Idealization XIV: Models in Science [1 ed.] 9789004318847, 9789004318298

Citation preview

Idealization xiv: Models in Science

Poznań Studies in the Philosophy of the Sciences and the Humanities Founding Editor Leszek Nowak (1943–2009) Editor-in-Chief Katarzyna Paprzycka (University of Warsaw) Editors Tomasz Bigaj (University of Warsaw) – Krzysztof Brzechczyn (Adam Mickiewicz University) – Jerzy Brzeziński (Adam Mickiewicz University) – Krzysztof Łastowski (Adam Mickiewicz University) – Joanna OdrowążSypniewska (University of Warsaw) – Piotr Przybysz (Adam Mickiewicz University) – Mieszko Tałasiewicz (University of Warsaw) – Krzysztof Wójtowicz (University of Warsaw) Advisory Committee Joseph Agassi (Tel-Aviv) – Wolfgang Balzer (München) – Mario Bunge (Montreal) – Robert S. Cohen (Boston) – Francesco Coniglione (Catania) – Dagfinn Føllesdal (Oslo, Stanford) – Jacek J. Jadacki (Warszawa) – Andrzej Klawiter (Poznań) – Theo A.F. Kuipers (Groningen) – Witold Marciszewski (Warszawa) – Thomas Müller (Konstanz) – Ilkka Niiniluoto (Helsinki) – Jacek Paśniczek (Lublin) – David Pearce (Madrid) – Jan Such (Poznań) – Max Urchs (Wiesbaden) – Jan Woleński (Kraków) – Ryszard Wójcicki (Warszawa)

VOLUME 108

Idealization The titles published in this series are listed at brill.com/idea

Idealization xiv: Models in Science Edited by

Giacomo Borbone Krzysztof Brzechczyn

LEIDEN | BOSTON

Poznań Studies is sponsored by the University of Warsaw. Cover illustration: Morguefile Library of Congress Cataloging-in-Publication Data Names: Borbone, Giacomo, 1981- editor. Title: Idealization XIV : models in science / edited by Giacomo Borbone, Krzysztof Brzechczyn. Other titles: Idealization 14 | Idealization fourteen Description: Boston : Brill-Rodopi, 2016. | Series: Poznań studies in the philosophy of the sciences and the humanities, ISSN 0303-8157 ; VOLUME 108 | Includes bibliographical references. Identifiers: LCCN 2016014301 | ISBN 9789004318298 (hardback : alk. paper) Subjects: LCSH: Social sciences--Philosophy. | Science--Philosophy. | Humanities--Philosophy. Classification: LCC H61 .I33 2016 | DDC 501/.1--dc23 LC record available at http://lccn.loc.gov/2016014301

Want or need Open Access? Brill Open offers you the choice to make your research freely accessible online in exchange for a publication charge. Review your various options on brill.com/brill-open. Typeface for the Latin, Greek, and Cyrillic scripts: “Brill.” See and download: brill.com/brill-typeface. issn 0303-8157 isbn 978-90-04-31829-8 (hardback) isbn 978-90-04-31884-7 (e-book) Copyright 2016 by Koninklijke Brill nv, Leiden, The Netherlands. Koninklijke Brill nv incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, ma 01923, usa. Fees are subject to change. This book is printed on acid-free paper and produced in a sustainable manner.

Contents List of Tables and Figures vii Notes on Contributors viii The Role of Models in Science: An Introduction 1 Giacomo Borbone and Krzysztof Brzechczyn

Part 1 Idealization: General Problems 1 On Fictions, Theoretical Entities, and Ideal Objects: Applying Zalta’s Abstract Objects Theory to Scientific Theories 13 Xavier de Donato Rodríguez and José L. Falguera 2 The Inherent Type of Scienific Law, The Idealized Types of Scientific Law 43 Igor Hanzel 3 On Deformational Modeling: Max Weber’s Concept of Idealization 63 Lidia Godek 4 On Reduction in the Idealizational Theory of Science: A Case Study of the Relationship between the Concept of Rational Act and the Concept of a Habitual-Rational Action 81 Mieszko Ciesielski

part 2 Idealization in the Social Sciences 5 Idealization in Economics: A Structuralist View 113 Adolfo García de la Sienra 6 Between Isolations and Constructions: Economic Models as Believable Worlds 130 Łukasz Hardt

vi

Contents

7

The Fallacy of the Reification of Idealization in Economic Research 161 Adam Czerniak

8

Strategies of Comparative Analysis in Historical Comparative Sociology: An Attempt at an Explication within the Conceptual Framework of the Idealizational Theory of Science 184 Krzysztof Brzechczyn

9

The Structure of Idealization in Noam Chomsky’s Generativist Theory 204 Barbara Konat

Part 3 Idealization in the Humanities 10

The Method of Idealization and Concretization on the Ground of Negativistic Unitarian Metaphysics 219 Krzysztof Kiedrowski

11

The Choice of the Principles of Justice in the Political Philosophy of John Rawls: An Idealizational Interpretation 244 Piotr Przybysz

12

Inductive Modeling and Discursive Idealization in Scenario Planning 270 Zenonas Norkus

13

Historical Narration in the Light of the Idealizational Theory of Science: A Recapitulation and Expansion 299 Piotr Szwochert

List of Tables and Figures Tables 4.1 A comparison of factors examined in the cra and ch-ra 99 5.1 Empirical structure ​D̂ ​ represented as a table 117 8.1 Explanation: sbm – a semi-bureaucratic absolute monarchy, hbas – a highlybureaucratic absolutist state, sba – a semi-bureaucratic absolutist state, hbs – a highly-bureaucratic state, nba – a non-bureaucratic state 189 10.1 The methods of unitarian metaphysics and their role in the system 226 10.2 The categorial and idealizational characteristics of the unitarian metaphysics’ models 227

Figures 1.1 The whole class of abstract objects and some its subclasses: numbers, fictional characters, stories and theoretical entities. As a part of the class of the stories, there is the class of scientific theories 29 4.1 A comparison of the cra and ch-ra essential structures, taking into ­account the conceptual compatibility; an arrow signifies the relationship of the ­explication of cra concepts in the language of ch-ra 101 4.2 A comparison of statements concerning particular types of action: in the ­concept of a rational act and in the concept of habitual-rational action 104 5.1 Approximated imbedding of empirical structure into a partial potential model 118 7.1 Illustration of the VaR metric 165 8.1 The model of successful revolution 188 9.1 Idealization thresholds in science (Nowak 1977, p. 60) 208 10.1 Structure of Nowak’s negativistic unitarian metaphysics 232 10.2 The structure of negativistic metaphysics’ system 241 11.1 Thresholds of idealization in science 266 12.1 David Holmgren’s global climate change and energy descent scenarios 281 13.1 → one-sided influence, ↔ mutual influence 312 13.2 Where → one-sided influence; ↔ two-sided influence 315

Notes on Contributors Giacomo Borbone is a research fellow at the Department of Human Sciences, Catania University, Italy. [email protected] Krzysztof Brzechczyn is professor at the Institute of Philosophy, Adam Mickiewicz University, Poznań, Poland. [email protected] Mieszko Ciesielski is a research fellow at the Institute of European Culture in Gniezno, Adam Mickiewicz University, Poznań, Poland. [email protected] Adam Czerniak is an assistant professor at the Department of Economics ii in the Warsaw School of Economics and a chief economist at the center for policy analysis Polityka Insight, Warsaw, Poland. [email protected] Xavier de Donato Rodríguez is temporary professor at the Faculty of Philosophy, University of Santiago de Compostela. Spain [email protected] José L. Falguera is associate professor at the Faculty of Philosophy, University of Santiago de Compostela. Spain [email protected] Adolfo García de la Sienra is a research professor at the Instituto de Filosofía of the Universidad Veracruzana, in Xalapa, Veracruz, Mexico. [email protected] Lidia Godek is a research fellow at the Institute of Philosophy, Adam Mickiewicz University, Poznań, Poland. [email protected] Igor Hanzel is a senior research worker at the Department of Logic and Methodology of Science, Comenius University in Bratislava, Slovakia. [email protected]

Notes on Contributors

ix

Łukasz Hardt is an associate professor at the Faculty of Economic Sciences, University of Warsaw, Poland. [email protected] Krzysztof Kiedrowski is a research fellow at the Institute of Philosophy, Adam Mickiewicz University in Poznań, Poland. [email protected] Barbara Konat is a postdoctoral researcher at the Institute of Philosophy, Adam Mickiewicz University in Poznań, Poland. [email protected] Zenonas Norkus is professor at the Department of Sociology, Faculty of Philosophy, Vilnius University, Lithuania. [email protected] Piotr Przybysz is an assistant professor at the Institute of Philosophy, Adam Mickiewicz University, Poznań, Poland. [email protected] Piotr Szwochert graduated philosophy from the Institute of Philosophy, Adam Mickiewicz University, Poznań, Poland. [email protected]

The Role of Models in Science: An Introduction Giacomo Borbone and Krzysztof Brzechczyn Many years ago Galileo Galilei wrote that “if a geometrical philosopher wants to find in the concrete the effects demonstrated in the abstract, he must ­deduct the impediments of matters” (Galilei, Opere, p. 449). In this statement, the ­Italian physician is telling us that in order to investigate scientific phenomena we must turn our gaze to a new theoretical procedure no longer based on Aristotelian abstraction but on what modern philosophers of science call ­idealization. This methodological and theoretical revolution allowed Galileo to overcome the limits and fallacies of the old Aristotelian-Ptolemaic system, but how was it possible? It was possible thanks to the systematic use of idealizations and models in science. In contrast to what has commonly been maintained by inductive philosophies or even by positivist and post-positivist approaches to science, idealization is the core of scientific method. However, we should differ between Aristotelian abstraction and idealization and in this respect the first philosopher of science which did that in a systematic way was the Polish philosopher and co-founder of the Poznań School of Methodology, Leszek Nowak: A scientific law is basically a deformation of phenomena. It resembles much more the logical structure of a caricature than that of the generalization of facts. The crucial point for a proper understanding of the Idealizational procedure is that it differs fundamentally from that of abstraction […]. Abstraction, i.e. the omitting of properties, leads from individuals to sets of individuals (and from sets of individuals to families of sets, etc.). Idealization does not do this. Omission of the dimensions of physical bodies does not yield any set of physical bodies but the masspoint. Abstraction is generalization. Idealization is not. nowak 1992, p. 10–11

This means that while abstraction is applied by the human intellect to obtain universal concepts from the knowledge of particular objects (by the generalization of empirical facts), whereas ‘idealization’ proceeds “enclosing in parenthesis” some aspects of phenomenal reality that we consider secondary, to take into consideration the essential factors of the phenomena under investigation. An application of idealizations and models in the works of Galileo, Isaac Newton and other scientists gave birth to the modern scientific revolution

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_002

2

Borbone and Brzechczyn

b­ ecause just simple observation and generalization of empirical facts were ­unsatisfactory for the description and explanation of scientific phenomena. Let’s analyze, for instance, the problem of movement in Aristotle’s physics: he used to state that we have two types of motion, that is to say natural and forced. Therefore, according to the common observation of moving objects, there are two factors that exert their influence on the movement of bodies, that is to say, external forces and resistance of the environment. But Nowak says, “the question of how objects behave when there is no resistance was never answered by Aristotle, for ‘first it should be made clear what a vacuum between the ­moving bodies is. Since for that matter it has never been noticed within this world’ (Aristotle, Physics, 216b)” (Nowak 2000b, p. 18). In fact, Aristotle said, “there is no necessity for there being a void if there is movement” (­Aristotle Physics, iv, 7, 214a). Galileo, in contrast, knew very well that a vacuum cannot be perceived, but nonetheless he asked himself a question which was senseless in Aristotle’s physics. Namely how will “the perfect round ball moves on the plane which is smoothly leveled in order to eliminate all external and accidental obstacles” (Galilei 1962, p. 155) upon an assumption that the “resistance arising when the ball makes its way and all other obstacles that could arise” (Galilei 1962, p. 155) are not considered at all? Galileo answered to this apparently senseless question postulating that: 1. the rolling ball is perfectly round; 2. the plane is ideally smooth; 3. the plane is perfectly spherical; 4. the resistance of the environment is equal zero.1 Thanks to these idealizing assumptions, Galileo formulated his famous law of inertia, applied to an obvious ideal case that does not exist in reality but idealization which, as we said previously, differs from abstraction because the former refers to ideal objects that allowed us to overcome the abyss e­ xisting between essence and appearance.2 Historically speaking, Aristotelian abstraction acted as a kind of epistemological obstacle to the birth of modern science but, as Amos Funkenstein wrote:

1 We are based on Nowak’s reconstruction (Nowak 2000, p. 18). 2 As McMullin wrote: “When Galileo was faced with complex real-world situations, however, he “idealized” in more specific ways. That is, he shifted the focus to a simpler analogue of the original problem, one that lent itself more easily to solution” (McMullin 1985, p. 254–255).

3

The Role of Models in Science

The strength and novelty of seventeenth-century science, both theoretical and experimental, was in its capacity to take things out of the ­context and analyze their relations in ideal isolation. It was a new form of abstraction, or generalization; and it was recognized by many who employed it as new, as the source of the advantage of the new science of nature over the old. funkenstein 1989, p. 75

However, idealization needs another fundamental procedure called concretization, or, as McMullin would say, de-idealization (McMullin 1985) but before to explain this procedure it’s necessary to understand what an idealizing assumption is; it consists in the following propositional function: ​p​(x)​ = 0​ If and only if 0 denotes the minimum value of magnitude p and it holds for any real object x, (R)x. Thus, an idealizational statement will be a general proposition with the following form: ​(​T​ k​)​: G​(x)​ ∧ ​p1​ ​(x)​ = 0 ∧ ​p2​  ​(x)​ = 0 ∧ ⋯ ∧ ​pk​  ​(x)​ = 0 → F​(x)​ = ​fk​ ​(H​(x)​)​ Where G(x) is a realistic statement, while pi(x) = 0 is a generic idealizing condition. In this case (Tk) determines how a given magnitude depends on its ­principal factors. We can find examples of idealizing assumptions in Das K ­ apital, where Marx postulated a society formed by only two social classes (capitalists and workers), while we (and Marx too) know that these are not the only existing classes in a society, in Noam Chomsky’s linguistics (Nowak 2000a; 2004), in Darwin’s in biology (Łastowski 1977) and so on. Even though an idealizational statement describes an ideal state and not a real one, it’s possible anyway to pass from idealizing conditions to less idealizing ones, closer to reality through the procedure of concretization. In order to do that, we must delete the assumption pk(x) = 0, and, introduce the realistic condition pk(x) ≠ 0. The result of this operation is what Nowak calls concreti­ zation of the statement Tk: ​(​T​ k−1​)​: G​(x)​ ∧ ​p1​ ​(x)​ = 0 ∧ ​p2​  ​(x)​ = 0 ∧ ⋯ ∧ ​pk​  ​(x)​ ≠ 0 →  F​(x)​ = ​fk−1 ​  ​(H​(x)​, ​pk​  (​ x)​)​ Obviously, we can further concretize this formula until we obtain, finally, a factual statement, that’s to say, a statement containing no idealizing assumptions:

4

Borbone and Brzechczyn (​ ​T ​ 1)​ ​: G​(x)​ ∧ ​p​ ​(x)​ = 0 ∧ ​p​  ​(x)​ ≠ 0 ∧ ⋯ ∧ ​p​  ​(x)​ ≠ 0 →  1

2

k

F​(x)​ = ​f1​ ​(H​(x)​, ​pk​  ​(x)​, …, ​p2​  ​(x)​)​ (​ ​T ​ 0)​ ​: G​(x)​ ∧ ​p​ ​(x)​ ≠ 0 ∧ ​p​  ​(x)​ ≠ 0 ∧ ⋯ ∧ ​p​  ​(x)​ ≠ 0 →  1 2 k F​(x)​ = ​f0​  (​ H​(x)​, ​pk​  (​ x)​, …, ​p2​  (​ x)​, ​p1​ (​ x))​ ​ Basing his analysis on (T 1) and (T 0), Nowak states that mature science proceeds by systematic idealizations and concretization. Thus in science we work in the following way: (1) we introduce idealizing assumptions, (2) we formulate idealizing law and (3) we gradually concretize this law until we obtain the factual law without idealizing assumptions.3 However, in the scientific practice, the procedure of concretization is interrupted at a certain moment and idealizational law is approximated to empirical dates. The approximation of idealizational law relies on a waiving of all idealizing assumptions and attributing a certain value to remaining secondary factors. This value a1 is different from a minimal value of the secondary factor that is allocated to it in idealizing assumptions and from the actual values adopted by it. However, this value should not exceed a certain threshold of accuracy ε accepted in a given science at a given time. (​ A​T ​ 0)​ ​: G​(x)​ ∧ ​p​ ​(x)​ ≤ ​a​ ​ ∧ ​p​  ​(x)​ ≠ 0 ∧ ⋯ ∧ ​p​  ​(x)​ ≠ 0 →  1

1

2

F​(x)​ ​≈ε  f​0  ​(H​(x)​, ​pk​  ​(x)​, …, ​p2​  ​(x)​, ​p1​ ​(x)​)​

k

In this way, it is easy to realize that idealizational laws are the ­fundamental types of law formulated in physics and in order to demonstrate this we can take from physics some examples of idealizational laws: the law of inertia, the law of free fall, Boyle-Mariotte and Gay-Lussac’s laws (for perfect gases), Ostrogradski-Gauss’s formula, Ohm’s law and so on. Unfortunately, a lot of scientists and philosophers of science never gave the right attention to the importance of idealizations and models in scientific practice and the case of neo-positivism, for instance, is quite emblematic. Some outstanding philosophers of science such as Evandro Agazzi (1974), Edward N. Zalta (1983), Ronald Giere (1988), Nancy Cartwright (1983), Francesco Coniglione (2010), Ignazio Licata (2008) and so on, have recently focused their attention on the importance of idealizations and models in ­science. Obviously, we must not forget the so-called Semantic View of ­Theories that Patrick Suppes formulated in order to criticize the Syntactic View. The latter tried to solve the relationship between theory and observation by the 3 About the procedures of idealization and concretization see: Nowak (1970; 1971; 1972, pp. 533–548; 1975a,b; 1980) and their presentations: Borbone (2011; 2013; 2014).

The Role of Models in Science

5

use of the correspondence rules, while the former states that the complicated relationship between theories and empirical facts is mediated by models. This means that on the one side there are the theories that specify or define the idealized systems and, on the other side, models that constitute a structure that satisfies these definitions. In the end, between models and real systems there is a relationship of similarity that is more or less sufficient but never perfect. The importance of models in science is the topic of this new volume of “Poznań Studies in Philosophy of the Sciences and the Humanities,” devoted to the main applications of models and idealization in scientific practice (both in natural and social sciences). This volume is divided into three parts: “Idealization. General Problems,” “Idealization in Social Sciences” and “Idealization in the Humanities”. In the first essay “On Fictions, Theoretical Entities, and Ideal Objects: Applying Zalta’s Abstract Objects Theory to Scientific Theories,” Xavier de Donato Rodríguez and José L. Falguera, apply Edward N. Zalta’s theory of abstract objects to scientific theories. In Zalta’s view an abstract object does not spatiotemporally exist and in this respect his theory (zaot) includes also fictional and impossible entities. According to this theory, an ordinary object exemplifies properties while an abstract objects encodes them. De Donato Rodríguez and Falguera aim to extend the zaot conceptual framework to scientific terms and scientific theories which constitute some types of abstract objects. Igor Hanzel in his essay, “The Inherent Type of Scientific Law, the Idealized Types of Scientific Law,” takes into account three types of idealizations: the pure idealized type of scientific law, the inherent type of scientific law and the inherent idealized type of scientific law. He uses the first type in order to improve Nowak’s reconstruction of Newton’s Laws, while the second one and the third one are employed by Hanzel in order to provide a different explanation and reconstruction of the Marxist theory of value. Lidia Godek in her paper “On Deformational Modeling. Max Weber’s Concept of Idealization,” considers the Weberian concept of idealization. As it is well known, there is a difference between Weberian and Nowakian concept of idealization: the first one is anti-naturalist and instrumentalist while Nowak’s idealization is essentialist and naturalist. Nonetheless, Godek tries to go deeper and she also analyzes Weber’s concept-construction procedure in Nowak’s typology of deformational procedures (Nowak 1990). In this typology strong (reduction and transcendentalisation) and weak deformational procedures (negative and positive potentialisation) are distinguished. Let us assume that we have an initial object O, provided with a certain set of properties, which are characterised by a certain intensity. As a consequence of a transcendentalisation, object O’ is provided with certain additional properties which the

6

Borbone and Brzechczyn

initial object does not have. The contrary one is a reduction, which relies on depriving object O” of certain properties in comparison with the initial object. As a result of the use of positive potentialisation, the properties of object O’ are characterised by a greater intensity than the properties of the antecedent object O. The procedure of negative potentialisation, instead, eventuates in an object, the properties of which are characterised by a lesser intensity than the intensities of the antecedent object’s properties. According to Godek, Weberian ideal-type method can be seen as a method of positive potentialization. The last essay of the first section is Mieszko Ciesielski’s “On Reduction in the Idealizational Theory of Science; a Case Study of the Relationship between the Concept of a Rational Act and the Concept of a Habitual-Rational Action”. Ciesielski examines the relations between the Concept of Rational Act (cra) and the Concept of Habitual-Rational Action (ch-ra) using Paprzycka’s formulation of reduction between scientific theories. Ciesielski concludes that the relations between two concepts do not satisfy the three criteria of reduction: explicatory, essentialist and derivative. In the explicatory dimension the condition of a reduction is fulfilled as the theories are conceptually compatible. In the essentiality dimension the condition of a reduction is not fulfilled as the theories are in essential disagreement. Finally, in the derivative dimension the condition of a reduction is partially fulfilled as not all ch-ra concretizations are formed and one can only derive the most idealized cra statements in the ch-ra terms. According to Ciesielski the concept of reduction proposed by Paprzycka is applied to exact sciences. If we want to apply the concept of reduction in the humanities and social sciences using mostly qualitative language, we should ‘weaken’ the criteria for relations of reduction. The second section includes the following essays: Adolfo García de la Sienra, “Idealization in Economics: A Structuralist View,” Łukasz Hardt, “Between Isolations and Constructions: Economic Models as Believable Worlds,” Adam Czerniak, “Fallacy of Reification of Idealization in Economic Research,” Krzysztof Brzechczyn, “Strategies of Comparative Analysis in Historical-Comparative Sociology. An Attempt at an Explication within the Conceptual Framework of Idealizational Theory of Science” and Barbara Konat, “The Structure of Idealization in Noam Chomsky’s Generativist Theory.” Adolfo de la Sienra analyzes Marxian views on the method of political economy and he states that what Marx calls “the concrete totality in thinking” (Gedankenkonkretum), should not be confused with a theoretical model because it is nothing more than a general concept. He illustrates his consideration by analyzing the model of the ideal consumer which has idealizational status. Łukasz Hardt criticizes and tries to overcome the dichotomist polarity between isolations and constructions and that is why he prefers to see models

