of the symbol; but therefore besides this it would contain “nothing,” i.e., 0. Hence 1 and 0 would make up the class of the objects that are to be equal to 1; and so we should have to admit not only: 1 = 1 but also: 0 = 1. For a predicate that applies to a class—in our case, the predicate: to be identically equal to 1—must also apply to every individual in the class. (Frege 1895, 91)

For Schröder, all predications are assertions of subsumptions, and in particular, the predication “is equal to 1,” which we would express as the formula “x = 1,” becomes an assertion about subsumption. Given that the empty set, represented by “0,” is subsumed by every class a, that is, 0 € a, it is also subsumed by that of the things equal to 0, hence we derive 0 = 1, a contradiction. This he takes to be a proof that there is no absolutely universal class 1. B. Rang and W. Thomas (1981) find an anticipation of Russell’s Paradox in a little-known exchange between Edmund Husserl and his former student Ernst Zermelo that emerged from a discussion of this argument by Schröder. Husserl had written a review of The Algebra of Logic that included the charge that Schröder’s argument had confused the subset and membership relations (Husserl 1979). Husserl pointed out that while it is true that the empty class 0 is

Russell’s Objections to Frege and Meinong in Context

79

a subset of every set, 0 is not a member of every set, and so the argument fails. Ernst Zermelo wrote about this review to his former teacher, Husserl, pointing out that “on the issue, not the method of proof, Schröder is right.” Translated from the original German Gabelsburger Stenographie shorthand, the notes announce the result as follows: A set M, which contains each of its subsets m, m’, .  .  .  as elements, is an inconsistent set, i.e., such a set, if at all treated as a set, leads to contradictions. PROOF. We consider those subsets m which do not contain themselves as elements. . . . These constitute in their totality a set M0. (Rang and Thomas 1981, 16–17)

As presented by Zermelo, this is a proof that no set contains all of its own subsets as members. A universal set of all things, however, would certainly include all of its subsets as members, as those are all sets of things. Zermelo considers a set M0, the set of all subsets of the universal set that do not contain themselves, which is simply the set of all sets that do not contain themselves. M0, in this case, is the “Russell set.” The proof that M0 leads to contradiction is the same argument that Russell gives, if M0 is a member of itself then it is not, and if it is not a member of itself then it is. We have the same contradiction as in Russell’s letter to Frege, and both come fairly directly from applying Cantor’s Theorem to a set of all sets (or in Zermelo’s case, a set which contains at least all its subsets). Russell came to his paradox independently, while writing The Principles of Mathematics. Russell had been dubious of the applicability of Cantor’s Theorem, that any class will have more subsets than members, to the special case of the class of all things, the “universe.”4 In the details of Cantor’s argument applied to a supposed class of all things (including all classes) one encounters a mapping from each class of classes to a unique class. It is then shown that one class of classes cannot have a value by that mapping, namely the class of all classes that are not members of themselves. This seems to be how Russell himself came upon his paradox.5 This is different from the Zermelo’s argument, which is presented as a proof that no set can contain all its subsets. Zermelo’s argument involves the same notions, and also relies essentially on Cantor’s “diagonal argument.” There is no doubt that in this sense Russell’s Paradox was “anticipated” by others. What is clear, however, is that the ideas involved in the argument, about the notion of their being a universal class, the “universe of discourse,” and of its conflict with the distinction between a set and its members, was common in

80

The Bloomsbury Companion to Bertrand Russell

the various discussions of Schröder’s logic at the time and was shared by Frege and Russell. Russell does not remark on Frege’s criticisms of Schröder’s argument, but there is clear evidence in the notes from 1902, and through his later logical writings, that he shared in the rejection of the conflation of membership and subset. In the spring of 1913 the mathematician Norbert Wiener had written a thesis at Harvard University comparing the logic of Whitehead and Russell with that of Schröder. Wiener then travelled to Cambridge and had several discussions with Russell about his thesis and the logic of Schröder. In a note exchanged after one of their meetings Russell objected to the conflation of subset and membership, and asked Wiener whether Schröder could understand “the difference between Peter and the class whose only member is Peter.” Wiener’s response, however, was that Schröder was “not concerned” with that distinction.6 Russell got nowhere with Wiener on this issue. Russell thus agreed with Frege’s criticism of Schröder. He differed with Frege, however, on the analysis of concepts as functions from objects to truth-values, and continued to disagree on through the writing of Principia Mathematica. This did not keep him from learning from Frege about the logical analysis of arithmetic.

2  What Russell learned from Frege Upon reading through Frege’s works in 1902, Russell found that Frege had not only anticipated many of his results in Principles, but also had much to teach him about carrying out the logicist program. In the Preface to Principia Mathematica, in 1910, Whitehead and Russell write: In all questions of logical analysis, our chief debt is to Frege. Where we differ from him, it is largely because the contradictions showed that he, in common with all other logicians, ancient and modern, had allowed some error to creep into his premises; but apart from the contradictions, it would have been almost impossible to detect this error. In Arithmetic and the theory of series, our whole work is based on that of Georg Cantor. (Whitehead and Russell 1910, viii)

A review of the records of Russell’s reading of Frege in 1902 reveals the attention Russell paid to the logical details of Grundgesetze der Arithmetik, as well as evidence of what Russell and Whitehead learned from Frege about “logical analysis.” It is possible to follow Russell’s thought as he came to the awareness that Frege was subject to the same contradiction that he himself had found over

Russell’s Objections to Frege and Meinong in Context

81

the preceding year. After praising Frege’s work in the 1902 letter to Frege, Russell introduces the paradox almost as an afterthought, saying “There is just one point where I have encountered a difficulty” (Russell 1902, 125). Russell realized that “the contradiction” was as much a problem for his own work as for Frege’s and that he owed a great deal to the rest of Frege’s work. Calling the paradox “just one point” of difficulty with Frege’s system was an appropriate description of its place in his appreciation of Frege. The remainder of this section will examine the origins of Russell’s appendix A to Principles of Mathematics, “The Logical and Arithmetical Doctrines of Frege,” and in particular Russell’s own marginal remarks and reading notes from his careful study of Frege’s works that he undertook in May 1902, immediately before reporting the paradox to Frege. Russell took note of points that he shared with Frege, and acknowledged numerous technical ideas that he had learned from him, and only slowly, and not as a direct point of investigation, discovered that Frege was subject to the same paradox that he had been considering for the last year. There is nothing like a marginal mark near the fatal Law V, which identifies the course of values of coextensive concepts, or anywhere else in Frege’s works, indicating that as the source of the contradiction. Russell’s Paradox assumes that one is committed to the existence of a set of all sets that are not members of themselves, and that commitment is only indirectly made in Frege’s system as it does not have an explicit “comprehension” principle. It was Frege himself, in his response to Russell’s letter, who points out the exact source of the contradiction in the Grundgesetze system. The marginalia and notes do not include any reference to the paradox at all. The marginalia in Grundgesetze clearly precede the writing of the letter to Frege, as Russell writes that he had already made a study of Frege’s writings. They may therefore precede Russell’s conclusion that the paradox would affect Frege. Russell notes the important passages, and often comments on some that have been especially significant in the history of Frege scholarship. One gets the impression of Russell working out Frege’s views as he read, for example, initially thinking that there is something wrong with Frege’s notorious Law V, then deleting the objection and citing the passage that resolves the difficulty. At other points he voices an objection that we appreciate all too well, having struggled with the very different views of Frege and Russell on some issues. Russell begins the notes on Grundgesetze with a series of objections to Frege’s exposition of the “Werthverlauf” or “course-of-values” of a function. Russell’s problems arise because of Frege’s view of concepts as functions from objects to

82

The Bloomsbury Companion to Bertrand Russell

truth-values, the True and the False. For Frege extension of a concept is its course of values. Other functions such as those expressed by “sin x” and “x + 4,” also have a course of values, an object which represents the arguments and values of the function in itself, much like the set of ordered pairs of argument and value to which functions are reduced in contemporary set theory. Russell seemed to think that Frege’s Axiom V, that identifies any two coextensive functions that map each argument to the same value, in fact only makes sense for the special case of concepts, not with arbitrary functions. His comment is “Df: bad if other than propositional functions included.”7 Russell’s notes report Frege’s distinction between a singleton class and its sole member. (“ɩx” in Peano’s notation. This is “{x}” in modern notation.) Indeed, it is arguments from Frege that seem to have persuaded Russell to change his mind on this issue. In the body of Principles Russell considers the “extensional” view of classes to be committed to identifying a class with its members, and so not distinguishing a singleton from its sole member. One such argument is that if a set z has two members x and y, then the singleton {z} would have one member and while z has two, and so they must be distinct. Russell must have been led to understand this point after struggling to give an account of the numbers 0 and 1 as equivalence classes like the other numbers. In appendix A (Russell 1903, 513), Russell acknowledges Frege’s arguments that this distinction must be made, even when one views classes extensionally.8 Russell’s interest at that time was in Frege’s analysis of mathematics, and so the philosophy of language that is so important to us, is only noted in passing in the comments. Thus while Russell makes extensive notes on Frege’s notions of Sinn und Bedeutung he does not notice the issues about the shift of Bedeutung of expressions in “indirect” contexts and whether objects are constituents of thoughts, which became of interest in the correspondence with Frege in 1903, and to philosophers of language since.9 Other issues looming large in contemporary discussions of Frege do not appear at all in the notes. In the “Introduction” to Foundations of Arithmetic (Frege 1959, x) Frege cites three “fundamental principles” which he will follow: “always to separate sharply the psychological from the logical, the subjective from the objective; never to ask for the meaning of a word in isolation, but only in the context of a proposition; never to lose sight of the distinction between concept and object.” Russell only mentions the last of these in his notes on that page but pays a good deal of attention to the distinction of concept and object in Frege’s thought. This includes his reading of Benno Kerry’s criticisms of Frege. Kerry argued that

Russell’s Objections to Frege and Meinong in Context

83

Frege must hold that concepts cannot be named by singular terms, hence that Frege was committed to the notorious paradoxical view that “The concept horse” is not a concept.10 Yet Russell notes Frege’s view, but did not agree with it. Russell thought though that everything, including concepts, or rather propositional functions, must be possible subjects of propositions, or terms in his own system. Thus the comment next to Grundgesetze §21, after Frege has painstakingly explained that a concept or function name cannot appear in the place where the name of an object can: He does not realize that everything is a gegenstand (Linsky 2004, 21) Frege’s third principle expressed as “never to ask for the meaning of a word in isolation” is not mentioned in the notes. Yet, with the name “the context principle” it is now seen as a key to Frege’s notion of reference, or Bedeutung, and to motivate the definition of number with Hume’s Principle. Russell’s primary interest in the notes is in those aspects of Frege’s logical views that have to do with the foundations of arithmetic, and in particular the principle of induction, that any property true of 0, and if true of a number n is true of its successor, n + 1, will be true of all numbers. The important theorems of elementary number theory are proved using this principle, called proofs “by induction.” For Frege’s account of induction, a property F is said to be inherited by the R relation if when some individual x has F then anything related by R to x also possesses F. Frege defines what Russell dubbed the “ancestral” of a relation R as holding between an individual x and any individual y which has all the properties possessed by x which are inherited by the relation R. The natural numbers are defined as those objects bearing the ancestral of the successor relation to 0, and so it is easy to prove that the principle of induction holds for the natural numbers: any property possessed by 0 and hereditary with respect to the successor relation will be possessed by all numbers. In the first set of notes on the Begriffschrift, on §26, where Frege had first given his definition Russell remarks: Frege says this defines “x precedes y in series generated by R.” The whole proposition amounts merely to a R b. (Linsky 2004–05, 169)

84

The Bloomsbury Companion to Bertrand Russell

Russell seems to suppose that R is a transitive relation, in which case it is identical with its ancestral. However, by the time of writing a later set of notes his view has changed. Here we find This relation may be expressed “x precedes y in the R-Series.” [It seems to be a non-numerical definition of RN and very ingenious: it is better than Peano’s mathematical induction.] (Linsky 2004–05, 161) Russell here thinks of the notion of a series of relations. If xRy we might write xR1y and if there is an intermediate z such that xRz and zRy, we write instead xR2y. This suggests a definition of xRny which might be defined by induction, thus xRx+1y means that there is a z such that xRny and yRz. It is easy to see this as a definition of the ancestral that presupposes numerical notions, including induction. This is likely what Russell here sees as the advantage of Frege’s account of induction over Peano’s. The interval between the writing of these notes bridges the moment when Russell came to appreciate Frege’s definition of the ancestral of a relation. This remarks suggests, however, that at this point Russell sees Frege’s approach as a superior alternative to simply adopting an axiom of induction as Peano proposed. This remark suggests that Russell does not see deriving the induction principle from logic alone as a primary goal for the logicist program. It is simply “better,” albeit “very ingenious.” Many scholars of the history of logic agree that Russell must have got the idea of the ancestral of a relation and its connection with mathematical induction from reading Richard Dedekind’s Was sind und was sollen die Zahlen?11 It appears, however, that Russell not only knew of Dedekind’s earlier analysis, but saw Frege’s as an improvement on it. Dedekind’s work followed Frege’s Grundlagen and Begriffschrift, but Russell was studying them all together to supplement and annotate his work in Principles. Russell made numerous marginal notes in his symbolism, and seven pages of notes on the proofs. In section 59 Dedekind proves a “Theorem of complete induction,” to the effect that in order to show that the chain A0 is part of any system ∑. . . it is sufficient to show that A ε ∑, and . . . that the transform of every common element of A0 and ∑ is likewise an element of ∑.

In section 60 Dedekind says that “The preceding theorem, as will be shown later, forms the basis for the form of demonstration known as complete

Russell’s Objections to Frege and Meinong in Context

85

induction (the inference from n to n + 1) . . .” To this Russell remarks in the margin: ? Does not the definition of a chain involve mathematical induction in many cases?

Russell does not see in Dedekind the reduction of induction to logical principles that he does identify in Frege’s work. In fact this remark suggests that Russell suspected that Dedekind’s informal method of proof was making tacit use of some sort of principle of induction. Russell was not ignorant of the earlier accounts of induction by Peano and Dedekind, but rather he knew them both, and preferred Frege’s as “non-numerical” and not presupposing any sort of induction. Russell’s notes follow the principal theorems of the first volume of Grundgesetze. The main result of part A is Theorem 32, what has come to be known as Hume’s Principle: if there is a one-to-one function mapping u onto v, then the number of u’s is the same as the number of v’s.12 Russell transcribes these theorems into his own notation, using occasional borrowings from Frege, and copies out selected lines from the proofs, again translating them. Russell’s notation is generally adequate to translate Frege’s, although around the notions of ancestral and number series he has to introduce some new defined expressions. The main result of the next section is Theorem 71: the successor relation is one-to-one. This is one of Peano’s postulates. The proof occupies Frege for over twenty-seven pages, but Russell devotes less than one page to it, merely stating the result and transcribing a few lines of the proof without citation. Only a few sections on, it appears that Russell has become familiar with the system and the notation. One major theorem that Russell notes is 155, that “the number of finite numbers up to and including b is b + 1.” This is a key step in proving that every number has a successor, another of Peano’s axioms. Whitehead and Russell were unable to prove this result this way in Principia Mathematica because the successor of a number would be of a higher type than the numbers that belong to it, and so the numbers would run through the whole of the “illegitimate totality” of types. As a result Whitehead and Russell were forced to add the notorious “Axiom of Infinity” (*120.03 Infin Ax), which simply asserts that every inductive class of individuals is non-empty, thus the infinity of the numbers is guaranteed by the infinity of individuals of a single type.13 It is clear, then, that when Whitehead and Russell went about correcting Frege’s logicist account of the numbers, Russell had in mind the details of Frege’s construction.

86

The Bloomsbury Companion to Bertrand Russell

Russell’s interest in the rest of notes on Frege’s project is not as intense as the attention that he pays to the theory of natural numbers in Volume I. There are only three pages of notes on Volume II that pick out the principal results and lemmas through to the end of the volume.14 Thus Russell studied the details of Frege’s construction of the numbers in Grundgesetze and read the entire book, although he would work out the details of those later accounts of real numbers very differently in Principia Mathematica. Russell came to appreciate Frege’s analysis of arithmetic during the period of intensive study in preparation for adding an Appendix to The Principles of Mathematics. He slowly came to appreciate Frege’s theory of extensions (courses of values) and so was able to see that Frege shared a susceptibility to the paradox. However, at the same time, he learned many lessons from Frege that were important for his own later development of arithmetic in Principia Mathematica. These points included a deep appreciation of Frege’s notion of the ancestral of a relation, and so of Frege’s account of mathematical induction, and also the elementary distinction between an object and the singleton set containing it, which was important for his accounts of the numbers 0 and 1 and also for sharpening his objection to Schröder’s earlier logic which had not distinguished clearly enough between a member of a set and its subsets. On other points Russell took note of Frege’s views on which he would later disagree, first in the foundational views about semantics such as the “context principle,” the distinction of sense and reference, and between concepts and objects.

3  Where Russell agreed with Meinong Russell shared a number of Meinong’s logical doctrines, even though he had serious objections to the idea of nonexistent objects, and it is much more difficult than with Frege to identify particular views that Russell learned from Meinong. In a letter dated December 15, 1904, however, Russell wrote to Meinong’s about his Über Annahmen: I find myself in almost complete agreement with the general viewpoint and the problems dealt with seem to be very important. I myself have been accustomed to use the name “Logic” for that which you call “Theory of Objects,” and the reasons that you cite against this use on p. 20ff appear to me to be hardly decisive. . . . I have always believed until now that every object must be in some

Russell’s Objections to Frege and Meinong in Context

87

sense, and I find it difficult to recognize nonexistent objects. In a case such as the golden mountain or the round square one must distinguish between sense and reference (in accordance with Frege’s distinction). (J. F. Smith 1985, 347)

It is only in “On Denoting” that Russell was finally able to complete his replacement of Frege’s notions of sense and reference with a uniform referential notion of “meaning.” In the same article, however, he rejected his own notion of “denoting concepts” as well as Meinong’s nonexistent objects as providing alternative accounts of denoting. Russell’s discussion of Meinong in his earlier 1904 paper “Meinong’s Theory of Complexes and Assumptions,” and the 1905 review of a collection of essays by Meinong and his school, attend primarily to the theory of assumptions, and only incidentally discuss the notorious “round square” and nonexistent objects. The example of the round square appears in a discussion of negative judgments, in particular judgments that such things are impossible or do not exist. In the review Russell’s opposes Meinong’s doctrine that distinguishes assumption as the relation to an objective or proposition from the relation of presentation of objects (Vorstellung) in perception or illusion. Russell counters that an assumption is a just the presentation of a proposition.15 This is the same relation of presentation which has as its objects the constituents of a proposition, and lies underpins what is known as “Russell’s Principle” from “On Denoting”: Thus in every proposition that we can apprehend (i.e., not only in those whose truth or falsehood we can judge of, but in all that we can think about), all the constituents are really entities with which we have immediate acquaintance. (Papers 4, 427)

Russell’s Principle expresses the view that the relation of apprehending or being able to think about a proposition, as in an assumption, is the same relation of acquaintance or presentation which relates us to the objects which are constituents of that proposition. It was given the name “Russell’s Principle” by Gareth Evans (1982, 65). Evans explains it this way: “Russell held the view that in order to be thinking about an object or to make a judgment about an object, one must know which object is in question—one must know which object it is that one is thinking about.” However, contrary to this widespread interpretation of Russell’s views, it is clear from examining Russell’s discussions and his notion of “presentation” that acquaintance is primarily an intentional relation to a “transcendent” object, and so a doctrine of the tradition of Brentano that Russell shared with Meinong. It is a direct relation between the mind or subject, and

88

The Bloomsbury Companion to Bertrand Russell

an object of thought, rather than the relation of a knower to sensory experience that is found the empiricist tradition. In the former role there is no intimate connection between acquaintance with an object and being able to “identify” or “re-identify” it, or indeed to “know which” thing it is to any extent. It has no consequences for knowledge of truths about the object, but is an unmediated, direct, relationship between a mind and an object. “Meinong’s Theory of Complexes and Assumptions” was published in successive numbers of Mind. One installment is devoted to Russell’s objections to Meinong’s view that the objectives which we are related to in assumption do not have an independent or “transcendent” existence. Meinong insists that instead they are sometimes only “immanent” to a mental act and so have the “Being” of a Meinongian object such as the round square. Russell starts with a statement of the view to which he objects: Meinong holds—so it would seem—that the object of a presentation is sometimes immanent, but at other times not so; while the object of a judgment—which he calls an Objective, and I call a proposition—is always purely immanent (p. 257) Now for my part I do not see how an immanent object differs from no object at all. The immanent object does not exist, according to Meinong, and is therefore not part of the mental state whose object it is; for this mental state exists. Yet, although not part of any mental state, it is supposed to be in some sense psychical. But it cannot be in any way bound up with any particular mental state of which it is the object; for other states, at other times and in other people, may have precisely the same object, since an object or a proposition can be presented or believed more than once. I confess these facts seem to me to show, without more ado, that objects and propositions must always have being, and cannot be merely imaginary relata for what appears as a relation of presentation or judgment. (Papers 4, 461)

Here is a clear statement of Russell’s view that in judgments the mind is related to a proposition that is an abstract entity with “being” at least, if not the concrete “existence” of particulars. The argument in the body of this part of Russell’s article addresses various possible views about the mind-independent, or “transcendent” existence of Meinong’s “Objectives,” relying crucially on the problem of accounting for false propositions. Russell describes the study of judgment, belief, and the weaker relation of assumption as the principal subject matter of the “Theory of Knowledge,” and lists the alternatives which he will discuss: 1. It may be urged that the knowledge does not differ from what is known, that is, there is no object of knowledge.

Russell’s Objections to Frege and Meinong in Context

89

2. We may admit the distinction of content and object, but hold that the latter is merely immanent. 3. We may hold that the object is immanent when false, transcendent when true. 4. We may hold that when a judgment is false there is no object; but when true, there is a transcendent object. 5. We may hold that the object is always transcendent. (Papers 4, 464) These theories are supposed to be possible with regard to both presentation and judgment, which is the version to which Russell pays most attention. The version of (3) for objects holds that “object is transcendent when it exists, or when it has being (these are the alternatives) and that otherwise is immanent…” Russell says that idealism “oscillates” between positions (1) and (2), that Meinong holds the appropriate variation of position (3) for presentations and (2) for judgments, while, he says, “common sense seems to prefer” (4). Finally, “Frege and Mr Moore advocate (5)” (Papers 4, 464). Russell then considers these positions in order, beginning with his familiar realist objections to idealism for (1). In the discussion of (2) we find a passing criticism of a correspondence theory of truth on Russell’s part. The objection is to a view that truth consists in the correspondence of the object immanent to a mental particular with their transcendent objects in the world. Russell holds that the objects of the judgment or belief are already transcendent. The immanent object of a judgment such as “Peter is walking along the road” when in fact it is John I see, is Peter, not John. Russell proposes instead a correspondence theory of truth in which it is a transcendent “fact” which makes the judged proposition true, thus presenting an early statement of the view that was to persist through his entire “logical atomism” period. Russell’s long-standing interest in the “objects” with which we are acquainted, and of the analysis of judgment as a relation of a judger to a proposition, have their origins in his sympathy for some aspects of Meinong’s “Object Theory.”

4  The origins of “The present King of France” objection In “On Denoting,” published in 1905, Russell presented his theory of descriptions, and argued for it both as a solution to his three “puzzles” about the logical form of sentences using definite descriptions, but also as an improvement on the rival

90

The Bloomsbury Companion to Bertrand Russell

theories of Gottlob Frege and Meinong. It is in only in this paper, published a year later than “Meinong’s Theory of Complexes and Assumptions,” that we first find the famous objections to nonexistent objects. The objections occupy just one paragraph and two passing remarks after that: The evidence for the above theory is derived from the difficulties which seem unavoidable if we regard denoting phrases as standing for genuine constituents of the propositions in whose verbal expressions they occur. Of the possible theories which admit such constituents the simplest is that of Meinong. This theory regards any grammatically correct denoting phrase as standing for an object. Thus “the present King of France,” “the round square,” etc., are supposed to be genuine objects. It is admitted that such objects do not subsist, but nevertheless they are supposed to be objects. This is in itself a difficult view; but the chief objection is that such objects, admittedly, are apt to infringe the law of contradiction. It is contended, for example, that the existent present King of France exists, and also does not exist; that the round square is round, and also not round; etc. But this is intolerable; and if any theory can be found to avoid this result, it is surely to be preferred. (Papers 4, 418)

The only other reference to Meinong occurs near the end of the paper in a comparison with MacColl, who distinguishes between real and unreal objects: This assumes that such phrases as “the present King of France,” which do not denote a real individual, do, nevertheless, denote an individual, but an unreal one. This is essentially Meinong’s theory, which we have seen to reject because it conflicts with the law of contradiction. (Papers 4, 426)

Some correspondence between Russell and Meinong, Rudolf Ameseder and Ernst Mally add to our understanding of the origins of these brief statements in “On Denoting.”16 Russell wrote a review of collection of essays by Meinong and his students, Investigations in Object Theory an Psychology, in the same issue of Mind in which “On Denoting” appeared. There he presents slightly different formulations of his objections to “non-subsistent” entities. It appears that Russell confirmed his assessment of Meinong’s position in correspondence before he wrote the review. In a letter to Russell of December 7, 1904, Meinong says that he will send the four chapters of the volume that were written by himself, Mally, and Ameseder in advance of the publication of the volume. Russell’s response, of December 15, 1904, quoted above thanks Meinong for the first three papers that had arrived at that point, and proposes Frege’s distinction between sense and reference [Sinn] and [Bedeutung] as a way out.17

Russell’s Objections to Frege and Meinong in Context

91

Russell must have written to Mally and Ameseder at the same time with questions about their articles, as he got responses within days that he made use of in his final composition. Ameseder replied in a letter of January 3, 1905. Ameseder’s letter is not mentioned explicitly in the review, or in the notes, but it is clearly the source for Russell’s attribution to Meinong of the view that an apparently non-denoting description such as “the present King of France” nevertheless denotes an unreal object. Ameseder’s letter contains mention of “the present King of France”:18 When I speak of the present King of France [vom,, jetzigen König von Frankreich”], I still speak of something [etwas], and admittedly, of something that is not. Such a something that is not, is a something in spite of that, and hence, an object [Gegenstand], and if that last is not allowed, then I would gladly speak of the “theory of somethings” [Theorie des Etwasse] if that is proper language. In any case, “something” seems completely indistinguishable from “object.”

Ameseder makes an explicit reference to “the present King of France,” asserting that it is a “something” [etwas]. Russell must have presented the example “the present King of France” as an example of a contingently non-existing object, like “the Golden Mountain.” This was different from the impossible “round square” that Meinong himself mentioned. Russell wanted to make sure that the Meinongian account applied to accidentally empty descriptions of ordinary objects as well. Russell’s famous example “the present King of France,” is an “empty,” or non-denoting, definite description. Its first occurrence in a work on logic is in Richard Whately’s Elements of Logic, although it does not serve as an example of an empty description. France was a monarchy in 1826, when Whately’s book was published, making the referent Charles X, the Bourbon king of France from 1824 to 1830. Whately intended the example as a typical singular term. Whately’s Logic was the leading text used in a revival of the study of syllogistic logic, first at Oxford, and then throughout the English-speaking world, being republished throughout the nineteenth century. The example of “present King of France” shows up in a discussion of subject terms, the subjects S of the four kinds of “categorical propositions”: All S are P, No S are P, and so on. Logicians since Russell and Frege have found these categories to be insufficient to handle “singular predications” such as a is P because the logic is based on concepts alone and does not recognize the role of individuals in logic. Whately’s account

92

The Bloomsbury Companion to Bertrand Russell

was that “singular terms” have their origin as special general terms arrived at by a process called “abstraction,” that is to be contrasted with the “generalization” which finds something in common to all entities in the extension by finding something they have in common. Singular terms are of the same logical form as general terms, but just happen to be true of only one individual. They are not arrived at by generalizing over many instances, but by abstracting to find what is called an “individual concept” in later logic: Generalization (as has been remarked) implies abstraction, but it is not the same thing; for there may be abstraction without generalization. When we are speaking of an individual, it is usually an abstract notion that we form; e.g. suppose we are speaking of the present King of France; he must be sitting, standing, or in some other posture; and in such and such a dress, etc. Yet many of these circumstances (which are separable Accidents . . . and consequently), which are regarded as non-essential to the individual, are quite disregarded by us; and we abstract from them what we consider as essential; thus forming an abstract notion of the Individual. Yet there is here no generalization. (Whately, 1827, Book II, Chap. 5, §2)

Whately’s view is that a singular expression such as “The present King of France is P” is of the form “All K are P” where “K” is a concept abstracted from the king leaving only “essential” properties, not, apparently including whether he is sitting or standing, or what he is wearing, but presumably instead only such properties which are “essential to him.” A sentence which may in fact be true, such as “the present King of France is sitting” will be true if it is the case that everything which has that property K, namely the one King of France, has the predicate “sitting.” Whately does not say what the truth-conditions will be if there is in fact no king of France, that is, if the subject-term is “empty.” When Whately’s Logic was first published in 1826, thi was not an example of a non-denoting description. However, the example remained in later editions of the text, well after 1848, when France no longer had a king, but was first a republic and then had an emperor!19 It should have been an issue for readers of the unmodified editions after France no longer had a king! This had been a significant issue on which the “traditional” logic and the new logic differed. Russell says that Leibniz erred in concluding that the syllogism Darapti was in fact valid by ignoring he possibility of empty subject terms, in which case the A proposition, All S are P, will be true. Russell says that his example “All chimeras are animals, all chimeras breath flame, therefore

Russell’s Objections to Frege and Meinong in Context

93

some animals breathe flame” shows this (PLA, 202). There was a long history of discussion of this issue among proponents of syllogistic logic, and, although the text does not mention the issue, it appears that Whately joined in making the “existence assumption” that logic was to deal with only non-empty terms. It seems quite likely that Russell came across the problem of empty singular terms when already considering empty general terms, and somewhat mischievously used Whately’s example when discussing the issue. When Russell composed “On Denoting” in a few days in July 1905, he was able to be confident of his attribution of this view to Meinong on the basis of Ameseder’s explicit statement. There is no mention in Ameseder’s letter of any acknowledgment that these “things that are not” produce violations of the law of contradiction. That was a new argument of Russell’s, and is the basis for the second of his three puzzles in “On Denoting,” the question of the truth-value of “The present King of France is bald.” Russell concludes that on Meinong’s theory the proposition will be both true and false, so the nonexistent king will lead to violations of the “law of contradiction.” Whately would be content to say that the king was neither bald nor not bald, presumably, without realizing that this indeterminacy would lead to a logical difficulty in the new logic with singular terms. The “present King of France” was thus a problem for the theories of definite descriptions of both Meinong and Whately. After the letters and this review, Russell does not seem to have communicated much with Meinong, or his school, about the issue of the nonexistent objects or his own later theories of the nature of judgment and propositions. Russell wrote a short review of Meinong (Russell 1907), and then nothing more. Unlike finding a resolution of the paradox for Frege, which guided Russell’s work for the next decade, the rejection of nonexistent objects was easy for Russell to separate from the range of doctrines upon which he agreed with Meinong.

Notes 1 Technically, a two-place relation R is functional if and only if for any x, y, and z, if Rxy and Rxz, then x = z. 2 See POM, 103 in Chapter X, “The Contradiction.” 3 Alexius Meinong’s student Ernst Mally formulated his logic based on “Object Theory” within Schröder’s system of logic. The issue of distinguishing concepts and objects seems to be behind Mally’s attempts to defend Meinong’s theory against Russell’s objections. See Linsky (2014a).

94

The Bloomsbury Companion to Bertrand Russell

4 The theorem asserts that there cannot be a one-to-one mapping from a set onto the set of its subsets. The proof is by contradiction, then we can find one subset R that is missed by the correlation f. R will be the set of all those objects x that are not members of the subset with which they are matched. We have supposed that there is some element r in X that is mapped onto R. Is r an element of R? This is Cantor’s argument. An application of the same argument to show that there are more classes of classes than classes has f as the identity function, fx = x, Which yields the Russell class as the one that is missed by the correlation. 5 See Moore’s introduction to Papers 3, xxxii. 6 See Grattan-Guiness (1975, 128). Michael Beaney points out that Russell was in fact convinced of this particular case of the distinction by reading Frege, rather than having learned it from Peano, as he later claimed in MPD (p. 52) 7 See Linsky (2004, 15). 8 As indicated above, Michael Beaney points out that many years later Russell inappropriately says that this is something he learned from Peano in My Philosophical Development (Russell 1959, 52). 9 (Frege 1980, 147–70). Frege raises these issues in letters beginning October 20, 1902. 10 See Frege’s “Concept and Object” in Frege (1984). 11 Richard Dedekind, Was sind und was sollen die Zahlen? (1887). The Bertrand Russell Archives also possesses Russell’s copy of Dedekind, dated March 1898, but probably read carefully also only in 1902. 12 Hume’s Principle is generally taken to be the biconditional rather than just this one direction. 13 It is notorious, however, as it was disputed whether it is a logical truth and so not appropriate in a reduction of arithmetic to logical notions. 14 See Dummett (1991, 285–91), for a summary of the technical contents of this section. 15 In notes on Frege’s paper “Function and Concept” made in 1902 in preparation for adding appendix A to The Principles of Mathematics, we find Russell remarking that Frege’s notion of a supposition as opposed to an assertion (a versus ⊢a ) reminds him of Meinong. The comparison appears in the final appendix A at §477. 16 See the head note to “On Denoting” (Papers 4, 14) and (Linsky 2014b) for an account of the composition of the review and surrounding correspondence. 17 Hylton (2010) emphasizes the difference between Frege’s two stage semantics of sense and reference and Russell’s direct semantics of propositional constituents which is made possible by the theory of descriptions in “On Denoting.” Russell was also disagreeing with Frege over this very point in correspondence with Frege.

Russell’s Objections to Frege and Meinong in Context

95

At least for explaining his position to Meinong, however, Russell was happy to urge that something like Frege’s distinction was necessary. 18 Nicholas Griffin has pointed out a reference to discussions with Whitehead over “whether the present King of France is bald” in a letter from Russell to Alys on April 9, 1904, so this is not the first recorded use of this example (Griffin 1992, 277). 19 In the second version, published in 1840 in London by Richard Griffin, at page 25, the example is corrected to “the present ex-King of France,” which was also a denoting description, as the king had abdicated and was still living as the only “ex-king” of France. Further editions of the logic text did not rely on a “factchecker” as had the Encyclopaedia Metropolitana.

References Beaney, Michael. 2003a. “Frege, Russell and Logicism.” In Gottlob Frege: Critical Assessments, edited by Michael Beaney and Erich Reck, vol. I, 213–40. London and New York: Routledge. Revised version of Beaney, M. Beaney, Michael. 2003b. “Russell and Frege.” In The Cambridge Companion to Bertrand Russell, edited by Nicholas Griffin, 128–70. Cambridge: Cambridge University Press. Dedekind, Richard. 1901. Was sind und was sollen die Zahlen?, Braunschweig: Biewig und Sohn, 1887, second edition 1893. Translated as “The Nature and Meaning of Numbers,” in W. W. Beman, ed., Essays on the Theory of Numbers, LaSalle: Open Court. Reprinted Dover Publications, NY, 1963. Dummett, Michael. 1991. Frege: Philosophy of Mathematics. Cambridge, MA: Harvard University Press. Evans, Gareth. 1982. The Varieties of Reference. Oxford: Oxford University Press. Frege, Gottlob. 1959. Die Grundlagen der Arithmetik. Breslau: Verlag von Wilhelm Köbner, 1884. Page references from translation as The Foundations of Arithmetic, by J. L. Austin, Oxford: Basil Blackwell. Frege, Gottlob. 1980. Philosophical and Mathematical Correspondence. Chicago: University of Chicago Press. Frege, Gottlob. 1984a. Collected Papers on Mathematics, Logic and Philosophy. Edited by B. McGuinness, Oxford: Basil Blackwell. Frege, Gottlob. 1984b. “A Critical Elucidation of some Points in E. Schröder, Vorlesungen ueber die Algebra der Logik (1895).” In Collected Papers on Mathematics, Logic and Philosophy, edited by B. McGuinness, 210–28. Oxford: Basil Blackwell. Frege, Gottlob. 2013. Grundgesetze der Arithmetik, Vol. I, Jena: Hermann Pohle, 1893. Reprinted Hildesheim: Georg Olms Verlag, 1998. Translated as Gottlob Frege: Basic

96

The Bloomsbury Companion to Bertrand Russell

Laws of Arithmetic, by Philip Ebert and Marcus Rossberg. Oxford: Oxford University Press. Grattan-Guinness, Ivor. 1975. “Wiener on the Logics of Russell and Schröder: An Account of his Doctoral Thesis, and of his Discussion of it with Russell.” Annals of Science 32: 103–32. Griffin, Nicholas, (ed.) 1992. The Selected Letters of Bertrand Russell, Volume I. Boston: Houghton Mifflin. Husserl, Edmund. 1979. “Review of Schröder (1890).” In Aufsatze und Rezensionen (1890-1910), edited by Bernhard Rang. Husserliana XXII. Springer. Hylton, Peter. 2010. “Frege and Russell.” In The Cambridge Companion to Frege, edited by Michael Potter and Tom Ricketts, 509–49. Cambridge: Cambridge University Press. Linsky, Bernard. 2004. “Russell’s Marginalia in his Copies of Frege’s Works.” Russell: The Journal of Bertrand Russell Studies n.s. 24 (1): 5–36. Linsky, Bernard. Winter 2004–05. “Russell’s Notes on Frege for Appendix A of The Principles of Mathematics.” Russell: The Journal of Bertrand Russell Studies n.s. 24 (2): 133–72. Linsky, Bernard. 2006–07. “Russell’s Notes on Frege’s Grundgesetze der Arithmetik, from §53,” Russell: The Journal of Bertrand Russell Studies n.s. 26 (2): 127–66. Linsky, Bernard. 2013. “Ernst Schröder and Zermelo’s Anticipation of Russell’s Paradox.” In La crise des fondaments: quelle crise? edited by François Lepage and Karine Fredet, 1–13. Montreal: Les Cahiers d’Ithaque. Linsky, Bernard. 2014a. “Ernst Mally’s Anticipation of Encoding.” Journal for the History of Analytical Philosophy 2 (5): 1–14. http://jhaponline.org Linsky, Bernard. Summer 2014b. “Notes and Correspondence for Russell’s 1905 Review of Meinong.” Russell: The Journal of Bertrand Russell Studies n.s. 34 (1): 35–62. Linsky, Bernard. Winter 2013–14. “Russell’s Notes for ‘Meinong’s Theory of Complexes and Assumptions’.” Russell: The Journal of Bertrand Russell Studies n.s. 33 (2): 143–70. Meinong, Alexius. 1983. Ueber Annahmen, Zeitschrift für Psychologie und Physiologie der Sinnesorgane, Ergänzungsband 2, Leipzig: Johann Barth, 1902. Sec. ed., 1910, reprinted as Volume 4 of Alexius Meinong Gesamtausgabe, Rudolf Haller and Rudolf Kininger, eds., Graz, Austria: Akademische Druck, 1977, translated by Jeames Heanue as On Assumptions, Berkeley, Los Angeles, and London: University of California Press. Peano, Giuseppe. 1967. Arithmetices principia, 1889, Translated as “The Principles of Arithmetic, Presented by a New Method,” In From Frege to Godel: A Source Book in Mathematical Logic, 1879-1931, edited by Jan Van Heijenoort, 83–97. Cambridge, MA: Harvard University Press.

Russell’s Objections to Frege and Meinong in Context

97

Rang, B., and W. Thomas. 1981. “Zermelo’s Discovery of the ‘Russell Paradox’.” Historia Mathematica 8: 15–22. Schröder, Ernst. 2001. Vorlesungen ueber die Algebra der Logik (Exakte Logik), Vol I. Leipzig: Teubner, 1890. Reprinted Bristol: Thoemmes. Smith, Janet Farrell. 1985. “The Russell-Meinong Debate,” Philosophy and Phenomenological Research XLV (3, March): 305–50. van Heijenoort, Jan. 1967. From Frege to Godel: A Source Book in Mathematical Logic, 18791931. Cambridge, MA: Harvard University Press. Whately, Richard. 1853. Elements of Logic, 2nd ed., 1827, London: J. Mawman, p. 61, and 9th ed. London: Parker and Son, p. 84. Whately’s Logic was originally published in the Encyclopaedia Metropolitana.

Works by Russell Russell, Bertrand. 1986. [Papers 8], The Philosophy of Logical Atomism and Other Essays 1914-19, The Collected Papers of Bertrand Russell, Vol. 8. Edited by John G. Slater. London: George Allen & Unwin. Russell, Bertrand. 1903. The Principles of Mathematics [POM]. Cambridge: Cambridge University Press, 1903. Second ed., 1937, London: George Allen & Unwin. Russell, Bertrand. 1904. “Meinong’s Theory of Complexes and Assumptions.” Mind 13 (April, July, October 1904): 204–19, 336–54, 509–24. Papers 4, 431–74. Russell, Bertrand. 1905. “On Denoting.” [OD] Mind 14 (October 1905), 479–93. Papers 4, 414–27. Russell, Bertrand. 1905. “Review of Meinong and Others, Untersuchungen zur Gegenstandstheorie und Psychologie.” Mind 14 (October 1905): 530–38. Papers 4, 595–604. Russell, Bertrand. 1907. “Review of: A Meinong, Uber die Stellung der Gegenstandstheorie in System der Wissenschaften.” Mind n.s. 16 (October 1907): 436–39. Papers 5, 689–92. Russell, Bertrand. 1919. “The Philosophy of Logical Atomism.” [PLA] The Monist 28 (October 1918), 495–527; 29 (January, April, July 1919): 32–63, 190–222, 345–80. Papers 8, (157–244). Whitehead, Alfred North, and Bertrand Russell. 1925–27. Principia Mathematica [PM]. Cambridge: Cambridge University Press, 1910–13. Russell, Bertrand. 1967. “Letter to Frege” June 16, 1902, translated in Jan van Heijenoort, ed., From Frege to Godel: A Source Book in Mathematical Logic, 18791931, 124–25. Cambridge, MA: Harvard University Press.

98

The Bloomsbury Companion to Bertrand Russell

Russell, Bertrand. 1992. [Papers 3], Toward the “Principles of Mathematics: 1900-02,” The Collected Papers of Bertrand Russell, Vol. 3. Edited by Gregory H. Moore. London and New York: Routledge. Russell, Bertrand. 1994. [Papers 4], Foundations of Logic 1903-05, The Collected Papers of Bertrand Russell, Vol. 4, Alasdair Urquhart, ed. London and New York: Routledge. Russell, Bertrand. 2014. [Papers 5] Toward “Principia Mathematica”: 1905-08, The Collected Papers of Bertrand Russell, Vol. 5. Edited by Gregory H. Moore. London and New York: Routledge.

4

Russell’s Initial Encounter with Wittgenstein 1911–14 Russell Wahl

1 Introduction Of all the philosophers that Russell interacted with over his life, Wittgenstein stands out the most. Russell’s relation with Wittgenstein was very intense during the years 1912 and 1913. Russell himself many years later said, “Wittgenstein’s doctrines influenced me profoundly,” but then added, “I have come to think that on many points I went too far in agreeing with him” (MPD, 83). Wittgenstein is reported to have said that the happiest hours of his life were spent discussing philosophy in Russell’s rooms. Much has been written on the interaction of Russell and Wittgenstein, with most of the discussion concerning how an understanding of Russell can help one make sense of some of Wittgenstein’s remarks in the Tractatus. Over the last several decades many of these studies have focused on Wittgenstein’s criticism of Russell’s multiple-relation theory of judgment, which Russell abandoned in the light of Wittgenstein’s criticisms. These studies have delved considerably into the details of Russell’s modifications to the multiple-relation theory and the possible meaning of Wittgenstein’s criticism.1 My concern will be less with how this prefigures what Wittgenstein later developed in the Tractatus and instead with the philosophical interaction between the two during the years 1912–13, and with what Russell took away from this interaction. The intense relation between Russell and Wittgenstein was interrupted by Wittgenstein’s departure to Norway in the fall of 1913. Wittgenstein returned briefly to dictate the Notes on Logic but then returned to Norway. Russell went to America in March 1914 and the two were further cut off by the First World War. After the war, Russell met with Wittgenstein in The Hague in December 1919

100

The Bloomsbury Companion to Bertrand Russell

to discuss the Tractatus, and he was instrumental in getting it published. They met again in Innsbruck in 1922. Prior to the end of the war Russell mentioned Wittgenstein several times in his own work, crediting him with making him rethink many of his ideas. However, after 1922 and their meeting in Innsbruck, the two became quite distant. There were brief encounters after that, particularly in 1929 and 1930 when Russell was an examiner on Wittgenstein’s “thesis” (which was actually the Tractatus written several years earlier), and then read some of Wittgenstein’s more recent work to help Wittgenstein get a fellowship. In 1935 Wittgenstein sent him a copy of the Blue Book, but otherwise there was not much interaction between them. Wittgenstein was involved with Russell’s work throughout the early years 1911–14, but was moving in a different direction even with the notes he dictated to Moore in 1914 and later with the Tractatus. In particular, the important distinction between saying and showing in the Tractatus does not stem from any of Russell’s work and was not accepted by Russell. Nonetheless, Russell worked with Wittgenstein in detail through the Tractatus during their meeting in The Hague in 1919, and he wrote a preface to the Tractatus to help it get published. There were some further encounters, and Wittgenstein’s later work often included explicit criticism not only of Russell’s earlier views, but also of positions he developed in his later work. Russell rather brusquely dismissed Wittgenstein’s later work, saying that his later doctrines “have not influenced me at all” (MPD, 83). Russell first encountered Wittgenstein in 1911 when he attended Russell’s lectures on the foundations of mathematics.2 The first indication we have of the content of Russell’s discussions with Wittgenstein revolves around a paper that Russell wrote in 1912, delivered first to an audience in Cardiff in May 1912 and later in Cambridge in October 1912. The paper, “On Matter,” occupied him throughout most of 1912 and into 1913 when he put it down to take up what we now have as the unpublished (until 1984) book, Theory of Knowledge. Wittgenstein and Russell discussed central issues of that work throughout 1913, and in October of 1913 Russell managed to get Wittgenstein to dictate parts of his notebooks, which we now have as the Notes on Logic. These notes captured the work in logic which Wittgenstein had been discussing with Russell, and Russell made use of them in his lectures at Harvard. It is to these that Russell refers when he mentions Wittgenstein in the 1914 Our Knowledge of the External World and his 1917–18 lectures on the Philosophy of Logical Atomism, published in 1918 and 1919 in a series of articles in the Monist.

Russell’s Initial Encounter with Wittgenstein 1911–14

101

Here is a brief time line of Russell’s first encounter with Wittgenstein: August 1911

Russell finishes the manuscript of Problems of Philosophy.

October 1911

Russell’s meets Wittgenstein who is attending his class.

Spring 1912

Russell begins working on “On Matter.”

May 1912

First draft of “On Matter” presented at Cardiff.

October 1912 Revision of “On Matter”; short notes, “What is Logic?” Russell writes reply to Dawes Hicks. January 1913 Wittgenstein writes Russell that he has changed his views on atomic complexes and that the theory of types should be replaced by a theory of symbolism. May 1913

Russell works on Theory of Knowledge.

June 1913

Wittgenstein’s criticism of Russell’s theory of judgment.

October 1913

Wittgenstein dictates the Notes on Logic.

Fall 1913 Russell writes the Lowell Lectures, later revised as Our Knowledge of the External World. January 1914

Russell writes “The Relation of Sense Data to Physics.”

March–June 1914 Russell in Cambridge. Delivers courses on Logic and Theory of Knowledge and delivers Lowell Lectures. April 1914

Wittgenstein dictates notes to G. E. Moore in Norway.

August 1914 Our Knowledge of the External World published. The First World War begins.

2  Wittgenstein and Russell 1911–12: The matter of matter We first see Wittgenstein’s influence on Russell in Russell’s 1912 manuscript, “On Matter.” Russell wrote the paper for a presentation he gave in Cardiff in May 1912 and then revised it over the course of 1912, but gave a different version in Cambridge in October. The manuscript was heavily edited and reveals a development of Russell’s thought. It began embracing a skepticism that Russell said he received from Wittgenstein and in its last draft ended with a suggestion

102

The Bloomsbury Companion to Bertrand Russell

to develop the logical constructions Russell later published in “The Relation of Sense Data to Physics” (RSDP). While working on “On Matter,” Russell wrote in his letters to Ottoline about Wittgenstein, “He is the only man I have ever met with a real bias for philosophical skepticism; he is glad when it is proved that something can’t be known” (BR to OM, May 02, 1912), and eighteen days later after expressing some dissatisfaction with his work on matter, questioned his own belief in the philosophical enterprise, saying “I couldn’t think it valuable. This is partly due to Wittgenstein, who has made me more of a sceptic; partly it is the result of a process which has been going on ever since I found you” (BR to OM, May 20, 1912). A clue to this skepticism may be in the earlier letter just at the onset of writing “On Matter”: I have got a number of new technical ideas from him, which I think are sound and important. . . . I argued about Matter with him. He thinks it a trivial problem. He admits that if there is no Matter then no one exists but himself, but he says that doesn’t hurt, since physics and astronomy, and all the other sciences could still be interpreted so as to be true. (BR to OM, April 23, 1912)

The next day, Russell, who in an important paper in Paris in 1911 had given the name “analytic realism” to his position, said, I expect my paper on matter will be a model of cold passionless analysis, setting forth the most painful conclusions with utter disregard of human feelings. I haven’t had enough courage hitherto about matter, I haven’t been sceptical enough. I want to write a paper which my enemies will call “the bankruptcy of realism.” (BR to OM, April 24, 1912)

Five days later, he said of the paper, “It will shock people, especially those who would like to agree with me—it is altogether too sceptical. Wittgenstein (who has just been here) is delighted, but no one else will be.”3 It is useful to reflect on just what Russell might have taken this skepticism to be. In Problems of Philosophy (POP), Russell had advanced an account of our knowledge of matter which was indirect: we have knowledge by acquaintance of sense-data, but only knowledge by description of matter. Knowledge by description was a species of knowledge of things, but it involved knowledge of truths, to the extent that to know a thing by description required knowing that one and only one item satisfied a description. Russell’s view in 1911 was that while strictly speaking we could not prove that matter exists, nor could we be acquainted with it, yet we could know of matter indirectly as the cause of the

Russell’s Initial Encounter with Wittgenstein 1911–14

103

order and regularity of our sense-data. Russell took the position that matter doesn’t exist to be a skeptical position. He also treated solipsism as though it is equivalent to skepticism. It then seems that the view that there is nothing existing independent of our private sensations, which he takes to be solipsism, that he now characterizes as skepticism. Russell saw “On Matter” as being itself the most skeptical thing he had written, and he reported that Wittgenstein read the beginning and end of his paper and liked it, thinking it “the best thing I have done” (BR to OM, May 22, 1912). He noted, though, that when Wittgenstein read the rest of it, he changed his mind (BR to OM, May 26, 1912). “On Matter” opens with the following bold lines: In what follows, I shall endeavor to maintain three theses: 1. That all the arguments hitherto alleged by philosophers against matter are fallacious; 2. That all the arguments hitherto alleged in favour of matter are fallacious; 3. That, although there may perhaps be reason to suppose that there is matter, yet we can have no means of finding out anything whatever as to its intrinsic nature. (Papers 6, 80)

I conjecture that the first draft of the paper ended in the same vein with this: To sum up what has been said so far: if we mean by “matter” something which exists independently of being perceived, but which can be inferred or known through sensation, then, though the idealist arguments against its possibility are fallacious, the counter-arguments in favour of its existence appear to be equally fallacious. If this is the case, it follows that we cannot know either the existence or the non-existence of matter. (Papers 6, 516)

In the manuscript that we have, this latter passage was crossed out as Russell ended up going in a different direction and not one that pleased Wittgenstein. In his April 24 letter to Ottoline Russell stated that Wittgenstein thought that if matter didn’t exist there would be no problem, since science could still be interpreted as true. This suggests that Wittgenstein, at least at this time, accepted a phenomenalist view where all the truths about what we take to be matter could be understood to be just about sense-data or at least series of sense-data. As the “On Matter” manuscript unfolds, we see Russell initially leaving open the possibility that matter can be defined entirely as a function of sense-data. This possibility is absent in the final text, since eight pages into it Russell asks,

104

The Bloomsbury Companion to Bertrand Russell

“Can Dynamics be defined in terms of sense-data alone? The answer, I think, must be negative.” But this section was added (after f. 18), and seven folios which had been clearly (given the numbering) written earlier were deleted. He eventually decided that dynamics cannot be defined in terms of sense-data (at least of one person in the manner suggested by Wittgenstein’s position) on the grounds that dynamics requires not only the two dimensions of the visual field, which he calls the “verifiable coordinates,” but also the third dimension of distance which he calls the “hypothetical coordinate.” He then clarified his approach by distinguishing two questions (Papers 6, 89): 1. Have we any ground for assuming the existence of anything except sensedata? And 2. Does physics require the existence of anything other than sense-data if it is to be true? Given the remark about the hypothetical coordinate, Russell thought he now had a definite answer to the second question, and that answer was that physics does require more than sense-data. The first question was given an affirmative answer in Problems and the answer was that the existence of matter was the simplest explanation of the order and regularity of our sense-data. “On Matter” clearly rejected this answer and sought without much success a different answer to question (1). It was the negative answer to this with which he began and originally ended the piece. The motivation for this is the fairly long discussion of a priori truths included in the deleted folios 19–26. This discussion may also be a hint to what was behind Wittgenstein’s own concern with the problem of matter and perhaps even why Russell thought Wittgenstein would be delighted with this while no one else would be. Here Russell rejected an extreme empiricism that holds that all knowledge is derived from experience alone. His reason is that we do possess what he calls “logical knowledge.” However, he now made explicit the position that all a priori knowledge is logical knowledge and these are “purely formal principles.” He suggested tentatively that this may include “the principle of induction,” but he argued that it does not include any causal principles, principles of uniformity of nature or “any other general laws which are supposed to distinguish our actual world from other logically possible worlds.” This new position on a priori knowledge undermines the affirmative answer to question 1: “If this view is adopted, nothing can be inferred from sense-data except what they imply logically, and we have seen reason to think that matter is not logically implied by sense-data” (Papers 6, 514).

Russell’s Initial Encounter with Wittgenstein 1911–14

105

Of concern here is what can be logically inferred from sense-data. In POP Russell had maintained that our knowledge of the existence of matter and those relations among items of matter that we can affirm are indirect. He argued that we needed synthetic a priori principles to move from claims about our sense-data to claims about physical objects. Wittgenstein was moving toward the view that logic was quite different from the other sciences, in fact moving slowly toward the position that logic was really tautological, although at this stage he had not quite worked out what that meant. Russell’s doubts about matter, then, can be in part attributed to his conversations with Wittgenstein. His newfound concern is with what constitutes a correct logical inference. In POP Russell had sought to ground a priori knowledge in acquaintance with universals. Wittgenstein’s conception of logic, it seems, is then the source of his doubts about his former solution. In POP Russell had stated that without such principles as the principle of induction, we would have no knowledge beyond our immediate experience. This would be for him a skeptical position. It is likely, then, that the main issue Wittgenstein and Russell discussed (or disputed) involved the nature of logic. Wittgenstein’s letters indicate that the view which he would articulate in the Notes on Logic one year later were already in embryonic form. On this view there would be no place for what Russell would later call “non-demonstrative inference.” Wittgenstein would be delighted with what he read at first from “On Matter,” as it did not countenance any of these inferences. But Wittgenstein had also held that there was no problem with the solipsistic view, since the truths of physics and astronomy could still be interpreted as coming out true. Russell, over the course of his attempting to develop “On Matter,” came to the conclusion that this was not true. He saw that he could not maintain a purely “no-matter” view and still have the sciences. In a part of the manuscript which was eventually omitted from the later draft, there is a deleted sentence: “The scientific and logical problem of matter would be completely solved if matter could be defined as a logical function of sensedata.” This was replaced by again asking the question whether the sense-data are signs of “the existence of something less dependent than they are upon the observer” (Papers 6, 513). This led Russell to return to a discussion of what we normally believe about ordinary physical objects and then to a discussion of the a priori and a critical examination of the attempt to make “same cause, same effect” an a priori principle. All of this was then jettisoned, however, probably because of the turn that Russell made at the end of the manuscript. After rejecting the position that we

106

The Bloomsbury Companion to Bertrand Russell

can find a principle which would allow us to move from sense-data to the matter of POP, Russell said, “What is self-evident is, I think, what crude realism affirms, namely that qualities which are or resemble sense-data, or at least those of sight and touch, exist at times when they are not given in sense” (Papers 6, 93). He ended then with the position on sense-data and sensibilia that he would later adopt in RSDP, although what he would later call sensibilia here he simply called “qualities of the nature of sense-data” which are not necessarily perceived by a single observer or by any observer at all. This view, he said (Papers 6, 94), will “preserve matter, and avoid conflicting with any very strong instinctive belief.” He said he did not yet know whether this view could be maintained, and he recognized that it still employs an a priori principle. The hesitations he mentions at the end of the paper were those that he felt later could be solved by adopting a technique from Whitehead. Had Wittgenstein seen this part of the manuscript, he would have been dismayed, for Russell is retreating from his flirting with Wittgenstein’s view that any a priori principle will be a logical principle in Wittgenstein’s sense, even though what that sense was had not yet been determined.4 What Wittgenstein couldn’t imagine, given his evolving view of logic, was working from sensedata forward to something else. Wittgenstein couldn’t imagine this not because he rejected sense-data but because he rejected the possibility of the inference from sense-data to anything else, in particular any inference to the existence of complexes, or even inferences to any future sense-data, since presumably that would involve a principle of induction Wittgenstein thought to be illegitimate. Russell’s struggle over the “On Matter” manuscript shows how much he was influenced by Wittgenstein, even though as the manuscript unfolds, he retreated from Wittgenstein’s austere position. At the end he did not abandon all nondemonstrative inference, although the principle he now accepted would be confined to inference to objects of the same kind as sense-data. Perhaps it was his unease with these issues that lead him in early 1913 to abandon the “On Matter” manuscript. In a letter to Ottoline Morrell he said that he felt that he must first tackle theory of knowledge.5

3 1913: Theory of knowledge The manuscript “On Matter” found Russell ultimately thinking about logic, and he and Wittgenstein had an ongoing discussion about its nature. Wittgenstein

Russell’s Initial Encounter with Wittgenstein 1911–14

107

had not yet worked out his own theory, but he seems to have come to the position that logic must be wholly different from any other science and in no way depend on something contingent for its grounding. Russell, on the other hand, was rethinking what logic is, but he still thought of logic as the study of the forms of complexes.6 His manuscripts from the time indicate that his views were still very much in flux. Russell began writing Theory of Knowledge in May of 1913. One impetus for the book was his forthcoming visit to Cambridge, Massachusetts, and the courses in theory of knowledge and logic he was to give at Harvard, as well as the Lowell Lectures.7 Another impetus may have been his discussions with Wittgenstein concerning the nature of logic. As Russell initially conceived the book, it would begin with a discussion of acquaintance, proceed to atomic propositional thought and molecular propositional thought and in a second part address knowledge with respect to logic, sense, and science. The book was never completed, and it is generally agreed that Wittgenstein’s criticism of the theory of judgment played a key role in Russell’s abandonment of the project. The reason for this is Wittgenstein’s remark in a letter to Russell in July 1913, “I am very sorry to hear that my objection to your theory of judgment paralyses you. I think it can only be removed by a correct theory of propositions” (LW to BR July 22, 1913, CL 33), and Russell’s recounting of that time to Lady Ottoline Morrell three years later, where he said that Wittgenstein’s criticism was “an event of first-rate importance in my life, and affected everything I have done since. I saw he was right, and I saw that I could not hope ever again to do fundamental work in philosophy” (BR to OM Quoted in Auto, II, 57). What Russell did was in fact abandon the middle part of the book. The first six chapters of the first part were published in the Monist over 1914 and 1915. As the first of these appeared in January 1914, it is unlikely that they were heavily revised, although there is no extant manuscript of them. The rest of that section and the section on atomic propositional thought were written but never published. Russell abandoned the book before part III on molecular propositions. The second section was not written as planned, but the ideas were incorporated into the Lowell Lectures, which he published in August 1914 as OKEW. Despite the remark in Russell’s 1916 letter to Ottoline, it is not easy to see exactly why Russell abandoned his theory of judgment and with it the central part of Theory of Knowledge. What is clearer is that Russell came to see that his

108

The Bloomsbury Companion to Bertrand Russell

project faced technical difficulties and that Wittgenstein’s views on logic ruled out the theory of judgment he was working on. In this section I will focus on Russell’s tinkering with his theory of judgment in Theory of Knowledge to see how Russell had already modified some of his positions from the published version of the theory, and in the next section I will take stock of Russell’s reaction to Wittgenstein. From at least 1910 on, Russell had advocated a multiple-relation theory of judgment. What he meant by this was that in the case of a judgment, a person was not in a dual relation with a proposition, such as J(S,p) but rather in a multiple relation with such items which he had taken to be the constituents of the proposition. Russell’s example illustrating the theory was that if Othello judged that Desdemona loves Cassio, then this fact should be understood as a multiple relation holding among Othello, loving, Desdemona, and Cassio: J(o, L, d, c). While Russell often presented the theory as a means of avoiding false propositions as entities, much of his motivation for the theory appears to be an avoidance of the paradoxes he continued to be troubled by with his substitution theory of classes and relations, which held propositions to be entities.8 The theory was also used in Principia to give an account of the levels of truth in the theory of types. In Theory of Knowledge Russell gave a spirited defense of the multiple-relation theory, focusing on a mistaken judgment that mercury is heavier than gold. The example is taken from Hume, and it is Hume he is criticizing. The passage ends with this: My judging obviously consists in my believing that there is a relation between the actual objects, mercury and gold, not in there being in fact a relation between my ideas of these two objects. Thus the whole nature of belief must necessarily be misunderstood by those who suppose that it consists in a relation between “ideas,” rather than in the belief of a relation between objects. (Papers 7, 140)

Despite this statement of confidence, difficulties emerged concerning the theory. If we take all the positions in the relation J to be positions for logical subjects, then, as Wittgenstein pointed out, the judgment relation in J(s, a,R,b) would permit such judgments as “this table penholders the book.” If we see the theory of types as an ontological theory of different types of entities and think of the places in this relation as being restricted to entities of a specific type, then this objection could be avoided.9 The fact that neither Russell nor Wittgenstein mentioned this seems to be an indication that they still thought of these variables as unrestricted

Russell’s Initial Encounter with Wittgenstein 1911–14

109

variables. In the course of Theory of Knowledge, though, Russell abandoned the view that every entity could be substituted for any other entity in the belief or judgment complex.10 Russell’s development of the multiple-relation theory (now of the relation understanding, which he took to be basic) in Theory of Knowledge is rather complicated. He introduced the logical form of what was being judged as a relatum of the relation so that the way the elements would be conjoined in a fact was included in the understanding. At this stage he didn’t hold that any entity could be substituted for any other entity in a complex, nor was he troubled by questions of the unity of the proposition. His trouble developed when it came to specifying which complexes made the beliefs or understandings true or false. To handle this, Russell complicated the theory even more. The problem he focused on were those beliefs such as that mercury is heavier than gold. What we may think of as the complex or fact which would make this true, he called “permutative,” because a different complex from the very same constituents could be formed by permuting mercury and gold. Such complexes as “A is similar to B” which would make someone’s judgment that A is similar to B true, Russell found unproblematic. These complexes he called “non-permutative.” To handle the permutative complexes, Russell complicated his theory by reflecting on the (non-permutative) relation that each constituent has toward the complex, namely the position in the complex. For example, in the complex mercuryheavier than-gold, mercury has the relation C1 to the complex, and gold has the relation C2 to the complex, whereas in the complex gold-heavier-than-mercury, gold has the relation C1 and mercury has the relation C2 to the complex. So Russell now said that when I have a belief whose objects appear to be heavier than, mercury and gold, what I am really believing is that there is a complex γ in which mercury C1 γ and gold C2 γ (Papers 7, 148). He says of the complex γ that it “cannot be directly named, and does not directly correspond with our belief, or with any possible belief ” (Ibid.). It cannot be named because even if we have acquaintance with it, “a complex name for it must be descriptive, not simply composed of the names of the constituents.” So then “belief only reaches it at the second remove, by corresponding with its associated non-permutative complex” (Ibid.). An interesting thing to note here is that Russell tied his difficulty in specifying the associated complex to the Wittgensteinian concern that the complex cannot be named. Yet Russell at this stage held there can be complex names for nonpermutative complexes, but not for the permutative ones. Grave difficulties for his proposed solution to permutative complexes crop up immediately. For

110

The Bloomsbury Companion to Bertrand Russell

one thing, what seems to be a nonmolecular belief (mercury is heavier than gold) now becomes a molecular belief, and in fact a belief involving a quantifier. Russell at this stage put off discussion of molecular and general beliefs, and discussed alternative theories of truth, but he quickly returned to recognizing problems with his theory: One special objection is that, in order to regard the associated complex as nonpermutative, we have to regard its atomic constituents, x1C1 γ, x2C2 γ, etc., as really its constituents, and what is more, we have to regard the corresponding propositions as constituents of the proposition “there is a complex γ in which x1C1 γ, x2C2 γ, etc.” This seems to demand a mode of analyzing molecular propositions which requires the admission that they may contain false atomic propositions as constituents, and therefore to demand the admission of false propositions in an objective sense. (Papers 7, 154)

It is likely that this difficulty was something Russell did not, in the end, see a way out of. Theory of Knowledge continued for two more chapters, one on selfevidence, which analyzed self-evidence in terms of acquaintance with facts, and another on degrees of certainty, which modified the account of memory in Problems and attempted to cash out uncertain judgments as certain judgments concerning memory images. The book ends there, before any attempt to handle molecular beliefs. Whether something in Wittgenstein’s criticism led him to stop at this point or whether this specific difficulty caused him to abandon the book is difficult to discern. Russell modified the theory from his old one to include a general form of a proposition as a relatum of the multiple-relation, and accepted that the position of the relation R as a relatum of the judgment relation was a different kind than the positions of the particulars. There is nothing in Theory of Knowledge to show that Russell had problems with the basic non-permutative case of someone judging, for example that a is similar to b. As we’ll see in the next section, Wittgenstein would have found difficulty with even that part of the project and Russell came to agree with him. In particular, he changed his position in important ways on two of these issues, acquaintance with universals and facts on the one hand, and the account of propositions and logic on the other. These changes are evident beginning in 1914, with the lectures given at Harvard, reported well by Victor Lenzen, and in Russell’s correspondence in 1914. The account of propositions ultimately is connected with Russell’s reaction to Wittgenstein’s views on logic as these are developed in the Notes on Logic, which Russell had on hand and which he made use of in his course at Harvard.

Russell’s Initial Encounter with Wittgenstein 1911–14

111

Eventually Russell developed a theory of judgment distinct from both his old view and Wittgenstein’s. However, this development didn’t occur until early 1919.

4  Russell’s reaction to Wittgenstein’s criticisms There are two sources of Wittgenstein’s criticisms of Russell that Russell was aware of: Wittgenstein’s letters to Russell and the Notes on Logic. Behind these criticisms lies Wittgenstein’s emerging view of logic as a science totally different from any other science. Wittgenstein was especially concerned with having a correct symbolism which would reveal the special place of logical propositions and would obviate a need for an explicit theory of types. Here is an early letter where Wittgenstein addressed the issue of types: I have changed my view on “atomic” complexes: I now think that Qualities, Relations (like Love), etc. are all copulae! That means I for instance analyze a subject predicate prop, say, “Socrates is human” into “Socrates” and “Something is human” (which I think is not complex). The reason for this, is a fundamental one: I think there cannot be different Types of things! In other words whatever can be symbolized by a simple proper name must belong to one type. (LW to BR, Jan. 1913, CL 24)

There is a sense in which Russell had previously embraced the view that there are not different types of things. As early as Principles of Mathematics he had embraced the position that everything whatsoever was a term, that is, could be the logical subject of a proposition (POM, 43). As he worked through the difficulty stemming from his contradiction, he toyed with the idea of thinking of classes as things of a different logical type, but during 1905–07 had moved in favor of a view that there was just one type of entity, and that the theory of types involved structured variables. This theory, which he called the “substitution theory,” treated all entities of a higher order as logical fictions. This substitution theory is mentioned in his 1908 paper “Mathematical Logic as Based on the Theory of Types,” but is not made explicit in PM. Instead Russell there developed a theory of the order of judgments constructed, it seems, from the bottom up, where the bottom consisted of elementary judgments of the sort “a in the relation R to b” or “a having the quality q” (PM I, 44). Whether the hierarchy of functions and propositions Russell constructs from this base should be considered an ontological theory of types remains controversial.

112

The Bloomsbury Companion to Bertrand Russell

Gregory Landini has argued that Russell did not think that the theory of types in Principia applied to things, but rather held there was an unrestricted variable that ranged over absolutely everything (including universals and relations), and that the type/order restrictions on propositional functions are not types of entities. Others have argued that Russell did have a type theory of relations and universals, although it is not clear that there are ontological differences corresponding to the typing of propositional functions.11 In support of the more austere view, there is a letter from Russell to Ralph Barton Perry, where Russell said with respect to the theory of types, “all except the first are fictions” (BR to RBP November 9, 1913). This letter is late enough that Russell could have changed his position since he composed PM. Even on the austere view, while in one sense Russell didn’t think there were different types of things, there seems to be another sense in which he did. Since POM, Russell held there was a logical difference between universals and particulars, in that universals were able to be predicates in a way particulars were not. Landini, for example, holds that the theory of orders and types of PM is not a theory of ontological types, but he does think that Russell did make this distinction. He has recently introduced the notion of types* which are type differences in things. Differences between complexes and universals and particulars are differences in type*, but this typing was not, according to Landini, coded into the grammar of PM.12 If we adopt the distinction between types and types* for the moment, we can see Wittgenstein as arguing that type* differences must be coded in the grammar too. If we understand “things” to be values of the term variable we can see Wittgenstein’s position as restricting these to objects of the lowest type. By the end of 1913 Russell embraced Wittgenstein’s view on this point and recognized that it was a change in his own position. Here is a remark from a letter he wrote to C. D. Broad before his trip to America and before Wittgenstein had even dictated his further notes to Moore in Norway: The fact is that “acquaintance” cannot be applied straight off except to particulars, K i.e. the only two-term cognitive relation of the form S → O has its converse domain confined to particulars. Knowledge by description is knowledge of a general proposition of the form (∃x)φx. In such cases we cannot properly speak of acquaintance with φx̂, “φx̂” must never be put in subject place, i.e. it must only occur in propositions where it is doing the proper work of a function. Universals, props., function, facts cannot be named, & cannot occur in subjectplaces; they are not “things.” The symbols which are concerned with them are

Russell’s Initial Encounter with Wittgenstein 1911–14

113

never simple, & do not name them. E.g. redness is introduced by “the meaning of ‘x is red’ whatever x may be.” This is Wittgenstein’s theory & I am sure it is right. (BR to CDB Feb. 10, 1914)

Russell went on to say, though, There is, of course, in all such cases, immediate knowledge of the sort one calls “understanding the indefinable,” which I formerly took in a lump as “acquaintance” but in fact it is necessary to distinguish. Immediate Knowledge of any indefinable other than a particular involves apparent variables and the propositional form.

Russell pursued this theme in his Harvard lectures. Victor Lenzen’s notes from March 17 suggest that Russell thought the acquaintance relation came in different sorts and that acquaintance with “actual particulars” was a different relation from acquaintance with facts, which he called perception, and acquaintance with universals. Lenzen’s notes also include the remark, “Never make a verb as subject only particulars subject.”13 Russell had adopted Wittgenstein’s position, and this position rules out anything like his earlier account of judgment. What is not clear is exactly what position Russell put in its place. He said his earlier view is inadequate, but he wanted to have something play the role of acquaintance with facts and universals. In fact, Russell played up in the lectures at Harvard the role of perception of facts. Facts are perceived, but the relation between the perceiver and the fact must be of a different kind from the relation (such as the acquaintance relation) between a subject and a particular. Somehow, also, the fact is not in subject position in the relation. Once one has that relation, presumably the acquaintance-like relation to universals can be retrieved, where there is a relation with a fact where the universal is occurring as verb. Russell didn’t give details as to how this might work.14 Wittgenstein’s criticism of Russell’s theory of judgment was tied to his developing views of logic and of propositions. Both of them had been working on a development of logic based on Sheffer’s work.15 The themes of truth functions and of the bipolarity of propositions are touched on many times in Wittgenstein’s Notes on Logic, and do seem to be at the heart of Wittgenstein’s cryptic remark about his objection to Russell’s theory of judgment.16 Wittgenstein concluded from the various truth table analyses that it was the truth combinations of atomic propositions that are at the heart of logic, and these could be expressed in a variety of truth-functional connectives, none, therefore, of which were sacrosanct, or, therefore, referring to pieces of the world or logical forms in the world.

114

The Bloomsbury Companion to Bertrand Russell

This point, especially given their newfound interest in Sheffer’s work, is then directly related to Wittgenstein’s developing view that the key to propositions, which are what the logical connectives connect, is their bipolarity. At the early stage Wittgenstein had a hard time expressing his view on this, perhaps in part because he was so interested in the development of his a–b diagrams, and in the fact that the diagrams could be set out without specifying which poles were the true poles or the false poles, that is, that the same diagram could represent one way a tautology and the other way a contradiction, that he did not want in advance of the diagrams to specify which poles were which.17 This could well explain why he did not immediately respond whether his a–b diagrams were the same as the true-false ones and why he responded to Whitehead “in a voice of thunder” that a and b were indefinable.18 Wittgenstein said that the Grundgedanke of the Tractatus was that the logical constants do not represent (Tractatus 4.0312). If there is a governing idea of Wittgenstein’s musings in 1913 it is the bipolarity of the proposition and the related view that logic can somehow all be represented by the a–b diagrams. These diagrams were to show that logic is tautologous and to give an account of the function of the logical constants. The logical constants are used to express these truth combinations and so do not represent logical objects. Thus the Notes on Logic are tied to the fundamental thought of the Tractatus. Alternative indefinability shows that indefinables have not been reached.19 Whatever his later connection of this view with the showing and saying distinction, it does not seem to be tied to this distinction in Wittgenstein’s earlier remarks on the position that there are no logical constants.20 Wittgenstein’s concern with the bipolarity of the proposition led him in the Notes on Logic to say that p and ~p have the same meaning (bedeutung) but opposite sense (N 95). It is for this reason that he says that propositions cannot be names. “Names are points, propositions arrows—they have sense. The sense of a proposition is determined by the two poles true and false” (N 101– 02). Propositions are, as Wittgenstein puts it even in the Notes on Logic, facts: “Propositions are themselves facts: that this inkpot is on this table may express that I sit in this chair” (N 97). These remarks about propositions have direct bearing on Wittgenstein’s view of Russell’s theory of judgment. In the “summary” of the Notes on Logic Wittgenstein says, “When we say A judges that etc. then we have to mention a whole proposition which A judges. It will not do either to mention only its constituents, or its constituents and form, but not in the proper order. This shows that a proposition itself must occur in the statement that it is judged”

Russell’s Initial Encounter with Wittgenstein 1911–14

115

(N 94). What we do not see, though, is a critique which involves getting rid of the judging subject as a constituent in the judgment fact. This move had to wait until the notes he dictated to Moore in April 1914.21 Wittgenstein sketched out a theory of judgment in the Notes on Logic which suggested that when A judged that p, there was a relation between A and the two poles required to understand what is the case when p is true and what is the case when p is false, and he had a little diagram with A related to the a–b poles: A ⧸⧹ a -- p -- b

In My Philosophical Development, Russell said, “Wittgenstein maintains that logic consists wholly of tautologies. I think he is right in this, although I did not think so until I read what he had to say on this subject” (MPD 88). What he read was the Notes on Logic, and what he adopted can be seen in his works from 1917 and 1918. In Introduction to Mathematical Philosophy, Russell suggested that what makes purely symbolic or formal propositions logical is that they are analytic or that their contradictories are self-contradictions. He recognizes that this is not fully specified, but calls whatever the characteristic is “tautology” (IMP 203). In a footnote (IMP 205) he says that the importance of “tautology” for a definition of mathematics was pointed out to him by Wittgenstein, but also says that he does not know how to define “tautology.” A similar point was made in the “Philosophy of Logical Atomism,” where he comes closer to Wittgenstein’s position where he says of propositions of logic: Everything that is a proposition of logic has got to be in some sense or other like a tautology. It has got to be something that has some peculiar quality which I do not know how to define, that belongs to logical propositions and not to others.  .  . . They have a certain peculiar quality which marks them out from other propositions and enables us to know them a priori. But what exactly that characteristic is, I am not able to tell you. (Papers 8, 211)

Wittgenstein had himself told Russell that he was not fully clear on what a tautology was: As to what tautologies really are, however, I myself am not yet able to say quite clearly, but I will try to give a rough explanation. It is the peculiar (and most important) mark of non-logical propositions that one is not able to recognize their truth from the propositional sign alone.  .  . . But the propositions of

116

The Bloomsbury Companion to Bertrand Russell

logic—and they only—have the property that their truth or falsity, as the case may be, finds its expression in the very sign of the propositions. (LW to BR, November 1913, CL, 59–60)

Russell no longer defended the multiple-relation theory of judgment after 1913. This issue is confusing, since Russell continued to state that judgment was not a dual relation, for example in his lectures at Harvard, in Our Knowledge of the External World, and later in the “Philosophy of Logical Atomism.” However, there is no continuation of the multiple-relation theory as it was conceived from 1910 until 1913. In his Harvard lectures Russell discussed the proposition “I believe Smith hates Jones.” After asserting that one can never make a verb a subject (Lenzen writes “only particulars subject”), Russell sketched the view that in such propositions as “I believe Smith hates Jones,” there are “two verbs which occur as verbs.” While there is not much to go on, it seems that while Russell recognized that this is not a simple two-place relation, he did not think the proposition should be analyzed as a relation holding between a subject and the constituents of a possible complex.22 He insisted that both verbs occur as verbs. What we have here in his 1914 lectures then is extremely close to the “new beast in the zoo” mentioned in the “Philosophy of Logical Atomism,” where Russell rejected both the view that belief was a two-place relation between a person and a proposition, and the multiple-relation theory, and denied that there was a map of the belief relation (Papers 8, 198). In many ways this theory could be seen as an attempt to capture Wittgenstein’s diagram from Notes on Logic, given above. One difficulty with that view is that he couples the remark with his claim that there is no map of the belief relation. It is unclear how the verb of the judged proposition actually occurs as a verb in the new beast.23 Wittgenstein did not pursue the analysis given above beyond the Notes on Logic as he already had the germ of the Tractatus picture in the notes he had dictated to Moore. Russell also abandoned this kind of analysis with his adoption of neutral monism and the accompanying rejection of the judging subject as a single mental entity. But it was his neutral monism, not Wittgenstein’s work, which led him to this point.

Notes 1 See, for example, Griffin (1985), Wahl (1986), Pears (1987), Landini (1991), Stevens (2005), Hanks (2007), Pincock (2008).

Russell’s Initial Encounter with Wittgenstein 1911–14

117

2 Most of what is known of these early encounters comes from Russell’s letters to Ottoline Morrell given in detail in Blackwell 1981. Russell’s encounter with Wittgenstein more or less corresponds with the onset of his relationship with Ottoline. This was fortunate for posterity, for had Russell not been writing letters every day to Ottoline during this period we would know much less about Russell and Wittgenstein’s early relation. 3 It is curious that when the early encounter with Wittgenstein is recounted, this discussion of skepticism is often omitted. Monk, for example, talks about these letters from Russell to Ottoline, but omits all reference to Wittgenstein’s skepticism. See Monk, 1990, 40–42, and 1999, 251, 263. Michael Potter is an exception here, making reference both to the skepticism and to Wittgenstein’s remark about interpreting physics and astronomy so that they come out true even if matter doesn’t exist. Perhaps Wittgenstein scholars wish to read back into Wittgenstein’s early views the attitude toward skepticism expressed in the late “On Certainty” or even that remark in the Tractatus (6.51) that skepticism is not “irrefutable” but rather “nonsensical” (unsinnig). 4 Here I disagree with McGuinness, who suggests that Wittgenstein was pleased with Russell’s remarks at the end attempting to construct matter out of sense-data and possible sense-data (McGuinness 1988, 107). My own view is that given his views on logic and inference, Wittgenstein could not have been pleased with this and I suggest that it was an earlier version ending with the same skeptical remarks as the paper had begun, with which Wittgenstein was pleased. The remark stemmed from a letter Russell wrote in May. However, it is not clear when all the changes to the manuscript were made, and when Russell read the paper to the Moral Science Club in Cambridge in October, he reported to Ottoline Morrill that the paper was “not a success, it was much too difficult. No one except Wittgenstein understood it all” (BR to OM, October 26, 1912). The next day, in a further letter he reported, “Just before I came to Lausanne, I got a central idea out of which a lot ought to come—I am pleased to find that Wittgenstein thinks just as well of that idea as I do” (BR to OM, October 27, 1912). Unfortunately, he omitted to say what this central idea was. Nonetheless, we can see that Wittgenstein was not entirely dismissive of Russell’s project including what must have been later additions. In January 1913 writing from Vienna while his father was dying, Wittgenstein wrote of his musing on logic (here very much in the spirit of Russell as he focuses on complexes). At the end of this same letter, though, was the remark I mentioned earlier: “I am very interested to hear your views about matter, although I cannot imagine your way of working from sense-data forward.” 5 BR to OM: “I have planned a book on theory of knowledge, then I thought I could do Matter first, but now I see that even apart from having to lecture, theory of knowledge must come first” (Quoted in Blackwell, 1981, 14).

118

The Bloomsbury Companion to Bertrand Russell

6 See “What is Logic” (1912) Papers 6, 55. The forms he has in mind are not just what we now think of as logical connectives, but such things as the subject-predicate form, the form xRy, etc. as well as such forms as (x).φx and (∃x). ϕx. 7 For this and other remarks about the origin of Theory of Knowledge I have used Elizabeth Eames’s Introduction to Papers 7, and Blackwell, 1981. 8 Russell toyed with views in which propositions were not themselves entities and so not values of the individual variable (as they were in his first substitution theory as early as the 1906 papers “Logic in which Propositions are not Entities” and “On the Functional Theory of Propositions, Classes and Relations”). A multiple-relation theory of belief is mentioned in the working notes “Fundamentals,” dated January 1907 (Papers 5, 543). 9 See Wahl 1986 for a way of avoiding Wittgenstein’s criticisms by taking the judgment relation to be a relation holding between different types of things. 10 This is the principle which Pincock (2008, 111) calls (SUB). Pincock correctly points out that Russell had rejected this principle by the time he was writing this part of Theory of Knowledge. In particular, Russell’s discussion of non-permutative heterogeneous complexes (see below) demonstrates that he rejected (SUB) with respect to items which were complexes and those which were not. It is important, as Landini (1998) and Pincock (2008) emphasize, to distinguish these categories of objects from the theory of types Russell developed in Principia Mathematica. For this reason I am not persuaded by the Griffin-Sommerfield account given in Griffin (1985). 11 See Linsky (1999, 32ff). Linsky sees propositional functions as constructed entities and does not identify them with universals. 12 See Landini (1998, 57). 13 These notes are in the Russell archives. The remark about different sorts of acquaintance occurs on p. 3 of the notes from Russell’s lectures. Russell still continued to speak of acquaintance with universals, though. The remark about not making a verb a subject is from p. 43 of the notes. Linsky thinks this note was taken in Russell’s logic class rather than the class in Theory of Knowledge. Russell had given the example “Drinking is agreeable,” and presumably held that a correct analysis would have the verb in predicate, not subject position. 14 Lenzen’s notes indicate that Russell still viewed the various acquaintance relations as fundamental to his epistemology. The notes for April 2 (p. 46 of the pages from the beginning of Russell’s lectures) give us “Perceiving a fact awfully like acquaintance.” 15 While it is hard to date exactly, Russell had written a later manuscript on matter, called “Matter, the Problem Stated.” While listed first of the nine short manuscripts on matter collected in Collected Papers 6, it contains a point Russell developed considerably in his January 1914 writing of “The Relation of Sense Data to Physics,” namely the distinction between the place a datum is and from where it is in that

Russell’s Initial Encounter with Wittgenstein 1911–14

16

17

18 19

20

119

place (Papers 6, 98). This suggests to me that it was written in 1913. In any case, Landini has found that on the verso of the first folio there are truth tables of the Sheffer stroke written in the handwriting of both Russell and Wittgenstein (Landini 2007, 109), and this is something that they were both very interested in. Russell had apparently been sent an advance copy of Sheffer’s paper and had it from April, 1913. I have in mind here Wittgenstein’s remark in his June 1913 letter “I believe it is obvious that, from the prop ‘A judges that (say) a is in the Rel R to b,’ if correctly analyzed, the prop ‘aRb.v.~aRb’ must follow directly without the use of any other premiss. This condition is not fulfilled by your theory” (CL, 29). While there are many different discussions of just what Wittgenstein had in mind here, many of them do agree that the essential point is that what is judged must be either true or false. Rosalind Carey, for example, finds the core problem of his theory to be “Russell’s inability to explain propositional bipolarity within the framework of his theory of judgment” (Carey, 95). Christopher Pincock also sees Russell and Wittgenstein’s commitment to what he calls (T/F), which he characterizes as: “Each judgment has exactly one of the following two properties: truth, falsity” (Pincock 2008, 108), as being the key to Wittgenstein’s objection (Pincock 2008, 132). See also the remark on N 97 that we can use not-q to express what we actually mean by q—presumably the a’s and b’s would be reversed and the schemas would give different values, but nothing corresponds in the world to “not.” The issue concerning whether the a–b poles are the same as the TF poles is born out in the later notes dictated to Moore from April 1914, where after giving a diagram where a is connected to agreeing poles to two p’s and b is connected to an a pole and a b pole : “This symbol might be interpreted either as a tautology or a contradiction. In settling that it is to be interpreted as a tautology and not as a contradiction I am not assigning a meaning to a and b; that is, saying that they symbolize different things but in the same way. What I am doing is to say that the way in which the a-pole is connected with the whole symbol symbolizes in a different way from that in which it would symbolize if the symbol were interpreted as a contradiction. And I add the scratches a and b merely in order to shew in which ways the connection is symbolizing” (N 115). Michael Potter has suggested that Wittgenstein used the a–b instead of W F because using W and F “might reify truth and falsity: it might encourage us to think that truth and falsity are logical objects” (Potter 2009, 174). Russell reports this incident in his Autobiography, Vol. II, 140. He oddly thinks the diagrams show that “or” and “not” are dependent on one another (N 95–96). Presumably what he meant to say is that they are both definable in terms of the Sheffer stroke. Michael Potter also doesn’t think that the showing/saying distinction is prominent in the “Notes on Logic” (see Potter 2009, 194).

120

The Bloomsbury Companion to Bertrand Russell

21 At the end of the notes dictated to Moore, Wittgenstein says that the relation of “I believe p” to “p” can be compared to the relation of “‘p’ says p” to “p.” He then draws the conclusion that I cannot be simple (N 119). 22 Bernard Linsky has told me that in T. S. Eliot’s notes on Russell’s course in logic taught during this same term, Eliot indicates Russell’s reference to “My old theory of judgment,” further confirmation that he had abandoned the multiple-relation theory by early 1914. The notes for the next date (April 11, 1914) also contain a sharp distinction between verbs and names and also Wittgenstein’s view that it is the fact that a verb is in a given relation to other symbols that symbolizes a relation. 23 David Bostock thinks Russell should not have abandoned the theory when he did, as the earlier criticisms do not apply to it. He glosses the new analysis of Othello believes Desdemona loves Cassio as Believes (Othello, [λxy: x loves y] (Desdemona, Cassio) One of the difficulties with using lambda-abstraction here is that this device is normally used to turn a predicate into a term. So it is unclear whether this analysis has kept the verb occurring as verb in the manner that both Russell and Wittgenstein emphasized (Bostock 2012, 215).

References Blackwell, Kenneth. 1981. “The Early Wittgenstein and the Middle Russell.” In Perspectives on the Philosophy of Wittgenstein, edited by I. Block. Oxford: Blackwell. Bostock, David. 2012. Russell’s Logical Atomism. Oxford: Oxford University Press Carey, Rosalind. Griffin, Nicholas. 1985. “Russell’s Multiple-Relation Theory of Judgment.” Philosophical Studies 47: 213–47. Hanks, Peter. 2007. “How Wittgenstein Defeated Russell’s Multiple Relation Theory of Judgment.” Synthese 154: 121–46. Landini, Gregory. 1991. “A New Interpretation of Russell’s Multiple-Relation Theory of Judgment.” History and Philosophy of Logic 12: 37–69. Landini, Gregory. 1998. Russell’s Hidden Substitution Theory. Oxford: Oxford University Press. Landini, Gregory. 2007. Wittgenstein’s Apprenticeship with Russell. Cambridge: Cambridge University Press. Linsky, Bernard. 1999. Russell’s Metaphysical Logic. Palo Alto: CSLI. McGuinness, Brian. 1988. Wittgenstein: A Life: Young Ludwig (1899-1921). London: Duckworth. Monk, Ray. 1990. Ludwig Wittgenstein: The Duty of Genius. New York: The Free Press. Monk, Ray. 1999. Bertrand Russell: The Spirit of Solitude 1872-1921. New York: The Free Press. Pears, David. 1987. The False Prison. Vol. I. Oxford: Oxford University Press.

Russell’s Initial Encounter with Wittgenstein 1911–14

121

Pincock, Christopher. 2008. “Russell’s Last (and Best) Multiple-Relation Theory of Judgment.” Mind 117: 107–39. Potter, Michael. 2009. Wittgenstein’s Notes on Logic. Oxford: Oxford University Press. Stevens, Graham. 2005. The Russellian Origins of Analytic Philosophy: Bertrand Russell and the Unity of the Proposition. New York: Routledge. Wahl, Russell. 1986. “Bertrand Russell’s Theory of Knowledge.” Synthese 68: 383–407. Wittgenstein, Ludwig. 1979. Notebooks 1914-1916 (N). Chicago: University of Chicago Press. Wittgenstein, Ludwig. 1997. Cambridge Letters (CL). Edited by McGuinness and von Wright. Oxford: Blackwell. Wittgenstein, Ludwig. 1922. Tractatus Logico-Philosophicus. Translated by C. K. Ogden. London: Routledge and Kegan Paul.

5

From Russell to the Vienna Circle and Back François Schmitz

1 Introduction In this chapter we shall take “Vienna Circle” in its strict sense, namely those doctrines elaborated in Vienna by the group of thinkers gathered around M. Schlick between 1924—the first year of the seminar directed by Schlick—and 1936. In June 1936 Schlick was murdered and due to the rise of Nazism, the group scattered into several countries, essentially in Great Britain and in the United States. It is during these years that Russell was repeatedly referred to by the members of the Circle, and it was the work of the Circle at that time that Russell came to criticize in the 1940s. It has been frequently said that the Circle’s doctrines were the result of the encounter between classical empiricism (Hume, but also Mach) and the “new logic,” which found its more complete expression in the Principia Mathematica of Whitehead and Russell, published in 1910 and 1913. It is enough to take a quick look at the writings of the most prominent members of the Circle in those years, for example, Schlick, Waismann, Carnap, Neurath, Hahn, to note that the references to Russell are very frequent, and that he was more or less considered as a sort of “spiritual father.”1 This is far from being false, but must be qualified. Most often, the name of Russell appears on lists of authors whom the Circle considered their forerunners, as do Mach, Einstein, Poincaré, and Wittgenstein, among others. However, the philosophical doctrines and conceptions that Russell developed in the years 1900–20 are seldom discussed, just alluded to.2 To understand this, we must remember what the main theses of the Circle are. The “logical empiricism” developed by the Circle rests upon two main theses. First, the so-called logical and mathematical “propositions” say nothing (hence

From Russell to the Vienna Circle and Back

123

they are neither true nor false); second, a proposition is meaningful only if it can, in principle, be “verified” in experience (the so-called “verificationist thesis” of propositional meaning). The first thesis obviates the worries that the existence of mathematics triggered in classical empiricism; the second is a particularly strong formulation of the empiricist requirement since it concerns not only knowledge, but also, and mostly, meaningfulness. These theses came mainly from Wittgenstein’s Tractatus, not from Russell. As is well-known, for Russell, in the first decades of the century, mathematics affords true knowledge about the world, even if very abstract; further, he was not at all prepared to endorse the verificationist thesis. It happens that some of the members of the Circle did not follow Wittgenstein in all of his claims and were even critical of some of the Tractatus’s tenets; this was the case of Carnap, H. Hahn and especially of O. Neurath. Others embraced, almost blindly, not only the doctrines of the Tractatus, but also the new ideas developed by Wittgenstein in the early 1930s: this was the case of M. Schlick and his assistant F. Waismann.3 This division was manifest in the relation they proposed to Russell: while Schlick and (especially) Waismann were ready to share the Wittgensteinian criticisms of Russell’s work, the others, essentially Carnap and Hahn, considered the contributions of Russell to their own thinking much more positively, and even took Russell (somewhat improved by Wittgenstein) as one of their major inspirations. H. Hahn, a prestigious mathematician who was equally interested in the theory of knowledge, was among the first to have recognized the chief importance of the Principia, not only for mathematics, but also for Erkenntnistheorie.4 He introduced the study of Principia in Vienna University during a “memorable” seminar in 1924–25. Carnap, for his part, had studied thoroughly the Principia in the year 1919 and planned immediately to provide the German philosophers and scientists an abridged version of this work. Moreover, the two had been very impressed by Russell’s Our Knowledge of the External World and its use of “Occam’s razor”; the major result of this interest was Carnap’s book Der logische Aufbau der Welt (1928), to which we shall return at some length.5 Concerning the “right wing,” Schlick, who studied Principia only in 1926, made few references to Russell; Waismann, for his part, following Wittgenstein’s criticisms of Russell, wrote some papers in the thirties: against the axiom of reducibility, against the definition of identity in Principia, against the axiomatic presentation of logic, and so on. These papers have the same depreciatory tone as Wittgenstein’s remarks on Russell at that time.6

124

The Bloomsbury Companion to Bertrand Russell

As for the remainder, if Russell is celebrated (particularly by Neurath, the “historian” of the Circle), it is primarily because he made available to the proponents of a “scientific philosophy” a symbolic and logically ordered tool that allows one to address more rigorously certain philosophical problems, and, above all, to show that a good deal of the traditional metaphysics was the result of “bad grammar” and so must be rejected as “non-sensical.”7 On the other hand, how did Russell react to this philosophical movement? Russell repeatedly expressed his sympathy for this movement. He attended the first Congress for the Unity of Science in Paris in September 1935 and wrote a brief report on it in 1936, which begins with these words: “The Congress of Scientific Philosophy in Paris in September 1935 was a remarkable occasion, and, for lovers of rationality, a very encouraging one. . . . My second impression, on hearing the papers and discussions, was one of further surprise, to find that the opinions advocated conformed to this rule [viz. that opinion should be based on evidence]” (Papers 10, 120). And, in the same year, he presented the Vienna Circle “as the youngest and most vigorous offspring of the marriage of empiricism and mathematical logic which took place at the beginning of the century” (ibid., 331). Since this “marriage” had been celebrated between Russellepistemologist and Russell-logician by Russell-philosopher, this appears to mean that Russell was recognizing the Vienna Circle as his “offspring”!8 But, as we shall see, Russell was not ready to embrace the kind of empiricism advocated by the Circle, and he had a completely different conception of what philosophy is. We shall first present the Russellian logical legacy, which concerns essentially Carnap. Second, we shall focus on Carnap’s Aufbau, which came directly from Russell’s OKEW and is the more Russellian (and celebrated) work of the Circle. And third, we shall consider how Russell (critically!) reacted to the two main theses of the Circle.

