Leibniz on Causation and Agency 1107192676, 9781107192676, 9781108131629

Table of contents :
Cover
Half-title page
Title page
Copyright page
Contents
List of Tables
Acknowledgments
Note on Translations and Citations
List of Abbreviations
Introduction
1 Monads and Their Actions
2 Spontaneity
3 Teleology
4 Attributability and Divine Concurrence
5 Freedom
6 Control, Weakness, and Compulsion
7 Moral Agency
References
Index

Citation preview

Leibniz on Causation and Agency

This book presents a comprehensive examination of Gottfried Leibniz’s views on the nature of agents and their actions. Julia Jorati offers a fresh look at controversial topics including Leibniz’s doctrines of teleology, the causation of spontaneous changes within substances, divine concurrence, freedom, and contingency, and also discusses widely neglected issues such as his theories of moral responsibility, control, attributability, and compulsion. Rather than focusing exclusively on human agency, she explores the activities of nonrational substances and the differences between distinctive types of actions, showing how the will, appetitions, and teleology are key to Leibniz’s discussions of agency. Her book reveals that Leibniz has a nuanced and compelling philosophy of action that has relevance for present-day discussions of agency. It will be of interest to scholars and students of early modern philosophy as well as to metaphysicians and philosophers of action. julia jorati is Assistant Professor of Philosophy at the Ohio State University. She has published numerous articles on Leibniz’s metaphysics, philosophy of mind, and ethics, in publications including the Journal of the History of Philosophy, The Leibniz Review, Philosophy Compass, and several edited volumes.

Leibniz on Causation and Agency Julia Jorati The Ohio State University

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 4843/24, 2nd Floor, Ansari Road, Daryaganj, Delhi – 110002, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107192676 DOI: 10.1017/9781108131629 © Julia Jorati 2017 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2017 Printed in the United Kingdom by Clays, St Ives plc A catalogue record for this publication is available from the British Library. ISBN 978-1-107-19267-6 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Contents

List of Tables Acknowledgments Note on Translations and Citations List of Abbreviations Introduction 1 Monads and Their Actions

page vi vii viii ix 1 8

2 Spontaneity

37

3 Teleology

59

4 Attributability and Divine Concurrence

92

5 Freedom

114

6 Control, Weakness, and Compulsion

148

7 Moral Agency

180

References Index

208 217

v

Tables

1 2 3

vi

Three Types of Spontaneity Types of Spontaneity and Teleology Types of Necessity

page 58 77 128

Acknowledgments

Many people have made direct or indirect contributions to this project. First and foremost, I am grateful to Michael Della Rocca, whose many comments over the years have shaped and sharpened my thoughts about Leibniz. I also thank Bernd Ludwig, who first introduced me to early modern philosophy and to Leibniz when I was an undergraduate. Marleen Rozemond, Paul Lodge, Donald Rutherford, and Stephan Schmid provided extensive comments on the book manuscript and prompted me to make many improvements, for which I am tremendously thankful. Moreover, I am grateful to all the others who have given me feedback at various stages: Chloe Armstrong, Christian Barth, Sebastian Bender, Gregory Brown, Julia Borcherding, John Carriero, Lisa Downing, Tom Feeney, Juan Garcia, Ursula Goldenbaum, Sean Greenberg, John Hare, Verity Harte, Glenn Hartz, Larry Jorgensen, Sukjae Lee, Antonia LoLordo, Jeffrey McDonough, Brian McLean, Sam Newlands, Robert Pasnau, Sydney Penner, Dominik Perler, Kristin Primus, Kelley Schiffman, Tad Schmaltz, Lisa Shabel, Stewart Shapiro, Sun-Joo Shin, Allan Silverman, Alison Simmons, Robert Sleigh, Neil Tennant, John Whipple, and Kenneth Winkler. Of course, I am solely to blame for any mistakes and imperfections in this book – much like Leibnizian agents are responsible for all privations in their actions, as we will see in Chapter 4. Finally, I thank my partner Hadi Jorati, who encouraged me every step of the way. There is some overlap between this book and material already published elsewhere: Chapters 2 and 3 contain some of the material from Jorati (2015) and (2013); Chapter 4 contains some material from von Bodelschwingh (Jorati) (2011). I thank the publishers for their permission to use this material.

vii

Note on Translations and Citations

If an English edition is explicitly cited (or associated with an abbreviation), translations are taken from that English edition unless otherwise specified. In all other cases, translations are mine. Quotations will include all italics from the original text, unless otherwise specified.

viii

Abbreviations

(a) A

AG Beeley C CD COE

CP

DM DPG Dut E ET FR

Texts by Leibniz and Editions of Leibniz Texts Sämtliche Schriften und Briefe. Ed. Deutsche Akademie der Wissenschaften. Darmstadt, Leipzig, Berlin: Akademie Verlag, 1923–2015. Cited by series, volume, page. Philosophical Essays. Ed. and transl. Roger Ariew and Daniel Garber. Indianapolis: Hackett, 1989. “Leibniz on Wachter’s Elucidarius cabalisticus” (1706). Ed. Philip Beeley. The Leibniz Review 12 (2002), 1–11. Opuscules et fragments inédits de Leibniz. Ed. Louis Couturat. Paris: F. Alcan, 1903. Causa Dei, appended to the Theodicy (1710). Cited by section number as in G 6:439–62. “Observations on the Book Concerning the Origin of Evil,” appended to the Theodicy (1710). Cited by section number as in G 6:400–36. Confessio Philosophi: Papers concerning the Problem of Evil, 1671–1678. Ed. and transl. Robert Sleigh. New Haven: Yale University Press, 2005. “Discourse on Metaphysics” (1686). Cited by section as in A 6.4.1529–88; translation from AG 35–68. Dissertation on Predestination and Grace. Ed. and transl. Michael J. Murray. New Haven: Yale University Press, 2011. Opera Omnia. 6 vols. Ed. Louis Dutens. Geneva: Fratres de Tournes, 1768. Cited by volume, part, and page. Opera Philosophica. 2 vols. Ed. Johann E. Erdmann. Berlin: Eichler, 1839–40. Cited by volume and page. “Excursus on Theodicy §392” (1711), G 6:347–50. “Preliminary Dissertation on . . . Faith and Reason,” published with the Theodicy (1710). Cited by section number as in G 6: 49–101.

ix

x

G

GLW GM

Gr

H L LC

LDB

LDV LGR LSC

M Mason Mollat MP NE

List of Abbreviations

Die philosophischen Schriften von Gottfried Wilhelm Leibniz. 7 vols. Ed. Carl Immanuel Gerhardt. Berlin: Weidmann, 1875–90. Reprinted Hildesheim: Georg Olms, 1978. Cited by volume and page. Briefwechsel zwischen Leibniz und Christian Wolff. Ed. Carl Immanuel Gerhardt. Halle: Schmidt, 1860. Leibnizens Mathematische Schriften. 7 vols. Ed. Carl Immanuel Gerhardt. Berlin: Asher, 1849–63. Reprinted Hildesheim: Georg Olms, 1971. Cited by volume and page. Textes inédits d’après des manuscripts de la Bilbliothèque provinciale d’Hanovre. 2 vols. Ed. Gaston Grua. Paris: Presses Universitaires de France, 1948. Theodicy. Transl. E.M. Huggard. La Salle: Open Court, 1985. Philosophical Papers and Letters. Ed. and transl. Leroy Loemker. Dordrecht: Reidel, 1969. The Leibniz-Clarke Correspondence (1715–16). G 7:352–440; cited by number of letter and number of paragraph [e.g., LC 5.2: letter 5, paragraph 2]; translation from Correspondence. Ed. and transl. Roger Ariew. Indianapolis: Hackett, 2000. The Leibniz-Des Bosses Correspondence. Ed. and transl. Brandon Look and Donald Rutherford. New Haven: Yale University Press, 2007. The Leibniz-De Volder Correspondence. Ed. and transl. Paul Lodge. New Haven: Yale University Press, 2013. Leibniz on God and Religion: A Reader. Ed. and transl. Lloyd Strickland. London: Bloomsbury, 2016. The Leibniz-Stahl Controversy. Ed. and transl. François Duchesneau and Justin E.H. Smith. New Haven: Yale University Press, 2016. “Monadology” (1714). Cited by section as in G 6:607–23; translation from AG 213–25. The Leibniz-Arnauld Correspondence (1686–1690). Ed. and transl. Haydn T. Mason. New York: Barnes & Noble, 1967. Mittheilungen aus Leibnizens ungedruckten Schriften. Ed. Georg Mollat. Leipzig: Haessel, 1893. Philosophical Writings. Ed. and transl. Mary Morris and George H.R. Parkinson. London: J. M. Dent, 1973. New Essays on Human Understanding (1704). Cited by page numbers from A.6.6; translation from New Essays on Human Understanding, transl. Peter Remnant and Jonathan Bennett. New York: Cambridge University Press, 1996.

List of Abbreviations

“On Nature Itself” (1698). Cited by section as in G 4:504–16; translation from AG 155–67. “Principles of Nature and Grace, Based on Reason” (1714). Cited by section as in G 6:598–606; translation from AG 206–13. Philosophical Texts. Ed. Roger Woolhouse and Richard Francks. New York: Oxford University Press, 1998. “Reflections on Hobbes,” appended to the Theodicy (1710). Cited by section number as in G 6:388–99. Political Writings. Ed. and transl. Patrick Riley. Cambridge: Cambridge University Press, 1972. Malebranche et Leibniz: Relations Personnelles. Ed. André Robinet. Paris: Vrin, 1955. The Shorter Leibniz Texts: A Collection of New Translations. Ed. and transl. Lloyd Strickland. New York: Continuum, 2006. Theodicy (1710). Cited by section number as in G 6:102–365. “Summary of the Controversy Reduced to Formal Arguments,” appended to the Theodicy (1710). G 6:376–87. A Collection of Several Pieces of Mr. John Toland, London: Peele, 1726. [Contains a short Leibniz text as an appendix.] Preface to the Theodicy (1710). G 6:25–48. Leibniz Selections. Ed. and transl. Philip Wiener. New York: Scribner, 1951. Leibniz’s “New System” and Associated Texts. Ed. and transl. Roger Woolhouse and Richard Francks. New York: Oxford University Press, 1997.

ONI PNG PT RH Riley Robinet SLT T Ta Toland Tp W WF

(b)

xi

Texts by Authors Other Than Leibniz

AT Concordia

CT

De Angelis

René Descartes, Oeuvres de Descartes. 11 vols. Ed. Charles Adam and Paul Tannery. Paris: Vrin, 1996. Cited by volume and page. Luis de Molina, Liberi arbitrii cum gratiae donis, divina praescientia, providentia, praedestinatione et reprobatione concordia. Ed. Johannes Rabeneck. Oniae: Collegium Maximum, 1953 [1588]. Cited by part, disputation, and paragraph. Thomas Aquinas, Corpus Thomisticum: Sancti Thomae de Aquino Opera Omnia. Ed. Enrique Alarcon. www.corpustho misticum.org/iopera.html Francisco Suárez, De Angelis. In Opera omnia. 28 vols. Ed. Carolo Berton. Paris: Vivès, 1856–78, vol. 2. Cited by chapter and section.

xii

DV MD

OCC

OEC

SCG ST

List of Abbreviations

Thomas Aquinas, De veritate. In CT. Cited by question and article. Francisco Suárez, Metaphysicarum disputationum. In Opera omnia. 28 vols. Ed. Carolo Berton. Paris: Vivès, 1856–78, vols. 25–26. Cited by disputation, section, and paragraph. Francisco Suárez, On Creation, Conservation, and Concurrence: Metaphysical Disputations 20, 21, and 22. Ed. and transl. Alfred J. Freddoso. South Bend: St. Augustine Press, 2002. Francisco Suárez, On Efficient Causality: Metaphysical Disputations 17, 18, and 19. Ed. and transl. Alfred J. Freddoso. New Haven: Yale University Press, 1994. Thomas Aquinas, Summa contra gentiles. In CT. Cited by book, chapter, and paragraph. Thomas Aquinas, Summa Theologiae. In CT. Cited by part, question, and article.

Introduction

Gottfried Leibniz left behind a corpus of writings that is impressive both in size and in breadth: it contains discussions of an enormous array of areas of inquiry, both inside philosophy and out. One comparatively neglected portion of this corpus is Leibniz’s contribution to what we would today call ‘the philosophy of action.’ That is unfortunate because Leibniz’s discussions of agency are sophisticated and often compelling. In fact, they are far more compelling than one would expect, given the notorious eccentricity of Leibniz’s metaphysics. To mention just a few examples of his eccentric metaphysical doctrines, take the claim that our minds do not, strictly speaking, interact with our bodies or the denial of causal interaction among finite substances. Or consider the doctrines that every substance perceives everything that happens in the entire universe and that the only ultimately real things are immaterial. Or take, finally, Leibniz’s rehabilitation of substantial forms and teleology, which are rejected almost universally by modern philosophers. All of these idiosyncrasies may suggest that Leibniz’s views on action are bound to be implausible and useless for advancing our understanding of agency. Yet, we will see that quite the opposite is the case. The philosophy of action, as practiced in the twentieth and twenty-first centuries, analyzes a broad range of philosophical issues surrounding the notion of agency. These issues include the freedom of the will, shared agency, moral responsibility, the distinction between things that we do and things that merely happen to us, what it means to possess control over one’s actions, and whether it is possible to act against one’s better judgment. As this book shows, Leibniz discussed all of these topics. Many of his views on agency are directly relevant to present-day debates, and we can learn a number of things about agency from him.1 Leibniz’s discussions of agency are more subtle and insightful than those of most (if not all) of his contemporaries. And this is no coincidence: his 1

Analytic philosophers working in the philosophy of action are not typically aware of the fact that Leibniz made important contributions to their field. For instance, a recently published companion to the philosophy of action (O’Connor and Sandis 2010) contains chapters on the views of six early modern figures, but Leibniz is not among them. A notable exception is Susan Wolf, who explicitly describes part of her project as inspired by Leibniz (1990: 103).

1

2

Introduction

metaphysical idiosyncrasies force him to pay particularly close attention to distinctions that his contemporaries simply take for granted. In many cases, this leads Leibniz to theories that are far less eccentric than their metaphysical foundations and that possess significant plausibility and explanatory power. One example is Leibniz’s compatibilist theory of freedom. While other early modern compatibilists pay very little attention to internal impediments to free agency,2 Leibniz’s denial of inter-substance causation prompts him to take this type of impediment extremely seriously. This, in turn, makes his account of freedom, and his moral psychology more generally, far more powerful. The resulting theory of freedom is an intriguing combination of agent-causal views, the doctrine that being free means being determined by the good, and the doctrine that free actions have to flow from the agent’s real self. There are other cases where Leibniz arrives in familiar territory from extremely eccentric points of departure. One is the distinction between what we today call ‘autonomous agency’ and ‘nonautonomous agency’; another is the closely related distinction between acting and being acted upon. Leibniz discusses these distinctions in terms of self-determination and end-directedness. Actions that we would describe as autonomous are self-determined and end-directed in a more demanding way than other actions. Similarly, active states differ from passive states in their self-determination and end-directedness. Leibniz’s starting point is his idiosyncratic claim that finite substances do not, in metaphysical strictness, interact causally with each other. Because this doctrine makes it difficult to distinguish between activity and passivity, Leibniz is forced to be particularly attentive to the different ways in which states can originate in a finite substance. That scrutiny pays off: Leibniz manages to isolate factors within an agent that can undermine agency and autonomy. In fact, Leibniz’s account resembles that of several prominent philosophers of action in our own day. Some advantages of Leibniz’s solutions arise directly from his peculiar metaphysical commitments and are hence unlikely to command broad appeal. For instance, some of the particularly attractive aspects of his theory of freedom depend on his doctrine that finite substances do not interact. Other advantages, however, do not depend on eccentric Leibnizian commitments. For example, Leibniz’s accounts of control, weakness of will, compulsion, and moral responsibility – or, at least, their most central features – are compatible with a wide range of metaphysical systems. These accounts are more likely to be appealing to contemporary philosophers of action. Yet, it is fascinating and useful to explore both types of advantages. Studying Leibniz’s theory of freedom, for instance, is helpful in part because it illustrates the costs of securing a particularly demanding type of independence from external determination. 2

See Gary Watson, who argues that this is a widespread problem among classical compatibilists (2004: 164).

Introduction

3

Let me make a few remarks on the book’s scope and methodology. First, it is not my goal merely to mine Leibniz’s writings for doctrines that might be useful to contemporary philosophers of action. Instead, the book attempts to understand Leibniz on his own terms and within his historical context, unconstrained by the prospective utility of the resulting interpretation for contemporary philosophy. This often requires looking at Leibniz’s predecessors and contemporaries. While I do think that many of Leibniz’s views on agency are promising and helpful, defending their viability is not part of my project. Relatedly, the book does not explore in detail the similarities and differences between Leibniz’s theory of action and the theories of more recent philosophers. Even though that is a worthwhile project, it would distract too much from my primary aim of providing an interpretation of Leibniz’s views on agency in their philosophical and historical context. Hence, the book merely mentions some connections between Leibniz’s doctrines and the contemporary philosophy of action, especially when these connections can help us understand Leibniz’s views. Given the book’s methodology, it may seem worryingly anachronistic to talk of a Leibnizian philosophy of action in the first place. After all, Leibniz does not appear to use the term ‘philosophy of action’ himself, nor was it a commonly acknowledged subfield of philosophy until the twentieth century. Yet, that does not make this book’s project anachronistic. It is clear, after all, that Leibniz answers many of the questions that we today associate with the philosophy of action; it is similarly clear that these questions are closely interrelated for Leibniz. Hence, treating Leibniz’s answers to those questions as a unit is not only legitimate but also natural and useful. The term ‘philosophy of action’ is merely a convenient way of referring to that set of questions. However, using that term should not make us lose track of the fact that, for Leibniz, there are intimate connections between human agency and the changes that occur in animals and even in plants and other inanimate things. All substances act, in a broad sense, and understanding more primitive kinds of activity can help us understand human agency. As a result, significant portions of Chapters 1, 2, and 3 are concerned quite generally with the causation of changes in Leibnizian substances. The book thus goes beyond what we would today categorize under ‘philosophy of action.’ One further methodological choice concerns the parts of Leibniz’s corpus that the book takes into consideration. Leibniz’s philosophical views undergo a number of changes in the course of his long career. While there is a substantial amount of controversy over the precise nature, significance, and timing of these changes, interpreters often divide Leibniz’s career into three broad stages: the early period, the middle period, and finally the late, mature, or monadological period. The central focus of this book is the late period, which I take to begin around the middle of the 1690s. This is mainly because texts from the mature

4

Introduction

period contain particularly sophisticated and detailed discussions of agency. It would, of course, be valuable and interesting to examine the development of Leibniz’s philosophy of action, starting with his earliest philosophical writings. Yet, it is not possible to delve into such developmental questions here, given the breadth of topics that this book discusses, as well as the sheer size and complexity of Leibniz’s corpus. While I occasionally bring up passages from earlier texts when they are particularly helpful – for instance, when they answer a question that the mature texts appear to leave open – my chief focus is on the late period. The structure of the book is the following. Chapter 1 introduces readers to the book’s protagonists: Leibniz’s simple and mind-like substances, or monads. Leibniz describes the fundamental nature of monads in several different ways – for instance, as substantial forms, entelechies, primitive forces, and laws of the series. Even more perplexingly, he talks in many different ways about the changing modifications of monads and the relation of these modifications to their subjects. As one might expect, scholars disagree widely about the correct interpretation of these fundamental building blocks of Leibniz’s philosophical system. Chapter 1 aims to untangle some of these issues in order to clear the way for the work of future chapters. It will also serve to introduce nonspecialists to the most central aspects of Leibniz’s ontology. The chapter pays particularly close attention to appetitions and perceptions, the two fundamental types of monadic states. It argues that the distinction between these two kinds of states is important to Leibniz, that every state of a monad is efficiently caused by that monad itself, and that a monad’s actions consist in its bringing about new perceptions. The chapter also examines the different types of appetitions and perceptions acknowledged by Leibniz, as well as the most plausible way to understand the influence of one monadic state on another. Then, in Chapter 2, I explore a central commitment of Leibniz’s metaphysics of action: the doctrine that all states of any substance originate within it, or arise “out of its own depths.” All monads possess a far-reaching independence from other things, which Leibniz calls ‘spontaneity.’ This doctrine is undeniably radical. Many who encounter it – be it today or in Leibniz’s own time – find it absurd, in part because it is difficult to square with the commonsensical distinction between acting and being acted upon. Yet, Leibniz can capture that difference by distinguishing three ways in which changes can originate in a subject; I call them ‘metaphysical spontaneity,’ ‘agent spontaneity,’ and ‘rational spontaneity.’ That threefold distinction, in turn, is tremendously important for Leibniz’s philosophy of action, quite apart from helping him to distinguish actions from passions. In particular, it allows him to claim that some of the desires and emotions that occur in our minds are not ours in an important sense. They are external to our true selves. As a result, Leibniz – like several prominent philosophers of action today – can distinguish between situations in

Introduction

5

which we act autonomously and situations in which we are controlled by desires or emotions that undermine our autonomy. Chapter 3 investigates a second aspect of monadic independence: monads are not only the sources of their actions but also set the ends of those actions. For Leibniz, all monadic activity is immanently end-directed, or an instance of what is traditionally called ‘final causation’ or ‘teleology.’ The chapter first explores Leibniz’s motivations for viewing teleology as ubiquitous. Next, it argues that there is a tight connection between spontaneity and teleology and that it is useful to distinguish three different types of teleology, parallel to the three types of spontaneity described in Chapter 2. I call them ‘metaphysical teleology,’ ‘agent teleology,’ and ‘rational teleology,’ respectively. This distinction, in turn, allows Leibniz to view end-directedness as ubiquitous without anthropomorphizing the least perfect monads and without trivializing the enddirectedness of the most perfect actions. In fact, it allows him to make teleology in its most demanding form a crucial component of his accounts of freedom, control, and moral responsibility, as I argue in later chapters. Finally, Chapter 3 examines the lowest type of teleology in more detail. It argues, against the overwhelming majority of interpreters, that monads perform many of their actions simply because their natures prescribe these actions, not because these actions are or appear good. Chapter 4 tackles another central issue in the philosophy of action: attributability, that is, the question of when an action is properly attributed to an agent. Here, one major obstacle for Leibniz is his endorsement of concurrentism: he accepts the traditional theistic doctrine that creatures require God’s cooperation for all of their actions. Yet, when God acts together with a creature, the resulting action is supposed to be the creature’s action alone, not God’s. This is particularly important for sinful actions, which for theological reasons must not be attributed to God. Chapter 4 examines how it is possible for actions to be attributable only to the created agent, even though God and the creature are acting together. It argues that we can answer that question by taking seriously the roles that final and formal causation play in Leibniz’s account of agency. The fifth chapter focuses on a type of agency that has long been a central topic in the philosophy of action: free agency. Leibniz’s theory of freedom is what we would today categorize as a version of agent-causal compatibilism. In other words, Leibniz believes that freedom is compatible with determinism, and he also holds that free actions are caused by agents rather than by events internal or external to agents. One goal of the chapter is to elucidate Leibniz’s compatibilism. I will show that Leibniz’s metaphysical commitments allow him to circumvent notorious shortcomings of other compatibilist theories. The chapter’s second goal is to take a fresh look at the sense in which Leibnizian free actions are contingent. I argue, against many other interpreters, that Leibniz

6

Introduction

does not ultimately take freedom to be incompatible with necessitarianism simpliciter. Instead, he takes it to be incompatible only with the kind of necessitarianism according to which everything is necessitated in a value-neutral way. Therefore, what Leibniz calls ‘contingency’ has an intimate connection to the most demanding type of final causation. Chapter 6 examines Leibniz’s responses to three notoriously difficult problems: (a) the problem of explaining in what sense free agents have control over their actions, (b) the problem of explaining ostensibly weakwilled actions, and (c) the problem of distinguishing weak-willed from compelled actions. Leibniz explicitly discusses the notion of control – or, as he usually calls it, ‘mastery’ – and, this chapter argues, he manages to make room for a meaningful and desirable type of control. For Leibniz, we possess control to the extent that our rational judgments and rational desires are able to influence our actions. He acknowledges that we sometimes lack the ability to control our actions – namely, when our passions are so powerful that they would outweigh even the strongest rational desire. Yet, Leibniz insists, there are indirect ways to make our rational desires succeed: we can take steps ahead of time that drastically reduce the influence of the passions. Some of the resources that allow Leibniz to give a convincing account of control also allow him to acknowledge a form of weakness of will. That is surprising because he holds that all intentional actions are determined by what the agent perceives as good. Moreover, Leibniz can capture the difference between weakness and compulsion – a hard problem for determinists. The seventh and final chapter addresses two questions concerning moral agency: what it takes to be a moral agent and what it takes to be morally responsible for particular actions. Moral agency is intimately connected to many of the concepts investigated in previous chapters, though these connections are less straightforward than one might initially think. One particularly important result of this chapter’s discussion is that teleology is central to Leibniz’s notion of moral agency. Another important result is that agents are morally responsible for some of their unfree actions. Finally, a particular kind of ability to do otherwise is required for moral blame but not for moral praise. Some general themes will emerge in this book. One important theme is that the will, appetitions, and teleology are key players in Leibniz’s discussions of agency. In order to understand what it means to act and what differentiates different types of activity, we need to look not just at cognition but also at appetition. This goes against existing scholarship on Leibniz’s moral psychology; with a few notable exceptions, other scholars focus almost exclusively on perceptions and the intellect. Another, related general theme of the book is that Leibniz is less of an intellectualist than commonly thought. While this comes up in several chapters, it becomes particularly clear in Chapter 6, where I argue

Introduction

7

that knowing what is best, all things considered, is often insufficient for doing the right thing. Leibniz acknowledges that taming our irrational passions requires us to be extremely resourceful. For instance, we sometimes need to distract ourselves, cultivate beneficial passions, and use sensory images to our advantage. He is far less optimistic than one might initially expect about our intellect’s ability to take on the passions directly.

1

Monads and Their Actions

The protagonists of this book are monads, or Leibnizian simple substances. Despite their simplicity, it is surprisingly difficult to figure out what exactly these monads are and how they act. Leibniz describes the fundamental nature of monads in a number of different ways: as substantial forms, entelechies, primitive forces, and laws of the series, to name just a few. And as if that were not confusing enough, it is unclear how exactly Leibniz understands the changing modifications of monads and the relation of these modifications to their subjects. It is no surprise, therefore, that there is no consensus about the correct interpretation of these fundamental building blocks of Leibniz’s philosophical system. The present chapter aims to address these issues in order to shed light on some of the most basic features of monadic agency and clear the way for the work of future chapters. Of course, I cannot – and do not need to – answer all of the numerous questions about Leibniz’s fundamental ontology. Instead, I focus on the questions that are directly relevant for my interpretation of Leibniz’s philosophy of action. One can learn a lot about what monads are and how they act by looking into the reasons that Leibniz cites for rejecting a purely mechanistic view of nature. These reasons include accounting for the reality, unity, and activity of natural things. Because understanding these reasons proves helpful for my interpretation of monadic agency, I will examine them briefly. Next, I turn to some of the ways in which Leibniz characterizes monads: his claims that they are similar to Scholastic substantial forms and Aristotelian first entelechies, that they consist in primitive force, that their only internal qualities are perceptions and appetitions, and that they contain their entire histories. I will also sketch some of the most important differences between the three types of monads that Leibniz distinguishes, that is, bare monads, nonrational souls, and minds. Finally, I consider the types of causation that are involved in monadic activity. All of these elements will become important later in the book, though they are also interesting in their own right. 8

The Fundamental Nature of Monads

1

9

The Fundamental Nature of Monads

Let us start at the beginning – the very beginning, in fact. For Leibniz, it all begins with God and his ideas. Leibniz’s God is all-powerful, all-knowing, and perfectly good: he can do whatever is metaphysically possible, knows everything that there is to know, and wills only what he recognizes as best. God’s power and goodness will take center stage in later chapters where I discuss divine freedom and the contingency of the created world. For now, however, let us focus on God’s knowledge, or his intellect. God’s omniscience, according to Leibniz, “encompasses every idea and every truth, that is, all things – simple or complex – that can be an object of the understanding” (CD 13). In other words, God eternally possesses ideas of all metaphysical or logical possibilities, as well as knowledge of all necessary truths (see COE 21; letter to Morell, September 29, 1698, A 1.16.164). As a matter of fact, Leibniz views the divine intellect as the source or ground of all possibilities and necessary truths: without God’s intellect, nothing would be possible or true (M 46; CD 7f.; T 184; 189). We must be careful, however, not to confuse this dependence on the divine intellect with the Cartesian doctrine that God creates eternal truths. Leibniz’s God does not ground these truths by willing that they be true, but merely by having an intellect that contains them. Because God’s intellect contains everything that can be known, it also contains knowledge of the goodness and badness of all different possibilities: the ideas in God’s intellect “represent to him the good and the evil, the perfection and the imperfection, the order and the disorder, the congruity and the incongruity of possibles” (COE 21). As a result, the divine intellect can also compare different possibilities and judge them with respect to their goodness. And that is precisely what God does in order to figure out what to create: he judges, based on his perfect knowledge of all possibilities, which possible world is the best.1 Because of God’s perfect goodness, the world that he judges to be best is the world that he subsequently creates. What exactly does God create when he creates a world? In Leibniz’s mature writings, the answer is that God creates an infinity of monads – that is, an infinity of simple, immaterial, mind-like substances.2 Monads are the fundamental building blocks of the created world, and everything else that has a place in Leibniz’s ontology depends on, or is grounded in, them. One crucial aspect of the relationship between the divine intellect and finite monads is that each finite monad corresponds exactly to an idea in God’s intellect. As already seen, the 1 2

How precisely God does this is controversial and complicated. Yet, for present purposes, we need not worry about the details. God himself is a monad, in fact (see a letter to Bierling, August 12, 1711, G 7:502; supplement to a letter to Des Bosses, February 15, 1712, LDB 233f.), though, of course, he differs from created monads in a number of important ways.

10

Monads and Their Actions

divine intellect eternally contains ideas of all possibilities, which include ideas of possible finite substances. God’s ideas of these possible substances, moreover, contain information about every change that these substances will (or would) undergo if they are created. In the middle period, Leibniz usually calls these ideas ‘complete concepts’; in his mature writings, he more frequently calls them ‘possibles’ or ‘essences.’ When God creates the world, he actualizes some of these possibles, namely the ones that together constitute the best of all possible worlds. One significant complication for understanding the fundamental nature of created monads is that the details of Leibniz’s theory of substance, or at least the terminology he uses to express it, are not entirely stable, even within the mature period. In some texts, Leibniz identifies monads with entelechies or substantial forms (e.g. M 18; 63; T 396); in others, he describes monads as made up of entelechies or substantial forms together with passive force, or with matter (e.g. ONI 11). Despite this apparent instability, however, a large portion of Leibniz’s ontology remains the same. As we will see later, he thinks throughout the mature period that finite substances possess, or are,3 primitive active and passive forces,4 from which all of their modifications arise, and that their only fundamental internal modifications are appetitions and perceptions. The ways in which he distinguishes rational monads from lower monads, as well as his characterization of different types of appetition and perception, appear to be stable as well, at least in the most important respects. 1.1

Reality, Unity, and Activity

The essay “New System of Nature” is Leibniz’s first published account of his mature views. It is an autobiographical account of the process that led him to realize that there must be something over and above matter, namely true unities.5 He lists three closely related reasons for positing immaterial unities: (a) matter is not in itself fully real, (b) matter lacks unity, and (c) matter lacks activity. The first two reasons are closely related; Leibniz argues that “a multitude can derive its reality only from true unities,” but that matter, since it is 3

4 5

Even though sometimes Leibniz talks of substances as possessing primitive forces, that is probably misleading. There is good evidence that, strictly speaking, substances are primitive forces. I argue this in Jorati (forthcoming b). However, I will bracket this complication for the purposes of this book because it does not appear to impact my interpretation of monadic agency. God is different from finite substances in this respect: he does not possess any passive force, but only active force. See, for instance, a letter to Remond, February 11, 1715 (G 3:636/L 659). Leibniz does not actually use the term ‘monad’ in “New System,” but he is clearly describing the entities that he elsewhere calls ‘monads.’ The first mention of the term ‘monad’ appears to be in an unfinished letter to the Marquis de l’Hôpital, dated July 22, 1695 (GM 2:295); see Rutherford (1995b: 166n24), Garber (2015: 165). The first published text in which the term is used is “On Nature Itself” (1698).

The Fundamental Nature of Monads

11

infinitely divisible, is devoid of true unities (“New System,” G 4:478/AG 139; see also G 4:482/AG 142 and ONI 11). Leibniz claims, in other words, that a purely material thing cannot possess genuine unity and that what lacks genuine unity is not real. The third reason Leibniz cites for introducing immaterial entities – namely, that matter alone cannot ground activity – is particularly crucial for Leibnizian metaphysics because he holds that activity is a necessary condition for substancehood (e.g. ONI 9; T 393; “New System,” G 4:485/AG 144; NE 65).6 Hence, bodies can be substances only if they are active. To show that mechanical philosophy cannot account for activity, Leibniz argues that activity requires active force, which matter does not possess (e.g. letter to de Volder, April 3, 1699, LDV 77; “Conversation of Philarete and Ariste,” G 6:587/AG 263). Elsewhere, Leibniz elaborates on this line of argument: action cannot be a modification of something passive because otherwise there would be more reality in the modification than in that which is modified. Thus, since matter is passive, action cannot be a modification of matter. If bodies act, they must possess something over and above matter, namely an active principle, of which actions are modifications.7 This argument is valid, but Leibniz’s opponents might simply grant that bodies are passive things that are moved around by other bodies. They could concede that bodies lack activity in any more demanding sense: bodies do not move themselves, but they move each other. Leibniz would have to say more to argue that there is something wrong with that picture of the natural world.8 The argument is nevertheless helpful because it sheds light on the relation between actions and the underlying active principle, which Leibniz typically identifies with primitive force – an issue to which I will return later. Leibniz’s critique of occasionalism contains another argument for the existence of active forces in nature. One problem with occasionalism, Leibniz insists, is that it undermines the absolute perfection of God. After all, when an omnipotent being decrees the laws of nature, that decree should leave a permanent impression in finite things.9 Instead of having created, by his original decree, substances that naturally follow the divine laws, the God of the occasionalists constantly needs to intervene to ensure compliance with his 6

7

8 9

For this reason, Leibniz says in several places that denying activity to created things is tantamount to substance monism; see, for instance, ONI 15; remarks on Lamy, G 4:590/WF 164; reply to Bayle 1705, G 4:568/PT 252; letter to Lelong, February 5, 1712, Robinet 421. Leibniz provides this argument in “On Body and Force” (G 4:397/AG 254), a letter to Jaquelot (March 22, 1703, G 3:457/WF 201), two letters to de Volder (June 30, 1704, LDV 307; April 3, 1699, LDV 75f.), and a letter to Bernoulli (November 18, 1698, A 3.7.944/AG 169). For a helpful discussion of the reasons Leibniz may have had for rejecting that kind of picture, see Carriero (2008). Among other texts, see ONI 6; “On Body and Force,” G 4:396/AG 253f.; letter to Basnage de Beauval, G 3:122/WF 65; letter to Hansch, July 25, 1707, E 2:446/L 593.

Monads and Their Actions

12

laws. In fact, he needs to intervene in ways that go beyond the natural powers of created things. That, Leibniz complains, is a very imperfect state of affairs. A truly perfect creator would make creatures that possess active forces and act in accordance with God’s decrees by themselves, requiring only divine concurrence.10 Finally, Leibniz sometimes argues for active forces from the Principle of Continuity, that is, the principle that everything in nature is uniform (e.g. NE 307). Based on this principle, Leibniz reasons that since there is activity in us – we know that we “elicit in ourselves many thoughts and volitions” (ONI 10) – we should conclude that there are similar active principles in all other created substances. As he puts it in a letter to Damaris Masham, “nature would show little consistency if this particle of matter which makes up the human body were the only thing endowed with something which would make it infinitely different from everything else . . . This leads me to think that there are such active beings everywhere in matter” (May 1704, G 3:339/WF 204; similarly in ONI 10; letter to de Volder, June 30, 1704, LDV 307; letter to Queen Sophie Charlotte, May 8, 1704, G 3:343f./WF 221; G 7:329/SLT 64). This short survey of Leibniz’s reasons for rejecting the mechanistic view and introducing monads shows that monads must have the following characteristics: they are true unities, indivisible, immaterial, and active. Insofar as a composite thing possesses any reality, unity, or activity, it must derive them from monads. Moreover, monads should be understood as in some ways analogous to the human soul. 1.2

Substantial Forms and Entelechies

Leibniz uses a wide variety of terms to explain what monads are. In “New System,” for instance, he describes the entities he is postulating as, among other things, Aristotelian first entelechies, primitive forces, and substantial forms (G 4:478f./AG 139). Given what I said about monads earlier, these designations make sense. First, when Leibniz compares monads to Aristotelian entelechies, he must have in mind the notion we find, for instance, in De Anima, where Aristotle claims that the soul is the first entelechy of the body (De Anima II.1, 412a27–28). For Aristotle, this appears to mean that it is a state of completion or perfection.11 In the “Monadology,” Leibniz states explicitly that he uses the term ‘entelechy’ for his simple substances because they contain “a certain perfection” (M 18; see also T 87).12 Since monads are supposed to be 10 11 12

I will say much more about divine concurrence in Chapter 4. See Johnson (2005: 85–90) for a discussion of the meaning and etymology of ‘entelechy’ in Aristotle. In other texts, Leibniz identifies ‘first entelechy’ with primitive active force; see e.g. “On Body and Force,” G 4:395/AG 252; NE 169.

The Fundamental Nature of Monads

13

analogous to souls, and since they are supposed to ground the reality of organic bodies, it makes sense for Leibniz to borrow the Aristotelian term. The reason why Leibniz also calls monads ‘primitive forces’ is presumably that they are supposed to ground actions or changing modifications. This fits well with Leibniz’s argument, sketched earlier, that actions must be modifications of some active principle and therefore cannot be modifications of matter. Finally, part of the reason why Leibniz calls his simple entities ‘substantial forms’ is presumably that he understands monads as the entities that – like Scholastic substantial forms – account for the unity and reality of things. Just as for some Scholastics, matter without form is mere potentiality and hence lacks reality, Leibniz argues that Cartesian matter is not real because it is entirely passive. To get something real, there must be monads.13 In several texts, in fact, Leibniz claims that simple substances consist of substantial forms together with prime matter. He often identifies the substantial form with primitive active force and prime matter with primitive passive force.14 According to these texts, Leibnizian monads are like Scholastic substantial forms in an additional respect: they combine with matter to form a complete substance.15 In addition to prime matter, or primitive passive force, Leibniz often acknowledges secondary matter, which is an organic machine or body composed of simple substances.16 Accordingly, a complete animal or plant consists of what Leibniz sometimes calls a central (or dominant) monad, or a primary entelechy – the substantial form or “soul” of the organism – and a multitude of subordinate monads that constitute the organism’s body (e.g. letter to de Volder, June 20, 1703, LDV 265; PNG 3f.; NE 220). Because Leibniz sometimes claims that every finite monad has a body (e.g. PNG 3; letter to Bierling, August 12, 1711, G 7:502), this is another way in which monads are analogous to substantial forms. They are the organizing and unifying principles of living things. What is not entirely clear, however, is whether an organism consisting of a central monad and a multitude of subordinate monads counts as a substance. The central monad does appear to convey a certain degree of unity on the body; this sets organisms apart from mere aggregates that lack central monads. Yet, it 13

14 15

16

Both monads and Scholastic substantial forms are also supposed to explain the identity of things over time as well as the characteristic properties and activities of things. I will come back to the latter when discussing final causation in Chapter 3. For the former, see footnote 23. Among other places, he identifies them in “On Body and Force,” G 4:395/AG 252; letter to Jaquelot, March 22, 1703, G 3:458/WF 201. For an excellent discussion of the status of prime matter, see Antognazza (2014). She argues convincingly that prime matter (or primitive passive force) is simply the imperfection or limitation in active force. I provide some additional reasons in Jorati (forthcoming b). This becomes very clear in a letter to de Volder, June 20, 1703, LDV 265; see also ONI 12; “Supplement to the Explanation of the New System,” G 4:572/WF 138; letter to Jaquelot, March 22, 1703, G 3:457/WF 200f.

Monads and Their Actions

14

is controversial whether that degree of unity is sufficient for substancehood.17 Leibniz, at times, suggests that such organisms are substances, and at other times, he insists that only individual monads count as substances. Fortunately, I do not need to resolve that controversy for present purposes. What matters here is merely that monads are substances and that there is a special relationship between the central and the subordinate monads in organisms. I will shed some light on the nature of that special relationship in Chapter 2. 1.3

Primitive Forces, Laws of the Series, and Natures

Let us look in more detail at ‘primitive force,’ one of the terms that Leibniz uses in describing monads, because it is crucial for understanding the way in which monads act. As already seen, Leibniz argues that activity must be a modification of an active principle in the agent. Without such an active principle, a substance would not be able to act. Leibniz typically calls this active principle ‘primitive active force’; a substance’s changing modifications – which Leibniz calls ‘derivative forces’ – inhere in a primitive active force that is unchanging (see letter to de Volder, June 30, 1704, LDV 307; T 87; 396). But what exactly is this primitive active force? In “On Nature Itself,” Leibniz says that it is “analogous to a soul” and that it is “a certain urge [nisus]” or “an inherent law, impressed by divine decree” (ONI 12). This calls to mind Leibniz’s argument against occasionalism that we considered earlier: a perfect God would give his creatures natural powers that enable them to follow the divine laws; the decree of a perfect God would leave a permanent impression in creatures. In “New System,” it becomes clear that primitive force is identical to the natural powers of creatures that God impressed on them at creation. Leibniz stresses that at creation, God gave each finite substance “a nature or an internal force that can produce in it, in an orderly way . . . everything that will happen to it . . . without the help of any created being” (G 4:485/AG 144; similarly in a reply to Bayle, G 4:548/PT 237; ONI 12). He sometimes calls this selfsufficiency of monads ‘spontaneity,’18 and the internal force, the ‘law of the series’ (e.g. in two letters to de Volder: January 21, 1704, LDV 287, and April 3, 1699, LDV 75).19 In a letter to Bayle, he adds that every substance must have “its own true and internal force of acting” and that “the nature of substance consists in this ordered tendency [tendence reglée] that it received at the start and from which the phenomena arise in order” (G 3:58). While the details are not entirely straightforward, the idea appears to be roughly the following. All of the natural – that is, nonmiraculous – changes that occur in a finite substance 17 18 19

For helpful discussions, see e.g. Lodge (2015) and Look and Rutherford (2007). I will say much more about spontaneity in Chapter 2. For helpful discussions of this law of the series, see Kulstad (1990: 139ff.), Rutherford (2013), and Adams (1994: 79ff.).

The Fundamental Nature of Monads

15

result from something internal to the substance. The internal principle of change, moreover, can be fittingly described as the substance’s nature, as a law, as an ordered tendency, or as primitive force. Leibniz also claims in several places that all future and past states of a substance are contained in it at the present time. Sometimes he expresses this doctrine by saying that the substance contains traces of all its past states and marks of all its future states; some interpreters call this the ‘doctrine of marks and traces.’ Leibniz’s favorite metaphor for this doctrine is that the present is “pregnant” with the future (e.g. M 22; PNG 13; a letter to de Volder, January 21, 1704, LDV 287). Yet, some texts show that this talk of “containing” future and past states should not be understood completely literally. For instance, he writes to Jaquelot that “[t]he future is in the past only as an inclination in that past towards the production of the future” (April 28, 1704, G 3:473/WF 182; see also September 4, 1704, G 6:559/WF 188).20 This doctrine should not come as a surprise. After all, Leibniz is a determinist and holds that finite substances do not interact.21 In a closed and perfectly deterministic system, it is always possible to predict all future states and retrodict all past states from the present state and the relevant laws. Leibniz describes his doctrine to Burcher de Volder in almost exactly these terms: “the law of the series shows where it must reach by continuing its progression, i.e., that, with the starting point and the law of progression given, the terms will be produced in order” (January 21, 1704, LDV 289). A substance’s law of the series, then, makes it possible to know all of the past and future states of the substance from its present state. As one might expect – given that ‘law of the series’ is simply a designation for the nature of the substance, or its primitive force – Leibniz sometimes makes an analogous point about creaturely natures: all of a substance’s past and future states can be derived from its nature. This is particularly clear in “On Body and Substance Truly One,” probably written in 1690: “In every substance there is nothing other than that nature or primitive force from which follows the series of its internal operations. This series, i.e. all of its past and future states, can be recognized from any state of the substance, i.e. from its nature” (A 6.4.1673/ SLT 53). While many questions remain unanswered about the different ways in which Leibniz describes the ground of monadic changes, the basic contours of his view appear to be the following. Natural changes in a monad arise from its 20

21

This fits well with my interpretation: a sufficiently intelligent being can see the future in the present state of a substance because it can see that the substance will, given its nature (which consists in force or inclination), produce that particular series of future states. See also a letter to Bayle, G 4:552/PT 239. I will say much more about this denial of interaction in Chapter 2.

Monads and Their Actions

16

primitive force, which is at least partially active22 and unchanging,23 and renders intelligible the entire series of changes occurring within the monad. 2

Monadic States

Having briefly examined the fundamental nature of monads, we can now explore the actions, or changing modifications, of these monads. Leibniz claims in several places that perceptions and appetitions are the only internal qualities of monads (e.g. M 17; PNG 2; letter to Remond, 1714, G 3:622; letter to de Volder, June 30, 1704, LDV 307), that monads are never without perceptions and appetitions, and that all monads have them (e.g. in a letter to Bourguet, December 1714, G 3:574f./L 662f.). Let us examine what Leibniz means by ‘perception’ and ‘appetition,’ what types of perceptions and appetitions there are, and how perceptions and appetitions relate to one another. 2.1

Perceptions

For Leibniz, a perception is simply “[t]he passing state which involves and represents a multitude in the unity or in the simple substance” (M 14; similarly in PNG 2; letter to Bourguet, December 1714, G 3:574f./L 662f.; letter to Remond, 1714, G 3:622; G 7:329f./SLT 65) or “the internal state of the monad representing external things” (PNG 4). An internal state of a monad represents an external thing when there is a particular type of correspondence between them: Leibniz writes to Rudolf Christian Wagner that perception consists in the “correspondence of internal and external, or representation of the external in the internal, of the composite in the simple, of multiplicity in unity” (June 4, 1710, G 7:529/W 505). In another text, Leibniz tells us even more explicitly what it is for one thing to express, or represent, another: “it is sufficient for the expression of one thing in another that there should be a certain constant relational law, by which particulars in the one can be referred to corresponding particulars in the other” (“Metaphysical Consequences” §11, C 15/MP 176f.). This suggests that perception involves an isomorphism. Whenever there is a lawful correspondence between the 22

23

In finite substances, primitive passive force is plausibly part of this nature as well. Accordingly, the natures of finite substances are only partially active, whereas God’s nature is fully active. Additional evidence that the primitive force is unchanging comes from a letter to de Volder: “nothing is permanent in [individual things] except the very law that involves the continued succession” (January 21, 1704, LDV 289; similarly LDV 287 and 291). That fits well with Leibniz’s claim that in order to account for the identity of a substance through change, we must postulate something like a substantial form that remains the same through that change (e.g. NE 231f.; ONI 8; see Adams 1994: 313f. and 79f., Di Bella 2015).

Monadic States

17

internal variety in a monad and external things, the monad perceives these things.24 Perceptions do not have to be conscious, however.25 Most of the perceptions of finite substances are unconscious as well as, in Leibniz’s terminology, “confused.” Only some are distinct. Leibniz appears to view distinct perceptions as perceptions that capture one’s attention and stand out from the infinite mass of simultaneous perceptions. Accordingly, it is helpful to think of distinct perceptions as distinctive.26 Leibniz sometimes uses the terms ‘distinct’ and ‘confused’ as relative notions. He says, for instance, that our perceptions “are sometimes more and sometimes less clear and distinct” (letter to Queen Sophie Charlotte, May 8, 1704, G 3:344/WF 221).27 He also compares degrees of distinctness across monads. For instance, he claims that finite monads are “differentiated by the degree of their distinct perceptions” (M 60) and that one substance can be said to act on another substance insofar as the perceptions of the former are more distinct (“A Specimen of Discoveries,” A 6.4.1620/MP 79). This leads us to another important feature of monadic perceptions. Monads perceive not only some external objects – they perceive the entire world. Leibniz, therefore, calls each monad a “mirror of the universe” (M 56). Yet, each finite monad perceives only a small portion of the universe distinctly, which constitutes its point of view. 2.2

Appetitions

Let us now turn to the second type of internal modification of monads, which Leibniz typically calls ‘appetition.’28 Because appetitions have received much less attention in the secondary literature than perceptions, and because appetitions feature prominently in Leibniz’s philosophy of action, it will prove helpful to investigate them more thoroughly than we have investigated perceptions. First of all, it is important to note the terminology that Leibniz employs to refer to appetitions. He often uses the terms ‘appetite’ (Latin: appetitus, French: appetit) or ‘appetition’ (Latin: appetitio, French: appetition), apparently interchangeably, for all tendencies in monads (“Table of Definitions,” C 472). At other times, he reserves the term ‘appetition’ for imperfect, unconscious 24 25

26 27 28

Simmons provides a more detailed discussion of Leibnizian perception as an isomorphism (2001: 67ff.; see also Jorgensen 2015, Swoyer 1995: 84f., Kulstad 1977: 73f., 2006: 414ff.). This becomes clear, e.g., in M 14 and PNG 4. It is controversial what exactly consciousness is for Leibniz. See, for instance, Jorgensen (2009, 2011b), Simmons (2001, 2011), Gennaro (1999), Kulstad (1991), and Barth (2011). This is suggested, e.g., in NE 53f.; PNG 13. See also Simmons (2001: 57) and Jorgensen (2011a: 192). Bolton’s definition captures this well (2008: 121). There is some overlap between my discussion of appetitions in this chapter and Jorati (forthcoming a).

18

Monads and Their Actions

inclinations and contrasts it with ‘volition’ (e.g. NE 173; 189; 194; revision note, A 6.1.286/L 92n18). To make things less confusing, I will use the terms ‘appetition’ and ‘appetite’ in the general sense for all monadic inclinations. Leibniz also uses other terms to refer to or describe appetitions: the French and Latin counterparts of the terms ‘tendency’ (letter to Wolff, GLW 56; letter to Remond, G 3:622; letter to Bourguet, G 3:575/L 663), ‘inclination’ (reply to Bayle, G 4:550/WF 105; Gr 480/SLT 97; CD 138; NE 351), and ‘desire’ (NE 192; “Definitions,” A 6.4.310; Beeley 11), the French term effort (NE 172f.; 192), and finally, the Latin terms conatus (‘Table of Definitions,’ C 491) and percepturitio (letter to Wolff, GLW 56). Leibniz defines appetitions as a monad’s “tendencies to go from one perception to another” (PNG 2) or as the “action of the internal principle which brings about the change or passage from one perception to another” (M 15).29 In one way or another, then, appetitions are supposed to help explain30 the changes in monadic perceptions. When a soul transitions naturally from pleasure to pain, for instance, this change is explained – at least in part – by the appetitions of that soul. That of course raises the question of how appetitions relate to primitive active force, which Leibniz also views as explaining the entire series of changes. The definition of ‘appetition’ from “Monadology” §15 can help us answer that question: appetitions are the actions of an “internal principle,” that is, presumably, of the monad’s primitive active force. This suggests that appetitions are derivative active forces: they are the changing modifications of the underlying, unchanging principle of action.31 One problem with understanding the way in which appetitions explain perceptual change is that Leibniz views perceptual transitions, like all natural transitions, as continuous (see e.g. M 10; 13; letter to Johann Bernoulli, September 30, 1698, A 3.7.912/LDV 11; NE 56f.). Just as a line is not composed out of points, a perceptual series is not composed out of instantaneous perceptions (see e.g. note on Foucher’s objection, G 4:491/AG 146; NE 152). Even though we may be able meaningfully to talk about instantaneous perceptual states, just as we can talk about points, there is strictly speaking no 29 30 31

Elsewhere, he defines ‘appetite’ as “the endeavour of acting tending towards new perception” (G 7:330/SLT 66); similarly in a letter to Bourguet, August 5, 1715, G 3:581/L 664. In Section 4, I will consider whether this explanation should be understood in terms of causation. See Rutherford (2005: 165), Adams (1994: 380), as well as Kulstad (1990: 136), who also think that appetitions are best understood as derivative active forces. Pauline Phemister disagrees, however: she claims that even though perceptions and appetitions are modifications of primitive forces, they are not what Leibniz calls ‘derivative forces’; instead, she claims, Leibniz uses the term ‘derivative force’ exclusively for the forces attributed to bodies (2005: 214 and 220). Yet, this seems to be merely a terminological issue; Phemister agrees that appetitions are modifications of primitive force. Moreover, as McDonough shows (2016a: 11), Leibniz does appear to use the term ‘derivative force’ for metaphysical forces in at least one text, namely a letter to de Volder (January 21, 1704, LDV 286).

Monadic States

19

such thing as a transition from one such instantaneous state to its immediate successor. After all, there are further perceptions between any two instantaneous perceptual states.32 As a result, appetitions cannot, strictly speaking, be the tendencies to transition from one instantaneous perceptual state to the next because there is never an immediately adjacent state.33 This means that appetitions cannot be what we might call ‘short-term tendencies’: tendencies to transition directly from one instantaneous state to the next. If an instantaneous appetition leads to an instantaneous perceptual state, that perceptual state must always be separated from the appetition by a series of other perceptions. There are also independent reasons for holding that at least some appetitions are what we might call ‘long-term tendencies’: tendencies to transition from one state to a later one that is not adjacent to it.34 Take for instance someone’s desire to lose weight, learn a foreign language, or earn a college degree. Those desires should presumably be some kind of appetition, but they are not simply appetitions for an immediate transition to a new perception. Instead, they are appetitions for a certain outcome that can be achieved only via a series of other transitions. Indeed, in several texts, Leibniz acknowledges the existence of long-term appetitions. He typically does this by drawing a distinction between the end and the means: an appetition for some end often gives rise to appetitions for the necessary means. This is particularly explicit in a table of definitions from the early 1670s, where Leibniz defines ‘end’ and ‘means’ as follows: “the END is that, an appetite for which brings about an appetite for another thing. This other thing is called the MEANS” (A 6.2.492). At least some appetitions, according to this early text, are directed at ends that they bring about not directly but via appetitions for means to those ends. In another table of definitions, written about 30 years later, Leibniz seems to paint a similar picture: “the end is that, an appetition for which is the sufficient cause of an endeavor [conatus] in the agent” (C 472). Because conatus is either a synonym for ‘appetite’ or refers to a kind of appetite (see e.g. NE 172; Beeley 11; G 7:330/SLT 66), this passage is further evidence that some appetitions aim not at a state to which the substance can transition immediately but rather at an end that can be brought about only indirectly, via other appetitions and hence other perceptions. Another helpful passage, which is further evidence for the existence of longterm appetitions, stems from a letter to Nicolaas Hartsoeker. Appetitions for actions that require the cooperation of one’s bodily organs can succeed, Leibniz tells Hartsoeker, only if “the appetitions, and consequently the perceptions to which they give rise, . . . go into perfect detail with respect to all that happens in 32 33 34

In Section 4.2, I will address the question of whether this undermines the reality of perceptions. Whipple (2010b) explores this issue in great depth and draws a rather radical conclusion; I will discuss his solution in Section 4.2. See Kulstad (1990: 145), who argues this as well.

20

Monads and Their Actions

the organs” (G 3:509). Whenever one’s appetition is not perfectly detailed, Leibniz goes on to say, it is not “a perfect and entire appetite for this whole object,” and therefore, it is not “an appetition that the body is obliged to follow and to execute.” He points out that a perfectly detailed appetition is one that tends not only toward the end but also toward “the means and the means of the means” (G 3:510). This makes sense: my appetition to raise my arm succeeds only if I also possess appetitions for the requisite means, such as particular muscle contractions, as well as any necessary means to those means, and so on. After all, appetitions are simply tendencies. If for some reason I lack the tendency for one of those means, I will not succeed in raising my arm. An equally crucial thing to note about appetitions is that Leibnizian substances have multiple simultaneous appetitions. This must be the case because, as we saw earlier, each monad constantly represents everything in the entire universe, and the universe is infinitely complex. As a result, a monad must have infinitely many perceptions at any time, or – what comes down to the same thing, for present purposes – one infinitely complex perceptual state. Because appetitions are tendencies to transition from one perceptual state to another, monads must also possess infinitely many appetitions, or an infinitely complex appetitive state, at any time.35 Leibniz says this explicitly in the New Essays (NE 110; see also a letter to Masson, G 6:627/AG 228). Let us now examine how Leibniz thinks of the conflict among simultaneous appetitions. This is significant for several aspects of his theory of monadic agency. That there must be some kind of conflict becomes clear in the “Monadology”: Leibniz says that “the appetite cannot always completely reach the whole perception toward which it tends, but it always obtains something of it, and reaches new perceptions” (M 15). According to this passage, then, there are cases in which an appetition does not fully succeed. The most promising explanation for this is the presence of at least one other incompatible appetition. Leibniz suggests this himself in another passage and employs a useful analogy: he compares the competition between different mental tendencies to the way that “in mechanics, the composite movement results from all the tendencies that occur simultaneously in one and the same moving thing, and satisfies each one equally, in so far as it is possible to do all at once” (T 22). That might explain what Leibniz means when he says that even though an appetite may fail to reach the entire perception at which it aims, “it always obtains something of it, and reaches new perceptions” (M 15). The picture, then, seems straightforward: when there are tendencies in a substance that cannot all be satisfied simultaneously, these tendencies 35

Jeffrey McDonough (2016b: 108f.) argues that there is only one infinitely complex global appetition. Yet, for our purposes, it does not matter how we individuate appetitions. To simplify things, I will talk of infinitely many simultaneous appetitions, but what I say works equally well for aspects or components of one global appetition.

Monadic States

21

compete with each other analogously to the way in which physical forces compete.36 Thinking of the competition among appetitions by analogy with conflicting physical forces certainly helps. Yet, due to the presence of long-term appetitions, the picture must ultimately be more complicated than that.37 After all, a long-term appetition can fail in at least two ways: either because of another long-term appetition with which it is in direct conflict or because the agent does not have successful shorter-term appetitions for all the requisite means. The latter, in turn, can happen either because the agent does not have appetitions for those means at all or because something prevents the agent’s inclinations toward the means from succeeding. So, for example, my long-term appetition to learn Farsi by the end of the year might fail for any of the following reasons: (a) because of my long-term appetition to write a book, which I make my top priority, (b) because I lack appetitions for some of the requisite means to learning Farsi (say, because I am unaware that they are requisite means), or (c) because my appetitions for some of the requisite means are outweighed by other appetitions. The competition between appetitions, then, is an incredibly complicated affair, not only because there are infinitely many simultaneous appetitions, but also because there are long-term appetitions whose success depends on a series of other shorter-term appetitions. It may appear that the inevitable consequence of this is utter chaos. It could seem almost miraculous that we ever succeed in doing what we want to do because it requires that none of the infinitely many other appetitions get in the way. Can Leibniz explain why we so often succeed in doing what we want to do? Clearly, what he needs is some account of how certain appetitions reliably trump other appetitions; there must be some sort of hierarchy that explains why certain types of appetitions, such as my appetition to raise my arm,38 usually succeed. One part of the solution to this puzzle is that appetitions differ in strength. In the Theodicy, for instance, Leibniz claims that when there are opposing tendencies in a mind, “the strongest [la plus forte] wins” (T 337).39 Similarly, Leibniz writes to Clarke that “if the mind should prefer a weak inclination to a strong one [l’inclination foible à la forte], it would act against itself” 36 37 38

39

For Leibniz, the interaction between physical forces may, of course, ultimately be phenomenal. Bolton has recently argued this as well (2013: 189). Strictly speaking, as we will see in Section 2.3, raising my arm ultimately consists in producing perceptions of my arm going up – perceptions that are mirrored by all other monads. Yet, that does not make it any less mysterious that my desire to raise my arm reliably results in such a perception. This does not always happen, he adds, because they might also “stop each other or some third option will result from them” (ibid.). That makes sense: in cases where there are two contrary appetitions of the same strength, for instance, they might impede each other or result in some compromise state.

22

Monads and Their Actions

(LC 5.15).40 Sometimes, Leibniz even compares the conflict among appetitions to weights in a balance. Like a balance, the mind will move toward the side on which the appetitions are strongest or heaviest (see e.g. LC 5.3; NE 116; Gr 480/SLT 96). This talk of strength fits well with the force analogy: the outcome of an interaction between forces depends not only on the directions of the forces but also on their strength. In Chapter 6, I will say more about the reasons why some appetitions are stronger than others. Relatedly, it is important to note that Leibniz acknowledges several types of appetitions. One distinction that Leibniz draws is between conscious and unconscious appetitions: he writes to Nicolas Remond that “our great perceptions and our great appetites of which we are conscious, are composed of innumerable little perceptions and little inclinations of which we cannot be conscious” (November 4, 1715, G 3:657/W 554). This passage is interesting because it tells us both that we are conscious of some but not all appetitions and that unconscious appetitions can compose conscious ones. In fact, as the passage indicates, this account is parallel to the one Leibniz sometimes gives of perceptions (e.g. NE 134). Leibniz also distinguishes conscious from unconscious appetitions – which he there calls ‘insensible’ – in a reply to Bayle (G 4:550/WF 105).41 It is independently plausible that there are conscious and unconscious appetitions. We are clearly aware of some of our desires or inclinations but not of all of our infinitely many simultaneous inclinations. For instance, you may be aware of your craving for coffee but unaware of the appetitions responsible for your heart’s beating. It is because of these insensible appetitions that we may be taken by surprise by some of the perceptions that arise in us: we do not see them coming because we are not aware that our soul has an inclination toward those perceptions. In some cases, these insensible appetitions even lead to perceptions that are unwelcome or unpleasant and that we would never pursue intentionally.42 In addition to distinguishing conscious from unconscious appetitions, Leibniz subdivides the conscious ones into rational and nonrational: [T]here are insensible inclinations of which we are not aware. There are sensible ones: we are acquainted with their existence and their objects, but have no sense of how they 40 41 42

See also LC 5.4, where he uses the term ‘large’ (grande) to describe the inclination that succeeds. Leibniz describes some appetitions as ‘insensible’ elsewhere as well, e.g. in a marginal note on Temmik (in Mugnai 1992: 163) and in T 310. See Leibniz’s example of a man who unexpectedly gets stung by an insect and feels pain, which Leibniz explains in terms of “insensible dispositions of the soul” (G 4:547/PT 237). This is an interesting example because it illustrates that unconscious appetitions can give rise to conscious perceptions. The reverse can plausibly happen as well: something that an agent consciously perceives can presumably give rise to an unconscious tendency in her. For instance, while watching an advertisement, an agent may develop an unconscious inclination to purchase the product.

Monadic States

23

are constituted . . . Finally there are distinct inclinations which reason gives us: we have a sense both of their strength and of their constitution [formation]. (NE 194)

According to this passage, there are three types of appetitions: (a) insensible or unconscious appetitions, (b) sensible or merely conscious appetitions, and (c) distinct or rational appetitions.43 A few pages earlier Leibniz provides an even more fine-grained account. Instead of dividing conscious appetitions merely into rational and sensible, he further subdivides sensible appetitions into those that are accompanied by pleasure or suffering and those that are not (NE 192). 2.3

Actions

We have already explored several aspects of the fundamental nature of monads and their changing modifications. Yet, one obvious question remains open: what exactly are the actions of monads? Given that monads have only two basic types of states – appetitions and perceptions – and given that finite monads do not produce any changes outside of themselves, there are two straightforward candidates. Monadic actions could be either appetitions, that is, the tendencies to transition to new perceptual states, or they could be (the transitions among) perceptual states. In my view, we should endorse the second option: a monad’s perceptions – or perhaps, more precisely, a monad’s giving rise to perceptions – are its actions. This may initially sound odd, but on reflection, it is the obvious choice. After all, Leibnizian perceptions are not passive states; a monad does not receive its perceptions passively but rather produces them actively. This must be the case because finite monads do not causally interact. Your action of raising your arm, for instance, consists merely – in metaphysical strictness – in your producing a series of perceptual states that represent your arm going up. Moreover, Leibniz sometimes describes appetitions as the endeavors for acting (e.g. G 7:330/SLT 66). Because appetitions tend toward the production of new perceptions, this production must be the action. This also means that appetitions are not good candidates for actions; they are merely tendencies for acting in a particular way. A monad’s actions consist, most plausibly, in its production of perceptual states.44 This suggestion finds 43 44

See Phemister, who also discusses these three types (2005: 247f.). The three Leibnizian types of appetition are similar to Aquinas’s three types (see e.g. SCG 2.47.2; ST I q59 a1 corp.). McRae puts forward a slightly different explanation, which is compatible with mine (1976: 63f.). Sukjae Lee provides yet another explanation. He thinks that Leibniz calls perceptions ‘actions’ because “a perception is a real cause of a subsequent state of the substance” (2004: 227). I am unconvinced by Lee’s explanation, and not only because I do not think that one perception is the active cause of another (see Section 4.1). What makes something an action of mine is not that it causes my subsequent states but rather that it has a special, immediate relationship to me. Leibniz appears to agree (see NE 210).

24

Monads and Their Actions

support in the draft of a letter to Des Bosses, in which Leibniz writes that “the operation proper to the soul is perception” (April 30, 1709, LDB 129), as well as in the New Essays, in which he claims that “thoughts are actions” (NE 86). An important caveat is that Leibniz sometimes reserves the term ‘action’ for transitions to more perfect, or more distinct, perceptions and opposes it to ‘passion,’ or to transitions to less perfect, or more confused, perceptions. One such passage is from the “Monadology”: “The creature is said to act externally insofar as it is perfect, and to be acted upon by another, insofar as it is imperfect. Thus we attribute action to a monad insofar as it has distinct perceptions, and passion, insofar as it has confused perceptions” (M 49; similarly NE 210f.). This is, however, clearly a different sense of ‘action,’ and not what Leibniz calls the “metaphysically rigorous sense” (NE 210). In the rigorous sense, all changes in a monad’s perceptions are its actions. In Chapter 2, I will describe the distinction between these two senses of ‘action’ with much more precision. 2.4

The Distinction between Appetitions and Perceptions

While I have so far discussed perceptions and appetitions separately, it is important to note that there is an intimate relation between them. In fact, it is somewhat controversial whether appetitions and perceptions are fundamentally distinct. Some passages suggest that the two are separate modifications of monads (e.g. PNG 2; letter to Bierling, August 12, 1711, G 7:501f.; G 7:330/SLT 66; letter to Bernoulli, July 1, 1704, LDV 311), but other passages make it sound as if they are ultimately one and the same thing. In this section, I argue that appetitions and perceptions are at least different modes of monadic states and that the distinction between them is important to Leibniz. Some passages appear to describe monads not as having perceptions and separate tendencies toward new perceptions but rather as having perceptions that themselves tend toward other perceptions. Leibniz writes, for instance, that “the soul . . . involves a compound tendency, that is to say a multitude of present thoughts, each of which tends towards a particular change, depending on what is involved in it” (G 4:562/WF 115). This passage seems to say that perceptions themselves, rather than appetitions that are distinct from perceptions, are or contain the monad’s tendencies to change its perceptions. Similarly, Leibniz writes to Samuel Masson that because the objects of our perceptions are composite, “these perceptions are called composite, as are their tendencies or appetites” (G 6:627/AG 228). Here again, Leibniz appears to suggest that perceptions themselves have tendencies. On the basis of passages like these,

Monadic States

25

some interpreters argue that perceptions and appetitions are not fundamentally distinct.45 There are at least three ways to deal with this apparent tension in Leibniz’s description of appetitions. One is to discount passages in which Leibniz appears to ascribe tendencies to perceptions and to take at face value passages in which he describes them as separate states. A second option is to conclude that appetitions and perceptions differ only conceptually: they are merely two different ways of understanding the same thing.46 A third possibility, which I prefer to the second, is to argue that appetitions and perceptions are separate modes of monadic states – separate modes of modes, if you will. This would explain why Leibniz sometimes describes them as two different things and sometimes as one thing. This third interpretation allows us to say that even though the same monadic state represents and strives, the distinction between the representation and the striving is more than merely conceptual.47 We can understand the difference between the second and the third interpretation through an admittedly imperfect analogy with motion. The third interpretation views the distinction between appetitions and perceptions as analogous to the distinction between the speed of an object’s motion and its direction: these are two independent modes of the object’s motion. On the second interpretation, on the other hand, the distinction between appetitions and perceptions is analogous to the distinction between an object’s motion away from its origin and its motion toward its terminus. It is simply the same motion conceived in two different ways. The difference between these two interpretations is important because the third one leaves room for the possibility that a monad might be inclined differently by a certain representation than it in fact is; the second interpretation rules that out. That possibility is crucial for Leibniz’s theory of freedom, as we will see in Chapter 5.48 For this reason, I prefer the third interpretation to the second one, and have no objection to the first one either, since it too is compatible with Leibniz’s views on freedom. There is further evidence that Leibniz does not want to treat perceptions and appetitions as merely conceptually distinct. He writes to Wolff, for instance, that “whatever one may understand generally in the soul can be reduced to two things: the soul’s expression of the present state of external things, in 45

46 47 48

McRae, for instance, writes that they are merely “the same modifications viewed differently” (1976: 60). Other interpreters who hold that appetitions and perceptions are not fundamentally distinct include Whipple (2010b: 407n56), Clatterbaugh (1973: 9), and Bolton (2011: 145). This appears to be McRae’s view (1976: 60). We could call it a modal distinction. This might be what Rutherford has in mind (1995b: 138). This becomes clear, e.g., in a letter to Jaquelot, G 3:468/WF 179; T 310. Additional evidence occurs in T 34, where Leibniz entertains the possibility of an agent who can judge but lacks the inclination to act accordingly. See Jorati (2014a: 759f.), where I argue that there is no metaphysically necessary connection between particular representational contents and particular tendencies toward new representations.