The Role of Models in Science

7

as believable worlds. The first polarity can be found in Nowak’s approach, while the second one treats models as constructions. According to Hardt, models are not merely idealizations and nor are they only constructions, but rather believable worlds that can provide us justifiable beliefs about the way the world works. Adam Czerniak in his essay (“Fallacy of Reification of Idealization in Economic Research”) investigates the relationship between idealization and economics. The author gives evidence of the crucial role that mathematical models played in the economic research. He also focusses his attention on the fact that a lot of economists often used abstract models to explain a social reality without applying the fundamental procedure of concretization and this is what is referred to as a fallacy of reification of idealization. According to Czerniak this methodological mistake of economists contributed to the recent global financial crisis. Krzysztof Brzechczyn explicates, in the light of idealization theory of science, the main types of comparative analysis used in historical-comparative sociology, namely the parallel, contrast-oriented, and a macro-causal comparative methods. In the latter part of the paper he summarizes the methodological discussion surrounding The State and Social Revolution authored by Theda Skocpol. According to him, the explication comparative method used by Skocpol in the notion apparatus of idealizational theory of science would allow for the avoidance methodological confusion and misinterpretations aroused around her work and to interpret the status of comparative historical sociology itself. The last essay of the second section authored by Barbara Konat is devoted to the methodological status of Chomsky’s generativist theory. As we know, Nowak applied a differentiation between the period of immature science and the period of theoretical science. In the first period science applies the induction as its fundamental scientific practice, while in the second one science applies the method of idealization. According to Nowak only with Chomsky did linguistics became a mature science and that was possible because the American linguist applied the method of idealization and concretization. Konat also extends the conceptual apparatus of the idealizational theory of science to Chomsky’s generativist theory and concludes, paraphrasing Nowak, that if Marx was the Galileo of the social sciences, Chomsky, for the same reasons, should be named the Galileo of linguistics. The third part of the book, entitled “Idealization in the Humanities,” includes the following essays: Krzysztof Kiedrowski, “The Method of Idealization and Concretization in the Negativistic Unitarian Metaphysics”, Piotr Przybysz, “The Choice of Principles of Justice in the Political Philosophy of John Rawls. An Idealizational Interpretation”, Zenonas Norkus, “Inductive Modeling

8

Borbone and Brzechczyn

and Discursive Idealization in the Scenario Planning” and Piotr Szwochert, “Historical Narration in the light of Idealizational Theory of Science. A Recapitulation and an Attempt at Expansion”. Kiedrowski’s essay is devoted to one of the main aspects of Nowak’s thought, that is to say Negativistic Unitarian Metaphysics, which is based on the same procedure of idealization and concretization that Nowak systematized in his early epistemological works. Nowak’s metaphysics criticizes the so-called Parmenine’s dogma, which affirms the identity between being and thought (positiveness of being). Nowak’s approach, in turn taken from Hegel, also attributes an important role to “negativity,” that in Nowak’s metaphysical conception is also part of being. Przybysz discusses the methodological elements that can be found in John Rawls’ Theory of Justice. Rawls is well known for his ideas on democratic liberalism but what is important here is the methodological reconstruction made by Przybysz. Through an analysis of Rawls’ research practice Przybysz states that the methodology used by Rawls corresponds to Nowak’s method of idealization and concretization. Zenonas Norkus (“Inductive Modeling and Discursive Idealization in the Scenario Planning”) explores the main features of modeling and idealization in so-called scenario planning. According to the author, scenarios are analytical narratives or complex causal stories about the distant future which are constructed in future studies and strategic management. Scenarios differ both from conventional scientific predictions and science fiction and social utopias because they are inductive models grounded in the discursive multi-model idealizations. The last essay in the volume is “Historical Narration in the light of Idealizational Theory of Science. A Recapitulation and an Attempt at Expansion.” Piotr Szwochert analyzes in it the relationship between historical narration and idealization and in this respect he takes into account two key theories of truth: the Aristotelian and the Hegelian. The former is based on common observation while the latter is characterized by a deformational procedure that enables us to catch the essential factors of a phenomenon in order to omit the secondary ones. Assumed by him idealizational theory of naration is based on hegelian notion of truth. Szwochert tries to implement this theoretical approach to historical narration to some pieces of historiography and modifies idealization theory of historical narration. In conclusion, we must be aware that both natural and social sciences work through models and idealizations and the history of mature science has also shown us that the characteristics of a system cannot be reduced to a single

The Role of Models in Science

9

formal model. Indeed, the opposite appears true, in that, a model-builder can use different models in order to provide different explanatory perspectives. References Agazzi, E. (1974). Temi e problemi di filosofia della fisica. Rome: Abete. Aristotle (1984). Physics, translated by R.P. Hardie and R.K. Gaye. In: Aristotle, Complete Works, vol. 1, edited by J. Barnes, Princeton: Princeton University Press. Borbone, G. (2011). The Legacy of Leszek Nowak. Epistemologia 34(2), 233–258. Borbone, G. (2013). Leszek Nowak e la scienza come idealizzazione. Saggio sull’approccio idealizzazionale alla scienza. Il Protagora 19, 91–144. Borbone, G. (2014). Metodologiczne dziedzictwo Leszka Nowakowa. Poznańskie Studia z Filozofii Nauki 23(1), 141–166. Cartwright, N. (1983). How the Laws of Physics Lie. Oxford: Oxford University Press. Coniglione, F. (2010). Realtà e astrazione. Scuola polacca ed epistemologica post-­ positivistica. Acireale-Rome: Bonanno. Funkenstein, A. (1989). Theology and the Scientific Imagination from the Middle Ages to the Seventeenth Century. Princeton: Princeton University Press. Galilei, G. (1811). Opere, vol. XI. Milan: Società Tipografica dé Classici Italiani. Galilei, G. (1962). Dialogue on Two Principal Systems of the World, Ptolemaic and C ­ opernican. Warsaw: PWN. Giere, R. (1988). Explaining Science: A Cognitive Approach. Chicago: University Chicago Press. Łastowski, K. (1977). The Method of Idealization in the Populational Genetics, Poznań Studies in the Philosophy of the Sciences and the Humanities 3(1–4), 201–212. Licata, I. (2008). La logica aperta della mente. Turin: Codice. McMullin, E. (1985). Galilean Idealization. Studies in History and Philosophy of Science 16, 247–273. Nowak, L. (1970). O zasadzie abstrakcji i stopniowej konkretyzacji [On the Principle of Abstraction and Gradual Concretization]. In: J. Topolski (ed.) Założenia metodologiczne “Kapitału” Marksa, pp. 123–218. Warszawa: KiW. Nowak, L. (1971). U podstaw Marksowskiej metodologii nauki [Foundations of the ­Marxian Methodology of Science]. Warszawa: PWN. Nowak, L. (1972). Laws of Science, Theories, Measurement, Philosophy of Science 39(4), 533–548. Nowak, L. (1975a). Abstraction, Idealization and Model, Teoria a Metoda 7(4), 23–26. Nowak, L. (1975b). Relative Truth, the Correspondence Principle and Absolute Truth, Philosophy of Science 42(2), 187–202.

10

Borbone and Brzechczyn

Nowak, L. (1980). The Structure of Idealization: Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht/Boston/London: Reidel. Nowak, L. (1990). Abstracts Are Not Our Constructs. The Mental Construct Are Abstracts. In: J. Brzeziński, F. Coniglione, T.A.F. Kuipers (eds.), Idealization I: ­General Problems. Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 16, pp. 193–206. Amsterdam-Atlanta: Rodopi. Nowak, L. (1992). The Idealizational Approach to Science: a Survey. In: J. Brzeziński and L. Nowak (eds.), Idealization III: Approximation and Truth. Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 25, pp. 9–63. Amsterdam-Atlanta: Rodopi. Nowak, L. (2000a). Ajdukiewicz, Chomsky and the Status of the Linguistic Theory of Natural Language. In: I. Nowakowa and L. Nowak, Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and in the Humanities, vol. 69, pp. 325–337. Amsterdam-Atlanta: Rodopi. Nowak, L. (2000b). Galileo-Newton’s Model of Free Fall. In: I. Nowakowa and L. Nowak, Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and in the Humanities, vol. 69, pp. 17–62. Amsterdam-Atlanta: Rodopi Nowak, L. (2004). L’unità nascosta di scienze sociali e scienze naturali. In: F. Coniglione and R. Poli (eds.) La realtà modellata. L’approccio idealizzazionale e le sue applicazioni nelle scienze umane, pp. 237–272. Milano: Franco Angeli. Zalta, E.N. (1983). Abstract Objects: An Introduction to Axiomatic Metaphysics. ­Dordrecht: Reidel.

Part 1 Idealization: General Problems

∵

chapter 1

On Fictions, Theoretical Entities, and Ideal Objects: Applying Zalta’s Abstract Objects Theory to Scientific Theories* Xavier de Donato Rodríguez and José L. Falguera Abstract In the present contribution, we apply Zalta’s axiomatic theory of abstract objects to scientific theories, theoretical entities (the referents of theoretical terms), and a special kind of theoretical entity we may call ‘ideal objects’. We begin by summarizing some problems of a certain realist approach to the reference of theoretical terms, namely, the causal theory of reference. We maintain that hybrid approaches mixing the causal and the descriptivist theories cannot really be applied to theoretical terms. Then we propose to understand scientific theories and their components (theoretical entities, ideal objects, models) as abstract objects and use Zalta’s theory as a framework to make the idea more precise. As it is shown in the paper, this proposal has several advantages.

1 Introduction As is well known, Hacking (1983) distinguished between realism about theories and realism about theoretical entities, both theses being independent of each another. Cartwright (1983) and Hacking himself both count as defenders of entity realism plus antirealism about theories. Though there are variations, the central idea for rejecting realism about theories whereas accepting entity realism is the same in both authors, namely that theories usually are complex constructs that typically contain highly idealized laws which are strictly speaking false (or true only of the entities in the model), whereas the entities which scientists are alleged to speak about are real in so far as they can be manipulated or can be made to causally interact in experimental processes: “If you can spray them, then they are real” – so says Hacking’s motto. Entity realism * This paper has received financial support from the research programs FFI2013-41415-P of the Spanish Ministry of Economy and Competitiveness and PICT-2012-2662 of the National Agency of Scientific and Technological Promotion, Argentina.

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_003

14

de Donato Rodríguez and Falguera

contrasts with van Fraassen’s constructive empiricism (van Fraassen 1980). While accepting a realist semantics, van Fraassen contends a form of epistemological antirealism, according to which the aim of science is searching for theories that are just empirically adequate without having a commitment either to the truth of the sentences of the theory that refer to unobservable entities or to the existence of unobservable (theoretical) entities themselves. As far as the existence of theoretical entities and the truth of theoretical s­ entences are ­concerned, van Fraassen remains agnostic. Leaving aside the very debatable issue whether constructive realism is really a convincing thesis,1 van Fraassen’s position may, prima facie, seem a more coherent position. As some authors have noticed,2 the idea of being a realist about theoretical entities while remaining skeptic about the (approximate) truth of scientific theories involves at first sight a tension that requires an appropriate explanation. By resorting to causal manipulation in experimental procedures, entity realists argue that, even if theories are false, we have good reasons to think that the entities that they postulate really exist. But the only unquestionable thing, at least if we are not radical skeptics, is that there is something out there which is probably causing such and such effects that are observable and measurable. We have no idea about whether the kind of entity out there causing those effects is really as the theory says it is. The only serious position for entity realists is to reject descriptivism and maintain some version of the causal theory of reference applied to theoretical terms.3 The bad news for the entity realist is that the causal theory of reference cannot be ­successfully applied in this context. We will argue this in the next section and, at the same time, we will advocate for a particular version of the descriptivist theory of reference. In van Fraassen’s approach, given that he accepts a realist semantics, theoretical terms may have reference (not all terms, of course; for example, phlogiston or ether would lack of reference). It is only that we cannot say that they have. But this does not matter, because we do not need to suppose that they refer to something outside, to real (though unobservable) parcels of the world. 1 See: Rosen (1994); Bandyopadhyay (1997); Ladyman (2000), and Sainsbury (2010, pp. 161–168) for some critical points related to van Fraassen’s position, mainly from the ontological point of view. 2 See, for example, Morrison (1990, p. 6 and ff.); Resnik (1998, p. 1179), and Chakravartty (1998, pp. 396–397; 2007, p. 31, and 2014, Section 2.3). 3 Entity realism does not imply per se the causal theory of reference. It is an independent (­compatible) thesis. But what we are saying is that entity realists should probably be ­committed to it if they want to have a coherent and solid position about reference.

On Fictions, Theoretical Entities, and Ideal Objects

15

For someone that defends (i) that scientific theories are constructions that involve many mathematical abstractions and formulate statements that hold only under very idealized situations, and also defends (ii) a d­ escriptivist theory of the reference of theoretical terms, it is difficult (if even possible) to still maintain that theoretical terms really refer to (kinds of) parcels of the world. In the present paper, we defend (i) as well as (ii), though we present the descriptivist theory in a new version in order to avoid the difficulties commonly associated with this thesis (see next section for details). We will offer an analysis based on Zalta’s axiomatic theory of abstract objects to account for the reference of theoretical terms and of what later will be called ‘ideal objects’ (see also de Donato Rodríguez and Falguera 2014b for details). So, the plan of the paper runs as follows. In the following section (2), we briefly comment on two main traditions that have dealt with the issue of the reference of theoretical terms and argue that one of them, the causal theory of reference (in fact, a hybrid version which combines the causal and the descriptivist theory), is basically wrong when applied to theoretical terms. In Section 3, we try to argue that a suitable version of the descriptivist theory, one which makes use of the notion of abstract object, can have many advantages when applied to scientific contexts. More particularly, we use Zalta’s theory as a framework to analyze the reference of theoretical terms and the nature of ideal objects. This will allow us to extract some interesting conclusions on the nature of idealization and scientific models and on its relation to fiction and abstract objects. 2

The Scientific Realism Debate and the Reference of Theoretical Terms

Let us begin by distinguishing two main approaches to the issue of the reference of theoretical terms in the philosophy of science. The first assumes some form of scientific realism and holds that the theoretical terms of (approximately) correct theories do really refer; that is, they designate (types of) spatiotemporally existing entities. Proposals in this tradition also seek to justify the continuity of reference for theoretical terms between rival theories in order to account for the idea of theoretical progress. Contenders of this approach usually appeal to some version of the Causal Theory of Reference (ctr). In fact, any acceptable version of ctr for theoretical terms establishes that in order to fix the reference of a theoretical term it is necessary to use a description, because there are no exemplars or samples of that reference which can be determined by ostension. For this reason, the causal theories

16

de Donato Rodríguez and Falguera

that are at issue are those that have been called “hybrid theories” (i.e. theories that combine ctr with the descriptivist account), as in the case of Kitcher (1993) or Psillos (1999). According to the second approach we want to consider, the descriptivist approach, the reference of theoretical terms is given by means of a description provided by the theory. The reference of a theoretical term, if the term has a reference at all (i.e. it is not just like ‘phlogiston’), is a (kind of) spatiotemporal entity which has those properties that the theory attributes to it. This approach faces at least two problems. The first is that, if theoretical descriptions are constitutive of the meaning of theoretical terms, what can guarantee that scientists advocating different versions of the atomic theory are really speaking about the same (type of) entity (i.e. the atom)? And how can Dalton really disagree with Bohr if they do not speak about (approximately) the same thing? The second problem is the lack of reference in the case of terms that belong to theories that are considered to be false (according to our current status of knowledge). If the descriptions associated with the term in question do not really apply to anything, then the term does not refer at all. This would be the case of terms like ‘phlogiston’, ‘caloric’, or ‘ether’. Of course, a descriptivist could easily accept this consequence, but the first problem seems (and is) much more delicate, because rational progress between theories and rational disagreement between scientists (about the same issue) are placed in question. Clearly, descriptivism needs to either join ctr in the form of a hybrid theory (as in the case of Psillos 1999) or find a more subtle formulation. In the first case, one remains a realist. In the second, the options that are available are typically non-realist (or even antirealist): on one hand, the Ramsey-Carnap formalist approach to theoretical terms, which influenced Sneed (1971) and the German Structuralist School ( Balzer, Moulines, and Sneed 1987), according to which the references of theoretical terms are mathematical entities that are connected with (observable) physical entities in some way that needs to be specified (see, for example, Carnap [1958] 1975); and on the other, Kuhn’s idea of constituted kinds, i.e. kinds that are constituted by human action though they may sometimes correspond to ways in which nature is actually carved out (Kuhn 1991 and 1993). The present contribution is somewhat related to this suggestion (though it has also some similarities with the Ramsey-Carnap approach). 2.1 The Difficulties of Applying ctr to Theoretical Terms Prominent hybrid versions of ctr for theoretical terms include Kitcher’s (1978 and 1993) and Psillos’ (1999) accounts. As is well known, ctr was initially formulated by Kripke ([1972] 1980) and Putnam (1975) for proper names and for some general terms whose corresponding references would be preserved

On Fictions, Theoretical Entities, and Ideal Objects

17

through chains of communication (each of these chains having at its origin the entity denoted by the term at issue). For most of the cases initially considered by ctr the reference would be fixed by ostension; that is, through the use of the expression in the presence of a particular which a proper name refers to or in the presence of an exemplar or a sample of the reference of a general term. This proposal is considered to be inadequate for (the vast majority of uses of) theoretical terms. But, as is well known, the initial contenders of ctr also considered the possibility that the fixation of reference can sometimes be established by means of a description (Putnam [1973] 1975, p. 200; Kripke [1972] 1980, p. 55 and ff., p. 96 n. 42). Moreover, if we accepted ctr as a theory of reference for theoretical terms, this would be the more natural way of thinking (Enç 1976; Nola 1980; Sankey 1994, p. 61). An advantage of ctr is arguably that, by means of ctr, we can account for the continuity (transtheoreticity) of reference. Authors of this approach believe that at least some uses of a term t in a theory that has proven to be ­incorrect (or is currently rejected) may refer to the same entity that is referred to by a homographic and homophonic term in a theory which is thought to be correct from our current perspective. Hence, for instance, the term ‘mass’ would have the same reference both in Relativistic and in Classical Mechanics (at least according to some uses of the term). The same would be the case of terms that are not used anymore, but which have been replaced in ­current ­theories by more appropriate terms. A well known case is the expression ‘dephlogisticated air’ as used by Priestley and other contenders of the Phlogiston Theory and the term ‘oxygen’ that was introduced by Lavoisier, the founder of modern chemistry. This is one of the case studies explored by Kitcher to illustrate his account (see Kitcher 1993, chap. 4).4 Kitcher contends that the reference of a theoretical type term is heterogeneous. According to him, different tokens of the same type term may have different references. There is a chain of events linking a token term with an introductory event for the reference of token term at issue. The introductory events for a given term determine the ‘modes of reference’ for that term and can be either of the kind in which the reference for a token term is fixed by means of a causal story or of the kind in which the reference is fixed by a description. The compendium of the modes of reference for a term is called the ‘reference potential’ of that term. Since the main problem with Kitcher’s account is similar to the problem with Psillos’, we will focus more on the latter in order to rule out this kind of hybrid versions of ctr (that in fact constitute the best versions of ctr available that are relevant 4 Kitcher has a series of papers in which he presents his account, advocating the idea of the heterogeneity of reference of type terms (Kitcher 1978; 1982; 1984; 1993).

18

de Donato Rodríguez and Falguera

to our problem) when applied to theoretical terms. But let us make a couple of remarks about Kitcher’s own proposal. From Kitcher’s analysis we cannot conclude that all references belonging to the reference potential of a type term are established from instances of ­entities or samples that are fixed by a series of causal relations, and we cannot then guarantee that these entities really exist. The only way of reference-fixing that may guarantee the connection with a spatiotemporally existing entity is the causal way of determination (by means of experimental manipulation). This may lead to the situation where a certain token term has a reference (by means of the causal story) while another token of the same type term lacks a reference. This is explicitly recognized by Kitcher and, in fact, describes the case of ‘dephlogisticated air’. The problem is that this account renders the texts of the Phlogistonians (in which this expression occurs) inconsistent. Phlogistonians not only thought that they were isolating a certain substance, but they also attributed to this substance certain properties that they considered to be essential to determine what kind of substance it was. And here we have the main problem, because Kitcher’s account does not take into account all the assumptions that are central to the theory at issue, i.e. it does not take into account all the ‘constitutive conditions’.5 Since this problem is explained more thoroughly in what follows (concerning Psillos’ account), let us continue the argument by presenting the other hybrid account. Psillos’ (1997; 1999) proposal, which is preceded by some similar a­ nalyses due to Enç (1976), Nola (1980) and Sankey (1994), accepts that a d­ escription providing constitutive conditions is playing a crucial role in fixing of the ­reference of a theoretical term. He also assumes that these conditions are theory-dependent. But he restricts such conditions to those properties that account for the causal role of the entity referred by the term at issue. He calls the set of these conditions the ‘core causal description’. More specifically, a term t refers to entity x if and only if x satisfies the core causal description associated with t. This also provides us with a criterion for co-referentiality, as a term t – belonging to a theory T1 – is co-referential with t′ – belonging to a different theory T2 – exactly when they have associated the same core causal description, that is, when their putative referents play the same causal role in relation to a given class of phenomena studied both by T1 and T2 (­Psillos 1999, p. 296). Psillos’s example of co-referentiality is the pair constituted by the expressions ‘luminiferous ether’ and ‘electromagnetic field’. Psillos points out that the core causal description does not need to make precise the specific 5 For further details see de Donato Rodríguez and Falguera (2014a), in which this argument is explained more carefully.