2  Carnap, the Principia, and logicism Carnap attended some of Frege’s courses on logic in the years 1910, 1913, and 1914, during which time it seems that Frege made various references to Russell’s works. So Carnap began to study Principia in these years.9 In the early twenties, he corresponded with Russell, asking him if he could supply a copy of Principia, which was not available in Germany and in the University of Fribourg’s library.

From Russell to the Vienna Circle and Back

125

In the course of this correspondence, Carnap was clear on an important point: he wished to give the German public an introduction to Russell’s new logic (particularly to the theory of relations), not only for its own sake, but “as a general tool for logicians and epistemologists, applicable in various areas” (ibid., 162). In 1924, he had written a first version of what became the Abriss der Logistik, published in 1929. This book is on the whole Russellian. As indicated by Carnap in its foreword, “it is on the work [of Russell and Whitehead], Principia Mathematica, that the following presentation rests.” And he immediately adds: “Furthermore, I must thank especially Mr. Russell who has personally supported my first studies in this area in an extraordinary serviceable and devoted manner.” Indeed, it is easy to notice that most of the definitions and theorems (given without proofs), as well as the symbolism, found in the Abriss are taken from Principia10 (though this does not mean that the order of exposition faithfully follows the Principia and that the Abriss covers all of Principia!). Three points are notable. In the paragraph dedicated to what is now called the propositional calculus (“deduction theory” in Russellian terms), Carnap, oddly, mixes the Wittgensteinian thesis that the “logical principles” are tautologies (which can be directly checked through truth tables) and Russell’s axiomatic presentation of logic (omitting the Pp *1.5, associativity of the disjunction, which can be deduced from the four other axioms). This is odd because Wittgenstein’s thesis is in strict opposition to the axiomatic presentation adopted by Frege and Russell. This thesis amounts to the claim that “all the propositions of logic are of equal status: it is not the case that some of them are essentially primitive propositions [Grundgesetze] and others essentially derived propositions.”11 More generally, for Wittgenstein to believe that there are “logical propositions” is fallacious: the important thing is not that an implication φ ⇒ ψ is a “logical law” (a tautology), it is that φ and ψ have some structural properties such that ψ follows from φ, that is, that the truth grounds of φ are contained in those of ψ; and in a suitable notation, this can be read from the symbols themselves (cf. Tractatus, 5, 131). If we want to continue to speak of “logical laws,” this can be expressed in a somewhat different way: all “logical laws” are rules of inference in disguise, or there is, in principle, no difference between a logical law and a rule of inference. So, when Carnap mixes the two points of view, this amounts to: “The usual distinction in logic between fundamental and derived propositions (Sätzen) is arbitrary. It is inessential whether a logical proposition is derived from other propositions. Its validity can be recognized from its own form.”12

126

The Bloomsbury Companion to Bertrand Russell

Second, the Abriss not only presents some results of the Principia; it includes a second part, which shows how the logical apparatus of the first part (especially, the theory of relations) can be applied to the axiomatization of various sciences (geometry, arithmetics, set theory, physics, etc.) and also to the theory of knowledge, that is to say to the constitution theory of concepts of the Aufbau. In this, it perfectly agrees with the Russellian claim that the new logic makes it possible to deal rigorously with scientific questions, but also with philosophical ones. In fact, the reader of the Aufbau would be well advised to read the Abriss first! Third, Carnap rejects the ramified type theory of Principia, but gives a prominent place to the simple theory, as he will continue to do thereafter (we shall come back to this topic below). This leads us to consider more generally the relation of Carnap to Principia and its “logicist” thesis as it is expressed in the papers of these years, around 1930.13 The title of the first paper is clear:14 Carnap endorses Frege’s thesis as it is expressed in the first sentence of the Introduction to the Grundgesetze: “In my Grundlagen der Arithmetik I aimed to make plausible that arithmetic is a branch of logic . . . [dass die Aritmetik ein Zweig der Logik sei . . .]” and which Russell expressed a little differently in his Introduction to Mathematical Philosophy: “Logic has become more mathematical and mathematics has become more logical. The consequence is that it has now become wholly impossible to draw a line between the two; in fact, the two are one.”15 In a very classical way, Carnap elucidates the logicist thesis in two points: “Mathematical concepts can be derived from logical concepts. .  .  . The mathematical theorems are part of the logical theorems.”16 However, although Carnap endorses, by and large, the logicist thesis, and, in particular, the logicist definition of the cardinal numbers, he recognizes that there are some still unresolved problems in Principia. Those problems are well-known: they are the presence of the axioms of infinity, of choice and of reducibility and the status of the impredicative definitions. More generally, he compares the logicist program with the other two: intuitionism and formalism, and tries to show that Principia can be saved from the difficulties that the above problems pose and that it is possible to reconcile, at least in part, these three schools.17 It is not possible to enter into all the details here, but Carnap’s position can be summarized in the following way: ●●

As already seen, it is possible, following Ramsey, to do without the ramified type theory, and so the axiom of reducibility may disappear; by contrast, “the simple type theory must be acknowledged.”18

From Russell to the Vienna Circle and Back ●●

●●

127

For the other two axioms (infinity and choice), it is possible to adopt the Russellian solution: when the proof of a theorem φ requires one of these axioms (Ax), the axiom is added as a condition: Ax ⇒ φ. Impredicative definitions are not logically faulty and may be accepted.

As to the constructivist requirement of intuitionism, it is by and large satisfied by logicism19 but must be better integrated. On the other hand, Carnap heavily insists that formalism must be corrected: considering numbers as mere marks on paper makes it impossible to justify logically the application of mathematics to the propositions about the world; for example, it is impossible logically to transform the sentence “Fritz and Karl are the only persons in this room” into the sentence “There are two persons in this room.”20 On this point, Carnap wholly endorses the “application constraint” (as it is now called), which Russell expressed in IMP (9) in this way: “We want our numbers not merely to verify mathematical formulae [as in the Peano axiomatization], but to apply in the right way to common objects. We want to have ten fingers and two eyes and one nose. . . . We want ‘0’ and ‘number’ and ‘successor’ to have meanings which give us the right allowance of fingers and eyes and noses.” Apart from these rather technical points, it must be noticed that Carnap did not limit himself to the logicist thesis; he was greatly piqued by the kind of investigation that the formalists made, and at the end of the twenties, he tried to define some important meta-logical concepts: completeness, categoricity, decidability.21 Even if these attempts were not successful, they foreshadowed what can be called his “syntactic turn,” which ended in The Logical Syntax of Language (1934), whose inspiration is not Russellian, but clearly Hilbertian. This means that he dropped the requirement (related to the “application constraint”) that “every sign of the language, and hence also of the mathematical symbolism, must possess a definite specifiable meaning.”22 Then, in 1934, on the contrary, he states “let any postulates and any rules of inférence be chosen arbitrarily; then this choice, whatever it may be, will determine what meaning is to be assigned to the fundamental logical symbols.”23 Russell in the Introduction to the second edition of POM (1938, xii) rejected this “conventionalist” trend: “Some writers, for example Carnap in his Logical Syntax of Language, treat the whole problem [i.e., what does it mean for a proposition to be true in virtue of its form, and is there a deductive system of logic in which all such propositions are included?] as being more a matter of linguistic choice than I can believe it to be.” Let us come back now to the importance of the simple type theory for Carnap. We cannot do better than to quote what he said in “Die Mathematik

128

The Bloomsbury Companion to Bertrand Russell

als Zweig . . .”: “The simple type theory must be acknowledged. The principal reason for this is less the fact that it enables us to avoid the antinomies, which nevertheless was, at first, the impetus for devising this theory, than the fact that it has as consequences that the concepts of the factual sciences are divided into types and that meaninglessness results from confusion of types.” Hence, two points: first, type theory plays a crucial role in the construction (constitution) of scientific concepts as presented by Carnap in the Aufbau; we shall come back to this in a moment. Second, given type theory, propositions are no longer divided simply into true and false ones, but a third category must be introduced: that of meaningless propositions, due to “confusion of types.” Carnap presented this point in this way in The Logical Syntax (138): “[Frege] was thus obliged to count as false, expressions in which an unsuitable argument was attributed to some predicate. It was Russell who first introduced the triple classification into true, false and meaningless expressions—a classification which was to prove so important for the further development of logic and its application to empirical science and philosophy.” What Carnap had in mind here is obviously the possibility to show that many metaphysical statements are meaningless because they violate the rule of types.24 In “The Elimination of Metaphysics . . .” he said: “Another very frequent violation of logical syntax is the so-called ‘sphere confusion’ of concepts. .  .  . Since the confusion of sphere causes no harm in conversational language, it is usually ignored entirely. This is indeed expedient for the ordinary use of language, but had disastrous consequences in metaphysics. Here the conditioning by everyday language has led to confusions of sphere which, unlike those in everyday language, are no longer translatable into logically correct form. Pseudo-statements of this kind are encountered in especially large quantity, e.g. in the writings of Hegel and Heidegger.”25

3  OKEW, Aufbau: Logical construction and type theory Now, what is the importance of type theory for the constitutional system of the Aufbau? Before answering this question, we must turn to the role of OKEW and Russell’s other papers at the same time (circa 1914), to understand what Carnap wanted to do in the Aufbau. In his “Autobiography,” (13) after having quoted the last lines of OKEW in which Russell expressed his conviction that “in order to secure for philosophy in the near future an achievement surpassing all that has hitherto been

From Russell to the Vienna Circle and Back

129

accomplished by philosophers, [the only condition] is the creation of a school of men with scientific training and philosophical interest,” Carnap adds: “I felt as if this appeal had been directed to me personally.” This took place in 1921. As a result of this appeal, Carnap devoted himself in the following years to the task of constructing all the concepts of daily life and of science on a very narrow basis. Russell in his book of 1914 tried to show how fruitful the use of the new logic was for the clarification and solution of some problems in the field of the theory of knowledge. The basic idea was that what had been done for mathematical concepts in Principia, could be done for concepts appearing in other sciences, especially in physics. In Principia, all the mathematical concepts had been “reduced” to a few logical concepts via “logical construction,” the paradigm of which is the definition of numbers as an equivalence class for the relation of “similarity” (one-one correspondence between classes). This definition is the most famous example of an application of the so-called “abstraction principle” (yet a principle “which dispenses with abstraction,” as Russell put it, OKEW, 51). In OKEW, as well as in the papers of 1914, this possibility of reduction is understood as an application of “Occam’s razor.” By reducing cardinal numbers to classes of classes, a class being considered as a mere “logical fiction” (“no class theory”), “we avoid the necessity of this metaphysical postulate [i.e., the postulate that a cardinal number exists], and thereby remove a needless element of doubt from the philosophy of arithmetic.”26 In the same way, it is possible to consider that the “thing” of common sense is “defined as a certain series of aspects, namely those which would commonly be said to be of the thing.”27 The idea behind this use of Occam’s razor is that we often admit entities as “existing” (for example, cardinal numbers, irrational numbers, constant things (substances), points (in space), instants (in time) and so on). Although these entities do not pertain to what is undeniably given, we feel that a series of aspects are aspects of one same “thing” because they are in some way related to each other. But we have no access to this “thing” apart from this series of aspects— that is, anything we say of it brings us back to those aspects. In Russellian parlance, the constant “thing,” for example, is an “inferred entity,” and in each such case we (philosophers) have to ask ourselves: Is it not possible to construct this inferred entity out of more basic, less hypothetical, entities? So, Occam’s razor is a way of clearing up our “ontological furniture,” and the task of the philosopher is to reduce as far as possible what has to be admitted as inferred entities;28 so “wherever possible, logical constructions are to be substituted for inferred entities.”29

130

The Bloomsbury Companion to Bertrand Russell

It is this formula that Carnap put as the epigraph to his Aufbau. Before comparing the Aufbau’s project with what Russell tried to do in OKEW, some words about the Aufbau are useful. We can, roughly, distinguish within two major trends: first, it is a theory of concept construction; second, it gives an example of what can be achieved using the tools of the first theory. It is this second aspect that has been the most discussed, celebrated or criticized, but it is the first that is more in line with Russell’s project. As seen above, one of the major logical tools of logical construction in Russell’s work is the so-called “abstraction principle,” which had been logically justified in Russell’s great article of 1901 on the theory of relations, published just after he encountered Peano at the Paris Congress of 1900.30 The important thing is that what for centuries had been taken for a psychological process can be treated logically, but with a quite different result. For Aristotle, if we notice that two things have the same “property” (the same “form”), we can conclude that they are similar on this point; for example: “Beings whose quality is one by the form, for example, white .  .  ., are called similar owing to the unity of their form.”31 In this case, we have first a property, being white, and second, the similarity of the beings having this property. For Locke, it was the reverse; for example, when a child looks at its mother, at its nanny, and so on, it can discard their differences and see their resemblances (similarities); then, it “abstracts” from these people their common property (idea): to be women. And, following this line of thought, Peano said, in his Calcolo Geometrico secondo l’Ausdehnunglehre di H. Grassmann: “Every equality between entities of a system other than identity, is equivalent to the identity that is obtained from those entities, abstracting from all the properties that distinguish one entity from the other ones to which it is equal.” For example, let two segments be superimposed on each other; we get, by abstraction, the notion of their magnitude: they have the same magnitude.32 But are we right to suppose that this magnitude is the only entity that we can so abstract from the relation “x is superimposable on y”? Put in more familiar words: if we have an equivalence relation R between a group of things, can we be sure that they share one and only one “common property” in relation to R? Russell discovered that this was not the case, that we must recognize that there is an infinity of entities which can play the role of this supposed “common property.” Consequently, those “definitions by abstraction” à la Peano are not logically legitimate; we are not entitled to affirm the existence of a unique entity (property) “shared” by elements which are in a given equivalence relation. The

From Russell to the Vienna Circle and Back

131

way out of this difficulty is well-known: we can choose one among those possible entities, namely, the equivalence class for R; and then Russell shows something which amounts to: let R be an equivalence relation on D, and S a relation that relates each element of D to its equivalence class for R; then S is many-one, and R  =  S|cS (i.e., for all x, y, xRy iff x and y are related to the same equivalence class, which is at the same time the equivalence class of x, and that of y).33 So we can regard this equivalence class as “the” common property of which we speak ordinarily. Carnap was very impressed with this result. Already, in a short book published in 1926, Physikalische Begriffsbildung, he had made extensive use of this kind of definition, showing how to define some physical magnitudes (for example, weight), starting from an equivalence relation. For the theory of knowledge, the benefit is this: by means of this kind of definition, it is possible to bring back “abstract” notions, which are not empirically given, to what is observable, or, in other words, to construct those notions out of what is empirically given, and so to secure their own empirical legitimacy. So, it is a powerful tool for an empiricist. But can this principle of abstraction be used in every case? No! To apply it we have to find a transitive and symmetrical empirical relation. In the case of weight, for example, the relation that holds between two bodies when they are in equilibrium when put on the scales is transitive and symmetric. But what if the empirical relation is only reflexive and symmetrical but not transitive? Let us take an example: if we have a collection of things, each of them having various colors, we have the relation C: “x is partially of the same color as y.” This relation is clearly reflexive and symmetric, but not transitive. The thing a may be, say, green and red, the thing b, red and blue, and the thing d, blue and yellow: so we have aCb and bCd, but not aCd. So we cannot extract from this relation C, via the principle of abstraction, the “circle” corresponding to green or to red, and so on. It was Carnap’s aim in the Aufbau, to enlarge the procedure of the definitions “by abstraction” to face such cases,—what he calls “quasi analysis.” Of course, he has to make use of the whole theory of relations to achieve such a task, and it is why he so much insisted on the importance of this theory, as we have already noted. In this way, the definitions by abstraction à la Russell become a particular case of a more general procedure. To give a rough idea of what this amounts to, we can quote the Abriss: “If R is a similarity relation [i.e., reflexive and symmetrical], a class α is said to be a ‘similarity circle’ associated with R, if

132

The Bloomsbury Companion to Bertrand Russell

each pair in α is an R-pair and if no element outside α bears the relation R to all the elements of α. . . . The similarity circles associated with an ‘equivalence relation’ are called ‘equivalence circles’ or ‘abstraction classes’.” The difference between these two kinds of “circles” can be put in this way: in the case of an equivalence relation R on a domain D, we define the equivalence class β of an element a of D as (E): β  =  {x  ∈  D: aRx} from which it follows logically that ●●

●●

(1) ∀x, y {(x ∈ β ∧ y ∈ β) ⇒ xRy} (2) ∀y{∃x (x ∈ β ∧ xRy) ⇒ y ∈ β}

But in the case of a similarity circle δ associated with a reflexive and symmetric relation S, we cannot define it in the same way as (E); we have to introduce directly something like the two clauses that Carnap gives above: ●●

●●

(1’) ∀x, y {(x ∈ δ ∧ y ∈ δ) ⇒ xSy} (2’) ∀y{∀x (x ∈ δ ⇒ xSy) ⇒ y ∈ δ}

One easily sees that (1) and (1’) are exactly the same, but not (2) and (2’). We cannot enter into more details here but this may suffice to give an idea of the kind of use Carnap made of the theory of relations in his theory of concept construction.34 This theory of concept construction is to be applied to the second, and more developed, project: to logically construct all the concepts of daily life as well as those of the sciences (not including mathematics). In a way, this project is reminiscent of what Russell tried to do in his writings in the years 1913–14: to construct out of sense-data, from what is given in perception, such concepts as point, instant, etc. In this sense, Russell appears to have endorsed a phenomenalist perspective not very far from that of Mach. In the same vein, Carnap wanted to show that all concepts can be reduced to a primitive relation, that of “recollection of similarity” among the “elementary experiences” of one person (“methodological solipsism”). But it is also clear that this project was much more ambitious than Russell’s, not only because it claims to cover all concepts, but because it aims to show that all concepts can be constructed out of a very limited (“auto-psychological”) basis, and, hence, in particular, that all the sciences constitute a unique body of knowledge: it is the thesis of “the unity of science.” We don’t have the space to enter into the details of this very sophisticated construction, but some of its features, in relation with Russell, can be indicated.35

From Russell to the Vienna Circle and Back

133

First, it is no accident that Carnap chooses as his basis a relation (“recollection of similarity”), the “elementary experiences” being derived from it as its field. The idea behind this is that we can only give a “structural description” of the “objects” that enter into his construction: it is only via such structural descriptions that we evade the purely subjective, uncommunicable, “content” of our experiences. This notion of “structural description” is directly derived from the Russellian notion of the “structure” of a relation (and of the notion of “relation number”). This “structuralism” goes back to Poincaré, but Russell emphasized it in IMP (62) when he noticed that the Kantian distinction between the phenomenal, knowable, world and the unknowable world of “things in themselves,” cannot be maintained, because  “if the hypotheses as stated are correct, the objective counterparts would form a world having the same structure as the phenomenal world,” and so “every proposition having a communicable significance must be true of both worlds or of neither: the only difference must lie in just that essence of individuality which always eludes words and baffles description, but which, for that very reason is irrelevant to science.”36 Second, the construction of the Aufbau can be compared to a “genealogical tree,” each concept being constructed (defined) out of previously constructed concepts, with the result that every proposition, including those constructed concepts, can be translated in a proposition about the sole primitive relation of “recollection of similarity” among elementary experiences. This reductionist trend could easily be misunderstood; does it mean that there is “nothing more” than elementary experiences in relation and so, we are not driven to negate that there are various autonomous kinds of objects: physical, psychological, cultural, and so on? Against this, it seems clear, for example, that we cannot say of a physical object what we say of a psychological one and vice versa: that it is nonsense to say that a stone is sad, or that a pleasure is three meters high. But this would be a misinterpretation of Carnap’s construction. Recall for example the principle of abstraction: we start from a set of “objects,” construct their equivalence classes for a given equivalence relation, and introduce, as an abbreviation, a name for each of these equivalence classes. However, those equivalence classes are not of the same type (in the sense of the Russellian type theory) as the objects out of which they have been constructed (defined). Therefore, we cannot say of those classes what we can say of the objects with which we started. So, when it is said that we can translate each proposition containing the names introduced as abbreviations into a proposition containing only logical symbols and the name of the primitive relation, this simply means

134

The Bloomsbury Companion to Bertrand Russell

that we eliminate those names, but not the constructions themselves. Hence, we can understand that this result of the type theory seemed the most important for Carnap.37 Third, the “phenomenalist” system built into the Aufbau is not the unique one contemplated by Carnap; he insists on the possibility of a physically based system (cf. §59, 62). His choice of a (auto)psychologically based system is motivated by the fact that for him the phenomenalist system respects the epistemic order, or in Russellian terms, by the fact that “verification [of physics] is only possible if physical objects can be exhibited as functions of sensedata.”38 But as Carnap put it (Aufbau, §3, 64), his “methodological solipsism,” the choice of an auto-psychological basis, is precisely what seemed to Russell not feasible (it is for this reason that Russell admitted as unavoidable “inferred entities” the sense-data of others, and the “sensibilia” in RSDP (116, cf. as well MDP, 79). In this respect, clearly, Carnap thought he had achieved a kind of tour de force, having devised “a pure language of the given” (Aufbau, §60), and, in this way, secured the possibility of verifying every proposition containing scientific concepts: “Only the construction formula of the object—as a rule of translation of statements about it into statements about the basic objects, namely, about relations between elementary experiences—gives a verifiable meaning to such statements [i.e., statements about the “constructed” objects]” (§179). All these points can give the impression of a great similarity between the 1914 Russell and Carnap’s Aufbau. And as this last book for some time has been taken as the most representative work of the Vienna Circle, one could think that Russell is the real father (or grandfather?) of logical empiricism. But this is not quite right. What Russell was aiming at in OKEW was to find the less dubious premises of our knowledge, in particular of our knowledge in physics. We start from common or scientific knowledge in order to “examine and purify” it, and soon “it appears . . . that some of it is derivative, while some is primitive.” In the primitive part, we have “what the senses give” (“immediate facts perceived by sight or touch,” 75) and the laws of logic. The task of philosophical analysis is first to discover “what is really given in sense” and then to consider “how the derivative parts of our common knowledge arise.” So, the philosophical work consists in discovering what are the simplest and most certain premises from which common and scientific knowledge can be derived, a work that supposes, in particular, that we pay attention to all the psychological or physiological facts that lead us to the derived part. Thus, the result of all this is rather hypothetical,

From Russell to the Vienna Circle and Back

135

or even artificial, but claims to secure our actual knowledge in bringing it back to what seems the less dubious pieces of knowledge (sense-data, sensibilia, and so on). All this is very far from what Carnap called “rational reconstruction” (rationale Nachkonstruktion): there is no claim here to be faithful to psychological or any other type of facts, even if all the results of actual sciences are used in constructing the concepts, in particular in choosing the suitable indicators for them. This is why the “phenomenalist” system of the Aufbau is no more than one among various possible systems. This rational reconstruction gives us no new knowledge of the world; it limits itself to organizing the concepts appearing in the science in the sense that it gives them their “epistemological” coordinates, or that it gives them their place in the genealogical tree of all the (constructed) concepts, and this can be done in various ways. The result is that concepts which had been historically introduced in the sciences with intuitive and more or less vague meanings are now logically constructed and so endowed with precise meanings; and, conversely, non-constructible concepts (as the “entelechy” of Driesch) are to be excluded from the science. Moreover, as a by-product, if this rational reconstruction is successfully made in a phenomenist system, empiricism could be vindicated. This leads us to more general remarks about Russell’s relations with the Circle.

4  Russell and the circle As we noted at the beginning of this paper, the main thesis of the Vienna Circle came, with some alterations, from Wittgenstein, even if, for example, the Aufbau, with its Russellian flavor, had been written independently of the Tractatus, and was on many points in agreement with it, or, rather, could easily be put in agreement with it. The first thesis, the tautological (or “analytic”) character of the propositions of logic and mathematics, has, as a direct consequence, the position that these propositions lack meaning: they are devoid of content, neither true nor false; they are mere rules of transformation of meaningful propositions into other (equivalent) meaningful propositions: for example, 2  +  3  =  5 is just a rule to transform the proposition “there are 2 apples and 3 pears on this table” into the proposition “there are 5 pieces of fruit on this table” (knowing, of course, that apples and pears are fruit!). Carnap expresses this thesis thus: “The formal science

136

The Bloomsbury Companion to Bertrand Russell

[viz. logic and mathematics] has no independent significance, but is an auxiliary component introduced for technical reasons which facilitates technically the linguistic transformations necessary for the factual sciences.”39 More generally, as Hahn puts it: “Logic thus says nothing whatever about the world, but refers only to the way I speak about the world.”40 It is obvious that this kind of “instrumentalism” is in perfect opposition with the “realist” view of the Principles; it is not then surprising that this Russellian “realism” was entirely rejected by the members of the Circle.41 On Russell’s side, it is well-known that, after “On Denoting,” following Wittgenstein,42 Russell had given up his luxuriant ontology of the Principles, and that, finally, in the “Introduction” to the second edition of POM (1938), he ended up recognizing that the so-called “logical constants” are not “names of objects, as ‘Socrates’ is intended to be,”43 and that “in this way, logic becomes much more linguistic than I believed it to be” (pp. xi–xii); but, in spite of this, at this date, he did not accept the purely instrumental (and conventionalist) view of logic and mathematics advocated by the Circle: he is still in search of a satisfactory account of what it is for a proposition to be “true in virtue of its form,” confessing however that he “is unable to give any clear account of what [this] means” (p. xii) ; and, as we saw, he rejects the purely conventionalist view of Carnap’s Logical Syntax of Language concerning what must be acknowledged as “logic.” It is not very clear (at least for me!) what Russell’s philosophy of logic and mathematics was after the Second World War. He seems to admit, in 1956, that what had been done in the Principia was still held, namely that “it was not now necessary to deny being to them [viz. the natural numbers], but only to abstain from asserting it, that is to say one could maintain the truth of pure mathematics with fewer assumptions than were formerly necessary” (MDP, 163). So, he seems to maintain that mathematical propositions are true44. But a little later, at the end of MDP (157), he says that he had, even if “reluctantly,” embraced the view that  “mathematics consists of tautologies” and, hence, that “to a mind of sufficient intellectual power, the whole of mathematics would appear trivial, as the statement that a four-footed animal is an animal.” It is odd and surprising that Hahn had said almost the same thing in the thirties: “It seems hardly credible at first sight that the whole of mathematics . . . could be dissolved into tautologies. But this argument overlooks just a minor detail, namely the circumstance that we are not omniscient. An omniscient being would, of course, know immediately all that is asserted in asserting several propositions: he would

From Russell to the Vienna Circle and Back

137

know immediately that . . . what is meant by 24 × 31 and by 744 is the same; for an omniscient being needs no logic and no mathematics.”45 Did Russell, in the end, adopt a view on mathematics akin to that of the Circle? Be that as it may, it must be noted that in the 1950 paper “Logical Positivism” (in LK, 367–82), Russell did not discuss the mathematical “instrumentalism” of the Circle, limiting himself to some remarks about the novelties introduced in the thirties by Carnap, Gödel, Tarski, essentially about the need for an hierarchy of languages, and concluded that those topics had become “so technical (. . .) that they can hardly be regarded as belonging to philosophy as formerly understood” (LK, 371). However, he was much more interested in discussing the kind of empiricism, and what it means for a proposition to be empirically meaningful, developed by the Circle, as he had himself been engaged in a reflection on this subject since the years 1935–40. The second thesis of logical empiricism is the principle that a sentence has a meaning if and only if it can be verified, or, more exactly, if we know how it can in principle be verified (verificationism). In its crude form, at the beginning of the thirties, this thesis was formulated by Waismann in this way: “The sense of a proposition is the method of its verification. . . . A statement which cannot be conclusively verified is not verifiable at all; it just lacks all sense.”46 Basically, this applies only to a very particular kind of propositions, the “elementary propositions” of the Tractatus for example, which were called “protocol sentences” in the Circle and are supposed to be directly compared with “the reality.”47 The other sentences which are not directly verifiable must be reduced one way or another, to those “protocol sentences.” For example, in the Aufbau, once sentences of higher levels had been “reduced” to sentences about the “basic objects” (relations between elementary experiences), these last sentences could be directly compared with what is “given” (to a subject), so that the first ones acquire an “empirical meaning.” All this poses a lot of problems which were discussed in the early thirties within the Circle. As a result, Waismann’s above formulation of the verificationist thesis was largely weakened, so that, Carnap, for example, came to substitute “confirmation” for “verification” in his great article of 1935–36, “Testability and Meaning.” Russell, who was a reader of the Circle’s review, Erkenntnis, and read “Testability and Meaning,” discussed this thesis. An Inquiry into Meaning and Truth can be read partly as a refutation of the major thesis of the phenomenal system of the Aufbau (that every (scientific)

138

The Bloomsbury Companion to Bertrand Russell

proposition can be translated without loss in a proposition about the relation of recollection of similarity between elementary experiences and that every “object” (pseudo-object) can be constructed from this basis)48. In Russellian terms, at that time, this means that knowledge is confined to what each of us has as “experiences,” and, even if it is the more genuine empirical knowledge, for Russell we cannot so strongly limit what counts as common or scientific knowledge. As is known, Russell’s method consists in starting from “common sense, except where there was some very cogent reason against it” (MDP, 143) and then in analyzing the kind of “inferences” that are involved in what “we all unquestioningly accept as knowledge” (ibid., 144). In this way, he was, in a sense, less interested in substituting logical constructions for inferred entities (Occam’s razor49) than in codifying the inferences that lead us to believe in the minds of others (and so to rely on their testimonies), to believe that there are “things” more or less permanent, that there are sound waves, and so on. He called those inferences “non-demonstrative inferences,” the principles of which are not grounded on experience, but are used in all knowledge accepted as reasonable.50 Even if Russell continued to admit that these inferred entities could be treated as mere (convenient) fictions, he rejected this proposal on the ground that it is not in this way that we commonly and reasonably proceed; hence it must be acknowledged that nobody believes in pure empiricism (IMT, 305). It is against this background that he criticized the verificationist thesis. Of course, Russell was in agreement with the idea that there are “basic propositions” caused by some perception and, in this way, proved to be true (IMT, 139), propositions without which empirical knowledge would be impossible; but he was not satisfied with the account he found in the Circle’s writings,—or, more accurately, by the absence of an account—of “the relation between a sensible occurrence and an ‘experiential proposition’”51 and more generally of what it is for a proposition to be verified (or verifiable). As is known, a good deal of Russell’s work in IMT was aimed at devising a satisfactory correspondence theory of truth.52 And Russell contended that such a theory “requires much that is psychological and demands some understanding of pre-linguistic mental process” (LK, 381). In other words, it is for lack of a precise answer to the question “What can be learnt from an observation?” that the verificationist thesis is not satisfactory. Their fault is to go “top-down,” starting from “general” sentences instead of starting from the most simple ones (those which record observations), and to

From Russell to the Vienna Circle and Back

139

investigate what knowledge these last ones can convey. If the logical positivists had seriously envisaged this question, they would have been aware that we can learn very little from observations alone and that what they count as a “verifiable” sentence on the basis of observations is not in fact verifiable in the sense that their procedure of verification does not really lead to what is learned from observations alone, but stops before it has reached them. So, as a result of the vagueness of their notion of verifiability, logical empiricists were apt to call “verifiable” sentences that a more precise investigation would show not to be.53 Moreover, if we speak of “verifiable” instead of “verified,” we speak of a future which is indefinitely long. Therefore that a sentence verifiable “is itself not verifiable” (LK, 376, HK, 466). On the other hand, on the basis of “non-demonstrative inferences,” there are many sentences which are obviously meaningful without being “verifiable” in the strict sense: for example, that there are sound waves, that the earth existed before there were human beings, that there are people with whom I shall never be acquainted, that my friend Jones has a mind, and so on. In a word, there are true sentences that are not verifiable, and these true sentences are obviously meaningful. It is in this sense that we must leave pure empiricism. But this does not at all mean that we must reject the somewhat loose claim that a proposition is true “in virtue of some (external) relation to one or more facts” (MDP, 140), and with this, Russell was in conflict with a thesis sustained by Neurath and in part by Carnap, in the early thirties. Neurath was fiercely opposed to the idea that a sentence or proposition could be compared with something like a “reality,” a “given,” or anything else. All we can do is compare sentences with sentences, not with the “reality” (which would be metaphysical absolutism).54 He was thus led to the thesis that what it means for a sentence, even a “protocol sentence,” to be “true,” is simply that it is incorporated without contradiction into a system of sentences already accepted. This does not mean that there are no differences between the protocol sentences and the others. However, the protocol sentences have no claim to be called “absolutely certain” or to be the “ultimate ground” of the truth of the other scientific propositions.55 This leads to a form of “sociologism,” where the community of scientists, at a given time, decides to accept or to reject the sentences (including the protocol sentences) which form the body of the science at that time; there is no external point of view (the “experience”) from which the choices made can be judged.56

140

The Bloomsbury Companion to Bertrand Russell

For his part, after his “syntaxic turn,” Carnap sustained the thesis that philosophy is no more than the logic of science, and the logic of science, no more than the syntax of the language of science, that is to say: the task of philosophy is to study the logical relationships of sentences belonging to a given (constructed) language and not to investigate what it is for a (real) sentence to be verifiable or what kind of (real) sentences are directly verifiable and so on. These last questions are outside the scope of philosophical investigation, even if Carnap obviously admits that, in fact, scientists agree to count such or such a sentence as an “observable sentence” (“protocol sentence”) directly confirmable. However, all this is a matter of convention57. Of course, one of the main tasks of the “logic of science” is to determine, in a given (constructed) language, what could be the relations between the sentences which are not “observable sentences” and those counted as the observable ones. However, given all this, the question of the nature of the “comparison” of a sentence with (empirical) “reality” is left aside. Against Neurath, Russell was clearly on the side of Schlick, and he devoted various writings58 to refute Neurath-Carnap’s thesis. First, he was firmly opposed to the idea that knowledge is only social phenomena, and contended that all (serious) theory of knowledge must start from the question “What do I know?” (a Cartesian question) and not from the question “What does humanity know?”; or to put it in other words: in contrast to Neurath, Russell “maintains that social knowledge is a structure built on the knowledge of individuals, and impossible except on this foundation” (Papers 10, 352). If we have to rely on the opinions of “the scientists of our culture circle” to determine if a given (protocol) statement is accepted or rejected, “the whole basis of empiricism, namely the appeal to experience, is gone” (Papers 10, 351). So Neurath’s thesis (equally endorsed by Hempel at this time) ends in the thesis that the empirical truth is determined by the police (IMT, 148)! One of Neurath’s arguments is that statements and facts are so different that it is impossible to compare them. However, Russell argues, a statement, spoken or written, is a fact as any other,59 and comparing a statement with another statement is comparing a statement with a fact; hence we have to admit either that the verbal facts have a special status among the facts, which is absurd; or that we can compare a statement with a fact, and Neurath’s thesis is rejected. This kind of coherentism, and more generally, the idea that philosophy deals only with language, not with the relation of language to what it signifies, appears to Russell as a new scholasticism as if we were confined in language, forgetting

From Russell to the Vienna Circle and Back

141

“the relation to facts that make a statement true” (LK, 380). He even says that this way of thinking “throws over empiricism and falls into a Platonic mystical belief in The Word” (Papers 10, 359; also, IMT, 23, 149, HK, 262, etc.). With this, we arrive at the last and most crucial point: for Russell, it seems obvious that the aim of philosophy is to “understand the world as well as may be, and to separate what may count as knowledge from what must be rejected as unfounded opinion” (MDP, 161), and if, as philosophers, we are interested in language, it is not for its own sake, but, in the end, to learn something about the most general features of the world: “For my part, I believe that partly by means of the study of syntax, we can arrive at considerable knowledge concerning the structure of the world” (IMT, 347). For the Vienna Circle, following Wittgenstein,60it is just the reverse: philosophy has nothing to say about the world—it is the task of science to do this—philosophy is only concerned with our way of speaking about the world, and to clarify the sense, or meaning, of what we say61. As Schlick once put it: “Science is the pursuit of Truth, and Philosophy is the pursuit of Meaning.”62 But the question of meaning is: What meaning have we given to our words? This question cannot be answered by inspecting the world, and, in answering it, we learn nothing about the world.

Notes 1 We shall see that it would be more appropriate to speak of a “spiritual uncle.” 2 Members of the Circle read AMi and AMa, but these were not important references. 3 It is customary to refer to the first group as the “left wing” of the Circle, and to the second, as the “right wing”; these expressions also have a political sense. 4 According to Menger, Hahn once suggested “that one day Russell might well be regarded as the most important philosopher of his time.” 5 In 1930, Hahn wrote a pamphlet entitled “Überflüssige Wesenheiten (Occams Rasiermesser)” (“Superfluous Entities, or Occam’s razor”) in which he closely follows OKEW in order to defend a strict empiricism. It is part of Russellian orthodoxy, so we shall not comment on it. 6 As those criticisms are essentially Wittgensteinian, we shall not present them in this short paper. 7 For Russell, as it is well known, the undue privilege of the subject-predicate form is the cause of all the metaphysics of “substance.”

142

The Bloomsbury Companion to Bertrand Russell

8 In his 1950 paper, “Logical Positivism” (in LK 381), Russell, after having made some criticisms against the “verificationist” view, ended this paper with these words: “Their procedure, in fact, is more analogous to that of science than to that of traditional philosophy. In this respect I am wholly at one with them. . . . I am more hopeful of results by methods such as theirs than by any that philosophers have employed in the past.” 9 In his “Autobiography” of 1963 (11), Carnap writes: “Around 1919, I studied the great work Principia Mathematica, by Whitehead and Russell, to which Frege had sometimes referred in his lectures.” But in a letter to Russell, on June 13, 1922, Carnap seems to have already studied the Principia when he was Frege’s student before the war: “I do have a relation to your work, however: it was brought to my attention when I was a student of Frege’s in Jena. At that time I studied Principia Mathematica which was available in the library there, and made excerpts from it that have often been of use to me.” cf. E. H. Reck, “From Frege and Russell to Carnap: logic and logicism in the 1920s,” in Carnap Brought Home, 161. 10 By way of example, compare §15 of the Abriss with PM, I, *31 and *35, or §17 with PM I, *38, and so on. 11 Tractatus, 6. 127 ; Wittgenstein will further develop this criticism of the axiomatic presentation of logic in what he dictated to Waismann at the beginning of the 1930s, a criticism that the latter faithfully conveyed in his 1938 article “ Is logic a deductive theory ?” 12 “Die alte und die neue Logik” (“The old and the new logic”), 21–22 (Engl. 141–42), It must be noted that in the thirties, Wittgenstein himself came to the idea that the distinction between the truth table presentation and the Russell’s presentation was a question of convenience. Later on, Carnap insisted that, even if this is not very convenient, we can construct languages in which no “analytic” propositions appear, but only rules of inferences (for example, instead of “p ⇒ (p ∨ q),” we have the rule: every proposition of the form φ ∨ ψ is an immediate consequence of φ); see “Formalwissenschaft und Realwissenschaft,” 34 sq. (Engl. 126). 13 “Die Mathematik als Zweig der Logik”; “Die alte und die neue Logik” (already quoted); “Die logizistische Grundlegung der Mathematik” (“The logicist foundations of mathematics”; see also “Diskussion zur Grundlegung der Mathematik.” 14 The formula “Die Mathematik / Arithmetik als Zweig der Logik” is used on several occasions by Carnap in these years: cf. Abriss, 2, 52, “Die alte . . .,” 20 (Engl. 140). 15 IMT, 194; Carnap quotes this passage at the end of §21 of the Abriss. 16 “Die Mathematik als Zweig . . .,” 298, cf. also, “Die logizistische Grundlegung,” 91–92 (Engl. 41), “Die alte . . .”, 20 (Engl. 140–41).

From Russell to the Vienna Circle and Back

143

17 This is characteristic of the ecumenical spirit of Carnap! 18 “Die Mathematik als Zweig . . .,” 308. 19 For Carnap the Russellian definition of real numbers, for example, is “constructivist”: “The essential point of this method of introducing real numbers [as the ‘lower classes’ of the Dedekindian cuts] is that here those numbers are not ‘postulated’ but ‘constructed’. . . . In similarly constructivist fashion, the logicist introduces the rest of the concepts of mathematics.” “Die logizistische Grundlegung,” 94 (Engl. 43–44). 20 “Die Mathematik als Zweig . . .,” 309 ; see also, “Diskussion zur Grundlegung der Mathematik,” 142; “Autobiography,” 48. The “application constraint” was first formulated by Frege, especially against the first formalists, see, for example, Grundgesetze, II §91–92. 21 Carnap was working on a book about these topics in the years 1928–29; he did not publish it because of some deficiencies pointed out to him by Tarski; but this book is now available: Untersuchungen zur allgemeinen Axiomatik. 22 “Diskussion . . .,” 141. From this requirement, it follows that one must not introduce mathematical entities through “creative definitions” or by “postulates,” see note 19 above. Let us recall Russell’s “bon mot”: “The method of ‘postulating’ what we want has many advantages; they are the same as the advantages of theft over honest toil” (IMP, 71). Carnap quotes it in “Die Mathematik als Zweig . . .,” p. 303. 23 The Logical Syntax of Language, Foreword, xv. From this there follows directly the famous “principle of tolerance,” but that does not mean that we are allowed to forget that mathematics must be applicable; but this question is resolved in considering logic and mathematics as part of a global language containing “synthetic” statements, cf. id, 327. 24 In the Aufbau, Carnap used the word “sphere” as a more general concept than “type,” including the case of nonlogical concepts. 25 “Überwindung der Metaphysik durch logische Analyse der Sprache,” 235 (Engl. 75). 26 “The relation of sense-data to physics” (1914), in ML, 115 (London, Unwin Books, 1963). And Russell continues: “A similar method, as I have shown elsewhere, can be applied to classes themselves, which need not be supposed to have any metaphysical reality, but can be regarded as symbolically constructed fictions.” 27 OKEW, 112. 28 In “The relation of sense-data . . .,” concerning the philosophy of physics, two kinds of inferred entities are admitted: the sense-data of other people and “sensibilia.” 29 Id, 115. 30 “La Logique des Relations,” Rivista di Mathematica, vol. vii, 1900–01, 115–48; English translation by R. C. Marsh, in LK, 3–38 (also, Papers 3, 314–49).

144

The Bloomsbury Companion to Bertrand Russell

31 Metaphysics, I, 3, 1054b 8. 32 It is in the same way that Cantor famously defined the “power” of a set: “We will call by the name ‘power’ or ‘cardinal number’ of M the general concept which, by means of our active faculty of thought, arises from the aggregate M when we make abstraction of the nature of its various elements m and of the order in which they are given.” 33 “S|cS” means the relative product of S and its converse. 34 Apart from these two types of “circle” (“similarity circles” and “equivalence circles”), Carnap defines a third kind of class based on the relation: x has a constituent that is similar to a constituent of y (for example: x has a color similar (but not identical) to a color of y). 35 For a detailed analysis of this construction and of its logical apparatus, see N. Goodman’s classic Structure of Appearance, chap. V. 36 Carnap refers to this text at the end of §16 of the Aufbau, adding that “it was not until Russell that the importance of structure for the achievement of objectivity was pointed out.” This Russellian structuralist remark was rather popular among the members of the Circle: Schlick quotes it at least three times, for example. 37 Carnap emphasized this very important point in §41 of the Aufbau. 38 RSDP, 109. 39 “Formalwissenschaft und Realwissenschaft,” 35 (Engl. 127). 40 “Diskussion zur Grundlegung der Mathematik,” 137. 41 Already, in his General Theory of Knowledge (1918–25), Schlick classified the Russell of The Problems of Philosophy among the Platonists, and criticized him as such. Hahn said in the paper quoted above (138): “Russell’s absolutist-realist point of view supposes that the world consists of individuals, properties of individuals, properties of such properties, and so on; and that the logical axioms are now assertions about this world. I have already said that this conception is incompatible with a consequent empiricism.” 42 But, as is well-known, this does not prevent Russell from contending in his Introduction to Mathematical Philosophy from 1919 (169) that logic is concerned with the real world just as truly as zoology, though with its more abstract and general features,” while he ends this book saying, following Wittgenstein, that all logical propositions “have the characteristic which, a moment ago, we agreed to call ‘tautology’” (204). But, he confesses that he does not know what the precise meaning of “tautology” is. 43 It was still Russell’s conviction in the (unpublished) Theory of Knowledge (1913) that “such words as or, not, all, some, plainly involve logical notions; and since we can use such words intelligently, we must be acquainted with the logical objects involved” (Papers 7, 99).

From Russell to the Vienna Circle and Back

145

44 However, in IMT (240), Russell says that “the meaning of truth as applied to tautologies is different from its meaning as applied to empirical sentences.” But what is the difference? 45 “Logik, Mathematik und Naturerkennen,” 160 (Engl. 36). Schlick, and Ayer made similar remarks. 46 “A logical analysis of the concept of probability,” 229 (Engl. 5). 47 As it is well-known from this thesis, it follows that all the “sentences” of traditional metaphysics are meaningless; they have the same status as “Caesar is and,” mere strings of sounds or of written marks. 48 However, the general aim of IMT is clearly to refute the verificationist thesis in its strong form, cf. IMT, 22. 49 After his “linguistic turn,” Russell put more emphasis on the question of “minimum vocabularies” than on the use of Occam’s razor. 50 In IMT Russell alludes to this kind of inferences, but it is in HK that he studies those inferences exhaustively, distinguishing five “postulates” involved in “nondemonstrative inferences,” cf. HK, part 6, MDP, chap. 16. 51 Papers 10, 332. Russell made this remark in his review of Ayer’s book. 52 Roughly: the truth of the basic propositions depends on a sensible occurrence, and that of the remaining propositions depends on their (complex) syntactic relations to the basic propositions. 53 See for example, LK, 373–74, HK, 463–64. It must be noted that some logical empiricists, like Hempel and even Ayer, were aware of this kind of difficulties. 54 See, for example, “Soziologie im Physicalismus,” 541 (Engl. 66): “Statements are compared with statements, not with “experiences,” not with a “world” nor with anything else. All these meaningless duplications belong to a more or less refined metaphysics and are therefore to be rejected.” 55 We have no room to develop those points, but Neurath’s conception is less strange than it may seem. 56 For Schlick (“right wing”), this is absurd; Neurath’s conception is no more than a “coherentist” theory of truth, and if we accept it, we can no longer distinguish a fairy tale from a scientific theory. As empiricists, we must acknowledge the possibility of comparing a sentence with an experience because it is the unique source of truth. But this leaves open the problem: What form must a “protocol” sentence” have to be directly comparable with something given in experience? Clearly, when I say: “The ashtray is on the table,” neither the ashtray, nor the table is “given” to me ; what is given is an aggregate of colors, odors, forms, etc. Admitting this, Schlick came to his notion of “Konstatierungen,” sentences of the form: “here, now, blue” pronounced at the very instant when blue is observed now and here. This sentence is “true” at this instant, but becomes only a hypothesis an instant later. However, these “Konstatierungen” are

146

The Bloomsbury Companion to Bertrand Russell

supposed to be the “absolute ground” of all knowledge. It is clear that he was here entangled in difficulties, precisely those difficulties which Russell tried to overcome in IMT. 57 See, for example, §82 of The Logical Syntax of Language. The difference between Carnap and Neurath is that Carnap does not deny that, in fact, we compare statements with “experiences” or with “facts” or anything else, but this is a “zoological fact,” falling outside the syntax of the language of science. 58 “On verification” (1938), in Papers 10, 345–59, chapter x of IMT, “Logical Positivism,” in LK, 380–81, MDP, 185–86, 272–74. 59 The strange thing is that Neurath was equally of the opinion that language is on the same footing as any kind of physical phenomena and that it can be “behaviorally” studied. 60 Cf. Tractatus, 4.112: “Philosophy does not result in ‘philosophical propositions,’ but rather in the clarification of propositions.” 61 And this can be done not only by “philosophers” but also by scientists interested in the foundations of their disciplines, the paradigmatic example being Einstein’s clarification of the notion of “simultaneity at distance.” 62 “Form and Content,” 367.

References Awodey, S., C. and Klein (eds.) 2004. Carnap Brought Home. Chicago, Open Court. Carnap, R. [1926] 1966. Physikalische Begriffsbildung. Darmstadt: Wissenschaftliche Buchgesellshaft. Carnap, R. [1928] 1967. Der logische Aufbau der Welt [1928], Hambourg, F. Meiner Verlag, 1966 (2nd ed.), English translation by R. A. George. The Logical Structure of the World, Berkeley and Los Angeles: University of California Press. Carnap, R. 2000. Untersuchungen zur allgemeinen Axiomatik. Edited by Th. Bonk and J. Mosterin. Darmstadt: Wissenschaftliche Buchgesellshaft. Carnap, R. 1929. Abriss der Logistik, mit besonderer Berücksichtigung der Relationtheorie und ihrer Anwendungen (A summary of logic, with a particular attention to the theory of relations and its applications). Vienna: J. Springer. Carnap, R. 1959. “Die alte und die neue Logik.” Erkenntnis, Bd. 1, H. 1, 1930, 12–26; English translation in Logical Positivism. Edited by A. J. Ayer, 133–46. New York: The Free Press [Ayer]. Carnap, R. 1930–31. “Die Mathematik als Zweig der Logik” (“Mathematics as a Branch of Logic”). Blätter für Deutsche Philosophie 4: 298–310. Carnap, R. 1983. “Die logizistische Grundlegung der Mathematik” (The Logicist Foundations of Mathematics), in Erkenntnis, vol. 2, 1931, 91–105; English translation

From Russell to the Vienna Circle and Back

147

by E. Putnam and G. J. Massey. In Philosophy of Mathematics, Selected Readings, edited by P. Benacerraf and H. Putnam, 41–52. Cambridge: Cambridge University Press. Carnap, R. 1931. “Diskussion zur Grundlegung der Mathematik.” Erkenntnis 2: 135–51. Carnap, R. 1932. “Überwindung der Metaphysik durch logische Analyse der Sprache” (“The Elimination of Metaphysics Through Logical Analysis of Language”). Erkenntnis, Bd. 2, H. 4, 219–41; English translation by A. Pap, in [Ayer], 60–81. Carnap, R. [1937] 1967. The Logical Syntax of Language. London: Routledge and Kegan Paul. Carnap, R. 1953. “Formalwissenschaft und Realwissenschaft” (“Formal and Factual Science”), Erkenntnis, Bd. 5, H. 1, 1935, 30–37; English translation by H. Feigl and M. Brodbeck. In Readings in the Philosophy of Science, edited by H. Feigl and M. Brodbeck, 123–28. New York: Appleton-Century-Croft. Carnap, R. 1963. “Autobiography.” In The Philosophy of Rudolf Carnap, edited by P. A. Schilpp, 3–84. La Salle, IL: Open Court. Frege, G. 2013. Grundgesetze der Arithmetik (I et II) (Basic Laws of Arithmetic), Hildesheim, G. Olms, 1966; English translation by P. A. Ebert and M. Rossberg. Oxford: Oxford University Press. Hahn, H. 1980. “Überflüssige Wesenheiten (Occams Rasiermesser)” [1930] (“Superfluous Entities, or Occam’s Razor”). In Empirismus, Logik, Mathematik, edited by B. McGuinness, 21–37. Stuttgart: Suhrkamp, 1988; English translation in H. Hahn, Empiricism, Logic, and Mathematics, edited by B. McGuinness, 199–217. Dordrecht: D. Reidel (Vienna Circle Collection, n 13). Hahn, H. 1987. “Logik, Mathematik und Naturerkennen” [1933] (“Logic, Mathematics, and Knowledge of Nature“), in id. 141–72; English translation by H. Kaal. In Unified Science, edited by B. McGuinness, 24–45. Dordrecht: D. Reidel (Vienna Circle Collection n 19). Neurath, O. 1983. “Soziologie im Physicalismus” [1931] (“Sociology in the Framework of Physicalism”). In Gesammelte philosophische und methodologische Schriften, edited by R. Haller et H. Rutte, 533–62. Vienne, Hölder-Pichler-Temsky, 1981; English translation by R. S. Cohen and M. Neurath in O. Neurath, Philosophical Papers, 1913-1946, edited by R. S. Cohen and M. Neurath, 58–90. Dordrecht: D. Reidel (Vienna Circle Collection, n 16). Schlick, M. 1974. Allgemeine Erkenntnislehre [2nd ed., 1925] (General Theory of Knowledge), Berlin: J. Springer, 1925; English translation by A. E. Blumberg, Vienna: Springer. Schlick, M. 1979. “Form and Content” [1932] in Philosophical Papers, vol II (1925–36), edited by H. Mulder and B. van de Velde-Schlick, 285–369. Dordrecht: D. Reidel (Vienna Circle Collection, n 11, vol. 2).

148

The Bloomsbury Companion to Bertrand Russell

Waismann, F. 1977. “Logische Analyse des Wahrscheinlichkeitsbegriffs” (“A Logical Analysis of the Concept of Probability“), Erkenntnis, Bd. I, 1930–31, 228–48; English translation in F. Waismann, Philosophical Papers, 4–21. Dordrecht: D. Reidel, ed. (Vienna Circle Collection, n 8). Wittgenstein, L. 1969. Tractatus Logico-Philosophicus (German text with English translation by D. F. Pears and B. McGuinness). London: Routledge and Kegan Paul.



Part Two

Philosophical Issues

150

6

Russell’s Logicism Kevin C. Klement

1 Introduction Logicism is typically defined as the thesis that mathematics reduces to, or is an extension of, logic. Exactly what “reduces” means here is not always made entirely clear. (More definite articulations of logicism are explored in Section 5.) While something like this thesis had been articulated by others (e.g., Dedekind 1888 and arguably Leibniz 1666), logicism only became a widespread subject of intellectual study when serious attempts began to be made to provide complete deductions of the most important principles of mathematics from purely logical foundations. This became possible only with the development of modern quantifier logic, which went hand in hand with these attempts. Gottlob Frege announced such a project in his 1884 Grundlagen der Arithmetik (translated as Frege 1950), and attempted to carry it out in his 1893–1902 Grundgesetze der Arithmetik (translated as Frege 2013). Frege limited his logicism to arithmetic, however, and it turned out that his logical foundation was inconsistent. Working at first in ignorance of Frege, Bertrand Russell’s interests in the fundamental principles of mathematics date back to the late 1890s. He published An Essay on the Foundations of Geometry in 1897 and soon thereafter began work on the nature and basis of arithmetic.1 He found his work hampered somewhat by a mismatch between his earlier generally Kantian views on mathematics and the staunch realist metaphysics he had adopted at the end of the 1890s largely under the in influence of G. E. Moore.2 This tension ended when Russell attended the International Congress of Philosophy in July 1900. There Russell became acquainted with Giuseppe Peano and his work, which Russell describes as one of the most significant events of his philosophical career (MPD, 11). After being exposed to and quickly mastering Peano’s style of

152

The Bloomsbury Companion to Bertrand Russell

symbolic logic, Russell became convinced that it was possible to analyze all the concepts of pure mathematics in logical terms, and provide deductions of them from logical axioms.3 The result was Russell’s first major work of lasting significance, The Principles of Mathematics (POM), published in 1903. In it, Russell laid out his views on philosophical logic and argued informally for logicism. He described the project as covering in part the proof that all pure mathematics deals exclusively with concepts definable in terms of a very small number of fundamental logical concepts, and that all its propositions are deducible from a very small number of fundamental logical principles. (POM, v; cf. §434)

His initial plan was to carry out the actual deductions symbolically in a second volume, and recruited his Cambridge colleague and former teacher, A. N. Whitehead, as a coauthor. By the time their technical work was ready for publication, it had grown so large, and their views had changed so significantly in response to certain paradoxes (see Section 3), that they decided to rename their work Principia Mathematica (PM). It was published in three large volumes from 1910 to 1913. (A proposed fourth volume, which was to be Whitehead’s responsibility primarily, was never produced.) Russell later describes the process as having been exhausting, and seems to have been ready to retire from technical work in mathematical logic after its publication. Nonetheless, in 1919, he published Introduction to Mathematical Philosophy, which provides an informal introduction to the theories and results of PM, and in the mid-1920s, Russell oversaw the publication of a second edition of PM, wherein he added a new introduction and various appendices exploring certain new theories and made note of what he saw as important contributions made by others in the intermediate decade. While Russell did not write any additional lengthy pieces on the philosophy or foundations of mathematics after the second edition of PM, comments on various issues concerning the nature of mathematics are scattered among his other writings. Russell’s attempt to establish logicism in PM, and the logical system developed therein for the purposes of the project, have had a remarkable impact on the history of logic and philosophy. Many of the most important figures in the history of analytic philosophy (from Wittgenstein and Carnap to Quine and Putnam, etc.) describe Russell’s logical work as a significant in influence. The logic of PM is perhaps the single greatest inspiration for the kind of predicate logic nearly all philosophy students are now required to learn. While it is not

Russell’s Logicism

153

widely accepted today that PM in fact succeeded in establishing logicism, I think a case that it did can still be made.

2  The regressive method and its prior successes According to Russell, progress in mathematics can proceed in either of two directions (IMP, 1–2). Most research proceeds in the “forward” or “constructive” direction, in which new results are proven from previously known findings. In this way, one attains the results of newer, more complex, and “higher” mathematics. Russell saw his own work, however, as proceeding in a different direction, employing a “regressive” or “analytic” method. Here, one begins with a body of knowledge already accepted as at least mostly true. The goal is to work backward from these known results to a deductive basis for them, a set of principles more general, less complex, and employing a smaller undefined vocabulary. If it can be shown that the original body of knowledge—or at least all of it deemed worth preserving—can be deductively recovered from this more austere basis, we gain insight into the real nature of the truths. Logical connections between the terms employed are revealed. The new basis might also remove puzzling or unwanted aspects of the original theory. Finally, the process organizes the new theory as a deductive system, which invites and facilitates the discovery of new results.4 Russell portrays the truth of logicism as something emerging from alreadyestablished results in the “regressive” or backward journey from accepted mathematical principles to their deductive origins. Russell situates his work as the next in a series of recent successes in that direction (see Papers 5, 574–78, IMP: chaps. 1–2). According to Russell’s story, a number of late nineteenth-century mathematicians, including Dedekind, Weierstrass, and the members of Giuseppe Peano’s Italian school, had succeeded in “arithmetizing” most of pure arithmetic, and even many areas of pure geometry. That is, they showed how these areas of mathematics could be deductively recovered starting only with the basic principles governing natural numbers. Moreover, Peano and his associates had made further “regressive” progress in number theory by reducing its basis to five principles, now known as the Peano-Dedekind axioms: 1. Zero is a natural number. 2. Every natural number has a successor, which is also a natural number.

154

The Bloomsbury Companion to Bertrand Russell

3. Zero is not the successor of any natural number. 4. No two natural numbers have the same successor. 5. Whatever holds of zero, and always holds of the successor of any natural number of which it holds, holds of all natural numbers. (The principle of mathematical induction.) At the time, it was widely believed that all of pure arithmetic could be derived from these principles together with the usual assumptions of logic and set or class theory (and even today, it is generally agreed that no axiomatic system for number theory represents an improvement on this). However, according to Russell, the work of Peano’s school had not completed the backward journey to the basic principles of mathematics, because an even more minimal basis is possible. Russell portrays Frege as having taken the next step in this direction, by showing how what amount to Peano’s “axioms” for number theory could be derived from logical axioms, removing the need for any basic assumptions beyond the purely logical. Achieving this next step required first a way of defining the nonlogical terminology used in Peano’s postulates: “zero,” “successor,” and “natural number.” There are in fact different ways of defining these, but let us first focus on natural numbers considered as cardinal numbers. Taking Frege’s talk of “extensions of concepts” in his Grundgesetze as interchangeable with talk of “classes,” Frege defined the cardinal number of class α as the class of all classes whose members could be put in 1–1 correspondence with the members of α. Zero could then be defined as the cardinal number of the null class. The relation of successor could be defined as holding between n and m when there is a class α and member of that class x such that n is the number of α and m is the number of the class consisting of all members of α except x. In that case, the members of n are all those containing one more member than contained in members of m. The natural numbers could be defined as all those (cardinal) numbers following zero in the series created by the successor relation: zero, zero’s successor (1), zero’s successor’s successor (2), and so on (Frege 2013, §§41–46; cf. Frege 1950, §§73–83). In particular, n is a natural number when it has every property possessed by 0 and always passed from something having it to that something’s successor. Frege also provided an axiomatic system for logic, including principles governing classes, in which Peano’s “axioms,” interpreted using these definitions, could be derived as theorems. Frege’s proposed basis for arithmetic, although fully logicist (in not employing any specifically nonlogical vocabulary or nonlogical principles) and clearly

Russell’s Logicism

155

more “regressive” or “analyzed” (in Russell’s sense of these words) than Peano’s basis, was flawed in that its assumptions about classes (or “extensions,” as Frege called them) led to contradiction. In particular, Russell discovered that Frege’s Basic Law V led to the antinomy now known as “Russell’s paradox” (discussed in Section 3). Russell communicated this problem to Frege in June 1902 (Frege 1980, 130–33), and Frege himself admitted the flaw in his system (Frege 2013, vol. 2, afterword). Russell sees his work as an attempt to improve upon Frege’s advance by taking yet another “regressive” step toward fundamentality. Russell sought to recover what was worthwhile in Frege’s proposed basis but at the same time purge it of inconsistency by making use of a yet even smaller logical basis, discarding Frege’s problematic assumptions about classes. Here we have an example of how the regressive method, as Russell sees it, does not necessarily preserve every aspect of the original body of knowledge being analyzed. It can reasonably be hoped that sometimes the analytic process might remove problems with the original theory or show dubious assumptions to be unnecessary. Changes to the original body of knowledge could be acceptable if what was theoretically important was preserved. Considered as anything like a historical record of the progress in research in the foundations of mathematics, this story Russell tells must be taken with several large grains of salt. Frege did not see himself as making further progress on a basis established by Peano; Frege’s core theses on the nature of arithmetic had been formed, and at least partially defended, before Peano’s major works were published. For his own part, Russell had formed the logicist thesis and had worked out logical definitions of the key concepts of number theory independently from Frege, and only discovered Frege’s work later (POM, xviii). While Russell himself discovered Peano before Frege, and no doubt found Frege to be an improvement over Peano once he had found him, this does not describe very accurately how anyone else would have surveyed the developments. Moreover, it has been suggested that Russell’s later depiction of his work on higher mathematics as merely exploiting the “arithmetization” already achieved in the late nineteenth century does not do adequate justice to the strides toward generality achieved even by Russell himself (and his coauthor Whitehead) in their technical work.5 Nonetheless, this story, as misleading as it is, tells us something about how he understood the achievement of his work and what he counted as “success” when it came to philosophical research in mathematics. It also underscores what for him would have been the most important part of mathematics to be shown to have a logical foundation: the basic principles of number theory.

156

The Bloomsbury Companion to Bertrand Russell

Russell’s regressive method also has implications for how he understood the epistemological dimensions of his project.6 Unlike Frege (at least as usually read), Russell was not attempting to argue that simple workaday mathematical truths such as 2 + 3 = 5 gained their epistemological justification from the more general logical axioms that could be used to deduce them. In fact, Russell saw the issue the other way around: the obviousness and certainty of such elementary mathematical truths provide “inductive support” for a more minimalist theory or list of basic assumptions from which they can all be derived (Papers 5, 572–77).

3  The paradoxes and the development of Russell’s logic Russell describes the period just after his discovery of Peano’s work in 1900 as a kind of intellectual honeymoon. Peano’s logic allowed him as he saw it to give precise definitions for all of the fundamental notions of mathematics and he “discovered what appeared to be definitive answers” (Auto I, 148) to the philosophical doubts he previously had entertained. By the end of this year, he reports (MPD, 73; Auto I, 148), he had finished an initial draft of what was to become POM. What ended the honeymoon was the discovery in 1901 that the assumptions he had been making about the existence of classes, and other entities such as predicates and propositions, were impossible because they led to contradictions (and that the assumptions made by Peano and Frege were similarly flawed). The best known difficulty is now known as “Russell’s paradox of classes.” If we consider it intelligible to ask whether or not a given class x is a member of itself, x ∈ x, and consider every intelligible condition (every open sentence) to define a class, then we can consider the class of all such x which are not members of themselves—which Russell writes as “ xˆ(x ~∈ x )”—and ask whether it is or is not a member of itself. By the naïve assumptions he made at the time, it would be a member of itself just in case it met its defining condition, that is, just in case it was not a member of itself. This is a contradiction. A similar problem threatens if one assumes that there is a quality or monadic universal, or “predicate” as Russell would say at the time, corresponding to every condition defined by an open sentence. It would seem then that there is a quality q which a predicate a has just in case a does not have quality a. But then we get a contradiction if we consider whether or not q has quality q (POM, §101). This version has come to be known as “Russell’s paradox of predication.”

Russell’s Logicism

157

Russell discovered these contradictions by considering Cantor’s result that every class has more subclasses than members (POM, §100; MPD, 75). Cantor’s reasoning for this result proceeded by showing that a contradiction results from the assumption that any class has as many members as subclasses as follows: Suppose there is a mapping from subclasses to members that yields a distinct member for every subclass. One can then define a “diagonal” subclass s of all those members which are not in the subclass mapped to them. Since s is a subclass, it itself should be mapped to some member, a. But then we can ask whether or not a is in s: since s is defined as the class of all those members not in the subclass mapped to them, a would be a member of s just in case it isn’t. Russell arrived at his contradiction more or less by applying Cantor’s argument to such classes as the universal class or the class of all classes, for which it would appear impossible that they would have more subclasses than members, since their subclasses are members.7 Russell also discovered more complicated versions of similar kinds of problems, such as a paradox of relations (Papers 5, 588), and another involving propositions (POM, §500). At the time of POM, Russell considered a proposition to be a mind-independent complex object somewhat like a state of affairs. If propositions are genuine objects, by Cantor’s result there ought to be more subclasses of the class of all propositions than members, that is, there ought to be more classes of propositions than propositions. However, it seems possible, for each class of propositions, m, to generate a distinct proposition, such as the proposition that every member of the class is true, that is, (p)(p ∈ m ⊃ p). Cantor’s reasoning would invite us to consider the class w of all propositions of the form (p)(p ∈ m ⊃ p) not in the class m they are “about.” Now we consider the proposition (p)(p ∈ w ⊃ p) and ask whether or not it is in w, and a contradiction arises from either answer. Russell mentioned these problems in POM, and made certain suggestions in the direction of solving some of them, but even by his own admission, he needed to do more work to provide a complete solution before the formal project of PM could proceed. Russell found this task more difficult than he anticipated, and attempting to solve such paradoxes was his chief intellectual preoccupation from 1903 to 1907. In his Autobiography (154), he describes himself as having spent the period largely staring at a blank page, failing to make progress. In fact, Russell considered a large variety of different solutions during this period, but usually found that his attempts were only partly successful, solving some but not all versions of the paradoxes. For example, the rather elegant theory found in his 1906 withdrawnfrom-publication paper “On the Substitutional Theory of Classes and Relations”

158

The Bloomsbury Companion to Bertrand Russell

(Papers 5, 236–61) solved the paradoxes involving classes and predicates, but fell prey to complicated versions of the kind of paradox of propositions mentioned above.8 Most of the manuscripts in which Russell explored various options are now published in Papers 4 and Papers 5. Many of his attempts are fascinating, and still worthy of further study, but they cannot be discussed in detail here. Central to the solution Russell settled on for PM was the suggestion that many apparent “abstract” entities such as classes and propositions are not genuine entities at all, but only “logical constructions” or “logical fictions.” Apparent terms for them are not genuine terms, but “incomplete symbols,” meaningful not by naming entities but rather by contributing to the meaning of statements in which they appear in a more complicated way. PM *20.01 offered the following contextual definition of a class abstract “zˆ z ” for a class of individuals: f { zˆ(y z )} =df (∃f )[(x )(f !x ≡ y x )& f (ϕ! zˆ)] (A similar contextual definition *20.08 was given for class abstracts for classes of classes.) Roughly, to say something f about the class zˆ(y z ) is really to say that there is a (predicative—see below) property ϕ coextensive with ψ of which f holds. It follows from this definition that claims made about classes of individuals must be reinterpretable as claims about properties of individuals rather than claims directly about individuals. A formula of the form x ∈ α would only be meaningful if x and α were of different logical types, and therefore the entire question of whether or not something, x, is a member of itself, x ∈ x, is syntactically illegitimate and unintelligible. Similarly, a discourse about a “class of all classes not members of themselves” is not possible, and thus the paradox cannot be formulated. Russell dubbed this method of eliminating commitment to classes in favor of higher-order quantification, somewhat oxymoronically, as the “no-classes theory of classes.” Russell’s higher-order logic made use of a hierarchy of variables of different logical types. Unfortunately, the exposition of the syntax of the logic of PM does not live up to contemporary standards of rigor, and as a result, there is a fair amount of disagreement about the precise nature of its type theory.9 At the bottom of the hierarchy, there are variables (x, y, z, …) for individuals. Russell speaks of the values of higher-order variables as “propositional functions,” or sometimes, less formally, as “properties” or “relations.” Specific propositional functions could be represented by open sentences, or formulas with one or more free

Russell’s Logicism

159

variables, “x is green,” “x is left of y,” or “x = x.” Part of the controversy, however, is whether Russell regarded propositional functions to be extralinguistic entities in some sense denoted by such expressions, or merely talked of “propositional functions” as a loose way of speaking about such open sentences themselves and their functioning in the logic in the “material mode,” to borrow a phrase from Carnap (1937, §64). At any rate, if the variable ϕ is used to quantify over (predicative) propositional functions of individuals, a formula such as “(ϕ) ϕ!a” is taken to entail the truth of every one of an entire class of propositions formed by replacing “ϕ” with an open sentence and letting a be taken as the value of its variable: “a is green,” “a is left of b,” and so on. There is then a new class of open sentences containing such variables as occupying the next highest type, and then variables for quantifying over such propositional functions, and so on. Moreover, variables for propositional functions with two arguments, or two variables, would be considered as a different logical type from those with one, and so on. Combined with the “no-classes” theory, this meant that one could introduce eliminable variables for classes of individuals, and variables for classes of classes of individuals, and so on, forming a hierarchy of types of class variables and class abstracts, but it would be impossible to have a single logical type for all classes. The foregoing is compatible with what is now known as “simple type theory.” However, in addition to the sorts of divisions mentioned above, Russell also regarded it as unintelligible for a higher-order quantifier to quantify over properties involving the same kind of quantification as itself or anything of a higher type or order. Consider our previous example of “(ϕ)ϕ!a.” Loosely, this can be interpreted as saying that “a has every property,” where properties are represented by open sentences such as “x is green,” “x is left of b,” and so on. But what of the open sentence, “(ϕ)ϕ!x” or “x has every property”: is this included among the properties a must have if “(ϕ)ϕ!a” were to be true? Arguably, the suggestion that it is so included leads to a vicious circle in the truth-conditions of “(ϕ)ϕ!a,” and Russell took such vicious circles to be what generate such “semantic” paradoxes as the Liar paradox and Richard’s Paradox (PM, 60–65). Hence Russell distinguished the “order-type” of propositional functions not involving quantification over their own or higher types—so-called “predicative” propositional functions—from those which do quantify over predicative functions of the same type, and those which quantify over those, and so on. This is known as Russell’s “ramified” hierarchy. There is some disagreement over how and to what extent this hierarchy is fully enshrined in the logic of

160

The Bloomsbury Companion to Bertrand Russell

PM itself.10 However, Russell found it necessary to add an assumption, the “axiom of reducibility” (*12.1), according to which every monadic propositional function ϕ is equivalent to a predicative one f, marking predicative ones with an exclamation point ! (as well as similar principles for higher types and twoargument functions):

( ∃f )( x ) (f x ≡ f ! x ) Notice that the quantified ϕ in Russell’s contextual definition of classes, above, is restricted to predicative values. If the function ψ were not equivalent to a predicative φ, then according to this definition, all statements f {zˆ(y z )} about the class zˆ(y z ) would come out as false. Thus, Russell took assuming the axiom of reducibility to be a necessity for recapturing an adequate class theory and sometimes also called it “the axiom of classes” (Papers 5, 540, 607). If one interprets the “ϕ” in the axiom as a schematic letter standing in for any complex open sentence, the axiom takes a form very similar to the standard “principle of comprehension” in simple type theory, a point to which we shall return later.

4  Logicism in Principia Mathematica The no-classes theory of classes, and PM’s similar treatment of “relations in extension” (*21), greatly simplified its syntactic primitives and reduced its basic assumptions. In fact, the only signs taken as primitive in PM are two statement operators, “~” (negation), “∨” (disjunction), the universal and existential quantifiers, and variables of the various logical types. Other truth-functional statement operators are defined from negation and disjunction in the usual ways. For instance, material implication p ⊃ q is defined as ~p ∨ q. The axioms of the system, apart from the various forms of the axiom of reducibility mentioned above, include only those for standard classical propositional and (higherorder) quantifier logic.11 From this basis alone, Whitehead and Russell sought to define all of the notions of pure mathematics and derive the laws governing them. (There are two other principles sometimes misleadingly called “axioms” in the discussions of Russell’s logicism, namely, the “axiom of infinity” and the “multiplicative axiom”; more on these in what follows.) Russell’s treatment of cardinal numbers is very similar to Frege’s, with the substitution of the no-classes theory of classes for Frege’s naïve theory of

Russell’s Logicism

161

extensions being the main point of departure. Cardinal numbers are treated as classes of classes alike in size, that is, which can be put in 1–1 correspondence with each other. The number 0 is then the class of all empty classes, of which there is of course only one (the null class). The number 1 is the class of all singlemembered classes; the number 2 the class of all couples, and so on. One caveat, introduced by Russell’s type distinctions and its consequences for the no-classes theory, is that numbers are duplicated across various types. For example, the null class of individuals must be treated differently in the no-classes theory from the null class of classes of individuals, and consequently, the “zero” which contains only the former is in a different type from the “zero” containing only the latter. In the summary of the formal treatment below, variables such as x, y and z are used for individuals, Greek letters such as α, β, and so on, as special variables for classes (eliminable in favor of propositional function variables on the no-classes theory), variables R, S, and so on, for relations-extension (also eliminable); however definitions for higher-type analogues of these concepts would be quite similar.12 identity non-membership universal class complement null class union intersection subset singleton y domain converse domain

x = y =df (f )(f ! x ⊃ f ! y ) x ~∈a =df ~ (x ∈a ) V =df xˆ(x = x ) −a =df xˆ(x ~∈ a ) Λ =df −V a ∪ b =df xˆ( x ∈ a ∨ x ∈ b ) a ∩ b =df xˆ( x ∈ a & x ∈ b ) a ⊂ b =df ( x ) ( x ∈ a ⊃ x ∈ b )

i‘y =df xˆ( x = y )

D‘R =df xˆ( ( ∃y ) xRy ) ‘R =df xˆ( ( ∃y ) yRx ) class of many-one relations Nc → 1= df Rˆ( ( x )( y )( z ) ( xRy & xRz ⊃ y = z ) ) class of one-many relations 1 → Nc= Rˆ( ( x )( y )( z ) ( yRx & zRx ⊃ y = z ) ) D

df

class of cardinal numbers zero successor of α class of natural numbers

1 → 1= df Nc → 1 ∩ 1 → Nc a sm b = df ( ∃R ) ( R ∈1 → 1& a = D‘R & b = ‘R ) Nc‘a =df bˆ( b sm a ) NC = bˆ( ( ∃x ) b = Nc‘a ) D

class of one-one relations cardinal similarity cardinal number of α

df

0 =df Nc‘Λ Succ‘a =df bˆ( ( ∃a ) ( x ∈ b & b ∩ −i‘x ∈ a ) )

( (

ℕ=df aˆ ( b ) ( 0 ∈ b & (g ) (g ∈ b ⊃ Succ‘g ∈ b ) ) ⊃ a ∈ b

))

162

The Bloomsbury Companion to Bertrand Russell

Notice here that the successor of a number α is the class of classes which are such that if a member is removed one gets a member of α: the class of all classes with one more member than the members of α. The class of natural numbers are those contained in all classes containing 0 and the successor of anything they contain, and hence would be the class containing exactly 0, 1, 2, 3, . . . . Using these definitions, formal versions of the Peano-Dedekind axioms could be stated as follows: 1. 0 ∈ ℕ 2. (α)(α ∈ ℕ ⊃ Succ‘α ∈ ℕ) 3. ~(∃α)(0 = Succ‘α) 4. (α)(β)((α ∈ ℕ & β ∈ ℕ & Succ‘α = Succ‘β) ⊃ α = β) 5. (α)((0 ∈ α & (γ)(γ ∈ α ⊃ Succ‘γ ∈ α)) ⊃ ℕ ⊂ α) Using nothing beyond the usual assumptions of higher-order quantifier logic (which also provide a proxy for class-logic via the no-classes theory), four of these five can be derived as theorems relatively straightforwardly. (An interested reader may consult PM, parts I–II, for details.) The problematic one of the bunch is the fourth, which states that no two natural numbers have the same successor. In order for this to hold, no natural number can be empty: that is, for every natural number there must be at least one class with that many things. Notice further that in order for this to be true for every natural number, the number of things in total must be infinite. Suppose instead that there were only definitely many individuals; for convenience, let us assume there are only three: a, b, and c. Then the class consisting of a, b, and c would be the only three-membered class, and the only member of the number 3. The successor of 3 (viz., 4) would be the class of all those classes which are such that, if you removed a member from them, you’d get a member of 3. But since a, b, and c are the only individuals there are, no class could meet this condition. Hence, 4 would be empty. The successor of 4 (viz., 5) would also be empty, since it is clearly impossible to get a member of the null class by removing an element from a class, as the null class has no members. Hence, 3 and 4 would both be natural numbers, and both would have the same successor (the null class), but they would not be identical to each other, and so the fourth Peano “axiom” would be false. During earlier periods of his career, when he took propositions and classes to be more than just “logical fictions,” Russell had regarded it as possible to deduce the existence of infinitely many individuals purely logically. By the time of PM, however, Russell rejected such deductions. Such proofs only appear to succeed

Russell’s Logicism

163

because of a failure to take type distinctions into account. Consider what is now called the Zermelo sequence: Λ, ɩ‘ Λ, ɩ‘ Λ, ɩ‘ɩ‘ Λ, . . . (the null class, its singleton, its singleton’s singleton, and so on). It might appear that this sequence is unending, and thus infinite, and that there must therefore be infinitely many things. However, notice that each member of this sequence is of a higher type than the previous member, and so it cannot be used to establish an infinity of individuals, or an in infinity of things in any one type at all. A similar difficulty makes it impossible for Russell to make use of Frege’s argument (Frege 1950, §82) alleging that all the numbers preceding a given natural number show that this number must apply to at least one collection (i.e., the class containing only 0 has one member, the class of 0 and 1 has two members, the class of 0, 1 and 2 has three, and so on). Russell also rejects an argument (adapted from Dedekind) that alleges that the series: Socrates, the idea of Socrates, the idea of the idea of Socrates, the idea of the idea of the idea of Socrates, and so on is infinite, on the grounds that the relevant sort of ideas can only be assumed to exist when actually thought by some mind, and hence we have no reason to think that this list goes on beyond some finite number of steps (IMP, 139). Russell therefore concludes that one cannot derive something like the fourth Peano axiom stated above without assuming that every natural number is nonempty, or what he sometimes calls axiom of infinity: (α)(α ∈ ℕ ⊃ (∃β) (β ∈ α))(Inf) Despite his use of the word “axiom” here, in PM itself this is not taken as an axiom in the usual sense, and is simply left as an undischarged antecedent on many results, including the fourth Peano axiom and all results depending on it. The question as to whether or not this compromises the success of Russell’s logicism is revisited in Section 6. For reasons mentioned in Section 2 above, natural number theory is perhaps the most important test case for Russell’s logicism, but of course, PM covers far more than simply the theory of natural numbers. In addition to the treatment of cardinal numbers just sketched, PM (parts IV and V; mostly in volume 2) also provides a treatment of ordinal numbers, which are taken as a subclass of a broader notion of “relation numbers,” defined as classes of relations which are structurally isomorphic to each other. Finite ordinal numbers provide yet another interpretation of “natural numbers” for the purposes of the PeanoDedekind axioms, though a similar problem with infinity arises there. For both cardinals and ordinals, they discuss infinite numbers (if there are any collections

164

The Bloomsbury Companion to Bertrand Russell

or relations exemplifying such). To capture multiplication for infinite numbers and achieve the expected results (that is, those suggested by Cantor’s work), Whitehead and Russell discuss an assumption they call the multiplicative axiom, which asserts that for every class of disjoint (non-overlapping) classes, there is a class which has among its members exactly one member from each member of the class of disjoint classes (PM, *88). This assumption is equivalent with Zermelo’s axiom of choice, which is notoriously independent of the other basic assumptions of set theory. Like the so-called “axiom” of infinity, the multiplicative axiom is not formally taken as an axiom or basic assumption in PM, but is simply abbreviated as “Mult ax” and left as an antecedent on the various results which require it, such as the principle that all classes can be wellordered. In Parts V and VI (mostly in volume III) of PM, Whitehead and Russell provide treatments of integers, ratios, complex numbers, and, finally, real numbers, in terms of what is in common between certain systems of relations, of which relationships between segments of ratios provide only a single instance.13 Along the way, they give formal logical definitions of such notions as limit and continuity. Part VI goes further yet and generalizes on these notions to provide a generic theory of quantity and measurement. The planned fourth volume of PM was slated to provide logicist accounts of many aspects of pure geometry as well, though one can glean certain aspects of how this might have proceeded from Russell’s chapters on projective, metric, and descriptive geometry which make up the later parts of his earlier POM.

5  Gödel’s results and the scope of Logicism We now turn to an evaluation of Russell’s logicism. Perhaps the most common problem one hears cited against it, or logicism in general, especially in casual conversation, appeals to Gödel’s famous incompleteness results. However, I think the relevance of these results for assessing logicism, at least on the most reasonable understandings of the logicist thesis, is greatly exaggerated. Using PM as an example,14 Gödel showed that any deductive system for mathematics having certain features contains undecidable sentences, that is, sentences such that neither they nor their negations can be derived as theorems. Especially when combined with well-known corollaries shown by Rosser, Tarski, and others,15 the features can for all intents and purposes

Russell’s Logicism

165

be reduced to consistency, having the requisite strength to capture the basic elements of number theory, and having a recursive axiomatization. Since recursive functions have been shown equivalent with Turing-computable functions and λ-definable functions, if we accept the Church-Turing thesis that these and only these functions are effectively computable, having a recursive axiomatization amounts more or less to there existing an effective decision procedure for determining what does or does not count as a deduction in this system. The feasibility or desirability of a non-recursively axiomatized system is basically nil: what good could a deductive system be if there was no effective way to tell what counts as a deduction in it and what doesn’t? Gödel’s “first incompleteness theorem” roughly establishes that systems with the desirable features must contain an undecidable sentence abbreviated G, which, roughly asserts of itself that it is not a theorem of the system (or really, asserts something analogous of a number which corresponds to it in an arithmetization of the system’s syntax). Clearly, it would be undesirable for G to be a theorem, as then one would have a false theorem in the system. But if it is not a theorem, it is true, and hence we have what appears to be a number-theoretic truth not derivable from the axioms. Gödel’s “second incompleteness theorem” establishes that in such systems one can form a sentence which roughly asserts that the system is consistent, that is, that no contradiction is derivable within, but that this sentence is also undecidable, at least if the system is consistent. How relevant such results are for evaluating Russell’s (or anyone else’s) form of logicism depends largely on how strong we take the logicist thesis to be. Suppose we define a “purely logical deductive system” as one whose axioms are all logical truths and whose inference rules are sound on logical grounds alone, and so cannot lead from a logical truth or truths to something that isn’t also a logical truth. A very strong interpretation of the logicist thesis (LT) would be the following: There is a practical purely logical deductive system, S, such that for every mathematical truth, p, p is a theorem of S.

(LT-a)

It is likely that Russell believed that (LT-a) was true early on, and that PM was such a deductive system. However, even if we interpret the “practical” in (LT-a) to restrict us to recursively axiomatized systems, additional assumptions would be required to argue that the Gödel results undermine (LT-a). Russell was aware, even in 1910, that there were undecidable sentences in PM. The (so-called) axiom

166

The Bloomsbury Companion to Bertrand Russell

of infinity, for example is neither derivable nor refutable therein. Russell would not have taken this to undermine his main thesis, or even (LT-a), since he would regard neither the axiom of infinity nor its negation as a “mathematical truth”; if it is true, it is contingently true (IMP, 141). Unless it could be shown that some of the undecidable sentences proven to exist by Gödel, or their negations, are mathematically true, there is no problem. Russell even suggested a response along these lines in 1963 when asked about the importance of Gödel’s results in a letter.16 Unfortunately, as such, the response seems a bit hasty. Gödel’s G and the consistency of PM itself, unlike the axiom of infinity (arguably), we do seem to know to be true, and on a priori grounds. If they are not mathematical truths, it is very hard to see what other kind of a priori truths they could be. And if they are mathematical truths, then this would seem to undermine PM’s claim to justify (LT-a), as they cannot be proven in PM. However, it is natural to wonder whether or not the heart of Russell’s logicism requires anything quite as strong as (LT-a). Russell saw himself as arguing against the Kantian thesis that mathematical truths were importantly different from logical truths, with the former understood as synthetic a priori and the latter as analytic.17 From this perspective, it seems that Russell’s principal thesis was that mathematical truths don’t have a special nature or essence distinct from that possessed by logical truths. Whatever it is that makes logical truths special or sets them apart from other kinds of truths is also possessed by mathematical truths: mathematical truths are a species of logical truths. It seems possible to hold that thesis without insisting on anything as strong as (LT-a). Unfortunately, the issue cannot be formally spelled out without saying exactly what it is that makes logical truth special, an issue Russell himself struggled with.18 Early on, he seems to have thought that any truth which could be stated using only logical constants counts as a “logical truth” (PM, §10); later on, however, he suggests that they must have a special “tautological” form, but neglects to provide an exact specification of what that amounts to (IMP, 203). In contemporary research in metalogic, it is often the case that two distinct criteria are given for what makes formulas logically necessary: a deductive or proof-theoretic one (derivability from the axioms), and a semantic one (usually amounting to something like truth in all acceptable models or interpretations). If a deductive system is both sound and complete, the two characterizations are equivalent: something is a logical truth in one sense if and only if it is in the other. However, some deductive systems are incomplete. For example, it is a wellknown corollary of Gödel’s results that second-order and higher-order logical

Russell’s Logicism

167

calculi (including PM), when interpreted according to “standard semantics,”19 cannot be complete when axiomatized recursively. There are formulas that are logically true according to the semantic criterion that cannot be derived as theorems. Obviously, this result should not be interpreted as showing that some logical truths are not logical truths, and similarly, it is unclear why the fact that certain apparently “mathematical” truths cannot be derived in a certain deductive calculus should be taken to establish that those truths are not logical truths. In other words, if the logicist thesis were interpreted along the lines of (LT-b) below,20 then it is not clear how or why Gödel’s results should be taken to pose a problem for it: Every mathematical truth has the semantic feature which sets apart logical truths semantically from others.

(LT-b)

Of course, PM does not explicitly provide a semantic characterization of logical truth nor a direct argument for anything like (LT-b). Nonetheless, the deductions of various mathematical truths provided in PM could be relevant for semantic characterizations of logicism, provided it could be established that the axioms of PM are logical truths in this sense and that the inference rules preserve this feature. Moreover, even for those primarily interested in a deduction-centered account of logical truth, it is not obvious that something as strong as (LT-a) needs to be maintained by a logicist. Consider the following similar, but slightly weaker, thesis: For every mathematical truth, p, there is a practical purely logical deductive system, S, such that p is a theorem of S.

(LT-c)

This is much like (LT-a) except that it does not require there to be a single logical system in which every mathematical truth is captured: it requires only that each such truth be captured in some such system or other. To my knowledge, there is nothing in Gödel’s results to suggest that (LT-c) must be false. Consider again the situation that we seem to know that Gödel’s G is true, despite its undecidability in PM. So long as our knowledge of this relies only on logical principles, perhaps there is some purely logical extension of PM in which G is derivable. (And if our knowledge of G requires something beyond logic, for this to pose a problem for Russell, it would need to be shown that it doesn’t require anything beyond mathematics, or his earlier response is sufficient.) If this new system is recursive, it too will have undecidable (true) sentences, but perhaps they are derivable in another logical system of the appropriate type. This is more

168

The Bloomsbury Companion to Bertrand Russell

or less in keeping with Russell’s “official” response to Gödel, which suggests that there is a hierarchy of logical languages or logical systems, and what cannot be captured in one is captured higher in the hierarchy (Papers 11, 159; MPD, 114). Indeed, when the issue is posed in terms of the special nature of mathematics and its relation to logic as a whole—the issue of what makes mathematical truths different from others, and whether or not this is the same as the special nature of logical truths—what PM does succeed in establishing seems far more important than any worries stemming from the few rather recondite undecidable sentences shown to exist by Gödel. Consider every mathematical truth taught in primary or even secondary school. Consider all the mathematics of quantity and measurement used in engineering. Consider Kant’s famous example of the allegedly synthetic 5 + 7 = 12. It appears that all of these can be captured in any higher-order system interpreting the number theory provided by Peano arithmetic. If PM or similar work succeeds in showing that this portion of mathematics consists in nothing other than logical truths, it becomes almost comical to suggest that the logicist thesis as a whole is mistaken. If nearly all of the mathematics anyone uses or knows about, the truths about numbers which most readily come to mind when considering arithmetic, reduce to logic, can it seriously be maintained that the essence of mathematics is extralogical, merely because there are certain highly complex, and application-free truths in certain formal systems which cannot be shown to be logical in precisely the same way within a (single) recursive axiom system? If mathematics has a special essence, a special nature, surely that nature expresses itself already in Peano arithmetic, so here again it seems right to maintain that capturing Peano arithmetic is the best test case for Russell’s logicism as a whole. Of course, doubts may still be raised as to whether or not PM adequately captures Peano arithmetic, but they are distinct from any stemming from Gödel’s results. Such doubts are considered in the next section.

6  The controversial “axioms” and if-then-ism Perhaps the next most commonly heard objection to Russell’s logicism—and the most common from specialists in the area—points to its employment of certain basic assumptions which do not seem, or at least do not obviously seem, like logical truths.21 There are three controversial (so-called) axioms here: reducibility, infinity, and the multiplicative axiom. Of course, it must be

Russell’s Logicism

169

remembered that the latter two were not treated as axioms in PM, but instead simply left as undischarged antecedents on results depending on them. This point, however, simply invites stating the criticism in a different form. The charge, called if-then-ism by the critics that offer it,22 is that Whitehead and Russell do not prove the basic principles of mathematics from logic outright, but merely derive conditionals with certain mathematical assumptions as antecedents and their logical consequences as consequents. For example, rather than proving that no two natural numbers have the same successor (the fourth Peano axiom), they only prove that this is true if there are in infinitely many individuals. In the eyes of these critics, this greatly compromises the success of PM’s logicism. After all, for any theory which can be organized as a deductive system, even an obviously empirical and nonlogical one such as a theory in physics, the conditionals from the axioms to their logical consequences can take the form of logical truths. This tells us nothing about the essence of its subject matter. For what it is worth, Russell himself admitted that the truth of the axiom of reducibility was less than fully obvious. Indeed, even when first stating it he admitted that the grounds for accepting it were “inductive,” in keeping with his general “regressive” method. It could be used as a premise from which many desirable results could be proven and he knew of no more plausible or no more general principle from which as many important results followed (PM, 59). Russell also dedicated much of the new material added to the second edition of PM to exploring a less than fully successful23 attempt to do away with the axiom by adopting instead a doctrine of extensionality he took to be motivated by the philosophy of Wittgenstein’s Tractatus (see PM 2, xiv and appendices). However, it should be remembered that Russell’s main motivation for adopting the axiom was to recover the ability to speak of classes, which for him amounted simply to things considered in groups, or plurally,24 without a stronger assumption of “real” classes taken as genuine things. It is now widely acknowledged that something like quantification over collections or pluralities, or in general, what one finds in a “standard” higher-order logic, is necessary for capturing some logical forms used in everyday reasoning. Perhaps the most famous example is the Kaplan-Geach sentence, “some critics only admire one another,” which cannot be interpreted in first-order logic. The history of logic does have its first-order purists, like Quine, who considered higher-order logic to be “settheory in sheep’s clothing” (Quine 1986, 66), and there are even holdouts among contemporary philosophers (e.g., Burgess 2005, 201–14). Still, I think it safe

170

The Bloomsbury Companion to Bertrand Russell

to say that these represent the minority. Something like the ability to consider objects in groups seems to be a necessary central element in reasoning about nearly every topic, and allowing for this is not to take on a special extralogical or specifically mathematical assumption. As noted above, the axiom of reducibility is roughly equivalent to the usual comprehension principles allowed in simple type theory or most plural logics (cf. Yi 2013).25 The ideal formulation of a logical apparatus that allows us to speak of classes, or collections, or pluralities, remains an active area of research, discussion, and controversy, but when compared to the alternatives, Russell’s adoption of something like the axiom of reducibility seems all things considered to be among the most restrained. The failure to prove the multiplicative axiom or obtain its results outright in PM also does not seem to pose a serious problem for the success of Russell’s logicism. While the very similar axiom of choice is perhaps widely accepted among mathematicians, and it is also known to be equivalent to a number of other principles with a similar status (the well-ordering principle, Zorn’s lemma, the multiplicative axiom itself, etc.), no one has been able to provide a proof of it from more obvious or more fundamental principles, and indeed, it is wellknown to be independent of the other axioms of ZF set theory. The multiplicative axiom is not needed for results in finite arithmetic, but only for certain results in less firmly settled infinite arithmetic. The axiom of choice and the multiplicative axiom hardly count as uncontroversial examples of definite mathematical truths. The only results which are considered definite are their mutual equivalence and their relationships with these other principles, which are captured in PM. It does not seem reasonable to argue that the logicism of PM is a failure merely because the multiplicative axiom, or its logical consequences, are not derived in nonconditional form. Similar concerns surrounding the axiom of infinity seem to pose a far greater challenge to a positive evaluation of Russell’s logicism. Without it, as we have seen, Russell cannot recapture Peano arithmetic in its usual form, as the axiom of infinity would be needed to obtain the fourth axiom. A defender of Russellian logicism seems to be forced into one of the following positions: (a) argue that the axiom of infinity can reasonably be taken as a basic logical truth, (b) provide a demonstration of the axiom of infinity from other plausible purely logical assumptions (not already included in PM), (c) deny that the fourth Peano axiom is, without qualification or antecedent, a mathematical truth, or (d) admit that certain very important mathematical truths cannot be derived from a logical starting place. Response (d) seems tantamount to abandoning logicism. Responses (a) and (b) are perhaps worthy of further investigation,

Russell’s Logicism

171

but (c) seems to have been Russell’s own considered view. Can anything be said in its favor? Proponents of the if-then-ist criticism of Russell sometimes point to early passages such as in POM (§1), where Russell claims that all mathematical propositions can be seen as taking the form of conditionals: Pure mathematics is the class of all propositions of the form “p implies q,” where p and q are propositions containing one or more variables, the same in the two propositions, and neither p nor q contains any constants except logical constants.