26

Monads and Their Actions

accordance with its body; and the tendency to a new expression” (GLW 56). If Leibniz thought that appetitions and perceptions were the same thing viewed differently, it would be misleading to also claim that they are “two things.” Perhaps more importantly, Leibniz sometimes says that appetition and perception in created substances correspond to will and intellect in God (e.g. M 48), which is helpful because Leibniz considers it crucial that God’s will and God’s intellect are distinct in a rather robust sense (COE 21; T 149; M 46). One potential problem for my interpretation is the following. As I argue in Section 4.2, there are good reasons to think that, strictly speaking, instantaneous monadic perceptions are merely abstractions from a continuous series. They do not ground that series – the series grounds them. The same might hold for appetitions. Yet, I do not think that this undermines my claim about the distinction between appetitions and perceptions. After all, when we abstract two different instantaneous entities from a continuous series – a representation and an inclination – it does not follow that these entities differ from each other only conceptually. The following analogy from geometry might help: when describing a curve, we can provide the coordinates for one point on the curve as well as specify the curve’s slope at that point. Even if we view those two entities as abstractions from the curve, as Leibniz appears to do, it does not follow that the distinction between them is merely conceptual. Instead, the distinction appears to be a modal distinction: passing through a particular point and possessing a particular slope at that point are two different modes of the curve. It is possible for a curve to pass through the same point at a different slope, and vice versa. 3

Types of Monads

Each monad differs from any other monad – as it must, in accordance with Leibniz’s Principle of the Identity of Indiscernibles. Yet, Leibniz holds that we can group monads into three broad types: bare monads, nonrational souls, and minds. Bare monads are the most common type of monad; they occur in plants and in the small organisms that make up inanimate things, as well as the bodies of animals and human beings. Nonrational souls, on the other hand, are the central monads of subhuman animals. Finally, minds are rational souls and include the souls of human beings and of angels. God is a mind as well but differs from other minds in lacking a body. The distinction between these three types of monads is important for future chapters because it has implications for the differences between rational and nonrational actions. Thus, it is helpful to examine the grounds of this tripartite distinction. First of all, we should note that there is a significant amount of continuity. All three types of monads have perceptions, for instance – even the dullest of monads. Yet, the least perfect monads have only comparatively confused

Causation in Monadic Actions

27

perceptions and lack memory. They resemble a human mind that is sleeping dreamlessly (M 20; PNG 4) or in a stupor (M 24; G 7:330/SLT 66). Animal souls are capable of more distinct perceptions and sensation, which is simply “perception accompanied by memory” or by memory and attention (PNG 4; similarly M 20; letter to Wagner, June 4, 1710, G 7:529; G 7:330/SLT 65). Finally, minds possess reason and their perceptions can amount to thought, which is perception with reason or with reflection (PNG 4f.; M 29; G 7:331/ SLT 66; NE 173). Of course, not all perceptions of minds are thoughts, just as not all perceptions of animal souls are sensations. Leibniz thinks that most of the perceptions in finite minds are exactly like the perceptions in animal souls; in fact, many of them are like the perceptions of bare monads. Only a small number of a finite mind’s infinitely many simultaneous perceptions can be rational, and at certain times – for instance, while asleep – none of these perceptions are rational. Thus, the perceptual differences between the three kinds of monads can be summarized as follows: bare monads have only lowlevel perceptions, animal souls additionally possess sensation and memory, and minds additionally possess reason and reflection. The three types of monads differ not only with respect to their perceptions but also with respect to their appetitions. In fact, there are good reasons to think that the three types of appetitions described in Section 2.2 correspond to the three types of perceptions just mentioned. Even though Leibniz does not say so explicitly, as far as I know, it seems clear that bare monads have only insensible appetitions, that animal souls additionally have sensitive appetitions, and that minds additionally have rational appetitions. After all, bare monads lack sensation, and only minds possess the power of reasoning. 4

Causation in Monadic Actions

The final question I would like to discuss in this chapter is how precisely monadic actions are caused. This question is surprisingly difficult to answer. Not only is it controversial what exactly causes monadic actions; it is also far from obvious which types of causation are involved in monadic activity. The present section aims to show that actions have efficient, final, and formal causes, perhaps in addition to something like occasional causes. Moreover, I will argue that the monads themselves – rather than their modifications – are the efficient and formal causes of their actions. Because Chapter 3 examines the role that final causation plays in monadic actions, I will not say very much about it here. 4.1

Efficient Causation

In a number of passages, Leibniz calls the monadic realm ‘the kingdom of final causes’ and the physical realm ‘the kingdom of efficient causes’ (e.g. M 79;

28

Monads and Their Actions

“Against Barbaric Physics,” G 7:344/AG 319; G 6:542/L 588; GM 6:243/ AG 126). On the face of it, this suggests that there are no efficient causes in the monadic realm and no final causes in the physical realm. In fact, if only monads are ultimately real, it suggests that efficient causation is merely phenomenal and that all real causation is final causation. Yet, that conclusion is problematic for a number of reasons. For instance, there are passages in which Leibniz seems to state that final causes can explain some physical events (e.g. “On Body and Force,” G 4:398/AG 254), as well as passages in which he says that there is efficient (or productive) causation in finite monads (e.g. letter to l’Hôpital, July 12, 1695, A 3.6.451/WF 56f.). There is no need to discuss the role of final causation in Leibniz’s natural philosophy here;49 we can instead focus on the role of efficient causes in monadic actions. In Chapter 3, I will then explore the role of final causation in monadic change. There are many good reasons for thinking that there must be efficient causation, in addition to final causation, in created monads. First of all, final causation, as traditionally understood, goes hand in hand with efficient causation: in order for a change to occur, there must be not only a purpose or end but also an efficient cause that acts for the sake of this end or produces something in order to achieve the end. An end or purpose on its own, without an efficient cause, does not lead to a change. The way Leibniz describes efficient causation suggests that he agrees: he equates the efficient cause with that which produces, or with the productive cause (NE 228). Moreover, in his “Table of Definitions” from the early 1700s, he defines the efficient cause as “the active cause” (C 472); in another table of definitions from the early 1670s, he defines it as “a cause through action” (A 6.2.490). Accordingly, an efficient cause in the most general sense seems to be that which produces some change through an action.50 It is plausible, then, that Leibniz agrees with mainstream medieval Aristotelians that without an efficient cause no change would be produced. He could, of course, locate this efficient causality entirely in God. This is what Sukjae Lee argues: Leibniz’s God is the sole productive cause of monadic changes (2004). Yet, this strikes me – along with many other interpreters of Leibniz – as too close to occasionalism. It also appears to make created substances more passive than Leibniz often insists they are (see McDonough 2007: 32). Formal and final causation – which Lee takes to be the only types of causation of which finite substances are capable – do not seem to be types of 49 50

For helpful discussions, see McDonough (2008, 2011: 197f.), Bennett (2005), Rutherford (2005: 166f.), and Hirschmann (1988). There are interesting parallels here with Suárez’s definition of the efficient cause as “that whence the effect exists by means of an action; that is, . . . a principle from which the effect flows forth, or on which it depends, through an action” (MD 17.1.6/OEC 10).

Causation in Monadic Actions

29

genuine activity.51 Hence, there must be efficient causation in the realm of finite monads. There is also more direct textual evidence that changes in created monads have created efficient causes. For instance, Leibniz writes to Guillaume François de l’Hôpital in 1695 that “every substance . . . produces for itself, internally, in order, everything that will ever happen to it” (A 3.6.451/WF 56f.). Since he equates the productive with the efficient cause (NE 228), this is strong evidence that there is efficient causation in all monads. Similarly, Leibniz tells Isaac Jaquelot, “I maintain that God gave the soul the power of producing its own thoughts” (February 9, 1704, G 3:464/WF 175). Furthermore, in a 1715 letter to Nicolas Remond, Leibniz first stresses – pace occasionalists – that finite substances must be active, then claims that secondary (that is, created) causes are efficacious, and finally states that “each substance or monad . . . follows its own laws in producing its actions” (November 4, 1715, G 3:657).52 And finally, in “On Nature Itself,” Leibniz states that God has placed “a certain efficacy [quandam efficaciam]” in created substances (ONI 6; similarly ONI 13). In fact, Leibniz’s main goal in that essay is refuting the occasionalist doctrine that creatures do not produce anything (see e.g. ONI 10). These are compelling reasons to believe that there is efficient causation in created monads. Yet, acknowledging that there must be something in the realm of finite monads that efficiently causes monadic states does not yet tell us where exactly to locate this efficient cause. One thing is clear, as already mentioned: created monads do not interact causally, according to Leibniz. Instead, all nonmiraculous changes in each created monad originate within it. Hence, the finite cause of changes in a monad cannot be anything external to it. Something in that monad must cause those changes. But what exactly? Most interpreters assume that substances underlie, in some sense, the states or changes that occur in them.53 On that kind of interpretation, there are at least two ways to understand Leibniz’s claim that all changes arise in a created monad spontaneously. First, it could mean that any state of a created substance is efficiently caused by a prior state of the same substance. Alternatively, it could mean that the substance itself efficiently causes all of its states.54 Each of these alternatives has 51 52

53 54

As seen, Leibniz defines the efficient cause as the active cause. The implication seems to be that other types of causes – including formal and final causes – are not active. There are other passages in which Leibniz talks of finite substances producing – that is, efficiently causing – their states. See, for instance, a letter to de Volder, June 30, 1704, LDV 309; T 298; “New System,” G 4:485/AG 144. There are notable exceptions. See especially John Whipple, who challenges this common assumption (2010b: 407). Martin Lin contends that the distinction between causation by the monad and causation by prior states might ultimately be a distinction without a difference. After all, saying that a perception causes another perception arguably means that the substance itself, insofar as it has a certain

30

Monads and Their Actions

adherents. I will briefly discuss the first of these readings before arguing for the second one. If all states of a finite monad are efficiently caused by prior states of the same monad, monadic states form causal chains, and efficient causation consists in state (or perhaps event) causation. Because Leibniz acknowledges two basic types of monadic states, there are two possibilities: either prior perceptions or prior appetitions are the efficient causes of changes in a monad.55 Following Bobro and Clatterbaugh, we can call the former possibility the “efficacious perception view” (1996: 408f.), and we can call the latter possibility, which Bobro and Clatterbaugh do not consider, the ‘efficacious appetition view.’ Each of these two possible interpretations has supporters.56 There is some textual evidence in favor of the efficacious perception view, which is endorsed explicitly by Nicholas Jolley (1998: 605). Leibniz says in his “Animadversions” against Georg Stahl, for instance, that “the representation of the end in a soul is the efficient cause of the representation of the means [to this end] in the same [soul]” (Dut 2.2.134/LSC 23). Here, Leibniz explicitly describes one representation in the soul as the efficient cause of another representation. A similar statement occurs in Leibniz’s reply to Pierre Bayle’s question of why a dog would transition to a painful state when beaten: “The representation of the present state of the universe in the dog’s soul produces in it the representation of the subsequent state of the same universe” (G 4:532/WF 78). Here again, Leibniz suggests that earlier representations – that is, perceptions – produce later perceptions.57 Several other interpreters endorse what I call ‘the efficacious appetition view,’ for instance, Donald Rutherford (2005: 166, 2013: 166f.), Laurence Carlin (2006: 231), and Martha Bolton (2013: 178). The main reasons in favor of the efficacious appetition view appear to be that Leibniz often describes changes in a monad as the consequences of prior states, as well as the fact that

55 56

57

perception, causes itself insofar as it has another perception (2014: 181). There is definitely some truth in this: modifications, plausibly, are simply the substance modified in a particular way (see Lin 2014: 172). Yet, as we will see below, the continuity of perceptual states makes it problematic to think of perceptions or appetitions as the causes of other states, even if we understand this in Lin’s way. Of course, one could also claim that perceptions and appetitions together produce new states in a substance. Yet, I am not aware of any commentators who explicitly argue this. In addition, there are some Leibniz scholars who claim that every state of a finite monad is efficiently caused by a prior state of the same monad but who do not specify whether this efficient cause is a prior perception or a prior appetition (e.g. Sleigh 1990b: 162, Kulstad 1993: 96). There are also numerous passages in which Leibniz describes the states of created monads as “consequences” or “results” of earlier states (e.g. M 22; letter to Foucher, 1686, G 1:382/WF 52; undated letter to Jaquelot, G 3:468/WF 179; remarks on Lamy, G 4:579/WF 154). Yet, most of these passages talk only of states generally, not of perceptions in particular, and more importantly, they do not employ explicitly causal language (see Bobro and Clatterbaugh 1996: 415). Hence, they are not extremely strong evidence for the efficacious perception view.

Causation in Monadic Actions

31

appetitions seem like better candidates than perceptions for something that is causally active. After all, Leibniz appears to view appetitions as forces or tendencies. Other Leibniz scholars – myself included – reject both the efficacious perception and the efficacious appetition view and argue instead that the substance itself must be the efficient cause of its states. On this interpretation, which Bobro and Clatterbaugh call the ‘monadic agency view,’ monadic states do not form causal chains: neither appetitions nor perceptions efficiently cause anything, and their efficient cause is always the monad whose states they are.58 This interpretation, then, views monadic efficient causation as agent causation, or substance causation. One reason to adopt the monadic agency view is that Leibniz sometimes claims that only substances can be causally active. In “On Nature Itself,” for instance, he says, “everything that acts is an individual substance” (ONI 9), and in the New Essays he notes, “[f]aculties or qualities do not act; rather, substances act through faculties” (NE 174). Another reason is that individual, instantaneous monadic states may not possess enough reality to be efficient causes. As we will see in the next subsection, they may be mere abstractions from the series of modifications, which is prior to them. There is also evidence that when Leibniz talks of the states of substances as causally active, he is talking loosely and at bottom holds that only the substances themselves can be efficient causes. He writes to Samuel Clarke, for instance, that “properly speaking, motives do not act on the mind as weights do on a balance, but it is rather the mind that acts by virtue of the motives, which are its dispositions to act” (LC 5.15). According to this passage, it is always the mind that acts, and appetitions are merely its dispositions or motives for acting. Along similar lines, Leibniz says that “[w]hen we say that an intelligent substance is moved by the goodness of its object . . . [the object’s] representation acts in the substance, or rather, the substance acts on itself, insofar as it is disposed and influenced by this representation” (COE 21). The expression ‘or rather’ suggests that describing the representation as acting in the substance is correct only loosely speaking. What is actually, or strictly speaking, acting is the substance itself, although the representation does influence it or dispose it to act. I take these passages to be convincing evidence that, strictly speaking, monads themselves – rather than prior states – are the efficient causes of their states. While perceptions do figure into the explanation, they are not properly speaking productive causes. Instead, the monad itself acts, influenced in some nonproductive way by its perceptions. The overall evidence, then, points not only to the presence of efficient causation in the realm of finite monads but also to the identification of all monadic efficient causation with agent causation. This will be significant in 58

Bobro and Clatterbaugh argue for this interpretation at length (1996: 416). See also Bobro (2008: 329) and Schmid (2011: 326f. and 340f.).

Monads and Their Actions

32

several future chapters, particularly in Chapter 5, where I argue that Leibniz’s adherence to agent causation makes his theory of freedom particularly powerful and intriguing. 4.2

Occasional, Formal, and Final Causation

So far I have argued that there is efficient, or productive, causation in the monadic realm and that monads themselves are the efficient causes of their states, though they are influenced by prior perceptions. Of course, this is extremely vague; it does not tell us the exact nature of this influence from prior perceptions. If it is not efficient causation, as I have argued, what is it? Leibniz is not very forthcoming about this in any text I have encountered; he often describes one perception as a consequence or result of a previous perception (e.g. DM 14; letter to Foucher 1686, G 1:382/WF 52; letter to Jaquelot, G 3:468/WF 179; remarks on a letter to Arnauld, G 2:47),59 or says things like “[t]he soul is stimulated [excitée] to its next thoughts by its internal object, that is to say, by its preceding thoughts” (letter to Jaquelot, February 9, 1704, G 3:464/WF 176). In the present subsection, I consider the most appealing ways to understand this influence. One promising proposal is that perceptions are occasional causes for further perceptions. They do not themselves produce any changes, but they are the occasions on which the substance produces new states (see Schmid 2011: 343). This interpretation can be seen as a way of spelling out Leibniz’s vague claim that substances are “stimulated” by previous perceptions. The proposal is attractive despite the fact that Leibniz vehemently objects to occasionalism, because he is not hostile toward all uses of occasional causation. For instance, in a note written around 1689, he describes the influence of one finite substance on another as occasional causation (A 6.4.1640f./MP 81n.1). He also includes definitions of ‘occasional cause’ in at least two of his discussions of causation.60 What Leibniz is opposing, then, is arguably not occasional causation generally but merely the doctrine that God is the sole efficient cause in the universe. One problem with this interpretation, however, is that Leibniz states in a passage from the late 1670s that occasional causation occurs only in voluntary agents.61 Similarly, in a mature text, Leibniz defines occasional causes as being “fitting to us for acting” (C 472; emphasis added). It is not at 59

60

61

See Bobro and Clatterbaugh (1996: 415), Bobro (2008: 326), and Lee (2004: 225f.) for a discussion of what Leibniz might mean by ‘consequence’ or ‘result.’ They all argue that these terms are unlikely to refer to efficient causation. See Leibniz’s “Table of Definitions:” “occasion is a state of affairs, fitting to us for acting, presented without our effort” (C 472). See also “Definitions” (1679): “An occasion is the existence of causal conditions external to the agent under which the agent is ready to act” (A 6.4.310). He says, to be precise, that “an occasion appears to have place only in voluntary actions” (A 6.4.310).

Causation in Monadic Actions

33

all clear, then, whether Leibniz thinks of the influence of perceptions on actions, even in nonrational agents, as occasional causation. Another candidate for describing the influence of perceptions might be formal causation: since Leibniz, as seen, believes that the present state of a substance specifies all future states, one might suspect that a substance’s present perceptions are the formal causes of its future perceptions. Formal causes, on this understanding, are the fundamental explanations for the specific properties of the effect; the formal cause of a perceptual transition explains why the new perception has the form that it does. There is even some textual evidence that Leibniz is happy to acknowledge formal causation in the monadic realm: in Causa Dei, he says that there is a harmony “of formal causes or souls with material causes or bodies” (CD 46). Yet, it is doubtful that a perceptual state could be the formal cause of a subsequent perceptual state. Strictly speaking, after all, perceptual states do not by themselves specify or explain what the subsequent perceptual states are going to be. Instead, current perceptual states specify subsequent states only in combination with the substance’s nature or dispositions.62 In fact, as already seen, monadic natures by themselves specify all of the monad’s states. Presumably, that is why in the passage from Causa Dei, Leibniz identifies formal causes with souls. That strikes me as much more plausible: the substance itself – or, what comes down to the same thing, its nature or substantial form – is the formal cause of its actions.63 Likewise, one perceptual state cannot be the final cause of a subsequent perceptual state.64 Final causation, after all, is the explanation of a change in terms of an end at which the agent aims, or toward which the agent is directed. Yet, my current perception is clearly not an end at which I aim in producing a subsequent perception. Once I attain a certain end, this end can no longer explain further changes since nothing can move me any closer to this end. The final cause is something toward which, rather than away from which, the agent moves. Even in cases where the agent perceives the end of a perceptual transition, it seems mistaken to say that this perception is the final cause of the subsequent state. Instead, the end itself – that is, the object of this perception – is plausibly the final cause.65 Moreover, Leibniz does not appear to hold that monads perceive the ends of all of their perceptual transitions, at least not in the requisite way.66 62 63 64 65

66

I argue this in Jorati (2014a: 759f.). Here I agree with Lee, who argues that created substances are the formal causes of their future states because they fully specify and determine these future states (2006: 447). Bobro suggests something along those lines, however (2013: section 8). This is the traditional understanding of final causation; see e.g. Des Chene (1996: 194) and Pasnau (2001: 303). Yet, admittedly, there are some medieval philosophers who think that the perception of the end is the final cause. True, Leibniz does say that a substance’s present state contains marks of all of its future states. Yet, as seen, he most plausibly means that the present state contains the explanation for, or inclination toward, future states. That does not imply that the agent already perceives these

34

Monads and Their Actions

What, then, is the nature of the influence of prior perceptual states on later perceptual states? Among the options considered – that is, efficient, occasional, formal, and final causation – the most plausible one is occasional causation. Despite evidence that Leibniz wants to restrict occasional causation to voluntary agents, one could argue that he should be happy to acknowledge occasional causation in a looser sense that characterizes the nonproductive influence of any creature’s prior perceptual states on its future states. Yet, I am inclined to go a different way.67 In fact, I suspect that the question itself may turn out to be somewhat wrongheaded. Recall that Leibniz, as seen earlier, often describes finite substances in terms of primitive and derivative forces. Talking about perceptions and appetitions is supposed to be a different way of characterizing this fundamental metaphysical reality. Moreover, as already seen, it is strictly speaking illegitimate to describe monadic activity as the transition from one perceptual state to the next, even though Leibniz himself at times talks like this.68 After all, Leibniz claims that perceptual transitions are continuous (e.g. M 10).69 He also categorically rejects the notion that a continuum can be composed of unextended parts, such as instantaneous entities in the case of temporally extended continua or points in the case of spatially extended continua (e.g. note on Foucher’s objection, G 4:491/AG 146; NE 152). Hence, Leibniz cannot hold that a monad’s series of perceptions is composed of instantaneous perceptual states.70 Nor can such a series be composed of temporally extended static perceptual states, because perceptions are supposed to change constantly. What, then, is the status of individual perceptions? I propose that the right way to think about individual perceptual states is as modalities abstracted from

67

68

69 70

future states now, at least not in a way that could serve as the fundamental explanation for the transition. Even if being inclined toward a future state is sufficient for representing that state in some sense, the inclination would be the ground of that representation. Hence, such a representation cannot at the same time ground the inclination by being the end at which the inclination aims. That would be circular. See Chapter 3 for a discussion of a similar circularity; but see Schmid, who endorses the kind of view I am criticizing in (2011: 350) but seems to have changed his views since (2016). One problem is that the occasional causation proposal is not an entirely satisfactory explanation of the precise nature of the causal relation. The occasional causation postulated by occasionalists is intelligible: x is the occasional cause of y just in case God brings about y because of his knowledge of x. But Leibniz clearly does not want the causation within finite monads to be mediated by God in this way. What, then, does it mean to say that one monadic state is the occasional cause of another? That ultimately sounds like a mere placeholder for a nonproductive causal relation that has yet to be specified. Leibniz uses that locution, e.g., in M 22; two letters to Jaquelot (G 3:464/WF 176 and G 3:468/ WF 179); “Animadversions,” Dut 2.2.134/LSC 23; G 1:382; G 4:579/WF 154; G 4:533/ WF 78. Richard Arthur provides a very helpful discussion of the continuity of monadic states (2001: lxxxviff.). See Whipple (2010b), who discusses this problem in detail.

Causation in Monadic Actions

35

a continuous series of modifications of primitive force.71 This makes perceptions analogous to points, which Leibniz views as mere modalities of lines (e.g. note on Foucher’s objection, G 4:491/AG 146; T 384). If this is correct, the perceptual series is not composed out of individual perceptual states but is rather prior to them. The substance itself, or primitive force, produces and explains this perceptual series as a whole. On this picture, it is not strictly speaking accurate to describe earlier states as giving genuine metaphysical explanations for later states. One modality of the series does not genuinely explain or ground a later modality. Yet, because the series of perceptions is perfectly lawful, it is possible to predict later states from earlier states and the law of the series, which might be the reason why Leibniz sometimes talks as if prior states explain later states. In accordance with the present proposal, a monad’s perceptions are coherent and intelligible as a series not because there are causal or explanatory relations between them but rather because they spring from one unified source. By way of analogy, it is like the coherence of a bird’s song, which is explained not by causal relations among different notes but rather by the fact that the song consists in a continuous stream of air that the bird produces and that is modified in a lawful fashion. However, there is a potential problem with my proposal. Leibniz suggests in a few places that whatever cannot be resolved into primary constituents, or whatever has indefinite parts, cannot be real (e.g. letter to de Volder, June 30, 1704, LDV 303; January 19, 1706, LDV 339). In a letter to Bartholomew Des Bosses, Leibniz puts this rather bluntly: “In actual things, simples are prior to aggregates; in ideal things, the whole is prior to the part” (July 31, 1709, LDB 141; see also a letter to de Volder, October 11, 1705, LDV 327). This might be a problem: arguably, the perceptual series is not a merely ideal thing, which suggests that it cannot be prior to individual perceptions. Yet, we can solve that problem. Leibniz does hold that a body can only be real to the extent that it is resolvable into simple, indivisible things (e.g. letter to de Volder, January 21, 1704, LDV 285f.; June 30, 1704, LDV 303), and he clearly holds the same about substances (e.g. M 2; PNG 1). Nevertheless, I see no reason to apply this to something that is neither a substance nor a body, such as a series of mental states. Leibniz in fact states explicitly in at least one passage that “in actual substantial things, the whole is a result . . . of a multitude of real unities” (note on Foucher’s objection, G 4:491/AG 146). In another passage, he adds that “in real things, namely bodies, the parts are not indefinite” (letter to de Volder, June 30, 1704, LDV 303). Insofar as perceptual series are neither “actual substantial things” nor bodies, I do not see why they could not be prior to 71

Rutherford defends a very similar interpretation (2008: 277ff.). An analogous solution appears to work for appetitions: as derivative forces, they are merely abstractions or modalities of the continuous series of modifications of primitive force.

36

Monads and Their Actions

instantaneous perceptual states. In other words, I do not see why perceptual series would have to be composed of discrete parts in order to be real. After all, a perceptual series is merely a complex modification of primitive force, which in turn is indivisible, unified, and real. Perceptual series do not need to derive their reality from constituent parts.72 5

Conclusion

Even though Leibniz is not always as explicit as one might hope, he identifies monads at the most fundamental level with primitive forces. The primitive forces are unchanging but they ground the entire series of continuous changes that each monad undergoes. The changing states of a monad, which Leibniz views as modifications of the monad’s primitive force, are that monad’s perceptions and appetitions. At any time during its existence, the monad represents external things – in fact, it represents the entire universe – and possesses inclinations toward new representations. Producing these representations constitutes acting. Furthermore, there are excellent reasons to understand Leibniz as endorsing agent causation: monads produce, or efficiently cause, all of their states. States or events themselves do not efficiently cause anything.73 In fact, individual states appear to be mere abstractions from the continuous series of modifications. Of course, this chapter has not answered all important questions about the fundamental nature of monadic change. The sketch it provides does, however, contain the essential building blocks of Leibniz’s philosophy of action. Hence, we can now proceed to a more comprehensive examination of monadic agency. Along the way, more details about the basic ontology of monadic change will emerge. For instance, Chapter 3 will acquaint us far more intimately with Leibniz’s understanding of final causation and its role in explaining monadic change. First, however, we will take a closer look at the spontaneity of monadic actions, that is, at the self-determination of Leibnizian substances.

72

73

Whipple proposes an alternative interpretation, on which it makes sense to talk of next and previous perceptual states, even though monadic perceptions are not composed of instants. His alternative solution is that individual perceptions are temporally extended intervals, whose temporal boundaries are reflectively specified by the monad (2010b: 400ff.). Some other authors discuss this problem as well (Arthur 2001: lxxxviii, McGuire 1976: 313, Rutherford 2008: 279 n58). Nevertheless, as we will see in Chapter 3, states or events can be final causes of monadic change. A monad may engage in a particular type of behavior for the sake of pleasure, for instance.

2

Spontaneity

As briefly mentioned already, it is a central commitment of Leibniz’s metaphysics that all substances or monads possess perfect spontaneity: all states of a given monad originate within it or arise “out of its own depths” (NE 210). Created monads do not strictly speaking interact with each other. Instead, each gives rise to all of its states single-handedly, requiring only God’s ordinary concurrence. This doctrine, which I shall call ‘the spontaneity thesis,’1 is undeniably radical. Many who encounter it – be it today or in Leibniz’s own time – find it absurd. Yet, as we will see in this chapter, Leibniz’s fundamental metaphysical commitments, in particular his commitments to the intelligibility of the created world and the independence of substances, give him excellent reasons for adopting this drastic thesis. One reason why the spontaneity thesis strikes many readers as absurd is that it is difficult to square with our experience. It sure seems that there is interaction between different things in the world; most centrally, we seem able to exert a genuine causal influence on external things, and external things sometimes seem to influence us. When a mosquito bites me, it surely is the cause of the itchiness and pain in my arm; I am not the one who causes my arm to itch and hurt. Likewise, when I consequently kill the mosquito, I am surely the one who causes its body to be crushed; it does not crush itself. Yet, Leibniz claims that his system can accommodate this way of talking; there is a sense in which I am the cause of the mosquito’s demise and in which the mosquito is the cause of my pain. According to Leibniz, we can distinguish different ways in which a change originates in a monad. This, in turn, allows us to distinguish between actions and passions, or between things that we do and things that happen to us.2 By doing 1

2

Note, however, that some authors use ‘spontaneity thesis’ to refer to a slightly different doctrine, namely the doctrine that prior perceptual states cause subsequent perceptual states (see e.g. Bobro 2008: 326, Lee 2004: 225, Sleigh 1990b: 162). As seen in Chapter 1, I believe that the monad itself, rather than its prior perceptual states, is the efficient cause of new perceptions. Yet, I intend the terms ‘spontaneity’ and ‘spontaneity thesis’ to be neutral about whether monadic states are caused by prior states or by the monad itself. I am using these terms merely to imply the absence of external causes. Harry Frankfurt calls the problem of explicating this distinction the “problem of action” (1988: 69).

37

38

Spontaneity

so, Leibniz can deny the absurdly implausible claim that writing a book and feeling pain are actions of mine in a univocal sense. Leibniz’s spontaneity thesis, then, forces him to examine with particular diligence the different ways in which changes can originate in a subject. Attending to those differences turns out to be tremendously beneficial for Leibniz’s philosophy of action, quite apart from the distinction between acting and being acted upon. In particular, as I will show, his strategies for distinguishing different ways in which a state can originate within a monad allow him to claim that some of the desires and emotions that occur in our minds are in an important sense not ours.3 They are external to our true selves, and our attitude toward these mental states is very much like our attitude toward the external world. This in turn allows Leibniz to distinguish between situations in which we act autonomously and situations in which we do not act autonomously because we are controlled by desires or emotions that are quasi external. I contend that Leibniz’s distinction between different ways in which a state can originate in a substance is best understood as a distinction between three types of spontaneity. The broadest and least demanding type, which I will call ‘metaphysical spontaneity,’ is present in all monadic states, even those that we would usually describe as passive. What I refer to as ‘agent spontaneity,’ on the other hand, is present only in a subset of monadic actions, namely those that are perfections of the agent’s nature. The third and final type, which I call ‘rational spontaneity,’ is even more demanding and occurs exclusively in free actions. With the help of these distinctions, which are implicit in Leibniz’s discussions of monadic agency, he can build a bridge between his spontaneity thesis and common sense by acknowledging different levels of independence from external determination. 1

The Spontaneity Thesis

Leibniz uses the term ‘spontaneity’ to characterize actions4 that originate in the agent instead of being imposed on the agent externally. For instance, he says that “an action is spontaneous when its source [principe] is in the agent” (T 301),5 and that the spontaneous actions of a substance arise “out of its own depths” (NE 210; similarly in “New System,” G 4:484/AG 143) or follow 3

4 5

See Frankfurt, who argues that philosophers of action should acknowledge that there is “a legitimate and interesting sense in which a person may experience a passion that is external to him, and that is strictly attributable neither to him nor to anyone else” (1988: 61). See also Watson (2004: 62). Leibniz also sometimes ascribes spontaneity to substances, rather than to actions (e.g. “Metaphysical Consequences” §8, C 14/MP 175). Similar descriptions can be found in CD 108; “Metaphysical Consequences” §8, C 14/MP 175; T 290; “Conversation about Freedom and Fate,” Gr 480/SLT 97; “Table of Definitions,” C 474.

The Spontaneity Thesis

39

from its own nature (e.g. “Conversation about Freedom and Fate,” Gr 481/SLT 98). Conversely, he associates spontaneity with the absence of external determination or impulse (e.g. “Conversation about Freedom and Fate,” Gr 478/SLT 95; Beeley 11). As we will soon see, there are several ways of understanding these general definitions: depending on how narrowly we construe the substance’s “own depths” or nature, and what we take to be internal and external to the agent, we can derive broader and narrower types of spontaneity from these definitions. Yet, for now, I will focus on the broadest type of spontaneity, which Leibniz claims to be present in all nonmiraculous monadic changes: on the fundamental metaphysical level, all natural states of a created substance are caused and explained by it alone, aided only by divine concurrence.6 No other created things are causally or explanatorily involved at this bottom level.7 On the face of it, the spontaneity thesis is extremely counterintuitive. What could possibly have driven Leibniz to adopt it? To answer this question, I find it helpful to consider the alternatives to this doctrine and to understand Leibniz as arguing for spontaneity indirectly, by ruling out all alternatives.8 Consider some creature C. The options for explaining a nonmiraculous state of C appear to be the following: a. Occasionalism: God is the only cause of C’s state b. Bruteness or randomness: C’s state does not have a cause or explanation c. Interaction: C’s state is caused (at least in part) by one or more created thing outside of C d. Spontaneity: C’s state is caused by C or something within C; no other creature is a cause of C’s state These options are exhaustive, because if C’s state has a cause at all, this cause must be either created or divine. If it is created, it is either within or outside of C. Hence, someone who rejects the first three options must embrace the last option, that is, the spontaneity thesis. It is clear that Leibniz rejects the first option. He repeatedly and mercilessly argues against occasionalism, as we saw in Chapter 1. Likewise, it is easy to rule out the second option – the claim that there is no cause or explanation for the states of creatures. After all, it violates the Principle of Sufficient Reason,

6 7

8

For a helpful discussion of the traditional notion of spontaneity and its different uses, see Murray (2005: 195f.) and Hoffmann (1998). God possesses spontaneity as well, as one might expect. Leibniz says so explicitly in “On God and Man,” G 3:32/LGR 291. For passages in which Leibniz explicitly denies interaction among created substances, see e.g. a letter to de Volder, June 20, 1703, LDV 263; NE 210; letter to Des Bosses, August 19, 1715, LDB 349; letter to Remond, November 4, 1715, G 3:657/W 555. According to Garber, the first occurrence of the Leibnizian thesis that created substances do not interact is a text called “De libertate et gratia” (1680–1684?), A 6.4.1458; see Garber (2009: 197). Leibniz employs this strategy explicitly, for instance, in a postscript to a letter to Basnage de Beauval, January 3/13, 1696 (G 4:498f./AG 148). Cover and O’Leary-Hawthorne reconstruct Leibniz’s argument for the spontaneity thesis in a similar way (1999: 181).

40

Spontaneity

which Leibniz treats as axiomatic. The only remaining alternative to the spontaneity thesis, then, is the interaction among creatures. Leibniz’s reasons for rejecting this option are more complicated than his reasons for rejecting occasionalism and randomness, and they are worth examining in some detail. There are some grounds for thinking that Leibniz originally came to deny the causal interaction among created substances because of his infamous complete concept doctrine. According to this doctrine, each finite substance has a complete individual concept or notion that contains everything that ever happens to the substance, or everything that one can predicate of it truly.9 A correlate of this doctrine is Leibniz’s claim that in every true affirmative proposition the predicate is contained in the concept of the subject.10 For instance, if it is true that a dog is in pain, ‘is in pain’ must be contained in the complete concept of this dog. In a few passages from the middle period, Leibniz suggests that his denial of creaturely interaction follows directly from his views on complete concepts and predication: because finite substances have complete concepts, they do not interact. He does this, for instance, in the 1680s text “Primary Truths” (C 521/AG 33; similarly in remarks on a letter from Arnauld, G 2:57/AG 76; a letter to Arnauld, G 2:75/AG 78; DM 14; 32; “Logical-metaphysical principles,” A 6.4.1647/SLT 51). Yet, there is something puzzling about this supposed entailment.11 It seems illegitimate to transition so seamlessly from logical notions such as ‘concept,’ ‘predicate,’ and ‘truth’ to metaphysical notions such as ‘substance,’ ‘action,’ and ‘interaction.’12 A concept is not a substance, and containing a predicate is not the same as causing a state. In fact, the claim that there is a complete concept for every substance seems perfectly compatible with the claim that some changes in this substance are caused by other substances. The concept of a dog, for instance, may contain the predicate ‘is in pain,’ but it is not obvious why having such a concept should rule out the possibility that another creature inflicts this pain. Indeed, since Leibniz is sometimes happy to say that the 9

10

11 12

This doctrine is prominent in the middle period but almost never comes up in mature texts. For examples of middle-period passages endorsing the doctrine, see e.g. DM 8; “A Specimen of Discoveries,” A 6.4.1618/MP 77; “Primary Truths,” C 520/AG 32. A rare mention in a late text occurs in Leibniz’s notes on Temmik (Mugnai 1992: 156 and 2012: 197). Again, Leibniz expresses this principle in many texts of the middle period, for instance DM 8; “Primary Truths,” C 518f./AG 31; “On Freedom,” A 6.4.1654/AG 95. It occurs in only a very small number of mature texts; see, for instance, “Metaphysical Consequences” §1, C 11/MP 172. Garber argues that Leibniz drops this type of argument after just a few years, perhaps because he realizes that it is problematic (2009: 198f.). Commentators often follow Leibniz in moving back and forth between logic and metaphysics in this way. I believe that such moves are ultimately legitimate in many instances (though not in all; see Bobro 2004: 63 for a problematic example). Yet, it is crucial to acknowledge that these transitions are not trivial. See Nachtomy, who stresses the importance of distinguishing between concepts and agents in the context of necessitation (2007: 148).

The Spontaneity Thesis

41

concept of each substance involves or depends on the concepts of all its worldmates (e.g. letter to de Volder, July 6, 1701, LDV 209), it might seem sensible for him also to claim that at the metaphysical level there are causal relations between substances that correspond to these relations among concepts. For instance, if the concept of the dog involves or needs the concept of the person beating it and causing it pain, one might expect the dog itself to depend on this person with respect to its pain. What Leibniz needs, then, is an additional premise that suitably connects the logical to the metaphysical level and justifies inferring the spontaneity thesis from the complete concept doctrine. Luckily, I believe we can supply such a premise, namely a metaphysical correlate of Leibniz’s claims about complete concepts and predication. This metaphysical correlate is the doctrine that just as the complete concept of a substance contains everything that can be predicated of the substance, so the substance itself must contain marks and traces of everything that ever happens to it (see Chapter 1). For instance, Leibniz holds that if it is true of Caesar the day he is born that he will die on the Ides of March, this must not only be part of Caesar’s complete concept; there must also be something about Caesar’s soul at birth in virtue of which it is true of him.13 The idea seems to be that if something is true about a substance today, something about the way this substance is today must ground this truth. Even at birth, then, Caesar’s soul must already contain marks of his death in such a way that a sufficiently intelligent mind could know this just by looking at Caesar’s soul at birth. This, in turn, means that in order to understand Caesar’s death or anything else about Caesar’s future, we do not need to look beyond Caesar himself. Unlike the doctrine of complete concepts, this metaphysical correlate does appear suitable for motivating the spontaneity thesis: because the explanation or truth-maker for Caesar’s death is already in Caesar’s soul, there is no need to posit an external cause – at least not a finite one – of Caesar’s death.14 As Leibniz asks rhetorically in a letter to Des Bosses, “why should a monad give to a monad what [the latter] already has?” (August 19, 1715, LDB 349).15 Hence, all changes in a monad come from its own depths, or arise spontaneously. 13

14

15

See Garber, who argues along similar lines: if a complete concept is predicated of an individual, there must be something in this individual in virtue of which the concept belongs to it (2009: 186). The relation between explanation and causation in Leibniz is complicated (see Di Bella 2008). Leibniz does not reduce causation to explanation. Yet, he does appear to hold that all causes provide explanations. He says, for instance, that “the cause is nothing other than the real reason” (G 7:289/C 533). See also “A Specimen of Discoveries,” A 6.4.1620/MP 79: “Causes are assumed, not from a real influx, but from the need to give a reason.” Similarly, in the draft of a letter to Arnauld, Leibniz claims that it would be pointless or useless (inutile) for one finite substance to have a real, physical influence on another (G 2:71); in the letter to Des Bosses, he calls it “superfluous” (superfluum; LDB 349). See also Cox, who points out that in order to argue against interaction from the doctrine of complete concepts Leibniz has to assume that there is no causal over-determination (2002: 189).

42

Spontaneity

In this way, Leibniz’s theory of complete concepts and predication does imply, via its metaphysical correlate, that creatures do not causally interact. After all, there is a close parallel between the logical and the metaphysical level: just as one can in principle derive everything that can be predicated of a substance from its complete concept, one can in principle explain all states of a substance from its nature, substantial form, or law of the series. I believe that this parallel between complete concepts at the logical level and natures or substantial forms at the metaphysical level is ultimately behind Leibniz’s claim that spontaneity is a consequence of his logical commitments.16 Leibniz does not always argue for the spontaneity thesis from his theory of complete concepts, and in fact, I believe that he has several other excellent reasons for denying the causal interaction of creatures. The reason he appears to cite most often is that there is something unintelligible about the genuine causal influence of one finite substance on another. In most passages he does not explain what exactly is supposed to be unintelligible about inter-creature causation but merely states that it is.17 Those passages can be read as challenging his opponents to provide an intelligible – that is, presumably, naturalistic – account of how one simple, unextended, soul-like substance could causally interact with another. An extremely interesting passage from an undated letter to Isaac Jaquelot might shed further light on the issue. There, Leibniz argues that while body–body interaction is at least somewhat intelligible, mind–body interaction is completely unintelligible. We can understand this argument as also explaining why mind–mind – and in fact, monad–monad – interaction is unintelligible.18 The passage runs as follows: There is immediate contact between bodies, and we understand how that can be, and how, since there is no penetration, their coming together must alter their movement in some way. But we see no such consequences with the soul and the body: these two do not touch, and do not interfere with one another in an immediate way which we can understand and deduce from their natures. (G 6:570/WF 199)

Plausibly, what Leibniz says in the last clause of this passage applies also to any two finite monads: because they are unextended and immaterial, there is nothing in their natures that would explain why they should immediately interfere with one another. What goes on in one mind or monad does not by itself put any constraints on what goes on in another, unless, of course, one presupposes that they have to harmonize. Yet, for Leibniz, inter-monadic 16

17 18

There is some textual support for this: in “A Specimen of Discoveries,” Leibniz states that positing the interaction of created substances is superfluous because all of the states of a finite substance must “flow from its nature and its notion” (A 6.4.1620/MP 79). Examples include the following texts: letter to de Volder, January 1705(?), LDV 319; letter to Des Bosses, August 19, 1715, LDB 349; “New System,” G 4:483/AG 142f.; M 11. In fact, in the mature period, mind–body interaction would come down to monad–monad interaction.

The Spontaneity Thesis

43

harmony is not required by the natures of created monads; instead, it is something that God brings about.19 As a result, this argument is quite persuasive: nothing that happens in one finite monad directly constrains what happens in another. Metaphysically speaking, any state in one finite substance is compatible with any state in another. Hence, a state in one substance cannot cause, or render intelligible, a state in another.20 Another reason why Leibniz calls causal interaction unintelligible might be that he is assuming – quite plausibly – that most of his readers would agree that created soul-like things cannot interact directly with one another. Even Cartesians, after all, do not appear to hold that created thinking substances interact with each other, except via their bodies.21 Telepathy, as it were, was not widely accepted, to the best of my knowledge. Another possibility is that such interaction would constitute a robust kind of relation, against whose reality Leibniz repeatedly argues elsewhere.22 There are also a few passages in which Leibniz explicitly provides reasons for the unintelligibility of causal interaction, most notably in the “Monadology”: There is . . . no way of explaining how a monad can be altered or changed internally by some other creature, since one cannot transpose anything in it, nor can one conceive of an internal motion that can be excited, directed, augmented, or diminished within it . . . The monads have no windows through which something can enter or leave. Accidents cannot be detached, nor can they go about outside of substances, as the sensible species of the Scholastics once did. Thus, neither substance nor accident can enter a monad from without. (M 7; see also letter to Basnage de Beauval, January 3, 1696, G 4:499/WF 63)

Leibniz’s point in the last two sentences of this passage appears to be that since accidents or modes cannot exist without substances, they cannot pass from one substance to another.23 After all, that would require them to become detached 19

20 21

22 23

As we will see in Section 3, there is a sense in which created substances interact, and as we will see in Chapter 3, there is even a sense in which merely possible substances constrain one another. Yet, on my interpretation, neither of these takes place merely in virtue of the natures of creatures; the constraints that created substances or possible substances place on each other stem ultimately from God’s preference for harmony. Interestingly, though, to the extent that they constrain each other, they do undermine each other’s agent spontaneity, as I will call it. The question of how God can interact with finite monads is fascinating and difficult, but I cannot discuss it here. Some Scholastics, in contrast, appear to hold that immaterial creatures can interact directly under certain circumstances. They claim, for instance, that angels – who do not have bodies – can directly communicate with each other (see e.g. Suárez, De Angelis 2.26). Likewise, Aquinas claims in at least one passage that higher angels can act on lower angels and on our souls (Quaestiones de quolibet III, q. 3 a. 2 co., in CT). For excellent recent discussions of Leibniz’s treatment of relations, see e.g. Jauernig (2010) and Mugnai (2012). Cover and O’Leary-Hawthorne helpfully spell this out by noting that if accidents could migrate from one substance to another, they would be separable. Separability, however, is supposed to be a mark of substances (1999: 181). This is how Leibniz appears to argue in a rare humorous passage from his last letter to Clarke: after comparing the view that properties can pass from one

44

Spontaneity

from the first and attached to the second. Intra-substantial causation, on the other hand, does not suffer from this problem and is perfectly intelligible; it does not require that accidents “go about outside of substances.” 24 This last argument, of course, only rules out inter-creature causation construed as a transfer of properties, or what Leibniz sometimes calls a “physical influx.” 25 Since this is not the way that most of Leibniz’s contemporaries and predecessors understand causal interaction (see O’Neill 1993), it may seem like a straw man. To avoid that charge, Leibniz has to show that all alternative accounts of inter-creature causation are also unintelligible. In fact, the first sentence of the above passage from the “Monadology” may be meant to rule out one alternative account of causation: Leibniz appears to argue that because monads are simple and unextended, they cannot interact by causing motion in one another, or by rearranging each other’s parts. In other words, a mechanistic account of their interaction is not possible. Of course, that still does not rule out all other possible accounts of interaction. Yet, if Leibniz can indeed appeal to a broad consensus among his readers about the impossibility of direct, quasitelepathic interaction between created incorporeal substances, he can safely assume that his readers will go along. Moreover, the argument we considered earlier – based on the claim that finite monads do not constrain one another – is very promising. The final reason that I would like to explore for Leibniz’s denial of creaturely interaction is the independence or self-sufficiency of substances, which is closely related to what I above described as the metaphysical correlate of the complete concept theory. Like Descartes and Spinoza, Leibniz holds that in order for something to count as a substance, it must be independent of other things, or at least of other created things (“Philarete and Ariste,” G 6:586/AG 262).26 For Leibniz, it seems, this means that a created substance cannot depend for its states on other creatures. As he puts it in a draft of a letter to Arnauld, “it would also be useless [for one substance to have a physical influence on another], since each substance is a complete being [un estre accompli], which itself is sufficient [se suffit luy même] for determining everything that must happen to it in virtue of its own nature” (1686, G 2:71). In other words, if a creature depended causally on other created things for some of its states, it would lack the independence and self-sufficiency required for

24

25 26

substance to another to the claim that “subjects will leave off their accidents, like clothes, so that other subjects may put them on,” Leibniz asks, “[a]t this rate how shall we distinguish accidents and substances?” (LC 5.39). In fact, as Leibniz sometimes points out, we experience this intra-substantial causation within ourselves when we produce volitions and thoughts, and hence, it is nothing mysterious (ONI 10; letter to Masham, May 1704, G 3:340/WF 206). For a thorough investigation of Leibniz’s reasons for thinking that inter-substantial causation would involve a transmission of accidents, or an influxus physicus, see O’Neill (1993). Of course, as seen in Chapter 1, independence is not the only mark of substancehood.

The Spontaneity Thesis

45

substancehood; it would be incomplete and dependent. After all, if any of its states even partially depended on, or had to be explained with reference to, another substance, it would need that substance (see Rutherford 1995a: 135). The fact that Leibniz ties substancehood so closely to self-sufficiency, independence, and spontaneity raises a serious worry, however. After all, Leibniz endorses the traditional theological doctrine that creatures depend on God for all of their states, albeit not as completely as occasionalists would have it: Leibniz holds that without God’s concurrence or cooperation, creatures would not be able to act at all. God needs to causally contribute to all of our actions. One might worry that Leibniz has no principled reason for claiming that this type of dependence is compatible with our spontaneity and substancehood, while simultaneously maintaining that a partial causal dependence on other created things would undermine our spontaneity and our status as substances. I will argue in Chapter 4, however, that Leibniz has a rather convincing response to this worry. So far, we have seen that Leibniz has excellent reasons for adopting the spontaneity thesis: some of his deepest metaphysical commitments suggest that there cannot be genuine interaction among created monads and that all natural actions of a monad must instead originate within it. In the remainder of this chapter, I will argue that the spontaneity thesis, though undeniably radical, is not quite as radical as it may initially seem. It leaves room for distinguishing different ways in which an action can originate within the agent. In fact, there are good reasons to interpret Leibniz as making a distinction between different kinds of spontaneity, at least implicitly. This renders the spontaneity thesis much more palatable because, as some commentators point out, the general type of spontaneity seems too broad in many contexts. After all, this general type is present in all monadic actions to the same degree (e.g. “New System,” G 4:484/AG 143; reply to Bayle, G 4:565/PT 251; T 65). There appears to be a need for a narrower type of spontaneity that allows Leibniz to distinguish constrained actions from unconstrained actions, voluntary from nonvoluntary actions, and actions from passions.27 Some passages suggest that there are more demanding types of spontaneity that, unlike the broad type of spontaneity, are neither ubiquitous nor always present perfectly. For instance, Leibniz talks of “the imperfection present in our knowledge and in our spontaneity” (T 288), which makes sense only if in addition to the broad type there is a more demanding type of spontaneity that we do not always possess to the highest degree.28 In a reply to Bayle, Leibniz 27 28

Other interpreters agree; see e.g. Murray (2005: 197f.), (1995: 80), Rutherford (2005: 157), Sotnak (1999: 216), but cf. Greenberg (2005: 227). See also Leibniz’s notes on Jaquelot’s letter of September 6, 1704, Gr 490/WF 193, where Leibniz explicitly talks of degrees of spontaneity, and “Commentary on Burnet” §11(e), where he says that there is some spontaneity in all substances (DPG 63).

46

Spontaneity

first states that the soul is like a little world that is “as imperturbable as the great world” because “there is spontaneity in the confused as well as in the distinct.” Yet, he then goes on to acknowledge another sense in which it is correct “to call those things that consist in confused thoughts . . . ‘perturbations’ or ‘passions’” (G 4:565). In this latter sense, one might infer, the soul is not always perfectly spontaneous. These and similar reasons have led some interpreters to distinguish two types of spontaneity. Yet, I do not think that goes far enough: the evidence points to three types, and we need these three types in order to distinguish the most important kinds of monadic actions. 2

Metaphysical Spontaneity

As already noted, Leibniz’s general definition of a spontaneous state as one arising from the agent’s “own depths” or one that follows from the agent’s nature allows for several readings. Depending on how much we take to be included in the agent’s nature or “depths,” the definition can yield more and less broad types of spontaneity. When Leibniz states that all monadic actions are perfectly spontaneous, he appears to intend the broadest reading of this definition. On that reading, the absence of real, physical influences from other created substances is sufficient for spontaneity. I will call this broadest and least demanding type of spontaneity ‘metaphysical spontaneity’29 because it is a consequence of the fact that, metaphysically speaking, created monads do not interact, and because Leibniz himself says that “in a certain metaphysical sense . . . there is spontaneity in everything which happens to us” (“Conversation about Freedom and Fate,” Gr 481/SLT 98). Since finite monads do not interact at all, their actions are always perfectly spontaneous in this metaphysical sense. Each state of a monad is caused and made intelligible by something inside the monad itself, namely by its perceptions and appetitions – or, at bottom, its nature or primitive force. Nothing but the monad itself and God’s ordinary concurrence is causally responsible for the monad’s states. 3

Agent Spontaneity

In addition to metaphysical spontaneity, Leibniz seems to implicitly recognize a more demanding type of spontaneity that I will call ‘agent spontaneity.’30 I am basing my distinction between metaphysical and agent spontaneity on 29 30

Murray calls this type “spontaneity-for-free” (2005: 198), and Rutherford calls it “monadic spontaneity” (2005: 157). I am borrowing this term from Rutherford, who however uses it more narrowly than I do. He reserves it for desire-based actions (2005: 169) and appears to understand desires as conscious appetitions. I am applying it even to some actions of bare monads who do not have conscious appetitions.

Agent Spontaneity

47

Leibniz’s distinction between actions in the broad sense and actions in a narrower sense: monads possess agent spontaneity whenever they are acting in the narrow sense, while they possess metaphysical spontaneity whenever they are acting in the broad sense. A passage from the New Essays explains the distinction between the broad and the narrow sense very clearly. In that passage, Leibniz’s spokesperson Theophilus starts out by defining actions in the broad sense, tying that notion to what I call metaphysical spontaneity: [A]nything which occurs in what is strictly a substance must be a case of ‘action’ in the metaphysically rigorous sense of something which occurs in the substance spontaneously, arising out of its own depths [de son propre fond]; for no created substance can have an influence upon any other, so that everything comes to a substance from itself. (NE 210)

Hence, every change that arises from a substance’s “own depths,” and thus every metaphysically spontaneous state, is an action in the broad sense. Because there is no true influence among finite substances, all natural changes within a monad are actions in this sense. Theophilus then points out that in another sense, only some of these changes are actions while others are passions: “But if we take ‘action’ to be an endeavour towards perfection, and ‘passion’ to be the opposite, then genuine substances are active only when their perceptions . . . are becoming better developed and more distinct, just as they are passive only when their perceptions are becoming more confused” (NE 210). According to this passage, then, there is a sense in which being an agent requires more than just metaphysical spontaneity: it additionally requires that one’s perceptions are becoming more distinct.31 When a monad’s perceptual distinctness decreases, it is not acting in this sense – it is passive. Theophilus then elaborates further on this distinction between actions and passions: “any change through which [a substance] comes closer to its own perfection” is an action of that substance and can be attributed to the substance itself, whereas “any change in which the reverse happens” is a passion of that substance and can be attributed to an outside cause (NE 211; see DM 15). Thus, passions occur – unsurprisingly – when a substance is subject to external influences, that is, when something outside of it is acting on it. Of course, metaphysically speaking, there is no interaction among created monads. From that point of view, monads are always agents and never patients. Yet, Leibniz frequently stresses that there is an important sense in which there is mutual influence among finite monads; he sometimes calls this “ideal influence” (French: influence idéale). In the “Monadology,” for instance, after once again distinguishing between acting and being acted upon in terms 31

The reference to “an endeavour towards perfection” in the passage already suggests a close connection between this kind of agency and teleology, which I will discuss in Chapter 3.

48

Spontaneity

of perfection and imperfection, or distinct and confused perception (M 49), he adds that “the influence of one monad over another can only be ideal, and can only produce its effect through God’s intervention, when in the ideas of God a monad reasonably asks that God take it into account in regulating the others from the beginning of things” (M 51; similarly in T 66). Hence, even though in the strict sense there is no interaction, Leibniz acknowledges another “ideal” sense in which there is, and this is the basis for his distinction between actions and passions.32 That the distinction between actions and passions can be drawn either in terms of the relative perfection of the substances involved or in terms of God’s regulation of things at creation should not be surprising. As Leibniz explains in the Theodicy, substances interact “ideally and in the reasons of things, in that God in the beginning ordered one substance on the basis of another, in accordance with the perfection or imperfection that there is in each” (T 66). God’s regulation of things thus corresponds to the perfection of these things. Their perfection, in turn, plausibly correlates with the distinctness of their perceptions. As Leibniz says in “Supplement to the Explanation of the New System,” “confused thoughts are a mark of our imperfection, passions, and dependence on the assemblage of exterior things or on matter, whereas the perfection, force, control, liberty, and action of the soul consists principally in our distinct thoughts” (G 4:574/WF 140). God has ordered created things in such a way that actions in one substance always correspond to passions in another, or in other words, in such a way that perfections or distinct perceptions in one substance always correspond to imperfections or confused perceptions in another.33 Leibniz apparently holds that we can adequately describe the order of divine reasoning as follows: insofar as one substance that God considers creating is perfect or active, it “asks” God – or more literally, gives God a reason – to also create another substance with a corresponding imperfection or passion (M 51). After all, God does not will imperfections for their own sake. He only permits imperfections to exist insofar as they are required for the overall harmony and perfection of the world.34 As a result, imperfections are in need of a special explanation, and 32

33 34

See Puryear 2010, who argues that ideal causation for Leibniz is not merely apparent causation but a genuine type of causal interaction. He claims that it is best understood as a form of final causation. I partly agree with Puryear: ideal causation is not merely apparent; it is well-founded in features at the fundamental metaphysical level, as we will soon see. I also agree that ideal causation can be understood as corresponding to a type of final causation, as I in fact argue in Chapter 3. Yet, I disagree with Puryear’s conclusion that ideal causation is a genuine type of action or activity. Final causation is genuine causation, but it is not a kind of activity. Hence, the ideal influence of one monad on another should not be called a genuine interaction. Moreover, as I argue in Chapter 3, the type of final causation corresponding to ideal causation is one step removed from the final causation at the fundamental metaphysical level. For Leibniz, “action and passion are always reciprocal in creatures” (T 66; see also M 52). See e.g. T 22ff. and CD 25, where Leibniz argues that God antecedently wills only good things.

Agent Spontaneity

49

in a sense they exist because of the perfections to which they correspond. This means that on the ideal level of explanation an action or perfection in one monad explains the occurrence of a corresponding passion or imperfection in another monad (see M 50; NE 211). As Leibniz puts it in one text, things that are more perfect “make the rule for other things in the harmony of the universe” and can therefore be said to act on other things (Gr 481/SLT 98). Consequently, the ideal interaction among creatures – that is, the exact correspondence between creaturely perfections and imperfections – is a result of the harmony that God has preestablished among his creatures. Hence, it is a crucial aspect of the world’s perfection. Leibniz makes this explicit, for instance, in “Explanation of the New System,” where he notes that it is perfectly acceptable to talk of causal interaction between body and soul as long as one is aware that this causation occurs only “in consequence of the laws of harmony” (G 4:495/WF 49) or “in virtue of a ‘pre-established harmony’” (G 4:496/WF 51; similarly in a letter to l’Hôpital, September 20, 1695, GM 2:298/WF 58). My notion of agent spontaneity, which builds on the distinction between ideal passivity and activity, therefore captures an immensely important feature of the world. It is clear, as seen, that Leibniz acknowledges two senses in which a finite monad can be independent of external influences: the metaphysical sense and the ideal sense. I propose to interpret Leibniz as acknowledging two types of spontaneity, corresponding to these two types of independence from external influences. As a matter of fact, there is some textual evidence for this interpretation. For instance, Leibniz says that “insofar as we are imperfect, we are said to suffer, and to be subject to external things, although in a certain metaphysical sense . . . there is spontaneity in everything which happens to us” (“Conversation about Freedom and Fate,” Gr 481/SLT 98). In this passage, Leibniz explicitly distinguishes spontaneity “in a certain metaphysical sense,” which belongs to us at all times, from something else – presumably from spontaneity in another sense, in which we are sometimes determined by external things. He makes a very similar point in the Theodicy: “spontaneity . . . belongs to us insofar as we have within us the source [principe] of our actions . . . It is true that the impressions of external things often divert us from our path, and that people commonly believed that – at least in this respect – some of the sources of our actions were outside of us” (T 290; see also T 65). He continues the passage by pointing out that even though it is not true in the strict sense that the sources of such actions are not within us, there is a certain sense in which it is true; in fact, he acknowledges that we sometimes have to speak in this way.35 Given that spontaneity means that the 35

In several texts, Leibniz compares this to the way we speak about the sun rising and setting in everyday conversations, even though we know that the Copernican account is strictly speaking correct (T 65; NE 74; “Explanation of the New System,” G 4:495/WF 49).

50

Spontaneity

source of the action is within the agent, this passage is an excellent support for a narrower type of spontaneity. The distinction between actions and passions, or between metaphysical and agent spontaneity, is also central to the way we understand the world and our place within it: we see ourselves as sometimes acting on other things and sometimes being acted upon.36 We experience actions of the former type as originating within us and actions of the latter type as originating outside of us. Leibniz thinks that based on preestablished harmony he can “give a true and philosophical sense” to this ordinary way of understanding the world (T 66): he can say that one substance is acting on another insofar as God has adjusted the latter to the former (e.g. “New System,” G 4:486/AG 145; M 52), or insofar as the former is more perfect than the latter with respect to that particular event. Interestingly, this also means that agent spontaneity corresponds to the way in which some of Leibniz’s contemporaries and predecessors use the term ‘spontaneous.’37 The concept of ideal influence, after all, is supposed to give a true and philosophical sense to ordinary statements about the causal influence of one finite substance on another. As a result, agent spontaneity consists in the absence of what we ordinarily view as the causal influence of other substances, and being exempt from such influence is one traditional conception of spontaneity.38 In short, my distinction between agent spontaneity and metaphysical spontaneity simply piggybacks on Leibniz’s distinction between actions and passions, or between states that are ideally influenced by other agents and states that are free from such external influence.39 Metaphysical spontaneity requires merely the absence of real external influences, while agent spontaneity additionally requires the absence of ideal external influences. Both actions and 36 37

38

39

See Seidler, who in passing distinguishes “monadic, metaphysical spontaneity” from “actual independence of action in the experienced, phenomenal world” (1985: 19). Leibniz apparently coined the Latin noun ‘spontaneitas,’ but the English term ‘spontaneity’ was already used by others (see Lærke 2009: 944n26). Moreover, the Latin terms ‘sponte’ and ‘spontaneus’ had been in use for centuries. To be precise, there are at least two traditional conceptions of spontaneity: a broad one that corresponds to Leibnizian agent spontaneity and that requires merely the absence of external determination and a narrower one that additionally requires that the agent have knowledge of what she is doing. For the former conception, see e.g. Molina, Concordia pt. 2, disp. 2, §2; Aquinas, Sentencia libri de anima, lib. 2 l. 7 n. 6, in CT; ST IaIIae q102 a6 ad2. However, the narrower conception was more common and goes back to Aristotle’s definition of hekousios (Nicomachean Ethics 1111a22–24). Note that this narrower conception of spontaneity, the way it is used by Scholastics like Aquinas, is still broader than what I will below call ‘rational spontaneity’ because it does not require rationality; animals and young children can act spontaneously without being rational. See e.g. Aquinas, DV q24 a2 ad1. For a good overview of the history of the terms ‘spontaneous’ and ‘spontaneity,’ see Hoffmann (1998). Murray makes a similar distinction: he claims that Leibniz acknowledges not only “spontaneityfor-free” – that is, metaphysical spontaneity – but also a narrower type not present in acts that “can be attributed to an external source or quasi-cause” (2005: 199).

Agent Spontaneity

51

passions are metaphysically spontaneous, whereas only actions are agentspontaneous. Before turning to the third and narrowest type of spontaneity, some features of agent spontaneity deserve to be stressed and some potential misunderstandings need to be cleared away. First of all, even though Leibniz sometimes talks of God adjusting one substance to another, it is important to note that this must not be understood literally. God finds the essences or complete concepts of possible substances in his intellect and actualizes some of them by creating substances that correspond perfectly to these concepts. He does not add or alter anything: he “leaves them exactly as they were in the state of pure possibility, that is, he does not change anything, neither in their essence or nature, nor even in their accidents” (T 52). Hence, when Leibniz talks of God adjusting one substance to another, he must be referring metaphorically to the order of reasons that led God to choose one particular creature for actualization over another (see T 66; remarks on Lamy, G 4:578/WF 153). As already seen, God chooses to actualize a creature with a particular imperfection because of a corresponding perfection in another creature. He does not literally give imperfections to particular creatures.40 Similarly, it is crucial to keep in mind that explaining an imperfection in one substance in terms of a perfection in another substance is not a bottom-level explanation. At the fundamental level, even passions are explained in terms of the substance that is undergoing this change. This must be the case because strictly speaking, as already seen, all substances are perfectly spontaneous and hence produce as well as explain all of their states themselves. It is only when we consider God’s reasons for combining these monads into a harmonious world that we can talk of one creature’s state as being the reason for another creature’s state. It is also important to note that what matters for this type of influence is not the overall perfection of a substance but rather its perfection with regard to some particular event. For instance, a mosquito can act on you even though it has less overall perfection. With respect to a mosquito bite, you are passive and the mosquito is active because your role in that event is less perfect than the mosquito’s: while the bite is a perfection for the mosquito, it is an imperfection for you. Furthermore, the discussion of one monad influencing another ideally brings to the forefront something already implicit in my description of metaphysical spontaneity: Leibnizian spontaneity has a causal dimension and an explanatory dimension. The explanatory dimension is particularly explicit in Leibniz’s characterizations of ideal influence and hence in my account of agent spontaneity. Leibniz says, for instance, that “there are certain things in the soul that can only be explained in an adequate way through external things, and to that 40

I will say more about this in Chapter 5, Section 1.

52

Spontaneity

extent the soul depends upon external things” (Beeley 11/AG 279; similarly in NE 211; M 52; “New System,” G 4:486/AG 145). Yet, this explanatory dimension is also present in metaphysical spontaneity, as seen earlier: Leibniz holds that a substance is metaphysically spontaneous when a complete explanation of its action can be given without invoking anything external to this substance.41 This should not be surprising given that Leibniz generally associates causation and explanation very closely.42 Finally, what I have said so far may sound as if agent spontaneity applied only to individual monads. Yet, interestingly, one can also apply this notion to organisms. Organisms are agent-spontaneous to the extent that they are not ideally influenced by other organisms or bodies.43 Talking that way can be extremely useful. For instance, it allows us to distinguish between someone being pushed into a swimming pool and someone jumping into the said pool. An individual monad, on the other hand – for instance, the central monad of an organism – is agent-spontaneous to the extent that it is free from the ideal influence of all other created external things, including its own body. This latter notion is useful because it allows us to describe cases in which a soul fails to act in accordance with its perfection because of its own body. Leibniz does, after all, sometimes talk about the ideal influence of the body on the soul and of the soul on the body. In the New Essays, for instance, Leibniz’s spokesperson says that “in involuntary actions the mind depends on the body (to put the matter accurately), in other actions the mind is independent and even makes the body depend upon it” (NE 177). He goes on to say that this dependence “consists in God’s taking account of one of them in regulating the other, or taking more account of one than of the other according to the inherent perfections of each” (NE 177; similarly in T 66). Hence, in some contexts it is important to determine whether the soul possesses agent spontaneity – that is, whether it is free even from the ideal influences of its own body – while in other contexts it is more helpful to view the organism as a unit and to determine whether it possesses this type of spontaneity. 4

Rational Spontaneity

The third and final type of spontaneity is even narrower than agent spontaneity, and I call it ‘rational spontaneity.’ It is quite easy to characterize this kind of 41 42

43

I argue in Chapter 4 that in the context of specifying the necessary conditions for substancehood, the explanatory dimension is particularly important. See Rutherford (1995a: 135) for a good discussion of the close connection between being the source of one’s actions and being the ultimate explanatory principle of one’s actions. But cf. footnote 14. For an interesting discussion of the question whether nonsubstantial bodies interact or possess spontaneity, see Brown (1992).

Rational Spontaneity

53

spontaneity: it belongs to actions to the extent that they are determined by rational rather than nonrational inclinations and perceptions. For Leibniz, as we will see in Chapter 5, this is equivalent to saying that rational spontaneity belongs to actions to the extent that these actions are voluntary or free. Thus, a monadic action possesses rational spontaneity insofar as it is exempt not only from the ideal influence of other substances but also from determination by factors external to the rational aspects of the agent’s nature. These rational aspects comprise both rational perceptions and rational appetitions. As a result, rational spontaneity is determination by the agent’s intellect and will. After all, the intellect is the capacity for rational thought and judgment, whereas the will is the rational appetite, or the inclination to act in accordance with the deliverances of the intellect.44 Rational spontaneity is hence rational self-determination. Similarly, agent spontaneity can be described as agential self-determination, or determination by perfections of the agent’s nature; metaphysical spontaneity can be described as metaphysical self-determination, or determination by any aspect of the agent’s nature. Any time an agent possesses rational spontaneity, she also possesses agent spontaneity because reason and freedom are perfections.45 Yet, many actions are agent-spontaneous but not rationally spontaneous. The mosquito that bites you, for instance, acts with agent spontaneity but without rational spontaneity. After all, mosquitos are not rational and hence incapable of free, rational agency. Likewise, instinctive human actions, such as scratching a mosquito bite in one’s sleep, are agent-spontaneous but not rationally spontaneous. Rational spontaneity is therefore narrower than agent spontaneity, which in turn is narrower than metaphysical spontaneity. Rational spontaneity is crucial for understanding Leibniz’s theory of freedom. In fact, he even uses the term ‘rational spontaneity’ in “Conversation with Steno” (1677) when approvingly describing what he takes to be the traditional, pre-Molinist understanding of freedom.46 That freedom consists in rational 44

45

46

See “Table of Definitions:” “Will is the conatus of the person who understands [conatus intelligentis]” (C 498). See also A 6.1.284/L 91n11: “To will is nothing but the striving [conatus] arising from thought, or to strive for something which our thinking recognizes as good.” Finally, see T 22. Leibniz says, for instance, that “we are free insofar as we are determined to follow the perfection of our nature, which is to say, reason” (“Conversation about Freedom and Fate,” Gr 481/SLT 98) and that in voluntary actions, “the soul is exercising a perfection of its nature, which is freedom” (remarks on Lamy, G 4:579/WF 154). In that text Leibniz says, “According to the ancients the free differs from the spontaneous as species from genus; surely, freedom is rational spontaneity [spontaneitas rationalis]” (A 6.4.1380/CP 123). In a similar vein, Leibniz says in much later texts that “freedom is the same as spontaneity with reason [spontaneum cum ratione]” (“Critical Thoughts on Descartes,” G 4:362/L 389), that “freedom is the spontaneity of the person who understands [spontaneitas intelligentis]” (G 7:108, ca. 1692), and that “freedom is the spontaneity of the person who deliberates [spontaneitas consultantis]” (“Table of Definitions,” C 498). He also describes freedom as deliberate spontaneity, or as spontaneity combined with the understanding of the

54

Spontaneity

spontaneity, or the agent’s self-determination by her will and intellect, also becomes clear in the New Essays, where Theophilus says, “[i]n reasonings about the freedom of the will, or about ‘free will,’ the question is . . . whether [the] will itself is sufficiently independent” (NE 181; similarly in “On Freedom from Necessity in Choosing” (1680–1684), A 6.4.1453/SLT 108). The independence of the will that is at issue here is, of course, not an independence from all determination but rather an independence from problematic, freedomundermining kinds of determination, or from undue influences. After all, Leibniz vehemently denies that the will can act without a sufficient reason (e.g. letter to Morell, September 29, 1698, Gr 139; LC 4.2). That kind of independence would be an even more demanding type of spontaneity that Leibniz unequivocally rejects.47 Instead, the will is “sufficiently independent” and free insofar as it is determined by rational factors, that is, by the intellect’s judgments concerning the relative goodness of different options. Accordingly, Leibniz opposes acting freely to being enslaved by the passions and by nonrational inclinations; in fact, he sometimes tellingly describes the passions as impositions or constraints (e.g. NE 175). Freedom, then, is determination by the agent’s rational faculties, independently of nonrational influences; it is rational spontaneity. This close connection between freedom and rational spontaneity is confirmed in an untitled text probably written in the early 1690s, where Leibniz ties freedom to a particularly demanding type of self-determination that intelligent creatures share with God: [T]he more [substances] approach the divine perfection the less need they have to be determined from the outside. For God, being the most free and most perfect substance, is also the most determined by himself to do the most perfect. But the more one is ignorant and impotent, the more one is indifferent . . . Now insofar as we have wisdom and act in accordance with reason, to that extent we will be determined by the perfections of our own nature, and consequently we will be all the more free as we will have fewer hindrances to our choice. (G 7:110f./SLT 95; similarly in G 7:109/SLT 94)

Here, Leibniz equates the most demanding type of self-determination – which he connects to freedom – with the determination by reason, and with the absence of factors that could interfere with or hinder rational choice. It is therefore precisely what I call ‘rational spontaneity.’ I will say much more about the connections between free agency and spontaneity in Chapter 5.