On Fictions, Theoretical Entities, and Ideal Objects

19

causal mechanism which is required to produce the relevant effects associated with the entity at issue, nor does it need include detailed conditions on the nature of this entity. The only properties that matter in order to have an appropriate core causal description for a given term are those which are ‘kind-­constitutive’ (to use Psillos’ term). Psillos says that the appeal to ‘kindconstitutive’ ­properties is necessary, because otherwise we would omit some property that is essential for fixing the reference. It is the task of the theory to specify which properties are kind-constitutive. The problem is that the properties that are essential for fixing the reference are not always primarily causal. Let us focus on this point more carefully. Psillos tries to argue that the expressions ‘luminiferous ether’ and ‘electromagnetic field’ are co-referential by pointing out that the luminiferous ether was postulated as a medium for the propagation of light and as a sort of repository for the kinetic and potential energy during the propagation of light, whereas the electromagnetic field was also postulated in association with these properties (though extended to electric and magnetic interactions). But this is not the point. What matters here for completely fixing the reference of ‘luminiferous ether’ is the property of ‘being a substance of a subtle kind that occupies those parts of the space which are apparently empty and, therefore, surrounds the material objects’. This condition is kind-constitutive and is not causal (at least not prima facie). The same goes for many properties (such as those provided by conservation laws) that are essential for fixing the reference of many central terms in physics such as mass and energy, although they are not strictly speaking causal. They can also specify certain ontological commitments (in the form of conditions about the nature of the entity at issue). In the case of the ether, the ontological property we have mentioned made ­possible to conceive the ether as a seat for electromagnetic fields, as having an absolute repose and thereby having a privileged status as inertial system. The conclusion seems to be that (i) we need some kind constitutive conditions, (ii) that these conditions must be provided by the relevant theory, and, most importantly, (iii) that these conditions have not be exclusively causal. This is of course not to say that causal descriptions do not play a role in fixing of the ­reference. In many cases, they will indeed play a role and will be counted among the constitutive conditions. Moreover, in some cases the core causal description will completely fix the reference of a theoretical term. But this will not generally be the case for every theoretical term.6

6 Ladyman (2002, p. 248 and ff.) sees another problem in Psillos’ account from the point of view of the realist.

20

de Donato Rodríguez and Falguera

2.2 A New Descriptivist Account Based on Zalta’s Abstract Object Theory Realists try to convince us that the referents of the theoretical terms in ­currently accepted theories (those which are considered to be approximately correct) are spatiotemporally existing objects. They also attempt to ensure continuity of reference as theories change (to a certain extent). If the argument of the last section is sound, realists lack of an appropriate way of fixing the referent of a theoretical term. The one proposed by the most credited versions of ctr is not acceptable. We may be luckier if we try a new approach, even if it is not a realist approach. The new option we are exploring has certain similarities with Kuhn’s account. As it is known, some of Kuhn’s last contributions deal with the semantics of theoretical terms and the issue of reference/meaning variance (incommensurability among rival theories). In the context of this discussion, Kuhn proposes that the referents of theoretical terms are kinds, not natural, but constituted by human action (Kuhn 1991; 1993, p. 315). ‘Constituted’ does not mean the same as ‘unreal’, for in some cases they may correspond to ways in which nature is really carved up. Though these kinds could also be understood in purely extensional terms (which is rather problematic), the best option is to understand Kuhn’s kinds as abstract objects. Following Bird’s (2000) interpretation, we may distinguish between the reference and the extension of a scientific term. The reference would be a constituted kind, i.e., an abstract object which is characterized by satisfying a certain set of (theoretical) conditions, while the extension is the set of instances to which the term is applied. The kind-constitutive conditions will depend on each theory which the term at issue is part of and will be all the conditions which, according to the relevant theory, are essential for fixing the referent. Contrary to the defenders of ctr, these conditions don’t need to be causal. Conservative and ontological principles are going to be typically involved in many cases. Some conditions may come even in the form of connecting principles which relate the term in question with other terms in the whole web of scientific theories (in this sense a certain level of holism should be accepted). For reasons that will be clear later, we will make use of Zalta’s Abstract Objects Theory (henceforth zaot) in order to formalize these ideas in an appropriate way.7 The main advantage of using this theory (apart from the formal precision and the possibility of deriving some interesting results) is that it provides us with a useful framework to deal with many problems related to the ontology of models, theoretical terms and scientific theories. Zalta (1983; 7 An extended version of the following considerations is to be found in de Donato Rodríguez & Falguera (2014b).

On Fictions, Theoretical Entities, and Ideal Objects

21

1988) applies his own theory not only to those entities that, like numbers or Fregean senses, are traditionally considered to be abstract, but also to fictional entities (fictional characters such as Sherlock Holmes, or mythological places such as the fountain of youth). In (1983), Zalta considers the possibility of extending his theory to scientific cases. What he considers explicitly is the concept of (absolute) simultaneity in classical mechanics. This is a fictional property (and, hence, on Zalta’s view, an abstract object), given that the simultaneity assumption is rejected by Einstein’s relativity theory. Given this interpretation, it makes sense to speak about the “non-existence” of simultaneity. According to Zalta, absolute simultaneity would simply be the abstract object that is referred to by the term ‘simultaneity’ in classical mechanics (see Zalta 1983, p. 146). Following a proposal made by Zalta himself,8 we can apply the same idea to other concepts that have been rejected or belong to theories that have been disconfirmed. This would be the case of the referents of terms like ‘ether’, ‘caloric’, ‘phlogiston’, or ‘Vulcan’. Our proposal is not only to apply zaot to terms that form part of rejected theories, but also to terms of theories that are currently accepted as approximately correct. That is, the idea is to apply zaot also to terms like ‘mass’, ‘energy’, ‘electron’, or ‘Higgs boson’. The immediate objection is that this account would automatically guarantee a referent for all theoretical terms (namely, an abstract object). Would this not trivialize scientific theories? The answer is: not necessarily. We could still account for the empirical character of scientific theories by resorting to connecting principles and empirical conditions that relate the abstract objects postulated by theories and certain observable and measurable features of the world. In the best cases and given certain conditions, we could say that to the abstract object there corresponds a concrete object (an individual, a property, a relation) which can be found in nature. Thus, the abstract objects would be the primary objects of science, whereas the concrete objects (spatiotemporally existing objects) that may (in the best cases) correspond to them are the secondary or mediate objects. Following Peirce’s semiotic ideas,9 we could distinguish between (i) the term or representamen (the linguistic expression), (ii) the content or meaning that it expresses (the ‘interpretant’), (iii) the referent or immediate object that the content determines and which the term refers to (that is, the abstract object), and (iv) the concrete (spatiotemporal) object which the abstract object may correspond to. 8 In one of his presentations at a colloquium held at the University of Santiago de Compostela in April 2012 (see: Zalta 2012). 9 See, in particular, Peirce (1933), p. 536 and Peirce (1958), pp. 314–315.

22

de Donato Rodríguez and Falguera

The immediate advantages of this account would be (i) that it renders c­ oherent and non-vacuous the texts of old (not completely accepted, superseded or simply rejected) theories, (ii) that – contrary to entity realism – it seems a more coherent position for those who are skeptic about scientific theories or who believe that theories are too highly idealized to be able to be true, and (iii) that, as we shall see, it constitutes a framework that allows us to find a ­suitable solution to many issues in the metaphysics of science. But let us begin the next section by presenting zaot. 3

The Nature of zaot and Its Application to Scientific Theories

As noted above, Edward N. Zalta has developed a powerful (axiomatic) theo­ry, called ‘Abstract Object Theory’, to give an account of entities that were traditionally considered as abstract (for instance, natural numbers), ­fictional entities (such as the references of expressions like ‘Sherlock Holmes’, or ‘the fountain of youth’) or even impossible entities (such as the reference of ‘the round square’).10 This theory is not restricted to particulars, but is also ­intended for abstract universals (as the reference of ‘unicorn’, which is a kind term). It is then a theory for objects that do not exist spatiotemporally. In fact, Zalta defines ‘abstract objects’, or A-objects, as those entities that cannot be spatiotemporal, thereby including not only those that are traditionally conceived as abstract, but also fictional and impossible entities. Let us now define some notions that are fundamental to zaot. That x is an A-object is expressed as ‘A!x’. Objects are divided in zaot into abstract objects and what is called ‘ordinary objects’. The class of ordinary objects, or O-objects, is disjoint with respect to the class of A-objects. That x is an O-object is expressed as ‘O!x’. O-objects are possibly spatiotemporal, or what amounts to the same, they are possibly concrete. In symbols: O!x = df [λy⬨E!y]x; A-objects cannot be concrete: A!x = df [λy¬⬨E!y]x (where ‘E!y’ is read as ‘y is concrete’). The existence of these two kinds of objects results in two ways of having properties. O-objects exemplify properties (or relations) and this is expressed in symbols by means of the usual predication: if a designates an O-object and F designates a property that a exemplifies, we write ‘Fa’ (analogously ‘Fna1…an’ is read as a1…an exemplify the n-ary relation ‘Fn’). An A-object is determined 10

We cannot reproduce a complete list of contributions. Zalta (1983) and Zalta (1988) are the most important and comprehensive works, but see also Zalta (1989; 2000; 2001; 2003 and 2006) for more recent developments with applications to different issues on philosophy of language and modality.

On Fictions, Theoretical Entities, and Ideal Objects

23

by a group of properties. These properties constitute the object in question. For instance, the object which is denoted by ‘Sherlock Holmes’ (the referent of this expression) is determined by the properties that are attributed to Sherlock Holmes in Conan Doyle’s novels and stories concerning this character (for instance, that he is a detective, or that he lives in London on Baker Street). A mathematical object is determined by a group of properties that are ­at­tributed to the object by a given mathematical theory. Zalta says that an A-­ object ‘encodes’ the properties that are attributed to that object. This is expressed by means of another kind of predication: if ‘a’ designates an A-object that encodes the property designated by ‘F ’, we write ‘aF ’. But, whereas Oobjects only exemplify properties, A-objects may encode as well as exemplify properties. For instance, Sherlock Holmes, the fictional character, exemplifies (but does not encode) the properties ‘being a fictional character’ or ‘being conceived by Conan Doyle’, etc. A mathematical object (such as number 2 or π) exemplifies (but does not encode) the properties of ‘being abstract’, ‘not having a mass’, ‘not being spatiotemporal’, etc. Let us make two final remarks on abstract objects. Firstly, A-objects may share the characteristics of being non-spatiotemporal entities and being the intentional object of some representation that occurs in a certain context (a ­fictional story or a given theory). A-objects may be individuals, properties or relations. In general, we have: xt is abstract⟨t⟩ [A!⟨t⟩ xt] = d f [λyt¬⬨E!⟨t⟩ y]xt (where ‘⟨t⟩’ is the appropriate type for properties that have objects of type t as instances).11 For example, the properties designated by expressions like ‘­unicorn’ or ‘hobbit’, given their fictional character, are A-objects. Analogously, scientific terms such as ‘phlogiston’ or ‘caloric’ would designate abstract objects (in this case, properties) as well. And secondly, A-objects are introduced in zaot by means of a comprehension axiom scheme that establishes that, for a given condition (or complex of conditions) φ defined for properties of type ⟨t⟩, there is an A-object that encodes exactly those properties. Symbolically, ∃xt(A!⟨t⟩x&∀F ⟨t⟩(xF≡φ)), where φ is any formula in which xt is not free. Furthermore, according to a uniqueness theorem established by zaot, for each condition (or complex of conditions) that A-object is unique.

11

Zalta (1983, p. 109 and ff.) introduces a set of types for categorizing the formal language: “i” is the type of individuals and “” is the type of relations having as arguments objects of type t1,…,tn. For example, “” is the type of properties of individuals, “” is the type of a two-place relation between individuals, “” is the type of properties of properties of individuals, and so on.

24

de Donato Rodríguez and Falguera

3.1 Scientific Theories as Stories As we have seen, with the help of zaot we can deal with the properties ­designated by terms like ‘phlogiston’, ‘caloric’, ‘simultaneity’, or even ‘mass’. In Section 3.2., we are going to see a method of characterizing the referents of these theoretical terms by means of zaot. But, before that, in this section we are going to apply zaot even to the concept of a scientific theory. In fact, this is a prerequisite for the further characterization of scientific terms as abstract objects, namely, to characterize scientific theories as abstract objects and, more particularly, as particular examples of ‘stories’ (which, in zaot, are regarded as abstract objects). Zalta’s starting point for defining the notion of a story is the concept of ‘situation’. A situation (the concept goes back to Barwise and Perry’s situation semantics) is in zaot simply defined as an abstract object that encodes only propositional properties. A property F is propositional if and only if for some proposition p, F is the property of being such that p. A proposition p is true in a situation s simply if and only if s encodes p, i.e., by encoding the propositional property “being such that p”. In symbols, s╞ p iff s[λyp]. We will use hereafter the expression “s╞ p”. A story is then defined as a particular case of situation, one in which the situation is identified by human intention (typically, a work of fiction, such as a novel). An example of story is Don Quixote (the novel, not the fictional character, which is an abstract object as well). Another one is Crime and punishment. Of course, we cannot confuse the novel type with a token of it (an exemplar of the book, which is a physical, spatiotemporal object). Only the type is an abstract object. So, we can say, that according to the novel, according to the story Don Quixote, Don Quixote thought the windmills were giants and fought against them. This is something which is true in the story. And the story is just an abstract object authored by a concrete object, a human, a Spanish writer of the 16th–17th centuries named Cervantes. Let us then pass to the definitions. A situation is thus an abstract object encoding only propositional properties (if a propositional property p is encoded by a situation, we shall say that p is true or holds in that situation): ​Situation​(x)​ ​=​ df​ A !x ∧ ∀F​(x F → Propositional ​(F)​)​ A story is a situation for which some concrete object (a human being or group of human beings) authors that situation: ​Story​(x)​ ​=​ df​  Situation​(x)​ ∧ ∃ y ​(E !y ∧ Ayx)​,​ where ‘A’ means ‘authors’. Now a scientific theory can be defined as a specific kind of story and, hence, of situation:

On Fictions, Theoretical Entities, and Ideal Objects

25

​STheory​(x)​ ​= ​ df​  Situation​(x)​ ∧ x contains projectable g­ eneralizations  ∧ ∃y ∃z (​ E!z ∧ SCz ∧ Azx ∧ y is a possible application for x ∧ z aims to apply x to y)​.​ What this definition says is that x is a situation authored by some concrete object – a scientist – belonging to a scientific community (this is what ‘sc’ means) and for which there is a possible application, a given set of phenomena, to which that scientist intends to apply the theory. So, this is an intentional, pragmatic concept of scientific theory (but in no case psychological, as a theory is an abstract object, not a mental entity). In terms of zaot, a s­ cientific theory becomes a particular kind of story that satisfies some further (empirical and pragmatic conditions) that distinguish theories from other (fictional) stories. We can also define: y​ is an intended application of x ​=​ df​  STheory​(x)​ ∧ y is a possible application of x ∧ ∃z ​(E !z ∧ SCz ∧ z aims to apply x to y)​.​ We conceive intended applications as concrete systems that are selected by scientists of a given community following pragmatic criteria. The idea of thinking that scientific theories are stories (or, better, particular kinds of stories and, thus, of situations) is not alien to Zalta’s work. Zalta (2007) contains an application of this idea to mathematics and he has extrapolated it to other fields (in more recent, still unpublished contributions). It is a new idea in neither the philosophy of science nor epistemology. For example, when van Fraassen characterizes scientific realism, a thesis he is trying to attack, he says “[s]cience aims to give us, in its theories, a literally true story of what the world is like” (van Fraassen 1980, 8). And, similarly, Peter Medawar argues that “[s]cientists are building explanatory structures, telling stories which are ­scrupulously tested to see if they are stories about real life” (Medawar 1996, 20). Or Jonathan H. Turner also considers that a “[t]heory is a ‘story’ about how and why events in the universe occur” (Turner 1991, i). And yet another interesting quote by Patricia Goodson, which perhaps points to a slightly different though related notion of theories as ‘stories’: When Edberg or I say that theories are stories, we are not saying that theories are the product of fantasizing about make-believe worlds. We mean this: Theories themselves are built following certain narratives structures, certain “story-building rules” and purposes. What these stories or theories “look like” depends on whether they are scientific theories, public policy theories, ethics theories, commonsense theories, etc. goodson 2010, p. 11

26

de Donato Rodríguez and Falguera

Let us now assume that a scientific theory is a kind of story in terms of zaot. Consider a scientific theory T. As a result of considering it as a story, we would obtain that the constituted kinds denoted by the T-terms will be determined by all the conditions that T attributes to them (just as the fictional characters are determined by those properties that, let us say, a novel attributes to them). Let now “ΣT” be an operator which is read “according to the story (theory) T…” (see: Zalta 1983, p. 91 and Zalta 1988, pp. 124–125), and by means of which we can characterize the entities (individuals, properties, relations) just as they are conceived internally to T. In that case, for a theory T, we will have that: if ​s​​ 1T​ ​ , …, ​s​ Tn​  denote the constituted kinds (i.e. abstract objects) that are referred to by the T-theoretical terms, then ​s​ Ti​ ​ may be identified as the abstract object that encodes exactly the properties attributed to si in theory T, i.e., ​​si ​ T ​ ​ = ιz (A !z ∧ ∀ G (zG ≡ ​ΣT ​ ​​ G​si ​ )​​ )​ where G ranges over properties and ΣTGsi is a formula asserting that the ­property or properties G are attributed to si in T. (We are making use here of the axiom of comprehension for A-objects and of the uniqueness theorem). But the above characterization requires us to take into account all the ­properties that are attributed to the entity in question by the theory and this may be too much (it is more than we need). In face of this, we have two options: either we assume all these properties are needed, and accept this consequence, or we restrict the axiom in order to take into account just those properties that are kind-constitutive (i.e., those required by the fundamental laws and main theoretical principles in which that entity is involved). Note that the fundamental conditions that are kind-constitutive would in general depend on each theory. This brings us to the next section. 3.2 Theoretical Terms as Abstract Objects As we have said, zaot can be applied to fictional characters, and Zalta himself considers applying his theory to cases like ‘phlogiston’ or ‘absolute simulta­ neity’. Our proposal is to extend zaot to any scientific term. One needs only to adopt a certain view in relation to scientific theories and scientific terms that in principle is neutral with respect to the scientific realism debate, a view that is non-realist though not necessarily antirealist. The view would hold that ­scientific terms always refer to A-objects and that scientific theories can in general (not only those theories that are rejected, but even those that are accepted) be seen as particular examples of stories. We have already said that the suggestion that scientific theories are stories is not new. Should one be

On Fictions, Theoretical Entities, and Ideal Objects

27

reluctant to accept that theoretical terms refer to abstract entities? We make a few remarks about this point. Every empirical theory is subject to be rejected and, if and when that occurs, the theory is no longer considered correct. For example, our best physical theories may be incorrect; electrons may in fact not exist (there may be no particle in nature with the properties physicists attribute to electrons). In that case, we would say that the term ‘electron’ refers to an A-object, but that it does not correspond to any O-object. We must also recognize that at best scientific theories are considered to be idealizations or approximations to the real world. Moreover, while there is no good reason to consider a theory completely correct, it can be considered as a useful fiction in its theoretical part. If so, we would be justified in regarding it as a fictional story and in applying the above treatment to denoting components of their theoretical terms. In these cases, the entities designated by theoretical terms are non-spatiotemporal properties or relations and, hence, they can be treated as A-objects in the framework of zaot. Of course, this does not mean that such A-objects may not correspond, given a certain relation of correspondence, to O-objects. For instance, physicists think that electrons (or Higgs bosons) do in fact exist and there are much empirical data and many experiments that lead physicists to think that such subatomic particles exist and that they have at least many, though perhaps not all, of the relevant properties (or conditions) that our best theory attributes to them. For example, let us assume that electrons do possess negative electric charge, but that when an electron is bound to the nucleus of an atom this is not by the attractive force governed by Coulomb’s law, but by a force which is governed by a more complicated law. In that case we will say – contrary to what is assumed – that there are not electrons as they are currently considered. We can say that electrons as they are currently considered are fictional entities and, because of this, they are abstract objects; and that the kind of electrons, so considered, is not a real natural kind. But it must be added that there may exist a kind of ordinary entities corresponding to the constituted kind of electrons. We would say that these O-objects, namely, the electrons that are found in nature (and which are recognized by experimental means, although they don’t have all the properties which are attributed to them) correspond to the A-objects that constitute the reference of ‘electron’. Let us now fix all these ideas in formal terms. Given a theory T, let FCT be a formula that expresses the fundamental conditions of T (it relates the theoretical entities with other entities that are not theoretical). The T-theoretical terms will then refer to certain abstract objects encoding just those properties that are attributed to them by T. These are ‘constituted kinds’ in a sense similar to Kuhn’s kinds. Now the result of replacing a T-theoretical term in

28

de Donato Rodríguez and Falguera

FCT by a variable yi, such that the variable is bound by a λ-operator applied to the open formula thus obtained, gives us the property (or complex of conditions) to be satisfied by each ​s​ ​ Ti​ ​ (​ 1≤i≤n)​​. Usually this property will be a complex of conditions. More specifically, for a theoretical term ​s​ ​ 1T​ ​​that property will be [λy1FC(r1,…,rm ,y1 ,s2 ,…,sn)], where r1 ,…,rm are the non-theoretical terms and s2 ,…,sn the rest of the theoretical terms different from s1. If we now represent the constitutive property for the term ​s​ Ti​ ​​by ‘Gi’, we have that Gi=[λyi FC(r1 ,…,rm ,…,yi ,…)] (1 ≤ i ≤ n) [where yi is replacing either s1 or s2 or…or sn , maintaining the rest of sj in the formula and being j≠i] and, hence, that ​s​ ​ Ti​ ​ = ιz ​(z ​Gi​ ​ & ​(​Σ​ T​ G​i ​ ​si​ )​ ​)​  ​(1 ≤ i ≤ n)​​. Of course, this is not a canonical way, according to zaot, of identifying an abstract object, as it does not result from the axiom of comprehension.12 (The canonical form is ​s​ Ti​ ​ = ιz(A !z ∧ ∀ G(zG ≡ ​ΣT ​ ​​ G​si ​ )​​ )​ as we saw at the end of Section 3.1.) The problem is that many of the properties attributed to si in the theory would be, let us say, accidental. They are not needed to fix the reference of si. Equally, the abstract object that encodes the constitutive property (Gi) may also encode other properties as well (properties that, for example, are also attributed to si in other theories). In consequence, we prefer to use only the (complex) property that is constitutive in order to determine the reference of a theoretical term. What property Gi then expresses is a complex of constitutive conditions that, according to T, serves to fix the ­reference of si. Each ​​s​ Ti​ (​ 1 ≤ i ≤ n)​​is a linguistic expression referring, by definition, to the abstract property (or relation) corresponding to the i-th T-theoretical term that is related to other A-objects denoted by the remaining terms (T-theoretical and T-non-theoretical) according to the fundamental conditions of T. The similarities between this approach, which uses zaot, and the well-known Ramsey-Lewis’ method for the definition of theoretical terms is clear (Lewis 1970). 3.3 Idealization, Ideal Objects and zaot We have seen that the idea of conceiving a rejected theory as a kind of fictional story can be extended to any scientific theory (or more exactly to the 12

As a consequence, someone could argue that we cannot guarantee the uniqueness of the abstract object denoted by siT. Zalta (1983, p. 35) can prove a result of uniqueness from his comprehension axiom scheme (Axiom 4), but as we are not making use of this axiom (in its canonical form) and we are not presenting a substitute of this axiom neither, we ­cannot prove an analogous uniqueness result. We are merely assuming that there is ­going to be just one abstract object satisfying the relevant constitutive conditions. Strictly speaking, we should take also certain empirical and pragmatic constraints into account, so that all these conditions together will be enough restrictive to ensure the uniqueness. In any case, the present formulation has been chosen for the sake of simplicity.