However, there are a number of reasons for being cautious about reading this in favor of if-then-ism. For one, in the logic of POM, all other connectives (negation, disjunction, conjunction, etc.) are defined in terms of material or formal implication (§§18–19), and so any logically complex proposition would be of the requisite form, and obviously a logical deduction of a statement relies on its having the kind of complexity that might make its derivation from basic principles possible. Russell’s example in this context (§6) in which he rewords “1 + 1 = 2” as “If x is one and y is one, and x differs from y, then x and y are two” helps explain how uncovering the true logical complexity of a proposition is necessary in order to see its real logical status. As stressed by Griffin (1982, 89), Russell (even later on) thought many mathematical truths took the form of conditionals not because he thought of them as representing the relationships between the axioms of a theory and the consequences, but rather because he took them as general truths with definite application conditions where those conditions were stated in the antecedents applied to variables. Applying mathematics involves finding values of the variables satisfying the antecedents, and therefore accepting the relevant instances of the consequents for those values. There is precedent in Russell’s earlier work for recognizing situations where required antecedents have been left out of stated theorems by practicing mathematicians because they took for granted the most likely practical application conditions for their work. Past geometers who assumed that actual (physical) space was Euclidian took for granted that they could presume features of Euclidian space when stating their results. As Russell argues, however, the real “pure” mathematical truth should be taken as the conditional from their assumptions about the structure of space to the consequences thereof (POM, §§353, 412), rather than the unqualified statements which only are true in certain application conditions. Hence, it appears Russell would argue that the true form of a given mathematical truth may reasonably be taken to be conditional in form if the antecedent involves

172

The Bloomsbury Companion to Bertrand Russell

special and restricted conditions for its application, and the consequent itself, even if true for a given instance, is not true for that instance on mathematical grounds alone, but only because the antecedent is satisfied. In the geometrical context, if the antecedent is not satisfied, we have a different kind of space, where a different kind of geometry may be applicable; this does not make it in any way a logically or mathematically defective space. Is the case of the relationship between the axiom of infinity and the fourth Peano axiom at all similar? The suggestion on the table then is that the fourth Peano axiom (and its consequences) should not be considered mathematical truths, full stop, just like geometrical statements are only true if space is Euclidian. Hence, it is not required for logicism that we derive them, full stop, from a logical basis. It is enough if we can demonstrate that they hold given the right conditions. Russell himself does not explicitly argue this way to my knowledge, and at first blush it seems a fairly radical suggestion. Should we really conclude that 4 is distinct from 5 only given certain application conditions? Yet, I think the proposal is not wholly without merit.26 Without presupposing infinity, our usual assumptions about the behavior of numbers will only start going “awry” for those greater than the number of individuals there in fact are. If we are applying the mathematics of PM to a domain of individuals with 3 members, then all the usual results for numbers up to including 3 would be what we expect. At higher types, where there are more and more classes formed of our three things or formed of classes of them, or classes of classes of them, and so on, the “well-behaving” numbers will reach higher and higher (IMP, 134). In short, nothing will go wrong in our actual conditions of application. Perhaps “non-infinite” arithmetic, where the series of numbers runs out after a certain stage, or goes in circles, much like non-Euclidian geometry, should be seen as a new area of study. Indeed, one could make the case that modular (“clock”) arithmetic already exemplifies this. Some philosophers take a hard conservative line and demur from any philosophical proposals which could be seen as suggesting “reforms” to mathematics (e.g., Lewis 1986, 109). Russell himself, as we have seen, thought it an advantage of the process of analysis that it did not necessarily preserve the analyzed body of knowledge as originally conceived. Whatever may be said for the above proposal, Russell himself seems to have been disappointed by his failure to establish in any definitive or a priori way the existence of the complete series of numbers. In fact, while the logic of PM presupposes there is at least one individual—something Russell eventually came to regard as “a defect of logical purity” (IMP, 203n)—there is no way in it to

Russell’s Logicism

173

show that there is any more than one, and therefore, no logical way to arrive at any numbers beyond one, at least at the lowest type. Ivor Grattan-Guinness provides us with a report, perhaps apocryphal, from an acquaintance of Russell who allegedly heard him bemoan this fact after having finished PM: Bertrand Russell called me aside as a mathematician I suppose and likely to appreciate the gravity of his statement—“I have just realised that I have failed— it is easy to establish the unit one but I have omitted to establish a second like unit”—(I won’t guarantee the precise wording but it’s not far off). He went on to say “I have finished.” (Quoted in Grattan-Guinness 2000, 401)

Did, however, Russell fail? When we consider the effects of Russell’s pursuit of logicism on the history of both logic and philosophy, there can be no question that it has had a lasting and positive impact. It shaped not only Russell’s philosophy but has made modern symbolic logic an irreplaceable part of almost every analytic philosopher’s toolkit. Even when it comes to the question of the logicist thesis for mathematics in particular, there is a lot more reason to think that Russell’s work was a success than is generally realized.

Notes 1 Early drafts of his work on this subject can be found in Papers 2 and Papers 3. 2 See Hylton (1990) and Griffin (1991) for discussion of Russell’s philosophical conversion to realism. 3 See Proops (2006) for an account of the philosophical motivations behind Russell’s initial acceptance of logicism. 4 It is worth comparing Russell’s description of philosophical progress in the study of mathematics, with his overall description of “analysis” as a method employed in other areas of philosophy; see, for example, Hager (2003) for discussion. 5 On this point, see especially Gandon (2012, chaps. 5–6) and note 13 of this chapter. 6 For further discussion on this point, see Irvine (1989). 7 For further historical details of the context of Russell’s discovery, see Griffin (2004). 8 For discussion, see Landini (1998, chap. 8). 9 For a variety of positions, see Church (1976), Hatcher (1981, chap. 4), Landini (1998, chap. 10), Linsky (1999, chap. 4), and Klement (2013). 10 In particular, Landini (1998, chap. 10, 2011b, chap. 3) has argued that PM should be interpreted to make use of no object-language variables except predicative variables, making the syntactic rules much simpler than what is suggested by the traditional interpretation found in for example, Church (1976), where variables

174

The Bloomsbury Companion to Bertrand Russell

of infinitely many orders within the same type are possible. Landini reads the non-predicative variable occurring in the axiom of reducibility as a metalinguistic schematic letter. 11 Strictly speaking, PM offers us two different treatments of the logic of quantification, one in *9 and one in *10, with the more austere *9 disallowing quantifiers subordinate to truth-functional operators except as abbreviation. For the sake of simplicity, I pass over such complications here. 12 I here try to make use of Whitehead and Russell’s own notation as much as possible, although note that they used a dot, or concatenation, rather than “&” for conjunction, and also used dots in place of parentheses for grouping; see PM, 9–10. I also here omit certain intermediately defined signs employed in PM itself, and simply write the unabbreviated forms where these are used later. 13 In IMP (chap. 7), Russell defines real numbers simply as segments of ratios or rational numbers according to the “Dedekind cut” method, and does little to explain the more general application of PM ’s theory of quantity; this oversimplification of the more general treatment provided in PM has been bemoaned by Gandon (2012, 9) among others. 14 It is likely that Gödel’s exposition of PM was not entirely faithful to the original; in particular, Gödel did not fully take into account that numerals, as signs for classes, in PM were incomplete symbols and not genuine terms. This issue, and the complications arising from it, cannot be explored in depth without providing a full reconstruction of PM, which cannot be attempted here. 15 Gödel’s original results were published in Gödel (1931), and a summary of them and the most important corollaries can be found in most textbooks on mathematical logic, for example, Mendelson (2010). 16 See Russell’s letter to G. Simons, 1 October 1963; document RA2.710.111928 at the Bertrand Russell Archives, McMaster University, and compare Papers 11: 65. 17 Interestingly, Russell’s exact attitude of where Kant had erred seems to have changed over the years; in POM (§434), Russell suggests that Kant was wrong not about mathematics but about logic, alleging that even logical truths can be understood as synthetic a priori in Kant’s sense. Later on, Russell claims instead that logicism shows that mathematics is analytic and not synthetic (e.g., HWP, 740). It is likely, however, that the different attitudes are as much the result of employing different definitions of “analytic” and “synthetic” as they are a reflection of a change of mindset; see Landini (2011b, 223–25) and Korhonen (2013, chap. 1). 18 For further discussion, see Klement (2015). 19 Standard semantics is often contrasted with “general” or “Henkin” semantics; the former, unlike the latter, requires that every subset of the domain of the first-order quantifiers be the extension of some value of the domain of quantification for second-order variables, and similarly for other types.

Russell’s Logicism

175

20 Rayo (2007, 240) provides a list of different interpretations of logicism, though curiously he does not bother to make distinctions similar to my (LT-a) and (LT-c) for forms focused on deductive consequence or derivability. 21 See, for example, Kneale and Kneale (1962, 669), Roselló (2012, 50), Tennant (2013, §1.3.1) and Soames (2014, 488). 22 See, for example, Putnam (1967), Musgrave (1977), Coffa (1981) and Boolos (1998, chap. 16). For a variety of responses, see Griffin (1982), Landini (2011b), Gandon (2012), Klement (2012), Galaugher (2013), and Kraal (2014). 23 The difficulties with the attempt were partly appreciated by Russell himself, partly not. For further discussion, see Landini (1996), Hazen and Davoren (2000), Landini (2007), and Linsky (2011). 24 For more on Russell’s attitude about the relationship between discourse about classes and about plurals, see Oliver and Smiley (2005), Bostock (2008) and Klement (2014). 25 Indeed, if Landini is right that the object-language variables of PM are only predicative variables (see note 10 above), then the system of PM is formally speaking the same as a simple type theory, and the axiom of reducibility just is its comprehension principle; see Landini (1998, 264–67). 26 See Landini (2011a) for a sympathetic treatment of such an approach.

References Boolos, George. 1998. Logic, Logic and Logic. Cambridge, MA: Harvard University Press. Bostock, David. 2008. “Russell on ‘the’ in the Plural.” In Russell vs. Meinong: The Legacy of “On Denoting,” edited by Nicholas Griffin and Dale Jacquette, 113–43. New York: Routledge. Burgess, John. 2005. Fixing Frege. Princeton: Princeton University Press. Carnap, Rudolf. 1937. The Logical Syntax of Language. Translated by A. Smeaton. London: Routledge & Kegal Paul. Church, Alonzo. 1976. “Comparison of Russell’s Resolution of the Semantical Antinomies with that of Tarski.” Journal of Symbolic Logic 41: 747–60. Coffa, J. Alberto. 1981. “Russell and Kant.” Synthese 46: 247–63. Dedekind, Richard. 1888. Was sind und was sollen die Zahlen? Braunschweig: Vieweg. Frege, Gottlob. 1950. The Foundations of Arithmetic. Translated by J. L. Austin. Evanston: Northwestern University Press. Frege, Gottlob. 1980. Philosophical and Mathematical Correspondence. Translated by H. Kaal. Edited by B. McGuinness. Chicago: University of Chicago Press. Frege, Gottlob. 2013. Basic Laws of Arithmetic. Translated by P. Ebert and M. Rossberg. Oxford: Oxford University Press.

176

The Bloomsbury Companion to Bertrand Russell

Galaugher, Jolen. 2013. Russell’s Philosophy of Logical Analysis 1897–1905. Basingstoke: Palgrave Macmillan. Gandon, Sébastien. 2012. Russell’s Unknown Logicism. Basingstoke: Palgrave Macmillan. Gödel, Kurt, 1931. “Über formal unentscheidbare Sätze der Principia Mathematica und verwandter Systeme I.” Monatshefte für Mathematik und Physick 38: 173–98. Grattan-Guinness, Ivor. 2000. The Search for Mathematical Roots 1870–1940. Princeton: Princeton University Press. Griffin, Nicholas. 1982. “New Work on Russell’s Early Philosophy.” Russell 2: 69–83. Griffin, Nicholas. 1991. Russell’s Idealist Apprenticeship. Oxford: Clarendon Press. Griffin, Nicholas. 2004. “The Prehistory of Russell’s Paradox.” In One Hundred Years of Russell’s Paradox, edited by Godehard Link, 349–71. Berlin: Walter de Gruyter. Griffin, Nicholas, and Bernard Linsky (eds.) 2013. The Palgrave Centenary Companion to Principia Mathematica. Basingstoke: Palgrave Macmillan. Hager, Paul. 2003. “Russell’s Method of Analysis.” In The Cambridge Companion to Bertrand Russell, edited by Nicholas Griffin, 310–31. Cambridge: Cambridge University Press. Hatcher, William. 1981. The Logical Foundations of Mathematics. Oxford: Pergamon Press. Hazen, Allen P., and J. M. Davoren, 2000. “Russell’s 1925 Logic.” Australasian Journal of Philosophy 78: 534–56. Hylton, Peter. 1990. Russell, Idealism and the Emergence of Analytic Philosophy. Oxford: Oxford University Press. Irvine, A. D. 1989. “Epistemic Logicism and Russell’s Regressive Method.” Philosophical Studies 55: 303–27. Klement, Kevin C. 2012. “Neo-logicism and Russell’s Logicism.” Russell n.s. 32: 127–59. Klement, Kevin C. 2013. “PM’s Circumflex, Syntax and Philosophy of Types.” In The Palgrave Centenary Companion to Principia Mathematica, edited by Nicholas Griffin and Bernard Linsky, 218–47. Basingstoke: Palgrave Macmillan. Klement, Kevin C. 2014. “Early Russell on Types and Plurals.” Journal for the History of Analytical Philosophy 2: 1–21. Klement, Kevin C. 2015. “The Constituents of the Propositions of Logic.” In Acquaintance, Knowledge and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 189–229. Stanford: CSLI Publications. Kneale, W., and M. Kneale. 1962. The Development of Logic. Oxford: Clarendon Press. Korhonen, Anssi. 2013. Logic as Universal Science: Russell’s Early Logicism and Its Philosophical Context. Basingstoke: Palgrave Macmillan. Kraal, Anders. 2014. “The Aim of Russell’s Logicism.” Synthese 191: 1493–1510. Landini, Gregory. 1996. “The Definability of the Set of Natural Numbers in the 1925 Principia Mathematica.” The Journal of Philosophical Logic 25: 597–615. Landini, Gregory. 1998. Russell’s Hidden Substitutional Theory. Oxford: Oxford University Press.

Russell’s Logicism

177

Landini, Gregory. 2007. Wittgenstein’s Apprenticeship with Russell. Cambridge: Cambridge University Press. Landini, Gregory. 2011a. “Logicism and the Problem of Infinity: The Number of Numbers.” Philosophia Mathematica 19: 167–212. Landini, Gregory. 2011b. Russell. New York: Routledge. Leibniz, G. W. 1666. Dissertatio de Arte Combinatoria. Leipzig: J. S. Fick and J. P. Seubold. Lewis, David. 1986. On the Plurality of Worlds. Oxford: Blakcwell. Linsky, Bernard. 1999. Russell’s Metaphysical Logic. Stanford: CSLI Publications. Linsky, Bernard. 2011. The Evolution of Principia Mathematica: Bertrand Russell’s Manuscripts and Notes for the Second Edition. Cambridge: Cambridge University Press. Mendelson, Elliot. 2010. Introduction to Mathematical Logic, 5th ed. Boca Raton: CRC Press. Musgrave, Alan. 1977. “Logicism Revisited.” British Journal for the Philosophy of Science 28: 99–127. Oliver, Alex, and Timothy Smiley. 2005. “Plural Descriptions and Many-valued Functions.” Mind 114: 1039–68. Proops, Ian. 2006. “Russell’s Reasons for Logicism.” Journal of the History of Philosophy 44: 267–92. Putnam, Hilary. 1967. “The Thesis that Mathematics Is Logic.” In Bertrand Russell: Philosopher of the Century, edited by Ralph Schoenman, 273–303. London: Allen & Unwin. Quine, W. V. 1986. Philosophy of Logic, 2nd ed. Cambridge, MA: Harvard University Press. Rayo, Agustín. 2007. “Logicism Reconsidered.” In The Oxford Handbook of Philosophy of Mathematics and Logic, edited by Stewart Shapiro, 203–35. Oxford: Oxford University Press. Roselló, Joan. 2012. From Foundations to Philosophy of Mathematics. Newcastle upon Tyne: Cambridge Scholars Publishing. Russell, Bertrand. 1897. An Essay on the Foundations of Geometry. Cambridge: Cambridge University Press. Russell, Bertrand. 1903. The Principles of Mathematics. Cambridge: Cambridge University Press. Russell, Bertrand. 1919. Introduction to Mathematical Philosophy. London: Allen & Unwin. Russell, Bertrand. 1946. History of Western Philosophy. London: Allen & Unwin. Russell, Bertrand. 1958. My Philosophical Development. London: Allen & Unwin. Russell, Bertrand. 1990. The Collected Papers of Bertrand Russell, vol. 2, Philosophical Papers 1896–99, edited by N. Griffin and A. C. Lewis. London: Unwin Hyman. Russell, Bertrand. 1993. The Collected Papers of Bertrand Russell, vol. 3, Toward the “Principles of Mathematics” 1900–02, edited by G. H. Moore. London: Routledge.

178

The Bloomsbury Companion to Bertrand Russell

Russell, Bertrand. 1994. The Collected Papers of Bertrand Russell, vol. 4, Foundations of Logic 1903–05, edited by A. Urquhart. London: Routledge. Russell, Bertrand. 1997. The Collected Papers of Bertrand Russell, vol. 11, Last Philosophical Testament 1943–68, edited by J. G. Slater. London: Routledge. Russell, Bertrand. 1998. Autobiography. London: Routledge. Russell, Bertrand. 2014. The Collected Papers of Bertrand Russell, vol. 5, Towards “Principia Mathematica” 1905–08, edited by G. H. Moore. London: Routledge. Soames, Scott. 2014. The Analytic Tradition in Philosophy, vol. 1, The Founding Giants. Princeton: Princeton University Press. Tennant, Neil. 2013. “Logicism and Neologicism.” The Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/logicism/, accessed March 19, 2016. Whitehead, A. N., and Bertrand Russell. 1925–27. Principia Mathematica, 2nd ed., 3 vols. Cambridge: Cambridge University Press. Yi, Byeong-uk. 2013. “The Logic of Classes of the No-Class Theory.” In The Palgrave Centenary Companion to Principia Mathematica, edited by Nicholas Griffin and Bernard Linsky, 96–129. Basingstoke: Palgrave Macmillan.

7

Russell on Denoting and Language Graham Stevens

Introduction Many people would credit Russell as the founder of analytical philosophy. Even those who express reservations about conferring that accolade on him will agree that his theory of descriptions, published in 1905s “On Denoting,” marks a pivotal moment in the establishment of analytical philosophy as we now know it. This is largely down to the enormous influence of that work in establishing analytical philosophy of language as a core element of modern philosophy. Yet, according to the majority of recent commentators, Russell had no more than a fleeting interest in the philosophy of language. Russell’s greatest philosophical legacy, on this view, was accidental. In this chapter, I will challenge this interpretation and offer in its place a radical reappraisal of the theory of descriptions and its place in Russell’s philosophy, arguing that it was always intended to be part of a wider project in the philosophy of language. That project places the analysis of propositional content at the very heart of philosophy. This, I argue, was Russell’s primary concern throughout his entire work, and the theory of descriptions was a crucial component of it. This is not to say that the many applications of the theory of descriptions to other areas such as epistemology and ontology were not also of central importance to the theory, but I will argue that these applications actually rely on the premise that the theory of descriptions is a theory of natural language semantics and thus serve to demonstrate the fundamental role played by the philosophy of language as the foundation for Russell’s philosophical project as a whole. As I will go on to show, there are some aspects of the theory which Russell prized highly that do pose an obstacle to the continued application of the theory to contemporary philosophy of language. However, I will argue that these aspects of the theory

180

The Bloomsbury Companion to Bertrand Russell

are expendable and that the theory of descriptions can, and should, be exorcised of them.

1  Russellian propositions In his book on Leibniz, written very early on in his philosophical career, Russell provided a clear statement of his approach to philosophical analysis that perfectly captured the key element that would drive his philosophical thinking for the next half century: “That all sound philosophy should begin with an analysis of propositions, is a truth too evident, perhaps, to demand proof ” (Russell, PL: 8). True to his word, Russell’s philosophy always reserved a central position for the analysis of propositional content, even in those periods where he sought to deny the reality of propositions. Russell’s first theory of propositions remains one of his greatest legacies. Russellian propositions, as we now call the entities he posited in the semantic theory underwriting 1903s Principles of Mathematics, are mind-independent and language-independent objects composed of the very things that our words and thoughts are about. The proposition that Socrates is mortal, for example, is an object whose constituents include the individual Socrates and the property of mortality.1 The proposition is both the meaning of the declarative sentence “Socrates is mortal,” and the object of various “propositional attitudes” toward it such as the belief, wish, hope, or fear, that Socrates is mortal. This semantic theory, Russellianism, has a number of highly desirable attributes that continue to make it appealing to contemporary philosophers of language.2 For one thing, it has the virtue of simplicity and elegance—meaning is a very simple sort of a thing on Russell’s theory: words stand for simple entities which are their meanings, and sentences composed from those words stand for complex entities composed of the simple entities. Furthermore, Russell’s key intuition here that (at least some) words stand directly for objects (i.e., the intuition that for many expressions meaning is no more than reference) resonates powerfully with more recent arguments for a direct-reference semantic theory for certain core expressions including proper names, indexicals, and demonstratives. However, despite these attractions, the theory has been plagued by difficulties since its inception. The difficulties divide roughly into metaphysical problems and semantic ones. The most pressing metaphysical problem is the problem of the unity of the proposition. The problem of the unity of the proposition is the

Russell on Denoting and Language

181

problem of how to explain the difference between a proposition and the mere collection of its constituents. Propositions have a distinctive kind of unity that is quite different to the unity had by objects like sets, and unless Russell can explain what that unity consists in, it seems as if the simplicity of his semantics is bought only at the expense of metaphysical mystery. I will not pursue the metaphysical problems here, as the focus of this chapter is Russell’s semantic theory.3 The semantic problems besetting Russellianism stem from Russell’s insistence that propositions are composed of the objects they are about, as opposed to representations of those objects. This yields two fundamental kinds of difficulty. Firstly, it becomes hard to explain how talk of nonexistent things can be meaningful; secondly, the theory faces a cluster of challenges revolving around its counterintuitive predictions about certain apparently intensional contexts. An intensional sentence is distinguished from an extensional sentence by the fact that some aspect of meaning distinct from reference of some or all of the (referring) expressions in the sentence appear to be significant in fixing its truth-conditions, whereas only the reference of its (referring) expressions is significant to a sentence’s truth-conditions in an extensional context. But, as noted above, Russell’s theory effectively eschews any notion of meaning distinct from reference. This, obviously, will make intensional contexts problematic for the Russellian. Consider the following two examples: a. The morning star and evening star are both the same object, namely the planet Venus, seen at different times of the day. If we assume that the expressions “the morning star” and “the evening star” simply have the object Venus as their semantic values, then the following two sentences mean the same thing (express the same Russellian proposition): i. The morning star is the evening star. ii. Venus is Venus Yet, intuitively, (i) is an informative identity statement, whereas (ii)is trivial. b. Assume that the Superman fiction is true. Then, Lois believes that Superman can fly. Superman is Clark Kent. If the meanings of the names “Superman” and “Clark Kent” are simply the individuals they refer to, then, as they refer to the same individual, the two names must contribute the same individual to the proposition expressed by any sentence using these names. In which case, the sentences “Superman can fly” and “Clark Kent can fly” express the same proposition. So Lois must believe that Clark Kent flies. Yet, intuitively, Lois does not believe this.

182

The Bloomsbury Companion to Bertrand Russell

Problems such as these had led Frege to adopt a more complicated semantic theory, Fregeanism, which divides semantic content into two distinct levels, which he termed the sense and reference of an expression. The sense is a mode of presentation of the referent. The difference in sense between the two expressions in the pairs “evening star, morning star” and “Clark Kent, Superman” are appealed to by Frege in explaining the above examples. In the case of (a) the sentence (i) is informative because it identifies two different modes of presentation as being modes of presentation of the same object, whereas (ii) identifies the same object by the same sense, hence it is trivial.4 These two competing semantic theories offer competing ontological accounts of what sorts of things propositions are. Russellianism treats propositions as composed of the individuals they are about; Fregeanism treats propositions as composed of representations of the individuals they are about. To a great extent, the choice between these two positions has defined discussions of the nature of propositional content ever since, at least for those who take propositions to be structured entities.5 Russell’s response to the semantic problems can easily be mistaken for just another version of Fregeanism.6 However, this would be to seriously misunderstand Russell’s position. Russell’s response comes in two parts, only one part of which retains its appeal to this day. The first part, and the one which is still popular, is his theory of descriptions. This theory, which I will explain in detail in the next section, treats definite descriptions as devices of quantification, not reference. The second part, which is now almost universally rejected by philosophers of language, argues that proper names are disguised definite descriptions. Thus, while their grammatical form makes them appear like simple referring expressions, they are in fact quantificational expressions which have no referential function. It is tempting to gloss this as a version of Fregeanism as it also responds to the semantic problems by denying that the expressions in question contribute objects (referents) to the propositions they are being used to express. However, there are some fundamental differences that should not be overlooked. Firstly, Russell is insistent that definite descriptions do not have any representational meanings akin to a Fregean sense. Indeed he goes to great lengths when introducing the theory to argue that the Fregean distinction between sense and reference is unworkable and offers the theory of descriptions in its place. Furthermore, the recent publication of his manuscripts from the period in which he devised the theory of descriptions shows clearly that the

Russell on Denoting and Language

183

theory arose in part out of his criticisms of Frege’s theory, not in response to the semantic problems. Russell repeatedly insists that definite descriptions are “incomplete symbols” which have no meaning at all in isolation. Secondly, Russell is modifying, but not rejecting, Russellianism. Although he is recognizing that not all propositions are as simple as the original version of Russellianism presented them as being, he is maintaining the core thesis of Russellianism that propositions contain the things they are about. However, he no longer thinks that sentence of the form “The F is G” are about the F. Rather he thinks, as we shall see in the next section, that they have on analysis a more complicated structure that reveals the propositions they express to be more complex also. Furthermore, despite his insistence that proper names are not devices of reference (and hence that the proposition that Socrates is mortal does not contain Socrates), he still maintains that Russellianism is strictly speaking correct, in the sense that every genuine device of reference will contribute its referent as its semantic value. He simply draws a distinction between logically proper names and grammatically proper names. The grammatically proper name “Socrates” is not a logically proper name. Other expressions, however, such as demonstrative pronouns, are still taken to be logically proper names and these receive the same treatment as in earlier versions of the theory. Thirdly, Russell’s quantificational analysis of descriptions is quite different to Frege’s claim that descriptions have a sense. We might helpfully think of the situation as follows. The presence of obviously quantificational expressions, like “all,” “every,” “no,” and so on, in English already demonstrates the need to recognize expressions that do more than just refer directly to objects. Both Frege and Russell do so by introducing a syntactic theory of quantification. By extending the treatment of quantifiers to embrace definite descriptions, Russell is attempting to explain the semantic puzzles by appeal to this syntactic mechanism rather than by appeal to an additional semantic mechanism like the sense/reference distinction.

2  Russell’s theory of descriptions According to Russell’s 1905 theory of descriptions, the grammatical similarity between definite descriptions and referring singular terms is an illusory accident of English (and, presumably, many other natural languages, although Russell never really address this question). The situation, as mentioned above, is

184

The Bloomsbury Companion to Bertrand Russell

somewhat complicated by Russell’s additional claim that many of the expressions we would think of as paradigmatic referring expressions, such as proper names, are what he calls “disguised” descriptions. It will help to elucidate Russell’s position if we overlook this additional thesis momentarily. Consider these two sentences: a. David Cameron is dishonest. b. The present prime minister of the United Kingdom is dishonest. Grammatically, both sentences seem to combine a noun phrase with a predicate expression. So they seem to share the same syntax. This might naturally lead us to expect a shared semantic structure also. However, Russell’s theory takes the surface grammar of English to conceal the true structure of the propositions these sentences express. The correct logical forms (i.e., the structures of the proposition each sentence expresses) are radically different, and can be paraphrased as follows: (LFa) David Cameron is dishonest. (LFb) There is one and only one x such that x is presently prime minister of the United Kingdom and x is dishonest. The definite description, in other words, has a very different semantic function to the name, despite the fact that both appear to be used to talk about the same person. The name introduces that person directly into a proposition, whereas the definite description introduces into a proposition a uniquely quantifying concept that picks out that person. The analysis of definite descriptions is just one part of the theory of descriptions developed by Russell in 1905. The complete theory is a general theory of quantification that locates definite descriptions within the class of firstorder quantifiers along with expression like “all Fs,” “every F,” “no F,” “some F(s),” “not all Fs,” and indefinite descriptions like “an F.” The publication of the first volume of Principia Mathematica in 1910 effectively established as orthodoxy the view that the meanings of these expressions can be captured within firstorder predicate logic, and their translation into that language is now standard fare for all students of the predicate calculus. The orthodoxy nonetheless has not gone unchallenged, and the inclusion of definite descriptions on this list of quantifier expression has been by far the most controversial claim. Perhaps its fiercest and most influential challenge came from Strawson (1950). Strawson’s general objection was summarized by his insistence that “ordinary language

Russell on Denoting and Language

185

has no exact logic” (1950, 27), an objection grounded in a series of more local objections to the theory of descriptions. Strawson objected to the theory of descriptions on both a negative and a positive front. The negative point consists in a famous problem that he raised for the theory called the problem of incompleteness. The positive point consists in his alternative proposal about the functioning of definite descriptions which draws extensively on pragmatics rather than semantics and which, he maintains, provides more reliable predictions concerning the behavior of “empty” descriptions like “the present king of France.” Both points have been highly influential and continue to be the focus of discussion to date. The problem of incompleteness arises because it seems that everyday uses of definite descriptions are far more relaxed with regard to the uniqueness of the thing described than Russell’s theory appears to predict. For example, it is perfectly natural to say things like “the book is over there” in a context where the speaker is communicating the location of a particular book. But Russell’s theory analyses this utterance as encoding the proposition that one and only one thing is a book and that that thing is in the demonstrated location. In all worlds, including the actual world, containing more than one book, this proposition is false. Yet, intuitively, the utterance has communicated something true. This must mean, so the objection runs, that Russell is wrong about which proposition is encoded by the utterance. The problem of incompleteness is best countered by the observation that it can be viewed as an instance of the general natural language phenomenon of contextually mandated quantifier restriction. All natural language quantifiers are subject to restriction within context to a narrowed range of variables. For example, when I am asked how my lecture on the theory of descriptions went, I might reply “everyone was bored,” but here it is obvious that the range of the quantifier “everyone” has been restricted to a contextually salient class of people (those who were present at the lecture). Close inspection reveals that such contextual restriction of quantification is routine in natural language. Competing semantic and pragmatic explanations of how exactly this phenomenon functions can be given, but the only point that really matters here is that, if we take the problem of incompleteness to be an instance of this general phenomenon, then definite descriptions are clustering with quantifier phrases more generally. Thus, viewed in this way, the problem of incompleteness is turned on its head, becoming evidence for, rather than against, Russell’s quantificational analysis of definite descriptions.

186

The Bloomsbury Companion to Bertrand Russell

Strawson’s positive proposal is usually seen as a contribution to pragmatics. Strawson argues that uses of definite descriptions introduce a “presupposition” of reference to contexts in which they feature. For example, interpretation of the sentence “the present CEO of Google stole my wallet” requires the presupposition that there is a present CEO of Google. Without this presupposition it is not clear that the sentence even expresses a proposition at all. Certainly, according to Strawson, it fails to express anything holding a truth-value. Thus, in cases of sentences displaying reference failure, such as we encounter with sentences like “the present king of France is bald,” Strawson predicts that the sentences will lack a truth-value. In fact, intuitions appear to diverge quite wildly with respect to this example. Some hear the sentence as straightforwardly false (as Russell predicts), whereas others find it odd to attribute a truth-value to the sentence (as Strawson predicts). Nonetheless, there are other cases where intuition does not seem to be on Strawson’s side at all. Stephen Neale (1990) points out that competent English speakers who know that France lacks a monarchy are in almost universal agreement regarding many utterances of sentences about the present king of France. Certainly, the current students on my Russell course at Manchester were in universal agreement that my claim that “I had dinner with the present king of France last night” is straightforwardly false when I tested this slight modification of Neale’s example out in a recent lecture. Hence, it seems, that the question of whether all uses of definite descriptions presuppose reference is harder to answer than Strawson thought. Regardless of whether his proposal should depose Russell’s quantificational analysis, however, there can be no doubt that Strawson’s work has revolutionized pragmatic theory, which has developed the theory of presupposition to areas far beyond the domain of definite descriptions. To a large extent, Strawson and Russell can be understood as championing the referential and the quantificational interpretations of descriptions respectively. The next great development in the debate came with Donnellan’s (1966) argument that both Strawson and Russell were partially correct and, at the same time, partially incorrect. Donnellan provided a series of examples demonstrating both referential and quantificational (or, as he termed them, attributive) functions. Imagine a case where detectives are called to the scene of a suspected murder. The body of Smith is found brutally butchered. Confronted by what appears to be an attack of extreme violence, one of the detectives says “Smith’s murder is insane.” This is an attributive use of the description “Smith’s murderer.” No

Russell on Denoting and Language

187

particular object is in the detective’s mind when uttering the sentence. Rather the detective means something like “whoever it was that did this must be insane.” Next, consider a case where a suspect has been arrested on suspicion of the murder and is being questioned in court. The suspect acts extremely unpredictably in the courtroom, talking to themselves, responding to questions with bizarre answers, and so forth. Someone in the courtroom watching this display whispers to the person next to them “Smith’s murderer is insane.” This is a referential use. What the speaker means is something like “that person over there is insane.” Indeed, they will have successfully communicated this information even if the person in question does not answer to the description (e.g., if they have been wrongfully accused of the murder). In each of these cases the Strawsonian and Russellian analyses make differing truth-conditional predictions about the content of the utterances. The Strawsonian correctly captures the truth-conditions of the referential use; The Russellian captures those of the quantificational/attributive use. Following Kripke’s influential (1977) response to Donnellan, debate has centered (inconclusively) around whether one or the other of these analyses should be given priority as the correct account of the literal meaning of the utterance, while the other account is accommodated within a pragmatic explanation of nonliteral uses of the sentences in question.

3  Logical form, metaphysics, and semantics Few debates have received as much attention in the philosophy of language as that instigated by Russell’s theory of descriptions. Yet, somewhat surprisingly, it is common among scholars of Russell’s philosophy to view this debate as no more than an accidental outcome of the theory. Russell himself, it is commonly claimed, was no philosopher of language and neither was the theory intended to bear on issues in the semantics of natural languages. Certainly when one looks at both the context in which Russell discovered the theory, and the ingenious applications he immediately put the theory to after its discovery, it appears that natural language semantics was not Russell’s primary concern. The theory was discovered and developed in response to the problems facing Russell’s logicist project. In particular, the theory played a crucial role in the philosophical explanation of the theory of logical types which was intended to preserve the formal system of Whitehead and Russell’s Principia Mathematica

188

The Bloomsbury Companion to Bertrand Russell

from the paradoxes of naïve set theory, most notably Russell’s own paradox of the class of all classes which are not members of themselves. On April 14, 1904, Russell wrote to his first wife that he and Whitehead “had a happy hour yesterday when we thought the present King of France had solved the Contradiction; but it turned out finally that the royal intellect was not quite up to that standard. However, we made a distinct advance” (Russell 1992, 277). Comments like this make it fairly plain that, right from the outset, Russell’s interest in descriptions stemmed from his ongoing battles against the contradictions. He later wrote that the theory of descriptions was “the first step toward overcoming the difficulties that had baffled me for so long” (Russell Auto 2, 152). The way that the theory helped was through the introduction of an analysis that allowed apparent logical subjects to be treated as “incomplete symbols,” showing that apparent reference to an entity (e.g., a class) could be explained without any ontological commitment to that object. This was crucial to the no-classes theory underlying the theory of types in Principia.7 Russell’s own aims for the theory of descriptions then were clearly directed at concerns in logic and metaphysics. But it would be a mistake to conclude from this that the theory was not intended to make a contribution to the semantic analysis of natural language definite descriptions. Russell often (particularly after 1905) talks of natural language being defective or deficient, urging its replacement with what he calls a “logically perfect language” (Russell 1918, 198), leading to his widespread characterization as the founder of so-called “Ideal Language Philosophy.” There are, I think, at least two ways in which this term is used. On the first interpretation of ideal language philosophy (IP1), natural languages are viewed as deficient in the sense that they simply represent the world in an inaccurate way. So, for example, English uses expressions which have exactly the same linguistic form as referring expressions when in fact reality does not contain any objects for those expressions to refer to. Thus, English misrepresents how things really are. It would be better, according to IP1, to abandon English (at least within certain domains which strive for an accurate representation of reality, such as precise philosophy) and replace it with a logically perfect language which did not misrepresent reality in this way. Notice that IP1 does not simply urge an alternative grammar to replace that of the natural language in question, but also urges the adoption of an alternative semantics or, more precisely, urges the adoption of a language with a fundamentally different grammar and semantics. An example of a philosopher who defends IP1 would be Quine. Quine’s proposal that all names should be

Russell on Denoting and Language

189

converted into predicates8 and then uniquely quantified over so as to reveal our ontological commitments through our existentially quantified commitments is an attempt to replace natural language with something else. It is not in any way intended to be a thesis about the meanings of natural language names, indeed it is intended on the contrary to urge a wholesale revision of those meanings—a self-conscious adoption of an alternative semantics, rather than a statement of the original semantics. On the second interpretation of Ideal Language Philosophy (IP2), what logical analysis does is to reveal a semantic structure which is superficially disguised by natural language. So, for example, English grammar seems to treat all noun phrases as having the same function. But, in fact, the noun phrase “a man” has a very different semantic function to the proper name “Socrates” when combined with the verb phrase “drank poison.” A logically perfect language will remove the disguise and unveil the true structure and semantic function of natural language expressions. An example of a philosopher defending IP2 would be Frege, whose development of predicate logic is, among other things, designed to reveal the difference in semantic function (or, as he would call it, logical form) of these sorts of expressions. Evidently, the defender of IP2 cannot maintain a position of neutrality on issues in the philosophy of language in the way that the defender of IP1 can. Despite maintaining the logico-semantic deficiency of natural language, IP2 is a philosophical claim about natural language. Which variety of ideal language philosophy, then, is Russell defending? There are two clearly distinct positions here. However, these were certainly not disambiguated by either Russell or his contemporaries and Russell slides quite ambivalently between the two. For example, Russell leans strongly toward IP1 when he explicitly warns against the dangers of expecting linguistic distinctions to mirror metaphysical ones in the chapter on “words and meaning” in The Analysis of Mind: The things that words mean differ more than words do. There are different sorts of words, distinguished by the grammarians; and there are logical distinctions; which are connected to some extent, though not so closely as was formerly supposed, with the grammatical distinctions of parts of speech. It is easy, however, to be misled by grammar, particularly if all the languages we know belong to one family. In some languages, according to some authorities, the distinction of parts of speech does not exist; in many languages it is widely different from that to which we are accustomed in the Indo-European languages. These facts have to be borne in mind if we are to avoid giving metaphysical importance to mere accidents of our own speech. (Russell, AMi, 191–92)

190

The Bloomsbury Companion to Bertrand Russell

Elsewhere Russell diagnoses this same structural mismatch between language and reality as the source of the ancient question concerning the nature of relations (which, it should be noted, had been a constant source of puzzlement throughout Russell’s own work in previous years): There is however, a complication about language as a method of representing a system, namely that words which mean relations are not themselves relations, but just as substantial or unsubstantial as other words. In this respect a map, for instance, is superior to language, since the fact that one place is west of another is represented by the fact that the corresponding place on the map is to the left of the other; that is to say a relation is represented by a relation. . . . I believe that this simple fact is at the bottom of the hopeless muddle which has prevailed in all schools of philosophy as to the nature of relations. (Russell 1923, 152–53)

Both of these passages are easily read as advocating something like IP1. However, it would be a mistake to ascribe the view to Russell as a universal philosophical attitude toward language. In particular, it would be wrong to interpret the theory of descriptions as part of this kind of project for two reasons: firstly, there is no compelling evidence in Russell’s writing on the theory to support such a view, and plenty of evidence against it; secondly the theory of descriptions becomes largely incoherent if interpreted in this way. I will expand on each reason in turn. Russell’s justifications for the theory of descriptions repeatedly appeal to arguments about the semantic function of definite descriptions in natural language. His classic statement of the theory in “On Denoting,” for example, adopts the strategy of posing three puzzles which, he maintains, must be solved by any viable theory of denoting. All three puzzles are straightforwardly semantic puzzles—puzzles about the meanings of denoting phrases, or the sentences containing them, in natural language. The first puzzle is a version of Frege’s puzzle concerning de dicto attitude reports involving expressions which denote the same object. This is simply a question about the semantics of propositional attitude reports involving denoting phrases. The second is the famous “present king of France” case, and concerns the apparent failure of the law of excluded middle which occurs when a sentence of the form “A is B” and its (apparent) negation “A is not B” has an empty definite description like “the present king of France” in place of “A.” This is simply a question about the truth-conditions of English sentences with non-referring terms in subject position. The third puzzle also concerns the semantics of non-referring expressions: we can form a definite description “the difference between A and B” regardless of whether A and B

Russell on Denoting and Language

191

actually differ. For example, if they do not differ then it is meaningful and true to say “the difference between A and B does not exist.” Russell, due to other commitments in his semantic theory which are not clearly expounded in “On Denoting,” thinks that there is something very paradoxical about a proposition that denies the existence of its own subject. This may not be as familiar a case of a semantic concern to those unfamiliar with Russell’s philosophy but it is a concern about semantics all the same. The theory of descriptions is forwarded as providing a complete solution to all three puzzles. It solves the puzzle of de dicto attitude reports which differ in truth-value despite ascribing seemingly the same attitudes about the same objects, by analyzing the propositions in question into complex quantificational forms in which they turn out not to be about the same objects after all. George IV, in Russell’s example, can wonder whether Scott is the author of Waverley without this collapsing into the trivial question of whether Scott is Scott because the proposition George actually stands in an attitude toward is the proposition that Scott is identical with one and only one thing which authored Waverley, whereas he does not stand in the same relation to the proposition that Scott is self-identical.9 This is clearly not any part of a defense of IP1. Russell is not arguing that English needs to be abandoned and replaced with a language with a different semantics, he is simply presenting an argument for ascribing a given semantic interpretation of part of English. The solution offered to the second puzzle is directly aimed at the semantic profile of the English sentence “The present king of France is not bald.” Russell’s argument here is that the sentence, intuitively, has two readings: 1. There is an entity which is presently king of France and it is not bald. 2. It is false that there is an entity which is presently king of France and bald. Russell’s theory, we know, predicts this brilliantly as the quantificational form of the sentence it ascribes permits a syntactic interplay between the existential quantifier and the negation operator which gives reading (1) or (2) depending on which of these has wide scope over the other: (1a) ∃x ((Fx & (∀y)(Fy ⊃ x = y)) & ~Gx) (2a) ~∃x ((Fx & (∀y)(Fy ⊃ x = y)) & Gx) This is a paradigm instance of IP2, and a world away from IP1. Russell is directly appealing to the semantic behavior of English sentences and to our intuitions qua English speakers about the truth-conditions of English sentences to argue for a given semantic interpretation of English definite descriptions.

192

The Bloomsbury Companion to Bertrand Russell

The third puzzle, concerning the formation of propositions seemingly about things which do not exist, is solved by the same mechanism of course—on the analyses given above, propositions about “the F” are not really about any entity after all, as a survey of their logical forms reveals. No expression referring to the present king of France can be found in (1a), for example. Again, we have a solution to the puzzle which relies on arguing for a certain semantic interpretation of the natural language expression in question. Another clear instance of IP2, in other words. The above considerations give us overwhelming evidence for interpreting the theory of descriptions as a contribution to a project along the lines of IP2, not IP1 (regardless of whether other aspects of Russell’s philosophy may be better interpreted as contributions to IP1). But there is an even stronger reason than this for situating the theory of descriptions within the framework of IP2, namely the theory is simply incompatible with IP1, as is made clear by considering the ways in which Russell appeals to the theory when applying it more widely within his philosophy. Consider, by way of an example, Russell’s application of the theory to provide a foundation for his epistemological distinction between knowledge by acquaintance and knowledge by description. The latter sort of knowledge allows us to have thoughts about objects that we are not acquainted with only because it is grounded in the theory of descriptions, according to which knowledge about such objects is accessed via certain concepts that they fall under (including the quantificational concept of uniqueness). It is crucial to this claim, however, that the theory of descriptions is a theory of the semantic content of natural language definite descriptions. The theory of descriptions is doing all of the hard work in this claim, and the epistemological thesis rides on the back of it. If definite descriptions were devices of direct reference in natural language, the proposal would not get off the ground at all. This is also why Russell’s extension of the distinction to the use of names forces him to endorse the semantic claim that proper names really do have a descriptive content. The Quinean (IP1) position whereby we replace referring expressions with quantificational ones in order to reveal ontological commitments more clearly would not establish the epistemological claim that Russell endorses.10 Russell’s position is that we cannot be acquainted with the objects seemingly referred to by most proper names, hence our understanding of those names must not be a matter of our apprehending those objects. The theory of descriptions explains how we can

Russell on Denoting and Language

193

understand them only if we take the extra semantic step of insisting that the contents of names are not objects at all, but that names have an entirely different, quantificational, function. The theory of descriptions is thus, at the most fundamental level, a theory of natural language semantics, and its role in Russell’s philosophy even when it is applied to issues beyond those directly pertaining to semantics remains firmly anchored in those semantic considerations.

4  Going forward: Two problems with the theory of descriptions Following on from the above discussion of the theory of descriptions as a theory of natural language semantics I now want to discuss two problems with the theory, which suggest, I think, a need for some modifications to Russell’s theory if it is to remain a viable account of the semantics of natural language quantification. The first problematic feature of the theory is Russell’s conflation of semantic and epistemological considerations. The second problematic feature is Russell’s insistence that descriptions are “incomplete symbols.”

4.1  Semantics versus epistemology The theory of descriptions is a semantic theory. Unfortunately, however, Russell ran together semantic and epistemological considerations right from the very inception of the theory. This conflation of semantics and epistemology may very well have been present in his thinking before the discovery of the theory in 1905. Indeed, the best explanation for a number of passages in 1903s Principles of Mathematics, requires the assumption that Russell is implicitly appealing to an epistemological principle to justify his semantic theory of denoting concepts.11 The epistemological principle is made explicit, however, in “On Denoting,” where Russell commends the theory on the grounds that it preserves the principle: Thus in every proposition that we can apprehend (i.e., not only those whose truth or falsehood we can judge of, but in all that we can think about), all the constituents are really entities with which we have immediate acquaintance. (Russell 1905, 56)

194

The Bloomsbury Companion to Bertrand Russell

This principle was later dubbed “Russell’s Principle” by Gareth Evans, who interpreted it as follows: In order to have a thought about a particular object, you must know which object it is about which you are thinking. (Evans 1982, 74)

For Russell, this epistemological component was a core commitment of the theory of descriptions that he recognized immediately upon devising the theory: This topic is very interesting in regard to theory of knowledge, because most things are only known to us by denoting concepts. Thus Jones = the person who inhabits Jones’s body. We don’t have acquaintance with Jones, but only with his sensible manifestations. Thus if we think we know propositions about Jones, this is not quite right; we only know propositional functions which he satisfies, unless indeed we are Jones. Thus there can be no such thing as affection for persons other than ourselves; it must be either their sensible manifestations or the concepts denoting them that we like. It cannot be the latter, for it would be absurd to say that we loved some of these and hated others. Denoted objects only known to us as denoted may be identified, without such great error as in other cases, with the sum of their predicates; for it is only their predicates that we know, and these (all or some of those we know) must be meant by us whenever we speak of such objects. But we can only know an object as denoted if we are acquainted with the denoting concept; thus immediate acquaintance with the constituents of the denoting concept is presupposed in what we may call denotative knowledge. (Russell 1905a, 369)

This distinction between knowledge by acquaintance and knowledge by description may be a perfectly valid and important one in epistemology, but Russell’s importation of the distinction into his semantic theory has a number of unfortunate consequences. For one thing, it makes the step toward a descriptive theory of proper names a natural one, for the same epistemological puzzles that are solved by appeal to the distinction in kinds of knowledge appear mirrored in the semantic puzzles discussed above. Consequently, Russell’s theory of descriptions has been unjustly blended in the eyes of many of Russell’s interpreters with a rather messy hybrid of the descriptive theory of names and the distinction between knowledge by acquaintance and knowledge by description. In doing so, they have only been following Russell’s own lead. But regardless of whether Russell himself saw these three things as essentially interconnected, and regardless of any evaluation of these three things, it is a mistake for us to view the theory of descriptions as committed to either of these extraneous additions about

Russell on Denoting and Language

195

names or knowledge. There are three distinct theories here, all of which are in fact entirely independent of one another: 1. The semantic theory that definite descriptions contribute quantificational, not referential, content to the propositions expressed by the sentences in which they feature. 2. The semantic theory that proper names have descriptive content. 3. The epistemological theory that knowledge of objects comes in two distinct varieties: (a) direct knowledge of (acquaintance with) objects; (b) indirect (descriptive) knowledge of objects. That these three are independent of one another is fairly evident: one could endorse (1) but not (2) (as, for example, Kripke does12); one could endorse (2) but not (1) (as, for example, Frege did13); one could endorse (3) without either (1) or (2) (as, arguably at least, Russell did in 1903). Aside from the messy interpretation of his theory that Russell’s conflation of these issues has resulted in, there is a more serious concern. There is good reason to doubt the truth of Russell’s principle. David Kaplan gives the following vivid example as a demonstration that epistemological access to the things one is talking about is not a necessary condition for referring to them: A kidnapped heiress, locked in the trunk of car, knowing neither the time nor where she is, may think “It is quiet here now” and the indexicals will remain directly referential. (Kaplan 1989, 536)

In Kaplan’s example, we have a very clear case of the paradigm of directly referential expressions, indexicals, successfully referring in the absence of Russell’s epistemological criterion. This shows that it is as mistaken to import (3) into the theory of descriptions as it is to import (2) into it. The theory of descriptions (1) must be permanently severed from (2) and (3) if it is to make a worthwhile contribution to continued theorizing about natural language quantification.14

4.2  Incomplete symbols Let us now turn to the second problem with Russell’s theory that I mentioned above, namely his insistence that descriptions are incomplete symbols. This view is crucial to the ends that he has in mind for the theory of descriptions, most notably it is this aspect of the theory which connects with the theory of types. Russell’s theory of types was, for many years, understood as heralding a fundamental change of heart on the part of its author about the nature of logic

196

The Bloomsbury Companion to Bertrand Russell

and of the interplay between logic and metaphysics. Russell’s original version of logicism was grounded in a simple metaphysics where all objects stood on a logical par. Russell endorsed a position which has been called “the doctrine of the unrestricted variable,” according to which the variables of pure logic range over all things without any division of those things into distinct logical categories. Logic, on this view, is independent of any metaphysical assumptions about the nature of entities (though, of course, it may not be independent of metaphysical assumptions about which, or how many, entities there are). The theory of types appears to be a straightforward rejection of this position, as it restricts the range of a variable in accordance with type indexes. In recent years, however, the discovery and publication of Russell’s manuscripts during the period in which he developed the theory of types has shown that this interpretation is, at best, overly simplistic.15 For at least some of the time during which Russell was developing the theory of types, he was intending to do so in a way which allowed him to maintain the doctrine of the unrestricted variable. The key to this was his “substitutional theory of classes and relations,” developed during the period 1905–1907, but eventually abandoned before the completion of Principia Mathematica. The substitutional theory provides the missing link between Russell’s theory of descriptions and his theory of types, making sense of his later claim, cited above, that it was the discovery of the former which held the key to the latter. The theory makes it very clear that the key feature of the theory of descriptions which paved the way for the theory of types was the notion of an incomplete symbol. By an “incomplete symbol,” Russell is best understood16 as meaning a linguistic symbol which has no meaning that can be assigned in isolation of a linguistic context, but which nonetheless makes a systematic contribution to the meanings of linguistic contexts in which it does feature. So “the present prime minister of the United Kingdom” is, strictly speaking, meaningless, but “the present prime minister of the United Kingdom is dishonest” is meaningful. Russell does offer arguments for this claim as a thesis regarding definite descriptions in natural language. However, his case is far stronger as an observation about the behavior of certain descriptive constructions within the formal system of his substitutional theory. The substitutional theory makes a wholesale revision of predicate logic (most notably higher-order predicate logic), to such a dramatic extent that it can seem quite alien on first acquaintance. In place of propositional functions of the sort familiar from predicate logics, Russell introduces constructions

Russell on Denoting and Language

197

which he calls matrices. The simplest kind of matrix is of the form: p/a, where “p” and “a” are understood as wholly unrestricted variables ranging over any entities whatsoever. Russell then introduces a primitive substitution operation, symbolized as p/a;x which can be read as a definite description “the result of replacing a in p by x.” We then have propositions describing the results of these substitutions of the form “p/a;x!q” which can be read: “q results from the substitution of a in p by x.” Although the substitutional theory only had very limited exposure in Russell’s published work,17 he devoted a great deal of effort to this new system of logic. In manuscripts he developed detailed axiomatizations of the system, proving many important theorems and clearly working toward an attempted demonstration of logicism within the system. For reasons that we will not go into here, that demonstration was not to be completed, and Russell was eventually persuaded to forego substitution as the formal language for Principia Mathematica.18 Nonetheless, we can see from his incomplete efforts, why it was that he felt that the justification for theory of types lay in the theory of descriptions. The real benefit of substitution becomes apparent when we see the application that Russell intends for the matrices of the theory: “The theory which I wish to advocate is that this shadowy symbol p/a represents a class” (Russell 1906b, 170). In other words, the idea behind the theory is that these matrices have all of the formal properties needed of classes. Yet, and this is the key point, they are not themselves entities. Russell defines “class” membership as follows: x is a member of the “class” p/a just in case the substitution of x for a in p results in a true proposition. Notice that neither p nor a must be propositions for the matrix to be well-formed; however, only in the case where p is a proposition, can the “class” be non-empty. Russell’s ontology must embrace propositions for the substitutional theory to have a coherent interpretation, but it does not need to embrace classes. Hence the substitutional theory is, in a perfectly clear way, Russell’s first no-classes theory. What makes the substitutional theory most impressive, however, is that once we have defined “classes” as derived wholly from matrices of propositions, we effectively gain a theory of logical types for free. This follows immediately from the fact that only entities may be substituted for entities. Not being itself an entity, a matrix cannot be substituted for an entity in another matrix. This immediately blocks any possibility of self-membership among “classes.” If, for example, p/a is a class, then any attempt to predicate self-membership of this class will be simply ungrammatical in the language of substitution. We cannot have matrices of

198

The Bloomsbury Companion to Bertrand Russell

the form p/a;(p/a) because only an entity can be substituted for a in p/a.19 As Russell puts it: But now “x is an x” becomes meaningless, because “x is an α” requires that α should be of the form p/a, and thus not an entity at all. In this way membership of a class can be defined, and at the same time the contradiction is avoided. (Russell 1906b, 172)

Russell’s theory of types has been the subject of a number of vehement criticisms over the past hundred years. The substitutional theory shows that many of these are unfair. For example, Russell’s claim that violations of type theory are “ungrammatical” or “nonsensical” has been dismissed as mere ad hoc stipulation. But we can see, within the framework of substitution, that the claim is simply justified. Furthermore, related claims like Whitehead and Russell’s insistence that they do not countenance the existence of classes, which seem quite cryptic when presented in Principia, make perfect sense when viewed in the light of the substitutional theory (though, whether Russell and Whitehead are still entitled to maintain the same claim in Principia where substitution has been dropped is another question, of course, which I will not go into here). Most importantly for our interests here, the connection between the theory of descriptions and theory of types is made explicit by the substitutional theory. It is only because matrices are incomplete symbols that they generate type distinctions. This is a remarkable achievement that demonstrates Russell’s technical genius. But we should be careful what we conclude from it. Whatever merits it may have for Russell’s mathematical logic, it is quite independent of the claim that definite descriptions in natural language are incomplete symbols. To demonstrate that, Russell needs to appeal to raw linguistic data; as we will now see, his attempts to locate such data are far less convincing. Russell’s favored argument for demonstrating that definite descriptions are meaningless in isolation occurs in a number of places, but the following passage from Principia probably puts the argument in its clearest form: It can be easily shown that (ix)(ϕx) is always an incomplete symbol. Take, for example, the following proposition: “Scott is the author of Waverley.” . . . This proposition expresses an identity; thus if “the author of Waverley” could be taken as a proper name, and supposed to stand for some object c, the proposition would be “Scott is c.” But if c is anyone except Scott, this proposition is false; while if c is Scott, the proposition is “Scott is Scott,” which is trivial, and plainly different from “Scott is the author of Waverley.” Generalizing, we see that the proposition

Russell on Denoting and Language

199

a = (ix)(ϕx) is one which may be true or may be false, but is never merely trivial like a = a; whereas, if (ix)(ϕx) were a proper name, a = (ix)(ϕx) would necessarily be either false or the same as the trivial proposition a = a. We may express this by saying that a = (ix)(ϕx) is not a value of the propositional function a = y, from which it follows that (ix)(ϕx) is not a value of y. But since y may be anything, it follows that (ix)(ϕx) is nothing. Hence, since in use it has a meaning, it must be an incomplete symbol. (PM, 67)

The problem with the argument is that Russell wants it to support a stronger conclusion than it really establishes, namely that definite descriptions are meaningless. The argument seems to support that conclusion because the term “nothing” in the penultimate sentence is ambiguous. Russell is talking about the range of values of the variable y in the function “Scott  =  y” can take. Now clearly only objects are going to be in this range of values. So what the argument (if it is valid) really establishes is that definite descriptions do not contribute objects to propositions. But that is not the same as their being meaningless. A weaker—and better—conclusion to draw from Russell’s argument is that definite descriptions (along with other quantifier expressions) make available a class of object-independent propositions.20 If we wish, we can put a metaphysical gloss on the distinction between object dependent and object-independent propositions by talking in terms of whether or not objects are contained in the propositions as constituents. But this metaphysical gloss is not essential. We can think of object-dependent propositions as propositions expressed by sentences whose truth-conditions depend on objects, and object-independent ones as expressed by sentences whose truth-conditions do not. For example, the proposition that Scott is happy (that is, the meaning of the sentence “Scott is happy”) is true just in case the object Scott has the property of being happy, and false just in case the object Scott lacks the property of being happy; while the proposition that the author of Waverley is happy is true just in case one and only one thing authored Waverley and is happy, and false (on the wide scope reading) just in case it is not the case that one and only one thing authored Waverley and is happy. The difference may seem subtle at first, but it is significant. The first proposition depends on Scott for its truth or falsehood; the second does not. The second in fact has precisely the same feature that Frege first identified as distinctive of quantificational propositions—it is best understood as making a claim about the concepts or properties, which we might (using the notation of

200

The Bloomsbury Companion to Bertrand Russell

lambda-abstraction) specify as λx (x authored Waverley) and λx (x is happy), and its truth depends on features of the instantiation of these concepts, rather than on any object. Frege himself gave the example of the quantificational sentence “All whales are mammals” to make this point: It is true that at first sight the proposition “All whales are mammals” seems to not be about concepts but about animals; but if we ask which animal then are we speaking of, we are unable to point to any one in particular. Even supposing a whale is before us, our proposition still does not state anything about it. . . . As a general principle, it is impossible to speak of an object without in some way designating or naming it; but the word “whale” is not the name of any individual creature. . . . However true it may be that our proposition can only be verified by observing particular animals, that proves nothing as to its content. (Frege 1884, 60–61)

If we are persuaded by the theory of descriptions we should be persuaded that, like all other quantified propositions, descriptive propositions are independent of particular objects in precisely the way that Frege is describing here. In other words, Russell’s claim that descriptions are “incomplete symbols” that are meaningless in isolation, as powerful a notion as that may have been in the context of his mathematical logic, is best replaced by the claim that descriptive propositions are object independent if the theory is to be taken seriously as an account of the meanings of natural language quantifiers.

5 Conclusion There can be no doubt that the theory of descriptions is located at the very core of Russell’s philosophy. It impacts dramatically on his metaphysics, his epistemology, and, of course, his mathematical logic. These impacts, however, have led many readers of Russell to mistake the theory itself as a contribution to one, or more, of these areas. Subsequently, they have concluded that the theory was never intended by Russell to be a contribution to the philosophy of language. Russell, on this interpretation, played the leading role in establishing philosophy of language as a core enterprise of twentieth-century analytical philosophy by accident. Many of Russell’s own comments on philosophy and the so-called linguistic turn taken in early part of the twentieth century are easily read as lending support to this interpretation. However, as I have argued at length elsewhere (Stevens 2011), the hostility displayed by Russell toward those who took this

Russell on Denoting and Language

201

turn is not really hostility against philosophers of language in general, but against the “ordinary language” philosophers inspired by the Later Wittgenstein, a school of thought dismissed by Russell as “completely unintelligible” (MPD, 160). Many, perhaps most, contemporary philosophers of language are likely to share Russell’s judgment of this school. Furthermore, the naturalistic approach to language adopted by Russell from 1919 onwards, finds many sympathizers among current philosophers of language. More importantly, as I have demonstrated above, the theory of descriptions itself should not be viewed through a lens distorted by its applications. The theory is only intelligible as a theory of the semantics of natural language quantification. It is this doctrine which lies at the heart of the many ingenious applications Russell put the theory of descriptions to. Far from being disinterested in the philosophy of language, therefore, Russell made it the essential foundation of his most significant philosophical works.21

Notes 1 Whether the proposition contains any other constituents proved to be a difficult question for Russell to answer, as he grappled with the question of whether some extra entity or collection of entities is required to unify Socrates and mortality into the proposition that Socrates is mortality (see Stevens 2005 for extended discussion). 2 See, for example, Kaplan 1989, Kripke 1980, Salmon 1986, King 2007, among many others. 3 I have pursued it in detail in several other places, including Stevens 2003, 2004, 2005, 2008. 4 An additional semantic problem arises when we consider what semantic value “empty” names like “Bilbo Baggins” or “Pegasus” have. This is a separate problem which I shall not consider here (see Stevens 2011 for discussion), hence our assumption that the Superman fiction is true in the above example. 5 Alternative, unstructured, accounts of propositions usually take them to be identified with sets of worlds at which they are true. 6 A third theory, which we will not consider here, was proposed in Russell’s (1903). This theory has some similarities with both Fregeanism and Russellianism, but is importantly distinct from either. Denoting phrases, on this theory, express denoting concepts which have often been construed as akin to Fregean senses, however the theory (like Russell’s later 1905 theory) treats denoting phrases as quantifier phrases (though his analysis of quantification is quite different to that provided by the

202

7 8 9

10

11

12 13 14

15 16

17

The Bloomsbury Companion to Bertrand Russell later theory). Russell himself helps to confuse things by describing his 1903 theory as ‘very nearly the same as Frege’s’ (Russell 1905, 42) when presenting his later alternative. This is a point I will return to in detail below. Quine 1948. Russell actually confuses the example by using a demonstrative when explaining the wide scope reading, which he says ‘would be true, for example, if George IV had seen Scott at a distance, and had asked “Is that Scott?”’ (Russell 1905, 52; see Kripke 2005, Stevens 2011 for discussion). This is precisely why Quine’s strategy of paraphrasing proper names as definite descriptions is immune to Kripke’s objections to descriptivism, as Kripke acknowkledges (Kripke 1980, 29). For example, it explains how we can apprehend propositions about infinite collections without requiring us to have acquaintance with an infinite number of things (see Russell 1903, 349). Kripke 1977; 1980. Frege 1892. It might be thought that Kaplan’s objection is better directed at Evans than Russell (Kaplan himself does not explicitly direct it at either, but simply takes it to show that no special epistemological access to an object is required in order to refer to it with an indexical). Evans effectively revises Russell’s principle to include the criterion that we must have discriminatory knowledge of an object in order to have a thought, or grasp a proposition, about it. This is clearly lacking in Kaplan’s example. But could Russell reply that the heiress is acquainted with the present moment but unaware of what time it is? Perhaps this sounds plausible for the temporal indexical “now,” but is she acquainted with the referent of “here”? Surely not, for her thought that “it is quiet here now” is about the location in which the car is parked, not the inside of the trunk in which she is secured, and it seems quite implausible to think that she is acquainted with that place. These manuscripts have now been published in Papers 5. Russell is famously rather sloppy in his demarcation of use and mention of symbols when employing this distinction, so this understanding does require us to be charitable with some of the things he says about incomplete symbols. I think most would agree with this interpretation of the concept, however. The theory was first published in the 1906 paper “On Some Difficulties in the Theory of Transfinite Numbers and Order Types,” which had been read before the London Mathematical Society in 1905. The theory (under the name ‘no-classes theory’) was tentatively suggested without full endorsement alongside alternative responses to the paradoxes in the paper, but by the time of publication, Russell

Russell on Denoting and Language

18

19

20 21

203

was prepared to add a note, dated February 5, 1906: “From further investigation I now feel hardly any doubt that the no-classes theory affords the complete solution of all the difficulties stated in the first section of this paper” (Russell 1906a, 164). However, he subsequently withdrew before publication a more elaborate statement of the theory titled “On the Substitutional Theory of Classes and Relations” after it had been accepted by the journal of the same society. He then published his most complete version of the theory in French, under the title “Les Paradoxes de la Logique,” later that year. This was later translated under the name “On ‘Insolubilia’ and their Solution by Symbolic Logic,” and published alongside the other two papers in Russell (1973). It does feature, however, in Russell (1908), although in such a cryptic form as to have gone more or less unnoticed for nearly seventy years. The recovery of the theory began with the publication of Russell (1973). However, it was the publication of Russell’s correspondence with Philip Jourdain in Grattan-Guinness (1977) that made the importance of the theory clear. Landini’s (1998) is the definitive study of the substitutional theory. Russell does allow matrices to be substituted for matrices in “higher-order” matrices. For example, q/(p/a);(r/s) is well-formed in the language. But here we understand the substitution as a “dual” substitution of r for p and s for a in p. Such matrices as seen here are used as the equivalents of both dyadic relational predicates, and classes of classes in the theory (see Russell 1906b, 176). See Neale (1990) for detailed elaboration of this point. I would like to thank Russell Wahl for helpful comments on an earlier draft.

References Donnellan, Keith. 1966. “Reference and Definite Descriptions.” Philosophical Review 75: 281–304. Evans, Gareth. 1982. The Varieties of Reference. Oxford: Oxford University Press. Frege, Gottlob. [1884] 1959. The Foundations of Arithmetic. Translated by J. L. Austin. Evanston, IL: North Western University Press. Frege, Gottlob. [1892] 1997. “On Sense and Reference.” In The Frege Reader, edited by M. Beaney. Oxford: Blackwell. Grattan-Guinness, Ivor. 1977. Dear Russell-Dear Jourdain. London: Duckworth. Kaplan, David. 1989. “Demonstratives.” In Themes from Kaplan, edited by Joseph Almog, John Perry, and Howard Wettstein, 481–563. Oxford: Oxford University Press. King, Jeffrey. 2007. The Nature and Structure of Content. Oxford: Oxford UniversityPress.

204

The Bloomsbury Companion to Bertrand Russell

Kripke, Saul. 1977. “Speaker’s Reference and Semantic Reference.” In Philosophical Troubles: Collected Papers Volume 1, edited by S. Kripke. Oxford: Oxford University Press. Kripke, Saul. 1980. Naming and Necessity. Oxford: Blackwell. Kripke, Saul. 2005. “Russell’s Notion of Scope.” In Philosophical Troubles: Collected Papers Volume 1, edited by S. Kripke. Oxford: Oxford University Press. Kripke, Saul. 2011. Philosophical Troubles: Collected Papers Volume 1. Oxford: Oxford University Press. Landini, Gregory. 1998. Russell’s Hidden Substitutional Theory. Oxford: Oxford University Press. Neale, Stephen. 1990. Descriptions. Cambridge, MA: MIT Press. Quine, Willard Van Orman. 1948. “On What There Is.” In From A Logical Point of View. edited by Quine, Cambridge, MA: Harvard University Press, 2nd ed., 1961. Russell, Bertrand. 1905a. “On Fundamentals,” in Papers 4. Russell, Bertrand. 1905b. “On Denoting,” in LK. Russell, Bertrand. 1906a. “On Some Difficulties in the Theory of Transfinite Numbers and Order Types,” in EA (reprinted in Papers 5). Russell, Bertrand. 1906b. “On the Substitutional Theory of Classes and Relations,” in EA(reprinted in Papers 5). Russell, Bertrand. 1906c. “On “Insolubilia” and their Solution by Symbolic Logic,” in EA(reprinted in Papers 5). Russell, Bertrand. 1918. “The Philosophy of Logical Atomism,” in LK. Russell, Bertrand. 1923. “Vagueness.” Australasian Journal of Philosophy and Psychology 1, 84–92, reprinted in Papers 9. Russell, Bertrand. 1992. Selected Letters of Bertrand Russell, Vol. 1. Edited by N. Griffin. London: Routledge. Salmon, Nathan. 1986. Frege’s Puzzle. Cambridge, MA: MIT Press. Stevens, Graham. 2004. “From Russell’s Paradox to the Theory of Judgement: Wittgenstein and Russell on the Unity of the Proposition.” Theoria 1: 28–61. Stevens, Graham. 2003. “The Truth and Nothing But the Truth, Yet Never the Whole Truth: Frege, Russell and the Analysis of Unities.” History and Philosophy of Logic 24: 221–40. Stevens, Graham. 2005. The Russellian Origins of Analytical Philosophy. London: Routledge. Stevens, Graham. 2008. “Russell and the Unity of the Proposition.” PhilosophyCompass 3 (3): 491–506. Stevens, Graham. 2011. The Theory of Descriptions. Basingstoke: Palgrave Macmillan. Strawson, Peter F. 1950. “On Referring.” Mind 59: 320–44, reprinted in Strawson,P. F., Logico-Linguistic Papers. London: Methuen, 1971.

8

Russell’s Logic as the Essence of Philosophy Gregory Landini

Abstract In Our Knowledge of the External World, Russell first advanced his thesis that logic is the essence of philosophy. This chapter identifies Russell’s logical atomism with this thesis and examines Russell’s conception of philosophy. It will be found that Our Knowledge is a natural sequel to Russell’s book The Problems of Philosophy. This chapter argues that in both books Russell holds that the nature and value of philosophy lies in its using our synthetic a priori knowledge of the logic of Principia Mathematica, together with empirical advances in physics, to investigate kinds of necessity and reveal that logical necessity is the only necessity. Philosophy is the study of necessity. Ontology is not the study of what there is; it is the study of what there has to be.

1 Introduction In Russell’s 1911 paper “Analytic Realism,” we find an attempt to articulate a research program in philosophy called “logical atomism.” This articulation awaited refinement and reached its mature state in Russell’s Our Knowledge of the External World as a field for Scientific Method in Philosophy (1914). The work heralds Russell’s thesis that logic is the essence of philosophy. In a chapter titled “Logic as the Essence of Philosophy,” the thesis is summarized as follows (OKEW, 35): Every philosophical problem, when it is subjected to the necessary analysis and purification, is found either to be not really philosophical at all, or else to be, in the sense in which we are using the word logical.

206

The Bloomsbury Companion to Bertrand Russell

To understand this thesis, we need to better understand Russell’s conception of philosophy. To what extent are empirical scientific discoveries relevant to philosophy for Russell? What did he mean by “logic” and by “scientific philosophy”? Why wasn’t the thesis advanced in Russell’s earlier 1912 book The Problems of Philosophy? In order to address these questions, let me demarcate a period of Russell’s works between 1910 and 1916 that form a unified period that I shall call the Principia era. The title is apt because Russell’s many books and articles in this time frame embody ideas and implications of Whitehead and Russell’s Principia Mathematica. Russell’s works include Principia (vol. 1, 1910; vol. 2, 1912; vol. 3. 1913), The Problems of Philosophy (1912), Theory of Knowledge (unfinished 1913), and Our Knowledge (1914). These works all hold that a fundamental relation of acquaintance that a subject can have to a universal is the foundation of our understanding logic and our knowledge of its synthetic a priori truths. Unfortunately, Whitehead’s fourth volume on geometry was delayed and delayed as he grappled with Einstein’s revolutionary ideas about space-time. It was slated for 1914, and Whitehead was still promising it as late as 1920. There are many controversies about Principia, and we cannot hope to engage them here. I hold that Principia set out a formal syntax involving the impredicative comprehension of attributes (properties and relation in intension) that I call “cp-logic” (comprehension principle logic).1 Principia’s formal language is that of simple-type scaffolded variables, with impredicative comprehension axiom schemata such as the following:

(∃f )(ϕx (t )

t

( ))

≡ t f (t ) x t , x

where f (t) is not free in the wff φxt.

(∃f

(t1 ,t 2 )

) ϕ ( x

t1

)

, y t2 ≡

xt1 , yt2

(

)

 f (t1 ,t2 ) x t1 , y t2  , 

where f (t1 ,t2 ) is not free in the wff ϕ(x t1 , y t2 ) . To say that comprehension is impredicative is just to say that the sorts of wffs ϕx t , ϕ(x t1 , y t2 ) involved may themselves contain bound predicate variables. The simple-type indices on the predicate variables are conveniently dropped in Principia. This requires that the genuine variables φ!, ψ! f!, g! and so on are made distinct from the schematic

Russell’s Logic as the Essence of Philosophy

207

letters φ, ψ, f, g and so on for wffs. This is done by adopting the exclamation. Thus, Principia, writes the above impredicative comprehension axiom schemas as follows:

∗12.1 (∃f )(ϕx ≡ f ! x ) x ∗12.11 (∃f )(ϕxy ≡ f ! xy ) xy It is precisely in virtue of impredicative comprehension that logic is a general synthetic a priori science. It was designed to support what Russell regarded as the Cantorian revolution in mathematics and what he regarded as Frege’s revolution in logic (i.e., the discovery of cp-logic).2 Whitehead and Russell’s informal description of its grammar, and the convention in Principia of dropping its type indices, opened the way for Carnap and Church to present the work as if it coded ramification into its syntax. Church imagines a Principia that adopts comprehension schemata3 set out with non-predicative predicates variables such as nφ(o), where n can be any numeral higher than the order (namely 1) of the simple type (o). A predicative variable is predicative when its order is the order of its simply type. Church then adds an axiom (not a schema) of reducibility, such as

( n ϕ o )(∃ ϕ o )( n ϕ o (xo ) ≡ xo ( )

1 ( )

( )

1

)

ϕ(o) (x o ) .

This avoids the deleterious consequences of the ramified theory. Church’s interpretation quickly became the orthodoxy. But nothing like this occurs in the actual Principia. What occurs is the formal system of simple-type theory cobbled together with an informal introduction by Whitehead and Russell that offers a semantic interpretation of the Principia’s bindable predicate variables— an interpretation based on a recursive and nominalistic definition of “truth.” The nominalistic semantics required imposes ramification into orders, and does not validate Principia’s impredicative comprehension axioms. Church’s approach endeavored to “fix” Principia, inventing a ramified syntax of non-predicative predicate variables that accords with a realist semantics that interprets predicate variables as ranging over ramified types of attributes (properties and relation) in intension. The differences are very important, but can be ignored for the purposes of this chapter.4 The agenda of Principia is to establish logicism—the thesis that all the branches of mathematics, including non-Euclidean geometries, study relational

208

The Bloomsbury Companion to Bertrand Russell

structures, independently of whether the relations in question are exemplified. Though they certainly thought it possible to achieve, no derivation of mathematics from a consistent axiomatization of cp-logic is part of Whitehead and Russell’s logicism.5 Nor is their logicism a thesis imposed on mathematics from outside. Logicism finds its foundation in what Russell regarded as a revolution coming from within mathematics initiated by Cantor and by Frege. Cantor’s work (which can be embellished by borrowing from the work of Dedekind, Weierstrass, Pieri, et al.) transformed mathematics into a science of relational order. Frege’s revolutionary idea that logic embodies the comprehension of functions makes it into an informative science capable of studying all relational structures. These, together with the revolutionary non-Euclidean geometries, are movements that reveal that all the branches of mathematics study relational structures. All the branches of mathematics get along perfectly well without the metaphysician’s assumption of abstract particulars—without numbers (natural, signed, rational, real, complex), without sets/classes, relations-in-extension, without propositions, without geometric spatial figures, or what have you. Whitehead and Russell’s logicism is the thesis that mathematicians are studying the logic of relations when they do mathematics (see Russell (1917)). Principia’s cp-logic has a privileged status and is independent of the civil wars between metaphysicans fighting over favorite kinds of necessity governing their special abstract particulars. Both cp-logic and its epistemology must retain this status if Russell’s scientific philosophy is to retain a lofty position that enables it to arbitrate between the metaphysical rivals who introduce kinds of necessity grounded in their special abstract particulars. In its effort to explain our knowledge and understanding of logic, the epistemology of the Principia era relied on Problems and its assumption that the foundation of synthetic a priori knowledge of cp-logic lies in a subject standing in an acquaintance relation to a universal. But Theory of Knowledge embellished its story of acquaintance by including acquaintance with abstract particulars that are logical forms (all, some, not, and, or). With such abstract particulars as logical forms in place, the epistemology for Principia was no longer above the metaphysical fray. Theory of Knowledge was abandoned. The Principia era ended. The era that followed it is Russell’s neutral monist era. Russell’s neural monist books of the 1920s abandon acquaintance and adopt sympathies with Watson’s behaviorism. Russell wrote books as replacements for the books that belong his Principia era. There is a nice match up: Problems is replaced by An Outline of Philosophy (1927), Our Knowledge is replaced by The

Russell’s Logic as the Essence of Philosophy

209

Analysis of Matter (1925), and the unfinished Theory of Knowledge is replaced by Russell’s The Analysis of Mind (1921). It was Russell’s sympathies with Watson’s behaviorism that brought him to believe that there are no relations of “acquaintance” and “sensation” between a subject and an object. That is, he used behaviorism to counter his argument from Theory of Knowledge that the “acquaintance” relation (or more exactly its converse “presentation”) is precisely what shows that neutral monism’s abandonment of the subject to be mistaken. The relation of “presentation” has a subject engaged in an act of selective attention enabling a this to be presented to it. The same point applies to the selective attention of the subject involved in the relation of “sensation.” In The Analysis of Mind, however, Russell imagines the contracting coil of a thermometer to be a “sensation,” and knowing becomes reacting “appropriately” to an environmental stimulus (AMi, 260, 254, respectively). We noted that Russell’s logical atomism was originally articulated in his 1911 paper “Analytic Realism” and reached its zenith in the thesis of Our Knowledge that logic (i.e., cp-logic) is the essence of philosophy. It belongs to the Principia era. But the 1918 Lectures on Logical Atomism depict Russell in transition between the Principia era and neutral monist era where his logical atomism takes on a quite different epistemology. The lectures on logical atomism belong to neither era. The shift to neutral monism was supposed to set things all right. The 1918 Lectures on Logical Atomism are tentative and often experimenting with the ideas Russell charitably attributes to Wittgenstein who was a student of Russell in 1912–14, whose well-being he was concerned about since he had left to fight with Austria during the First World War. Russell entertains abandoning acquaintance as a relation between a subject and an object (universal or particular). He imagines that, his arguments in Theory of Knowledge notwithstanding, behaviorist ideas might well convince him that indexicals (the presentation of this) in thought are not telling against neutral monism. Russell’s neutral monist monograph Logical Atomism (1924) replaces the experimental and often fumbling 1918 Logical Atomism Lectures. Curiously, in this later work, Russell retains the privileged status for logic, writing: I hold that logic is what is fundamental in philosophy, and that schools should be characterized rather by their logic than by metaphysic.

It is difficult to understand how Russell can maintain this position while embracing his new found behaviorist approach to our understanding of logic. In

210

The Bloomsbury Companion to Bertrand Russell

Outline, Russell’s behaviorist sympathies seem to infest logic itself. He imagined inference as a habit. He wrote (OOP, p. 86): When we “understand” a mathematical expression, that means we can react to it in an appropriate matter, in fact, that it has “meaning” for us. . . . When we come to algebra, and have to operate with x and y, there is a natural desire to know what x and y really are. That, at least was my feeling: I always thought the teacher knew what they really were, but would not tell me. To “understand” even the simplest formula in algebra, say (x + y)2 = x2 + 2xy + y2, is to be able to react to two sets of symbols in virtue of the form which they express, and to perceive that the form is the same in both cases. . . . We may sum this discussion by saying that mathematical inference consists in attaching the same reactions to two different groups of signs, whose meanings are fixed by convention in relation to their constituent parts, whereas induction consists, first in taking something as a sign of something else and later, when we have learned to take A as a sign of B, in taking A also as a sign of C.

We can only look upon this passage with dismay. Without understanding of cp-logic, I dare say that there could never be any conventions set out for anything, to say nothing of conventions governing the meaning of the expressions of elementary algebra. One might habituate a rat to pressing a lever marked “x2 + 2xy + y2,” upon electrical stimulation in the presence of an image marked “(x  +  y)2” but to proclaim that this stimulus-response cycle should be called “knowing” seems to little more than just an equivocation. Knowledge cannot be constituted by reacting to stimuli in accordance with the Law of Effect. Knowing is certainly not reacting “appropriately” to an environmental stimulus, if the only criteria for being appropriate is to survive long enough to sire viable offspring. The theory of the evolution by natural selection of cognitive protocols of behavior can have no bearing whatsoever on the content of mathematical logic, and thus it is difficult to see how it can have no bearing of nature of our knowledge of mathematical logic. Russell seemed to understand this well enough in Our Knowledge. In Problems and his unfinished Theory of Knowledge, he hoped to found knowing mathematical logic on acquaintance with universals. This is the right road to the solution of the problem.6 In what follows I will argue on behalf of the conception of philosophy Russell was articulating in Our Knowledge. The distinctive kind of criticism unique to Russell’s new scientific philosophy is made possible by cp-logic, which frees the mind from the prisons produced by dogmatisms parading as if they are necessities. My thesis is that Russell’s Our Knowledge should be included in the

Russell’s Logic as the Essence of Philosophy

211

Principia era in spite of the fact that the unfinished 1913 Theory of Knowledge was abandoned. It will be argued that in Problems Russell was on the cusp of holding the fundamental thesis that logic, that is, cp-logic, is the essence of philosophy. Russell’s thesis is that philosophy, when properly conceived, is the partly empirical, partly synthetic a priori, science that studies all the kinds of necessity that there are. Russell conception of a scientific philosophy requires that both cp-logic and our knowledge of it be privileged if cp-logic is to be the essence of philosophy. The only necessity is logical necessity. It is the task of Russell’s research program in philosophy to reveal just that.

2  Logic as the essence of philosophy: Problems In Problems Russell endeavored for the first time to offer an epistemology for the cp-logic of Principia. The sequel to Problems was originally to have been a large volume which was tentatively entitled Theory of Knowledge. It contained an in-depth study of the relation of acquaintance between a subject, which is mental and capable of selective acts of attention, and an object which may or may not be mental. The objects of acquaintance are universals, logical forms, and transient physical particulars (as opposed to material continuants persisting in time obeying the laws of physics). The part of Theory of Knowledge dedicated to acquaintance aimed at establishing a foundation of our synthetic a priori knowledge of mathematical logic. The second part of Theory of Knowledge was on matter. The second part became the third and fourth chapters of Our Knowledge and it was published as the sequel to Problems. It continued the discussion in Problems of the nature and value of philosophy and presented a construction of matter, space-time and continuity. Some aspects of the first parts of the Theory of Knowledge project concerning the relation of acquaintance were published separately as papers. The epistemology Russell planned for the cp-logic of Principia relied on a multiple-relation theory. It was first endorsed in Principia and then elaborated in some detail in Problems. The multiple-relation theory maintains that truth is necessarily correspondence between a belief-fact and one or more facts. Principia applied the multiple-relation theory only to the base case of its recursive definition of truth. Theory of Knowledge was to work out an extension of the multiple-relation theory to the kinds of general belief what we encounter when we believe, say, that all men all mortal, or when we believe that two plus two are

212

The Bloomsbury Companion to Bertrand Russell

four. It led Russell to the thesis that there is a special faculty of acquaintance with logical forms such as “all,” “not,” and “or” grounding our understanding of logic. Russell eventually confessed that he reached an impasse when it came to the difficulties of adapting his multiple-relation theory to the complications arising from considerations of “molecular” and general belief. What, after all, is before the mind when, for example, when we believe that (x)Fx v (∃y)Gy? This question is of utmost importance to Russell since his multiple-relation theory is supposed to be adequate to provide for our understanding of the cp-logic Principia. It is often thought that we can reliably date the abandonment of the Theory of Knowledge project as 1914, when Russell published some of what had been its opening chapters. But this cannot mark its abandonment since it may well have been Russell’s intent to reorganize the material and publish a self-standing book focusing on solving the problems of the multiple-relation theory. Moreover, a multiple-relation theory of judgment is advocated in Our Knowledge itself— though Russell doesn’t give any details. He writes (OKEW, 61): The case of judgment demands the admission of more complicated forms. If all judgments were true, we might suppose that judgment consisted in apprehension of a fact, and that the apprehension was a relation of a mind to a fact. From poverty in the logical inventory, this view has often been held. But it leads to absolutely insoluble difficulties in the case of error. . . . It is therefore necessary, in analyzing belief, to look for some other logical form than a two-term relation. Failure to realize this necessity has, in my opinion, vitiated almost everything that has hitherto been written on the theory of knowledge, making error insoluble and the difference between belief and perception inexplicable.

It is difficult, therefore, to say precisely when Russell abandoned the multiplerelation theory and came to believe that acquaintance with logical forms is not the way to explain our understanding of Principia’s cp-logic. In adopting neutral monism, however, Russell abandoned the relation of “acquaintance” and the relation of “sensation.” These relations require a subject as one of the relata and neutral monism rejects the subject. Since the multiple-relation theory requires the theory of acquaintance, we can be sure that 1918 marks its abandonment (see Russell (1919a)). It is safe to post 1918 as the date of the abandonment of Russell’s theory that there is a relation of acquaintance with universals. In fact, Russell himself gives this date.7 With Theory of Knowledge shelved in 1914, Our Knowledge belongs as the rightful sequel to Problems. It is a nice sequel, much better than Theory of

Russell’s Logic as the Essence of Philosophy

213

Knowledge would have been. It is important not to think of it as some sort of change of heart for Russell. The chapters of Our Knowledge sketching a construction of matter (as continuants persisting through time) in terms of physical transient particulars available for sensation comprise just two of the eight chapters of the work. The more descriptive title is Scientific Method in Philosophy, which was on the cover of the original 1914 publication. The book is devoted to illustrations of Russell’s applications of the new logical forms of cp-logic to continuity, the Zeno paradoxes, infinity, the Cantorian and Fregean revolutions in mathematical logic, and notions of cause, change, and even free will. Most of the book consists in chapters devoted to revealing how the new cp-logic avoids the infamous muddles of metaphysicans, showing that discovering and applying the right logical form is the key feature of scientific philosophy. Indeed, Our Knowledge is a rather seamless sequel since a great many theses of Problems appear wholly intact in the work. This includes Russell’s discussion of sense-data. The only important difference is that in Our Knowledge Russell follows Whitehead’s suggestion to offer a construction of matter (i. e, a physical thing obeying the laws of physics) as a series of transient physical particulars some of which are data to sense (sense-data) and others of which are not (sensibilia). In Problems, Russell had imagined that we hypothesize the existence of matter as the best explanation of the cause of sense-data. It is quite unfortunate that after converting to neutral monism in 1918 Russell came to revise Our Knowledge twice. He removed passages and added passages to make the work appear as though it is compatible with his neutral monism. This inappropriately distances Our Knowledge from Problems. The relation of “sensation,” discussed in Our Knowledge, is regarded as mental though the sensedata that is its object may well not be. Russell writes (OKEW, 1914): A patch of colour, even if it only exists when it is seen, is still something quite different from the seeing of it: the seeing of it is mental, but the patch of colour is not.

This quote is from the first edition of 1914. It does not appear in the 1929 edition or in later editions which Russell revised to try to make consistent with holding his neutral monist thesis that there is no relation of sensation (or acquaintance) at all. But Russell’s original position in Our Knowledge is exactly the same as in Problems. Russell writes (POP, 41): Berkeley’s view, that obviously the color must be in the mind, seems to depend for its plausibility upon confusing the thing apprehended with the act of

214

The Bloomsbury Companion to Bertrand Russell

apprehension. . . . Berkeley’s argument is seen to be wrong in substance as well as in form, and his grounds for supposing “ideas”—i.e., objects apprehended— must be mental, are found to have no validity whatever. It will be seen that among the objects with which we are acquainted are not included physical objects (as opposed to sense-data), nor other peoples’s minds. These things are known by what I call “knowledge by description.”

The last part of the above is quite revealing if we recall that Russell held a fourdimensionalist philosophy of time. The transitory physical particular (sensedatum) that is the object of sensation is not to be conflated with physical continuants such as the sun or a table or a tree that are studied in physics. We cannot sense physical continuants, but physical transient particulars can be objects of the relation of sensation. Problems allows that a patch of color (i.e., the physical sense-datum that is surface reflectance) is the object of a mental act of sensation. In both works, Russell is in earnest to refute Berkeley’s argument that whatever stands in a relation of sensation (acquaintance, perception) to a subject must be mental. The key to the refutation is that “sensation” and “acquaintance” are relations involving a subject, which is mental, and an object which may or may not be mental. Indeed, the selective attentiveness of a subject in acquaintance, and the indexical (“this”) that seems to reflect this selective attention is given in Theory of knowledge as the most substantive reason for rejecting of neutral monism.8 In Problems, Russell tells us that what makes mathematical logic epistemically accessible is not a Rationalist faculty of intuition connecting the mind to a realm of abstract particulars each sort of which has its own kind of ontological necessity. What makes the synthetic a priori accessible is a faculty for an acquaintance with universals and the perception of relations between universals—an acquaintance that occurs even when nothing at all exemplifies the universals. He writes (POP, 106): Thus, the difference between an a priori general proposition and an empirical generalization does not come in the meaning of the proposition; it comes in the nature of the evidence for it. In the empirical case, the evidence consists in the particular instances.

Russell contrasts the empirical generalization “all men are mortal” with the mathematical truth “two plus two are four.” The difference, of course, cannot be merely epistemic. He means to say that while both involve relations between universals, the difference lies in that the former requires the contingent exemplification of the universals for it to be true, while the latter is true even

Russell’s Logic as the Essence of Philosophy

215

if no property whatever is exemplified by exactly two entities. The ground of the mathematical truth is the existence of a relation between the universal “two” and the universal “four” that obtains even if these universals are not exemplified. It takes a good deal of insight to reveal, for a given true statement of mathematics, what relation of universals holds independently of whether the universals are exemplified. Of course, in Principia the expression of “two plus two are four” is couched in a “no-classes” theory. Russell did not explain the no-classes theory in Problems, and he endeavored to express the point in a very simple way, assuming for the sake of discussion that we are acquainted with the universals “two” and “four.” Russell’s point, however, is that we can directly perceive the relationship between the universals independently of whether they are exemplified. To glimpse this, let us express 2+2 = 4 as follows:

(

4 x θ! x ≡θ 2 y ϕ! yϕ+,ψ 2 y ψ ! y

) x θ! x .

The upshot is that we can break up any group of four in different ways into two groups each containing exactly two. More formally it says: For all properties θ!, there are exactly four entities that exemplify θ! if and only if there are properties φ! and ψ! such that there are exactly two entities that exemplify φ! and exactly two other entities that exemplify ψ!, and the entities that are θ! are in one-to-one correspondence with entities that either exemplify φ! or exemplify ψ!. It is a mouthful to say. Can we directly perceive it? No. But we can perceive an instance of it. We have only to notice pictorial representation (below fig. 1) to see one way to break a group of four into two independent groups of two. It should be noted that cp-logic is required for this to hold since, for any θ! exemplified by exactly four entities, it is comprehension that assures that there are universals φ! and ψ! each exemplified by exactly two entities and such that no φ!’s are ψ!’s. Without the comprehension of universals afforded by cp-logic, “two plus two are four” would not be a relation between universals. φ!

ψ!

θ!

Figure 1

216

The Bloomsbury Companion to Bertrand Russell

Russell’s conclusion is that all a priori knowledge deals exclusively with relations of universals. He explains (POP, 103): Let us revert to the proposition “two and two are four.” It is fairly obvious, in view of what has been said, that this proposition states a relation between the universal “two” and the universal “four.” This suggests a proposition which we shall now endeavor to establish: namely, All a priori knowledge deals exclusively with the relations of universals. This proposition is of great importance, and goes a long way toward solving our previous difficulties concerning a priori knowledge.