47

agent (spontaneum deliberatum, seu cum intellectu agentis; marginal note on Temmik’s Philosphia vera, Mugnai 1992: 163). All of these texts are excellent evidence that freedom is a type of spontaneity for Leibniz, and not only in the early years. See T 404: “one would be unhappy if one were sufficiently master in one’s domain to be able to will without grounds [sujet], without rhyme or reason.” This passage is interesting because, as we will see in Chapter 6, Leibniz often uses the terms ‘master’ and ‘mastery’ for our control over nonrational inclinations, which corresponds to rational spontaneity.

Rational Spontaneity

55

There is some additional indirect textual evidence that, for Leibniz, free actions possess a special type of spontaneity. See, for instance, Theodicy §64: “All that happens to the soul depends on it, but depends not always on its will.” The most straightforward way to understand this short remark is the following: while souls are self-determining in all of their actions – that is, while they always possess what I call ‘metaphysical spontaneity’ – they do not always determine themselves through their wills; that is, they do not always possess what I call ‘rational spontaneity.’ Similarly, Leibniz says later in the Theodicy, “we share a spontaneity in the exact sense [une spontaneité exacte]48 with all simple substances, and . . . in the intelligent or free substance it becomes a mastery over its actions” (T 291). Saying that the spontaneity common to all monads becomes mastery in intelligent or free monads can be read as claiming that intelligent substances possess a more demanding form of spontaneity that consists in mastery.49 An even more telling passage occurs in a text written around 1692. There Leibniz says that “what is called spontaneity in beasts and other substances lacking in intelligence, is elevated in man to a higher degree of perfection, and is called freedom” (G 7:108/SLT 94). As in “Conversation with Steno,” then, Leibniz here appears to be equating freedom with a particularly elevated or demanding type of spontaneity.50 If it is still unclear why what I have described should be viewed as a type of spontaneity, the following way of looking at it might be helpful. When a rational creature’s action is determined by factors outside of its will and intellect – for instance, by instincts or passions – this action is determined by factors of which the agent tends to be unaware and which it does not fully understand. Moreover, because reason and freedom are the perfections of human nature, and insofar as we, qua human beings, ought to act freely and in accordance with reason, the rational aspects of our mind are the aspects with which we ought to identify most strongly. Any nonrational, nonvoluntary factors are thus undue influences. While these factors are strictly speaking internal to our souls, they can be entirely unwelcome, and we in fact often ought 48

49 50

The term ‘exact’ is referring back to the preceding sentence, which is the last sentence of T 290. In that preceding paragraph, Leibniz distinguishes between the type of spontaneity we must invoke when we need to be exact (“quand il s’agit de s’expliquer exactement”) – that is, metaphysical spontaneity – and the ordinary way of speaking, according to which the origin of some of our actions is outside of us. See Kaphagawani, who also claims that mastery is a particularly demanding form of spontaneity (1999: 55). He reaffirms this later in the same text (G 7:111/SLT 95). Note that Leibniz’s description of freedom as the spontaneity of brute animals “elevated . . . to a higher degree of perfection” does not imply that the spontaneity of animals is merely a very low degree of freedom. Leibniz in fact makes it clear elsewhere that brute animals lack freedom altogether (e.g. in “Commentary on Burnet” §11(e), DPG 63; similarly in “Dialogue on Human Freedom,” Gr 362/AG 112).

56

Spontaneity

to fight them – as if they were foreign elements – to keep them from determining our actions.51 Leibniz’s views, on my interpretation, are similar to what contemporary philosophers of action sometimes call an ‘endorsement theory of autonomy,’ or more specifically, the ‘Platonic Model’ (e.g. Mitchell-Yellin 2014, 2015, Watson 2004: 17ff. and 37f.). According to that type of theory, the autonomy or self-governance of an agent can be undermined not only by other agents or external factors but also by appetites or impulses that are internal to the agent’s mind. To simplify a bit, autonomous actions are the ones that are the outcomes of desires that the agent endorses or values, or of desires whose objects the agent judges to be good (e.g. Watson 2004: 17, 25, 62). Not all of our desires or motivations are like that, however. An agent may find desires within herself whose objects she does not judge to be good or of which she even disapproves. Such desires are disjoint from the agent’s values; they originate in the nonrational part of her mind rather than from her rational judgments about the good (e.g. Watson 2004: 17, 37). When an agent acts on this kind of desire, she is not autonomous or self-governed and may in extreme cases even be said to act under compulsion. That is because the Platonic Model identifies the agent with her values or the rational part of her soul. Therefore, when a nonrational desire leads to an action, there is a sense in which the agent is determined by something external to her real self. Like present-day proponents of the Platonic Model, Leibniz sometimes talks of nonrational influences as quasi-external factors that we need to tame or overcome. For instance, in a passage from the Theodicy, he acknowledges a sense in which our own inclinations are external to us insofar as they are unwelcome and insofar as they are not rational: he says that “the force of others and our own passions enslave us” and that this kind of enslavement or servitude “comes from outside, it leads to that which displeases, and especially to that which displeases reason” (T 228; see also “Conversation about Freedom and Fate,” Gr 481/SLT 98; NE 196). Later in the Theodicy, Leibniz not only associates mastery with the power of reason to subdue the passions but even employs Plato’s chariot analogy to describe the way in which reason ought to control the passions (T 326). On this way of looking at it, nonrational inclinations are undue influences, external to us qua rational or autonomous agents, that can prevent us from acting as we should and as we reflectively want.52 Hence, only 51

52

Even though we cannot get rid of our passions altogether, that would be best – it would make us godlike. Because we cannot achieve this, keeping our irrational passions at bay sometimes requires using other less dangerous passions against them. That does not mean, however, that some passions are parts of our true (or rational) selves. They are more like external tools that reason can employ. I will discuss this in much more detail in Chapter 6. In a paper on self-determination in early modern philosophy, Vere Chappell notes that an action caused by “something that is within the agent in some way but is distinguishable from

Rational Spontaneity

57

voluntary actions are self-determined and spontaneous in the strictest sense, since to the extent that they are voluntary, they are free from determination by nonrational inclinations, or generally by factors external to the agent’s intellect and will and thus to her true self.53 One final thing to note is that even though both rational and agent spontaneity come in degrees, it is not the case that they differ from one another merely in degree and not in kind. The reason they both come in degrees is that monadic perceptions are infinitely complex and determined by an infinity of inclinations, as seen in Chapter 1. Hence, it can happen that external ideal influences play a role in somebody’s action without taking away that person’s agent spontaneity entirely. An agent’s perceptions can be very distinct and she can be more perfect with respect to a change than all other directly affected substances while still being ideally influenced in one small respect. For instance, when somebody hits a golf ball, she is ideally active to a high degree despite the fact that the golf ball is also acting on her, insofar as it resists. Likewise, nonrational factors can play roles in an action without taking away the agent’s rational spontaneity entirely. In the voluntary actions of human beings, in fact, there are always some nonrational influences because we never understand every aspect of the action perfectly and thus do not consciously will every aspect of a subsequent state (see e.g. a letter to Hartsoeker, October 30, 1710, G 3:509f.). As a result, freedom itself comes in degrees for Leibniz, as does rational spontaneity.54 This does not mean, however, that rational spontaneity is simply a high degree of agent spontaneity. Rational spontaneity, after all, requires the involvement of intellect and will, and the involvement of these faculties – rational deliberation concerning the best course of action and a subsequent volition – either does or does not occur. Consequently, there is a difference in kind between rational spontaneity and mere agent spontaneity. A monad can possess perfect agent spontaneity without being rationally spontaneous to any degree; a perceptual transition of a brute animal, for instance, might conceivably be so perfect that it is prior to all the corresponding perceptions of other substances in God’s reasoning, yet it would not thereby possess any rational spontaneity.

53

54

himself – that is, from his real or true self” is often viewed as an example of non-selfdetermination (2005: 127). There is an intriguing parallel here to Thomism: for Aquinas, the only truly human actions are rational actions, that is, actions that are actualizations of the specific potentiality of human nature (see e.g. ST IaIIae q1 a1 corp. and ad 3). Leibniz explicitly acknowledges degrees of freedom in several texts. For instance, in a letter to Bayle, he says that “the more perfect one is, the more one is determined to the good, and also freer [plus libre]” (G 3:59). See also “Dialogue on Human Freedom,” Gr 362/AG 112; NE 175; 181; G 4:362/L 388f.

58

Spontaneity

Table 2.1 Three Types of Spontaneity Type of Spontaneity

Type of Monadic State

Examples

Metaphysical Spontaneity: The agent is exempt from external influences in the metaphysical sense Agent Spontaneity: The agent is exempt from external influences both in the metaphysical and in the ideal sense Rational Spontaneity: The agent is also exempt from nonrational influences or from influences external to her real self

All monadic states, that is, actions in the loose sense, which include passive states

Your perception of pain upon being bitten by a mosquito

Active monadic states, that is, actions in the strict sense, which exclude passive states

The mosquito’s action of biting you

Voluntary monadic states, that is, actions chosen on the basis of the intellect’s judgment

Your action of designing a mosquito trap

5

Conclusion

For Leibniz, all finite monads possess perfect spontaneity, or are immune from the causal influence of other finite monads. Leibniz’s reasons for accepting this eccentric thesis appear to include his commitment to the independence of substances and his doctrine that creaturely interaction is unintelligible. Yet, there are good reasons to distinguish between three types of spontaneity. All monads possess the broadest type of spontaneity – metaphysical spontaneity – at all times and to the highest degree. Agent spontaneity, in contrast, comes in degrees and is not present in all monadic change. Finally, rational spontaneity, or the narrowest type of spontaneity, is reserved for rational and free actions; the freer or more rational an action is, the higher its degree of rational spontaneity. These distinctions are summarized in Table 2.1.

3

Teleology

Armed with an understanding of what monads are, how they act, and in what senses their actions are spontaneous, we can now turn to teleology and its role in monadic activity. There are good reasons for thinking that teleology plays a role in all monadic actions. I examine first some indirect evidence that Leibniz views teleology as ubiquitous and then turn to more direct textual evidence. The indirect evidence has to do with the ways in which Leibniz partially aligns himself with – and shares some of the central philosophical motivations of – the Aristotelian tradition, for which teleology is key to understanding all natural change. This serves to bolster the direct textual evidence, which is independently quite robust. Examining both types of evidence, furthermore, gives us a preliminary understanding of what teleology is for Leibniz. I also argue that there is a tight connection between spontaneity and teleology and that it is useful to distinguish three types of teleology, parallel to the three types of spontaneity described in Chapter 2. This, in turn, allows Leibniz to acknowledge important differences between the ways in which different kinds of actions are end-directed. In fact, I argue that the least perfect monadic actions are end-directed without being goodness-directed, while the most perfect actions are always goodness-directed. Two caveats are in order before we proceed. First, I will be using ‘teleology,’ ‘final causation,’ and ‘end-directedness’ – as well as their cognates – interchangeably. Second, I am concerned mainly with what is sometimes called ‘immanent teleology.’1 An action is teleological in this sense if and only if it is end-directed due to intrinsic features of the agent – that is, if and only if this end-directedness obtains simply in virtue of the agent’s nature or will. For the most part, I will bracket extrinsic teleology, such as the kind of end-directedness that creaturely actions have in virtue of serving God’s ends. Of course, there is teleology in this other sense. Yet, I am primarily interested in something more fundamental: the internal end-directedness of creatures, that is, the end-directedness that creaturely actions have independently of the 1

For discussions of the differences between immanent and non-immanent teleology, see Osler (2001: 152f.) and Carlin (2012: 54).

59

60

Teleology

ways in which God uses them to achieve his ends. The purposes that these actions serve due to the role they play in God’s providential design, on my view, are a higher-level ordering of ends. God imposes this ordering on things that are already naturally or immanently end-directed. 1

The Ubiquity of Teleology

Before considering the textual evidence that there is final causation in all monadic actions, let us take a quick look at the historical background. This is helpful for two reasons. First, it shows that Leibniz’s terminology is teleologically loaded. By calling monads ‘entelechies’ or ‘substantial forms’ and explaining their activities in terms of appetitions, Leibniz appears to align himself partially with a tradition that views final causation as central to understanding natural change. While it is implausible that Leibniz agrees with this tradition in all respects, it would also be highly misleading for him to use terms like ‘entelechy,’ ‘substantial form,’ and ‘appetition’ if he wanted to eschew all of their traditional teleological connotations. Second, there are arguments – formulated, for instance, by Thomas Aquinas – that all genuine activity must be teleological, and it is plausible that Leibniz shared the motivation behind these arguments. Examining this background has a further payoff that goes beyond determining whether teleology is ubiquitous for Leibniz: it helps us understand some of the connections between teleology and spontaneity. We already noted in Chapter 1 that Leibniz explains all nonmiraculous monadic changes in terms of appetitions – that is, strivings for, or tendencies toward, new perceptions. Such tendencies or strivings toward new states, in turn, are quite naturally understood as teleological. And indeed, many of Leibniz’s predecessors tie the term ‘appetite’ intimately to final causation. Aquinas, for instance, makes that connection explicit in the following passage: “The influence of an efficient cause is to act; the influence of a final cause is to be sought [appeti] and desired” (DV q22 a2 corp.). Elsewhere, Aquinas describes the end as “that in which the appetite of an agent or mover, and of that which is moved, finds rest” (SCG 3.3.3), again associating appetitions with final causation.2 In fact, Aquinas views substantial forms or creaturely natures as the sources of some of these end-directed appetites and hence as the ultimate sources of many instances of teleology. That is helpful because Leibniz, as seen in Chapter 1, frequently describes substances as being or possessing substantial forms. Aquinas claims that natural appetitions – that is, strivings whose source is neither the creature’s reason nor its sensation – are strivings for ends that are determined 2

Robert Pasnau describes Aquinas’s view as follows: “The purpose of appetite is to achieve a certain end . . . Final causality . . . plays its role through the mechanism of appetite” (2002: 209).

The Ubiquity of Teleology

61

directly by the God-given natures or substantial forms of things: “in those [things] that lack cognition one finds only a form determining each to its one proper being, the being that is natural to each thing. Therefore a natural inclination, which is called ‘natural appetite,’ follows from this natural form” (ST I q80 a1 corp.). Thus, natural appetite is an inclination toward ends that are specified by the thing’s nature. As Aquinas puts it elsewhere, “natural appetite is nothing but the ordination of things to their end in accordance with their proper natures” (In Physic., lib. 1 l. 15 n. 10, in CT; similarly in ST I q78 a1 ad 3).3 It is true quite generally for Aquinas that “upon the form follows an inclination to an end, or to an action, or to something of that kind; for everything, insofar as it is in act, acts and tends towards that which is fitting for it, in accordance with its form” (ST I q5 a5 corp.). One central role of the substantial form, after all, consists in specifying what kinds of activities are natural for the substance, that is, what kinds of actions it must perform to be perfect of its kind. The substantial form of the human being, for example, specifies rational activity as that which makes a human being perfect of its kind. Even though there is a sense in which such end-directedness ultimately comes from God, who created these substantial forms and directed them to certain ends, Aquinas often stresses that it is nonetheless natural. God directs creatures to these ends by building the corresponding appetitions into their natures rather than by imposing actions on them that are not in accordance with their natures.4 As a result, substantial forms have built-in teleology, according to Aquinas. These considerations not only show how intricately final causation is traditionally tied to the notions ‘appetition’ and ‘substantial form,’ but they also point to important similarities between Aquinas’s and Leibniz’s views: both authors believe that created things are endowed with natures and that in virtue of having these natures – or in virtue of having a substantial form – they possess appetitions for actions that are in accordance with their natures. All of this suggests that Leibniz intended the terms ‘appetition’ and ‘substantial form’ to be understood teleologically and that there is end-directedness in all monadic activity. Even the least perfect substances possess substantial forms that direct them to ends. Let us now turn to another piece of indirect evidence that Leibniz views all monadic actions as teleological: he has excellent reasons for agreeing with Thomists that final causation is required for all genuine activity. For Aquinas, 3

4

Aquinas defines ‘appetite’ as follows: “To have an appetite [appetere] is nothing else but to strive for something [petere], stretching [tendere], as it were, toward something that is ordained for oneself” (DV q22 a1 corp.; see also ST IaIIae q8 a1 corp.). For the connection between creaturely natures and immanent teleology in Leibniz, see also Madden (2003: 183). I will soon say more about the distinction between natural and nonnatural (or violent) enddirectedness. See footnote 61 for a brief discussion of whether the actions of nonrational creatures are instances of genuine immanent teleology for Aquinas.

62

Teleology

the end or final cause is “the cause of the other causes” (e.g. De principiis naturae 4.22, in CT). Without end-directedness, he appears to be saying, an efficient cause would not be able to effect anything. There is some controversy over why precisely Aquinas and other Thomists endorsed this doctrine. Without entering too far into this exegetical question, however, we can distinguish two plausible interpretations of the motivation behind the doctrine, both of which fit well with Leibniz’s commitments. The first one is roughly Paul Hoffman’s interpretation of Aquinas’s version of the doctrine, while the second one is roughly John Carriero’s. The first option is that philosophers who view final causation as a prerequisite for all genuine efficient causality are working with a very thin and undemanding notion of final causation. Suppose that whenever a thing has a specific tendency or disposition, there is final causation.5 On this construal, the fact that fire has the tendency to ignite pieces of dry paper that come in contact with it is sufficient for describing fire as end-directed. Fire tends toward paper-ignition – or, more generally, fire tends toward the production of more fire. If that kind of disposition is sufficient for teleology, there are good reasons to think that all efficient causation is teleological. After all, on an understanding of efficient causation that was widespread in the early modern period, a cause must render its effect intelligible. There must be something about the cause that explains why this effect rather than some other effect takes place. If we fully understood the cause, we would – in Michael Della Rocca’s words – be able to “see the effect coming” (2013: 53). That means that on this understanding of causation every efficient cause in some sense tends toward its effect. Every efficient cause has a tendency – in a very undemanding sense of ‘tendency’ – to produce a particular effect.6 As a result, all efficient causation would presuppose enddirectedness in the thin sense that we are currently considering: in order for x to efficiently cause y, x must make y intelligible and hence in a broad sense be determined toward, or tend toward, y.7 Since Leibniz is among those who hold that causes render their effects intelligible and that every cause is determined to a specific effect, he would clearly grant that all actions are end-directed in the thin sense just sketched. He does, after all, think that a sufficiently intelligent being is able to predict all of a substance’s future states from its present state, or from its nature. But may there be a more demanding sense in which all actions are teleological on Leibniz’s 5 6

7

This proposal is based on Hoffman (2009). This fits well with what Aquinas says: “the first of all causes is the final cause . . . For if the agent were not determined to some particular effect, it would not do this thing rather than that: consequently in order that it produce a determinate effect, it is necessary for it to be determined to some certain one, which has the nature of an end” (ST IaIIae q1 a2 corp.). This, as Hoffman points out, can be backed up with a version of the Principle of Sufficient Reason: if the cause supplies the sufficient reason for the effect, the cause must be determined to this specific effect; it cannot be indifferent to several possible effects (Hoffman 2009: 309).

The Ubiquity of Teleology

63

view?8 This brings us to the second reason why many Thomists might believe that all efficient causality presupposes end-directedness. The basic idea here is that these philosophers are working with a very demanding notion of efficient causation rather than with a very thin notion of final causation.9 On this more demanding understanding of efficient causation, a billiard ball that hits another billiard ball is not a genuine efficient cause of the second ball’s motion. That is because this first ball is not the originator of its own motion toward and collision with the second ball; nothing in the first ball’s nature explains that motion. Instead of following from its nature, the motion is imposed on the ball by the billiard player. More generally, on this more demanding understanding of efficient causation, an efficient cause must be the originator of its activity in the way that, for Thomists, substances are the originators of their natural activities. John Carriero spells out this demanding notion of efficient causality and its relation to final causation in a very helpful way. It is crucial for Thomists to distinguish between natural and violent motions, or between natural and violent changes. Violent motions are motions imposed on a thing from the outside, going against or beyond the thing’s nature (Carriero 2005: 120). The billiard ball is a good example of a violent motion. Something external, namely the cue stick, explains the ball’s motion; the motion does not follow from the billiard ball’s nature. Aquinas often uses the example of an arrow to illustrate violent motion (e.g. ST I q103 a1 ad3; DV q22 a1): when an archer shoots an arrow at a target, she directs the arrow toward that target. However, flying toward the target is not a motion that is natural for the arrow. Instead, the archer imposes this motion on the arrow. When a substance moves or changes naturally, by contrast, its motion is not externally imposed.10 Natural changes are consequences of the substance’s nature. For instance, consider a flower turning toward the sun: nothing outside of the flower is imposing this motion on it. The distinction between violent and natural changes gives us a good way to spell out the demanding understanding of efficient causality. A thing is an efficient cause in this sense only when its motion or change is natural. The 8

9 10

See Carlin, who discusses three different ways in which final causes have primacy for Leibniz (2006: 219). Yet, Carlin understands Leibnizian final causation very differently – ultimately as a species of efficient causation (Carlin 2006: 228) – which affects his interpretation of the primacy of final causes. Hoffman in fact characterizes the difference between his and Carriero’s interpretation in this way (2009: 198). Note that I am using ‘natural’ in a broad sense here, namely in the sense of ‘nonviolent.’ For Aquinas, not all nonviolent motions are strictly speaking natural motions because not all internal ends are what he calls natural ends. Aquinas believes that rational agents can direct themselves towards ends freely, instead of having all their ends determined by their natural inclinations; he also holds that creatures with sensation can direct themselves toward something perceived by their senses, albeit not freely. These are clearly not instances of violent motion or violent enddirectedness, however.

64

Teleology

arrow is not the originator of its motion toward the target, whereas the flower is the originator of its motion toward the sun. The connection between teleology on the one hand and efficient causality understood as natural motion on the other hand is straightforward. Natural motions are natural because they are explained by natural strivings. In order for the flower to turn toward the sun naturally – or in order for the flower to be the originator of its motion – it must have a natural tendency or striving to move in this way. The billiard ball, by contrast, does not have a natural tendency to roll across the table toward the other ball; it is the billiard player rather than the ball’s nature that directs the ball toward that particular end. As a result, we can say that the ball is not the originator of its motion because the end-directedness is imposed on the ball by something external. Natural motions, then, are the results of natural tendencies or strivings. These tendencies or strivings, in turn, are teleological in a more robust way than the thin end-directedness that we considered earlier. Take a purely mechanistic world, for instance, in which all changes are instances of bodies getting pushed or pulled in a deterministic way by other bodies. In that kind of world, there would be end-directedness in the thin sense: each cause – that is, each moving portion of matter – is determined toward a particular effect in the sense that we can see the effect coming. Yet, because there are no natural ends in such a world, there is no teleology in the more demanding sense. All motion is violent motion because all motion is imposed by something external. In other words, no event or body is a genuine originator of end-directed motion in such a world; one thing only pushes or pulls other things insofar as it is itself pushed or pulled.11 One possible motivation for claiming that all genuine efficient causation requires end-directedness, then, is the doctrine that being the originator of a change requires possessing an internal, natural tendency toward that change. When something external makes you change in a certain way, you are not the agent of that change. This motivation for holding that all genuine activity is end-directed would clearly appeal to Leibniz. After all, as seen in Chapter 2, Leibniz insists that all monadic actions are spontaneous, that is, they originate in the agent’s own depths and are made intelligible by the agent’s nature. Something within the agent, as we have seen, must prompt her to act in the way that she acts. Genuine activity cannot be imposed from the outside, going against or beyond the agent’s nature. Leibniz seems to endorse the doctrine that ownership of an action requires being a genuine originator of the action. This is an excellent reason to view all activity as end-directed: the agent’s nature directs her toward all of her actions and prescribes what actions she is supposed 11

Carriero imagines a similar world and argues that it would not contain genuine agency (2008: 122f.).

The Ubiquity of Teleology

65

to perform.12 These considerations indicate that all monadic actions are teleological not only in the thin sense we considered earlier but also in this more robust sense. Here, incidentally, we can already see that spontaneity and teleology are closely connected: being the originator of one’s actions plausibly requires being internally or naturally directed toward those actions. I will return to this connection in Section 2. There is one difference between the Leibnizian and the Thomistic account that is worth stressing: for Leibniz, substances are never subject to violent motion in the normal course of nature.13 Created substances do not interact, after all, and thus do not impose changes on one another. Instead, each always acts with perfect metaphysical spontaneity. God is the only substance with the power to make other substances act in ways that are not natural to them, and he never – or, at least, almost never – exercises this power.14 As a result, Leibnizian substances exhibit the kind of immanent teleology that is necessary for genuine agency not only when they are agent-spontaneous but at all times – even when they are merely metaphysically spontaneous. This will be crucial for my distinction between three types of teleology in Section 3. Based on this brief overview of the historical background, it should already be at least somewhat plausible that there is final causation in all monadic actions. If that is indeed the case, we already have a rough, preliminary understanding of what Leibnizian final causation consists in: it is a causality present in the spontaneous actions of substances, that is, in changes that arise in a substance in virtue of its nature. More precisely, monads exhibit immanent final causation insofar as they strive for each new state naturally, or transition to each new state because of their natural appetitions. The ends they strive for are specified by their natures. Section 4, in which I examine the role of goodnessdirectedness in monadic teleology, will yield a deeper understanding of Leibnizian final causation. Let us now turn to the more direct textual evidence for thinking that Leibniz views all monadic actions as instances of immanent final causation. There are at least three categories of textual support. First, as already mentioned in Chapter 1, Leibniz often claims that the realm of monads is the kingdom of final causes, 12 13 14

See Schmid (2011: 383f.), who explains the essential end-directedness of monads in a similar way. At least strictly or metaphysically speaking. We will see in Section 3.2 that Leibniz can capture the notion of violence in the same way in which he captures the notion of interaction. In some texts, Leibniz suggests that genuine miracles never occur once the world is created; what does happen, however, are events that human beings are unable to predict but that nevertheless arise naturally (see e.g. DM 16). In other texts, he at least considers it possible that in addition to the original creation of substances, it might take a miracle to elevate monads that are destined to become human beings to the level of reason at conception (CD 81; letter to Des Bosses, April 30, 1709, LDB 127; but see T 397, where he says he prefers to think this elevation happens without a miracle). Moreover, he says that the Incarnation involved a genuine miracle (T 249). For a detailed discussion of the elevation to reason, see Jorgensen (2013).

66

Teleology

while the realm of bodies is the kingdom of efficient causes. Second, there are passages in which Leibniz stresses the similarity between the agency of higher monads and that of lower monads, based on the Principle of Continuity: if nature is uniform and there is final causation in human actions, there ought to be something analogous in all other monadic actions. And third, Leibniz often attributes appetitions to all monads and associates appetitions closely with final causation. Yet, before discussing the textual evidence for the ubiquity of teleology further, I will take a look at two passages that may seem to point in the opposite direction.15 The perhaps most striking such passage is from Leibniz’s remarks on Lamy: [T]here is a law of order among perceptions just as there is with movements . . . [I]t is sufficient to say that the perceptions which express the laws of motion are linked together just like those laws, which they express according to the order of efficient causes. But the order of voluntary perceptions, which is that of final causes, is in conformity with the nature of the will. (G 4:580/WF 154f.)

At least on the face of it, this passage suggests that only changes in voluntary perceptions are teleological, whereas nonvoluntary perceptions – which include all the perceptions of nonrational monads and most perceptions of rational creatures, as we saw in Chapter 1 – are governed only by laws of efficient causation. Two years later, in another response to Lamy, Leibniz appears to restate this idea: “Confused perceptions are ordered just like the laws of the motions which they represent. The motions of bodies are explained by efficient causes, but in the distinct perceptions of the soul, where there is liberty, final causes reappear” (G 4:592/WF 166f.). Only free actions, these passages appear to suggest, are instances of final causation.16 The Lamy passages are, however, clear outliers.17 Numerous other texts strongly suggest that there is final causation in all monadic actions, even the least perfect ones. Let us start with the first category of textual evidence: passages that associate the monadic realm in general with final causation. In “Principles of Nature and Grace,” for instance, Leibniz states explicitly that all monadic changes are governed by laws of final causation while all bodily motion is governed by laws of efficient causation: [T]he perceptions in the monad arise from one another by the laws of appetites, or by the laws of the final causes of good and evil, which consist in notable perceptions, ordered or disordered. Similarly, changes in bodies and external phenomena arise from one another by the laws of efficient causes, that is, the laws governing motions. Thus there is perfect harmony between the perceptions of the monad and the motions 15 16 17

See McRae (1976: 67), who also points to this apparent inconsistency in Leibniz. On that reading, Leibniz would agree with later Scholastics that immanent final causation requires intelligence, so there is no immanent final causation in nonrational creatures. I will later suggest a way to reconcile the Lamy passages with the other texts.

The Ubiquity of Teleology

67

of bodies, pre-established from the first between the system of efficient causes and that of final causes. (PNG 3)

This passage refers to the perceptions of monads generally as a system of final causes – that is, it explicitly says that all perceptual transitions are teleological. Similar passages abound. Admittedly, the terminology that Leibniz employs in many of these texts does not completely settle the question whether the realm of final causes includes bare monads or whether it is limited to minds and souls.18 Yet, as we saw in Chapter 1, many changes in minds and animal souls are exactly like the perceptual transitions in bare monads. Hence, if all perceptual transitions in minds and souls are teleological, then so too are the transitions in bare monads. It would be odd for Leibniz to claim that although we are very much like bare monads when we are unconscious – and go through countless unconscious changes even when we are awake – all of our perceptual transitions are teleological while those of bare monads are not. Let us now turn to the second category of textual evidence. In a letter to Queen Sophie Charlotte, Leibniz employs the Principle of Continuity to argue for the ubiquity of final causation in monadic action. He writes that since we are familiar with souls following “moral laws of good and evil” in deliberation, we should explain all other actions of souls analogously and assume that “it’s all the same as here” (May 8, 1704, G 3:346/WF 224).19 From this we can conclude, he says, that “everything in the soul happens morally, or in accordance with perceived good or evil . . . [E]ven in our instinctive or involuntary actions, where it seems only the body plays a part, there is in the soul a desire for good or an aversion to evil which directs it” (G 3:347/WF 224). This passage is strong evidence for thinking that according to Leibniz a desire for the good directs even involuntary actions; I will say more about that in Section 4. What is relevant for our purposes here, however, is that the text explicitly argues, based on the Principle of Continuity, that “moral laws” – which in this context is presumably a synonym for ‘laws of final causation’ (see e.g. CD 46; COE 20) – direct instinctive or involuntary actions. In fact, this is true for “everything [that happens] in the soul.” Even though the passage is strictly speaking only about “our” – that is, human – involuntary actions, it would again be exceedingly implausible to claim that the same does not apply to the instinctive actions of animals20 and even to bare monads. The Principle of Continuity, after all, applies across the board. 18

19

20

This is partly because Leibniz sometimes uses the term ‘soul’ only for monads with sensation, and sometimes for all monads. This ambiguity affects passages like M 79; T 62; “Against Barbaric Physics,” G 7:344/AG 319. Leibniz uses “it’s all the same as here,” or something analogous, in a few other places as a slogan for his Principle of Continuity; see e.g. letter to Masham, May 1704, G 3:340/WF 205f.; “Considerations on Vital Principles,” G 6:546/L 590. Leibniz sometimes stresses that some human actions are just like animal actions; e.g. T 250.

68

Teleology

Finally, let us turn to the third and strongest category of direct textual evidence: the fact that Leibniz thinks that all monadic changes are explained by appetitions, which he understands in teleological terms. There are not only historical reasons for taking appetitions to be teleological, which we already considered earlier, but also some textual evidence. For instance, in an untitled text probably composed in 1710, Leibniz defines ‘appetite’ as “the endeavour of acting tending towards new perception” (G 7:330/SLT 66), which has clear teleological overtones. Appetitions, Leibniz tells us, are tendencies or strivings toward something, which means that they are end-directed states. Similarly, he says in the “Monadology” that souls “act according to the laws of final causes, through appetitions, ends, and means” (M 79), which suggests that acting through appetitions, ends, and means is an instance of final causation. In fact, Leibniz’s association of appetitions with ends and means is another piece of evidence, since ‘ends and means’ carries teleological connotations that are at least as strong as those of ‘appetition.’ Leibniz says, for instance, that “since the nature of a simple substance consists of perception and appetite, it is clear that there is in each soul a series of appetites and perceptions, through which it is led from the end to the means, from the perception of one object to the perception of another” (“Metaphysical Consequences” §8, C 14/MP 175).21 Similarly, in a table of definitions from the mature period, Leibniz defines ‘end’ as “that, an appetition for which is a sufficient cause of an endeavor [conatus] in the agent” (C 472). This terminology, again, is clearly teleologically loaded.22 It seems safe to conclude, based on the direct and indirect evidence, that all monadic changes are teleological. 2

Connections between Spontaneity and Teleology

Before turning to the distinction between different types of teleology, let us take a general look at the relation between spontaneity and teleology.23 Based on what we have seen thus far, the most fundamental kind of teleology that Leibniz ascribes to all monadic actions is roughly the following. Monads are immanently end-directed in virtue of possessing natures that specify all states that 21 22

23

Aquinas also associates appetition with ends and means (e.g. DV q22 a5 corp.). Interestingly, he defines ‘means,’ in the same text, as “a cause, which the efficient cause, . . . aiming at the end, causes to be a cause” (C 472). This definition strongly suggests that final and efficient causation always, or at least typically, go together in monadic agency. A few other interpreters mention a connection between spontaneity and teleology. See, for instance, Rutherford, who distinguishes two types of spontaneity and teleology, and closely associates what he calls ‘agent spontaneity’ with what he calls ‘desire teleology’ (2005: 169). See also Murray, who also argues for a type of Leibnizian spontaneity that can distinguish coerced from uncoerced actions; he ties this more demanding type of spontaneity to moral necessity, which is “intrinsically teleological” (2005: 207). Yet, neither of these two interpreters fleshes out the connection between spontaneity and teleology in detail.

Three Types of Teleology

69

this monad will ever have. These natures, which consist in primitive force, give rise to strivings for particular actions. Leibniz views all monadic actions as teleological, on my view, because of these strivings that arise naturally in the monad, dictated by its nature. What does this kind of teleology have to do with spontaneity? First of all, note that one can view both spontaneity and immanent teleology as kinds or aspects of independence: a substance is spontaneous when it is itself the source of its actions or changes, that is, when it is not determined from the outside. Likewise, a substance exhibits immanent teleology when it acts in accordance with its own ends, that is, when external things are not forcing it to act in ways that are not demanded by its natural strivings. Violent changes, or changes imposed on a substance from the outside that are not in accordance with its own ends or strivings, would not count as immanent teleology. Likewise, such changes would not be spontaneous. Both spontaneity and immanent teleology are therefore aspects of monadic immunity from external influence – that is, from the violent imposition of a change contrary to the nature of the monad. Thus, there is at least one fundamental similarity between the two notions. Yet, there is an even closer connection between spontaneity and teleology. As I argued in Section 1, there are good reasons for believing that immanent teleology is a necessary condition for spontaneous activity on Leibniz’s view. In order to act spontaneously – that is, in order to be a genuine originator of an action – a substance must be immanently end-directed. On that interpretation, the connection between immanent teleology and spontaneity is very close for Leibniz: the former is a necessary condition for the latter. Likewise, as seen, the substantial form of the agent is simultaneously the source of the agent’s end-directedness and spontaneity. It not only prescribes all changes and directs the agent toward them but also spontaneously gives rise to these changes.24 Thus, the fact that monads have the principle of all of their actions within them means both that they act spontaneously and that they are internally end-directed. 3

Three Types of Teleology

In Chapter 2, I distinguished three types of spontaneity: (1) metaphysical spontaneity, which belongs to all monadic actions; (2) agent spontaneity, which belongs to actions insofar as they are immune from the ideal influences of external things; and (3) rational spontaneity, which belongs to actions to the extent that they are immune from nonrational influences. Given the tight 24

The latter may sound un-Aristotelian, but it is in line with the way many scholastics understood substantial forms. See van Ruler (1995: 69), Schmaltz (2012: 128f.), and Pasnau (2011: 557–65).

Teleology

70

connection between spontaneity and teleology pointed out in Section 2, it is natural to wonder whether there is a parallel threefold distinction with respect to teleology. I will argue that there is. This is important because it reinforces the distinctions between the three corresponding types of monadic activity. For instance, as I will elaborate further in Chapter 5, these distinctions allow us to understand free agency as a particularly demanding type of spontaneity as well as a particularly demanding type of teleology. Likewise, it adds another dimension to Leibniz’s distinction between actions in the broad sense – that is, metaphysically spontaneous actions – and actions in the stricter sense – that is, agent-spontaneous actions. 3.1

Metaphysical Teleology

We saw earlier that all monadic actions are teleological because of the role that appetitions or strivings play in these actions, or – to put this slightly differently – because these actions are demanded by the agent’s nature. Hence, there is an undemanding and ubiquitous type of immanent teleology that corresponds to metaphysical spontaneity; we can call it ‘metaphysical teleology.’ Monads always possess this type of teleology because they naturally strive for all of their states. No other agent ever imposes actions on them violently, or against their natures, in the natural course of events. God is the only agent capable of interfering with a creature’s natural end-directedness, but he never makes use of this ability, except perhaps in extremely rare cases of miraculous interventions.25 In the natural course of events, created monads always perform the actions toward which they are directed by their natures. 3.2

Agent Teleology

Let us now turn to actions in the strict sense – that is, to agent-spontaneous actions. These actions are teleological in a more demanding sense than those that are merely metaphysically spontaneous. To illustrate, consider again the mosquito example from Chapter 2. We saw that when you feel pain upon being bitten by a mosquito, your state is spontaneous in the metaphysical sense but not in the agential sense. Moreover, the painful state must be dictated by your natural strivings: because you produce this state spontaneously, you must be 25

As already mentioned, Leibniz sometimes suggests that genuine miracles never occur once the world is created, but he acknowledges their occurrence in other texts (see footnote 14). It is important to note, however, that if miracles do occur, the creaturely states produced by God’s extraordinary intervention possess neither metaphysical teleology nor metaphysical spontaneity. After all, these miraculous states by definition do not arise from the creature’s own depths, and they are not demanded by the creature’s nature. The parallel between spontaneity and teleology thus holds even in these cases.

Three Types of Teleology

71

internally directed toward that painful sensation.26 Since your natural strivings are the bottom-level metaphysical explanation for your transition to the painful state, this transition is an instance of immanent metaphysical teleology. Yet, as seen in the previous chapter, we can say that in the ideal sense, the mosquito’s nature is the reason why you feel this pain: your painful state ideally depends on the mosquito. The bite constitutes a decrease of perfection for you, after all, but an increase in perfection for the mosquito – that is how Leibniz explains ideal interaction. This means that the mosquito’s action is teleological in a more demanding sense than yours: the mosquito not only acts in accordance with its natural strivings but also in accordance with a perfection of its nature. To describe this more demanding type of end-directedness, we can speak of agent teleology. Just as external things can undermine your agent spontaneity by exerting an ideal influence on you, they can undermine your agent teleology by ideally imposing a state on you that is an imperfection for you, such as pain. In doing so, they limit the extent to which your actions are based on perfection-directed strivings. In the mosquito example, you lack immanent teleology in the agential sense because the painful state, while natural, is an imperfection of your nature.27 The mosquito, on the other hand, possesses agent teleology, just as it possesses agent spontaneity, because it is actualizing a perfection of its nature. We could even say that your painful sensation occurs for the sake of the mosquito because, as already seen, Leibniz says that the perfection of the agent explains (or is God’s reason for allowing) the imperfection of the patient.28 There is some indirect textual support for distinguishing metaphysical from agent teleology. First of all, in a passage already quoted in Chapter 2 that distinguishes a broad sense from a narrower sense of ‘action,’ Leibniz defines the latter as “an endeavour towards perfection” (NE 210). This indicates that there is a teleological dimension to the distinction between actions and passions. Elsewhere, Leibniz draws a distinction between two senses of ‘natural’ that suggests that there are multiple types of teleology: “when I say that everything that happens to a substance can be seen to be in some sense natural to it, or to be a consequence of its individual nature, I mean the complete nature, which includes all that belongs to that individual” (remarks on Lamy, G 4:582/ WF 157). He fully acknowledges that there is another legitimate sense of ‘natural’ in which some things that happen to us are not natural but rather 26

27 28

See Leibniz’s note on Bayle, in which he says the following about a dog’s soul transitioning from pleasure to pain: “The principle of change is in the dog, the disposition of its soul moves imperceptibly towards giving it pain” (G 4:532/WF 78). In Section 4, I will address the worry that what is natural may always be a perfection in some sense. This kind of teleology, of course, is no longer immanent teleology. Moreover, the ideal influence is not the bottom-level explanation for the change.

72

Teleology

“accidental or violent” (ibid.).29 These accidental or violent changes are presumably changes ideally caused by external things. Since immanent teleology consists in acting on the basis of natural strivings, this acknowledgment indicates that Leibniz should be willing to recognize at least two types of teleology: one that encompasses all actions dictated by the “complete nature” of the substance and one that encompasses only non-accidental or nonviolent actions. These two types appear to line up perfectly with metaphysical and agent teleology. Moreover, Leibniz says that to the extent to which substances are ideally determined from the outside, they are “proportionally obliged to serve external things” (“On Freedom and Spontaneity,” G 7:110/SLT 94). This phrase has teleological overtones: serving external things is most naturally interpreted as acting for the sake of the perfections or ends of external things rather than for the sake of one’s own perfections or ends. Hence, this passage also supports my distinction. A final piece of evidence for a type of teleology corresponding to agent spontaneity is yet another passage from Leibniz’s remarks on Lamy. There, Leibniz explicitly connects the ideal interactions among created substances to final causation: [I]n the intentions of God and in the order of final causes, one substance depends on another; for God considered one when producing the other, even though so far as physical influence, or efficient causation, goes, they have as little dependence on each other as if each were alone in the world with God. (G 4:578/WF 153; emphasis added)

Leibniz here describes the ideal dependence of one finite substance on another as an instance of final causation.30 This makes perfect sense because, as already seen, when an imperfect state in one creature ideally depends on another creature, we can say that the imperfect state occurs not only because of but also for the sake of the creature that is more perfect in this respect. After all, God created a substance with this imperfection in order to be able to include the substance with the corresponding perfection in his harmonious design. Therefore, it is highly plausible that ideal interaction and consequently agent spontaneity are connected to final causation. In the agential sense, the patient’s transition is not an instance of immanent teleology, because instead of acting for the sake of its own perfection, the patient acts for the sake of another monad’s perfection. The transition of this other monad, by contrast, is an instance of immanent teleology in the agential sense. 29

30

Similarly, Leibniz writes earlier in the same text that there are “degrees of naturalness among modifications” and that “the most natural is that which is entirely in conformity with the perfection of the nature which produces it” (remarks on Lamy, G 4:582/WF 157). Accordingly, some actions express our nature better than others, namely insofar as they constitute a perfection of our nature rather than an imperfection. See draft of letter to Arnauld, G 2:69 and 2:71, where Leibniz also connects ideal dependence to final causation; see also Puryear (2010: 783).

Three Types of Teleology

73

In fact, actions that are teleological in this more demanding sense are presumably exactly the ones that are traditionally viewed as instances of immanent teleology. After all, on the traditional view, creatures lack immanent teleology when their actions do not aim at their own perfection. For Thomists, for instance, all actions that are prescribed by a creature’s nature are perfections of its nature, and consequently, all natural actions aim at this perfection.31 Hence, just as agent spontaneity is a Leibnizian way to capture one traditional understanding of spontaneity, agent teleology is a Leibnizian way to capture the traditional understanding of immanent teleology. It therefore also gives him a way to capture the notion of violent change. Even though metaphysically speaking there are no violent changes in Leibniz’s system (miracles aside), there is a sense in which there are violent changes – namely, when a change lacks agent teleology. As seen, Leibniz even says so explicitly in one passage (G 4:582/WF 157). We should also note that, as in the case of agent spontaneity, agent teleology can be applied either to entire organisms or to individual substances. An organism, after all, can act either for the sake of its own perfection or for the sake of another creature’s perfection; the mosquito example is a good illustration of this.32 Likewise, one monad that belongs to an organism can employ other monads within that organism to achieve its ends. In fact, this is a helpful way to understand the subordination of monads within an organism: the monads that make up my heart serve my central monad, or soul, as long as the heart is functioning properly, and generally, the monads in my body serve the perfection of my soul.33 Finally, it is interesting to note that there is a connection between the notion of agent teleology and Leibniz’s doctrine of striving possibles. This is the doctrine that possibles in some sense strive for existence in proportion to their perfection or degree of reality (“Ultimate Origination of Things,” G 7:303f./AG 150f.), and, because of their incompossibility, struggle or compete with one another for existence (T 201). Bertrand Russell, with characteristic wittiness, calls it “a tug of war . . . of ghosts all hoping to become real” 31 32

33

In Section 4.3 I explain why and how Leibniz departs from this traditional understanding of teleology. In other cases – for instance, my stepping on an insect unintentionally – it may seem less obvious that one organism acts for the sake of the perfection of another organism. Yet, insofar as Leibniz wants to generally account for the apparent interaction between creatures in terms of the agent’s perfection and the patient’s imperfection, even such cases fall under the same description. When I accidentally crush an ant with my foot, it is a perfection of mine that I am able to perform that motion without being significantly impeded by the ant’s body. My foot is pushing the ant’s body out of its way and emerges from the encounter unscathed. Hence, even here we can say that the ant’s transition to a less perfect state occurs for the sake of my perfection. A great way to describe the unity of organisms is Glenn Hartz’s term “teleological superharmony” (1996: 79). See Carriero, who also suggests that end-directedness can unify organisms (2008: 136f.).

74

Teleology

(1903: 186). One promising way of understanding the strivings of possibles is, as one interpreter puts it, in terms of “the degree of attractiveness that they have to a God who is disposed to create a world” (Blumenfeld 1973: 170; see also Anfray 2011: 56).34 Since, on what I take to be the standard interpretation, these possibles are ideas in the divine mind, this is plausibly the only way in which they can be said to strive for something.35 Other interpreters argue, however, that the striving of possibles ought to be understood quite literally, not merely metaphorically as the weight they carry in divine deliberations (e.g. Shields 1986, Griffin 2013: 52ff., Wilson 1989: 278f.). Yet, on either interpretation, the agent teleology present in the actual world is a correlate of the strivings of possibles before creation. After all, as seen, agent teleology corresponds to the order of reasons in God’s mind when he decided which world to actualize, and this order is based on the relative perfection and compatibility of the possible substances God considered. That, in turn, is precisely what the striving of possibles corresponds to, on both interpretations: possibles either carry weight in God’s deliberation in proportion to their reality or perfection, or they literally strive for existence and struggle with one another in proportion to this reality or perfection. The struggle of possibles explains why there exists not only a mosquito whose perfection it is to bite larger animals but also larger animals (like us) who experience being bitten. The perfection of the mosquito took precedence over the perfection of some possible larger animals that never experience mosquito bites. Hence, the mosquito’s perfection prevented these other possible animals from coming into existence while admitting into existence some animals that do experience being bitten. In this way, agent teleology is a correlate of the struggle of possibles before creation.36

34

35

36

Strong textual support for a metaphorical reading is a passage from T 201, which is particularly interesting for my purposes because Leibniz there describes the struggle of possibles as “ideal.” The most relevant portion of the passage is the following: “as soon as God has decreed to create something, there is a struggle among all the possibles, all of them laying claim to existence . . . It is true that this entire struggle can only be ideal, that is, it can only be a conflict of reasons in the most perfect understanding.” See also a dialogue from 1679: “as possible things have no existence at all, they have no power to make themselves exist, and consequently the choice and the cause of their existence has to be sought in a being whose existence is already established” (A 6.4.2232/LGR 130). Of course, they are the ideas of creatures that would, if actualized, strive for and spontaneously produce all of their states. Yet, it would be improper to say that the ideas themselves strive for these states, just as it would seem improper to say that they produce their own states. Only actual substances can truly produce or strive for anything. One indication for this connection is that in M 51, which introduces the notion of ideal influence among finite substances, Leibniz refers the reader, among other passages, to T 201, which explains the doctrine of striving possibles in terms of an ideal struggle or a conflict of reasons in the divine understanding.

Three Types of Teleology

3.3

75

Rational Teleology

In addition to metaphysical and agent teleology, we can distinguish an even more demanding type of teleology that is present only in voluntary actions and that corresponds to rational spontaneity. This type of teleology, which I call ‘rational teleology,’ differs from agent teleology in the following way: a rationally teleological transition is not only a perfection of the agent’s nature but it is the highest perfection a creature can possess (see e.g. “Critical Thoughts on Descartes,” G 4:362/L 388). Such a transition is based on what the agent explicitly judges to be best; as we will see in Chapter 5, voluntary (or free) actions are by definition chosen on the basis of the agent’s understanding of the good. This is precisely the way in which divine actions aim at the good, and it is hence the most perfect way in which any agent can direct herself to the good (see e.g. G 7:111/SLT 95; Gr 481/SLT 98; LC 5.7; letter to Bayle, G 3:58f.; Ta, response 8, G 6:385/H 386). This most perfect way of directing oneself toward the good plausibly manifests a more demanding kind of teleology that corresponds to rational spontaneity.37 To illustrate the distinction between agent and rational teleology, we can again use the mosquito example. When you instinctively and unthinkingly squash a mosquito that is biting you, you are actualizing a perfection of your nature. Your action is therefore an instance not only of metaphysical teleology but also of agent teleology. Yet, when you design a mosquito trap to prevent future bites, you are actualizing a much higher perfection of your nature: your capacity for rational, free agency. Hence, that action is an instance of rational teleology. You are striving for a state based on your rational judgment concerning the good; you are not acting merely on the basis of nonrational strivings that lead you to more perfect states. Again, there is some indirect textual evidence for this distinction. Leibniz says, for instance, “we are free insofar as we are determined to follow the perfection of our nature, which is to say, reason” (“Conversation about Freedom and Fate,” Gr 481/SLT 98). Similarly, he claims that “voluntary actions are no less natural to the soul than any others; one can even say they are more natural, because in them the soul is exercising a perfection of its nature, which is freedom” (remarks on Lamy, G 4:579/WF 154). Thus, voluntary actions are expressions of special perfections of the agent’s nature, and in fact the most crucial perfections of which creatures are capable: rationality and freedom. As a result, these actions are natural and teleological in a particularly demanding sense. While metaphysical teleology consists in acting in accordance with one’s natural strivings, and agent teleology in actualizing a perfection of one’s nature, rational teleology consists in 37

Interestingly, Aquinas also appears to acknowledge a close parallel between the highest type of teleology and the highest type of self-determination when discussing the ways in which rational creatures are closer to God than nonrational creatures (DV q22 a4 corp.).

76

Teleology

actualizing a very special perfection. This special perfection is the ability to knowingly pursue courses of action that one has judged to be best. Further indirect textual evidence for rational teleology is found in passages in which Leibniz appears to equate teleology with voluntariness or freedom, that is, in which he appears to say that only free actions are teleological. One such text is the passage I already quoted in Section 1 from the remarks on Lamy: “Confused perceptions are ordered just like the laws of the motions which they represent. The motions of bodies are explained by efficient causes, but in the distinct perceptions of the soul, where there is liberty, final causes reappear” (G 4:592/WF 166f.).38 This passage is extremely puzzling because it seems to directly contradict other passages in which Leibniz describes all monadic activity – not just free actions or distinct perceptions – as teleological. My interpretation provides an attractive solution to this puzzle: when Leibniz appears to restrict teleology to free actions, he must have in mind the most demanding kind of teleology – rational teleology – which indeed is found only in free activity. Yet, the main reason that I propose the category of rational teleology is not the textual evidence just considered. Instead, it is the fact that Leibniz’s remarks about voluntary actions imply that these actions are teleological in a different sense than nonvoluntary actions. After all, the teleology at work in voluntary actions is such that the agent knows what she is doing and why she is doing it;39 the agent strives for an action by choice, rather than nonvoluntarily. In fact, this type of teleology is often viewed as the paradigm of final causation:40 an intelligent agent’s intentional, deliberate action. Moreover, the distinction between rational and nonrational strivings, and hence between rational and nonrational end-directedness, fits extremely well with the so-called Platonic model of agency which, as I argued in Chapter 2, Leibniz appears to embrace. The real self, on this model, is the rational aspect of the soul, which comprises desires that depend on the agent’s judgments about the good. To the extent that an action springs from “blind” desires – that is, from desires that are independent of what the agent judges to be good and that are not included in her real self – the agent’s autonomy is compromised (see e.g. Watson 2004: 37). For Leibniz, the sharp distinction between those two types of desires or appetitions plausibly yields a special type of teleology that is at work only in autonomous actions. This type of teleology is completely different from the teleology Leibniz recognizes in unreflective, instinctive actions. Animals, lacking reason, are not capable of the most elevated type of 38 39 40

As mentioned earlier, a very similar passage occurs in another remark on Lamy, written two years before (G 4:580/WF 155). Incidentally, this is a requirement for moral agency, as we will see in Chapter 7 (see e.g. CD 98). That is true for philosophers of action today who are sympathetic to teleology; it was also the case among many medieval philosophers: see Des Chene (1996: 187).

Three Types of Teleology

77

Table 3.1 Types of Spontaneity and Teleology Type of Spontaneity

Type of Teleology

Type of Monadic State

Examples

Metaphysical Spontaneity: The agent is exempt from external influences in the metaphysical sense Agent Spontaneity: The agent is exempt from external influences both in the metaphysical and in the ideal sense Rational Spontaneity: The agent is also exempt from nonrational influences or from influences external to her real self

Metaphysical Teleology: The agent is acting in accordance with its natural strivings

All monadic states, that is, actions in the loose sense, which include passive states

Your perception of pain upon being bitten by a mosquito

Agent Teleology: The agent is acting in accordance with strivings for perfections of its nature Rational Teleology: The agent is acting in accordance with its rational strivings

Active monadic states, that is, actions in the strict sense, which exclude passive states Voluntary monadic states, that is, actions chosen on the basis of the intellect’s judgment

The mosquito’s action of biting you

Your action of designing a mosquito trap

teleology: they cannot deliberate about the best course of action, and they do not know what they are doing or why they are doing it. Instead, they are driven entirely by their passions, instincts, and limited recognition of sources of immediate pleasure. A final indication that rational teleology differs in kind from agent teleology is that Leibniz holds that agents are in control of their strivings when acting voluntarily in a way that they are not when acting nonvoluntarily. Like Aquinas, he often describes this by saying that they have mastery over their inclinations or actions.41 Such agents can choose their actions instead of being determined by their insensible, nonvoluntary inclinations: because of their rationality, they can knowingly and deliberately strive for ends that they recognize as good rather than “blindly” – that is, unknowingly – following the ends set by their natures. This strongly suggests that there are deep and significant differences between the kind of teleology present in voluntary actions and the kinds of teleology found in nonvoluntary actions. Table 3.1 summarizes the distinctions for which I have argued in this and the previous chapter. 41

I will spell out Leibniz’s views on mastery in much more detail in Chapter 6. For Aquinas’s discussion of mastery, see e.g. DV q22 a5 ad7.

78

Teleology

4

Neutral Teleology

We saw earlier in this chapter that Leibniz’s world is replete with teleology: there is immanent end-directedness in all nonmiraculous monadic changes. In the present section, I will explore the role that goodness plays in Leibniz’s account of teleology, focusing particularly on the question of whether goodness plays any role in actions that possess only metaphysical teleology. The role of goodness in rational teleology is quite straightforward: rationally teleological actions are performed for the sake of what the agent judges to be best. We can to some extent understand the final causation present in nonrational but agentteleological actions as analogous to rational teleology. When a cat catches a mouse, it might be acting for the sake of pleasure. But can we stretch the analogy to cover actions that are teleological only in the metaphysical sense? Does goodness-directedness govern even the mosquito’s perception of getting squashed? An overwhelming majority of interpreters think that Leibniz views even the least perfect monadic changes as goodness-directed, and that this is the reason why these changes are instances of teleology. Explaining an action teleologically, according to these interpreters, simply means explaining it in terms of the goodness, or apparent goodness, of the state at which the agent aims. The present section challenges the widespread assumption that there is a necessary connection between end-directedness and goodness-directedness. On my interpretation, which I call the ‘Neutral Teleology Interpretation’ (‘NTI,’ in short), Leibnizian teleology does not always consist in acting for the sake of the good. Instead, created monads perform many of their actions simply because their natures prescribe these actions, not because these actions are or appear to be good.42 I argue that teleology as such is independent of – and hence neutral with regard to – the good. What I just suggested may very well strike some readers as a complete nonstarter. Is teleology not by definition goodness-directedness? Or, at the very least, does teleology not traditionally consist in a striving for the good? If so, is it not extremely implausible to suppose that Leibniz would depart from this traditional definition? And finally, are there not many passages in Leibniz’s 42

Larry Jorgensen argues something similar in his unpublished paper “Leibniz’s Appetite.” Yet, he holds that in order for a bare monad to strive for some action A, it must perceive that A expresses its nature. I disagree with him: I do not think it is necessary for the monad to perceive this. In fact, I consider it no more plausible to claim that bare monads can perceive that something expresses their nature than to claim that they can perceive something as good. Stephan Schmid also sketches an interpretation similar to what I call ‘NTI’; he contends that nonrational monads confusedly and unconsciously represent their future states, since those future states are implicit in the contents of their present states. These future states, toward which the monads are unconsciously disposed, are the final causes of their perceptual changes. Schmid stresses that nonrational monads do not consider those activities good (2011: 350). For my criticism of Schmid’s interpretation, see Chapter 1, footnote 66.

Neutral Teleology

79

writings that identify – or at least very closely associate – end-directedness with goodness-directedness? It is true that there is a long philosophical tradition that views all final causation as goodness-directedness and that Leibniz often says things that appear to indicate his adherence to this tradition. Yet, a thorough examination of the relationship between monadic teleology and goodness makes two things clear. First, Leibniz’s writings do not in fact provide us with strong evidence that goodness plays the role in the least perfect monadic actions that most interpreters claim it does. Second, Leibniz has excellent reasons for departing from the traditional understanding of teleology as goodness-directedness. In fact, rather surprisingly, his spontaneity thesis requires him to reject the Aristotelian identification of ends with goods. First, it will help to state explicitly what I, presumably along with other interpreters, mean when I talk of ends and end-directedness. When I say that a monad performs an action for the sake of some end or acts in a certain way because it strives for some end, I mean that this end is the final cause of the action. In other words, the end, or some aspect of the end, is a metaphysical explanation for the action. A final cause, after all, is supposed to be a type of cause or explanation. Hence, when other interpreters argue that teleology is always goodness-directedness, I understand them as claiming that the goodness or apparent goodness of the state aimed for is an explanation for the action. A monad strives for certain ends, according to these interpreters, because those ends appear good to the monad or because they are good. I take it that teleology is usually understood in the way I just explained. If those who claim that all teleology for Leibniz is goodness-directedness mean merely that all ends are goods, even though this goodness is not always an explanation for the action, I do not have any objections to their view. Yet, that would not be a very interesting claim. For Leibniz, all events in the actual world do in fact contribute to the perfection of the world; hence, all strivings that succeed are strivings for some aspect of the best possible world. I presume that most if not all of these interpreters intend to make a stronger claim: the claim that the goodness or apparent goodness of the end is part of the explanation for all monadic actions. That is the claim I will challenge. This explication of teleology immediately makes unavailable one strategy for connecting teleology with goodness. The strategy I have in mind consists in arguing that an end is by definition a good because whatever an agent strives for is ipso facto good (or ipso facto seems good to the agent). In other words, what an agent strives for, on this view, is good (or seems good) simply because the agent strives for it. Let us call this type of goodness ‘ipso facto goodness.’ One might think that if one acknowledges ipso facto goodness, the attempt to divorce goodness from teleology must fail because it is conceptually necessary that the object of a striving be a good. Yet, given my explication of teleology, this strategy cannot succeed because its account of strivings is viciously

Teleology

80

circular: it cannot be the case both that what an agent strives for is good because the agent strives for it and that the agent strives for it because it is good. In order to maintain that goodness is the final cause of all monadic actions, one therefore needs a different explanation for the goodness or apparent goodness of the end. If the ends of some actions are or appear good only in the ipso facto sense,43 goodness cannot be the final cause of all actions. 4.1

Standard Interpretations

There are two standard interpretations of monadic teleology that I want to reject. I call them ‘the Apparent Good Interpretation’ (AGI) and ‘the Objective Good Interpretation’ (OGI). AGI is by far the most popular way of interpreting Leibnizian teleology. According to this interpretation, every monadic action is performed for the sake of what appears good to the agent, and every monadic striving is a striving for an apparent good. One prominent proponent of this interpretation is John Carriero, who describes the end-directedness of monads by saying that every monad perceives the entire universe, and its appetitions “are responsive to what appears best” in its perception (2008: 134). This, according to Carriero, “holds as much for bare monads . . . as it does for spirits,” and it holds for all appetitions and perceptions of spirits, whether they are conscious or unconscious. Appetition, he says, “always tends to an apparent good” (2008: 135). Martha Bolton also endorses AGI: she argues that the ends of any monad “consist in attaining the things which it perceives to be imminent and valuable given its perspective on the world, and each of its acts implements its ends as well as possible under the circumstances” (2013: 191, similarly 194). A further example is Jeffrey McDonough, who argues that for Leibniz there are always two equally valid ways to explain a monadic action. One type of explanation that works for all monadic states is viewing them as the results “of a teleological process governed ultimately by the creature’s perception of the good” (2011: 197).44 Pauline Phemister (2005: 223), Sukjae Lee (2014),45 and Nicholas Jolley (1998: 607, 2005: 68) also endorse AGI. 43

44

45

I do not mean to deny that ipso facto goodness can legitimately be viewed as a type of goodness, nor even that Leibniz acknowledges it as a type of goodness. In fact, I will later suggest that Leibniz does acknowledge it. What I do mean to deny is that ipso facto goodness can be a final cause. McDonough does admit elsewhere that this kind of teleology is “less intuitive” in the case of bare monads and that one could interpret Leibniz as applying this teleology only in the case of higher monads. Yet, McDonough thinks that Leibniz is committed to applying it across the board (2016b: 101n28). Lee’s interpretation is quite different from the others. He argues that there are two separate “strands of appetitive activity” in monads: one consists in “law-following and preserving the harmoniousness and excellence of the law of the series,” the other in “striving for the desirable state of being more completely at ease” (2014: 143). Both of these, however, appear to be types

Neutral Teleology

81

This brings us to the other standard interpretation: the Objective Good Interpretation, or OGI. The main proponents of OGI are Donald Rutherford (2005, 2013) and Marleen Rozemond (2009). Like AGI, OGI maintains that teleology is always goodness-directedness (see Rutherford 2005: 166, 2013: 170). Yet, unlike AGI, it holds that some monadic strivings aim not at what seems good to that monad but rather at what is objectively best, or at what God knows to be best. According to Rutherford, there are two types of final causation, namely what he calls “natural teleology” and what he calls “desire teleology” (2005: 167f.). All monadic actions follow the laws of natural teleology: they are “movements toward the best next state of the universe” because God has created the world in accordance with the Principle of Goodness (2013: 170; see 2005: 167). Actions are teleological in this sense because they are explained in terms of what God knows to be best; the finite agent need not perceive anything as good in order to be subject to natural teleology (2005: 174; see Rozemond 2009: 293f.). While all monadic actions are subject to natural teleology, according to Rutherford, some are also subject to desire teleology: they result from strivings not just for what is best objectively but also for what seems best to the agent (2005: 174). Thus, OGI fleshes out the claim that all creaturely teleology consists in striving for the good by distinguishing between (a) striving for what seems best to the creature and (b) striving for what is objectively best. All monadic actions are teleological in sense (b), and some are additionally teleological in sense (a). Elsewhere, I argue against the two standard interpretations at length (Jorati 2013). My objections to AGI are roughly the following. First, there are good reasons to think that bare monads are not capable of perceiving anything as good. Since there is nevertheless teleology in bare monads, AGI cannot be the correct model. Second, AGI does not have a good explanation for transitions to less perfect states. Even if bare monads were able to perceive things as good, how would their perceptions of the good explain changes that are extremely bad for them, such as the destruction of most of a monad’s organic body?46 One major problem with OGI, on the other hand, is that it views only some monadic actions as immanently teleological. For OGI, after all, some monadic actions are teleological only because of extrinsic factors, namely the objective goodness of subsequent states, which is not represented in the monad. That is a problem, as we will see in Chapter 4, because it makes it difficult to attribute those actions to created agents and to deny that God is the author of evil.

46

of goodness-directedness, even though they involve two distinct kinds of goodness; the first one involves the goodness consisting in the “harmoniousness of the world,” the second involves the goodness consisting in “being more completely at ease” (2014: 144). Lee takes monadic actions generally to be appetitive or teleological in the sense of “pursuing what is desirable” (2014: 143). For more details about these two objections, see Jorati (2013: 47ff.)

Teleology

82

Furthermore, the objective goodness that OGI invokes does not provide a true, bottom-level explanation for the states of created substances and thus does not meet the criteria for teleology sketched earlier. After all, as seen, creaturely essences are prior to God’s decision to create this world; he merely discovered them in his intellect and selected the ones for actualization that together constitute the best possible world. As a result, it would be wrong to say about a created substance that it acts in a particular way, or that something is included in its nature, because of God’s plan.47 God simply actualized the possible substances that, as he foreknew, would be naturally disposed to produce the states that accord with his plan.48 4.2

An Alternative Interpretation

For the reasons just cited, we should prefer an interpretation on which there is immanent teleology in all monadic actions. At the same time, there are also strong reasons for rejecting interpretations on which all teleology, even that of the least perfect monadic actions, consists in strivings for the apparent good. My alternative, the Neutral Teleology Interpretation (NTI), achieves both of these desiderata. On my interpretation, all monadic actions are teleological because they are the outcomes of the agent’s natural strivings, whether or not these strivings aim at the good. In other words, NTI holds that there can be final causation that is not goodness-directedness.49 Some final causation consists in neutral end-directedness. As seen in Chapter 1, Leibniz distinguishes three types of appetitions: rational appetitions, sensible appetitions, and insensible appetitions. Rational appetitions are similar to the strivings described by AGI, and to Rutherford’s desire teleology: when I deliberately choose to do something, I strive for it because I judge it to be best. These are the strivings at work in rational teleology. Most of our strivings, however, are nonrational and not aimed at the good in this way; they belong into one of the other two categories. Sensible appetitions can be strivings for a type of goodness: pleasure is a kind of goodness, after all, and sensible appetitions often aim at pleasure. Yet, insensible appetitions such as the strivings for less perfect states that are characteristic of merely metaphysical teleology – for instance, the strivings leading to a monad’s perception of major damage to its body – are not goodness-directed, on my view. Their end-directedness is neutral. Even many agent-teleological 47 48 49

See Madden (2003: 183), who also argues that the ends of substances that they have in virtue of their natures are prior to the ends of the created world as a whole. For more on my objections against OGI, see Jorati (2013: 54f.) The only other interpreters who argue for an interpretation like this, to the best of my knowledge, are Schmid and Jorgensen. Yet, neither of them spells it out or defends it in detail, and I disagree with parts of their accounts; see footnote 42.

Neutral Teleology

83

strivings are neutrally end-directed – for example, those that occur in bare monads. A monad strives for these states not because they seem good or appear to be a means to some good, nor because of God’s ends, but simply because that is the monad’s nature. To see why what I just described should count as final causation, recall that there are traditional accounts on which final causation is a presupposition of all efficient causality. In Section 1, we considered two such accounts – a more demanding one and a less demanding one. Neutral teleology qualifies as final causation on both of these accounts.50 Moreover, there is an important sense in which a monad is functioning properly when it is acting in accordance with its nature and in which the specification of states by a monad’s nature is normative, or a prescription. One can say that given its nature, this is how the monad is supposed to act. If God were to produce a state in a monad miraculously, that would be improper and against its nature. Thus, to the extent that God allows monads to act in accordance with their natures, there is a kind of normativity and teleology. Still, I do not think that we should equate this properness or naturalness with goodness or perfection: Leibniz needs a strict distinction between natural states in which a monad becomes more perfect or better and natural states in which the monad becomes less perfect or worse.51 Take the example of the central monad of a plant that perceives getting eaten by a cow. Even though it is natural for this monad to lose most of its organic body in the encounter with the cow, and even though it is in one sense what this monad’s nature dictates should happen, it is not good for this monad, and it is not a perfection of its nature.52 The natural and the good, and hence the normative and the good, must sometimes come apart for Leibniz. That is precisely where he disagrees with medieval Aristotelians, as we will soon see. 4.3

Potential Obstacles

My discussion so far has shown, I hope, that NTI has several advantages vis-àvis the two standard interpretations. Yet, there are also potential obstacles to adopting NTI: (a) the tradition of equating end-directedness with goodnessdirectedness and (b) textual evidence suggesting that Leibniz adheres to this tradition. I will examine these obstacles and argue that neither of them is 50

51 52

Note, however, that Carriero – on whose interpretation the more demanding account is based – interprets Leibniz in accordance with AGI, as seen earlier. Even though he does not appear to think that teleology generally has to be goodness-directedness, he thinks that this is the case for Leibniz. He needs this, for instance, to account for the ideal influence of one monad on another, as seen in Chapter 2. At least, it is not good for the monad in any relevant sense. It is good in the ipso facto sense, but this type of goodness is much thinner than the goodness with which Leibniz is typically concerned, and, as already seen, it cannot be a final cause. See footnote 53.