On Fictions, Theoretical Entities, and Ideal Objects

29

formulation of any scientific theory), no matter whether we are pondering a rejected theory or a theory accepted at a given particular time. This allows us to understand not only the referent of theoretical terms as abstract objects (in Zalta’s sense), but also to consider as abstract objects other entities that are characteristic constituents of empirical theories, like models, idealized applications or idealized entities of an application; for example, simple pendula, ideal gases, or harmonic oscillators. Furthermore, all these objects can be identified as a particular kind of abstract object that we understand as idealized (fictional) objects, including the referents of theoretical terms. This strategy also allows us to give a precise understanding of the idea that scientific models and theoretical entities can be seen, in virtue of being the product of multiple idealizations of different degrees, as fictions or as having a fictional nature (Duhem [1916] 1977; van Fraassen 1980; Cartwright 1983; Fine 1993; Barberousse and Ludwig 2009; Elgin 2009). According to this, the entities that are denoted by theoretical terms can be considered as useful fictions that eventually, in optimal cases, could correspond to spatiotemporally existing objects. Nonetheless, as we are going to see, there are substantial differences between the approach which we defend (based on the idea of abstract objects) and the fictionalist approaches (which mainly rely on the theory of make-believe). Let us now, through the following figure (Fig. 1.1), represent the whole class of abstract objects and some its subclasses: numbers, fictional characters, ­stories and theoretical entities. As a part of the class of the stories, there is abstract objects

stories scientific theories

numbers

fictional characters

theoretical entities ideal objects

Figure 1.1 The whole class of abstract objects and some its subclasses: numbers, fictional characters, stories and theoretical entities. As a part of the class of the stories, there is the class of scientific theories.

30

de Donato Rodríguez and Falguera

the class of scientific theories. And as a part of the class of theoretical entities, we have the class of ideal objects. The sizes of the circles are not indicative of the real size of the classes. In this way, we can then easily visualize some of the different kinds of abstract objects which can be studied by zaot: Hence, among theoretical terms there are typically some of them that are introduced by highly idealized assumptions so that their application ­conditions hold only in ideal situations that cannot be achieved in the real world (i.e. their application conditions are then counterfactual). This is the case with terms or expressions like ‘ideal gas’, ‘rigid body’, ‘perfect crystal’, ‘harmonic oscillator’, ‘simple pendulum’, ‘center of mass’, etc. As their application conditions are typically counterfactual, the laws in which these terms appear can only hold approximately. That their application conditions are counterfactual essentially means that they contravene some accepted truth. For example, the term ‘ideal gas’ applies only when the temperature remains constant, but the temperature does not usually remain constant either in nature or in the laboratory (there is always some fluctuation). So, we will define a proposition p as ideal when there is some proposition q accepted by the scientists in the community such that p contravenes q. This means, using modal language: ​Ideal​(p)​ ​=​ df​  ∃ q ∀ z​(SCz ∧ Accepts ​(z, q)​ ∧ q is empirical ∧ ¬⬨ ​(p ∧ q)​)​ At the same time, a world in which the conjunction between p and q holds should be seen as impossible. In zaot, a (possible) world is a particular kind of situation, one in which all and only the propositions that are true in that situation are propositions that are true (Zalta 1993; 1997). Let us quantify over worlds by using the variable w and let “╞ w” mean “to be true in world w”. Now in zaot, an impossible world is defined in terms of the notion of a ‘maximal situation’ (one situation s is maximal iff, for every proposition p, either s╞ p or s╞ ¬ p) and the notion of ‘actual’ (a situation s is actual iff every proposition p such that s╞ p is true). An impossible world would be a maximal situation which is not possibly actual (a possible situation is one which is possibly actual). We can use this definition perhaps without requiring the maximality condition (as we may want to allow for incomplete situations). In any case, if we define the predicate ‘ImpossibleWorld’ as in Zalta (1997), i.e. as a (­maximal) not possibly actual situation, we can then say that every world in which the conjunction of p and q holds, where p and q contradict each other, is an ­impossible world. Or, being p and q contradictory propositions: ​∀ w​(​⊨​ w​(p ∧ q)​ → ImpossibleWorld​(w)​)​

On Fictions, Theoretical Entities, and Ideal Objects

31

When this is the case, we will say that q is a defeater of p (a proposition such that contradicts p or, what amounts to the same, that makes from the world where both p and q hold an impossible world). Now, for each α, if q is a defeater of p and ╞α q, then ╞α¬ p. In particular, we can assert this for every possible world w: ​∀ w​(​⊨​ w​ q → ​⊨​ w​¬p)​ Moreover, if q is a defeater of p and q is a nomic fact, ◽N¬ p, i.e. according to the “nomological facts”, it is necessary that non-p. Later on we will say something more about defeaters. Now we can define, in terms of zaot, an idealized scientific theory in the following way: ​IdSTheoryof​(x,y)​ = df Situation​(x)​ ∧ ∃ z​(E !z ∧ Azx ∧ SCz ∧ ∃ y ​(y is an intended application of x ∧ z aims to apply x to y))​ ​ ∧ ∃ p​(x ⊨p ∧ Ideal​(p)​)​ Of course, virtually all scientific theories are going to be ideal, as all of them contain one or another ideal proposition. Idealization is something ubiquitous in science. As has been said, idealization is a common procedure in model construction. In constructing models, we usually assume conditions that in the best cases can only be approximated by real systems, as the ideal conditions in question are assumed by the ­scientists for convenience: in order to simplify the calculations, or because they want to test simple causal hypotheses in an initial stage of the research. As many authors have pointed out, the procedure of introducing idealizing conditions is common to the many faces of scientific activity (construction of data and theoretical models, formulation of laws, simulations, thought experiments, experimentation design,…) at theoretical as well as at practical level (see, for example: Jones and Cartwright 2005; Weisberg 2007; Barberousse and ­Ludwig 2009; de Donato Rodríguez 2011; de Donato Rodríguez and ArroyoSantos 2012). Though idealization can be considered at least from two points of view: (i) as a mental process or a certain procedure of reasoning (of counter­ factual nature), and (ii) as the product of that process (de Donato Rodríguez and Arroyo-Santos 2012, p. 12), we refer here more to the second point of view. So, in this sense, we say that models, laws, theories, or (kinds of) objects (such as the ideal pendulum) are idealized, or are idealizations (because they are the product of certain idealizing assumptions). Of scientific models it has been said that they are ‘works of fiction’. As Nancy Cartwright has put it:

32

de Donato Rodríguez and Falguera

A model is a work of fiction. Some properties ascribed to objects in the model will be genuine properties of the objects modelled, but others will be merely properties of convenience. […] Some of the properties and relations in a model will be real properties, in the sense that other objects in other situations might genuinely have them. But they are introduced into this model as a convenience, to bring the objects modelled into the range of the mathematical theory. Not all properties of convenience will be real ones. There are the obvious idealizations […] some properties are not even approached in reality. They are pure fictions. cartwright 1983, p. 153

This has given raise to different proposals maintaining that models and ideal objects are fictions or fictional systems/objects (see: Godfrey-Smith 2009; Súarez 2009; Barberousse & Ludwig 2009; Frigg 2010a; 2010b; Contessa 2010; Leng 2010 and Toon 2010). The nature of fiction is, by itself, a philosophical problem which deserves and has deserved much attention (see, for example: Currie 1990; Walton 1990; Thomasson 1999; Kalderon ed. 2005, and Sainsbury 2010). Among fictionalists, there are different approaches as well. A brief look into the book edited by Suárez (2009) may serve for giving us an idea of this. Whereas Morrison (2009), Teller (2009) or Giere (2009) view fictions in scientific modeling as false assumptions about the systems scientist want to investigate, authors like Rouse (2009), Elgin (2009) or Winsberg (2009) focus more on the functions that fictions play both in scientific modeling and practice. Barberousse and Ludwig (2009), Suárez (2009) or Knuttila (2009) maintain that fictive or idealizing assumptions and representations are fictions by themselves. Many authors have followed Cartwright in arguing that models are fictions and have resorted to Walton’s pretense or make-believe theory to account for them (see, for example: Toon 2010). According to Walton’s theory adapted to scientific contexts, scientists do not exactly assert what they say when they are idealizing, they merely pretend to do it. In this sense, the objects they are allegedly referring to are not real objects, but fictional. They are, like the fictional characters of a novel, non-existing objects. Even if we agree with Cartwright in saying that models and ideal objects like simple pendula or rigid bodies are works of fiction, this does not mean per se that the products of our fictions (of our idealizing procedures) are fictions themselves. A novel is the result of a process of fictive imagining, but it is not a fiction in and of itself. Like novels, theories and their components are abstract objects, that is, intensional entities (neither physical nor mental) which are the intentional referent of the imagination of scientist and which, in the best cases, may even (approximately) correspond to concrete objects and, hence, represent them in a(n) (approximately) correct way. If we follow Thomasson (1999),

On Fictions, Theoretical Entities, and Ideal Objects

33

we could say that they are ‘abstract artifacts’ in the sense of being the result of human intention, having thereby a representational character. This way of putting the things could explain the role models and ideal objects play in ­scientific explanation and could also account for usual linguistic uses such as negative existential statements of the kind “According to Einstein’s Relativity Theory, (absolute) simultaneity does not really exist” or “Lavoisier discovered that phlogiston does not exist”. Consequently, given the approach we are defending in this paper, we contend that scientific models, at least according to a certain meaning of ‘model’, can be understood as abstract objects. In this sense, we have: • models in the sense of mathematical structures (they would be abstract objects a fortiori, because of their mathematical nature) (see: Suppes 1960; Przełęcki 1969; Sneed 1971; van Fraassen 1980; and Balzer, Moulines and Sneed 1987); • models in the sense of fictions or mathematical abstractions (so highly ­idealized constructions that it becomes impossible for them to be even ­spatiotemporally realizable) (see: Vaihinger 1911; Fine 1993; Cartwright 1983). But ‘model’ has also other meanings: • models in the sense of material models (Drosophila melanogaster). • scale models of real things (a scale model of a given crystal). These would not be examples of abstract objects if we think in the concrete exemplars that are used by scientists. Of course, we can distinguish the material support of the model (the particular instance we use to modelize something) from the intensional structured entity expressed by that material support which constitutes the model per se (and which is an abstract entity). Let us now focus on ideal objects such as ideal pendula, rigid bodies or harmonic oscillators. Following zaot, we can define an ideal object as a particular case of abstract object that encodes every property that, according to some scientific theory, exemplifies and such that al least some of the propositional properties they encode is ideal. In symbols: ​IdealObject​(x)​ ​=​ df​  A !x ∧ ∃ s​(STheory​(s)​ ∧ ∀ F(​ xF ↔ s ⊨ Fx)​)​  ∧ ∃ p (​ x ⊨ p ∧ Ideal​(p))​ ​ Intuitively, this means that an ideal object is that abstract object, introduced by a given scientific theory, such that is characterized in terms of some ideal conditions.

34

de Donato Rodríguez and Falguera

For example, the expression ‘simple pendulum’ would denote an abstract (ideal) object that, according to classical mechanics (cm), satisfies the ­so-called law of the simple pendulum holding some ideal conditions, for example, that the action of the friction force is negligible or that x is a mass point suspended from a rod of negligible mass. More specifically, we could write something like: __ ​SimplePendulum​(x)​ = IdealObject​(x)​ ∧ CM ⊨​[λy T = 2π ​√ __​Lg ​ ]​x  ∧ x has no friction ∧ … ​In fact, we can write: ​SimplePendulum​(x)​ = ιx​(A !x ∧ ∀ F​(xF ↔ CM ⊨ Fx)​)​,​ where F stands for the propositional property that summarizes all the defining conditions of the simple pendulum. An analogous definition could be given for other examples of ideal objects (rigid body, ideal gas, mass-point, etc.). It does not follow from the definition of Ideal, but in the case of natural sciences it seems reasonable to require that: ​Ideal​(p)​ → ⬨ p, i.e. in principle there may be a world in which p is true ​ (in the sense that it is logically and metaphysically possible, though not physically possible)​.​ Now, it is not possible that the ideal pendulum exists (in the sense of E!), as it is an abstract object and, hence, a not possibly concrete object. Nevertheless, from the previous considerations, we can make sense of the intuitive modal truth that there might have existed the simple pendulum. This would mean that there is at least a possible world in which a correlate of the simple pendulum exists. If a denotes the simple pendulum: ​⬨ ∃ x​(E !x ∧ ∀ F ​(aF → Fx)​)​​ Another application of our approach is the following. Since defeaters can be of different modal strengths (ranging from contingently true defeaters to fundamental laws), we can make sense of the intuition that ideal propositions can exhibit different degrees of idealization. That is, we can meaningfully say that a certain proposition p is more idealized (in a given context) than another proposition q. There seems to be two different criteria to capture this intuition:

On Fictions, Theoretical Entities, and Ideal Objects

35

(i) the number of defeaters: if |Dp| ≤ |Dq|, then q is more idealized than p; (ii) the quality of the defeaters: if Dq but not Dp contains some crucial theoretical law, then we also say that q is more idealized than p. Between the ideal propositions that are encoded by a certain ideal object, there usually are nomic counterfactual relationships of the following kind: some encoded property p may be such that if p were the case, it would be the case that r, where typically p is an idealized condition for r (this would be represented by an idealized law). In the case of the simple pendulum, we could say that, according to classical mechanics (mc), if a pendulum x has no friction (and satisfies other equally idealized conditions), x satisfies the law of the simple pendulum: __ ​MC ⊨ ∀ x (x has no friction ∧ … ◽→ ​[λy T = 2π ​√ __​Lg ​ ]​x )​ This counterfactual character of our objects is also captured by Suppe’s notion of ‘replication’ of a phenomenal system: Scientific theories do not describe the behavior of phenomenal systems in all their complexity, but rather attempt to characterize the p ­ henomena in terms of a few selected parameters. For example, in classical particle mechanics, one treats mechanical phenomena involving interacting ­bodies as if the phenomena only involved certain specified bodies and only involved the positions and momenta of these bodies […] Under certain circumstances this fiction can be approximately realized […] In most circumstances, however, the fiction is not realized […] In such cases, the values of the parameters of the characteristic of the state s that the physical system S is in at t are not the actual parameter values characteristic of the phenomenal system P at t. Rather, they stand in the following ­replicating relation to the values in P: If P were an isolated phenomenal system in which all other parameters exerted a negligible influence, then the physical quantities characteristic of those parameters abstracted from P would be identical with those values characteristic of the state at t of the physical system S corresponding to P, [in which case we can say that S is a “replica” of P]. suppe 1989, p. 94–95

In the same passage, Suppe (1989) also provides us with an example of what we meant above by saying that p, our ideal assumption, could be nomologically

36

de Donato Rodríguez and Falguera

impossible (i.e., not compossible with some q that is considered to be a law or a scientific fact that was previously established, and therefore constituting part of the background knowledge of the scientists): sometimes it is impossible for P to realize the conditions such that P’s behavior would be as S indicates. For instance, in classical particle mechanics, position is a parameter and the possible values for this parameter are the coordinates of points in space. This imposes the requirement that, at each time, a body must have an idealized unique point in space as its spatial location, which in turn and strictly speaking requires that the body be extensionless or dimensionless, something that is causally impossible, since it would require that bodies have an infinite gravitational potential, which is simply impossible if classical particle mechanics is true. According to a fictionalist like Elgin, an ideal gas representation is a fiction characterizing a putative gas that would behave exactly as the ideal gas law says. The scientist represents the putative gas as if it were composed of perfectly spherical particles of negligible volume and exhibiting no intermolecular forces (Elgin 2009, 84). From our perspective, the ideal gas is just an abstract object exemplifying certain properties attributed to it in the kinetic theory of gases. Some of these properties, as we already saw in Cartwright’s text, are so highly idealized (such as being a perfectly spherical particle) that they cannot be even approximated by real gases. Other properties can indeed be approximated by real gases (for example, their particles are really very small, too small to have to take into account their volumes, though this does not mean that they do not have any extension at all). If some properties can only be attributed to “objects in the model” (to use well-known Cartwright’s expression: Cartwright 1983, p. 17) while others are only true of real objects or systems the objects in the model are alleged to represent, then if we want to take scientific theories to be genuinely referential, it seems we have (i) to posit the existence of abstract entities for theories to be about, and (ii) to ­establish some kind of suitable relation of correspondence between the abstract entities and the real entities they represent. As Giere (2004, p. 745) has put it, “[i]f we insist on regarding principles as genuine statements, we have to find something that they describe”. And what they firstly describe are Cartwright’s “objects in the model”, i.e., abstract entities. Leng (2010, p. 135–137) argues against this move and assumes a ­fictionalist position. Though she devotes her book mainly to mathematics, she also considers scientific models and the indispensability of idealizations in empirical sciences. She seems to invoke Quine’s general skepticism about abstract entities to maintain that the use of idealization in empirical science does not

On Fictions, Theoretical Entities, and Ideal Objects

37

commit us to the existence of abstract objects.13 As it is well known, Quine ([1953] 19612, p. 44) said that physical objects (in the sense of objects posited by physical theories) could be compared to Homer’s gods, noting that their ­difference was a matter of degree. According to our approach, there is not only a sharp difference between concrete and abstract objects, but also a difference between abstract objects posited by scientific theories and Homer’s gods. We totally agree with Giere (2009) when he points out that it would be a dangerous move to put fictions and the entities posited by science on the very same level. But it is precisely here where abstract objects theory could help. The utility of abstract objects in order to make predictions and provide explanations and the correspondence that could be established between abstract and concrete objects prevent us from an undesirable identification between abstract objects and mere fictions.14 Leng (2010, p. 112) argues that, when highly idealized assumptions are involved, the truth of scientific theories is not confirmed. We agree if by ‘truth’ we mean exact truth or truth in the sense a traditional (ontological) scientific realist could admit. But confirmation of theories need not necessarily have this aim (i.e. to confirm the exact truth of theories in the sense a traditional scientific realist could admit). It has been argued that idealizations cannot be tested, as their antecedents are never realizable (at least not in this world). But, as Laymon (1982 and 1985) has seen, idealizations are justified by showing how they lead to more accurate (approximately correct) theories.15 If we admit a certain notion of approximation, we could resort to a more reliable notion of confirmation by which we can ­establish the approximate confirmation of our theories. In this sense, our positing theoretical entities is not a mere make-believing game dangerously

13

14

15

Leng accepts Quine’s naturalism, but does not accept that naturalism requires us to be r­ealists about mathematical entities nor abstract ideal objects (such as mass points or frictionless planes). In general, “[t]rusting science to tell us what there is should not require us to believe in all of the objects posited by our successful scientific theories” (Leng 2010, pp. 125–126). So, she emphasizes the limits of the indispensability argument as an argument to believe in the objects posited by our best scientific theories. This goes together with her rejecting of Quine’s confirmational holism. To turn back to van Fraassen, a nice argument showing that his constructive empiricism is committed to abstract objects of some sort (at least the existence of models) has been provided by Rosen (1994). This is also basically the sense of the idealization-concretization process in Nowak’s ­approach and the Poznań School. See Nowak (1980 and 1992) and Nowak and Nowakowa (2000).

38

de Donato Rodríguez and Falguera

similar to a literary fictionalizing. But we will focus more on the advantages of appealing to abstract objects in another paper.16 4 Conclusion Among the advantages of taking the referents of theoretical terms to be abstract objects, we can count that (i) the proposal avoids the problems of the realist accounts that we have seen exemplified through two ctr-inspired approaches; (ii) it is a proposal that allows for a theory of reference which is adequate for the supposition that scientific theories are works of fiction, though without identifying them just with fictions; (iii) it gives sense to the idea that, through the development of a theory, its theoretical terms preserve their reference, because of the fact that the individuating conditions for theoretical entities (and of ideal objects) as abstract objects are the fundamental, kind-constitutive conditions associated to them by each relevant theory; (iv) it is a non-realist interpretation, but need not necessarily count as antirealist (at least it is not an ontological antirealist approach), as it does not exclude that the referent of a theoretical term, which is an abstract object, may correspond to a kind in the real world, a natural kind; and (v) it provides us with a theoretical framework in which it seems natural to treat as abstract objects other components of scientific products, such as idealized applications, the ­idealized entities of an application, scientific models, the laws of a given theory and the theories themselves. References Balzer, W., C.U. Moulines and J.D. Sneed (1987). An Architectonic for Science. The Structuralist Program. Dordrecht: Reidel. Bandyopadhyay, P. (1997). On an Inconsistency in Constructive Empiricism. Philosophy of Science, 64(3), 511–514. Barberousse, A. and P. Ludwig (2009). Models as Fictions. In: M. Suárez (ed.). Fictions in Science, pp. 56–73. London: Routledge. Bird, A. (2000). Thomas Kuhn. Chesham: Acumen Pub. Ltd.