Russell goes on to say (POP, 105): Thus the statement “two and two are four” deals exclusively with universals, and therefore may be known by anybody who is acquainted with the universals concerned and can perceive the relation between them which the statement asserts. It must be taken as a fact, discovered by reflecting on our knowledge, that we have the power of sometimes perceiving such relations between universals and there of sometimes knowing general a priori propositions such as those of arithmetic and logic

As we noted, the metaphysical difference between “all men are mortal” and “two plus two are four” lies in the fact that the former’s truth is not grounded by a relation between the universal “human” and the universal “mortal” that holds independently of whether they are exemplified—independently of whether there are mortal men. In contrast, the latter relation holds between the universals “two” and “four” independently of whether these universals are exemplified (i.e., there needn’t be any properties that are exemplified by exactly two entities or by exactly four entities). As we case see, the synthetic a priori nature of Principia’s cp-logic is essential to Russell’s account, for this alone assures us of the existence of the universals in question. Those who read Problems without any inkling of cp-logic will readily fall into misunderstanding. Worse, such a misunderstanding militates against understanding the import of what Russell had to say about the nature and value of philosophy. The nature of philosophy as a distinctive science lies in the distinctive kind of criticism it affords and that criticism comes solely from cp-logic. When this is realized, we can see that in Problems Russell was very close to the thesis in Our Knowledge, that logic (i.e., cp-logic) is the essence of philosophy. To reveal this, we have only to examine why it is that we don’t find Russell espousing the thesis that logic is the essence of philosophy in Problems. The

Russell’s Logic as the Essence of Philosophy

217

answer lies, somehow, with his account of the synthetic a priori as involving relations between universals that hold independently of whether the universals are exemplified. To flush out the answer, let us turn to Russell’s discussion of the value of philosophy. The Russellian conception of philosophy as the study of necessity explains the wonderfully inspiring passages of Problems concerning the way philosophy, when properly conducted, is liberating. It frees the mind, as Russell put it, from a kind of prison.9 What is this prison? In Problems, Russell makes this point rather clear but to understand his examples readers have to know something about the role the new cp-logic played in undermining the bad indispensability arguments so commonly found among the metaphysicians. The prison is this: Dogmatisms of one’s culture, religious or otherwise, insidiously parade themselves as if they were necessities. It is these pseudo-necessities that imprison the mind. Social institutions of government, religion, and so on have to respond to the crisis of finding legitimation. The way to legitimate is to appeal to this or that as an essence of something which grounds a necessity. In promoting and legitimating a given social organization, women may be stereotyped as essentially emotive, nonrational, domestic, care givers; in contrast, men may be stereotyped as essentially coldly rational, aggressive, unfaithful, and so on. It is a matter of biological essences, say Aristotelian essentialists. These stereotyped essences are then used to try to legitimate social practices and institutions of subjugation. For example, children in divorce must be in the custody of the mother, education of women in mathematics is futile, and so on. These are simply dogmatisms parading, as if necessities. In the quest to legitimate social institutions (oppressive or otherwise), the existence of nonlogical kinds of essences, entelechies, teleological processes offers a bountiful resource. The earth is essentially at rest, animals essentially function in a food chain and a hierarchy of teleological processes contributing to life on earth, and so on. Obviously, there is no chance that one could give a cogent argument that any of these kinds of necessities are logical necessities. Thus, the walls of the prison that entraps a mind that believes in such kinds of necessities are instantly removed by Russell’s thesis that the only necessity is logical necessity. Russell’s eloquent statement in Problems of the value of philosophy reveals that he envisions just this situation. We should not be put off by his remark: “Thus utility does not belong to philosophy” (POP, 153). Indeed, the word “thus” is a misprint of later editions and was originally the word “this,” referring to the kinds of instrumental value found in the medical and technical achievements of

218

The Bloomsbury Companion to Bertrand Russell

physical empirical science.10 Philosophy has a quite unique utility. Let us offer a few revealing quotes. We find (POP, 157): The value of philosophy is, in fact, to be sought largely by its very uncertainty. The man who has no tincture of philosophy goes through life imprisoned in the prejudices derived from common sense, from the habitual beliefs of his age or nation, and from convictions which have grown up in his mind without the co-operation or consent of his deliberate reason. . . . Thus while diminishing our feeling of certainty as to what things are, it greatly increases our knowledge as to what they may be; it removes the somewhat arrogant dogmatisms of those who have never travelled into the region of liberating doubt, and it keeps alive our sense of wonder by showing familiar things in an unfamiliar aspect.

The “uncertainty” arises because a claim of a kind of necessity has been removed. Among the “habitual beliefs” and “prejudices derived from common sense” Russell clearly includes the indoctrinations of religious authority, for example, Aristotelian essences and entelechies (natural states and places) grounding the religious proclamation that Copernican heliocentrism is a challenge to the authority of the Holy See in matters of metaphysics (ethical and physical)—a doctrine whose denial brought Galileo under the Inquisition. Russell’s point is clear: Philosophy is distinctive in its ability to remove dogmatisms and promote “liberating doubt.” Philosophy has great utility in, as Russell put it, “showing unsuspected possibilities” (POP, 157). It shows them because it analyzes away nonlogical pseudo-necessities. The possibilities are limitless (as it were)—for they are “limited” only by logic. Now Russell thinks that, for the most part, the dogmatisms parading as necessities are products of the effort to tailor a metaphysics suited to human desires, aspirations, the hope of justice, of good prevailing over bad, the fear of death and meaninglessness. Let me quote at length: Philosophical contemplation does not, in its widest survey, divide the universe into two hostile camps—friends and foes, helpful and hostile, good and bad—it views the whole impartially. Philosophic contemplation, when it is unalloyed, does not aim at proving the rest of the universe is akin to man. (POP, 158) There is a widespread philosophical tendency towards the view which tells us that Man is the measure of all things, that truth is man-made, that space and time and the world of universals are properties of the mind, and that, if there is anything non created by the mind, it is unknowable and of no account to us. This view, if our previous discussion were correct, is untrue; but in addition to being

Russell’s Logic as the Essence of Philosophy

219

untrue, it has the effect of robbing philosophical contemplation of all that gives it value, since it fetters contemplation to the Self. (POP, 159) By thus making a barrier between subject and object, such personal and private things become a prison to the intellect. The free intellect will see as God might see, without a here and now, without hopes and fears, without the trammels of customary beliefs and traditional prejudices, calmly, dispassionately, in the sole and exclusive desire of knowledge. (POP, 160)

The free man’s worship is a worship of knowledge for knowledge’s sake—a kind of objectivity given by a contemplation of the world sub specie aeternitatis. If the only necessity is logical necessity, a prison that fetters the mind is removed. This brings us, however, to the important question of what, in Russell’s view, is the tool for removing the prison opening the mind so that the only “limit” is the limits of logic. We needn’t agree with Russell insinuation in Problems that the only source of the metaphysical Prisons—the source of the belief in nonlogical necessities—is the anthropic instinct to tailor a metaphysics to one’s aspirations. The source of the belief in nonlogical necessities may well be, in some cases, as intellectually honest and contemplative an application of inquiry as any other. We shall find that Russell came to see this more clearly in Our Knowledge. The problem is not merely the influence of instincts which favor a metaphysic tailored to human aspirations and fears, the problem is also a failure to understand (imagine, conceive of) and apply relevant logical forms. Without the logical forms that become known to us through cp-logic, we are prone to fall into nonlogical necessities, metaphysical ontologies, and the muddles they often bring. “Existence” becomes a property, ontological arguments seem valid, abstract particulars that are numbers, spatial figures, points, classes, sets, mereological sums, gunk, Meinongian golden mountains and round-squares and the Homeric gods seem indispensable. The tool for removing the prisons, according to Russell, is Principia’s cp-logic. Of course, this is precisely the answer we expect from Our Knowledge, since in that work he explicitly maintained that logic is the essence of philosophy. But it is quite important to realize that the same position is to be found in Problems. Logic, that is the new cp-logic of relations exemplified in Principia, is the liberating tool. We had better offer a quote to establish this result. We find (POP, 147): Most of the great ambitious attempts of metaphysicians have proceeded by the attempt to prove that such and such apparent features of the actual world were self-contradictory, and therefore could not be real. The whole tendency of

220

The Bloomsbury Companion to Bertrand Russell

modern thought, however, is more and more in the direction of showing that the supposed contradictions were illusory. . . . Formerly it appeared that experience left only one kind of space to logic, and logic showed this one kind to be impossible. Now logic presents many kinds of space as possible apart from experience, and experience only partly decides between them. Thus, while our knowledge of what is has become less that it was formerly supposed to be, our knowledge of what may be is enormously increased. Instead of being shut in within narrow walls, of which every nook and cranny could be explored, we find ourselves in an open world of free possibilities, where much remains unknown because there is so much to know. What happened in the case of space and time has happened, to some extent, in other directions as well. The attempt to prescribe to the universe by means of a priori principles as broken down; logic, instead of being, as formerly, the bar to possibilities, has become the great liberator of the imagination, presenting innumerable alternatives which are closed to unreflective common sense.

These passages make it clear that in Russell’s view a new logic is afoot and it is this new logic that is playing the central role in liberating the mind from the prisons produced by dogmatisms. It is the new mathematical logic that is the driving force of Russell new conception of philosophy. The new logic is the essence of the “criticism” that Russell says, even in Problems, is “the essential characteristic of philosophy” He writes (POP, 149): Philosophical knowledge, if what has been said above is true, does not differ essentially from scientific knowledge; there is no special source of wisdom which is open to philosophy but not to science, and the results obtained by philosophy are not radically different from those obtained from science. The essential characteristic of philosophy, which makes it a study distinct from science, is criticism.

The close of this passage may seem difficult to understand. Russell uses the word “science” in two ways in the last two sentences of the quote. Empiricism notwithstanding, Russell holds that not all science is a posteriori (empirical science). He maintains that mathematics is an a priori science. Indeed, Russell’s logicism regards it as synthetic a priori. Russell holds that the new mathematical logic of Principia is a synthetic a priori science. If the “criticism” that is in Russell’s view the essential characteristic of philosophy embodies the new mathematical logic, we can see precisely why he regards philosophy as a genuine,

Russell’s Logic as the Essence of Philosophy

221

partly a priori, science. The results of philosophy do not differ essentially from scientific knowledge because they involve both synthetic a priori knowledge (coming from cp-logic) and empirical scientific knowledge. In short, Problems maintains that what makes philosophy different from empirical science is its use of cp-logic to uncover the logical forms which unravel the muddles that have led metaphysicians to promote the existence of nonlogical necessities. Now if Problems maintains that the tool of “criticism” that is distinctive of philosophy is the application of the forms studied by cp-logic, then Russell is on the cusp of the thesis, advanced explicitly in Our Knowledge, that logic (the new synthetic a priori mathematical cp-logic) is the essence of philosophy. But why does it not appear in Problems itself? The answer is clear if we know where in Problems to look for it. But we cannot know where to look without asking the following key question: Are all synthetic a priori propositions those of cp-logic? If Russell’s answer were “yes” then we would expect Problems to advance the thesis that logic is the essence of philosophy. Hence, Russell answer in Problems must have been “no.” But we have yet to discover the cases in particular. In Problems, Russell unequivocally maintains that some fundamentals of probability theory, namely the inductive principle, are a priori and part of logic. He writes (POP, 73): In addition to the logical principles which enable us to prove from a given premise that something is certainly true, there are other logical principles which enable us to prove, for a given premise, that there is a greater or less probability that something is true. An example of such principles—perhaps the most important example—is the inductive principle.

Russell says that the inductive principle is an a priori principle of logic in other passages as well (See POP, 67, 112, 149). It is clear that probability theory does not give us an example of something Russell thought in Problems to be synthetic a priori but which is not part of the new cp-logic. Happily Russell does give an explicit example of a priori knowledge from a field he thinks is outside of logic. In Problems, Russell allows that there may well be synthetic a priori knowledge in Ethics. He writes (POP, 75–77): A priori knowledge is not all of the logical kind we have hitherto been considering. Perhaps the most important example of non-logical a priori knowledge is knowledge of ethical value. . . . I am speaking of judgments as to the intrinsic desirability [good] of things.

222

The Bloomsbury Companion to Bertrand Russell

Knowledge as to what is intrinsically valuable is a priori in the same sense in which logic is a priori, namely in the sense that the truth of such knowledge can neither be proved nor disproved by experience. All pure mathematics is a priori like logic.

A priori knowledge is what Russell calls “self-evident.” Such self-evidence, he goes on to say comes in degrees. But he writes (POP, 112): It would seem also, though this is more disputable, that there are some selfevident ethical principles, such as “we ought to pursue what is good.”

Now if some fundamental principles of ethics are synthetic a priori and yet not part of cp-logic then there is a special kind of necessity unique to ethics. And philosophy, understood as the study of the nature of all the purported kinds of necessity, must certainly include ethics. Logic would not be the essence of philosophy if ethics is part of philosophy. Ethics is an outlier. But this raises a very difficult position for Russell to take. It raises the question for Russell’s of where Ethics fits into his distinction in Problems between truths knowable a priori such as “Two plus two are four” and truths such as “All men are mortal.” According to Problems, all a priori knowledge concerns the perception of relations between universals. If some of our a priori knowledge includes ethical principles, then these ethical principles must concern relations between universals. Why then does Russell exclude them from cp-logic? There is only one answer available to him. The relations between universals grounding cp-logical truths do not require exemplification, but the relations between universals involved with ethics do require exemplification. But if this is correct, how does Russell distinguish the ethical truth such as “All pain is intrinsically bad” from a contingent generality such as “All men are mortal,” which surely can only be known a posteriori? Perhaps Russell’s answer is that the ethical truth that pain is intrinsically bad, requires that the universal “pain” (being in pain) be exemplified. On this view, sensory/qualitative universals exist only if exemplified. Moreover, unlike the universals “human” and “mortality,” in order to become acquainted with the universals “pain” and “intrinsic badness” one has to experience them instantiated in states directly in oneself. This is very special sort universal which requires for its very existence a first person exemplification. Russell never says such things in Problems. But it would seem that it is the only option left to him—given his account of a priori knowledge. This approach would also explain what Russell had in mind in admitting tentatively that we have an immediate a priori perception

Russell’s Logic as the Essence of Philosophy

223

of a greater similarity between two (phenomenal) shades of green than a shade of red and a shade of green (POP, 103). If so, it would give yet another case of a priori knowledge outside of logic. It is outside of logic because the universals must be exemplified in order for the relation between them to hold. It is nonetheless known a priori because of the very special nature of the sort of first person exemplification of the distinctively qualitative universals involved. It is interesting, however, that Russell leaves the color question open. In fact, he says that ethics is the “most important example of nonlogical a priori knowledge.” This comment suggests the trajectory he was on. After all, any nonlogical a priori knowledge is ipso facto important. Russell is talking about synthetic a priori knowledge that is outside of the new cp-logic. So the comment strongly suggests that Russell is entertaining the notion that ethics is the only significantly worrisome outlier he knows of, though he admits that there might be others.

3  Logic as the essence of philosophy: Our Knowledge Our Knowledge is a very natural sequel to Problems. The two belong together and should be read together. In Our Knowledge we find an engaging continuation of Russell’s discussion in Problems concerning the nature and value of philosophy. He writes (OKEW, 261): When everything has been done that can be done by method, a stage is reached where only direct philosophic vision can carry the matter further. Here only genius will avail. What is wanted, as a rule, is some new effort of logical investigation, some glimpse of possibility never conceived before, and then the direct perception that this possibility is realized in the case in question. Failure to think of the right possibility leaves insoluble difficulties, balanced arguments pro and con, utter bewilderment and despair. But the right possibility, as a rule, when once conceived, justifies itself swiftly by its astonishing power of absorbing apparently conflicting facts. From this point onward, the work of the philosopher is synthetic.

Russell makes comments very similar to those in Problems concerning the characteristic feature of philosophy being a unique kind of criticism, and he emphasizes how the logical forms of the new logic open the mind to innumerable new possibilities (OKEW, 259). Thus, we can see that it is quite natural to regard Our Knowledge as beginning precisely where Problems left off. But now, of course, Russell heralds his thesis that logic is the essence of philosophy. The status

224

The Bloomsbury Companion to Bertrand Russell

of ethics, which was holding him back in Problems, has been resolved. Either Ethics has synthetic a priori principles that are part of logic (as is the case of mathematics and probability), or it consists entirely of empirical matters and has no proper place in philosophy. Russell rejects the first disjunct. In order to see this, we must first show that Our Knowledge does not differ from Problems concerning the a priori status of the foundations of probability. In Our Knowledge, Russell is sometimes a bit cagey in asserting that the foundations of probability are, in fact, known. Having set out a principle of induction, he writes (OKEW, 39): Hence, if it is known, it is not known by experience, but independently of experience. I do not say that any such principle is known: I only say that it is required to justify the inferences from experience which empiricist allow, and that it cannot itself be justified empirically. A similar conclusion can be proved by similar arguments concerning any other logical principle. Thus logical knowledge is not drivable from experience alone, and the empiricist’s philosophy can therefore not be accepted in its entirety, in spite of its excellence in many matters which lie outside logic.

He continues this hypothetical later in the work. He writes (OKEW, 241): The principle involved is the principle of induction, which, if it is true, must be an a priori logical law, not capable of being proved or disproved by experience. . . . It may well be question whether this proposition is true; but if we admit it, we can infer that any characteristic of the whole of the observed past is likely to apply to the future and to the unobserved past. This proposition, therefore, if it is true, will warrant the inference that causal laws probably hold at all times, future as well as past; but without this principle, the observed cases of the truth of causal laws afford no presumption as to the unobserved cases, and therefore the existence of a thing not directly observed can never be validly inferred.

But in other passages he is more confirmed in his belief the inductive principle is known. He writes (OKEW, 60): The above conclusion, of which we had an instance in the case of the inductive principle, is important, since it affords a refutation of the older empiricists. . . . We must therefore admit that there is a general knowledge not derived from sense, and that some of this knowledge is not obtained by inference but is primitive. Such general knowledge is to be found in logic. Whether there is any such knowledge not derived from logic, I do not know; but in logic, at any rate, we have such knowledge.

Russell’s Logic as the Essence of Philosophy

225

The status of the inductive principle cannot very well afford a refutation of empiricism in Russell’s view if he thought that it is not a case of knowledge (synthetic a priori). No one can reasonably deny that the principle of induction is a synthetic principle. Thus, we have every good reason to hold that Russell agrees with his stand in Problems that the inductive principle is a part of a logic and knowable a priori. It seems fair to conclude that Russell is consistent and indeed that he is simply agreeing with his thesis in Problems that there are degrees of self-evidence and that even in dealing with something that is, in fact, a logical truth known a priori, our confidence in it comes in degrees.11 He writes (POP, 117): The inductive principle has less self-evidence than some other of the principles of logic, such as “what follows from a true premise must be true.” . . . The truths of logic and mathematics have (broadly speaking) less self-evidence as they become more complicated. Judgments of intrinsic ethical or aesthetic value are apt to have some self-evidence, but not much.

Russell’s own degree of self-evidence assigned to the principle of induction is perhaps less than the degree he has assigned to some other logical principles. But complete certainty is not required for knowledge—not even for a priori knowledge of the laws of cp-logic. Having established that Russell’s views on probability in Our Knowledge are the same as those of Problems, we are in positon to see that it was only just his position on Ethics that changed. In an unpublished manuscript “On Matter,” Russell’s still seems to have held that ethics is altogether outside of philosophy.12 In a passage that he ended up deleting, there is a long discussion of a priori principles. Russell remarks: “It is questionable however whether our a priori knowledge is wholly logical. Ethical knowledge, to begin with, if it exists, must be in part a priori, and is certainly not logical” (Papers 6, 514). But clearly Russell was on the cusp of changing his mind. Interestingly, Russell comments in Problems concerning prisons are very similar to what is found in Our Knowledge. Part of what he says ties in nicely to his comments in Problems concerning a free man’s contemplation of the universe unfettered to any anthropic self-interest motivating one to manufacture a metaphysics to assuage one’s fears of meaninglessness of life and one’s hope for one’s conception of good to prevail over evil. We find: “The desire to establish this or that result or generally to discover for agreeable results, of whatever kind, has of course been the chief obstacle to honest philosophizing”

226

The Bloomsbury Companion to Bertrand Russell

(OKEW, 258). It is in this context that he critiques Bergson’s work, writing as follows (OKEW, 28): The hope of satisfaction to our more human desires—the hope of demonstrating that this world has this or that desirable ethical characteristic—is not one which, so far as I can see, philosophy can do anything whatever to satisfy. The difference between a good world and a bad one is a difference in the particular characteristics of the particular things that exist in these worlds; it is not a sufficiently abstract difference to come within the province of philosophy. Love and hate, for example, are ethical opposites, but to philosophy they are closely analogous attitudes toward objects. The general form and structure of those attitudes toward objects which constitute mental phenomena is a problem for philosophy; but the difference between love and hate is not a difference of form or structure, and therefore belongs rather to the special science of psychology than to philosophy.

In Problems Russell thought that some parts of Ethics are synthetic a priori and not part of logic. In Our Knowledge he maintains that no ethical proposition is known a priori. What, we may naturally ask, prompted the change concerning Ethics? The wrong answer, I believe, is to think that Russell came to be influenced by Wittgenstein’s thesis that logic consists of tautologies and generalized tautologies. (An embryonic form of this thesis was communicated to Russell in letters of 1913 from Norway). Such an interpretation would have Russell abandoning Principia’s cp-logic. Instances of the comprehension axiom schemas for cp-logic cannot be construed as tautologies or generalized tautologies. It is the precisely the new logical forms made available in Principia’s cp-logic that are essential for the study of time, change, continuity and the solution of Zeno’s paradoxes that Russell sets forth in Our Knowledge. It is cp-logic that is the key component of the kind of criticism Russell thinks distinctive of his new scientific philosophy. Our Knowledge does contain two enthusiastic allusions to Wittgenstein’s work in progress. The more important is that Russell writes (OKEW, 208): If the theory that classes are merely symbolic is accepted, it follows that numbers are not actual entities, but that propositions in which numbers verbally occur have not really any constituents corresponding to numbers, but only a certain logical form which is a part of propositions having this form. This is in fact the case with all the apparent objects of logic and mathematics. Such words as or, not, if, there is, identity, greater, plus, nothing, everything, function, and so on,

Russell’s Logic as the Essence of Philosophy

227

are not names of definite objects. . . . All of them are formal, that is to say, their occurrence indicates a certain form of proposition, not a certain constituent. “Logical constants,” in short, are not entities; the words expressing them are not names, and cannot significantly be made into logical subjects. . . . This fact has a very important bearing on all logic and philosophy, since it shows how they differ from the special sciences.

The most interesting feature of this passage is that what is on display it is almost exclusively a list of constructions of Principia, not a list from Wittgenstein’s work in progress. Principia took an eliminativistic stance not only against classes but also against or, not, if, there is (i.e., some), greater, plus, nothing, everything and function. Russell had long abandoned an ontology of functions in favor of manyone relations. And in abandoning his pre-Principia ontology of propositions in 1908, Russell had rid himself of his view that logical particles stand for properties and relations.13 The only clear addition to Russell’s list belonging to Wittgenstein is that of identity. To be sure, that is a substantive addition—if Wittgenstein could only have made it workable. Looking back, Russell recalls in My Philosophical Development that for a short time he had been captivated by Wittgenstein’s concern to eliminate identity but he came to see that the implications of the elimination make mathematical logic impossible (MPD, 115ff). Wittgenstein himself felt that the elimination of identity requires that Principia be “done afresh” by appealing to recursive operations characterized by using signs with numeral exponents.14 Ramsey, who had initially also been enthusiastic, came to agree with Russell and concluded that the elimination of identity undermines even elementary arithmetic since one cannot emulate quantification over natural numbers (Ramsey, 1931, 17). In light of this, there is no good reason to hold that Russell’s positions in Our Knowledge were due to Wittgenstein’s views about logic consisting of tautologies and generalized tautologies. Russell’s new stand against Ethics as (synthetic a priori) comes from another quarter.15 We have seen that in Problems Russell has a difficulty which arises from his thesis that all synthetic a priori knowledge concerns relations between universals—universals that need not be exemplified. The difficulty was how to exclude certain synthetic truths of ethics, which he takes to be known a priori, from the province of logic. Thus, to see the transition to Our Knowledge, we need only imagine Russell reviewing the difficulty and maintaining that ethics is known a posteriori after all. In short, Russell has taken on a new thesis, namely this: All necessity is structural.

228

The Bloomsbury Companion to Bertrand Russell

This is not the Tarskian thesis the logical truths are true in virtue of their structure. It is the thesis that logical truths are synthetic and yet logically necessary truths about structural relationships—truths that can be expressed by quantification alone. Now anything concerning structure is amenable to study within cp-logic, which, as we have seen, is the very thesis of Principia. Russell’s logicism regards cp-Logic as the general synthetic a priori science that studies all the kinds of structures that there are—structures realized by relations (whether they are exemplified or not) ordering their fields. In Our Knowledge, Russell holds that considerations of intrinsic value and other allegedly synthetic a priori issues in ethics are, if true, not truths about kinds of structure. But if they are not truths about structure, then they cannot, after all, be necessary truths and thus they are not in the province of his new scientific philosophy understood as the study of necessity.

4  Reconstruction in philosophy The essence of Russell’s philosophy is its cp-logic, and thus his reconstruction is not offering philosophy as wholly empirical science. In Our Knowledge, he explained “Philosophy does not become scientific by making use of other sciences, in the kind of way in which (e.g.) Herbert Spencer does” (OK, 256). The evolutionary approach of Spencer was followed by Dewey’s The Influence of Darwin on Philosophy (1910) and it led to his Reconstruction in Philosophy (1920). This offers a striking contrast. Dewey thinks that the model of natural selection of social behaviors legitimates his reconstruction of philosophy. The interaction of organism and environment resulting in some adaptation is the primary tool of his explanations. Knowledge is not the acquisition of truth for Dewey. It is a process of engagement with an environment whereby the senses lose their place as gateways to knowing and become mechanisms to trigger behavior protocols for engagement. The entire dispute between empiricists and rationalists is, thereby, supposed to be dispatched (op cit., 83). “Intelligence,” Dewey tells us, “is not something possessed once and for all. It is in constant process of forming, and its retention requires constant alertness in observing consequences, an open-minded will to learn and courage in re-adjustment” (op. cit., 90). Dewey thinks that the origin of philosophy grows within “a background of an authoritative tradition; a tradition originally dictated by man’s imagination working under the influence of love and hate and in the interest of emotional excitement and satisfaction”

Russell’s Logic as the Essence of Philosophy

229

(op. cit., 44). With this supposed source of philosophical speculation laid bare, Dewey continues (op. cit., 43): Philosophy has arrogated to itself the office of demonstrating the existence of a transcendent, absolute or inner reality and of revealing to man the nature and features of this ultimate and higher reality. . . . Only comparatively recently has another conception of the proper office of philosophy arisen.

And what is this new conception that he commends? Dewey tells us that it is the empirical psychological and anthropological study of human civilization and culture. The proper subject for study is anthropology, primitive life, the history of religion, literature and development of social institutions. Dewey asks no philosophical questions about whether Newton, Darwin, Watson and Einstein were philosophers, questions about how to go about studying something in a field such as anthropology or psychology, or about whether mathematics is committed to numbers as necessary abstract particulars. He asks no philosophical questions about what methods are properly empirical or about what (if anything) demarcates science from religion, politics. As a result, reconstruction is an endeavor more susceptible than ever to being manipulated by the political and cultural dogmatisms that Dewey laudably deplores. Dewey cannot exempt an evolutionary reconstruction of Newton’s philosophical introduction of force at a distance into physics. He cannot exempt Einstein’s philosophy of general relativity which made mass and acceleration into kinematic metrical geometrical notions. He must even reconstruct Darwin’s a philosophical revolt against the Aristotelian essentialism of fixed specie. And if we follow Dewey’s plan, Dewey’s own philosophical reconstructions in philosophy are to be reconstructed as unsuccessful evolutionary behavior protocols. What could Dewey say in response to such an obvious self-refutation? Russell saw this clearly enough in the 1940s when he disparaged Dewey’s pragmatic form of evolutionary naturalism and demanded a special place for reason, truth, and logic (see Russell 1919b). Russell doesn’t mention Dewey in Our Knowledge, but does include among “current tendencies” a rejection of the work of James and Bergson which had set forth what they thought to be the transformative impact that adherence to the theory of evolution has on the nature of the philosophical enterprise. Russell writes (OK, 16): What biology has rendered probable is that the diverse species arose by adaptation from a less differentiated ancestry. This fact is in itself exceedingly interesting but it is not the kind of fact from which philosophical consequences follow.

230

The Bloomsbury Companion to Bertrand Russell

Russell was quite opposed to those advocating the thesis that philosophy should be “naturalized,” i.e., subsumed into empirical sciences of psychology, anthropology, evolutionary biology and the like. Russell’s work in Problems and Our Knowledge and the unfinished Theory of Knowledge endeavored to keep cp-logic and our knowledge of it privileged in a way a metaphysics of abstract particulars is not. Russell’s approach preserves philosophical questions as distinctive, and offers a distinctively scientific philosophy to investigate them. The synthetic a priori science of cp-logic is central to uncover the needed relational structures to unravel riddles about this or that purported necessity. It is in this sense that Russell’s thesis is that Philosophy, properly understood, is the scientific study (partly empirical, partly synthetic a priori) of necessity. Once we see this, Russell’s alliance with Wittgenstein and his Tractatus take on a new clarity. Wittgenstein also held that logic and mathematics, and our knowledge of them, is privileged. He held that philosophy is not an empirical science, that evolution is irrelevant to its reconstruction, that philosophy unravels riddles about necessity. He held that logical necessity is the only necessity. These and many more are Russellian ideas lifted into the Tractatus. Consider the following: 4.111 Philosophy is not one of the natural sciences. (The word “philosophy” must mean something whose place is above or below the natural sciences, not beside them.) 4.1122 Darwin’s theory as no more to do with philosophy than any other natural science. 4.113 Philosophy sets limits to the much disputed sphere of natural science. 4.114 It must set limits to what can be thought; and, in doing so, to what cannot be thought. It must set limits to what cannot be thought by working outwards from what can be thought. 4.115 It will signify what cannot be said, by presenting clearly what can be said. 4.1212 What can be shown, cannot be said. 6.37

… The only necessity that exists is logical necessity.

What is distinctive about Wittgenstein’s Tractatus is its attempt to apply itself to logic itself. The centerpiece of this radical elimination is its doctrine of showing. Logical necessity cannot be said with predicates, as Russell’s conception of

Russell’s Logic as the Essence of Philosophy

231

philosophy allowed; it must be shown. Logical necessity, and only logical necessity, is shown by building its distinctions in to the structured variables of the language of empirical science.16 Mathematical logic and our knowledge of it, remains privileged, for it alone is shown. The ouroboric Tractatus does not remove the very tool, the distinctively philosophical criticism afforded by cp-logic, which Russell so lauded. We are to climb a Russellian ladder (use the tool of mathematical logic) and then, adopting a technique of showing which builds logic/mathematics into the syntax of every empirical statement, throw away the ladder. Collecting cryptic remarks elucidates none, no matter how nuanced may be their presentation in a map.17 Wittgenstein’s Tractarian remarks on showing (his N-operator,  his recursive functions characterized with the help of numeral exponents, his exclusive variables, his picturing) remain to be deciphered. Wittgenstein’s reconstruction in philosophy leave us paralyzed. Russell’s reconstruction in philosophy offered a science and thereby allows us to move forward. What then are the problems of Russellian Philosophy? The problems of philosophy are obviously the problems that concern the existence and ground of purported kinds of necessity. Of course, Russell didn’t speak of “necessity” the way metaphysicians do today. Russell thought that the use of the word “necessary” by philosophers to be deplorable.18 It has many different meanings and the equivocations it bring on lends itself to many muddles. So let me clarify the point. Many philosophers imagine necessities to come in kinds: logical, mathematical, physical/causal, biological, ethical, metaphysical, and so on. Philosophy, according to Russell, is engaged in the study of such necessities, promoting, rejecting, identifying them. The case of mathematics is paradigmatic of an identification: mathematical necessity is logical necessity. Indeed, for Russell logical necessity is the only necessity. It is the task of philosophy, in Russell’s program of logical atomism, was to show how all other candidates are will o’the wisp. Ontology follows suit. If Russell is correct that logical necessity is the only necessity, then the right ontology is the ontology of cp-logic. If there are other kinds of necessity, then there are nonlogical ontologies. Either way, matters of ontology are distinctly philosophical and concern the ground of some or other kind of necessity. Only those facts (if any) that ground necessity can be properly said to be posited as a part of an ontological theory. Ontology is not the study of what there is; it is the study of what there has to be.19 In his History of Western Philosophy, Russell includes Einstein ideas as an example of logical analysis (HWP, 832). This is no quixotic exaggeration. The philosophers according to Russell’s conception of philosophy as the study of

232

The Bloomsbury Companion to Bertrand Russell

necessity include many who have been set aside as if exclusively empirical scientists—for example, Newton, Huygens, Maxwell, Einstein, Heisenberg, Pasteur, and Darwin. All and only those working on the foundations (pro or con) of some form of necessity are doing philosophy. The Russellian reconstruction in philosophy does not exclude the study of epistemology, the study of truth (and falsehood), the study of mind.20 Russell’s own multiple-relation theory held that in the simple cases, truth is necessarily a correspondence relation between a fact containing a universal and a belief-fact whose constituents include that very universal. Erroneous belief is possible only because universals, which exist of necessity, are often contingently exemplified. Views about truth engage with claims of what is necessary. In contrast, naturalized epistemology21 does not belong to philosophy. It offers contingent natural histories explaining the happenstance occurrences of organism-world relationships. This belongs to empirical psychology, chemistry and molecular biology. Dewey’s pragmatist theory of truth as adaptive organism-world engagement at the limit of some evolutionary process is not a philosophical theory. Dewey is a philosopher only because he denies the thesis that a correspondence relation is necessary to the nature of truth—offering as a replacement his contingent organism-environmental engagements wrought by the accidents of co-evolutionary natural selection of behavior protocols which attune organisms with their environmental niches. Russell’s program for a scientific philosophy enjoins us to be concerned with structure—logical form—since this is the nature of necessity. It is a vain hope to put aside such concerns. It is a vain hope that we may marginalize the “Whyquestions” and be left with the genuine “How-questions” which give the empirical scientific questions worth asking. For one thing, questions such as “How is -1 a number?” cannot be addressed by empirical science no matter what the story of evolution of behaviors protocols may be. The answer lies with the a priori science of the mathematical logician. For another, “How-questions” that are pseudoquestions empty as the worst teleological why-question pervade empirical science. The question, “How did the universal come to exist?” may well be among them. In his debate with Copleston, Russell dismissed the cosmological argument, arguing a priori that there is no Universe. The word “nothing” does not name a substance and neither does the word “everything.” If you want to investigate how we have a universe from nothing, you had better start with a synthetic a priori investigation into not only whether “nothing” names a vacuum field but also whether “everything” names a variable curvature metrical field.22 These are issues of logical form—issues of structure, of what is necessity. Current problems of the philosophy of mind continue to offer examples that reflect a failure of logical form.

Russell’s Logic as the Essence of Philosophy

233

It is only by recognizing a priori the failure of logical form, , not by an empirically robust sense of reality, that we understand that we cannot infer that there is something for which Ponce de Leon searched, on grounds that he searched for the fountain of youth. The distinctly philosophical problems of the nature of mind remain paralyzing because we do not know the logical forms required to solve the vexing problem of qualia, in which a mind might phenomenally experience a qualitative way, without there being a way (a quale) which the mind phenomenally experienced. Appeal to Russell’s famous “a robust sense of reality” is no help. We need the logical forms that enable us to avoid robust prejudices. Russell’s lesson is not that that his scientific philosophy admonishes us to employ Occam’s Razor as an empiricist’s supreme maxim of empirical science. The Russellian supreme maxim is not a maxim of empiricism, it is this: Complexity of logical form is inversely proportional to ontological baroqueness.

Metaphysical commitments to ontologies and kinds of necessities increase in inverse proportion to complexity of logical form. The most harmful ideology is one that excludes even the articulation of a rival. The search for new logical forms is the distinctive task of Russell’s logic as the essence of philosophy and the goal of his reconstruction in philosophy.

Notes 1 The discovery of both simple-types and impredicative comprehension began with Frege’s Begriffsschrift (1879). 2 Since mathematics concerns structures and not particulars (abstract or otherwise) that may exemplify the structures, we can see that simple-type theory in no way incapacitates mathematical studies. The sets Zermelo regarded as natural numbers, the sets Von Neuman regarded as natural numbers, and so on, violate simple-types. But mathematics is not interested in such sets. It is interested in the kind of structure that is a Progression. Every kind of structure can be studied within simple-typetheory. 3 Church prefers an axiomatization with axioms and a rule of uniform substitution. We present his position here for convenience so as to compare it with our interpretation of Principia. 4 See Landini (1998) and Landini (2007). 5 Gödel showed that there can be no consistent recursive axiomatization that captures all and only logical truths of even a second-order language where predicate variables may occur and be bound only in predicate positions.

234

The Bloomsbury Companion to Bertrand Russell

6 It remains an open question whether the cognitive innateness of our understanding of quantification may be analyzable in terms of the theory of acquaintance. I think it can, with the help of Quine’s variables explained away. See Landini (2015). 7 See Russell MPD, p. 134. When his 1911 paper “Knowledge by Acquaintance and Knowledge by Description” was republished in the 1917 collection Mysticism and Logic, Russell added a footnote that the multiple-relation theory is “somewhat unduly simple.” This comment is repeated in his 1918 Logical Atomism Lectures. 8 In his 1918 Logical Atomism Lecture Russell alludes to this without mentioning Theory of Knowledge. He had published the view in 1914 paper “On the Nature of Acquaintance.” See ONA. 9 There is evidence that Russell and Lady Ottoline Morrell imagined a writing a book called Prisons on the subject of a free man’s worship of the intellectual love of wisdom. But when their relationship ended, so also ended any plans for such a book. 10 For a discussion of how “this” became “thus,” see eds., Bernard Linsky and Donovan Wishon, Acquaintance, Knowledge and Logic: New Essays on Bertrand Russell’s Problems of Philosophy (Stanford: CSLI Publications, 1012), 20. 11 In fact, this is a thesis that Russell held as early as his 1906 paper “On ‘Insolubilia’ and Their Solution by Symbolic Logic”. 12 This passage was brought to my attention by Russell Wahl. 13 Perhaps Russell had in mind that Wittgenstein was working on a solution to the problem he had encountered in his Theory of Knowledge manuscript of 1913 concerning what we may be acquainted with in understanding the meaning of the logical particle signs. 14 See Ramsey, Letter to his mother September 20, 1923, in von Wright (1973, 77). 15 Russell’s appreciation for Spinoza’s views on ethics do not help to explain the change. Spinozistic views fit nicely with Russell’s idea that philosophic contemplation is the worship of a free man, once out of prison, who has thrown off the yoke of dogmatisms parading as if they were metaphysical necessities. Such a free man’s worship considers the world sub specie aeternitatis. This attitude of a scientific contemplation free of self-interest is expressed in Problems and in Russell’s work “The Essence of Religion” (1911). 16 See Landini (2007). 17 See http://tractatus.lib.uiowa.edu/ 18 See Russell’s “On the Nature of Cause,” Proceedings of the Aristotelian Society 13 (1912–13), 1026. Pagination is to its reprint in MyL, 180–208. 19 Russell’s conception of ontology is strikingly different from that of Quine. Quine’s naturalism is not evolutionary and, unlike Dewey’s, he attempted to face the issue of demarcation between philosophy and other forms of inquiry head on. Quine maintains that no questions are distinctly philosophical. All questions are on a

Russell’s Logic as the Essence of Philosophy

235

par—questions to be settled by empirical methods. This includes questions about existence of dinosaurs, comets, bacteria, quarks, qualia, numbers, sets, mereological sums and any abstract particular of any metaphysician’s reverie. Ontology for Quine is the empirical study of what there is. Quine is a relativist about what there is. See Quine (1969). 20 Some philosophers, for example, Joachim, argued that truth is necessarily coherence (since, in Joachim’s view, Bradley had shown that external relations, and hence an external correspondence relation, is impossible.) Others, such as Aristotle, held that truth is necessarily correspondence. When he embraced necessary existing propositions in his pre-Principia era, Russell was a truth-primitivist. 21 See e.g., Kornblith (2002). 22 Krauss (2012) hopes his book A Universe from Nothing shows how to separate scientific matters from religious dogmatisms by appeal to the scientifically acceptable “How-questions” and what he regards as the empty teleological “Why-questions.” This is untenable.

References John, Dewey. 1910. The Influence of Darwin on Philosophy. Bloomington: Indiana University Press. Kornblith, Hilary. 2002. Knowledge and its Place in Nature. Oxford: Oxford University Press. Kraus, Lawrence. 2012. A Universe From Nothing. New York: Atria Books. Landini, Gregory. 1998. Russell’s Hidden Substitutional Theory. New York: Oxford University Press. Landini, Gregory. 2007. Wittgenstein’s Apprenticeship with Russell. Cambridge: Cambridge University Press. Landini, Gregory. 2015. “Types* and Russellian Facts.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s the Problems of Philosophy, edited by B. Linsky and D. Wishon, 231–73. Stanford: CSLI Publications. Quine, Willard Van Orman. 1969. “Ontological Relativity.” In Ontological Relativity and Other Essays, 26–68. New York: Columbia University Press. Ramsey, Frank Plumpton. 1931. “The Foundations of Mathematics.” In The Foundations of Mathematics and Other Essays by Frank Plumpton Ramsey, edited by R. B. Braithwaite. New York: Harcourt, Brace and Co. Russell, Bertrand. 1914. Our Knowledge of the External World as a Field for Scientific Method in Philosophy (OKEW). London: Allen and Unwin, 1926. Pagination is to the first edition (Chicago: Open Court). Russell, Bertrand. 1917. Mysticism and Logic and Other Essays (ML). London: Longmans, Green and Co.

236

The Bloomsbury Companion to Bertrand Russell

Russell, Bertrand. 1919a. “On Propositions: What They Are and How They Mean,” in LK. Russell, Bertrand. 1919b. “Professor Dewey’s ‘Essays in Experimental Logic’.” Journal of Philosophy, Psychology and Scientific Methods 16: 5–26. Wittgenstein, Ludwig. 1922. Tractatus Logico-Philosophicus. Edited by D. F. Pears and B. F. McGuinness. London: Routledge & Kegan Paul. Wittgenstein, Ludwig. 1973. Ludwig Wittgenstein Letters to C. K. Ogden. Edited by G. H. von Wright, 77. Oxford: Basil Blackwell.

9

Sense-Data, Sensibilia, and Percepts Russell Wahl

1 Introduction “Let us give the name ‘sense-data’ to things that are immediately known in sensation: such things as colours, sounds, smells, hardnesses, roughnesses, and so on” (POP, 12). So Russell introduced sense-data in Problems of Philosophy. Russell used the term “sense-data,” as did G. E. Moore, as a neutral term. In fact they both introduced the term so as to distinguish what was sensed from the mental act of sensation.1 In contemporary philosophy, sense-data are generally taken to be phenomenal particulars, dependent on the mind, and contrasted with the physical world. The examples that both Russell and Moore give of sense-data might naturally lead one to think that they also hold this picture. This standard view of sense-data is nicely captured by Michael Huemer’s characterization in the Stanford Encyclopedia of Philosophy: Sense data are the alleged mind-dependent objects that we are directly aware of in perception, and that have exactly the properties they appear to have. For instance, sense data theorists say that, upon viewing a tomato in normal conditions, one forms an image of the tomato in one’s mind. This image is red and round. The mental image is an example of a “sense datum.” (Huemer, 2011)

Huemer sums up the standard view as holding (i) that sense-data are the kinds of things we are directly aware of by perception, (ii) that they are minddependent, and (iii) that they have the properties that perceptually appear to us. However, during the time Russell subscribed to the view that there are sensedata, he held (i), but not the others. There are passages where it seems that he held (ii), but many others which suggest otherwise. There are also arguments he gave which might seem to suggest that he held (iii), but other passages again

238

The Bloomsbury Companion to Bertrand Russell

which suggest that he did not accept this. The sense-data theory is now often linked with epistemological foundationalism and a concern for an absolutely certain foundation. Points such as (iii) above are important to this kind of foundationalism. Russell’s arguments for sense-data in POP have often been seen as of a piece with this position, and some of the things he says in Our Knowledge of the External World (OKEW) have also suggested this was his aim. But Russell did not hold (iii), and in fact his reasons for accepting sense-data are not linked to the foundational project of finding guaranteed truths as a foundation for all other knowledge nor as a need to combat skepticism.2 Russell did hold that sense-data are in some sense the basis for our knowledge of what he called “the external world.” In POP he held that the ordinary physical objects of the external world are known by description as the causes of our sense-data, so that our knowledge of the external world is based on an inference from our knowledge of sense-data. Starting in 1912 he no longer saw material objects as a distinct kind of entity inferred from the existence and organization of sense-data, but rather saw them as logical constructions of sense-data. In order to make his constructions such that the analysis would fit the requirements of a continuous public space and allow for the truth of the laws of physics, he introduced objects just like sense-data with the exception that they are not being sensed. These he termed “sensibilia.” The actual sense-data he still considered to be the ground of our knowledge of the external world. In My Philosophical Development (MPD) he reported abandoning sense-data in his 1921 Analysis of Mind. Yet while he gave up sense-data with his adoption of neutral monism, he had other entities serve many of the roles sense-data had played on his earlier views. In what follows I will look at the role sensedata played in Russell’s philosophy, the status of sensibilia, his reasons for abandoning sense-data, and the role sensations and percepts, the substitutes for sense-data, played in his later philosophy. Despite what appear to be radical shifts in view, we shall see that the modifications he made are less radical than they appear.

2  Sense-data introduced Russell’s introduction of sense-data stems from two developments in his philosophy. The first is the distinction he made between those items known directly in acquaintance and those only known by denoting concepts, or in his later views, “by description,” and the second was his anti-idealism. We can see

Sense-Data, Sensibilia, and Percepts

239

the roots of the first distinction as early as his 1903 manuscript, “Points about Denoting,” where he made the distinction between knowledge by acquaintance and knowledge by description (Papers 4, 306), and the manuscript from the same year, “On Meaning and Denotation,” where he connected “having an idea designating” something with having an immediate acquaintance with it (Papers 4, 324). However, the application of this theory to matter first appeared in “On Denoting,” where he enunciated the principle of acquaintance implicit in “Points about Denoting” but added, “Now such things as matter (in the sense in which matter occurs in physics) and the minds of other people are known to us only by denoting phrases, i.e. we are not acquainted with them” (Papers 4, 427). Yet while the principle of acquaintance was in place, the role of sense-data was not there yet, for Russell at this stage seemed to have in mind the contrast between ordinary objects and the material particles of physics. While Russell mentioned sense-data in 1910,3 his first real discussion of them appeared in his March 1911 lecture, “Knowledge by Acquaintance, Knowledge by Description,” which he read before the Aristotelian Society, and “Analytic Realism,” which he read originally as “Le réalisme analytique” to the Societé Française de Philosophie that same month. These papers had rather different focuses. While the first is concerned with epistemology, its focus is on the distinction between knowledge by acquaintance and knowledge by description rather than the issue of justification. The second espoused Russell’s general approach to philosophy and really didn’t discuss the epistemological role of sense-data. The Aristotelian Society paper was expanded in POP, where the distinction between acquaintance and description is important. In both of these papers a sharp distinction is made between the mental act of awareness and the object of that awareness. The first Russell called a “sensation” and the second, the “sense-datum.” In these papers he also held that we are directly acquainted with sense-data but not with enduring physical objects. In “Analytic Realism,” Russell emphasized the realism of his view, yet he also characterized the sense-data as “subjective.” He contrasted this subjectivism about particulars with a strong realism about concepts or universals, prompting one in his audience to see his view as “the inverse of Kant” (Papers 6, 427), as the raw data are considered subjective and the concepts which applied to them were perfectly objective. Russell clarified his view of sense-data as subjective in the following way: There is no general reason to reject naïve realism, i.e., the realism which claims that sense-data are identical with physical objects and that they subsist unchanged when not perceived. There are, however, special reasons .  .  . for

240

The Bloomsbury Companion to Bertrand Russell

believing that sense-data do not depend (causally) only on the object, but also on the subject. In other words, one is forced to believe that sense-data are entities which exist only when there is a certain relation between the subject and the physical object—a relation partly spatial and partly physiological.4 (Papers 6, 136)

Russell’s view that sense-data should not be identified with everyday physical objects has often been taken to be an affirmation that they are in some sense mental. Indeed, some of his French audience for “Analytic Realism” took him to hold that they are in some sense mind-dependent.5 However, the emphasis in “Analytic Realism” is on the realism: Russell was adamant that what is known does not have to be mental: “Knowledge of an object is a direct relation between the mind and the object: there is no corresponding change in the mind, but only a relation. . . . Knowledge does not imply that the mind and what it knows have a common nature” (Papers 6, 134). The subjectivity of sense-data is a matter of a causal interaction between the body and the input. The view that sensedata are not mental and that the subjectivity of sense-data is a function of the physiological is reiterated in Russell’s response to an article by G. Dawes Hicks, “The Nature of Sense Data.” On the question of whether the sense-datum is in the mind Russell had this to say: I hold that the sense-datum is certainly something other than the subject, something to which the subject’s relation is just as “external” as to the physical object. The only point where I part company from the out-and-out realist is in holding that, for various empirical reasons of detail, it is not certain that the quality which is the sense-datum ever exists at times when it is not a sensedatum. (Papers 6, 186)

The reasons of detail are then spelled out: “The sense-data . . . are not determined by the object alone but also by the physical intermediaries, including the senseorgans. This is one fact which makes it very difficult to identify the physical object with the sense-data to which it corresponds” (Papers 6, 188). This position is not different from the one he articulated in “Analytic Realism” before he wrote POP. These remarks, like those in “Analytic Realism,” stress what we might call the scientific argument for the claim that what we immediately experience cannot be an ordinary physical object. However, in POP Russell introduced sense-data after a classic relativity argument of the kind found in Locke, Berkeley, and Hume. The focus now was epistemological: sense-data must be the immediate objects of acquaintance because the table “appears to be different colours from

Sense-Data, Sensibilia, and Percepts

241

different points of view and there is no reason for regarding some of these as more really its colour than others” (POP, 29). He used this argument for the claim that what we immediately perceive cannot be physical objects, as what we perceive changes with perspectives and viewers and we do not hold that the physical objects so change. So the argument employs ordinary beliefs about physical objects to contrast “what is immediately known” from the ordinary physical objects. “What is immediately known in sensation” are the sensedata, such things as colors, sounds, and smells. The sense-data of POP are characterized as private and fleeting, which has suggested to some people that at least in POP Russell was unsure whether to characterize sense-data as mental or physical. But in chapter three, on idealism, Russell contrasted his position with Berkeley’s, first giving Berkeley’s view that the colors and other things that we immediately perceive are ideas, and then rejecting that conclusion from the  claim that sense-data in some sense depend on us and wouldn’t exist if we were not perceiving them (POP, 41). In fact, his argument for their being unlike the physical objects that we ordinarily think are there again involved the scientific argument: I know that, if I move, the parts that reflect the light will be different, so that the apparent distribution of colours on the table will change. It follows that if several people are looking at the table at the same moment, no two of them will see exactly the same distribution of colours, because no two can see it from exactly the same point of view. (POP, 8)

While Russell likened his method to that of Descartes, in fact he did not employ the technique of Descartes’s first Meditation. Russell’s arguments employed background scientific knowledge and background views of what ordinary physical objects are supposed to be like.

3  Logical constructions and sensibilia After the arguments against idealism and the introduction of the distinction between knowledge by acquaintance and knowledge by description, the focus of POP turned to the issue of our knowledge of the external world. Russell argued that we must use a synthetic a priori principle to conclude that a world other than that we immediately perceive causes our sense-data, and that therefore physical objects are known by description as the causes of our sense-data.

242

The Bloomsbury Companion to Bertrand Russell

The account given of a priori knowledge involved acquaintance with universals, although which universals might be involved in the grounding of the induction principle was never made clear. In all this, Russell drew a sharp contrast between ordinary physical objects and those with which we are immediately acquainted. Russell considered sense-data to be private to individuals, existing in “private spaces” (POP, 29). In contrast to the kind of things he thought sensedata were, he thought that physical objects serve as the common referent of our experiences and so must exist in an inferred public space and are “over and above the private and particular sense-data which appear to various people” (POP, 21). He argued for the acceptance of matter on the basis of the simplicity of the explanation of our sense-data and those of other people. He also argued that the belief in matter was “instinctive,” and while that did not necessarily mean the belief was true, he held that instinctive beliefs should be rejected only for positive reasons (which should also be based on instinctive beliefs) and there were none in this case. Russell’s confidence in this position waned as he wrote “On Matter” in 1912. In that paper Russell rejected the inference from sense-data to physical objects as he reflected on the nature of a priori truth.6 The position tentatively suggested at the end of “On Matter” involved substituting series of sense-data and qualities just like sense-data but not actually being sensed by anyone at the moment. This position was reworked in late 1913 and early 1914 and presented in “The Relation of Sense Data to Physics” (RSDP) and also in OKEW. There are two related points, likely prompted by his conversations with Wittgenstein, which caused Russell to rethink what he had done in POP, but neither of these had to do with the nature of sense-data per se. Russell reported that Wittgenstein made him more skeptical and so led him to think that the inference, argued for in Problems, from sense-data to enduring physical objects was unwarranted. Wittgenstein also appears to have thought that the nonexistence of matter would not be a problem for science as all the propositions of science could still be interpreted as true. Presumably, on Wittgenstein’s view, the truths of science could be understood simply as truths about what an individual immediately experiences. While “On Matter” started with the thought that physics could be constructed from one person’s sensedata alone, Russell abandoned this position as a failure. What he needed in the constructions were series of items just like sense-data, but which were not the objects of anyone’s actual experience. So at the end of “On Matter” he bluntly affirmed that “qualities which are or resemble sense-data … exist at times

Sense-Data, Sensibilia, and Percepts

243

when they are not given in sense” (Papers 6, 93). These he termed “sensibilia” in RSDP. Those who see Russell’s project as one of finding foundational propositions which are immune from doubt will see the introduction of sensibilia as a betrayal of the project, and completely wrongheaded. Those who hold that sense-data are to be understood as mental entities find the very concept of sensibilia incoherent. Yet while Russell gave lip service in RSDP to the attempt to do without sensibilia, and made some remarks in OKEW that the ideal project would be a construction of physical objects using only actual sensedata and even the actual sense-data of just one person, he proceeded with sensibilia. In RSDP and his 1915 paper “The Ultimate Constituents of Matter” (UCM) he held that qualities just like sense-data, but unsensed by any mind, exist unperceived. The one place he rehearsed the position rejected in “On Matter” is in OKEW. In Chapter III of that work Russell outlined the issue of knowledge of the external world as that of inferring the existence of anything other than “our own hard data” from that data (OKEW, 73).7 Hard data are confined to our own sense-data, the laws of logic, and some facts of memory and observed relations among the data. After pointing out the meagerness of what can be inferred from this, Russell adopted the viewpoint of RSDP including sensibilia and other people’s sense-data.8 However, in Chapter IV when returning to these issues Russell said: “If physics is to consist wholly of propositions known to be true . . . [sensibilia and other people’s sense-data] must all be capable of being exhibited as logical functions of sense-data” (OKEW, 111). Russell then labeled these other perspectives “ideal” and only the observer’s actual sense-data are “real.” Scott Soames sees these remarks as evidence that Russell’s primary concern was a quest for certain foundations, and that Russell has rejected the notion of sensibilia as they appear in RSDP. Soames thinks that ultimately Russell wanted to cash out what had been reference to sensibilia in counterfactual terms—that is, what actual sense-data would appear to an observer in such circumstances. But Soames then rightly points out some extreme difficulties with Russell’s project as he had characterized it. As Russell realized in RSDP and continued to hold in his later work, the constructions require all the perspectives. Soames is aware of that, and seeing Russell’s project as developing a foundation of knowledge from the first person perspective sees the glaring problem: But surely, if Russell can’t explain how we know other people exist, then he can’t explain our alleged knowledge of what everyone else’s sense data would be like

244

The Bloomsbury Companion to Bertrand Russell

under all relevant conditions. For this reason, his analysis of material objects as logical constructions out of everyone’s sense data doesn’t solve the problem it was designed to solve. (Soames 2014, 566)

Soames sees Russell’s project, then, as similar to Carnap’s project in the Aufbau, where Carnap explicitly argued for what he called autopsychological constructions.9 OKEW appears to advocate this position more than Russell’s other work at this time. What is puzzling is why Russell returned to talk of constructing everything out of the sense-data of a single person when he had already rejected that possibility. Perhaps Russell was rehearsing the project as outlined in the earlier drafts of “On Matter.” There is some evidence that Russell wrote some of the lectures of OKEW even before he wrote RSDP.10 In any case, Russell’s talk of exhibiting other people’s sense-data as a logical function of our own has to be taken somewhat loosely. Russell seems to have had in mind filling in gaps of our own sense-data with items just like them so as to constitute a continuous series. It is then these series that go into the logical constructions of public space and public objects. The series of sensedata is known by description in the way Russell had claimed physical objects are known in POP. Russell had already argued that what A. J. Ayer later called a “horizontal inference” (i.e., an inference to an entity of the same kind) was superior to the vertical inference to enduring physical objects, of which his own views about physics conflicted.11 Sense-data or items like them that are unsensed—that is, they are what Russell called the ultimate constituents of matter: I maintain . . . that if physics could describe truly and fully all that occurs in the physical world when there is a flash of lightning, it would contain as a constituent what I see, and also what is seen by anybody else who would commonly be said to see the same flash. What I mean may perhaps be made plainer by saying that if my body could remain in exactly the same state in which it is, although my mind had ceased to exist, precisely that object which I now see when I see the flash would exist, although of course I should not see it, since my seeing is mental.12 (Papers 8, 78)

Russell’s sense-data were foundational units in the sense that they are the objects of immediate acquaintance, and all understanding ultimately rests on acquaintance. But Russell had made a distinction between knowledge by acquaintance and knowledge by description and he never thought that the only things we should admit into our understanding of the world are objects of acquaintance. After rejecting the inference to enduring physical objects as

Sense-Data, Sensibilia, and Percepts

245

known by description Russell substituted logical constructions out of sensedata and objects like them for the enduring physical objects. He recognized that our knowledge of the existence of these series is not guaranteed, but the hypothesis of these series of such “ultimate constituents of matter” enabled an analysis where the sentences of physics come out as true. It is important to realize that there is no ontological difference between sensibilia and sense-data. In RSDP he held that “the relation of a sensibile to a sense-datum is like that of a man to a husband: a man becomes a husband by entering into the relation of marriage, and similarly a sensibile becomes a sense-datum by entering into the relation of acquaintance” (Papers 8, 7). Given that Russell maintained a view like this from “On Matter” through this work and even after he had replaced sense-data with percepts, as we shall see, it is hard to think that Russell really thought he could dispense with sensibilia, even in OKEW.13

4  Sense-data rejected Later in life, Russell said that he abandoned sense-data in his 1921 Analysis of Mind.14 Recall that the sense-data of 1911–12 were introduced primarily for two reasons: (1) to distinguish the objects of awareness from the mental acts of sensation, and in (2) to distinguish what we are immediately aware of from what we take to be ordinary physical objects. By 1919 Russell embraced neutral monism, and it is in the context of this embrace that he rejected sense-data. As a neutral monist, he now rejected the mental subject which was supposed to have the sensations and be directly acquainted with the sense-data. He still maintained that something like acquaintance was involved in understanding, but he said that if we regard the sensation as itself a cognition, we must distinguish the seeing from what is seen: we must say that, when we see a patch of color of a certain shape, the patch of color is one thing and our seeing of it is another. This view, however, demands the admission of the subject. (AMi, 141)

It is this admission which he found objectionable. When we reject the subject, the possibility of distinguishing the sensation from the sense-datum vanishes.  .  .  . Accordingly the sensation that we have when we see a patch of colour simply is that patch of colour, an actual constituent of the physical

246

The Bloomsbury Companion to Bertrand Russell

world, and part of what physics is concerned with. A patch of colour is certainly not knowledge, and therefore we cannot say that pure sensation is cognitive. (AMi, 142)15

The distinction between the sensation as mental act and the sense-datum as object of that act is abandoned. Russell’s neutral monism took as its basic elements particulars which are in themselves neither mental nor physical. What constitutes the mental or physical will be the way these various particulars are grouped. In saying we cannot distinguish the sensation from the sense-data on the neutral monist view, and in saying that the sensation simply is the color, Russell could easily be construed as rejecting not the sense-datum of old, but  the sensation. What he had characterized as the sense-data, the patches of color, etc. remain. It is the mental acts of awareness of these, the sensations, which are called into question: “A patch of colour is certainly not knowledge, and therefore we cannot say that the pure sensation is cognitive.” A. J. Ayer took the position that the best way to read Russell’s later philosophy was to understand Russell as relabeling sense-data: Though he subsequently speaks of himself as having “emphatically abandoned” sense data at this time, the change in his view is much less radical than this would suggest. He did cease to employ the term “sense-datum,” but he continued to speak of percepts, to which he attributed the same properties as he had attributed to sense-data, except that of being correlative to sensory acts. (Ayer 1972, 71)

Ayer’s view was that what was rejected in the earlier position was not the sensedatum but the mental sensation that was associated with it. What is especially confusing is that in AMi Russell continued to use the word “sensation” for such things as the patch of color, while rejected the view that these should be understood as mental entities. He rejected the word “sense-datum” since it was a mistake to think of these in isolation as data for a mind. Some support for Ayer’s view can be found in Russell’s later works. He called patches of color, etc., “sensations” in AMi, and “percepts” in Analysis of Matter (AMa). For example, in AMa, he characterized percepts as “patches of color, noises, smells, hardness, etc.” (AMa, 257). In An Outline of Philosophy, he suggested that percepts on the new theory play a role much like sense-data did on the old view. In the chapter on “Physics and Perception” he discussed what we

Sense-Data, Sensibilia, and Percepts

247

see when we see stars, both when we are looking at the sky and when someone hits us in the head. Using the relativity argument, Russell said, The sun looks red in a London fog, grass looks blue through blue spectacles, everything looks yellow to a person suffering from jaundice. But suppose you ask: What colour are you seeing? The person who answers, in these cases, red for the sun, blue for the grass and yellow for the sick-room of the jaundiced patient, is answering truly. And in each of these cases he is stating something that he knows. What he knows in such cases is what I call a “percept.” I shall contend later that, from the standpoint of physics, a percept is in the brain; for the present, I am only concerned to say that a percept is what is most indubitable in our knowledge of the world. (OOP, 139)

Clearly, Russell held that percepts are those things of which we are most directly aware. In AMa they and entities like them are the building blocks of the constructions. Another place where we might see a continuity from sense-data in the old view to sensations or percepts in the new view is in Russell’s revision of Our Knowledge of the External World. Russell revised this for the 1926 reprint and it is interesting to compare the two editions. In fact, not many changes were made between the editions, and most of them in Lecture IV, “The World of Physics and the World of Sense.” Most of these additions involved adding remarks on current scientific theory (OKEW 2nd, 108), further discussion of space, and Whitehead’s construction of points (OKEW 2nd, 119). Following the exposition of the earlier work, Russell continued to refer to sense-data even in the changes. What most interests us is a paragraph added to Lecture III, “On Our Knowledge of the External World”: According to some authors—among whom I was formerly included—it is necessary to distinguish between a sensation, which is a mental event, and its object, which is a patch of colour or a noise or what not. If this distinction is made, the object of the sensation is called a “sense-datum” or a “sensible object.” Nothing in the problems to be discussed in this book depends upon the question whether this distinction is valid or not. If it is not valid, the sensation and the sense-datum are identical. If it is valid, it is the sense-datum which concerns us in this book, not the sensation. (OKEW 2nd, 83)

Russell didn’t want to include his arguments for neutral monism at this stage, so he presented OKEW as neutral between using sensations or sense-data as the building blocks. This remark may suggest that while Russell had thought

248

The Bloomsbury Companion to Bertrand Russell

sense-data were nonmental and thus might exist unperceived, that in fact all the constructions could have been done with entities which cannot exist unperceived, such as sensations.16 But Russell’s “sensations” of 1926 should definitely not be equated with the “sensations” of 1914. The latter were clearly mental items which could not exist without a mind, the “sensations” of the new view are the elements, the ultimate constituents of both matter and mind. No item of this kind is either mental or material in itself on this view, but only as it enters into constructions of one or the other. It is not impossible for them to exist apart from minds. While this passage does not support the view that sense-data could just as well have been mental items in 1914, it does appear to buttress Ayer’s view that Russell merely relabeled sense-data. Others, on the contrary, have thought that sense-data were essentially defined by their relation to a subject and thus had no place in the new theory.17 However, Russell when he still adhered to the distinction of sense-data and sensations held that “the existence of sense-data is logically independent  of the existence of mind, and is causally dependent upon the body of the percipient rather than his mind” (UCM Papers 8, 85). Readers of Russell who thought that he was committed to the position that our knowledge of the immediate properties of sense-data was infallible also see a change in position from the old theory of sense-data. But we have already seen that this was not Russell’s considered view even when he held on to sense-data.18 However, there is another reason to think that the relabeling view is mistaken. In Lecture VII of AMi Russell defined a perception as “the appearance of the object from a place where there is a brain” (AMi, 131). The presence of a brain will involve certain other features of these particulars, namely that they are remembered and associated in certain ways. “It is these . . . characteristics .  .  . that distinguish perceptions from the appearances of objects in places where there is no living being” (AMi, 131). Sensations by themselves are not cognitive. Perceptions, which are cognitive, are sensations which are connected with other particulars in certain ways, and are influenced by them. Those influences, which he called mnemic influences, that is influences involving remembering and association, can theoretically be separated out, and the “sensation” is what would be left over: “The sensation is a theoretical core in the actual experience; the actual experience is the perception” (AMi, 132). This suggests that the sensation should not have been thought of as

Sense-Data, Sensibilia, and Percepts

249

relabeled sense-data, for they are even experienced. Russell emphasized this approach the most in his later HK: When, say, we see and hear an airplane, we do not merely have the visual sensation and the auditory sensation of a whirring noise; spontaneously and without conscious thought we interpret what we see and hear and fill it out with customary adjuncts. (HK, 167)

After describing a view of a wall which was easily mistaken for a road going up a hill, he continued: The right interpretation as a house and the wrong interpretation as an uphill road were both, in a sense, inferences from the sensational datum, but they were not inferences in the logical sense, since they occurred without any conscious mental process. (ibid.)

Russell called this sort of inference “animal inference” and then talked of such  inferences filling in the sensational core until it becomes a perception (HK, 169). This talk of sensation as a “theoretical core” may make one think that knowledge of sensations is somehow inferred from other experiences. Using Russell's earlier terminology sensations would be known only by description as contributing causes of our experiences. Seen this way, the new view is a complete rejection of the old view that sense-data are known directly. On this picture the sensations are part of an unconscious core and what anyone is conscious of is already a filled-in perception. We should be a little cautious of this interpretation, even though it is true that in HK Russell moved away more from even the modest foundationalism of POP and his earlier works.19 First, Russell did not see the sensational core as something that was necessarily beyond experience. It is clear that in HK he thought that the sensational core could become an object of attention, and he came close to reverting back to POP when he suggested that we need to strip away both the conscious and the animal inferences in order to get to genuine data: “Only sensations and memories are truly data for our knowledge of the external world” (HK, 170). When emphasizing why perceptions can be misleading, Russell often contrasted them with mere sensations. In MPD he went so far as to call the “stimulus alone” the “theoretical core” of a total experience and said this latter “is always an interpretation in which the sensational core has accretions embodying habits” (MPD, 106) Nonetheless, most of his discussion

250

The Bloomsbury Companion to Bertrand Russell

in MPD is taken from IMT, and in the middle of that discussion we again have a discussion of such propositions as “this is warm” and “this is next to that.” Of these he stated, The important point is that such propositions .  .  . are still directly based on experience. .  .  . This is quite a different matter from Gestalt as dealt with in Gestalt-psychology. Take (say) perception of the ten of clubs. Any person used to cards sees at once that it is the ten of clubs, and sees it by a perception of Gestalt, not analytically. But he can also see that it consists of ten similar black patterns on a white ground. (IMT, 56)

The second point I should mention is that Russell made very similar remarks even in the heyday of his acceptance of sense-data. In POP he was clearly aware that we don’t always notice the features of sense-data and pointed out that it often takes the training of an artist to be able to notice these things, a point reiterated in similar terms in IMT. Take the following remarks from the 1915 UCM to see that Russell was quite sensitive to these points during his embrace of sense-data: Psychologists point out how much of what we think we see is supplied by association or unconscious inference, how much is mental interpretation, and  how doubtful is the residuum which can be regarded as crude datum. From these facts it is argued by psychologists that the notion of a datum passively received by the mind is a delusion, and it is argued by the physiologists that even if a pure datum of sense could be obtained by the analysis of experience, still this datum could not belong, as common sense supposes, to the outer world, since its whole nature is conditioned by our nerves and sense organs, changing as they change in ways which it is thought impossible to connect with any change in the matter supposed to be perceived. (Papers 8, 76)

Russell used these points to call into question the naïve realist (or common sense) view that ordinary things are exactly as they seem, and to set the stage of a skepticism he wishes to challenge, but in this paper he did not reject the findings of psychologists or physiologists except with respect to the remark that the datum could not belong to the outside world. This passage could easily have come from later critiques of sense-data, but Russell didn’t see it that way, because the question of the certainty of the qualities of what we have immediately before our minds was not really an issue in UCM. What was at issue is how, given these observations, one can make sense of the relation of mind to matter.

Sense-Data, Sensibilia, and Percepts

251

5  Russell’s epistemology after the rejection of sense-data Russell’s philosophy continued to evolve after his rejection of sense-data. The major change was that he no longer thought of acquaintance as a simple unanalyzable relation between a mind and a sense-datum. There is no talk of bare acquaintance with facts of the kind he had given in POP. There are other positions developed in his later philosophy which have suggested to some commentators that a more radical change took place with the advent of AMa. In particular Russell advocated a causal theory of perception in AMa which some have taken to be incompatible with Russell’s earlier project of constructing physical objects out of sense-data and sensibilia. Those who see things this way also see Russell as rejecting the constructions of 1914–18 and a return to the previously abandoned view that physical objects are known by inference rather than understood as constructions.20 It is true that in AMa Russell emphasized that percepts should be understood as being caused by physical objects, and this was a position he had held in POP. There the fact that physical objects were taken to be the causes of sense-data was used to argue for a vertical inference from sense-data to objects of a completely different kind. In AMa, however, we do not have anything like this vertical inference. We have instead constructions, now from percepts. Percepts, like all the elements used in constructions in AMa, are events. Russell ended his chapter on the causal theory of perception with the following: I shall assume henceforth not only that there are percepts which I do not perceive, connected with other people’s bodies, but also that there are events causally connected with percepts, as to which we do not know whether they are perceived or not. (AMa, 215–16)

These two assumptions correspond exactly to the assumption of other people’s sense-data and the unsensed sensibilia of RSDP. This is not a radical change, but a continuation of the same project.21 The inference to the physical objects will still be a horizontal inference. There is no reason to think that these other events are of a different kind from percepts. The causal theory of perception Russell advocated in AMa still considered physical objects to be constructions. The events out of which they were constructed were not different in kind from percepts. This is actually a position Russell also advocated in RSDP, where he held that “we want to be able to express the fact that the appearance of a thing in a given perception is causally affected by the matter between the thing and the perspective” (Papers 8, 16–17).

252

The Bloomsbury Companion to Bertrand Russell

Russell thought at the time that traditional causal talk was mistaken and that the concept of a scientific law was best understood simply as differential equations which displayed relations of some values to others.22 By the time he came to write HK he was more cautious about this matter, seeing the drawbacks of holding that causation simply is “invariable concomitance or succession” (HK, 455). He did not define causality in HK, but he talked of a “primitive concept of causality” which he said was “not used in any advanced science.” But “still has importance as the source of approximate generalizations and pre-scientific inductions, and as a concept which is valid when suitably limited” (HK, 453). In HK Russell was willing to consider that the causes of percepts are radically different from percepts. “The physical causation of perception,” he said, “makes it probable that physical objects differ greatly from percepts” (HK, 209). To an extent, he had moved away from at least part of the motivation of the constructions, that entities of a wholly different kind from those we have as data should be avoided. Consequently, Ayer’s distinction between horizontal and vertical inferences would no longer be applicable. Physical objects in HK are considered series of events, and causal lines from these extend to our brains, where the events which are percepts occur. In HK Russell saw physical objects in part as constructions, but also as entities inferred. In a sense this was also true of the constructions out of sense-data during the 1914–18 period. The full constructions were not themselves data, but were inferred from the immediate sense-data. So in the end Russell still maintained part of the epistemological role of sensedata up to HK. He also kept the spirit of construction.23 But the constructions of 1914–18 and even those of AMa were done in terms of items that could be objects of experience or were ontologically the same as such objects. This is no longer a requirement for the constructions in HK. HK begins not with the epistemology, as we have in OKEW, but with the science. Percepts don’t make an entrance until Part III after the discussion of science and language. But they are still there, in a role which has evolved from the role of sense-data in the early philosophy rather than as a result of a radical break.

Notes 1 See Moore 1953, 44. While Some Main Problems in Philosophy was published in 1953, it was a set of lectures Moore gave in 1910–11, and Moore did not modify the lecture, but instead indicated in footnotes where he disagreed with it.

Sense-Data, Sensibilia, and Percepts

253

2 See Linsky 2015, Proops 2015 and Wahl 2015, all in Wishon and Linsky 2015, for arguments that Russell did not hold (iii). 3 This was in the revision of the second part of “The Nature of Truth,” which he included in Philosophical Essays (PE). 4 Note the translation in Papers 6 is in error here—his actual (French) text says “physiological” but the translation has “psychological.” See Papers 6 413 for the correct French text. The mistranslation could easily make one think sense-data are mind-dependent, but Russell is clear in this work that the dependence he has in mind is physiological, not psychological. 5 Gaston Milhaud remarked that for Russell that concrete object depends somewhat on the mind (Papers 6 144, 428). He asked that if we get rid of all humanity, would Russell say that there would be no sounds, no color, but that mass, potential energy, and kinetic energy would continue to exist? Russell answered “yes,” and Milhaud said he must disagree, but did not further state his position. Russell’s answer could easily be taken to see these properties as mind-dependent, but his other remarks suggest the dependence is on the physiological rather than the psychological. 6 See Chapter 4 of this book for a discussion of Russell’s rethinking of a priori inferences as a result of his discussions with Wittgenstein. 7 References to OKEW are to the first edition unless otherwise specified. Russell revised the work in 1926 after he had changed some of his views, particularly after he had abandoned sense-data and the view of acquaintance as a two-place relation between a mind and the item with which one is acquainted. 8 In OKEW Russell does not use the term “sensibilia,” but instead talks of a system of perspectives of “all views of the universe, perceived and unperceived” (OKEW 88). These perspectives play the role of sensibilia and other people’s sense-data. 9 Similar in the sense of taking an autopsychological point of view. Of course Carnap’s elements are not at all like Russell’s sense-data. 10 See Blackwell 1973 for reference to the composition of OKEW. The breakthrough Russell spoke about in the beginning of January 1914 appears to be a reference to RSDP. 11 See Ayer 1972, 34–35. Ayer went so far as to suggest that Russell would not call entities made by a horizontal inference “inferred entities” as he would reserve that usage for vertical inferences. Of course Russell realized that there is a clear sense in which sensibilia and other people’s sense-data are inferred entities, but Ayer is on the whole correct that when Russell called into question “inferred entities,” he was usually speaking of entities inferred by what Ayer calls vertical inference. 12 A similar point was made earlier in RSDP (Papers 8, 8). 13 My own view in contrast to Soames’s is that Russell’s counterfactual talk should probably be understood as a way of referring to sensibilia. Russell thought we construct (actually posit) a continuous series of sense-data like objects from our

254

The Bloomsbury Companion to Bertrand Russell

gappy actual sense-data. We then construct other series from perspectives other than our own. It is only in this sense that the sensibilia and the sense-data of others can be seen as a function of actual sense-data. 14 See MPD, 101. He also said this in an unpublished paper on perception which he apparently wrote for A. J. Ayer, after reading Ayer’s Problem of Knowledge (Papers 11, 305) as well as a published review of Gilbert Ryle’s Concept of Mind. In this otherwise critical review of Ryle, Russell said that he agrees with Ryle’s criticism of sense-data and will say no more on the topic (Papers 11, 635). 15 It is these pages from AMi that Russell cited at length in MPD when recounting his rejection of sense-data. 16 Soames appears to take this passage this way in Soames (2014, 540, 557). 17 Sainsbury thinks this, and thinks there would be something circular in constructing minds out of these since they are themselves essentially related to minds. See Sainsbury 1979, 192. 18 See the references in note 2. This topic was also dealt with in Savage, 1989, in a thorough discussion of Salmon, 1974. Salmon in that piece does seem to hold that Russell held a more traditional view of sense-data. 19 See Chapter 11 of this volume by Griffin and Olson for a further discussion of Russell’s epistemology in HK. 20 For articulations of this view, see Sainsbury 1979, 53, and Bostock 2012, 192–96. Bostock is aware that Russell himself did not think of AMa as involving a radical change in his view. 21 Of course if one thinks the project really didn’t involve sensibilia, or that they were ideal in some sense, one might see this as a change. But a careful reading of RSDP should dispel this. Russell recognized that things changed with his neutral monism, but he held in his 1922 “Physics and Perception” that “I have never been a phenomenalist.” In his response to Stace in the Schilpp volume he denied that there was a radical change between AMi and AMa (Schilpp 1944, 707). 22 He argued for this position in “On the Notion of Cause” (1913), reprinted in Papers 6. 23 The constructivist aspect is further complicated in HK by Russell’s new view that particulars themselves are bundles of qualities. See HK Part IV Chapter 8 for this discussion.

References Ayer, A. J. 1972. Russell. New York: Viking. Blackwell, Kenneth. 1973. “Our Knowledge of ‘Our Knowledge.’” Russell o.s. 12: 11–13. Hicks, G. Dawes. 1912 Mind n.s. 21. Reprinted in Papers 6 as Appendix 6.

Sense-Data, Sensibilia, and Percepts

255

Huemer, Michael. 2011. “Sense-Data.” In The Stanford Encyclopedia of Philosophy. Edited by Edward N. Zalta, http://plato.stanford.edu/archives/spr2011/entries/sense-data/. Linsky, Bernard. 2015. “Acquaintance and Certainty in The Problems of Philosophy.” In Aquaintance, Knowledge, and Logic, edited by Donovan Wishon and Linsky Bernard. Stanford: CSLI. Moore, G. E. 1953. Some Main Problems in Philosophy. London: George Allen and Unwin. Proops, Ian. 2015. “Certainty, Error, and Acquaintance in The Problems of Philosophy.” In Aquaintance, Knowledge, and Logic, edited by Donovan Wishon and Linsky Bernard. Stanford: CSLI. Sainsbury, R. Mark. 1979. Russell. London: Routledge. Salmon, Wesley. 1974. “Memory and Perception in Human Knowledge.” In Bertrand Russell’s Philosophy, edited by George Nakhnikian. New York: Barnes & Noble. Savage, C. Wade. 1989. “Sense-Data in Russell’s Theories of Knowledge.” In Rereading Russell, edited by C. Wade Savage and C. Anthony Anderson. Minneapolis: University of Minnesota Press. Schilpp, Paul A. (ed.) 1944. The Philosophy of Bertrand Russell. Chicago: Open Court. Soames, Scott. 2014. The Analytic Tradition in Philosophy. Volume 1: The Founding Giants. Princeton: Princeton University Press. Wahl, Russell. 2015. “Sense-Data and the Inference to Material Objects: The Epistemological Project in Problems and Its Fate in Russell’s Later Work.” In Aquaintance, Knowledge, and Logic, edited by Donovan Wishon and Linsky Bernard. Stanford: CSLI. Wishon, Donovan and Linsky Bernard (eds.) 2015. Aquaintance, Knowledge, and Logic. Stanford: CSLI.

10

Russell on Introspection and Self-Knowledge Donovan Wishon

1 Introduction This chapter examines Russell’s developing views—roughly from 1911 to 1918— on the nature of introspective knowledge and subjects’ most basic knowledge of themselves as themselves. One reason for doing so is that the details of Russell’s views on introspection have largely been neglected or misunderstood, despite the sizeable interest in his epistemology and metaphysics of the self. Another reason is that doing so helps shed additional light on other aspects of his thought at the time, such as his broader acquaintance-based theory of knowledge, his preference for logical constructions over inferred entities, and his gradual progression toward neutral monism. This chapter argues that Russell’s theory of introspection distinguishes between direct awareness of individual psychological objects, the presentation of psychological complexes involving those objects, and introspective judgments that aim to correspond to them. It also explores his transition from believing subjects enjoy introspective self-acquaintance, to believing they only know of themselves by self-description, and eventually to believing that self-knowledge is a logical construction. It concludes by sketching, in broad outline, how Russell’s views about introspection and self-knowledge change as a result of his adoption of neutral monism.

2  Knowing things and knowing truths Russell’s early views (from around 1911 to 1913) about introspective knowledge and self-knowledge center on his well-known notion of knowledge by acquaintance. Because there are many misconceptions about its nature and role

Russell on Introspection and Self-Knowledge

257

in Russell’s theory of knowledge, this section briefly considers how it fits into his overall epistemology. While some of the elements of his view appear earlier, it achieves its fullest articulation in his 1911 “Knowledge by Acquaintance and Knowledge by Description” (hereafter KAKD), 1912 The Problems of Philosophy (hereafter POP), and abandoned 1913 Theory of Knowledge manuscript (hereafter TK). Thus, they are the primary focus of this discussion. In these works, Russell’s theory of knowledge—inspired by the writings of William James (1890/1950)—draws a fundamental distinction between knowledge of things and knowledge of truths. Like James, Russell sees this distinction as corresponding imperfectly with an ordinary distinction drawn in many natural languages, including French and German, between two uses of the term “know” (POP, 69–70).1 We sometimes talk about a subject knowing that such and such is the case regarding some subject matter, but we also talk about a subject simply knowing of a person, place, or thing, either by encountering it firsthand or in virtue of some less direct informational connection to it. However, many interpreters have pointed out that Russell’s theory of knowledge departs from ordinary usage in numerous ways (Bostock 2012; Crane 2012; Kremer 2015). There is a great deal of ongoing controversy concerning how exactly Russell understands these two kinds of knowledge and their relationship to one another. As I interpret him, a subject has knowledge of a thing if and only if the subject is in some way aware of it such that he or she is in a position to think and talk about it.2 In contrast, a subject has knowledge of truths about something if and only if the subject bears a cognitive attitude toward it which can be evaluated in terms of truth or falsity (or success conditions more generally). Paradigmatically, such knowledge involves conceptually articulated beliefs and is propositional (or multiple-relational) in character. Knowledge of truths does not, however, require the exercise of reflective judgment or reasoning, as it often involves “instinctive,” “psychological,” or “physiological” judgments and inferences. For Russell, there are two ways a subject can have knowledge of something. Subjects can know of something by becoming directly aware of it “without the intermediary of any process of inference or any knowledge of truths” because it is presented to them in experience (POP, 73). Russell calls this kind of knowledge of things knowledge by acquaintance. Alternatively, subjects can know of something indirectly in cases where “in virtue of some general principle, the existence of a thing answering to this description can be

258

The Bloomsbury Companion to Bertrand Russell

inferred from the existence of something with which [they are] acquainted” (POP, 71). Russell calls this kind of knowledge of things knowledge by description.3 During this period, Russell conceives of acquaintance as a special epistemic relation holding between two distinct relata, where one constituent of the relational fact is a mental subject and another is an object of the subject’s awareness. In TK, he identifies this epistemic relation with direct experiential awareness: Now, since we have decided that experience is constituted by a relation, . . . we shall employ synonymously the two words “acquaintance” and “awareness,” generally the former. Thus when A experiences an object O, we shall say that A is acquainted with O. (35)

Russell holds that acquaintance is “the simplest and most pervading aspect of experience. .  .  . All cognitive relations—attention, sensation, memory, imagination, believing, disbelieving, etc.—presuppose acquaintance” (TK, 5).4 In fact, he maintains, “the faculty of being acquainted with things other than itself is the main characteristic of a mind. Acquaintance with objects essentially consists in a relation between the mind and something other than the mind; it is this that constitutes the mind’s power of knowing things” (POP, 66–67). Indeed, Russell follows Brentano (1874/2009) and James (1890/1950) in thinking it vital to draw a distinction between the psychological acts involving acquaintance and the objects toward which those acts are directed.5 At this time, Russell holds that subjects can be acquainted with relatively few kinds of mind-independent objects. Through sensation, subjects can be acquainted with external “sense-data” consisting of particular sensible objects, qualities, and relations presented within the egocentric space of their sensory experience (POP, 12 and 29–31).6 Through memory, subjects can be acquainted with previously experienced sense-data as well as various temporal relations (POP, 76). Through imagination, subjects can be acquainted with sense-data that need not be experienced as having occurred at any time whatsoever (TK, 59–63).7 And through conceiving, subjects can be acquainted with abstract universals and general principles, including the objects and relations of logic (POP, 81; TK, 97–101).8 When it comes to the relationship between knowing objects by acquaintance and knowing truths about them, there is compelling (though controversial) evidence that Russell endorses what Proops (2014) calls “the independence thesis.”9 According to the independence thesis, a subject’s possession of

Russell on Introspection and Self-Knowledge

259

knowledge of something by acquaintance neither presupposes nor entails that the subject knows any truths about it. In one clear statement of this thesis, Russell asserts: Knowledge of things, when it is of the kind we call knowledge by acquaintance, is essentially simpler than any knowledge of truths, and logically independent of knowledge of truths, though it would be rash to assume that human beings ever, in fact, have acquaintance with things without at the same time knowing some truth about them. (POP, 72)

Later in POP, he reiterates this claim: We may have knowledge of a thing by acquaintance even if we know very few propositions about it—theoretically we need not know any propositions about it. (225, emphasis added)10

In contrast, a subject’s possession of knowledge of something by description requires “as its source and grounds” antecedent possession of knowledge of truths about objects of his or her acquaintance, including truths about general principles (POP, 73). Many take Russell’s theory of knowledge to be motivated largely by a Cartesian quest for certainty. Such readings appear to be supported by his claims that subjects know the objects of acquaintance “perfectly and completely” and “just as they are” such that “no further knowledge of [them] is even theoretically possible” (POP, 73–74). Russell also says that it is not possible, without absurdity, to doubt the existence of such objects and that there is “no error involved” even in dreams or hallucinations unless subjects “go beyond the immediate object” (POP, 74, 172, and 235). These and other passages lead many to read him as holding that subjects can (at least sometimes) acquire absolutely certain knowledge about the existence, identity, features, and/or nature of the objects of immediate acquaintance.11 As I have argued elsewhere, there is strong evidence that such Cartesian interpretations are misguided.12 However, it is beyond the scope of the present chapter to rehearse those arguments. Instead, I will simply point to Russell’s remarks that “it is of course possible that all or any of our beliefs may be mistaken, and therefore all ought to be held with at least some slight element of doubt” and that “a theory which ignored this fact would be clearly wrong” (POP, 39–40 and 210). In any event, one should not uncritically presume that Russell’s accounts of introspection and self-knowledge are motivated by broader Cartesian concerns.

260

The Bloomsbury Companion to Bertrand Russell

But what then is motivating Russell’s acquaintance-based epistemology during this period? As Hyton (2003) notes, one of the main factors is his commitment to direct realism—the view that subjects can attain direct empirical knowledge of mind-independent reality without the mediation of conceptual structures (such as mental contents, ideas, or Sinne). Russell sees such direct realism as essential for challenging the Monistic Idealist claim that the conceptual structure of human thought modifies all empirical knowledge so that it provides only imperfect, partial, and at best approximately “true” knowledge of Absolute Reality (Hylton 2003, 207–09). Russell also sees it as necessary for explaining how human thought (and talk) can manage to connect up with, and provide piecemeal knowledge about, mind-independent reality. Indeed, at this time he is committed to what is often called Russell’s Principle, according to which “every proposition which we can understand must be composed wholly of constituents with which we are acquainted” (POP, 58). In defense of this thesis, he insists that “it is scarcely conceivable that we can make a judgement or entertain a supposition without knowing what it is that we are judging or supposing about. We must attach some meaning to the words we use, if we are to speak significantly and not utter mere noise” (POP, 91). Thus, the job of acquaintance is to determine (and put subjects in a position to know) the reference of thought at its most fundamental level, and thereby to supply subjects with the materials needed for thinking indirectly about the world beyond their personal experience (as well as about putative entities which do not, in fact, exist). In addition, Russell maintains that an acquaintance-based theory of knowledge is required for giving an adequate analysis of various cognitive phenomena such as consciousness, attention, belief, memory, imagination, and selectively based egocentric and demonstrative thought, among others (TK, 5, 31–32, and 39–41).

3  Knowing the mind by acquaintance Though much of Russell’s interest around 1911 to 1913 concerns how subjects are able to acquire knowledge of, and about, the mind-independent objects of sensation, memory, and conception, he also seeks to explain how they can have knowledge concerning the mind itself. As with the other sources of knowledge, his account is ultimately based on subjects having direct awareness of their

Russell on Introspection and Self-Knowledge

261

psychological acts through a distinctive form of introspective acquaintance. In POP, he introduces it as follows: We are not only aware of things, but we are often aware of being aware of them. When I see the sun, I am often aware of my seeing the sun; thus “my seeing the sun” is an object with which I have acquaintance. When I desire food, I may be aware of my desire for food; thus “my desiring food” is an object with which I am acquainted. Similarly we may be aware of our feeling pleasure or pain, and generally of the events which happen in our minds. (76–77)

Elsewhere, he says subjects can be introspectively acquainted with their own acts of sensation, attention, memory, imagination, believing, disbelieving, feeling, desiring, and willing, among others (POP, 80; TK, 5 and 34). When it comes to the minds of other people, in contrast, subjects can only know them by description based on indirect sensory evidence of others’ verbal and nonverbal behavior and similarity to themselves. Russell appeals to a number of different considerations in making the case for subjects having introspective acquaintance with their own minds. To begin with, he thinks it is supported by an argument to the best explanation of how knowledge concerning minds is possible at all. Given that subjects are not acquainted with the minds of others, he argues, such knowledge must be based largely on their acquaintance with their own minds. Without such awareness, he maintains, subjects would not be in a position to understand the minds of others even by description. Consequently, “we should be unable to imagine the minds of others, and therefore we could never arrive at the knowledge that they have minds” (POP, 77). In fact, he questions whether subjects could know they have minds—or so much as consider whether they exist—if not for introspective acquaintance (POP, 78).13 In addition, Russell takes the existence of introspective acquaintance to be an “obvious” or “plain” fact of experience (POP, 76–77, KAKD, 110, and TK, 33). In fact, he admits to not knowing how to prove it, “for I cannot think of anything more evident” (TK, 31). His only (slight) hesitation concerns its relational character, since neutral monists such as James (1912), Mach (1889/1984), and Perry (1912) deny the distinction between psychological acts and their objects (TK, 33). Even still, he insists that there is a need to explain why introspective awareness at least appears to be relational. Moreover, he believes that neutral monism faces daunting challenges which cast it in serious doubt (Pincock, this book). Among them is the fact that “it is obliged to have recourse to extraneous considerations, such as the nervous system, in order to explain the difference

262

The Bloomsbury Companion to Bertrand Russell

between what I experience and what I do not experience, and this difference is too immediate for any explanation that neutral monism can give” (TK, 31). One thing to notice about Russell’s view is that subjects do not invariably have introspective awareness of their psychological episodes. Rather, introspective awareness is something subjects “can” and “often” enjoy with respect to their thoughts, feelings, desires, and so on, through an act of “what may be called inner sense” (POP, 80).14 On such occasions, subjects are not only conscious of mind-independent objects, but also achieve a state of “self-consciousness” through which they are directly aware of the psychological episodes themselves (POP, 77–81). But, Russell suggests, such self-consciousness is seemingly absent in cases in which subjects do not attend to the objects of their experience (TK, 121). Furthermore, he argues that “it is logically evident that there must be instances [of being aware of something without being aware of such awareness], since otherwise every acquaintance would entail an infinite introspective series, which is absurd” (TK, 121). The issue of determining when exactly subjects enjoy self-consciousness is complicated, however, by the fact that Russell also thinks subjects can be directly aware of their psychological episodes without directing introspective attention to them. For instance, he says that subjects can have introspective awareness of faint and peripheral sensations, dim thoughts and desires, and various other conscious episodes at the margins of their attentive focus (TK, 8–9). In fact, he argues, the field of introspective acquaintance must extend beyond the limited range of introspective attention since such acts involve the selection of objects from a larger, already present field of awareness “out of which attention chooses what it wants” (TK, 9). In addition to being acquainted with particular psychological episodes, subjects can be introspectively aware of their qualities and various relations that hold between them. For instance, subjects can be introspectively aware of relations of similarity and difference between particular psychological acts. This allows subjects to become directly aware, through acts of conceiving, of the psychological kinds they fall under and various relational universals they instantiate.15 With the help of memory, subjects can also be directly aware of the temporal relations of the psychological episodes to themselves (in their private space of time) as well as their duration, simultaneity, and succession (TK, 64–78). Perhaps most importantly, subjects can be introspectively aware of the relation of “being experienced together” which holds between the various sensory, cognitive, and conative episodes that make up their psychological life

Russell on Introspection and Self-Knowledge

263

(TK, 8). For Russell, the awareness of such “being experienced together” relations explains the unity that subjects experience their minds as having at any one time. By extension, the experience of psychological unity across time is achieved by combining their present experience of psychological unity with memories of similarly unified thoughts and experiences at earlier times (TK, 12–13). It must be emphasized, however, that the direct introspective awareness subjects enjoy with respect to (at least some of) their psychological episodes and their features does not constitute knowledge of any truths about them. Nor does such introspective acquaintance, all by itself, entail any knowledge of truths about them. Rather, introspective acquaintance presents psychological episodes to the subjects who have them and thereby put those subjects in a position to think, talk, and acquire knowledge about them by exercising relevant cognitive capacities.