84

Teleology

insurmountable. In fact, surprisingly, a closer look at the differences between Leibnizian and medieval Aristotelian teleology will strengthen my interpretation: it will show that Leibniz’s spontaneity thesis forces him to depart from the traditional understanding of the relationship between teleology and goodness. There is some textual support for a close connection between goodness and Leibnizian teleology. Leibniz writes, for instance, that “our souls . . . can be moved only by some reason of good or evil, even when no distinct knowledge can be extracted from them” (COE 3). Likewise, in the “Principles of Nature and Grace,” Leibniz states that “the perceptions in the monad arise from one another by the laws of appetites, or by the laws of the final causes of good and evil” (PNG 3; similarly in Beeley 11/AG 279). These passages can be seen as evidence for either one of the standard interpretations because they point to a close relation between final causation and goodness without specifying what precisely this relation is. Leibniz’s corpus also contains a few passages that, taken at face value, seem to support AGI in particular. For instance, he writes to Queen Sophie Charlotte in 1704 that “everything in the soul happens morally, or in accordance with perceived good or evil . . . [E]ven in our instinctive or involuntary actions . . . there is in the soul an appetition for good or an aversion to evil which directs it” (G 3:347/WF 224; translation altered). A similar passage occurs in a letter to Damaris Masham, also from 1704: “Everything . . . comes down to a present state combined with a tendency towards changes, changes which are brought about . . . in the soul by perceptions of good and evil” (G 3:341/WF 206). These passages do indeed seem to suggest that appetitions are always directed toward apparent goods and that perceived goodness is the final cause of any monadic action. There are several things to say about those texts. Most importantly, it is possible to read these passages as invoking only goodness in the sense in which every object of an appetition or desire is ipso facto a good for the agent.53 Calling something a “perceived good,” accordingly, could simply be a way of saying that the agent is inclined toward the object of a perception or that she is inclined by a certain prior perception. The locution ‘x appears good to monad m’ would then just be another way of saying “m strives for x.” Acknowledging this as a type of goodness is rather common among Leibniz’s predecessors; Aquinas endorses it (e.g. ST IaIIae q94 a2 corp; SCG 1.37.4; 3.3.3), as do 53

It is clear that Leibniz acknowledges different kinds of goodness. In several places, he distinguishes three types: physical, metaphysical, and moral (see CD 30ff.; T 21). The least demanding type of goodness, metaphysical goodness, consists in perfection. Yet, perhaps there are two senses of ‘perfection’: a loose sense of perfection, according to which acting naturally is always a perfection, and a stricter sense, according to which only some aspects of one’s nature are perfections. As seen, Leibniz often uses the term in the latter sense, which he needs for his account of ideal influence. Yet, he may well acknowledge the loose sense as well, a sense coextensional with ipso facto goodness.

Neutral Teleology

85

Hobbes (e.g. Leviathan Ch. 6, §7) and Spinoza (e.g. Ethics 3p9s), to name just a few. The sense in which monads strive for the good during very imperfect changes could, in other words, be a striving for the good simply in the sense in which all objects of appetitions are ipso facto good. Yet, goodness of this type cannot be the final cause of a striving, as already argued earlier. Thus, I can grant that whatever a monad strives for is something that is good for that monad – in this derivative and very thin sense – and even that it is a perceived good in the sense just described. This does not mean that goodness or apparent goodness is a final cause of all actions. At bottom, something other than ipso facto goodness must explain why the action occurs – for instance, the agent’s nature, as NTI maintains. A passage from the New Essays may appear to support AGI by suggesting that there are insensible perceptions of the good that can incline us. In the context of a discussion of pleasure and happiness, Leibniz’s spokesperson Theophilus explains, [F]undamentally pleasure is a sense of perfection, and pain a sense of imperfection, each being notable enough for one to become aware of it. For the minute insensible perceptions of some perfection or imperfection, which I have spoken of several times and which are as it were components of pleasure and of pain, constitute [forment] inclinations and propensities but not outright passions. (NE 194)

This passage does appear to say that at least some insensible inclinations are inclinations toward a good, or perfection, that is perceived insensibly. Yet, even if that is the correct reading, it would not get proponents of AGI very far. After all, Theophilus is not stating here that all monads possess this kind of perception and inclination, nor is he stating that all insensible inclinations are inclinations toward an apparent good. He is, rather, discussing a very particular type of perception: the minute perceptions that make up a conscious perception of pain. As seen in Chapter 1, Leibniz holds that every monadic state is infinitely complex and that distinct sensitive perceptions and inclinations are “composed of innumerable little perceptions and little inclinations of which we cannot be conscious” (letter to Remond, November 4, 1715, G 3:657/W 554). Hence, a conscious feeling of pain, or a conscious perception of imperfection, must have insensible perceptions as components. Those might arguably, in some sense, be perceptions of imperfection as well. Yet, this does not mean that bare monads are capable of perceiving perfection and imperfection, nor that the minute perceptions in minds or animal souls that do not compose pleasure and pain are perceptions of perfection and imperfection. Moreover, it is not even clear that the passage under discussion is claiming that the perfection that is the object of insensible perceptions functions as a final cause. After all, it is again possible to understand the references to perfection or goodness in what is roughly the ipso facto sense. This reading receives support

86

Teleology

from another passage from the New Essays. Theophilus there explains that “nature has given us the spurs of desire in the form of the rudiments or elements of suffering [douleur], semi-sufferings one might say, or (to put it extravagantly for the sake of emphasis) of minute sufferings of which we cannot be aware [petites douleurs inapperceptibles]” (NE 165). These minute pains or “spurs of desire” are, presumably, natural or instinctive appetitions.54 Leibniz states that it is a good thing that we cannot be aware of them. After all, if we perceived them distinctly, we “would always be miserable when looking forward to something good” (NE 165; see also NE 188f.). This is presumably because we would then constantly be aware of what we are lacking.55 Interestingly, he then tells us that the satisfaction of these spurs of desire “provides us with many semi-pleasures,” which, when they continue and accumulate, can become “genuine pleasure” (NE 165). As a matter of fact, Theophilus stresses, “without these semi-sufferings there would be no pleasure at all” (NE 165; see also NE 56). What he appears to be arguing here is that pleasure or semi-pleasure, at least in the cases under discussion, is not something that explains why we strive for certain states but is instead something that results from the satisfaction of strivings. The fundamental strivings, then, are not explained by what we find pleasing; rather, they are provided to us by nature. While these strivings might aim at something independently good, they do not arise because we perceive their objects to be good. Rather, they arise because that is the nature of the agent. Someone may object here that goodness-directedness still plays a role at the fundamental level on the view described in NE 165. After all, the objection goes, the monad fulfills its desires, or overcomes its semisufferings, only because it strives for a better state, a state in which those desires are satisfied. The goodness of that state, one might be tempted to claim, is the final cause of the transition. Another passage from the New Essays may seem to bolster such an objection. That passage claims that we are usually spurred on by “minute impulses [which] consist in our continually overcoming small obstacles – our nature labours at this without our thinking about it . . . [W]e are never without some activity and motion, simply because nature continually labours to be more completely at ease” (NE 188). Theophilus even goes on to say that “nature’s accumulation of continual little triumphs, in which it puts itself more and more at ease – drawing closer to the good and enjoying the image of it, or reducing the feeling of suffering – is itself a considerable pleasure, often better than the 54

55

Later in this passage, Leibniz calls them “imperceptible little urges,” “confused stimuli,” and “impulses [which] are like so many little springs trying to unwind and so driving our machine along” (NE 166). This is how Leibniz distinguishes semi-sufferings from passions: in the case of the latter, “we at least know what we want” (NE 166).

Neutral Teleology

87

actual enjoyment of the good” (NE 189).56 Theophilus appears to be suggesting that what drives transitions in the monad is, at bottom, the monad’s desire to be more completely at ease, to overcome semi-sufferings and convert them into semi-pleasures.57 Yet, I do not ultimately find this objection compelling. After all, our nature’s continual laboring “to be more completely at ease” arguably does not constitute genuine immanent teleology. Instead, it merely appears to be a way of referring to the general disposition that monads have of finding a balance between all of their inclinations – or, in fact, a way of referring to the laws that govern the conflict of inclinations. While this disposition or law may be good, it is an innate aspect of the monad’s nature rather than something it obeys because it recognizes it as good. As a matter of fact, Leibniz appears to ascribe this very same tendency toward a state of being more at ease to bodies, which clearly do not have the capacity of acting for the sake of the apparent good. In the paragraph from which I quoted above, in which Theophilus discusses the “spurs of desire” (NE 165), he compares the uneasiness in our minds to the disquiet of “our bodies, which can never be perfectly at their ease. For if one’s body were at ease, some new effect of objects . . . would at once alter the balance and compel those parts of the body to exert some tiny effort to get back into the best state possible” (NE 166; similarly in T 325). Just as minds have the tendency to restore equilibrium or find a balance between all their competing inclinations, bodies have the tendency to balance the forces to which they are subject and find an equilibrium. Because the latter should not be understood as a striving for what appears good, the former need not – and should not, especially given the other problems with AGI – be understood in that way either. Finally, there is at least one text that supports OGI: a note scribbled into Leibniz’s copy of Aloys Temmik’s 1706 book Philosophia vera. In his chapter on sensible beings, Temmik claims that “it is better to say that a brute animal is led toward the end [agatur ad finem] than that it acts [for the end], or it acts for the end intended by God, not intended by it” (Temmik 1706: 181). Leibniz underlined the phrase “acts for the end intended by God, not intended by it” and wrote “correct” (recte) above “not intended.”58 This supports OGI insofar as Leibniz here appears to agree with Temmik that some creaturely actions are teleological only in the sense that God has directed them toward the good. 56 57

58

See also NE 166: “insensible stimuli . . . make us find one direction of movement more comfortable than the other.” This passage supports Lee’s version of AGI more than other versions. Lee’s interpretation, after all, is based on the idea that appetites tend “toward the desirable state of being more at ease” (2014: 143). A facsimile of Temmik’s work with Leibniz’s notes is appended to Mugnai (1992); the page under discussion is reproduced on page 279 of Mugnai’s book and the transcription of Leibniz’s note is on page 164.

88

Teleology

Hence, it also constitutes evidence against AGI. Yet, the passage is not a perfect fit with OGI because Temmik is discussing animals, which, according to OGI, are capable of desire teleology and thus direct themselves toward ends in some of their actions. As a result, the passage is problematic for OGI, AGI, and NTI. In light of the reasons for rejecting AGI and OGI that I mentioned earlier, I find it most plausible to interpret Leibniz’s note as indicating not that animals lack immanent teleology altogether but only that they cannot intend ends in the strong sense required for rational teleology. Even if this may not be the most straightforward reading of the note, it is less far-fetched than it may initially seem, given the context in which Temmik’s statement occurs. Immediately before the sentence quoted earlier, Temmik says, “the actions of brute animals prove the remarkable wisdom and power in the creator, who entirely contains their ends; he directs them to these actions” (1706: 180f.). As a result, Leibniz’s note may simply be an indication that Leibniz agrees with Temmik that animal actions display divine wisdom because they serve God’s ends. He does hold, after all, that God’s wisdom is evident in the natural order of the world and that God is, in some sense, the last final cause of the world (e.g. “Against Barbaric Physics,” G 7:344/AG 319; “Ultimate Origination of Things,” G 7:305/AG 152; M 90). There appears to be sufficient evidence elsewhere in Leibniz’s corpus that he is committed to another, and more fundamental, kind of teleology in the actions of all creatures, namely immanent teleology. The overall textual evidence, then, does not give us strong reasons to reject NTI. Of course, there is a long tradition of equating final causation with goodnessdirectedness. In particular, Aristotelians – starting with Aristotle himself and continuing in medieval Scholasticism – typically see all ends as goods. For our purposes here, it makes sense to focus on the Scholastics because their version of Aristotelianism was especially prominent in Leibniz’s time.59 Many major Scholastic thinkers hold that ends are always goods, and thus all end-directedness is goodness-directedness. On this point, Leibniz differs from the Scholastic tradition, on my interpretation. But it may seem puzzling why Leibniz would depart from the tradition in this respect since he follows this tradition in many other respects. Yet, on reflection, this departure is not surprising at all: it is in fact required by Leibniz’s spontaneity thesis. He disagrees with Scholastics concerning the causal interaction of created substances, and this requires him to also disagree with them concerning the goodness-directedness of these substances. To see why Leibniz’s spontaneity thesis should have such surprising consequences, we need to take a short detour and examine the Scholastic view more closely. I will concentrate on Aquinas because he provides several particularly instructive arguments for the claim that all ends are goods. Some 59

With respect to Aristotle, see Physics ii.3, 195a23–26 and Johnson (2005: 91f.).

Neutral Teleology

89

of these arguments invoke ipso facto goodness, which as shown earlier cannot be the final cause of actions.60 Yet, Aquinas also provides arguments that can be interpreted as ascribing immanent goodness-directedness to all creatures.61 These arguments have to do with his understanding of creaturely natures and substantial forms. In rough outline, he holds that whatever is specified by a substance’s substantial form is good, and that “everything, insofar as it is in act, acts and tends towards that which is fitting for it, in accordance with its form” (ST I q5 a5 corp.). For Aquinas, every substance has a substantial form in virtue of which it belongs to a certain lowest species and has the characteristics proper to that species (see ST IaIIae q85 a4 corp.). The characteristics in virtue of which a substance belongs to a particular species include powers or potentialities for activities that are essential to that species. For instance, the characteristics proper to human beings include the power for rational activity. Since the substantial form of a substance specifies the characteristics that are proper for the kind of thing that it is, the substance is perfect of its kind to the extent that it actualizes those potentialities. Human beings, for instance, are perfect human beings to the extent that they actualize their power of reasoning. Thus, Thomistic substantial forms include only characteristics that constitute the substance’s perfection.62 On one interpretation of this type of argument, goodness is the final cause of all natural actions. The perfection that is specified by a creature’s substantial 60

61

62

Aquinas argues for instance that “the end is that in which the appetite of an agent or mover, and of that which is moved, finds rest. But the nature of the good is that it provides a terminus for appetite: the good is that which all desire. Therefore, every action and motion is for the sake of a good” (SCG 3.3.3). See also ST I q5 a6 corp. Another reason Aquinas offers is that while rational agents direct themselves to the good, nonrational agents are directed to the good by rational agents (see SCG 3.3.7; DV q22 a1 corp. and a5 corp.). This is similar to Rutherfordian natural teleology: such actions are goodness-directed only because God, or some other rational agent, knows the end to be good and directs the nonrational agent to this end. There is one crucial difference between Aquinas and Leibniz in this respect, however: Aquinas’s God literally builds up creaturely natures by giving them certain inclinations. These creatures therefore owe even their natural end-directedness to God. Leibniz’s God, on the other hand, finds creaturely essences ready-made in his intellect; he does not add any appetitions or inclinations. Rather, he merely actualizes some of the essences he finds in his intellect – inclinations and all. God makes sure his creatures have the inclinations he wants not by bestowing these inclinations on creatures but rather by actualizing the right ones. Carlin argues, against many other interpreters, that Aquinas did not believe that there is immanent teleology in nonrational creatures (2012). I disagree, and I read Aquinas, like Leibniz, as acknowledging different types of immanent end-directedness, corresponding to different types of appetition (see e.g. DV q22 a4 corp.). In the passages Carlin cites, Aquinas is, on my interpretation, denying only that lower creatures possess immanent teleology of the more demanding types. For the relation between the Scholastic understanding of creaturely natures and teleology, see MacDonald (1991: 20) and Carriero (2005: 120). When making this point, Aquinas uses ‘perfection’ in what we today would think of as two senses: the substantial form of a human being not only specifies what a good, or perfect, human being ought to do, but it also specifies a state of perfection in the sense of completion.

90

Teleology

form is also the immanent final cause of the corresponding actions. Note, however, that this only works because Thomistic natures contain nothing but perfections, so that whatever a substance naturally strives for is a state of increased perfection or being. As a result, whenever a substance becomes less perfect, it is because something interfered with, or prevented the efficacy of, its natural strivings. As Aquinas says in the Summa Theologiae, “that anything depart from its natural and due disposition can only happen through some cause that draws it out of its proper disposition” (ST I q49 a1 corp.). In other words, the loss of perfection is always violent.63 Yet, this cannot be how things work for Leibniz. After all, Leibniz denies that there is any true causal interaction among created substances and that God ever prevents creatures from acting in accordance with their natures (except perhaps in very rare cases of miraculous interventions). As a result, on Leibniz’s view, a finite substance’s loss of perfection cannot strictly speaking be explained through the interference of something else with its nature. For instance, when I feel pain or lose some perfection, the explanation for this must ultimately lie in myself, that is, in my own natural dispositions. Other created substances can only be ideal causes of my suffering, as already explained. Somehow, then, it must be in my nature to feel this pain or lose this perfection; after all, I am producing the painful state spontaneously, without anybody interfering with my natural inclinations. What Thomists would describe as the interference of external factors with my strivings must, for Leibniz, take place entirely within me and be written into my very nature. This, of course, constitutes a radical departure from the Aristotelian worldview on which whatever is part of a thing’s nature, and whatever it naturally strives for, is also a perfection of its nature. For Leibniz, the natures of creatures sometimes prescribe changes that are imperfections for that creature, and as a result, creatures are sometimes naturally inclined toward less perfect states. Leibniz’s spontaneity thesis has far-reaching consequences, and one of these is that not all teleology is goodness-directedness. 5

Conclusion

We saw in this chapter that teleology, like spontaneity, is ubiquitous for Leibniz: all natural changes in the monadic realm are instances of immanent final causation. Yet, we can distinguish three types of immanent teleology that are exactly parallel to the three types of spontaneity distinguished in Chapter 2. Metaphysical teleology – the broadest type – requires merely that the substance 63

Aquinas says elsewhere, “that which is in a thing against the motion of its natural appetite is violent and unnatural. Hence, evil in each thing is violent and unnatural, insofar as it is an evil for that thing” (SCG 1.39.7).

Conclusion

91

is acting in accordance with its natural strivings. Agent teleology, on the other hand, is more demanding: in order for a transition to be agent-teleological, it must not only be dictated by the agent’s nature but aim at a perfection of the agent’s nature. Rational teleology, finally, is the most elevated kind of teleology: it requires that the transition be voluntary, which means that it aims at what the agent judges to be, or rationally recognizes as, best. I have also argued that many instances of Leibnizian teleology are not instances of goodness-directedness. A large number of monadic actions are simply based on natural strivings, and these natural strivings are sufficient for making the resulting actions teleological; there is no additional requirement that the finite agent or God perceive the objects of these strivings as good. I have also shown that the two main obstacles to adopting my interpretation are surmountable. Even the most problematic texts can be read in a way that is consistent with my interpretation, and there is compelling evidence that goodness cannot be a final cause of the least perfect monadic actions. A look at the reasons cited by Aquinas for believing that all ends are goods supports my interpretation, because it reveals that Leibniz cannot agree with him, given his spontaneity thesis. He cannot reasonably maintain that all monadic strivings aim at the good because he holds that all states of a monad – including losses of perfection – arise spontaneously from the monad’s nature, and can be explained teleologically.

4

Attributability and Divine Concurrence

Previous chapters explored monadic agency generally and distinguished between three distinct types of agency, corresponding to three types of spontaneity and teleology. The present chapter tackles a further issue that is central to the philosophy of action: the attribution of actions to agents. As even a cursory glance at the recent philosophical literature reveals, this is a complicated issue. Leibniz – along with many other historical figures – faces an additional complication: he holds that God concurs, or cooperates, with every creaturely action. Created substances cannot do anything without divine concurrence. Hence, in contemporary parlance, all creaturely actions are cases of shared agency:1 God and the creature act jointly; a creature never acts on its own. As one might expect, this doctrine is problematic in a number of ways. Here, I focus on two closely related worries. First, it seems that when God concurs with a creature, the resulting action is properly attributed to both God and the created agent. Second, it seems that concurrence undermines creaturely spontaneity. These are serious worries.2 It is important to Leibniz that when God concurs with a creature’s activity, the resulting action is the creature’s action alone, not God’s. Likewise, as already seen, spontaneity is a necessary condition for substancehood and freedom; a threat to the spontaneity of creatures is a threat to their status as substances and as free agents. This chapter shows that Leibniz can solve both of these problems. By invoking final and formal causation, he can reconcile divine concurrence with creaturely spontaneity and with the attribution of actions solely to creatures. After all, these types of causation allow us non-arbitrarily to ascribe an action to one agent, even if there are other concurring efficient causes. They also enable us to see that divine concurrence does not compromise the type of 1 2

In at least one text, Leibniz describes God and the creature as acting together (coagere) when God concurs with a creature (A 6.4.1605/LGR 263). They are not, to be sure, the only worries arising from concurrentism. Another worry – which has received much more attention in the literature – concerns the intelligibility of this type of shared agency. Concurrentists hold that God and the creature produce an effect jointly; both contributions are necessary for the effect, but it is not the case that each agent produces only part of the effect. How precisely is this supposed to work? I will not discuss this issue here; see McDonough (2007) for a helpful discussion.

92

Two Problems

93

spontaneity that is crucial for substancehood; I will call that type ‘explanatory spontaneity.’ Because Leibniz, on my interpretation, locates the formal and final causes of creaturely actions in creatures,3 there is a sense in which these actions are made intelligible by that creature alone. That, I claim, is the kind of spontaneity that Leibniz requires for substancehood. Moreover, we can understand rational spontaneity – the type of spontaneity that is present only in free actions, as argued in Chapter 2 – in a corresponding way and hence reconcile creaturely freedom with divine concurrence. 1

Two Problems

There are several passages in which Leibniz explicitly endorses concurrentism. He says in Causa Dei, for instance, that “things depend on God with respect to their actions, since God concurs in the actions of things insofar as there is some perfection in these actions” (CD 10; emphasis omitted; similarly in “On God and Man,” G 3:29f./LGR 289). Similarly, Leibniz claims elsewhere that “God produces substances, but not their actions, with which he only concurs” (Beeley 13/AG 281).4 Without divine concurrence, Leibniz states, creatures would not be able to act at all: “God’s concurrence [is] so necessary that, whatever creaturely power is assumed, no action would follow if God were to withdraw his concurrence” (letter to Des Bosses, February 2, 1706, LDB 11). Leibniz’s endorsement of concurrentism gives rise to what I will call ‘the spontaneity problem.’ This is the problem of explaining how creatures can be spontaneous if God concurs with all of their actions. It is true that in formulating his spontaneity thesis, Leibniz sometimes explicitly exempts divine influence: spontaneity requires only the absence of causal influences from other created substances (e.g. T 59). Yet, it may seem that this exemption is ad hoc unless Leibniz can explain why God’s causal influence is unproblematic while the causal influence of other creatures would be problematic.5 The problem is important because, as mentioned, spontaneity is a necessary condition for 3

4

5

By saying that the final cause is located in the creature, or that the creature is contributing the final cause, I mean simply that the final cause of the action is an end for which the creature strives or toward which it is naturally directed. Of course, the creature can strive for something that is not strictly speaking inside the creature; I merely use that locution as a convenient shortcut. This passage and others like it are evidence against Whipple’s claim that, for Leibniz, concurrence is not ontologically distinct from creation or from conservation (2015: 216f., similarly 2010a: 874). There is, of course, precedent for exempting divine influence in this way, and some philosophers do not consider it to be ad hoc. Descartes, for instance, famously claims that ‘substance’ does not apply univocally to creatures and to God, because only God is independent of all other substances (Principles I.51, AT 8:24). Yet, Leibniz not only has stricter requirements for substancehood than Descartes but is also more of a naturalist: he is less willing to allow for exceptions. For these reasons, I think that Leibniz would not – or at least should not – make an exception for God without a satisfactory explanation of why God’s influence is unproblematic.

94

Attributability and Divine Concurrence

substancehood and for freedom. If Leibniz cannot solve the spontaneity problem, major elements of his philosophical system – namely, his commitments to created substances and human freedom – would collapse. Concurrentism gives rise to a further problem, which I will call ‘the attributability problem.’ Despite God’s direct involvement in creaturely actions, it is important to Leibniz that these actions are properly attributed to creatures rather than to God. Unfortunately, it is difficult to see why this should be the case. God’s contribution to such an action is so similar to the creature’s contribution that there may not seem to be any basis for attributing the action only to the creature. This is of course particularly worrying with respect to evil or sinful actions: attributing evil actions to God would undermine his perfect goodness. Yet, the problem is not limited to evil actions; Leibniz is arguably committed to attributing some non-sinful actions to creatures and not to God. When Leibniz talks about the attribution of actions to agents, he often uses the term ‘author’ (Latin: autor; French: auteur). For instance, he repeatedly says that God is not the author of evil actions (e.g. CD 64; 68; “Commentary on Burnet” §56(d), DPG 137 and §39(b), DPG 117; CD 64; 68). Using this terminology, we can describe the attributability problem as follows. Leibniz needs a theory of attributability that allows him to deny that when God concurs with a creaturely action, the creature and God are coauthors of the action – or, even worse, that God is the sole author of the action and that creatures are merely instruments of divine agency.6 In other words, God’s role in creaturely actions cannot be analogous to your role when you and I are jointly carrying a heavy box, nor can it be analogous to your role when you are driving a car. In the former example, you and I are coauthors of the action, and it would be wrong to ascribe the action to only one of us. In the latter example, both you and your car are contributing to the effect but in such a way that you are the author while the car is, intuitively, merely an instrument that you employ. What Leibniz needs, then, is an account on which we can non-arbitrarily attribute actions to creatures, and not to God, even though God and the creature coproduce these actions. If such an account is not available to Leibniz, God would be a coauthor of all creaturely actions and would plausibly bear at least as much moral responsibility for these actions as his creatures.7 Before proceeding, let me say a few more words about the term ‘author.’ To the best of my knowledge, Leibniz never defines or explicates this term in his mature works. There is an explicit definition in an early text, however: in the 6

7

Some Scholastics embrace the view that creatures are divine instruments (see e.g. Aquinas, De potentia III.7; see also Schmaltz 2008: 17). Yet, other Scholastics reject this doctrine (see e.g. Suárez, MD 17.2.15; 17/OEC 27ff.), at least in part because they acknowledge that in genuine instrumental causation, the action is not attributed to the instrument but only to the agent who acts through the instrument (see Suárez, MD 17.2.8/OEC 19f.). See Chapter 7 for a discussion of the relationship between authorship and moral responsibility.

Two Problems

95

“Confession of a Philosopher,” a work in dialogue form from the early 1670s, Leibniz’s spokesperson says, “To be the author [autorem esse] is by one’s will to be the ground of something else” (A 6.3.127/CP 55; similarly in A 6.2.490). According to this definition, authorship implies voluntariness. Hence, creatures without wills cannot be authors at all. Yet, for the purposes of this chapter, I will use ‘author’ in a broader sense. I will, that is, use it to refer to any agent to whom the action in question is properly attributed, whether the action possesses – to borrow the terminology from Chapter 2 – rational spontaneity or merely agent spontaneity.8 After all, I believe that the problem of the authorship of voluntary creaturely actions has a close analogue in merely agent-spontaneous actions, even if the former is more urgent.9 Suppose, for instance, that a cat is chasing a mouse. According to Leibniz, animals cannot act voluntarily and do not have wills. Yet, he would undoubtedly still want to say that the action of chasing is properly attributed to the cat, and the action of running away is properly attributed to the mouse.10 It would go against Leibniz’s anti-occasionalist metaphysics to claim that these actions ought to be attributed to God alone: it is important to Leibniz that creatures are genuinely active and are not merely acted upon by God. Likewise, it would be exceedingly odd to have to say that, strictly speaking, both God and the cat are chasing the mouse and that both God and the mouse are running away. And when the cat eventually kills the mouse, it is even less appealing to say that this is God’s action as well. Hence, even for actions that merely possess agent spontaneity, it is important to be able to say that the action belongs to the creature rather than to God. In fact, the term ‘author’ could be extended to states that are spontaneous merely in the metaphysical sense. After all, Leibniz holds that even passions are, in an important sense, actions that are properly attributed to their subjects. When you start feeling pain, this change is, strictly speaking, something that you bring about. As mentioned earlier, Leibniz’s concurrentism makes it particularly difficult to claim that creatures are the sole authors of their actions. But why? We should not assume that the author of an action is always the only efficient cause of that 8

9

10

In fact, the main reason why Leibniz defines the term ‘author’ in terms of willing may be simply that Leibniz is focusing on divine authorship. His main goal in many passages is to argue that God is not the author of things that follow from his intellect, such as the eternal truths (see A 6.3.121/CP 41). Because anything that depends on God depends either on his will or on his intellect, one way to make this point is to define authorship in terms of willing. For creatures, there is a third option: changes in their states can also depend on nonrational appetitions. Of course, one cannot presuppose that the solution to one of these problems will also constitute a solution to the other. Yet, if I am correct, this will turn out to be the case, which is another reason to use the term ‘author’ in this broader sense. In the New Essays, Leibniz even draws the distinction between ideal activity and ideal passivity in terms of attributability: “we can . . . attribute to the substance itself, any change through which it comes closer to its own perfection; and . . . attribute to an outside cause . . . any change in which the reverse happens” (NE 211).

96

Attributability and Divine Concurrence

action. Thinking about it pre-theoretically (and setting aside Leibniz’s claim that creatures do not interact), there appear to be cases in which an action has multiple efficient causes but only one author. Take, for instance, the example of the car that I mentioned earlier: it seems that both the driver and the car are efficient causes of the car’s motion, but only the driver is the author of it. Or, perhaps more interestingly, suppose that a CEO asks her assistant to buy flowers for the CEO’s husband. In that example, the CEO is plausibly the author of the action of giving flowers to her husband, even though the assistant is an efficient co-cause of that action. Hence, God’s causal involvement in creaturely actions does not automatically make him a coauthor. Even though Leibniz owes us an account of attributability that assigns authorship in the correct way, it may seem unclear why concurrentism should be a particular problem. One factor that makes it especially difficult for Leibniz to solve the attributability problem is that he appears to endorse a version of concurrentism according to which the causal contributions of God and of the creature are quite similar. The following passage from Causa Dei brings this out: God’s concurrence “is special [specialis] because it is directed not only at the existence of the thing and its actions, but also at the mode and qualities of existing insofar as there is some perfection in these” (CD 12; similarly in “On God and Man,” G 3:30/LGR 289). By saying that divine concurrence is “special” and aims at the creature’s mode of existing, Leibniz is most plausibly rejecting the model of concurrence endorsed by the Jesuit Luis de Molina, among others. For Molina, God’s concurrence is general, that is, indifferently directed at various effects. It is only the creature’s contribution that determines it to a particular effect (see Freddoso 1988: 17). As an analogy, Molina repeatedly uses the sun’s indifferent concurrence with various biological processes: “God’s general concurrence is determined by the particular concurrence of secondary causes similarly to the way in which the influence of the sun, which is also universal, is determined by the influence of a human being so that a human being is produced and by the influence of a horse so that a horse is created” (Molina, Concordia, pt. 2, disp. 26, §11). For Molina, then, a strong asymmetry is built into divine concurrence: God’s contribution is general, while the creature’s contribution consists in determining or specifying the resulting action. Leibniz, on the other hand, is apparently denying that God’s contribution is general in this way.11 Divine concurrence is special: it aims at the particular qualities of the creaturely action, at least insofar as they contain 11

Apparently, one reason why Leibniz prefers special concurrence is his skepticism concerning general actions. See LC 5.17, where Leibniz insists on “every action being individual and not general, nor abstract from its circumstances, but always needing some particular way of being put in execution”; see also T 27.

Privation and Attributability

97

perfection.12 As a result, God’s contribution looks quite similar to the creature’s contribution: both aim at a specific outcome. This makes it difficult to explain why the resulting actions should not be attributed to God. If God and the creature are productive co-causes of the action in this apparently symmetrical way, why is only the creature the author of the action, and not God? What accounts for the asymmetry in attributability?13 2

Privation and Attributability

Before proposing my own solution for reconciling divine concurrence with creaturely spontaneity and authorship, it is instructive to consider some other interpretations of Leibniz’s concurrentism and the difficulties they face. Understanding these difficulties will pave the way for my interpretation. Hence, I will briefly discuss what I take to be the two main alternatives to my solution. We can call them ‘simple privation model’ and ‘sophisticated privation model.’ Both of these interpretations find some support in Leibniz’s writings, but neither of them is a satisfactory and complete solution to our two problems. Then, I advance my own solution, which I call the ‘final and formal causation model,’ and which solves both the attributability problem and the spontaneity problem. In several texts, Leibniz seems to propose the following solution to the attributability problem: the cooperation between God and creatures is asymmetrical and creatures are the authors of their sins because God produces only what is positive in the actions, while the creatures are responsible for the actions’ limitations or imperfections. In the Theodicy, after agreeing with Augustine that “evil is a privation of being, whereas the action of God tends to the positive” (T 29), Leibniz argues that “God is the cause of the perfection in the nature and in the actions of the creature, but the limitation of the receptivity of the creature is the cause of the defects there are in its action” (T 30). Hence, in some sense that will need to be spelled out further, the sinfulness of a sinful action is due to the creature, and whatever is good in the action is due to God. Only the creature, therefore, is properly the author of sin. Roughly the same account occurs in several other passages (e.g. T 377; CD 69; “On God and Man,” G 3:34/LGR 294; “On Freedom, Fate, and God’s Grace,” A 6.4.1605/LGR 262f.). 12 13

Later in this chapter, I will say more about what Leibniz might mean in CD 12 by “insofar as there is some perfection in these.” One might try to introduce an asymmetry by pointing out that God is the primary cause and the creature with whose action he is concurring is merely a secondary cause; they belong to different causal orders. It is true that this is an asymmetry, but it seems unsuitable for explaining why creatures are the sole authors of their actions. If anything, it appears to make God’s causality more fundamental and more important. See e.g. Aquinas, De potentia 3.7.

98

Attributability and Divine Concurrence

Leibniz’s favorite way to illustrate this solution is the analogy of the boat, which he uses in at least six different texts in the mature period.14 The analogy goes as follows: when a heavily laden boat floats down a river, one can say that “the speed comes from the river, but the retardation which restricts this speed comes from the load” (Ta, response 5, G 6:383/H 384). Or, as he puts it elsewhere, “the current . . . is the cause of the boat’s speed without being the cause of the limits of this speed” (T 30). The river’s contribution to the boat’s motion is supposed to be analogous to God’s contribution to the creaturely action, and the load’s contribution is analogous to the creature’s contribution: the river (or God) is the source of what is positive, while the boat’s load (or the creature) is the source of the limitation. On what I take to be the most straightforward reading, Leibniz appears to argue that anything that is real in a creaturely action is due to God, while the privation or limitation of that action is due to the creature (see ET, G 6:348/H 389ff.; Ta, response 5, G 6: 383/H 384; CD 69). I will call this reading the ‘simple privation model.’ On this model, we need not worry that God is the author of human sins because the sinfulness of the action is a mere privation and due to the creature alone. There are at least two ways to understand the simple privation model, but I think that neither of them ultimately works. The first is that the creature produces some aspects of the action, whereas God produces other aspects of the action. On this interpretation, which Robert Sleigh endorses, God produces what is real or perfect in the action while the creature produces the limitations of the action (1990a: 185). Let us call this the ‘divided effort interpretation.’15 The second way to understand the simple privation model is slightly different. Rather than saying that God and the creature produce different aspects of the action, this interpretation holds that the creature merely limits God’s action, or channels it in an imperfect way. This interpretation is captured quite well by Aquinas’s analogy of a limping leg: when an animal limps, he explains, “limping arises in a leg not insofar as it is moved by the locomotive power, but insofar as it fails, through its deficiency, to receive the influx of the locomotive power. Therefore, the limping is not caused by the animal’s locomotive power” (De malo, q3 a1 ad4, in CT). Just like the imperfect leg, on this interpretation, a creature receives the divine influx, which is not defective, in a deficient manner. This is another way to understand the claim that the limitation in the action is due to the creature while everything that is real is due to God. Let us call this the ‘defective reception interpretation.’ 14

15

These texts are CD 71; ET, G 6:349/H 390; “Commentary on Burnet” §56(d), DPG 137; T 30; Ta, response 5, G 6:383/H 384; “On God and Man,” G 3:34/LGR 294. In a text from the middle period, Leibniz uses a different analogy: throwing a feather with great force (“On Freedom, Fate, and God’s Grace,” A 6.4.1605/LGR 263). Whipple, who rejects this interpretation, calls it the “Divided Labour Model” (2015: 209).

Privation and Attributability

99

Can the simple privation model, interpreted in either of these two ways, work for Leibniz? Like other scholars who have written on this,16 I have several worries about the simple privation model, on either reading. One problem with the first way to understand privation – that is, the divided effort interpretation – is that it does not explain how we can ascribe good or neutral actions to creatures. In other words, it seems to imply that the only actions of which creatures are capable are sins or defects, because everything that is positive or real has to be attributed to God.17 Thus, while the solution may allow us to call creatures the authors of sin, it apparently does not allow us to call them the authors of anything else. Moreover, the claim that there is an aspect of the action – the negative aspect – of which the creature is the sole cause goes against one of the most fundamental tenets of concurrentism: the doctrine that creatures cannot produce anything on their own. Finally, and most importantly, the divided effort interpretation strikes me as extremely implausible because it assumes that a privation is a separate part of the action and can have a separate productive cause. Leibniz himself relentlessly pokes fun at this type of argument in several early writings. In a text from the early 1670s, for instance, Leibniz argues, “[t]his is as though someone were a cause of the number three and wanted to deny that he was a cause of its oddness” and that one could claim along the same lines that “a bad musician is only the cause of the violin bowings and drumbeats and not the resulting dissonance” (A 6.1.544/CP 23). In another text from the same period, he calls this maneuver “a manifest illusion” and “a subterfuge with which a reasonable person will never be satisfied” (A 6.3.150f./CP 111). The point of these criticisms appears to be the following: producing all the positive aspects of a defective effect just is producing the defective effect. The fact that the producer of the effect has not produced the defect directly does not exculpate her, nor does it make sense to claim that another agent is contributing the deficiency. Take the example of a three-legged dog: missing a leg is a privation. However, if God creates everything that is real or positive about that dog, he will have created a three-legged dog. It sounds absurd to say that because the absence of the fourth leg is not something real and positive, some other cause can produce this absence, and therefore God is not responsible for it. The same applies to defective or evil actions. If Leibniz in his later writings understands privation along the lines of the divided effort 16

17

One of these other scholars is Sukjae Lee (2004: 208f.). Even Robert Sleigh, who accepts this model, says that Leibniz would need to say more to convince him that this is compatible with Leibniz’s metaphysical account of creaturely actions (1990a: 185). It may be the case for Leibniz that even the best creaturely actions are limited and imperfect in some way (see e.g. T 377). Even if that is so, on the divided effort interpretation, we would be able to ascribe only the negative aspect of the action to the creature and the rest to God, which still seems problematic.

100

Attributability and Divine Concurrence

interpretation, he is not only committed to a wildly implausible view but also vulnerable to his own earlier criticism. In fact, Leibniz does say in at least one mature text that privations do not have efficient causes (T 20). Given these problems with the divided effort interpretation, it may seem tempting to reject it as naïve and take refuge in the defective reception interpretation. After all, it seems that the latter interpretation is immune from most of the criticisms just described, including the scathing remarks from Leibniz’s early writings. It may also seem like a more straightforward interpretation of Leibniz’s boat analogy, and of his claim that he wants to explain sin through the “limitation of the receptivity of the creature” (T 30; see also M 47). Yet, the defective reception interpretation also faces serious problems. First of all, the first criticism that I levelled at the divided effort interpretation applies here as well: on the defective reception interpretation, it is also difficult to say why virtuous or neutral actions can be ascribed to finite substances rather than to God, since finite substances can at most limit what God provides. Second, and more importantly, it does not fit very well with the Leibnizian doctrine that finite substances possess both active and passive forces (see Adams 1994: 96, Lee 2004: 208f., Sleigh 1990a: 185). If creatures only limit divine activity and everything that is real in the action comes from God, creatures are purely negative, and it seems that the relation between God and creatures is like that between active and passive forces. Yet, as seen in Chapter 1, the claim that finite substances themselves possess active force is crucial for Leibniz.18 Finally, the defective reception interpretation also seems to be at odds with Leibniz’s claim that divine concurrence is special. It seems to turn God’s contribution into a general contribution, which is then specified or limited by the creature. After all, the river analogy, when interpreted like this, suggests that the positive contribution is indifferent to the particular outcome: the river concurs in the same way with all things that are floating on it, regardless of their weight.19 Jeffrey McDonough, in passing, sketches a solution to the attributability problem that sounds very similar to the defective reception interpretation but that seems to extend it to non-sinful actions. McDonough claims that Leibniz agrees with late Scholastics on the authorship issue: the reason why the creature rather than God is the author of its actions is that some aspects of these actions are explained only by the creature’s causal powers. To illustrate this, 18

19

In fact, passive force is arguably just a limitation in active force; see Antognazza (2014) and Jorati (forthcoming b). Moreover, if the perfection of our actions were entirely due to God, it would be difficult to avoid occasionalism (see Scott 1998: 84) as well as substance monism, as Leibniz points out in a letter to Johann Christian Schulenburg (March 29, 1698, GM 7:239/ SLT 38). Perhaps the river’s contribution is not entirely general either because its water flows at a specific speed. Yet, what is important for current purposes is that it is not “special” in Leibniz’s sense. The river concurs in the same way with the movements of a floating feather and a heavy barge.

Privation and Attributability

101

McDonough mentions Molina’s claim that when God concurs with a fire, the fire’s causal powers explain why heat, rather than coldness, is produced (2007: 45).20 McDonough also provides a more up-to-date analogy: we can say that electricity concurs with a toaster’s toasting, and that the reason why the electricity results in toasting rather than blending is that the electricity is paired up with a toaster rather than a blender (ibid.). The main problem with this suggestion is that the Molinist solution presupposes general concurrence. According to Molina, God’s concurrence with the fire’s heating is not qualitatively different from his concurrence with, say, the water’s cooling. McDonough’s analogy suggests the same: the electricity flowing into the toaster is not qualitatively different from the electricity flowing into the blender. On that picture, it may indeed make sense to say that the specific nature of the action is due to the creature and not to God.21 Yet, it is not at all clear how that solution would work for someone who is committed to special concurrence.22 If God’s contribution is special – that is, already tailored to that particular effect – there seems to be perfect symmetry between God and the creature.23 Since concurrence is special for Leibniz, he will have to come up with a different way to account for the asymmetry in attributability; the defective reception interpretation is not available to him. 20 21

22

23

Molina says this in Concordia, pt. 2, disp. 26, §12; he explains how this solution exculpates God from human sins in, for instance, disp. 27, §3, disp. 32, §10, and disp. 32, §18. Even on that picture, however, solving the attributability problem for non-sinful actions is not trivial. When I plug in a toaster and a coffeemaker, for instance, it is indeed because of the toaster’s and the coffeemaker’s contribution that toast and coffee are produced. My contribution may be exactly the same in both cases. Yet, that does not mean that I am not the one who is making breakfast. Likewise, the mere fact that the specific nature of an action is due to the creature rather than to God does not imply that the creature is the author of the action and that God is not. That fact is completely consistent with instrumental causation, as illustrated by the toaster and coffeemaker example, and thus with divine authorship. McDonough mentions that divine concurrence is special for Leibniz (2007: 43) and acknowledges this even more explicitly with respect to conservation (2007: 50). Yet, McDonough does not say how, given Leibniz’s commitment to special concurrence, something like the Molinist solution to the attributability problem can work for Leibniz. Suárez, whom McDonough also mentions in this context, has a completely different solution: for him, concurrence is special (see MD 22.4.4/OCC 218; 22.4.8/OCC 220; 22.4.32/OCC 233), and therefore Molina’s solution is not available. Instead, he claims that God’s concurrence with unfree (or natural) causes is different from his concurrence with free causes. In the case of natural causes, divine concurrence limits the creature to one specific act and necessitates that act (MD 22.4.10/OCC 222; 22.4.21/OCC 227f.). In the case of free secondary causes, on the other hand, divine concurrence includes an implicit condition; it is in the power of the creature to act or refrain from acting, so the creature possesses liberty of exercise (MD 22.4.14/OCC 225). Moreover, he says that when God concurs with a free cause, he offers concurrence to more than one act and thus grants the creature liberty of specification (MD 22.4.21/OCC 227). This solution allows Suárez not only to ascribe freedom of indifference to creatures but also to say that God is not the author of free creaturely actions. Yet, it is rather implausible that this is Leibniz’s solution. After all, Leibniz does not agree with Suárez that freedom requires indifference (see Chapter 5). What I take to be Leibniz’s solution is, however, somewhat similar to Suárez’s account of divine concurrence with unfree causes (see footnote 30).

102

Attributability and Divine Concurrence

So far, I have argued that the simple privation model does not solve the attributability problem in a satisfactory way. Luckily, there is a more promising privation model, which has been put forward by Tad Schmaltz (2014) and which I call the ‘sophisticated privation model.’ For reasons similar to the ones we examined earlier, Schmaltz argues that Leibniz’s view cannot be that creatures cause only the limitations of their actions. Instead, he claims, creatures must positively contribute to the production of their actions and be efficient co-causes: they possess active as well as passive forces (2014: 147f.). Why, then, does Leibniz repeatedly say that the deficiency of a sinful action comes from the creature while God contributes only the positive aspects? Schmaltz gives a plausible answer, and it is one of the building blocks for my own solution. It is true, he points out, that the deficiency of evil actions comes only from the creature and not from God. Yet, the reason for this is not that the creature somehow produces the defective aspect of the action while God produces the perfect aspect; that cannot be how deficient causation works for Leibniz. Nor does the creature merely channel divine activity in a deficient manner. Instead, Leibniz most plausibly views deficient causes as “nonproductive reasons that explain the limitations present in evil” (2014: 150). In the case of sinful actions, the deficient cause is the lack of adequacy among creaturely perceptions, which present something as good that in fact is not good (2014: 150f.). Hence, when Leibniz says that the limitations of the action stem from the creature, he merely means that an imperfection in the creature’s perceptions explains why the action is evil; it would be a mistake to think of deficient causes as producing some reality. In short, Schmaltz argues that Leibnizian creatures are both efficient and deficient causes of their actions, and that God concurs only with what the creatures efficiently cause, that is, with the positive aspects of the action (2014: 151). As a result, God is not the author of human sins. Excellent support for this interpretation comes from “On God and Man,” where Leibniz tells us that “privation arises from the limitation of creatures, which means [facit] that their knowledge does not extend to all things, and that their will, which should be directed at the highest good, i.e. God, instead remains fixed on inferior goods” (G 3:34/LGR 294; see also “Rationale of the Catholic Faith,” A 6.4.2322/LGR 78). The limitation of human knowledge, then, explains some of the bad choices that human beings make. But the way in which our limitations explain the badness of our choices is not by efficiently causing the lack of goodness. In the Theodicy, Leibniz even explicitly says that privations do not have efficient causes; instead, they have merely deficient causes: “the formal aspect [le formel] of evil does not have any efficient [cause], for it consists in privation, . . . that is, in that which the efficient cause does not bring about. That is why the Scholastics have the custom of calling the cause of evil deficient” (T 20).

The Final and Formal Causation Model

103

The sophisticated privation model provides very appealing answers to the worries raised earlier about the simple privation model. In particular, it shows how a creature and God can be efficient co-causes of an action while the creature alone is responsible for the deficiency of the action: the imperfect perceptions of the creature can be the reason for the privation. This is compatible with special divine concurrence, because the positive aspects of the action that God coproduces fully specify the action. A defect is merely a privation and as such does not require a productive cause. And, as we saw in the example of the three-legged dog, producing all the positive aspects just is producing that particular defective effect. The deficient cause of a sin, on this interpretation, is not a productive cause but merely the reason for the deficiency of the effect. The inadequate perceptions of the creature are the reason why the creature is producing an evil action. They are also – and this is crucial – the reason why God concurs with the creature in this particular way. We can thus say that on this model the deficient cause is explanatorily prior both to the creature’s efficient contribution to the action and to God’s efficient contribution. God contributes in this particular way because of the inadequate prior perceptions of the creature. As a result, the sophisticated privation model can explain why evil actions are properly attributed to creatures and why God’s concurrence does not make him a coauthor of evil. Schmaltz’s sophisticated privation model, then, goes a long way toward solving the attributability problem. Yet, this model cannot be the entire solution, because it is not clear how it helps us with the spontaneity problem. Moreover, it is not even a complete solution to the attributability problem: as mentioned earlier, a complete solution should explain the authorship of not only sinful but also non-sinful actions. The sophisticated privation model thus shares one weakness of the other models: it does not straightforwardly apply to non-sinful creaturely actions. 3

The Final and Formal Causation Model

Let us finally turn to my own solution, which is structurally similar to the sophisticated privation model. Yet, by invoking formal and final causation rather than deficient causation, it applies to all creaturely actions – not only to sinful ones. It thus constitutes a complete solution to the attributability problem. As we will see, it also opens up a way to solve the spontaneity problem. Moreover, it is not subject to any of the objections to the simple privation model that I discussed earlier. To understand the final and formal causation model, it helps to first consider Sukjae Lee’s claim that creatures are the formal and final causes of their actions, while God is the sole productive cause (see Lee 2004, 2006). Even though I reject the second component of Lee’s interpretation, mainly for textual

104

Attributability and Divine Concurrence

reasons, the first component is crucial for my interpretation.24 Lee explains his claim that creatures are the formal and final causes of their modifications as follows: “the nature of a creature cooperates in the production of the modification by setting bounds to God’s production in the sense that it provides the reasons that determine and demand which modifications God is to produce among infinite possibilities” (2004: 221). Creatures provide the final causes for their actions by demanding (or striving for) those actions, and the formal causes by determining (or specifying) those specific actions.25 On this interpretation, there is an asymmetry between God’s contribution and the creature’s contribution: God is the only productive or efficient cause of the creature’s state; the creature, or its nature, is the final and formal cause of its state. The creature determines and demands this state and hence determines the specific characteristics of God’s efficient contribution. While Lee does not explicitly use his interpretation of concurrence to address the attributability problem,26 this seems like a promising approach to it. Unlike the various privation models, Lee’s suggestion that the creature, or its nature, rationally determines and demands God’s contribution appears to give us a plausible way to ascribe even non-evil actions to creatures. Like deficient causes, formal and final causes are nonproductive and can be construed, the way Lee does, as a type of “rational determination” or reason-giving (2004: 221). The final cause is also traditionally considered explanatorily prior to the efficient cause, as seen in Chapter 3. Thus, if we can make a good case for the claim that creatures, rather than God, contribute the formal and final causes of the action – similar to Schmaltz’s case for claiming that creatures alone are the deficient causes of the action – we have an elegant and non-arbitrary way to attribute actions to creatures and not to God. This solution clearly needs to be spelled out in much more detail. Yet, it is worth pausing to note how antecedently plausible it is that final and formal causation can help us attribute actions to agents. By way of illustration, take the example I briefly mentioned earlier in this chapter. When you drive somewhere, you and your car efficiently co-cause the car’s motion. Yet, we typically describe the situation by saying that it is your action, not the car’s, and that you are merely employing the car as an instrument. One rather satisfactory explanation of why we describe the situation in this way invokes final and formal causation. Because the action aims at your ends rather than at the car’s, 24

25 26

Lee is, of course, not the only commentator who attributes final and formal causation to monads. Yet, he is the only one, as far as I know, who argues that these types of causation are central for understanding Leibniz’s doctrine of divine concurrence. Elsewhere, Lee describes formal causation as the “formal determination of the particular states” of the creature (2006: 444). He does address the question whether the creature can truly be said to act (2004: 223f.), but it is not at all clear whether he is addressing the attributability problem or merely the potential objection that creatures do not seem active at all.

The Final and Formal Causation Model

105

and because you determine where you are going, which route you are taking, and so on, you are the author of the action; the car is merely an instrument. A similar explanation seems to work in the example of the CEO who asks her assistant to buy flowers. There are excellent reasons to believe that Leibniz views creatures, rather than God, as contributing the final and formal causes to their actions. Final and formal causes play crucial roles in Leibniz’s account of creaturely agency, as seen in Chapters 1 and 3. Because Leibnizian substantial forms both specify all of the substance’s actions and direct the substance toward those actions, they provide both the formal and the final causes of such actions. Yet, it is not enough to say simply that the actions of finite substances are explained by their substantial forms or by their internal end-directedness. In order to show that the ends of actions that are jointly produced by a creature and God truly belong to the creature, rather than God, we arguably need to be able to claim that God has not imposed those ends on that creature. And Leibniz can indeed claim that: as seen in Chapter 1, all of a creature’s actions and strivings are already specified by the creature’s essence or concept that God finds in his intellect. When God subsequently creates that substance, he merely actualizes its essence. He does not impose ends and forms on a creature but rather actualizes a possible creature whose concept already specifies all of its strivings and dispositions for action. The creature’s concept, then, is prior in the order of explanation both to God’s choice to create this creature and, a fortiori, to God’s act of concurring with the creature’s actions. Leibniz says this explicitly and even uses it to explain why God is not the author of sin in a text from the early 1680s: “the idea or concept of a creature (involving possible general divine decrees) [is] prior to the decree of the divine will (actual and particular to the creature itself), as it [i.e., the creature’s concept] is established in his understanding.” This priority, he goes on to say, allows us to see “how God produces every reality in evil actions and yet is not the cause of evil” (“On Freedom and Grace,” A 6.4.1459). Even though this passage employs the priority of creaturely concepts to argue only that God is not the author of evil, the same line of argument can easily be extended to non-evil actions. All of a creature’s actions, whether evil or not, are specified by the concept of the creature. Hence, this strategy allows us to say that non-evil natural actions of creatures should be ascribed not to God but rather to the creature. They are included in the creature’s essence or concept that in turn is prior to any of God’s choices or actions and explains why God efficiently co-causes that specific action. In fact, this fits well with one of the ways in which Leibniz argues that God is not the author of sin: “God . . . produces [a creature’s] essence before its accidents, its nature before its operations, following the priority of their nature, and in the sense of the priority of reason [in signo anteriore rationis]. From this,

106

Attributability and Divine Concurrence

one sees how the creature can be the true cause of sin” (T 390). What Leibniz seems to suggest here is that since God produces the creature’s nature or essence prior to its actions – taking ‘prior’ in a nontemporal sense, of course – it is possible for the creature’s nature to ground its actions in such a way that the creature, and not God, is their author. The most plausible way to spell this out is to say that the creature’s nature can supply the most fundamental reasons, or the final and formal cause, for the action that God and the creature jointly produce.27 Just as the creature’s nature is prior to its actions in the order of explanation, the final and formal cause are prior to the efficient cause; hence, God can concur with the efficient cause without also concurring with, or cocontributing, the final and formal cause. On this interpretation, God’s concurrence is special because the specification is prior to God’s act; God does not perform a general act that the creature then determines. The account under consideration is even able to explain the difference between God’s creating the world, of which he is the author,28 and God’s coproducing a creature’s action, of which the creature is the author. God’s creation of the world is a miraculous divine act that is not explained by the essences of creatures; creaturely essences, after all, do not include existence. The actions of creatures, on the other hand, are explained by their essences or concepts. When God concurs with these actions, he does so because the creatures’ concepts – or their substantial forms, which God created in exact correspondence to the concepts – specify them and because these are the actions that the creatures’ natures demand (see Lee 2004: 224). By creating a substance, God commits himself, as it were, to letting it act according to its own nature.29 Hence, we can say that the creature’s concept and the corresponding substantial form contribute the formal cause to the effect jointly produced by God and the creature. Likewise, the final cause of the action is also coming from the creature. After all, the creature specifies its future actions in virtue of its internal end-directedness; dictating future states through one’s nature introduces a type of teleology. As a result, neither the formal nor the final cause of these actions comes from God; instead, God merely contributes to the production of an effect that the creature specifies and strives for, and he does so because the creature specifies and strives for it.30 27

28 29 30

This strikes me as more plausible than Robert Adams’s suggestion, also based on this passage, that God produces “the creature’s nature ‘operating’ and thus producing its affections and actions” (1994: 97). Adams’s suggestion appears to remain at the level of efficient causation, and does not seem like a fully satisfactory solution to the attributability problem or to the spontaneity problem. Leibniz says this, for instance, in NE 165, M 83; 90; letter to Des Bosses, May 29, 1716, LDB 369. See “On God and Man”: “God . . . does not determine any one to sin but only decrees never to prevent someone who is going to sin” (G 3:33/LGR 293). This solution is in fact somewhat similar to the way Suárez explains divine concurrence with natural (that is, unfree) causes: “it is because the secondary cause has power of a certain sort that God has decided to give it a certain species of concurrence. And in this respect . . . the secondary

The Final and Formal Causation Model

107

Someone might object that the essences of finite substances in God’s intellect are the formal causes even of God’s creation of these substances and that their perfection might in a sense be the final cause. That would undermine my claim that God is the author of the created world but not of the actions of creatures. Yet, I do not think this objection ultimately succeeds. While Leibniz may sometimes be happy to talk of creaturely essences as the final and formal causes of their creation, there are important differences between that and the way in which creatures supply the final and formal causes of their actions. First of all, as argued in Chapter 1, we must distinguish between the concepts and the natures or substantial forms of creatures. While it may be the concept (or essence) that provides the ultimate formal cause even of the actions of a creature (see A 6.4.1458), the final cause of creaturely actions seems to be located in the substantial form, not the concept. As mentioned in Chapter 3, I find it implausible to say that possibles in the divine intellect literally strive for anything. Only actual things can strive for something in any genuine sense. Second, in the creation of substances, all of the efficient causality is coming from God; the creature cannot contribute anything. As Leibniz writes to Jaquelot, “the soul certainly did not contribute to its original constitution, but it will contribute to the actions which arise from it in the course of time” (G 3:472/WF 181).31 This is in fact one reason to think, pace Lee, that being an efficient cause is a necessary condition for authorship; final and formal causation is insufficient. Moreover, the final cause of God’s creation of any particular creature is, in a way, the goodness of the world as a whole or God’s glory, rather than the ends or the perfection of that individual creature. Based on Leibniz’s doctrine that the nature and the strivings of a creature are the reason why God co-causes the actions that follow from them, we now have a good way of solving the attributability problem. A creature is the sole author of an action if and only if the creature contributes the final and formal cause and efficiently co-causes the action.32 At the metaphysical level, that is the case for all nonmiraculous changes in created substances, so creatures are the sole authors of these changes despite the fact that God concurs “specially” with what they efficiently cause. This is of course a very broad construal of ‘authorship.’ Interestingly, if we take seriously Leibniz’s notion of ideal influence, we can also introduce a more demanding sense of ‘authorship,’ which requires

31

32

cause is said to determine God’s concurrence as regards its specific nature” (MD 22.4.5/OCC 218f.). See also T 391, where Leibniz says that creatures do not concur with their conservation – that is, with God’s action of keeping them in existence – but that they do concur with, or contribute to, their actions. This makes sense because conservation is a continued creation for Leibniz; the creature contributes neither to its creation nor to its conservation. See John Carriero, who in a different context argues that ascribing internal, natural ends to creatures allows Leibniz to view them as genuine agents, or originators of activity, rather than mere instruments of God (2008: 119, 123).

108

Attributability and Divine Concurrence

agent spontaneity and agent teleology. In this sense, as seen in Chapter 3, imperfections in one creature occur for the sake of perfections in another creature. We can therefore say that insofar as the more perfect creature supplies the ideal end for the imperfection, this creature is the author of that imperfection. For instance, we can say that the mosquito is the author of your pain. In fact, this is exactly how Leibniz talks about the attribution of actions and passions in the New Essays (NE 211). There are some passages in the Theodicy that further support my interpretation, even though they focus on the authorship of evil actions rather than on authorship generally. In these passages, Leibniz claims that the forms or ideas of creatures are the origins of evil and that because God is not the author of these forms he is not the author of evil. For instance, he argues that “the cause of evil . . . must be sought in the ideal nature of the creature, insofar as this nature is contained in the eternal truths which are in the understanding of God, independently of his will” (T 20; see also “Rationale of the Catholic Faith,” A 6.4.2322/LGR 78). Because this creaturely essence specifies not only the evil actions of that creature but also the neutral or good ones, it makes sense to extend that solution to creaturely actions generally and use it to argue that God is the author of none of these actions. In fact, later in the same section of the Theodicy, Leibniz acknowledges that the divine understanding contains “the primitive form of good” as well as “the origin of evil” and that the region of eternal truths is therefore “the ideal cause of evil, so to speak, as well as of good” (T 20). Hence, insofar as all actions of a creature are explained by the creaturely essence, of which God is not the author because he merely found it in his intellect, we can attribute these actions to the creature rather than to God, even though God concurs with them. Later in the Theodicy, Leibniz makes the same point in slightly different terminology: “Evil comes . . . from the forms themselves, but abstracted, that is, from the ideas, which God has not produced by an act of his will” (T 335). There are also passages suggesting that Leibniz views deficient causation as a type of formal causation, which in turn makes formal causation an obvious candidate for the attempt to expand the sophisticated privation model. Just as a creature can be the nonproductive reason for a deficiency in an effect that it causes together with God, a creature can be a nonproductive reason for the form of the effect more generally. One such passage is from a letter to Isaac Jaquelot: “[t]he form of sin is a voluntary privation of the relevant perfection, and that voluntary privation comes only from us” (G 6:568/WF 197). Similarly, Leibniz points out in several other texts that God is not the formal cause – or is not responsible for the formal aspect – of sin but that creatures are, in virtue of their limitations (CD 68; “On God and Man,” G 3:34/LGR 294; T 30; “On Freedom, Fate, and God’s Grace,” A 6.4.1605/LGR 262f.).

The Final and Formal Causation Model

109

Another passage that supports my solution to the attributability problem is from a letter to the Marquis de l’Hôpital. There, Leibniz writes that in the ordinary course of nature, we can explain all events in terms of creaturely natures, without reference to any new divine operations, because created substances spontaneously produce all of their actions and passions. He goes on to say that “God intended to concur [s’estant proposé de . . . concourir] with this only in conformity with . . . the primitive nature of the thing” (July 12, 1695, A 3.6.451).33 This passage suggests that the nature of the creature determines or specifies God’s act of concurrence. The letter is hence additional evidence for my claim that when we focus on final and formal causation, there is an asymmetry between God’s and the creature’s contribution, and this makes it plausible to attribute actions to the creature and not to God. More textual evidence for the final and formal causation model can be found in a text from the early 1680s called “On Freedom and Grace,” from which I already quoted earlier. Leibniz there discusses the relation between the causality of God, who is the primary cause, and the causality of creatures, which are secondary causes. With regard to that great question whether the second cause determines the first, or the first the second, it is to be answered that the first is determined by the second taken ideally, that is, the idea of the second as perceived in the divine intellect determines the first will. But the second taken actually is determined by the first, i.e. each takes its own entity [entitas] from it. (A 6.4.1458)

In this passage, Leibniz says explicitly that creatures determine the divine will through their concepts or ideas, which reside in the divine intellect. This kind of determination, I think, is best understood as formal causation. The ideas of creatures provide the formal causes for some of the things that God produces or coproduces, namely finite substances and their actions. Note that Leibniz’s claim that secondary causes determine God “taken ideally” (sumta idealiter) cannot mean that this influence is exactly like the ideal influence of one creature on another. We saw in Chapter 2 that one creature influences another ideally insofar as a perfection of the former was God’s reason for creating a substance like the latter that possesses a corresponding imperfection. We can therefore explain the imperfection in the latter creature by referencing the perfection in the former creature. Yet, this is not a complete explanation. To explain a creaturely imperfection fully, we must invoke not only the other creature that has a corresponding perfection but also God, who is the fundamental reason for the harmony among all created 33

WF contains a translation of this text, but it unfortunately obscures the point about concurrence I take Leibniz to be making, since it uses ‘sustain’ for the French term concourir (WF 57). Leibniz clearly uses the verb concourir in other French texts to refer to concurrence; see e.g. DM 30.

110

Attributability and Divine Concurrence

substances. Yet, when Leibniz talks about the ideal influence of a creature on God, he is describing precisely that fundamental level because he is referring to the fact that the idea of the creature in the divine intellect determines the divine will. The idea of a creature has a genuine, direct influence on God’s choice, whereas one creature does not have any genuine, direct influence on another creature. Hence, the fact that the passage under consideration describes the influence of creatures on God as ideal does not mean that this influence is not fundamentally real; it merely means that it is the influence of an idea in God’s mind. The passage also claims that “taken actually,” God determines the creature; but this does not mean that, at bottom, the determination goes only in this direction. Rather, it must mean that creatures owe their existence to God – after all, Leibniz glosses this phrase by saying that each creature “takes its own entity” from God. So far, we have seen how the final and formal causation model can solve the attributability problem: because the creatures are, at bottom, the providers of the formal and final causes of their actions, God’s concurrence in the production of these actions does not compromise creaturely authorship. I will now show how the final and formal causation model can likewise solve the spontaneity problem. Recall the worry about spontaneity: if Leibniz thinks that interaction among creatures would undermine their substancehood because substances have to be perfectly independent, why would divine concurrence not have similar consequences? And how can creatures be free, or possess what I call ‘rational spontaneity,’ if their actions are co-caused by God? It seems ad hoc to solve this problem by simply making an exception for God’s causal influence. Interestingly, invoking final and formal causation gives Leibniz a way to reconcile creaturely spontaneity with divine concurrence. Remember that one important reason why Leibniz denies interaction among finite substances is that substances must be able to make intelligible all of their states. Leibniz’s spontaneity thesis, as seen in Chapter 2, has a crucial explanatory dimension. For instance, it cannot be the case that in order to explain why you are currently feeling pain, we must invoke the mosquito that has just bitten you. Instead, we must in principle be able to understand your pain just in terms of your own nature and previous states; otherwise you would not be a substance. As we saw in Chapter 2, Leibniz admits that there is a sense in which we can explain your pain in terms of the mosquito’s bite. Yet, that is not the explanation at the metaphysical ground floor. If created things truly did interact with one another, however, a full explanation of the state of the patient would require invoking the agent. But this is relevantly different from the way in which God causally acts on creatures when concurring with their actions. Even though divine concurrence is a necessary condition for the production of the action, God is

The Final and Formal Causation Model

111

not involved in the final and formal causation of the action. Hence, God’s contribution does not ultimately explain why the action is the way it is.34 Consequently, Leibniz can solve the spontaneity problem as follows. Because of divine concurrence, creatures are never spontaneous in the sense that they are the sole efficient causes of their states. Yet, they are spontaneous in the sense that (i) they are the only finite efficient causes of their actions, and (ii) they are sufficient explanations for the qualities of their actions. That is, they are spontaneous in the sense of being the sole providers of the formal and final causes of their actions, in addition to being efficient co-causes. We could say that while they do not have complete productive spontaneity, they possess complete explanatory spontaneity. In virtue of providing the final and formal cause of the action, they single-handedly explain the properties of the action. As a result, Leibniz’s doctrine that divine concurrence is compatible with creaturely substancehood is not ad hoc. After all, it is plausible that Leibniz’s necessary conditions for substancehood include explanatory spontaneity and causal activity, but not complete productive spontaneity. This fits well with the reasons that Leibniz typically cites for rejecting occasionalism and interactionism, which we discussed in Chapters 1 and 2. If this is correct, Leibniz’s concurrentism does not undermine the substancehood of creatures because creatures are causally active and possess explanatory spontaneity. Even though what I have called ‘explanatory spontaneity’ may seem contrived, the term describes something that is of central importance to Leibniz: the notion distinguishes nonmiraculous from miraculous events in the created world. After all, as Leibniz repeatedly stresses, a miraculous event is an event that goes beyond the natures of created substances, or that is not explainable in terms of these natures (e.g. “New System,” G 4:483/AG 143). Natural events, on the other hand, are intelligible in virtue of the natures of finite substances. This, according to Leibniz, is an enormously crucial difference. As a matter of fact, Leibniz sometimes connects the ability of a thing to render its states intelligible to its substancehood. In a letter to Pierre Bayle, for instance, he claims that unless creatures possess states “from which it can be seen that there will be a change of modifications, . . . God would have produced nothing, and there would be no substances other than God” (G 4:568/WF 122). Elsewhere, he stresses the importance of activity: the creatures described by occasionalists would not be substances because they do not act (ONI 15). Hence, there is good evidence that the requirements for substancehood include explanatory spontaneity and causal activity, but not complete productive spontaneity. Both 34

Of course, God’s contribution is part of the explanation for the occurrence of the action; without divine concurrence, the action would not take place. Yet, to explain why the action has the properties that it does – for instance, why the action is a perception of my arm going up rather than of my arm waving – we do not need to invoke God. The creature’s nature is sufficient for explaining all of the properties of the action.

112

Attributability and Divine Concurrence

explanatory spontaneity and causal activity are compatible with divine concurrence, if the final and formal causation model is correct. These insights about explanatory spontaneity do not require a significant revision of the schema from Chapter 2. That schema – which distinguishes between metaphysical, agent, and rational spontaneity – is still adequate. After all, my descriptions of those three types of spontaneity followed Leibniz in making an explicit exception for divine concurrence: spontaneity is the absence of external, non-divine influences. What we have learned now is that we need this exception only for efficient causation, but not for formal and final causation. In other words, as long as we are talking about what I have called the ‘explanatory dimension’ of spontaneity, we do not need to make any exceptions. That is important because complete independence in this sense – even from God – is a necessary condition for substancehood. When we are talking about the efficient-causal or productive dimension of spontaneity, on the other hand, we need to make an exception for divine concurrence. Yet, as just seen, this exception is not ad hoc. Distinguishing carefully between the explanatory and productive dimension of spontaneity also allows us to understand how creatures can be free despite divine concurrence. A finite mind possesses rational spontaneity in its free actions not in the sense that nothing except its rational faculties produces the actions,35 but in the sense that nothing but its will and its rational judgments about the good provides the final and formal cause of the actions. This means that to the extent that an action is free, its ends have to be ends that the agent pursues deliberately and voluntarily, rather than instinctively or unconsciously. Likewise, to the extent that an action is free, its properties have to be specified by the agent’s rational faculties. God’s concurrence with the production of the action does not compromise these two features of free agency.36 4

Conclusion

As we saw in this chapter, Leibniz has excellent solutions to the attributability problem and the spontaneity problem. Even though God concurs with every creaturely action by special concurrence, Leibniz has principled ways of attributing a jointly produced action only to the creature. Likewise, he can claim that despite God’s intimate involvement, the creature possesses the kind of spontaneity that is required for substancehood and, in some cases, the more demanding kind of spontaneity that is required for freedom. There ultimately is no tension between Leibniz’s concurrentism on the one hand and creaturely 35 36

This way of defining rational spontaneity is dubious anyway, since on my interpretation only substances can be productive causes (see Chapter 1). For my response to the objection that concurrence undermines divine explanatory spontaneity, see von Bodelschwingh (Jorati) (2011: 293f.).