16

For a comparison between zaot and the pretense theory see: Zalta (2000). See also Zalta (2003).

On Fictions, Theoretical Entities, and Ideal Objects

39

Carnap, R. ([1958] 1975). Observational Language and Theoretical Language. In: J. ­ Hintikka (ed.) Rudolf Carnap, Logical Empirist. Materials and Perspectives, pp. 75–85. Dordrecht: Reidel. Cartwright, N. (1983). How the Laws of Physics Lie. Oxford: Clarendon Press. Chakravartty, A. (1998). Semirealism. Studies in History and Philosophy of Science 29, 391–408. Chakravartty, A. (2007). A Metaphysics for Scientific Realism: Knowing the ­Unobservable. Cambridge: Cambridge University Press. Chakravartty, A. (2014). Scientific Realism. In: E.N. Zalta (ed.). Stanford Encyclopedia of Philosophy (2014 edition), [URL = http://plato.stanford.edu/entries/scientificrealism/ May 15, 2014]. Contessa, G. (2010). Scientific Models and Fictional Objects. Synthese 172 (2), 215–229. Currie, G. (1990). The Nature of Fiction. Cambridge: Cambridge University Press. De Donato Rodríguez, X. (2011). Idealization within a Structuralist Perspective. Metatheoria 1(2) 65–90. De Donato Rodríguez, X. and A. Arroyo-Santos (2012). The Structure of Idealization in Biological Theories: The Case of the Wright-Fisher Model. Journal for General Philosophy of Science 43, 11–27. De Donato Rodríguez, X. and J.L. Falguera (2014a). Why the Causal Theory of Reference Cannot be Applied to Theoretical Terms. Manuscript. De Donato Rodríguez, X. and J.L. Falguera (2014b). Theoretical Terms, Ideal Objects and Zalta’s Abstract Object Theory. Manuscript. Duhem, P. ([1916] 1977). The Aim and Structure of Physical Theory. New York: Atheneum. Elgin, C. (2009). Exemplification, Idealization, and Scientific Understanding. In: M. Suárez (ed.). Fictions in Science, pp. 77–90. London: Routledge. Enç, B. (1976). Reference of Theoretical Terms. Noûs, 10, 261–282. Fine, A. (1993). Fictionalism. Midwest Studies in Philosophy, 18, 1–18. Frigg, R. (2010a). Models and Fiction. Synthese 172, 251–268. Frigg, R. (2010b). Fiction in Science. In: J. Woods (ed.). Fictions and Models: New Essays, pp. 247–287. Munich: Philosophia Verlag. Giere, R. (2004). How the Models Are Used to Represent Reality. Philosophy of Science 71, 742–752. Giere, R. (2009). Why Scientific Models Should Not Be Regarded as Works of Fiction. In: M. Suárez (ed.). Fictions in Science, pp. 248–258. London: Routledge. Godfrey-Smith, P. (2009). Models and Fictions in Science. Philosophical Studies, 143, 101–116. Goodson, P. (2010). Theory in Health Promotion Research and Practice. London: Jones and Bartlett Publ. Hacking, I. (1983). Representing and Intervening, Cambridge: Cambridge University Press.

40

de Donato Rodríguez and Falguera

Jones, M. and N. Cartwright (eds.) (2005). Idealization XII: Correcting the Model. Poznań Studies in the Philosophy of Science and the Humanities, vol. 86. Amsterdam: Rodopi. Kalderon, M.E. (ed.) (2005). Fictionalism in Metaphysics. Oxford: Clarendon. Kitcher, P. (1978). Theories, Theorist and Theoretical Change. The Philosophical Review, 87, 519–547. Kitcher, P. (1982). Genes. British Journal for the Philosophy of Science, 33, 337–359. Kitcher, P. (1984). Species. Philosophy of Science 51, 308–333. Kitcher, P. (1993). The Advancement of Science. Oxford: Oxford University Press. Knuuttila, T. (2009). Representation, Idealization, and Fiction in Economics: From the Assumptions Issue to the Epistemology of Modeling. In: M. Suárez (ed.). Fictions in Science, pp. 205–231. London: Routledge. Kripke, S. ([1972] 1980). Naming and Necessity. Oxford: Basil Blackwell. Kuhn, T.S. (1991). The Road Since Structure. In: A. Fine, A. Forbes and L. Wessels (eds.) PSA 1990, vol. 2, pp. 2–13. Michigan: East Leasing. Kuhn, T.S. (1993). Afterwords. In: P. Horwich (ed.), World Changes. Thomas Kuhn and the Nature of Science, pp. 311–341. Cambridge, Mass.: The MIT Press. Ladyman, J. (2000). What’s Really Wrong With Constructive Empiricism? Van F­ raassen and the Metaphysics of Modality. British Journal for the Philosophy of Science, 51, 837–856. Ladyman, J. (2002). Understanding Philosophy of Science. London: Routledge. Laymon, R. (1982). Scientific Realism and the Hierarchical Counterfactual Path from Data to Theory. In: P. Asquith, and T. Nickles (eds.), Proceedings of the 1982 Biennal Meeting of the Philosophy of Science Association, vol. 1, pp. 107–121. Michigan: East Lansing. Laymon, R. (1985). Idealizations and the Testing of Theories by Experimentation. In: P. Achinstein, and O. Hannaway (eds.). Experiment and Observation in Modern S­ cience, pp. 147–173. Boston: MIT Press and Bradford Books. Leng, M. (2010). Mathematics and Reality. Oxford: Oxford University Press. Lewis, D. (1970). How to Define Theoretical Terms. Journal of Philosophy, 67, 427–446. Medawar, P. (1996). The Strange Case of the Spotted Mice and other Classic Essays on Science. Oxford: Oxford University Press. Morrison, M. (1990). Theory, Intervention and Realism. Synthese, 82, 1–22. Morrison, M. (2009). Fictions, Representations, and Reality. In: M. Suárez (ed.): Fictions in Science, pp. 110–135. London: Routledge. Nola, R. (1980). Fixing the Reference of Theoretical Terms. Philosophy of Science 47, 505–531. Nowak, L. (1980). The Structure of Idealization: Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht: Reidel. Nowak, L. (1992). The Idealizational Approach to Science: A Survey. In: J. Brzeziński and L. Nowak (eds.), Idealization III: Approximation and Truth. Poznań Studies in the

On Fictions, Theoretical Entities, and Ideal Objects

41

Philosophy of the Sciences and the Humanities, vol. 25, pp. 9–63. Amsterdam-Atlanta: Rodopi. Nowak, L. and I. Nowakowa (2000): Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 69. Amsterdam/ Atlanta, GA: Rodopi. Peirce, Ch.S. (1933). Collected Papers, vol. IV, ed. by Ch. Hartshorne and P. Weiss, ­Cambridge, MA: Harvard University Press. Peirce, Ch.S. (1958). Collected Papers, vol. VIII, ed. by A.W. Burks., Cambridge, MA: ­Harvard University Press. Przełęcki, M. (1969). The Logic of Empirical Theories. London/New York: Routledge & Kegan Paul. Psillos, S. (1997). Kitcher on Reference. International Studies in the Philosophy of Science 11 (3), 259–272. Psillos, S. (1999): Scientific Realism. How Science Tracks Truth. London: Routledge. Putnam, H. ([1973] 1975). Explanation and Reference. In: idem, Mind, Language and Reality. Philosophical Papers, vol. II, pp. 196–214. Cambridge: Cambridge University Press. Putnam, H. (1975). The Meaning of ‘Meaning’. In: idem, Mind, Language and Reality. Philosophical Papers, vol. II, pp. 215–271. Cambridge: Cambridge University Press. Quine, W.V.O. ([1953] 1961). From a Logical Point of View. New York: Harper & Row. Resnik, D.B. (1998). Hacking’s Experimental Realism. In: M. Curd and J.A. Cover (eds.). Philosophy of Science: The Central Issues, pp. 1169–1185. New York/London: W.W. ­Norton and Co. Rosen, G. (1994). What is Constructive Empiricism? Philosophical Studies 74(2), 143–178. Rouse, J. (2009). Laboratory Fictions. In M. Suárez (ed.). Fictions in Science, pp. 37–55. London: Routledge. Sainsbury, R.M. (2010). Fiction and Fictionalism. London: Routledge. Sankey, H. (1994). The Incommensurability Thesis. Avebury: Aldershot. Sneed, J.D. (1971). The Logical Structure of Mathematical Physics. Dordrecht: Reidel. Suárez, M. (2009). Scientific Fictions as Rules of Inference. In: M. Suárez (ed.): Fictions in Science, pp. 158–178. London: Routledge. Suárez, M. (ed.) (2009). Fictions in Science. London: Routledge. Suppe, F. (1989). The Semantic Conception of Theories and Scientific Realism., Urbana & Chicago: University of Illinois Press. Suppes, P. (1960). A Comparison of the Meaning and Uses of Models in Mathematics and the Empirical Sciences. Synthese, 12, 287–301. Teller, P. (2009). Fictions, Fictionalization, and Truth in Science. In: M. Suárez (ed.): Fictions in Science, pp. 235–247. London: Routledge. Thomasson, A.L. (1999). Fiction and Metaphysics. Cambridge: Cambridge University Press.

42

de Donato Rodríguez and Falguera

Toon, A. (2010). The Ontology of Theoretical Modelling: Models as Make-Believe. ­Synthese 172, 301–315. Turner, J.H. (1991). The Structure of Sociological Theory, Wadsworth, Belmont, CA. Vaihinger, P. (1911). Die Philosophie des Als Ob. Berlin: Verlag von Reuther & Reichard. Van Fraassen, B.C. (1980). The Scientific Image. Oxford: Clarendon Press. Walton, K.L. (1990). Mimesis as Make-Believe. Cambridge: Harvard University Press. Weisberg, M. (2007). Three Kinds of Idealization. Journal of Philosophy 104 (12), 639–659. Winsberg, E. (2009). A Function for Fictions: Expanding the Scope of Science. In: M. Suárez (ed.) Fictions in Science, pp. 179–189. London: Routledge. Zalta, E.N. (1983). Abstract Objects: An Introduction to Axiomatic Metaphysics. ­Dordrecht: Reidel. Zalta, E.N. (1988). Intensional Logic and the Metaphysics of Intentionality. Cambridge, MA and London: MIT Press. Zalta, E.N. (1989). Singular Propositions, Abstract Constituents, and Propositional ­Attitudes. In: J. Almog, J. Perry and H. Wettstein (eds). Themes from Kaplan, pp. 455– 78. Oxford: Oxford University Press. Zalta, E.N. (1993). Twenty-Five Basic Theorems in Situation and World Theory. Journal of Philosophical Logic, 22, 385–428. Zalta, E.N. (1997). A Classically-Based Theory of Impossible Worlds. Notre Dame J­ ournal of Formal Logic, 38 (4), 640–660. Zalta, E.N. (2000). The Road Between Pretense Theory and Abstract Object Theory. In: A. Everett and T. Hofweber (eds.). Empty Names, Fiction, and the Puzzles of Non-Existence, pp. 117–147. Stanford: CSLI Publications. Zalta, E.N. (2001). Fregean Senses, Modes of Presentation, and Concepts. Philosophical Perspectives (Noûs Supplement) 15, 335–359. Zalta, E.N. (2003). Referring to Fictional Characters. Dialectica, 57(2), 243–254. Zalta, E.N. (2006). Essence and Modality. Mind 115 (459), 659–693. Zalta, E.N. (2007). Reflections on Mathematics. In: V.F. Hendricks and H. Leitgeb (eds.), Philosophy of Mathematics: 5 Questions, pp. 313–328. New York, London: Automatic Press/VIP. Zalta, E.N. (2012). A Solution to the Problem of Abstract Objects That Even a Naturalist Could Love. Talk delivered at the University of Santiago de Compostela in April 2012.

chapter 2

The Inherent Type of Scientific Law, the Idealized Types of Scientific Law1 Igor Hanzel Abstract The paper provides a methodological explication of the structure of three types of scientific law: the pure idealized type, the inherent type and the inherent idealized type. It provides also a reconstruction of the explanation procedures based on them. Utilizing an analysis of both Newton’s Principia and of the first five chapters in the first volume of Marx’s Capital it shows that while the former utilizes the first type of idealized law, the latter involves both the inherent type and the inherent idealized type.

1 Introduction The aim of this paper is threefold. It aims, first, at providing a methodological explication of the structure of two types of scientific laws involving idealization (idealized types of scientific law, for short) as well as of the explanation procedures based on them. Second, its aim is historical in the sense that it will relate this explication to Marx’s Capital as well as to Newton’s Principia. And, third, its aim is reconstructive since it explicates the structure of these theories which can be viewed as paradigmatic examples of similar but at the same time in certain respects different approaches to issues like theory construction, formulation of scientific laws and scientific explanation. I start with an analysis of Leszek Nowak’s reconstruction of the structure of the idealized laws and explanation based on then by means of the method of gradual concretization and show that, contrary to his efforts, they correspond neither to the structure of the law of value as stated by Marx in the beginning of the first volume of Capital, nor to second law of motion of classical mechanics in the context of Newton’s Principia. Then, I show that in order to grasp methodologically these laws one has to provide a differentiated typology of scientific laws, namely, the so-called pure idealized type of scientific law, the inherent type of scientific law, and the inherent idealized type of scientific 1 This paper was written with the support of the vega research grant, grant number 1/0221/14.

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_004

44

Hanzel

law. Next, based on my reconstruction of the structure of these types of laws I develop the methods of explanation based on them. Finally, based on my typology of scientific laws, I suggest a typology of quantification based them. 2

The Standard Account

Leszek Nowak proposed in his (1971; 1972; 1974; 1980) a highly innovative reconstruction of the structure of idealized laws involving k idealizations, which symbolically can be expressed as follows: ​​L​​ ​(k)​​ :  ∀ x​{U​(x)​ & ​p1–k ​  ​(x)​ = 0 → ​E​ ​(k)​​(x)​ = ​fk​ ​[H​(x)​]​}​​

(1)

Here “U” denotes the universe of discourse over which the individual variable x ranges; “&” stands for the sentential connective “and”; “px = 0” denotes a propositional function expressing an idealizational assumption; “0” denotes the minimum value of the magnitude denoted by “p”; “p1–k x = 0” denotes the conjunction of k idealizational assumptions (idealizations, for short) which should hold for x; “E” stands for a factor under examination, the so-called ­phenomenal form; “E(k)” stands for the phenomenal form in the kth idealization, that is, when the 1st, …, kth idealizations hold; “H” denotes the principal factor, or ground of the phenomenal forms and the whole expression on the right side of the symbol “→” expresses a functional dependence of the phenomenal form on its ground.2 Based on the structure (1), Nowak constructs a model of scientific explanation which views scientific explanation as involving a gradual removal of the involved idealizations which in turn requires the introduction of corrections in the understanding of the functional dependence of the phenomenal form on the respective factors. Symbolically this can be expressed for the case when one abolishes the kth idealization, that is, assuming the validity of “pk ≠ 0,” as follows: ​L​ (​ k−1)​​ :  ∀ x​{U​(x)​ & ​p​ 1−​(k−1)(​​ x)​ = 0 & ​pk​  (​ x)​ ≠ 0 → ​E​ (​ k−1)(​​ x)​ = ​fk−1 ​  ​[H​(x)​, ​pk​  (​ x)​]​}​​

(2)

Here “p1–(k–1) = 0” expresses that the 1st through (k–1)th idealizations still hold while the “pk” introduced on the right side should express the impact of the parameter it denotes on the phenomenal form under investigation. 2 For the sake of simplicity I presuppose only the existence of one principal factor; L. Nowak presupposes several of them.

The Inherent Type of Scientific Law

45

L. Nowak claimed that the structure of (1) can be viewed as a methodological key for the reconstruction of the second law of motion in classical mechanics as well as of the law of value as given in the beginning of the first volume of Capital. The former is expressed by him as follows (Nowak 1974, p. 29, p. 46): ​​L​​ ​(2)​​ :  ∀ x​[O​(x)​ & D​(x)​ = 0 & ​Fe ​ ​​​(x)​ = 0 → ​Fr ​ ​( ​ 2)​​(x)​ = m​(x)​a​(x)​]​​

(3)

“Ox” states that x is a physical object located in a physical system; the first idealization that all spatial dimensions of x are equal to zero; the second idealization that the system in which x is placed is not subjected to an external forces Fe; “Fr x” states that x is subjected to the action of the force Fr; “mx” and “ax” stated that m and a are x’s mass and acceleration, respectively. For the law of value as given in the beginning of the first volume of Capital should hold (Nowak 1980, pp. 8–10): ​​L​v ​( ​ 8)​​ :  ∀ x​{C​(x)​ & ​p1–8 ​  ​(x)​ = 0 → ​P​ ​(8)​​x = l​[V​(x)​]​}​​

(4)

Here “C(x)” stands for “x is a commodity,” “p1–8 = 0” expresses a conjunction of eight idealizations like, for example, the difference between rate of surplusvalue characteristic of the capital used to produce x and the average rate of surplus-value of the capital engaged in the economic system in question as a whole where x is produced equals zero; the difference between the organic composition of agricultural capital in the economy producing x and the average organic composition of the remaining capital in this economy equals zero; the difference between the organic composition of capital in the branch of production turning out x and the average organic composition of the capital engaged in the economic system where x is being produced equals zero; the difference between the organic composition of capital used in the production of x and the average organic composition of capital in the branch of production turning out x is equal to zero, etc.; the expression on the right side states the dependence of the price P of commodity x on its value V if the eight idealizations hold. 3

Possible Objections and Proposals for Revisions

Even if Nowak’s reconstruction of the structure of the idealized law and of the explanation based on it has, especially when compared to Hempel’s D-N model, indisputable merits, still it can be further improved.3 Let me therefore 3 For a detailed analysis of these merits see: Černík (1978), Viceník (1988) and Hanzel (2007).

46

Hanzel

state certain objections against the reconstruction of the second dynamic law of classical mechanics in (3) as well as against the reconstruction of Marx’s law of value from the beginning of the first volume of Capital in (4). 3.1 Classical Mechanics in the Context of Newton’s Principia Nowak maintains that “[a]ccording to the proper scientific method an ­investigator should separate the principal and secondary factors for a given phenomenon and abstract from them establishing the law connecting the ­phenomenon with the principal determinants of it” (Nowak 1980, p. 37). The result of the application of this method can by symbolically expressed as E(k) = fk(H) as given in the consequent of (1). Nowak supposes that this expressions can be applied to the structure of the second dynamic law of mechanics, that is, the equation involved in it could be restated into the form Fr = f2(m, a). This would entail that mass and acceleration are the principal factors, with force being their phenomenal form which they together produce. According to classical mechanics, however, this does not hold. It is force that is the ground, and the phenomenal form (effect) it produces on a body with mass m is the ­acceleration a of this body. From the point of view of the structure of the formula (1), an adequate representation would thus be ​a​ (​ 2)​​ = ​__  Fr ​. Here, however, m one has to take into account a special feature of classical mechanics which goes back as far as Newton’s Principia. Newton employs in the Principia a bi-directional method of thought-­ movement: from effects of forces to the forces themselves, and from the forces to their effects. He declares in its 1687 preface: “The basic problem of philosophy seems to be to discover the forces of nature from the phenomena of m ­ otion and then to demonstrate the other phenomena from these forces” (1999, p. 382). The cyclical nature of thought movement in the Principia can be readily seen when one compares, for example, Definitions 4 and 8 with the Laws (Axioms) 1 and 2 (1999, p. 405, p. 407, p. 416): Definition 4 Impressed force is the action exerted on a body to change its state either of resting or of moving uniformly straight forward.

Law 1 Every body preserves in its state of being at rest or of moving uniformly straight forward, except insofar as it is compelled to change its state by forces impressed.

Definition 8 The motive quantity of centripetal force is the measure of this force that is ­proportional to the motion which it ­generates in a given time.

Law 2 A change of motion is proportional to the motive force impressed and takes place along the straight line in which that force is impressed.