4  Inner perception and introspective judgments Russell does not (at least to my knowledge) make explicit how exactly subjects acquire knowledge of truths about their psychological episodes on the basis of being introspectively acquainted with them. However, he asserts in KAKD that “from the point of view of theory of knowledge, introspective knowledge is exactly on a level with knowledge derived from sight or hearing” (111). Given similar remarks Russell makes elsewhere, it is reasonable to reconstruct his view based on what he says concerning how subjects derive knowledge of truths from other forms of acquaintance, such as sensation, memory, and conceiving. When it comes to sensory knowledge, Russell draws a distinction between having direct awareness of objects through sensation, having direct awareness of complexes involving those objects and their features, and making judgments about the complexes of which subjects are directly aware.16 Regarding the latter two categories of sensory knowledge, Russell distinguishes between “truths of perception” and the “judgements of perception” which aim to express them (POP, 177–78).17 Both are kinds of “intuitive knowledge” which are noninferentially derived from sensory experience, but there are fundamental differences between them. The perception of truths is a matter of subjects being directly presented in experience with facts involving the existence of sense-data, the features they possess, and their relations to other sense-data (POP, 178). As such, the

264

The Bloomsbury Companion to Bertrand Russell

perception of truths should properly be considered a species of knowledge of things—in this case of sensory complexes—by acquaintance. Indeed, Russell remarks, “we may distinguish sensation from perception by saying that the former gives particulars while the latter gives facts” (TK, 37). He also makes this point clear in POP: Thus in regard to any complex fact, there are, theoretically, two ways in which it may be known: (1) by means of a judgment, in which its several parts are judged to be related as they are in fact related; (2) by means of acquaintance with the complex fact itself, which may (in a large sense) be called perception, though it is by no means confined to objects of the senses. (211, second emphasis added)18

In POP, he tentatively assumes that all such perception of truths is what he later calls “complex perception” (TK, 125). In complex perception, subjects are directly aware of both the complexes presented to them in perception and (at least some of) the constituents of those complexes. This contrasts with what he calls “simple perception” in which subjects are presented with complexes as wholes without being directly aware of any of their constituents individually. When it comes to judgments of perception, more is required of subjects than the receptive presentation of sensory complexes. Subjects must attend selectively to these complexes, in some cases analyze them, and ultimately form non-inferential judgments which aim to correspond to them. On Russell’s view, analyzing complexes requires selectively attending to both their constituents and the relations between them (TK, 123). Such analysis is a precondition for subjects being able to judge, explicitly or implicitly, that sense-data have particular sensible qualities (such as “this is round” or “this is red”) or stand in particular sensible relations to each other (such as “this is to the right of that”) (POP, 179). In contrast, Russell suggests that judgments about the existence of experienced sense-data (such as “there is that”) do not require analysis (POP, 179).19 Thus, subjects can make existential judgments about them without at the same time making any judgment about what kind of thing they are, what qualities they have, or what relations they bear to other things (even if subjects nearly always do so). Altogether, this account of sensory knowledge provides a well-developed model for how, at this time, Russell likely conceived of introspective knowledge. Such an account would draw a distinction between having direct awareness of psychological acts, having direct awareness of psychological complexes, and making judgments about them. Both of the latter kinds of knowledge

Russell on Introspection and Self-Knowledge

265

would be cases of intuitive knowledge that are non-inferentially derived from introspective experience. The first of these, which we might call “truths of inner perception,” would involve the receptive presentation of mental complexes to the subject through “inner perception.”20 As such, it is best thought of as introspective knowledge of mental complexes by acquaintance, rather than as a kind of knowledge of truths about them. If such inner perception directly presents them to subjects as wholes (without at the same time presenting their constituents individually), the result will be “simple inner perception.” On the other hand, if it also presents at least some of the complex’s constituents, the result will be “complex inner perception.” Russell initially supposes that subjects have complex inner perceptions of their psychological episodes, but gradually comes to think that subjects only have simple inner perceptions due to growing worries about whether subjects enjoy introspective self-acquaintance. As in the case of sensory knowledge, arriving at “judgments of inner perception” would involve the active exercise of various cognitive capacities. Subjects would have to attend selectively to the perceived mental complexes, in some cases analyze them in terms of their constituents, and issue non-inferential judgments which aim to correspond to them.21 The logically simplest of such judgments would concern the existence of these mental complexes and would take the form of an introspectively directed “there is this” (or perhaps even “lo!”). Such judgments would not, by themselves, logically entail more complex ones about what kinds of objects are being experienced, what features they have, or what relations they bear to other things. Subjects could only make judgments of this kind after selectively attending to (at least some of) the constituents of the mental complexes and analyzing them in terms of their features and relations to each other. Before moving on, it is worth noting that Russell’s account of introspective knowledge thus far does not exhaust the knowledge subjects can have regarding their psychological lives. For subjects can also acquire indirect knowledge of their psychological episodes by description, as well as derive additional knowledge of truths about them through the application of self-evident logical principles. The former of these is particularly vital for knowing of, and thereby being in a position to know about, unremembered past thoughts and experiences, predicted or imagined thoughts and experiences, and unconscious psychological episodes, among other things.22 What is more, the capacity for subjects to know of mental complexes and their constituents by description comes to play a key role in Russell’s developing views about self-knowledge during this period.

266

The Bloomsbury Companion to Bertrand Russell

5  Is there acquaintance with the self? On Russell’s early views (1911 to 1913), there are a number of different ways that subjects can think and talk about (and thereby be in a position to acquire knowledge about) themselves. First, subjects can think (or talk) about themselves by means of a description that they uniquely satisfy. For example, the right subjects can think (or talk) about “the philosopher who received the Nobel Prize in Literature in 1950,” “the future president of the United States in 2030,” “the featherless biped who wrote this chapter,” and so on, and manage to think about themselves because of things they have done or features they uniquely possess at some time. When subjects have this kind of knowledge of themselves, they have what Russell calls “merely descriptive knowledge” of the persons they happen to be (KAKD, 113). Second, subjects can think (or talk) about themselves using ordinary proper names that, as matters of linguistic convention, designate them. During this period, Russell treats such ordinary names as expressing, in a particular context, truncated definite descriptions that uniquely pick out their bearers. Sometimes these descriptions will be analogous to those mentioned above and include (somewhat vague and varying) unique historical or personal information about the persons they designate. However, in the most basic case ordinary names can designate subjects by means of descriptions of the form “the person named such-and-such” (KAKD, 119). In this way, Russell can think about himself as “the person called Bertie,” Bernie Sanders can think about himself as “the person called Bernie,” and so on. Because ordinary names express (contextually determined) truncated definite descriptions, subjects who have knowledge of themselves by means of ordinary names once again have merely descriptive knowledge of the persons they happen to be. On Russell’s view, neither of these kinds of knowledge captures the most basic, intimate knowledge that subjects have concerning themselves— knowledge which is best expressed in ordinary language with the contextsensitive expression “I.” For one thing, anyone with the appropriate descriptive knowledge of someone is on equal footing in thinking of them using descriptions such as these. In doing so, subjects are thinking of someone (who they happen to be) in a way that is characteristically used for thinking of persons other than themselves (and is, in fact, the only way of doing so). For another, subjects seem to have ways of thinking of themselves as themselves that are more basic than, and do not require, descriptive knowledge of this

Russell on Introspection and Self-Knowledge

267

kind. For example, John Perry can have knowledge of himself as himself even if he does not realize that he is, in fact, “the person leaving a trail of sugar in the grocery store,” or wrongly thinks he is “the brooding vigilante superhero of Gotham City.” Likewise, he can think of himself as himself even if he forgets that the description “the person named John Perry” designates him, or acquires the delusional belief that the description “the person named Bruce Wayne” does.23 In KAKD, Russell defends the view that the most basic form of self-knowledge subjects possess is introspective knowledge of themselves by acquaintance. In particular, he maintains that subjects can be directly aware of themselves through acts of complex inner perception directed at their own psychological episodes. He does hesitate, however, when it comes to the question of whether subjects can have introspective acquaintance with themselves as bare particulars, independent of such psychological episodes. Indeed, he says, “it is hard to discover any state of mind in which I am aware of myself alone, as opposed to a complex of which I am a constituent” (KAKD, 110). But, he insists, it is “very difficult to account for plain facts if we assume that we do not have acquaintance with ourselves” (KAKD, 110).24 The chief difficulty arises when trying to account for how subjects can understand their own introspective judgments. Recall that at this time, Russell is committed to Russell’s Principle, according to which “whenever a relation of supposing or judging occurs, the terms to which the supposing or judging mind is related by the relation of supposing or judging must be terms with which the mind in question is acquainted” (KAKD, 118). So given that psychological episodes are complexes involving a subject, a psychological act, and an object, subjects can understand their introspective judgments only if they are acquainted with themselves or with all of the elements of an introspective-based description of themselves (KAKD, 110). However, he argues that there is a formidable problem with supposing that, upon analysis, the introspective judgments subjects make involve only self-descriptions, rather than self-acquaintance. The problem is this: If we wished to maintain the view that there is no acquaintance with Self, we might argue as follows: We are acquainted with acquaintance, and we know that it is a relation. . . . Hence we know that this complex must have a constituent which is that which is acquainted, i.e. must have a subject-term as well as an object-term. This subject we define as “I.” Thus “I” means “the subject-term in awareness of which I am aware.” (KAKD, 110)

268

The Bloomsbury Companion to Bertrand Russell

Russell’s worry here is that if subjects only know themselves by description roughly as “the subject-term of the psychological episode with which I am introspectively acquainted,” they will in turn need to understand the meaning of the “I,” properly analyzed, as it occurs in this description. But if the subject designated by that descriptive element is known only by an analogous description, and so on, then understanding an introspective judgment would involve an infinite regress. The best way to block that regress, Russell reasons, is if at some level of analysis the descriptive elements eventually bottom out in introspective self-acquaintance (KAKD, 110).25 By the time of POP, however, Russell becomes less confident that subjects do, in fact, enjoy introspective self-acquaintance. His growing hesitation stems largely from Humean worries about whether subjects ever directly experience themselves in introspection. Indeed, he remarks, “when we try to look into ourselves we always seem to come upon some particular thought or feeling, and not upon the ‘I’ which has the thought or feeling” (POP, 78). Thus, when it comes to acquaintance-based introspective knowledge, “it is our particular thoughts and feelings that have primitive certainty,” rather than knowledge of the self (POP, 30). Nevertheless, Russell asserts, “there are some reasons for thinking that we are acquainted with the ‘I’, though the acquaintance is hard to disentangle from other things” (POP, 78). To begin with, he suggests that it simply seems evident, upon careful reflection, that subjects can be acquainted with themselves in inner perception given the relational character of psychological acts (POP, 79). What is more, there is no serious question of who the subject of the psychological episode is: When a case of acquaintance is one with which I can be acquainted (as I am acquainted with my acquaintance with the sense-datum representing the sun), it is plain that the person acquainted is myself. Thus, when I am acquainted with my seeing the sun, the whole fact with which I am acquainted is “Self-acquaintedwith-sense-datum.” (POP, 79, emphasis added)

What Russell seems to be suggesting here is that the high degree of certainty subjects have that they, rather than anyone else, are the subject of the introspected psychological acts is difficult (if not impossible) to explain on the assumption that their self-knowledge is merely descriptive. Russell also gestures at something like the regress argument he offers previously in KAKD, though with somewhat less confidence in its persuasiveness.

Russell on Introspection and Self-Knowledge

269

Indeed, he argues, “we know the truth ‘I am acquainted with this sense-datum’. [But it] is hard to see how we could know this truth, or even understand what is meant by it, unless we were acquainted with something which we call ‘I’” (POP, 79–80). Once again, the worry concerns how Russell’s Principle can be satisfied if subjects lack introspective self-acquaintance (POP, 58). However, he grants, “the question [of whether subjects can be acquainted with themselves] is difficult, and complicated arguments can be adduced on either side. Hence, although acquaintance with ourselves seems probably to occur, it is not wise to assert that it undoubtedly does occur” (POP, 79–80). Despite his continuing belief in the likelihood of self-acquaintance, he becomes increasingly less sure that these selves are long-lived, persisting things such as persons. Answering the regress argument, he notes, does not require “that we are acquainted with a more or less permanent person, the same to-day as yesterday, but it does seem as though we must be acquainted with that thing, whatever its nature, which sees the sun and has acquaintance with sense-data” (POP, 80). Similarly, he denies that acquaintance with a persisting self, rather than a momentary one, is essential for explaining the high degree of confidence subjects have that they, rather than anyone else, are the subjects of their introspective judgments (POP, 29–30). In fact, he argues that the lack of introspective evidence for the existence of a persisting self blunts some of the force of Descartes’ well-known cogito argument: “I think, therefore I am” says rather more than is strictly certain. It might seem as though we were quite sure of being the same person to-day as we were yesterday, and this is no doubt true in some sense. But the real Self is as hard to arrive at as the real table, and does not seem to have that absolute, convincing certainty that belongs to particular experiences. (POP, 29)26

So while Russell continues to think that subjects likely have self-acquaintance in POP, he is inclined to think that persons know of themselves as persisting things only by description—roughly as those things (whatever their nature may be) that have the continuing psychological lives they, as momentary subjects, are aware of through introspection and memory. By the time of TK, however, “the elusiveness of the subject in introspection” convinces Russell that subjects likely lack introspective awareness of both the persisting and momentary self (36). Instead, subjects only know of them­ selves as themselves by description roughly as “the subject of such-and-such

270

The Bloomsbury Companion to Bertrand Russell

introspected psychological episodes” (TK, 35). Put somewhat differently, he says: Our theory maintains that the datum when we are [introspectively] aware of experiencing an object O is the fact “something is acquainted with O.” The subject appears here, not in its individual capacity, but as an “apparent variable”; thus such a fact may be a datum in spite of incapacity for acquaintance with the subject. (TK, 37)

Of course, given the relational character of all acquaintance, the act of selfconsciousness itself will involve both a subject and the psychological episode experienced. Thus, Russell remarks, an act of self-consciousness can be represented symbolically as “S′—A—[(∃S) • (S—A—O)],” where S is the subject of the introspected episode, O is the object of that episode, A is acquaintance, and S′ the subject introspecting the episode (TK, 38).27 There are a number of issues raised by Russell’s newly adopted denial of selfacquaintance. One might worry that while the subject S of the introspected psychological episode is known only by description, the subject S′ doing the introspecting does not appear in this way in the represented complex. Thus, it might strike one that self-acquaintance is required after all for subjects to make introspective judgments about their psychological episodes. However, Russell is quick to reiterate his claim in POP that episodes involving acquaintance need not themselves be known by acquaintance, on pain of an infinite regress (TK, 39). Russell must also answer his former objection that introspective selfknowledge in the absence of self-acquaintance would require the use of the regress-producing self-description “the subject of the psychological episode I am aware of having.” He addresses this challenge largely by reconceiving the relationship between the meaning of “I” and the subject’s present psychological episodes. Indeed, he argues, “we might suppose that ‘my present experience’ might be defined as all the experience which ‘I’ have ‘now’. But in fact we shall find that ‘I’ and ‘now’, in the order of knowledge, must be defined in terms of ‘my present experience’, rather than vice versa” (TK, 8). As Russell now sees things, the meaning of “I” derives from the subject’s ability to be directly aware of his or her psychological episodes as wholes (via simple inner perception) and to designate them with proper names (TK, 39). Thus, the meaning of “I” as used by a subject in an introspective judgment is roughly “the subject of this presented psychological episode (and others co-experienced with it)” where the meaning of the “this” is simply given to the subject without any need for him or her to

Russell on Introspection and Self-Knowledge

271

understand that it is so given (TK, 40). In this way, the alleged regress involved in introspective self-description is blocked.28 Russell’s view that subjects lack self-acquaintance, despite having introspective acquaintance with their ongoing conscious mental lives, has far-reaching consequences for his views about what subjects can know about themselves. For one thing, he reasons: If it is true, as it seems to be, that subjects are not given in acquaintance, it follows that nothing can be known as to their intrinsic nature. We cannot know, for example, that they differ from matter, nor yet that they do not differ. They are known merely as referents for the relation of acquaintance, and for those other psychical relations—judging, desiring, etc.—which imply acquaintance. (TK, 37)

What is more, he continues to hold that introspection cannot reveal whether the successive psychological episodes of a single mind share a single persisting subject or different momentary ones (TK, 35). In fact, he argues that there is no good reason to think that the subject of an act of self-consciousness is identical with the subject of the psychological episode being introspected, even when they are contemporaneous (TK, 38). Indeed, he continues, “the one ‘self ’ or ‘mind’ which embraces both may be a construction, and need not, so far as the logical necessities of our problem are concerned, involve any identity of the two subjects” (TK, 38–9). This suggestion, that selves and minds might be logical constructions, soon becomes central to Russell’s thought on these topics.

6  Constructing the self Shortly after completing POP, Russell begins to explore how the techniques of logical construction, which were central to the logicist project of Principia Mathematica, might also be fruitfully applied to physics and the empirical sciences, more broadly.29 Russell’s change in course undeniably results from his growing unease about inferring material objects from sense-data, but there is much debate about what exactly is behind his preference for logical constructions over inferred entities. Some think that he sees logical constructions as providing a better tool for answering traditional skepticism (Sainsbury 1979; Graham forthcoming). Others interpret him as adopting logical constructions to explain how knowledge of physics is possible while better respecting the principle of acquaintance (Pears 1987; Hylton 1990; Baldwin 2001). Still others take his

272

The Bloomsbury Companion to Bertrand Russell

theory of logical constructions to be a linguistic doctrine according to which the meaning of all talk ostensibly about physical objects is really about sensedata and nothing more (Soames 2003 and 2014). However, as Wahl has cogently pointed out, none of these interpretations is plausible given Russell’s appeal to both unsensed sensibilia and the sense-data of others in the logical construction of matter (2015, 98–100). A more compelling interpretation is that Russell prefers the ontological parsimony of a physics logically constructed from sense-data and unobserved entities of the same kind to one requiring inferred entities of a fundamentally different, and unobservable, kind (Bostock 2012; Wahl 2015). The advantage of logical constructions, then, is that they supply a means for paring down the kinds of entities to which physics is ontologically committed. This reading also receives support from his frequent designation of the objects of logical construction as “logical fictions,” suggesting a penchant on his part to deny their existence.30 To take one notable example from his 1918 Philosophy of Logical Atomism (hereafter PLA), Russell asserts that statements ostensibly about Piccadilly, when properly analyzed, will not express propositions including it as a constituent, but will instead include the series of suitably related sense-data and unsensed sensibilia which are normally presumed to be appearances caused by a certain region of the Earth’s surface (51). But, he remarks, “I believe that series and classes are of the nature of logical fictions: therefore that thesis, if it can be maintained, will dissolve Piccadilly into a fiction” (PLA, 51). If the elimination of unnecessary kinds of inferred entities is indeed the primary motivation behind Russell’s logical constructions, it would have radical consequences for his views about knowledge of the “self ” and “mind.” As noted above, he suggests, but does not endorse, the possibility of logically constructing both in TK. But by the time of his 1915 “The Ultimate Constituents of Matter” (hereafter UCM), he embraces this view. In this paper, Russell relays his first cinematic experience of seeing what appears to be a persisting man but which, in reality, is “a [continuous] succession of films, each with a different momentary man” (UCM, 129). Praising the cinema in this respect as “a better metaphysician than common sense, physics, or philosophy,” Russell asserts that “the real man too . . . is really a series of momentary men, each different one from the other, and bound together, not by numerical identity, but by continuity and certain intrinsic causal laws” (UCM, 129). Thus, persons, like other ordinary objects, can be treated as series of suitably arranged momentary particulars “collected together on account of some property which makes it convenient to be able to

Russell on Introspection and Self-Knowledge

273

speak of them as wholes, [and] are what I call logical constructions or symbolic fictions” (UCM, 129). These remarks, and similar ones made elsewhere, lead some to read Russell as flatly denying the existence of the self (whether momentary or persisting), and ipso facto any possibility of genuine self-knowledge (Sainsbury 1979; Olson 2007). One problem for reading Russell as eliminating the self, as Sainsbury (1979) points out, is that it conflicts with his ongoing commitment to the relational nature of psychological acts and mental facts (192). Thus, either Russell does not recognize this tension in his views about the self and mind, or his description of them as logical constructions is not motivated primarily by considerations of ontological parsimony. While Sainsbury sees the former as more likely, there is good reason to prefer the latter interpretation. As I read him, Russell’s logical constructions are largely guided by epistemological concerns other than a desire to secure certain knowledge in the face of skepticism. Rather, the aim of logical construction is to reinterpret a body of knowledge ostensibly about unobservable kinds of entities entirely in terms of a smaller number of observable kinds of entities (Hylton 2015). For Russell, there are several important benefits of doing so. First, to the extent that a body of knowledge can be reinterpreted in terms of observable kinds of entities (even if they are not, in fact, observed by anyone), it becomes possible, in principle, for the truth of that body of knowledge (so interpreted) to be verifiable.31 Second, it also explains how the body of knowledge (so interpreted) can be knowable, in principle, to subjects, even if it does not describe its actual psychological origin.32 Third, it reduces the size of the apparatus needed to account for this knowledge, thereby diminishing (but not eliminating) its risk of error (PLA, 154). At the same time, it reduces the extent to which the truth of the body of knowledge is held hostage to the existence of unobservable kinds of entities.33 Thus, it makes it possible to accept the body of knowledge (so interpreted) while sidestepping contentious metaphysical debates about the nature of unobservable kinds of entities and the legitimacy of the principles used to infer their existence.34 In doing so, it renders possible the elimination of the logically constructed entities, but does not, all by itself, give grounds for doing so.35 When it comes to the logical construction of the mind and self, then, Russell’s primary aim is not to deny (nor affirm) the existence of the metaphysical subject or ego. Rather, it is to reinterpret the body of knowledge subjects have concerning themselves and other people into a smaller apparatus that is less risky, more apt for empirical verification, and less prone to entanglement in contentious

274

The Bloomsbury Companion to Bertrand Russell

metaphysical disputes. On Russell’s view, empirical knowledge of other persons can be logically constructed from series of suitably related sensory appearances they present, rather than in terms of a persisting ego since “whether there be such a persistent subject or not, [that] is certainly not a datum” (PLA, 149). In a similar manner, subjects’ empirical knowledge of their own minds can be logically constructed from series of suitably related thoughts, memories, sensations, and other psychological episodes presented introspectively without any need to assume a metaphysical self (or selves) to whom they are presented (PLA, 149).36 Lastly, when it comes to these subjects’ “selves,” such knowledge can be logically constructed from series of suitably related psychological episodes and sensory appearances together (PLA, 149–50). In a subject’s own case, then, “you have a much richer material and are therefore much less likely to be mistaken as to your own identity than as to someone else’s” (PLA, 149). So while Russell notes that “there are [sometimes] mistakes even as to one’s own identity, in cases of multiple personality and so forth,” he also asserts that “as a rule you will know … that it is you, not by consciousness of the ego at all but by all sorts of things, by memory, by the way you feel and the way you look and a host of things” (PLA, 149). But despite the fact that the self and mind can be treated as logical constructions for most empirical purposes, Russell insists that “we shall not deny that there may be a metaphysical ego” (PLA, 150).37 In fact, there is reason to think that he continues to accept, however cautiously, the existence of the metaphysical ego (whether momentary or persisting) as an inferred entity. Russell’s primary grounds for doing so are the difficulties, noted in sections two and three above, in accounting for the difference between what someone experiences and does not experience, the selectiveness of attention, and egocentric and demonstrative thought, among other things (PLA, 86–87 and 153). However, he eventually confesses to the possibility that these worries might “be solved by ingenuity” and that these issues are “so delicate and so subtle that I cannot feel quite sure whether [this set of objections] is a valid one or not” (153). Thus, while Russell is arguably not completely agnostic about the existence of the metaphysical self or ego by the time of PLA, as Pears (1967) seems to suggest, he is (at least) very nearly so.

7  Toward neutral monism Taking stock, Russell’s theory of introspection from roughly 1911 to 1918 is grounded in the threefold distinction between direct awareness of psychological objects, the presentation of psychological complexes, and introspective

Russell on Introspection and Self-Knowledge

275

judgments about these complexes. When it comes to self-knowledge, his initial view is that subjects enjoy self-acquaintance, but he gradually comes to think subjects merely know themselves by introspective self-description, and he ultimately arrives at the view that “selves” and “minds” are, as far as our evidence goes, logical constructions. There is good reason to think that these changes are not driven by a Cartesian quest for certainty and that his preference for logically constructing the self and mind over inferring them is not primarily motivated by a taste for ontological parsimony. Indeed, Russell seemingly continues to cautiously accept the existence of the metaphysical self in 1918 due to the difficulties in explaining the selectiveness of the mind and egocentric thought, and because the relational character of psychological acts is introspectively observable even if the self is not. Russell’s views about introspection and self-knowledge begin to undergo significant changes in his 1919 “On Propositions” (hereafter OP) and 1921 The Analysis of Mind (hereafter AMi). In OP, Russell at last comes to the view that neither mental subjects nor psychological acts are discoverable in introspection (25). At the same time, he grows increasingly confident that neutral monism can account for both the selectivity of attention and egocentric and demonstrative thought (Landini 2011; Pincock, this book). As a result, he concludes that there are neither empirical nor theoretical grounds for drawing a distinction between subjects, psychological acts, and their objects. Instead, he embraces the view that both “minds” and “matter” are constructions composed of transitory elements that are intrinsically neither mental nor physical but are rendered one or both depending on whether they are parts of physical and/or psychological processes. Consequently, he comes to believe that “persons” and “selves” are also bundles of transitory neutral elements bound together by the right kinds of spatiotemporal and causal relations.38 It is beyond the scope of this chapter to survey the many changes that result from Russell’s adoption of neutral monism. However, it is worthwhile to mention, all too briefly, some of the more significant ones regarding introspection and selfknowledge. Whereas he previously conceived of sensory experience, for instance, as a matter of subjects engaging in an act of sensation directed toward mindindependent sense-data, he comes to see it as the occurrence of qualitied sensory events (or “percepts”) in the brain that are causally connected to physical events outside the brain. As a result, he comes to deny that there is a sharp distinction between awareness of sensory experience and the perception of the external world, the key difference only being the degree of privacy involved and their position in the causal process (AMi, 118–21). As for introspective judgments,

276

The Bloomsbury Companion to Bertrand Russell

Russell adopts the view that they are a matter of belief-image events in the brain being causally connected to (and often in some way resembling) sensory events elsewhere in the brain (AMi, 116–22).39 In this respect, introspective judgments are not essentially different in kind from other sorts of judgments and are “not, in isolation, in any degree more trustworthy than ‘external’ perception” (AMi, 123). Thus, not only does Russell’s neutral monism aim at closing the gap between “mind” and “matter,” it also aims at doing so regarding introspective knowledge and knowledge of the world more generally. Much more can be said about each of these issues, but that must be a project for another time (though see Pincock, this book).40

Notes 1 In TK, Russell acknowledges that “a certain difficulty as regards the use of words is unavoidable. . . . The meanings of common words are vague, fluctuating and ambiguous, like the shadow thrown by a flickering streetlamp on a windy night; yet in the nucleus of this uncertain patch of meaning, we may find some precise concept for which philosophy requires a name. . . . Sometimes it will be well to introduce a new technical term, sometimes it will be better to polish the common word until it becomes suitable for technical purposes” (6). 2 In saying that knowledge of something puts one in a position to think and talk about it, I do not mean to imply that all other preconditions for thought and talk about it are met. A subject might also need to attend to it, demonstrate it with a proper name, analyze it, or otherwise exercise relevant cognitive capacities. The key (trivial) point is that a subject cannot in any way cognize something of which he or she has no knowledge whatsoever. 3 Proops (2014) suggests that Russell’s key divergence from how James (1890/1950) understands the distinction between knowledge of things and knowledge of truths is that he allows, whereas James does not, that subjects can have genuine knowledge of, and therefore be in a position to know about, objects outside of their acquaintance. While James grants that subjects can in some sense conceive of things beyond their experience, he insists that such conceptions are “hollow and inadequate” and generate a “false conceit of knowledge” (1890/1950, II, 7). 4 Interestingly, later in TK Russell seems to leave open the possibility of both acquaintance and mental subjects being further analyzable, saying only that he has “no analysis to suggest, and therefore formally both will appear as if they were simple, though nothing will be falsified if they are found to be not simple” (45).

Russell on Introspection and Self-Knowledge

277

5 James remarks that it is a shortcoming of the English language that it does not explicitly draw “the generic distinction between the-thing-thought-of and thethought-thinking-it, which in German is expressed by the opposition between Gedachtes and Gedanke, in Latin by that between cogitatum and cogitation” (1890/1950, I, 195). This shortcoming, of course, is in addition to the failure of English to mark the difference between knowledge of things and knowledge about things. For an excellent discussion of Russell’s introduction to Brentano through his teacher G. F. Stout, see Nasim 2008. 6 Russell is clear in POP that sense-data are external to the mind but is frustratingly silent about their nature. Early in this work, he says that sense-data are at most “signs of some property [inherent in a physical object] which perhaps causes all the sensations, but is not actually apparent in any of them” (16). He goes on to suggest that they likely correspond in some way to the relational structure of physical objects in physical space, but are seemingly themselves neither mental nor physical (49–50). However, it could be that Russell already implicitly sees sense-data as transitory physical particulars that are signs of the unobserved material continuants of physics, as Landini 2011 claims (238–39). 7 See Carey 2015 for an in-depth discussion of Russell’s changing thought about acquaintance via memory, hallucination, dreams, and imagination and how they influence his views about physical knowledge. 8 This is by no means an exhaustive list of the ways subjects can be acquainted with external sensory and abstract objects. Subjects can also fear them, desire them, love them, and hate them, among many other possibilities. See Klement 2015 for an excellent discussion of Russell’s views about logical objects. 9 Also see Amijee 2013, Linsky 2015, Proops 2015, and Wishon 2017. 10 As Amijee and others have noted, Russell also expresses this claim in a 1911 letter to Gilbert Murray, the Home University Library editor for POP: “Acquaintance with a thing does not (theoretically) involve any knowledge of truths about the thing, and in practice involves often very little such knowledge” (Amijee 2013, 1183). 11 Among the many interpreters who read Russell this way are Baldwin 2001, Bostock 2012, Campbell 2009, Evans 1983, Geach 1957, Jeshion 2010, Ludlow 2013, Irvine 2015, and Soames 2010. 12 See Wishon 2015 and 2017. Also see Landini 2011, Linsky 2015, Proops 2015, and Wahl 2015. To my knowledge, Russell’s earliest statement that subjects cannot achieve absolute certainty even with perceptual judgments about what is currently given in sensory experience occurs in his 1910 Philosophical Essays (hereafter PE) (182–83). 13 Russell goes on to suggest that introspective awareness of psychological episodes is perhaps what distinguishes conscious human persons from sentient nonhuman animals (POP, 77–78).

278

The Bloomsbury Companion to Bertrand Russell

14 One should resist the temptation to understand “inner sense,” as Russell is here loosely using the phrase, as either a literal act of looking into the theater of the mind (since psychological episodes are relations between subjects and things outside the mind) or in terms of causally mediated internal scanning or tracking mechanisms. This passage also shows that Russell rejects the view that whenever subjects bear an acquaintance relation to anything, they are at the same time acquainted with their acquaintance. This view is defended by Brentano (1874/2009), Meinong (1910), and many in the phenomenological tradition (most notably Husserl 1913/1982 and Sartre 1943/1948). 15 See KAKD, 111; POP, 158–61; and TK, 79–89. We must be careful to note that Russell does not think that knowledge of universals can be acquired from acquaintance with particular instances of psychological relations and their particular similarities and differences alone. As his regress argument against nominalism demonstrates, he thinks subjects must also be directly acquainted with the bare universal relation of resemblance at the very least (POP, 149–51). In TK, Russell suggests that we are likely directly aware of other bare universal relations as well (79–85). For more on Russell’s regress argument for universals, see Perovic 2015. 16 Russell leaves open whether the correspondence of complexes with facts or truths is a matter of identity or of some other one-one relation (TK, 79–80). I here follow Russell in treating them as identical for ease of exposition. 17 To my knowledge, Russell first draws this distinction in his 1910 “On the Nature of Truth and Falsehood” in PE, but he does not therein discuss its applicability to introspective knowledge (181–83). 18 Similarly, Russell claims that subjects can have “perceptions” of past complexes presented in memory as well as of abstract complexes presented in a kind of intellectual experience, among other possibilities. 19 Russell is here following James who also suggests that the most basic judgments of perception have roughly the form of “there is this.” However, James argues that even perceptual judgments of this kind “would perhaps be too discriminative, and the intellectual acknowledgement [of the existence of the objects of sensation] would [in its most basic form] be better expressed by the bare interjection ‘lo!’” (1890/1950, II, 8). 20 Once again, we must be careful not to understand “inner perception” here as a literal act of perception directed inwards (given the relational character of psychological episodes) or in terms of causally mediated internal scanning or tracking mechanisms. 21 As with all judgments, Russell allows that these introspective judgments can fail to correspond to the mental complexes they are about, and so they do not enjoy absolute certainty. It is worth noting how Russell’s views about introspective

Russell on Introspection and Self-Knowledge

22

23 24

25 26

27 28

279

knowledge here differ from Brentano’s (as I read him). First, as noted above, Brentano thinks that every psychological act involves, in itself, both awareness of something and awareness of the act itself (1874/2009, 107). Second, Brentano does not distinguish between inner perception and introspective judgments since, on his view, the former are already cognitions (111). Third, he argues that every act of awareness is accompanied by an inner perception of it (in addition to the reflexive awareness involved in that very act) (111). And fourth, Brentano holds that it is immediately evident that inner perceptions are infallible and absolutely certain (107–11). Russell’s views at this time concerning the full scope and nature of unconscious psychological episodes, if there are any, are somewhat unclear. But he does suggest that subjects, like chickens and horses, engage in unconscious reasoning (POP, 97–98 and 175). By the time of his 1915 “The Ultimate Constituents of Matter,” Russell is more explicit about the existence of unconscious psychological phenomena: “Psychologists point out how much of what we think we see is supplied by association or unconscious inference, how much is mental interpretation, and how doubtful is the residuum which can be regarded as crude datum” (126–27). See Perry 1979, 2009, and 2012. Russell amends this passage in the revised version of KAKD that appears in his 1919 Mysticism and Logic and Other Essays to say, “It is difficult, but probably not impossible, to account for plain facts if we assume that we do not have acquaintance with ourselves” (Papers 6, 149). As Pincock (this book) notes, Russell adds “or to find some other analysis of selfconsciousness” in the revised version of KAKD (Papers 6, 149). As noted previously, I think there is good reason to resist interpreting Russell’s remarks here, as elsewhere, as part of a Cartesian defense against skepticism. See Landini 2011, 300–02 and Wishon 2017. As will be discussed below, Russell warns against assuming that the subjects S and S′ are one and the same. Bostock (2012) challenges Russell’s view about self-knowledge in TK from a somewhat different direction. His question is this: “What grounds does one have for supposing that acquaintance is a two-term relation, if one of its two terms is never experienced?” (172). Russell’s response at the time would be that subjects can be introspectively acquainted with their psychological acts, and thereby be in a position to know their relational character, even if they lack self-acquaintance. Indeed, in a review of James’ Essays in Radical Empiricism, Russell remarks: “On the grounds of the purest empiricism, from mere inspection of experience, I for my part should hold it obvious that perception is in its intrinsic nature a fact of relation, involving an act as well as an object” (Papers 6, 303).

280

The Bloomsbury Companion to Bertrand Russell

29 Russell’s initial foray into this topic occurs in his unpublished 1912 manuscript “On Matter” (Papers 6). 30 See Klement 2013 and Linsky 2003 and 2014 for more detailed discussions of Russell’s notion of “logical constructions” and its relation to the notions of “incomplete symbols” and “logical fictions.” 31 See Russell’s “The Relation of Sense-Data to Physics” (hereafter RSP), 145–47; Our Knowledge of the External World (hereafter OKEW), 110–11; and PLA, 144. One might challenge Russell on the issue of whether the verifiable truth of the reinterpreted body of knowledge has any bearing on the truth of the original body of knowledge. Settling those issues, however, is beyond the scope of this chapter. 32 See OKEW, 105 and 140–41; and PLA, 146. 33 See RSP, 155–56; OKEW, 107; and PLA, 144. 34 Of course, it leaves untouched the unavoidable metaphysical and epistemological debates about whether the unobserved world actually consists of the same kinds of entities as those subjects observe, assuming it exists at all. 35 Russell is explicit on this point in PLA when, after proposing the possible logical construction of material objects, he asserts, “I want to make clear that I am not denying the existence of anything [in calling it a logical fiction]; I am only refusing to affirm it. I refuse to affirm the existence of anything for which there is no evidence, but I equally refuse to deny the existence of anything against which there is no evidence” (146). It is also worth bearing in mind Russell’s warning in “On Matter” that “if it were known that the universe had been created for the purpose of delighting mathematicians, there would be some reason to suppose that, of two hypotheses which both fit the facts, the simpler is more likely to be true. As, however, there is no evidence that this is the purpose of the universe, there is no reason whatever to expect the [universe and its workings to be parsimonious]” (Papers 6, 86). 36 Presumably, a subject’s knowledge of the minds of others would be logically constructed from the hypothetical series of suitably related thoughts and experiences the subject would attribute to them based on sensory appearances (including of observable behavior) and analogy with the subject’s own case (OKEW, 96). For a recent criticism of Russell’s views about inferring or constructing knowledge of others, see Kremer 2015. 37 There is one notable possible exception to the empirical adequacy of logical constructions of the self and mind: if introspection reveals psychological episodes to be relational, then any logical construction of them that ignores the role of the subject will leave this empirically discoverable fact unaccounted for. Incidentally, it is Russell’s coming to doubt the availability of introspective evidence for the relational character of psychological episodes (in addition to the subject) in his 1919 “On Propositions” that finally convinces him to embrace (a partial) neutral monism. See Wishon 2015.

Russell on Introspection and Self-Knowledge

281

38 For more on Russell’s treatment of objects as bundles, see Maclean 2014. 39 It is not quite accurate to say that image events are in the brain in OP and AMi since at the time Russell thinks they are parts of psychological processes but not physical ones. But within just a few years he adopts, and never thereafter abandons, his notorious belief that both images and percepts are brain events. It is also worth noting that Russell comes to believe that subjects do not enjoy introspective awareness of their desires or impulses and that “the discovery of our own motives can only be made by the same process by which we discover other people’s, namely, the process of observing our actions and inferring the desire which could prompt them” (AMi, 31). For more on Russell’s neutral monism, see Bostock 2012; Landini 2011; Pincock, this book; Stubenberg 2015 and 2016; Tully 2003; and Wishon 2015. For more on Russell’s neutral monist theory of desire, see Griffin 2015. 40 Many thanks to Matt Duncan, Sabastien Gandon, Kevin Morris, John Perry, Galen Strawson, Leopold Stubenberg, Russell Wahl, and anyone else I might be forgetting for helpful discussions of various topics in this chapter. I am also grateful to the audience at Rhode Island College and the graduate students in my 2017 seminar on Russell. Research for this chapter was supported by a College of Liberal Arts Summer Research Grant at the University of Mississippi.

References Amijee, Fatema. 2013. “The Role of Attention in Russell’s Theory of Knowledge.” Sartre British Journal for the History of Philosophy 21 (6): 1175–93. Baldwin, Thomas. 2001. “Bertrand Russell (1872–1970).” In Blackwell Companions to Philosophy: A Companion to Analytic Philosophy, edited by A. P. Martinich and David Sosa, 21–44. Malden, MA: Blackwell. Bostock, David. 2012. Russell’s Logical Atomism. Oxford: Oxford University Press. Brentano, Franz. 1874/2009. Psychology from an Empirical Standpoint. Translated by A. C. Rancurello, D. B. Terrell, and L. L. McAlister. New York: Routledge. Campbell, John. 2009. “Consciousness and Reference.” In The Oxford Handbook to Philosophy of Mind, edited by Brian McLaughlin, Ansgar Beckermann, and Sven Walter, 648–62. Oxford: Oxford University Press. Carey, Rosalind. 2015. “Seeing, Imagining, Believing: From Problems to Theory of Knowledge.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited byDonovan Wishon and Bernard Linsky, 107–27. Stanford, CA: CSLI Publications. Crane, Tim. 2012. “Tye on Acquaintance and the Problem of Consciousness.” Philosophy and Phenomenological Research 84 (1): 190–98. Evans, Gareth. 1983. The Varieties of Reference. New York: Oxford University Press.

282

The Bloomsbury Companion to Bertrand Russell

Geach, Peter. 1957. Mental Acts. London: Routledge and Kegan Paul. Graham, Peter. forthcoming. “Russell’s Logical Construction of the External World.” In Skepticism: From Antiquity to the Present, edited by Diego Machuca and Baron Reed. New York: Bloomsbury. Griffin, Nicholas. 2015. “Russell’s Neutral Monist Theory of Desire.” Russell: The Journal of Bertrand Russell Studies 35 (1): 5–28. Husserl, Edmund. 1913/1982. Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy—First Book: General Introduction to a Pure Phenomenology. Translated by F. Kersten. The Hague: Nijhoff. Hylton, Peter. 1990. Russell, Idealism and the Emergence of Analytic Philosophy. Oxford: Oxford University Press. Hylton, Peter. 2003. “The Theory of Descriptions.” In The Cambridge Companion to Bertrand Russell, edited by Nicholas Griffin, 202–40. New York: Cambridge University Press. Hylton, Peter. 2015. “Problems of Philosophy as a Stage in the Evolution of Russell’s Views on Knowledge.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 25–44. Stanford, CA: CSLI Publications. Irvine, Andrew D. 2015. “Bertrand Russell.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Winter 2015 Edition. Available online: https://plato. stanford.edu/archives/win2015/entries/russell/. James, William. 1890/1950. The Principles of Psychology, vols. I and II. New York: Dover. James, William. 1912. Essays in Radical Empiricism. New York: Longmans, Green. Jeshion, Robin. 2010. “Singular Thought: Acquaintance, Semantic Instrumentalism, and Cognitivism.” In New Essays in Singular Thought, edited by Robin Jeshion, 105–40. New York: Oxford University Press. Klement, Kevin. 2013. “Russell’s Logical Atomism.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Spring 2016 Edition. Available online: https://plato.stanford.edu/archives/spr2016/entries/logical-atomism/ Klement, Kevin. 2015. “The Constituents of the Propositions of Logic.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 189–229. Stanford, CA: CSLI Publications. Kremer, Michael. 2015. “Russell on Acquaintance, Analysis, and Knowledge of Persons.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 129–52. Stanford, CA: CSLI Publications. Landini, Gregory. 2011. Russell. New York: Routledge. Linsky, Bernard. 2003. “The Metaphysics of Logical Atomism.” In The Cambridge Companion to Bertrand Russell, edited by Nicholas Griffin, 371–91. New York: Cambridge University Press.

Russell on Introspection and Self-Knowledge

283

Linsky, Bernard. 2014. “Logical Constructions.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Winter 2016 Edition. Available online: https://plato.stanford.edu/archives/win2016/entries/logical-construction/. Linsky, Bernard. 2015. “Acquaintance and Certainty in The Problems of Philosophy.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 65–85. Stanford, CA: CSLI Publications. Ludlow, Peter. 2013. “Descriptions.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta, Fall 2013 Edition. Available online: https://plato.stanford.edu/ archives/fall2013/entries/descriptions/. Mach, Ernst. 1889/1984. The Analysis of Sensations and the Relation of the Physical to the Psychical. Translated by C. M. Williams. La Salle: Open Court. Maclean, Gülberk K. 2014. Bertrand Russell’s Bundle Theory of Particulars. New York: Bloomsbury. Meinong, Alexius. 1910/1983. On Assumptions. Translated by James Heanue. Berkeley, CA: University of California Press. Nasim, Omar W. 2008. Bertrand Russell and the Edwardian Philosophers: Constructing the World. New York: Palgrave Macmillan. Olson, Eric. 2007. What Are We?: A Study in Personal Ontology. New York: Oxford University Press. Pears, David F. 1967. Bertrand Russell and the British Tradition in Philosophy. New York: Random House. Pears, David F. 1987. The False Prison: A Study of the Development of Wittgenstein’s Philosophy, Volume One. Oxford: Clarendon Press. Perovic, Katarina. 2015. “The Importance of Russell’s Regress Argument for Universals.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 171–88. Stanford, CA: CSLI Publications. Perry, John. 1979. “The Problem of the Essential Indexical.” Noûs 13 (1): 3–21. Perry, John. 2009. “Diminished and Fractured Selves.” In Personal Identity and Fractured Selves, edited by Debra J. H. Mathews, Hilary Bok, and Peter V. Rabins, Baltimore: The Johns Hopkins University Press. Perry, John. 2012. “Thinking about the Self.” In Consciousness and the Self: New Essays, edited by JeeLoo Liu and John Perry, 76–100. New York: Cambridge University Press. Perry, Ralph B. 1912. Present Philosophical Tendencies: A Critical Survey of Naturalism, Idealism, Pragmatism and Realism together with a Synopsis of the Philosophy of William James. London: Longmans, Green. Pincock, Christopher (this book, Chapter 12). “Neutral Monism.” In The Bloomsbury Companion to Bertrand Russell, edited byRussell Wahl. New York: Bloomsbury Academic. Proops, Ian. 2014. “Russellian Acquaintance Revisited.” Journal of the History of Philosophy 52 (4): 779–811.

284

The Bloomsbury Companion to Bertrand Russell

Proops, Ian. 2015. “Certainty, Error, and Acquaintance in The Problems of Philosophy.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 45–63. Stanford, CA: CSLI Publications. Russell, Bertrand. 1910. Philosophical Essays. New York: Longmans, Green. Russell, Bertrand. 1911. “Knowledge by Acquaintance and Knowledge by Description.” Proceeding of the Aristotelian Society 11: 108–28. Russell, Bertrand. 1911/1992. “Knowledge by Acquaintance and Knowledge by Description.” In The Collected Papers of Bertrand Russell, Volume 6: Logical and Philosophical Papers, 1909–13, edited by John G. Slater, 147–62. New York: Routledge. Russell, Bertrand. 1912. The Problems of Philosophy. London: Williams and Norgate. Russell, Bertrand. 1912/1992. “Review of James’s Essays in Radical Empiricism.” In The Collected Papers of Bertrand Russell, Volume 6: Logical and Philosophical Papers, 1909–13, edited by John G. Slater, 298–304. New York: Routledge. Russell, Bertrand. (1914. Our Knowledge of the External World: As a Field for Scientific Method in Philosophy. Chicago: Open Court. Russell, Bertrand. 1914/1919. “The Relation of Sense-Data to Physics.” In Mysticism and Logic and Other Essays, 145–79. New York: Longmans, Green. Russell, Bertrand. 1915/1919. “The Ultimate Constituents of Matter.” In Mysticism and Logic and Other Essays, 125–44. New York: Longmans, Green. Russell, Bertrand. 1918/1985. “The Philosophy of Logical Atomis.” In The Philosophy of Logical Atomism, edited by David F. Pears, 35–155. Peru, IL: Open Court. Russell, Bertrand. 1919a. Mysticism and Logic and Other Essays. New York: Longmans, Green. Russell, Bertrand. 1919b. “On Propositions: What They Are and How They Mean.” Proceedings of the Aristotelian Society, Supplementary Volumes, Vol. 2, Problems of Science and Philosophy, 1–43. Russell, Bertrand. 1921. The Analysis of Mind. London: George Allen and Unwin. Russell, Bertrand. 1992a. “On Matter.” In The Collected Papers of Bertrand Russell, Volume 6: Logical and Philosophical Papers, 1909–13, edited by John G. Slater, 77–95. New York: Routledge. Russell, Bertrand. 1992b. Theory of Knowledge: The 1913 Manuscript. Edited by Elizabeth Ramsden Eames. New York: Routledge. Sainsbury, Mark. 1979. Russell. London: Routledge and Kegan Paul. Sartre, Jean-Paul. 1943/1948. Being and Nothingness. Translated by Hazel E. Barnes. New York: Philosophical Library. Soames, Scott. 2003. Philosophical Analysis in the Twentieth Century, Volume I: The Dawn of Analysis. Princeton: Princeton University Press. Soames, Scott. 2014. The Analytic Tradition in Philosophy, Volume I: The Founding Giants. Princeton: Princeton University Press.

Russell on Introspection and Self-Knowledge

285

Stubenberg, Leopold. 2015. “Russell, Russellian Monism, and Panpsychism.” In Consciousness in the Physical World: Perspectives on Russellian Monism, edited by Torin Alter and Yujin Nagasawa, 58–90. New York: Oxford University Press. Stubenberg, Leopold. 2016. “Neutral Monism.” In The Stanford Encyclopedia of Philosophy, edited byEdward N. Zalta, Winter 2016 Edition. Available online: https:// plato.stanford.edu/archives/win2016/entries/neutral-monism/. Tully, Robert E. 2003. “Russell’s Neutral Monism.” In The Cambridge Companion to Bertrand Russell, edited by Nicholas Griffin, 332–70. New York: Cambridge University Press. Wahl, Russell. 2015. “Sense-Data and the Inference to Material Objects: The Epistemological Project in Problems and Its Fate in Russell’s Later Work.” In Acquaintance, Knowledge, and Logic: New Essays on Bertrand Russell’s The Problems of Philosophy, edited by Donovan Wishon and Bernard Linsky, 87–105. Stanford, CA: CSLI Publications. Wishon, Donovan. 2015. “Russell on Russellian Monism.” In Consciousness in the Physical World: Perspectives on Russellian Monism, edited by Torin Alter and Yujin Nagasawa, 91–118. New York: Oxford University Press. Wishon, Donovan. 2017. “Russellian Acquaintance and Frege’s Puzzle.” Mind 126 (502): 321–70.

11

Russell’s Bridge Dustin Olson and Nicholas Griffin

Abstract This chapter presents Russell’s proposed structure for rational belief, here understood in terms of epistemic justification. Russell likens this structure of justification to a bridge, wherein the individual data, or intrinsically credible propositions, make up the piers, and inferred propositions make up the connecting girders which, in virtue of their coherence, stabilize and strengthen the entire structure including the piers themselves. Each individual pier serves as both epistemic premise and conclusion on this model. We show how one might characterize this theory using the familiar “foundationalist” and “coherentist” characterizations of the structure of epistemic justification. On this front, this chapter suggests that Russell anticipates a number of more recent advances in epistemology that blur the foundationalist-coherentist distinction. From this brief discussion of Russell’s view of epistemic structure, we broaden the discussion to Russell’s philosophical method, showing that this epistemic theory is consistent with Russell’s regressive method from earlier philosophical programs. In employing this method, it is again argued that Russell anticipates significant moves made in later-twentieth-century philosophy, not least of which includes the method of reflective equilibrium.

1  Building Russell’s Bridge Russell’s later philosophy (1914–48) is motivated by a predominantly epistemic program, a program he characterizes as that of philosophy itself: “an unusually obstinate attempt to arrive at real knowledge” (OOP, 2). He later refines this

Russell’s Bridge

287

program as a response to the question: “How comes it that human beings, whose contacts with the world are brief and personal and limited, are nevertheless able to know as much as they do know?” (HK, 5). For the paragon of human knowledge, Russell suggests we needn’t look further than the success of science. Scientific knowledge, however, is not based on those methods from which we can be most confident—a priori methods of deduction without inferences beyond those things with which we are directly acquainted. Rather, we come to scientific knowledge through perception augmented by “non-demonstrative,” that is, non-monotonic, inference. And herein lies the rub: the world as revealed by science (especially physics, the most advanced of the sciences) appears to be vastly different from that revealed by ordinary perception. How, then, can we come to know physics, since the first step to acquiring such knowledge rests on our ordinary experience of the external world? The empirical foundations of the system seem to be at odds with the scientific doctrines which are based on them. So, the major problem of philosophy, for Russell, is to account for the relation between individual experiences and the general body of scientific knowledge.1 Russell admits that a precise analysis of “knowledge” isn’t possible.2 Moreover, all knowledge is to a greater or lesser extent uncertain. Russell suggests that “the supposed absolute concept ‘knowledge’ should be abandoned for the concept ‘knowledge with degree of certainty p’, where p will be measured by mathematical probability when this can be ascertained” (HK, 517).3 Accordingly, “knowledge” can be measured within the boundaries “certainly true” and “certainly not-true.” It is consistent with Russell’s view to take these degrees of certainty as degrees of epistemic justification—or in Russell’s words, degrees of rational credibility. In cases of knowledge, known propositions must be accepted on rationally credible grounds; our beliefs require justification. Russell, therefore, provides a nuanced interpretation of the traditional account of knowledge (TAK), viz. justifiedtrue-belief: one knows that p to the degree of rational credibility one has for accepting that p. Despite accepting that something better than TAK is unavailable, Russell acknowledges that it is not without its difficulties. First, not all true belief, even if believed for good reasons, is knowledge.4 Second, and more importantly, TAK faces the problem of accounting for how beliefs are justified. It cannot simply be by further justified beliefs, for that results in an endless regress. The options for this familiar problem are four: we may suppose that justification can be achieved by a (potential) infinity of justified beliefs (infinitism), or by some self-justifying beliefs (foundationalism), or by a process of circular justification (coherentism),

288

The Bloomsbury Companion to Bertrand Russell

or that it cannot be achieved at all (skepticism).5 Historically, foundationalism and coherentism have been the two most popular responses. Russell’s response is, in effect, to combine these two. He treats knowledge, not as a set of propositions based on some set of self-evident, indubitable truths, nor simply as a coherent system of beliefs, but rather as a coherent body of beliefs some of which have an inherent degree of rational credibility. Here the distinction between premisses and conclusions is blurred: All, or nearly all, of what passes as knowledge is in a greater or less degree uncertain; that, if principles of inference are among the prima facie materials of knowledge, then one piece of prima facie knowledge may be inferrible from another, and thus acquires more credibility than it had on its own account. (HK, 173)

Of particular importance for Russell’s view, then, is explaining how it is that a proposition has any degree of rational credibility and what the prima facie materials for knowledge are. A degree of rational credibility, according to Russell, can be attached to any datum or to a proposition that is supported or undermined by a datum or data. Russell holds that a datum has intrinsic credibility in virtue of its status as a datum, and that data can provide inferential support for other propositions and the inferences themselves. Regardless of a proposition’s inferred credibility, after any analysis of rational belief, accepting a proposition is “rooted in data, and data alone” (HK, 401). As the base for any reason for belief, data serve as the prima facie materials for knowledge—that is, data are epistemic premisses.6 Matters become somewhat complicated at this point, however: data are always to some degree uncertain. So, independent credibility does not guarantee that one knows the independently credible propositions simply in virtue of their status as epistemic premisses; it follows that one’s conclusions on the basis of these premisses also cannot be known with any degree of certainty. Skepticism seems inevitable. This conclusion needn’t follow, however: It is obvious that the conclusion of an argument cannot derive from the argument a higher degree of credibility than that belonging to its premisses; consequently, if there is such a thing as rational belief, there must be rational beliefs not wholly based on argument. It does not follow that there are beliefs which owe none of their credibility to argument, for a proposition may be both inherently credible and also a conclusion from other propositions that are inherently credible. (HK, 409)

Russell’s Bridge

289

Applying this reasoning to the concern that data are insufficient for knowledge, recall Russell’s conclusion that a whole body of knowledge may derive more credibility than any individual proposition within that body. If the body was built exclusively out of defeasible premisses independently of each other, then the whole body would only be as strong as the defeasible foundation. Accepting that a proposition qua epistemic premise can garner further credibility as proposition qua epistemic conclusion, however, enables a whole stronger than its independent parts. To see how a proposition can attain a higher degree of credibility from a set of independent data, where each datum is less credible than all the data taken together, consider the following case. Jones is trying to recall a pub that she enjoyed while vacationing in Manhattan with her friend Smith a number of years ago. Jones faintly recalls that it was called The Whitehorse Tavern. After inquiring with Smith, Smith reports that it seems to her that it was called The Whitehorse Tavern. Later, Jones discovers her travel diary with an entry, and accompanying Polaroid of her and Smith, extolling the atmosphere of the Whitehorse Tavern. In this case, each datum that Jones has provides a degree of independent credibility to the proposition, “The pub I enjoyed on my vacation in Manhattan was the Whitehorse Tavern.” Independently, however, each datum has less credibility than all the data taken together.7 The foregoing development of Russell’s thought concerning the rational credibility of accepted propositions given the inevitable uncertainty attached to those propositions presents a theory of justification wherein the whole is more credible than any one part. Russell summarizes his view thus: Given a number of propositions, each having a fairly high degree of intrinsic credibility, and given a system of inferences by virtue of which these various propositions increase each other’s credibility, it may be possible in the end to arrive at a body of interconnected propositions having, as a whole, a very high degree of credibility. Within this body, some are only inferred, but none are only premisses, for those which are premisses are also conclusions. The edifice of knowledge may be compared to a bridge resting on many piers, each of which not only supports the roadway but helps the other piers to stand firm owing to interconnecting girders. The piers are the analogues to the propositions having some intrinsic credibility, while the upper portions of the bridge are the analogues of what is only inferred. But although each pier may be strengthened by the other piers, it is the solid ground that supports the whole, and in like manner it is intrinsic credibility that supports the whole edifice of knowledge. (HK, 413)

290

The Bloomsbury Companion to Bertrand Russell

This is Russell’s Bridge. In what follows, we discuss how to characterize Russell’s prescient epistemic views and the bridge’s place in epistemology, which is then shown to extend to his overall and again prescient philosophical method.8

2  Russell’s Bridge and the structure of epistemic justification Returning to the epistemic project from whence the bridge metaphor derives, recall the regress problem for justification. As we have said, responses to this challenge have tended in two directions: foundationalism, which accepts that there are some beliefs that are self-justified; and coherentism, which denies that there are any privileged foundational beliefs, maintaining that beliefs are justified only in so far as they are members of a coherent whole. Russell’s view does not neatly fit in either of these camps. While there are certainly foundationalist aspects to his view, coherence plays an equally important role. In juxtaposition to the two traditional approaches, recall Russell’s employment of data, those propositions with independent credibility, as the basis for knowledge: I propose to advocate an unorthodox opinion, namely, that a datum may be uncertain. There have been hitherto two views: first, that in a proper articulation of knowledge we start from premisses which are certain in their own right, and may be defined as “data”; second, that since no knowledge is certain, there are no data, but our rational beliefs form a closed system in which each part lends support to every other part. . . . The view which I set forth is a compromise, but one somewhat more in favour of the [former]. (HK, 409)

If Russell is proposing a compromise, how then should we understand his model? During much of the twentieth century the view seems to have been quite common that not only were coherentism and foundationalism the two dominant approaches to epistemology, but that essentially everyone did epistemology in one of these two ways or the other. There is no evident justification for this view, except perhaps for the fact that coherentism and foundationalism are the two most promising of the four responses to the regress problem (Williams 2001: 81; DePaul 1999:154; Aristotle Posterior Analytics, 71b–74a). It is certainly possible to formulate foundational and coherentist theories in such a way as to make them mutually exclusive and exhaustive: Pollock (1986: 20–21) does so, insisting that foundational theories privilege some beliefs and coherence theories do not. But this simple dichotomy, though perhaps taxonomically useful, does little

Russell’s Bridge

291

justice to the ways in which foundationalism privileges certain beliefs and the purpose for which it does so, nor to the respects in which a cautious coherentism might want to privilege some beliefs. Epistemologists and Russell scholars alike overwhelmingly fit Russell into this ham-fisted classification as a straightforward foundationalist.9 In traditional strong foundationalism, the notion of a basic belief requiring no further justification is based on something analogous to “self-evidence,” where a basic belief is self-evident to the believer and can therefore be held as certain.10 Candidate self-evident beliefs might include beliefs from introspection, or immediate perception, or elementary analytical truths. All nonbasic beliefs accrue their justification by being deduced from the incorrigible basic beliefs. To suggest that this is Russell’s view is a clear misunderstanding. Russell believes that some propositions have a very high degree of intrinsic credibility and is sympathetic to the notion of self-evidence: “If you slip on a piece of orange peel and hit your head with a bump on the pavement, you will have little sympathy with the philosopher who tries to persuade you that it is uncertain whether you are hurt” (HK, 172). Despite this sympathy, however, Russell highlights two troubling aspects of self-evidence as the key epistemological feature of basic beliefs. First, those propositions that we feel are self-evident may only have a subjective resemblance to truly self-evident beliefs and thus do nothing more than satisfy a psychological need rather than an epistemic one. Second, and more importantly, self-evidence itself comes in degrees and is therefore subject to doubt. To this second point, consider one’s varying perceptions of the planet Venus from just prior to dawn, when it is clearly visible, until it has completely disappeared sometime after dawn. At first, one is confident in one’s perception of Venus, but as the light of the sun intensifies, Venus appears more and more faintly, thereby diminishing the self-evidence of the perception. Likewise for analytical truths, such as simple logical proofs over more complicated ones. On the one hand, simple proofs seem self-evident, but a complicated, difficult-to-see step in a proof is far from self-evident. Yet if the steps have been made as small as possible, both must be basic. Likewise, even the analyticity of sufficiently complex sentences is subject to doubt.11 Thus, no datum can be treated as certain according to Russell. It follows, then, that foundations, or premises for knowledge, are not incorrigible. Russell is therefore not a defender of strong foundationalism. Although clearly not strong foundationalism, might Russell’s view be a version of moderate foundationalism, according to which basic beliefs are

292

The Bloomsbury Companion to Bertrand Russell

non-inferential and have independent, knowledge-level justification that can fall short of certainty?12 For similar reasons provided against a strongfoundationalist reading of his view, we can also conclude that Russell is not a moderate foundationalist. Recall that on Russell’s view any datum has some degree of intrinsic credibility attached to it and acts as a premise for knowledge. Recall further, however, that because credibility works in degrees, it is unclear, independent of other data, what degree of credibility should be attached to a datum. But if data are supposed to act as the basic beliefs, in a moderately foundational sense, then these degrees of certainty must necessarily be high enough to confer knowledge. Russell’s view neither guarantees nor requires such a basis.13 Although, to our knowledge, no one has made out a serious case that Russell defends coherentism, given the above binary characterization of epistemologies as “either foundationalist or coherentist” over most of the past century, and that Russell is not a foundationalist in any traditional sense, it is worth highlighting that Russell is also not defending the traditional caricature of coherentism.14 This caricature holds that no beliefs are privileged, that any justified beliefs are justified only to the extent that they cohere within a large enough system of belief, and that coherence is a property of a set of beliefs and not its members. On Russell’s view, of course, there is a privileged set of beliefs—all of one’s data that make up the piers of the bridge. This latter feature alone reveals that Russell’s theory is not overtly coherentist. Although Russell’s view is not a straightforward version of coherentism, it is worth highlighting that Russell considers it a “modified” coherence theory— modified in the sense that it is a coherence theory of knowledge rather than truth (HK, 173). Russell recognizes, however, that there are problems with pure coherentism that necessitate some privileged beliefs, or content, grounding a coherent belief set. Consider, for example, the “many-systems” objection (Williams 2001: 19): two incompatible belief sets could be equally justified in virtue of their respective coherence.15 While this is not completely damning for a coherence theory of justification, it does highlight a further more considerable problem: a system could have wildly implausible members, but could maintain its justification simply in virtue its coherence. There needs to be some means of avoiding coherent fictions if justification is to have the right kind of truthconnection. Another challenge to coherentism concerns isolated beliefs. We all have experiences where the accuracy of our judgment given that experience has no bearing whatever on the truth or falsity of the vast majority of our extant

Russell’s Bridge

293

beliefs. Jones’s belief that a stranger is walking a dog in front of her house right now has no bearing on her belief about who will win Super Bowl 100 or when she should next change the oil in her car, two beliefs that also have no clear connection. These types of challenges prompt a more modest endorsement of coherence—the type of coherentism anticipated by Russell, whether or not we would label his view as such. We find, then, that Russell’s view is neither a straightforward foundationalism nor a straightforward coherentism. Arguably, this inability to easily characterize Russell’s view is one of its most significant merits. Epistemic justification is not simply a matter of independent experiences or coherent belief sets; it is a conjunction of the two. This recognition is motivated in part by fallibilism and the difficulties facing strong foundationalism while also acknowledging the importance of experience in grounding beliefs, all of which Russell anticipates. Consider the engineering involved in a bridge, which highlights Russell’s compromise between foundationalism and coherentism. On the foundationalist side, for a bridge to be grounded, it must have piers built on solid ground. Alone, piers can stand unsupported; alone, they can also easily be knocked over. A bridge’s true strength is found in its girders, that is, through coherence. It is possible that a girder on any superstructure may be weakened and must be replaced. Such a girder can be replaced to strengthen that section of the bridge, while also not requiring a refabrication of the entire bridge. Indeed, even a pier  can be added or removed in isolation without the entire superstructure being affected. Applying the metaphor to epistemology, the piers represent intrinsically credible, yet defeasible, propositions. These propositions can increase their credibility by cohering with other intrinsically credible beliefs that are connected within a cohering body of beliefs. If one finds that a belief is incoherent with the relevant set of one’s beliefs, it can be replaced with neither the entire structure being undermined nor the set losing its justification. Finally, there is nothing inconsistent with Russell’s view that an individual may have more than one set of cohering beliefs. Sticking with the metaphor, one can build a number of bridges. Of course, the more the bridges connect, the stronger the system is on the whole, and too many small bridges would be unwieldy. Nonetheless, there are certainly some independent sets of beliefs with no relevant connections. Current trends in epistemology recognize the importance of coherence for epistemic justification—there are very few strong foundationalists left.16

294

The Bloomsbury Companion to Bertrand Russell

In response to this recognition, a number of contemporary theories offer similarly difficult-to-label bridge-style structures of epistemic justification. Among these views are Susan Haack’s foundherentism (1993, 1997), Earl Conee’s foundational coherentism (1988), Kevin McCain’s explanationist evidentialism (2014), Catherine Elgin’s holism (1996, 2005), and William Lycan’s explanationist coherentism (1988, 2012). Obviously these theories differ in the details, but as to the nature of epistemic justification they share a number of commonalities. First, they each accept a fallibilist conception of knowledge: we may never have certainty, but we can have knowledge. Furthermore, they each recognize that the two traditional theories needn’t be at odds concerning their core contributions to justification. On the one hand, they all adopt the foundationalist’s contentions that justified beliefs must be suitably attuned to one’s experiences and that a coherent set of beliefs cannot be justified if that set doesn’t cohere with one’s experiences; on the other side, they adopt the coherentist’s contentions that coherence contributes to a belief ’s justification and that no autonomously justified belief is required for coherence to provide this justificatory work.17 Each contemporary theorist admits, then, that while the traditional dialectic offers important insights into a suitable theory of epistemic justification, neither theory with its accompanying label is a sufficient account of justification. Consider Conee’s and Haack’s contentions concerning what an acceptable theory of justification requires, in terms of the traditional dialectic. According to Conee, “In order for a theory [of justification] to capture all that’s right about both coherentism and foundationalism, it must imply that both coherence with some system of belief and a foundation in experience are separately necessary and jointly sufficient for the justification of a belief ” (1988: 46). Haack has similar sentiments: “Neither of the traditionally rival theories can be made satisfactory without sacrificing its distinctive character. The obvious conclusion . . . is that we need a new approach which allows the relevance of experience to empirical justification, but without postulating a class of basic beliefs or requiring that relations of support be essentially one directional” (1997: 239).18 Russell anticipates just the type of theory Conee and Haack call for. As he is adamant that all justified belief is based on data, Russell recognizes the importance of the empirical. While data alone are where one’s reasons for belief are found, an individual datum provides little by way of certainty—Russell’s gauge for degrees of credibility, or justification. One can achieve near certainty, however, when one’s system of belief forms a coherent whole. The structural strength afforded

Russell’s Bridge

295

by coherence results from Russell’s insight that data can serve as both premises and conclusions, none of which are solely foundational and most of which are firmly rooted in the empirical. Theorists wed to the traditional foundationalist-coherentist distinctions have often labeled these types of hybrid views as a weak foundationalism.19 On this view, basic beliefs provide minimal justification that by themselves are neither sufficient for knowledge nor can act as acceptable premises for justifying further beliefs. The main contrast between this versus traditional versions of foundationalism occurs in response to the regress problem addressed above. Rather than appealing to self-justifying foundations or coherent belief sets, this view recognizes the importance of both of these characteristics for epistemic justification. Given this view, Laurence BonJour contends that “weak foundationalism represents a kind of hybrid between moderate foundationalism and .  .  . coherence theories” (1985: 23). This general description certainly captures Russell’s view. It is important to note, however, that this is not the version of foundationalism that that majority theorists attribute to Russell; those theorists have a more extreme foundationalism in mind.20 Furthermore, as this view is a hybrid between foundationalism and coherentism, the nomenclature is a bit misleading—it could just as easily be referred to as a weak coherentism.21 Also, if foundationalism merely requires that some class of beliefs have a degree of intrinsic credibility, then Russell and other hybrid theorists are foundationalist. If, on the other hand, foundationalism requires that there are properly basic beliefs that can independently provide inferential beliefs with justification, then neither Russell nor the contemporary hybrid views are foundationalist.22 In the end, the above views are probably best described as holistic, as their central features are that beliefs are justified within the context of a system or structure of beliefs, and not simply in isolation. Epistemic holism is traditionally associated with coherentism. However, it is neither obviously coherentist nor obviously foundationalist: there is nothing in holism that divorces a comprehensive system from experience, or requires that any datum is limited to a one-directional support relation with other data, or that certain data must remain in stasis for the whole system to garner justification. Paradigmatic of epistemic holism is Russell’s view, where individual data, each with a degree of intrinsic credibility, can acquire more credibility in relation to each other than they have individually. Thus justification is not simply a product of a oneway relation of support, but rather one’s beliefs are given further support, or

296

The Bloomsbury Companion to Bertrand Russell

credibility, by the support provided by the structure as a whole: “It may thus happen that a body of propositions, each of which has only a moderate degree of credibility on its own account, may collectively have a very high degree of credibility” (HK, 173).

3  Russell’s regressive method To this point we have discussed only Russell’s later epistemology, indeed only his very last major philosophical book. It might be thought that Russell’s epistemic holism was merely a late turn in his famously convoluted philosophical development. Principia Mathematica (1910–13) and the logicist project from which it emerged are often cited as clear evidence of Russell’s foundationalism, at least early in his career.23 But axiomatization is not an infallible sign of foundationalism and a moment’s thought should be sufficient to convince that the point of the logicist project was not to place mathematics upon a secure logical foundation. There is little, indeed essentially nothing, about justification or certainty in Principia Mathematica or even in The Principles of Mathematics (1903), originally conceived as its philosophical prolegomenon. In those works it was the logical structure of mathematical theory rather than the epistemology of mathematics which was Russell’s concern. Had it been undertaken for epistemological purposes, Principia would have been a quixotic venture to put it mildly. It would be hard for anyone to suppose that 1 + 1 = 2 owed its rational credibility to the proof it eventually receives at PM*110.643, well into the second volume. And indeed Russell said as much: There is an apparent absurdity in proceeding, as one does in the logical theory of arithmetic, through many rather recondite propositions of symbolic logic, to the “proof ” of such truisms as 2 + 2 = 4: for it is plain that the conclusion is more certain than the premisses, and the supposed proof therefore seems futile. (Russell 1907, 572)

Unfortunately the paper in which he said it, “The Regressive Method of Discovering the Premises of Mathematics,” was not published until 1974 (in EA) and the important lessons it makes explicit were therefore often missed. In it Russell makes it clear that, as far as rational credibility24 is concerned, the dependency runs the other way: the axioms of the system get their credibility from the fact that propositions like 2 + 2 = 4 can be deduced from them.

Russell’s Bridge

297

In this, Russell maintains, logic and mathematics are not different from any other science. “In any science,” Russell says, we start with “what are commonly called the facts,” “a body of propositions of which we feel fairly sure” (1907, 573; rearranged). He calls these the “empirical premisses” of the science.25 By contrast, the general laws (or “logical premisses”) of a science are propositions, “logically simpler26 than the empirical premisses of the science, but such that the empirical premisses, or some of them, can be deduced from these laws” (ibid.). These general laws, or logical axioms, are discovered through what Russell calls “the regressive method.” It is unfortunate that the “Regressive Method” paper remained unknown for so long.27 Russell wrote the paper for the Cambridge Mathematical Club where he read it in March 1907. He seems to have made no attempt to publish it, and he made no use of the text itself in other writings. Nor did he write any other paper so exclusively focused on the methodology which underlay Principia Mathematica. The paper itself does not give any indication as to why Russell chose that particular topic for a paper to the CMC, but it hard to resist the conclusion that he was provoked to write it by his fierce debate about logicism with Henri Poincaré, which had culminated a few months before with the publication of Russell’s “Les paradoxes de la logique” in the Revue de métaphysique et de morale (Russell 1906). Poincaré (1905, 1906, 1906a) had attacked logicism on essentially epistemological grounds. His most important objection concerned mathematical induction. It was, he held, viciously circular to define numbers using mathematical induction (as Peano had done) and then to derive the principle of induction from the logic of ancestral relations (as Frege and Russell had done). None of these three concepts, he held, was more fundamental than the others, indeed from an epistemological point of view they were all equivalent, amounting to an a priori arithmetic intuition as to how to generalize or how to continue a series.28 It was an interesting criticism, but it mistook the logical purpose of logicism for an epistemological one. Russell’s point was not that the concept of an ancestral relation was epistemologically more basic than that of induction or of number, nor that one must (or even could) grasp the first before one could grasp the other two. Russell was not making an epistemological point at all. His point was simply that the concept of an ancestral relation was the most general notion of the three and that the other two could be derived from it. We suspect that the care Russell took in “Regressive Method” to distinguish the empirical order from the logical was influenced by his duel with Poincaré, and his  direct

298

The Bloomsbury Companion to Bertrand Russell

response (Russell, 1906) to Poincaré supports this. Concerning Russell’s insistence on using explicitly formulated logical rules rather than intuition, Poincaré (1906a: 295) asked rhetorically whether the rules must be applied blindly or not: If not, then only intuition could distinguish acceptable from unacceptable uses of them; but if they were to be applied blindly, then the rules would have to be infallible. Russell remarks that this embodies a “misconception of the claims of logistic” (Russell 1906, 280) and then, to make logistic’s claims clear and to explain how infallible first principles were not required, he embarks on a very brief statement of the point of view to be found in “Regressive Method,” to our knowledge the first time he had put this position on paper (Russell 1906, 280–82). He rejects the notion of infallible rules, pointing out that it is unnecessary to assume them; emphasizes the methodological continuity between logistic and the other sciences; and even concedes the use of the term “intuition” to explain the initial credibility of the initial data.29 Though Russell’s most focused discussion of these matters is to be found in the posthumously published “Regressive Method” paper of 1907, the same point of view can also be found in later works which were widely available during Russell’s lifetime. Thus, for example, the idea that the logical axioms from which mathematics can be deduced are more abstract, general and logically simple than the “obvious,” “ordinary” and “easy” things which form our “starting-point” appears at the very beginning of An Introduction to Mathematical Philosophy (Russell 1919, 1–2). And pretty much all the major themes developed in the 1907 paper are briefly stated in a widely reprinted paper from 1924, “Logical Atomism.” There, for example, Russell says: When pure mathematics is considered as a deductive system .  .  . it becomes obvious that, if we are to believe the truth of pure mathematics, it cannot be solely because we believe in the truth of the set of premisses. Some of the premisses are much less obvious than some of their consequences, and are believed chiefly because of their consequences. (Russell 1924, 163–64)

Russell does not invoke the bridge metaphor in these earlier writings, but it is nonetheless applicable. The important point in making use of it is not to treat the logical premises as the piers from which the rational credibility of the other parts of the system is derived. That would be the traditional foundationalist understanding of an axiomatic system. It is, rather, the “inherently obvious” epistemic premises which form the piers and which supply rational credibility to the logical premises. The latter may themselves be inherently obvious, but

Russell’s Bridge

299

that is by no means essential. The account combines both foundationalist and coherentist elements: the epistemic premisses are privileged by their inherent credibility, but, overall, justification within the system is achieved by its logical coherence. Certainty is hard to come by in these early works, as it is in the later ones, notwithstanding that the former are concerned primarily with the mathematical sciences. The empirical premisses may be inherently obvious, but that does not make them absolutely certain; and the general laws which the regressive method reveals, at least those of them that are not themselves inherently obvious, derive what certainty they have from that of the empirical premises that can be deduced from them. Russell writes: The [general] laws only become as certain as the empirical premisses if we can show that no other hypotheses would lead to the empirical premisses, or if (what may happen in mathematics) the [general] laws, once obtained, are found to be themselves obvious, and thus to be capable of themselves becoming empirical premisses. (1907, 573)

By way of illustration, Russell compares the Peano postulates for arithmetic with the basic laws and rules of Frege’s Grundgesetze, from which the Peano postulates can be derived. The Peano postulates get their certainty both from the fact that arithmetic follows from them and from their “inherent obviousness” (1907, 575). As a result, Peano’s postulates become in turn the empirical premisses for the Grundgesetze, where they are to be derived from more general laws, in this case the basic logical laws of Frege’s system. Not surprisingly, since the Grundgesetze system was known to be inconsistent, Russell finds some of Frege’s logical premisses “more intrinsically obvious than others” (1907, 576). Among those which are “not very obvious” Russell cites the premiss that “whatever can be said significantly of an individual can be said significantly . . . of a class” (i.e., that individuals and classes are both objects, the principle which allows the formulation of Russell’s paradox). “Such principles,” he says, “are premisses in the sense of being logically simple propositions from which the desired consequences follow, but they are not empirical premisses, in the sense of being intrinsically obvious” (1907, 577). While what we may call the “progressive method,” which deduces consequences from the logical premisses of a successfully developed science, will confer on the propositions deduced the same degree of certainty possessed by the logical premisses from which they are deduced, the regressive method cannot be assumed to confer on the logical premisses the same degree of certainty that attends the empirical premisses. Nonetheless, Russell’s holist

300

The Bloomsbury Companion to Bertrand Russell

model of justification does allow the inherently obvious, interconnected parts of a successfully developed science to achieve a greater degree of certainty than would accrue to them as inherently obvious truths standing alone. In the “Regressive Method” paper, Russell says three things about obviousness which are worth quoting at length: (1) It is a matter of degree, and, in the case of conflict, the more obvious is to be preferred, caeteris paribus, to the less obvious. (2) Even where there is the highest degree of obviousness, we cannot assume that we are infallible—a sufficient conflict with other obvious propositions may lead us to abandon our belief. . . . (3) Assuming the usual laws of deduction, two obvious propositions of which one can be deduced from the other both become more nearly certain than either would be in isolation: and thus in a complicated deductive system, many parts of which are obvious, the total probability may become all but absolute certainty. (1907, 577)

In this the natural sciences are on par with the mathematical sciences. The differences between the two are that in the natural sciences the inherent obviousness of the empirical premisses comes from the senses, whereas in the mathematical sciences they have an “a priori obviousness, such as that of the law of contradiction” (ibid.). The methodology which Russell proposes in 1907, with the mathematical sciences in mind, is essentially the same as that which he advocates forty years later, when he’s primarily concerned with the natural sciences.

4  The relevance of Russell’s method Russell’s regressive method has a familiar feel to it. This familiarity is most likely a result of Russell’s anticipation of the more familiar methodology dubbed reflective equilibrium, which we suggest is regressive in ways similar to Russell’s method. Indeed, Russell’s regressive method might actually be characterized as a nascent version of reflective equilibrium. Reflective equilibrium’s introduction is most often attributed to Nelson Goodman (1955) and finds its popularity in John Rawls’s work (1971).30 In its broadest application, reflective equilibrium is a method by which one examines a number of propositions, principles, and general judgments in light of each other, seeking to reach equilibrium among them. Equilibrium is achieved when each proposition is reasonable in light of the others after consideration and revision. It is just this

Russell’s Bridge

301

type of examination that Russell defends in (1907) and (HK). It is also the method, once appreciated, readers can observe in action in much of Russell’s philosophy. We can observe the extent to which the regressive method is a form of reflective equilibrium when comparing Goodman’s description of the latter with Russell’s description of the former. In the interest of explaining and justifying both deductive and inductive logical reasoning, Goodman suggests that our accepted deductive and inductive rules of inference are justified by their conformity with and production of inferences that we would be unwilling to reject. Likewise, our accepted deductive and inductive inferences are justified by their conformity with rules that we are unwilling to amend. In one of the more famous passages of twentieth-century philosophy, Goodman makes the point thus: Principles of deductive inference are justified by their conformity with accepted deductive practice. Their validity depends upon accordance with the particular deductive inferences we actually make and sanction. . . . An inductive inference, too, is justified by conformity to general rules, and a general rule by conformity to accepted inductive inferences. Predictions are justified if they conform to valid canons of induction; and the canons are valid if they accurately codify accepted inductive practice. . . . The point is that rules and particular inferences alike are justified by being brought into agreement with each other. A rule is amended if it yields an inference we are unwilling to accept; an inference is rejected if it violates a rule we are unwilling to amend. (1955, 63–64; rearranged; his emphasis)

Consider next Russell’s proposal for how one would apply his method to deal with contradictions in symbolic logic: the method to be pursued must be, first, to make a kind of hierarchy of obviousness among the results to which our premisses ought to lead, then to isolate, if possible, the premisses from which contradictions flow, and the kind of reasoning which gives rise to the contradictions, and then to invent various modifications of the guilty premisses, applying to such modifications the two tests (1) that they must yield the more obvious of the results to be obtained, (2) that they must not yield any demonstrably false results. . . . [I]f we seem to have discovered precisely why our previous premisses led to contradictions . . . and if the whole kind of reasoning from which the contradictions sprang is ruled out by our new premisses, we may have reasonable confidence that we have made the right kind of modification. (1907, 577–78)

302

The Bloomsbury Companion to Bertrand Russell

Russell’s proposed method is essentially the method Goodman advances nearly half a century later. Russell’s anticipation of reflective equilibrium is not limited to Goodman’s logical program. In much the same way that Rawls recognizes this method’s applicability to alternative areas of inquiry—in his case moral inquiry—Russell supposes that the regressive method is analogous to any scientific inquiry, as discussed in §2 and §3 above. And as scientific knowledge is the paradigm for human knowledge, we can infer that most rational inquiry requires a similar process on Russell’s view. Indeed, many contemporary theorists recognize and appeal to this method across numerous philosophical domains and indeed all theoretical inquiry.31 Consider, for example, Rawls’s method for achieving justification for one’s moral judgments (1971, §9; 2001, §10). We begin with extant accepted moral judgments that are made in a sober and reflective way— Rawls refers to these as “considered judgments”—and weigh them against other such judgments and the motivations for them. In so doing we will discover higher-level principles that agree with our general moral conception and our particular moral judgments. From there we seek to make all of these judgments coherent across levels. At times our principles will be fixed, for example, slavery is wrong, such that any particular judgment that opposes that principle is wrong and should be rejected. Likewise, some particular judgments carry such force that any moral principle allowing an alternative judgment in that particular case must be amended. Once in equilibrium these judgments have more support than they would independently. Reflective equilibrium is a notably regressive method. We are not here contending that Russell is the motivating figure behind the method of reflective equilibrium.32 What is notable, however, is if his regressive method is a form of reflective equilibrium—and as the above examples show there are compelling reasons to accept this conclusion—then how we interpret Russell’s philosophical programs later on is significantly impacted. In very broad outline we can see how reflective equilibrium offers a solution to his initial problem of the apparent clash between our ordinary perception of the world from which science starts and the very different world which science ultimately reveals. Ordinary perception provides the data for physics, but under reflective equilibrium the data are not immune to revision: they may be reinterpreted, reevaluated, or even rejected in the light of subsequent reflection. The overall aim, as Russell said in 1907, is “avoiding the false and keeping what we cannot but think true” (1907, 578). Under this interpretation of Russell’s philosophical

Russell’s Bridge

303

method, in inquiry coherence is equally as important as foundations for generating justification; conclusions can be afforded more certainty than the premises from which they are derived; any isolated proposition is defeasible; all inquiry is fundamentally inductive. Such a holistic fallibilism is rarely attributed to Russell, but it should be. Accepting as much will provide Russell scholars with new and exciting ways in which to appreciate Russell’s metaphilosophy and philosophical methodology, which will inevitably affect extant and influence future readings of his works.

Notes 1 Providing the details to Russell’s response to this problem is beyond our objective here, but the general theory is that all of what we take to be the objects and relations in the external world are inferred from the spatiotemporal location, in our own individual perceptual spaces, of the events that make up our brains/minds. See Olson (2011) for a summary of Russell’s later theory of perception and Olson (n.d.) for a discussion on Russell’s thought development concerning inference and the necessary conditions for any knowledge whatever. 2 On his first attempts at defining “knowledge” in his HK, Russell suggests that “there is a certain inevitable vagueness and inexactitude in the conception. . . . Nevertheless, it is well to be as precise as possible about the unavoidable lack of precision in the definition of which we are in search” (170). By the end of this work, Russell concludes: “‘Knowledge’, as we have seen, is a term incapable of precision” (516). 3 Russell acknowledges, however, that it is rarely, if ever, possible to attach a mathematical probability to one’s degree of certainty. The reason for this limitation results from the limitations in our premises for knowledge, most notably the knowledge of particular facts rooted in our empirical reasoning. 4 Russell provides a number of cases that we would now term “Gettier Examples” in response to the true-belief theory of knowledge: the clock, the lottery, the prime minister with the initial B (HK, 170–71). It’s worth highlighting that these examples anticipate Gettier. 5 For an account of the regress problem for justification, traditional responses to this problem, and challenges for these responses, see Feldman (2003, 49–78). 6 Russell defines epistemic premiss and data nearly synonymously: “An epistemological premiss may be defined as a proposition which has some degree of rational credibility on its own account, independently of its relation to other propositions” (HK, 401); “A datum [is] a proposition which has some degree of

304

7

8

9

10

11

12 13

14

The Bloomsbury Companion to Bertrand Russell rational credibility on its own account, independently of any argument derived from other propositions” (HK, 409). We find, then, that included in the basic materials for knowledge are data, and that these data serve as epistemic premisses in virtue of their intrinsic credibility. This example employs memorial and testimonial-based belief, but the structure is equally applicable to any combination of all types of belief—perceptual, a priori, testimonial, memorial, and so on. The bridge metaphor was not original to Russell. It is to be found in William James who, comparing conceptual knowledge with that derived from sensation, said: “Conceptual systems which neither began nor left off in sensations would be like bridges without piers. Systems about fact must plunge themselves into sensation as bridges plunge their piers into rock” (James1890, II: 7) But James, more foundationalist than Russell, thought of sensations as “the stable rock, the terminus a quo and the terminus ad quem of thought.” Russell had read the book as long ago as September 1894 and commented favorably in the margin against this passage: “applies to Metageometry” (i.e., to non-Euclidean geometry, on which Russell was then working for his fellowship dissertation). See Brennan and Griffin (1997). Among epistemologists, see: Moser (1999, 321); Foley (2000, 8); Pastin (1975, 142); Hamlyn (1970, 289); and among Russell scholars: Pears (1967, Ch. 3); Eames (1969, 83); Sainsbury (1979, 193–200; 1989, 200). Ayer (1971, 12) falls into both groups. This version of foundationalism is often dubbed classical and Cartesian foundationalism, which is synonymous with strong foundationalism concerning the incorrigible status of basic beliefs. We will stick with strong when referring to this brand of foundationalism. Russell (HK, 410–12) uses these types of examples in defense of the uncertainty of data while responding to Keynes’s claims that beliefs acquired directly by observation are certain. Cf. IMT, 124–25 for a similar argument. See BonJour (1985, 26–7). Issues are somewhat complicated by Russell’s esoteric use of “knowledge” in different contexts. Some passages strongly suggest a moderate foundationalist reading of Russell’s view. The claim that “sensation, immediate expectation, immediate memory, and true memory all give knowledge which is, in some degree and with appropriate limitations, independent of extraneous evidence” (HK, 112), seems to jibe with this form of foundationalism. As we highlight, however, Russell’s foundations are not properly basic and thus cannot be foundationalist in this sense. Defenders of coherence theories of epistemic justification include, BonJour (1985), Sellars (1973), Lehrer (1974), Dancy (1985), Kvanvig (1997), Rescher (1974, 1977]), and Lycan (1988, 2012). Although, especially in Lycan’s and Rescher’s cases, the coherence theory is modified and strictly speaking not a pure coherence theory.

Russell’s Bridge

305

15 The many-systems objection, so far as we know, though Williams did not acknowledge its provenance, originated with Russell (1907a, 443) who argued against the coherence theory of truth on the grounds that two mutually inconsistent sets of beliefs may be equally coherent. 16 Some notable exceptions include BonJour (2010), Neta (2011), McGrew (1997), and (Moser 1989). Unsurprisingly, these authors also defend an infallibilist conception of knowledge. 17 Along the lines described here, these theories share important features with Russell’s. Russell’s conception of data is cashed out in terms of evidence in Conee’s, Haack’s, and McCain’s theories, as deliverances in Elgin’s, and as spontaneous beliefs in Lycan’s. Building from Conee and Feldman’s (2004) framework, McCain highlights the importance of one’s total background evidence at play anytime one acquires a justified belief; Lycan argues that a comprehensive coherent system of belief is best explained by that system being true, and thus the stronger the coherence, the more explanatory force produced by the system and the more justified the members of that system are; Haack and Conee contend that there is a pervasive mutual support relation among one’s justified beliefs, highlighting that it makes little sense to have entirely independently justified beliefs in utter isolation of, or inconsistent with, one’s background beliefs; and Elgin holds that a belief ’s justification is contingent on its membership in an “epistemically acceptable system,” which is a belief system in reflective equilibrium. 18 We might understand Conee’s contribution as a “both p and q” approach, offering what each theory gets right, whereas Haack’s account as a “neither p nor q” approach that seeks to avoid the problems of the respective views. Nevertheless, the two positions are very similar. 19 This form of foundationalism is both introduced and attributed to Russell in (BonJour 1985, 25). See also (Feldman 2003, 70–75). 20 Here are two paradigmatic attributions from an epistemologist and a Russell scholar respectively: “To avoid the problems of this sort is to embrace an especially extreme version of foundationalism, one that insists that the conditions of rational belief are conditions to which we always have immediate and unproblematic access. Bertrand Russell defends such a view” (Foley 2001, 23). “When Russell was planning his reconstruction of empirical knowledge, he felt himself pulled in two opposite directions. On the one hand he wanted to save as much as possible. On the other hand he wanted what was saved to be firmly based on sound foundations, and well and truly saved” (Pears 1967, 23). 21 Indeed, BonJour suggests as much. It is telling that he classifies (Rescher 1977), a staunch defender of coherentism, as possibly a weak foundationalist. Similar claims can be made for Lycan’s view.

306

The Bloomsbury Companion to Bertrand Russell

22 It is our contention that weak foundationalism is not, strictly speaking, foundationalism. If basic beliefs are allowed to have defeasible justification, then, by the weak foundationalist’s own admission, their justification does depend upon other beliefs, for it can be undermined by defeaters (and restored or strengthened by the undermining of those defeaters). A defeasible belief does not stand alone as a basic belief is supposed to in foundationalism; it gains its justification in part from other beliefs in the system. The clear foundationalist distinction between beliefs which derive their justification from other beliefs and those which do not has been lost. In other words, it’s unclear that the notion of weak foundationalism makes sense, since by its own lights basic beliefs are not, properly speaking, basic. And if the foundations are not sole proprietors of justification from which all other beliefs are justified, then we do not have, properly speaking, foundationalism. 23 Hamlyn (1970, 289), e.g., cites PM as evidence of Russell’s foundationalism. See also Lakatos (1962, 16–18); Parsons (1967, 193); Steiner (1975, 17–24). 24 Russell didn’t use the term “rational credibility” in 1907—it derives from Keynes (1921). Rather he spoke of the “certainty” or “obviousness” of a proposition or belief. 25 Russell’s choice of the term “empirical premisses” is not a happy one. He originally introduced it to mean the premisses “from which we are actually led to believe the proposition in question” (1907, 572). In this way, he thinks, we are led to believe that 2 + 2 = 4 from such empirical premisses as “2 sheep + 2 sheep = 4 sheep.” But when we come to the treatment of arithmetic itself, the sheep are long forgotten, and the empirical premisses are now propositions such as “2 + 2 = 4,” the “propositions of which we feel fairly sure.” Of course, in the empirical sciences, the empirical premisses will typically be empirical propositions, which get their credibility from the senses; but in the mathematical sciences, the empirical premisses may get their credibility a priori (1907, 577). For the more general use, the term “epistemological premisses” which he used in Human Knowledge is much preferable. 26 Simplicity is measured “roughly speaking” by the number of constituent-types in the proposition (1907, 572). The simplicity of a proposition will, other things being equal, be inversely proportional to the number of its constituent-types. 27 Andrew Irvine (1989) deserves the credit for first drawing attention to its importance. See also Godwyn and Irvine (2003, 193–201). In Irvine (2004), however, he sees a tension between Russell the foundationalist and Russell the fallibilist, with the latter finally gaining ascendancy in An Inquiry into Meaning and Truth (IMT). We think this conflict is over-drawn, but Russell did have a relatively brief deductivist phase (1914–19) during which he held that the “supreme maxim” was to replace inferred entities by logical constructions (wherever possible) (Russell 1914, 11). We think this looks more like hard-core foundationalism than it actually is. There is no space to examine the issue here, but see Olson (n.d.).

Russell’s Bridge

307

28 See Folina (1992: Chs 4 and 5) for sympathetic commentary. 29 Ever since he had abandoned his early Kantianism, Russell had had little good to say about intuition and it is surprising to see him making this concession to Poincaré, though he uses “intuition” in exactly the (non-Kantian) way that many later analytic philosophers did (and still do). We suspect that he used the term “empirical premisses” (rather than say “intuitive” or “intuited” premisses) in the “Regressive Method” paper precisely to avoid any suggestion of Kantianism. 30 The term reflective equilibrium was coined by Rawls (1971, §9). 31 Such appeals are made in the following areas—at times only implicitly: ethics and moral inquiry (Daniels 1996), (Rawls 1971), epistemology (Elgin 1996; 2005), (Goldman 1986), metaphilosophy (DePaul 1998), (Walden 2013), science (Duhem 1906), (Gillies 1993). 32 Tracing the history of reflective equilibrium is beyond our scope. It is notable, however, that in its earliest forms, viz. (Duhem 1906) and (Russell 1907), reflective equilibrium was motivated as a response to Poincaré. We have found no evidence that Russell was ever aware of Duhem’s work. Duhem, of course, could not have known of the “Regressive Method” paper, though he may have seen the earlier brief account of the method in the Revue de métaphysique et de morale (Russell, 1906), though only after his own work was complete. How much Goodman was influenced by either of these two authors is, as of yet, unclear, although we know that there is a high probability that he discussed Duhem with Quine in their correspondence, and that he would have crossed paths with Russell at Harvard University during Russell’s William James Lectures in 1940, which are reproduced in IMT. While there is neither an explicit appeal to the regressive method nor the bridge, Russell does seem to gesture toward this method in parts of Chapters IX and XX.

References Audi, Robert (ed.) 1999. The Cambridge Dictionary of Philosophy, 2nd ed. Cambridge: Cambridge University Press. Ayer, Alfred J. 1971. Russell and Moore. The Analytical Heritage. London: Macmillan. BonJour, Laurence. 1985. The Structure of Empirical Knowledge. Cambridge, MA: Harvard University Press. BonJour, Laurence. 2010. “The Myth of Knowledge.” Philosophical Perspectives 24: 57–84. Brennan, Frances, and Nicholas Griffin. 1997. “Russell’s Marginalia in his Copy of William James’s Principles of Psychology.” Russell: The Journal of Bertrand Russell Studies NS 17: 123–70.