Conclusion

113

spontaneity, freedom, and authorship on the other. The attributability problem, which seems intractable when focusing on efficient causality, dissipates when we invoke formal and final causation. Similarly, what is most important to Leibniz about the spontaneity and freedom of creaturely substances is fully compatible with divine concurrence.

5

Freedom

In previous chapters we explored the self-determination and end-directedness that is at work in different types of agency, as well as the attributability of actions. This puts us in a position to look more closely at one particularly important type of agency: free agency.1 Here again, Leibniz employs his notions of spontaneity and teleology to craft a subtle and powerful response to a long-standing philosophical problem. We already saw in Chapters 2 and 3 that, for Leibniz, free actions are instances of what I call ‘rational spontaneity’ and ‘rational teleology.’ When agents act freely – or more precisely, to the extent that they act freely – their actions are determined solely by the rational aspects of their nature. These rational aspects comprise the agents’ will and intellect, or more specifically, the agents’ rational inclinations and intellectual judgments. In fact, we saw that actions are free if and only if they exhibit these most demanding types of spontaneity and teleology. Through this chapter and the next two, I will explore some of the unique advantages of this characterization of free agency. Some of these advantages depend crucially on Leibniz’s monadological metaphysics; others are independent of it and can hence be adopted by philosophers who subscribe to different ontologies. Leibniz’s views on freedom of the will have received a considerable amount of attention in the secondary literature. The same is true for Leibniz’s views on contingency, which are closely connected to his theory of freedom. Yet, this literature almost never gets at Leibniz’s most central motivations for insisting that there is free and contingent agency nor at what I take to be the biggest strengths of Leibniz’s account. In fact, some interpreters conclude that Leibniz’s attempts to make room for contingency and avoid necessitarianism are ultimately failures.2 My interpretation, in contrast, reveals that Leibniz’s aim is not what people usually assume it is and that he succeeds in achieving this aim. Building on the work of a few other scholars, I will argue that Leibniz’s ultimate goal was not avoiding necessitarianism simpliciter but rather 1 2

A few paragraphs in this chapter are modified versions of Jorati (2017). Russell talks of “the hopeless inconsequence involved in Leibniz’s doctrine of contingency, with its tiresome progeny of final causes, liberty, and optimism” (1903: 183). Other examples include Carriero (1991: 83, 1996: 63f.), Riley (1996: 85–88), and more tentatively, Sleigh (1999: 276).

114

Freedom

115

avoiding the kind of necessitarianism according to which everything is necessitated through value-neutral principles. By insisting that at least some things are determined in a value-sensitive way, Leibniz is insisting on a kind of modality and freedom that turns on goodness-directed teleology. Because of the novel interpretation of teleology and spontaneity that I have laid out in previous chapters, I am able to pinpoint Leibniz’s concerns and describe his solution far more accurately than other commentators. In the course of arguing for this reading, I will also show that Leibniz’s theory of freedom has some surprising advantages over rival theories; these advantages have not been sufficiently appreciated. Moreover, I will show that a number of Leibniz’s strategies to make room for contingency are at bottom the same. Leibniz’s theory of freedom is what we would today describe as a version of agent-causal compatibilism. In other words, Leibniz believes that freedom is compatible with determinism, and he holds that free actions are caused by agents rather than by events internal or external to agents. This should not be surprising, given the previous chapters. As seen in Chapter 1, agent (or substance) causation is the only type of efficient causation acknowledged by Leibniz. Finite substances are efficient causes of all natural changes in their own states; God is the efficient cause of any nonnatural events and an efficient co-cause of all natural changes in the created world. At the same time, all of these changes are clearly deterministic: there is a sufficient reason for any action or change, and agents always follow their strongest inclination. As a result, it is already clear that Leibniz’s theory of freedom must be a version of agent-causal compatibilism. The main goal of this chapter, then, is to elucidate Leibniz’s agent-causal compatibilism. I will show that his theory of freedom has two notable and often overlooked features. First, in virtue of Leibniz’s views about spontaneity, the creation of finite substances, and agent causation, his theory of freedom circumvents one of the notorious shortcomings of other compatibilist theories. This shortcoming is the concession, which classical compatibilists are forced to make, that free actions are parts of causal chains that originate outside of the agent and, in fact, begin before the agent comes into existence.3 Second, Leibniz’s views on teleology allow him to respond to the worry that determinism turns free agents into mere automata. I will show in this and the next two chapters that the role played by teleology in Leibniz’s theory of freedom places free actions in an entirely different category than the activity of unfree substances. This allows Leibniz to say that our actions are not necessitated in any problematic sense, that we can exert control over our actions in a way that 3

Of course, classical compatibilists can also say that these causal chains go back infinitely and do not have beginnings. Yet, that view has the same shortcoming: free actions are parts of deterministic causal chains that already existed before the agent came into existence.

116

Freedom

unfree creatures cannot, and that we can be held morally responsible for our deeds. In discussing Leibniz’s views on freedom of the will, it is convenient to start with his definition of freedom in terms of three necessary and jointly sufficient conditions. These three conditions are described, among other places, in the following passage from the Theodicy: [F]reedom . . . consists in intelligence, which includes a distinct knowledge of the object of deliberation, in spontaneity, by which we determine ourselves, and in contingency, that is, in the exclusion of logical or metaphysical necessity . . . The free substance determines itself by itself, following the motive of the good recognized by the understanding, which inclines it without necessitating it; and all the conditions of freedom are contained in these few words. (T 288; see also CD 20)

Intelligence, spontaneity, and contingency are the three components of Leibniz’s account of freedom, according to this passage. The passage even explains what, roughly, these three terms are supposed to mean. In fact, we learn quite a lot about Leibniz’s views on freedom from this short passage. We learn that only deliberate actions – that is, actions based on deliberation – are free, that freedom requires some kind of spontaneity, that free actions aim at something that the agent takes to be good, and that freedom is compatible with inclination but not with logical or metaphysical necessitation. The passage also raises a number of questions, however. What, for instance, is the difference between inclining and necessitating? And what kind of spontaneity does Leibniz have in mind? I will discuss the three conditions one by one to answer these questions and get a fuller picture of Leibniz’s account of freedom. Because I have already explored many relevant aspects of spontaneity and intelligence in previous chapters, my discussion of those two conditions will be relatively short. I will spend most of the present chapter on the third condition: contingency. Perhaps surprisingly, it will turn out that contingency is intimately connected to teleology. In fact, when Leibniz insists that free actions must be contingent, I think we should understand him as insisting that free actions must be instances of rational teleology. My analysis will reveal that Leibniz’s theory of freedom is significantly more compelling and sophisticated than widely thought and that teleology plays a crucial role in it. Before proceeding, it is important to note that Leibniz’s definition captures not only the freedom of rational creatures but also God’s freedom. After all, Leibniz holds that divine and human freedom are of the same kind and differ only in degree (see e.g. “On Freedom,” A 6.4.1452; NE 175; 181; T 337; “Conversation about Freedom and Fate,” Gr 480f./SLT 98). In fact, this similarity is an important component of Leibniz’s doctrine that humans are made in the divine image and that they ought to imitate God in their actions

Spontaneity

117

(e.g. response to Fardella, February 5, 1691, A 2.2.382; LC 5.7; CD 98; “On Freedom and Spontaneity,” G 7:111/SLT 95). God is free in the same sense in which human beings are free: his actions are based on his knowledge concerning the good,4 they are spontaneous, and they are contingent. Yet, God’s freedom is perfect because he has complete knowledge of the good and possesses perfect rational spontaneity. In contrast, our freedom is never perfect because we are always subject to passions. 1

Spontaneity

The first of Leibniz’s three conditions for freedom that I would like to discuss is spontaneity. We already saw in Chapter 2 that Leibniz ties freedom very closely to rational spontaneity. That some kind of self-determination should be a necessary condition for freedom also strikes me as immensely compelling. In fact, characterizing free agency in terms of self-determination is widespread among Leibniz’s contemporaries and predecessors.5 The basic idea, which is common to a wide range of theories of freedom, is that at least certain kinds of determination by people or things that are external to the agent (or to the agent’s “real self”)6 undermine this agent’s liberty. What is required for free agency, conversely, is that the agent is in some sense determining her actions herself. As one might suspect, there are many different ways of spelling out the relevant sense of self-determination. For instance, philosophers disagree about whether external things can indirectly determine an agent’s actions under some conditions without thereby rendering these actions unfree. Compatibilists traditionally allow such indirect determination because on most versions of determinism all actions are parts of deterministic causal chains that originate outside of the agent, at a time before the agent came into existence. If freedom is compatible with that form of determinism, such causal chains do not generally undermine free agency. Incompatibilists, on the other hand, typically object that if an action is completely determined by something external to the agent or prior to the agent’s existence, then the action cannot be free.7 Instead, 4 5 6

7

One difference is that divine deliberation is not discursive. Yet, he can still be said to weigh reasons and to compare different options. See footnote 34. For a helpful discussion of the role of self-determination in early modern theories of freedom, see Chappell (2005). Of course, what this “real self” is varies from theory to theory. As already seen in Chapter 2, Leibniz appears to hold that an agent’s nonrational inclinations are not part of her real self but are in some sense external. For helpful discussions of “real self” views from a contemporary perspective, see Wolf (1990: 23ff.) and Watson (2004: 270ff.). Gary Watson puts it as follows: “Historically . . . the main inspiration for libertarianism is an understanding of self-determination that is incompatible with determinism; roughly, determinism means determination by something other than the self, and hence heteronomy” (2004: 163). It will become clear that Leibniz would disagree with the second half of Watson’s statement: for Leibniz, determinism is completely compatible with self-determination or autonomy.

118

Freedom

they often insist that free agency requires the capacity for beginning a new causal chain, that is, for acting in a way that is not necessitated by earlier events or external factors. Because of his idiosyncratic metaphysical commitments, Leibniz can achieve something that many have thought impossible: without jettisoning his determinism, he can agree with incompatibilists that even indirect external determination is incompatible with freedom. This means that Leibniz endorses a much more stringent form of self-determinism than any other compatibilist I have so far encountered. Leibniz himself sometimes advertises his theory of freedom in those terms: “there is in free actions a perfect spontaneity, beyond all that has been conceived up to now” (Tp, G 6:37/H 61). What allows him to pull off such a remarkable feat is his acceptance of three doctrines that we have already encountered in earlier chapters: (a) his spontaneity thesis, (b) his claim that all efficient causation is agent causation, and (c) his account of the creation of rational agents. The first two of these doctrines are closely connected. Created monads do not strictly speaking interact with one another; instead, each monad is an efficient cause of all of the changes occurring within it. This means that actions are not parts of causal chains originating outside of the agent (at least, setting aside God’s influence): all creaturely actions, whether free or not, are agent-caused by the agent herself, and no other finite substance causally determines this action.8 Even though Leibniz’s endorsement of agent causation and spontaneity means that human actions are self-determined in a very robust sense, it does not undermine his commitment to determinism, since all actions of an agent are completely determined by the agent’s nature. Of course, what I have said so far leaves out one other potential source of external determination: God. Yet, Leibniz has an answer there as well, and this is where his views about creation become crucial. We already saw in Chapter 4 that God’s ordinary concurrence does not jeopardize the kind of spontaneity that is arguably crucial for authorship and substancehood. Finite substances themselves provide the final and formal causes of their actions – that is, they specify the properties of these actions and are internally directed toward them. The explanation of the action is within them, and God merely concurs with the production of whatever action the finite substance specifies and strives for. Moreover, God cannot be said to impose natures or natural inclinations on his creatures. Instead, God finds the essences of these creatures – essences that 8

Endorsing agent causation instead of the efficacious perception or efficacious appetition view (as described in Chapter 1) has advantages, even though all three theories can rule out external determination. One advantage is that on the latter two interpretations, your voluntary actions today are parts of deterministic causal chains going back to a time when you were not yet rational. Even though these chains are not external to you, that seems more problematic than the agent-causal interpretation, on which your voluntary actions depend directly on you, rather than on a chain of states that started when you were first created.

Spontaneity

119

already specify everything – in his intellect and then actualizes them without adding anything to them. If I have an inclination toward impatience, it is not because God somehow gave it to me but simply because that is part of who I am. God gave it to me only in the sense that he discovered my essence in his intellect and decided to actualize it; he made it the case that there be an individual with this inclination, but he did not literally give this inclination to some individual.9 While this account of creation may be problematic in other ways, it is extremely convenient for Leibniz’s theory of freedom. It allows him to say that the agent’s nature determines her actions, but her nature is not in turn determined by anything else, not even by God. Being able to describe finite substances as so radically independent of each other and even – with respect to their natures and actions – from God’s agency is a huge advantage. It allows Leibniz to alleviate a common worry about compatibilism without compromising his thoroughgoing determinism. Spontaneity, then, is clearly a central component of Leibniz’s account of freedom. Given my discussion of spontaneity in Chapter 2, it is natural to ask the following: which of the three types of spontaneity might Leibniz be invoking in his definitions of freedom? Of course, as already seen, all free actions are instances of the most demanding type of spontaneity: rational spontaneity. An action is free to the extent that it is not determined by anything external to the rational aspect of the agent’s nature. As a result, it is tempting to conclude that in his definitions of freedom, Leibniz is invoking rational spontaneity rather than metaphysical or agent spontaneity. Yet, there are a few problems with that interpretation. First of all, as we saw in Chapter 2, Leibniz does not appear to distinguish explicitly among the three types, and I am in fact not aware of any passages in which he uses the term ‘spontaneity,’ without further qualification, for what I call ‘rational spontaneity.’ When he defines or describes ‘spontaneity’ he apparently always – including in the passage from T 288 in which he lists the three conditions for freedom – identifies it with self-determination broadly construed. Hence, it would be highly misleading if Leibniz were using this term in a narrower sense without explicitly telling his readers. Second, if rational spontaneity were one of the three conditions for freedom, the intelligence condition would be entirely redundant. In fact, even the contingency condition might be redundant in that case, as we will see in Section 3. An action is free if and only if it is an instance of rational spontaneity, after all, which means that there are no additional, distinct necessary conditions. 9

It is for this reason that Leibniz sometimes says we cannot rightfully complain to God for our shortcomings. See e.g. a short dialogue from 1689/90: “You insist: ‘You can inquire why God has not given you more powers.’ I respond: ‘If he had done that, you would not be you, because then God would have produced not you, but another creature’” (A 6.4.1639; similarly in T 416).

120

Freedom

For these reasons, it appears more plausible to understand Leibniz’s definition as invoking either metaphysical or agent spontaneity. As a matter of fact, it might initially seem most plausible that he is invoking metaphysical spontaneity because that is what he typically appears to mean when using the term ‘spontaneity’ without qualification. One problem with this, as seen in Chapter 2, is the following: metaphysical spontaneity is ubiquitous. Even the least perfect monadic actions are spontaneous in this broad sense. As a result, making this type of spontaneity a condition for freedom may seem utterly uninformative; it does not tell us anything that is unique to free agency. By way of analogy, suppose that upon asking what it means for a car to be fuel efficient, you are told that one of the necessary conditions is having an engine. Since all cars have engines, this would be extremely unhelpful. Likewise, since all actions are metaphysically spontaneous, it may seem that the notion of metaphysical spontaneity cannot be part of an informative explanation of free agency. Hence, the best way of interpreting Leibniz’s usage of ‘spontaneity’ in his definition of freedom may well be as referring to agent spontaneity.10 In the end, however, I suspect that it does not truly matter which of the three interpretive options we choose, because it does not ultimately affect our understanding of Leibniz’s theory of freedom. As a matter of fact, I think that the spontaneity condition makes sense and is helpful for describing free agency on any of the three interpretations. Saying that freedom requires metaphysical spontaneity is useful – despite the worry just raised – because it rules out cases in which an action is directly determined, in the strict metaphysical sense, by something external. Leibniz, of course, thinks that God is the only external substance capable of this kind of violent determination and that God exercises this power only very rarely, if ever.11 If he does, however, the resulting change is ipso facto not a free action. Thus, even if we interpret Leibniz as invoking only metaphysical spontaneity in his definition, he is saying something substantive and informative because he is ruling out God’s interference. Moreover, the definition understood in this way can be used to show that in rival metaphysical systems in which all actions are ultimately caused by something external, there can be no freedom. Agent spontaneity, on the other hand, is helpful for understanding freedom because it rules out the ideal influence of external things, or the determination by internal imperfections. Finally, it is obviously helpful to say that freedom requires rational spontaneity, since an action is free if and only if it is spontaneous in this strictest sense. 10 11

See Murray (1995: 80), who draws a similar conclusion. As seen in Chapter 4, God’s concurrence with human actions does not amount to a violent determination. Yet, genuine miracles – that is, changes that cannot be explained in terms of the natures of created things – would be a violent determination. For more on the possibility of miracles, see Chapter 3, footnote 14.

Contingency

2

121

Intelligence

Let us turn next to another one of Leibniz’s three conditions for freedom: intelligence. From what we have so far learned about the Leibnizian theory of freedom, it is already quite clear what role the intellect plays for him and why he would list intelligence – or “a distinct knowledge of the object of deliberation” as well as an intellectual recognition of “the motive of the good” (T 288) – as a necessary condition for freedom. In order to be free, the agent must act in accordance with her judgment regarding the best course of action, and this judgment in turn must be the outcome of deliberation.12 Nonrational agents cannot act freely, and agents who are generally capable of rational action are unfree when their actions are not determined by the outcome of their deliberation concerning the good. Intelligence is hence immensely important for freedom. In fact, directly after listing his three conditions, Leibniz refers to intelligence as “the soul of freedom” and to the remaining two conditions as freedom’s “body and foundation” (T 288). However, this cannot mean that intelligence is the only thing that genuinely matters for freedom.13 Judging that some course of action is best is not sufficient for freedom, after all, since one might lack the inclination to act accordingly. Leibniz tells us earlier in the Theodicy not only that “when there is no judgment in the agent there is no freedom” but also that “if we had a judgment not accompanied by any inclination to act, our soul would be an understanding without will” (T 34). And without a will, or the inclination to act in accordance with our intellectual judgments, there would be no freedom.14 Even though the will is determined by the intellect, the will is an important faculty and should not be neglected. Both intellect and will – or, in other words, both intelligence and spontaneity – are crucial for free agency. 3

Contingency

The third and last condition for freedom is contingency. Some of Leibniz’s early writings sound surprisingly close to Spinozism. Yet, from the 1680s onward, he is clearly eager to distance himself from necessitarianism (see e.g. “On Freedom,” A 6.4.1653/L 263; NE 73) and makes contingency central 12

13 14

See T 34, where Leibniz explicitly says that “judgment [jugement]” is a necessary condition for freedom. For additional evidence that free actions must be based on deliberation, see CD 20; NE 175f.; “Conversation about Freedom and Fate,” Gr 479f./SLT 96f.; a letter to Jaquelot, G 6:569/ WF 198; Marginal Notes on Temmik, in Mugnai (1992: 163); “On God and Man,” G 3:36/ LGR 297. Some interpreters do appear to draw that conclusion. See e.g. Greenberg (2005: 227) and Carlin (2000: 142n33). There are interesting connections here with Gary Watson’s view; he argues that “[t]he free agent has the capacity to translate his values into action” (2004: 26).

122

Freedom

to his theory of freedom. In a text from the early 1680s, he claims that being exempt from necessity is “the root of freedom” (A 6.4.1456), and he also stresses the centrality of contingency in many passages from the mature period (e.g. CD 20f.; 101f.; LC 5.8f.; NE 175). The contingency condition is by far the most controversial of Leibniz’s three conditions for freedom. The reasons for this controversy are not difficult to find. Even though Leibniz repeatedly insists that free actions must be contingent, it is not at all clear that he is entitled to the claim that any actions – creaturely or divine – are genuinely contingent. His adherence to the Principle of Sufficient Reason and his doctrine of complete concepts appear to commit Leibniz to necessitarianism. As a result, some interpreters think that his attempts to distance himself from Spinozism do not ultimately succeed. Some argue, for instance, that what Leibniz calls ‘contingency’ is something that even Spinoza would be happy to acknowledge or that Leibniz’s strategies have internal flaws (e.g. Newlands 2010, Garrett 1990: 39f., Pickup 2014: 521, Carriero 1991: 83, 1996: 63f.). In fact, even during his lifetime, Leibniz was attacked for the perceived fatalist consequences of his doctrine of complete concepts – for instance, by Antoine Arnauld. I am going to take a fresh look at Leibniz’s account of the contingency of free actions and propose an interpretation according to which Leibniz does succeed in avoiding Spinozistic necessitarianism. In order to do so, I find it helpful to approach Leibniz’s views on necessity from a different direction than other interpreters, who typically frame the issue in terms of Leibniz’s logical or truththeoretic commitments (e.g. Sleigh 1990a: 48ff., Baxter 2000: 192, Garrett 1990).15 Initially bracketing Leibniz’s complete concept theory and his definition of truth, I will start by examining the interplay between teleology and what Leibniz calls ‘contingency.’ I aim to argue that even though Leibniz is committed to a form of necessitarianism, he succeeds admirably in his stated goal of distancing himself from Spinoza’s necessitarianism. After all, on my view, the type of necessitarianism he is eager to avoid and that he attributes to Spinoza is the type according to which everything is governed by what Leibniz sometimes calls “blind” or “brute” necessity. Leibniz does manage to avoid this kind of necessity, I contend, by giving teleology a central role in his theory of freedom – that is, by tying freedom to rational teleology.16 It is not Leibniz’s goal, if I am correct, to make room for contingency as it is commonly understood. In fact, I interpret Leibniz as rejecting the need for contingency in the traditional sense. What Leibniz calls ‘contingency’ is a type of necessity, but it is not a type of necessity that Spinoza would be happy to acknowledge. After all, Spinoza 15

16

There are some notable exceptions, including Mogens Lærke, who argues explicitly against approaching the topic of contingency from an almost exclusively logical standpoint (2007: 40); see also Nachtomy (2007: 148). See Chapter 3 for my discussion of rational teleology.

Contingency

123

rejects the kind of teleology that is at the heart of Leibnizian contingency. That free actions are necessary only in this way is crucial for understanding Leibniz’s views on freedom and the special moral status of free agents. It also enables Leibniz to say that there is a sense in which we have the ability to do otherwise than we freely choose to. I am by no means the first to suggest that teleology is the most promising way for Leibniz to avoid Spinozism.17 Yet, nobody, to the best of my knowledge, has so far explored it in much detail, or examined how precisely teleology relates to what Leibniz calls ‘contingency.’ That is what I will do here. I first explore the ways in which Leibniz employs teleology to make room for contingency. I will argue that his various strategies are, at bottom, identical and that they succeed in drawing a legitimate distinction. After that, I investigate the relation between Leibniz’s views on necessity and Spinoza’s necessitarianism. At the end of the section, I briefly revisit the question of whether Leibniz’s logical and truth-theoretic commitments undermine his more straightforwardly metaphysical strategies for distancing himself from Spinozism. I will argue that they do not and that the metaphysical strategies have logical correlates. One caveat is that because the present chapter is about freedom, my discussion of contingency will focus almost exclusively on the modal status of free actions. Leibniz sometimes calls unfree creaturely actions ‘contingent’ as well (e.g. T 34), but, I think, they are contingent in a different sense. In fact, their contingency might be derivative; it might derive from the contingency of the laws that govern them or from the contingency of God’s free decrees.18 They might be contingent because God could have actualized a different world. Alternatively, their contingency might be a consequence of the teleology – metaphysical teleology and, in some cases, agent teleology – that these actions possess. I will largely bracket that issue here, however, and concentrate on free actions, which are the paradigm of contingency for Leibniz. 3.1

Teleology and Contingency

Let us take a look at the ways in which Leibniz connects contingency to teleology. For the time being, I will follow Leibniz in using the term ‘contingency’ for the special metaphysical status of the actions under discussion. In the next subsection, I will argue that even though teleology may not give Leibniz what most philosophers would recognize as genuine contingency, it 17

18

My interpretation builds, in particular, on work by Jalabert, Adams, and Lærke (see Jalabert 1960: 218f., Adams 1996: 110, Lærke 2007: 61, but see also Wilson 1999: 437, Garber 2009: 230ff., Schmid 2011: 354f.). Leibniz appears to suggest this in “On Freedom and Grace” (early 1680s, A 6.4.1456). For a helpful discussion of the contingency of the laws of nature, see Garber (2009: 244ff.).

124

Freedom

nevertheless enables Leibniz to distance himself meaningfully from Spinoza; it makes room for a type of necessity not found in Spinoza’s system. Moreover, this type of necessity gives Leibniz everything that a compatibilist might want. When Leibniz describes contingency, particularly in the mature period, he employs a number of teleological terms. More specifically, he frequently connects contingency very closely to the kind of teleology that is present in rational, free agency – that is, to rational teleology. We can identify at least four major Leibnizian strategies that tie contingency to rational teleology: a. He says that that free actions are contingent because motives or reasons do not necessitate but merely incline free agents. I will call this the ‘inclination versus necessitation strategy.’ b. Leibniz also describes “moral necessity” – which is, roughly, the persuasive force of reasons in a wise, free mind – as a kind of contingency, contrasting it with absolute or metaphysical necessity. I will call this the ‘moral necessity strategy.’ c. He furthermore claims that free actions are contingent because free agents have the power to do otherwise, even if they are determined to will what they judge to be best. It is crucial, he stresses, to distinguish carefully between what an agent wills to do and what she has the power to do. I will call this the ‘will versus power strategy.’ d. Finally, Leibniz also argues that free actions are contingent because they originate in a choice between different options, none of which are made impossible by failing to be chosen. Following Robert Adams (1994: 20), I will call this the ‘reality of choice strategy.’ I interpret each of these four strategies as connecting contingency to teleology – to rational teleology, in particular. After all, they all concern the influence of motives on rational minds or the free choices of rational agents who act for the sake of the good. Because each of the four strategies is helpful for understanding Leibniz’s claim that freedom requires contingency, I will briefly explore all of them.19 Even though I initially discuss them separately, it will turn out that they are closely connected – indeed, they are different ways of describing the same thing.20 Inclination versus Necessitation Let us start with the inclination versus necessitation strategy. When arguing that free actions are contingent, Leibniz frequently employs this strategy. For instance, he uses it in the definition of freedom from the Theodicy discussed at the beginning of this chapter 19 20

Unfortunately, I can provide only a rough sketch, leaving many important questions unanswered. In some passages, Leibniz even combines them; see e.g. LC 5.9.

Contingency

125

(T 288).21 One particularly explicit text is from Leibniz’s final letter to Clarke, where he claims that “good, either true or apparent – in a word, the motive – inclines without necessitating, that is, without imposing an absolute necessity” (LC 5.8). This passage is helpful because it seems to apply to both human and divine free actions: it applies to cases in which an agent is inclined by merely apparent goods and to cases in which an agent is inclined by the true good. An even more crucial feature of this passage is that Leibniz tells us explicitly what he means by saying that motives incline without necessitating: motives do not impose an absolute necessity (similarly in NE 199 and T 132). This shows that the inclination versus necessitation strategy is closely linked, or even identical, to the moral necessity strategy that I will discuss next. Leibniz connects these two strategies in another important passage: If by ‘necessity’ we understand a man’s being inevitably determined . . . it is certain that every free act would be necessary; but we must distinguish what is necessary from what is contingent though determined. . . . Geometrical and metaphysical “followings” necessitate, but physical and moral ones incline without necessitating. (NE 178; similarly in a letter to Coste, December 19, 1707, G 3:402/AG 194f.; T 367)

Here, it is clear that the distinction between necessitation and inclination is not supposed to mark the distinction between inevitable determination and the lack thereof. Instead, it is supposed to mark the difference between geometrical or metaphysical necessity, on the one hand, and physical as well as moral necessity, on the other. In some passages, Leibniz uses the term ‘force’ instead of ‘necessitate,’ and in others, he uses ‘determine’ instead of ‘incline.’ Elsewhere in the New Essays, for instance, Leibniz’s spokesperson claims that “a thinking being . . . is only inclined and not forced by considerations of good and bad” (NE 177; similarly in NE 178), and in the Theodicy, Leibniz insists that “[a] very clear knowledge of the best determines the will; but this is not properly speaking necessity” (T 310).22 I take such passages to be making the same point as the ones in which Leibniz contrasts inclination with necessitation. Force and necessitation are incompatible with freedom, Leibniz appears to be saying, while the inclination or determination to do what one judges best is perfectly consistent with freedom. The terms ‘inclination’ and ‘determination’ are interchangeable in this context because Leibniz thinks that we are always determined to follow the strongest inclination – that is, that our inclinations determine our choices (see e.g. NE 199; “Conversation about Fate and Freedom,” Gr 479/SLT 96f.; a letter 21 22

In addition to the passages I discuss in this section, see COE 14; letter to Jaquelot, G 3:468/WF 179; T 43; 45; NE 175. See also T 367, where Leibniz argues that contingency is compatible with inclinations, determination, certainty, moral necessity, and being constrained to choose the best; he contrasts it with necessity – in particular, metaphysical necessity.

126

Freedom

to Coste, December 19, 1707, G 3:401f./AG 194). This is not a problem, according to the inclination versus necessitation strategy, because this infallible inclination or determination does not amount to an absolute or metaphysical necessity. Moral Necessity As just seen, Leibniz closely links the inclination versus necessitation strategy to the moral necessity strategy: saying that motives or the good incline a free agent without necessitating her means that this agent is necessitated only morally, rather than metaphysically or absolutely. In other words, both strategies consist in the claim that what undermines freedom is necessitation in the strict sense – that is, absolute or metaphysical necessitation – rather than moral necessitation, or determination by motives and by the good. As Leibniz puts it in a text from the early 1680s, “the root of freedom consists in this: that a mind does not choose by reasons of necessity but by reasons of goodness, true or apparent, by which it is inclined” (A 6.4.1456). It is important to note that even though Leibniz is happy to use the term ‘moral necessity’ for the kind of determination present in free actions, he sometimes stresses that moral necessity is a type of contingency. For instance, Leibniz notes that “one can say, in a certain sense, that it is necessary . . . that God . . . choose the best [and] that man follow the side which after all strikes him the most. But this necessity is not opposed to contingency” (T 282; see also T 367). This willingness to describe free actions as subject to a type of necessity and also as contingent will become crucial later. Let us take a look at the various types of necessity acknowledged by Leibniz. First, he distinguishes hypothetical from absolute necessity. An instance of hypothetical necessity is merely necessary on some hypothesis. For instance, something is hypothetically necessary when it is necessary only on the hypothesis that God has resolved to do it or foreseen it (e.g. T 37; 53; RH 3). Another particularly important example of hypothetical necessity is that an agent must act in a certain way on the hypothesis that she follows the strongest inclination (e.g. T 53). An instance of absolute necessity, on the other hand, is necessary independently of any hypotheses; it is necessary tout court. Mathematical truths, for instance, are absolutely necessary, as is God’s existence. Whatever is not absolutely but merely hypothetically necessary, Leibniz sometimes says, is contingent (T 37; 52). In addition to absolute and hypothetical necessity, Leibniz also draws a distinction between metaphysical, physical, and moral necessity. Something is metaphysically necessary when its opposite implies a contradiction; metaphysical necessity is thus founded on the Principle of Contradiction (e.g. LC 5.10). For instance, ‘A=A’ is metaphysically necessary because ‘A≠A’ is a contradiction. In fact, in several passages, Leibniz equates metaphysical necessity with logical, geometrical, arithmetical, and “blind” necessity (e.g. T 282; 346; 349;

Contingency

127

351; LC 5.4). Because its necessity is not based on some hypothesis, something that is metaphysically necessary is also absolutely necessary, and it is thus incompatible with freedom and contingency (e.g. Tp, G 6:37/H 61). An event is physically necessary for Leibniz, on the other hand, if its occurrence is determined by natural laws (FR 20; see also comments on Toland, August 8, 1701, Toland II, appendix: 66/LGR 215). He says, for instance, that “physical necessity . . . produces the order of nature and consists in the rules of motion and in some other general laws which it pleased God to give to things when giving them existence” (FR 2).23 Thus, it is physically necessary for a stone to fall, since it must fall according to the laws of nature. A stone’s rising or hovering in midair unsupported, in contrast, is physically impossible. Yet – and this is crucial – it does not imply a contradiction either. In fact, God can bring about events that go against the laws of nature, such as the levitation of a stone (see e.g. T 207; FR 3). Therefore, the stone’s falling is physically necessary, though it is not absolutely or metaphysically necessary. It is also hypothetically necessary because it happens on the hypothesis that the laws of nature require it and that nature runs its normal course.24 As we will see later, physical necessity plausibly governs not only the phenomena but also unfree monadic changes. Finally, an action is morally necessary if its occurrence is determined by what Leibniz sometimes calls the ‘Principle of Wisdom and Goodness’ (T 174);25 such actions are determined by what the agent’s intellect judges to be the greatest good.26 Moral necessity, then, “comes from the free choice of wisdom with respect to final causes” (T 349). For example, it is morally necessary for God to choose the best. Likewise, it is morally impossible for the blessed to sin or for “a wise and serious magistrate, who has not lost his 23 24

25

26

If these “other general laws” include the laws governing the choices of finite minds, free creaturely actions are both physically and morally necessary. The question whether (and if so, how) the contingency of events like the stone’s falling differs from the contingency of free actions is interesting and important, but I will not address it in this chapter. I cannot here discuss the fascinating and complex history of the term ‘moral necessity.’ Instead, I focus on how Leibniz uses and defines it. For helpful discussions of the usage of ‘moral necessity’ among late medieval and early modern Jesuits, see Murray (2004), Knebel (1991, 2000), and Anfray (2011: 57ff.). See also Anfray (2011: 55), who understands moral necessity in a similar way. Leibniz also uses the term ‘moral necessity’ in a different sense in a few texts, namely in the sense of ‘morally obligatory’ (see e.g. A 6.1.301; G 3:386). Some interpreters think that this is Leibniz’s understanding of moral necessity generally (e.g. Adams 1994: 22, Nachtomy 2007: 150, Rescher 2001: 146). Yet, that goes against some textual evidence. Leibniz appears to think it is morally necessary for the damned to sin, for instance (T 282). Adams in fact acknowledges this in a later article (2005: 181, see also Murray 2004: 2). More generally, Leibniz defines moral necessity as the determination by one’s judgments about the good, without specifying that these judgments have to be about the moral good or cannot be misguided. See e.g. LC 5.4. That also counts against Sleigh’s interpretation, which ties moral necessity to a type of final causation involving the agent’s motives and moral character (2009: 270).

128

Freedom

mind, to publicly perform some reckless action, . . . for instance, to run in the streets entirely naked in order to make people laugh” (T 282). Given their character and wisdom, the blessed and the serious magistrate cannot freely perform evil or reckless actions. Because of the way Leibniz defines freedom, all free actions – and only free actions – are morally necessary (see e.g. T 237). Like the physical laws, the Principle of Wisdom and Goodness does not necessitate absolutely or metaphysically. For instance, Leibniz sometimes claims that it is metaphysically possible for God to create a sub-optimal world, even though it is certain, and morally necessary, that he does not (see e.g. T 234). Moreover, like physical necessity, moral necessity is only hypothetical. Just as a physically necessary event must occur on the hypothesis that certain conditions obtain and that certain laws of nature are operating, a morally necessary action must occur on the hypothesis that it seems best to the agent and that she is following the Principle of Wisdom and Goodness (see e.g. T 53). The different types of necessity and their characteristics can be summed up as follows: Table 5.1 Types of Necessity

Type of Necessity Metaphysical Physical Moral

Absolute or Hypothetical?

Contingent?

Law Governing It

Absolute

No

Hypothetical

Yes

Principle of Contradiction Laws of nature Principle of Wisdom and Goodness

We are now in a position to understand Leibniz’s moral necessity strategy: Free actions, he contends, are not absolutely or metaphysically necessary but merely morally necessary. They are necessitated not by the Principle of Contradiction but merely by the Principle of Wisdom and Goodness.27 Or, in other words, they are necessitated only by the goodness or value that the agent sees in this course of action, combined with her tendency to pursue what she views as best. The moral necessity strategy, then, enables Leibniz to say that free actions are contingent in one sense – namely in the metaphysical sense – but not in another sense, because they are necessitated morally.28 27

28

Here I agree with Kenneth Seeskin, who says, “while my will is always disposed to choose in accordance with the best, it is logically possible that it might do otherwise . . . From the standpoint of logic alone, the will is free to choose anything” (1977: 92). See also Nachtomy (2007: 151f.). Rutherford, interestingly, thinks that viewing Leibnizian substances as dynamic entities helps us understand the sense in which changes in their states are contingent (that is, not metaphysically

Contingency

129

Will versus Power I turn next to the will versus power strategy: the strategy that draws a distinction between what an agent freely chooses to do, which is determined in part by the Principle of Goodness, and what the agent has the power to do. What God has the power to do is determined solely by the Principle of Contradiction, since he is omnipotent. From what we have already seen, it is clear that the Principle of Contradiction is associated with absolute or metaphysical necessity. If God’s choice were governed only by the Principle of Contradiction – that is, if God were determined to create everything that is not contradictory, regardless of its goodness – there would not be any contingency in God’s actions. Leibniz thinks it is crucial that this is not the case: God’s choice is determined by what is best and not just by what is possible. Even though God is omnipotent and hence has the power to do whatever is metaphysically possible, he chooses to do only what is best. In order to understand what God chooses, we need to invoke the Principle of Goodness, not just the Principle of Contradiction. As a result, Leibniz claims, non-actual possible worlds fail to exist only because God follows the Principle of Goodness; they are metaphysically possible, which means that God has the power to create them. Hence, there is a sense in which God could have done otherwise: God had the power to create something different. He refrained from creating something different not because he lacked the power, but because he lacked the will. One place in which Leibniz uses the will versus power strategy very explicitly is his last letter to Clarke. He accuses Clarke of confounding “what God will not do with what he cannot do” (LC 5.73), or “the will of God with his power” (5.76), and argues against this conflation that “God can produce everything that is possible or whatever does not imply a contradiction, but he wills only to produce what is the best among things possible” (5.76). This means, Leibniz claims, that “God is not . . . a necessary agent in producing creatures” (5.77). In the Theodicy, he also invokes the will versus power strategy several times. For instance, he stresses that “[p]ower and will are different faculties, and their objects also are different; one confounds them when one says that God can do only that which he wills. On the contrary, among various possibles, he wills only that which he finds the best. For all the possibles are regarded as objects of power” (T 171). Similarly, he calls it an abuse of terms to claim “that one cannot do a thing, simply because one does not will it” (T 228). The idea appears to be the following: there are things that an agent does not do because she is not able to, and there are things that an agent does not do because she does not want to. God, for instance, does not create four-sided triangles because he necessary). As Rutherford puts it: “Leibniz is entitled to think of modal change as contingent. From the point of view of dynamics, no logical, metaphysical, or conceptual necessity determines that a monad should pass from state n to state n + 1” (2005: 165).

130

Freedom

cannot, and he does not create a nonoptimal world because he does not want to. Conflating these two things means conflating an important distinction.29 Even though Leibniz employs this strategy mainly in the context of God’s actions, it can also be applied to the actions of rational creatures. After all, it is plausible that free creatures have the power to perform actions they do not want to perform and, as a result, will never perform. For instance, the “wise and serious magistrate” in Leibniz’s example has the power to run through the streets naked (T 282).30 Yet, he does not want to do that and hence will not do it. In at least one passage, Leibniz in fact applies something like the will versus power strategy to human beings: he says that Adam had “the power of not sinning” even though he willed to sin (T 369; similarly in letter to Jaquelot, April 28, 1704, G 3:472/WF 182). The will versus power strategy strikes me as convincing. One might wonder, however, whether Leibniz ultimately has the resources for drawing this prima facie plausible distinction. I think that he does, and that they are in fact the very resources we have already encountered. Here is one way to draw the distinction. In a case in which God can do something but does not do it because he does not want to, we can say that it is only God’s goodness, or the Principle of Wisdom and Goodness, that is preventing God from performing this action.31 There is no logical or metaphysical obstacle to God’s doing it. The only “obstacle” – if we can call it that – is that this action is not sufficiently good. In other words, the will versus power strategy comes down to the same thing as the moral necessity strategy: it introduces a distinction between two ways in which something can be necessitated or determined. Something can be necessitated either by the Principle of Goodness or by some value-neutral principle, such as the Principle of Contradiction or the “blind” laws that govern unfree creaturely actions. In the Theodicy, Leibniz even explicitly connects the moral necessity strategy with the will versus power strategy: it is true, he says there, that God was determined to act in the way that he did. Nevertheless, we can say that “the thing might have gone otherwise.” After all, “metaphysically speaking, he could have chosen or done what was not best; but he could not have done this morally speaking” (T 234). In other words, there is one sense – the metaphysical sense – in which God had the power to do otherwise, and another sense – the moral one – in which God could not have acted otherwise because 29

30

31

Gary Watson is in agreement with Leibniz here: “the ordinary notion of ability is in some way relative to attributions of desire or will. It is simply a misuse of ‘able to’ to say: ‘S is unable to leave the room because he doesn’t want to (or has no reason to)’” (2004: 175). What finite agents have the power to do must presumably be spelled out in terms of their general capacities: a human agent lacks the power to fly or to jump over a building but possesses the power to do many other things. See Adams, who also mentions the distinction between what God cannot do merely because of his goodness and what God cannot do for some other reason (1994: 49).

Contingency

131

his goodness determines him always to choose the best. The laws of logic alone do not dictate this choice; the laws of goodness play a central role. God, as a perfectly rational and benevolent agent, is subject both to the demands of logic and to the demands of goodness. Logic constrains God by ruling out selfcontradictory options, but it still leaves many different options that are on a par, logically speaking.32 Goodness, in contrast, determines God to choose the best of these options. The Reality of Choice The last strategy I would like to consider is the reality of choice strategy.33 The idea is roughly this. For Leibniz, choices must be real: in order for there to be a choice, there must be multiple options that are possible and within the agent’s power. Volitions are always preceded by deliberation,34 according to Leibniz, so that there are always different options under consideration. If there were only one possible course of action, there would be no genuine deliberation and, hence, no freedom.35 As Leibniz writes to Gerhard Wolter Molanus in 1699, “[w]hen there are several paths, there is freedom to choose . . . But if one found oneself in a narrow street, between two high walls, there would only be one possible path, and that represents necessity” (A 1.17.611; similarly in CD 21; T 367; a letter to Bayle, 1702, G 3:59). This allows Leibniz to claim that free actions are contingent in the sense that at least one alternative to what was actually chosen is possible. Leibniz says, for instance, that the Principle of Wisdom and Goodness prompts God to create the best “without necessitating him, for it does not render impossible” the options that are not best (T 228). The reality of choice strategy is hence closely related to the will versus power strategy: when applied to free divine choices, both strategies rely crucially on the fact that there are metaphysically possible essences or worlds that God did not choose to actualize. They fail to exist not because God lacks the power to actualize them but because he does not want to actualize them, or because it is not fitting, good, or wise to do so.36 Yet, the fact 32

33 34

35

36

Ohad Nachtomy expresses this point as follows: “pure logic cannot decide between possibilities; rather, logic can fix possibilities according to the principle of contradiction.” As a result, the choice between possibilities must be based on “the principle of sufficient reason and a modality of picking the best” (2007: 151f.). For other discussions of this strategy, see Adams (1994: 20), Lærke (2007: 60), and Lagerlund and Myrdal (2006). In God’s case, of course, there is no discursive deliberation and, more generally, no temporal priority. Yet, God’s choices also depend on a comparative judgment about the goodness of the available options. Hence, there is a priority of reason and something closely analogous to deliberation (see e.g. RH 12). Even though one can presumably deliberate as long as there are at least two epistemically possible courses of action, Leibniz appears committed to a stronger claim: at least two options must be metaphysically possible – otherwise, the agent is not free and the choice is not a genuine choice; see Ta, response 8, G 6:386/H 387. As Adams explains, this might in fact show that the teleological account of contingency comes down to the same thing as another famous Leibnizian characterization of contingency in terms

132

Freedom

that these options are not sufficiently good does not render them metaphysically impossible. Something analogous is true for free human choice. When you pick a movie from an online streaming service, you have a real choice between a number of different movies, each of which it is possible for you to watch. When you reject all but one, it is not because you lack the power to watch them – you simply lack the will. Likewise, the reality of choice strategy is a correlate of the moral necessity strategy. A choice is real if there are multiple metaphysically possible options among which the agent chooses the one she judges to be best – that is, if the choice is necessitated only morally but not metaphysically (see “On God and Man,” G 3:32/LGR 291). In fact, as already hinted, all four strategies are rooted in the distinction between determination by the good and determination by some other, value-neutral principle.37 3.2

Is This Genuine Contingency?

The four strategies just sketched appear to draw a legitimate distinction: there is value-neutral necessitation, on the one hand, and determination of a will by the apparent or actual good, on the other. This distinction is legitimate because these two types of determination rely on very different principles or laws; the source or ground of the necessitation is different for each type. Moreover, moral necessitation is teleological in the strongest sense while nonmoral necessitation is at most teleological in a thinner sense.38 Yet, of course, saying that the distinction is legitimate does not mean that it is a distinction between necessity and contingency, nor does it mean that it is any kind of modal distinction. One might still worry, as John Carriero does, that when Leibniz claims that these strategies give him contingency, he is merely “relabeling . . . [the] traditional categories” of contingency and necessity, failing to make room for anything that his contemporaries would have recognized as contingency (1991: 83, similarly in 1993: 3).39 Pointing to different sources of necessitation or determination may not seem sufficient in order to accommodate contingency.

37 38

39

of what is not per se necessary: “God’s choice is an indispensable link in the chain of explanation for the actuality of this world” because other worlds are per se possible (1994: 21). God’s goodness explains which of the per se possible worlds becomes actual. Leibniz sometimes associates contingency with determination by the good generally, without invoking one of the specific strategies I have explored. See e.g. DM 13. More precisely, morally necessary actions are rationally teleological while other actions are not. Leibniz might describe the actions of Spinoza’s God as agent-teleological because of God’s complete independence from other things and because God’s nature dictates all of his actions. Yet, Spinoza’s God does not act for the sake of the good, and his actions are therefore not rationally teleological. Other interpreters express this worry as well (e.g. Sleigh 2005: xxvi, Adams 1996: 110).

Contingency

133

Is what Leibniz describes as contingency really contingency? That question is surprisingly difficult to answer. Carriero may be right that Leibniz’s contemporaries would not have considered it to be genuine contingency. Yet, that might simply be due to the fact that Leibniz is doing something radically new; it is not entirely clear to me whose definition of contingency we should use in evaluating Leibniz’s claims. So, instead of pursuing that line of inquiry, let us ask a different question that may ultimately turn out to be more helpful: Can Leibniz distance himself meaningfully from Spinoza’s necessitarianism? As hinted earlier, I will answer this question in the affirmative. Moreover, I will show that this is what Leibniz ultimately cares about when insisting on contingency: he wants to avoid Spinozist versions of necessitarianism, according to which goodness or value do not play any genuine explanatory role at the fundamental level.40 Leibniz’s system, because of its reliance on rational teleology, does manage to do that, and Leibniz thinks that is an extremely important achievement. On my interpretation, Leibniz is claiming that what ultimately matters for freedom and responsibility is not contingency as traditionally understood; rather, it is moral necessity. Leibniz sometimes calls it ‘contingency,’ perhaps in order to signal that it is intended to play the role that contingency plays in other systems.41 Yet, he is perfectly happy to call it ‘necessity’ as well, as already noted. It is clear that Leibniz believes that in one sense, everything – including free actions – is necessary. The fact that he describes free actions as morally necessary is one important piece of evidence: free actions are not metaphysically necessary, but they are nevertheless necessary in some sense. Even more importantly, Leibniz often calls the necessity present in free actions a “happy necessity,” stressing that it is desirable to be moved irresistibly by the good (e.g. T 175; 191; 344; Ta, response 8, G 6:386/H 387). In addition, Leibniz sometimes states explicitly that there are different legitimate ways to use the term ‘necessity,’ and that in one of these senses, free actions are necessary. For instance, in a passage from the New Essays quoted in part earlier, Leibniz’s spokesperson admits: 40

41

See Lærke (2007: 61), who says something along very similar lines; see also Adams (1996: 110). Even Michael Griffin agrees with the core of my claim: he says that “[w]hat Leibniz is objecting to . . . is not Spinoza’s necessitarianism, but the fact that the necessity that applies to all things is ‘blind,’ unguided by divine wisdom and goodness” (2013: 81, 59). Yet, Griffin does not say much more about this difference between Leibniz and Spinoza, focusing instead on their similarity. According to Griffin, Leibniz and Spinoza agree that only some things exist by what Griffin calls “intrinsic necessity” (that is, exist simply in virtue of their essence) but that all other things exist by “extrinsic necessity” (that is, in virtue of the essence of God; 2013: 58). That, however, is not the kind of distinction crucial for understanding freedom. For a more detailed criticism of Griffin’s approach, see Jorati (2014b). In addition, Leibniz might be using that term because he is building on a medieval tradition of closely tying final causation (and hypothetical necessity) to contingency. See Schmid (2011: 354ff.) for a helpful discussion.

134

Freedom

If by ‘necessity’ we understand a man’s being inevitably determined, as could be foreseen by a perfect Mind provided with a complete knowledge of everything going on outside and inside that man, then, since thoughts are as determined as the movements which they represent, it is certain that every free act would be necessary; but we must distinguish what is necessary from what is contingent though determined. (NE 178)

I understand this passage as granting that there is a legitimate sense in which free actions are necessary but insisting that it is important to distinguish two types of necessity, or two ways in which something can be determined or inevitable. The broad sense glosses over a distinction that Leibniz finds crucial: the distinction between the morally necessary and the absolutely or metaphysically necessary. This distinction, however, need not be a modal distinction. Leibniz’s point could be that the modal status of these actions is not what ultimately matters; rather, what matters is the way in which these actions are determined. As Leibniz puts it, the realm of the necessary is different from the realm of the contingent because “the kind of determining that is involved is not the same” (NE 178). This way of looking at the issue can help us understand what Leibniz dislikes about Spinozism and why he thinks that he can successfully avoid it. In Spinoza’s system – at least on Leibniz’s interpretation – value and goodness do not play fundamental explanatory roles; there is no genuine rational teleology. Spinoza’s God does not act for the sake of the good.42 To know what exists, all we need at bottom is the Principle of Contradiction. Leibniz sometimes calls this a brute or “blind” necessity – that is, it is a necessitation by value-neutral principles, or a necessitation by something other than the motive of the good. In the Theodicy, for instance, he sometimes contrasts moral necessity with “metaphysical and brute necessity [necessité metaphysique et brute]” (T 174; similarly in T 349; letter to Remond, June 22, 1715, G 3:645).43 Similarly, he argues in a letter to Clarke that while it is true that God’s choice is determined by his wisdom, “it is not this fatality (which is only the wisest order of providence) but a brute fatality or necessity [une fatalité ou necessité brute] void of all wisdom and choice, which we ought to avoid” (LC 3.8, translation altered). In an appendix to the Theodicy, among other places, Leibniz explicitly associates this kind of brute necessity with Spinoza: 42 43

See Lin (2012: 435), who describes along similar lines one crucial difference between Spinoza’s and Leibniz’s views on necessity. See also Adams, who characterizes moral necessity as opposed to “blind necessity,” or to “any value-free mechanism determining choice” (2005: 183). My interpretation is close to Adams’s, though I reject his claim that there may well be moral necessity in nonrational monads (2005: 184). I disagree with this because on my interpretation, as seen in Chapter 3, there are fundamental differences between the final causation involved in different types of agency; not all monadic actions are goodness-directed, on my view, and free actions are goodness-directed in a special way.

Contingency

135

[T]here are reasons to draw a sharp distinction between the necessity that compels the wise to do good, which is called ‘moral,’ and which takes place even with respect to God, and that blind necessity, by which – as Epicurus, Strato, Spinoza, and perhaps Mr. Hobbes believed – things exist without intelligence and without choice. (RH 3)

Similarly, in the Theodicy proper, Leibniz says that Spinoza . . . appears to have explicitly taught a blind necessity, having denied the Author of Things understanding and will, and imagining that good and perfection pertain only to us, and not to him . . . [H]e does not acknowledge any goodness in God, properly speaking, and teaches that all things exist through the necessity of the divine nature, without God making any choice. (T 173; similarly T 371f.; Tp, G 6:43f./H 67f.)

In an earlier text, Leibniz accuses Spinoza of sharing the views of the “new Stoics,” as Leibniz calls them, who reject final causation and hold that God is determined by a “blind necessity,” lacks intellect and will, and is neither just nor benevolent (“Two Sects of Naturalists” [1678–1680?], A 6.4.1385/AG 282). As a matter of fact, in all passages that I know in which Leibniz criticizes necessitarianism or Spinozism, he specifically says that what he is objecting to is absolute, brute, or “blind” necessity. To name just one other example, when Leibniz’s spokesperson admits in the New Essays that he himself was once close to Spinozism, he says that he “once strayed a little too far in another direction, and began to incline to the Spinozist view which allows God infinite power only, not granting him either perfection or wisdom, and which dismisses the search for final causes and explains everything through brute necessity” (NE 73; similarly in “On Freedom,” A 6.4.1653/AG 94). Importantly, these passages not only state explicitly that we have to avoid brute, absolute, or “blind” necessity; some also say that we need final causation or goodness and choice in God. My suggestion, then, is the following: what Leibniz is eager to avoid is not necessitarianism generally but the kind of necessitarianism according to which everything is subject to an absolute, “blind,” or brute necessity.44 What we need, according to Leibniz, is a world in which only some things exist or are true because of the value-neutral laws of logic and metaphysics, and other things exist or are true because of the laws of goodness. And that is exactly what the four strategies considered earlier are supposed to accomplish. To put this more precisely, when Leibniz claims that free actions must be contingent, what he means is that they must be instances of rational teleology; they must not be determined in a brute or value-neutral way. This applies, in fact, both to 44

Lagerlund and Myrdal describe Leibniz’s reasons for rejecting Spinoza’s and Hobbes’s accounts in similar terms (2006: 172). See also Adams: “[Leibniz’s] principal reason for insisting on some sort of contingency in connection with free action seems to have been to insure the reality of choice – to insure that what happens is really influenced by final causes and judgments of value” (1994: 20, similarly 1996: 110).

136

Freedom

divine choices and to the choices of finite minds. For finite minds as well, the determination by what they think is best is a happy necessity, while determination by value-neutral principles is an imperfection. I suspect that some of the reasons why avoiding Spinozism is so important to Leibniz are the following. He thinks it matters for freedom, moral responsibility, and divine goodness that some things are determined by value-laden rather than by value-neutral principles. It matters that there is rational teleology at the fundamental level, in divine actions. If we deny this, we end up with a system in which we cannot say that God is wise, good, or free – a system in which it does not make sense to praise or blame God for any of his actions.45 As Leibniz says in his Comments on Spinoza’s Philosophy, “there is nothing loveable in a God who produces all good and evil things indiscriminately, by necessity” (Beeley 13; similarly in RH 12). A system in which God’s actions are governed by “blind” necessity is also a system without providence, which, for Leibniz, would mean the loss of a central feature of Christianity (see e.g. Tp, G 6:30f./H 54f.; “Two Sects of Naturalists,” A 6.4.1385/AG 282; RH 12).46 In fact, God would be superfluous in such a system (T 350). When Leibniz objects to Spinozism, he usually focuses on the lack of teleology and of concern for goodness at the divine level. And that is certainly bad enough, for Leibniz. Yet, there is good evidence that he also has significant reservations about Spinoza’s account of creaturely agency, and for parallel reasons. Leibniz appears to think that we cannot hold finite agents morally responsible in a Spinozistic world because there is no genuine teleology or freedom in their actions. After all, Spinoza’s finite agents are not spontaneous in any of Leibniz’s three senses. That is a problem because, as we saw in Chapter 3, Leibniz holds that only substances that possess metaphysical spontaneity can be subject to genuine immanent teleology. Likewise, as seen in this chapter, freedom requires spontaneity and teleology. Because spontaneity, teleology, and freedom are requirements for moral agency,47 Spinoza’s finite agents are not moral agents. When Leibniz claims that freedom requires contingency, then, he is saying that freedom is incompatible with value-neutral determination and that free actions must be determined by the agent’s inclination toward what she judges to be good. This, in fact, is also how Leibniz can avoid the charge that given his determinism, free agents resemble automata: unlike machines, and even unlike 45 46

47

For more on the connection between freedom and moral responsibility, see Chapter 7. Similarly, see “Specimen of Catholic Demonstrations” (1683–1686?), where Leibniz says that “if all things are directed by mind, then the production and preservation of all things will not be necessary and blind, but free and full of purpose. From this the elements of piety and justice immediately proceed” (A 6.4.2325/LGR 106). I will argue this in Chapter 7. See also Bobro, who argues that, for Leibniz, thinking machines would not be moral agents because they lack genuine spontaneity (2004: 67ff.). The same, arguably, applies to Spinoza’s finite agents.

Contingency

137

animals, rational agents are not subject to a “blind” or brute necessity to the extent that they are free. Instead, they determine themselves based on their rational grasp of the good – something that brute animals cannot do and, as we will see in the next two chapters, something that Leibniz thinks is crucial for control and moral agency. It is interesting to note that if my interpretation is correct, Leibniz is agreeing with Aquinas in a central respect, while disagreeing with later Scholastics like Luis de Molina and Francisco Suárez. After all, Aquinas also appears to hold that necessitation does not generally undermine freedom and responsibility; what matters instead is that the agent is not subject to what Aquinas sometimes calls ‘necessity of force.’ It is not always a problem for freedom or moral responsibility to be subject to what he calls ‘necessity of natural inclination,’ such as in cases where an agent is irresistibly drawn to something that she recognizes as good in all ways (e.g. SCG 3.138.2; ST I q82 a1 corp.; II.II q58 a3 ad2; III q46 a1 corp.; and DV q22 a5 corp.). 3.3

A Residual Metaphysical Worry

Before turning to the worry that Leibniz’s logical commitments make everything absolutely or metaphysically necessary, I should address a potential objection of a more metaphysical nature. I claimed earlier that morally necessary actions are only hypothetically and not absolutely necessary. Yet, one might worry that at least God’s morally necessary actions are also absolutely necessary. After all, the objection goes, it is absolutely necessary that God is perfectly good, and being perfectly good just means always trying to do what one believes to be best. That God follows the Principle of Wisdom and Goodness, then, itself seems absolutely necessary.48 After all, if a thing is necessary on a hypothesis that is absolutely necessary, that thing itself plausibly is absolutely necessary. To put this point slightly differently, something that is metaphysically necessary arguably does not count as a hypothesis for the purposes of establishing hypothetical necessity. Otherwise, God’s existence would be merely hypothetically necessary because it is necessary only on the hypothesis that existence is included in his essence. In fact, all metaphysical necessities would be hypothetically necessary because their necessity depends on the hypothesis 48

In the case of rational creatures, it is at least somewhat more plausible to say that following the Principle of Wisdom and Goodness is merely hypothetically necessary, since God could have created creatures following different principles. See Kulstad, who makes room for possible substances with very different laws of appetites, e.g. finite spirits that always do what appears worst (1990: 143). Moreover, God could intervene miraculously and force a creature to act contrary to its nature. Of course, if all divine actions were absolutely necessary, creaturely hypothetical necessity would likewise be jeopardized. This is all the more reason to preserve a sense in which God’s actions are not absolutely necessary.

138

Freedom

that the Principle of Contradiction holds. That cannot be how hypothetical necessity is supposed to work; the hypothesis must, it seems, be something that is not itself metaphysically necessary. But then, just as God’s existence is necessary regardless of any hypothesis, God’s goodness, or God’s adherence to the Principle of Wisdom and Goodness, seems necessary regardless of any hypothesis. After all, God is necessarily and essentially good. That, in turn, may seem to imply that all of God’s actions are absolutely necessary because they are necessitated by God’s goodness. A related worry is the following: if it is metaphysically necessary that God exists and is perfectly good, and if it is metaphysically necessary that a perfectly good being wills what is best, it seems to follow that it is metaphysically necessary for God to will what is best (see e.g. Russell 1937: 39, Lovejoy 1936: 172f.). More formally, the worry is that Leibniz accepts the premises of the following argument but wants to deny its conclusion:49 1. It is metaphysically necessary that God is perfectly good. 2. It is metaphysically necessary that any perfectly good being wills the best.50 3. Therefore, it is metaphysically necessary that God wills the best. In response to both worries, Leibniz could simply admit that God’s choice of the best is indeed absolutely and metaphysically necessary. That admission would not undercut the gist of the strategies I discussed above, since Leibniz could still draw the distinction between what is morally necessary and what is not morally necessary. He could simply view moral necessity as a sub-category of metaphysical necessity and insist that the distinction between moral and nonmoral necessitation is of the utmost importance. After all, even if God’s choice of the best is absolutely and metaphysically necessary, its necessity still depends on the Principle of Wisdom and Goodness. Hence, there could still be a meaningful distinction between what is necessitated by goodness and what is necessitated by value-neutral principles. Yet, this solution goes against some of the textual evidence considered earlier: Leibniz sometimes explicitly states that free actions are not subject to absolute or metaphysical necessity (e.g. CD 21). Let us therefore treat this solution as a last resort. Alternatively, Leibniz could deny Premise 2: he could insist that it is not absolutely or metaphysically necessary that a perfectly good being choose the best. And, indeed, this does appear to be the route Leibniz wants to take in at least some mature texts. As already seen, he says in the Theodicy that “metaphysically speaking, [God] could have chosen or done what was not best; but he could not have done this morally speaking” (T 234; similarly in T 237; letter to 49 50

For different formulations of this type of argument, see Blumenfeld (1988: 83), Griffin (2013: 59f.), and Anfray (2011: 49). Strictly speaking, this premise would have to say either that any perfectly good and omniscient being wills the best or that any perfectly good being wills what seems to this being to be best. To keep things simple, however, I will omit those qualifications.

Contingency

139

Coste, December 19, 1707, G 3:402/AG 195). To evaluate this proposal, it is helpful to distinguish the de re reading of ‘It is metaphysically necessary that God chooses the best’ from the de dicto reading.51 According to the former, the thing that is best is such that it is metaphysically necessary for a perfectly good being to choose it. According to the latter, in contrast, it is metaphysically necessary that a perfectly good being choose whatever is best. In order to deny that it is metaphysically necessary that a perfectly good being choose the best, Leibniz could do either of two things: (a) he could argue that it is metaphysically necessary only on the de dicto but not on the de re reading, or (b) he could deny that it is metaphysically necessary on either reading. As Robert Adams points out (1994: 25), there is some textual evidence that Leibniz employs the first of these two options. In his 1706 reply to Bayle, Leibniz writes: It is true that this proposition, ‘God wills the work most worthy of him,’ is necessary. But it is not true that he wills it necessarily. For this proposition, ‘this work is the most worthy,’ is not a necessary truth, it is an indemonstrable, contingent truth of fact. I believe that it can generally be said that this proposition is necessary: ‘his will will act following the greatest inclination.’ But it does not follow that it will act necessarily. (Gr 493/SLT 114; similarly in “On Contingency,” A 6.4.1652/AG 30)

According to this passage, the general claim that God wills whatever is best is indeed necessary – presumably, metaphysically or absolutely necessary. Yet, it is not necessary that God choose the actual world, since it is not necessary that the actual world is best. Even though Leibniz does not put it that way, his point appears to be that the statement ‘It is metaphysically necessary that God wills the best’ is true on the de dicto but not on the de re reading. That would allow him to likewise deny the de re reading of Premise 2 in the argument under consideration: it is not metaphysically necessary that any perfectly good being wills the option that in fact is the best. As a result, Leibniz could deny that God’s choice of the actual world was metaphysically necessary. Why would the bestness of the actual world fail to be absolutely or metaphysically necessary, on my interpretation? Here, I have to slightly tweak my categories because we are now talking about the contingency of states of affairs or facts, rather than actions. It cannot quite be moral necessity that makes this bestness contingent since moral necessity concerns the necessitation of a rational mind by reasons. For Leibniz, the bestness of the actual world is not the outcome of anybody’s choice. Yet, my general strategy might nevertheless apply: the bestness of the actual world might depend on the Principle of Wisdom and Goodness, not just on the Principle of Contradiction. Purely logical principles, plausibly, do not tell us which world is best, since logically 51

Some other interpreters also distinguish these two readings: Adams (1994: 25, 36), Anfray (2011: 50), and Pickup (2014: 511).

140

Freedom

speaking, all possible worlds are on a par. A world that is sub-optimal does not thereby violate the Principle of Contradiction. I find it plausible that the best world is best not because of the Principle of Contradiction but because of some principle or standard of goodness, which is not itself a logical principle. Leibniz might hold that facts about goodness supervene on more basic facts but in a way that presupposes or relies on value-laden, nonlogical principles. The Principle of Contradiction alone does not give us goodness.52 This would allow Leibniz to claim that the bestness of the actual world is not metaphysically necessary: the Principle of Contradiction does not by itself rule out the bestness of some other world. Likewise, it enables Leibniz to deny that the bestness of our world is absolutely necessary; this world is best only on the hypothesis of a particular standard of goodness.53 The strategy just described for denying the absolute or metaphysical necessity of God’s choice of the best strikes me as quite promising and interesting. It is, after all, supported by at least one very explicit passage, and it is independently appealing. Yet, I would briefly like to explore the other option – namely, the proposal that even de dicto, it is not absolutely or metaphysically necessary for God to choose the best. Given what I said earlier about Leibniz’s notion of contingency, this option is more plausible than it may otherwise appear to be. Suppose that calling an agent perfectly good means that she has a perfectly good character or perfectly good inclinations.54 In order to know what this agent will do, however, we need to know not only what her character is but also that she will not act out of character, or against her inclinations. How would we know that? Well, the Principle of Sufficient Reason requires it: there is no sufficient reason for the agent to act against her strongest inclination or character. But if violations of the Principle of Sufficient Reason do not amount to violations of the Principle of Contradiction – and there are good reasons to think that this is the case for 52

53

54

This consideration is connected to something I argue elsewhere: a bare monad’s perceiving the entire universe does not imply its perceiving things as good, because goodness is a higher-level property (Jorati 2013: 48f.). On this interpretation, it is not metaphysically necessary that something is the correct standard of goodness; supposing that some other standard of goodness is correct does not imply a logical contradiction. This may be controversial and I cannot argue for it here, but I find it plausible. In what exactly would that contradiction consist? Arguably, the correctness of the correct standard is grounded in the Principle of Sufficient Reason, not in the Principle of Contradiction alone. See Pickup (2014: 513ff.), who explores this issue at length. He argues that ultimately Leibniz cannot view the standard of goodness as metaphysically contingent because that would commit him to a brute fact: the correct standard of goodness would be ungrounded. Yet, if my interpretation is correct, Leibniz can avoid this consequence. The truth of the correct standard of goodness might be grounded in God’s nature and the Principle of Sufficient Reason in such a way that it is not logically necessary. Leibniz, in fact, suggests that this is his understanding of goodness in “Meditation on the Common Concept of Justice”: “Goodness is simply the inclination to do good to everyone” (Mollat 48/Riley 50).

Contingency

141

Leibniz55 – that is not enough to get us metaphysical or absolute necessity. Logically speaking, we cannot rule out the agent’s acting out of character. It is not a logical contradiction for a perfectly good being to fail to do what is best, even though it would be a violation of the Principle of Sufficient Reason. Logically speaking, we cannot rule out any substance’s acting without any reason whatsoever. There is another plausible way for Leibniz to conclude that it is not metaphysically necessary for God to choose the best: perhaps the correct way to understand a substance’s nature is teleologically, analogous to Scholastic substantial forms.56 Natures specify not what properties the substance must have but rather what properties the substance ought to have, or properties that the substance aims at. This would again make it metaphysically possible for a substance to have a particular nature but to lack some of the states that this nature specifies. On the Scholastic picture, after all, it is possible for agents to act against their natures. For instance, it is possible for a rational animal not to exercise her rationality and for an acorn not to become an oak tree. Analogously, we might be able to say that there is nothing metaphysically impossible about God failing to act in the way that a perfectly good being ought to act. There would not be a sufficient reason for this failure, but again, that might not mean that it is metaphysically impossible. There is some textual evidence that for Leibniz, it is not metaphysically or absolutely necessary that a perfectly good being choose whatever it judges to be best. In the Theodicy, for instance, he says, [A]s to the connection between causes and effects, it only inclined the free agent without necessitating him . . . [T]hus it does not produce even a hypothetical necessity, except in conjunction with something from outside, namely, this very maxim, that the prevalent inclination always wins. (T 53)

What Leibniz is saying here entails, I think, that it is not absolutely or metaphysically necessary for the prevalent inclination in a mind to win. After all, he calls it “something from outside [quelque chose de dehors]” and implies that whatever follows from this maxim is only hypothetically necessary. Plausibly, the idea is that violations of this maxim are metaphysically possible: there is nothing contradictory about a case in which the prevalent inclination does not win.57 Hence, there is nothing contradictory about a being with perfectly good inclinations who does what is not best. 55 56

57

Note that Leibniz sometimes distinguishes contingency from necessity on the basis of the Principle of Sufficient Reason (e.g. LC 5.9f.). See Nachtomy, who argues that we should view “the individual’s complete concept as a prescription for action” and that this allows us to deny that those actions are metaphysically necessary (2007: 156). See also a letter to Jaquelot, September 4, 1704: “our future actions are in us, but only as a kind of inclination, which carries no necessity with it” (G 6:559/WF 188). This, of course, is the same point discussed earlier: Leibniz distinguishes between inclination and necessitation; motives or

142

Freedom

As a result, Leibniz has a number of good ways of responding to the worry that all divine actions are metaphysically and absolutely necessary. He could deny that the bestness of the best option is metaphysically necessary and hence deny that it is metaphysically necessary for a perfectly good being to choose this option. Or he could deny that it is metaphysically necessary for perfectly good beings to choose whatever is best. The same goes, mutatis mutandis, for absolute necessity. Alternatively, Leibniz could bite the bullet and admit that God’s choices are absolutely and metaphysically necessary. That bullet is not as Spinozistic as one might think, because Leibniz would still be able to draw a distinction between moral necessity and nonmoral necessity, as seen earlier. On any of these options, Leibniz can distance himself meaningfully from Spinoza’s necessitarianism, according to which no divine action is morally necessary. Yet, given Leibniz’s repeated insistence that free actions are not metaphysically necessary, it seems that he did not want to bite the bullet but thought that one of the other options would work. It is not clear to me which of the other options Leibniz endorses; he appears to try out different strategies at different times and it is possible that he endorses them both. I do think that Leibniz has the resources for claiming that even de dicto, it is not absolutely or metaphysically necessary for God to choose the best, and I like that solution. 3.4

Logical Commitments and Contingency

Let us now turn to the question of whether Leibniz’s logical and truth-theoretic commitments entail necessitarianism.58 The relevant commitments are the following. First, Leibniz holds that in any true affirmative proposition, the predicate is contained in the notion of the subject. Second, he holds that every substance has a complete concept that individuates it and contains everything that can ever be truly predicated of this substance. These commitments may seem to entail necessitarianism: if Julius Caesar’s concept specifies that he crosses the Rubicon, and if any man who does not cross the Rubicon is ipso facto not Caesar, it looks like it is necessary for Caesar to cross the Rubicon.59 Or, more precisely, it seems necessary that if Caesar exists, he crosses the Rubicon. The only way for Caesar’s action to be contingent, it may seem, is for his existence to be contingent. Yet, that does not seem to be the kind of

58

59

the good do not necessitate absolutely or metaphysically but only morally. Now, we have a deeper understanding of what that might mean. Failing to act in accordance with one’s inclinations is metaphysically possible. As mentioned in Chapter 2, Leibniz expresses these commitments mostly in texts from the middle period. Because this book focuses on the mature Leibniz, they are hence not of very central concern to us here. Yet, since these commitments do occasionally show up in mature works – that is, since Leibniz does not appear to have abandoned them entirely – it seems fair to ask whether they are compatible with Leibniz’s mature theory of contingency. This is an example Leibniz uses himself, for instance in DM 13.