The Inherent Type of Scientific Law

47

Each of these two laws is a converse of its respective definition, or, to be more precise, a special type of “converse.” In the definitions we identify (discover) the cause (impressed force on a body, centripetal force) by means of its effects (change of state of this body, the motive quantity generated proportional to the motion generated in a given time), that is to say, we proceed from the effects of forces to the forces themselves. In both laws, we proceed from the forces (their non-action on a body; motive force) to their respective effects (the preservation of the state of the body; change of motion in time). Accordingly, the question that should be posed here is then: how is the movement characterized by Newton as that of the type phenomena … → … ­other phenomena accomplished and how are the effects-phenomena which constitute the as yet unexplained/non-derived, presupposed points of departure ­related to the derived/explained effects-phenomena? In order to answer this question, from the point of view of the Principia, let us turn to the formulas for the time-effects and path-effects of force F, ∫Fdt and ∫Fds. We discover that in both derived/explained effects is hidden, behind the symbol “F,” the accelerative measure of the centripetal force which is delineated in Definition 7 as follows “The accelerative quantity of centripetal force is the measure of this force that is proportional to the velocity it generates in a given time” (1999, p. 407). To understand this, one has to take into account that up to section 11 of Book i of the Principia Newton deals only with the accelerative measure of force, and considering the fact that the meaning of the term “mass” in the Principia is in fact derived on the basis of considerations of the force of gravity as given in Book iii, then F can be expressed only by its proportion to acceleration, that is, as F ∝ a. This means that the phenomenon of acceleration is only the unexplained/non-derived presupposition and never the explained/ derived result of thought-movements in the Principia. Because in the Principia acceleration has only the status of an unexplained/ non-derived presupposed phenomenon, it is possible to view Newton’s cyclical movement from Definition 8 to the Law 2 as partially negatively closed. This is readily seen in the mutual relation between that definition and that law. Because nowhere between this definition and this law does Newton give a measure of force other than the change of motion, what one can substitute for the phrase “motive force impressed” in this law is simply the phrase “the motion that it generates in a given time” from this definition, and so the whole circular movement ends up in the following claim, “A change of motion is proportional to the motion generated in a given time.” The reason why the Principia display this feature can be understood through an understanding of the type of knowledge which the Principia provides with respect to forces in general and to the force of gravity in particular. Regarding the former, Newton speaks about “considering in this treatise not the species

48

Hanzel

of forces and their physical qualities but their quantities and mathematical proportions, as I have explained in the definitions” (1999, p. 588), and in the System of the World he declares that: our purpose is only to trace out the quantity and properties of this forces from phenomena, and to apply what we discover in some simple cases as principles, by which, in a mathematical way, we may estimate the effects thereof in more involved cases… We said, in a mathematical way, to avoid all questions about the nature or quality of this force. newton 1946, p. 550

In a similar manner he states, regarding the force of gravity that, “thus far I have explained the phenomena of the heavens and of our sea by the force of gravity, but I have not yet assigned a cause to gravity” (1999, p. 943). So, in the Principia we will not find any explanation of the very forces and of the very force of gravity, and thus not the explanation why it generates acceleration as one of its phenomenal forms. The conclusion one can state here then is that Newton’s cycle phenomena … → … other phenomena is incomplete because the phenomenon-effects labeled by him as “velocity generate in a given time,” which is the point of departure of this cycle, is embedded into the end-point of this cycle; so as the former is left ­unexplained, so are the latter left partially unexplained. Based on this analysis of the Principia, I can state with respect to Nowak’s reconstruction (3) of the structure of the second dynamic law of classical mechanics the following. First, because classical mechanics is a neither a science about the origin of forces in general, nor a science about the origin of the force of gravity in particular, the equation ​a​ ​ ​(2)​​ = ​__  Fr ​​, even if it corresponds to Nowak’s formula m E(k) = fk(H), in terms of this lack of knowledge about the origin of forces, has to be restated in the form Fr =  Df ma(2) in the sense that the force as the cause of acceleration is defined by means of its effect acceleration as its phenomenal form. This, in turn requires that with respect to the second dynamic law of classical mechanics Nowak’s formula E(k) = fk(H) must be changed into the formula fk(H) = E(k), that is, the ground H is cognized (known) by means of one of its effects given in the kth idealized form. Second, someone could object that the distinction between “E(k) = fk(H)” and “fk(H) = E(k)” is only apparent (i.e., that it is not a real distinction) and can be eliminated by simply switching the sides in the latter. But such a reformulation cannot work with respect to the second dynamic law because here the ground

The Inherent Type of Scientific Law

49

cannot be expressed in it independently of its effect-phenomenon. One could here replace “H” in “E(k) = fk(H)” only by means of “​​f ​ *k ​(​E​ ​(k)​​)​” where “​​f k​ *​ ” denotes a function inverse to that denoted by “​​f k​ *​ ,” but such a substitution would end up in “​​E​ ​(k)​​ = ​f​ k​[ ​f ​ *k (​ ​E​ ​(k))​​ ​]​” and thus in the triviality “E(k) = E(k)”. By comparing the formulas “E(k)  =  fk(H)” and “fk(H)  =  E(k)” in the context of classical mechanics I am led to a distinction not accomplished in Nowak’s works (1971; 1972; 1974; 1980). Phenomenal forms obtained, cognized before the knowledge of the ground was obtained, and by means of which it is possible to move to the latter, I label forms of appearance of the ground. Phenomenal forms obtained after and derived from, that is, explained on the basis of knowledge of the ground, I label forms of manifestation of the ground.4 In the context of classical mechanics, for example, the path-effect as well as the time-effect of forces has the status of manifestations of force; contrary to this, acceleration in this context only has the status of a presupposed appearance of force, but never that of a derived manifestation. Stated otherwise: classical mechanics does not explain why force, once acting on a body necessarily manifests itself as the acceleration of this body. Nowak’s formula (1) can thus, with respect to the second dynamic law of classical mechanics, be restated as follows: ​​L​​ ​(k)​​ :  ∀ x​{U​(x)​ & C​mod​1–k   ​(x)​ = 0 → ​f1​ ​[H​(x)​]​ = ​E​ ​(k)​​(x)​}​​

(5)

I label the type of scientific law with the structure corresponding to that of (5) the pure idealized type of scientific law. Here “Cmod1–k” stands for a conjunction of k modification conditions, that is, conditions which – once they are at work – are causally relevant for the phenomenal forms; “E(k)” stands the form of appearance in the kth degree of idealization and “f1(H)” stands for a certain function of the ground H. The latter is viewed a ground of the phenomenal form denoted by “E(k)”, so that the former is cognized by the latter. By comparing the structure of formula (5) with that of (1) stated by Nowak it becomes readily seen that while the antecedent of (1) can be viewed as grasping the structure of the second dynamic law, its antecedent cannot. Now I will show that while the consequent of both (1) and (4) really corresponds to Marx’s understanding of the relation of price and value of a commodity, neither the antecedent of (1) nor that of (4) does. In this way I will also justify why in (5) the term “modification condition” is used instead of Nowak’s term “secondary factor.” 4 For a detailed account of this with respect to physics see: Hanzel (2008; 2012; 2013).

50

Hanzel

3.2 Marx’s Law of Value in the Context of the First Volume of Capital When carrying on the reconstruction of Marx’s law of value Nowak asserted the following: “since Marx has been keeping the law of value immutable throughout the two opening volumes of Capital until Chapter 8 of the third one, we shall enumerate assumptions on which the cited law is formulated” (Nowak 1971, p. 30). Yet the following two objections can be raised against this claim. First, Marx states in the manuscript “Chapter 6. Results of the immediate process production,” planned by him to be the conclusion of the first volume of Capital: The commodity that emerges from capitalist production is different from the commodity we began with as the element, the precondition of capitalist production. We began with the individual commodity viewed as an autonomous article in which a specific amount [Quantum] of labor time is objectified and which therefore has an exchange-value of a definite amount [Größe]. The commodity may now be further defined as follows: (1) What is objectified in it – apart from its use-value – is a specific quantum [Quantum] of socially necessary labor. But whereas in the commodity regarded by itself it remains quite undecided (and is in fact a matter of indifference) from whom this objectified labor is derived, the commodity as the product of capital can be said to contain both paid and unpaid labor. … (2) The individual commodity does not only appear [erscheinen] as a part of the total produce of capital, but as an aliquot part of the total produced by it. We are now no longer concerned with the individual autonomous commodity, the single product. The result of the process is [erscheinen] not individual goods, but a mass of commodities. marx 1988, p. 33; marx 1976b, pp. 953–954

This means that contrary to Nowak’s claim, the law of value at the end of the first volume of Capital cannot have the structure of (4) because in the latter both price and value are related to an individual commodity. In fact, when reflecting on the value at the beginning of the first volume of Capital, Marx considers just one commodity understood as an “average sample of its kind” ([1872] 1987, p. 73; 1976, p. 130). Second, a closer look at the first three chapters of Capital’s first volume, where such terms like “value,” “exchange value,” “price,” etc. are used, discloses the fact that the terms “capital,” “surplus-value,” and “organic composition” are not employed. In order for Nowak’s reconstruction of the law of value to hold already for those chapters, Marx would have had to employ in them the

The Inherent Type of Scientific Law

51

term “variable capital” understood as the value which the worker obtains in exchange for the time in which he/she performs labor.5 Marx, however, states there that “The reader should note here that we are not speaking here of the wages or the value the worker receives (e.g.,) a day’s labor… At this stage of our investigation, the category of wage does not exist at all” (Marx [1872] 1987, p. 78; 1976, p. 135). This then means that from the idealizations given in Nowak’s formula (4) those I listed above certainly cannot be given in first three chapters of the first volume of Capital.6 Against Nowak’s (4) one can state an even stronger objection, namely, that while its consequent relates price to its value, it does not state that law of value employed by Marx before he derives the price from value in beginning of the first volume of Capital. In this law the very value of a commodity is related to the labor (L for short) expended on it in its production; that is, it contains the relation expressed as “V = k(L),” where k expresses a functional relation. Here, what becomes apparent is one crucial difference between the second dynamic law in the context of Newton’s Principia and the law of value in the beginning of the first volume of Capital, namely, that while in the former the cause of the ground (i.e., the origin of forces producing acceleration) is not known, in the latter it is; the labor embodied in a commodity is the cause of the value which, once interacting with other commodities in exchange, acquires the phenomenal form of price. One could, of course, replace the expression “P(8)  =  l[V]” in (4) by the ­expression “V = k(L),” thus obtaining (here m[L] stands for “l(k(L))”): ​​L​ v​(8 ​ )​​ :  ∀ x​{C​(x)​ & ​p1–8 ​  ​(x)​ = 0 → V​(x)​ = m​[L​(x)​]​}​​

(6)

However, (6) is plagued by several problems. First, the idealization involved here are completely irrelevant for the relation between labor and value; they are relevant only for the deviations of quantitative characteristics of the price of a commodity from quantitative characteristics of its value. By removing the expression for the idealizations one obtains: ​​L​ v​​ :  ∀ x​{C​(x)​ → V​(x)​ = m​[L​(x)​]​}​​

(7)

5 Under the term “labor” I understand here and in the rest of this paper what Marx labeled as “socially necessary abstract labor.” 6 In (Hanzel 1999) I show why some of the remaining idealizations also cannot be given in these three chapters.

52

Hanzel

Such a reconstructed law of value, however, contradicts Marx’s understanding of it. For Marx a product turns into a commodity once it is produced by labor and thus has the property of value; it is in fact a “value-thing.” (7), however, does not express this relation; given the nature of the sentential connective symbolized as “→”; it cannot exclude the case that there can be an entity which has value but is not a commodity. In addition, the expression on the right side does not correspond, from the point of view of the origin of value in the production where labor is expended, to the left side. About the former we suppose that x is already given as commodity, while about the latter that value of x is related to the labor expended in the process of production. The final objection is that all the above given reconstruction of the law of value share one and the same deficit: they do not reconstruct the conditions under which work becomes labor so that the product is in fact a commodity endowed with value.7 From Marx’s reflection in the first chapter in the first I ­reconstruct these conditions as follows: the producer is a private (separate) producer and his/her subjective aim is the exchange of commodities as u ­ se-values with other commodity producers. The law of value can then be ­expressed as follows: ​​L​ v​​ :  ∀ x ∀ y​{P​(x, y)​ & E​(x, y)​ ⎯  ​  n   →​  V​(y, x)​ = g​[L​(y, x)​]​}​​

(8)

Here “P(x, y)” stands for “x is the private owner of the enterprise producing the product y” “E(x, y)” for “x’s subjective aim is the exchange of y for products produced in enterprises owned by other private producers,” “V(y, x)” for “the value of y produced in x’s enterprise” “L(y, x)” for “labor expended on y in x’s enterprise” “​⎯  ​  n   →​”  for the modal sentential connective “if __, then necessarily comes into being…” and “g” denotes a function. By employing the symbol “​⎯  ​  n   →​​”  I depart in yet another aspect from Nowak’s reconstruction (3) of the law of value. There Nowak used the symbol “→” for the sentential connective “If __ then…” and which, he claims, should appear in the language stating the relation between the price and value of a commodity as given in the first volume of Capital. In distinction to Nowak, I opt for a modal connective because Marx employs in his linguistic expressions for the relation of value to price as well labor and value employs alethic modalities. So, for example, he states: “Money as measure of value is the n ­ ecessary form of manifestation of the immanent measure of value of commodities, the labor time” (Marx [1872] 1987, p. 121; 1976, p. 188), and “The 7 The term “work” denotes what Marx labels by the German “konkrete Arbeit.”

The Inherent Type of Scientific Law

53

­magnitude of value expresses…a necessary, to its process of production inherent relation to the social labor-time” (Marx [1872] 1987, p. 128; 1976, p. 196). The type of necessity being here at work is such that everywhere where the conditions expressed in the antecedent of (8) are given, everywhere there necessarily the relation expressed in its consequent comes into being.8 What Marx meant by “­immanent measure” and “manifestation of immanent m ­ easure” I will e­ xplicate below. In (8) the meaning of the term “labor” stands for the cause of the value which, as shown above, has the status of the ground of the price. In addition, the conditions stated in the antecedent of (8) stand for the two conditions which once given together turn work into labor and the product produced by it into a commodity. Following (Černík 1978) I label the conditions under which the cause (C, for short) of a ground (H, for short) of certain phenomena exist the inherent conditions of the cause of the ground (Cin, for short). The type of law in which it appears and which can be viewed as a methodological generalization of (8) I label as the inherent type of scientific law and its structure is as follows: ​ ​L​ i​ :  ∀ x​{C​in​1–l   ​(x)​ ​ ⎯  n   →​  H​(x)​ = ​f2​  ​[C​(x)​]​}​​

(9)

In the antecedent is stated the conjunction of the 1st through lth ­inherent ­conditions under which the cause related to the ground by the relation ­expressed in the consequent exists. As I will show below, the methodological reconstruction of the inherent conditions enables one to reconstruct methodologically how Marx accomplishes the explanatory move from the law of value given in the first three chapters of Capital, Volume i, to the law of value given later in Chapter 5. Because Marx derives – after stating at the beginning of the first volume of Capital, the law of value relating value and labor – in the third chapter of this volume the meaning of the term “price,” with respect to a certain amount of weight of the commodity gold, the law of price as given in this chapter can tentatively be reconstructed as follows: ​L​ p​ :  ∀ x ∀ y ∀ w​{P​(x, y)​ & E​(x, y)​ & ​[D​(y)​ − S​(y)​ = 0]​ & ​[D​(w)​ − S​(w)​ = 0]​ ⎯  ​  n   →​   P​(y, w)​ = m​[L​(y)​, L​(w)​]​}​​

(10)

8 In the terminology of possible worlds this means that it is not important if these conditions are given in all possible worlds, or in some of them, or just in one possible world.

54

Hanzel

Here “w” denotes the commodity gold; a certain amount of its weight enters into the exchange ratio to the product y, and where this ratio is given by the ratio of the labors expended on the production of y and w, respectively, while the two idealizations state that the demand for y and w, respectively, equals their supply. This reconstruction I regard as tentative because, in order not to overburden the formula, I do not take into account the existence of the producer of the gold as a commodity, while the number of idealizations at work in the third chapter of the first volume of Capital could be higher. Formula (10) can be methodologically generalized. I propose the following formula for the structure of the type of idealized law, which I label the inherent idealized type of scientific law: ​ ​L​ ​(iik​ )​​ :  ∀ x​{C​in​1–l   ​(x)​ & C​mod​1–k   ​(x)​ = 0 ​⎯    n   →​  ​E​ ​(k)​​(x)​ = ​f3​  ​[H​(x)​]​}​​

(11)

It can be viewed as a generalization of the situation when cognition acquires knowledge why the ground H, which exists once the inherent conditions ­Cin1–l are given, produces, once the idealizations Cmod1–kx  =  0 hold, the form of ­manifestation E(k). Thus, since Marx’s derivation of the price from value is based on antecedent grounding of value in the expended labor or, stated in more general methodological vocabulary, the derivation of the form of manifestation E(k) from its ground H is based on the antecedently given derivation of the latter from its cause C, it is possible to substitute into the expression “E(k) = f3[H]” from (11) for “H” the expression “H = f2[C]” from (9). One then obtains (“f4[C]” denotes the composed function denoted by “f3[ f2(C)]”): ​ ​L​ ​(iik​ )​​ :  ∀ x​{C​in​1–l   ​(x)​ & C​mod​1–k   ​(x)​ = 0 ​⎯    n   →​  ​E​ ​(k)​​(x)​ = ​f4​  ​[C​(x)​]​}​​

(12)

By comparing the type of idealized law with the structure (5) with the type of idealized law with the structure (11) it becomes readily seen that Nowak’s reconstruction of the structure of idealized law as given in (1) is in fact a fallacious “unification” of the antecedent and the sentential connective given in what I labeled above as pure idealization type of scientific law with the consequent given in what I labeled above as inherent idealized type of scientific law. So, at least in my view, the antecedent and the consequent in Nowak’s reconstruction of the structure of idealized law (1) do not match each other; the antecedent requires a different consequent and the consequent requires a different antecedent together with the reconstruction of another type of sentential connective.

The Inherent Type of Scientific Law

4

55

Explanations Based on the Inherent Type of Scientific Law and the Idealized Types of Scientific Law

Based on the above given differentiation between three types of scientific law I can now propose three different models of explanation based on them. What is common to all of them is the fact that they are based on the employment of knowledge about the respective conditions being at work. With respect to the inherent type of scientific law one can reconstruct how Marx moves from the law of value as given in the first three chapters to the law of value as given in Chapter 5 of the first volume of Capital. This movement, which I view as that from an explanans to an explanandum is mediated by Marx’s reflection – in fourth chapter of this volume – on the inability to explain on the basis of the law of value as given in the first three chapters the phenomenon of the production of the money-profit in the economic cycle ­expressed by him as “M → C → M’”, where “M” and “C” stand for money and commodity, respectively, while the profit is given by the difference expressed as “M’ – M”. In order to overcome the explanatory limitation of the law of v­ alue as given in the first three chapters, he brings in two new inherent conditions: the (subjective) condition that the aim of the producer is to obtain a profit, symbolized as “Cin1”, and the (objective) condition that there exists a free owner of his own capability to work selling this capability to the owner of the means of production and subsistence from which he himself/herself is separated, symbolized as “Cin2”. Marx’s explanatory move from the law of value as given in the first three chapters of the volume to that given in the fifth chapter can then be expressed as follows: ​ ​L​ v​ & C​in​​ 1,2​  ╡​L​ v​(sv)​​​

(13)

Here “╡” stands for the explanatory derivation while “Lv(sv)” stands for the ­scientific law of value once the meaning of the term “surplus-value” is integrated into it. Its structure in the fifth chapter I reconstruct as follows: ​L​ v​(sv)​ ​ :  ∀ x ∀ y ∀ z​{P​(x, y)​ ∧ M​(x, y)​ ∧ W​(z, x, y)​ ⎯  ​  r   →​  ​VSV ​  ​(z, y, x)​  ​  ​(z, y, x)​]​}​​ = h​[​LSL

(14)

Here “M(x, y)” stands for “x’s subjective aim is to make profit by selling the product y”; “W(z, x, y)” for “z is the owner of workforce being sold to x and then consumed in the production of y”; “VSV(y, z, x)” for “the value (­involving

56

Hanzel

­surplus-value) of y produced by z for x”; “LSL(y, z, x)” for “labor (involving ­surplus-labor) expended by z on the production of y for x” and “h” denotes a function. Here “VSV” denotes only the newly created value and not the total value of a commodity. Into the latter enters also the value of the material inputs of its production process, and where the value of these inputs is labeled by Marx in the sixth chapter as “constant capital.” Having distinguished between the two types of idealized scientific law (5) and (12), we face now the following question: Is the course of scientific explanation proceeding from the former identical or different to the course of explanation proceeding from the latter? As shown above, the crucial difference between these two types of idealized law is that the inherent idealized one, contrary to the pure idealized one, draws on the knowledge of the cause of the ground which is expressed in the inherent type of scientific law, that is, the origin of the ground in its cause is already known in both inherent types of scientific law. Therefore, the question stated above can be answered once we determine the origin of the modification conditions (Nowak’s secondary factors) as they are reconstructed in Nowak’s reconstruction of the law of value (3). The modification condition relating the difference between rate of surplusvalue characteristic of the capital used to produce x and the average rate of surplus-value of the capital engaged in the economic system in question as a whole where x is produced, has its origin in the difference between their respective value-composition, and thus also in the quantitative difference of the living labor commanded in them respectively. The same holds also for the remaining modification conditions. The organic composition of capital is, according to the law of value given in the fifth and subsequent chapters of the first volume of Capital, a function of the labor embodied in it. Thus, what holds for all these modification conditions is that they are functions of the cause in which the ground of value has its origin. This functional dependence I express generally as Cmodk = Cmodk(C).9 The method of explanation by gradual concretization whose explanans contains a scientific law of the inherent idealized type (12) can then be expressed symbolically as follows: ​ ​L​ (​iik​ )​​ & C​mod​ k​ ≠ 0⊣ ​L​ (​iik−1 ​  )​​​

(15)

Here “​⊣​” stands for the explanatory procedure of gradual concretization, while Lii(k-1) stands for an explanandum law whose structure is as follows: ​L​ (​iik−1 ​  )​​ :  ∀ x​{C​in​1–l   ​(x)​ & C​mod​1–​   (k–1)​[​ C​(x)​]​ = 0 & C​mod​k  ​[C​(x)​]​ ≠ 0 ​⎯    n   →​   ​(k–1)​ ​E​  ​ x = ​f5​  ​[C​(x)​, C​mod​k  ​(C​(x)​)​]​}​​ (16) 9 For a similar analysis of the remaining modification conditions stated by Nowak in (1980) see (Hanzel 1999, pp. 73–74).

The Inherent Type of Scientific Law

57

What is the situation in the case when explanation is based on the explanans law of the pure idealized type of scientific law (5)? Because it stands for a type of cognition which expresses the knowledge of the ground based on the knowledge of its form appearance, here the knowledge of the very ground which would be based on the knowledge of its cause is not given. Therefore, the only possible way for the derivation/explanation of the forms of manifestation E(k-1),…, E(k-j) by means of gradual concretization is in fact based on the knowledge about the form of appearance E(k). The method of explanation by gradual concretization whose explanans contains a scientific law of the pure idealized type (5) can then be expressed symbolically as follows: ​ ​L​ ​(k)​​ & C​mod​ k​ ≠ 0⊣ ​L​ ​(k–1)​​​

(17)

Here L(k-1) stands for an explanandum type of law whose structure is as follows: ​L​ ​(k–1)​​ :  ∀ x​{U​(x)​ & C​mod​1–​   (k–1)​(​ x)​ = 0 & C​mod​k  ​(x)​ ≠ 0​ E​​ ​(k–1)​​(x)​  = ​f6​  ​[​E​ ​(k)​​, C​mod​k  ​(x)​]​}​​

(18)

Here I take into account that the ground denoted as “H” is known only by means of the form of appearance denoted by “E(k)”; so the former symbol can be substituted by the latter. The formula (18) once united with the formula (5) on which it draws, I view as the methodological reconstruction of Newton’s claim about the method he applies in his Principia, namely “to discover the forces of nature from the phenomena of motion and then to demonstrate the other phenomena from these forces” (1999, p. 382). While in the pure idealized type of law (5) the appearance E(k) is the phenomenon serving as the point of departure enabling to identify its ground, the forms of manifestation E(k-1),…, E(k-j) are Newton’s “other phenomena” which are the end-points of derivation/explanation based on the knowledge of the ground which is already given and based on the knowledge of the form of appearance E(k). Newton’s movement from E(k) to E(k-1),…, E(k-j) has an analogue in Marx’s movement in the beginning of the first volume of Capital, where he departs from the phenomenon as appearance labeled “exchange-value” and then is able to derive another phenomenon he labels as “price”.10 There exists, however, an important difference in their respective movements of the type phenomena → … → phenomena. Newton, as shown above, succeeds only partially in the movement of the type phenomena → other phenomena because he is not able to perform the movement of the type phenomenon → the same phenomenon. 10

On a detailed reconstruction of this movement in all chapters of the first volume of C ­ apital see (Halas 2012a, 2012b, 2012c, 2013).