308

The Bloomsbury Companion to Bertrand Russell

Conee, Earl. 1988. “The Basic Nature of Epistemic Justification.” In Evidentialism: Essays in Epistemology, edited by Earl Conee and Richard Feldman, 37–52. Oxford: Oxford University Press. Conee, Earl, and Richard Feldman. 2004. Evidentialism: Essays in Epistemology. Oxford: Oxford University Press. Dancy, Jonathan. 1985. “A Defense of Coherentism.” In The Theory of Knowledge: Classic and Contemporary Readings, edited by Louis Pojman, 2nd ed., 206–15. Belmont, CA: Wadsworth. Daniels, Norman. 1996. Justice and Justification: Reflective Equilibrium in Theory and Practice. Cambridge: Cambridge University Press. DePaul, Michael R. 1998. “Why Bother with Reflective Equilibrium?” In Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry, edited by Michael R. DePaul and Ramsey William, 293–309. Lanham, MD: Rowman & Littlefield. DePaul, Michael R. 1999. “Coherentism.” In The Cambridge Dictionary of Philosophy, edited by Robert Audi, 2nd ed., 154–55. Cambridge: Cambridge University Press. DePaul, Michael R., and William Ramsey (eds.) 1998. Rethinking Intuition: The Psychology of Intuition and Its Role in Philosophical Inquiry. Lanham, MD: Rowman & Littlefield. Duhem, Pierre. 1906. The Aim and Structure of Physical Theory. Translated by Philip P. Wiener. Princeton NJ: Princeton University Press. Eames, Elizabeth R. 1969. Bertrand Russell’s Theory of Knowledge. London: Allen and Unwin. Elgin, Catherine Z. 1996. Considered Judgments. Princeton, NJ: Princeton University Press. Elgin, Catherine Z. 2005. “Non-foundationalist Epistemology: Holism, Coherence, and Tenability.” In Contemporary Debates in Epistemology, edited by Matthias Steup and Ernest Sosa, 156–67. Malden, MA: Blackwell. Feldman, Richard. 2003. Epistemology. Upper Saddle River, NJ: Prentice Hall. Foley, Richard. 2000. “The Fall of Foundationalism and the Rise of Externalism.” Crítica. Revista de Filosofia 6: 7–18. Foley, Richard. 2001. Intellectual Trust in Oneself and Others. Cambridge: Cambridge University Press. Folina, Janet. 1992. Poincaré and the Philosophy of Mathematics. Basingstoke: Palgrave. Gillies, Donald. 1993. Philosophy of Science in the Twentieth Century. Oxford: Oxford University Press. Godwyn, Martin, and Andrew D. Irvine. 2003. “Bertrand Russell’s Logicism.” In The Cambridge Companion to Bertrand Russell, edited by Nicholas Griffin, 171–201. Cambridge: Cambridge University Press. Goldman, Alvin. 1986. Epistemology and Cognition. Cambridge, MA: Harvard University Press.

Russell’s Bridge

309

Goodman, Nelson. 1955. Fact, Fiction, and Forecast. Cambridge, MA: Harvard University Press. Haack, Susan. 1993. “Double-Aspect Foundherentism: A New Theory of Empirical Justification.” Philosophy and Phenomenological Research 53: 113–28. Haack, Susan. 1997. “A Foundherentist Theory of Epistemic Justification.” In The Theory of Knowledge: Classic and Contemporary Readings, edited by Louis Pojman, 2nd ed., 237–47. Belmont CA: Wadsworth. Hamlyn, David W. 1970. The Theory of Knowledge. New York: Anchor. Irvine, Andrew D. 1989. “Epistemic Logicism and Russell’s Regressive Method.” Philosophical Studies 55: 303–27. Irvine, Andrew D. 2004. “Russell on Method.” In One Hundred Years of Russell’s Paradox. Mathematics, Logic, Philosophy, edited by Godehard Link, 481–500. Berlin: de Gruyter. James, William. 1890. The Principles of Psychology. London: Macmillan, 2 vols. Keynes, John M. 1921. A Treatise on Probability, The Collected Wiritings of John Maynard Keynes, vol. 8. London: Royal Economic Society. Kvanvig, Jonathan. 1997. “In Defense of Coherentism.” Journal of Philosophical Research 22: 299–306. Lakatos, Imre. 1962. “Infinite Regress and Foundations of Mathematics.” In Mathematics, Science and Epistemology, edited by John Worral and Gregory Currie, 3–23. Cambridge: Cambridge University Press.. Lehrer, Keith. 1974. Knowledge. Oxford: Oxford University Press. Lycan, William G. 1988. Judgement and Justification. Cambridge: Cambridge University Press. Lycan, William G. 2012. “Explanationist Rebuttals (Coherence Defended Again),” The Southern Journal of Philosophy 50 (1). McCain, Kevin. 2014. Evidentialism and Epistemic Justification. New York, NY: Routledge. McGrew, Timothy. 1997. “A Defense of Classical Foundationalism.” In The Theory of Knowledge: Classic and Contemporary Readings, edited by Louis Pojman, 2nd ed., 194–206. Belmont CA: Wadsworth. Moser, Paul K. 1989. Knowledge and Evidence. New York: Cambridge University Press. Moser, Paul K. 1999. “Foundationalism.” In The Cambridge Dictionary of Philosophy, edited by Robert Audi, 2nd ed., 321–23. Cambridge: Cambridge University Press. Neta, Ram. 2011. “A Refutation of Cartesian Fallibilism.” Nous 45 (4): 658–95. Olson, Dustin. 2011. “Bertrand Russell on Perception and Knowledge (1927–59).” M.A. Thesis, McMaster University. Olson, Dustin. n.d. “From ‘the Maxim’ to ‘the Postulates’: The Role of Inference in Russell’s Scientific Philosophy.” Parsons, Charles. 1967. “Mathematics, Foundation of.” In The Encyclopedia of Philosophy, vol. 5, edited by Paul Edwards, 188–213. New York: Macmillan.

310

The Bloomsbury Companion to Bertrand Russell

Pastin, Mark. 1975. “Modest Foundationalism and Self-Warrant.” In Studies in Epistemology, American Philosophical Quarterly Monograph Series, No. 9. edited by Nicholas Rescher, 141–49. Oxford: Blackwell. Pears, David F. 1967. Bertrand Russell and the British Tradition in Philosophy. London: Fontana. Poincaré, Henri. 1905. “Les mathématiques et la logique.” Revue de Métaphysique et de Morale 13: 815–35. Poincaré, Henri. 1906. “Les mathématiques et la logique,” Revue de Métaphysique et de Morale 14: 17–34. Poincaré, Henri. 1906a. “Les mathématiques et la logique,” Revue de Métaphysique et de Morale 14: 294–317. Pojman, Louis (ed.) 1999. The Theory of Knowledge: Classic and Contemporary Readings, 2nd ed. Belmont CA: Wadsworth. Pollock, John. 1986. Contemporary Theories of Knowledge. Savage, MD: Rowman and Littlefield. Rawls, John. 1971. A Theory of Justice, 2nd ed. Cambridge, MA: Harvard University Press, 1999. Rawls, John. 2001. Justice as Fairness: A Restatement. Cambridge, MA: Harvard University Press. Rescher, Nicholas. 1974. “Foundationalism, Coherentism, and the Idea of Cognitive Systematization.” Journal of Philosophy 71: 695–708. Rescher, Nicholas. 1977. Methodological Pragmatism. New York: New York University Press. Russell, Bertrand. 1903. The Principles of Mathematics, 2nd ed. London: Allen and Unwin. Russell, Bertrand. 1906. “The Paradoxes of Logic.” In The Collected Papers of Bertrand Russell, vol. 5: Toward “Principia Mathematica” 1905–08, edited by G. H. Moore, 278–96. London: Routledge. Russell, Bertrand (1907), “The Regressive Method of Discovering the Premises of Mathematics.” In The Collected Papers of Bertrand Russell, vol. 5: Toward “Principia Mathematica” 1905–08, edited by G. H. Moore, 572–80. London: Routledge. Russell, Bertrand (1907a), “On the Nature of Truth.” In The Collected Papers of Bertrand Russell, vol. 5: Toward “Principia Mathematica” 1905–08, edited by G. H. Moore, 439–54. London: Routledge. Russell, Bertrand. 1914. “The Relation of Sense-data to Physics.” In The Collected Papers of Bertrand Russell, vol. 8, The Philosophy of Logical Atomism and Other Essays, 1914–19, edited by G. Slater, 5–26. London: Allen and Unwin, 1986. Russell, Bertrand. 1919. An Introduction to Mathematical Philosophy. London: Allen and Unwin. Russell, Bertrand. 1924. “Logical Atomism.” In The Collected Papers of Bertrand Russell, vol. 9, Essays on Language, Mind and Matter, 1919–26, edited by John G. Slater, 162–79. London: Unwin Hyman.

Russell’s Bridge

311

Russell, Bertrand. 1927a. An Outline of Philosophy. London: Allen and Unwin. Russell, Bertrand. 1927b. The Analysis of Matter. London: Allem and Unwin. Russell, Bertrand. 1940. An Inquiry into Meaning and Truth. London: Allen and Unwin. Russell, Bertrand. 1948. Human Knowledge: Its Scope and Limits. London: Allen and Unwin. Russell, Bertrand. 1956. Logic and Knowledge. Edited by R. C. Marsh. London: Allen and Unwin. Russell, Bertrand. 1959. My Philosophical Development. London: Allen and Unwin. Russell, Bertrand. 1974. Essays in Analysis. Edited by Douglas Lackey. London: Allen and Unwin. Russell, Bertrand. 2012. The Collected Papers of Bertrand Russell, vol. 5: Toward “Principia Mathematica” 1905–08. Edited by G. H. Moore. London: Routledge. Sainsbury, Richard M. 1979. Russell. London: Routledge and Kegan Paul. Sainsbury, Richard M. 1989. “On Induction and Russell’s Postulates.” In Rereading Russell: Essays in Bertrand Russell’s Metaphysics and Epistemology, edited by C. Wade Savage and C. Anthony Anderson, 200–19. Minneapolis: University of Minnesota Press. Sellars, Wilfred. 1973. “Givenness and Explanatory Coherence.” Journal of Philosophy 70: 612–24. Steiner, Mark. 1975. Mathematical Knowledge. Ithaca, NY: Cornell University Press. Walden, Kenneth. 2013. “In Defense of Reflective Equilibrium.” Philosophical Studies 166: 243–56. Whitehead, Alfred N., and Russell Bertrand. 1910–13. Principia Mathematica, 2nd ed., 3 vols. Cambridge: Cambridge University Press. Williams, Michael. 2001. The Problems of Knowledge. Oxford: Oxford University Press.

12

Neutral Monism Christopher Pincock

Abstract Neutral monism is the view that each particular is intrinsically neither mental nor physical, but is instead a member of some neutral metaphysical category. In this chapter I will reconstruct Russell’s changing views on neutral monism. Initially Russell accepted the existence of the self, and took it to be distinctively mental. However, by Problems of Philosophy (1912) Russell admitted that we seem to lack direct acquaintance with the self. Nevertheless, the self continues to play a central theoretical role in Russell’s account of knowledge and judgment. This led Russell to an extended reflection on neutral monism in places like The Theory of Knowledge (1913) and the “Philosophy of Logical Atomism” (1918). In 1918 Russell endorses neutral monism as a goal to be achieved, but remains puzzled about belief. While the 1921 book The Analysis of Mind comes close to accepting neutral monism, there is still a residual resistance to it when Russell posits images that appear to be distinctively mental. By 1925, though, Russell had seen how to eliminate even these particulars, and as a result became a committed neutral monist.1

1 Introduction Russell’s 1914 article “Neutral Monism” begins with a clear statement of this position: “Neutral monism”—as opposed to idealistic monism and materialistic monism—is the theory that the things commonly regarded as mental and the things commonly regarded as physical do not differ in respect of any intrinsic

Neutral Monism

313

property possessed by the one set and not by the other, but differ only in respect of arrangement and context. (Papers 7, 15)2

The things Russell has in mind are what he elsewhere calls particulars. For the neutral monist, particulars are neither intrinsically physical nor intrinsically mental. What makes a particular either physical or mental (or both) is the relations that it stands in to other particulars. Russell began his philosophical career by more or less presupposing that neutral monism is false. However, after the completion of Principia Mathematica, and the more resolute turn to epistemology and metaphysics, Russell took neutral monism more seriously. In this 1914 article Russell argues quite carefully against neutral monism. By the 1918 logical atomism lectures, he admits to great difficulties in deciding whether or not neutral monism is correct. In the Analysis of Mind, published in 1921, Russell seemingly has become a wholehearted advocate of neutral monism. This marks one of the more puzzling shifts in Russell’s philosophical outlook. I will reconstruct this philosophical evolution by distinguishing three epistemic attitudes that one could take on a neutral monist metaphysics. The most qualified endorsement of neutral monism says that no intrinsic difference between particulars is known to us directly. A more confident attitude agrees that no intrinsic difference is given in experience, but goes further and affirms neutral monism on the basis of systematic considerations. In Russell’s case, the shift from a qualified attitude to a more confident attitude toward neutral monism occurs after 1921, largely based on his interpretation of Occam’s Razor. Finally, there is a more strident attitude toward neutral monism that maintains that every particular actually is known to be both mental and physical. In Russell’s terms, this involves showing that every particular is part of a logical construction that leads to a mental object (such as a mind) and also part of a logical construction that leads to a physical object. As I see it, Russell never embraced this strident attitude to neutral monism, and remained with his confident attitude. Russell’s considerations for and against neutral monism focus largely on its ability or inability to make sense of what is distinctive of our mental experiences and activities. In 1898 Russell rejected neo-Hegelian idealism in favor of an extreme form of realism that Hylton has dubbed “platonic atomism” (Hylton 1990, 108). A central plank of platonic atomism is the existence of intrinsically mental particulars, although Russell does not often emphasize this aspect of his realism. In Principles of Mathematics Russell aims to “see clearly, and to make others see clearly, the entities concerned, in order that the mind may have

314

The Bloomsbury Companion to Bertrand Russell

that kind of acquaintance with them which it has with redness or the taste of a pineapple” (Russell 1903, xv). Minds are just those things that can stand in relations of acquaintance to other things. These varying relations are central to our best account of what is responsible for our sensory and intellectual experiences. Russell seems to have initially thought that each mind is acquainted with itself, although no mind is acquainted with any other mind. This is the best interpretation of Russell’s somewhat condensed remarks in “On Denoting”: “Now such things as matter (in the sense in which matter occurs in physics) and the minds of other people are known to us only by denoting phrases, i.e., we are not acquainted with them, but we know them as what has such and such properties” (LK, 56). Here Russell clearly commits himself to the existence of a special sort of thing, minds, and distinguishes them from the matter found in physics. He does not come out and say that each of us is acquainted with our own mind. However, two considerations indicate that Russell thought this at this time. First, if Russell believed that we were not acquainted with our own minds, then he would have said so here. Instead, he carefully restricts his claim to other minds. Second, there are internal philosophical pressures that should strongly incline the defender of platonic atomism to the view that each mind is acquainted with itself. In 1911 Russell argues that his theory of acquaintance requires acquaintance with the self. The argument does not proceed by any principle of transparency. Russell allows that I may be acquainted with X and yet be uncertain that I am acquainted with X. So, the fact that I can doubt that I am acquainted with my self does not settle whether or not I am acquainted with my self. Russell instead proceeds by considering propositions that can be known by introspection. In “Knowledge by Acquaintance and Knowledge by Description” (1911) acquaintance is presented as the basis of our entire mental life. It involves a “dualism of subject and object” that is a “fundamental fact concerning cognition” (Papers 6, 148; 1911, 109). The only alternatives to this dualism that Russell allows in 1911 are “the view that there is no subject, whence we arrive at materialism” or “the view that what is presented is part of the subject, whence we arrive at idealism” (Papers 6, 148; 1911, 109).3 While seeing the sun, I may know that I am seeing the sun. For Russell this requires that I be acquainted with whatever it is that I am seeing. Russell calls these objects sense-data. In some cases, through introspection, I know that I am acquainted with this sense-datum. When I know that I am acquainted with this

Neutral Monism

315

sense-datum, Russell claims that “the complex has been analysed” (Papers 6, 149; 1911, 110).4 The complex here is the whole composed of my self standing in the relation of acquaintance to this sense-datum. And, if this whole has been analyzed, then Russell concludes that I am also here acquainted with my self. He goes on to consider an alternative analysis in terms of a description. However, the only analysis that he mentions is that “I” be “the subject-term in awareness of which I am aware” (Papers 6, 149; 1911, 110). As this circular proposal does not mark any progress, Russell concludes, “It would seem necessary, therefore, to suppose that I am acquainted with myself, and that ‘I’, therefore, requires no definition, being merely the proper name of a certain object” (Papers 6, 149; 1911, 110).5 A few months later when Russell was working on Problems of Philosophy he already noted the limitations of this argument. The same introspection argument now leads to a more tempered conclusion: “In some sense it would seem we must be acquainted with our Selves as opposed to our particular experiences. But the question is difficult, and complicated arguments can be adduced on either side. Hence, although acquaintance with ourselves seems probably to occur, it is not wise to assert that it undoubtedly does occur” (Russell 1912,  51). A more searching investigation of acquaintance and its features was needed to resolve this issue.

2  Three arguments against neutral monism A likely proximate cause of Russell’s second thoughts about acquaintance with the self is his reflections on Moore’s Some Main Problems of Philosophy.6 An additional reason to consider the attractions of neutral monism is Perry’s Present Philosophical Tendencies (Perry 1912). However, the overriding philosophical influence on Russell at this time is Wittgenstein. From their meeting in October 1911 through to the beginning of the First World War, Wittgenstein challenged Russell’s philosophical commitments and methods in fundamental ways. Among other effects, Wittgenstein led Russell to abandon his ambitious Theory of Knowledge manuscript. Russell worked on this project from April through June of 1913. The first six chapters of this book were published as articles in the Monist under the heading “The Nature of Acquaintance.” The second article, corresponding to the second chapter of the book, is the 1914 “Neutral Monism” article.7

316

The Bloomsbury Companion to Bertrand Russell

Russell notes his agreement with neutral monism on the central point that knowledge of objects in experience is direct and unmediated by any idea or content (Papers 7, 22). However, where neutral monism goes wrong is in supposing that “if anything is immediately present to me, that thing must be part of my mind” (Papers 7, 22). In Theory of Knowledge Russell offers a wide range of arguments against this claim, and he takes the failures of neutral monism to offer indirect support for his fundamental cognitive relation of acquaintance. Distinctively mental facts involve acquaintance, and subjects are defined as what stands in acquaintance relations to objects (Papers 7, 35). Three kinds of worries about neutral monism are raised. The weakest argument that Russell gives against neutral monism develops one aspect of his earlier discussions of introspection and self-awareness. Russell first assumes that there are objects that can be experienced by multiple subjects. So, we must distinguish A’s experiencing of O from B’s experiencing of O. Furthermore, A may experience A’s experiencing of O, but B is not capable of experiencing A’s experiencing of O. Crucially, Russell adds that “A can experience his experiencing of O without logically requiring any other experience” (Papers 7, 35). Russell concludes that A’s experiencing of O is different from O. This follows from Russell’s assumptions because if A’s experiencing of O were identical with O, then B’s experiencing of O would just be B’s experiencing of A’s experiencing of O, but this is said to never occur. The argument is a weak one because the neutral monist has no reason to grant these assumptions, and it is not even clear how to square these assumptions with the account of introspection that Russell develops two pages later. There Russell analyzes a subject’s awareness of the fact that it is acquainted with some object using the existential quantifier: S’ – P – [(∃S). S – A – O] (Papers 7, 38) S’ is presented with the fact that some subject S is acquainted with O. This must be Russell’s account of what is going on when we ordinarily say that A is experiencing his experiencing of O. Russell adds that “there is no good reason why the two subjects S and S’ should be numerically the same: the one ‘self ’ or ‘mind’ which embraces both may be a construction” (Papers 7, 38–39). This undermines the above argument against neutral monism as the neutral monist can say exactly the same thing. For the neutral monist what makes A’s experiencing of O different from B’s experiencing of O is the difference between one series that links O to other objects, and which constitutes A, and a second

Neutral Monism

317

series that links O to other objects, which constitutes B. A at t2’s experience of A at t1’s experiencing of O obtains when the right kind of relation obtains between O and some other object O’.8 Russell also argues that the neutral monism developed by James and others has serious difficulties with propositional attitudes. He focuses on the problem of false beliefs and how to distinguish knowing some proposition from other cognitive attitudes. Russell first argues that “belief involves a different kind of relation to objects from any involved in sensation and presentation” (Papers 7, 25). For A to sense O, O must exist. However, for A to believe that O is red, there does not have to be any fact or state of affairs of O’s redness. That is, some beliefs are false. The problem, according to Russell, is that neutral monists assimilate belief to sensation, and so are forced to ignore this basic difference. Related issues arise for James’ view that S’s knowing that p is just S being part of a “process of leading” that links S’s experiences: for James “knowing can easily be explained as a particular sort of relation toward one another into which portions of pure experience may enter” (given at Papers 7, 17). However, Russell argues that many processes leading from S’s desire to find his dog to his finding his dog would fail to constitute knowing where the dog is, for example, if S found the dog by falling into a cellar (Papers 7, 26). The basic problem is that the “logical relation” between a propositional attitude and its object has been lost (Papers 7, 27). The natural response is that neutral monists do not need to assimilate belief to sensation or to identify states of knowing in these simple terms. Russell might counter by noting that neutral monists arrive at their position by supposing that what is present to the mind is a part of the mind. As beliefs are present to the mind, then they are parts of the mind, just as with sensations. However, the determined neutral monist could reject this assumption, and argue for their neutral monism on some different basis. Russell himself here mentions the importance of Occam’s Razor as “the supreme methodological maxim in philosophizing” (Papers 7, 21). As we will see, this is a crucial factor in Russell’s later adoption of neutral monism. So, at best, what these objections based on propositional attitudes show is that the proposals offered by neutral monists are inadequate. In 1913 Russell left open the possibility of acceptable neutral monist accounts of propositional attitudes. Russell’s third and “most conclusive” (Papers 7, 41) consideration against neutral monism emphasizes the differences within the objects of experience at any given time. It is these differences that provide the best sort of indirect

318

The Bloomsbury Companion to Bertrand Russell

evidence for his fundamental cognitive relation of acquaintance. Russell now admits that we are not acquainted with our selves, even when we are presented with the fact that something is acquainted with O. This is why he opts for the analysis of these cases in terms of S’ – P – [(∃S). S – A – O]. How, then, he asks, can we use the word “I,” if it is not a genuine proper name for an object of our acquaintance? The answer turns on a tight series of connections between the words “this,” “I,” and “now.” As Russell summarizes his position, “such “emphatic particulars” as “this” and “I” and “now” would be impossible without the selectiveness of mind” (Papers 7, 40–41). Russell’s positive proposal is that at any given time a subject is acquainted with many objects. There is a second relation of attention that the subject bears to one of the objects of his current acquaintance.9 A subject can name that object using “this.” Once “this” has acquired its referent via attention, “I” and “now” are analyzed descriptively: “The subject attending to ‘this’ is called ‘I’, and the time of the things which have to ‘I’ the relation of presence is called the present time” (Papers 7, 40–41). This is how Russell proposes to make sense of our ability to think of ourselves and our experiences despite our lack of acquaintance with ourselves. In 1913 Russell thus endorses a descriptive analysis of “I” and overcomes his 1911 objections to such an approach. In this way, he frees himself from the uncomfortable assumption that we are acquainted with the self. Now that he admits that we not acquainted with ourselves, he is happy to admit that we lack knowledge of the intrinsic character of the self. Subjects are whatever stand in the acquaintance relation to objects: “We reach subjects only by description, and cannot know whether they are among objects are not” (Papers 7, 35). Mental facts are those that involve the relation of acquaintance, but Russell is unsure about the existence of intrinsically mental particulars. In retrospect, this must be seen as a decisive move toward a neutral monist metaphysics. However, in 1913 Russell used this very position to present his “emphatic particulars” argument against neutral monism: What I demand is an account of that principle of selection which, to a given person at a given moment, makes one object, one subject and one time intimate and near and immediate. . . . In a world where there were no specifically mental facts, is it not plain that there would be a complete impartiality, an evenly diffused light, not the central illumination fading away into outer darkness, which is characteristic of objects in relation to mind? (Papers 7, 40)

This problem does not concern how to make sense of the difference between those objects that make up A’s experience and those objects that make up

Neutral Monism

319

B’s experience. An allocation of objects to subjects does not yet address the problem of emphatic particulars. Russell supposes that each subject’s experience at each moment has an internal structure that involves a center and increasingly peripheral bands of objects. Russell claims that the neutral monist is unable to make sense of this partiality within the experience of a given subject. The significance of this objection is made clear in a letter that Russell wrote to Carr in April 1918 as Russell prepared to enter prison. Now, five years after Theory of Knowledge, Russell distinguishes three issues for the neutral monists: sensation, belief, and emphatic particulars. This [neutral monism] works admirably so far as sense is concerned, but raises great difficulties as regards Belief. If Belief could be dealt with on James’s lines, I think desire, volition and feeling could be dealt with. I do not see clearly whether a Theory of Belief on such lines is possible or not, but I wish to find out. There is a special problem which raises difficulties, namely the problem of “emphatic particulars,” such as “this,” “here,” and “now.” (Thompson 1975, 18)

The next three sections will consider how Russell overcame these three problems in his logical atomism lectures and the work leading up to and including 1921’s Analysis of Mind. Once these problems were solved, Russell was in a position to give a qualified endorsement to the neutral monist metaphysics that he had previously rejected. However, even then he did not express much confidence in neutral monism based on systematic reasons until considerations from physics gave the final push.

3 Sensation From January through March 1918 Russell gave eight lectures that were published with the title The Philosophy of Logical Atomism. Neutral monism is presented in these lectures as a serious alternative to Russell’s own metaphysics. At one point Russell mentions the earlier discussion of neutral monism that we have just reviewed, but cautions his audience that “some of the arguments I used against neutral monism are not valid” (Papers 8, 195–96). He does not say which arguments he has in mind, but we can infer that he now rejected the arguments tied to the difference between some object, A’s experiencing that object and B’s experiencing that object. Worries about belief and emphatic particulars are

320

The Bloomsbury Companion to Bertrand Russell

raised in the lectures, while the earlier worries concerning sensation are absent. What had prompted Russell to withdraw this line of attack on neutral monism? Russell’s last lecture presents two “purpose[s]” for the lectures as a whole. The first motivation is the difference between simples and facts. Simples include not only particulars, but also qualities and relations (Papers 8, 234). These simples are the “atoms” that give “logical atomism” its distinctive metaphysics. The only genuinely complex entities that remain are facts. Russell now endorses a metaphysical distinction between simples and facts based on the logical point that you cannot name facts. This insight is credited to Wittgenstein (Papers 8, 199), and is absent from the Theory of Knowledge manuscript.10 The second purpose that Russell notes here is Occam’s Razor. As Russell interprets it, this maxim requires one to minimize the number of kinds of “undefined things” and “undemonstrated premisses” that are needed to recover the collection of propositions that make up a science like physics (Papers 8, 235). Some investigations along these lines were developed in Russell’s 1914 Our Knowledge of the External World and related papers. The basic idea is to replace a posited entity with a class (or series of classes) whose members are sense-data or potential sense-data (“sensibilia”). This replacement restricts one’s posits to entities whose kind overlaps with entities that are to be found in one’s experience. It is surely not a coincidence that the “ultimate constituents” (Papers 8, 236) that appear in Russell’s logical constructions of ordinary physical objects have the simplicity required by the categorical distinction between logical simples and facts. The two aims for his lectures that Russell here identifies work together to force his dramatic reinterpretation of physics. In addition, Russell applies the very same technique to persons (Papers 8, 241).11 The basic idea is to take whatever relation one uses to collect experiences as the experiences of a single person, and to use that relation to define a person as a series of classes of sense-data. There is thus no need to posit “a sort of pin-point ego” (Papers 8, 239) that unites all of these experiences. In addition, many of the elements that we collect in one way in the logical construction of ordinary physical objects like chairs will also appear when we collect elements in the logical construction of persons. These two sorts of logical constructions serve to effectively rebut Russell’s argument from 1913 that experience involves a relation between a subject and an object: “That actual appearance that the chair is presenting to me now is a member of me and a member of the chair, I and the chair being logical fictions. That will be at any rate a view that you can

Neutral Monism

321

consider if you are engaged in vindicating neutral monism. There is no simple entity that you can point to and say: this entity is physical and not mental” (Papers 8, 241). One component of his earlier resistance to neutral monism has been removed. I conjecture that it is the surprisingly wide scope for this project of logical construction that led Russell to take neutral monism much more seriously. However, it is a mistake to think that Russell endorses neutral monism simply because he rejects the “pin-point ego.” The worries tied to belief and emphatic particulars lead Russell to stop short of endorsing neutral monism in 1918. There is also a new and increasingly urgent problem for neutral monism tied to experiences that are not sensations. Recall that in the letter to Carr Russell says only that neutral monism “works admirably so far as sense is concerned.” However, in Theory of Knowledge and Philosophy of Logical Atomism Russell emphasizes that our experience is made up of more than just sense-data. In addition, there are images, or what Russell sometimes calls phantoms or hallucinations (Papers 8, 224). What distinguishes a sensation from an act of imagination is not the reality of the objects. Russell insists that both objects, the sense-datum and the image, are real. The only difference is that the phantom lacks the usual relations to other sense-data: “The phantom is in itself just as much part of the world as the normal sense-datum, but it lacks the usual correlation and therefore gives rise to false inferences and becomes deceptive” (Papers 8, 225). This way of distinguishing between sense-data and images basically guarantees that a genuine sense-datum will appear in both the logical constructions of physical objects and of persons. This is why Russell admits that one will be unable to point to a particular that is physical, but not mental. However, Russell has no reason to accept the converse claim. If some of my experiences are images, then I may very well be able to point to particulars that are mental, but not physical. The “usual correlation” that groups sense-data are the relations used in the logical constructions that make sense-data physical. Images are special because they fail to stand in those very relations. They should, of course, stand in the relations that make them part of a person’s experience, for example, memory. As such, images stand as a stark counterexample to the strident neutral monist claim that every particular is known to be both mental and physical. A more qualified attitude to neutral monism need not include this strong assertion and can allow particulars that fail to be part of physical objects. The core claim is just that there is nothing special intrinsically about these particulars.

322

The Bloomsbury Companion to Bertrand Russell

The assumptions needed for this sort of objection to strident neutral monism are present in Theory of Knowledge, but it took Russell until 1919’s “On Propositions: What They Are and How They Mean” to present the argument explicitly. There Russell singles out “images of public sensations, i.e., especially visual and auditory images” (Papers 8, 286) such as imagining a friend in a chair. These images are not physical because they cannot be located in physical space: you cannot locate the image in the body because it is visual, nor (as a physical phenomenon) in the chair, because the chair, as a physical object, is empty. Thus it seems that the physical world does not include all that we are aware of, and that introspection must be admitted as a source of knowledge distinct from sensation. (Papers 8, 287)12

In this way, introspection still serves to block Russell’s acceptance of strident neutral monism, even though there are no longer any problems with sensedata and sensation. I conclude that even if the problems for belief and emphatic particulars had been resolved, the problem of locating images would have remained. In fact, as we will see in the next two sections, images became Russell’s means of solving his own difficulties with belief and emphatic particulars. The viability of neutral monism thus comes to turn on the possibility of a neutral monist account of images.

4 Judgment One of the more puzzling questions for Russell scholars is what convinced Russell to abandon the theory of judgment from Theory of Knowledge. It is clear that Wittgenstein’s objections to this theory in 1913 were a decisive factor, but debates continue concerning exactly what the objection was and how Russell understood it.13 According to the multiple-relation theory of judgment there are no freestanding objectively existing propositions composed of the objects of the judgment. The act of judgment takes center stage. It involves the subject standing in a complicated relation (or series of relations) to the objects that the judgment is about. For example, when Othello judges that Desdemona loves Cassio, Othello bears a relation to Desdemona, loving and Cassio. In this way, Russell is able to avoid freestanding false propositions. Still, the judgment fact contains the objects that the judgment is about, and it is true just in case a certain fact obtains in the world.

Neutral Monism

323

In Theory of Knowledge when Russell wrote that neutral monists ignore the “logical relation” between a belief and what makes that belief true or false, he must have thought that his own multiple-relation theory avoided this problem. However, Wittgenstein convinced Russell that the theory was flawed. The flaw turns either on what I have called the proposition problem or the correspondence problem. The proposition problem involves saying which complexes involving the subject count as judgments, and which judgments they involve. The correspondence problem is to say what fact in the world must obtain for a given judgment to be true. In The Philosophy of Logical Atomism lectures Russell advances a peculiar two-verb fact account of judgment that preserves some elements of the earlier multiple-relation theory. He admits that “the theory of judgment which I set forth once in print some years ago was a little unduly simple, because I did then treat the object verb as if one could put it as just an object like the terms, as if one could put ‘loves’ on a level with Desdemona and Cassio as a term in the relation ‘believe’” (Papers 8, 199). In 1911 Russell had indeed supposed that Othello stands in a relation to the objects Desdemona, loving and Cassio, and had not distinguished any special way for the relation loving to occur in the judgment complex. In 1918, however, Russell insists that “when A believes that B loves C, you have to have a verb in the place where ‘loves’ occurs” (Papers 8, 198). So, a fact of this sort not only has a relation or “verb” for believing, but another relation or “verb,” occurring as a relation, for loving. This is so even when B does not love C. This is what Russell calls “the puzzle about the nature of belief ” (Papers 8, 198). Russell tentatively suggests that we can resolve this puzzle by positing a new sort of fact over and above the atomic facts such as that A loves B and that A is not to the right of B. Wittgenstein is said to have discovered this new sort of fact (Papers 8, 198), but unfortunately Russell does not say what convinced either Wittgenstein or Russell to countenance these facts. It is likely to have to do with either the proposition problem or the correspondence problem. The significance of neutral monism for understanding this shift is that Russell sees neutral monism as a way to avoid two-verb facts. As we have seen, James proposed an account of belief and knowledge that tied these propositional attitudes to the behavior of individuals. In the 1918 lectures Russell supposes that neutral monism mandates a thoroughgoing behaviorism for such states.14 For example, S’s belief that a train leaves at a certain time is just S’s acting in a such and such a way. The truth of the belief is analyzed in terms of the satisfaction

324

The Bloomsbury Companion to Bertrand Russell

of S’s aims. Russell’s description of the view anticipates his own later account of belief in Analysis of Mind: The logical essence .  .  . will be a relation between two facts having the same sort of form as a causal relation, i.e. on the one hand there will be your bodily behaviour which is one fact, and on the other hand the fact that the train starts  at  such and such a time, which is another fact, and out of a relation of those two the whole phenomenon is constituted. . . . It is quite a different logical form from the facts containing two verbs that I am talking of today. (Papers 8, 195)

Russell sees here that one could try to make sense of the truth or falsity of a belief in ordinary causal terms. The difficulty for this proposal is that the belief is no longer a fact that has what it is about as its parts. In 1918 Russell is unwilling to give up this core commitment. In the lectures Russell notes that the behaviorist strategy of the neutral monist requires that beliefs never occur as an “isolated phenomenon” (Papers 8, 193). What it is to have a belief, and the truth or falsity of that belief, are analyzed in terms of the person’s behavior. But this “is a subject belonging to psychology” (Papers 8, 193). While in prison Russell read some psychology, with a special emphasis on behaviorist approaches to belief. He found behaviorism to be inadequate, primarily because of its inability to deal with images. However, Russell pressed forward with this attempt to concoct an account of belief that would avoid two-verb facts. He hit on the plan of identifying a belief-fact with a fact composed of mental images standing in relations. The belief ’s truth or falsity is thus determined by what these images are about, and whether or not these targets stand in the appropriate relation. These image propositions mark a decisive rejection of the core platonic atomist assumption that a belief has as parts the objects that the belief is about. Now the belief only has images as parts, and the intentionality of the belief is secured via the links between those images and what they are about. This new approach to belief is unveiled in “On Propositions” and occupies a central place in 1921’s Analysis of Mind. As Russell summarizes the view, The most important thing about a proposition is that, whether it consists of images or of words, it is, whenever it occurs, an actual fact, having a certain analogy of structure—to be further investigated—with the fact which makes it true or false. A word-proposition, apart from niceties, “means” the corresponding imageproposition, and an image-proposition has an objective reference dependent upon the meanings of its constituent images. (Papers 8, 297)

Neutral Monism

325

All Russell says here to account for this shift is that his earlier multiple-relation view “is rendered impossible by the rejection of the subject” (Papers 8, 295). However, it is not clear if Russell has independent motivations for rejecting the subject or if his main reason for rejecting the subject is this shift in his view of propositions. He does insist that “the act, or subject, is schematically convenient, but not empirically discoverable. It seems to serve the same sort of purpose as is served by points and instants, by numbers and particles and the rest of the apparatus of mathematics” (Papers 8, 294). A determined defender of a multiplerelation approach could admit the need to logically construct the self, but view this construction as merely an addition to the account of judgment.15 Of course, this move would not address whatever worried Russell about the multiplerelation theory. A reasonable conjecture, then, is that Russell felt pressures of two sorts: first, to revise his account of propositions and judgment and, second, to extend his logical constructions into the domain of the mental. He certainly does emphasize the virtues of his new approach to propositions in particular: image propositions have “the advantages .  .  . derived from the rehabilitation of the content, making it possible to admit propositions as actual complex occurrences, and doing away with the difficulty of answering the question: what do we believe we believe falsely?” (Papers 8, 296). This is what I have called the correspondence problem. Image propositions also seem well-placed to address Russell’s judgment problem. Clearly, if Russell thought that neutral monist had an adequate account of images, then he would have had a satisfactory response to his earlier worries about belief and neutral monism. But, as we have seen, in this paper Russell argues that images are distinctively mental entities (Papers 8, 295). The subject has been eliminated and plays no role in belief, but new mental entities have taken its place. Russell thus rejects neutral monism in 1919.

5  Images and emphatic particulars It is commonly supposed that Russell accepted neutral monism in his 1921 book The Analysis of Mind. As noted in the introduction, Russell only affirms a qualified form of neutral monism in 1921: no intrinsic difference is known directly. The more confident attitude is based on Occam’s Razor. Russell explicitly defends the claim that every particular is neither mental nor physical starting only in 1925. Famously, the conclusion of The Analysis of Matter says: “There is therefore no

326

The Bloomsbury Companion to Bertrand Russell

ground for the view that percepts cannot be physical events” (OOP, 384). Prior to 1925 there is little evidence that Russell supported anything more than the qualified view that no intrinsic difference is given in experience. As we will see, there are even passages in 1921’s Analysis of Mind that suggest a rejection of neutral monism. Russell’s concluding discussion in that book fits best with the qualified attitude as he there sketches out a future program of research that he suggests would be needed to be anymore confident in the neutral monist metaphysics. As one would expect from “On Propositions,” the viability of neutral monism turns on making sense of images. In the first chapter of Analysis of Mind Russell says, “I should say that images belong only to the mental world” (AMi, 21, noted at Papers 8, xxii). As in “On Propositions,” images obey different laws from sensations, and so resist incorporation into the logical construction of physical objects and matter. However, in 1921 Russell is less enamored of the location argument that he deployed in 1919: If regarded as a sensation, my image has all the marks of the supernatural. My image, therefore, is regarded as an event in me, not as having that position in the orderly happenings of the public world that belongs to sensations. By saying that it is an event in me, we leave it possible that it may be physiologically caused: its privacy may be only due to its connection with my body. (AMi, 153)

Elsewhere I have argued that Russell is here influenced by the materialistic account of memory proposed by Semon (Pincock 2006). Semon claimed that sensations leave a trace or engram in the brain that is then capable of generating images on later occasions. This special sort of “mnemic” causation suggests that images may be parts of physical objects after all, namely brains. More generally, as Russell puts the point, “it is by no means certain that the peculiar causal laws which govern mental events are not really physiological” (AMi, 139). Russell’s concluding chapter “Characteristic of Mental Phenomena” is meant to draw out the conclusions of his earlier analysis for the mental as a whole. Russell notes “the suggestion that many things which seem essentially mental are really neural” (AMi, 295). He thinks he has shown that beliefs, desires and other mental states are all just “complex phenomena consisting of sensations and images variously related” (AMi, 300). As sensations are distinguished from images in a way that guarantees that each sensation is part of a physical object, “it is the causation of images that is the vital problem” (AMi, 303). Here Russell considers two options. One is that Semon’s materialistic analysis of mnemic causation is correct, and so images will bear

Neutral Monism

327

causal relations of the sort found in physiology. If so, then every image will be part of the logical construction of some physical object. A second option is that “mnemic causation is ultimate.” What this means is that the trace left by a past sensation is causally efficacious in generating a present memory image, but this causation is not mediated by any ordinary physiological process. Understandably Russell says that Semon’s “materialistic answer” is more likely to be true (AMi, 303). However, Russell also raises severe reservations about materialism about minds. This is the view that the subject matter of psychology is merely tacked on to the subject matter of physics. Instead, what would be preferable would be a science that dealt directly with the particulars encountered in the logical constructions that both of these sciences rely on: if our scientific knowledge were adequate to the task, which it neither is nor is likely to become, it would exhibit the laws of correlation of the particulars constituting the momentary condition of a material unit, and would state the causal laws of the world in terms of these particulars, and not in terms of matter. (AMi, 306)

These laws will either be differential equations or “finite-difference equations, if the theory of quanta should be correct” (AMi, 306n). Russell is thus wellmotivated to turn to a detailed engagement with the physics of his time. This project culminated in 1927’s Analysis of Matter, which we turn to in the next section. Given the prominence that the argument from emphatic particulars had received in 1914 and 1918, it is somewhat disappointing to find Russell offering no explicit discussion of these issues in 1921. However, a passage in this last chapter of Analysis of Mind shows how Russell may have thought he could use images to solve the problem. Recall that the challenge is to single out one sensation that is at the “center” of one’s experience. In certain circumstances, Russell says that an image will perform this function: “When a sensation is followed by an image which is a ‘copy’ of it, I think it may be said that the existence of the image constitutes consciousness of the sensation” (AMi, 289). Thus the selective role that was previously performed by the self is now performed by an appropriately caused image. However, Russell is quick to add that the image alone is not sufficient for this sort of consciousness: this image constitutes consciousness only “provided it is accompanied by that sort of belief which, when we reflect upon it, makes us feel that the image is a ‘sign’ of something other than itself ” (AMi, 289). The link to the emphatic

328

The Bloomsbury Companion to Bertrand Russell

particular “this” is then made explicit: “This is the sort of belief which, in the case of memory, we expressed in the words ‘this occurred’” (AMi, 289). Other images can be referred to the present. The combination of images and beliefs about those images is needed to make sense of our ability to single out “this” sensation. This is what consciousness amounts to in 1921. Russell does not here talk of “emphatic particulars” or use the evocative metaphor of selective illumination to characterize our experience. However, he remained attached to this conception of experience. The 1940 book Inquiry into Meaning and Truth contains an entire chapter on “egocentric particulars.” By this point Russell uses the term “percept” to refer to what were formerly either sensations or images. This is a consequence of his more resolute embrace of neutral monism around 1925. After this shift Russell is still happy to talk of the contrast between physics and perception: “Physics views space-time impartially, as God might be supposed to view it; there is not, as in perception, a region which is specially warm and intimate and bright, surrounded in all directions by gradually growing darkness” (IMT, 102). As in 1914, this aspect of our experience is tied up with our use of the words “this” and “I.” Now Russell’s account of “this” is explicitly marked as physiological. In certain cases a “verbal reaction to a stimulus” is “immediate” in the sense that “the afferent current runs into the brain and continues along the efferent nerve until it affects the appropriate muscles, and terminates in ‘this is’.” In that situation the speaker’s words obtain the “temporary uniqueness” that is distinctive of “this” and the puzzle about egocentric particulars is resolved (IMT, 105–06). “I-now” is then “a set of occurrences” that co-occur with “this,” and “I” is fleshed out by the appropriate causal relations to this “I-now” (IMT, 107). The upshot is that egocentric particulars “are not needed in any part of  the description of the world, whether physical or psychological” (IMT, 108). This marks Russell’s considered response to his most long-standing objection to neutral monism.

6  Neutral monism and physics Russell takes the development and confirmation of the general theory of relativity (GTR) as the decisive factor in favor of his considered support for a more confident attitude to neutral monism. One implication of GTR for Russell is that particulars are now thought of as events. Some events are perceptions, or

Neutral Monism

329

what Russell calls percepts. Around 1925 Russell concludes that percepts should be identified with events that occur in the space-time of physics, and be located in the perceiver’s brain. Thus Russell arrives at the peculiar position defended in the Analysis of Matter: “The space of physics is connected with causation in a manner that compels us to hold that our percepts are in our brains, if we accept the causal theory of perception, as I think we are bound to do” (AMa, 383). On the causal theory, a percept is a perception of some event, such as a flash of lightning, when there is an appropriately related series of events that begins with the flash and ends with the percept in the perceiver’s brain. As all percepts are events in a person’s brain, there is no difficulty locating them in the spacetime of physics. Every known psychological event is a percept, and so every known psychological particular is also a physical particular. At earlier stages, a sensation could have as its object something that was spatio-temporally distant from the person, as in Russell’s famous discussion of the table in Problems of Philosophy. Developments in physics thus convince Russell that every known mental particular is also physical. But is there any reason to think that every physical particular is also mental, or at least that there is no intrinsic difference between them? This worry is now addressed from two directions. First, Russell is now explicit that events where there are no brains are not likely to have exactly the same qualitative character as the percepts that each of us is aware of. This is because the presence of brains is likely to change the character of percepts. Still, there is a “principle of continuity” that we should use to relate percepts to other events. Second, Russell now argues that we know only the structural relations that obtain between events that are not percepts and that this structural character is all that is required for the truth of physical theories like GTR. The denial of knowledge of the intrinsic qualities of events that are not percepts is a crucial ingredient in Russell’s confidence in neutral monism. For given that we do not know the qualitative character of these events, there is no barrier to ascribing them a similar qualitative character to percepts. As there is no barrier, the principle of continuity counsels us to make this ascription. The result is that there are “no good grounds for excluding percepts from the physical world” (AMa, 384). As there is no reason to exclude them, and some reason to include them, the balance of reasons supports a neutral monist metaphysics. The core argument for admitting events that are not percepts is that this expanded class of events provides the best overall interpretation of the laws of physics. Ultimately, then, Russell is deploying his version of Occam’s Razor. This

330

The Bloomsbury Companion to Bertrand Russell

is most clear when Russell considers sparser alternatives that try to get by just with the percepts of one individual or the percepts of all individuals. Against these alternatives, Russell presents a causal theory of perception that adds additional events, and supposes that these events are “causally continuous with the visual percept” (AMa, 216). Summing up, he says that The essential points are (1) the arrangement about a centre, (2) the continuity between percepts and correlated events in other parts of the space derived from percepts and locomotion. This first is a matter of observation; the second is a hypothesis designed to secure simplicity and continuity in the laws of correlation suggested by the grouping of percepts. It cannot be demonstrated, but its merits are of the same kind as those of any other scientific theory, and I shall henceforth assume it. (AMa, 217)

It is difficult to pinpoint when Russell arrived at this position. In the 1924 paper “Logical Atomism” this version of neutral monism in terms of events is presented as “an outline of a possible structure of the world” (Papers 9, 177). The unpublished 1925 “Philosophical Analysis of Matter” seems to go a step further by ending with a discussion of neutral monism and calling it the view that “does justice to the strong points” of both idealism and materialism (Papers 9, 284). In a summary of lectures published in 1926 called “Perception,” Russell finally comes out and says neutral monism is the theory that “I believe to be true” (Papers 9, 187). The exact same language is used in Outline of Philosophy from 1927, which presents these lectures. Russell’s route to a confident endorsement of neutral monism, then, is somewhat convoluted and elaborate. The overriding consideration is Occam’s Razor and developments in physics. However, the way was surely prepared by the problems that Russell ran into in making sense of acquaintance with the self and his multiple-relation theory of judgment. Once Wittgenstein convinced him that particulars must be simple, and that only facts can correspond to facts, Russell was motivated to find simple, neutral particulars as the sole ingredients of his logical constructions of everything else. Given these twists and turns it is somewhat surprising to find a newly energized debate about “Russellian monism” (Wishon 2015). Philosophers of mind, in particular, wish to find out if Russell’s position offers a viable alternative to physicalism or whether it is perhaps just a version of physicalism in disguise. Our reconstruction of Russell’s arguments has emphasized Russell’s determination to solve his philosophical dilemmas, even at the cost of radically revising his metaphysics. This is a lesson that contemporary metaphysics is all too eager to learn.

Neutral Monism

331

Notes 1 I am grateful to the editor and Thomas Pashby for very helpful comments on an earlier draft of this chapter. There is no consensus on when and why Russell adopted neutral monism. For alternative accounts and additional references, I would recommend Tully 2003, Landini 2011, ch. 6, Bostock 2012, ch. 10–11 and Wishon 2015. 2 This article was an early chapter in Russell’s abandoned Theory of Knowledge manuscript. I consider this manuscript in section II. 3 It is difficult to know why Russell thought the options were dualism, materialism, and idealism. Cf. Ward’s Naturalism and Agnosticism, lecture 18: “Neutral or agnostic monism tends to degenerate into materialism; but it might logically advance to idealism” (Ward 1899, Vol. II, 205). Russell is thanked for comments on proofs in the preface to the first edition. However, in Analysis of Matter Sheffer is credited with the phrase “neutral stuff ” (AMa, 10). 4 Russell distinguishes between an acquaintance with the complex as a whole and the acquaintance with the complex’s parts. However, he struggles to decide if acquaintance with the whole involves acquaintance with the parts. See esp. “Analysis and Synthesis” (Papers 7, 119–28). 5 In 1917 Russell adds “or to find some other analysis of self-consciousness” (Papers 6, 149). 6 See, for example, Moore on “acts of consciousness” (Moore 1953, 4). Lectures 1–10 are the unpublished writings mentioned in the preface to Problems of Philosophy, as noted in Moore’s preface to Moore 1953. Russell 1911 was read in March 1911, while Problems of Philosophy was completed in Aug. 1911 (Papers 6, lxv). 7 We lack manuscripts for these articles, so it is not possible to know if anything was changed when Russell published these articles (Papers 7, xxviii). 8 Russell’s account of consciousness in Analysis of Mind has exactly this structure, but with images playing the role of object O’. It is also worth noting Russell’s bold assertion that a mind could exist for a “fraction of a second” and experience only one object (Papers 7, 23). However, this statement is not much of an objection to neutral monism. 9 Russell adds the puzzling qualifier “or at most some very small number of objects” (Papers 7, 40). This may be proposed to make sense of a subject’s attending to both a present and past particular. See Papers 7, 73 on immediate memory. 10 For example, “there is certainly a one-one correspondence of complexes and facts, and for our present purposes we shall assume they are identical” (Papers 7, 80). 11 In 1914’s “Relation of Sense-Data to Physics” Russell still speaks of “a mental particular as a constituent” of the world (Papers 8, 8, noted by Wishon 2015). This suggests that he is not yet ready to view the subject as a logical construction,

332

12 13 14

15

The Bloomsbury Companion to Bertrand Russell even though he is here showing how to construct physical objects from sense-data and sensibilia. However, in a note added in 1917 to Russell 1911, Russell says that he uses the phrase “a mind” “merely to denote the something psychological which enters into judgment, without intending to prejudge the question of what this something is” (Papers 6, 161). So, minds or subjects may now be targets for logical construction. This argument can be traced to Russell’s prison manuscripts. See Papers 8, 260, 322. Russell is especially focused on the work of Knight Dunlap on introspection. See Pincock 2008, Bostock 2012, ch. 12 and MacBride 2013 for some recent discussions. Russell admits that neutral monism does not entail behaviorism. One suspects that even at this stage Russell was considering non-behaviorist approaches to belief that could vindicate neutral monism. See footnote 11 for some hints of this strategy in 1917.

References Bostock, David. 2012. Russell’s Logical Atomism. Oxford: Oxford University Press. Hylton, Peter. 1990. Russell, Idealism and the Emergence of Analytic Philosophy. Oxford: Oxford University Press. Landini, Gregory. 2011. Russell. New York: Routledge. MacBride, Fraser. 2013. “The Russell-Wittgenstein Dispute: A New Perspective.” In Judgement and Truth in Early Analytic Philosophy, edited by M. Textor, 206–41. Basingstoke, UK: Palgrave Macmillan. Moore, George E. 1953. Some Main Problems in Philosophy. London: Allen and Unwin. Perry, Ralph B. 1912. Present Philosophical Tendencies. New York: Longmans Green. Pincock, Christopher. 2006. “Richard Semon and Russell’s Analysis of Mind.” Russell 26: 101–25. Pincock, Christopher. 2008. “Russell’s Last (and Best) Multiple-Relation Theory of Judgement.” Mind 117: 107–39. Russell, Bertrand. 1903. Principles of Mathematics. Cambridge: Cambridge University Press. Russell, Bertrand. 1911. “Knowledge by Acquaintance and Knowledge by Description.” Proceedings of the Aristotelian Society 11: 108–28. Russell, Bertrand. 1912. Problems of Philosophy. Oxford: Oxford University Press. Russell, Bertrand. 1921. The Analysis of Mind. London: Allen and Unwin. Russell, Bertrand. 1927. An Outline of Philosophy. London: Allen and Unwin. Russell, Bertrand. 1927–54. The Analysis of Matter. New York: Dover. Russell, Bertrand. 1940–62. An Inquiry into Meaning and Truth. London: Penguin. Russell, Bertrand. 1956–92. Logic and Knowledge. New York: Routledge.

Neutral Monism

333

Russell, Bertrand. 1984. Theory of Knowledge: The 1913 Manuscript. Collected Papers of Bertrand Russell, vol. 7. Edited by E. R. Eames. London: Allen and Unwin. Russell, Bertrand. 1986. The Philosophy of Logical Atomism and Other Essays: 1914–19. Collected Papers of Bertrand Russell, vol. 8. Edited by J. G. Slater. London: Allen and Unwin. Russell, Bertrand. 1988. Essays on Language, Mind and Matter: 1919–26. Collected Papers of Bertrand Russell, vol. 9. Edited by J. G. Slater. London: Unwin Hyman. Russell, Bertrand. 1992. Logical and Philosophical Papers: 1909–13. Collected Papers of Bertrand Russell, vol. 6. Edited by J. G. Slater. New York: Routledge. Thompson, Michael. 1975. “Some Letters of Bertrand Russell to Herbert Wildon Carr.” Coranto 10: 7–19. Tully, Robert E. 2003. “Russell’s Neutral Monism.” In Cambridge Companion to Russell, edited by N. Griffin, 332–70. Cambridge: Cambridge University Press. Ward, James. 1899. Naturalism and Agnosticism. Two volumes. London: Macmillan. Wishon, Donovan. 2015. “Russell on Russellian Monism.” In Consciousness in the Physical World: Perspectives on Russellian Monism, edited by T. Alter and Y. Nagasawa, 91–118. Oxford: Oxford University Press.

13

Bertrand Russell’s Moral Philosophy Ray Perkins, Jr.

Introduction Contrary to long-standing philosophical consensus, Bertrand Russell did do serious work in ethics (though often unnoticed) throughout his long and distinguished career. And thanks largely to Charles Pigden that fact is becoming recognized, and deservedly so, given its rich variety, high quality, and often remarkable originality.1 This chapter on Russell’s moral philosophy will focus mainly on its theoretical side, although in the last section we’ll glimpse an instructive example of its practical application as well. Shortly after Russell and Moore liberated themselves from the neo-Hegelian confines at Cambridge in the late 1890s, Russell adopted—though with important critical differences—the ideal utilitarianism of Moore’s Principia Ethica (1903)2 and defended it until 1913 when he was persuaded by Santayana’s Winds of Doctrine to move away from moral objectivism toward a more subjective theory of value.3 Over the next two decades Russell developed an original version of emotivism (in its early, rudimentary form merely that moral judgments are matters of feeling/desire; there is no moral knowledge) which he gradually brought to a mature form in 1935 in a single (and little-noticed) chapter of Religion and Science (RS IX). Russell was one of the theory’s earliest pioneers. His mature theory was in several ways more sophisticated than those of later and much better-known emotivists, for example, Ayer and Stevenson. He was never quite satisfied with the subjectivity of value and even tried to give it a naturalistic overhaul in Human Society in Ethics and Politics (HSEP, 1954, ch. 9). But he soon returned to emotivism at a time when his chief concern was not moral theory but the practical moral problem of war and peace in a nuclear age and the need for deep sociopolitical reform.

Bertrand Russell’s Moral Philosophy

335

There was another earlier, and much briefer, deviation from his emotivist development. In 1922, he presented a brilliant little paper (two pages!) to the secret Cambridge Apostles society, “Is There An Absolute Good?”—one of the earliest versions of what J. L. Mackie made famous decades later as “error theory”: value judgments are objective assertions purporting to be about moral facts; but all such statements are false. The paper, first published in 1987, is still little known.4 We’ll get a more detailed and critical look at Russell’s theoretical ethics—both its normative and metaethical features—in parts I and II below. Although our focus is on theory, practical application was always of paramount importance to Russell. In part III, as a case study, we’ll look at his public repudiation of nuclear deterrence (and its practitioners, both East and West)—remarks widely criticized at the time, though a model of courage and insight in the art of moral persuasion.

1  A decade with the Moorean good Moore’s highly influential Principia appeared in 1903 (along with Russell’s Principles of Mathematics, both widely praised as two of the earliest exemplars of the new twentieth-century analytic philosophy). Russell wrote two reviews of Moore’s book in 1903 and 1904 (Papers 4, 568–75) and a long essay in 1905, “The Elements of Ethics” (publ. 1910), largely based on it.5 “Elements” develops the main themes of Moore’s utilitarianism with the analytic clarity that Russell became known for. “Elements,” and his reviews of Principia, made original criticisms that anticipated those later made by wellknown ethicists (often without knowledge, or mention, of Russell).6 He accepted the main features of Moore’s ethics, including much of the Platonic metaphysics that underpinned the “nonnatural” character of good. Good, unlike the property of (say) pleasure or red, was not an empirical (“natural”) property; it was a universal; it had being, but not existence in time.7 There are two main areas where one finds important differences between Moore and Russell: the definability of moral terms, and the nature of moral obligation.

1.1  Analysis and moral concepts In his 1904 review of Principia Russell praises Moore for his discovery of what Moore called the “naturalistic fallacy,” according to which it’s a mistake to try

336

The Bloomsbury Companion to Bertrand Russell

to define good: Good is simple; good is good; that’s it. The main argument that good can’t be defined is what has come to be known as Moore’s “open question argument.”8 A definition (in the required sense) is an analysis of a concept (property) into its constituent parts and their relations. The meaning of a concept-word is the property it names. Consider: Brother = df male sibling

If the definition gives the true meaning of the concept/word to be defined, then the given meaning will be synonymous with that of the word (or concept) being defined, and the statement “A brother is a male sibling” will be trivially true (for competent English speakers), just as “A brother is a brother” is. And a question like “Is a male sibling really a brother?” will not be a question conducive to differing opinions, anymore than will “Is a brother really a brother?,” that is, it will not be an “open” question. Proposed definitions that raise open questions are incorrect definitions. This definitional error, exposed by Moore’s open question argument, was called the “naturalistic fallacy”—a fallacy that hedonistic utilitarians like Mill and Bentham (who identified goodness with pleasure) had allegedly committed. Russell accepted Moore’s open question argument which both took (somewhat hastily I think) as conclusive evidence that good was simple and indefinable. In fact he used it against Moore’s own Principia definition of “ought to do” (as “the action that will produce the most good”), on the grounds that it is not a genuine definition since one can significantly ask “But ought one really to do such an action?”9 In his later Ethics (1912), Moore was careful not to call his characterization of duty (act productive of the most good) a “definition”; he calls it a “criterion.” Two years earlier in “Elements” (in PE, 33), Russell had clarified for the reader that what he (Russell) had just called (32) a “definition” of right was not a proper definition in the sense of an analysis of the meaning of the word; it was a “mark” of the features necessary and sufficient for an action’s rightness. It does appear that Russell was correct, given their shared ideas of analysis and meaning. But before we turn to the closely related issue on the multiple senses of “duty,” we must ask: Is the open question argument a fair test of a good definition? It’s doubtful. Concepts are notoriously unclear—that, after all, is part of the rationale for doing analysis—and the question “But is an X really a Y?” may reasonably elicit conflicting answers from competent speakers. Philosophy

Bertrand Russell’s Moral Philosophy

337

seeks real definitions, not unlike science. Reportative dictionary definitions must reflect common usage, but it’s not clear that philosophical or scientific definitions must do so per the open question test. By that test, a definition of a cube as a solid figure with exactly 12 equal sides would fail. The concepts are logically equivalent, although apparently not identical since the concept-words are not synonymous (whatever that means), as Moore and Russell seemed to require. And worse, even the scientific definition of water as H2O would fail. A lesson here, as Kripke reminded us, is that property P can be identical with property Q although their ordinary designators are not synonymous.10 In ethics, as we’ll see, Russell’s notion of definition as analysis came to take on the same revisionist features that had characterized it in his early work in the philosophy of mathematics and physics, for example, numbers as classes of similar classes, and physical objects as classes of sensibilia. In short, the open question argument is of doubtful cogency; it doesn’t clearly prove the indefinability of good (as Moore later admitted).11 Still, of course, philosophy needs analysis and can reap its benefits even if that analysis doesn’t meet strict Principia Ethica standards.

1.2  Russell’s senses of moral obligation Russell saw other difficulties with the notion of duty that called for recognition of several distinct senses of “ought to do.” In Russell’s 1904 review he says that the Principia definition of “ought to do” (as that action productive of the most good) not only commits the naturalistic fallacy, but also contradicts Moore’s own advice for practical action in cases where actions’ consequences are unclear or unknowable: In such cases we ought to follow the social/moral rules known to be generally useful and generally obeyed (e.g., “Thou shalt not kill”). As Moore put it: “Though we may be sure that there are cases where the rule should be broken, we can never know which cases those are, and ought, therefore, never to break it” (Principia, 163–64). Suppose a case where the consequences of my killing X would be optimific and require that I ought to kill X. But because we can’t know the optimific consequences, Moore says, I ought not to kill X. Now, unless Moore is using “ought” in two different senses, he has a contradiction. Russell thinks this requires a different sense of “ought to do,” viz. those actions that will probably (based on available evidence) have the best consequences. It’s an insightful discovery and a valuable suggestion, but one Moore never quite adopted.12

338

The Bloomsbury Companion to Bertrand Russell

In “Elements” Russell distinguishes several senses of “right” under two divisions: objective and subjective.13 Under the first division, he has (a) “the most fortunate act” (essentially Moore’s “duty”). This sense of “right” obligates an agent to do X even when its optimal consequences are unknowable or the evidence favors a different act. An act “which, of all that are possible, will probably have the best consequences,” he calls (b) “the wisest act”—his second and preferred sense of “objectively right” (PE, 32). Russell also wants to distinguish acts that are merely mistaken from those that are immoral or blameworthy; “subjectively right” acts, what he calls “moral acts,” are those which the agent would judge to be right (including those judged by nonutilitarian theories) after considering the action candidly and with intent of discovering what was right (35). This allows that in a subjective sense of “right” there can be acts which ought to be done even though objectively wrong. For example in his 1904 review: It is certain that some people . . . might with advantage to the world have been strangled in infancy; but we cannot blame the good women who brought them up for having omitted this precaution. (Papers 4, 573)

The strangulation may well have been objectively right (qua the most fortunate act), but omitting it was surely not blameworthy (“subjectively wrong”); and doing it would hardly have been “moral” or objectively right (qua the wisest act).

2  Three decades with ethical subjectivism 1913–44 2.1  Early efforts Soon after his Santayana-inspired doubts about moral objectivism, Russell adopted what Charles Pigden has called a “proto-emotivism” (RoE, 106); but he notes that Russell did not yet clearly distinguish between simple subjectivism and genuine emotivism. (The former typically regards moral judgments, e.g., “X is good,” as statements about the speaker’s personal taste, e.g., “I like X,” which are only trivially true (if sincere) reports about the speaker;14 the latter, as expressions of feeling/desire, e.g., “Hooray for X!” or “Would that everybody liked X” [Russell’s later optative version] which are truth-valueless. Both are subjective theories that deny moral facts and moral knowledge. See II. D below.)

Bertrand Russell’s Moral Philosophy

339

The earliest published statements of his new moral thinking were during the years of the First World War.15 In these Russell expresses the view that ethical statements concern feelings/desires and can’t be known to be true or false; they are not amenable to proof by scientific reasoning or rational argument, at least not if they concern intrinsic value. As he says in a public 1916 letter, they are all part of “the subjectivity of ethics.”16 In that letter, responding to a critic’s complaint that he hadn’t proved his case against the war, Russell explains what he takes to be the rational limits of ethical argument: (1) showing that the effects of some kind of action are different from what the opponent supposes; (2) . . . altering the desires . . . of the opponent. . . . I cannot imagine any argument by which it could be shown that something is intrinsically good or intrinsically bad.

Ethical disputes—at least those over intrinsic value—can’t be settled by rational argument, but only by emotive persuasion (“altering the desires of the opponent”). That an action will have different effects than supposed, as mentioned in (1), can be shown probable by empirical evidence and argument. Let’s consider the sort of thing he seems to have in mind. If X and Y disagree about the death penalty—X favors it, Y doesn’t—they might be able to settle their dispute if it turns out that X’s support is based mainly on his belief that the death penalty will cost the taxpayer less than life in prison. But if there are other underlying differences involving basic values regarding (say) the purpose of punishment—for example, X holds that the execution of a murderer is always a good and just action, regardless of cost, pain, or deterrent efficacy, but Y does not—then it seems that, as Russell says, there’s not much to be done other than try to persuade X (an appeal to emotion?) to feel as she (Y) does, viz. disapproval. But I don’t believe basic ethical disagreements—those irresolvable by rational/scientific argument—are really all that common. Almost all ethical issues have a hidden consequential complexity which, when discovered, can alter moral judgment, even for those people who aren’t consequentialists. In the example above, X might change his mind if Y could be made aware (by empirical evidence) that the death penalty may actually increase homicides. Notice also, that this resolution of their disagreement needn’t affect the question of the intrinsic worth of executing a murderer: X might still hold that such executions are of intrinsic worth (or prima facie right) but might also believe the social effects include enough additional intrinsic disvalue to render capital punishment less desirable overall than life in prison.

340

The Bloomsbury Companion to Bertrand Russell

I’m suggesting two points: (1) that disagreements over intrinsic value needn’t mean that we can’t settle some or most of our moral disputes with the aid of cogent reasoning, deductive and inductive; and (2) that even if some disagreements aren’t settled by cogent argument, that doesn’t show that there’s no truth on one side or the other, nor that there is no justification for a belief that there is. These are important points against the subjectivity of value, not unlike some made by Russell himself.17 Part of the attractiveness of the subjectivity of ethics for Russell was in his conviction that belief in the objectivity of value was a cause of international conflict. Combined with nationalistic hubris, it affords an easy rationalization for “knowing” that one’s own nation is righteous and one’s enemy wicked. Like Santayana, he thought the prospects for avoiding war would be greatly improved if the ethics of moral objectivity could be supplanted by that of the subjectivity of values.18 And so it was for the next three decades—but with a brief interruption in 1922.

2.2  A brief affair with error theory Russell’s 1922 Apostles paper was astonishingly succinct and philosophically rich.19 In the course of 800 words he argues that Moore’s (and his own earlier) belief in an absolute good rested on a fallacy spawned by a naïve theory of meaning: “good” was meaningful because it was assumed to be a name standing for a unique property that was its meaning. But, he insists, “good” is not a name; it is a (disguised) description—a phrase of the form “the so-and-so.” Sentences containing “good” are meaningful, and do make objective statements (Moore was correct), but their meaningfulness doesn’t require the existence of such a property anymore than that of “The present king of France is bald” requires the existence of a current French king. He sketches the genesis of the error: We come to think we perceive goodness by mistaking the similarity of our feelings of approval toward various things A, B, C . . . for a special property (called “good”) which they all seem to possess in common. Since the things we call “good” are those we feel approval toward, we can fashion a definite description purporting to denote the supposed property. If we replace “good” by a description like “the so-and-so property common to A, B, C . . . ” in a judgment like “M is good,” we get (roughly): (1) The so-and-so property common to A, B, C . . . is possessed by M. By Russell’s theory of descriptions, this becomes:

Bertrand Russell’s Moral Philosophy

341

(1’) There is exactly one such property . . ., and it is possessed by M. If there is no such property, (1’) is false (not meaningless). Russell claimed there was no such property, not just because “good” is a description not a name (some descriptions do denote real things) but because he thought there was no need, semantic or otherwise, to assume such a thing as the Moorean good to account for the intelligibility of moral discourse. The practical usage of moral language is the same whether there is such a special property or not: We judge M good just in case we feel approval toward M. If we don’t need to assume an absolute property to account for either the objective meaning of our moral judgments or their practical use, we ought not to do it. The Moorean good was ripe for Occam’s Razor: Entities should not be multiplied beyond necessity.20 Pigden (RoE, 119–24) has leveled several objections against Russell’s 1922 theory to which I have responded.21 We don’t have space here to do justice to the objections. But most arise from failure to heed Russell’s instruction that the description be fashioned “as nearly as possible in accordance with the usage of absolutists” (RoE 122). For example, we could have: “The absolute nonnatural property called ‘good’ that we find as common to all and only our objects of approval.” If there is such a property, (some) moral judgments will be true; and, pace Pigden (RoE, 22), moral disagreements over M’s goodness will be genuinely contradictory; if there is not such property, all moral judgments (in which the description has “primary occurrence”) will be false. But why did Russell abandon his theory so soon? Pigden suggests that Russell likely saw the theory at odds with his principle of acquaintance which disallows, as unintelligible, words not definable in terms of acquaintance. Since the Moorean good was simple, it was not definable at all. But Russell wasn’t seeking a definition, but rather a “description.” Putatively simple objects/properties can easily be described by definite descriptions (e.g., “the color of ripe tomatoes”). Even Russell’s physical objects, unknowable by acquaintance,22 could be described by words so definable (e.g., “the unknowable causes of sense-data”). Pigden later (BRS Bulletin, Fall, 2014) gives a more plausible reason for the theory’s short life: Russell’s 1920 witness to the amoralistic leaders of the Russian Revolution and its brutal aftermath—an aftermath of the sort he may have thought error theory would promote. I think this a plausible factor. And he may have had a related concern, viz. it made our ordinary moral language untrue and most ordinary users unwitting purveyors of falsehoods.

342

The Bloomsbury Companion to Bertrand Russell

And there’s another possible worry. The analysis may have seemed a spurious use of Occam’s principle: It was a normative principle and may itself have been false, or, qua imperative, truth-valueless. Of course, it’s not a moral principle, and perhaps Russell would deny that it is amenable to his error theory; we don’t know. Also, in his previous uses of the principle (e.g., numbers and physical objects), he officially remained agnostic regarding the existence of the preanalytic “metaphysical monsters” supplanted by the analysis.23 Perhaps a better analysis would recognize the practical role of emotive feeling without claiming the existence (or nonexistence) of an absolute good, and thus avoid making moral discourse a factory of falsehood. Dropping the new theory and continuing work on a viable emotivism may have seemed the best choice.

2.3  A more mature emotivism (1925–34) In both “What I Believe” (WIB)24 and Outline of Philosophy (OOP) we find Russell retaining the main ingredients of his early proto-emotivism: values are a function of human emotion; and judgments of intrinsic worth are not amenable to proof or knowledge. What’s new is an innovative quasi objectivity based on his notion of the harmony of human desires which enhances their satisfiability due to their compatibility with our other desires and with those of other people (in the family, group . . . world). Roughly speaking, the greater the satisfiability of our desires, the better; and a necessary condition for that is their harmony—their “compossibility” as he later says (HSEP, 47). The good life in both works (WIB and OOP) is “. . . one inspired by love and guided by knowledge.” It is, ideally, a life which promotes, and depends on, a maximal satisfaction of desires—but not any desires, compossible desires, especially those inspired by love and guided by knowledge (WIB, 57–60; OOP, 234–35).25 Desires are central to Russell’s moral concepts; they are the primary motivators for all behavior and, hence, moral conduct. The primary importance of moral language lies in its power to influence conduct. But nobody does anything knowingly unless they desire to do it, an idea in keeping with Hume’s aphorism that “Reason is the slave of the passions.” (He still has, and will continue to have, his earlier utilitarian conviction that right conduct is that likely to maximize value; and, like Moore’s act consequentialism, is aided by socially useful rules.) Russell’s moral concepts are essentially social-political notions and his moral theory has important practical moral and political implications for human

Bertrand Russell’s Moral Philosophy

343

society, nationally and internationally. The function of morality is to enable people to live together in society in spite of the possibility of conflicting desires. One tool is the criminal law and the moral code, combined with social censure. And there’s another: to alter men’s characters and desires . . . as to minimize occasions of conflict. . . . If . . . the good life is inspired by love and guided by knowledge, . . . the moral code of any community is not ultimate .  .  . but must be examined .  .  . [to see  that] .  .  . it is such that wisdom and benevolence would have decreed. (WIB, 64)

Despite morality’s essential dependence on human desires, there are objective and impersonal features in place: the harmony of those desires and the wisdom and benevolence of the examiners who judge that the moral code needs modification.26 The harmony/disharmony of desires is a more or less objective, measurable thing. But wisdom and benevolence seem less clear. Russell is presumably thinking of the improvement of defective codes, which are discriminatory and increase discordant desires, for example, the disfranchisement of women. But might we not have a community with largely harmonious desires and only disharmony vis a vis a small subgroup subject to discrimination? If so, it looks like the harmony of desires is not, by itself, sufficient for the good life or a moral society. Of course desires over time can be altered by impartial laws and education inspired by love and guided by knowledge. And Russell leaves the door open for nonconformists who can challenge the status quo and move it in a more harmonious (a better harmony?) direction. Still, love and knowledge seem to get their value as a means to an end they enhance, viz. the harmony of desires, and ultimately to their satisfiability.27 (The satisfaction of any desire seems to be a kind of intrinsic good, although he doesn’t explicitly say that.) But “a single desire, in isolation, is no better and no worse, than any other” (WIB 85). (This needn’t be a problem for a utilitarian; the intrinsic value of the satisfaction of what we might call a “bad” desire, e.g., for a heroin fix, is likely to be outweighed by the intrinsic disvalue of its effects.) But groups of desires need not be of equal value. A group of harmonious desires is better (as a means) than one with “inconsistent” desires—“that is why love is better than hatred” (86). But does it make no difference what kind of desires harmonious groups have? Consider a society S1 of near maximal harmony populated mainly by superstitious jingoists united in their love of country and hatred of foreign

344

The Bloomsbury Companion to Bertrand Russell

groups. Now compare it with an equally harmonious society S2, one “inspired by love and guided by knowledge,” and with much less superstition and xenophobia.28 By the harmony criterion it seems that S2 is no better than S1. But only in “isolation,” not if we consider these societies in a larger context, for example, as nation-states, members of the international community. Despite their “moral equality,” a world containing S2 (without S1) would almost certainly be better (more harmonious) than one containing S1 (without S2). In the mid-1920s the international context was very much on Russell’s mind. In OOP (234) he says that “a harmonious world is a better one”; and, thus, we should embrace “the supreme moral rule: Act so as to produce harmonious rather than discordant desires.” Notice that the italicized sentence is not in the indicative mood and, so, is neither true nor false. Yet it does seem, in a sense, rationally supportable (as a means), provided humanity’s well-being is generally accepted as good. It’s at least a persuasive imperative, and it anticipates his later explicit adoption of the optative mood for analyzing moral judgments.29

2.4  Russell’s mature emotivism (1935–45) His mature theory is stated clearly in RS IX.30 It has two important points, especially the second: 1. There is no ethical knowledge; and 2. Ethical judgments are expressions of desire in the optative mood. (1) is so, not because we can’t settle some disputes over X’s goodness, but because of (2), that is, the sentence “X is good” is analyzed as an optative (“Would that everyone desired X”); it can’t have a truth-value, and so, can’t be known to be true. This is radically different from a simple subjectivism whereby moral judgments are indicative statements, about the speaker’s feelings toward X. Russell’s mature theory retains the “objective” character that we see in “What I Believe” and An Outline of Philosophy. The business of ethics is “to seem to give universal importance to our desires”; there are mainly two ways—those of the “legislator” and the “preacher.” The former establishes a moral/legal code aimed at social order. The latter typically uses appeal to emotions to cause similar desires in others. Moral progress is made through modification of the moral/legal code conducive to the universality of desires fashioned with the help

Bertrand Russell’s Moral Philosophy

345

of impartiality, intelligence, and love. He stresses the relative unimportance of theory in that progress: [It]  .  .  . is not by ethical theory, but by the cultivation of large and generous desires . . . that men can be brought to act . . . in a manner that is consistent with the general happiness of mankind.

And he reminds us (a la Hume) that: Whatever our definition of the “Good,” whether . . . subjective or objective, those who do not desire the happiness of mankind will not endeavour to further it. (RS, 242–43)

1.  Advantages of the Theory a. Both universal and impersonal desires are undoubtedly important moral features for any emotivism, and Russell’s inclusion does seem to make his version richer and more attractive than some later, and better known, emotive theories—for example, Ayer’s (1936) and Stevenson’s (1944)—wanting in both.31 Pigden (RoE, 131–33) observes that both emotivists’ analyses of “X is good” (partly in terms of personal approval) presuppose a feeling that X is good, and are, therefore, circular—a flaw Russell’s analysis in terms of desire avoids.32 Stevenson (1944) also fails to distinguish carefully between expressing an approval and stating that one has it.33 b. Russell’s version can also meet certain logical objections long leveled against emotivism: (1) emotivism makes it impossible for moral judgments (S and S’, e.g., “X is good” and “X is not good”) to contradict one another; and (2) emotivism can’t accommodate valid moral reasoning since the concept of validity is defined in terms of truth. Regarding (1), Russell’s idea of compossibility of desires allows us to define a sense of “contradiction” for moral disputes (e.g., A says “X is good,” B says “X is not good”) as follows: Sentences S and S’ are “contradictory” just in case the desire expressed by S is incompossible with that expressed by S’ (as with “Would that everyone desired X” and “Would that no one desired X”); both can’t be (simultaneously) satisfied. Objection (2) is a tougher nut to crack. It seems that simple categorical syllogisms with moral sentences (optatives) can be handled in terms of “contradiction” in the above sense; that is, such arguments are valid just in case the conjunction of their premises is contradictory (incompossible) with the

346

The Bloomsbury Companion to Bertrand Russell

denial of the conclusion. But Pigden (p. 134) raises another problem originally raised by P. T. Geach (via Frege), viz. the problem of unasserted contexts— notorious for emotivists’ arguments with conditionals that appear to generate equivocations. Consider the argument: P1 Happiness is good. (Would that all desired happiness) P2 If happiness is good, then war is bad. C War is bad. (Would that no one desired war) It appears to be a valid argument—an instance of modus ponens. But what does “happiness is good” mean in P2 qua its optative analysans? While we can consider the fulfillment of the desires of P1 and not-C, what counts as the fulfillment of the unexpressed desires in the conditional P2? We seem forced to take the desires of P2 as statements of desire, so that we have: If I desire that everyone desire happiness, then I desire that no one desire war. But apart from the fact that on the emotive theory the expressions of desires in moral judgments are not properly statemental, “Happiness is good” in P1 won’t mean the same as what it means in P2, and the argument’s validity is wrecked by equivocation. The problem is ongoing and may be insoluble.34 But the validity issue need not be a complete spoiler. The primary function of ethical discourse from Russell’s mature perspective is the practical one of persuasion, that is, getting people to cultivate and alter desires in ways that increase their harmony and, hence, the cooperative action that furthers the general well-being. And this can be done with the aid of both the emotive force of language and statements/arguments that support means to ends generally accepted as good, within the framework of virtually any theory. (See the applied nuclear deterrence case in III below.)35 2. Disadvantages of the Theory a. In RS we do find a dubious rationale for the adoption of the subjectivity of values, namely, “the complete impossibility of finding any arguments to prove that this or that has intrinsic value” (238). As noted earlier, in his Moorean period Russell answered this criticism arguing that there was actually little disagreement about intrinsic value—much less than about what is good as a means (n.17). There would seem to be relatively little disagreement that (say) the satisfaction of desire has intrinsic value, that is, is valued for its own sake, apart from its conduciveness to other things of value or disvalue. In fact, as we saw in part C, Russell appears to assume the satisfaction of desire as an intrinsic

Bertrand Russell’s Moral Philosophy

347

value in his view that the good life is good because such a life is conducive to the general harmony of desires (and, hence, to their satisfaction).36 Of course, as he often reminds us, when it comes to desired ends reasons give out, and desires are without any rational justification (RS, 242). Interestingly, right after his “complete impossibility” statement (238), he seems to leave room for such justification: “If we all agreed [my italics], we might hold that we know intrinsic values by intuition.” Those familiar with Problems (1912) may recall that in addition to a few logical principles, some ethical truths were also said to be “self-evident” and knowable a priori by intuition (76, 109, 112). And some intuitive knowledge was a posteriori, as with (some) sense perception. Could Russell’s RS statement bespeak a hope for some kind of “scientific” moral knowledge (a posteriori)? In fact we shall find something along those very lines in HSEP (ch. 9). Let’s note some other criticisms perhaps felt by Russell who confessed that he was not satisfied with his theory (though even less with those of others) (See Schilpp, 724). b. Is his consequentialism adequate? At times he himself seems to have had doubts. For reasons of space I’ll mention but one case, which arises in his famous 1948 BBC debate with Copleston (Papers 11, 521–41).37 Russell says he opposes the Nazis’ actions on consequentialist grounds. But, he’s asked, would he still oppose those actions if they led to a “general balance of happiness.” He says he “can’t imagine circumstances” in which they could have such effects. But if there were such, he “might be obliged” to acquiesce (Papers, 539–40).38 Does Russell’s consequentialism need something like a deontological principle of justice/human rights? Some have thought so.39 The adequacy of consequentialism is an ongoing debate which best not be pursued here. But at least in his reply to Copleston, Russell might have clarified his “might be obliged” response: that even if there were such effects, his consequentialism would not oblige him—unless such effects were, on evidence, virtually certain. Short of that bare possibility, the (utility-based) moral/legal code against such atrocity would over-ride any “obliged” acquiescence in Nazis’ actions.40 c. Some critics have insisted that Russell’s frequent resort to “vehement ethical judgments” was incompatible with his theory.41 Russell replied that moral matters are often concerned with emotion, and in those cases where one feels passionately, why not resort to “vehement” moral language to persuade others, especially those in positions to make socially desired changes? (As we’ll see in part III, Russell’s use of emotive language, especially concerning nuclear

348

The Bloomsbury Companion to Bertrand Russell

weapons, could be shockingly vehement. But that’s hardly incompatible with the theory.) d. The American philosopher Brand Blanshard made a related, but much more serious, objection: Russell’s emotivism was a threat to world peace. Its general acceptance would . . . be an international disaster. . . . It would be meaningless to call these nations [Germany and Japan] mistaken; if they felt approval for what they did, then it was right.42

But this misunderstands the theory. A right action would be (roughly) one conducive to more satisfiability of desires in harmony with “the interests of society . . . and with the general happiness of mankind” than that of any other possible action (RS, 241–43).43 The theory does allow reasons regarding means to agreed-upon ends, for example, means conducive to international harmony such as compliance with international laws regarding boundaries, restrictions on use of force and the treatment of civilians. Axis’ actions would be legally wrong and, on Russell’s theory, properly deemed “wrong” on utilitarian grounds. e. Finally, there are objections stemming from the theory’s radical departure from the common morality and its language.44 Consider: 1. Intense pain is intrinsically bad. 2. I believe that intense pain is intrinsically bad. 3. I know that intense pain is intrinsically bad. Sentences (1–3) would almost certainly be endorsed as true by those with normal pain perception and an understanding of the language.45 Yet on Russell’s theory, (1) is an optative and truth-valueless; (2) is either false (its optative clause cannot truly be believed) or truth-valueless (the clause renders the sentence grammatically senseless, as in “I believe that would that no one desired intense pain”); and (3) is also either false or truth-valueless for similar reasons. Such paradoxes may have been the main motivators for Russell’s naturalistic “improvements” in HSEP. 46

2.5  Human Society in Ethics and Politics: An “Objective” improvement? The book is as much about the prospects for international peace in the nuclear age as about moral theory. Russell says he originally wrote the ethics chapters of HSEP in 1945–46 intending to include them as part of Human Knowledge (1948),

Bertrand Russell’s Moral Philosophy

349

but unsure in what sense ethics was “knowledge,” he decided against it (HSEP, vii). A principal aim in the ethics part (esp. chs. 4 and 9) was to make moral judgments indicatives while still retaining his view of ethics as based on interpersonal emotions and feelings; and, of course, his utilitarian theory of right, along with his compossibility of desires as promoting their interpersonal satisfiability and the prospects for world peace (HSEP, 5–7). The innovative theoretical feature is the definability of “good” which, after several chapters of exploratory discussion, he thinks achievable in naturalistic terms—emotions and feelings (100). The satisfaction of desire (enjoyment) is  said to be an intrinsic good (103), that is, a thing “valued for its own sake” (39). Before he offers his naturalistic definitions and statements (105), he quickly sketches a plan to use the “emotion of approval” to help identify right acts and discover the quality common to those acts which elicit the emotion. With this sketch he proceeds to define “good,” and “right” in terms of it, and claims to render moral language empirically confirmable like that of science: 1. Surveying acts which arouse emotions of approval or disapproval, we find that, as a general rule, the acts approved of are those believed likely to have, on balance, effects of certain kinds, while opposite effects are expected from acts that are disapproved of.47 2. Effects that lead to approval [of acts] are defined as “good,” and to disapproval “bad.” 3. An act which, on the available evidence, is likely to have better effects than those of any other possible act is defined as “right”; any other act is “wrong.” We ought (by definition) to do the act which is right. 4. It is right to feel approval of a right act and disapproval of a wrong act. Russell thinks that (1) may be thought a problem because there have been acts at times widely approved but with bad effects, for example, the burning of witches. But approval was based on mistaken beliefs as to what the effects really were, viz. not ridding the world of a dangerous source of evil but causing the suffering and deaths of innocents. The beliefs may have been based on the evidence available, but the evidence was flawed by limited scientific knowledge and theological superstition. Notice in (1) that Russell says the approved acts “as a general rule” are those “believed likely” to have effects of certain kinds. This seems to allow Russell’s definition of “right act” in (3) to yield the result that an act of a certain sort (e.g., burning witches) can be right at some times and places and wrong at others, depending on the nature of the approvers,

350

The Bloomsbury Companion to Bertrand Russell

including their state of knowledge—a feature which threatens to make his “objective” ethics a form of moral relativism, not something he may have been wholly satisfied with. Pigden (RoE, 184) suggests that (2) should be modified to qualify the nature of the approvers, for example, as impartial and informed. I agree (and one in keeping with Russell’s earlier insistence on wisdom, benevolence, and impartiality in the making and scrutiny of the moral code).48 In fact he may have intended his definition to be taken as describing acts that are only prima facie right and, like the hypotheses of science, subject to modification in light of new evidence. This seems likely given his description of the discovery of “exceptional acts,” that is, those with effects once “believed likely” to be good, but discovered not to be of the kind that generally lead to an act’s approval. Such an act is not right (by 3) and its approval would also be wrong (by 4) since its effects would not have the quality possessed by those of most approved acts per the criterion of “right” acts (in 3). Russell says that once an exceptional act is generally recognized and no longer approved, we can speak “in a certain sense . . . of ethical error” (107). That the theory allows for a kind of ethical error is an important feature in making possible the idea of moral progress, and it militates against relativism. Pigden thinks (3), as a naturalistic definition of “right,” is false because it fails the open question test (RoE, 101). But that test is hardly conclusive (see I. A above), a point Pigden himself seems to acknowledge (RoE, 11–12). And the “definition”49 is almost certainly offered as a proposal—a nonarbitrary stipulation (but partly reportative), and to be judged by its “usefulness”: its conformity with the emotion/feeling basis of moral language, its empirical confirmability and its avoidance of a priori intuitions of nonnatural properties. If one seeks an ethical theory grounded on emotion but with statements capable of empirical truth yet without resort to mysterious intuitions, it’s not a bad effort, although some have not thought of its descriptive innovations as “improvements.”50 Undoubtedly there are those (my former self among them) who would argue that Russell’s “improvements” might have been even better had he opted for a nondefinist nonnatural (normative) notion of intrinsic goodness. That is, not the quality of being “valued (desired) for its own sake,” but of being worthy of being so valued (desirable). But any advantages of that move would surely have been seen by Russell in the 1940s as a slide back to the metaphysical monsters of his Moorean past and a betrayal of the “supreme maxim” of the new scientific/ analytic philosophy—Occam’s Razor (n. 23).51

Bertrand Russell’s Moral Philosophy

351

Russell wasn’t satisfied with his HSEP naturalistic definist theory, nor with  emotivism to which he soon returned.52 His ongoing discomfort was theoretical,53 viz. his lack of rational grounds for desires concerning ends. In  his Auto (3, 29–30) he says that although his doctrine of compossibility was  “approximately rational .  .  ., it assumes that happiness is better than unhappiness . . . [a] principle incapable of proof.”54 Interestingly, a half-century earlier he also cites the very same principle (POP, 76) as having no proof—but because it is known (a priori) without proof, that is, as a “self-evident truth” (109). But even if it’s not self-evident (and we know that the “mature” Russell had long since repudiated knowledge by moral intuition), the principle doesn’t need proof; it has general agreement among rational observers. And the same holds for one of its necessary conditions— the continued existence of our species, that is, the nonoccurrence of nuclear annihilation.55 So even if Russell has, in a sense, failed the theoretical objective (wholly grounding morality in rationality), the practical one of changing behavior through rational argument and moral persuasion was very much alive, as we shall see.56

3  Russell’s ethics in practice: The nuclear peril 3.1  Some background After the onset of the Great War, Russell saw international harmony as a necessary condition for human happiness.57 And at the end of the Second World War, the advent of the Bomb made it clear to a few (notably Russell and Einstein) that the chances of human well-being—indeed, the continued existence of the species—depended on “new thinking” regarding the means to avoid nuclear holocaust and achieve world peace. As Russell had emphasized, ethical theory was not really important for the practical arguments to promote the general good,58 and even less to persuade world leaders and fellow citizens that a nuclear war would likely destroy hundreds of millions, if not humankind itself—a catastrophe growing more probable as the Cold War intensified. Russell’s antinuclear activism from 1955 through the mid-1960s59 was of vital importance in fostering a US-Soviet awareness of their common danger and the need for the cooperation that eventually ended the Cold War a quarter century later. Several items should be kept in mind:60

352

The Bloomsbury Companion to Bertrand Russell

* The 1955 Russell-Einstein Manifesto brought to world attention the dangers of nuclear war, the need to abolish nuclear weapons and eventually war itself. It called on international scientists to help educate their citizens and national leaders on the need for East-West cooperation and arms control. * The Pugwash Conferences on Science and World Affairs emerged with Russell’s guidance (1957; Russell as first president), as did the Campaign for Nuclear Disarmament (1958; Russell as first president) which mobilized millions in the UK and around the world. Pugwash was instrumental in making possible the first agreements on nuclear arms control, including the Partial Test Ban Treaty (1963) and the Nuclear NonProliferation Treaty (1968); and the later START treaties which have eliminated 75 percent of the superpowers’ nuclear stockpiles. Pugwash and its cofounder (J. Rotblat) received the Nobel Peace Prize in 1995. Russell undertook to inform the general citizenry regarding the nuclear peril and persuade them (and their leaders) to action in the interests of humanity. The example below concerns his controversial moral condemnation of the leaders of East and West regarding nuclear weapons policy. (Note what seem to be both utilitarian and deontic reasons underlying his condemnation.)

3.2  The wickedest people who ever lived On April 15, 1961, at the age of 89, Bertrand Russell gave a speech calling for nonviolent civil disobedience in his campaign for British unilateralism.61 At the end he added the following extemporaneous remarks: We used to think that Hitler was wicked when he wanted to kill all the Jews, but Kennedy and Macmillan and others both in the East and in the West pursue policies which will probably lead to killing not only all the Jews but all the rest of us too. . . . And I will not pretend to obey a government which is organizing the massacre of the whole of mankind. . . . They are the wickedest people that ever lived. (Auto. 3, 203–04)

Russell’s angry words were picked up by the press and widely dismissed as irresponsible nonsense. But I think, given the magnitude of the risk and nature of the threat posed by nuclear weapons at the time (about which Russell was well informed), his “nonsense” can be defended. A close look at his remarks suggests a sense of “wicked” with a clear consequentialist import. Hitler killed millions, perhaps tens of millions, of innocent people, including six million Jews. But a thermonuclear war between

Bertrand Russell’s Moral Philosophy

353

the nuclear superpowers would easily kill hundreds of millions, if not humankind itself. And, short of some change of attitudes and beliefs regarding our common peril, chances of catastrophe approach near certainty.62 By this “harm likely to be done” criterion, we have a sense of “wicked” (call it wicked1) by which nuclear policy and its national architects (Kennedy, et al.) were several times more wicked than Hitler. Yet Kennedy et al. never actually killed any innocents (unless you count Hiroshima or the several hundred thousand fatal cancers caused by eighteen years of atmospheric atomic testing).63 But Hitler did, and he intended to do it— something very different from Kennedy et al. This intention to harm innocents is another sense of “wicked” (wicked2), and one which applies even if the agent does no actual harm. Suppose I intend to kill my neighbor’s little girl; I buy a gun, take aim at the child, but the gun fails to fire and the child is unharmed. Am I not just as wicked2 as I would be if I had blown her head off? There are hints of this second sense as well in Russell’s claim that the EastWest leaders are “organizing the massacre of the whole of mankind.” Were Kennedy et al. wicked in this sense? It’s arguable. Russell’s moral insight here is his understanding that nuclear deterrence accepts—even requires—the preparation for, and the willingness to commit, the killing of very large numbers of innocent human beings.64 And this preparation and willingness to murder—which Russell described as “a willingness to commit genocide”65—is the profoundly immoral core of nuclear deterrence and its policy makers. During most of the Cold War, official nuclear war plans, on both sides, followed the formula of the early US doctrine of “massive retaliation”: thousands of nukes would be used against cities in response to attack; and, for the United States, even to a nonnuclear attack. (The United States, citing conventional disadvantage, claimed— and still does—the “right” to nuclear first use.) Mutual fears of a decapitating firststrike made “launch-on-warning” the only “rational” choice, and with it came the danger of accidental nuclear war. Russell was keenly aware of this, and of the scores of accidents that had already occurred, and said so publicly.66

4  Final thought: A call to duty In a little-known essay (1964),67 Russell appeals to philosophers—owing to their skills of persuasion, skeptical inquiry, and impersonal outlook—to rise to the call of duty in the nuclear age to “persuade mankind that human life is worth

354

The Bloomsbury Companion to Bertrand Russell

preserving and that an opposite view is only open to fanatics.” Russell would surely have been immensely gratified by the Cold War’s demise, but no less saddened by its threatened revival today—yet ever hopeful of a world without war, “a hope as ardent as the Christian hope of heaven” (ibid.). Russell asked then, as he would again, whether there will be sufficient time for that hope to be realized.