Contingency

143

contingency required for freedom;60 it does not, for instance, allow us to say that Caesar could have refrained from acting in this way. Moreover, if God’s actions are necessary, Caesar’s existence is also necessary. Of course, concepts cannot make an agent act.61 When Caesar crosses the Rubicon, it is not because of his complete concept; rather, it is because of his inclinations. And that concept is his concept in part because he is a substance with those inclinations.62 Still, that does not make the problem go away entirely. One can still worry that Caesar necessarily has this concept and that it is therefore metaphysically necessary for Caesar to do whatever the concept specifies.63 At the very least, Leibniz must distinguish between moral and nonmoral necessity at the level of complete concepts. The contingency or moral necessity of Caesar’s free actions must be captured by his complete concept. If at the concept level there is no distinction between contingent and necessary properties, it is unclear how that distinction could hold for the actualized Caesar. As already seen, Leibniz insists that God creates substances with all and only the properties that are contained in their concepts. In order to address this worry, let us try to translate the metaphysical strategies discussed earlier into solutions at the concept level. First of all, it must be possible to distinguish, by looking at an agent’s complete concept, between states that the agent has because of what she perceives as good and all other states. After all, no matter how one views complete concepts, they must contain information about the psychological processes in a substance that lead up to an action. Otherwise, they would be missing important predicates. That allows us to distinguish, at the level of complete concepts, between actions that are morally necessary and actions that are not. This solution is the logical analogue of the bullet-biting metaphysical solution I offered earlier: Leibniz could draw the distinction between morally and nonmorally necessary states even if moral necessity turns out to be a sub-category of absolute or metaphysical necessity. Yet, there are also logical analogues of the more ambitious metaphysical solutions. The first step is to claim that complete concepts are best viewed not as lists or sets of predicates but rather as entities with a hierarchical structure. There is, in fact, textual evidence that this is what Leibniz has in mind. First of all, instead of saying that concepts contain all predicates, Leibniz sometimes 60 61 62 63

But see Nelson (2005: 298), who argues that for Leibniz, this type of contingency is sufficient for human freedom. Ohad Nachtomy makes this point as well (2007: 161). As we saw in Chapter 2, there must be something in the substance that explains why its complete concept belongs to it; see Chapter 2, footnote 13. Incidentally, this problem is a creaturely analogue of the problem of explaining divine freedom that we discussed earlier. Just as one might worry that God, given his essence, must do the good by metaphysical necessity, one might worry that Caesar, given his concept, cannot but cross the Rubicon and is therefore subject to metaphysical necessity.

144

Freedom

talks of concepts being such that one can deduce all predicates from them.64 This suggests that some work is necessary in order to find certain predicates in a complete concept, presumably because these predicates need to be derived from more fundamental predicates that are explicitly included in the concept. Relatedly, Leibniz in at least one text claims that the complete concept of an individual substance contains only “primitive predicates,” that is, those predicates that are fundamental and do not depend on other predicates (remarks on Arnauld’s letter, May 1686, G 2:44/AG 74). This indicates that we should think of complete concepts as containing non-fundamental predicates only virtually – namely, by containing their grounds, from which we can derive them.65 Alternatively, we can distinguish between the core of a complete concept, which contains the primitive or fundamental predicates, and the rest of the concept, which contains the non-primitive predicates that follow from the core.66 Even independently of the textual evidence, it makes sense to think that complete concepts have some kind of a hierarchical internal structure. First of all, it seems superfluous for a concept to include predicates explicitly that can be derived from other predicates in the same concept. For instance, if your concept includes the predicates ‘reads the novel Waverley’ and ‘reads the novel Wuthering Heights,’ it would be redundant to also include ‘reads at least two novels.’ More importantly, since a concept is supposed to capture everything that is true of the substance whose concept it is, it must contain information about what the metaphysically most foundational features of the substance are, as well as how other features relate to this foundation. In order to have a perfect concept of the cactus in my office, I would need to have perfect knowledge of the cactus’s nature, that is, of its explanatorily fundamental metaphysical properties. In fact, when God created the central monad of this cactus, he created a substance with a particular nature; God foreknew that this substance would, in virtue of its nature, spontaneously produce a particular series of states. The cactus’s nature is explanatorily prior to its modifications. It would make sense, then, that the complete concept of this cactus – which is, after all, God’s idea of the cactus – is primarily the concept of a substance with this particular nature. Just as God foreknows what modifications the cactus will 64 65

66

He does this, for instance, in DM 8 when he introduces his complete concept theory. See also remarks on Arnauld’s letter, May 1686, G 2:42/AG 73; G 2:46/AG 76. Leibniz himself distinguishes between sentences whose subjects explicitly contain the predicate – presumably, sentences like ‘the round table is round’ – and sentences whose subjects contain the predicate only virtually. In the latter cases, we need to analyze the subject in order to find the predicate. See DM 8; “On Freedom,” A 6.4.1656/AG 96. Vailati draws a similar distinction between predicates that belong to the essence of the substance and qualities that do not belong to its essence (1986: 201f., similarly in Lin 2012: 428ff.). Even more similar to my proposal is Hawthorne and Cover’s (2000: 157). Finally, see Sleigh (1990a: 78f.).

Contingency

145

have by knowing its nature, he would then know the non-fundamental predicates by knowing the fundamental ones. It would make sense for complete concepts to have a structure that mirrors the ontological structure of the corresponding substance.67 The second step is the following. If complete concepts have the hierarchical structure just described, non-fundamental predicates could follow contingently from the core of the concept.68 In fact, if the structure of the concept mirrors the structure of the corresponding substance, we can employ an analogue of the last solution from Section 3.3. I argued there that even though all of God’s actions follow from his nature or essence, that need not mean that it is metaphysically impossible for God to perform sub-optimal actions. Having a perfectly good nature, after all, might just mean having only the best inclinations. It might be metaphysically possible for an agent to act contrary to her inclinations. In order to know that God will not act out of character, or against his inclinations, we arguably need to rely on the Principle of Sufficient Reason in addition to the Principle of Contradiction.69 There might not be anything metaphysically impossible about someone acting without any reason whatsoever. Let us translate this strategy into the complete concept level and apply it to created agents: perhaps we need more than the Principle of Contradiction in order to derive the predicate ‘crosses the Rubicon’ from the core of Caesar’s concept. We may also need to presuppose that Caesar will not act contrary to his prevalent inclination, and in fact we need to rule out the possibility that God intervenes miraculously and prevents Caesar from crossing the Rubicon. Metaphysically speaking, it might be possible that there is a substance with Caesar’s nature, and hence with Caesar’s core concept, which does not cross the Rubicon.70 In fact, Leibniz suggests something along these lines in his “Dialogue on Human Freedom” from the mid-1690s: the sins of human beings are “grounded in their essence, without, however, resulting from it as a necessary consequence” (Gr 368/AG 117). If this is correct, there is a sense in which ‘Caesar does not 67

68

69 70

I take this to be compatible with the claim that in God’s decision about which substances to actualize, the non-fundamental properties or predicates might be prior. After all, God wants the perceptions of all substances to harmonize, so in his deliberation he might start with harmonious series of perceptions and then work his way back to the complete concepts of substances whose natures are such that they will issue the requisite perceptions. Thinking about God’s reasoning in this way does not mean, however, that the complete concepts God finds in his intellect do not mirror the explanatory structure of the corresponding substances. It merely means that the derivative predicates enter into God’s deliberation at a more fundamental level. See Baxter, who also distinguishes between two different ways in which a concept can contain a predicate: logically and causally (2000: 203ff.). See also Grimm, who suggests that a subset of the predicates true of a thing form the essence of the thing, and they are the predicates necessarily included in the thing’s concept (1970: 217). See Dicker (1982: 230f.), who argues for a similar reading. This would mean that we could individuate substances either merely in terms of their core concepts or in terms of the core concept plus whatever (contingently) follows from that core. Leibniz usually appears to talk about individuation in the second way.

146

Freedom

cross the Rubicon’ does not imply a contradiction. To show that this sentence is false, we need to rely on the Principle of Sufficient Reason as well as on the contingent (or morally necessary) fact that God will not intervene. Caesar’s action of crossing the Rubicon is grounded in his complete concept, but it does not follow from the core of his concept as a necessary consequence. The solution just sketched works for free and unfree actions: all properties that are not fundamental metaphysical properties of a substance correspond to non-fundamental and hence noncore predicates in the substance’s complete concept. Insofar as Leibniz believes that all creaturely actions are contingent, this is not a problem. Moreover, we can distinguish between free and unfree actions – or morally necessary actions and other actions – at the logical level: just as free actions depend on the will, predicates describing free actions must derive from predicates describing the agent’s tendency to act in accordance with her rational judgments about the good. Hence, while everything a creature does and every fact about the creature is necessary in some sense – namely in the sense that it follows from, or is included in, its concept – there are meaningful distinctions even at the logical level between things that a creature does because it believes them to be best and things that it does “blindly.” Leibniz seems happy to grant that these are both necessary in some sense; he merely insists that these are different types of necessity that we do well to distinguish. This, I think, he can do both at the logical and at the metaphysical level of description. At the metaphysical level, free actions are special because of the explanatory role that goodness plays in them. Similarly, at the level of complete concepts, the predicates corresponding to free actions are special because of the way they depend on more fundamental predicates about goodness and goodness-directedness. Likewise, if violations of the Principle of Sufficient Reason are metaphysically possible, we can resist metaphysical necessitarianism on both levels. Incidentally, viewing complete concepts as hierarchically structured also fits well with another prominent Leibnizian account of contingency: the infinite analysis account, which Leibniz advances repeatedly in the middle period.71 On that account, for all necessary propositions, there is a finite analysis demonstrating that the predicate is contained in the concept of the subject. For contingent propositions, in contrast, there is no such finite analysis. If complete concepts are structured in the way I have suggested, and if all predicates describing actions depend on more fundamental predicates describing inclinations, it makes sense that there is no finite analysis that would take us from the agent’s complete concept to an action predicate.72 After all, as we saw 71 72

In the late period, Leibniz almost never mentions that account of contingency; the mature Leibniz appears to prefer the more metaphysical strategies that I outlined in Section 3.1. This might even solve the so-called “lucky proof” objection; see Hawthorne and Cover (2000).

Conclusion

147

in Chapter 1, each state of a Leibnizian substance is infinitely complex, as is the explanation of each such state in terms of prior inclinations or in terms of the substance’s primitive force. Since, according to my proposal, an analysis of the substance’s complete concept has to mirror that metaphysical or explanatory structure, it is plausible that the analysis cannot yield an action predicate in a finite number of steps. 4

Conclusion

Leibniz argues that there are three necessary and jointly sufficient conditions for freedom: spontaneity, intelligence, and contingency. Free agents are immune from external determination, know the object of deliberation, and are not subject to metaphysical or absolute necessity. Even though Leibniz uses the term ‘contingency,’ it is quite plausible to read Leibniz as embracing a type of necessitarianism and rejecting only “blind” necessitarianism. In other words, he rejects a necessitarianism like Spinoza’s that leaves no room for goodness or value to do any genuine explanatory work at the fundamental level. If Spinoza were right, Leibniz thinks, the world would be an enormously undesirable place – a place lacking a good, lovable, free, providential deity and even finite moral agents. By making rational teleology a fundamental feature of his system, Leibniz manages to avoid this kind of necessitarianism. Leibniz’s theory of freedom, which is a form of agent-causal compatibilism, has several unique advantages. It allows Leibniz to deny that free actions are determined by anything external to the agent without abandoning the Principle of Sufficient Reason or determinism. Likewise, it allows him to distinguish sharply between free actions and the activity of automata or nonrational substances, which are not goodness-directed in the way required for freedom. In the next two chapters, I will explore two further advantages of Leibniz’s theory of freedom. First, it provides him with a compelling account of the control that free agents possess over their actions. Second, it allows him to claim that free agents have a special moral status.

6

Control, Weakness, and Compulsion

We saw in Chapter 5 that Leibniz is a determinist and a compatibilist: he holds that all actions are determined and that some actions are nevertheless free. We also saw that Leibniz’s theory of freedom is a form of agent-causal compatibilism with some unique advantages. For Leibniz, free actions are determined only by the agent; they are not even indirectly determined by anything external. The present chapter examines Leibniz’s response to three problems that plague many determinists, namely (i) the problem of explaining the sense in which free agents have control over their actions,1 (ii) the problem of explaining ostensibly akratic or weak-willed actions, and (iii) the problem of distinguishing between weakness of will and compulsion. The question of whether Leibnizian agents possess control over their free actions goes beyond what I have said so far. One may worry, after all, that even if an agent’s free actions are determined only by her nature, she has genuine control only if she has a say in what her nature is. But agents clearly have no such say, on Leibniz’s view. Free actions, it may seem, are just as determined by factors beyond the agent’s control as are unfree actions. As a result, Leibniz needs to say more to address the worry that his brand of determinism leaves no room for control. And indeed, he does: he explicitly discusses the notion of control – or, as he usually calls it, ‘mastery.’ The present chapter’s first aim is to show that Leibniz manages to make room for a meaningful and desirable type of control. The second aim of this chapter is to explore a notion that is closely related to control: akrasia, or weakness of will.2 Akrasia is particularly problematic for determinists who hold that all intentional actions are determined by the agent’s perceptions of or judgments about the good.3 And Leibniz does appear to 1 2 3

For discussions of the difficulties that determinists face in accounting for control, see Dennett (1984: 51) and Fischer and Ravizza (1998). I am using ‘weakness of will’ and ‘akrasia’ interchangeably in this chapter. Akrasia is, of course, also problematic for many libertarians, especially those who endorse guiseof-the-good views. Yet, determinists are arguably under particular pressure to say that agents are determined by their judgment that something is best. Libertarians can simply say that an agent can freely choose a less good option without a sufficient reason.

148

Control

149

endorse this view. It may thus seem as if Leibniz’s moral psychology implies that we can never knowingly act against our better judgment. Yet, I argue that the resources that allow Leibniz to give a powerful account of control also allow him to acknowledge a form of akrasia.4 Finally, the chapter briefly examines whether Leibniz can capture the difference between akrasia and compulsion – another notoriously difficult feat for determinists. One upshot of the present chapter is that Leibniz’s moral psychology is less intellectualist and less similar to the Socratic view than often assumed. It is, in fact, much closer to Plato’s or Aristotle’s moral psychology. What I mean by that is the following: Leibniz does not hold that what we knowingly do is simply a function of which option we judge to be best, all things considered. Nor does he hold, conversely, that failures to do what is best are always due to ignorance or cognitive error.5 Instead, nonrational inclinations can influence our intentional actions in many different ways – not just by giving us a distorted picture of what is genuinely good. This, in fact, already came up in previous chapters: for Leibniz, the mind is somewhat like Plato’s divided soul, in which the rational element must struggle to gain the upper hand over the nonrational elements. Thinking about Leibnizian minds in this way makes the connection between control, weakness, and compulsion much clearer. As we will see, the agent is in control when the rational elements of her soul manage to subdue the nonrational elements; when the rational elements fail to do so, the agent is sometimes akratic and sometimes compelled, depending on the reasons for this failure. 1

Control

Let us first look in more detail at the reasons why it is difficult for Leibniz, or any philosopher with a similar moral psychology, to accommodate the notion of control. For Leibniz, as already seen, all actions are determined by prior perceptions and appetitions, and ultimately by the agent’s nature. In a particularly colorful passage, Leibniz stresses that it is a mistake to view the will as a queen seated on her throne; the understanding is her minister of state and the passions are her courtiers or favorite ladies, who by their influence often prevail over the counsel of the minister . . . [T]he understanding speaks only at this queen’s command; she can vacillate between the reasons provided by the minister and the suggestions of the

4

5

It is not entirely surprising that he can use the same resources because the notions of control and of weakness of will can be viewed as two sides of the same coin. See e.g. Watson, who argues that the fault of weak-willed agents is their lack of control (2004: 50f.). For a helpful discussion of the differences between intellectualism and anti-intellectualism, or between Socratic views and Plato’s as well as Aristotle’s views, see Irwin (1992: 453ff.).

150

Control, Weakness, and Compulsion

favorites; she can even reject the ones and the others; finally, she can silence them or make them speak and give them audience or not, as seems good to her. (COE 16)6

The will does not control the other faculties in the way that a queen controls her courtiers and ministers, Leibniz insists. In fact, in order to do that, the will would “need another understanding in itself, to understand what is presented to it” (COE 16) and subsequently to decide to which adviser she ought to listen. Endorsing that picture, Leibniz appears to be saying, amounts to viewing the will as a homunculus that can judge and decide. For Leibniz, in contrast, the will is determined by the input it receives,7 and this input is not under its ultimate control. He explicitly acknowledges this: “we will what we find good, which depends on our taste and the objects, and not on our choice” (“Conversation about Freedom and Fate,” Gr 482/SLT 99). More generally, Leibniz’s determinism may seem to rule out any meaningful control over our voluntary actions.8 If everything you do is already settled by the way you were the day you were created, what kind of control can you possibly have over any of your actions?9 While it is true, as seen in Chapter 5, that nothing external to you determines your actions, this does not significantly assuage the worry that you are not the one who is in control. After all, you have no control whatsoever over your nature; you are simply born (or created) with it. Moreover, it may seem that you have no more control over your voluntary actions than over your nonvoluntary actions, and it could seem that rational creatures have no more control than nonrational creatures. Both types of actions and both types of creatures are determined by prior states, or the agents’ natures. The only difference, one might conclude, is that free agents are aware of, or have more distinct perceptions of, what they are doing and why they are doing it.10 Yet, this awareness clearly does not afford them any additional control; if they feel more in control, it is an illusion. One might think that, for Leibniz, the difference between voluntary and nonvoluntary actions is rather like the difference between a nearsighted person watching a movie with her glasses on and without her glasses. Having a better idea of what is going on in the movie does not give her any additional control over the plot, even though it might enable her to predict more accurately what will happen next.

6 7

8 9 10

This is not as much of a straw man as it may at first appear. John Bramhall, for instance, endorses a very similar picture (A Defence of True Liberty §7, in Hobbes and Bramhall 1999: 46). That the will is determined by its input does not mean, however, that the will is unimportant. By way of analogy, the verdicts of an ideally just and rational judge may be completely determined by the input she receives, but that does not make this judge insignificant or superfluous. This in turn may seem to undermine our moral responsibility, as Riley worries (1996: 49). See Seidler, who contends that given Leibniz’s commitment to determinism, he is not entitled to the notion of control (1985: 19f., 34). Some interpreters do read Leibniz along these lines; see e.g. Greenberg (2005: 224).

Control

151

Nevertheless, Leibniz repeatedly stresses that agents capable of voluntary action have a special kind of control or mastery over their actions.11 The fact that we are determined by how things seem to us and by the “prevalence of inclinations,” as Leibniz sometimes puts it, “does not prevent man from being master in his domain [maître chez lui]” (T 326). In virtue of man’s capacity for willing and acting rationally he has a type of control over his actions that lower animals lack: by willing, a man “determines his thoughts by his own choice instead of being determined and swept along by involuntary perceptions” (NE 180). In such passages, Leibniz seems to be denying that the difference between voluntary and nonvoluntary actions consists merely in a more distinct awareness of what one is doing. He claims that we are masters of the former but not the latter.12 I will argue that, despite appearances to the contrary, Leibniz is entitled to this claim. Even though every Leibnizian action is determined by the agent’s nature, there is a crucial type of control over one’s actions that only an agent with a will can possess, and this type of control is worth having.13 Before turning to the details of Leibniz’s discussion of control, let me say a few words in general about what we are looking for when we seek a satisfactory account of control. In the contemporary philosophy of action, some philosophers distinguish between two different types of control. Alfred Mele, for instance, distinguishes between the type of self-control that is opposed to akrasia and the type that is often called ‘autonomy’ (1995, Preface). Along very similar lines, Gary Watson distinguishes between the ability to act in accordance with one’s rational desires, ends, or values and the ability to control one’s values or to adopt certain ends (2004: 271f.). For Watson, the former ability is connected to attributability: only actions that are in accordance with the agent’s values can be attributed to the agent’s real self. The latter ability, in contrast, has to do with accountability: in order to hold an agent accountable for an action, it is not enough that the action can be attributed to the agent’s real self. The agent must also be responsible for her ends or values (2004: 272f., 282). Finally, Susan Wolf suggests that in order to be morally responsible, an agent must not only have control in the sense of having 11

12 13

A terminological note: Leibniz sometimes says that we possess ‘mastery’ (empire), and sometimes he claims that we are ‘masters’ (mai[s]tres) of our actions or inclinations. That Leibniz uses empire as the abstract noun corresponding to mai[s]tre becomes clear, for instance, in T 404: “our mastery [empire] over volitions can be exercised only in an indirect way, and . . . one would be unhappy if one were sufficiently master [maître] in one’s domain to be able to will without grounds [sujet].” More precisely, we are not directly masters of nonvoluntary actions. As we will soon see, Leibniz holds that we sometimes have indirect control over nonvoluntary actions. Surprisingly, mastery is hardly discussed in the secondary literature on Leibniz. Rutherford mentions it in passing (2005: 177), and so does Phemister (1991: 30f., 2005: 249). I am not aware of any scholar who discusses it in detail; Seidler (1985) is an exception, but he is interested almost exclusively in what I call ‘indirect mastery’ (he calls it “moral therapy”), and he does not ultimately think this kind of mastery is compatible with Leibniz’s determinism.

152

Control, Weakness, and Compulsion

a potentially effective intelligent will or in the sense of acting “from a fully developed real self.” The agent must also possess a more fundamental control: she must be “responsible for that real self” or “in control of who she ultimately is” (1990: 44, 10). Possessing only the former means possessing “a merely superficial species of control” that is not sufficient for genuine or “deep” responsibility (1990: 14, 43). I understand Wolf, Watson, and Mele as drawing roughly the same distinction: an agent has control in the first sense whenever her actions reflect her real self or her rational desires and values, and she has control in the second sense if she is additionally responsible for, or in control of, her real self or her rational desires. It will become clear in the present chapter that Leibniz is often concerned only with control in the first sense: the ability to act on the basis of one’s rational desires. Yet, he has an account of control in the second sense as well. In addition to the ability to act in accordance with our values, he thinks we also at times possess the ability to shape our moral character and adopt more rational ends. The former is what Leibniz calls ‘direct control,’ while the latter is one aspect of what he calls ‘indirect control.’ 1.1

Direct Control

Let us turn to Leibniz’s positive account of control. As already mentioned, Leibniz distinguishes between direct and indirect control, which he sometimes describes as the control that we have over our actions and the control that we have over the will itself, respectively. Possessing mastery over one’s actions means, for Leibniz, “that we choose the thing that we will [nous choisissons ce que nous voulons]” (“Conversation about Freedom and Fate,” Gr 482/SLT 99, translation mine). This kind of mastery is direct: we directly choose our actions. Our control over our will, on the other hand, cannot be direct in this way because what we directly choose are actions, not choices. As Leibniz points out in the same text, “we do not will to will, but to do and have; we do not choose our wills, because this would be on the basis of other wills, and so on to infinity, but we choose the objects” (Gr 482/SLT 99; translation modified; see also T 301). Hence, if possessing direct control over something means choosing it by an act of will, then we cannot have direct control over all of our acts of will, on pain of an infinite regress. Yet, Leibniz does think that we sometimes possess indirect control even over our volitions, as we will see later. To better understand Leibniz’s notion of direct control, note that he often associates it closely with spontaneity. He says, for instance, that our mastery “results from spontaneity joined to intelligence” (T 301; similarly in T 34; 291; “Conversation about Freedom and Fate,” Gr 482/SLT 99). That seems antecedently plausible. When we are determined by something else, we are clearly not in control; spontaneity or self-determination is therefore a necessary

Control

153

condition for control. Since the passage describes mastery as resulting from spontaneity combined with intelligence, the type of spontaneity on which we need to focus here is rational spontaneity.14 To see the connection between rational spontaneity and mastery, we can simply look at what Leibniz opposes to mastery: he frequently contrasts it with servitude (French: servitude, Latin: servitus) or bondage (French: esclavage) to one’s passions. He tells us, for instance, that “there is in the soul not only an order of distinct perceptions, which forms its mastery, but also a series of confused perceptions or passions, which forms its bondage” (T 64). Accordingly, Leibniz sometimes talks of “mastery over our actions” (e.g. T 34) and sometimes of “mastery over the passions” (e.g. T 337). It is by exerting control over our passions that we are in control of our actions. In another text, he describes this bondage or servitude in more detail: The more we act according to reason, the freer we are, and there is so much more servitude the more we act in accordance with the passions. For the more we act according to reason, the more we act according to the perfections of our own nature, and insofar as we allow ourselves to be carried away by passions, we are slaves to external things, which act upon us. (“On Freedom and Spontaneity,” G 7:110/SLT 94; emphasis omitted)

We are hence in servitude to the extent that passions determine our actions, and we are masters to the extent that we are acting on the basis of our rational inclinations. Mastery just is rational spontaneity.15 It is important that in the passage just quoted, Leibniz associates being “carried away by passions” with serving external things. This is a claim that Leibniz repeats in other texts, for instance in the Theodicy: servitude “comes from outside, it leads to that which displeases, and especially to that which displeases reason: the force of others and our own passions make us slaves” (T 228). As we already saw in Chapter 2, Leibniz frequently talks of passions and nonrational inclinations as, in a sense, external; they often prevent us from acting the way we reflectively want to act. We need to overcome these inclinations in order to act rationally. In yet other texts, he notes that mastery requires finding “ways of fighting against” our passions (NE 196); insofar as we follow passions, customs, and unthinking impulses, we are “subject to external things” (“Conversation about Freedom and Fate,” Gr 481/SLT 98). Even though all of 14 15

See Chapter 2 for an in-depth discussion of this type of spontaneity. One important difference, however, is that ascribing mastery to an agent implies that the agent possesses passions, while ascribing rational spontaneity to an agent has no such implications. As a result, we can ascribe rational spontaneity to God, but not mastery, at least not in the way described here. God does not have passions, and therefore we cannot describe him as exercising control over his passions. Leibniz raises this issue himself in T 337. Yet, perhaps we can describe God as controlling his actions rationally and ascribe mastery in that sense to him. In at least one text, Leibniz in fact states that even divine freedom is opposed to bondage (“Conversation about Freedom and Fate,” Gr 481/SLT 98).

154

Control, Weakness, and Compulsion

these inclinations originate within the agent’s soul, some are external in a certain sense, namely by being external to the agent’s real or rational self. This is further evidence that mastery consists in rational spontaneity, or in selfdetermination by one’s rational self. Leibnizian mastery is a type of control that is worth having. To see this, imagine a situation in which a rational agent’s deliberations and judgments are entirely inefficacious. Take, for instance, a woman whose friends convince her to go bungee jumping. Immediately after jumping, while plummeting, she starts wondering whether it was really such a good idea.16 Of course, these second thoughts will not make a difference. Once the woman has jumped, she is no longer in control of what happens to her; her reasoning and deliberation are no longer capable of making a meaningful difference. Such lack of control is unfortunate, but luckily, we do not find ourselves in that kind of predicament all the time: in many situations, our reasoning does have the power to influence our fate. There could be rational agents who never have control. Imagine an agent who is always entirely at the mercy of her unconscious, nonrational inclinations, even though she possesses an intellect and hence has beliefs about what actions are good. This agent would be in the extremely unfortunate position of making judgments about the best course of action but being unable to act on those judgments. She would be completely controlled by her passions and nonrational inclinations. There may not be a creature like this in the actual world, but such creatures seem possible. One could describe them as creatures with perfectly weak – or, rather, completely inefficacious – wills: their wills are so weak that these agents never act in accordance with what they judge to be best, except accidentally, when their passions happen to be aligned with their rational judgments. Alternatively, and perhaps more accurately, we could describe these creatures as lacking wills altogether – that is, as lacking the general inclination to act in accordance with reason. Leibniz does appear to acknowledge that a creature with an intellect but without a will is conceivable. He says in Theodicy §34 that “if we had a judgment that was not accompanied by any inclination to act, our soul would be an understanding without will.” Moreover, as we saw in Chapter 5, he often stresses that reasons generally only incline the will without necessitating it. I take this to mean that there is no metaphysically necessary connection between perceptions of reasons to act and the corresponding appetitions (see Jorati 2014a: 785ff.); instead, it is a contingent fact, grounded in our natures, that certain perceptions incline us in particular ways. There is nothing logically or metaphysically incoherent about a creature that is inclined in a completely 16

I am basing this on Dennett’s much more morbid example of a person jumping off the Golden Gate Bridge to commit suicide (1984: 104).

Control

155

different way by the same perceptions. More specifically, it is logically or metaphysically possible for a creature to make rational judgments about what it should do without developing rational appetitions for those actions; this creature might be motivated only by its nonrational perceptions. If God did not create any such creatures, it is merely because they are not members of the best possible world, not because they are metaphysically impossible. Even more importantly, with respect to some of our actions, we are like these creatures. Sometimes we know what is right but find ourselves under the influence of such strong or vivid passions that we fail to do what we know to be right. This can take at least two different forms for Leibniz: weakness of will and compulsion. I will discuss both of them in detail later. But let me give a few examples that may throw light on the notion of mastery. An extreme example of a rational agent who has very limited control over her actions is someone suffering from a phobia or another anxiety disorder. Such an agent may be fully aware that her behavior is irrational but nevertheless cannot do anything about it. Similarly, an agent may be so consumed by anger, grief, fear, or some other passion that she is impervious to rational considerations. In yet other cases, agents may be constrained physically from acting in accordance with their rational judgments, for instance when someone is locked into a room and cannot keep a promise to a friend. Because of Leibniz’s spontaneity thesis, even these physical constraints must at bottom consist in unconscious perceptions and insensible inclinations within the agent herself. In fact, for Leibniz, all of these examples are instances of the same pattern:17 a failure of rational control due to the relative strength of nonrational inclinations. In these examples, the agents are in extremely frustrating situations precisely because they lack control over their actions. They are, instead, controlled by nonrational and usually unconscious inclinations, with which they do not identify and of which they do not approve. As already seen in Chapter 2, Leibniz seems to subscribe to something like the Platonic model of the self – a theory that identifies the agent with her values or with the rational elements of her soul. In the examples under consideration, the rational aspect of the agent’s soul, which is her real self, is taken hostage by the imperfect, nonrational elements of her soul. When agents act voluntarily, in contrast, they are in control – the rational aspect of the soul is setting the agenda. All actual human beings possess this kind of control to some degree because they have wills and intellects. This also sets them apart from lower animals, which are completely determined by their passions or, as Leibniz puts it, “swept along by involuntary perceptions” (NE 180). We have the power to resist our passions in part because reason allows us to take into consideration not only 17

There are, of course, other crucial differences between these cases. These differences, as we will see in Chapter 7, matter for ascribing moral responsibility.

156

Control, Weakness, and Compulsion

current or momentary goods but also future or lasting ones; animals are motivated only by the former. We grasp goodness in general and can employ this notion in deliberating about the best course of action. Instead of “rushing straight at a present pleasures,” reason makes it possible to do what we judge to be best overall (NE 189). For instance, even when I am extremely hungry, I may be able to refrain from eating the junk food in front of me because I know that eating it will be bad for me in the long run. A nonrational animal, in contrast, cannot do that. Leibnizian mastery is, of course, not a kind of control that allows us to rise above all determination. Leibniz thinks that the ability to act in an undetermined fashion is not only something we lack but would not even be worth having. Why would any rational creature desire to be able to act contrary even to the very best reasons that are available to it? For Leibniz, the influence of reasons or rational judgments is not something over which we should want to have control; that ability would not be an advantage. As seen in Chapter 5, Leibniz calls the determination by rational judgments a “moral necessity” and stresses that it is “happy and desirable” (RH 3; see T 175; 191). God is subject to this kind of happy necessity, and the more we are like God in this respect – or in any respect, for that matter – the better. What is worth having, then, is not the ability to act contrary to good reasons but rather the ability to act rationally and to fight any nonrational inclinations that one may possess.18 For instance, being able to resist a very strong urge to insult or punch someone who is provoking you is a good thing. This, Leibniz would say, is the kind of control that we should want to have. Luckily, we frequently have it: we can often exercise control over our passions and do the rational thing. 1.2

Indirect Control

All human beings have wills and hence the capacity to be masters over their passions in the sense that they can act in accordance with reasons, against their passions. Yet, human beings differ from one another with respect to how often and how well they exercise this valuable capacity. As Leibniz’s spokesperson in the New Essays points out, “one [person] may be able to rise above impressions by which another would be swept along” (NE 177). This, in turn, might make you wonder whether we have any control over this ability. It may seem, after all, that it is not enough to possess the general capacity for governing one’s actions rationally if it is not up to the agent whether she 18

As seen in Chapter 5, it is central to Leibniz’s theory of freedom that an alternative action is metaphysically possible. In that sense, it is possible even for God to act contrary to reason. However, this is an extremely thin type of ability to do otherwise and should not be conflated with the ability to control the influence of rational judgments. See below and see Chapter 7 for a more detailed discussion of the different types of ability to do otherwise.

Control

157

exercises that capacity in any given situation. Possessing the general capacity may be a good thing, but it is not of much use to an agent who rarely acts rationally and usually follows her nonrational inclinations. Nor does it seem to benefit an agent who is currently so consumed by passions that she would not be swayed by even the best reasons. Should an agent not also have some control over whether her action is rational or not – that is, over whether she succeeds or fails in controlling her passions? Or, to put it differently, should agents not also have control over their characters and how good they are at keeping the passions at bay? This is where Leibniz’s claim that we possess indirect control over our will – and hence over our voluntary actions – becomes relevant. Among other places, he affirms this indirect mastery in the following passage from the Theodicy: “It is true that we are not directly the masters of our will . . . Yet, we do have a certain power over our will, because we can contribute indirectly to willing another time that which we would like to will now” (T 301; see also T 404; NE 197; RH 5). Later in the same text, Leibniz tells us more about how we can indirectly influence our volitions: in order to be “master in his domain . . . [man] has only to prepare himself in good time to oppose the passions, and he will be able to stop the vehemence of the most furious” (T 326; similarly in a letter to Coste, December 19, 1707, G 3:403/AG 195). We can take steps ahead of time, then, that will help us resist even extremely powerful passions. By taking these steps, we can bring it about that “when the time comes we shall will . . . according to reason’s decrees” (NE 196; see also NE 182). Indirect mastery can be a long and strenuous process. In the Theodicy, Leibniz tells us that it is “not in a moment and by a simple act of will that one corrects oneself and that one acquires a better will” (T 328). Yet, by finding the right strategies and working hard on oneself, it is possible to act rationally much more frequently. This means that we can often be said to possess indirect control over whether we act rationally. After all, how we act in the present is in part a consequence of the ways in which we have acted, or failed to act, in the past. In many cases, an agent fails to control her passions in the present because she failed to acquire the requisite dispositions or habits in the past. For instance, she might lose her temper today because she failed earlier to develop adequate strategies for dealing with provocation. Had she exercised indirect control, she would have remained calm. Indirect control, then, is something we achieve by developing certain character traits and habits. We control our actions indirectly by controlling our character. As Leibniz puts it in a letter to Electress Sophie, agents can “adapt themselves little by little . . . by a carefully chosen and regulated practice” to good principles. In this way, Leibniz claims, it is possible to “change even one’s temperament” (February 6, 1706, G 7:569/SLT 84). This ability to indirectly control our actions by acquiring better dispositions is extremely important for

158

Control, Weakness, and Compulsion

accountability, as we will see in more detail in Chapter 7. Leibniz’s account of indirect mastery therefore constitutes an account of control in the second sense discussed earlier: it gives agents control over who they ultimately are and what their values are. Steps we can take in order to exercise indirect control include “choosing to be attentive to certain objects and . . . accustoming ourselves to certain ways of thinking” (letter to Coste, December 19, 1707, G 3:403/AG 195). Moreover, Leibniz’s spokesperson in the New Essays tells us that we need to become “accustomed to proceeding methodically and sticking to sequences of thoughts for which reason . . . provides the thread” (NE 196). When we are “in a good frame of mind,” he advises, we ought to make firm resolutions for the future (NE 187). We can benefit from good company, reflection, and “a lively representation of good and evil” (“Memoir for Enlightened Persons” §14, A 4.4.615/Riley 106); we can also get used to “touching on certain topics only in passing” (NE 196). In addition, we can acquire “contrary desires and inclinations” and even create “diversions” for ourselves – that is, we can occupy ourselves with other things (NE 196). These diversions can help us avoid potentially corrupting influences as well as situations in which the temptation to follow the passions is particularly strong. In a wonderful passage from the New Essays, Leibniz’s spokesperson gives some instructive examples of diversions that can keep us on the path of reason: To dangerous interests we will oppose innocent ones like farming or gardening; we will avoid idleness, will collect curiosities, both natural and artificial, will carry out experiments and inquiries, will take up some compelling occupation if we do not already have one, or engage in useful and agreeable conversation or reading. (NE 187)

It is clear, then, that there is a whole range of things one can and should do when one is not currently affected by powerful passions in order to exercise indirect mastery over one’s less noble impulses at future times. It may seem that this is not ultimately a satisfactory response to the objection that direct mastery is not very valuable unless we also possess control over our exercise of that mastery. After all, the same problem seems to reappear with respect to the measures we could have taken in the past to control our present action. If you took steps in the past to ensure that your rational inclinations are stronger than your passions today, you took those steps only because the inclinations for those steps happened to be sufficiently strong. Leibniz even acknowledges this in the Theodicy: “often creatures lack the means of giving themselves the will that one ought to have; often they even lack the will to make use of the means which indirectly give a good will” (T 120). Leibniz may seem to be left with a vicious regress and, hence, without a promising account of mastery. Yet, this objection is based on a misunderstanding of what the notion of indirect control is supposed to accomplish. It is clearly not Leibniz’s goal to

Control

159

show that we can ultimately escape determination. Perhaps, instead, it is merely his goal to show that we can often rise above nonrational inclinations and act contrary to them, even though this “rising above” is itself determined. Leibniz is an unapologetic determinist, after all, and he holds that the power to act against all motives would not be worth having. According to him, “one would be unhappy if one were sufficiently master in one’s domain to be able to will without grounds [sujet], without rhyme or reason” (T 404). In fact, he continues, complaining about one’s determination by reasons “would be to reason like Pliny, who finds fault with the power of God because God cannot destroy himself” (T 404). For Leibniz, neither the power to destroy oneself nor the power to act against the best available reasons is a perfection. It is, however, desirable for one’s actions to be determined by what one judges to be best rather than by other factors that are independent of one’s rational judgments. The reason why indirect control over our future actions is crucial, then, is not that it allows us to escape determination altogether. Instead, the reason must be that it gives us a way to widen the reach of our mastery and therefore of our rational self-determination. Once you are in the grip of powerful passions, your rational faculties may be impotent; the situation might be such that no matter how good your reasons are to perform action x, you will perform y instead because y is backed by extremely strong passions. However, there may have been a prior time at which you were not yet in the grip of these passions and were able, based on reasons, to prevent this situation from occurring in the first place. When our minds are calm, after all, it is easy to act in accordance with reason. What is valuable about indirect control, then, is that it allows us to act rationally more often. Indirect control gives us a way to avoid at least some situations in which we are controlled by our passions. If this is correct, it makes Leibniz’s theory similar to Susan Wolf’s theory: for Wolf, what matters is not just that we can act in accordance with our values but also that we can form our values “on the basis of what is True and Good” (1990: 75). We can hold agents morally responsible when they fail to act in accordance with reason if they could have acted in accordance with reason (1990: 68f.), which often means that they could have acquired better values or a better character (1990: 75). To understand Leibniz’s account of mastery better, it is useful to compare it to John Martin Fischer and Mark Ravizza’s notion of reasons-responsiveness. According to them, moral responsibility requires what they call ‘guidance control,’ which in turn requires that the agent is acting – directly or indirectly – on a mechanism that is at least moderately responsive to reasons.19 We do not need to get into the technical details here; a rough sketch will suffice. 19

More precisely, the agent must be acting on a mechanism that is “regularly receptive to reasons” (that is, exhibiting an understandable pattern of recognizing reasons; see 1998: 81) and that is at least weakly reactive to reasons (that is, capable of translating reasons into choices and subsequently behavior in some possible scenario; see 1998: 69, 63, 81).

160

Control, Weakness, and Compulsion

A moderately reasons-responsive mechanism is a process, such as practical reasoning or deliberation,20 that is both receptive and reactive to reasons: the mechanism can register reasons and be influenced by them. Examples of mechanisms that are not adequately reasons-responsive, on the other hand, include post-hypnotic suggestion (1998: 35), literally irresistible urges (1998: 48), direct stimulation of the brain by another agent to elicit certain types of behavior or emotion (1998: 258), and even extremely strong virtuous dispositions (1998: 50). In those examples, the mechanism leading to the action is not receptive and reactive to reasons in the requisite manner. The agent is going to act in a certain way no matter whether she has strong reasons to do otherwise. Translated into Leibniz’s terminology, such an agent lacks direct control. Fischer and Ravizza argue that when we are acting on a mechanism that is not even moderately reasons-responsive – that is, when we lack direct control – we are nevertheless morally responsible for the action if we previously had control over something in the history of the action that is “suitably related to the action” (1998: 50, emphasis removed). They call this “a ‘tracing’ approach” (ibid.). For instance, if someone has developed such a strong virtuous disposition that the mechanism leading to her virtuous actions is no longer sufficiently reasons-responsive, she is nevertheless responsible “insofar as reasonsresponsive sequences issued in [her] cultivation of the virtue” (1998: 50). The same holds, mutatis mutandis, with respect to responsibility for morally wrong actions (1998: 257). Hence, Fischer and Ravizza acknowledge the importance of something analogous to Leibniz’s indirect mastery. Consider again the case of an agent who is in the grip of extremely strong passions – so strong, in fact, that she will act in accordance with these passions no matter how good her reasons are for doing something else. Suppose, for instance, that she is completely consumed by rage. Fischer and Ravizza would describe this situation as a case in which the agent lacks guidance control because she lacks even moderate reasons-responsiveness. Yet, it matters greatly whether there was a point in the agent’s history at which she was able to prevent that situation, a point at which she was responsive to reasons and not completely at the mercy of her rage. If there was an earlier action that arose from a reasons-responsive mechanism and that contributed in a foreseeable way to her current situation, or if she earlier possessed guidance control but failed to perform an action that could have prevented the all-consuming rage, she is responsible. Leibniz, similarly, would describe the situation of the agent consumed by rage as a case in which the agent lacks direct control because her passions are too strong. Moreover, like Fischer and Ravizza, Leibniz claims that it matters 20

Note however that for Fischer and Ravizza, deliberation or practical reasoning is not the only moderately reasons-responsive mechanism (1998: 86).

Control

161

greatly whether at an earlier time she could have exercised indirect mastery and prevented this powerful rage from controlling her. Perhaps she could have started a stamp collection or a vegetable garden, which would have made her more even-tempered. I will argue in Chapter 7 that if the agent could have exercised indirect mastery, we can hold her morally responsible for the present action even though she lacks direct control. This solution, both for Leibniz and for Fischer and Ravizza, relies on a distinction between different senses in which a contrary action is possible.21 Leibniz, as seen in Chapter 5, distinguishes between moral, physical, and metaphysical possibility. He holds that even though it is metaphysically possible for you to act contrary to all of your inclinations, it is not physically possible to do so because it is in your nature to follow the strongest inclination. Likewise, for a perfectly wise mind, it is metaphysically but not morally possible to go against what it judges to be best. Fischer and Ravizza’s theory introduces an additional sense in which one can (or cannot) do something, and this sense also seems to be present in Leibniz: when an agent is acting on a mechanism that is not even moderately reasons-responsive, she cannot respond to reasons in the way that she can at other times. If the mechanism is leading her toward an irrational action, we can say that she cannot act rationally. She will act irrationally no matter how good her reasons are to do something else. Leibniz appears to acknowledge something similar: passions can become so powerful that the agent will not do the rational thing, no matter how strong her reasons are.22 Consider two cases. First, take an agent who acted out of anger but was only somewhat angry. Like all natural actions, this action was physically necessary: the anger was in fact the strongest inclination and thus, given the laws that govern her inclinations, the agent had to act in accordance with it. Yet, Leibniz holds that this agent may have had the ability to control her anger directly, through a sufficiently powerful rational appetition. If the agent had tried harder at the time, she would have acted differently.23 In many situations 21

22 23

Note that Fischer and Ravizza think (and Leibniz agrees) that in one crucial sense, alternative actions are never possible. They do not want to deny any premise in the so-called Consequence Argument. Their theory is supposed to be compatible with causal determinism. What is required for responsibility, on their view, is simply that the mechanism that was in fact operating could have led to a different action instead – not in the same situation, but in some possible situation (1998: 51). That is the sense in which an alternative action is possible. This might be what Leibniz has in mind in a passage from the Theodicy, already quoted earlier: “often creatures lack the means of giving themselves the will that one ought to have” (T 120). Leibniz claims that responsibility requires that it is possible to “abstain even from that sin which we actually commit, if we summon a sufficiently strong effort” (CD 98). Counterfactuals like this one should not be a major problem for Leibniz, even if he is ultimately a necessitarian of sorts. After all, he does not seem committed to a possible-world-semantics for counterfactuals, nor to a counterpart theory. Instead, he can for instance evaluate counterfactuals on the basis of general, deterministic laws that govern the unfolding of the actual world. Even if, taking

162

Control, Weakness, and Compulsion

in which an agent does not exercise her direct mastery, she could have done so because her passions were not too strong to be controlled directly. Second, take an agent who acts out of an all-consuming rage. This second agent, once possessed by rage, may entirely lack the ability to control her passions directly; even if she were aware of extremely good reasons to do otherwise, she would nevertheless act on her rage. The only remedy in such cases, as already seen, is to take steps beforehand to prevent such situations from arising – that is, the only remedy is to exercise indirect mastery. The sense in which the agent consumed by rage cannot do the rational thing, then, is stronger than the sense in which the somewhat angry agent cannot do the rational thing. The somewhat angry agent would do otherwise if her rational inclinations were stronger. The agent consumed by rage, on the other hand, would not do otherwise even if her rational inclinations were stronger.24 The notion of indirect mastery, like Fischer and Ravizza’s tracing approach, appears to introduce yet another kind of possibility. These notions give us a way to say that even in extreme cases like the all-consuming rage, there is sometimes a sense in which the agent could have done otherwise.25 For instance, suppose that the agent has been aware for a long time that she is terrible at anger management but has nevertheless not taken any steps to remedy this. Suppose further that if she had taken those steps, she would not have lost her temper today. In this indirect sense, then, it was possible for her to do the rational thing.26 If she had taken steps to become better at controlling her anger, she would not now be so consumed by rage that her rational self is temporarily powerless.27 2

Akrasia

Let us now turn to Leibniz’s views on weakness of will, or akrasia. This notion is closely connected to control. As we will see, akrasia is a particular type of

24 25 26

27

everything into consideration, it was not possible for an agent to have different inclinations, there are facts, grounded in the general laws that govern inclinations, about the influence that different inclinations would have had. By way of analogy, a necessitarian can observe an interaction between two billiard balls and calculate based on the impact laws what would have happened if – per impossibile – the first ball had moved faster. I will spell out this distinction in much more detail in Section 3. Of course, there are some situations in which the agent would not have acted rationally even if she had taken steps beforehand to become as rational as she can. Interestingly, Aquinas also appears to acknowledge these different senses of the ability to do otherwise. He says that there are some cases in which “the passion does not take away the use of reason altogether” and in which “reason can drive the passion away, by turning to other thoughts.” Yet, there are also cases in which the passion does “take away the use of reason altogether.” Even in the latter cases, however, agents might be blameworthy if they are responsible for letting the passion get so strong in the first place – that is, “if the will could have prevented it, but did not” (ST IaIIae q77 a7 corp.). In Chapter 7, I will explore the question of whether the ability to do otherwise in this sense might be sufficient for moral responsibility.

Akrasia

163

failure to control one’s passions and to act in accordance with one’s reasoned judgment, which means that akrasia is a failure of mastery and hence an instance of what Leibniz calls ‘bondage.’ In fact, akrasia and control are often viewed as two sides of the same coin.28 Aristotle, for instance, contrasts akrasia, which one can translate as ‘lack of mastery,’ with enkrateia (‘mastery’; see e.g. Nicomachean Ethics 1145a35ff.). However, it will turn out to be useful to distinguish between two types of cases in which the agent fails to master her passions: one is more properly called ‘weakness of will’; the other one is more properly called ‘compulsion.’ The agent who acts out of an anger that she had the power to control is an example of the former. The agent who acts out of an all-consuming rage that she did not have the power to control – at least if she could not have controlled it indirectly – is an example of the latter. Some interpreters think that Leibniz’s moral psychology makes akrasia impossible.29 A Leibnizian agent, they think, can never knowingly act in a way that goes against what the agent takes to be best, all things considered. And it is easy to see why someone might think that. On the standard interpretation, which I called the ‘apparent good interpretation’ in Chapter 3, Leibnizian agents always act in accordance with the balance of all of their appetitions for the apparent good. On this reading, what an agent does is simply a function of what seems good to her. And even on the second, less widely accepted interpretation I considered in Chapter 3 – that is, the objective good interpretation – all desire-based actions are governed by “desire teleology,” and hence in accordance with what seems best to the agent (see Rutherford 2005). Both interpretations, then, strongly suggest that akrasia is impossible: whatever an agent does knowingly or based on her conscious appetitions, she does because it seems best to her. In other words, both traditional interpretations suggest that Leibniz’s moral psychology bears a strong resemblance to Socrates’s intellectualist position according to which what we typically call ‘weakness of will’ is in fact a type of ignorance. The agent is mistakenly judging something to be best that is not in fact best, often because the judgment does not take the value of future or long-term goods into consideration sufficiently and overvalues immediate goods (see Watson 2004: 37). For instance, at the moment that the recovering alcoholic takes a drink, she judges having alcohol to be better than continued sobriety, even though she knew better right beforehand and regrets the action soon thereafter. Even though some passages may suggest otherwise, I contend that Leibniz does not endorse the Socratic position. It is possible, on my reading, for a Leibnizian agent to knowingly act against her better judgment; she can 28 29

For contemporary examples, see Mele (1995: 5) and Watson (2004: 41). Examples of such interpreters include Imlay (2002: 85), Jolley (2005: 178), Craig (1996: 56), Nachtomy (2007: 161), Hostler (1975: 65 and 31f.), and Youpa (2013, §3).

164

Control, Weakness, and Compulsion

knowingly act against what seems best to her, all things considered. While a few interpreters have already put forward some elements of my interpretation of Leibnizian akrasia (especially Davidson 2005, Vailati 1990), I can provide a much fuller and more systematic picture based on my account of appetitions from previous chapters and based on my discussion of control from the current chapter. 2.1

Defining Akrasia

First of all, we need to determine what exactly akrasia is supposed to be. So far I have described it as knowingly or intentionally acting against one’s better judgment, or against what seems best all things considered. This is slightly different from the way in which some contemporary philosophers characterize akrasia: they sometimes require that the akratic action be a free action (e.g. Mele 2012: 8, Davidson 2001 [1970]: 22n1, Stroud 2014). If that is the correct definition, however, there is no prospect whatsoever for a Leibnizian account of akrasia,30 since Leibnizian free actions must by definition be in accordance with the agent’s current judgment concerning what is best. Leibniz must, it seems, deny the possibility of what we might call ‘free akrasia’ because any actions that go against the agent’s current judgment concerning the good must ipso facto be unfree actions. For instance, on Leibniz’s view, if I eat a large chocolate bar while judging it best not to, I am not eating it freely. For a number of reasons, in the Leibnizian context, I do not think we should require akratic actions to be free. First of all, discussions of akrasia in ancient philosophy do not define it in terms of freedom31 or choice. Take Aristotle, for instance, for whom akrasia consists in acting against reason as a result of feelings or passions. A weak (as opposed to an impetuous) akratic agent is an agent who makes a choice (prohairesis) based on deliberation but who acts contrary to this choice, typically because of a conflicting appetite for pleasure (see Nicomachean Ethics 1150b19ff.). Such an agent knows what she is doing and for what end she is doing it – and thus the action is what Aristotle calls ‘voluntary’ (hekōn) – but she is not acting in accordance with her choice (1152a15ff.). It would be odd to claim that Leibniz is committed to denying akrasia even if he can accommodate akrasia as Aristotle understood it. Second, I suspect that the main reason why some contemporary philosophers require akratic actions to be free is that agents are typically thought to be morally responsible for their weak-willed actions. A compulsive or 30 31

At least there would be no prospect for synchronic akrasia, which is often viewed as the only genuine kind of akrasia. More below. That might turn out to be trivially true because these ancient figures may not have had any analogue to our modern notion of freedom. The fact that Aristotle denies that the akratic action is an action that the agent chooses, however, does seem relevant.

Akrasia

165

inadvertent action, after all, would not count as akratic since akrasia is supposed to be a moral failing. But in the eyes of many contemporary philosophers, uncompelled intentional actions for which the agent can be held responsible just are free actions. Hence, it makes sense for them to define ‘akrasia’ as free action against the agent’s better judgment.32 Yet, it is not obvious that, for Leibniz, uncompelled, intentional actions for which one can be held responsible are always free. I will argue in Chapter 7 that Leibniz is willing to hold agents morally responsible for actions that they did not perform freely. Moreover, Leibniz does not appear to think that all actions that we would call intentional and uncompelled are free. The Leibnizian definition of freedom, after all, is extremely demanding: I φ freely only if I deliberated about whether to φ and concluded that it would be best to φ. Hence, very few of our actions are free, on Leibniz’s definition. If I go to the kitchen to get a glass of water because I am thirsty, I am not acting freely unless I deliberated about whether getting the water would be the best thing to do. Yet, we would typically describe my action as intentional and uncompelled, even if it was not preceded by deliberation. Hence, for historical as well as conceptual reasons, I think the question we should ask in the Leibnizian context is not, “Can we freely act against our better judgment?” but rather something like, “Can we knowingly and culpably act against our better judgment?”33 This captures the concern that is behind traditional discussions of akrasia. Those who disagree with my understanding of akrasia can simply view the remainder of this section as an exploration of the ways in which agents can or cannot act against their better judgment. That is a legitimate endeavor whether or not we use the term ‘akrasia.’ 2.2

The Case against Akrasia

Let us now examine whether it is possible for Leibnizian agents to act akratically – that is, to act knowingly and culpably against their better judgment. One of the passages often cited as evidence that Leibniz rejects weakness of will is from the Theodicy. In this passage, Leibniz discusses the stock example of an akratic agent: Medea, a character from Ovid’s Metamorphoses. Leibniz says,

32

33

I find it telling that some contemporary philosophers define akratic actions not as free actions against one’s better judgment, but as intentional or uncompelled actions against one’s better judgment; see e.g. Audi (1979: 180, 1993: 321). Donald Davidson defines it that way as well (2001 [1970]: 22) but connects it to freedom in a footnote (2001 [1970]: 22n1). I am not putting this question in terms of intentional agency because Leibniz does not, as far as I know, talk about intentional actions. Leibniz does, however, talk about cases in which an agent knows what she is doing and why she is doing it, and he even ties that to moral responsibility (see e.g. CD 98). This suggests that knowingly acting against one’s better judgment is the relevant category, which is also the way Aristotle defines ‘akrasia.’

166

Control, Weakness, and Compulsion

[F]ree will tends to the good; if it meets with evil it is by accident – it is because this evil is concealed beneath the good, and masked, as it were. These words, which Ovid ascribes to Medea, “video meliora proboque, deteriora sequor,”34 mean that the morally good [le bien honnête] is outweighed by the agreeably good [le bien agréable], which makes more impression on souls when they find themselves agitated by the passions. (T 154)

It is possible to read this as a rejection of weakness of will: apparent cases of akrasia are simply cases in which what seems wrong to the agent in one respect seems good to her in another respect, and the agent does what she judges, at the time of action, to be best all things considered. However, the passage need not be read in this way. First of all, Leibniz might merely be denying that there is akrasia in free actions. Second, the passage does not explicitly say that the agent views the action she performs as best. All Leibniz is saying is that the agent acts in accordance with the agreeably good, rather than the morally good, because the former made a stronger impression on her. Making a stronger impression does not necessarily mean seeming better, as we will soon see. Similar ambiguities affect another very rich passage that is sometimes interpreted as a general denial of akrasia. This passage is from a 1707 letter to Pierre Coste, in which Leibniz spells out his objections to the libertarian conception of freedom or, as he calls it, “freedom of indifference.” At the beginning of the passage, Leibniz says something that is perfectly consistent with the acceptance of akrasia: “God or a perfectly wise person will always choose the best that they know of [le meilleur connu] . . . The passions often take the place of reason in other intelligent substances, and we can always assert, with respect to the will in general, that choice follows the greatest inclination (by which I understand both passions and reasons, true or apparent)” (G 3:401f./AG 194). What Leibniz is saying here is that only perfectly wise agents always do what they think is best, and imperfectly wise agents, who are often influenced by their passions, follow the balance of all of their inclinations. His main point appears to be that there is determinism in all activity and, hence, no freedom of indifference. In fact, it is telling that he distinguishes between the determination of perfectly wise agents by the good and the determination of imperfect intelligent agents by their passions and reasons;35 it is not at all clear that the passions influence our actions only by making something seem better to us. Leibniz then considers a few potential objections: some people claim that “sometimes we set ourselves for the lesser option, that God sometimes chooses 34 35

The quotation is from Ovid’s Metamorphoses, book VII, lines 20f.; it means “I see and approve the better; I follow the worse.” It would have been very easy for Leibniz to say that finite intelligent agents are determined by what seems best to them. That he chose not to put it in those terms may be evidence that his picture is more complex.

Akrasia

167

the lesser good, everything considered, and that a person sometimes chooses without grounds [sujet] and against all his reasons, dispositions, and passions, and finally, that we sometimes choose without any reason determining the choice” (G 3:402/AG 194). Yet, Leibniz replies, those scenarios are clearly absurd and violate the Principle of Sufficient Reason. He concludes: Thus when God chooses, it is by reason of the best, and when a man chooses, it is the option that struck him the most. If he chooses what he sees as less useful and agreeable in some respects, perhaps it becomes more agreeable to him through a whim, or contrariness, or for similar reasons which belong to a depraved sense; these are determining reasons . . . And we will never be able to find a contrary example. (G 3:402/AG 194; translation altered)

The first sentence is clearly not a rejection of the possibility of akrasia; in fact, it again makes a distinction between perfect and imperfect agents, stating that the latter choose what strikes them most forcefully. But how about the rest of the quotation? There, Leibniz lists some reasons why an agent might choose what she views as less agreeable in some respects. These reasons are different ways in which the less agreeable option can become more agreeable to the agent. While this does admittedly suggest that Leibniz is describing cases in which an agent chooses an otherwise less appealing option because it seems more agreeable to her at the time, it does not conclusively rule out akrasia as I defined it earlier. After all, the passage is about choices, which means that it is about free actions. As I argued earlier, saying that we cannot freely do what currently seems worse to us, all things considered, does not imply that there is no akrasia. There are other passages in which Leibniz quite explicitly makes room for cases in which an agent knowingly does something that seems worse to her, as we will see. And the references in the Coste letter to strongest inclinations, as well as to options that strike the agent most forcefully, already point in that direction as well. We should not forget, finally, that Leibniz’s main goal in this letter is not to explicate his views about akrasia, but rather to argue against liberty of indifference – that is, against the possibility of actions that go against the balance of the agent’s inclinations. On my interpretation of Leibniz, there are at least two ways to act against one’s better judgment. At least one of them deserves to be called ‘akrasia.’ One way is diachronic; the other one is synchronic. The difference between synchronically and diachronically acting against one’s better judgment is the following. An action is synchronically contrary to an agent’s better judgment when the agent acts at time t1 contrary to what she judges to be best at t1. In contrast, an action is diachronically against an agent’s better judgment when the agent acts at time t2 contrary to what she judged to be best at an earlier time t1.36 To be more 36

I am borrowing this distinction from Davidson (2005: 250) and Penner (1990: 46).

168

Control, Weakness, and Compulsion

precise, diachronic akrasia is a temporary mind-change or “lapse” at the moment of action that the agent later regrets (see Penner 1990: 46f.). After all, we do not typically call an agent ‘akratic’ for correcting or improving a mistaken prior judgment or for acting on the basis of a bad judgment without later regretting it. 2.3

Diachronic Akrasia

It is not difficult to see that Leibnizian agents can diachronically act against their better judgment. Consider, for instance, the following passage from the New Essays: “It is a daily occurrence for men to act against what they know; they conceal it from themselves by turning their thoughts aside, so as to follow their passions. Otherwise we would not find people eating and drinking what they know will make them ill or even kill them” (NE 94). In the types of cases invoked here, it seems, the agents shift their attention away from their knowledge of the bad long-term consequences of the action and instead focus on the pleasure that the action will bring them in the short term. Consequently, at the time of action, their judgment that the action is not the best available option is temporarily forgotten or pushed aside. Because they are focusing entirely on present pleasures, the action seems most appealing, and they pursue it. Leibniz describes similar diachronic cases in the Theodicy: [W]hatever perception one may have of the good, the effort to act on the basis of the judgment, which in my opinion forms the essence of the will, is distinct from it. Thus, since it takes time to raise this effort to its peak, it can be suspended and even changed by a new perception or inclination which gets into its way, which turns the mind away from it, and which even sometimes causes the mind to make a contrary judgment. As a result, our soul has so many means of resisting the truth that it knows, and the distance from mind to heart is so long. (T 311)

All manner of things can interfere with the execution of a judgment, and hence it is possible to act against what one perceived or even knew to be best just a short time before. It is interesting that the passage describes cases in which the mind forms a new, contrary judgment before acting against the original judgment, because that means that diachronically akratic actions can be free. Presumably, though, they need not be free. An agent can push aside a prior judgment and then act on a passion without deliberating again and without judging that it is best to act in that way. It is controversial whether diachronically akratic actions deserve to be called ‘akratic’ at all.37 And even if acting against one’s better judgment diachronically 37

Davidson and Penner think that this is a genuine type of akrasia (Davidson 2005: 250, Penner 1990: 46f., 1997). But see Audi, who holds that it is a necessary condition for acting against one’s better judgment (and hence for weakness of will) that the agent has not abandoned the original judgment (1993: 320).

Akrasia

169

does qualify as a type of akrasia, it may not seem like a very interesting type. That we can abandon or change our judgments may seem trivial; this type of akrasia does not typically seem to be at issue when philosophers debate whether akrasia is possible. Yet, the category is important and interesting in at least one respect: it is arguably a shortcoming for an agent to easily abandon or revise her rational judgments under the influence of passions. Take, for instance, an agent who makes very admirable resolutions and decisions but forsakes them at the slightest temptation.38 That does seem like a type of weakness – a morally relevant one, even. 2.4

Synchronic Akrasia

As seen, Leibniz can easily handle diachronically akratic actions. But are synchronically akratic actions also possible for him? It may seem that they are not possible, even if it is true that not all monadic actions are explained by the agent’s perceptions of the good, as I argued in Chapter 3. After all, appetitions that do not aim at what seems good to the agent cannot give rise to akratic actions, since the changes to which those appetitions lead are not actions that the agent performs knowingly and culpably. If somebody slips you a drug that makes you fall asleep and thereby keeps you from helping a friend, as you had originally planned, you are not akratic. An akratic action has to be based on a goodness-directed appetition, such as a desire for pleasure or a desire to avoid displeasure. And if we are bracketing appetitions that are not goodnessdirected, then it may seem that what we do must, after all, be what we perceive as best. As a result, synchronic akrasia would be impossible. Yet, I contend that for Leibniz, there is a form of synchronic akrasia. Take the following example. Carla, who is afraid of heights, goes hiking in the mountains by herself and comes across a young mountain goat that is trapped on a cliff. After deliberation, Carla decides that there is an entirely safe way for her to rescue the goat, that it is irrational to be afraid of heights in this situation, and that it is best, all things considered, to help the goat. Yet, fear washes over her as she approaches the cliff. She eventually walks away, deeply ashamed of herself, aware that she should have rescued the goat. Her passions prevented her from doing what she judged to be best, even though her judgment did not change. The inclination arising from her rational judgment was not strong enough to overcome her nonrational inclinations; her volition lost the battle against countervailing inclinations. Whether or not we classify this as akrasia, it is definitely a case in which an agent acts contrary to her better judgment. Moreover, as she walks away, Carla knows 38

Aristotle compares the akratic agent to “a city that votes for all the right decrees and has excellent laws, but does not apply them” (Nicomachean Ethics, 1152a20f.).

170

Control, Weakness, and Compulsion

what she is doing and why: she is acutely aware that her fear is the reason why she is not helping the animal. Can Leibniz describe Carla’s situation in the way that I just did? It may seem puzzling that distinct and rational inclinations can ever be weaker than confused and nonrational inclinations.39 Likewise, it may seem mysterious how a competing desire, such as Carla’s desire to avoid heights, can overpower a rational judgment if the competing desire was already taken into consideration during the deliberation.40 Carla knew about her fear of heights and factored that in when she decided to save the goat: she judged that her fear was irrational and that the discomfort and effort involved in overcoming her fear was a small price to pay for the life of an innocent animal. Yet, I think Leibniz can account for Carla’s action. He holds that the strength or motivational force of a desire does not always correspond to the amount of goodness that the agent sees in the desire’s object or to the distinctness of the agent’s perception of that goodness. In other words, the conjunction of the following three states of affairs is possible: (a) an agent desires two incompatible things, x and y; (b) the agent judges x to be better than y, all things considered; but (c) the agent is nevertheless more motivated to pursue y than she is to pursue x. That allows Leibniz to provide a surprisingly elegant account of weakness of will, structurally similar to Alfred Mele’s (1987: 37, 1995: 7). The most explicit passage in which Leibniz distinguishes between the motivational force of a desire and the degree to which the desire’s object seems good to the agent is from the New Essays, where Leibniz engages with Locke’s account of weakness of will. After acknowledging some merit in Philalethes’s – that is, Locke’s – account, Leibniz’s spokesperson Theophilus says that we should not therefore reject “the old axioms that the will pursues the greatest good, and flees the greatest evil, of which it is sensible.” When we neglect the good, Theophilus continues, it is not because the will perversely pursues something bad but rather because our knowledge of the good is unable to properly motivate us in those situations (NE 185). The problem with our motivation is the following: [O]n topics and in circumstances where our senses are not much engaged, our thoughts are for the most part what we might call ‘deaf’ [sourdes] – in Latin I call them cogitationes caecae [that is, blind thoughts]. I mean that they are empty of perception and sensibility, and consist in the wholly unaided use of symbols, as happens with those who calculate algebraically with only intermittent attention to the geometrical figures which are being dealt with. Words ordinarily do the same thing, in this respect, as do the 39

40

One way to solve this problem is simply to say that while rational inclinations are stronger than individual nonrational ones, a large number of nonrational ones taken together can overbalance rational inclinations (see NE 193). That appears to be McDonough’s view (2016b: 104n31). Yet, as I will soon show, there is an even better solution. Mele also describes the puzzle of akrasia in this way (1995: 7).

Akrasia

171

symbols of arithmetic and algebra. We often reason in words with the object itself virtually absent from our mind. But this sort of knowledge cannot influence us – something vivid [vif] is needed if we are to be moved. (NE 185f.; translation altered)41

The reason why some of our knowledge about the good does not motivate us strongly, according to this passage, is that it lacks vivacity. Our thoughts about the good are often so abstract or so couched in symbols that they do not affect us very powerfully. We can illustrate Leibniz’s point with an example: take a commander in chief who is deliberating about sending ground troops into a warzone. She considers the estimated number of casualties and weighs them against other national and humanitarian interests. It makes a large difference, Leibniz seems to be saying, whether she is thinking about the casualties in abstract terms – such as, in terms of the sheer number of civilians and soldiers who are expected to lose their lives – or in much more concrete terms. If, for instance, she were to vividly imagine the faces of the people who would be killed as well as the pain that their death would cause their friends and families, it would presumably become more difficult to deploy the troops. The New Essays passage shows that, for Leibniz, rational appetitions can be weaker than nonrational ones. As a matter of fact, “[c]onfused thoughts often make themselves vividly sensed, whereas distinct ones are usually only potentially vivid” (NE 186f.). This means that the motivational force of a desire corresponds neither to the amount of goodness that the agent perceives in its object nor to the distinctness of the perception. Consequently, the fact that an agent judges some course of action to be best, all things considered, does not guarantee that she is most strongly motivated to pursue it. Even thinking distinctly about the bestness of that course of action does not necessarily help, because a confused thought might be more vivid. A nonrational desire can outweigh an agent’s rational appetite, as in the Carla example; the option that the agent judges to be best may motivate her less strongly than some other option. As a result, it is possible knowingly to act contrary to what one judges to be best, all things considered. Synchronic akrasia is possible for Leibniz. Why exactly are the appetitions arising from abstract reasoning so weak? One possibility is the following. As seen in Chapter 1, Leibniz writes to Hartsoeker that an appetition for an end can succeed only if it is accompanied by appetitions for all of the requisite means (October 30, 1710, G 3:510). Yet, 41

This passage is also discussed in the context of weakness of the will by Davidson (2005: 250, 247), Jolley (2005: 178ff.), and Vailati (1990: 219f.). Leibniz appears to hint at the same problem in T 311, in the continuation of a passage already quoted: “the distance from mind to heart is so long, especially when the understanding to a large extent proceeds only from deaf thoughts, which have little power to affect us, as I have explained elsewhere.” Even the Medea passage I discussed earlier contains a hint at this doctrine: in some cases “the agreeably good . . . makes more impression on souls” than the morally good (T 154).