58

Hanzel

Marx, in distinction to Newton, succeeds in performing completely the movement of the type phenomena → … → other phenomena because he succeeds in movement of the type phenomenon → … → the same phenomenon. This holds for what he labels as the “exchange value of commodities” on which he declares the following: The progress of our investigation will lead us back to exchange-value as the necessary mode of expression, or form of manifestation, of value… In fact we started from exchange-value, or the exchange relation of commodities, in order to track down the value that lay hidden within it. We must now return to this form of manifestation of value. marx [1872] 1987, p. 72; p. 80; 1976, p. 128; p. 139

5

A Typology of Measures

There exists yet another methodological difference between Marx’s approach to value and Newton’s approach to force, namely, that which pertains to the epistemological category of measure. As shown above, Newton defines the quantitative characteristic of the magnitude of centripetal force in Definitions 7 and 8 by means of the quantitative characteristics of two different magnitudes, namely, of the changes of velocity in time and of motion in time. These two magnitudes thus acquire with respect to the first in cognition the status of a measure. And, so as these two quantities stand to the first quantity in a relation which is the of type phenomena as appearances to their own ground, the category of measure being here at work is that of an external measure. Based on this category it is possible to make more precise the claim given above, namely, that the phenomenal forms with the status of appearances enable to obtain knowledge about their ground. Knowledge about the quantitative characteristics of phenomena as appearances enables to obtain knowledge about the quantitative characteristics of their ground. And since, as shown above in formula (18), the knowledge about the form of appearance denoted as “E(k)” is involved in the knowledge about form of manifestation denoted as “E(k-1)”, the latter’s form quantification is also based on the knowledge of the external measure provided by the knowledge of that form of appearance. A more differentiated typology of measures can be discerned in the first three chapters in the first volume of Capital. Here, Marx, like Newton, draws initially on the category of external measure in the sense that the quantitative proportion in which commodities are mutually exchanged can be viewed as

The Inherent Type of Scientific Law

59

pointing to a quantitative characteristic which could be inherent to each commodity independently from their mutual proportions in which they are mutually exchanged; it is the quantitative characteristic of a commodity’s value. In a next step, Marx grounds the quantitative characteristic of value of a commodity in the quantitative characteristic of labor (i.e., in the time in which it is) expended on the production of this commodity. So, this grounding stands for the quantification of a ground by the quantity of its cause; the latter quantity acquires, with respect to the former, the status of immanent measure. Finally, by applying the category of measure to the formulas (12) an (16), it is obvious that Marx derives the quantitative characteristics of the various forms of manifestations like exchange value and price on the basis of the knowledge of the quantitative characteristics of value as their ground, after he grounded the latter in the quantitative characteristic of its cause – the quantitative characteristic of labor – its time-duration. This intertwining of the knowledge about these three quantitative characteristics means that knowledge of the quantitative characteristics of the manifestations is in the long run derived from the knowledge about the quantitative characteristics of the cause of their ground; the latter knowledge acquires with respect to the former the status of the knowledge about the manifestation of the immanent measure of their ground’s cause. Based on the differentiation between the three varieties of the category of measure (i.e., external measure, immanent measure and manifestation of immanent measure), it is possible to delineate in a more exact way the paradigmatic status of classical mechanics in the context of the Principia and of the first volume of Capital. The former is a paradigmatic example of an empirical science which has no knowledge about the quality of the ground in the entities under investigation, but has already knowledge about its quantities derived from and measured by the quantities of certain kind of its effects, and about the quantities of its other effects derived from and measured by the quantities of this ground as measured in the here indicated way. On the other hand, the first volume of Capital can be viewed as a paradigmatic example of an empirical science which has knowledge about the quality and quantities of the ground in the entities under investigation based on the knowledge about the quality and quantities of the cause of the ground, as well as knowledge about the quality and quantities of the effects derived from and ­measured by the quality and quantity of ground’s cause.11

11

For a reconstruction of the qualitative aspect of Marx’s approach to value see (Hanzel 2010b).

60

Hanzel

6 Conclusions Based on the above given reconstruction of the three types of scientific law and of the methods of explanation based them I can now delineate the directions of future methodological research which could be based on this reconstruction. First, as one possible direction worth pursuing is the methodological reconstruction of the actual measurement procedures based on these three types of scientific law. Here I mean the actual procedures employed in astronomy for the computation of the quantities of magnitudes like mass and density of planets and of their moons, of asteroids, etc., as well as the computations of the quantities of magnitudes like production prices, average rates of profit, value and surplus-value in the world of economics.12 Second, the reconstruction of the inherent type of scientific law displays a potential worth mentioning here, namely, that it reconstructs the conditions relevant for the very existence of the ground of certain phenomena. This predetermines it to be involved in a type of scientific explanation which can be viewed as a type of historical type of explanation. The meaning of this becomes more understandable if one turns to the article (Glennan 2010). Glennan employs here the category mechanism – whose role in the methodological analysis of scientific explanation comes quite close to that of ground employed by me here – and differentiates on its bases between the historical type of explanation and the ahistorical type of explanation as follows (2010, pp. 252–253): [Historical type of explanation]

[Ahistorical type of explanation]

I will take it to be the defining characteristic of an historical explanation that it explains the occurrence of some particular event or state of affairs by describing how it came to be.

Ahistorical explanations abstract from particular events at particular places and times in order to explain recurrent patterns of phenomena […] one does not explain a particular event, but some sort of repeatable phenomenon […] the same explanation works for potentially countless instances of that kind of phenomenon.

Worth to be noted here is that when Glennan speaks about the “­historical explanation,” he bypasses the issue how to deal methodologically with ­

12

For a tentative reconstruction of these computations as given in the Principia as well of the econometric implications of Marx’s economic works see (Hanzel 2010a).

The Inherent Type of Scientific Law

61

e­ xplanatory procedures targeting the history of the very mechanism, or in my terminology, of the very ground. The category history of the very ground is ­understand here as the formation of a ground with its constituent entities and activities out of another, declining ground with its own specific constituent ­entities and activities, and the decline and disappearance of a ground due to its transformation into another ground. This means that under the historical type of explanation one can understand not only, as Glennan does, a movement from the explanans-description of the working of a ground to the ­explanandum-description of a particular event produced by this ground under certain particular conditions, but also as a thought movement from an explanans to an explanandum where both stand for a description of a ground, mutually, of course, differing in the described entities, activities, and existing under different inherent conditions. This predetermines the inherent type of scientific law to be involved in an explanation type where in the explanans appears an inherent type of law pertaining to a ground while in the explanandum would be given an inherent type of law pertaining to a ground different from the former one. This type of ­explanation would enable to reconstruct history as a transformation of a ground into a different ground; each being at work under different inherent conditions and being grounded in different causes. The methodological reconstruction of this type of scientific explanation would thus bring to the fore a potential of the inherent type of scientific law which is not given in the pure idealized type, namely, that the latter provides the basis only for the type of ­explanation by gradual concretization in the course of which the ground and the respective universe of discourse are always same; this feature is readily seen when one compares the structure of the explanandum type of law L(k-1) with that of the explanans type of law L(k) as given above in (18) and (5). References Černík, V. (1978). Problém zákona v marxistickej metodológii vied. [The Problem of Scientific Law in the Marxist Methodology of Science]. Bratislava: Vydavateľstvo Pravda. Glennan, S.S. (2010). Ephemeral Mechanisms and Historical Explanation. Erkenntnis 72, 251–266. Halas, J. (2012a). Kategoriálny rámec Marxovej teórie nadhodnoty (1. časť) [The Categorial Framework of Marx’s Theory of Surplus Value (Part 1)]. Filozofia 67, 93–103. Halas, J. (2012b). Kategoriálny rámec Marxovej teórie nadhodnoty (2. časť) [The Categorial Framework of Marx’s Theory of Surplus Value (Part 2)]. Filozofia 67, 208–220. Halas, J. (2012c). Ryle and Marx on Absurdities. Organon F 19, 338–360.

62

Hanzel

Halas, J. (2013). Metodológia kritickej sociálne vedy [The Methodology of Critical Social Science]. Bratislava: Vydavateľstvo Univerzity Komenského. Hanzel, I. (1999). The Concept of Scientific Law in Philosophy of Science and Epistemology. Dordrecht: Kluwer. Hanzel, I. (2007). Leszek Nowak on Scientific Laws and Scientific Explanation. In: J. Brzeziński et al. (eds.), The Courage of Doing Philosophy, pp. 129–140. Amsterdam: Rodopi. Hanzel, I. (2008). Idealizations, Concretizations in Laws and Explanations in Physics. Journal for the General Philosophy of Science 39, 273–301. Hanzel, I. (2010a). Studies in the Methodology of Natural and Social Sciences. Frankfurt am Main: Peter Lang. Hanzel, I. (2010b). Mistranslations of “Schein” and “Erscheinung.” Science and Society 74, 509–537. Hanzel, I. (2012). Causation, Principle of Common Cause and Theoretical Explanation. Journal for the General Philosophy of Science 43, 29–44. Hanzel, I. (2013). Data, Phenomena and Mechanism. Organon F 20, 302–322. Marx, K. (1976). Capital, Vol. 1. Harmondsworth: Penguin. Marx, K. ([1872] 1987). Das Kapital, Erster Band. MEGA II/6. Hamburg and Berlin: Otto Meissner and Dietz. Marx, K. (1988a). Resultate des unmittelbaren Produktionsprozesses. In: Marx (1988b), pp. 24–135. Marx, K. (1988b). Ökonomische Manuskripte 1863–1867. Marx-Engels Gesamtausgabe (MEGA) II/4.1. Berlin: Dietz Verlag. Newton, I. (1946). Mathematical Principles of Natural Philosophy and the System of the World. Berkeley: University of California Press. Newton, I. (1999). The Mathematical Principles of Natural Philosophy. Berkeley: University of California Press. Nowak L. (1971). U podstaw Marksowskiej metodologii nauk [Foundations of Marxian Methodology of Science]. Warszawa: PWN. Nowak L. (1972). Laws of Science, Theories, Measurement. Philosophy of Science 39, 533–548. Nowak L. (1974). Zasady marksistowskiej filozofii nauki [The Principles of Marxist ­Philosophy of Science].Warszawa: PWN. Nowak L. (1980). The Structure of Idealisation. Dordrecht: Kluwer. Viceník, J. (1988). Spory o povahu metodológie vied [Disputes over the Nature of ­Methodology of Science]. Bratislava: Vydavateľstvo Pravda.

chapter 3

On Deformational Modeling: Max Weber’s Concept of Idealization Lidia Godek Abstract The article aims to reconstruct the idealization procedure understood as a deformational means of concept modeling in the theory proposed by Max Weber. The ideal type represents the basic form of the deformational transformation. Deformational modelling refers to a strategy of conscious and deliberate distortion of an object of empirical reality in varied and consequently contrafactual ways. The method essentially seeks to account for a concept by highlighting significant characteristics of the empirical content of investigated socioeconomic phenomena at the expense of their actual exemplification. The ideal type is a deformed means of representing a selected real-life phenomenon or object, oriented towards the fulfilment of specific cognitive goals while taking into account all methodological conditions involved in the process of its construction. By reference to Leszek Nowak’s typology of deformational procedures, it is possible to characterize the Weberian ideal-type method as representing an example of quantitative deformation (positive potentialization).

1

Introduction: Delineation of the Problem

The present study seeks to provide a few preliminary critical remarks regarding the discussion which has been revolving around idealization and Max ­Weber’s ideal types. The problems are interesting not only in the aspect of correct ­interpretation of Weberian thought but also, or perhaps predominantly, from the viewpoint of clear and adequate self-knowledge of the idealization method. What I claim is that Weber’s position, and therefore also the critique of his concept for the idealization model formulated in Poznań School of Methodology, have always carried a fundamental methodological and philosophical significance.1 This is especially evident if one takes a look at problems taken up by idealization theory in the context of the dispute with the positivist concept 1 It is generally recognized in literature dealing with the topic that Poznan School of Methodology emerged with the publications Leszek Nowak (Nowak 1970; 1971; 1980).

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_005

64

Godek

of science.2 The most important shared area is undoubtedly the domain of the method of exploring and theoretical modeling of theoretical notions for the cognition of sociohistorical reality (Weber [1904] 1949, pp. 91–92).3 It is quite lucid today that studies into Weberian thought, particularly in the field of methodology of human sciences, have brought a far-reaching consensus with respect to two central components of Weber’s framework. Nearly everyone seems to concede that: 1. The main points of Weber’s concept pertain to the method and s­pecific problems involved in exploring the reality of culture (which is ­documented in the anti-positivist streak of the Weberian system), i.e. what was ­integrated into Weber’s works as a method of ideal-typical ­construction of scientific notions (Rogers, 1969). 2. The keynote of Weber’s explorations is the principal question about the logical function and structure of theoretical concepts that are specific to sciences that investigate culture (Weber [1904] 1949, pp. 83 and 85). Major differences in opinion and methodological controversies come into view in the interpretation of the latter issue. What the majority of prominent authors bring to the fore here – in diverse aspects and with varying degrees of rejection – is the inadequacy (or apparentness) of proposals put forward by Weber to justify the exclusivity of ideal-typical notions for the modeling of theoretical concepts in humanist and social sciences. In particular, it is argued that Weber’s account of ideal types as a specific method of cultural cognition has proven unsubstantiated in theoretical terms (Nowak 1971, p. 86), and i­ llusory in practice (Lachmann 1971). What also draws attention here is the 2 It is worthwhile to note at this point that Weber’s view on positivism was complex and largely determined by problems that needed to be addressed. For example, in the “dispute about method” (Methodenstreit) Weber often reached for the argumentation formulated by ­positivists (e.g. Carl Menger). However, in matters related to conceptual realism his views were quite critical. Philosophical reflection over science finds its destiny in methodological explorations and ultimately concretizes itself in attempts to precisely define the concept of ideal type and outline its functionality in studying culture (the society). Here, Weber e­ merges primarily as a critic of methodological naturalism and conceptual realism. ­Furthermore, when Weber argues against positivism – his campaign essentially directed against the ­positivist cumulative concept of science and concept of objectivity – he still proposes the development of an “intermediate road” between the greed of the senses (hermeneutics) and the taste of the facts (positivism); compare Weber ([1907] 1982, pp. 489 – 540). 3 Problems related to Weber’s ideal types in the context of Poznań School of Methodology were addressed by Nowak (1971, 1973 and 1978), Nowakowa (2007) and Brzechczyn (2012).

On Deformational Modeling

65

fact that most debates which are centered around ideal types as a procedure of constructing scientific concepts expose their problematic cognitive status, pointing mainly to difficulties with their universal applicability and empirical verifiability.4 This stern view on Weber’s proposed conception, however, seems rather rash and imprudent without a prior clear and distinct discussion of some of its underlying beliefs. The relevance of such discussion lies in the fact that what can be unanimously assumed with regard to Weber’s methodological reflection as a whole is the problem-free, fairly self-evident and – crucially – ­homogeneous scope and substance of the idealization procedure as a deformational means of concept modeling. The line of thought assumes, in conformity with Leszek Nowak’s tenets, that the idealization method is anti-positivist to the extent to which it is deformational (Nowak 1997, pp. 235–236). In order to gain a detailed insight into Weber’s method of deformational modeling, it is necessary to reconstruct the method of concept creation taking into account the elements enumerated below: 1. Deformed means of representing a selected phenomenon: in Weber’s framework the basic form of transformation is the ideal type (Weber [1904] 1949, p. 83). 2. Determination of cognitive goals of modeling: with an aid of ideal types ­Weber exposes certain prominent constituents of the empirical content, documenting the “specificity” of the empirical phenomenon under study (ideal types are used for examining the historical uniqueness of studied phenomena) (Weber [1904] 1949, p. 101). 3. Methodological conditions associated with the process of model construction: in Weber’s proposed system they are rendered by a special abstraction procedure (Weber [1904] 1949, p. 100). 2

Cognitive Goals of Weberian Idealization

Weber outlined his conception of cultural sciences in the text Objectivity in Social Science published in 1904. Every domain of knowledge, Weber argues, 4 Weber himself was aware of these weaknesses, stressing that the validity of ideal types is inherently relative and problematic always when ideal types are regarded exclusively as historical presentations of empirical data. However, if this characterization of ideal types is dropped, ideal types have an enormous heuristic value as an instrument for measuring and comparing empirical content (Weber [1904] 1949, p. 97).

66

Godek

makes use of a specific method and means of acquiring knowledge, and a set of tools for organizing that knowledge. It seems that the most crucial of all preliminary problems associated with Weber’s idealization pertains to the type of method that needs to be embraced by the reader. The core of the concept is the assumption that the specificity of the cognitive goal of each study conducted in the field of social sciences lies in its dependence on the “one-sidedness” of viewpoints which have an impact not only on the selection of the study object but also on its analysis and manner of presentation. An analysis of the way in which Weber outlines the domain of culture cognition produces a scheme of research procedure comprising three main steps. The first cognitive task facing the investigator, Weber claims, is to isolate an interesting phenomenon from a chaos of diverse phenomena with an aid of value ideas embraced by the investigator and his times. Order is brought into this chaos only on the condition that in every case only a part of concrete reality is interesting and significant to us, because only it is related to the cultural values with which we approach reality. Only certain sides of the infinitely complex concrete phenomenon, namely those to which we attribute a general cultural significance – are therefore worthwhile knowing. weber, 1949 (1904), p. 70

Consequently, Weber posits, before embarking upon any cognitive activity the investigator – confronted with the chaos of phenomena – seeks to isolate from the general reality those phenomena which are of interest to him. So the first stage of the research process consists of the determination and delineation of the scope of study: There is no absolutely “objective” scientific analysis of culture – or put perhaps more narrowly but certainly not essentially differently for our purposes – of “social phenomena” independent of special and “one-­sided” viewpoints according to which – expressly or tacitly, consciously or unconsciously – they are selected, analyzed and organized for expository purposes. The reasons for this lie in the goal of understanding of all research in social science which seeks to transcend the purely formal treatment of the legal or conventional norms regulating social life. weber [1904] 1949, p. 72

Viewpoints represent the fundamental criterion for choosing the object of ­scientific investigation, analyzing and describing it in an organized fashion.

On Deformational Modeling

67

The one-sidedness of cognition affects not only the choice of the object of study but also the perspective from which it is examined. Also, it is no secret that Weber was chiefly interested in the economic dimension of cultural and social phenomena: Insofar as it traces a specific element of cultural life (the economic e­ lement in its cultural significance) through the most diverse cultural contexts, it is making an historical interpretation from a specific point of view, and offering a partial picture, a preliminary contribution to a more complete historical knowledge of culture. Social economic problems do not exist everywhere that an economic event plays a role as cause or ­effect – since problems arise only where the significance of those factors is problematical and can be precisely determined only through the application of the methods of social-economics. But despite this, the range of social-economics is almost overwhelming. weber [1904] 1949, p. 66

The second stage of investigation commences with the reconstruction of ­factors which are significant for the phenomenon under study. In other words, the investigator recognizes hypothetical factors which affect the constellations of events: When we require from the historian and social research worker as an elementary presupposition that they distinguish the important from the trivial (italics by the author – L.G.) and that he should have the necessary “point of view” for this distinction, we mean that they must understand how to relate the events of the real world consciously or unconsciously to universal “cultural values” and to select out those relationships which are significant for us. weber [1904] 1949, p. 81

The investigator gives prominence to relationships which he finds meaningful (important), and their structure is always subordinate to the goal and starting assumptions defined by the researcher (“value axioms”). At this stage of the research procedure, the investigator performs an indexation of significant and insignificant (accidental) characteristics which ­contribute to the cultural sense and meaning of the phenomenon under analysis. The stage can be referred to as the reconstruction of areas of hypothetical characteristics of the studied phenomenon. The investigator assumes them to be able to determine what types of factors can be relevant for the phenomenon concerned.

68

Godek

As the next step, the investigator performs a hierarchization of factors within the identified set of characteristics. He ascribes certain factors with the most significant status, recognizing them as primary for the analysis, and overlooks other factors as accidental (secondary). In this way, the investigator creates a simplified representation of the studied reality, which comprises only primary factors. In other words, factors are classified into those that are deemed important for the study and those that are considered accidental. A construct which makes it possible to capture significant characteristics of the studied phenomenon is Weber’s ideal type. The goal and course of the procedure contribute to its construction. The third (final) stage consists of the comparison of the empirical case with the ideal type and performing a classification on the basis of distance identified between the empirical case and the ideal type (i.e. the divergence between characteristics accommodated within the scope of a specific concept and characteristics of the empirical reality which never perfectly corresponds to the concept but only approximates it to a greater or lesser extent) (Weber [1904] 1949, p. 67). The above assumptions of the scientific procedure form the core of the approach referred to as the idealizing construction of scientific concepts. It is derived from the description of the investigation procedure given above. The process of constructing ideal types consists of two elements: 1. Creating a mental image of the studied phenomenon or object via accentuation (amplification, exaggeration) of its empirical characteristics. This gives rise to a mental construct (image) unifying in a non-contradictory way specific relations and processes (Weber [1904] 1949, p. 90). 2. Comparison of the ideal type with the empirical case and performing a classification on the basis of distance between that empirical case and the ideal type (i.e. the divergence between characteristics accommodated within the scope of a specific concept and characteristics of the empirical reality which never perfectly corresponds to the concept but only approximates it to a greater or lesser extent) (Weber [1904] 1949, p. 93). 3

Methodological Conditions Associated with Model Construction: Abstraction

The key aspect for understanding Weber’s idealization is the distinction between two forms of abstraction. However, even the most cursory reading of Weber’s writings reveals that the scholar uses the term “abstraction” in two disparate meanings. Abstraction can be construed as generalization ­(generalized observation) or idealization (deformation) (Weber [1904] 1949,

On Deformational Modeling

69

p. 100).5 Generalization opens up the way for a higher level of reflection thanks to adopting an empirically generalized perspective. The other interpretation of abstraction does not involve isolating a certain class of characteristics of objects of cognition. Instead, it involves deformation of these characteristics – or, more specifically, maximization of their intensities – which ultimately leads to the emergence of an emptily satisfied ideal type. Weber illustrates the two types of abstraction with two concept formulas: the former gives rise to real types, i.e. abstract generic concepts, and the other – ideal types, i.e. abstract typical concepts (Weber [1904] 1949, p. 100). Real types come about through the inductive generalization of characteristics of observable phenomena belonging to a particular class. Real types relate to reality in such a manner that they synthetically represent that what is shared by many concrete phenomena (Weber [1904] 1949, p. 93) – consequently also hypotheses accommodating concepts which denote real types are substantive statements. Essentially, the real type is the name of a selected object which is a distinctive example representative for objects of a specific kind. It can, therefore, be said that Weber’s abstract generic concept is, for all intents and purposes, an example of a classificatory notion which makes it possible to assign a specific characteristic to a concrete object using a predicate, and on that basis ascertain whether a given set of objects possesses – or does not possess – a particular characteristic (Weber [1904] 1949, p. 94). In formal terms, the real type has the following definition: the real type of object x is the actual object x’ which shares all its characteristics with object x at an adequate intensity. Weber highlights, however, that this type of concept is wholly useless in ­cultural studies because: The more comprehensive the validity – or scope – of a term, the more it leads us away from the richness of reality since in order to include the common elements of the largest possible number of phenomena, it must necessarily be as abstract as possible and hence devoid of content. In the cultural sciences, the knowledge of the universal or general is never ­valuable in itself. weber [1904] 1949, p. 80

This is why Weber argues for the construction of ideal typological notions which do not seek to explicitly reveal elements that are generic but instead

5 A different position is formulated by Leszek Nowak who asserts that abstraction and generalization are two disparate study procedures. See the discussion between Nowak (1989) and Renata Zielińska (1989).