Notes 1 See Pigden, ed., Russell on Ethics: Selections from the writings of Bertrand Russell [RoE] (London: Routledge, 1999). Also see “Bertrand Russell: Moral Philosopher or Unphilosophical Moralist?” in N. Griffin, ed., The Cambridge Companion to Bertrand Russell (Cambridge: Cambridge University Press, 2003), pp. 475–506; and “Russell’s Moral Philosophy” (2007, revised 2014), in Stanford Encyclopedia of Philosophy, http://plato.stanford.edu/entries/russell-moral/, sec. 10; and ‘Emotivism, Error and the Metaphysics of Bolshevism’, Bertrand Russell Society Bulletin (Fall 2014), pp. 23–24. 2 Principia Ethica (Cambridge: Cambridge University Press, 1903).   Pigden (RoE, Part I) includes, with critical commentary, Russell’s pre-Principia ethical writings (five for Sidgwick’s course, and five for the secret Cambridge Apostles society). See also Papers 1. 3 G. Santayana, “The Philosophy of Bertrand Russell,” in Winds of Doctrine (London: J. M. Dent and Sons, 1913). See Russell’s brief account (OOP, 230). 4 Russell: the Journal of Bertrand Russell Studies 6 (1987), pp. 144–9; Papers 9: 345–46. 5 Russell, Philosophical Essays [PE] (New York: Simon and Schuster, 1968; first pub. 1910), pp. 13–59. 6 E.g. D. Parfit, J. J. C. Smart and J. O. Urmson. See W. Shaw, Moore on Right and Wrong (London: Kluwer, 1995), pp. 119–20. 7 Moore, Principia, pp. 10–12. R. P. Sylvester criticizes scholars for inattention to Principia’s ontology. He follows H. Hochberg in holding that Moore’s simple natural properties are simple particulars. See Sylvester, The Moral Philosophy of G.E. Moore, eds. R. Perkins and R. Sleeper, Forward by Tom Regan (Philadelphia: Temple University Press, 1990), pp. 137–38. 8 Pigden (RoE, 10–13) has an informative account of Moore’s argument and of an earlier, weaker (pre-Principia) variation; Moore apparently fashioned his stronger Principia version (§13) to refute a “plausible” definition (“what we desire to desire”) proposed by Russell in an 1897 Apostles society paper. 9 Russell’s 1904 review, in RoE, 101; Papers 4:573.

Bertrand Russell’s Moral Philosophy

355

10 See S. Kripke, Naming and Necessity (Cambridge, MA: Harvard University, 1972). 11 See Moore, “Is Goodness A Quality?” (1932) Philosophical Papers, 98. Cf. Pigden (RoE, 10–11). 12 But many others did. See note 6 above. In his Ethics (London: Oxford University Press, 1912), pp. 114–16, Moore does allow that a wrong action done from good motives can also be praiseworthy. 13 In a 1905 letter to Moore concerning a draft of “Elements,” Russell says “as for right . . . I must think it out afresh with more kinds of right.” Papers 6: 215. 14 In “Elements” (PE, 20–21), Russell had rejected this view because it rendered genuine moral disagreement impossible (“no subject of debate”). 15 See e.g., “Scientific Method in Philosophy” (1914) in Papers 8: 62–64; “The Ethics of War” (1915) in Papers 13: 63; and his response to R. B. Perry, 13: 186–87; RoE, 105–18, includes well-chosen extracts from these early emotivist writings. 16 “‘North Staffs’ Praise of War,” Cambridge Magazine (March 11, 1916), 386; in Yours Faithfully, Bertrand Russell, R. Perkins, ed. (Chicago: Open Court, 2002), 59; also RoE, 116–18. 17 See “Elements” (PE, 20–21, 54–55). 18 Russell’s intimate relationship with Ottoline Morrell was also undeniably important in his retreat from the Moorean good. See M. Potter, Bertrand Russell’s Ethics (London: Continuum, 2006), pp. 6–12. 19 Interestingly, several philosophical fellow Apostles were in attendance, including R. B. Braithwaite, F. Ramsey, and G. E. Moore. There’s no known record of Moore’s reaction. See Perkins, “Was Russell’s Error Theory A Mistake?,” Russell [journal] 32 (Summer 2012), p. 32, note 3. 20 He lists the “Razor” as his fifth reason for accepting his analysis (RoE, 124). Cf. his 1916 reply to Hulme: “Occam’s Razor . . . leads me to discard the notion of absolute good if ethics can be accounted for without it,” in Yours Faithfully, op. cit., 59). 21 Perkins (2012), op. cit., 35–40. 22 POP, ch. 3; cf. HK, 229–31. 23 See RSDP (1914) Papers 8: 11 24 “What I Believe” (1925), reprinted in Why I Am Not A Christian and Other Essays, ed. Paul Edwards (New York: Simon and Schuster, 1957). 25 Potter (24), introduces a phrase with Kantian import—the “universalizability of desire.” Desires are universalizable when in harmony with the general desires of most human beings. Cf. n. 29 below. 26 See L. Aiken (1963), chs. 8–12 for a critical but sympathetic account of Russell’s emotive theory. She emphasizes the theory’s unique impersonal and interpersonal “objective” features. See note 31 below. 27 See Perkins, “Russell’s Metaethics” [review of Potter, op. cit.] in Russell [Journal], 181f.

356

The Bloomsbury Companion to Bertrand Russell

28 In WIB Russell gives many examples of moral codes based on superstition, including some connected with the dogma of the Catholic Church. 29 Later, explaining his mature theory, he says: “I do not think that an ethical judgment merely expresses a desire; I agree with Kant that it must have an element of universality. I should interpret ‘A is good’ as ‘Would that all men desired A’.” Schilpp, RTC (1944), 722. 30 Cf. his lucid explication to critics, RTC (1944), 719–25. 31 L. Aiken notes that Russell’s version included “objective” features “entirely overlooked by the ordinary emotive theories” (op. cit. p. x). Russell gave her book high praise, her criticisms notwithstanding. See Dear Bertrand Russell . . ., eds. B. Feinberg and R. Kasrils (Boston: Houghton Mifflin, 1969), pp. 97–98. 32 A similar circularity appears in W. H. F. Barnes, “A Suggestion about Value,” Analysis I (1933), p. 45. 33 Potter (43–44) astutely notes that Ayer’s emotivism (which renders moral judgments “meaningless”) is based on his verification principle—which, by its own criterion of meaning, is meaningless. 34 See Pigden’s short but helpful exposition (RoE, 133–34). Potter (46–47) suggests that a purely syntactic (formal) notion of validity could solve the problem. Yes, but apparently only by “extracting” a valid argument form from an argument ‘instance” which commits the informal fallacy of equivocation. 35 The validity problem was not to my knowledge discussed by Russell in his emotivist phase. But it may well have been a factor in his HSEP shift (1945–54) toward naturalism. 36 See WIB (PE, 85–86). 37 At this time Russell had already written his naturalistic ethics chapters later to appear in HSEP. But his theory of “right” remained utilitarian. 38 Russell’s uneasiness doesn’t show in the published version, but Copleston (1993, p. 137), later recalls that Russell admitted “off the record” that his theory put him in a discomforting “dilemma.” 39 See Ayer Russell (London: Fontana/Collins, 1972), pp. 126–27. 40 Cf. Power (1938), 166–67, where Russell falls back on such codes in a similar example; also “Philosophy and Politics” (1947) in UE (1950), p. 18, where he appeals to a (utility-based) “empirical principle”: “[It’s] not worth while to inflict a comparatively certain present evil for a comparatively doubtful future good.” 41 See J. Buchler on Russell’s ethics in Schilpp (1944) and Russell’s reply, pp. 720–24. 42 Blanshard (1949), reprinted in A. Minton, ed. Philosophy: Paradox and Discovery (New York: McGraw-Hill, 1976), pp. 237–38. Cf. remarks in M. Ruse’s apparent misunderstanding of Russell’s emotive theory in his otherwise excellent Introduction to the 1997 edition of RS, p. xix. 43 Implied; but not explicitly stated.

Bertrand Russell’s Moral Philosophy

357

44 See Perkins, “Russell’s Metaethics” (2007), op. cit., p. 183; cf. Pigden (RoE, 135). 45 I.e. “intrinsically bad” should be understood as roughly: “Apart from consequences, its nonexistence is preferable to its existence.” 46 In his 1922 “error theory,” he says he has “no doubt that our moral judgments claim objectivity” (RoE, 123). It’s the absence of such a claim in moral judgments qua Russell’s emotivist analysis that gives rise to these paradoxes. 47 These unspecified “effects” are apparently enjoyment or satisfaction. See two pages later (107) where he says the definitions are based on “the emotion of approval and the feeling of enjoyment or satisfaction, the former in the definition of ‘right’ . . ., the latter in that of ‘intrinsic value’”; and he adds, “aesthetic sensibility” and “intelligence” (106). 48 See WIB, 64 and RS, 232, 242. It also gets closer to Hume’s theory, apparently favored by Russell at the time, in which the approver is an ideal (enlightened, impartial) observer. See Hume’s Enquiry Concerning the Principles of Morals, op. cit., p. 289; cf. Pigden (RoE, 151–52, 166). 49 He speaks of the effects which “mark” most approved acts as “the ‘criterion’ of what is ‘right’” (HSEP, 107), not as the “definition.” 50 Ayer (in Russell, op. cit., pp. 126–27) says the theory “falls short as a descriptive theory”; it needs an emotive definition of “good” and, to improve the consequential theory of obligation, “a general rule that one should aim at maximizing satisfaction,” but allowing other intrinsic values like “justice and kindness . . . over-riding value.” Cf. L. Aiken (op. cit., Part III) who thinks the earlier theory already sufficiently “objective” and “closer to the truth of moral discourse” (162–63). 51 Interestingly, Russell’s moral naturalism in HSEP shares some features with R. Brandt’s later ideas on rationality in ethics, especially in his later A Theory of the Good and the Right (Prometheus Books, 1998; Forward by P. Singer). Brandt, like Russell, eschews moral intuition, the limits of ordinary usage and, like Russell, broadens Hume’s narrow limits on ways that reason can curb “irrational” desires/ impulses with the aid of foresight and “cognitive psychotherapy” (Brandt’s term).   To my knowledge Brandt says nothing about HSEP but was aware of Russell’s work in ethics, and that he had, over his lifetime, held and defended several very different moral theories—“always sanely and skillfully, although sometimes not in great detail.” See Value and Obligation, ed. Brandt (Harcourt, Brace & World, Inc., 1961), p. 376. 52 In a 1954 review of Ayer’s Philosophical Essays, Russell says he’s “in pretty complete agreement” with Ayer’s emotivism. See RoE, 164–65. 53 He says he has no problem with practical moral judgments, “which . . . I make on a roughly hedonistic basis” (Philosophy, 1960, 35, 146; Papers 11, 311). 54 Volume 3 of the Auto. was published in 1969, but these passages were apparently written in the early 1960s.

358

The Bloomsbury Companion to Bertrand Russell

55 Of course, not complete agreement. At the height of the nuclear peril some (fanatics?) insisted “Better dead than red”—and, no doubt, its equivalent on the other side: “Better corpses than capitalists.” See Russell (1961), p. 89. 56 Without pressing the analogy with Russell’s treatment of the justification of science too hard, it’s worth recalling that in his HK, 503–07) he candidly admits that knowledge of causal laws depends on five postulates that are synthetic a priori, while necessary for the justification of causal laws, can’t themselves be empirically justified. 57 See PSR (1916), Chap VIII. 58 See RS, 242–43. 59 As early as 1963 Russell was becoming concerned with the US “war of annihilation” in Vietnam. See his letter to NYT (8 April) in Yours Faithfully (2002), 360ff. 60 See Perkins, “Russell’s Legacy,” BRS Bulletin (Fall, 2013), pp. 6–7. 61 Also called “British neutralism,” by which the UK would unilaterally give up its nukes and NATO membership (but remain under the US nuclear umbrella) as a bona fide action to jump-start US-USSR disarmament negotiations. See Russell’s letters, Yours Faithfully, 221ff. 62 This was eighteen months before the Cuban Missile Crisis. Kennedy (who knew nothing of the 100 tactical nukes already deployed in Cuba with pre-delegated authority to repel a US invasion) put the chances of nuclear war between 1/3 and 1/2. See J. Trumpbour, ed., How Harvard Rules (Boston: South End Press, 1989), p. 77; Perkins, “Accidental Nuclear War” (2014), p. 87. 63 See J. Rotblat and R. Hinde, War No More (2003), p. 15. 64 Even a nuclear exchange of 100 Hiroshima-size nukes between India and Pakistan would likely bring on a nuclear winter and world famine—with two billion deaths. See I. Helfand, Nuclear Famine (Boston: Physicians for Social Responsibility, 2014). 65 Letter to the editor of Maariv, January 26, 1963, first publ, Bertrand Russell Society Quarterly, 2005. 66 See Perkins (2014), p. 88. 67 “The Duty of A Philosopher in This Time,” in Essays in Honor of Paul Arthur Schilpp, The Abdication of Philosophy: Philosophy and the Public Good, ed. Freeman (Open Court, 1976); reprinted in Papers 11, 457–63.

References Aiken, L. 1963. Bertrand Russell’s Philosophy of Morals. New York: Humanities Press. Ayer, A. J. 1936. Language Truth and Logic. New York: Dover Publications, 1946; first pub., 1936. Ayer, A. J. 1972. Russell. London: Fontana/Collins.

Bertrand Russell’s Moral Philosophy

359

Barnes, W. H. F. 1933. “A Suggestion about Value.” Analysis 1: 45. Baylis, C. 1958. Ethics: The Principles of Wise Choice. New York: Henry Holt. Blackwell, K. 1985. The Spinozistic Ethics of Bertrand Russell. London: Allen and Unwin. Blanshard, B. 1949. “The New Subjectivism in Ethics.” Philosophy and Phenomenological Research 9, p. 510, reprinted in A. Minton (ed.), Philosophy: Paradox and Discovery, 237–38. New York: McGraw-Hill, 1976. Brandt, R. 1959. Ethical Theory. Englewood Cliffs: Prentice-Hall, Chapter 9. Brandt, R. (ed.) 1961. Value and Obligation. Harcourt: Brace & World, Inc. Brandt, R. 1979. A Theory of the Good and the Right; Forward by Peter Singer. Prometheus Books, 1998; first publ. 1979. Copleston, F. 1993. Memoirs of A Philosopher. Kansas City: Sheed &Ward. Garner, T., and B. Rosen (eds.) 1967. Moral Philosophy. New York: Macmillan. Griffin, N. (ed.) 2003. The Cambridge Companion to Bertrand Russell. Cambridge: Cambridge University Press. Hare, R. M. 1952. The Language of Morals. Oxford: Oxford University Press. Helfand, I. 2014. Nuclear Famine: Two Billion People at Risk? 2nd ed. Boston: Physicians for Social Responsibility. Hochberg, H. 1969. “Moore’s Ontology and Nonnatural Properties.” In Studies in the Philosophy of G.E. Moore, edited by E. D. Klemke, 95–127. Chicago: Quadrangle Books. Hume, D. 1975. Enquiries Concerning Human Understanding and Concerning the Principles of Morals. Edited by P. H. Nidditch and L. A. Selby-Bigge, 3rd ed. Oxford: Clarendon Press. Mackie, J. L. 1946. “The Refutation of Morals.” Australasian Journal of Psychology and Philosophy 24: 77–90. Moore, G. E. 1903. Principia Ethica. Cambridge: Cambridge University Press. Moore, G. E. 1912. Ethics. London: Oxford University Press. Moore, G. E. 1932. “Is Goodness A Quality?” In Philosophical Papers. New York: Collier, 1962. Parfit, D. 1962. Reasons and Persons. Oxford: Oxford University Press. Perkins, R. (ed.) 2002. Yours Faithfully, Bertrand Russell: A Lifelong Fight for Peace, Justice and Truth in Letters to the Editor. Chicago: Open Court. Perkins, R. 2007. “Russell’s Metaethics.” Russell: The Journal of Bertrand Russell Studies 26: 179–84. Perkins, R. 2012. “Was Russell’s Error Theory A Mistake?” Russell: The Journal of Bertrand Russell Studies 32: 30–41. Perkins, R. 2014. “Accidental Nuclear War and Russell’s ‘Early Warning’.” Russell: The Journal of Bertrand Russell Studies 34: 84–91. Perkins, R., and R. Sleeper (eds.) 1990. R.P. Sylvester, The Moral Philosophy of G.E. Moore. Philadelphia: Temple University Press. Pigden, C. (ed.) 1999. Russell on Ethics. London: Routledge.

360

The Bloomsbury Companion to Bertrand Russell

Pigden, C. 2003. “Bertrand Russell: Moral Philosopher or Unphilosophical Moralist?” In The Cambridge Companion to Bertrand Russell, edited by N. Griffin, 475–506. Cambridge: Cambridge University Press. Pigden, C. 2014a. “Russell’s Moral Philosophy.” http://plato.stanford.edu/entries/russellmoral/, sec.10. Pigden, C. 2014b. “Emotivism, Error, and the Metaethics of Bolshevism.” The Bertrand Russell Society. Bulletin 150: 19–24. Potter, M. 2006. Bertrand Russell’s Ethics. London: Continuum. Rotblat, J., and R. Hinde. 2003. War No More. London: Pluto Press. Russell, B. 1910. “The Elements of Ethics,” in PE. Russell, B. 1925. “What I Believe.” In Why I Am Not A Christian and Other Essays, edited by P. Edwards. New York: Simon and Schuster, 1957. Russell, B. 1935. Religion and Science. Oxford: Oxford University Press., 1997; first pub. 1935. Russell, B. 1954. Human Society in Ethics and Politics. London: Routledge, 2010. Russell, B. 1959. Common Sense and Nuclear Warfare. New York: Simon and Schuster. Russell, B. 1961. Has Man A Future? New York: Simon and Schuster. Russell, B. 1969. Dear Bertrand Russell . . ., Edited by B. Feinberg and R. Kasrils. Boston: Houghton Mifflin. Santayana, G. 1913. Winds of Doctrine. London: Dent and Sons. Schilpp, A. (ed.) 1942. The Philosophy of G.E. Moore. Evanston, IL: Northwestern University Press. Schilpp, A. (ed.) 1944. The Philosophy of Bertrand Russell. Evanston, IL: Northwestern University Press. Shaw, W. 1995. Moore on Right and Wrong. Dordrecht, The Netherlands: Kluwer Academic Publishers. Smart, J. J. C., and B. Williams. 1973. Utilitarianism: For and Against. Cambridge: Cambridge University Press. Stevenson, C. L. 1944. Ethics and Language. New Haven: Yale University Press. Sylvester, R. P. 1990. The Moral Philosophy of G.E. Moore. Edited by R. Perkins, Jr. and R. Sleeper. Philadelphia: Temple University Press. Urmson, J. O. 1970. “Moore’s Utilitarianism.” In G.E. Moore: Essays in Retrospect, edited by A. Ambrose and M. Lazerowitz. London: Allen and Unwin.

14

Russell’s Literary Approach to History1 Peter Stone

But speaking as one of the general public, I cannot forbear from saying what very great pleasure and interest I have derived from his books, and how profoundly I agree with the point of view expressed in his essay “Clio, a Muse.” It seems to me that the reading of history is an essential element in the formation of a wide outlook and that this demands the writing of history for nonhistorians. History is not, like mathematics, a subject only for specialists, and those who, like George Trevelyan, enable the general public to enjoy wide surveys are performing a function of the utmost importance, especially in this age when rapid progress is producing a widespread ignorance of the past —Bertrand Russell, “My Recollections of George Trevelyan” (Papers 29, 54)

1 Introduction “History,” Bertrand Russell wrote late in life, “has always interested me more than anything else except philosophy and mathematics” (BW, 21). This intense interest in history led Russell to write many historically oriented works, most notably his monumental A History of Western Philosophy (1945). And yet none of Russell’s works on history have left the sort of mark that Russell’s purely philosophical works have left. Even A History of Western Philosophy, for example, is widely regarded by philosophers as an uneven and idiosyncratic work.2 Russell acknowledged as much. Even late in life, after writing A History of Western Philosophy and numerous other historically oriented works, Russell still described himself as primarily a consumer of historical research and not a producer (BW, 511). It is therefore worth asking what, if any, value remains in Russell’s approach to history.

362

The Bloomsbury Companion to Bertrand Russell

This chapter examines, not Russell’s historical writings per se, but his approach to the study of history. It argues that Russell identified with the so-called literary, as opposed to the scientific, approach to history. This approach was famously advocated by Russell’s longtime friend George Trevelyan in his classic essay “Clio, A Muse” (1913).3 Russell, following Trevelyan, opposed the effort to turn history into a science—an effort he associated with philosophers of history, such as Hegel and Marx. In defending the literary approach, Russell ably catalogued the failure of the scientific approach to produce the grand laws of history it promised. Russell thus rejected the attempt to identify such laws, although he did endorse the pursuit of practical maxims based upon history. (In doing so, he went beyond the limitations of the literary approach.) But the literary approach has problems of its own. In particular, it is difficult to see why it should matter whether history produced according to this approach is true or not. Both Russell and Trevelyan struggled with this problem in articulating and defending the literary approach, and while neither was able to mount a perfect defense, Russell, I shall argue, succeeded better than Trevelyan at this task. The reason for this is that Russell was able to go beyond the literary approach in several significant respects, even while remaining largely within the confines of that approach.

2  The scientific and the literary approaches to history Russell commented on the study of history numerous times over the course of his long life. But while these writings span several decades, Russell’s position did not shift much from beginning to end. There were differences in emphasis, but not in the general conclusions reached. This fact is worth noting for its own sake. Many of Russell’s views changed dramatically over the course of his life; as a mature adult, for example, he vocally criticized Hegel in philosophy and imperialism in politics, although as a young man he was a defender of both. These dramatic changes in Russell’s perspective considerably complicate the study of his ideas. But Russell’s approach to the study of history does not suffer from this complication; the approach he embraced at the start of his life (or at least the first time he articulated any approach at all) was the same as the approach he embraced at the end. This makes it meaningful to speak of Russell having a single coherent approach; it remains, then, to articulate how Russell understood this approach.

Russell’s Literary Approach to History

363

Russell’s perspective on the study of history was strongly influenced by methodological debates among English historians that took place at the dawn of the twentieth century. These debates centered upon the question of whether history should be considered a science. In 1903, J. B. Bury gave his inaugural lecture as Regius Professor of Modern History at Cambridge University, in which he argued in the affirmative—“that history is and should be a science, neither more nor less” (Stunkel 2001–02, 145). Bury’s argument provoked strong opposition from, among others, George Trevelyan, a historian who had been good friends with Russell since their time together as students at Cambridge (Moorehead 1993, 36). Trevelyan responded to Bury in the first issue of the Independent Review, a liberal magazine that Trevelyan helped to launch (Goodlad 2015, 215). His response was entitled “The Latest View of History,” and it appeared in 1903. Trevelyan subsequently made extensive revisions to the essay before republishing it as “Clio, A Muse,” the title essay in a collection of essays that appeared in 1913. It is as “Clio” that Trevelyan’s critique of the scientific approach to history is best known. In “Clio,” Trevelyan contrasts the “scientific” and the “literary” approaches to history (Trevelyan 1913, 3). Trevelyan never lays out the two distinct approaches under these names, but the contrast he seeks to draw between the approaches is clear from the text. “The root questions,” he writes, can be put in these terms:—“Ought history to be merely the Accumulation of facts about the past? Or ought it also to be the Interpretation of facts about the past? Or, one step further, ought it to be not merely the Accumulation and Interpretation of fact, but also the Exposition of these facts and opinions in their full emotional and intellectual value to a wide public by the difficult art of literature?” (emphasis in original; ibid., 5)

Here Trevelyan identifies the three essential features of his approach to history— the identification of facts, the interpretation of those facts, and the use of those facts to tell a story. Trevelyan later elaborates: To my mind, there are three distinct functions of history, that we may call the scientific, the imaginative or speculative, and the literary. First comes what we may call the scientific, if we confine the word to this narrow but vital function, the day-labour that every historian must well and truly perform if he is to be a serious member of his profession—the accumulation of facts and the sifting of evidence. . . . Then comes the imaginative or speculative, when he plays with the facts he has gathered, selects and classifies them, and makes his guesses and generalisations. And last but not least comes the literary

364

The Bloomsbury Companion to Bertrand Russell

function, the exposition of the results of science and imagination in a form that will attract and educate our fellow-countrymen. (emphasis in original; ibid., 30–31)

It is only with the exposition of historical research—with the telling of stories— that history really becomes complete. For “history is not merely the accumulation and interpretation of facts .  .  . but involves besides the whole art of book composition and prose style. Life is short, art is long, but history is longest, for it is art added to scholarship” (ibid., 34). “The art of history,” Trevelyan concludes, “remains always the art of narrative. That is the bed rock” (ibid., 14). For this reason, it is perfectly fair to call Trevelyan’s preferred approach the literary approach to history, even if Trevelyan himself did not quite put it like that. For similar reasons, it makes sense to refer to the position Trevelyan opposed as the scientific approach. Trevelyan contrasts the creation of stories about history with the search for causal laws. He explicitly argues that the latter—which Trevelyan associates with those who wish to make history a science—is an inappropriate task for the historian. The “alleged ‘science of cause and effect in human affairs’ . . . does not exist, and cannot ever exist in any degree of accuracy remotely deserving to be described by the word ‘science’” (Trevelyan 1913, 6). “The law of gravitation,” he writes, “may be scientifically proved because it is universal and simple.” In contrast, historical events are not simple, nor can they be generalized into universal rules connecting one factor to another. “You cannot so completely isolate any historical event from its circumstances as to be able to deduce from it a law of general application.” As a result, “no causal laws of universal application can be discovered in so complex a subject” as history (ibid., 7). What causal connections can be drawn in history hardly deserve the name of science; “common sense sometimes points to an obvious causal connection,” even though such connections can never become “scientific” or “exact” (ibid., 20). Because of the inherent complexity built into the field, Trevelyan concludes that “in the most important part of its business, history is not a scientific deduction, but an imaginative guess at the most likely generalisations” (Trevelyan 1913, 9). Those who pretend that history could become scientific, in the sense of generating universal causal laws, are deluding themselves. They abandon the tasks that history can perform—the tasks that Trevelyan associates with the literary approach—in favor of tasks at which it must fail. “It is the business of the historian,” he writes, “to generalise and to guess as to cause and effect, but he should do it modestly

Russell’s Literary Approach to History

365

and not call it ‘science’, and he should not regard it as his first duty, which is to tell the story” (my emphasis; ibid., 13). In the conflict between the scientific and the literary approach, Trevelyan’s sympathies are almost entirely on the side of the latter. Admittedly, he does express the hope for “a reconciling process; of a synthesis of the scientific to the literary view of history” (Trevelyan 1913, 3). Such a synthesis should be possible, he believes, because “in the vexed question whether history is an art or a science, let us call it both or call it neither. For it has an element of both.” But this concession to the scientific approach is minimal. For Trevelyan defines the legitimate “scientific” part of historical study so narrowly as to subsume it under the literary. “It is not in guessing at historical ‘cause and effect’ that science comes in; but in collecting and weighing evidence as to facts, something of the scientific spirit is required for an historian, just as it is for a detective or politician” (ibid., 30). Trevelyan denies any connection between the “scientific” part of history and the study of cause and effect; without that, there really isn’t anything particularly “scientific” about the field at all. One hardly has to be a scientist to accumulate a lot of facts. And so Trevelyan’s professed desire to reconcile the scientific and literary approaches is mere pretense; there is nothing meaningful in the scientific approach that Trevelyan wishes to retain. For Trevelyan, the literary approach to history requires the pursuit of a mass audience. There is no reason, after all, to tell a story if nobody is going to listen to it. This forms another point of contrast between history, when properly practiced, and science. Trevelyan identifies two functions with physical science— “direct utility in practical fields; and in more intellectual fields the deduction of laws of ‘cause and effect’.” But “history can perform neither of these functions” (Trevelyan 1913, 6). Trevelyan later elaborates: I conclude . . . that the analogy of physical science has misled many historians during the last thirty years right away from the truth about their profession. There is no utilitarian value in knowledge about the past, and there is no way of scientifically deducing causal laws about the action of human beings in the mass. In short, the value of history is not scientific. Its true value is education. It can educate the minds of men by causing them to reflect on the past. (ibid., 12)

Scientists can do their job even if the masses have no idea what they are doing; the latter can simply enjoy the fruits of the scientific project. But there are no such fruits from historical study; only by learning about history can the masses benefit from such study. The literary approach to history is thus inherently connected to education.

366

The Bloomsbury Companion to Bertrand Russell

What benefits can the general public obtain by “reflecting on the past?” Trevelyan identifies three benefits associated with the storytelling undertaken by the historian for the sake of the public good, which he describes as follows:4 It is the tale of the thing done, even more than its causes and effects, which trains the political judgment by widening the range of sympathy and deepening the approval and disapproval of conscience; that stimulates by example youth to aspire and age to endure; that enables us by the light of what men once have been, to see the thing we are, and dimly to descry the form of what should be. (Trevelyan 1913, 13–14)

Trevelyan elaborates upon each of these three benefits. First, history can “train the mind of the citizen into a state in which he is capable of taking a just view of political problems.  .  . . It can mould the mind itself into the capability of understanding great affairs and sympathising with other men.” The narratives told by history accomplish this, not so much by supplying facts, as by producing “a new state of mind” (ibid.,19). Second, history can “help to teach political wisdom” (ibid., 20). It can do this because “a review of the process of historical evolution teaches a man to see his own age, with its peculiar ideals and interests, in proper perspective as one among other ages.” When someone comes to understand just how many different ideals and interests have been defended by people in different places and times, “his mind will have veritably enlarged” (ibid., 21). Third, and perhaps most importantly, history should “breed enthusiasm” even as it removes “prejudice” (ibid., 22). It provides an almost endless repertoire of ideals and idea that may provide inspiration. “This presentation of ideals and heroes from other ages,” Trevelyan concludes, “is perhaps the most important among the educative functions of history”— precisely because the presentation may lead people today to emulate these heroes and pursue these ideals (ibid., 23). It is worth stressing that Trevelyan associates these benefits with “the tale of the thing done” much more than with “its causes and effects.” He argues that the benefits of the stories told by historians do not depend upon any causal connections drawn by those stories. Such causal connections, assuming they could be drawn, “still would not be the most interesting part of human affairs.” For the “deeds themselves are more interesting than their causes and effects, and are fortunately ascertainable with much greater precision” (Trevelyan 1913, 12). True, such causal connections as can be drawn may help history to provide these beneficial effects. But any such assistance is distinctly secondary. “The study of

Russell’s Literary Approach to History

367

cause and effect is by no means the only, and perhaps not the principal means, of broadening the mind” (ibid., 21). For Trevelyan, then, historians must first and foremost look to the past as a source for stories which may benefit the public. They must compile the facts about past events, put them together intelligently and persuasively, and then present them as stories for the interested reader. The historian who pursues the fool’s errand of causal laws is liable to neglect these indispensable tasks. “Since history has no properly scientific value, its only purpose is educative. And if historians neglect to educate the public, if they fail to interest it intelligently in the past, then all their historical learning is valueless except in so far as it educates themselves” (Trevelyan 1913, 18–19).

3  Russell on the literary approach Russell had written about history before Trevelyan launched his attack on the scientific approach to history. His first book, German Social Democracy (1896), examined the development of the socialist movement in Germany.5 But Russell did not reflect upon the proper approach to history in his writings until “On History.” Russell published this essay in 1904 in the Independent Review. It was Trevelyan who persuaded Russell to write “On History” (Stunkel 2001–02, 145). It is easy to understand why Trevelyan would do this, as Russell largely endorsed Trevelyan’s position. Specifically, he endorsed both the conflict between the literary and the scientific approaches to history and the overall superiority of the literary approach.6 Russell would, with minor variations, endorse the position he staked out in “On History” for the rest of his life. Russell subsequently produced, along with numerous works of a historical nature (such as A History of Western Philosophy), three additional “reflective essays on history and its uses” (Stunkel 2001–02, 129). “The Materialistic Theory of History,” which was primarily a critique of the Marxist approach to history, first appeared in The Practice and Theory of Bolshevism in 1920. “How to Read and Understand History” was originally published by Haldeman-Julius in 1943. And “History as an Art” was delivered as a public lecture in 1954; it was published later that year and then anthologized in Portraits from Memory in 1956. In all three essays, Russell continues to stake out a position similar to that offered by Trevelyan in Clio. It is hardly surprising that Russell should do so; Clio

368

The Bloomsbury Companion to Bertrand Russell

was the “only work on historiography mentioned in his writings on history” (Stunkel 2001–02, 145).7 Russell’s shared position with Trevelyan raises an obvious question. Who influenced whom? Did Russell adopt Trevelyan’s position, did Trevelyan adopt Russell’s position, or did the two friends influence each other? There is no easy answer to this question. The obvious assumption would be that Trevelyan, as a professional historian, influenced Russell, the amateur history enthusiast. But Russell told a somewhat different story in his private journal. He notes going for a walk with Trevelyan shortly after Bury’s inaugural lecture defending the scientific approach. Russell writes that Trevelyan abused Bury’s inaugural lecture, because it said the documentary style was alone legitimate. Then he talked of his own views, which are virtually that history should consist of political pamphlets. I talked of history as an art, but he was not interested, because, as he himself said, he couldn’t do that sort of thing. So I pointed out that he was committing a fault just like Bury’s. (Papers 12, 19)

Here Russell seems to claim some sort of influence on Trevelyan, and possibly some authorship for the literary approach defended by both men.8 But even the editors of Russell’s collected papers describe Russell’s account of this meeting as “biased and self-important” (Papers 12, 73). So the specific share of credit due to each man for the literary approach they both endorsed remains unresolved. What is beyond dispute, I will argue, is that the approach to history that Russell defended throughout his life was largely the same as the literary approach defended by Trevelyan in “Clio.” Like Trevelyan, Russell took up the question “as to whether history is a science or an art.” Russell agreed with Trevelyan that it was “entirely obvious that it is both,” but in a manner that rendered the “scientific” component of history negligible. For Russell, history could be scientific only in the “comparatively pedestrian sense in which science is involved in ascertaining historical facts” (BW, 512). Beyond this limited “scientific” task of getting the facts right, historians must devote themselves to interpreting the facts and then presenting them to a reading public as compelling stories. This is in accord with the three components of historical work identified by Trevelyan. History, according to Russell, must be “interesting,” and “not only to men who for some special reason wish to know some set of historical facts, but to those who are reading in the same spirit in which one reads poetry or a good novel” (ibid., 516). When it comes to history that aims to be scientific in a more ambitious sense,  Russell is generally as dismissive as Trevelyan. He shares Trevelyan’s

Russell’s Literary Approach to History

369

attitude, expressed in “Clio,” that “history cannot, like physical science, deduce causal laws of general application. All attempts have failed to discover laws of ‘cause and effect’ which are certain to repeat themselves in the institutions and affairs of men” (Trevelyan 1913, 7). For Russell, this does not mean that prediction in human affairs is impossible, but prediction only counts as scientific if it is “made explicitly by means of a more or less general law obtained inductively from observed facts.” People can make predictions without being scientific; if they could not, much human activity would be impossible. “A good horseman,” Russell writes, “can predict the behaviour of his horse, but he does so by sympathetic imagination rather than by science. A  skilled negotiator uses the same kind of faculty in foreseeing the response to a proposal that he thinks of making. Successful politicians similarly divine mass responses. All this is prediction, but it is not science” (Russell 1951, 735).9 No general laws, no science.10 According to Russell, the effort to produce laws of history has failed due to the inherent complexity and contingency built into history, as opposed to the physical sciences, where control of such factors is possible. “Whether there will ever be a science of history,” Russell writes, “it is quite impossible to guess; at any rate it is certain that no such science exists at present, except to some slight degree in the province of economics” (BW, 500).11 Any science worthy of the name will involve the discovery of generally applicable causal laws. “But,” Russell writes, it will hardly be maintained that history has reached, or is soon likely to reach, a point where such standards are applicable to its facts. History, considered as a body of truth, seems destined long to remain almost purely descriptive. Such generalizations as have been suggested—omitting the sphere of economics—are, for the most part, so plainly unwarranted as to be not even worthy of refutation. (ibid., 501)12

“I do not deny,” Russell later writes, “I do not mean to deny that it is a good thing to discover causal sequences in history when it is possible, but I think the possibility exists only in rather limited fields,” primarily economics (ibid., 513).13 Throughout Russell’s long life, he remained critical of the ambition of making history a science. The only variation in his opinion was in his level of certainty that this ambition could never be achieved. He most commonly argued that human affairs were sufficiently complex as to render it impossible to articulate general laws of cause and effect in history. But occasionally, he would concede

370

The Bloomsbury Companion to Bertrand Russell

the possibility that such laws might one day be discoverable in principle. In “How to Read and Understand History,” for example, he writes that while some aspects of history can be made more or less scientific, and while it is important to do this wherever it is possible, the material is too complex to be reduced to scientific laws at present, and probably for centuries to come. There is too much that, to our ignorance, appears as chance, and too great a likelihood of the intrusion of incalculable forces. There is nothing genuinely scientific in a premature attempt to seem scientific. (emphasis in original; Russell 1957, 38)14

Here Russell admits to the possibility of a genuine science of human affairs, but argues against “premature” pretenses of the attainment of such a science. But variations in Russell’s opinion such as this one do not affect Russell’s general attitude toward history. History cannot be a science now, and it cannot be a science for a very long time to come, regardless of whether or not it might ever be a science (cf. Willis 1987, 121). Russell’s criticisms of the scientific approach (in the non-minimalist sense) apply with especial force to those large schemes of historical development which have fascinated many eminent men from St Augustine to Professor Toynbee. In modern times, the most important inventors of general theories as to human development have been Hegel and his disciple Marx. Both believed that the history of the past obeyed a logical schema, and that this same schema gave a means of foretelling the future. (emphasis in original; BW, 513–14)

Russell refers to large schemes of this sort as philosophies of history. Elsewhere, Russell elaborates: If it is possible to distinguish between a philosophy and a science of history, I should say that, while certain departments of history can already be made more or less scientific, and one may hope that many more may be, the attempt to create a philosophy of history is a mistake. I should regard men like Hegel, Marx, and Spengler as having a philosophy of history, in the sense that they believe in sweeping laws of historical development, either progressive or cyclic. For such vast laws, I should say, there is not, and never can be, any adequate evidence; they are reflections of our own moods upon the cosmos. (emphasis in original; Russell 1951, 734)

The philosophers of history thus demonstrate an extreme form of the scientific approach to history, one that Russell dismissed as counterproductive.15 Much of Russell’s essay “How to Read and Understand History” is taken up with criticizing those who believe that the study of history necessitates the

Russell’s Literary Approach to History

371

development of a “philosophy of history.” Indeed, the contrast between Russell’s position and that of the leading philosophers of history—G. W .F. Hegel and Karl Marx—lies at the heart of the essay. When it comes to Hegel and Marx—here as in so many other writings—Russell is unsparing in his criticisms. He lumps both thinkers together with the all-but-forgotten Oswald Spengler and the cranks who sought a “divine message” in the Great Pyramid. In particular, he writes that “Hegel’s theory of history is not a whit less fantastic” then the nonsense produced by the pyramidologists (Russell 1957, 15).16 Marx is dismissed in much the same tone. “Like the early Christians,” Russell writes, “Marx expected the millennium very soon; like their successors, his have been disappointed—once more, the world has shown itself recalcitrant to a tidy formula embodying the hopes of some section of mankind” (ibid.,16). Given Russell’s well-known attitude toward Christians and the accuracy of their prognostications, this comparison is hardly flattering to Marx. The effort to identify causal laws in history is not only wrongheaded because it is almost guaranteed to fail, at least in the short term. It is dangerous because it can cause historians to neglect the importance of the historical events themselves. Following Trevelyan, Russell recognizes that chronicles of events can be valuable in themselves, without regard to their causes or effects. In “On History,” Russell contrasts the physical sciences to history in these terms: Where our main endeavour is to discover general laws, we regard these as intrinsically more valuable than any of the facts which they interconnect. In astronomy, the law of gravitation is plainly better worth knowing than the position of a particular planet on a particular night, or even on every night throughout a year. There are in the law a splendour and simplicity and sense of mastery which illuminate a mass of otherwise uninteresting details. … But in history the matter is far otherwise. In economics, it is true, the data are often subordinate to the attempts at science which are based upon them; but in all other departments, the data are more interesting, and the scientific superstructure less satisfactory. Historical facts, many of them, have an intrinsic value, a profound interest on their own account, which makes them worthy of study, quite apart from any possibility of linking them together by means of causal laws. (BW, 501)

Decades later, in “History as an Art,” Russell expresses the same sentiment as follows: It may be, as some have suggested, that Napoleon lost the battle of Leipzig because he ate a peach after the battle of Dresden. If this is the case, it is no doubt

372

The Bloomsbury Companion to Bertrand Russell

not without interest. But the events which it connects are on their own account much more interesting. In physical science, exactly the opposite is true. Eclipses, for example, are not very interesting in themselves except when they give fixed points in very early history. . . . But although most eclipses are not interesting in themselves, the laws which determine their recurrence are of the very highest interest, and the discovery of these laws was of immense importance in dispelling superstition. . . . But history is not like this. Most of the value of history is lost if we are not interested in the things that happen for  their own sakes. In this respect history is like poetry. There is a satisfaction to curiosity in discovering why Coleridge wrote “Kubla Khan” as he did, but this satisfaction is a trivial affair compared to that which we derive from the poem itself.

Russell even explicitly evokes “Clio” as the source for his thoughts on this subject (ibid., 513). Passages such of these help to clarify Russell’s meaning when he declares, in “History as an Art,” that “scientific laws in history are neither so important nor so discoverable as is sometimes maintained” (BW, 513). Russell makes very plain why he believes that scientific laws are not very “discoverable.” But why should such laws not be “important?” Surely scientific laws of history, if they could be discovered, would be of the greatest importance. Russell’s concern here seems to be that a focus upon the identification of scientific laws in history—a focus Russell regards as almost guaranteed to be fruitless—can generate a neglect for other aspects of historical study that are of great value. Put another way, the study of history cannot offer us causal laws, but it can offer us other things, and if we spend all of our time studying history in a vain attempt to obtain the former, we will risk missing out on the latter, entirely to our own detriment. So what can historical study offer, apart from causal laws? What benefits follow from the consideration of historical events in themselves? Here again, Russell follows Trevelyan in finding history valuable because of its educative effects. And like Trevelyan, Russell believes that the educative importance of history necessarily makes it an activity aimed at the general public. The community of physicists can have a great impact even if the public remains ignorant of its findings. But the community of historians can only do good by inculcating certain attitudes in a mass readership. Russell expresses this idea with particular eloquence in the opening to “History as an Art.” “It would never do,” he writes, “for a mathematician to try to please the general reader. The physical sciences in their serious aspects must be addressed primarily to specialists.” Indeed, “there have to be mathematicians until calculating machines become cheaper, but when that happy consummation

Russell’s Literary Approach to History

373

has been reached, there will be no point in teaching anybody to do sums, and the multiplication table can be placed alongside the birch as an out-of-date instrument of education.” The general public simply does not need much knowledge of the results of science.17 “But history,” he continues, seems to me to be in a different category. The multiplication table, though useful, can hardly be called beautiful. It is seldom that essential wisdom in regard to human destiny is to be found by remembering even its more difficult items. History, on the other hand, is—so I shall contend—a desirable part of everybody’s mental furniture in the same kind of way as is generally recognized in the case of poetry. If history is to fulfil this function, it can only do so by appealing to those who are not professional historians. (BW, 511–12)

There is thus an intrinsic link between the positive effects of history as a field and the education of the general public in this field, a link that does not exist in fields such as mathematics or physics. And because these positive effects are so important, “history should not be known only to historians,” anymore than “poetry should only be read by poets” (ibid., 515). But what precisely are the positive effects Russell identifies with history written for the general public? Russell recognizes more benefits from the study of history than does Trevelyan—“The value of history,” he writes, “is so multiform that those to whom some one of its sides appeals with especial force are in constant danger of forgetting all the others” (BW, 499). Nevertheless, some of the benefits Russell identifies clearly echo those offered by Trevelyan in “Clio.” In particular, Russell echoed Trevelyan in believing that history “fills our thoughts with splendid examples, and with the desire for greater ends than unaided reflection would have discovered” (BW, 503). This is an advantage to the study of history to which Russell would devote attention in both “How to Read and Understand History” and “History as an Art.” In the latter work, for example, Russell points out that “Heroic lives are inspired by heroic ambitions, and the young man who thinks that there is nothing important to be done is pretty sure to do nothing important. For such reasons I think the kind of history that is exemplified by Plutarch’s Lives is quite as necessary as the more generalized kind.” This benefit does not require that all of the examples history provides be splendid. We can learn not only from “men whom we may regard as models to be imitated,” but also from “all those who afford new material for imagination” (ibid., 519). History lets people know that there are many heroes one might emulate and many ideals to which one might devote oneself; once people learn this lesson, they are better equipped to find examples they find truly admirable.

374

The Bloomsbury Companion to Bertrand Russell

It would be a mistake to say that Russell’s approach to history was identical with that of Trevelyan. Such an identity would be surprising if it existed, given the differences in education and temperament between the two men. Russell did disagree with Trevelyan on a number of points. For example, Trevelyan stressed the need for the historian to get her hands dirty with the retrieval of historical facts. Russell, in contrast, argued for a “division of labour.” He based his argument for this upon literary considerations. The good writer, Russell notes in “History as an Art,” “needs a certain freshness of feeling which is apt to be destroyed by fatigue.” Russell continues: If expository prose is to be interesting, there has to be a period of incubation, after the necessary knowledge has been acquired, when the bare facts will become clothed with such associations as are appropriate, of analogy or pathos or irony or what not, and when they will compose themselves into the unity of a pattern as in a play. This sort of thing is hardly likely to happen adequately unless the author has a fair amount of leisure and not an unfair amount of fatigue. (BW, 517)

None of this is possible in the modern world, Russell believes, without the division of labor. There is simply too much evidence to sift and digest. The historian who collects original data will struggle to do more than produce esoteric reports for other historians—precisely the kind of work not aimed at a mass audience, which for the literary approach is the field’s raison d’être. And so “the archaeologist or the man who delves in unpublished manuscript material is likely to have neither the time nor the energy for large-scale history. The man who proposes to write large-scale history should not be expected himself to do the spade work.” There is nothing shameful in such a division; the natural sciences make use of them all the time, as the example of Kepler and Tycho Brahe demonstrates. (ibid., 518)18 But disagreements like this between Trevelyan and Russell are rather small. The question is whether there are disagreements to be found of a more important nature. This question matters because of the possible objections one might raise against the literary approach. The next section considers one of the most important of these objections.

4  Trevelyan and Russell on truth Both Trevelyan and Russell ably demonstrate the difficulties facing the scientific approach to history. In particular, Russell is relentless at deflating the pretensions

Russell’s Literary Approach to History

375

of philosophers of history, as they try to formulate general laws for the human world. Such laws will not be forthcoming anytime soon, if ever, and the failed attempts can have catastrophic consequences, as the history of Marxism demonstrates. But a catalogue of the problems with the scientific approach is not sufficient to vindicate the literary approach—it may well face more serious problems of its own. In this section, I will consider one of the more serious potential objections to the literary approach to history. What role does historical truth play in the literary approach? Why should truth matter to the historian? Both Trevelyan and Russell grappled with this question. It was impossible for them to avoid it, given the central role both of them assigned to truth in the study of history. But neither author, I will argue, was able to offer a truly satisfying answer to the question (although Russell’s answer was more satisfying than that of Trevelyan). In “Clio,” Trevelyan stresses that the careful collection of facts is an indispensable part of the historian’s job. It is only after this collection has been accomplished that the historian can craft uplifting and educative narratives to share with a mass audience. “To give a true picture of any country, or man or group of men in the past requires industry and knowledge, for only the documents can tell us the truth, but it requires also insight, sympathy and imagination of the finest, and last but not least the art of making our ancestors live again in modern narrative” (Trevelyan 1913, 17). As important as the narrative functions of history may be, they can only be served after the facts are right. Moreover, he warns against the “undoubted temptation to the artist to neglect . . . small, inconvenient pieces of truth” (ibid., 35).19 And he praises Gibbon for his extraordinary ability to unite “accuracy with art” (ibid., 38). The question, however, is why precisely should accuracy be so important to Trevelyan, given his commitment to the literary approach. If the goal of historical scholarship is to create educational stories, why should it matter whether or not the stories are true? There is no doubt that Trevelyan wished for historians to create fact-based, and not simply inspirational, stories; he surely would have recognized the line between history and propaganda. And yet Trevelyan does not always seem to put the truth regarding the past front and center. He occasionally seems willing to countenance history that falls short of the truth, so long as it provides the valuable narratives he seeks. For example, Trevelyan praises Thomas Carlyle as a historian even while acknowledging the ways that Carlyle “sometimes neglected the accumulation of facts and the proper sifting of evidence” (Trevelyan 1913, 11). This admittedly makes Carlyle less than a model historian, but it also suggests that the purely

376

The Bloomsbury Companion to Bertrand Russell

literary qualities of a work of history can to some extent make up for its shortcomings as fact. Later on in “Clio” Trevelyan praises Carlyle’s account of a  critical night during the French Revolution. “Whether or not” Carlyle’s account “is entirely accurate in detail, it is true in effect,” Trevelyan concludes (ibid., 17). Even more troubling is Trevelyan’s effusive praise for Sir Walter Scott, a man who Trevelyan sees as correcting for some of the limitations demonstrated by Gibbon. And yet while Scott wrote some nonfiction, it is for his novels that he has always been best known. Does it make sense to praise a novelist for being a good “historian?” Trevelyan seems to think so. Contra Gibbon, Scott demonstrated that “a man is not so much a human being as a type produced by special environment whether it be a border-farmer, a medieval abbot, a cavalier, a covenanter, a Swiss pikeman, or an Elizabethan statesman” (Trevelyan 1913, 39). The fact that Scott accomplished all of this while writing fiction does not concern Trevelyan. “Both as literature and as social history his Scotch novels are his best,” he writes (my emphasis; ibid., 39, n. 1). The line between pure literature and the literary approach to history that Trevelyan endorses is not as firm as one might perhaps desire. In “On History,” Russell similarly claims that “history is valuable, to begin with, because it is true; and this, though not the whole of its value, is the foundation and condition of all the rest” (BW, 499). He goes on to say “that the writing of history should be based on the study of documents is an opinion which it would be absurd to controvert. For they alone contain evidence as to what really occurred; and it is plain that untrue history can have no great value” (ibid., 500). Moreover, in his otherwise-glowing review of Clio, Russell chides Trevelyan for not doing enough justice to “the patient discoveries of those who, following the ‘scientific’ impulse, have made it possible to know the truth about the past as it could not be known through the historians of a less laborious school.” For “history which is not true, however splendidly written, has an essential lack” (Papers 12, 407). Russell here seems to suggest that Trevelyan must take more seriously the importance of truth in history. But Russell tends to assume the relevance of truth to history rather than argue for it. In “History as an Art,” for example, he writes, “We all think it worth while to know about the great heroes of tragedy—Agamemnon, Oedipus, Hamlet and the rest—but there have been real men whose lives had the same quality as that of the great tragic heroes, and had the additional merit of having actually existed” (BW, 519). But why precisely should it matter that these great men were real? How should that fact add or detract from the benefits that the study of their

Russell’s Literary Approach to History

377

lives provides to us? Why, in short, should Russell, anymore than Trevelyan, attach such importance to historical truth, given his commitment to the literary approach? Russell does offer a few suggestions as to the importance of truth to historical study. In “On History,” he writes “That all knowledge, as such, is in some degree good, would appear to be at least probable; and the knowledge of every historical fact possesses this element of goodness, even if it possesses no other” (BW, 499). But this is a rather slim reed on which to hang a connection between truth and the literary approach. It would render truth an added side benefit of a historical work written primarily to accomplish other ends. Moreover, Russell is quite contemptuous in “On History” of the view (which he associates, possibly unfairly, with the scientific approach) that all historical facts are of equal importance. This raises the question for Russell of whether some historical errors are more serious than others, and if so, why. And this question would arise even if one conceded that all truth has some value. In “History as an Art,” Russell offers a different argument regarding truth and the study of history. He writes: History, however much it may be pursued as an art, has to be controlled by the attempt to be true to fact. Truth to fact is a rule of the art, but does not in itself confer artistic excellence. It is like the rules of the sonnet, which can be scrupulously observed without conferring merit on the result. But history cannot be praiseworthy, even from the most purely artistic point of view, unless the historian does his utmost to preserve fidelity to the facts. Science in this sense is absolutely essential to the study of history. (ibid., 512)

The difficulty with this position is that the rules of a sonnet are to a significant extent arbitrary. Admittedly, such arbitrary rules or restrictions can play a  critical part in artistic or literary production (Elster 2000, ch. III). But surely Russell is not claiming that truth functions as a purely arbitrary “rule  of  the art.” He must believe that truth is essential to the function of historical writing. The question, however, is how Russell could justify making such a link. Some of Russell’s efforts to defend a link between a commitment to truth and the literary approach to history are thus less than successful. Moreover, some of the examples Russell provides—examples that emphasize the literary nature of the enterprise—display a shockingly casual attitude toward the truth. (I say “shockingly,” given Russell’s reputation as a fearless freethinker, a man committed to the pursuit of truth whatever the consequences.) Consider, for

378

The Bloomsbury Companion to Bertrand Russell

example, Russell’s discussion of what historians have written about Agathocles, the tyrant of Syracuse: At one time I read what Diodorus Siculus has to say about Agathocles, who appeared as an unmitigated ruffian. I looked up Agathocles afterwards in a modern reference book and found him represented as bland and statesmanlike and probably innocent of all the crimes imputed to him. I have no means of knowing which of these two accounts is the more true, but I know that the whitewashing account was completely uninteresting. I do not like a tendency, to which some modern historians are prone, to tone down everything dramatic and make out that heroes were not so very heroic and villains not so very villainous. No doubt a love of drama can lead an historian astray; but there is drama in plenty that requires no falsification, though only literary skill can convey it to the reader. (my emphasis, BW, 516)

Here Russell seems completely uninterested in whether he is being told truths or fictions about Agathocles—he just knows which story he likes better. And admittedly, to the extent that history is supposed to inspire feelings of admiration or revulsion—to provide heroes and villains, ideals and enemies—it is hard to see why he should need to know more. But giving up completely on historical accuracy here simply effaces the line between history and fiction.20 Russell’s evaluation of Gibbon in “History as an Art” displays a similar attitude. Gibbon, Russell writes, offers us a stately procession of characters marching through the ages, all in court dress and yet all individual. Not long ago I was reading about Zenobia in the Cambridge Ancient History, but I regret to say that she appeared completely uninteresting. I remembered somewhat dimly a much more lively account in Gibbon. I looked it up, and at once the masterful lady came alive. Gibbon had had his feelings about her, and had imagined what it would be like to be at her Court. He had written with lively fancy, and not merely with cold desire to chronicle known facts. It is odd that one does not more resent the fact that his characters all have to be fitted into an eighteenth-century mould. I remember that somewhere in dealing with the Vandals after the time of Genseric he speaks of “the polished tyrants of Africa.” I am quite unable to believe that these men were polished, though I have no difficulty in believing that they were tyrants. But somehow, in spite of such limitations, Gibbon conveys an extraordinarily vivid sense of the march of events throughout the centuries with which he deals. (BW, 517–18)

Russell’s Literary Approach to History

379

Russell here frankly admits that Gibbon transforms Zenobia and other figures of the ancient world into members of the eighteenth-century gentility, and yet he has little complimentary to say for a more accurate but “completely uninteresting” contemporary source (This example is all the more striking given Trevelyan’s praise of Gibbon for his commitment to historical accuracy). A commitment to the literary approach to history seems to leave Russell ill-equipped to distinguish between history and propaganda. In “History as an Art,” Russell condemns the way that many German historians have, “under the influence of Hegel … combined brilliance and wickedness in equal proportions.” “When we speak of the importance of history,” Russell admits, “we must admit its importance for evil as well as for good.” But the grounds for condemning German uses of history as propaganda are not completely clear. Are these German historians not providing interesting stories for their readership? Are they failing to make vivid ideals and heroes that may inspire people to noble action? If not, then Russell may condemn the specific choices of ideals and heroes identified by German historians like Heinrich von Treitschke (who Russell singles out for “spreading a pernicious myth”), but this is a purely literary judgment, and it is not clear how the literary approach could offer more (BW, 520–21).21

5  Beyond the literary approach I have suggested that Russell, to the extent that he was committed to the literary approach to history he shared with Trevelyan, was ill-equipped to explain the importance of truth to historical study. But as noted before, Russell sometimes went beyond the literary approach. This suggests that Russell may have had ways of explaining the connection between truth and history to which Trevelyan may not have had access. In this concluding section, I shall explore two of those ways. First, Russell was more willing than Trevelyan to countenance the possibility that some (nonobvious) generalizations might become possible thanks to the study of history. History may, as Russell notes in “On History,” “suggest minor maxims, whose truth, when they are once propounded, can be seen without the help of the events that suggested them” (Again, Russell believes this to be particularly true of economics).22 But such maxims “will only apply where

380

The Bloomsbury Companion to Bertrand Russell

the end is given, and are therefore of a technical nature. They can never tell the statesman what end to pursue, but only, within certain limits, how some of the more definite ends, such as wealth, or victory in war, are to be attained” (BW, 502).23 This is a limited form of generalization, to be sure, but Russell himself was quite willing to make use of it. His own writings on human affairs are littered with maxims generated via the study of history. And some of the lessons Russell learned from history remain of interest to social science today (e.g., Hardin 1996).24 Second, Russell identified a number of moral lessons that history can provide that are inseparable from the truth of that history. This is particularly true in “On History.” This essay was written at a critical point in Russell’s life. His first marriage, to Alys Pearsall Smith, had collapsed in 1902, although the couple remained married for almost twenty years after that (Moorehead 1993, 115–17). For months thereafter, Russell suffered from deep melancholia to which he strove to give expression through writing.25 This melancholia drove him toward a form of mysticism, a position from which Russell would later distance himself.26 Russell expressed himself at this time through work on an essay entitled “The Pilgrimage of Life,” which he never completed (Papers 12, ch. 2), as well as the more famous “A Free Man’s Worship,” which was published in 1903 in the Independent Review—just as “On History” would be  a  few months  later (Papers 12, ch. 4). The attitude of despair expressed in “A Free Man’s Worship” is quite evident at the end of “On History,” and it provides another reason for caring about history, one heavily dependent upon historical truth. At the end of “On History,” after cataloging a number of more mundane advantages to the study of history (most of which stemmed from the literary approach), Russell identifies another, more profound advantage. “The record of great deeds,” he writes, “is a defeat of Time, for it prolongs their power through many ages after they and their authors have been swallowed by the abyss of the non-existent.” By reviewing the past, “we see, more clearly than in the lives about us, the value for good and evil, of the aims men have pursued and the means they have adopted.” This makes it possible for us “to view the present as already past, and to examine what elements it contains that will add to the world’s store of permanent possessions, that will live and give life when we and all our generation have perished.” We can then see ourselves as part of a single large project—the human project—one that has preceded us for ages and that will continue long after we are dead (BW, 503–04).

Russell’s Literary Approach to History

381

This perspective—a very mystical perspective, it may be noted, whereby people come to perceive the oneness of the human race—in turn has several valuable effects. For one thing, “by the contemplation of great lives, a mystic communion becomes possible, filling the soul like music from an invisible choir.” The benefits of this communion are difficult to describe and yet unmistakable for Russell: To join in this glorious company, to swell the immortal paeon of those whom fate could not subdue—this may not be happiness; but what is happiness to those whose souls are filled with that celestial music? To them is given what is better than happiness: to know the fellowship of the great, to live in the inspiration of lofty thoughts, and to be illumined in every perplexity by the fire of nobility and truth. (BW, 504)

For another, this perspective makes evident to us the drama of the species. Russell writes that “it is the province of history to tell the biography, not only of men, but of Man; to present the long procession of generations as but the passing thoughts of one continuous life; to transcend their blindness and brevity in the slow unfolding of the tremendous drama in which all play their part” (ibid., 504). And just like the communion with great lives that history can inspire, so the perception of this drama has a value all its own. “We, too, in all our deeds, bear our part in a process of which we cannot guess the development: even the obscurest are actors in a drama of which we know only that it is great. Whether any purpose that we value will be achieved, we cannot tell; but the drama itself, in any case, is full of Titanic grandeur” (ibid., 505). When history is studied properly, it helps people to feel like a part of this unfolding drama.27 Russell did not maintain this mystical attitude throughout his life; his relationship later in life with “A Free Man’s Worship,” for example, was decidedly ambiguous (Papers 12, 63–64). And yet echoes of the mystical perspective Russell expressed in “On History” persisted throughout his later work. In “History as an Art,” for example, Russell claims that “the reading of history” has “first and foremost” provided him with “something like a new dimension in the individual life, a sense of being a drop in a great river rather than a tightly bounded separate entity” (BW, 515). And it would not be too much of a stretch to see the same attitude, or something like it, expressed in Russell’s famous antiwar admonition, “Remember your humanity, and forget the rest.” Even if people do not need the mystical union with the entire human race that the early Russell

382

The Bloomsbury Companion to Bertrand Russell

praised in “On History,” they benefit from an ability to distance themselves from their particular time, place, and circumstances (e.g., the particular nationality to which they belong). The study of history can help people to achieve this critical distance. But presumably it can only attain ends such as these if the history truly chronicles the history of humanity. A false-but-interesting story about human beings simply would not do.28 If this argument is correct, then Russell possessed good reasons to want the history presented to the general public to be accurate, reasons that Trevelyan may have lacked. The reason for this is that while Russell shared with Trevelyan a literary approach to history, this approach did not capture all of Russell’s beliefs about history and its value. Russell, for example, believed that useful lessons could be drawn from history, even if those lessons reliably failed to involve general laws of cause and effect. He also believed that human beings benefited from an ability to distance themselves from their own time and space and view themselves in relation to the rest of humanity, although his understanding of these benefits may have changed over time. These are surely reasons significant enough to justify a strong attachment to historical truth. In his “Good Citizen’s Alphabet,” Russell defined a pedant as “a man who likes his statements to be true.” As a consumer of history, Russell was definitely a pedant, and more importantly, he had good reasons to be one.

Notes 1 Based upon remarks made at “Understanding History: 50 Years Later,” a panel discussion held at the 2007 Annual Meeting of the Bertrand Russell Society, Monmouth University, June 8–10, 2007. My thanks to everyone who participated in the panel. 2 See Stunkel (2001–02, 142–44) for some critical responses to A History of Western Philosophy. 3 Numerous scholars, including Russell himself, have connected Russell’s approach to history with Trevelyan (e.g., Willis 1987; Stunkel 2001–02). And yet none, to the best of my knowledge, have explored in detail just how similar the two men’s approaches really are. This will be one of the primary goals of this essay. 4 Trevelyan identifies a few other benefits to be gained from the study of history—it can, for example, greatly increase the “value and pleasure of travel” (Trevelyan 1913, 25)—but these are of distinctly secondary importance.

Russell’s Literary Approach to History

383

5 German Social Democracy, and the respectable-if-critical reception it received, clearly demonstrates that Russell had some skills as a historian. On the book’s reception, see Willis (1976) and Stunkel (2001–02, 132–34). 6 Russell subsequently had the opportunity to review Trevelyan’s book. In his review he described “Clio” as a “truly remarkable essay” (Papers 12, 407). 7 Like Trevelyan, Russell did not explicitly refer to the “literary” and “scientific” approaches I am describing. Russell himself drew a number of different distinctions between ways of studying history. In a 1919 book review, he distinguished “two radically different ways of studying history,” the “dramatic” and the “sociological” (Papers 15, 87). And in a short 1924 essay entitled “How to Read History,” Russell wrote that “History may be written as drama, as epic, or as science; as a rule, it combines elements of all three” (Russell 1924, 10). 8 In addition, it appears that Russell began writing “On History” before, or at least at the same time, as Trevelyan began writing his own essay. At some point in 1903, Trevelyan wrote to Russell that “Your ‘History’ makes me feel ashamed of the essay I am attempting on the same subject” (Papers 12, 74–75). 9 For this reason, Russell would have seen no inconsistency between denying the possibility of historical laws and making political forecasts. But this defense might not suffice to save Russell from the charge of inconsistency. Sidney Hook (1951, 652–54) ably documents numerous instances where Russell appears to appeal to lawlike generalizations regarding history. He argues that Russell does not count these as laws of history due to his exceptionally (and unnecessarily, in Hook’s view) stringent standard of what counts as a scientific law. “When he takes distance to the subject matter of history,” Hook concludes, “Russell seems to be skeptical about the existence of historical laws. But in his actual historical writing he invokes them constantly, sometimes with fruitful results” (ibid., 652). 10 “Of course, in his post-Second World War writings, Russell himself often predicted the likely outcome of the human species—complete and utter annihilation. But he did not claim that this was a scientific prediction, as it was not inevitable, but only very likely given current circumstances. He was willing to grant that human attitudes could change, thereby changing the likely outcome, something which philosophers of history such as Hegel, Marx or Spengler would never admit” (Madigan 2010, 65). 11 Russell persistently suggests that in the realm of economics, the study of human affairs more closely approximates a real science than in any other. For more on Russell’s views on economics, see King (2005). 12 This may explain Kirk Willis’s judgment that “Russell’s explicit analyses of such bogus, ‘scientific’ philosophies of history—specifically those of Marx, Hegel, Spengler, and Toynbee—were, unfortunately, breezily perfunctory at best and

384

13 14

15

16

17 18

19

20

The Bloomsbury Companion to Bertrand Russell irresponsibly slapdash at worst. Indeed, he preferred to dismiss rather than to refute them” (Willis 1987, 132). This is perhaps unfair to Russell with regard to Marx and Hegel, about whom he had much to say. As for Spengler and Toynbee, scholars today would probably agree with Russell’s “breezily perfunctory” dismissal. Cf. Stunkel (2001–02). Cf. Russell’s Freedom versus Organization: “History, in short, is not yet a science, and can only be made to seem scientific by falsifications and omissions” (Russell 1934, viii). It is therefore unfortunate that the Basic Writings of Bertrand Russell collects “On History,” “The Materialistic Theory of History,” and “History as an Art” together in a section entitled “The Philosopher of History” (BW, Part XIII). This is a title that Russell would surely never have applied to himself. But the title certainly has been applied to him. Sidney Hook, for example, entitles his essay in The Philosophy of Bertrand Russell “Bertrand Russell’s Philosophy of History” (Hook 1951). Interestingly, while Russell dismisses Hegel’s philosophy as ridiculous, he does not argue that it yields objectionable political conclusions. This is an argument, however, that Russell did pursue elsewhere (Russell, 1947). In his writings on education, Russell adopts a more favorable attitude toward scientific education. Space prohibits further discussion of this problem here. This may be why Russell seems more willing than Trevelyan to countenance history produced purely for the scholarly community. When Russell distinguishes between “what history can do and should do for the general reader” and “what history does for historians,” he does not seem to regard the latter type of history as inherently suspect, as Trevelyan does (Russell BW, 515). At the same time, Trevelyan fears that “In history, as it is now written, art is sacrificed to science ten times for every time that science is sacrificed to art” (Trevelyan 1913, 35). According to Kenneth Stunkel, Russell believed that the “proper function of the historian is to balance a scientific obligation of truthfulness and accuracy with an artist’s sense of drama that reaches the imagination” (Stunkel 2001–02, 150). On the one hand, it is hard to imagine Russell wanting to “balance” truthfulness and accuracy against other considerations in any other area of life, except of course for the world of fiction. (Russell would not, for example, have demonstrated much respect for a theologian who wanted to “balance” the truth-value of religious claims—claims regarding the existence of God, for example—against other considerations, such as the usefulness of those claims for public morality.) On the other hand, to the extent that Russell was indeed committed to the literary approach, it is difficult to say just why any value must be assigned to the “scientific obligation of truthfulness and accuracy” at all.

Russell’s Literary Approach to History

385

21 Russell could, of course, condemn German authors inspired by Hegel—just as he condemned Hegel himself—for claiming to identify “laws of history.” “I do not think,” Russell wrote, “it would be unfair to Mommsen [the German classics scholar] to say that history had two themes: one, the greatness of Caesar because he destroyed liberty; the other, that Carthage was like England and Rome was like Germany and that the future Punic Wars to which he looked forward would have an outcome analogous to that of their predecessors” (BW, 521). To the extent that Mommsen et al. claimed to predict the future based on lawlike generalizations from ancient history, they would be guilty in Russell’s eyes of the wrongheaded scientific approach. 22 Russell seems to have in mind the deductive models that have become the mainstay of economic theorizing. “Wherever,” he writes, “out of the facts, a simple deductive argument from indubitable premises can be elicited, history may yield useful precepts” (BW, 502). 23 One could draw parallels between Russell’s minor maxims of history and Kant’s imperatives of skill (Kant 1997, 26). Both are “hypothetical imperatives” that inform an agent how to attain previously specified ends. 24 From this perspective, even Marxism has interesting lessons to offer. “Treated as a practical approximation,” Russell writes, “and not as an exact metaphysical law, the materialistic conception of history has a very large measure of truth” (BW, 507). 25 Russell used writing for a great many purposes. This may in part explain why he felt such an intense need to write throughout his life. Cf. Stone (2016). 26 On Russell and mysticism see Andersson (1994). 27 Russell adopts a similar perspective in “How to Read and Understand History.” “Our bodily life,” he writes, “is confined to a small portion of time and space, but our mental life need not be thus limited. What astronomy does to enlarge the spatial habit of the mind, history does to increase the temporal domain. Our private lives are often exasperating, and sometimes almost intolerably painful. To see them in perspective, as an infinitesimal fragment in the life of mankind, makes it less difficult to endure personal evils which cannot be evaded” (my emphasis; Russell 1957, 54; cf. Russell 1924, 12). 28 To some extent, Russell’s later position resembles Trevelyan, insofar as the latter thought that history “teaches a man to see his own age, with its peculiar ideals and interests, in proper perspective as one among other ages” (Trevelyan 1913, 21). Space prohibits any effort here to develop a distinction between the two. I should add, however, that this position is some distance away from the mystical position of the early Russell, and that the connection of the former to historical truth is not as direct as the connection of the latter. One can, after all, derive some ability to see one’s own age in perspective by reading science fiction or the like.

386

The Bloomsbury Companion to Bertrand Russell

References Andersson, Stefan. 1994. In Quest of Certainty: Bertrand Russell’s Search for Certainty in Religion and Mathematics up to The Principles of Mathematics (1903). Stockholm: Almqvist & Wiksell. Elster, Jon. 2000. Ulysses Unbound. New York: Cambridge University Press. Goodlad, Lauren M. E. 2015. The Victorian Geopolitical Aesthetic: Realism, Sovereignty, and Transnational Experience. New York: Oxford University Press. Hardin, Russell. 1996. “Russell’s Power.” Philosophy of the Social Sciences 26 (3): 322–47. Hook, Sidney. 1951. “Bertrand Russell’s Philosophy of History.” In The Philosophy of Bertrand Russell. 3rd ed., edited by Paul Arthur Schilpp. New York: Tudor. Kant, Immanuel. 1997. Groundwork of the Metaphysics of Morals. Translated by Mary Gregory. New York: Cambridge University Press. King, J. E. 2005. “Bertrand Russell on Economics, 1889-1918.” Russell n.s. 25 (1 Summer): 5–38. Madigan, Timothy J. 2010. “Six Degrees of Bertrand Russell.” Russell n.s. 30 (1 Summer): 63–67. Moorehead, Caroline. 1993. Bertrand Russell: A Life. New York: Viking. Russell, Bertrand. 1924. “How to Read History.” The Bermondsey Book 1 (2 March): 10–13. Russell, Bertrand. 1934. Freedom versus Organization 1814-1914. New York: Norton. Russell, Bertrand. 1947. Philosophy and Politics. New York: Cambridge University Press. Russell, Bertrand. 1951. “Reply to Criticisms.” In The Philosophy of Bertrand Russell. 3rd ed., edited by Paul Arthur Schilpp. New York: Tudor. Russell, Bertrand. 1957. “How to Read and Understand History.” In Understanding History and Other Essays. New York: Philosophical Library. Stone, Peter. 2016. “Russell Needed to Write.” Russell n.s. 36 (1 Summer): 92–94. Stunkel, Kenneth R. 2001–02. “Bertrand Russell’s Writings and Reflections on History.” Russell n.s. 21 (Winter): 129–53. Trevelyan, George Macaulay. 1913. “Clio, A Muse.” In Clio, A Muse, and Other Essays Literary and Pedestrian. New York: Longmans, Green and Co. Willis, Kirk. 1976. “The Critical Reception of German Social Democracy.” Russell 21–22: 35–45. Willis, Kirk. 1987. “Bertrand Russell on History: The Theory and Practice of a Moral Science.” In At the Nexus of Philosophy and History, edited by Bernard P. Dauenhauer. Athens, GA: The University of Georgia Press.

Bibliography Works by Russell Many of Russell’s papers have been published in the Collected Papers of Bertrand Russell. As of this writing Volumes 1 through 15 and Volumes 21 and 28 have been published. The full bibliographic references to these volumes are given in the section on abbreviations.

1  Books by Russell in chronological order This list just includes the first editions. Books which later became chapters of other books or were collected in essays are omitted. 1896. German Social Democracy. London: Longmans, Green. 1897. An Essay on the Foundation of Geometry. Cambridge: Cambridge University Press. 1900. A Critical Exposition of the Philosophy of Leibniz. Cambridge: Cambridge University Press. 1903. The Principles of Mathematics. Cambridge: Cambridge University Press. 1910–13. (with A. N. Whitehead) Principia Mathematica. Cambridge: Cambridge University Press. 3 vols. 1912. The Problems of Philosophy. London: Williams and Norgate. 1914. Our Knolwedge of the External World as a Field for Scientific Method in Philosophy. Chicago: Open Court. 1916. Principles of Social Reconstruction. London: Allen and Unwin. 1917. Political Ideals. New York: Century. 1918. Roads to Freedom. London: Allen and Unwin. 1919. Introduction to Mathematical Philosophy. London: Allen and Unwin. 1920. The Practice and Theory of Bolshevism. London: Allen and Unwin. 1921. The Analysis of Mind. London: Allen and Unwin. 1922. The Problem of China. London: Allen and Unwin. 1923. (with Dora Russell) The Prospects of Industrial Civilization. London: Allen and Unwin. 1923. The ABC of Atoms. London: Kegan Paul. 1924. Icarus, or the Future of Science. London: Kegan Paul. 1925. The ABC of Relativity. London: Kegan Paul. 1926. On Education, Especially in Early Childhood. London: Allen and Unwin.

388

Bibliography

1927. The Analysis of Matter. London: Allen and Unwin. 1927. An Outline of Philosophy. London: Allen and Unwin. 1929. Marriage and Morals. London: Allen and Unwin. 1930. The Conquest of Happiness. London: Allen and Unwin. 1931. The Scientific Outlook. London: Allen and Unwin. 1932. Education and the Social Order. London: Allen and Unwin. 1934. Freedom and Organization, 1814-1914. London: Allen and Unwin. 1935. Religion and Science. London: Thornton Butterworth. 1936. Which Way to Peace? London: Michael Joseph. 1937. (with Patricia Russell) The Amberly Papers. New York: Norton. 2 vols. 1938. Power: A New Social Analysis. London: Allen and Unwin. 1940. An Inquiry into Meaning and Truth. New York: Norton. 1945. A History of Western Philosophy. New York: Simon and Schuster. 1948. Human Knowledge: Its Scope and Limits. London: Allen and Unwin. 1949. Authority and the Individual. London: Allen and Unwin. 1951. The Impact of Science on Society. New York: Columbia University Press. 1952. New Hopes for a Changing World. London: Allen and Unwin. 1954. Human Society in Ethics and Politics. London: Allen and Unwin. 1959. Common Sense and Nuclear Warfare. London: Allen and Unwin. 1959. My Philosophical Development. London: Allen and Unwin. 1959. Wisdom of the West. New York: Doubleday. 1960. Bertrand Russell Speaks His Mind. New York: World Publishing. 1961. Has Man a Future? London: Allen and Unwin. 1963. Unarmed Victory. London: Allen and Unwin. 1967–69. The Autobiography of Bertrand Russell. London: Allen and Unwin. 3 vols.

2  Collections of essays and stories 1910. Philosophical Essays. London: Longmans, Green. 1918. Mysticism and Logic. London: Longmans, Green. 1927. Selected Papers of Bertrand Russell. New York: Modern Library. 1928. Skeptical Essays. London: Allen and Unwin. 1935. In Praise of Idleness and Other Essays. London: Allen and Unwin. 1941. Let People Think. London: Watts. 1950. Unpopular Essays. London: Allen and Unwin. 1953. Satan in the Suburbs and Other Stories. London: Bodley Head. 1954. Nightmares of Eminent Persons. London: Bodley Head. 1956. Portraits from Memory and Other Essays. London: Allen and Unwin. 1956. Logic and Knowledge (Robert Marsh, ed.) London: Allen and Unwin. 1957. Why I Am Not a Christian (Paul Edwards, ed.) London: Allen and Unwin.

Bibliography

389

1961. The Basic Writings of Bertrand Russell (Robert Egner and Lester Denonn, eds.) London: Allen and Unwin. 1961. Fact and Fiction. London: Allen and Unwin. 1967. War Crimes in Vietnam. London: Allen and Unwin. 1972. The Collected Stories of Bertrand Russell (Barry Feinberg, ed.) London: Allen and Unwin. 1973–83. Bertrand Russell’s America (Barry Feinberg and Ronald Kasrils, eds.) London: Allen and Unwin. 2 vols. 1973. Essays in Analysis (Douglas Lackey, ed.) London: Allen and Unwin. 1975. Mortals and Others (Harry Ruja, ed.). 1986. Bertrand Russell on God and Religion (Al Seckel, ed.) Buffalo: Prometheus. 1987. Bertrand Russell on Ethics, Sex and Marriage (Al Seckel, ed.) Buffalo: Prometheus. 1999. Russell on Religion (Louis Greenspan and Stephan Andersson, eds.) London: Routledge. 1999. Russell on Ethics (Charles Pigden, ed.) London: Routledge. 2003. Russell on Metaphysics (Stephen Mumford, ed.) London: Routledge.

3  Collections of letters 1969. Dear Bertrand Russell (Barry Feinberg and Ronald Kasrils, eds.) London: Allen and Unwin. 1977. Grattan-Guinness, Ivor. Dear Russell, Dear Jourdain: A Commentary on Russell’s Logic. Based on His Correspondence with Philip Jourdain. New York: Columbia. 1992–2001. The Selected Letters of Bertrand Russell (Nicholas Griffin, ed.) Vol 1: The Private Years 1884–1914. London: Penguin; Vol. 2, The Public Years 1914–70. London: Routledge. 2001. Correspondance sur la philosophie, la logique et la politique avec Louis Couturat (1897-1913) (Anne-Françoise Schmid, ed.) Paris: Kimé. 2 vols. 2002. Yours Faithfully, Bertrand Russell (Ray Perkins, Jr., ed.) Chicago: Open Court.

4  Papers by Russell Russell wrote over 2000 articles. A full bibliography of Russell’s articles is given in the following volumes: Blackwell, Kenneth and Henry Ruja. A Bibliography of Bertrand Russell. 3 vols. London: Routledge, 1994. Following is a list of his most well-known articles including all of the works mentioned in the essays in this Companion. Most of these have been reprinted in the Collected

390

Bibliography

Papers. They are ordered by date either of publication or (if not published prior to inclusion in the Collected Papers) the date they were written. 1901. “Sur la logique des relations avec des applications à la théorie des series.” [“The Logic of Relations with Some Applications to the Theory of Series.”] Revue de mathématiques, 7. Papers 3, LK. 1903. “The Free Man’s Worship.” The Independent Review. Papers 12, ML, BW, WNC. 1904. “Meinong’s Theory of Complexes and Assumptions.” Mind 13. Papers 3, EA. 1905. “On Denoting.” Mind 14. Papers 3, LK, EA. 1906. “On Some Difficulties in the Theory of Transfinite Numbers and Order Types.” Proceedings of the London Mathematical Society 4. Papers 5, EA. 1906. “Les Paradoxes de la Logique.” [“On ‘Insolubilia’ and their Solution by Symbolic Logic.”] Revue de métaphysique et de morale, 14. Papers 5, EA. 1906. “On the Substitution Theory of Classes and Relations.” Papers 5, EA. 1906. “The Paradox of the Liar.” Papers 5. 1907. “The Regressive Method of Discovering the Premises of Mathematics.” Papers 5, EA. 1908. “Transatlantic ‘Truth’.” [also titled “William James’s Conception of Truth”] Albany Review n.s.2, Papers 5, PE. 1908. “Mathematical Logic as Based on the Theory of Types.” American Journal of Mathematics 30. Papers 5, LK. 1909. “Pragmatism.” The Edinburgh Review 209. Papers 6, PE. 1910. “La Théorie des Types Logiques.” [“The Theory of Logical Types.”] Revue de métaphysique et de morale, 18. Papers 6, EA. 1910. “On the Nature of Truth and Falsehood.” PE, Papers 6. 1911. “Le Réalisme analytique.” [“Analytic Realism.”] Bulletin de la société française de philosophie. Papers 6. 1911. “Knowledge by Acquaintance and Knowledge by Description.” PAS 11. Papers 6, ML. 1912. “On the Relation of Universals to Particulars.” PAS 12, Papers 6, LK. 1912. “The Philosophy of Bergson.” Monist 22, Papers 6. 1912. “The Essence of Religion.” The Hibbert Journal 11, Papers 12. 1912. “On Matter.” Papers 6. 1913. On the Notion of Cause. PAS 13. Papers 6, ML. 1914. “The Relation of Sense-Data to Physics.” Scientia 16. Papers 8, ML. 1914. “Mysticism and Logic.” The Hibbert Journal 12. Papers 8, ML. 1914. “On Scientific Method in Philosophy.” (originally a pamphlet) Papers 8, ML. 1915. “The Ultimate Constituents of Matter.” Monist 25. Papers 8, ML. 1918. “The Philosophy of Logical Atomism.” Monist 28 and 29. Papers 8, LK. 1919. “Professor Dewey’s Essays in Experimental Logic.” Journal of Philosophy, Psychology and Scientific Methods 16. Papers 8. 1919. “On Propositions: What They Are and How They Mean.” PAS Supplementary Volume 2. Papers 8, LK.

Bibliography

391

1922. “Is There an Absolute Good?” Papers 9. 1924. “Logical Atomism.” (Muirhead, ed.) Contemporary British Philosophy. Papers 9. LK. 1925. “What I Believe.” (originally a booklet) BW, WNC, to appear in Papers 17. 1927. “Why I am not A Christian.” (originally a pamphlet.) Papers 10, BW, WNC. 1931. “Review of Ramsey, Foundations of Mathematics.” Mind 40, Papers 10. 1934. “The Philosophy of Communism.” Papers 10. 1936. “The Limits of Empiricism.” PAS 36. Papers 10. 1938. “On Verification.” PAS 38. Papers 10. 1939. “Dewey’s New Logic.” (Schilpp, ed.) The Philosophy of John Dewey, Papers 10. 1944. “My Mental Development.” (Schilpp, ed.) The Philosophy of Bertrand Russell. Papers 11, BW. 1944. “Reply to Criticism.” Schilpp, ed. The Philosophy of Bertrand Russell. Papers 11. 1945. “Logical Positivism.” Polemic 1. Papers 11. 1950. “Is Mathematics Purely Linguistic?” Papers 11, EA.

5  Secondary Sources 1  Monographs on Russell Prior to 2000 Aiken, Lillian W. Bertrand Russell’s Philosophy of Morals. New York: Humanities Press, 1963. Ayer, Alfred J. Russell. London: Fontana, 1972. Blackwell, Kenneth. The Spinozistic Ethics of Bertrand Russell. London: Allen and Unwin, 1985. De Almeida, Claudio. Russell on the Foundations of Logic. Porto Alegre: PUCS, 1998. De Rouilhan, Philippe. Russell et le cercle des paradoxes. Paris: PUF, 1996. Eames, Elizabeth R. Bertrand Russell’s Theory of Knowledge. London: Allen and Unwin, 1969. Eames, Elizabeth R. Bertrand Russell’s Dialogue with his Contemporaries. Carbondale: Southern Illinois University Press, 1989. Fritz, Charles Andrew. Dr. Bertrand Russell’s Construction of the External World. London: Routledge and Kegan Paul, 1952. Garciadiego, Alejandro. Bertrand Russell and the Origins of the Set-Theoretic “Paradoxes.” Basel: Birkhauser Verlag, 1992. Grayling, Anthony C. Bertrand Russell. London: Oxford University Press, 1996. Greenspan, Louis. The Incompatible Prophecies: An Essay on Science and Liberty in the Political Writings of Bertrand Russell. Oakville, ON: Mosaic, 1978. Griffin, Nicholas. Russell’s Idealist Apprenticeship. Oxford: Clarendon, 1991.

392

Bibliography

Hager, Paul J. Continuity and Change in the Development of Russell’s Philosophy. Dorfrecht: Kluwer, 1994. Hardy, Godfrey H. Bertrand Russell and Trinity. Cambridge: Cambridge University Press, 1970. Hylton, Peter W. Russell, Idealism, and the Emergence of Analytical Philosophy. Oxford: Oxford University Press, 1990. Ironside, Philip. The Social and Political Thought of Bertrand Russell. London: Cambridge University Press, 1996. Jager, Ronald. The Development of Bertrand Russell’s Philosophy. London: Allen and Unwin, 1972. Kilminster, Clive W. Russell. Brighton: Harvester, 1984. Landini, Gregory. Russell’s Hidden Substitution Theory. Oxford: Oxford University Press, 1998. Linsky, Bernard. Russell’s Metaphysical Logic. Stanford: SSLI, 1999. Pears, David F. Bertrand Russell and the British Tradition in Philosophy. London: Collins, 1967. Rodrigues-Consuerga, Francisco. The Mathematical Philosophy of Bertrand Russell: Origins and Development. Basel: Birkhauser, 1991. Ryan, Alan. Bertrand Russell: A Political Life. New York: Hill and Wang, 1988. Sainsbury, R. Mark Russell. London: Routledge and Kegan Paul, 1979. Slater, John. Bertrand Russell. Bristol: Thoemmes, 1994. Vernant, Denis. La philosophie mathématique de Russell. Paris: Vrin, 1993. Vuillemin, Jules. Leçons sur la première philosophie de Russell. Paris: Armand Colin, 1968. Watling, J. Bertrand Russell. Edinburgh: Oliver and Boyd, 1970.

Since 2000 Bostock, David. Russell’s Logical Atomism. Oxford: Oxford University Press, 2012. Carey, Rosalind. Russell and Wittgenstein on the Nature of Judgement. London: Continuum, 2007. Denton, Peter. The ABC of Armageddon: Bertrand Russell on Science, Religion, and the Next War, 1919-1938. Albany: SUNY. 2001. Galaugher, Jolen. Russell’s Philosophy of Logical Analysis 1897-1905. Basingstoke: Palgrave Macmillan, 2013. Gandon, Sébastien. Russell’s Unknown Logicism. Basingstoke: Palgrave Macmillan, 2012. Hylton, Peter. Propositions, Functions, and Analysis: Selected Essays on Russell’s Philosophy. Oxford: Clarendon Press, 2005. Korhonen, Anssi. Logic as Universal Science: Russell’s Early Logicism and Its Philosophical Context. Basingstoke: Palgrave Macmillan, 2013. Landini, Gregory. Russell. Oxford: Routledge, 2011. Landini, Gregory. Wittgenstein’s Apprenticeship with Russell. Cambridge: Cambridge University Press, 2007.

Bibliography

393

Linsky, Bernard. The Evolution of Principia Mathematica. Cambridge: Cambridge University Press, 2011. Maclean, Gülberk Koç. Bertrand Russell’s Bundle Theory of Particulars. London: Bloomsbury, 2014. Miah, Sajahan. Russell’s Theory of Perception (1905-1919). London: Continuum, 2006. Ongley, John, and Rosalind Carey. Russell: A Guide for the Perplexed. London: Bloomsbury, 2013. Potter, Michael K. Bertrand Russell’s Ethics. London: Continuum, 2006. Stevens, Graham. The Russellian Origins of Analytic Philosophy. Oxford: Routledge, 2005. Steven, Graham. The Theory of Descriptions: Russell and the Philosophy of Language. Basingstoke: Palgrave MacMillan, 2011. Vernant, Denis. Bertrand Russell. Paris: Flammarion, 2003. Weidlich, Thom. Appointment Denied: The Inquisition of Bertrand Russell. Amherst, NY: Prometheus Books, 2000.

2  Other books including work on Russell’s philosophy Before 2000 Ayer, Alfred J. Russell and Moore: The Analytic Heritage. Cambridge, MA: Harvard University Press, 1971. Cocchiarella, Nino B. Logical Studies in Early Analytic Philosophy. Columbus: Ohio State University Press, 1987. Coffa, Alberto. The Semantic Tradition from Kant to Carnap. Cambridge: Cambridge University Press, 1991. Dejnozka, Jan. The Ontology of the Analytic Tradition and its Origins: Realism and Identity in Frege, Russell, Wittgenstein and Quine. Lanham, MD: Littlefield Adams, 1996. Evans, Gareth. The Varieties of Reference. New York: Oxford University Press, 1982. Hochberg, Herbert. Thought, Fact, and Reference. Minneapolis: University of Minnesota Press, 1978. Linsky, Leonard. Referring. London: Routledge, 1967. Neale, Stephen. Descriptions. Cambridge, MA: MIT Press, 1990. Ostertag, Gary (ed.), Definite Descriptions: A Reader. Cambridge, MA: MIT Press, 1998. Tait, William (ed.), Early Analytic Philosophy. Chicago: Open Court, 1997. Urmson, J. O. Philosophical Analysis. Oxford: Oxford University Press, 1956.

Since 2000 Bonino, Guido. The Arrow and the Point: Russell and Wittgenstein’s Tractatus. Frankfurt: Ontos Verlag, 2008. Candlish, Stewart. The Russell/Bradley Dispute and its Significance for the Twentieth Century Philosophy. Basingstoke: Palgrave Macmillan, 2007.

394

Bibliography

Cocchiarella, Nino B. Formal Ontology and Conceptual Realism. Dordrecht: Springer, 2007. Potter, Michael. Reason’s Nearest Kin. Oxford: Oxford University Press, 2000. Potter, Michael. Wittgenstein’s Notes on Logic. Oxford: Oxford University Press, 2009. Sackur, Jérôme. Formes et faits: Analyse et théorie de la connaissance dans l’atomisme logique. Paris: Vrin, 2005. Soames, Scott. The Analytic Tradition in Philosophy, Vol. 1, The Founding Giants. Princeton: Princeton University Press, 2014. Soames, Scott. Philosophical Analysis in the Twentieth Century: Vol. 1: The Dawn of Analysis. Princeton: Princeton University Press, 2003. Sullivan, Arthur. Logicism and the Philosophy of Language: Selections from Frege and Russell. Peterborough: Broadview Press, 2003.

3  Biographies of Russell Clark, Ronald. The Life of Bertrand Russell. London: Jonathan Cape, 1975. Monk, Ray. Bertrand Russell 1872-1921: The Spirit of Solitude. London: Jonathan Cape, 1996. Monk, Ray. Bertrand Russell 1921-1970: The Ghost of Madness. London: Jonathan Cape, 2000. Moorehead, Caroline. Bertrand Russell. New York: Viking, 1992. Wood, Alan. Bertrand Russell: The Passionate Sceptic. London: Allen and Unwin, 1957.

4  Anthologies of articles on Russell Before 2000 Clementz, François, and Anne-Françoise Schmid (eds.), Bertrand Russell, de la logique à la politique. Hermès 7, 1990. Devaux, P. (ed.), Bertrand Russell 1872-1972. Revue Internationale de Philosophie, 26, 1972. Hintikka, J. (ed.), Bertrand Russell’s Early Philosophy Parts 1 and 2. Synthese 45, 6 and 46, 2, 1980–81. Irvine, Andrew D., and G. A. Wedekind (eds.), Russell and Analytic Philosophy. Toronto: University of Toronto Press, 1993. Klemke, E. D. (ed.), Essays on Bertrand Russell. Urbana: University of Illinois Press, 1970. Monk, Ray, and Anthony Palmer (eds.), Bertrand Russell and the Origins of Analytic Philosophy. Bristol: Theommes Press, 1996. Moran, Margaret, and Carl Spadoni, (eds.), Intellect and Social Conscience: Essays on Bertrand Russell’s Early Work. Hamilton: McMaster University Press, 1984. Nakhnikian, George (ed.), Bertrand Russell’s Philosophy. London: Duckworth, 1974.

Bibliography

395

Pears, David (ed.), Bertrand Russell: A Collection of Critical Essays. New York: Doubleday, 1972. Roberts, George W. (ed.), Bertrand Russell Memorial Volume. London: Allen and Unwin, 1979. Savage, C. Wade, and C. Anthony Anderson (eds.), Rereading Russell: Essays on Bertrand Russell’s Metaphysics and Epistemology. Minneapolis: University of Minnesota Press, 1989. Schilpp, P. A. (ed.), The Philosophy of Bertrand Russell. Chicago: Northwestern University Press, 1944. Schoenman, Ralph (ed.), Bertrand Russell: Philosopher of the Century. London: Allen and Unwin, 1967. Thomas, J. E., and Kenneth Blackwell (eds.), Russell in Review. Toronto: Samuel Stevens, Hakkert and Co., 1976. Winchester, Ian, and Kenneth Blackwell (eds.), Antinomies and Paradoxes: Studies in Russell’s Early Philosophy. Hamilton: McMaster University Press, 1988 and Russell 8, 1988.

Since 2000 De Rouilhan, Phillippe (ed.), Russell en héritage/Le centenaire des Principles. Revue Internationale de Philosophie, 58, 2004. Griffin, Nicholas (ed.), The Cambridge Companion to Bertrand Russell. Cambridge: Cambridge University Press, 2003. Griffin, Nicholas, and Bernard Linsky (eds.), The Palgrave Centenary Companion to Principia Mathematica. Basingstoke: Palgrave MacMillan, 2013. Griffin, Nicholas, Bernard Linsky, and Kenneth Blackwell (eds.), After “On Denoting”: Themes from Russell and Meinong. Hamilton: McMaster University Press, 2007 and Russell 27, 2007. Griffin, Nicholas, Bernard Linsky, and Kenneth Blackwell (eds.), Principia Mathematica at 100. Hamilton: McMaster University Press, 2011 and Russell 31, 2011. Guay, Alexandre (ed.), Autour des Principia Mathematica de Russell et Whitehead. Dijon: Editions Universitaires de Dijon, 2012. Hare, Anthony (ed.), Bertrand Russell and Critical Thinking. Inquiry: Critical Thinking Across the Disciplines, 20, 2001 Jacquette, Dale, and Nicholas Griffin (eds.), Russell vs Meinong: the Legacy of “On Denoting,” London: Routledge, 2009. Lapointe, Sandra, Matti Ekland, and Amie Thomasson (eds.), The Baltic International Yearbook of Cognition, Logic and Communication, 4, 2008. [Most of the articles] Link, Godehard (ed.), One Hundred Years of Russell’s Paradox. Berlin: de Gruyter, 2004. Marqueze, J. Rodriguez (ed.), Commemoration of the Centenary of the Publication of Russell’s “On Denoting.” Teorema 24, 2005. Neale, Steven (ed.), 100 Years of “On Denoting.” Mind 114, 2005.

396

Bibliography

Principia: An International Journal of Epistemology (Brazil) 5, 2001. Schwerin, Alan (ed.), Bertrand Russell on Nuclear War, Peace, and Language: Critical and Historical Essays. Westport: Praeger, 2002. Wishon, Donovan, and Bernard Linsky (eds.), Acquaintance, Knowledge and Logic. Stanford: CSLI, 2015.