172

Control, Weakness, and Compulsion

when I am thinking about the end in a very abstract way, without paying attention to all of its concrete properties, my mind may neglect not only those properties but also the means to them. As a result, my appetition for the end will be ineffectual unless I happen to have appetitions – and sufficiently powerful ones, at that – for the means. Another possibility has to do with the fact that Leibniz closely ties motivation to expected pleasure and displeasure (e.g. NE 163; see Jorati 2014a). Perhaps my judgment that something is bad or good does not result in a strong motivation unless I am acutely aware of the pain or pleasure that it will bring me. For example, suppose a woman is about to light a cigarette. She knows perfectly well that smoking is harmful, but she nevertheless goes ahead and smokes. The problem here, Leibniz might say, is that her knowledge of the harm that smoking might cause her is so abstract and hence so far removed from actual pain that it does not have a strong effect on her. If instead she were to imagine vividly what it would be like to suffer from a smoking-related disease, she would be much more likely to refrain from smoking. In fact, the graphic pictures of smoking-related diseases that cigarette companies have to print on their packages in some countries are presumably supposed to play precisely that role: they are supposed to deter people through vivid images of the negative effects that smoking can have. That abstract reasoning does not motivate us as strongly as nonrational but vivid perceptions is a major problem because it affects our moral reasoning. “[T]he finest moral precepts and the best prudential rules in the world have weight only in a soul which is as sensitive to them as to what opposes them” (NE 186), Theophilus says. Unfortunately, many souls are not sufficiently sensitive to those rules; the appeal of actions that violate the moral rules is often more vivid. The same holds for thoughts concerning the future: the future usually strikes us less forcefully than the present (NE 95).42 In general, absent goods or evils often have very little influence on us (NE 186), which means that we are liable to pursue immediate pleasures over greater goods in the distant future. Likewise, when human beings think about God, virtue, and happiness, they typically do so without explicit ideas (NE 186). As a result, the inclinations arising from such abstract or “deaf” thoughts are not very strong and can easily be overwhelmed by nonrational inclinations.43 42

43

This makes Leibniz’s views somewhat similar to Spinoza’s. According to Della Rocca, Spinoza also holds that we sometimes act irrationally because anticipated pain in the distant future motivates us less strongly than anticipated pain in the near future (1996: 239f.). It is interesting that Jolley, who thinks that Leibniz denies the possibility of akrasia, points out that invoking blind thoughts is unsatisfactory “as a strategy for denying weakness of will” because “it seems in danger of conflating the question of whether an agent has a belief, with the separate issue of how the agent holds the belief” (2005: 179). As a result, Jolley says, “it seems that Leibniz may have abandoned the principle that if an agent judges x to be better than y, he will choose x rather than y; he has retreated to the weaker principle that if the agent judges x to be better than y, and this judgment or belief is sufficiently vivid, then he will choose x rather than

Akrasia

173

Leibniz’s explanation rings true.44 A present pain is much more likely to motivate us to act than the idea of a pain in the distant future, even if it is the idea of a future pain much greater than the present one. Likewise, knowing that millions of people are suffering in a distant country and that we have a duty to help them may not be sufficient to motivate us to do anything. In contrast, seeing pictures of them, meeting them, or talking with them might. Hence, we often act contrary to our rational knowledge because our minds are more easily influenced by vivid sensory images than by rational knowledge.45 Leibniz realizes this and is, hence, less of an intellectualist than commonly thought. Luckily, according to Leibniz, we have ways to avoid situations in which our rational inclinations are thwarted by nonrational inclinations. One way is to make every effort to analyze the words and symbols we use, “getting through to [their] senses,” and thus making them as vivid as our “lively sentiments” (NE 187). One can also take steps in education to make “true goods and evils . . . as thoroughly sensible as they can be, by clothing one’s notions of them in details which are more appropriate to this end” (ibid.). In his “Memoir for Enlightened Persons,” Leibniz says that through “a lively [vive] representation of good and evil” we can come “to love the one and hate the other” (§14, A 4.4.615/Riley 106). It is perhaps by these means that we can learn to will more “vigorously” and ensure that our choices determine our thoughts “instead of being determined and swept along by involuntary perceptions” (NE 180). Moreover, we can resist inclinations and propensities that are contrary to our rational desires by turning the mind in a different direction and thereby diminishing their force (NE 195f.). As already seen, Leibniz also advises us to exercise our indirect mastery in order to prevent ourselves from acting in accordance with our passions (e.g. T 326). The solution just sketched is similar to the ones proposed by Jack Davidson and Ezio Vailati. Both of them base their interpretations on Leibniz’s notion of vivacity. Let me say a few quick words about the differences between my interpretation and theirs, which will further clarify my reading. Davidson describes Leibniz as holding that all moral evil involves “cognitive error” (2005: 251, similarly 1998: 405) or “a mistaken practical conclusion” (2005: 246). When our perception of the good is not sufficiently vivid and we do something wrong, it is because contrary passions led us to “mistake a temporary pleasure for a lasting one” (2005: 245). Moreover, Davidson

44 45

y” (2005: 179f.). Hence, Jolley appears to think that Leibniz, despite himself, acknowledges something rather close to akrasia, if not a form of akrasia, in the blind thoughts passage. Because Jolley believes that for Leibniz all actions are explained by what appears best to the agent (see Chapter 3), he cannot find this satisfactory; vivacity can be a factor, on his interpretation, only if it is translatable into degrees of apparent goodness. For a discussion of some empirical support for Leibniz’s views, see Vailati (1990: 222f.). Vailati makes a similar point (1990: 219f.).

174

Control, Weakness, and Compulsion

holds that Leibniz denies synchronic akrasia and acknowledges only diachronic akrasia, because Leibnizian agents always act in accordance with what they currently perceive to be best (2005: 250, 243). I disagree. On my interpretation, Leibniz does acknowledge synchronic akrasia, and he explains akratic actions not through cognitive mistakes but through the insufficient motivational force of the better judgment. An agent can perform a perfectly correct piece of abstract reasoning, without any error whatsoever, and the judgment can nevertheless fail to motivate the agent sufficiently. The conflicting desire can have a stronger motivational force even if the agent correctly judges the object of that desire to be less worthy of pursuit, all things considered. Vailati’s interpretation is more similar to mine than Davidson’s. He argues that Leibniz acknowledges “an essential connection between evaluation and motivation, but . . . such connection is conditional upon the presence of an adequate amount of sensitivity to the relevant evaluative judgment” (1990: 221f.). When we have only “blind” or abstract thoughts of a good, according to Vailati, we are not sufficiently sensitive to it. Our sensitivity to something good, then, does not necessarily correlate with how much we value it (1990: 222). I disagree with Vailati’s interpretation in at least two related ways. First, I do not think that there is an essential or necessary connection between motivation and evaluation, though I have not argued for it here.46 Second, Vailati appears to hold that there is a threshold of sensitivity to a good such that once we cross the threshold, evaluation and motivation match perfectly.47 On my interpretation, in contrast, there is no such threshold. The vivacity of an agent’s perception always makes a difference to the degree of their corresponding motivation. Vivacity is not just a background condition that we can neglect when it is present, but it generally has a direct impact on motivation. This difference is important for deciding whether Leibniz is an externalist about judgment and motivation.48 3

Compulsion

If Leibnizian agents can act against their better judgment, we are faced with the following question: is there a meaningful distinction between akratically yielding to a desire that is contrary to a rational judgment, and being compelled by such a desire? If there is no meaningful distinction, Leibniz’s account is much less plausible.49 Take again the example of Carla’s failure to rescue the 46 47 48 49

See Jorati (2014a), where I argue that Leibniz is an externalist about judgments and motivation. Vailati interprets Leibniz as a “moderate” internalist (1990: 221). That, at least, strikes me as the straightforward reading of the passage just quoted. See Footnote 46. See Watson, who argues that any successful account of weakness of will must be able to differentiate weakness from compulsion (2004: 44ff.). He also notes that this is particularly difficult for determinists (2004: 14).

Compulsion

175

mountain goat. If her fear of heights was so strong that it genuinely compelled her to walk away from the cliff, we would not typically describe her action as weak-willed. After all, we typically hold agents responsible for their weakwilled actions but not for actions that are genuinely compelled. A moral psychology that cannot differentiate between those two types of cases is arguably missing something very important. As a result, it makes sense to briefly investigate whether Leibniz can distinguish compulsion from akrasia. Some passages sound as if Leibniz simply denies that we are ever genuinely compelled to do anything. In Causa Dei, for instance, he tells us, In us, passion [affectus] or appetite is never so strong that an action follows from it with necessity. For as long as a man is in possession of his mind, he can always find some reason for stopping the urge, even if he is roused most vehemently by anger, thirst, or similar causes. Sometimes it is even enough to think of exercising one’s freedom and one’s power over the passions. (CD 105)

On the most straightforward reading of this passage, Leibniz is claiming that it is always possible for human beings in possession of their faculties to resist their passions, no matter how strong those passions are. If that is correct, there are no cases of compulsion, at least not in healthy adults. We are never completely at the mercy of our passions. Even when subject to an extremely strong passion, we have the power to exercise our mastery and prevent the passion from leading to an action. If that is correct, every action that is contrary to a better judgment is an instance of akrasia, provided that the agent knows what she is doing. In some ways, denying the existence of compelled actions makes sense for a rationalist and determinist.50 Yet, a flat-out denial of genuine compulsion is implausible.51 It seems clear that some passions or nonrational appetites can be so strong that we lack the power to resist them and that this should at least sometimes affect attributions of moral responsibility. This is particularly obvious in the context of Leibnizian metaphysics, where what we typically view as external forces are reduced to appetitions or passions in our own minds. The bungee jumper who has just jumped off the platform does not have the power to prevent herself from plummeting, no matter how good her reasons or how strong her desires are, and for Leibniz, this is strictly speaking because of very powerful appetitions within her own mind.52 Yet, her continued fall should surely not count as a case of akrasia. Moreover, as we saw earlier in this chapter, sometimes indirect control is the best we can get; in those situations, 50 51 52

After all, a staunch rationalist may insist that reason can in principle overcome any obstacle, and a staunch determinist may insist that all actions are equally determined. See Watson for some worries (2004: 46f.). This example nicely illustrates the extremely counter-intuitive consequences of Leibniz’s spontaneity thesis.

176

Control, Weakness, and Compulsion

we lack the power to control our passions directly. And at other times, even indirect control is not enough. As a result, there should be room for some kind of distinction between passions that we can resist and passions that we cannot resist, and therefore between uncompelled and compelled actions. One of the most intriguing passages in which Leibniz makes room for something like compulsion is from the New Essays, where Theophilus describes cases just like the bungee jumper’s: [I]t is useful to distinguish two sorts [of constraint (contrainte)]: physical, as when a man is carried to prison against his will or thrown off a precipice; and moral, as for example the fear of a greater evil, in which case the action, although forced in a way [forcée en quelque façon], is nevertheless voluntary. One can also be forced by the thought of a greater good, as when a man is tempted by the offer of a too great benefit, although this is not usually called constraint. (NE 179; translation altered)

This passage is helpful because it says that there are constrained actions, and we can presumably understand these as compelled. However, neither physical nor moral constraint is exactly what we need. Physically constrained actions appear to be actions in which the agent lacks agent spontaneity and in which the agent’s body is (ideally) forced by something external to move in a certain way. This seems to be Leibniz’s way to capture the notion of external compulsion. While this helps with the bungee jumper example, it does not appear to include cases in which an agent acts against her better judgment agent spontaneously because of an irresistible desire. Morally constrained actions, on the other hand, seem to be voluntary actions chosen under what we might describe as duress, such as the action of giving in to a blackmailer because the alternative seems much worse. This, again, is not the kind of compulsion in which we are interested in the present context; we are interested not in voluntary actions chosen under duress but rather in cases in which powerful passions or desires make it impossible for the agent to determine herself freely and rationally. In fact, Leibniz does not even appear to think of morally constrained actions as compelled in the strict sense: he calls them “forced in a way,” after all. This makes sense because he elsewhere writes that constrained actions are actions in which the principle comes from the outside and in which the agent therefore lacks spontaneity (G 7:110/SLT 94), by which he presumably means that the agent lacks agent spontaneity. By this standard, what Leibniz calls ‘moral constraint’ in the passage above is not genuine constraint. Earlier in the New Essays, however, Theophilus does mention a type of internal constraint or compulsion: he describes servitude, or the imperfection that is opposed to freedom as “an imposition or constraint [coaction ou contrainte], though an inner one like that which the passions impose,” and tells us that when one’s mind is “possessed by a great passion, . . . one cannot will as

Compulsion

177

one should, i.e. with proper deliberation” (NE 175). This passage suggests, like some other passages I examined in Section 1, that we are not always able to exercise direct control over our passions. Some passions are so strong that we can at most control them indirectly, by preventing them from getting so strong in the first place. In the present passage, Theophilus describes situations in which a strong passion prevents us from acting rationally as coaction or constraint, and it is clear that he is not limiting himself to what he later calls ‘physical constraint’ – that is, compulsion by something that is external to the agent and her body. If we take seriously the New Essays passages – even though Causa Dei seems incompatible with them – Leibniz does acknowledge that some actions are compelled. The question now becomes whether Leibniz can adequately distinguish compelled actions from weak-willed actions. One might worry that because of his determinism, Leibniz lacks the resources for such a distinction. All unfree actions are physically necessary – that is, they are determined by the agent’s nonrational strivings. No matter whether the passion is far stronger or only very slightly stronger than the countervailing rational inclination, the outcome is perfectly determined. As a result, it seems difficult for Leibniz to say that we cannot resist the passions in compelled actions but that we can in weak-willed actions. Yet, building on what we already learned about Leibnizian control, it is possible to distinguish akrasia from compulsion. We saw earlier that Leibniz thinks that in some cases we are able to exercise direct control, while in other cases we can exercise only indirect control. In fact, Leibniz seems happy to say that an agent who did not exercise direct control could have done so in some cases but not in others. The former cases, it seems, are instances of akrasia, while the latter may be instances of compulsion. What exactly is the difference between cases in which an agent could have exercised direct control, and cases in which she could not have done so? I do not know of any passage in which Leibniz explicates this, so some degree of speculation is necessary. One solution that I believe to be very much in Leibniz’s spirit is the following. There are presumably limits to how clear and vivid an agent’s perceptions can get, and to how strong their rational appetitions can be.53 Suppose you act on the basis of a very strong passion. Could you have controlled this passion and acted rationally in spite of it? That presumably depends on the strength that your rational appetition would need to have to overwhelm the passion and on your ability to summon a rational appetition of the requisite strength. If you are generally capable of summoning 53

Leibniz says explicitly that different people differ in the strength of their rational appetitions (see NE 177). Presumably, these limits are not completely fixed – we can improve our abilities through training.

178

Control, Weakness, and Compulsion

rational appetitions strong enough to outweigh this passion, there is a sense in which you could have controlled it. If this picture is correct, there are cases in which an agent has the ability to summon a rational appetition of the requisite strength but fails to do so. Of course, given Leibniz’s determinism, there is a reason for her failure to do so. Nevertheless, we can say that the agent ends up following the passion not because the passion was too strong but because her effort was too weak.54 If she had tried harder, she would have mastered the passion.55 This is different from cases in which no matter how strong a rational appetition the agent had summoned, she would not have mastered the passion. The latter might be an instance of compulsion while the former is an instance of akrasia. There is, however, another possible way to understand Leibnizian compulsion: we could reserve the term ‘compulsion’ for cases in which the agent could not even have prevented the action through indirect control. After all, if indirect control would have helped, the agent may be morally responsible for the action and may hence not be compelled, strictly speaking. In other words, if the agent could have exercised indirect control to prevent the passion from getting strong enough to overpower her rational inclinations, she might be culpable for the resulting action. The situation would be analogous to that of a drunk driver who crashes into an oncoming car that, given her intoxication, she was unable to avoid. Plausibly, the driver’s inability to avoid the other vehicle at the time does not excuse her, because she should either have stayed sober or refrained from driving. There are steps she ought to have taken earlier to prevent the situation in which she cannot help but crash into another car. The same might be true for a Leibnizian agent who cannot currently control her passions, no matter how hard she tries, but who could have taken steps earlier to prevent her passions from getting too powerful to control. And perhaps genuine compulsion occurs only when the agent could not have controlled her passions even indirectly. Both of these accounts of compulsion build on the distinction from Section 1.2 between different senses in which an alternative action is possible. According to my first suggestion, compulsion occurs when there is nothing that the agent could have done in that situation in order to bring about the alternative action. For instance, no matter how much the agent concentrates on her rational judgment or how hard she tries to act in accordance with that judgment, she does not manage to do what she judges best. It was not a situation in which the agent could have exercised her direct control. According to my second suggestion, in contrast, compulsion occurs when there is nothing the agent could have 54 55

This is roughly how Watson describes the difference between compulsion and weakness (2004: 45). See once more the passage from Causa Dei, in which Leibniz says that it is possible to “abstain even from that sin which we actually commit, if we summon a sufficiently strong effort” (CD 98).

Conclusion

179

done even beforehand, by way of indirect control, in order to bring about the alternative action. On this second interpretation, the alternative action was impossible in a stronger sense than on the first interpretation. 4

Conclusion

We saw in the present chapter that, according to Leibniz, rational creatures can exercise control over their actions. Possessing an intellect and a will enables them to pursue what they judge to be best overall and, thus, to direct themselves to actions independently of their nonrational inclinations. This, if my analysis is correct, is a type of control that is worth having and available even if volitions and actions are completely determined. To be masters in our domain, we do not need to be able to directly control what seems good to us, nor do we need to be able to act contrary to the balance of our inclinations. It is enough that we are able to determine our actions based on our rational understanding of each option’s goodness and take steps to lessen the impact of our passions. Moreover, I have argued that Leibniz’s moral psychology can accommodate akrasia surprisingly well. That is because Leibniz appears to hold that one desire can have a higher degree of motivational force than another desire, even though the agent judges the object of the second desire to be better, all things considered, than the object of the first desire. Vivid perceptions – such as perceptions that include sensory images – motivate human beings more strongly than abstract perceptions. As a result, Leibnizian agents sometimes knowingly act contrary to their better judgments. Leibniz can even distinguish between weak-willed and compelled actions because he can differentiate between situations in which an agent could have exercised control and situations in which she could not have.

7

Moral Agency

This chapter is an attempt to make sense of Leibniz’s views on moral agency and its relationship to other key notions in Leibniz’s philosophy of action. Issues surrounding moral agency have already come up a few times in this book. In Chapter 4, we examined how Leibniz employs the notion of authorship to argue that God is not morally responsible for human sins. In Chapter 5, we investigated free agency, which is closely related to moral agency. And in Chapter 6, we took a look at the ways in which rational agents can exercise control over their actions, which has several implications for moral agency. In fact, moral agency will turn out to be intimately connected to many of the concepts investigated in previous chapters – namely, to freedom, teleology, spontaneity, authorship, and control. These connections, however, are less straightforward than one might initially think. There are two separate questions that I would like to address. First, what does it take to be a moral agent? Second, what does it take to be morally responsible for particular actions? These are clearly different because a moral agent need not be responsible for all of her actions. My answer to the first question will be, roughly, that moral agency requires the possession of both an intellect and a will, which implies that only rational souls who can act freely are capable of moral agency. More specifically, a moral agent must be able to know what is right and to determine herself accordingly, motivated by the rightness of the action – she must be capable of moral knowledge and of rational teleology. The latter is often neglected. When interpreters of Leibniz discuss moral agency at all, they typically focus exclusively on the intellect or on cognitive requirements. Yet, I contend that teleology – that is, appetitive or conative aspects of activity – is an extremely important element of Leibniz’s notion of moral agency. My answer to the second question will be that in order to be morally responsible for an action, the agent must first of all be the author of that action. This, as seen in Chapter 4, can be spelled out in terms of final and formal causation. The action must also be spontaneous at least in the metaphysical sense and metaphysically contingent: it must be metaphysically possible for the agent to do otherwise. Yet, importantly, an action does not need to be free in 180

What Is a Moral Agent?

181

order to be morally evaluable, on my interpretation. While an agent is responsible for all of her free actions, she is arguably accountable for some unfree actions as well. To be blamed for an unfree action, however, the agent must have been able, through direct or indirect control, to perform a different, permissible action in that situation. This yields an asymmetry similar to the one that Susan Wolf endorses: the ability to do otherwise, in the sense of direct or indirect control, is required for moral blame but not for moral praise. Moral praise merely requires metaphysical contingency; unlike moral blame, it does not require an ability to do otherwise in the more demanding sense. 1

What Is a Moral Agent?

Let us, then, turn to the question of what it takes to be a moral agent, according to Leibniz. First, a preliminary terminological note: to the best of my knowledge, Leibniz does not use terms that could adequately be translated as ‘moral agent’ or ‘moral agency.’ As a result, someone might worry that using those terms to describe Leibniz’s views is dangerously anachronistic. Yet, Leibniz does frequently talk of substances who possess a special moral status1 and whose actions are subject to praise, blame, reward, or punishment. As we will see in Section 1.1, Leibniz holds that all minds – that is, all rational souls – are members of a moral community that is governed by God, which Leibniz calls (among other things) the City of God, the general republic of spirits, the divine society of minds, and the moral kingdom of grace.2 Animal souls do not meet the membership requirements for this moral community and thus differ radically from minds in a way that Leibniz repeatedly describes in moral terms. It therefore seems legitimate to use the terms ‘moral agent’ and ‘moral agency’ to refer to the special status and the special capacities possessed by members of that moral community. 1.1

Membership in the City of God

For Leibniz, asking what it takes to be a moral agent is equivalent to asking what it takes to be a member of the City of God, the moral community governed by God of which all minds are members. After all, Leibniz holds that all agents that possess a moral standing – and only those agents – are members 1

2

Leibniz refers to the “moral status and position [rang et employ . . . moral]” of minds in a reply to Bayle (G 4:528); in another text, he says that because God is the most perfect mind, he has a moral quality (qualitatem moralem; A 6.4.2361/SLT 204). See also a letter to des Billettes, December 4, 1696, G 7:452/WF 56; “New System,” G 4:481/AG 141. Leibniz discusses this moral community in many texts; see e.g. M 85ff.; PNG 15; “New System,” G 4:481/AG 141 and G 4:486/AG 145; “On Sin and Original Sin,” A 6.4.2361/SLT 204; letter to Wagner, June 4, 1710, G 7:531/W 507; letter to Arnauld, October 9, 1687, G 2:125/Mason 159; G 7:332/SLT 67; letter to Placcius, 1697, A 2.3.383/LGR 109.

182

Moral Agency

of God’s community. In fact, they are members precisely because of their moral standing. He writes in one of his replies to Bayle, for instance, that “rational substances have a double status or position [rang et employ]: one physical, . . . and the other moral, as a result of which they are in society with God, as citizens of the city of God” (G 4:528/WF 75). Nonrational animals, on the other hand, are excluded from this community because they lack this moral status and are not capable of society with God (letter to Wagner, June 4, 1710, G 7:530f./W 507). As a matter of fact, the laws of justice, which govern the City of God, cannot apply to nonrational animals (letter to Arnauld, October 9, 1687, G 2:126/Mason 160f.). Membership in the City of God and the capacity for moral agency are therefore intimately connected; Leibniz even goes so far as to say that “[God’s] social bond with us is the cardinal point of morality” (NE 237). Hence, it makes sense to take a brief tour of this divine city before discussing moral agency more directly. What exactly is the City of God? We saw earlier that all minds are members of this kingdom and that God is its sovereign.3 This already sets Leibniz’s City of God apart from Augustine’s, for whom only a small number of rational creatures – those who are predestined – are citizens of the City of God (see Augustine’s City of God XIV.28 and XV.1). But what precisely does being a citizen entail, for Leibniz? First, it does not entail that one no longer inhabits the same world that nonrational substances inhabit. In the “Monadology,” Leibniz helpfully describes the city as “a moral world within the natural world” (M 86). Membership in the moral kingdom and membership in the kingdom of nature, as a result, are not mutually exclusive. What constitutes Leibniz’s City of God, I contend, is the special relationship that its members have to God. Leibniz usually spells this out in terms of the laws of that divine kingdom – that is, in terms of the way in which God governs it. God governs the natural world the way that an engineer governs his machines, that is, through the laws of mechanics. In contrast, God governs the moral world like a prince or even a father, that is, through the laws of justice.4 Being a member of the divine city, it seems, just means being subject to that special set of laws. As a result, membership in the City of God – like most other memberships – brings with it both benefits and responsibilities. The laws of justice, after all, demand rewards for good actions as well as punishment for bad actions. God’s perfectly just government guarantees that all creatures who are capable of happiness and moral agency will eventually receive happiness in proportion to their moral goodness. On the flip side, they will suffer in 3 4

For references to texts in which Leibniz says this, see footnote 2. Leibniz makes this point in a large number of texts; see e.g. letter to Arnauld, October 9, 1687, G 2:124f./Mason 159; draft of “New System,” G 4:475/WF 25; PNG 15; “Specimen Dynamicum,” GM 6:243/AG 126; “Ultimate Origination of Things,” G 7:307/AG 154; “Considerations on Vital Principles,” G 6:545/L 590; letter to Bernoulli, January 13/23, 1699, A 3.8.39/AG 171.

What Is a Moral Agent?

183

proportion to their sinfulness. As Leibniz puts it in “Principles of Nature and Grace,” there is “no crime without punishment, no good action without proportionate reward, and finally, as much virtue and happiness as possible” in the City of God (PNG 15).5 Since the City of God is subject to special laws, it is natural to wonder about the relation between those laws, on the one hand, and the laws that govern bodies and nonrational substances, on the other. Understanding that relation is also important for grasping the special status of moral agents and, more generally, the status of the laws of justice. In some texts, Leibniz seems to suggest that members of the City of God are exempt from the laws of physics: “God has particular laws [for the rational soul] which make it exempt from mechanical operations of matter . . . The intellectual world (which is nothing else than the republic of the universe, or the city of God) is not subject to the inferior laws of the corporeal realm” (draft of “New System,” G 4:475/WF 25; see also “New System,” G 4:480/AG 140). Yet, it is implausible that on Leibniz’s considered view, finite minds are exempt from the laws that govern bodies in the sense that they can bring about changes that violate those laws. That would go against what Leibniz says in several other texts. Nevertheless, it is possible that Leibniz toyed with the idea at times. That would explain why he claims in a somewhat controversial text from the mid-1680s that “free or intelligent substances . . . are not bound by any certain subordinate laws of the universe, but act as it were by a private miracle” (“Necessary and Contingent Truths,” A 6.4.1519/MP 100). Yet, in many mature texts, Leibniz explicitly states that the two sets of laws are in harmony: whatever finite minds do, and whatever punishments or rewards they receive on the basis of the laws of justice, is in accordance with the laws of physics. In “Specimen Dynamicum,” for instance, Leibniz claims that God has pre-established things in such a way that the two sets of laws never come into conflict – that is, the laws of physics never dictate anything that would violate the laws of justice, and vice versa (GM 6:243/AG 126f.). Leibniz makes an analogous claim in “Principles of Nature and Grace”: in the City of God, justice in the form of punishment for all evil actions and reward for all virtuous actions “is accomplished without disordering nature (as if what God prepared for souls disturbed the laws of bodies), but through the very order of natural things, in virtue of the harmony pre-established from all time between the kingdoms of nature and grace, between God as architect and God as monarch” (PNG 15; similarly CD 46). 5

See also “Ultimate Origination of Things”: the law of justice “dictates that everyone should take part in the perfection of the universe and in his own happiness in proportion to his own virtue and to the extent that his will has thus contributed to the common good” (G 7:307/AG 154; similarly in “New System,” G 4:480/AG 140).

184

Moral Agency

If the two sets of laws are compatible with each other – which does appear to be Leibniz’s considered view – the order of the natural world and the order of the moral world are two different perfections of the one world that God has created. The world possesses one kind of perfection in virtue of the physical laws that make natural changes intelligible. And the world possesses another kind of perfection in virtue of the laws of justice that hold, without exception, for all creatures capable of being subject to them. Leibniz himself characterizes these two types of perfection in just this way in “Ultimate Origination of Things”: [T]he world is not only the most perfect naturally or if you prefer, metaphysically . . . but also . . . the most perfect morally . . . Hence the world not only is the most wonderful mechanism but is also, insofar as it consists of minds, the best commonwealth, through which there is conferred on minds as much felicity or joy as possible; it is in this that their natural perfection consists. (G 7:306/L 489)

Through pre-established harmony, God was able to create a world that is in accordance with both sets of laws and that is therefore simultaneously a perfectly intelligible mechanism and a perfectly just kingdom. There was no need for compromise – or, at the very least, almost no need. As already seen, Leibniz sometimes acknowledges occasional miracles, or divine interventions into the natural order, which God performs for the benefit of rational creatures (see e.g. LC 1.4). A closely related question about the two sets of laws, or two kingdoms, is whether the laws of justice took precedence in God’s reasoning over the laws that govern nonrational creatures. Even if the two sets of laws are compatible, after all, it is possible that the laws of the physical world are subordinate to the laws of the City of God. That is, it is possible that God instituted the former for the sake of the latter. That would mean that the laws of morality are the most important or most fundamental laws. And Leibniz suggests in several texts that this is in fact the case. In “New System,” for instance, Leibniz says that “we can say that everything else is made only for [minds], and that these tumultuous motions [of matter] themselves are adjusted for the happiness of the good and the punishment of the wicked” (G 4:480/AG 140).6 Similarly, he writes to Arnauld on October 9, 1687, that “the whole universe was created only so as to contribute to the embellishment and the happiness of that city of God” (G 2:125/Mason 160). This kind of subordination makes sense because, as Leibniz often stresses, the existence of rational creatures is an especially important part of God’s creation; without created minds who can know and admire God’s goodness, God would not possess glory (M 86). Hence, it is not 6

Later in “New System,” Leibniz also says that “bodies are made only for minds capable of entering into community with God and celebrating his glory” (G 4:485/AG 144). See also an earlier draft of “New System,” G 4:475/WF 25.

What Is a Moral Agent?

185

surprising that Leibniz sometimes describes finite minds as “the most important part of the universe” and says that “everything was established for their sake; that is, in choosing the order of things, the greatest account was taken of them” (“A Specimen of Discoveries,” A 6.4.1624/MP 83). Nevertheless, in some of his mature works, Leibniz qualifies the subordination of natural perfection to moral perfection, or the subordination of nonrational to rational creatures.7 In the Theodicy, for instance, he grants that “the happiness of intelligent creatures is the principal part of God’s design,” but he denies that it is God’s sole aim. Even though the natural order serves the moral order, there is also a sense in which the reverse is true: “since everything is connected in God’s great design, we must believe that the kingdom of grace is also in some way accommodated to that of nature, in such a way that the kingdom of nature preserves the most order and beauty, in order to make the combination of the two the most perfect that it can be” (T 118). In fact, Leibniz continues, “there is no room for judging that God, in order to reduce moral evil, would reverse the entire order of nature” (T 118). The following picture emerges from this text: God’s aim was to maximize both metaphysical or natural perfection and moral perfection; he did not create the natural order merely to serve the moral order. The last passage quoted even suggests that God could have created a world containing a smaller amount of moral evil – that is, a smaller amount of sin8 – but that God did not do this because such a world contains less metaphysical perfection (see Adams 2014: 201f.). Of course, this does not mean that God had to compromise the laws of justice, as we are currently understanding them. Those laws merely demand that no sin go unpunished; they do not demand that there be as little sin as possible.9 7 8

9

See Blumenfeld, who also discusses this shift in Leibniz’s views (1995: 402f.). For discussions of moral evil as sin, see T 21; CD 32; FR 43. Note, however, that in another mature text, Leibniz says that God chose the world with the fewest possible sins (Gr 374/ LGR 286). Leibniz appears to define ‘justice’ in a number of different ways. In addition to the sense that we already discussed – the sense in which justice is punishment and reward proportionate to desert – Leibniz sometimes suggests that someone is just if and only if nobody has a reason to complain to them (see “Meditation on the Common Concept of Justice,” Mollat 53/Riley 53, where Leibniz calls this a ‘nominal definition’; see also Mollat 58/Riley 56f.). God is perfectly just in that sense as well, even if he could have created happier individuals, because no individual has the right to complain about being created. Moreover, Leibniz sometimes defines ‘justice’ as charity – or universal benevolence – conformed to wisdom (e.g. letter to Nicaise, August 19, 1697, A 2.3.368/LGR 158f.) or as goodness and wisdom combined (“Meditation on the Common Concept of Justice,” Mollat 48/Riley 50; PNG 9; RH 12). Finally, he also describes it as “a reasonable advancement of the good of others” (“Meditation on the Common Concept of Justice,” Mollat 58f./Riley 57). Since these characterizations seem compatible with one another, it is possible that they are simply three aspects of justice, or three different ways to characterize the same notion of justice. That is particularly plausible given that several of these characterizations occur in the same text. None of these definitions requires that God create the happiest of all possible creatures or the least sinful.

186

Moral Agency

The passage under consideration does not imply that God needed to make a trade-off between this kind of moral order and the natural order.10 One final aspect of the City of God that I should mention is the following: Leibniz often states that finite minds are members of this moral community because they are particularly similar to God. The fact that Leibniz likens God’s relationship with finite minds to that of a father with his children already suggests a special kind of similarity, but he often spells this out explicitly. In a number of passages, Leibniz says that finite minds are like “little gods” (e.g. “New System,” G 4:479/AG 140; M 83; T 147; letter to Jaquelot, February 9, 1704, G 3:465/WF 176; letter to Arnauld, October 9, 1687, G 2:125/Mason 159). And, in fact, Leibniz thinks they are similar enough to God that we can apply the same term – ‘mind’ or ‘spirit’ – to God without equivocating.11 In “Discourse on Metaphysics,” moreover, he claims that “the quality that God has of being a mind himself takes precedence over all the other considerations he can have toward creatures” and that finite minds “are, as it were, of his race or like children of his household” (DM 36). Leibniz even writes to Antoine Arnauld that “created minds differ from God only in degree, from finite to infinite” (October 9, 1687, G 2:125/Mason 160; similarly in “Dialogue Between Theophile and Polidore” [1679], A 6.4.2234/LGR 132). God’s kinship with finite minds, then, is clearly enormously significant, and several texts make it clear that it is important, first and foremost, for membership in the City of God. For instance, in the letter to Arnauld from which I just quoted, Leibniz says that God “takes on another role respecting minds, . . . since he is himself a mind and as it were one amongst us, to the point of entering with us into a social relationship in which he is the leader” (G 2:125/Mason 159).12 Very similar statements occur in much later texts, for instance in the “Monadology”: “minds are . . . images of the divinity itself, . . . each mind being like a little divinity in its own realm. That is what makes minds capable of entering into a kind of society with God” (M 83f.).13 Leibniz then goes on to identify this special relationship with the City of God. There is strong textual 10

11

12 13

In order to answer the question of whether there is any trade-off between natural perfection and justice, we would need to investigate passages in which Leibniz acknowledges that genuine miracles are sometimes necessary in order to ensure the moral perfection of the world. That, however, would take us too far away from the topic of the present chapter. In a letter to Wolff, Leibniz calls God “the supreme mind” (May 18, 1715, GLW 172/AG 234; similarly in “On Sin and Original Sin” [c. 1686], A 6.4.2361/SLT 204). See also his notes for a letter to Des Bosses, February 15, 1712, LDB 233; a letter to Morell, May 4/14, 1698, A 1.15.561/SLT 39; T 186; a letter to Arnauld, October 9, 1686, G 2:125/Mason 159f.; DM 35; a letter to Bierling, August 12, 1711, G 7:502. See “Specimen of Discoveries”: “God is . . . the king of minds, and since he himself is a mind, he has a special association with them” (A 6.4.1624/MP 83). Similar statements occur, e.g., in “New System,” G 4:479f./AG 140 and in “Ultimate Origination of Things,” G 7:307/AG 154. Sometimes Leibniz merely says that minds are especially important to God because of their similarity to him (e.g. T 118).

What Is a Moral Agent?

187

evidence, then, that our similarity to God qualifies us for membership in God’s moral community and that at least some of the properties that we share with God are also the properties that make us moral agents. 1.2

Cognitive Requirements

It is no mere nepotism that the kinship between finite minds and God is a major reason – if not the reason – for their inclusion in the City of God, and thus also a major reason for their status as moral agents. The qualities that make finite minds like God turn out to be qualities that, for independent reasons, have moral import. What exactly are these qualities? I will argue that our similarity to God has a cognitive and an appetitive component; more specifically, finite minds are godlike mainly because they can both reason and will. Accordingly, I am going to show that there are both cognitive and appetitive requirements for moral agency.14 In order to be a moral agent, or an agent whose actions can be subject to moral evaluation, one must possess an intellect and a will. In other words, one must be able to grasp as well as deliberate about what is right, and one must be able to determine oneself in accordance with one’s intellectual judgments. Let us first look at the cognitive requirements, which are more straightforward and less controversial than the appetitive requirements that I discuss in the next section. A good starting point is a letter to Wagner, in which Leibniz tells us that one of the reasons why man is radically different from brute animals is that “from the use of reason he is capable of society with God, and thus of reward and of punishment in the divine government” (June 4, 1710, G 7:530/W 507). The use of reason, according to this passage, is crucial for moral agency. But what precisely must the cognition of a moral agent be like? Leibniz mentions a number of different cognitive abilities. It seems the two most important ones are (a) reflection and self-knowledge, especially the ability to remember one’s actions as one’s own, and (b) knowledge of eternal truths, in particular truths concerning morality and knowledge of God. I will briefly examine each in turn. There is strong textual evidence that Leibniz considers self-knowledge and memory of one’s past deeds a necessary condition for moral agency. Many such passages stem from Leibniz’s mature years, though he expresses the doctrine in the middle period as well. One particularly interesting discussion of this requirement, in fact, occurs in section 34 of “Discourse on Metaphysics.” The main difference between animal souls and intelligent souls, Leibniz says, 14

Some interpreters claim that certain cognitive abilities are sufficient for moral responsibility; see for instance Phemister, who argues that the ability to reflect and to recognize one’s actions as one’s own makes monads responsible for their actions (1991: 31).

188

Moral Agency

is that animal souls “do not know what they are nor what they do.” The fact that they lack this kind of self-knowledge, Leibniz continues, means that “they have no moral qualities” because “it is memory or the knowledge of [one’s] self that renders [one] capable of punishment or reward.” Without memory and selfknowledge, Leibniz explains, a substance cannot remain the same person over time; it can persist metaphysically, in the sense of remaining the same substance, but it cannot persist morally. Indeed, when animal souls change over time, “from the moral or practical point of view, the result is as if they had perished,” precisely because they lack the kind of self-awareness that unites different actions into one personal life (DM 34; similarly in a letter to Arnauld, October 9, 1687, G 2:125/Mason 160). A number of mature texts make analogous claims about the connection between self-knowledge and moral agency. The New Essays contain a particularly extensive and helpful discussion (see NE 233–47),15 but several other texts mention it as well. What Leibniz stresses again and again is that there is a crucial difference between (a) the persistence of persons or moral agents, which he calls either ‘moral identity’ or ‘personal identity,’ and (b) the persistence of substances, which he often refers to as ‘physical identity’ (e.g. NE 233; 236f.; letter to Wagner, June 4, 1710, G 7:531/W 507). Moral identity, moreover, is a prerequisite for reward and punishment; it requires reflective self-knowledge and the memory of past actions as one’s own.16 As Leibniz puts it in the Theodicy, what persists in a man is not only the soul but also “that which makes it the case that he is the same person, who retains his moral qualities by conserving the consciousness or the reflective inward sentiment of what he is, that which makes him capable of punishment and reward” (T 89). This type of self-knowledge, then, is clearly a necessary condition for moral agency. Self-knowledge and the ability to reflect are important not only for reward and punishment, but also for happiness and suffering. Animals, Leibniz holds, are not able to be happy or to suffer – though they are capable of pleasure and pain – precisely because they cannot reflect. Without reflection, he argues, a creature is “susceptible neither to the sorrow [chagrin] that accompanies pain, nor to the joy [joie] that accompanies pleasure” (T 250). The connections between reflection, happiness, and moral agency become clear in a draft of the “New System”: “the sensibility and pain of animals is of a completely 15

16

This is to be expected, of course, since the New Essays is a section-by-section commentary on Locke’s Essay, which in turn famously contains a chapter that accounts for personal identity in terms of memory. Note, however, that Leibniz acknowledges in the New Essays that gaps in one’s memory can be filled by the testimony of others, and on the basis of such testimony, one can justly be punished for something one does not remember doing (NE 236f.). Yet, since we are discussing moral agency, rather than responsibility for particular actions, we can bracket this complication. Presumably, an agent who generally lacks the ability to remember her own actions cannot be a moral agent.

What Is a Moral Agent?

189

different nature from our own, and, since they have no reflection, does not make them capable of being unhappy” (G 4:475/WF 25). This, he goes on to say, means that God is not unjust to animals by treating them so differently from minds.17 If a creature is not capable of happiness or suffering, God obviously cannot be blamed for failing to make the creature happy in proportion to its virtue or for failing to compensate it for its undeserved suffering. A further cognitive requirement for moral agency is the ability to know God and eternal or necessary truths.18 This ability, as Leibniz frequently points out, is one of the distinguishing features of minds. As he writes to Electress Sophie, [O]f all souls there are none more elevated than those that are capable of understanding the eternal truths, and not just of representing the universe in a confused manner, but also of understanding it and of having distinct ideas of the beauty and grandeur of the sovereign substance. That is, those that are capable of being the mirror . . . of what is best in the universe, that is, of God himself; and this is what is reserved for minds or intelligences. (November 11, 1696, A 1.13.90/SLT 79f.; similarly M 29)

Only minds, he stresses, can understand God and the eternal truths. Both of these types of understanding, moreover, are important prerequisites for moral agency. The ability to understand eternal truths, first of all, is required for understanding moral truths. After all, moral truths for Leibniz are eternal truths that are innate to the human mind, similarly to arithmetical truths.19 He sometimes expresses the moral difference between animals and human beings by saying that nonrational substances are “incapable of the divine . . . law” (letter to Wagner, June 4, 1710, G 7:531/W 508) or of “the spiritual laws of justice” (letter to Arnauld, October 9, 1687, G 2:124/Mason 159), which is arguably at least in part because they cannot understand those laws. In another text, in fact, he says explicitly that while animals can “distinguish good from bad . . . they are certainly not capable of moral good and bad, which presupposes reason and consciousness” (remarks on M. Foucher’s Objections, G 4:492/WF 46). And without the ability to distinguish moral evil from moral good – which presumably presupposes an understanding of moral goodness and badness – they cannot be genuine moral agents.20 17

18

19 20

He also writes to Arnauld that God is not doing any injustice to animals because “they have been created without the capacity for reflexion or consciousness, and consequently [are] not capable of happiness and unhappiness” (October 9, 1687, G 2:126/Mason 161). See also a letter to Hansch, July 25, 1707, E 2:447/L 595. As a matter of fact, this second requirement is closely related to the first because Leibniz sometimes claims that we have access to eternal and necessary truths through reflection (see e.g. DM 34; letter to Masham, May 1704, G 3:339/WF 205; but cf. M 30). Yet, it is helpful to examine them separately. See e.g. NE 92, where Leibniz says that “moral knowledge is innate in just the same way that arithmetic is, for it too depends upon demonstrations provided by the inner light.” See e.g. “Meditation on the Common Concept of Justice,” where Leibniz ties justice to wisdom and goodness, and defines wisdom as “knowledge of the good” (Mollat 48/Riley 50). Without

190

Moral Agency

The ability to understand or know God, which Leibniz also mentions in the letter to Electress Sophie, is important for a number of reasons. First and foremost, Leibniz holds that members of the City of God must be able to love God, which requires that they be able to know God. In one text, for instance, Leibniz contrasts God’s relationship to animals with God’s relationship to minds in the following way: “With creatures which are without reason God acts simply as creator and master, but with souls which can know and love him he acts as father and as leader” (draft of “New System,” G 4:475/WF 25; similarly in DM 35). The City of God, according to this passage, can contain only agents who have the capacity for loving God. This makes sense, since Leibniz ties genuine virtue closely to the love of God. The morally good, Leibniz tells us, are those “who love and imitate the author of all good, as they should, finding pleasure in the consideration of his perfections” (M 90).21 Because a moral agent – that is, a substance fit for inclusion in the City of God – must arguably be able to be morally good, such an agent must be able to know and love God. An additional reason why the ability to know God is important for moral agency is that, as Leibniz sometimes says, “certain rules of justice can be demonstrated in their full extent and perfection only if we assume the existence of God” (NE 89; see also NE 96; 201). If someone cannot know that God exists, she cannot acquire complete moral knowledge, which might in turn mean that she is not a moral agent. The reason why knowledge of God is required for complete moral knowledge appears to be the way in which Leibniz ties moral obligation to motivation. Leibniz is a psychological egoist: human beings cannot knowingly act against their self-interest (see e.g. NE 163; letter to Nicaise, August 19, 1697, A 2.3.369/LGR 159; preface to the Mantissa Codicis Iuris Gentium Diplomaticus, A 4.8.53/L 424). Yet, ‘ought’ implies ‘can’ for Leibniz.22 Hence, human beings cannot be obligated to act against their self-interest. Since God’s existence radically changes the motivations that human beings can have to act benevolently, it has important implications for human obligations.23 Thus, complete knowledge of human obligations presupposes knowledge of God.

21

22 23

the ability to know moral principles, an agent can at best act in accordance with some moral principles through an instinct. Yet, since animals possess that kind of instinct as well (see e.g. NE 93), genuine moral agency must require something more. See “Opinion on the Principles of Pufendorf,” where Leibniz says that someone who is truly just “does good out of love for God or of his neighbor” (Dut 4.3.280/Riley 72); see also a letter to Nicaise: “our true happiness essentially embraces the knowledge of God’s happiness and of the divine perfections, that is, the love of God” (August 19, 1697, A 2.3.370/LGR 160; similarly in Tp, G 6:27/H 51). See e.g. Tp: “nobody can be obliged to do the impossible” (G 6:33/H 57). In a text from 1680, Leibniz says this very explicitly: “if God did not exist, the wise would be obligated to be benevolent only insofar as it is to their advantage . . . in this short life” (A 6.4.2871); hence, he goes on to say, obligations would be very limited. Leibniz also says

What Is a Moral Agent?

1.3

191

Appetitive Requirements

Having considered some of the central cognitive requirements for moral agency, I now turn to evidence that Leibniz views the will and its activity as an additional requirement for moral agency. Let us first investigate which appetitive traits finite minds share with God. Then we can ask why those abilities might be required for moral agency. Consider the following passage from “Principles of Nature and Grace”: minds are superior to nonrational souls because they are “not only a mirror of the universe of created things, but also an image of the divinity. The mind not only has a perception of God’s works, but it is even capable of producing something that resembles them, although on a small scale” (PNG 14). Similarly, Leibniz says in “Meditation on the Common Concept of Justice” that “spirits . . . are the substances most similar to God because they are themselves capable of recognizing and inventing order and craftsmanship” (Mollat 50/L 565; emphasis added). He makes comparable points elsewhere, for instance in a letter to Queen Sophie Charlotte (G 6:507/AG 192). In addition to perceiving and partially understanding the world that God has created, these texts suggest, finite minds can themselves be creators on a smaller scale by producing or ordering things.24 In the section of “Principles of Nature and Grace” from which I quoted earlier, we find a clear description of the way in which finite minds imitate God’s creative acts: “For . . . our soul is also like an architect in its voluntary actions; and in discovering the sciences according to which God has regulated things . . . it imitates in its realm and in the small world in which it is allowed to work, what God does in the large world” (PNG 14). The human products that most resemble God’s work, Leibniz says here, are voluntary actions. This is presumably because when acting voluntarily, we produce something in accordance with our rational perceptions, which is precisely what God did when he created the world. The “sciences according to which God has regulated things” must, I think, be the Principle of Wisdom and Goodness, which – as seen in Chapter 5 – is the principle that governs voluntary actions. This principle allows finite minds to act, as God does, based on their rational judgments concerning the good. In fact, knowingly acting in accordance with reason is generally godlike, as Leibniz writes to Electress Sophie in 1706: “rational souls, which think about what they do . . . are also imitations of the

24

elsewhere that knowledge of God provides us with motives for being moral (e.g. “Meditation on the Common Concept of Justice,” Mollat 61f./Riley 58f.; NE 432). That is not only because of divine punishments and rewards, but also because the knowledge and love of God’s perfections make virtue its own reward. See Brown (2016), who makes a convincing case for the close connection between motivation and obligation. Elsewhere, Leibniz suggests that activity in general or the production of one’s states is an imitation of God, e.g. in a reply to Bayle, G 4:564/L 580. See also Jolley (1998: 598) and Phemister (2005: 235).

192

Moral Agency

divinity . . . [I]t is good and satisfying to know that one acts in accordance with reason: nothing is further removed from the beast state, and nothing approaches divinity more closely” (G 7:568f./SLT 84). There are passages that mention appetitive similarities between finite minds and God even more explicitly. In “On God and Man,” for instance, Leibniz tells us that “the vestiges of the image of God . . . consist in two things: in the light of the understanding and the freedom of the will” (G 3:36/LGR 296; see also LC 5.7). Leibniz makes an analogous point in Causa Dei: “The traces of the divine image consist in the innate light of the intellect as well as in the inborn freedom of the will” (CD 98). This strongly suggests that for Leibniz, we are images of the appetitive aspects of God’s nature, not only the cognitive ones. There is also strong textual evidence that the appetitive capacities that we share with God are important requirements for moral agency. First, Leibniz writes in the “Meditation on the Common Concept of Justice” that in order to be just, we need not only “knowledge of the good” but also “the inclination to do good to everyone, and to arrest evil,” which means that justice is both in the understanding and in the will (Mollat 48/Riley 50). Likewise, in the passage from Causa Dei quoted earlier, Leibniz says that our similarity to God consists in intellect and will, and that both of these are required for virtue and vice (CD 98). That in turn suggests the following. Suppose there were rational creatures without wills.25 These creatures would be able to understand that certain things are good, but they would have no inclination whatsoever to actualize these goods. Lacking wills altogether, they would not be inclined in the least by the judgments of their intellects concerning the good; instead, their perceptual transitions would be governed only by nonrational inclinations. Because they lack free will, even though they possess reason, they would not be moral agents. Hence, just like nonrational animals, they would be excluded from the moral kingdom of grace, and they would not deserve reward and punishment for their actions. It is not difficult to see why Leibniz would think that rational creatures that lack wills would not be moral agents. Such will-less rational creatures would be in the unfortunate situation of knowing what is good but being unable to direct their actions accordingly; they would be entirely at the mercy of their nonrational inclinations, and their reason would be completely powerless. Hence, they would never have even an indirect ability to control their actions rationally. They would be unable, no matter how clearly they understand that it would be best, to acquire the habit of acting in accordance with reason and of

25

If you separate intellect and will, as Leibniz does, there is no reason to think that such creatures are not possible; see Chapter 6 and Jorati (2014a: 758ff.).

What Is a Moral Agent?

193

becoming masters of their actions.26 In fact, Leibniz defines ‘virtue’ in the New Essays as “a general disposition to moderate the passions by means of reason, or more simply still a disposition to act in accordance with reason” (NE 98), which is further evidence that only an agent who has a will, or the capacity for controlling her passions, is capable of being virtuous.27 What I have said so far is enough to show that appetition is a crucial component of moral agency. A monad lacking any inclination to act in accordance with reason, no matter how intelligent it is, cannot be a moral agent. After all, such an inclination is required for the ability to act morally. Rationality is a necessary condition for moral agency, but it is not sufficient.28 The membership conditions for the City of God, then, include not only intelligence but also will. Based on my discussion in previous chapters, we can phrase this requirement even more precisely: moral agency requires the capacity for rational teleology and rational spontaneity. In other words, members of the City of God must be able to determine themselves in accordance with what they judge to be best, motivated by their intellection of the good. There is another related appetitive capacity that is crucial for moral agency: the capacity for loving God and one’s neighbors. The importance of the ability to love God already came up in the context of the cognitive requirements. Members of the City of God, as seen, ought to love their sovereign and creator above all else. God, according to Leibniz, “ought to be the whole aim of our will” (M 90). Since ‘ought’ implies ‘can’ for Leibniz, as already seen, this means that members of the City of God must be able to love their creator. And the capacity for loving God, while it arguably presupposes the ability to know God, is mainly an appetitive capacity.29 The same is true for the ability to love one’s fellow human beings, which for Leibniz is the key to moral virtue. After all, he holds that it is only through pure or disinterested love of others that one can, as morality often requires, act for the sake of their good instead of merely for the sake of one’s own good.30 For Leibniz, this pure kind of love, which we ought to have both for our creator and for our fellow citizens in the City of God, 26

27 28 29

30

Susan Wolf agrees: she argues that in order to be held morally responsible, an agent must possess “a will whose content can be informed and governed by the relevant considerations” (1990: 8). See also “Meditation on the Common Concept of Justice,” where Leibniz describes virtue and vice as “the perfection and imperfection of the will,” respectively (Mollat 61/L 569). See Jorati (2014a), where I argue that there is no metaphysically necessary connection between moral knowledge and moral motivation. Leibniz does say that those who know God as they ought will automatically love God, and those who love God will be “virtuous and just by inclination” (“Meditation on the Common Concept of Justice,” Mollat 62/Riley 59; similarly in Tp, G 6:27/H 51). Yet, loving God is only a natural consequence of knowing God in agents who possess a will that responds to perceptions of the good through love. The will, then, is crucial. In a letter to Claude Nicaise, for instance, Leibniz writes that justice requires charity, which in turn is universal benevolence, which in turn is “a disposition or inclination to love” (A 2.3.368/LGR 159).

194

Moral Agency

consists in taking pleasure, or finding happiness, in the happiness of the beloved (e.g. NE 163; M 90; letters to Nicaise, August 19, 1697, A 2.3.369/LGR 159 and May 4/14, 1698, A 2.3.441/W 566). Or, as Leibniz puts it in one text, it consists in accepting “the happiness of another as one’s own” (Preface of the Codex Juris Gentium Diplomaticus [1693], G 3:387/L 421).31 What it means to love other people in a disinterested way, then, is to be motivated by their happiness in the way in which one is typically motivated by one’s own happiness. This clearly falls under the job description of the will. 2

Moral Responsibility

Moral agency, as seen in the previous section, requires the general capacity to know oneself, God, and eternal truths. It also requires the capacity to determine oneself in accordance with one’s intellection of the good. Yet, while this may be a complete description of moral agency, it does not yet tell us what is required for an agent to be morally responsible for a particular action. After all, it would be implausible to claim that moral agents are automatically responsible for all of their actions. The present section aims to answer this further question. 2.1

Freedom

We already saw in Chapter 5 that for Leibniz, freedom is incompatible with metaphysical and absolute necessity. Perhaps unsurprisingly, Leibniz says something analogous about moral responsibility: “justice and injustice, praise and blame, punishment and reward cannot attach to necessary actions, and nobody can be obliged to do the impossible or to abstain from doing what is absolutely necessary” (Tp, G 6:33/H 57).32 According to this passage, an agent is not morally responsible for anything that is absolutely necessary.33 That does not entail, of course, that agents are morally responsible only for their free actions – or, what comes down to the same thing, for their morally necessary actions. After all, unfree monadic actions are not absolutely necessary either. 31

32

33

Because a disinterested lover finds happiness in the happiness of others, he can act for the sake of the good of others while still being motivated by his own happiness, as Leibniz’s psychological egoism requires. For a more detailed discussion, see Brown (2011), Rateau (2015: 230ff.), Jorati (2014a), and Youpa (2013, §3). To be more precise, Leibniz holds that retributive justice (justice vindicative; T 74) is incompatible with absolute necessity; what Leibniz calls “corrective justice” (justice corrective; T 74; or justice médicinal, COE 17), on the other hand, is compatible with absolute necessity (T 67–75). I will bracket that latter type of justice, however, and focus on retributive justice. As seen in Chapter 5, this is important for understanding Leibniz’s rejection of Spinozism: because the actions of Spinoza’s God are metaphysically necessary, he is not subject to genuine praise.

Moral Responsibility

195

Let us, then, examine the connection between moral responsibility, on the one hand, and freedom or moral necessity, on the other.34 While Leibniz would be in good company if he thought that moral responsibility is limited to free actions, there is evidence that this is not his considered view. In the Theodicy, for instance, Leibniz describes man as like a little god in his own world, or microcosm, which he governs in his own way. He sometimes does very well there . . . But he also makes grave mistakes, because he abandons himself to the passions, and because God leaves him to his own devices. God also punishes him for this, sometimes like a father or tutor . . . sometimes like a just judge, punishing those who abandon him. (T 147)

This passage’s talk of abandoning oneself to the passions suggests that it is referring to actions that are entirely dictated by passions and that are therefore not free. If that is correct, the passage states that it is just to punish human beings for some unfree actions.35 Other passages support this reading. In one text, for instance, Leibniz acknowledges that “whoever is not well-intentioned will often sin, at least by omission,” (“Opinion on the Principles of Pufendorf,” Dut 4.3.278/Riley 69). This strongly suggests that something can be a sin, and hence morally culpable, even if it is not free, since omissions of the type described in that passage are not typically the outcome of deliberation. The context, after all, is a discussion of cases in which an agent, out of fear or hope, represses her “wicked thoughts, so that they do no harm.” Leibniz’s point appears to be that even if such an agent can prevent herself from acting on her wicked thoughts by repressing them, she will nevertheless often fail to do what is required because she lacks inclinations for performing certain morally obligatory actions. Such omissions would plausibly not be based on a free decision. Moreover, when discussing different types of sin in Causa Dei, Leibniz not only acknowledges sins of commission and of omission, but also states that sins can be “culpable through an infirmity of nature” and “wicked through a depravity of the soul” (CD 92).36 Here, again, Leibniz appears to allow for sinful and culpable behavior that is not free. There are other reasons to think that an action can be culpable without being free. Leibniz’s notion of freedom, as seen in previous chapters, is extremely demanding. Only actions that are preceded by explicit, rational deliberation about the best course of action are free.37 Thus, if moral responsibility were 34 35 36 37

Some interpreters believe that there is a close connection in Leibniz between freedom and responsibility. See e.g. Lin (2012: 443). The last clause of the passage, I take it, suggests that this is also retributive justice, not merely corrective justice. See footnote 32. See also T 264, where Leibniz claims that punishing an immoral person is justified regardless of the origin of their badness. See Chapter 5, footnote 12 for textual evidence.

196

Moral Agency

restricted to free actions, human beings would be responsible for only very few of their actions.38 Among other things, they would not be responsible for their synchronically akratic actions or for their bad habits. In fact, the best way to avoid acting in a morally reprehensible way would be to deliberate as little as possible, and instead to cultivate one’s passions. If that were Leibniz’s view, it would not only be extremely implausible, but it would also conflict with his insistence that we have a moral duty to increase our mastery and become more rational.39 For Leibniz, as already seen, virtue is “a general disposition to moderate the passions by means of reason, or more simply still a disposition to act in accordance with reason” (NE 98) or “the habit of acting according to reason” (“Memoir for Enlightened Persons” §12, A 4.4.615/Riley 105). Moreover, he holds that in the moral kingdom of minds, “every virtue . . . is comprehended among the obligations of universal justice” (“Opinion on the Principles of Pufendorf,” Dut 4.3.281/Riley 73).40 We have a moral obligation, then, to control our passions. Hence, in at least some circumstances, we can presumably be held responsible for failing to do so. As a result, we are sometimes morally blameworthy for actions or omissions that are unfree. The fact that Leibniz encourages us to acquire virtuous habits is also evidence that we can be worthy of moral praise for actions that are not strictly speaking free. While some virtuous habits – for instance, the habit of deliberating before doing something risky – lead to free actions, other habits do not. For instance, if you have acquired the habit of coming to a full stop at every stop sign, you will do so without deliberating about it every single time. Yet, it seems that both types of habits are central to human virtue, for Leibniz. In one of his discussions of mastery, for instance, he states that “we are slaves insofar as we follow passions and customs or the unthinking impulses that reason has not beforehand formed into a habit of doing well” (Gr 481/SLT 98). This passage appears to say that reason can turn some unthinking impulses into virtuous habits and that when we follow those habits, we are not slaves but possess at least some mastery over the resulting actions. Leibniz sometimes, in fact, describes the ideally virtuous human agent as someone for whom virtue has become a second nature,41 which presumably means that such an agent 38

39 40

41

In fact, it may be implausible to make freedom a necessary condition for moral responsibility even on less demanding definitions of freedom, as some contemporary authors argue (see e.g. Smith 2005). See e.g. NE 88: “in so far as one is capable of knowledge, it is a sin to neglect to acquire it.” He also says earlier in the same text that “we owe it not only to ourselves, but also to society, above all to the one that we inhabit with God, . . . that we have a soul imbued with true thoughts and a will which tends constantly toward the just” (Dut 4.3.278/Riley 69; translation altered). See e.g. a text from the 1690s, in which Leibniz states that virtue requires not only wisdom but also habituation, because it is in the latter way that “the exercise of good actions becomes easy and natural for us, and passes into habit, because custom is another nature” (“Happiness,” Gr 581/SLT 169). Similarly, he says elsewhere that a good education “ought to consist in making virtue agreeable, and in making it a second nature” (“Memoir for Enlightened Persons of Good

Moral Responsibility

197

often does the right thing automatically, without deliberation. It would be strange, however, if we could not justly praise such an agent for the good actions that result from the habit.42 Something analogous can be said, mutatis mutandis, about the wicked, for whom sin has become a second nature. Another problematic consequence of tying moral responsibility to freedom is that it would then also make sense to say that we are less responsible for actions that are free to a lower degree. Yet, that would mean, at least in the human case, that agents typically bear only a very mild responsibility for their most evil actions, because those actions are usually free to a much lower degree than virtuous actions.43 After all, as already seen, the influence of nonrational inclinations is typically a major factor in evil actions, and this influence also makes those actions less free.44 While that might be a bullet that some philosophers are willing to bite, the considerations previously mentioned apply here as well: Leibniz clearly thinks that we can sometimes be blamed for our failure to control our passions, which suggests that we are not automatically excused to the extent that our passions influence our action. The connection between freedom and responsibility, as a result, cannot be so straightforward. Admittedly, there are also passages that suggest that freedom is a necessary condition for moral responsibility. Before examining some such passages, it is helpful to note a number of factors that make it difficult to interpret Leibniz’s statements about the relationship between freedom and responsibility. One complicating factor is that Leibniz uses the terms ‘freedom’ or ‘free will’ in different ways: in addition to using them to refer to the rational selfdetermination that I described in Chapter 5, he sometimes uses them merely to signal the absence of absolute necessity.45 I also suspect that in a few cases, he uses those terms to refer to the capacity to rationally control one’s actions, or the general capacity for free agency. Requiring freedom in that sense for moral responsibility is of course completely different from requiring rational selfdetermination, and it is in fact compatible with my claim that agents can be

42

43

44 45

Intentions” §14, A 4.4.615/Riley 106). See also NE 188; “Opinion on the Principles of Pufendorf,” Dut 4.3.278/Riley 69. Sometimes the habit might be acquired through voluntary actions of the agent, and in that case, we could of course still praise her for those voluntary actions. In many cases, however, virtuous agents acquire their habits through education in childhood and hence, presumably, not by choice. See T 75, where Leibniz discusses situations in which it is appropriate to praise agents for their good nature. Phemister goes even further and claims that actions that aim at something mistakenly thought to be good by the agent are not free at all; she holds that “the distinctive character of free action consists in the recognition or knowledge of the true good and acting accordingly” (2005: 230, similarly 248). Yet, it is implausible to interpret Leibniz in this way; Leibniz appears to hold that an agent is free to the degree that she is doing what her intellect judges to be best, even if that judgment is mistaken. Davidson discusses this problem as well (2005: 250). For an explicit discussion of those two senses, see NE 175.

198

Moral Agency

responsible for unfree actions.46 Moreover, Leibniz might sometimes use the term ‘freedom’ to refer to the ability to do otherwise, which as we will soon see is a requirement for moral responsibility. All of these factors put together make it extremely difficult to evaluate the evidence for the connection between moral responsibility and freedom. Among the passages that are worrying for my interpretation is the following: “Fallen and unregenerate man is under the domination of sin and of Satan because it pleases him so to be; he is enslaved voluntarily [esclave volontaire] through his evil concupiscence. Thus it is that free will [franc arbitre] and the enslaved will [serf arbitre] are one and the same thing” (T 277). Given what I said about mastery in Chapter 6, this passage – particularly the last sentence – is quite startling. While it certainly seems possible for an agent to freely or voluntarily abandon herself to her passions, it would be rather odd to insist that a free choice in Leibniz’s sense is required for such abandonment. Some agents may freely choose to be “enslaved” by their passions, but others may simply fail to exercise their mastery without ever making an explicit choice about it. For these reasons, I suspect that in this passage, Leibniz is not distinguishing as carefully as he should between blameworthy behavior and free actions; I suspect that it is not his considered view that the failure to exercise mastery is always a choice. What he should say, in my view, is not that people always give in to their passions freely but rather that they are morally responsible for failing to control their passions. In several other problematic passages, Leibniz justifies divine punishment by invoking the freedom of creatures, suggesting that only free will justifies such punishment. For instance, he calls free will “the proximate cause of the evil of guilt, and hence of the evil of pain” (T 288).47 Leibniz goes on to say that “the original imperfection of creatures, which is represented in the eternal ideas, is the first and most remote cause” of the evil of guilt, but he nevertheless appears to be claiming that divine punishment is always a response to free creaturely actions. Another troubling passage occurs in Causa Dei, where Leibniz states that “man’s corruption and depravity . . . do not render him excusable, or exempt him from culpability, as if he were not acting with sufficient freedom and spontaneity” (CD 97). This appears to indicate that acting without sufficient freedom automatically exempts agents from moral blame. Even worse, Leibniz then goes on to say that God can justly punish sinners because man possesses “traces of the divine image [which] consist in the innate light of the intellect as 46

47

Saying that only the general capacity for free agency is required for moral responsibility would simply be another way to make my point from Section 1 – namely, that only agents capable of free agency are moral agents. See also “Dialogue on Human Freedom,” where Leibniz’s spokesperson says that the sin of finite minds “arises from their will” (Gr 368/AG 117).

Moral Responsibility

199

well as in the inborn freedom of the will. Both are necessary to make actions virtuous or vicious; we must know and will what we are doing” (CD 98). On the most straightforward reading, the last sentence of this quotation states that an action is morally evaluable only if the agent willed the action. If the agent did not perform the action voluntarily and hence freely, it seems, the action cannot be called virtuous or vicious. I am not entirely sure what to make of these passages. One possibility is that Leibniz is primarily interested in the requirements for moral agency rather than in moral responsibility for particular actions. This reading makes sense for the Causa Dei passage, given its context. Leibniz’s point there could be that God is justified in punishing human beings in spite of their fallen nature because they are moral agents. Despite our depravity, we possess a sufficient amount of freedom and spontaneity to count as moral agents. We do, after all, possess the general capacity for recognizing the good and acting on the basis of that recognition. The strategy just sketched does not work for all passages, however. For instance, Leibniz says that “[t]rue retributive justice . . . assumes something more [than merely corrective justice], that is, the intelligence and freedom of him who sins, because the harmony of things demands a satisfaction, an evil of suffering, which makes the mind feel its error after the evil of the voluntary action to which it has given its consent” (COE 17). Even though the first half of this quotation could be read as requiring merely the general capacity for freedom and intelligence, the second half clearly specifies that the mind is punished for a voluntary action. Because I cannot explain away all of the problematic passages, they need to be taken as evidence that Leibniz, at least sometimes, views freedom as a necessary condition for moral responsibility. Yet, one must balance this evidence against the contrary evidence that I examined earlier. It is of course possible that Leibniz did not fully realize that he was committed to saying that agents can be morally responsible for some unfree actions. Yet, I find it more likely that he did realize this but was not always sufficiently careful when discussing moral responsibility. On balance, it seems that Leibnizian agents can be responsible for unfree actions. 2.2

Authorship, Spontaneity, and Teleology

If an action need not be free in order to be subject to moral evaluation, what are the requirements for moral responsibility? Later, I will argue that the requirements for moral praise or reward are different from the requirements for moral blame or punishment. They do have something in common, however: in order for the agent to be morally responsible for an action, it must be the case that the action is natural, or nonviolent. In other words, to use the terminology

200

Moral Agency

introduced in Chapters 2 and 3, the action must be metaphysically spontaneous and metaphysically teleological. Moreover, the agent must be the author of the action. This should not be surprising, given what I said in Chapter 4: being the author of an action simply means being the agent to whom the action is properly attributed. And, as also seen in Chapter 4, that requires that the agent supply the final and formal cause of the action. If God were to miraculously cause a change in you toward which you are not naturally directed and whose properties cannot be explained in terms of your nature, you would not be the author of that change. Instead, the change would be violent – that is, not in accordance with your internal or natural ends.48 And if you are not the author of a particular change, then, a fortiori, you cannot justly be held responsible for it. That is clear from the way Leibniz discusses authorship: he aims to show that God is not the author of creaturely sins and therefore is not morally responsible for those sins.49 So, the necessary conditions for moral responsibility include authorship, metaphysical teleology, and metaphysical spontaneity. In Leibniz’s system, metaphysical spontaneity and teleology are, of course, a very low bar. Even though God could cause changes in us that do not follow from our natures, he never or almost never does so. Likewise, as already seen, no finite substance has the power to violently impose changes on another substance. When one substance appears to impose a change on another – for example, when someone pushes you into your colleague’s porcelain collection – nobody’s metaphysical spontaneity or teleology is undermined. As a result, it may seem that a more demanding type of spontaneity and teleology is required for moral responsibility. It does seem unfair, after all, to hold you responsible for breaking your colleague’s porcelain figurines if someone else pushed you into them. One natural suggestion is that moral responsibility requires agent spontaneity and agent teleology. That would, after all, yield the intuitively plausible result in the porcelain figurine example. Could that be Leibniz’s view? Unfortunately, Leibniz does not seem to tell us explicitly. It seems clear that moral responsibility requires at least metaphysical spontaneity and teleology, and it plausibly does not require rational spontaneity and teleology, as seen in the previous section. Yet, it is not obvious that it requires agent spontaneity and teleology. To speculate a little, it would appear that at least the consequences of an agent-spontaneous and agent-teleological action can be culpable in certain 48

49

See Bobro (2004), who also argues that spontaneity is required for moral responsibility. He does not say what kind of spontaneity is required because he is mainly interested in the difference between genuine substances and machines on which God has miraculously imposed thought. Thinking machines, Bobro contends, cannot be moral agents because they lack spontaneity. They do not truly produce their own actions; the thoughts occurring in them cannot be attributed to them. In fact, if such a machine did something bad, God would be the author of that bad action (2004: 67ff.). See Chapter 4. See also ONI 10: if our spontaneity were called into doubt, “the cause of evil things [would] be thrust into God.”

Moral Responsibility

201

situations – even if those consequences are not themselves agent-spontaneous and teleological. For example, suppose I open the neighbor’s garden gate, their dog escapes, and it bites a child. My perception of the dog biting the child is not agent-spontaneous. Yet, I presumably bear some responsibility for the child’s injury because it is a foreseeable consequence of something I did. As a result, it would be plausible for Leibniz to claim that an agent can be morally responsible for a state with respect to which she lacks agent spontaneity. Before moving on, I need to address one minor problem with my claim that authorship, metaphysical spontaneity, and metaphysical teleology are required for moral responsibility. As already seen, Leibniz sometimes acknowledges that agents can be responsible for omissions, not just for actions. Yet, it is unclear what it would mean to be the author of an omission or what it would mean for an omission to be metaphysically spontaneous and teleological. Take, for instance, an agent who is so selfish that it never even occurs to her to offer help to others who are in need. Her failure to help others is, plausibly, an example of an omission for which she can justly be held responsible. But can we say that she is the author of this omission or that the omission is spontaneous and teleological? Arguably not.50 If that is so, it is not strictly speaking correct that authorship, spontaneity, and teleology are necessary conditions for moral responsibility; they are necessary conditions only if we bracket omissions. Yet, some parts of my discussion do apply to omissions. Omissions, like actions, must be explained by the agent’s nature in order to be attributable to this agent. After all, suppose that God miraculously impedes someone’s inclination to help a friend. If that happened, the person’s failure to help the friend would surely not be culpable. As a result, I think we can stretch the notion of authorship and even of spontaneity and teleology to cover omissions: if an omission is explained by the agent’s nature – that is, if it is not the result of a violent interference by another agent – we can ascribe that omission to the agent. We can likewise say that the agent’s self-determination and natural enddirectedness were not violated. This is admittedly a stretch, but not an unmotivated one. After all, as we saw in Chapters 2 and 3, spontaneity and teleology are at bottom aspects of independence. As long as a substance is allowed to unfold naturally, without external interference, it possesses metaphysical spontaneity and teleology. And just as we can distinguish between actions that are in accordance with the substance’s nature and those that are not, we can distinguish between omissions that are natural and those that are unnatural. 50

In fact, saying that an agent brings about an omission spontaneously appears to be an instance of the same mistake as saying that an agent produces the privative aspect of an action. See Chapter 4, where I discuss Leibniz’s criticism of philosophers who talk about privation in those ways.

202

Moral Agency

2.3

The Ability to do Otherwise

So far, I have suggested that authorship, metaphysical spontaneity, and metaphysical teleology are necessary conditions for moral responsibility, but freedom is not. That, of course, does not get us very far toward formulating sufficient conditions for moral responsibility. What will get us much closer, I think, is an investigation of the question of whether an agent must be able to do otherwise in order to be morally responsible for her action. There are a few passages in which Leibniz mentions that ability in connection with moral responsibility. For instance, he says that “he who causes evil by necessity is not culpable” (COE 15) and that in order for us to be blameworthy, it is necessary that “we can abstain even from that sin which we actually commit, if we summon a sufficiently strong effort” (CD 98). Even an inhabitant of hell, he insists, retains “a freedom which renders him culpable and a power, albeit remote, of recovering himself, although it never passes into action” (T 269).51 Leibniz hence appears to think that the ability to do otherwise, in some sense, is a requirement for moral responsibility.52 But in what sense must an agent be able to do otherwise in order to be morally responsible? We can of course immediately rule out a libertarian interpretation of ‘ability to do otherwise.’ Leibnizian moral responsibility obviously cannot require that the agent be exempt from all determination, even from physical and moral necessitation; it cannot require that the agent be able to do otherwise physically and morally speaking. After all, that would entail that no action of any actually existing agent meets the requirements for moral responsibility, which is definitely not Leibniz’s view. Yet, given what I said in Chapter 5, there is another option:53 moral responsibility might require merely metaphysical contingency. In order to be able to do otherwise, it simply needs to be metaphysically possible that one does otherwise. And this is in fact a sense of ‘ability to do otherwise’ that Leibniz 51

52

53

See also T 369: “the certain determination to sin that there is in man does not deprive him of the power of not sinning (absolutely speaking) and, since he does sin, it does not deprive him of being culpable and meriting punishment.” There might be exceptions. Leibniz sometimes makes room for cases in which an agent is justly held responsible not for an action but for an attitude or a passion. He says, for instance, that “there are cases where a man cannot do otherwise and yet could be guilty before God – e.g. where he is very pleased to have the excuse that he cannot help his neighbor” (NE 483). According to this passage, God judges human beings not just based on their external actions and omissions but also on their attitudes. Most of the time, however, Leibniz concentrates on responsibility for actions and omissions, and I will follow him in doing so. Moreover, God might hold us responsible only for sinful attitudes that we were able, in some sense, to prevent. The man who is pleased to be unable to help his neighbor might be blameworthy because he could – and ought to – have cultivated love for his neighbor. An additional option is suggested by Frankel: freedom and responsibility might require merely an epistemic possibility of having done otherwise, that is, “the conceivability of having done otherwise” (1984: 58).

Moral Responsibility

203

sometimes acknowledges, for instance when he says that “metaphysically speaking, [God] could have chosen or done what was not best” (T 234). As already seen, the ability to do otherwise in this sense is extremely important to Leibniz, and it is closely connected to freedom as well as moral responsibility. He argues, for instance, that if there were no other possibilities that God could have actualized, God would not deserve praise for his choice to create the actual world (CD 21). Yet, if only the ability to do otherwise in the metaphysical sense is required, we have not made much further progress toward formulating sufficient conditions for moral responsibility, since metaphysical contingency is common to all actions of Leibnizian substances. There is another option, however, based on the notion of control that I introduced in Chapter 6. Leibniz holds, as we saw there, that we are often able to control our passions directly, simply by focusing on rational considerations. When I give in to my desire to watch TV instead of keeping my promise to a friend, for example, I may have been able to do otherwise in the following sense: if I had tried harder and focused more on my obligation toward my friend, which is something I generally can do, I would have done otherwise. That is, I could have controlled these passions directly, by opposing stronger rational inclinations to them. The notion of control thus gives us an additional way to understand the ability to do otherwise: saying that an agent had the ability to do otherwise in this sense means that she would have done otherwise if she had exercised her capacity for direct control. Yet, as also seen in Chapter 6, our passions can sometimes get so strong that we cannot control them directly. When you are consumed with rage, for example, the rational aspect of your soul may temporarily be powerless; you may be unable to act rationally in the presence of this powerful passion. Leibniz insists, however, that there is often an indirect way to control such passions: there are steps we can take at an earlier time, when we are in a calm frame of mind, to prevent situations in which we cannot directly control our passions. For instance, take an agent who snaps at a waiter but who would have held her tongue had she started working on her anger management problems long ago. In that sense, losing her temper is something that she could have avoided – if not by controlling the anger directly, then by controlling it indirectly. This, then, provides us with yet another way of understanding the ability to do otherwise: when an agent does something that she could have prevented through indirect control, she was able to do otherwise. I find it plausible to combine the last two senses into one: an agent is able to do otherwise if and only if she would have done otherwise had she exercised her capacity for direct or indirect control.54 54

Davidson proposes a somewhat similar interpretation of Leibniz, even though he does not extend it to unfree actions (2005: 251).

204

Moral Agency

If the ability to do otherwise in the sense of direct or indirect control is required for moral responsibility,55 we may have something that helps us formulate necessary and sufficient conditions for responsibility. After all, in many cases, it is plausible that someone is responsible for an action if and only if they are the author of that action and had the ability to do otherwise, in the sense under consideration. That you broke your colleague’s figurines because someone pushed you, for instance, is not something you could have prevented through direct or indirect control.56 The same is true if you accidentally trip and fall into the figurines, despite your best efforts to walk past them carefully, or if you knock them over either because of a sudden involuntary twitch in your arm or because you are suddenly and unexpectedly consumed with an entirely uncharacteristic aggressiveness toward kitschy objects. If, on the other hand, you are aware that every time you see the figurines, you get an extremely strong urge to punch them, and you nevertheless do not make any effort to avoid going near them or develop a higher tolerance for kitsch, you are responsible for breaking them, even if the urge is so strong that you cannot resist it directly. An ability to do otherwise in the sense of control might even help us decide when omissions are culpable: perhaps you can be blamed for your failure to do what is right only if, through direct or indirect control, you could have done the right thing. For instance, we might be able to blame the selfish person who fails to help someone in need on these grounds: she could have taken steps to cultivate her love for other people and thereby acquired a sufficiently strong inclination to help. In this way, the necessary and sufficient conditions under consideration would make it much more difficult for an agent to get off the moral hook by invoking her irrational inclinations or her lack of virtuous inclinations. Whether Leibniz understands moral responsibility in the way just sketched is not as clear as one might wish. He does not, I think, say so explicitly. Yet, in addition to its plausibility, there is some indirect evidence for my proposal. First of all, a passage already quoted may suggest something like it: in order for our actions to count as vicious, Leibniz says, it must be the case that “we can abstain even from that sin which we actually commit, if we summon a sufficiently strong effort” (CD 98). Summoning an effort, quite plausibly, means exercising one’s rational control. Or consider the following passage: “the necessity contrary to morality, which must be avoided and which would 55

56

This would be similar to Aquinas’s view. Aquinas claims that in order for an agent to be excused for a bad action, the action must be “altogether involuntary.” Situations in which a passion “is so strong as to take away the use of reason altogether” do not excuse an agent if the agent is responsible for ending up in that situation or if she could have prevented that situation. Such an action can be what Aquinas calls “voluntary in its cause” or “indirectly voluntary” (ST IaIIae q77 a7 corp.). This seems true at least under the assumption that being pushed is not something for which you are indirectly responsible, or which was foreseeable and preventable.

Moral Responsibility

205

make punishment unjust, is an insuperable necessity, which would render all opposition useless, even if one wishes with all one’s heart to avoid the necessary action and makes all possible efforts” (Ta, response 3, G 6:380). Here again, Leibniz claims that one cannot justly be blamed for actions that one could not have avoided even if one had made all efforts one is capable of making, which suggests something like the reading I have proposed. Second, Leibniz evidently believes that our ability to control our actions is of the utmost moral importance because it makes it possible for us to become more virtuous. In a letter to Coste, for instance, he states that “we can contribute to making ourselves will what we should” by exercising indirect control (December 19, 1707, G 3:403/AG 195). Indeed, when reading passages in which Leibniz discusses strategies for avoiding situations in which we are under the sway of powerful irrational passions, it is difficult not to interpret him as claiming that we ought to use those strategies and that we are morally amiss if we do not. We have at our disposal “remedies for our ills [that] are . . . easily available” (NE 191); not using those remedies should, it seems, make us subject to blame. My interpretation implies that Leibniz is committed to an asymmetrical theory of moral responsibility, reminiscent of Susan Wolf’s theory.57 If what I have said so far is correct, we are responsible for sins only when we have the ability to do otherwise in the sense of the ability to exercise rational control. Yet, we can be responsible for morally good actions in the absence of that ability. After all, an agent has the ability to do otherwise by exercising her rational control only in situations in which she is acting in accordance with her passions. An agent who is doing the rational thing does not have the ability to do otherwise through rational control. This is the reason why the notion of control does not apply straightforwardly to God, who always does what is rational and has no passions to control. God, as Leibniz says in the Theodicy, “has no need of the power to change the most desirable volitions” (T 327) and does not have “the freedom to act unreasonably,” which is a “false freedom” (T 339) – that kind of freedom, or the ability to will something other than the best, is not a perfection and is not required for moral responsibility. God deserves our praise and admiration for acting in the most perfect manner possible even though there is a sense in which God cannot act otherwise. In fact, God deserves special praise because he is always morally necessitated to do what is best. As a result, an agent can be morally responsible for a good action even when she was unable (in the sense under discussion) to act badly, but an agent can be responsible for a bad action only if she could have acted reasonably instead. Or, more precisely, moral responsibility for bad actions requires the ability to do otherwise in the sense of control, whereas 57

For Wolf’s asymmetry, see 1990: 79ff.

206

Moral Agency

responsibility for good actions requires the ability to do otherwise only in the sense of metaphysical contingency. Like Wolf, then, Leibniz endorses an asymmetrical theory of moral responsibility. What matters to both Leibniz and Wolf is, as Wolf puts it, “the availability of one very particular option, namely, the option to act in accordance with Reason. If, on this view, the agent exercises this option, then it is irrelevant whether the agent might not have exercised it” (1990: 68). What matters, then, is not so much the ability to do otherwise generally, but the ability to do the rational thing. Both Leibniz and Wolf believe that we often possess that ability, even when we do not exercise it. There are also some important differences between Leibniz’s and Wolf’s views on responsibility. Most importantly, Wolf holds that being psychologically determined to do something wrong undermines an agent’s blameworthiness; physical determination, on the other hand, does not (1990: 79, 101, 112ff.). Leibniz, in contrast, does not think that either type of determination is problematic. For him, all actions are psychologically determined. As seen, he spells out the ability to do otherwise not in terms of psychological contingency, but in terms of the availability of direct or indirect methods for controlling the passions. In this respect, Leibniz’s approach strikes me as more promising than Wolf’s. For Wolf, human actions must have a psychological level of explanation that cannot be reduced to, or eliminated by, the physical level of explanation. Moreover, the psychological level of explanation cannot be deterministic (1990: 109). These are extremely controversial claims, as Wolf acknowledges. Leibniz’s account of moral responsibility, in contrast, does not rely on these controversial assumptions. His theory is compatible with the reducibility of psychological to physical explanations,58 as well as with complete psychological determinism. This, it seems, makes Leibniz’s account superior to Wolf’s account. 3

Conclusion

In order to be a moral agent and a member of the City of God, an agent must possess both an intellect and a will. Possessing an intellect is required because moral agents must not only be able to know what morality demands but also be capable of reflection and self-knowledge. More specifically, they must be able to remember their past actions as their own. Yet, possessing a will or particular appetitive capacities is at least as important as possessing an intellect: moral agents must be able to translate their knowledge about the good into actions, 58

To be clear, I am not claiming that Leibniz wants to reduce psychological to physical explanations. My point is merely that Leibniz’s theory is compatible with such a reduction, which should make it more attractive to many contemporary philosophers of action.

Conclusion

207

and they must possess the general capacity for exerting rational control over their passions. An agent who has no inclination whatsoever to do what she judges to be best – that is, an agent who lacks a will and who is completely at the mercy of her passions – would not be a moral agent. Another crucial appetitive requirement for moral responsibility is the ability to love God and the other members of the moral community. Someone who is incapable of that kind of love, after all, is unable to do what morality demands. While only free agents can be moral agents, I have argued that Leibniz does not restrict moral blame and praise to free actions. An agent can be morally responsible for an action, on this interpretation, even if she does not perform the action freely. In order to be morally responsible for an action, the agent must be the author of that action and the action must be metaphysically contingent. Moreover, in the case of moral blame, the agent must also have been able to do the rational thing by exerting either direct or indirect control over her passions. These conditions are excellent candidates for jointly sufficient conditions for moral responsibility, at least in the case of actions. The account can also be extended to omissions. An agent is morally responsible for failing to do what is required if and only if the agent’s self-determination and natural enddirectedness were not violated and if the agent could have done the right thing through direct or indirect control.

References

Adams, Robert Merrihew. 2014. “Justice, Happiness, and Perfection in Leibniz’s City of God.” In New Essays on Leibniz’s Theodicy, eds. Larry M. Jorgensen and Samuel Newlands, 197–217. Oxford: Oxford University Press. 2005. “Moral Necessity.” In Leibniz: Nature and Freedom, eds. Donald Rutherford and Jan A. Cover, 181–93. Oxford: Oxford University Press. 1996. “Response to Carriero, Mugnai, and Garber.” Leibniz Review 6: 107–25. 1994. Leibniz: Determinist, Theist, Idealist. New York: Oxford University Press. Anfray, Jean-Pascal. 2011. “Leibniz, le choix du meilleur et la nécessité morale.” In Lectures et interprétations des Essais de théodicée de G. W. Leibniz, ed. Paul Rateau, 49–68. Stuttgart: Steiner. Antognazza, Maria Rosa. 2014. “Primary Matter, Primitive Passive Power, and Creaturely Limitation in Leibniz.” Studia Leibnitiana 46 (2): 167–86. Arthur, Richard. 2001. “Introduction.” In The Labyrinth of the Continuum: Writings on the Continuum Problem, 1672–1686, ed. Richard Arthur, xxiii–lxxxviii. New Haven: Yale University Press. Audi, Robert. 1993. Action, Intention, and Reason. Ithaca: Cornell University Press. 1979. “Weakness of Will and Practical Judgment.” Noûs 13 (2): 173–96. Barth, Christian. 2011. “Leibnizian Conscientia and Its Cartesian Roots.” Studia Leibnitiana 43 (2): 216–36. Baxter, Donald. 2000. “Leibniz on Contingent Conceptual Truths in the Arnauld Correspondence.” Studia Leibnitiana 32 (2): 191–214. Bennett, Jonathan. 2005. “Leibniz’s Two Realms.” In Leibniz: Nature and Freedom, eds. Donald Rutherford and Jan A. Cover, 135–55. Oxford: Oxford University Press. Blumenfeld, David. 1995. “Perfection and Happiness in the Best Possible World.” In The Cambridge Companion to Leibniz, ed. Nicholas Jolley, 382–410. Cambridge: Cambridge University Press. 1988. “Freedom, Contingency, and Things Possible in Themselves.” Philosophy and Phenomenological Research 49 (1): 81–101. 1973. “Leibniz’s Theory of the Striving Possibles.” Studia Leibnitiana 5 (2): 163–77. Bobro, Marc E. 2013. “Leibniz on Causation.” In The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Summer 2013 Edition) https://plato.stanford .edu/archives/sum2013/entries/leibniz-causation/. 2008. “Leibniz on Concurrence and Efficient Causation.” Southern Journal of Philosophy 46 (3): 317–38. 2004. Self and Substance in Leibniz. Dordrecht: Kluwer Academic Publishers. Bobro, Marc E. and Kenneth Clatterbaugh. 1996. “Unpacking the Monad: Leibniz’s Theory of Causality.” Monist 79 (3): 408–25. 208

References

209

Bolton, Martha Brandt. 2013. “Change in the Monad.” In The Divine Order, the Human Order, and the Order of Nature: Historical Perspectives, ed. Eric Watkins, 175–94. Oxford: Oxford University Press. 2011. “Leibniz’s Theory of Cognition.” In The Continuum Companion to Leibniz, ed. Brandon Look, 136–58. London: Continuum. 2008. “Locke and Leibniz on the Structure of Substance and Powers: The Metaphysics of Moral Subjects.” In Studies on Locke: Sources, Contemporaries, and Legacy, eds. Sarah Hutton and Paul Schuurmann, 107–26. Dordrecht: Springer. Brown, Gregory. 2016. “Leibniz on the Ground of Moral Normativity and Obligation.” Leibniz Review 26: 11–62. 2011. “Disinterested Love: Understanding Leibniz’s Reconciliation of Self- and Other-Regarding Motives.” British Journal for the History of Philosophy 19 (2): 265–303. 1992. “Is There a Pre-established Harmony of Aggregates in the Leibnizian Dynamics, or Do Non-substantial Bodies Interact?” Journal of the History of Philosophy 30 (1): 53–75. Carlin, Laurence. 2012. “Boyle’s Teleological Mechanism and the Myth of Immanent Teleology.” Studies in History and Philosophy of Science 43 (1): 54–63. 2006. “Leibniz on Final Causes.” Journal of the History of Philosophy 44 (2): 217–33. 2000. “Leibniz’s Great Chain of Being.” Studia Leibnitiana 32 (2): 131–50. Carriero, John. 2008. “Substance and Ends in Leibniz.” In Contemporary Perspectives on Early Modern Philosophy: Essays in Honor of Vere Chappell, eds. Paul Hoffman, David Owen, and Gideon Yaffe, 115–40. Peterborough: Broadview Press. 2005. “Spinoza on Final Causality.” In Oxford Studies in Early Modern Philosophy, eds. Daniel Garber and Steven Nadler. Vol. 2, 105–47. Oxford: Clarendon Press. 1996. “Symposium on Leibniz: Determinist, Theist, Idealist by Robert Merrihew Adams.” Leibniz Review 6: 61–77. 1993. “Leibniz on Infinite Resolution and Intramundane Contingency – Part One: Infinite Resolution.” Studia Leibnitiana 25 (1): 1–26. 1991. “Spinoza’s Views on Necessity in Historical Perspective.” Philosophical Topics 19 (1): 47–96. Chappell, Vere. 2005. “Self-Determination.” In Early Modern Philosophy: Mind, Matter, and Metaphysics, eds. Christia Mercer and Eileen O’Neill, 127–41. Oxford: Oxford University Press. Clatterbaugh, Kenneth. 1973. Leibniz’s Doctrine of Individual Accidents. Studia Leibnitiana Sonderheft 4. Wiesbaden: Steiner. Cover, Jan A. and John O’Leary-Hawthorne. 1999. Substance and Individuation in Leibniz. Cambridge: Cambridge University Press. Cox, Donovan. 2002. “Leibniz on Divine Causation: Creation, Miracles, and the Continual Fulgurations.” Studia Leibnitiana 34 (2): 185–207. Craig, Edward. 1996. The Mind of God and the Works of Man. Oxford: Oxford University Press. Davidson, Donald. 2001 [1970]. “How is Weakness of the Will Possible?” In Essays on Actions and Events. 2nd edn., 21–42. Oxford: Clarendon Press.

210

References

Davidson, Jack. 2005. “Video Meliora Proboque, Deteriora Sequor.” In Leibniz: Nature and Freedom, eds. Donald Rutherford and Jan A. Cover, 234–52. Oxford: Oxford University Press. 1998. “Imitators of God: Leibniz on Human Freedom.” Journal of the History of Philosophy 36 (3): 387–412. Della Rocca, Michael. 2013. “Striving, Oomph, and Intelligibility in Spinoza.” In Judgement and the Epistemic Foundation of Logic, ed. Maria van der Schaar, 49–65. Dordrecht: Springer. 1996. “Spinoza’s Metaphysical Psychology.” In The Cambridge Companion to Spinoza, ed. Don Garrett, 192–266. Cambridge: Cambridge University Press. Dennett, Daniel C. 1984. Elbow Room: The Varieties of Free Will Worth Wanting. Cambridge: MIT Press. Des Chene, Dennis. 1996. Physiologia: Natural Philosophy in Late Aristotelian and Cartesian Thought. Ithaca: Cornell University Press. Di Bella, Stefano. 2015. “Hylomorphism (Even) without Matter? Transtemporal Sameness and the Rehabilitation of Substantial Form in Leibniz’s Theory of Substance.” In Leibniz’s Metaphysics and Adoption of Substantial Forms: Between Continuity and Transformation, ed. Adrian Nita, 43–58. Dordrecht: Springer. 2008. “Causa Sive Ratio: Univocity of Reason and Plurality of Causes in Leibniz.” In Leibniz: What Kind of Rationalist? ed. Marcelo Dascal, 495–509. Dordrecht: Springer. Dicker, Georges. 1982. “Leibniz on Necessary and Contingent Propositions.” Studia Leibnitiana 14 (2): 221–32. Fischer, John Martin and Mark Ravizza. 1998. Responsibility and Control: A Theory of Moral Responsibility. Cambridge: Cambridge University Press. Frankel, Lois. 1984. “Being Able to Do Otherwise: Leibniz on Freedom and Contingency.” Studia Leibnitiana 16: 45–59. Frankfurt, Harry G. 1988. The Importance of What We Care About: Philosophical Essays. Cambridge: Cambridge University Press. Freddoso, Alfred J. 1988. “Introduction.” In Luis de Molina, on Divine Foreknowledge (Part IVof the Concordia), ed. Alfred J. Freddoso, 1–81. Ithaca: Cornell University Press. Garber, Daniel. 2015. “Monads on My Mind.” In Leibniz’s Metaphysics and Adoption of Substantial Forms: Between Continuity and Transformation, ed. Adrian Nita, 107–76. Dordrecht: Springer. 2009. Leibniz: Body, Substance, Monad. New York: Oxford University Press. Garrett, Don. 1990. “Truth, Method, and Correspondence in Spinoza and Leibniz.” Studia Spinozana 6: 13–41. Gennaro, Rocco J. 1999. “Leibniz on Consciousness and Self-Consciousness.” In New Essays on the Rationalists, eds. Rocco J. Gennaro and Charles Huenemann, 353–71. Oxford: Oxford University Press. Greenberg, Sean. 2005. “Leibniz against Molinism: Freedom, Indifference, and the Nature of the Will.” In Leibniz: Nature and Freedom, eds. Donald Rutherford and Jan A. Cover, 217–33. Oxford: Oxford University Press. Griffin, Michael V. 2013. Leibniz, God and Necessity. Cambridge: Cambridge University Press.

References

211

Grimm, Robert. 1970. “Individual Concepts and Contingent Truths.” Studia Leibnitiana 2 (3): 200–23. Hartz, Glenn A. 1996. “Exactly How Are Leibnizian Substances Related to Extension?” In Leibniz’s “New System” (1695), ed. Roger S. Woolhouse, 63–81. Florence: Olschki. Hawthorne, John and Jan A. Cover. 2000. “Infinite Analysis and the Problem of the Lucky Proof.” Studia Leibnitiana 32 (2): 151–65. Hirschmann, David. 1988. “The Kingdom of Wisdom and the Kingdom of Power in Leibniz.” Proceedings of the Aristotelian Society 88: 147–59. Hobbes, Thomas and John Bramhall. 1999. Hobbes and Bramhall: On Liberty and Necessity, ed. Vere Chappell. Cambridge: Cambridge University Press. Hoffman, Paul. 2009. “Does Efficient Causation Presuppose Final Causation? Aquinas vs. Early Modern Mechanism.” In Metaphysics and the Good: Themes from the Philosophy of Robert Merrihew Adams, eds. Samuel Newlands and Larry M. Jorgensen, 295–312. Oxford: Oxford University Press. Hoffmann, Thomas S. 1998. “Spontaneität.” In Historisches Wörterbuch der Philosophie, ed. Joachim Ritter. Vol. 9, 1424–34. Basel: Schwabe. Hostler, John. 1975. Leibniz’s Moral Philosophy. London: Duckworth. Imlay, Robert A. 2002. “Leibniz on Freedom of the Will: A Vindication.” Studia Leibnitiana 34 (1): 81–90. Irwin, Terence H. 1992. “Who Discovered the Will?” Philosophical Perspectives 6: 453–73. Jalabert, Jacques. 1960. Le Dieu de Leibniz. Paris: Presses Universitaires de France. Jauernig, Anja. 2010. “Disentangling Leibniz’s Views on Relations and Extrinsic Denominations.” Journal of the History of Philosophy 48 (2): 171–205. Johnson, Monte Ransome. 2005. Aristotle on Teleology. Oxford: Clarendon Press. Jolley, Nicholas. 2005. Leibniz. London: Routledge. 1998. “Causality and Creation in Leibniz.” Monist 81 (4): 591–611. Jorati, Julia. Forthcoming a. “Leibniz on Appetitions and Desires.” In History of Philosophy of Mind, eds. Christopher Shields and Becko Copenhaver. New York: Routledge. Forthcoming b. “Leibniz’s Ontology of Force.” In Oxford Studies in Early Modern Philosophy, eds. Daniel Garber and Donald Rutherford. Vol. 8. Oxford: Clarendon Press. 2017. “Gottfried Leibniz [on Free Will].” In The Routledge Companion to Free Will, eds. Kevin Timpe, Meghan Griffith, and Neil Levy, 293–302. New York: Routledge. 2015. “Three Types of Spontaneity and Teleology in Leibniz.” Journal of the History of Philosophy 53 (4): 669–98. 2014a. “Leibniz’s Twofold Gap between Moral Knowledge and Motivation.” British Journal for the History of Philosophy 22 (4): 748–66. 2014b. “Review of Michael V. Griffin, Leibniz, God and Necessity.” Journal of the History of Philosophy 52 (1): 172–3. 2013. “Monadic Teleology without Goodness and without God.” The Leibniz Review 23: 43–72. Jorgensen, Larry M. 2015. “Leibniz on Perceptual Distinctness, Activity, and Sensation.” Journal of the History of Philosophy 53 (1): 49–77.

212

References

2013. “By Leaps and Bounds: Leibniz on Transcreation, Motion, and the Generation of Minds.” The Leibniz Review 23: 73–98. 2011a. “Mind the Gap: Reflection and Consciousness in Leibniz.” Studia Leibnitiana 43 (2): 179–95. 2011b. “Leibniz on Memory and Consciousness.” British Journal for the History of Philosophy 19 (5): 887–916. 2009. “The Principle of Continuity and Leibniz’s Theory of Consciousness.” Journal of the History of Philosophy 47 (2): 223–48. Kaphagawani, Didier Njirayamanda. 1999. Leibniz on Freedom and Determinism in Relation to Aquinas and Molina. Brookfield: Ashgate. Knebel, Sven K. 2000. Wille, Würfel und Wahrscheinlichkeit: Das System der moralischen Notwendigkeit in der Jesuitenscholastik 1550–1700. Hamburg: Meiner. 1991. “Necessitas moralis ad optimum: Zum historischen Hintergrund der Wahl der besten der möglichen Welten.” Studia Leibnitiana 23 (1): 3–24. Kulstad, Mark. 2006. “Leibniz on Expression: Reflections after Three Decades.” In Einheit in der Vielheit: Vorträge; VIII. Internationaler Leibniz-Kongress, Hannover, 24. bis 29. Juli 2006, eds. Herbert Breger, Jürgen Herbst, and Sven Erdner. Vol. 1, 413–9. Hannover: Gottfried Wilhelm Leibniz Gesellschaft. 1993. “Causation and Preestablished Harmony in the Early Development of Leibniz’s Philosophy.” In Causation in Early Modern Philosophy: Cartesianism, Occasionalism, and Preestablished Harmony, ed. Steven Nadler, 93–118. University Park: Penn State University Press. 1991. Leibniz on Apperception, Consciousness, and Reflection. Munich: Philosophia. 1990. “Appetition in the Philosophy of Leibniz.” In Mathesis Rationis: Festschrift für Heinrich Schepers, eds. Albert Heinekamp, Wolfgang Lenzen, and Martin Schneider, 133–51. Munster: Nodus. 1977. “Leibniz’s Conception of Expression.” Studia Leibnitiana 9: 55–76. Lærke, Mogens. 2009. “The Problem of Alloglossia: Leibniz on Spinoza’s Innovative Use of Philosophical Language.” British Journal for the History of Philosophy 17 (5): 939–53. 2007. “Contingency, Necessity, and the Being of Possibility: Leibniz’s Modal Ontology in Relation to His Refutation of Spinoza.” Revue Roumaine de Philosophie 51 (1–2): 39–62. Lagerlund, Henrik and Peter Myrdal. 2006. “Possible Worlds and the Nature of Choice in Leibniz.” Studia Leibnitiana 38 (2): 156–76. Lee, Sukjae. 2014. “Toward a New Reading of Leibnizian Appetites: Appetites as Uneasiness.” Res Philosophica 91 (1): 123–50. 2006. “Leibniz on Spontaneity: A Sketch of Formal and Final Causation.” In Einheit in der Vielheit: Vorträge; VIII. Internationaler Leibniz-Kongress, Hannover, 24. bis 29. Juli 2006, eds. Herbert Breger, Jürgen Herbst, and Sven Erdner. Vol. 1, 440–7. Hannover: Gottfried Wilhelm Leibniz Gesellschaft. 2004. “Leibniz on Divine Concurrence.” Philosophical Review 113 (2): 203–48. Lin, Martin. 2014. “Efficient Causation in Spinoza and Leibniz.” In Efficient Causation: A History, ed. Tad M. Schmaltz, 165–91. Oxford: Oxford University Press. 2012. “Rationalism and Necessitarianism.” Nous 46 (3): 418–48. Lodge, Paul. 2015. “Corporeal Substances as Monadic Composites in Leibniz’s Later Philosophy.” In Leibniz’s Metaphysics and Adoption of Substantial Forms:

References

213

Between Continuity and Transformation, ed. Adrian Nita, 107–24. Dordrecht: Springer. Look, Brandon and Donald Rutherford. 2007. “Introduction.” In The Leibniz-Des Bosses Correspondence, eds. Brandon Look and Donald Rutherford, xix–lxxix. New Haven: Yale University Press. Lovejoy, Arthur O. 1936. The Great Chain of Being: A Study of the History of an Idea. Cambridge: Harvard University Press. MacDonald, Scott. 1991. “The Relation between Being and Goodness.” In Being and Goodness: The Concept of the Good in Metaphysics and Philosophical Theology, ed. Scott MacDonald, 1–28. Ithaca: Cornell University Press. Madden, James D. 2003. “Leibniz on Teleology and the Intelligibility of Nature.” Proceedings of the American Catholic Philosophical Association 77: 173–88. McDonough, Jeffrey K. 2016a. “Leibniz and the Foundations of Physics: The Later Years.” Philosophical Review 125 (1): 1–34. 2016b. “Leibniz on Monadic Agency and Optimal Form.” In Leibniz’s Experimental Philosophy, ed. Arnaud Pelletier, 93–120. Stuttgart: Steiner. 2011. “The Heyday of Teleology and Early Modern Philosophy.” Midwest Studies in Philosophy 35: 179–204. 2008. “Leibniz’s Two Realms Revisited.” Noûs 42 (4): 673–96. 2007. “Leibniz: Creation and Conservation and Concurrence.” The Leibniz Review 17: 31–60. McGuire, James E. 1976. “‘Labyrinthus Continui’: Leibniz on Substance, Activity, and Matter.” In Motion and Time, Space and Matter, eds. Peter K. Machamer and Robert G. Turnbull, 290–326. Columbus: The Ohio State University Press. McRae, Robert. 1976. Leibniz: Perception, Apperception, and Thought. Toronto: University of Toronto Press. Mele, Alfred R. 2012. Backsliding: Understanding Weakness of Will. Oxford: Oxford University Press. 1995. Autonomous Agents: From Self-Control to Autonomy. New York: Oxford University Press. 1987. Irrationality: An Essay on Akrasia, Self-Deception, and Self-Control. New York: Oxford University Press. Mitchell-Yellin, Benjamin. 2015. “The Platonic Model: Statement, Clarification and Defense.” Philosophical Explorations 18 (3): 378–92. 2014. “In Defense of the Platonic Model: A Reply to Buss.” Ethics 124 (2): 342–57. Mugnai, Massimo. 2012. “Leibniz’s Ontology of Relations: A Last Word?” Oxford Studies in Early Modern Philosophy 6: 171–208. 1992. Leibniz’ Theory of Relations. Studia Leibnitiana Supplementa 28. Stuttgart: Steiner. Murray, Michael J. 2005. “Spontaneity and Freedom in Leibniz.” In Leibniz: Nature and Freedom, eds. Donald Rutherford and Jan A. Cover, 194–216. Oxford: Oxford University Press. 2004. “Pre-Leibnizian Moral Necessity.” The Leibniz Review 14: 1–28. 1995. “Leibniz on Divine Foreknowledge of Future Contingents and Human Freedom.” Philosophy and Phenomenological Research 55 (1): 75–108. Nachtomy, Ohad. 2007. Possibility, Agency, and Individuality in Leibniz’s Metaphysics. Dordrecht: Springer.

214

References

Nelson, Alan. 2005. “Leibniz on Modality, Cognition, and Expression.” In A Companion to Rationalism, ed. Alan Nelson, 282–301. Malden: Blackwell. Newlands, Samuel. 2010. “The Harmony of Spinoza and Leibniz.” Philosophy and Phenomenological Research 81 (1): 64–104. O’Connor, Timothy and Constantine Sandis. 2010. A Companion to the Philosophy of Action. Malden: Wiley-Blackwell. O’Neill, Eileen. 1993. “Influxus Physicus.” In Causation in Early Modern Philosophy: Cartesianism, Occasionalism, and Preestablished Harmony, ed. Steven Nadler, 27–56. University Park: Penn State University Press. Osler, Margaret J. 2001. “Whose Ends? Teleology in Early Modern Natural Philosophy.” Osiris 16: 151–68. Pasnau, Robert. 2011. Metaphysical Themes 1274–1671. Oxford: Clarendon Press. 2002. Thomas Aquinas on Human Nature: A Philosophical Study of Summa Theologiae Ia, 75–89. Cambridge: Cambridge University Press. 2001. “Intentionality and Final Causes.” In Ancient and Medieval Theories of Intentionality, ed. Dominik Perler, 302–23. Leiden: Brill. Penner, Terrence. 1997. “Socrates on the Strength of Knowledge: Protagoras 351B–357E.” Archiv für Geschichte der Philosophie 79 (2): 117–49. 1990. “Plato and Davidson: Parts of the Soul and Weakness of Will.” Canadian Journal of Philosophy, Supplementary Volume 16: 35–74. Phemister, Pauline. 2005. Leibniz and the Natural World: Activity, Passivity and Corporeal Substances in Leibniz’s Philosophy. Dordrecht: Springer. 1991. “Leibniz, Freedom of Will and Rationality.” Studia Leibnitiana 26 (1): 25–39. Pickup, Martin. 2014. “Leibniz and the Necessity of the Best Possible World.” Australasian Journal of Philosophy 92 (3): 507–23. Puryear, Stephen. 2010. “Monadic Interaction.” British Journal for the History of Philosophy 18 (5): 763–96. Rateau, Paul. 2015. Leibniz et le meilleur des mondes possibles. Paris: Classiques Garnier. Rescher, Nicholas. 2001. “Contingentia Mundi: Leibniz on the World’s Contingency.” Studia Leibnitiana 33 (2): 145–62. Riley, Patrick. 1996. Leibniz’s Universal Jurisprudence: Justice as the Charity of the Wise. Cambridge: Harvard University Press. Rozemond, Marleen. 2009. “Leibniz on Final Causation.” In Metaphysics and the Good: Themes from the Philosophy of Robert Merrihew Adams, eds. Samuel Newlands and Larry M. Jorgensen, 272–94. Oxford: Oxford University Press. Russell, Bertrand. 1937. A Critical Exposition of the Philosophy of Leibniz. 2nd edn., London: George Allen & Unwin. 1903. “Recent Work on the Philosophy of Leibniz.” Mind 12 (46): 177–201. Rutherford, Donald. 2013. “Laws and Powers in Leibniz.” In The Divine Order, the Human Order, and the Order of Nature: Historical Perspectives, ed. Eric Watkins, 149–74. Oxford: Oxford University Press. 2008. “Leibniz on Infinitesimals and the Reality of Force.” In Infinitesimal Differences: Controversies between Leibniz and His Contemporaries, eds. Ursula Goldenbaum and Douglas Jesseph, 255–80. Berlin: de Gruyter.

References

215

2005. “Leibniz on Spontaneity.” In Leibniz: Nature and Freedom, eds. Donald Rutherford and Jan A. Cover, 156–80. Oxford: Oxford University Press. 1995a. Leibniz and the Rational Order of Nature. Cambridge: Cambridge University Press. 1995b. “Metaphysics: The Late Period.” In The Cambridge Companion to Leibniz, ed. Nicholas Jolley, 124–75. Cambridge: Cambridge University Press. Schmaltz, Tad M. 2014. “Moral Evil and Divine Concurrence in the Theodicy.” In New Essays on Leibniz’s Theodicy, eds. Larry M. Jorgensen and Samuel Newlands, 135–52. Oxford: Oxford University Press. 2012. “Substantial Forms as Causes: From Suárez to Descartes.” In Matter and Form in Early Modern Science and Philosophy, ed. Gideon Manning, 125–50. Leiden: Brill. 2008. Descartes on Causation. New York: Oxford University Press. Schmid, Stephan. 2016. “The Intrinsic Directedness of Leibnizian Forces.” In‘Für unser Glück oder das Glück anderer’: Vorträge des X. Internationalen Leibniz-Kongresses Hannover, 18.–23. Juli 2016, ed. Wenchao Li. Vol. 5, 131–41. Hildesheim: Olms. 2011. Finalursachen in der frühen Neuzeit: eine Untersuchung der Transformation teleologischer Erklärungen. Berlin: de Gruyter. Scott, David. 1998. “Leibniz’s Model of Creation and His Doctrine of Substance.” Animus 3: 73–88. Seeskin, Kenneth R. 1977. “Moral Necessity.” The New Scholasticism 51: 90–101. Seidler, Michael J. 1985. “Freedom and Moral Therapy in Leibniz.” Studia Leibnitiana 17 (1): 15–35. Shields, Christopher John. 1986. “Leibniz’s Doctrine of the Striving Possibles.” Journal of the History of Philosophy 24 (3): 343–57. Simmons, Alison. 2011. “Leibnizian Consciousness Reconsidered.” Studia Leibnitiana 43 (2): 196–215. 2001. “Changing the Cartesian Mind: Leibniz on Sensation, Representation and Consciousness.” The Philosophical Review 110 (1): 31–75. Sleigh, Robert C. 2009. “Moral Necessity in Leibniz’s Account of Human Freedom.” In Metaphysics and the Good: Themes from the Philosophy of Robert Merrihew Adams, eds. Samuel Newlands and Larry M. Jorgensen, 252–71. Oxford: Oxford University Press. 2005. “Introduction.” In Confessio philosophi: Papers Concerning the Problem of Evil, 1671–1678, ed. Robert C. Sleigh, xix–xli. New Haven: Yale University Press. 1999. “Leibniz on Freedom and Necessity: Critical Notice of Robert Adams, Leibniz: Determinist, Theist, and Idealist.” The Philosophical Review 108 (2): 245–77. 1990a. Leibniz & Arnauld: A Commentary on Their Correspondence. New Haven: Yale University Press. 1990b. “Leibniz on Malebranche on Causality.” In Central Themes in Early Modern Philosophy: Essays Presented to Jonathan Bennett, eds. Jan A. Cover and Mark Kulstad, 161–93. Indianapolis: Hackett. Smith, Angela M. 2005. “Responsibility for Attitudes: Activity and Passivity in Mental Life.” Ethics 115 (2): 236–71. Sotnak, Eric. 1999. “The Range of Leibnizian Compatibilism.” In New Essays on the Rationalists, eds. Rocco J. Gennaro and Charles Huenemann, 200–23. Oxford: Oxford University Press.

216

References

Stroud, Sarah. 2014. “Weakness of Will.” In The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Spring 2014 Edition) https://plato.stanford.edu/archives/ spr2014/entries/weakness-will/. Swoyer, Chris. 1995. “Leibnizian Expression.” Journal of the History of Philosophy 33 (1): 65–99. Temmik, Aloys. 1706. Philosophia vera theologiae & medicinae ministra, in theatrum mundi dijudicanda. Cologne: W. Michahelles & J. Adolph. Vailati, Ezio. 1990. “Leibniz on Locke on Weakness of Will.” Journal of the History of Philosophy 28 (2): 213–28. 1986. “Leibniz on Necessary and Contingent Predication.” Studia Leibnitiana 18 (2): 195–210. van Ruler, Johannes A. 1995. The Crisis of Causality: Voetius and Descartes on God, Nature, and Change. Leiden: Brill. von Bodelschwingh (Jorati), Julia. 2011. “Leibniz on Concurrence, Spontaneity, and Authorship.” The Modern Schoolman 88 (3/4): 267–97. Watson, Gary. 2004. Agency and Answerability: Selected Essays. Oxford: Clarendon Press. Whipple, John. 2015. “Leibniz on Substance and Causation.” In Locke and Leibniz on Substance, eds. Paul Lodge and Tom Stoneham, 203–30. New York: Routledge. 2010a. “Leibniz on Divine Concurrence.” Philosophy Compass 5 (10): 865–79. 2010b. “The Structure of Leibnizian Simple Substances.” British Journal for the History of Philosophy 18 (3): 379–410. Wilson, Catherine. 1989. Leibniz’s Metaphysics: A Historical and Comparative Study. Princeton: Princeton University Press. Wilson, Margaret Dauler. 1999. “Leibniz’s Dynamics and Contingency in Nature.” In Ideas and Mechanism: Essays on Early Modern Philosophy, 421–41. Princeton: Princeton University Press. Wolf, Susan R. 1990. Freedom within Reason. New York: Oxford University Press. Youpa, Andrew. 2013. “Leibniz’s Ethics.” In The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta (Summer 2013 Edition) https://plato.stanford .edu/archives/fall2013/entries/leibniz-ethics/.

Index

ability to do otherwise, 123 and determinism, 124, 161–62 and moral responsibility, 6, 180–81, 198, 202–6 and the reality of choice, 131 connected to control, 203–6 in God, 129–31, 138, 156, 203 in human beings, 130 accidents, 43, 44, see also modes accountability. See moral responsibility action, 23–24 deliberate. See action, voluntary difference between voluntary and involuntary, 45, 150–51 free, 53, 66, 114–47 in animals, 3, 67, 95 in broad sense, 46–47, 70, 77 in narrow sense, 46–47, 70, 71, 77 instinctive, 53, 67, 76, 84 intentional, 148, 165 involuntary, 66–67, 84, 204 voluntary, 53, 57, 66, 75–76, 77, 191–92, 204 activity. See also primitive force and causation, 28–29, 48 and occasionalism, 95 cannot be grounded in matter, 11–12 ideal. See ideal influence in experience, 50 in substances, 14, 31 required for substancehood, 29, 100, 111 vs. passivity, 2, 4, 37, 47–49 Adams, Robert Merrihew, 14, 16, 18, 123, 131 on choice, 124 on concurrence, 106 on contingency, 131, 132, 133, 135 on divine power, 130, 139 on moral necessity, 127, 134 on perfection, 185 on privation, 100 akrasia, 6, 148–49, 162–75 and control, 151, 162–63

and freedom, 164–65, 168 and moral responsibility, 196 and regret, 168 as ignorance, 163, 173 definition, 164–65 diachronic, 168–69 diachronic vs. synchronic, 167–68 synchronic, 169–74 vs. compulsion, 174–76 alternative possibilities, 161–62, 178–79, see also ability to do otherwise Anfray, Jean-Pascal, 74, 127, 138, 139 angels, 26, 43 animals, 13, 26–27, 67 and attributability, 95 lack capacity for happiness and suffering, 188–89 lack control, 151, 155–56 lack freedom, 55 lack moral agency, 181–82, 187–89 lack moral knowedge, 189 lack rational teleology, 76, 137 lack self-knowledge, 187–88 possess immanent teleology, 88 Antognazza, Maria Rosa, 13, 100 Apparent Good Interpretation (AGI), 80, 84, 85 and akrasia, 163 problems, 81, 88 appetitions, 4, 6, 16, 17–23 and goodness, 85, 169 as teleological, 60–61, 68 conflict, 20–22, 87 conscious vs. unconscious, 22 definition, 17–18, 68 distinct, 23 for ends vs. means, 18–20, 171 global, 20 long-term, 18–20, 21 natural, 61, 86 rational vs. non-rational, 22–23 relation to perceptions, 23, 24–26 sensible vs. insensible, 22–23

217

218

Index

appetitions (cont.) strong vs. weak, 21–22, 177–78 terminology, 17–18 Aquinas, Thomas, 204 on angels, 43 on appetitions, 23, 68 on control, 77 on freedom and responsibility, 137, 204 on goodness, 84 on human agency, 57 on instrumental causation, 94 on privation, 98 on secondary causation, 97 on spontaneity, 50 on substantial forms, 60–61, 89–90 on teleology, 60–64, 75, 88–90 on the ability to do otherwise, 162 on voluntariness, 204 Aristotle, 12, 88 on akrasia, 163, 164, 169 on voluntariness, 50 Arnauld, Antoine, 122 Arthur, Richard, 34, 36 attributability, 5, 81, 95 and concurrence, 92, 94, 97 and control, 151 and efficient causation, 96 and ideal influence, 95 and omissions, 201 of passions, 95 attributability problem, 92, 94, 96–105 solution, 97, 104, 107–10, 112 Audi, Robert, 165, 168 Augustine, 97, 182 author of an action, 94–96, 97–103, 105–7, see also attributability and moral responsibility, 180, 199–201 broad vs. narrow sense, 95, 107 vs. instrument, 105 automata, 136, 147, 200 autonomy, 2, 5, 38, 56, 76, 117, 151 Barth, Christian, 17 Baxter, Donald, 122, 145 Bennett, Jonathan, 28 best possible world, 9, 10, 79, 82, 155, 184–86 contingently best, 139–40 blame, 181, 196–99, 202, 204–5, see also moral responsibility, punishment blind thoughts, 170, 172, 174 Blumenfeld, David, 74, 138, 185 Bobro, Marc, 30, 31, 32, 33, 37, 40, 136, 200 body, 11, 13, 26, 35, 42, 49, 52, 87 Bolton, Martha, 17, 21, 25, 30, 80 bondage, 153–54, 163, see also servitude

Bramhall, John, 150 Brown, Gregory, 52, 191, 194 brute facts, 39, 140 Carlin, Laurence, 30, 59, 63, 89, 121 Carriero, John on arguments against materialism, 11 on contingency, 114, 122, 132 on goodness-directedness, 80 on teleology, 62, 63, 64, 73, 83, 89, 107 on violent change, 63 causal chains, 30, 115, 117–18 causal interaction, 29, 37–46, 48, 90, 110 causation agent, 2, 5, 31, 115, 118 deficient, 102–3, 104, 108 efficient, 27–32, 63, 66, 95, 102–3, 106, 107, 115 efficient presupposes final, 61–65, 69, 83, 104 event, 5, 30, 36, 115 final. See teleology formal, 5, 28, 33, 92–93, 103–12, 118 instrumental, 94, 101, 104, 107 occasional, 32–33, 34 primary vs. secondary, 97, 109 productive. See causation, efficient state, 30, 36 substance, 31, 115 Chappell, Vere, 56, 117 choice, 134–35, 138, 142, 152, 164, 167, 198, see also freedom, will real, 124, 131–32 City of God, 181–87, 190, 193–94 Clatterbaugh, Kenneth, 25, 30, 31, 32 coaction. See compulsion compatibilism, 2, 5–6, 115, 117–18, 119, 124, 148 complete concept and causal interaction, 40–42 and contingency, 122, 141, 142–47 and creation, 10, 51, 105–7, 109 and moral necessity, 143–46 core, 143–47 hierarchical structure, 143–47 compossibility, 73 compulsion, 6, 56, 149, 174–79 external vs. internal, 176–77 vs. akrasia, 163, 164, 174–76 conatus, 18, 19, 53, 68 concurrence, 5, 45, 92–113, 118, 120 general, 96–97, 100, 101 special, 96–97, 100, 101, 103, 107, 112 consciousness, 17, 22, 67, 85–87 Consequence Argument, 161

Index conservation, 93, 101, 107 constraint, 45, 54, 155, 176–77, see also compulsion contingency, 5–6, 114–15, 121–47, 154 and complete concepts, 142–47 and freedom, 121–47 and moral necessity, 126, 133 and moral responsibility, 202–6 and teleology, 116, 122–42 in unfree actions, 123, 146 infinite analysis account, 146–47 of states of affairs vs. actions, 139 continuity, 12, 18–19, 26, 34–35, 66, 67 control, 6, 148–62 and akrasia, 162–63 and compulsion, 177–79 and freedom, 197 and God, 153 and moral responsibility, 178, 181, 203–6 and spontaneity, 54, 152–54 and the ability to do otherwise, 203–6 and voluntary agency, 77 as worth having, 154–56, 158–59 can be lacking in rational agents, 154–55, 192–93 direct, 152–56, 177 direct vs. indirect, 152 failures of, 155, 197 indirect, 151, 156–62, 175, 177, 178, 203–6 over one’s actions, 152 over one’s character, 157 over one’s nature, 148, 150 over one’s passions, 153 over one’s will, 152, 156–62 terminology, 151 counterfactuals, 161 counterpart theory, 161 Cover, Jan, 39, 43, 144, 146 Cox, Donovan, 41 Craig, Edward, 163 creation, 9–10, 48, 144, 184–86 and authorship, 106 and contingency, 131 and striving possibles, 74 as actualization of essences, 51, 82, 105, 118–19 formal and final cause of, 107 Leibniz vs. Aquinas, 89 Davidson, Donald, 164, 165 Davidson, Jack, 164, 167, 168, 171, 173–74, 197 defective reception interpretation, 98, 100–1

219 deliberation and freedom, 53, 57, 116, 121, 131, 165, 195–97 and reasons-responsiveness, 160 and the reality of choice, 131 divine, 73–74, 117, 131, 145 lacking in animals, 77, 156 Della Rocca, Michael, 62, 172 Dennett, Daniel C., 148, 154 derivative force, 14, 18, 34–35 Des Chene, Dennis, 33, 76 Descartes, René, 93 desert, 182–83, 185, 192, 203, 205 desire, 18, 46, 56, 86, 87, 169, see also appetitions blind, 76, 77 strength or motivational force of, 170–73, 174 determination. See also spontaneity by the good. See necessity, moral divine, 118–19 external, 39, 49, 69, 117–20, 147, 153–54 formal, 104 indirect, 117 inevitable, 125 rational, 53, 54, 104 value-neutral, 115, 130, 132–37, 138 value-sensitive, 115, 132–37 violent. See violent change vs. necessitation, 125 determinism, 15 and akrasia, 148, 178 and compulsion, 174–79 and control, 150–51, 156, 159 and teleology, 136 and the Consequence Argument, 161 compatible with freedom, 5, 115, 117–19, 147, 148–49 psychological, 166, 206 Di Bella, Stefano, 16, 41 Dicker, Georges, 145 divided effort interpretation, 98, 99–100 efficacious appetition view, 30, 118 efficacious perception view, 30, 118 egoism, 190, 194 end-directedness. See teleology endeavor. See appetitions, conatus endorsement theory of autonomy, 55–57, 76 entelechy, 8, 10, 12–13, 60 essences and attributability, 108 and contingency, 131, 133, 142–47 in divine intellect, 10, 51, 82, 89, 105–7, 118 eternal truths, 9, 95, 108, 187, 189–90

220

Index

evil, 94, 185, 197, 198, see also privation, sin explanation, 33, 35, 41, 52, 79, 82 externalism about judgment and motivation, 174 final and formal causation model, 97, 103–12 Fischer, John Martin, 148, 159–62 Frankel, Lois, 202 Frankfurt, Harry, 37, 38 Freddoso, Alfred, 96 freedom, 2, 5–6, 114–47 and akrasia, 164–65, 168 and concurrence, 92, 101, 112, 118 and contingency, 25, 121–47 and control, 148, 150 and determinism, 5, 115, 117, 147, 148 and moral necessity, 126–28, 133 and spontaneity, 53–55, 92, 94, 117–20 and teleology, 66, 75–76 conditions, 116, 147 definition, 116 degrees, 57, 116, 153, 197 divine vs. human, 116–17 not required for moral responsibility, 180–81, 194–99 of indifference, 101, 166, 167 Garber, Daniel, 10, 39, 40, 41, 123 Garrett, Don, 122 Gennaro, Rocco, 17 God, 9, 26 able to do otherwise, 129–31, 138, 156, 203 and moral motivation, 190 as final cause of the world, 88, 107 as mind or spirit, 186 as sole efficient cause, 32, 103 lacks passivity, 10, 16 relationship to finite minds, 182, 186–87, 190, see also City of God relationship to natural world, 182 role in creaturely actions, 92–113 goodness. See also perfection and teleology, 78–80, 82–90, 132 divine, 136, 138, 140 in agent teleology, 78, 82 in metaphysical teleology, 78, 82 in rational teleology, 78, 82 ipso facto, 79–80, 83, 84–85, 89 of character, 140 standard of, 139–40 types, 84 goodness-directedness and contingency, 115 and rational teleology, 75–76

in the least perfect changes, 67, 78 vs. teleology, 59, 78–80, 82–83, 91 Greenberg, Sean, 45, 121, 150 Griffin, Michael, 74, 133, 138 Grimm, Robert, 145 guidance control, 159–62 habits, 157–58, 192, 196–97 happiness, 182, 188, 193–94 harmony, 43, 48–49, 66, 72, 109, 183 Hartz, Glenn, 73 Hirschmann, David, 28 Hobbes, Thomas, 85 Hoffman, Paul, 62, 63 Hoffmann, Tobias, 39, 50 Hostler, John, 163 ideal influence, 47, 90, 109 and attributability, 95, 107 and spontaneity, 46–52, 57 and teleology, 71–73 identity over time, 13, 16, 188 image of God, 116, 186–87, 191–92 Imlay, Robert, 163 inclination, 18, 126, see also appetitions prevalent, 141–42, 145, 151 vs. necessitation, 124–26, 141–42 infinite analysis, 146–47 intellect, 6, 53, 114, 155 and moral agency, 180, 187–90 divine, 9 role in free actions, 121 intellectualism, 6, 149, 163, 173 intelligibility, 42, 43, 62, 110 internalism about judgment and motivation, 174 ipso facto goodness, 79–80, 83, 84–85, 89 Irwin, Terence, 149 Jalabert, Jacques, 123 Jauernig, Anja, 43 Johnson, Monte Ransome, 12, 88 Jolley, Nicholas, 30, 80, 163, 171, 172, 191 Jorgensen, Larry, 17, 65, 78, 82 joy, 188 judgment, 53, 54, 77, 114, 121, 127, 154, 168 justice, 182–86, 189, 192, 193 corrective vs. retributive, 194, 199 different characterizations, 185 to animals, 189 Kaphagawani, Didier, 55 kingdom of efficient causes, 27

Index kingdom of final causes, 27 kingdom of grace. See City of God Knebel, Sven, 127 Kulstad, Mark, 14, 17, 18, 19, 30, 137 Lærke, Mogens, 50, 122, 123, 131, 133 Lagerlund, Henrik, 131, 135 law of the series, 8, 14–15, 35, 42 laws of justice, 182–86, 190 laws of physics, 183–86 Lee, Sukjae, 32, 37, 99, 100 on agency, 23 on efficient causation, 28, 107 on final and formal causation, 33, 103–4, 106 on teleology, 80, 81, 87 libertarianism, 117, 148, 166, 202 liberty. See freedom limitation. See privation Lin, Martin, 29, 134, 144, 195 Locke, John, 170, 188 Lodge, Paul, 14 Look, Brandon, 14 love, 190, 193–94 Lovejoy, Arthur, 138 lucky proof objection, 146 MacDonald, Scott, 89 machines, 136, 200 Madden, James, 61, 82 marks and traces, 15, 41 mastery. See control matter, 11–13, see also body McDonough, Jeffrey, 28 on appetitions, 20, 170 on concurrence, 92, 100, 101 on derivative force, 18 on efficient causation, 28 on teleology, 80 McGuire, James, 36 McRae, Robert, 23, 25, 66 mechanism, 8, 11, 64 Medea, 165 Mele, Alfred, 151, 152, 163, 164, 170 memory, 26–27, 187 minds, 26–27, 67 have a special status, 184–87, 191–92 miracles, 137, 145, 183, 186 and violent change, 70, 73, 83, 90, 120 definition, 111 Mitchell-Yellin, Benjamin, 56 modes, 10, 11, 16, 17, 24–26, 27, 30, 34–36, 96, 144 modifications. See modes Molina, Luis de, 50, 96, 101, 137 monadic agency view, 31

221 monads, 4, 8–16, 26–27, see also souls, minds bare, 26–27, 67, 81, 83, 85, 140 central or dominant, 13–14, 26, 52, 73 contain past and future, 14–15, 62, 106 subordinate, 13–14, 73 terminology, 10 windowless, 43 monism, 11, 100 moral agency, 6, 136, 180–94, 199 appetitive requirements, 191–94 cognitive requirements, 187–90 requires capacity for rational spontaneity, 193 requires capacity for rational teleology, 193 requires capacity to love God, 190, 193–94 requires knowledge of eternal truths, 189–90 requires knowledge of God, 189–90 requires reflection or self-knowledge, 187–89 terminology, 181 moral knowledge, 180, 189–90 moral obligation, 127, 190, 196 moral psychology, 2, 6, 149, 163 moral responsibility, 6, 136, 194–207 and akrasia, 164–65, 175 and compulsion, 175 and control, 151–52, 159, 160–61, 203–6 and freedom, 194–99 as asymmetrical, 181, 205–6 for attitudes, 202 for omissions, 201, 204 requires ability to do otherwise, 202–6 requires contingency, 194–95, 202–6 motives for action, 125, 126, 170–73, 190, 194 Mugnai, Massimo, 22, 40, 43, 54, 87 Murray, Michael, 39, 45, 46, 50, 68, 120, 127 Myrdal, Peter, 131, 135 Nachtomy, Ohad, 40, 122, 127, 128, 131, 141, 143, 163 natural change, 63–65 natures and attributability, 105–6 and complete concepts, 42, 144–45 and efficient causation, 63–65 and formal causation, 33 and primitive force, 14–15 and spontaneity, 39, 46 and teleology, 68, 70, 77, 83, 90, 141 broad vs. narrow, 63–65, 75 in Aquinas, 89–90 include imperfections, 90 independent of God’s plan, 51, 82, 89, 105, 118

222

Index

necessitarianism and complete concepts, 142–47 and counterfactuals, 161 incompatible with freedom, 6, 116 Leibniz’s attitude toward, 114–15, 121–23, 133–37, 142, 147 necessity absolute, 124–28, 137–42 blind or brute, 122, 126, 134–37, 147 extrinsic, 133 hypothetical, 126, 128, 137–42 intrinsic, 133 metaphysical, 124, 125, 126–28, 137–42 moral, 124, 125, 126–28, 130–31, 132, 133, 137, 139, 142 moral as sub-category of metaphysical, 138, 143 physical, 127, 128, 161, 177 Nelson, Alan, 143 Neutral Teleology Interpretation (NTI), 78, 82–91 Newlands, Samuel, 122 normativity, 83, 141 O’Neill, Eileen, 44 O’Connor, Timothy, 1 O’Leary-Hawthorne, John, 39, 43, 144, 146 Objective Good Interpretation (OGI), 81–82, 87, 163 obligation. See moral obligation occasionalism, 11–12, 14, 32–33, 39, 95, 111 omissions, 195, 201, 204 omnipotence, 129 omniscience, 9, 138 organism, 13–14, 26, 52, 73 Osler, Margaret, 59 ought implies can, 190, 193 Ovid, 165 pain, 85–87, 188–89 Pasnau, Robert, 33, 60, 69 passions, 7, 54, 77, 86, 95, 153–56, 195, see also passivity and indirect control, 156–62 as external, 153–54 passivity, 11, 24 in experience, 50 vs. activity, 2, 4, 37, 47–49 Penner, Terry, 167, 168 per se necessity, 132 perceptions, 4, 16–17, 36 abstract, 171, 172, 174 as actions, 23–24 as isomorphism, 16 confused, 17, 24, 66, 153, 171

definition, 16–17 distinct, 17, 47 distinct vs. vivid, 171 of entire world, 17, 20 of pain, 85–87 relation to appetitions, 24–26 role in monadic change, 32–36 symbolic, 171 vivid, 170–74, 177, 179 percepturitio, 18 perfection, 11–12 and ideal influence, 48, 49, 51, 72, 109 and spontaneity, 53 and striving possibles, 73–74 and teleology, 71, 72, 73, 75–76, 83 in Aquinas, 89 moral, 184, 185 natural or metaphysical, 184, 185 perceptions of, 85 strict vs. loose sense, 84 personal identity, 188 Phemister, Pauline, 18, 23, 80, 151, 187, 191, 197 philosophy of action, 1, 3, 92, 151–52 physical forces, 21 physical influx, 44 Pickup, Martin, 122, 139, 140 plants, 3, 13, 26, 83 Platonic model of agency, 55–57, 76, 149, 155 pleasure, 85–87, 172, 188–89 possibility, 161–62, see also necessity possible worlds, 129, 131, 140 possibles, 10, 129, see also essences strive for existence, 73–74, 107 possible-world-semantics, 161 power divine, 129–31 human, 130 to do otherwise. See ability to do otherwise vs. will, 124, 129–31 praise, 181, 199, 205, see also moral responsibility, reward for unfree actions, 196–97 predicate contained in notion of subject, 142–47 essential, 144 fundamental vs. non-fundamental, 143–47 prime matter, 13 primitive force, 8, 10, 12–16, 18, 34–36, 46, 69 Principle of Continuity, 12, 66, 67 Principle of Contradiction and necessity, 126, 128, 129–31, 139–40, 145–46 in Spinoza, 134

Index Principle of Sufficient Reason and causation, 62 and contingency, 140–41, 145–46 and freedom, 147 and necessitarianism, 122 violations, 39, 140, 145, 146, 167 Principle of the Identity of Indiscernibles, 26 Principle of Wisdom and Goodness, 81, 127, 128–31, 137–40, 191 privation, 97–103, 201 and deficient causation, 102–3, 108 has no efficient cause, 102 proper function, 83 providence, 60, 136 punishment, 182–83, 185, 188, 195 and freedom, 198–99 corrective vs. retributive, 194, 199 Puryear, Stephen, 48, 72 Rateau, Paul, 194 rationalism, 175 Ravizza, Mark, 148, 159–62 real self, 2, 4, 38, 55–57, 76, 117, 151–52, 154, 155 reason, 26–27, 187, see also intellect reasons-responsiveness, 159–62 reflection, 26–27, 187–89 relations, 43 Rescher, Nicholas, 127 responsibility. See moral responsibility retribution, 194, 199 reward, 182–83, 188 Riley, Patrick, 150 Rozemond, Marleen, 81 Russell, Bertrand, 73, 114, 138 Rutherford, Donald, 10, 14, 25, 28, 36, 45, 46, 151 on appetitions, 18, 30 on contingency, 128 on perceptions, 35 on spontaneity, 46, 52 on teleology, 68, 81, 82, 89, 163 Sandis, Constantine, 1 Schmaltz, Tad, 69, 94, 102–3 Schmid, Stephan, 31, 32, 34, 65, 78, 82, 123, 133 Scott, David, 100 Seeskin, Kenneth, 128 Seidler, Michael, 50, 150, 151 self-determination. See spontaneity self-knowledge, 187–89 sensation, 26–27, 63, see also pain, pleasure servitude, 56, 153–54, 176, 196, 198 shared agency, 92, see also concurrence

223 Shields, Christopher John, 74 Simmons, Alison, 17 simple privation model, 97, 98–102 sin, 185, 198 and attributability, 5, 94, 97–103, 106 and formal causation, 108 and punishment, 183 and the ability to do otherwise, 202 can be unfree, 195 resulting from habit, 197 Sleigh, Robert, 30, 37, 98, 99, 100, 114, 122, 127, 132, 144 Smith, Angela, 196 sophisticated privation model, 97, 102–3, 108 Sotnak, Eric, 45 souls, 26–27, 67, see also monads species, 89 Spinoza, Baruch de and blind necessity, 122 and teleology, 122, 132, 134–36 on goodness, 85 on motivation, 172 Spinozism, 121–23, 133–37, 142, 194 spontaneity, 4–5, 14, 37–58, 92–93 agent, 4, 38, 46–52, 53, 57, 58, 71, 77, 95, 120 and autonomy, 2 and control, 152–54 and freedom, 114, 117–20 and omissions, 201 connected to teleology, 65, 68–69, 136 degrees, 45, 57 divine, 39, 54, 117 explanatory, 52, 110, 111–12 metaphysical, 4, 38, 46, 52, 53, 58, 65, 70, 77, 95, 120 productive, 111, 112 rational, 4, 38, 52–58, 77, 112, 114, 119–20, 152–54 required for moral responsibility, 199–201 spontaneity problem, 93–94, 97, 112 solution, 110–11 spontaneity thesis, 37–46, 84, 88, 91, 93 striving possibles, 73–74, 107 Stroud, Sarah, 164 Suárez, Francisco, 28, 43, 106, 137 on concurrence, 101 on instrumental causation, 94 substance, 13–14, 43, see also monads as dynamic, 128 substancehood, 11, 14, 92, 94, 111–12 substantial form as normative, 141 in Aquinas, 60–61, 89–90

224

Index

substantial form (cont.) relation to formal and final causation, 33, 69, 105, 106 relation to monads, 8, 10, 13 vs. complete concept, 42, 107 Swoyer, Chris, 17 teleology, 5, 6, 28, 33, 48, 59–91, 118 agent, 5, 70–74, 75, 77, 91, 132 and autonomy, 2 and concurrence, 5, 92–93, 103–12 and contingency, 6, 122–42, 147 and explanation, 34, 79 and freedom, 114 and goodness, 78–80, 82–90, 91, 132 and moral agency, 180 and omissions, 201 as ubiquitous, 59, 60–68 connected to spontaneity, 65, 68–69, 136 desire, 81 extrinsic, 59, 71 immanent, 59, 60, 65, 69, 71, 72–73, 81, 88, 90 metaphysical, 5, 70, 75, 77, 90 natural, 81 neutral, 78, 82–91 rational, 5, 75–77, 91, 114, 124, 132 required for moral responsibility, 199–201 thick sense, 64 thin sense, 62 telepathy, 43, 44 Temmik, Aloys, 87 tendency, 18, 24, see also appetitions, primitive force thought, 27 tracing approach, 160, 162 truth, 40, 142–47

unity, 10–11, 13, 73 Vailati, Ezio, 144, 164, 171, 173, 174 van Ruler, Johannes, 69 violent change, 61, 63–65, 69, 72, 73, 90, 120 and moral responsibility, 199–200 virtue, 193, 196, 205 vivacity, 170–74, 177 volitions. See will voluntariness. See action, voluntary Watson, Gary, 38, 117, 163 on akrasia, 163, 174 on autonomy, 56, 76 on compulsion, 175, 178 on control, 149, 151–52 on freedom, 2, 117, 121 on the ability to do otherwise, 130 weakness of will. See akrasia Whipple, John, 19, 25, 29, 34, 36, 93, 98 will, 6, 53, 114, 121, 155 and moral agency, 180, 191–94 can be lacking in intelligent creatures, 154–55, 192–93 determined by intellect, 121, 134, 149–50 importance of, 121, 150 vs. power, 124, 129–31 Wilson, Catherine, 74 Wilson, Margaret Dauler, 123 Wolf, Susan, 1 on control, 151–52, 159 on moral responsibility, 181, 193, 205–6 on the real self, 117 Youpa, Andrew, 163, 194