5 Journals The following journals are devoted to articles on Bertrand Russell: Russell: The Journal of Bertrand Russell Studies The Bertrand Russell Society Quarterly

6  Articles on Russell (Selections) Note: I have not listed separately the articles included in the journals listed above, or any of the above collections.

Before 2000 Almog, Joseph. “Nothing, Something, Infinity.” Journal of Philosophy, 96, 1996. Anderson, C. Anthony. “Russellian Intensional Logic.” In Almog, Joseph, John Perry, and Howard K. Wettstein (eds.), Themes from Kaplan. New York: Oxford University Press, 1989. Anderson, C. Anthony. “Some Difficulties Concerning Russellian Intentional Logic.” Nous, 20, 1986. Austin, James. “Russell’s Cryptic Response to Strawson.” Philosophy and Phenomenological Research, 38, 1978. Avery, Albert. “Russell’s Quest for Objectivity.” Philosophy and Phenomenological Research, 2, 1942. Ayer, A. J. “Bertrand Russell on Meaning and Truth.” Nature, 148, 1941. Bach, Kent. “Ramachandran versus Russell.” Analysis, 54, 1994. Bacon, John. “An Alternative Contextual Definition for Descriptions.” Philosophical Studies, 16, 1965. Bar-Elli, Gilead. “Constituents and Denotation in Russell.” Theoria, 46, 1980. Beck, Lewis. “Constructions and Inferred Entities.” Philosophy of Science, 17, 1950. Bell, David. “The Revolution of Moore and Russell: A Very British Coup?” In O’Hear (ed.), German Philosophy Since Kant. New York: Cambridge University Press, 1999. Bell, John L., and William Demopoulos. “Elementary Propositions and Independence.” Notre Dame Journal of Formal Logic, 37, 1996. Bergmann, Gustav. “The Revolt Against Logical Atomism.” Philosophical Quarterly, 7, 1957. Part II Philosophical Quarterly, 8, 1958. Bergmann, Gustav. “Russell on Particulars.” Philosophical Review, 56, 1947.

Bibliography

397

Berka, K. “Russell’s Theory of Quantity and Magnitude.” Teorie a Metoda, 2, 1970. Beyer, Christian. “Russell’s Principle Considered from Both a Neo-Fregean and a Husserlian Viewpoint.” Acta Analytica, 20, 1998. Blackburn, Simon and Alan Code. “The Power of Russell’s Criticism of Frege.” Analysis, 38, 1978. Blackburn, Simon and Alan Code. “Reply to Geach.” Analysis, 38, 1978. Blackman, Larry. “Russell on the Relations of Universals and Particulars.” Philosophy Research Archives, 9, 1983. Blackwell, Kenneth. “The Early Wittgenstein and the Middle Russell.” In Block, Irving (ed.), Perspectives on the Philosophy of Wittgenstein. Oxford: Blackwell, 1981. Blanshard, Brand. “Bertrand Russell in Retrospect.” Dialogue, 7, 1969. Boër, Steven. “Russell on Classes as Logical Fictions.” Analysis, 33, 1973. Boolos, George. “The Advantages of Honest Toil over Theft.” In Alexander, G. (ed.), Mathematics and Mind. Oxford: Oxford University Press, 1993. Bradie, Michael. “The Development of Russell’s Structural Postulates.” Philosophy of Science, 44, 1977. Bradley, Michael. “Russell and the Identity of Indiscernibles.” History of Philosophy Quarterly, 3, 1986. Braithwaite, Richard. “Bertrand Russell as Philosopher of Science.” British Journal for the Philosophy of Science, 21, 1970. Braun, David. “Understanding Belief Reports.” Philosophical Review, 107, 1998. Brinton, Alan. “Uses of Definite Descriptions and Russell’s Theory.” Philosophical Studies, 31, 1977. Broad, C. D. “Bertrand Russell, as Philosopher.” Bulletin of the London Mathematical Society, 5, 1973. Broad, C. D. “Is There ‘Knowledge by Acquaintance’?” Aristotelian Society Supplementary, 2, 1919. Burge, Tyler. “Russell’s Problem and Intentional Identity.” In Tomberland, James (ed.), Agent, Language, and the Structure of the World. Indianapolis: Hackett, 1983. Burke, Tom. “Dewey and Russell on the Possibility of Immediate Knowledge.” Studies in Philosophy and Education, 17, 1998. Butchvarov, Panayot. “Russell’s Views on Reality.” Grazer Philosophische Studien, 32, 1988. Butler, Ronald. “The Scaffolding of Russell’s Theory of Descriptions.” Philosophical Review, 63, 1954. Cargile, James. “The Problem of Induction.” Philosophy, 10, 1998. Carney, James. “Russell’s ‘Proof,’ Again.” Canadian Journal of Philosophy, 10, 1980. Carney, James, and Gregory Fitch. “Can Russell Avoid Frege’s Sense?” Mind, 88, 1979. Carruthers, Peter. “Russellian Thoughts.” Mind, 96, 1987. Cassin, Chrystine. “Existential Quantification in Russell’s Analysis of Definite Descriptions.” Mind, 80, 1971.

398

Bibliography

Casullo, Albert. “The Contingent Identity of Particulars and Universals.” Mind, 93, 1984. Casullo, Albert. “Particulars, Substrata, and the Identity of Indiscernables.” Philosophy of Science, 49, 1984 Casullo, Albert. “Russell on the Reduction of Particulars.” Analysis, 41, 1981. Chihara, Charles S. “Ramsey’s Theory of Types: Suggestion for a Return to Fregean Sources.” In Mellor, D. H. (ed.), Prospects for Pragmatism. Cambridge: Cambridge University Press, 1980. Church, Alonzo. “A Comparison of Russell’s Resolution of the Semantical Antinomies with That of Tarski.” Journal of Symbolic Logic, 41, 1976. Church, Alonzo. “Russell’s Theory of Identity of Propositions.” Philosophia Naturalis, 21, 1984. Cocchiarella, Nino B. “Frege, Russell and Logicism: A Logical Reconstruction.” In Haaparanta, L. (ed.), Frege Synthesized. Dordrecht: Reidel, 1986. Cocchiarella, Nino B. “Meinong Reconstructed Versus Early Russell Reconstructed.” Journal of Philosophical Logic, 11, 1982. Crody, W. Stephen. “Russell on the Meaning of Descriptions.” Notre Dame Journal of Formal Logic, 17, 1976. Dau, Paulo. “The Complex Matter of Denoting.” Analysis, 45, 1985. Dau, Paulo. “Russell’s First Theory of Denoting and Quantification.” Notre Dame Journal of Formal Logic, 27, 1986. Davant, James B. “Wittgenstein on Russell’s Theory of Types.” Notre Dame Journal of Formal Logic, 16, 1975. Dejnozka, Jan. “The Ontological Foundations of Russell’s Theory of Modality.” Erkenntnis, 32, 1990. Demopoulos, William. “Critical Notice: Russell’s Analysis of Matter.” Philosophy of Science, 52, 1985. Demopoulos, William. “On the Theory of Meaning of ‘On Denoting.’” Nous, 33, 1999. De Rouilhan, Philippe. “Russell and the Vicious Circle Principle.” Philosophical Studies, 65, 1992. Donagan, Alan. “Recent Criticisms of Russell’s Analysis of Existence.” Analysis, 12, 1952. Donnellan, Keith S. “Genuine Names and Knowledge by Acquaintance.” Dialectica, 44, 1990. Donnellan, Keith S. “Reference and Definite Descriptions.” Philosophical Review, 75, 1966. Donnellan, Keith S. “Substitution and Reference.” Journal of Philosophy, 63, 1966. Dukelow, Owen. “The Problem of Negative Facts in Russell’s Logical Atomism.” Southwestern Journal of Philosophy, 7, 1976. Dukran, Jane. “Russell on Names.” Philosophy Research Archives, 13, 1987. Eames, Elizabeth R. “The Consistency of Russell’s Realism.” Philosophy and Phenomenological Research, 27, 1967.

Bibliography

399

Eames, Elizabeth R. “Bertrand Russell’s Philosophical Method.” Midwestern Journal of Philosophy, 4, 1976. Eames, Elizabeth R. “Russell and the Experience of Time.” Philosophy and Phenomenological Research, 46, 1986. Edgell, Beatrice. “Is There ‘Knowledge by Acquaintance’?” Aristotelian Society Supplementary, 2, 1919. Edwards, Paul. “Bertrand Russell’s Doubts about Induction.” Mind, 58, 1949. Elugardo, Rinaldo. “Descriptions, Indexicals, and Speaker Meaning.” Protosociology, 10, 1997. Feser, Edward. “Can Phenomenal Qualities Exist Unperceived?” Journal of Consciousness Studies, 5, 1998. Findley, J. N. “Is There Knowledge by Acquaintance?” Aristotelian Society Supplementary, 23, 1949. Fitch, Frederic. “Propositions as the Only Realities.” American Philosophical Quarterly, 8, 1971. Fleischhaker, L. E. “Is Russell’s Vicious Circle Principle False or Meaningless?” Dialectica, 33, 1979. Fogelin, Robert. “Negative Elementary Propositions.” Philosophical Studies, 25, 1974. Fortier, Evelyn. “Was the Dispute Between Russell and Bradley about Internal Relations?” In Mander, W. J. (ed.), Perspectives on the Logic and Metaphysics of F. H. Bradley. Bristol: Thoemmes, 1996. Ganeri, Jonardon. “Contextually Incomplete Descriptions – A New Counterexample to Russell?” Analysis, 55, 1995. Garciadiego, Alejandro R. “The Principle of Mathematics of Bertrand Russell.” In Ramirez, Santiago, and Robert S. Cohen (eds.), Mexican Studies in the History and Philosophy of Science. Dordrecht: Kluwer, 1995. Gaskin, Richard. “Fregean Sense and Russellian Propositions.” Philosophical Studies, 86, 1997. Gastwirth, Paul. “The Hypothesis of Reducibility.” Monist, 42, 1932. Geach, P. T. “Russell on Denoting.” Analysis, 38, 1970. Geach, P. T. “Russell’s Theory of Descriptions.” Analysis, 10, 1950. Geach, P. T. “Two Paradoxes of Russell’s.” Journal of Philosophy, 67, 1970. Geirsson, Heimir. “True Belief Reports and Sharing of Beliefs.” Journal of Philosophical Research, 23, 1998. Gerrard, Steve. “Desire and Desirability: Bradley, Russell, and Moore versus Mill.” In Tait, William W. (ed.), Early Analytic Philosophy: Frege, Russell, Wittgenstein. Chicago: Open Court, 1997. Giaretta, Pierdaniele. “Analysis and Logical Form in Russell: the 1913 Paradigm.” Dialectica, 51, 1997. Goldfarb, Warren D. “Poincaré against the Logicists.” In William, Asprey, and Philip Kitcher (eds.), Essays in the History of Mathematics. Minneapolis: University of Minnesota Press, 1987.

400

Bibliography

Gorman, Jonathan. “On Hedgehogs and Foxes.” Philosophical Inquiry, 21, 1999. Graham, George. “Russell’s Deceptive Desires.” Philosophical Quarterly, 36, 1986. Grandy, Richard E. “Forms of Belief.” Synthese, 46, 1981. Grattan-Guinness, Ivor. “Bertrand Russell on His Paradox and the Multiplicative Axiom.” Journal of Philosophical Logic, 1, 1972. Grattan-Guinness, Ivor. “Georg Cantor’s Influence on Bertrand Russell.” History and Philosophy of Logic, 1, 1980. Grattan-Guinness, Ivor. “How Bertrand Russell Discovered His Paradox.” Historia Mathematica, 5, 1978. Grattan-Guinness, Ivor. “Notes on the Fate of Logicism from Principia Mathematica to Gödel’s Incompletability Theorem.” History and Philosophy of Logic, 1, 1980. Grattan-Guinness, Ivor. “On the Development of Logics between the Two World Wars.” American Mathematical Monthly, 88, 1981. Grelling, Kurt. “Realism and Logic: An Investigation of Russell’s Metaphysics.” Monist, 39, 1929. Griffin, Nicholas. “Did Russell’s Criticisms of Bradley’s Theory of Relations Miss Their Mark?” In Stock Guy (ed.), Appearance versus Reality: New Essays on Bradley’s Metaphysics. Oxford: Clarendon Press, 1998. Griffin, Nicholas. “Russell’s Critique of Meinong’s Theory of Objects.” Grazer Philosophische Studien, 25–26, 1985. Griffin, Nicholas. “Russell’s Multiple Relation Theory of Judgment.” Philosophical Studies, 47, 1985. Griffin, Nicholas. “Why did Russell think ‘E!a’ Was Meaningless?” In Stewart, Donald (ed.), Entities and Individuation. Studies in Ontology and Language in Honour of Neil Wilson. Queenston: Edwin Mellen, 1989. Griffin, Nicholas, and Gad Zak. “Russell on Specific and Universal Relations.” History and Philosophy of Logic, 3, 1982. Grossman, Reinhardt. “Definite Descriptions.” Philosophical Studies, 27, 1975 Grossman, Reinhardt. “Russell’s Paradox and Complete Properties.” Nous, 6, 1972. Gyekye, Kwame. “An Examination of the Bundle: Theory of Substance.” Philosophy and Phenomenological Research, 34, 1973. Hailperin, Theodore, and Hughes Leblanc. “Nondesignating Singular Terms.” Philosophical Review, 68, 1959. Halbasch, Keith. “A Critical Examination of Russell’s View of Facts.” Nous, 5, 1971. Hambourger, Robert. “A Difficulty with the Frege-Russell Definition of Number.” Journal of Philosophy, 74, 1977. Hampshire, Stuart. “Russell, Radicalism, and Reason.” In Held, Virginia, Nielsen, Kai, and Charles Parson (eds.), Philosophy and Political Action. New York: Oxford University Press, 1972. Hark, Michael Ter. “Cognitive Science, Propositional Attitudes, and the Debate between Russell and Wittgenstein.” In Meggle, Georg, and Ulla Wessels (eds.), Analyomen 1:

Bibliography

401

Proceedings of the First Conference: Perspectives in Analytic Philosophy. Berlin:Walter de Bruyter, 1994. Hart, H. L. A. “Is There Knowledge by Acquaintance?” Aristotelian Society Supplementary, 23, 1949. Hart, Wilbur D. “Russell and Ramsey.” Pacific Philosophical Quarterly, 64, 1983. Hartshorne, Charles. “Russell on Causality.” In Hartshorne, Charles (ed.), Beyond Humanism. Lincoln: University of Nebraska Press, 1968. Hausman, David B. “Russell on Negative Facts.” Southwestern Journal of Philosophy, 12, 1974. Hawkins, Benjamin. “Peirce and Russell: The History of a Neglected Controversy.” In Houser, Nathan, Don D. Roberts, and James Van Evra (eds.), Studies in the Logic of Charles Sanders Peirce. Bloomington: Indiana University Press, 1962. Hawkins, Denis John Bernard. “Moore, Russell and Sense Data.” In Hawkins, Denis John Bernard (ed.), Crucial Problems of Modern Philosophy. Notre Dame: University of Notre Dame Press, 1962. Hay, William H. “Bertrand Russell on the Justification of Induction.” Philosophy of Science, 17, 1950. Hayner, Paul. “Knowledge by Acquaintance.” Philosophy and Phenomenological Research, 29, 1969. Heath, A. E. “Logical Atomism and the Law of Parsimony.” Monist, 30, 1920. Hellman, Geoffrey. “How to Gödel A Frege-Russell: Gödel’s Incompleteness Theorems and Logicism.” Nous 15, 1981. Hempel, Carl G. “Russell’s Phenomenological Construction.” Journal of Philosophy, 63, 1966. Henkin, Leon. “Are Mathematics and Logic Identical?” Science, 138 (no. 3542 16 November), 1962. Henkin, Leon. “Completeness in the Theory of Types.” Journal of Symbolic Logic, 14, 1949. Hicks, G. Dawes. “Is There Knowledge by Acquaintance?” Aristotelian Society Supplementary, 2, 1919. Hintikka, Jaakko. “Existential Presuppositions and Existential Commitment.” Journal of Philosophy, 63, 1959. Hintikka, Jaakko. “On Denoting What?” Synthese, 46, 1981. Hintikka, Jaakko, and Jack Kulas. “Russell Vindicated: Towards a General Theory of Definite Descriptions.” Journal of Semantics, 1, 1982. Hiz, H. “Descriptions in Russell’s Theory and in Ontology.” Studia Logica, 36, 1977. Hochberg, Herbert. “Abstracts, Functions, Existence, and Relations in the RussellMeinong Dispute, the Bradley Paradox and the Realism Nominalism.” Grazer Philosophische Studien, 50, 1995. Hochberg, Herbert. “Causal Connections, Universals, and Russell’s HypotheticoScientific Realism.” Monist, 77, 1994.

402

Bibliography

Hochberg, Herbert. “Particulars as Universals: Russell’s Ontological Essay of Particularity and Phenomenological Space-Time.” Journal of Philosophical Research, 20, 1996. Hochberg, Herbert. “Particulars, Universals, and Russell’s Late Ontology.” Journal of Philosophical Research, 21, 1996. Hochberg, Herbert. “Russell, Ramsey, and Wittgenstein on Ramification and Quantification.” Erkenntnis, 27, 1987. Hochberg, Herbert. “Russell’s Attack on Frege’s Theory of Meaning.” Philosophica, 18, 1976. Hochberg, Herbert. “Russell’s Proof of Realism Reproved.” Philosophical Studies, 37, 1980. Hoensbroech, F. Graf. “On Russell’s Paradox.” Mind, 48, 1939. Hoernlé, Reinhold F. A. “The Religious Aspect of Bertrand Russell’s Paradox.” Harvard Theological Review, 9, 1916. Honderich, Ted. “On the Theory of Descriptions.” Proceedings of the Aristotelian Society, 69, 1969. Hooker, C. A. “Definite Descriptions.” Philosophical Studies, 23, 1972. Hope, V. “The Picture Theory of Meaning in the Tractatus as a Development of Moore’s and Russell’s Theories of Judgement.” Philosophy, 44, 1969. Hughes, George E. “Is There Knowledge by Acquaintance?” Aristotelian Society Supplementary, 23, 1949. Hurley, Patrick J. “Russell, Poincaré, and Whitehead’s ‘Relational Theory of Space.’” Process Studies, 9, 1979. Hylton, Peter W. “Logic in Russell’s Logicism.” The Analytic Tradition: Philosophical Quarterly Monographs, Vol. 1. Cambridge: Blackwell, 1990. Hylton, Peter W. “The Nature of the Proposition and the Revolt Against Idealism.” In Rorty, Richard, J. B. Schneewind, and Quentin Skinner (eds.), Philosophy in History. Cambridge: Cambridge University Press, 1984. Hylton, Peter. “The Vicious Circle Principle: Comments on Philippe de Rouilhan.” Philosophical Studies, 65, 1992. Iglesias, M. Teresa. “Russell and Wittgenstein: Two Views of Ordinary Language.” Philosophical Studies (Irl.), 28, 1981. Iglesias, M. Theresa. “Russell’s Theory of Knowledge and Wittgenstein’s Earliest Writings.” Synthese, 60, 1984. Irvine, A. D. “Epistemic Logicism and Russell’s Regressive Method.” Philosophical Studies, 55, 1989. Irvine, William B. “Russell’s Construction of Space from Perspectives.” Synthese, 60, 1984. Jadacki, Jacek Juliusz. “Leon Chwistek-Bertrand Russell’s Scientific Correspondence.” Dialectics and Humanism, 13, 1986. Jeffreys, Harold. “Bertrand Russell on Probability.” Mind, 59, 1950. Jourdain, Philip E. B. “Mr Bertrand Russell’s First Work on the Principles of Mathematics.” Monist, 22, 1912.

Bibliography

403

Judson, Lindsay. “Russell on Memory.” Proceedings of the Aristotelian Society, 88, 1987. Jung, Darryl. “Russell, Presupposition, and the Vicious-Circle Principle.” Notre Dame Journal of Formal Logic, 40, 1999. Kaplan, David. “How to Russell a Frege-Church.” Journal of Philosophy, 72, 1975. Kaplan, David. “The Logic of Demonstratives.” In French, Peter A, Theodore E. Uehling, Jr., and Howard K. Wettstein (eds.), Contemporary Perspectives in the Philosophy of Language. Minneapolis: University of Minnesota Press, 1979. Keane, E.F. “Bertrand Russell and the Emotive Theory.” Indian Journal of Philosophy, 3, 1961. Kemp, Gary. “Propositions and Reasoning in Russell and Frege.” Pacific Philosophical Quarterly, 79, 1998. Kennedy, H. C. “Nine Letters from Giuseppe Peano to Bertrand Russell.” Journal of the History of Philosophy, 13, 1975. Kennedy, H. C. “What Russell Learned from Peano.” Notre Dame Journal of Formal Logic, 14, 1973. Kenyon, Timothy A. “Russell on Pastness.” Dialogue, 33, 1991. Kleinknecht, Reinhard. “Zeitordnung and Zeitpunkte.” Erkenntnis, 53, 2001. Kline, A. David. “Humean Causation and the Necessity of Temporal Discontinuity.” Mind, 94, 1985. Kneale, William C. “Is Existence a Predicate?” Aristotelian Society Supplementary, 5, 1936. Kneale, William C. “Methods of Designation.” Proceedings of the Aristotelian Society, 68, 1968. Koehler, Conrad J. “Studies in Bertrand Russell’s Theory of Knowledge.” Revue Internationale de Philosophie, 26, 1972 Kohl, Marvin. “Bertrand Russell on Vagueness.” Australasian Journal of Philosophy, 47, 1969. Kremer, Michael. “The Argument of ‘On Denoting.’” Philosophical Review, 103, 1994. Kripke, Saul. “Naming and Necessity.” In Davidson, Donald, and Gilbert Harman (eds.), Semantics of Natural Language. Dordrecht: Reidel, 1972. Kultgen, J. H. “Operations and Events in Russell’s Empiricism.” Journal of Philosophy, 53, 1956. Labso, Sam. “Bertrand Russell and the Scientific Spirit.” Philosophy in Science, 7, 1997. Lackey, Douglas P. “Russell’s 1913 Map of the Mind.” Midwest Studies in Philosophy, 6, 1981. Lackey, Douglas P. “Russell’s Unknown Theory of Classes: The Substitutional System of 1906.” Journal of the History of Philosophy, 14, 1963. Lambert, Karel. “Russell’s Theory of Definite Descriptions.” Dialectica, 44, 1990. Lambert, Karel. “Russell’s Version of the Theory of Definite Descriptions.” Philosophical Studies, 65, 1992. Lambert, Karel. “What Is Russell’s Theory of Descriptions? An Addendum.” Pacific Philosophical Quarterly, 65. 1984.

404

Bibliography

Landini, Gregory. “The ‘Definability’ of the Set of Natural Numbers in the1925 Principia Mathematica.” Journal of Philosophical Logic, 25, 1996. Landini, Gregory. “How to Russell Another Meinongian: A Russellian Theory of Fictional Objects versus Zalta’s Theory of Abstract Objects.” Grazer Philosophische Studien, 37, 1990. Landini, Gregory. “Logic in Russell’s Principles of Mathematics.” Notre Dame Journal of Formal Logic, 37, 1996. Landini, Gregory. “A New Interpretation of Russell’s Multiple-Relation Theory of Judgment.” History and Philosophy of Logic, 12, 1991. Landini, Gregory. “Russell’s Intensional Logic of Propositions: A Resurrection of Logicism?” In Orilia, F., and W. J. Rapaport (eds.), Thought, Language, and Ontology. The Hague: Kluwer, 1998. Landini, Gregory. “Russell’s Substitutional Theory of Classes and Relations.” History and Philosophy of Logic, 8, 1987. Lejewski, Czeslaw. “A Re-Examination of the Russellian Theory of Descriptions.” Philosophy, 35, 1960. Leonard, Henry. “The Logic of Existence.” Philosophical Studies, 7, 1956. Levine, James. “Acquaintance, Denoting Concepts, and Sense.” Philosophical Review, 107, 1998. Levine, James. “From Absolute Idealism to ‘The Principles of Mathematics.’” International Journal of Philosophical Studies, 6, 1998. Lewis, Clarence I. “The Issues Concerning Material Implication.” Journal of Philosophy, Psychology, and Scientific Methods, 14, 1917. Lindberg, Jordan J. “From Russell to Quine: Basic Statements, Foundationalism, Truth and Other Myths.” Dialogue, 33, 1990. Linsky, Bernard. “Propositional Functions and Universals in Principia Mathematica.” Australasian Journal of Philosophy, 66, 1988. Linsky, Bernard. “Russell’s Logical Constructions.” Studies in Dialectics of Nature (Beijing) Supplementary, 11, 1995. Linsky, Leonard. “Russell’s ‘No-Classes’ Theory of Classes.” In Thomson, Judith Jarvis (ed.), On Being and Saying. Cambridge, MA: MIT Press, 1987. Linsky, Leonard. “Terms and Propositions in Russell’s Principles of Mathematics.” Journal of the History of Philosophy, 26, 1988. Linsky, Leonard. “The Unity of the Proposition.” Journal of the History of Philosophy, 30, 1992. Lipkind, Donald. “Russell on the Notion of Cause.” Canadian Journal of Philosophy, 9, 1979. Lockwood, Michael. “What Was Russell’s Neutral Monism?” Midwest Studies in Philosophy, 6, 1981. Lovejoy, Arthur O. “Mr. Bertrand Russell and the Unification of Mind and Matter.” In Lovejoy, Arthur O. (ed.), The Revolt Against Dualism. Chicago: Open Court, 1929.

Bibliography

405

Ludlow, Peter. “Indefinite Descriptions: In Defense of Russell.” Linguistics and Philosophy, 14, 1991. Lycan, William G. “Transformational Grammar and the Russell-Strawson Dispute.” Metaphilosophy, 1, 1970. McCawley, J. D. “Actions and Events Despite Bertrand Russell.” In Lapore, Ernest, and Brian P. McLaughlin (eds.), Actions and Events. Oxford: Blackwell, 1988. McDermott, Michael. “A Russellian Account of Belief Sentences.” Philosophical Quarterly, 38, 1988. McGrew, Timothy, and Lydia McGrew. “Internalism and the Collapse of the Gettier Problem.” Journal of Philosophical Research, 23, 1998. McGuinness, Brian. “Bertrand Russell’s and Ludwig Wittenstein’s ‘Notes on Logic.’” Revue Internationale de Philosophie, 26, 1972. McKenney, John I. “Concerning Russell’s Analysis of Value Judgement.” Journal of Philosophy, 55, 1958. McKeon, Matthew. “Bertrand Russell and Logical Truth.” Philosophia, 27, 1999. McKinsey, Michael. “The Semantics of Belief Ascription.” Nous, 33, 1999. McLendon, Hiram J. “Has Russell Answered Hume?” Journal of Philosophy, 49, 1952. McLendon, Hiram J. “Has Russell Proved Naïve Realism Self-Contradictory?” Journal of Philosophy, 53, 1957. McMahon, M. Brian. “Russell’s Denoting Relation.” The Personalist, 57, 1976. Magnell, Thomas. “The Extent of Russell’s Modal Views.” Erkenntnis, 34, 1991. Makin, Gideon. “Making Sense of ‘On denoting.’” Synthese, 102, 1995. Makin, Gideon. “Why the Theory of Descriptons?” Philosophical Quarterly, 46, 1996. Marcus, Ruth Barcan. “On Some Post-1920’s Views of Russell on Particularity, Identity, and Individualism.” In Marcus, Ruth Barcan (ed.), Modalities. Oxford: Oxford University Press, 1993. Martin, Richard M. “On the Berkeley-Russell Theory of Proper Names.” Philosophy and Phenomenological Research, 13, 1952. Martin, Richard M. “The Philosophic Import of Virtual Classes.” Journal of Philosophy, 61, 1964. Martinich, Aloysius P. “Russell, Frege, and the Puzzle of Denoting.” International Studies in Philosophy, 14, 1976. Martinich, Aloysius P. “Russell’s Theory of Meaning and Descriptions.” Journal of the History of Philosophy, 14, 1983. Mates, Benson. “Descriptions and Reference.” Foundations of Language, 10, 1973. Maxwell, Grover. “Russell on Perception and Mind-Body: A Study in Philosophical Method.” In Cheng, Chung-Ying (ed.), Philosophical Aspects of the Mind-Body Problem. Honolulu: University of Hawaii Press, 1975. Meyers, Robert G. “Knowledge by Acquaintance: A Repy to Hayner.” Philosophy and Phenomenological Research, 31, 1970. Miah, Sajahan. “Constructionism: Russell’s Resolution of Realism-Empiricism Dilemma.” Indian Philosophical Quarterly, 24, 1997.

406

Bibliography

Michell, Joel. “The Origins of the Representational Theory of Measurement: Helmholtz, Hölder, and Russell.” Studies in History and Philosophy of Science, 24, 1993. Michell, Joel. “Bertrand Russell’s 1897 Critique of the Traditional Theory of Measurement.” Synthese, 110, 1997. Mijuskovic, Ben. “The Simplicity Argument in Wittgenstein and Russell.” Critica, 8, 1976. Moorcroft, Francis. “Why Russell’s Paradox Won’t Go Away.” Philosophy, 68, 1993. Moore, George E. “Is Existence a Predicate?” Aristotelian Society Supplementary, 15, 1936. Moore, George E. “Is There Knowledge by Acquaintance?” Aristotelian Society Supplementary, 2, 1919. Morris, William Edward. “Moore and Russell on Philosophy and Science.” Metaphilosophy, 10, 1979. Moss, J.M.B. “Some B. Russell’s Sprouts (1903-1908).” In Hodges, Wilfrid (ed.), Conference in Mathematical Logic – London ’70 (Lecture Notes in Mathematics, 255). Berlin: Springer Verlag, 1972. Moulder, James. “Is Russell’s Paradox Genuine?” Philosophy, 49, 1974. Muehlmann, Robert. “Russell and Wittgenstein on Identity.” Philosophical Quarterly, 19, 1969. Mulligan, Kevin. “Relations-Through Thick and Thin.” Erkenntnis, 48, 1998. Musgrave, Alan. “Logicism Revisited.” British Journal for the Philosophy of Science, 28, 1977. Myhill, John. “Report of Some Investigations Concerning the Consistency of the Axiom of Reducibility.” Journal of Symbolic Logic, 16, 1951. Nagel, Ernest. “Basis of Human Knowledge.” In Nagel, Ernest (ed.), Sovereign Reason. New York: Free Press, 1954. Nelson, John O. “On Sommer’s Reinstatement of Russell’s Ontological Program.” Philosophical Review, 73, 1964. Nelson, Mark T. “Bertrand Russell’s Defence of the Cosmological Argument.” American Philosophical Quarterly, 35, 1998. Newman, J. H. A. “Mr. Russell’s ‘Causal Theory of Perception.” Mind, 37, 1928. Nielson, Kai. “Bertrand Russell’s New Ethic.” Methodos, 10, 1958. Oaklander, I. Nathan. “Does the Russellian Theory of Time Entail Fatalism?” The Modern Schoolman, 59, 1923. Oaklander, I. Nathan, and Silvano Miracchi. “Russell, Negative Facts, and Ontology.” Philosophy of Science. 47, 1980. Over, D. E. “Russell’s Hierarchy of Acquaintance.” Philosophical Papers, 16, 1987. Pakaluk Michael. “The Doctrine of Relations in Bertrand Russell’s Principles of Mathematics.” Topicos, 2, 1992. Pape, Helmut. “Peirce and Russell on Proper Names.” Transactions of the Charles S. Peirce Society, 18, 1992. Parker, DeWitt H. “Knowledge by Acquaintance.” Philosophical Review, 54, 1945.

Bibliography

407

Parker, DeWitt H. “Knowledge by Description.” Philosophical Review, 54, 1945. Parris, Henry. “The Political Thought of Bertrand Russell.” Durham University Journal, 58, 1965. Parsons, Terence D. “Russell’s Early Views on Demoting.” In Austin David F. (ed.), Philosophical Analysis. Dordrecht: Kluwer, 1988. Patzold, Detlev. “Wandlungen des Kausalitätsbegriffs.” Dialektik, 2, 1998. Pears, David F.. A Comparison between Ayer’s Views about the Privileges of SenseDatum Statements and the Views of Russell and Austin.” In Macdonald, G. F. (ed.), Perception and Identity. New York: Cornell University Press, 1979. Pears, David F. “The Relation between Wittgenstein’s Picture Theory and Russell’s Theories of Judgement.” Philosophical Review, 86, 1977. Pelham, Judy. “Russell, Frege, and the Nature of Implication.” Topoi, 18, 1999. Peressini, Anthony. “Cumulative Versus Noncumulative Ramified Types.” Notre Dame Journal of Formal Logic, 38, 1997. Perkins, Ray, Jr. “Russell, Frege, and the Meaning of the Theory of Descriptions (or): Did Russell Know his Frege?” Journal of the History of Philosophy, 20, 1992. Perkins, Ray, Jr. “Russell on Memory.” Mind, 82, 1973. Perkins, Ray, Jr. “Russell’s Realist Theory of Remote Memory.” Journal of the History of Philosophy, 14, 1976. Pineau, Lois. “Russell on Ordinary Names and Synonymy.” History of Philosophy Quarterly, 7, 1990. Pitt, Jack. “Russell on Religion.” International Journal for Philosophy of Religion, 6, 1975. Popper, Karl R., and F. Hansdorff. “Scientific Inference According to Bertrand Russell.” Hibbert Journal, 47, 1949. Prior, Arthur N. “Existence in Lesniewski and Russell.” In Crossley, John N., and Michael Dummett (eds.), Formal Systems and Recursive Functions. Amsterdam: North Holland, 1965. Quine, Willard Van Orman. “Designation and Existence.” Journal of Philosophy, 36, 1939. Quine, Willard Van Orman. “New Foundations for Mathematical Logic.” American Mathematical Monthly, 44, 1937. Quine, Willard Van Orman. “On the Theory of Types.” Journal of Symbolic Logic, 3, 1938. Quine, Willard Van Orman. “On What There Is.” Review of Metaphysics, 2, 1948. Radford, Colin. “MacColl, Russell, the Existential Import of Propositions, and the Null Class.” Philosophical Quarterly, 45, 1995. Ramachandran, Murali. “A Strawsonian Objection to Russell’s Theory of Descriptions.” Analysis, 53, 1993. Ramachandran, Murali. “Bach on Behalf of Russell.” Analysis, 55, 1995. Ramsey, Frank Plumpton. “The Foundations of Mathematics.” Proceedings of the London Mathematical Society ser. 2, 25, 1925. Richards, John. “Propositional Functions and Russell’s Philosophy of Language, 19031914.” Philosophical Forum, 11, 1980.

408

Bibliography

Richardson, Alan. “How Not to Russell Carnap’s Aufbau.” Proceedings of the Philosophy of Science Association, 1, 1990. Rieber, Steven. “Could Demonstratives be Descriptive?’ Philosophia, 26, 1998. Riska, Augustin. “Knowledge by Acquaintance Reconsidered.” Grazer Philosophische Studien, 11, 1980. Riverso, Emanuele. “A. J. Ayer’s Treatment of Russell.” In Hahn, L. E. (ed.), The Philosophy of A. J. Ayer. La Salle, IL: Open Court, 1992. Roberts, Lawrence. “Russell on the Semantics and Pragmatics of Indexicals.” Philosophia, 14, 1984. Rodrigues-Consuerga, Francisco. “Bertrand Russell and Bradley’s Ghosts.” In Hintikka, J., and K. Puhl (eds.), The British Tradition in Twentieth Century Philosophy. Vienna: Hölder-Pichler-Tempsky, 1995. Rodrigues-Consuerga, Francisco. “Russell, Gödel and Logicism.” In Czermak, J. (ed.), Philosophy of Mathematics. Vienna: Hölder-Pichler-Tempsky, 1993. Rodrigues-Consuerga, Francisco. “Russell’s Logicist Definitions of Numbers, 18981913: Chronology and Significance.” History and Philosophy of Logic, 8, 1987. Rodrigues-Consuerga, Francisco. “Russell’s Theory of Types, 1901-1910: Its Complex Origin in the Unpublished Manuscripts.” History and Philosophy of Logic, 10, 1989. Rolf, Bertil. “Russell’s Thesis on Vagueness.” History and Philosophy of Logic, 3, 1982. Rolston, John. “Bradley on Russell and Relations.” Philosophical Forum, 4, 1973. Rosenberg, Alexander. “Russell versus Steiner on Physics and Causality.” Philosophy of Science, 56, 1989. Rosenberg, Jay. “Russell on Negative Facts.” Nous, 6, 1972. Ruffino, Marco Antonio. “The Context Principle and Wittgenstein’s Criticism of Russell’s Theory of Types.” Synthese, 93, 1994. Saarinen, Esa. “How to Frege a Russell-Kaplan.” Nous, 16, 1982. Sainsbury, R. Mark. “Russell on Acquaintance.” In Vesey, G. (ed.), Philosophers Ancient and Modern. Cambridge: Cambridge University Press, 1986. Sainsbury, R. Mark. “Russell on Construction and Fictions.” Theoria, 46, 1980. Salmon, Nathan. “Nonexistence.” Nous, 32, 1998. Schock, Rolf. “Some Remarks on Russell’s Treatment of Definite Descriptions.” Logique et Analyse, 5, 1962. Schultz, Bart. “Bertrand Russell in Ethics and Politics.” Ethics, 102, 1992. Schutte, Karl. “Syntactical and Semantical Properties of Simple Type Theory.” Journal of Symbolic Logic, 35, 1960. Schwerin, Alan. “Some Remarks on Russell’s Account of Vagueness.” Contemporary Philosophy, 21, 1999. Searle, John R. “Russell’s Objections to Frege’s Theory of Sense and Reference.” Analysis, 18, 1958. Sellars, Wilfrid. “Acquaintance and Descriptions Again.” Journal of Philosophy, 46, 1949. Sellars, Wilfrid, “Classes as Abstract Entities and the Russell Paradox.” Review of Metaphysics, 17, 1963.

Bibliography

409

Shanker, Stuart. “Wittgenstein versus James and Russell on the Nature of Willing.” In Canfield, John, and Stuart Shanker (eds.), Wittgenstein’s Intentions. New York: Garland, 1993. Sharvy, Richard. “A More General Theory of Definite Descriptions.” Philosophical Review, 89, 1980. Shearn, Martin. “Russell’s Analysis of Existence.” Analysis, 11, 1951. Siitonen, Arto. “Logical Atomism Reconsidered.” In Puhl, Klaus (ed.), Wittgenstein’s Philosophy of Mathematics. Vienna: Hölder-Pichler-Tempsky, 1993. Simons, Peter. “On What There Isn’t: The Meinong-Russell Dispute.” In Simons, Peter (ed.), Philosophy and Logic in Central Europe From Bolzano to Tarski: Selected Essays. Dordrecht: Kluwer, 1992. Skyrms, Brian. “Logical Atoms and Combinatorial Possibility.” Journal of Philosophy, 90, 1993. Smart, Jack C. C. “Whitehead and Russell’s Theory of Types.” Analysis, 10, 1950. Smullyan, Arthur F. “Incomplete Symbols.” Philosophical Review, 67, 1958. Soloman, Graham. “An Addendum to Demopolous and Friedman (1985).” Philosophy of Science, 56, 1989. Stock, G. “Russell’s Theory of Judgment in Logical Atomism.” In Vesey, Godfrey (ed.), Understanding Wittgenstein. New York: St. Martin’s Press, 1974. Stroll, Avrum. “Four Comments on Russell’s Theory of Descriptions.” Canadian Journal of Philosophy, 8, 1978. Tatievskaia, Elena. “On Logical Form.” In Falguera, Jose I. (ed.), La Filosofia Analitica en el Cambio de Milenio. Santiago de Compostela: S.I.E.U., 1999. Taylor, Gladys, G. “The Analytic and Synthetic in Russell’s Philosophy of Mathematics.” Philosophical Studies, 39, 1981. Taylor, Elena. “The Double Entente of Russell’s Thesis that Mathematics is Logic.” International Logic Review, 15, 1984. Taylor, Kenneth A. “François Recanati’s ‘Direct Reference: From Language to Thought.’” Nous, 31, 1997. Tibbetts, Paul. “Phenomenological and Empirical Inadequacies of Russell’s Theory of Perception.” Philosophical Studies (Ireland), 20, 1972. Urmson, J. O. “Russell on Acquaintance with the Past.” Philosophical Review, 78, 1969. Urmson, J.O. “Russell on Universals.” In Vesey, Godfrey (ed.), Philosophers Ancient and Modern. Cambridge: Cambridge University Press, 1986. Urmson, J. O. “Russell’s Incomplete Symbols.” Analysis, 33, 1973. van Patten, James. “Some Reflections on Bertrand Russell’s Philosophy.” Educational Theory, 15, 1965. Vardy, P. “Some Remarks on the Relationship between Russell’s Vicious-Circle Principle and Russell’s Paradox.” Dialectica, 33, 1979. Vernant, Denis. “Le statut de la vérité dans le calcul implicationnel de Russell en 1903.” Revue Internationalede Philosophie, 51, 1997. Wahl, Russell. “Bertrand Russell’s Theory of Judgement.” Synthese, 68, 1986.

410

Bibliography

Wahl, Russell. “Russell’s Theory of Meaning and Denotation and ‘On Denoting.’” Journal of the History of Philosophy, 31, 1993. Weiss, Bernard. “On Russell’s Argument for Restricting Modes of Specification and Domains of Quantification.” History and Philosophy of Logic, 15, 1994. Welding, S. O. “Frege’s Conception of Sense and Reference Related to Russell’s Theory of Definite Descriptions.” Revue Internationale de Philosophie, 25, 1971. Welding, S. O. “Russell’s Theory of Definite Description as Opposed to Quine’s Singular Terms.” Revue Internationale de Philosophie, 26, 1972. Wettstein, Howard K. “Demonstrative Reference and Definite Descriptions.” Philosophical Studies, 40, 1981. Wettstein, Howard K. “Frege-Russell Semantics?” Dialectica, 44, 1990. Wetzel, Thomas. “Possible States of Affairs.” Philosophical Studies, 91, 1998. White, Alan R. “Knowledge, Acquaintance, and Awareness.” Midwest Studies in Philosophy, 6, 1981. Wilson, W. Kent. “Incomplete Symbols and Russell’s Proof.” Canadian Journal of Philosophy, 10, 1980. Woodger, J. H. “Russell’s Theory of Perception.” Monist, 40, 1930. Yagisawa, Takashi. “A Somewhat Russellian Theory of Intensional Contexts.” In Tomberlin, James E. (ed.), Philosophical Perspectives, II, Mind, Causation and World. Boston: Blackwell, 1997. Yourgrau, Palle. “Russell and Kaplan on Denoting.” Philosophy and Phenomenological Research, 46, 1985. Zimmerman, David. “Born Yesterday: Personal Autonomy for Agents without a Past.” Midwest Studies in Philosophy, 23, 1999.

Since 2000 Amijee, Fatema. “The Role of Attention in Russell’s Theory of knowledge.” British Journal for the History of Philosophy, 21, 2013. Anellis, Irving. “Some Views of Russell and Russell’s Logic by His Contemporaries.” Review of Modern Logic, 10, 2005. Banks, Eric C. “Neutral Monism Reconsidered.” Philosophical Psychology, 23, 2010. Basile, Pierfrancesco. “Leaning from Leibniz: Whitehead (and Russell) on Mind, Matter, and Monads.” British Journal for the History of Philosophy, 23, 2015. Beany, Michael. “The Analytic Revolution.” Philosophy Supplement, 79, 2016. Beany, Michael. “Conceptions of Analysis in Early Analytic Philosophy.” Acta Analytica, 15, 2000. Beruman, Michael. “Bertrand Russell and the ‘Conquest of Happines.’” Free Inquiry, 26, 2006. Blanchette, Patricia. “The Breadth of the Paradox.” Philosophia Mathematica, 24, 2016. Blendea, Sevastian. “Russell on Referring.” Analysis and Metaphysics, 42, 2005. Boër, Steven E. “On the Multiple Relation Theory of Judgment.” Erkenntnis, 56, 2002.

Bibliography

411

Bostock, David. “Russell’s Early Theory of Denoting.” History and Philosophy of Logic, 30, 2009. Brogaard, Berit. “The ‘Gray’s Elegy’ Argument, and the Prospects for the Theory of Denoting Concepts.” Synthese, 152, 2006. Cargile, James. “On Russell’s Argument against Resemblance Nominalism.” Australasian Journal of Philosophy, 81, 2003. Champagne, Marc. “Russell and the Newman Problem Revisited.” Analysis and Metaphysics, 11, 2012. Chateaubriand, Oswoldo. “Descriptions: Frege and Russell Combined.” Synthese, 130, 2002. Clapp, Lenny, and Robert Stainton. “‘Obviously Propositions Are Nothing’: Russell and the Logical Form of Belief Reports.” In Preyer, Gerhard (ed.), Logical Form and Language. Oxford: Clarendon Press, 2002. Cocchiarella, Nino B. “Denoting Concepts, Reference, and the Logic of Names, Classes As Many, Groups, and Plurals.” Linguistics and Philosophy, 28, 2005. Cocchiarella, Nino. “On the Logic of Classes as Many.” Studia Logica, 70, 2002. Connelly, James. “Russell and Wittgenstein on Logical Form and Judgement: What did Wittgenstein Try That Wouldn’t Work?” Theoria, 80, 2014. D’Cruz, Mark. “A Theory of Ordinary Proper Names.” Mind, 109, 2000. Demopoulos, William. “The 1910 Principia’s Theory of Functions and Classes and the Theory of Descriptions.” European Journal of Analytic Philosophy, 3, 2007. Engel, Pascal. “Bertrand Russell: An Inquiry into Meaning and Truth.” In Shand, John (ed.), The Twentieth Century: Moore to Popper. Montreal: McGill-Queens University Press, 2006. Engelmann, Mauro. “Wittgenstein’s New Method and Russell’s ‘The Analysis of Mind.’” Journal of Philosophical Research, 37, 2012. Fuhrmann, André. “Russell’s Way Out of the Paradox of Propositions.” History and Philosophy of Logic, 23, 2002. Gandon, Sébastien. “‘Principia Mathematica.’” Logique et Analyse, 54, 2011. Gandon, Sébastien. “Quelle Philosophie pour Quelle Mathématique?” Archives de Philosophie, 76, 2013. Goldstein, Laurence. “A Unified Solution to Some Paradoxes.” Proceedings of the Aristotelian Society, 100, 2000. Graff, Delia. “Descriptions as Predicates.” Philosophical Studies, 102, 2001. Griffin, Nicholas. “What Did Russell Learn from Leibniz?” Journal for the History of Analytical Philosophy, 2, 2013. Halimi, Brice. “The Versatility of Universality in ‘Principia Mathematica.’” History and Philosophy of Logic, 32, 2011. Hanks, Peter W. “How Wittgenstein Defeated Russell’s Multiple Relation Theory of Judgment.” Synthese, 154, 2007. Haque, Nayeema. “Russell’s Argument against Fregean Sense.” Journal of Indian Council of Philosophical Research, 17, 2000. Hare, William. “Bertrand Russell on Critical Thinking.” Journal of Thought, 36, 2001.

412

Bibliography

Hazen, Alan, and J. M. Davoren. “Russell’s 1925 Logic.” Australasian Journal of Philosophy, 78, 2000. Hillman, T. Allan. “The Early Russell on the Metaphysics of Substance in Leibniz and Bradley.” Synthese, 163, 2008. Hochberg, Herbert. “Facts, Truths and the Ontology of Logical Realism.” Grazer Philosophische Studien, 58–59, 2000. Hochberg, Herbert. “Propositions, Truth and Belief: The Wittgenstein-Russell Dispute.” Theoria, 66, 2000. Hochberg, Herbert. “Ramsey vs. Russell: Particulars, Universals and Truth.” Metaphysica, 3, 2005. Hochberg, Herbert. “Russell and Ramsey on Distinguishing between Universals and Particulars.” Grazer Philosophische Studien, 67, 2004. Hochberg, Herbert. “Russell and Ramsey on Facts and Forms.” Logical Analysis and History of Philosophy, 9, 2006. Holman, Emmett. “Panpsychism, Physicalism, Neutral Monism and the Russellian Theory of Mind.” Journal of Consciousness Studies, 15, 2008. Hornsby, Jennifer. “Neale, Russell and Frege on Facts.” Protosociology, 23, 2006. Hylton, Peter. “Frege and Russell.” In Potter and Ricketts (eds.), The Cambridge Companion to Frege. Cambridge: Cambridge University Press, 2010. Jorgenson, Larry M. “Russell’s Leibnizian Concept of Vagueness.” History of Philosophy Quarterly, 28, 2011. Kamareddine, Fairous, Twan Laan, and Rob Nederpelt. “Types in Logic and Mathematics before 1940.” Bulletin of Symbolic Logic, 8, 2002. Kitchener, Richard F. “Bertrand Russell’s Flirtation with Behaviorism.” Behavior and Philosophy, 32, 2004. Klement, Kevin. “Early Russell on Types and Plurals.” Journal for the History of Analytical Philosophy, 2, 2014. Klement, Kevin. “The Function of Russell’s No Class Theory.” The Review of Symbolic Logic, 3, 2010. Klement, Kevin. “The Paradoxes and Russell’s Theory of Incomplete Symbols.” Philosophical Studies, 169, 2014. Klement, Kevin. “Putting Form Before Function: Logical Grammar in Frege, Russell, and Wittgenstein.” Philosophers’ Imprint, 4, 2004. Klement, Kevin. “Russell, His Paradoxes, and Cantor’s Theorem.” (2 parts) Philosophy Compass, 5, 2010. Klement, Kevin. “Russell’s 1903-1905 Anticipation of the Lambda Calculus.” History and Philosophy of Logic, 24, 2003. Klement, Kevin. “Russell’s Paradox in Appendix B of the Principles of Mathematics: Was Frege’s response adequate?” History and Philosophy of Logic, 22, 2001. Knight, Gordon. “Idealism, Intentionality, and Nonexistent Objects.” Journal of Philosophical Research, 26, 2001.

Bibliography

413

Kraal, Anders. “The Aim of Russell’s Early Logicism: A Reinterpretation.” Synthese, 191, 2014. Kripke, Saul. “Frege’s Theory of Sense and Reference: Some Exegetical Notes.” Theoria, 74, 2008. Kumar Sen, Amit. “Strawson on Presupposition.” Indian Philosophical Quarterly, 27, 2000. Lambert, Karel. “Set Theory and Definite Descriptions ‘Four Solutions in Search of a Common Problem.’” Grazer Philosophische Studien, 60, 2000. Landini, Gregory. “Quantification Theory in *9 of ‘Principia Mathematica’.” History and Philosophy of Logic, 21, 2000. Landini, Gregory. “Typoes of ‘Principia Mathematica.’” History and Philosophy of Logic, 34, 2013. Levine, James. “Analysis and Decomposition in Frege and Russell.” Philosophical Quarterly, 52, 2002. Levine, James. “On the ‘Gray’s Elegy’ Argument and Its Bearing on Frege’s theory of Sense.” Philosophy and Phenomenological Research, 69, 2004. Levine, James. “Russell, Particularized Relations and Bradley’s Dilemma.” Dialectica, 68, 2014. Linsky, Bernard. “Leon Chwistek on the No-Classes Theory in ‘Principia Mathematica.’” History and Philosophy of Logic, 25, 2004. Linsky, Bernard. “Russell’s Logical Form LF and Truth-Conditions.” In Preyer, Gerhard (ed.), Logical Form and Language. Oxford: Clarendon Press, 2002. Livingston, Paul M. “Russellian and Wittgensteinian Atomism.” Philosophical Investigations, 24, 2001. Los, Paul. “Wittgenstein’s Objection to Russell’s Theory of Judgment.” Philosophical Writings, 30, 2005. Mancosu, Paulo. “The Russellian Influence on Hilbert and His School.” Synthese, 137, 2003. Mares, Edwin. “The Fact Semantics for Ramified Type Theory and the Axiom of Reducibility.” Notre Dame Journal of Formal Logic, 48, 2007. Martens, David B. “Knowledge and True Belief in Early Analytic Philosophy.” South African Journal of Philosophy, 31, 2012. Martin, Michael G. F. “Old Acquaintance: Russell, Memory and Problems with Acquaintance.” Analytic Philosophy, 56, 2015. Masciarelli, Pasqualino. “On a Notion of Extra-Essential Identity: Critical Notes on an Objection of Russell’s to Hegel.” Topoi, 19, 2000. McBride, Fraser. “The Cambridge Revolt Against Idealism: Was There Ever an Eden?” Metaphilosophy, 43, 2012. McBridge, Fraser. “Whence the Particular-Universal Distinction?” Grazer Philosophische Studien, 67, 2004. Milkov, Nikolay. “The History of Russell’s Concepts ‘Sense-Data’ and ‘Knowledge by Acquaintance.’” Archiv fuer Begriffsgeschichte, 43, 2001.

414

Bibliography

Milkov, Nicolay. “Lotze and the Early Cambridge Analytic Philosophy.” Prima Philosophia, 13, 2000. Milkov, Nikolay. “Russell’s Debt to Lotze.” Studies in History and Philosophy of Science, 39, 2008. Miller, Alexander. “Russell, Multiple Relations, and the Correspondence Theory of Truth.” Monist, 89, 2006. Milne, Peter. “Russell’s Completeness Proof.” History and Philosophy of Logic, 29, 2008. Morris, Sean. “Quine, Russell, and Naturalism: From a Logical Point of View.” Journal of the History of Philosophy, 53, 2015. Moulines, C. Ulises. “Die Mathematisierung der Erfahrung: Vorganger zu Carnaps ‘Aufbau.’” Erkenntnis, 54, 2001. Nasim, Omar. “The Spaces of Knowledge: Bertrand Russell, Logical Constructions, and the Classification of the Sciences.” British Journal for the History of Philosophy, 20, 2012. Ostertag, Gary. “A Problem for Russellian Theories of Belief.” Philosophical Studies, 146, 2009. O’Sullivan, Michael. “The Visual Field in Russell and Wittgenstein.” Philosophical Investigations, 38, 2015. Peregrin, Jaroslav. “‘Fregean’ Logic and ‘Russellian’ Logic.” Australasian Journal of Philosophy, 78, 2000. Pincock, Christopher. “Carnap, Russell and the External; World.” In Friedman, Michael, and Richard Creath (eds.), The Cambridge Companion to Carnap. Cambridge: Cambridge University Press, 2007. Pincock, Christopher. “Russell’s Influence on Carnap’s Aufbau.” Synthese, 131, 2002. Pincock, Christopher. “Russell’s Last (and Best) Multiple-Relation Theory of Judgement.” Mind, 117, 2008. Plourde, Jimmy. “Wittgenstein et les théories du jugement de Russell et Meinong.” Dialogue, 44, 2005. Prasad, B. Sambsiva. “Sahajan Miah on Russell’s Constructionism.” Indian Philosophical Quarterly, 27, 2000. Predelli, Stephano. “Russell-Names: An Introduction to Millian Descriptivism.” Journal of Philosophical Logic, 45, 2016. Predelli, Stephano. “The Teetotaler and His Martini.” Mind and Language, 15, 2000. Proops, Ian. “Russell and the Universalist Conception of Logic.” Nous, 41, 2007. Proops, Ian. “Russellian Acquaintance Revisited.” Journal of the History of Philosophy, 52, 2014. Proops, Ian. “Russell on Substitutivity and the Abandonment of Propositions.” Philosophical Review, 120, 2011. Proops, Ian. “Russell’s Reasons for Logicism.” Journal of the History of Philosophy, 44, 2006. Proops, Ian. “Soames on the Metaphysics and Epistemology of Moore and Russell.” Philosophical Studies, 129, 2006.

Bibliography

415

Psillos, Stathis. “Carnap, the Ramsey-Sentence and Realistic Empiricism.” Erkenntnis, 52, 2000. Pulkkinen, Jarmo. “Russell and the Neo-Kantians.” Studies in History and Philosophy of Science, 32, 2001. Pulkkinen, Jarmo. “Russell and the Neo-Kantians.” Studies in History and Philosophy of Logic, 32A, 2001. Pupa, Francesco. “On the Russellian Reformation.” Philosophical Studies, 147, 2010. Ray, Nicholas. “Interpreting Russell’s Gray’s ‘Elegy’ Argument.” Dialogue, 51, 2012. Ricketts, Thomas. “Wittgenstein against Frege and Russell.” In Reck, Erich (ed.), From Frege to Wittgenstein: Perspectives on Early Analytic Philosophy. New York: Oxford University Press, 2002. Rodriguez-Pereyra, Gonzalo. “Resemblance Nominalism and Russell’s Regress.” Australasian Journal of Philosophy, 79, 2001. Rosenkrantz, Max. “The King of France Restored.” Metaphysica, 8, 2007. Rosenkrantz, Max. “The Ontological Motivations for the Theory of Descriptions.” Pacific Philosophical Quarterly, 86, 2005. Rosenkrantz, Max. “The Problem of False Belief and the Failure of the Theory of Descriptions.” Theoria, 82, 2016. Ross, Don, and David Spurrett. “Notions of Cause: Russell’s Thesis Revisited.” British Journal for the Philosophy of Science, 58, 2007. Ryan, Alan. “Bertrand Russell and Ralph Barton Perry on War and Nonviolence: On Russell’s ‘The Ethics of War.’” Ethics, 125, 2014. Salmon, Nathan. “About Aboutness.” European Journal of Analytic Philosophy, 3, 2007. Schwerin, Alan. “Metaphysics, Mysticism and Russell.” Contemporary Philosophy, 24, 2002. Smith, Sheldon R. “Resolving Russell’s Anti-Realism about Causation: The Connection between Causation and the Functional Dependencies of Mathematical Physics.” Monist, 83, 2000. Soames, Scott. “No Class: Russell on Contextual Definition and the Elimination of Sets.” Philosophical Studies, 139, 2008. Sosa, David. “Rigidity in the Scope of Russell’s Theory.” Nous, 35, 2001. Stevens, Graham. “From Russell’s Paradox to the Theory of Judgement: Wittgenstein and Russell on the Unity of the Proposition.” Theoria, 70, 2004. Stevens, Graham. “Russell and the Unity of the Proposition.” Philosophy Compass, 3, 2008. Stevens, Graham. “Russell’s Ontological Development Reconsidered.” British Journal for the History of Philosophy, 18, 2010. Stevens, Graham. “Russell’s Repsychologising of the Proposition.” Synthese, 151, 2006. Stevens, Graham. “The Truth and Nothing But the Truth, Yet Never the Whole Truth: Frege, Russell and the Analysis of Unities.” History and Philosophy of Logic, 24, 2003. Szabo, Zoltan Gendler. “Descriptions and Uniqueness.” Philosophical Studies, 101, 2000.

416

Bibliography

Textor, Mark. “Vacuous Names in Early Analytic Philosophy: Frege, Russell, and Moore.” Philosophy Compass, 11, 2016. Urquhart, Alasdair. “Russell and Gödel.” Bulletin of Symbolic Logic, 22, 2016. Wilson, Mark. “Frege and Russell: Does Science Talk Sense?” European Journal of Analytic Philosophy, 3, 2007.

Index Abriss der Logistik (Carnap)  7, 125 Absolute, the  29, 42 Bradley’s view  23 McTaggart’s view  23 number  45 abstraction  92 classes  132 principle  129, 130 Acero, Juan Jose  72 n.10 acquaintance  36–8, 112–13 introspective  261 knowing mind by  260–3 knowledge by  10, 67, 102, 192, 194, 239, 241, 244, 256, 257, 314 with objects  87–8 post-peano philosophy and  46–7 as relation  209 self  266–71 Acquaintance, Knowledge and Logic: New Essays on Bertrand Russell’s Problems of Philosophy  234 n.10 advanced logic  64 Aiken, L.  355 n.26, 356 n.31, 357 n.50 The Algebra of Logic  78 Almog, Joseph  201 n.2 Alter, Torin  281 n.39, 331 n.11 Ameseder, Rudolf  90 Amijee, Fatema  277 nn.9–10 The Analysis of Matter  209, 246–7, 251, 252, 325–7, 329, 331 n.3 The Analysis of Mind  11, 68, 69, 71, 189, 209, 238, 245–6, 248, 275–6, 281 n.39, 312, 313, 324–7, 331 n.8 analytic philosophy  2 analytic realism  102 analytic realists  68 Anderson, C. Anthony  254 n.18, 304 n.9 Andersson, Stefan  385 n.26 animal inference  249 Appearance and Reality (Bradley)  25 application constraint  127

a priori knowledge  37, 104, 105, 216, 221–3, 225 a priori principle  51, 105–6, 225 a priori properties of space  31 arithmetic, foundations of  83 Armstrong, D. M.  51 n.2 Audi, Robert  304 n.9 Aufbau (Carnap)  124, 126, 128–35, 137, 143 n.24 Austin, J. L.  82, 151, 154, 163, 200 Autobiography of Bertrand Russell  13 n.1, 14 n.3, 22, 61, 107, 156, 157, 188, 351, 352, 357 n.54 (auto) psychologically based system  132, 134, 244 awareness  262 direct  263–4, 267, 274–5 experiential  258 introspective  261, 262–3, 269 mental act of  239, 246 of objects  27, 239, 245, 256, 263 by subject  27 Awodey, S.  142 Axiom of Infinity  85 Ayer, Alfred J.  136, 244, 246, 253 n.11, 254 n.14, 304 n.9, 345, 356 n.39, 357 nn.50, 52 Baldwin, Thomas  72 n.11, 271, 277 n.11 Barnes, Hazel E.  278 n.14 Barnes, W. H. F.  356 n.32 Basic Laws of Arithmetic (Frege)  2 Beacon Hill School  3 Beaney, Michael  52 n.19, 94 n.6, 202 n.13 Beckermann, Ansgar  277 n.11 Bedeutung  82, 83 Begriffschrift (Frege)  83, 84, 233 n.1 behavior  71–2 beliefs  66–72 consequences of  60–1 meaning  65

418 Index rational, structure for  286–307 self-evidence  291 Beman, W. W.  94 n.11 Benacerraf, P.  137 Berkeley, G.  51 n.2 Berkeley’s idealism  19 Bertrand Russell Peace Foundation  3 Blackwell, Kenneth  14 n.5, 117 nn.2, 5, 118 n.7, 253 n.10 Blanshard, Brand  348, 356 n.42 Block, I.  117 nn.2, 5, 118 n.7 Blue Book (Wittgenstein)  100 Blumberg, A. E.  144 n.41 Bok, Hilary  279 n.23 BonJour, Laurence  304 nn.12, 14, 305 nn.16, 19 Bonk, Th.  143 n.21 Boole, George  77 Boolos, George  175 n.22 Bostock, David  120 n.23, 175 n.24, 254 n.20, 257, 272, 277 n.11, 279 n.28, 281 n.39, 331 n.1, 332 n.13 Bowers, F.  66 Bradley, F. H.  22, 24, 51 nn.6, 8 ontological argument  28–9 theory of judgment  25, 28–9 Bradley’s regress  24 Braithwaite, Richard  62, 355 n.19 Brandt, R.  357 n.51 Brennan, Frances  304 n.8 Brentano, Franz  258, 278 n.14 Broad, C. D.  112 Brodbeck, M.  137 Buchler, J.  356 n.41 Burgess, John  169 Burkhard, F. H.  66 Burks, A.  65 Bury, J. B.  363 business of philosophy  39–40 Campbell, John  277 n.11 Candlish, Stewart  53 n.22 Cantor, Georg  80 theorem  79 theory  44 cardinal numbers  45–6, 154, 160–1 Carey, Rosalind  119 n.16, 277 n.7 Carnap, Rudolf  123, 136, 159 logicism  124–8

Principia  124–8 Carnap Brought Home  142 causal theory of perception  251–2 Chance, Love and Logic  65 Church, Alonzo  173 nn.9–10 Clark, Ronald  13 n.1 Coffa, J. Alberto  175 n.22 Cohen, M. R.  65 Cohen, R. S.  140 coherentism  11, 140–1, 287–8, 290–5 common property  130–1 common sense  43 complete induction, theorem of  84–5 concept construction, theory of  132 Concept of Mind  254 n.14 concepts versus objects  76–7 Conee, Earl  11, 294, 305 n.17 Congress of Scientific Philosophy in Paris, September 1935  124 The Conquest of Happiness  3 considered judgments  302 Copleston, F.  356 n.38 cosmological arguments  29 Costello, Harry  64, 65 Couturat, Louis  4 cp-logic  8, 206, 208–13, 211, 215–16, 221, 226, 230 Crane, Tim  257 Currie, Gregory  306 n.23 Dancy, Jonathan  304 n.14 Daniels, Norman  307 n.31 Darapti  92 Dauenhauer, Bernard P.  383 n.12 Davoren, J. M.  175 n.23 Decline and Fall of the Roman Empire (Gibbon)  12 Dedekind, Richard  84, 94 n.11, 151 deduction theory (see propositional calculus) definite descriptions  184, 185–6, 190–2 semantic theories of  195 denoting concepts  75–6, 87 DePaul, Michael R.  290, 307 n.31 derived propositions  125 Der logische Aufbau der Welt (Carnap)  123 descriptions analysis of  183 definite  184, 185–6, 190–2

Index  419 disguised  184 empty  185 incomplete symbols  195–200 knowledge by  10, 67, 102, 112, 192–4, 214, 239, 241, 244, 257–8, 314 problems with  193–200 quantificational interpretations of  186 referential interpretations of  186–7 Russell’s theory of  183–7 semantics versus epistemology  193–5 theory of  183–7 descriptive knowledge  266–7 Dewey, John  14 n.4, 61, 228–9, 232 diagonal argument  79 dialectical method McTaggart’s view  22–3, 24 Russell’s view  28, 30–1 disharmony, of desires  343–4 Donnellan, Keith  186 Dreben, B.  52 n.12 dualism  34–5 Duhem, Pierre  307 nn.31–32 Dummett, Michael  94 n.14 Dunham, J.  52 n.13 Dunlap, Knight  332 n.12 Eames, Elizabeth  118 n.7, 304 n.9, 313, 316, 318 Ebert, P. A.  126, 151, 154, 155 Edwards, Paul  306 n.23, 355 n.24 elementary number theory  83 Elements of Logic  91 Elgin, Catherine  11, 294, 307 n.31 Eliot, T. S.  64 Eliot, Valerie  65 Ellington, J. Q.  51 n.4 Elster, Jon  377 emotivism  334, 342–8 empirical knowledge  37 empty descriptions  185 entities (related), intrinsic properties of  32–3 epistemic holism  11, 295, 296 epistemic justification  286 building bridge  286–90 coherence for  293–4 coherentism  11, 140–1, 287–8, 290–5

foundationalism  11, 238, 249, 287–8, 290–1, 293–6 regressive method  296–300 relevance of regressive method  300–3 structure of  290–6 epistemic program  286–7 epistemology  36–8 coherentism  11, 140–1, 287–8, 290–5 direct knowledge of  195 foundationalism  11, 238, 249, 287–8, 290–1, 293–6 indirect knowledge of  195 versus semantic  193–5 equivalence circles  132 Erkenntnis  137 Erkenntnistheorie  123 An Essay on the Foundations of Geometry (Frege)  1, 151 Essays in Analysis  296 Essays in Experimental Logic (Dewey)  68 Essays in Radical Empiricism-for Mind  66 ethical subjectivism (1913-44) early efforts  338–40 emotivism (1925–34)  342–8 error theory  340–1 Human Society in Ethics and Politics  348–51 ethics  221–8 of moral objectivity  340 in practice  351–3 subjectivity of  340 Ethics (Moore)  336 Evans, Gareth  87, 190–2, 277 n.11 experience Bradley’s view  24 “subject” and “object” in  36 explanationist coherentism  294 explanationist evidentialism  294 extensional sentence  181 Feigl, H.  137 Feinberg, B.  356 n.31 Feldman, Richard  11, 303 n.5, 305 nn.17, 19 Fick, J. S.  151 Foley, Richard  304 n.9, 305 n.20 Folina, Janet  307 n.28

420 Index foundational coherentism  294 foundationalism  11, 238, 249, 287–8, 290–1, 293–6 The Foundations of Mathematics  62 foundherentism  294 Frápolli, M. J.  72 n.10 Frege, Gottlob  75, 82, 94 nn.9, 10, 151, 154, 155, 163, 200, 202 n.13, 299 Axiom V  82 learning from  80–6 objections to  76–80 Fregeanism  182 function, mathematical notion of  76–7 Galaugher, Jolen  175 n.22 gamma (λ)-definable functions  165 Gandon, Sébastien  52 n.17, 173 n.5, 174 n.13, 175 n.22 Geach, Peter  277 n.11, 345 genealogical tree  133 generalization  92 General Theory of Knowledge (Schlick)  144 n.41 general theory of relativity (GTR)  328–9 geometry, Russell’s work on  31 George, R. A.  123 Geraets, T. F.  21 German Social Democracy  367, 383 n.5 Gibbon, Edward  375, 376, 378, 379 Gillies, Donald  307 n.31 Gödel, Kurt  174 nn.14–15 logicism and  164–8 Godwyn, Martin  306 n.27 Goldman, Alvin  307 n.31 Goodlad, Lauren M. E.  363 Goodman, Nelson  300 Graham, Peter  271 Grattan-Guinness, Ivor  94 n.6, 173, 203 n.18 Gregory, Mary  385 n.23 Griffin, Nicholas  14 n.5, 52 nn.10, 15, 17–19, 54 nn.36, 39, 72 n.11, 94 n.6, 95 nn.18, 19, 116 n.1, 118 n.10, 173 nn.2, 4, 7, 175 nn.22, 24, 254 n.19, 260, 281 n.39, 304 n.8, 306 n.27, 354 n.1 Grundgesetze der Arithmetik (Frege)  80, 126, 151, 299

Grundgesetze system  81, 86 Grundlagen  84 Grundlagen der Arithmetik (Frege)  151 GTR (see general theory of relativity) Guevara, Che  3 Guyer, P.  51 n.3 Haack, Susan  11, 294 Hager, Paul  173 n.4 Hahn, H.  123, 141 n.5, 144 n.41 Haller, R.  140 Hamlyn, David W.  304 n.9, 306 n.23 Hanks, Peter  116 n.1 Hardin, Russell  380 harmony of desires  343–4 Harris, H. S.  21 Hartshorne, C.  65 Hatcher, William  173 n.9 Haughton, Hugh  65 Hazen, Allen P.  175 n.23 Heanue, Jeames  93 n.3, 278 n.14 Hegel, G. W. F.  21, 371 Hegelianism  28–9 Hegel’s idealism  21 Helfand, I.  358 n.64 Hicks, G. Dawes  240 Hinde, R.  358 n.63 Hintikka, J.  54 nn.42–43 history as art  367–8, 371–4, 378–9 benefits  366–7 beyond literary approach  379–82 causal laws in  371 economic and literary approaches to  362–7 as education  373 essays on  367 facts and  363–4 as field  373 for general public  373 philosophy of  371 physical sciences to  371 Russell’s approach to  361–2 into science  362 scientific approach to  363–5, 368–70 on truth  374–9 A History of Western Philosophy  3, 4, 174 n.17, 231, 361, 367, 382 n.2 Hochberg, H.  354 n.7 holism  294

Index  421 Hook, Sidney  383 n.9, 384 n.15 Huemer, Michael  237 human behavior  71–2 Human Knowledge  2, 4, 11, 139, 141, 145 n.145, 249, 252, 254 n.23, 287–91, 296, 301, 303 nn.4, 6, 304 nn.10, 11, 348, 355 n.22, 358 n.56 Human Society in Ethics and Politics  12, 334, 342, 347, 348–51 Hume’s principle  85, 94 n.12 Husserl, Edmund  78, 278 n.14 Hylton, Peter  53 n.21, 94 n.17, 173 n.2, 260, 271, 273, 313 idealism  19 Berkeley’s  19 Hegel’s  21 Kant’s  19–21 post-peano philosophy and  43–51 Russell’s  17–18, 30–1 ideal language philosophy  188–9 if-then-ism  168–73 images  321 immediate experience  24 impredicative comprehension axioms  207 incompleteness problem of  185 theorem  165 incomplete symbols  188, 195–200 individual concept  92 induction, principle of  83–6, 104, 105 inferred entities  129 The Influence of Darwin on Philosophy (Dewey)  228 inner perception  10, 263–5 An Inquiry into Meaning and Truth  7, 133, 250, 306 n.27, 328 intelligence  228 intensional sentence  181 interpretation  64 Introduction to Mathematical Philosophy  115, 144 n.42, 152, 298 introspective attention  262 introspective judgments  263–5 introspective knowledge  256–60, 263 acquaintance  261–2, 267–8

awareness  262–3 inner perception  263–5 introspective judgments  263–5 knowing mind by acquaintance  260–3 intuitionism  127 intuitive knowledge  263 Investigations in Object Theory and Psychology  76, 90 Irvine, Andrew D.  173 n.6, 277 n.11, 306 n.27 Jacquette, Dale  14 n.5, 175 n.24 James, William  52 n.13, 59, 66, 257, 258, 261, 304 n.8, 307 n.32 Jeshion, Robin  261, 277 n.11 Jourdain, Philip  203 n.18 judgments  25–6, 29 (see also relational experience) considered  302 introspective  263–5 moral  302 neutral monism  322–5 Kaal, H.  155 Kallen, Horace  14 n.4 Kant, Immanuel  51 nn.3–5, 385 n.23 Kant’s idealism  19–21 Kaplan, David  195, 201 n.2 Kaplan-Geach sentence  169 Kasrils, R.  356 n.31 Keeling, S. V.  51 n.6 Keene, Carol A.  24 Kennedy, H. C.  353 Kerry, Benno  82 Kersten, F.  278 n.14 Keynes, John Maynard  63, 306 n.24 King, Jeffrey  201 n.2, 383 n.11 Klein, C.  142 Klement, Kevin  7, 14 n.5, 173 n.9, 174 n.18, 175 nn.22, 24, 277 n.8, 280 n.30 Kneale, M.  175 n.21 Kneale, W.  175 n.21 knowledge by acquaintance  10, 67, 102, 192, 194, 239, 241, 244, 256, 257, 314 by description of matter  10, 67, 102, 112, 192–4, 214, 239, 241, 244, 257–8, 314

422 Index descriptive  266–7 introspective  256–60, 263 intuitive  263 scientific  287 self  256–60 sensory  263, 264–5 theory of  64, 88–9, 257 of truths  102, 257 Korhonen, Anssi  174 n.17 Kornblith, Hilary  235 n.21 Kraal, Anders  175 n.22 Krauss, Lawrence  235 n.22 Kremer, Michael  257, 280 n.36 Kripke, Saul  187, 201 n.2, 202 nn.9, 10, 12, 355 n.10 Kvanvig, Jonathan  304 n.14 Lakatos, Imre  306 n.23 Landini, Gregory  8, 112, 116 n.1, 118 nn.10, 12, 15, 173 nn.8, 9, 10, 174 n.17, 175 nn.23, 24–25, 203 n.18, 233 n.4, 234 nn.6, 16, 275, 277 nn.6, 12, 279 n.26, 281 n.39, 331 n.1 language, analytical philosophy of  179 Law V  81 Lehrer, Keith  304 n.14 Leibniz, G. W.  151 Lenzen, Victor F.  63, 65, 112 Levine, James  52 n.9, 72 n.11 Lewis, C. I.  72 n.7 Link, Godehard  173 n.7, 306 n.27 Linsky, Bernard  6, 14 n.5, 72 n.6, 93 n.3, 94 nn.7, 16, 118 n.11, 120 n.22, 170, 173 n.9, 175 n.23, 234 nn.6, 10, 253 n.2, 257, 273, 277 nn.7, 9, 12, 278 n.15, 280 nn.30, 36 Liu, JeeLoo  279 n.23 logic advanced  64 axiomatic presentation  125 as essence of philosophy introduction  205–11 Our Knowledge  223–8 problems  211–23 reconstruction in philosophy  228–33 logical forms of  208, 212–13, 219

propositions of  115, 135–6 logical analysis  80–1 logical atomism  205, 209 Logical Atomism  209 logical constants  136 logical constructions advantage of  272 body of knowledge  273 of mind  273–4 self  271–4 sense-data  241–5 logical empiricism  122–3 logical forms of logic  208, 212–13, 219 philosophy of language  187–93 logical laws  125 logical positivism  2 logical propositions  125 The Logical Syntax of Language (Carnap)  127, 128, 136, 143 n.23, 146 n.57 logical truth  166, 168 logicism  124–8, 151–75 axioms  168–73 definition  151 foundation  208 Gödel’s results and scope of  164–8 if-then-ism  168–73 paradoxes and development of  156–60 and philosophy  152–3 in Principia Mathematica  160–4 regressive method  153–6 logicist thesis (LT)  165–7 Logic (Whately)  92 Logische Aufbau der Welt (Carnap)  7 Lowell Lectures  107 Ludlow, Peter  277 n.11 Lycan, William  294, 304 n.14 MacBride, Fraser  332 n.13 Mach, Ernst  261 Machuca, Diego  271 Mackie, J. L.  335 Maclean, Gülberk K.  281 n.38 Madigan, Timothy J.  383 n.10 Majer, U.  71 Mally, Ernst  90, 93 n.3 Marriage and Morals  3

Index  423 Marsh, R. C.  143 n.30 Martinich, A. P.  271, 277 n.11 Marx, Karl  371 Massey, G. J.  137 mathematical functions  76–7, 81–2 mathematical logic  50–1 mathematical truths  166–8, 171–2 Mathews, Debra J. H.  279 n.23 McAlister, L. L.  258, 278 n.14 McCain, Kevin  11, 294, 305 n.17 McGrew, Timothy  305 n.16 McGuinness, B.  94 n.10, 117 n.4, 142 n.11, 144 n.41, 155, 231 McLaughlin, Brian  277 n.11 McTaggart, J. E. M.  22 McTaggart, J. M. E.  51 n.6 Meinong, Alexius  75, 93 n.3, 278 n.14 Russell’s agreement with  86–9 Mellor, D. H.  62–3 Mendelson, Elliot  174 n.15 mental prejudices  50–1 mental sensation  237, 239, 245–50 metaphysics, post-idealist  33–6 methodological solipsism  132 Milhaud, Gaston  253 n.5 Mill, John Stuart  1 Mind  75, 76 mind-dependent entities  37, 273–4 Minton, A.  356 n.42 Misak, Cheryl  6, 72 n.3 Moggridge, D. E.  63 monadism  31 monism  31, 67–8 The Monist  70 Monk, Ray  13 n.1, 117 n.3 Moore, G. E.  2, 12, 14 n.5, 17, 51 n.1, 53 nn.23, 25–26, 54 nn.29–31, 33, 55 n.55, 62, 237, 252 n.1, 307 n.32, 331 n.6, 355 nn.11, 19 Moore, G. H.  54 nn.42–43 Moorean philosophy  33 analysis  39 business of  39–40 epistemology  36–8 meaning  38 post-idealist metaphysics  33–6 on time, magnitude, and number  40–3 understanding  38–9

Moorean realism  5–6 Moorehead, Caroline  13 n.1, 363, 380 moral judgments  302 moral obligation  337–8 moral philosophy  334–5 analysis and moral concepts  335–7 ethical subjectivism (1913-44) early efforts  338–40 emotivism (1925–34)  342–8 error theory  340–1 Human Society in Ethics and Politics  348–51 moral obligation  337–8 Russell’s ethics background  351–2 wickedest people  352–3 Morrell, Ottoline  61, 102, 106, 117 nn.2, 4 Moser, Paul K.  304 n.9, 305 nn.15, 16 Mosterin, J.  143 n.21 Mulder, H.  146 n.62 multiplicative axiom  164, 170 Murray, Gilbert  4, 277 n.10 Musgrave, Alan  175 n.22 My Philosophical Development  66, 115, 227, 238, 249–50, 253 n.14 Mysticism and Logic  234 n.7 Nagasawa, Yujin  281 n.39, 331 n.11 Nakhnikian, George  254 n.18 Nasim, Omar W.  277 n.5 Naturalism and Agnosticism (Ward)  331 n.3 naturalistic fallacy  335–6 natural languages  188, 190 natural numbers  154 Neale, Stephen  186, 203 n.20 Neta, Ram  305 n.16 Neurath, M.  140 Neurath, O.  123, 140 neutral monism  10–11, 67–8, 274–6 arguments against  315–19 images and emphatic particulars  318, 325–8 introduction  312–15 judgments  322–5 and physics  328–30 sensation  319–22 non-demonstrative inferences  138, 139

424 Index non-existential propositions  37 Notes on Logic  100, 105, 110, 111, 113–16 Nuclear Nonproliferation Treaty (1968)  352 number and analysis  45–6 object-dependent propositions  199 object-independent propositions  199 objects versus concepts  76–7 Occam’s razor  123, 129, 145 n.49, 233, 313, 317, 320, 325, 329, 330, 341, 350, 355 n.20 Ogden, C. K.  72 n.13 Oliver, Alex  175 n.24 Olson, Dustin  303 n.1 Olson, Eric  254 n.19, 273, 306 n.27 “On Denoting”  75, 76, 90, 93, 239 “Russell’s Principle” from  87–8 target of  75 “On Matter,” manuscript  101–6 ontological arguments  28–9 open question argument  336 ordinal numbers  163–4 organic unities  33–4, 36 “ought to do”  337 Our Knowledge of the External World as a Field for Scientific Method in Philosophy (1914)  8, 63, 100, 123, 128–35, 143 n.27, 205, 210–11, 212–13, 223–8, 229–30, 238, 242–3, 253 n.8, 280 nn.31–33, 36, 320 Outline of Philosophy  47, 50, 208, 210, 247, 281 n.39, 286, 330, 342, 344 paradox of classes (Russell)  155, 156–60 Parfit, D.  354 n.6 Parsons, Charles  306 n.23 Partial Test Ban Treaty (1963)  352 Pashby, Thomas  331 n.1 Pastin, Mark  304 n.9 Peano, Giussepi  2, 17, 77, 151 axiom  171–2 notation  77 postulates  154 Peano-Dedekind axioms  153–4, 162–4 Pears, David  116 n.1, 142 n.11, 231, 271, 274, 304 n.9, 305 n.20

Peirce, Charles Sanders  61, 65, 77 perceptions  64, 66–72, 251–2 inner  263–5 pragmatism  66–72 truths versus judgments  263–4 of Venus  291 Perkins, Raymond  12, 354 n.7, 355 nn.16, 21, 27, 357 n.44, 358 nn.62, 66 Perovic, Katarina  278 n.15 Perry, John  201 n.2, 279 n.23, 315 Perry, Ralph Barton  52 n.13, 65, 66, 112, 261 Philosophical Essays  6, 338 philosophical method, Ward’s view on  26–7 philosophical skepticism  102–3 philosophy of language  179 descriptions incomplete symbols  195–200 problems with  193–200 Russell’s theory of  183–7 semantics versus epistemology  193–5 ideal  188–9 logical form, metaphysics, and semantics  187–93 Russellian propositions  180–3 Philosophy of Logical Atomism  272–4, 280 nn.31–33, 319, 321, 323 Physikalische Begriffsbildung (Carnap)  131 Pigden, Charles  334, 338, 341, 345, 346, 350, 354 nn.1–2, 8, 355 n.11, 356 n.34, 357 nn.44, 48 Pincock, Christopher  10, 116 n.1, 118 n.10, 119 n.16, 261, 275, 276, 279 n.25, 281 n.39, 332 n.13 Poincaré, Henri  4, 297, 298 Pojman, Louis  304 n.14 Pollock, John  290 Portraits from Memory  367 positivism  2 Posterior Analytics (Aristotle)  290 post-idealist metaphysics  33–6 post-peano philosophy idealism, critique of  47–51 logical construction and acquaintance  46–7 number and analysis  45–6

Index  425 set theory and justification  43–5 Potter, Michael  94 n.17, 117 n.3, 119 nn.17, 20, 355 nn.18, 25, 356 nn.33, 34 Power: A new Social Analysis  356 n.40 The Practice and Theory of Bolshevism  3, 367 pragmatism  59–72 antipathy to  61–2 behavior  71–2 belief  66–72 perception  66–72 Ramsey  62–3 works and  60 Pragmatism: A New Name for Some Old Ways of Thinking (James)  59 preceding  34–5 “The present King of France” objection  89–93 Present Philosophical Tendencies (Perry)  315 Principia  123, 124–8, 129, 198, 206–9, 226–7, 335, 337 Principia Ethica  334, 354 n.2 Principia Mathematica  2, 11, 62, 76, 80, 85, 86, 122, 125, 142 n.9, 152, 184, 196, 205, 206, 271, 296, 313 logicism in  160–4 Principles of Mathematics  2, 79, 81, 86, 111, 152, 174 n.17, 180, 193, 296, 313–14, 335 Principles of Psychology (James)  66 Principles of Social Reconstruction  358 n.57 problem of incompleteness  185 Problems  347 Problems of Philosophy  7, 9, 102, 105, 205, 206, 214–23, 237, 241–2, 257–64, 268–71, 277 nn.6, 13, 278 n.15, 279 n.22, 312, 315, 331 n.6, 351, 355 n.22 Proops, Ian  54 n.32, 72 n.9, 173 n.3, 253 n.2, 258, 276 n.3, 277 nn.9, 12 propositional calculus  125 propositional functions  159 propositions  34–6 of logic  115, 135–6 of mathematics  135–6 object-dependent  199

object-independent  199 Russellian  180–3 unity of  180–1 protocol sentences  137, 139, 140 proto-emotivism  338, 342 Putnam, E.  137 Putnam, Hilary  137, 175 n.22 qualia  233 quantification, syntactic theory of  183 quasi analysis  131 Quine, W. V.  2, 169, 188, 202 n.8, 234 n.19 Rabins, Peter V.  279 n.23 Ramsey, Frank  62–3, 71, 355 n.19 Rancurello, A. C.  258, 278 n.14 Rang, Bernhard  78 rational reconstruction  135 Rawls, John  300, 307 nn.30, 31 Rayo, Agustín  175 n.20 reality, Ward’s view on  26–7 Reck, Erich  94 n.6, 142 n.9 Reconstruction in Philosophy (Dewey)  228 recursive functions  165 Reed, Baron  271 reflective equilibrium, regressive method  300–3 Regan, Tom  354 n.7 regressive method, logicism  153–6, 296–303 Reidel, D.  123 relational experience  24–6 relation numbers  163 “The Relation of Sense-Data to Physics”  242–3, 245, 253 n.12 relativity, contradiction of  31–3 Religion and Science  12, 334 Rescher, Nicholas  71, 304 nn.9, 14, 305 n.21 Revue de métaphysique et de morale (Russell)  297 Ricketts, Tom  94 n.17 “right,” sense of  338 Roselló, Joan  175 n.21 Rossberg, M.  126, 151, 154, 155 Rotblat, J.  352, 358 n.63 Ruse, M.  356 n.42

426 Index Russell, Bertrand  14 n.5 advocate  3 agreement with Meinong  86–9 approach to history  361–82 awards  3 on denoting and language  179–203 dialectic and  30–1 disagreements with Frege  76–80 ethics  351–3 at Harvard (1914)  63–6 as idealist  30–1 on introspection and selfknowledge  256–81 learning from Frege  80–6 logicism  151–75 Moorean philosophy  33–43 moral philosophy  334–58 Paradox  76–81 in philosophical circles  4 philosophy  4–5 post-peano philosophy and idealism  43–51 and pragmatists  59–72 propositions  180–3 public intellectual  2–3 radical  3 in the Soviet Union  3–4 theory of descriptions  183–7 on truth  374–9 to Vienna Circle and back  122–46 with Wittgenstein  99–120 works  1–14, 97–8 Russell, Carr  321 Russellianism  180–1, 182, 183 Rutte, H.  140 Ryle, Gilbert  254 n.14 Sainsbury, R. Mark.  253 n.17, 254 n.20, 271, 273, 304 n.9 Salmon, Nathan  201 n.2, 254 n.18 Santayana, G.  354 n.3 Sartre, Jean-Paul  278 n.14 Savage, C. Wade  254 n.18 scepticism  288 Schiller, F. C. S.  61 Schilpp, Paul Arthur  142 n.9, 254 n.21, 347, 356 nn.29, 41, 383 n.9 Schlick, M.  122, 123, 140, 146 n.62 Schmitz, François  7

Schoenman, Ralph  175 n.22 Schröder, Ernst  77, 78 scientific knowledge  287 Scientific Method in Philosophy  213 scientific methods, in philosophy  49–50 scientific philosophy  124 self logical constructions  271–4 and mind  273–4 self-acquaintance  266–71, 275 “self and not-self ” presentation  24, 26, 36, 46–7 self-consciousness  262, 270 self-descriptions  267–8, 270–1, 275 self-evidence  44, 291 self-evident propositions  27 self-knowledge  256–60, 266–7, 275 Sellars, Wilfred  304 n.14 semantic theories  180–2 of definite descriptions  195 of descriptive content  195 versus epistemology  193–5 natural language  187–8 sensation  237, 239, 241 neutral monism  319–22 sense-data  9–10, 103–5, 106, 272 acquaintance and  315 acquaintance versus description  239 epistemological role of  239 external  258 images versus  321 introduction of  238–41 logical constructions  241–5 as logical function  244 mental sensation and  237, 239, 245–50 mind and  251 physical objects  241–5, 251 rejection  245–50 Russell’s epistemology after rejection of  251–2 sensibilia versus  245 standard view of  237–8 as subjective  239–40 sense versus reference, of expression  182–3 sensibilia  238, 243, 272 sense-data versus  245 set theory and justification  43–5

Index  427 Seubold, J. P.  151 Shapiro, Stewart  175 n.20 Shaw, W.  354 n.6 Sidgwick, Henry  22 Simons, G.  174 n.16 simple type theory  159 Singer, P.  357 n.51 singular terms  92 Skrupskelis, I. K.  66 Slater, John G.  298, 306 n.27, 323–6 Sleeper, R.  354 n.7 Smart, J. J. C.  354 n.6 Smeaton, A.  159 Smiley, Timothy  175 n.24 Smith, Alys Pearsall  1, 380 Smith, G.  65 Smith, Janet Farrell  87 Soames, Scott  175 n.21, 243, 253 n.16, 277 n.11 sociologism  139 solipsism  103 Some Main Problems of Philosophy (Moore)  315 Sosa, David  271, 277 n.11 Sosa, Ernest  307 n.31 Spadoni, Carl  52 n.15 Spencer, Herbert  228 Spengler, Oswald  371 Stanford Encyclopedia of Philosophy  237 START treaties  352 Steiner, Mark  306 n.23 Steup, Matthias  307 n.31 Stevens, Graham  116 n.1, 201, 201 nn.1–4, 202 n.9 Stevenson, C. L.  345 Stock, G.  52 n.9 Stone, Peter  12, 385 n.25 Stout, G. F.  22, 277 n.5 structural description  133 Stubenberg, Leopold  281 n.39 Studies in the Hegelian Dialectic (1896)  22 Stunkel, Kenneth R.  363, 367, 368, 382 nn.2–3, 383 n.5, 384 nn.13, 20 “subject” and “object” presentation  24, 27, 36, 37 subjectivism early efforts  338–40

emotivism (1925–34)  342–8 error theory  340–1 Human Society in Ethics and Politics  348–51 subsets  79 substitution theory  111, 196–8 subsumption notion  78 successor  154 Suchting, W. A.  21 supra-relational experience  26 (see also absolute) Sylvester, R. P.  354 n.7 symbolical entities  36 TAK (see traditional account of knowledge) tautology  115–16, 125 Tennant, Neil  175 n.21 Terrell, D. B.  258, 278 n.14 Textor, M.  332 n.13 theorem of complete induction  84–5 theory of concept construction  132 of judgment Bradley’s  25, 28–9 existential  29, 35, 52 n.16 Wittgenstein’s criticism  99, 107–8, 111, 113–16 of knowledge  64, 88–9, 257 Theory of Complexes and Assumptions  87, 88 Theory of Knowledge  6, 10, 100, 106–11, 144 n.43, 206, 209–13, 234 nn.8, 13, 257, 258, 260–4, 269–71, 276 nn.1, 4, 278 nn.15–16, 312, 315, 316, 321, 323, 331 n.2 A Theory of the Good and the Right  357 n.51 Thomas, W.  78 Thompson, Michael  319 Tractatus (Wittgenstein)  2, 62, 72, 99, 100, 114, 123, 137, 142 n.11, 169, 230 traditional account of knowledge (TAK)  287–8 Trevelyan, George  12, 362–9, 371–82, 382 nn.3–4, 383 nn.6–8, 384 nn.18–19, 385 n.28 Trumpbour, J.  358 n.62

428 Index truth-functional statement operators  160 truths history on  374–9 knowledge of  257 pragmatist theory of  59–61, 232 Tully, Robert E.  281 n.39, 331 n.1 Turing-computable functions  165 Über Annahmen  86–7 “The Ultimate Constituents of Matter”  243, 250 A Universe from Nothing (Krauss)  235 n.22 universe of discourse  78–9 Untersuchungen zur allgemeinen Axiomatik (Carnap)  143 n.21 Urmson, J. O.  354 n.6 utilitarianism  335 Value and Obligation  357 n.51 van de Velde-Schlick, B.  146 n.62 Van Heijenoort, Jan  77 Venus, perceptions of  291 verificationism  137–40 Vianelli, G.  55 n.51 Vienna Circle Aufbau  128–35 logical empiricism  122–3, 124 logicism  124–8 mathematical logic  124 members  122 OKEW  128–35 Principia  124–8 protocol sentences  137, 139, 140 Russell and  135–41 von Wright, G. H.  234 n.14 Wahl, Russell  116 n.1, 118 n.9, 234 n.12, 253 n.2, 261, 272, 277 n.12, 281 n.39 Waismann, F.  123

Walden, Kenneth  307 n.31 Walter, Sven  277 n.11 Ward, James  22, 52 nn.11, 14, 331 n.3 Watson, J. B.  70 Weidlich, Thom  14 n.4 Weiss, P.  65 Welby, Lady Victoria  63 Wettstein, Howard  201 n.2 Whately, Richard  92 Whitehead, Alfred North  1, 2, 14 n.5, 152 Wiener, Norbert  80 Wiener, Philip P.  307 nn.31–32 William, Ramsey  290 Williams, C. M.  261 Williams, Michael  290, 291 Willis, Kirk  382 n.3, 383 nn.5, 12 The Will to Believe (James)  59 will-to-believers  68 Winds of Doctrine (Santayana)  334 Wishon, Donovan  9, 14 n.5, 234 nn.6, 10, 253 n.2, 257, 273, 277 nn.7, 9, 12, 278 n.15, 279 n.26, 280 nn.36, 37, 281 n.39, 330, 331 nn.1, 11 Wittgenstein, Ludwig criticisms of Russell  111–16 matter of matter (1911–12)  101–6 Russell with  99–120 Theory of knowledge (1913)  106–11 Wood, A.  51 n.3 Woods, James  65 works, pragmatism and  60 Worral, John  306 n.23 Yi, Byeong-uk  170 Zalta, Edward N.  237, 277 n.11, 281 n.39 Zermelo, Ernst  77, 78–9 Zermelo sequence  163, 164 zero  154

429

430