70

Godek

focus on those that are specific to concrete cultural phenomena (Weber [1904] 1949, p. 101). The other type of abstraction makes it possible to create concepts through the application of the idealization method. Ideal types (i.e. typically ideal concepts) arising from this procedure do not denote actually existing objects but rather constitute constructs delineated by certain limiting characteristics that are never found in actuality: It is a conceptual construct which is neither historical reality nor even the “true” reality. It is even less fitted to serve as a schema under which a real situation or action is to be subsumed as one instance. It has the significance of a purely ideal limiting concept with which the real situation or action is compared and surveyed for the explication of certain of its significant components. Such concepts are constructs in terms of which we formulate relationships by the application of the category of objective possibility. By means of this category, the adequacy of our imagination, oriented and disciplined by reality, is judged. weber [1904] 1949, p. 93

Weber’s ideal type represents the extreme concept because the organization of characteristics encompassed by the notion includes all possible traits at their maximum levels of intensity. The extreme empty type emerges when there is not a single actual object embodying all dimensions at their maximum. Phrased differently, an ideal type is an unreal object or phenomenon which has been constructed in such a manner as to satisfy the following two conditions: they belong to the “extreme” outcome of systematization and conjunction performed on a set of characteristics of real objects or phenomena, and the “extreme” position is an emptily satisfied set. The difference between the two concept categories can be illustrated with ontological and epistemological criteria. On their basis, it is possible to establish that ideal types are ontologically and epistemologically open, whereas real types are ontologically and epistemologically closed: 1. Ontological criterion. An ideal type is ontologically open because it refers to the characteristics of phenomena, of which not all have been observed. In contrast, a real type is ontologically closed because it always relates to cases which have already occurred, with all exemplificatory concretizations having been observed (with observations forming the foundation for generalizations). 2. Epistemological criterion. An ideal type is epistemologically open because it concerns cases of which not all have been investigated. C ­ onversely, a real

On Deformational Modeling

71

type is epistemologically closed because it refers to cases which have been cognized in their entirety (Weber [1904] 1949, p. 90). On the basis of the above criteria differentiating two types of concepts one can assert that Weber rejects the realistic framework of concepts, replacing it with an instrumental approach: “The construction of abstract ideal-types recommends itself not as an end but as a means” (Weber [1904] 1949, p. 92). An ideal type is not capable of capturing the internally material structure of ­reality, while ideal types themselves (as opposed to real notions) do not represent study hypotheses (Weber [1904] 1949, p. 43). Ideal types are a specific type of heuresis; they mark out the direction for the formulation of hypotheses. Therefore, they do not describe but only provide a tool of description: “The ideal ­typical concept will help to develop our skill in interpretation in research: it is no ‘hypothesis’ but it offers guidance to the construction of hypotheses. It is not a description of reality but it aims to give unambiguous means of expression to such a description” (Weber [1904] 1949, p. 92). All p ­ resentations of the “essence” of social and cultural phenomena can be expressed with ideal types whose validity is quite relative and problematic always when an attempt is made to treat them as an enumerative presentation of empirical data r­epresenting a historical synthesis (both qualitatively and quantitatively). In contrast, high heuristic value of ideal types can always be claimed when they are seen as conceptual means for measuring and comparing the historical and cultural reality. It needs to be noted that the ideal image of phenomena and processes which arises from abstraction of this type is not normative in nature but entirely explanatory. A concept is a mental image which cannot be equated with historical reality or, even less so, with actual reality. Ideal types are not models of how things should be but rather templates which can be used for a theoretical description of phenomena. It is thus necessary to differentiate between ideal types in their logical sense and value judgments. Ideal types are completely indifferent towards value judgements and have nothing to do with any “excellence” aside from pure logic (Weber [1904] 1949, p. 91). 4

Ideal Type: Reconstruction

This way of constructing ideal concepts allows Weber to maintain flexibility in the procedure of assigning a type to an actual object.6 Ideal concepts 6 The reconstruction proposed here is epistemological and methodological in nature. Therefore, it does not address directly the actual and practical applications of the ideal type within Max Weber’s social theory. However, it should also be noted that Weber proposed

72

Godek

can be determined with respect to objects that possess defining characteristics to a lesser or greater extent – as long as they are sufficiently similar to the type. They also make it possible to compare real objects in terms of the degree to which they possess certain characteristics, and to rank them accordingly. Through the comparison with concrete phenomena, ideal types help to identify differences arising between them and, on that basis, allow to derive explanatory hypotheses and thus achieve a more thorough understanding of the historical phenomena under study (idiographism). In formal terms, an ideal type has the following definition: an ideal type of object x is a possible object x’ which shares at least one property with object x at a maximum intensity. Weber offers the following procedure for ideal type construction: An ideal type emerges: by the one-sided accentuation of one or more points of view (emphasis by the author L.G.) and by the synthesis of a great many diffuse, discrete, more or less present and occasionally absent concrete individual phenomena, which are arranged according to those one-sidedly emphasized viewpoints into a unified thought construct. weber [1904] 1949, p. 90

Characteristics occurring in the notion of ideal type may appear in their maximum intensities. What more, the notion of ideal type may include certain characteristics which do not appear in empirical case although it falls within a given ideal type. It is apt to illustrate an ideal type with a tangible example: a reconstruction of the concept of exchange economy which is introduced by Weber in the text Objectivity in Social Science to characterize the idealization procedure:7 an ­instrumental and historical interpretation of the ideal type which is to execute concrete research goals, namely enable the understanding (Verstehen) of social phenomena (Weber [1917] 1982, p. 532). This thought can be clearly seen in the introduction to Society and Economy, in which Weber formulates and defines the ideal type of goal-oriented rational action (Weber [1920] 1978, pp. 4–22). The ideal type of goal-oriented rational action serves as an instrument for the analysis and description of social reality. 7 Interpreters and investigators specializing in Weber’s methodological output have accustomed us to a specific manner of presentation of the ideal-typical method which directly realizes its pragmatic aspects. Namely, they always reach for the ideal-typical concept of goal-oriented rational action, and the ideal type of bureaucracy. This is understandable, taking into account the importance of these two categories for the concept of “understanding” (verstehende) sociology (understanding sociology is ‘rationalist’ in the scope in which it recognizes the unquestionable value of the concept of goal-oriented rational action for

On Deformational Modeling

73

We have in abstract economic theory an illustration of those synthetic constructs which have been designated as ideas of historical phenomena. It offers us an ideal picture of events on the commodity-market under conditions of a society organized on the principles of an exchange economy, free competition and rigorously rational conduct. This conceptual development brings together certain relationships and events of historical life into a complex, which is conceived as an internally consistent system. Substantively, this construct in itself is like a utopia which has been arrived at by the thinking accentuation of certain elements of reality. Its relationship to the empirical data consists solely in the fact that where market-conditioned relationships of the type referred to by the abstract construct are discovered or suspected to exist in reality to some extent, we can make the characteristic features of this relationship pragmatically clear and understandable by reference to an ideal-type. This procedure can be indispensable for heuristic as well as expository purposes. weber [1904] 1949, pp. 89–90

The scientific concept of exchange economy is always a kind of synthesis that is performed to achieve specific cognitive goals. The concept of exchange economy is abstracted from syntheses of traits that can be found in concrete historical periods. The investigator autonomously decides on the selection of characteristics (properties) ascribed to a studied phenomenon. In the process, the investigator focuses on the characteristics affecting the phenomenon’s assigned sense and cultural meaning. As the next step, the characteristics are extrapolated to their extreme form. The procedure gives rise to a conceptual construction, a mental construct which has no real-life embodiment. An ideal concept combines an aspect of experience which is available “here and now” into a whole, and involves a certain mental synthesis. This is not, however, a receptive arrangement but a constitutive condition for all content of experience. An ideal type is thus an effective tool for imposing order on the chaotic reality of human activities and, in this sense, introduces the cognizing entity into reality and, at the same time, makes it possible to structure that reality. the examination of social phenomena or, more precisely, actions taken by individuals) – on the one hand, and for the sociology of politics and theory of modernization, for which the ideal type of bureaucracy represents the criterion of measurement of social changes – on the other (Weber [1920] 1978, p. 6). In order to satisfy the methodological goal, however, I will refer to the concept of exchange economy because it forms the substantive background for the presentation of the idea of ideal type that can be found in the treatise Objectivity in Social Science.

74

Godek

To systematize manifestations of properties, the investigator delineates a certain temporal and spatial scope from which he draws knowledge about the characteristics of the phenomenon under study. Based on views prevailing in a given temporal and spatial scope, the investigator synthesizes characteristics in their maximum intensity in the form of a mental representation which is not an exemplificatory concretization, but “a purely ideal limiting concept to which reality is applied to accentuate certain important constituents of its empirical content to which it is compared” (Weber [1904] 1949, p. 92). The ideal type of exchange economy comes about through combining into a coherent structure episodic properties which occur in a diffuse manner in the economic activity across countries and ages. It should be noted that whenever one asks what in the empirical reality corresponds to the category of “­economy”, the answer encompasses an infinity of scattered human activities, legal standards, relationships, unique (one-off) or repeated relations, belief in the actually valid norms and rules of action. The scientific concept of exchange economy, regardless of how it is formulated, is always a kind of synthesis performed to accomplish specific cognitive goals. However, the concrete content which is historically assumed by “economy” in these syntheses, can be made evident only through reference to its ideal type. In the quote given above, Weber formulates a set of characteristics that can be used to describe the specific nature of economic activities. The economy demonstrates the following characteristics (Weber [1904] 1949, p. 89): A. High level of social organization. B. Domination of forms of goods exchange, commodity trade. C. Free competition. D. Economic entities are governed by rationality. The ideal type of exchange economy signifies a complex system whose ­content is constituted by a conjunction of properties A, B, C and D in their maximum intensity, and relations of assignment which make it possible to relate the ideal type to empirical cases. An empirical case falls under a given ideal type to a lesser or greater extent by possessing particular properties adopting appropriate values and situates them in a scale determined by relations of assignment. Characterizing a concrete case of “economy X” in the four-element assignment involves the specification of place which it holds in each dimension of a given property. This way of constructing concepts enables comparisons of historical phenomena on the basis of the degree to which they possess specific properties.

On Deformational Modeling

75

Matching ideal concepts to reality is inextricably linked to the need of identification, i.e. the development of an indicator which creates a basis for evaluating the difference in “behavior” between the ideal concept and the actual object. Matching an ideal type to reality is to be understood as verification or resolution of an identification task which is aimed at developing an ideal type that can serve as a starting point for a conceptual rationalization of a selected fragment of reality whose results would be a basis for drawing inferences about specific and unique characteristics of the phenomenon under study. The approach seems quite precise and cogent, however in practice the construction of ideal types is fraught with problems related to the assessment of accuracy (and adequacy) of ideal notions. The issue, however, cannot be unambiguously resolved because social sciences do not have a scale that would allow a straightforward comparison of differences between various degrees of intensity of characteristics. 5

Idealization and Deformational Modeling

In his outline of the broad area of applicability of idealization in social sciences, Leszek Nowak distinguishes two dominant paradigms: neo-Weberian, in which idealization is seen as a procedure of constructing scientific concepts; and neo-Hegelian, in which idealization requires separating significant and insignificant (accidental) aspects of the phenomenon under analysis (Nowak 1992, pp. 9–10). Modeling inspired by the Hegelian tradition, considerably closer to the idealization method applied within the Poznań School of Methodology paradigm, entails a construction of an abstract model of a studied phenomenon with a view to exposing factors which exert a significant corrective effect. Idealization in the Weberian spirit, on the other hand, requires an accentuation (exaggeration) of characteristics of the empirically cognized status of a study object in order to build an analytical conceptual construct encompassing all characteristics of the antecedent object at their maximum intensity. What is shared by both these concepts is the way of employing idealization which is concealed under the strategy of consistent deformational modeling. In Hegelian tradition, deformation involves incorporating into a reconstructional account of a given phenomenon such aspects of that phenomenon which are considered by the investigator as salient and relevant (i.e. the investigator disregards all secondary traits of the phenomenon under study). In contrast, Weber’s deformation includes an accentuation of characteristics of a studied phenomenon in order to account for that phenomenon in its entirety and

76

Godek

complexity. In both frameworks, scientific cognition does not entail a faithful reproduction of reality but rather its deformation in such a manner as to be able to identify its most essential relationships and correlations (Hegelian tradition developed in Poznań School of Methodology) and all potential characteristics of its empirical content (Weber). It can, therefore, be concluded that scientific cognition does not constitute a “presupposition-free” copy of “objective” facts (Weber [1904] 1949, p. 92). In this way idealization explicitly reveals itself as belonging to one of the groups of modeling procedures classified as deformational (Nowak 1990, p. 197). Such models comprise an intentional “deformation” of an object (empirical reality) into theoretical models in possibly diverse and, at the same time, contrafactual ways. Given the fact that every object is equipped with a ­certain group of characteristics, and each of these characteristics assumes a ­certain level of intensity, it is possible to distinguish two basic deformational procedures: 1. Quantitative (weak) deformation: when an object is assigned a selected properties in a degree which differs from that which is exhibited by that object empirically. Procedures of this type include positive potentialization (when a characteristic attributed to an object is greater than actual), and negative potentialization (vice versa, i.e. characteristics attributed to an object are established at their minimal levels). Based on L. Nowak, maximization of a property assigned to a selected object is referred to as mythologization (the extreme case of positive potentialization), while its minimalization is termed ideation (the extreme case of negative potentialization); 2. Qualitative (strong) deformation: when an object is contrafactually ­endowed with a certain property it actually does not have or, vice versa, it is deprived of a certain property which it in fact possesses (in this case, no ­account is taken of the intensity of the characteristic concerned). The ­former procedure is called transcendentalization, and the latter – ­reduction. The extreme case of reducts of an empirical object is the empty object, devoid of any characteristics. By analogy, the extreme transcendental is an object that is complete, i.e. devoid of any defects (Nowak 1990, pp. 198–200). Outlined above typology of deformational procedures allows for a preliminary analysis of Max Weber’s ideal type method. Weber asserts that ideal type is “like a utopia which has been arrived at by the thinking accentuation (emphasis by the author – L.G.) of certain elements of reality” (Weber [1904] 1949, p. 89). Let us repeat that construction of ideal type emerges: “by the one-sided

On Deformational Modeling

77

accentuation of one or more points of view [emphasis by the author – L.G.] (Weber [1904] 1949, p. 90)”. Weberian “accentuation” may be interpreted as Nowakian positive potentialization. Construction of ideal type includes also “the synthesis of a great many diffuse, discrete, more or less present and occasionally absent concrete individual phenomena, which are arranged according to those one-sidedly emphasized viewpoints into a unified thought construct” (Weber [1904] 1949, p. 90). The method of ideal type formulated by Weber is a positive potentialization (i.e. its absolutization) leading to a construction of a complete object of a given kind. For example, the ideal type of market economy – in comparison with empirical cases subsuming under this conceptual category – possesses all constitutive properties with a maximal intensity. It is worth noting that such a complete object, just like an empty object, is emptily satisfied in the real world.8 Weber’s deformational modeling shall then be defined a strategy of conscious and deliberate distortion of an object of empirical reality in possibly varied and hence contrafactual ways. The aim of the deformation method essentially comes down to accounting for a concept by giving prominence to significant characteristics of the empirical content of investigated socioeconomic phenomena – at the expense of their actual exemplification. Importantly, it is assumed that a model is always an idealized (deformed) means of representing a selected real-life phenomenon or object, oriented towards the accomplishment of specific cognitive goals while taking into account all methodological conditions related to the process of its construction. Consequently, it turns out that there is no isomorphic equivalence between perception and scientific representation, and cognition takes the form of transformation which enables the highlighting of the most salient features of a phenomenon under study. The legitimacy or accuracy of a model in idealization theory is not measured by how faithfully it reflects reality but by how aptly it deforms the investigated phenomenon. However, constructing a model is associated with 8 A slightly different interpretation of Weberian ideal types as a method of deformation is ­presented by I. Nowakowa (2007). She points out that the construction of Weber’s ideal types is a specific kind of potentialization – in more detailed terms, a type of negative potentialization (weak deformation). Weber repeatedly stated that the ideal type represents a ­certain utopian idea which combines in an internally coherent ideal image certain properties that are “one-sidedly accentuated” in their manifestations. The properties reveal themselves in the expression of thought (Gedankenausdruck) (Weber [1904] 1949, pp. 90–91). It seems, therefore, that the maximization rather than minimization of properties is involved. The creation of such an ideal type enables the researcher to analyze reality on the basis of the criterion of distance separating real phenomena from the complete ideal type (also when the latter is emptily satisfied in the world) (Weber [1920] 1978, p. 6).

78

Godek

the need to reconcile logical coherence and the “importance” of incorporating into a model a maximum number of characteristics of a real phenomenon or object at their maximum level of intensity [Weber, 1949 (1904), p. 90]. 6

Concluding Remarks: Weber’s Epistemological Phenomenalism

Both idealization interpreted as discrimination between significant and trivial aspects in a studied phenomenon, and idealization understood as the construction of ideal notions, are not alternative to deformational modeling but clearly correspond to its dynamic and static perspectives.9 Idealization acquires the status of a generally valid principle in the scientific cognition of cultural reality. The concept realized within the theoretical framework of Poznań School of Methodology and proposed in Weber’s writings cannot, therefore, be viewed either as alternative accounts of the modeling method or as alternative accounts of idealization. Quite the opposite: deformation which is realized in both these forms is rooted in the same belief about the nature of human cognition and instruments used for the analysis and description of the investigated reality. What Nowak presents and discusses as “idealization” is the construction of an abstract model of the studied phenomenon. For a phenomenon under review, the investigator prepares a hierarchy of primary (significant) and secondary (less significant) factors. The researcher thus indexes those characteristics which he regards as fundamental and, in this fashion, identifies relationships existing between the studied phenomenon and primary factors impinging on that phenomenon. What Weber presents and describes as “idealization” is not limited to the deformational modeling of notions (an abstract conceptual construct) but also incorporates a research procedure based on the hierarchization of factors by their significance. Even though in operative terms both positions make use of the method of idealization (at the methodological level), there is a marked difference between them at the level of higher-order (philosophical) assumptions. While Nowak’s framework is, by his own admission, essentialistic (both along the ontological and epistemological planes) (Nowak 1977). Weber’s views have a phenomenalistic character. Nowak prioritizes factors on the basis of their importance for the 9 Weber’s conception is static in nature – the greater number of empirical cases considered, the more common the coexistence of characteristics distinguished in an ideal type; the ­structure is static because it only accommodates the coexistence of certain properties of empirical reality, however no mutual interactions are accounted for.

On Deformational Modeling

79

objective structure (“substantive interconnection of objects”), whereas Weber assumes that differentiations of importance are not objective but belong to the nature of human discursive cognition (“conceptual interconnection of problems”). Social s­ ciences, which are focused on reality, offer an opportunity to understand the studied reality in terms of a constant yield of new results and a never-ending task presented to our cognitive and epistemological faculties and – crucially – practical activities. Constructing the ideal type on the basis of “one-sided accentuation” and differentiation of attributes of sociocultural phenomena, Weber creates an importance-based structure of studied phenomena (Nowak proposes an essential structure) which is variable over time and depends on the goal and baseline assumptions adopted by the investigator. Therefore, the importance-based structure is not objective but intersubjective in nature. The ideal type is hence supposed to serve the function of a new method that will enable the emergence of new science. I would like to finish my consideration quoting Weber: “It is not the ‘actual’ interconnections of ‘things’ but the conceptual interconnections of problems which define the scope of the various sciences. A new ‘science’ emerges where new problems are pursued by new methods and truths are thereby discovered which open up significant new points of view” (Weber [1904] 1949, p. 68). References Brzechczyn, K. (2012). Varieties of Idealization and Strategies of Modification of ­Social Theory. The Case of the Totalitarian Syndrome. Człowiek i Społeczeństwo 34, 235–247. Lachmann, L.M. (1971). The Legacy of Max Weber. Berkeley: The Glendessary Press. Nowak, L. (1970). O zasadzie abstrakcji i stopniowej konkretyzacji [On the Principle of Abstraction and Gradual Concretization]. In: J. Topolski (ed.) Założenia metodologiczne Kapitału Marksa, pp. 123–218. Warszawa: KIW. Nowak, L. (1971). U podstaw marksowskiej metodologii nauk [Foundations of the M ­ arxian Methodology of Sciences]. Warszawa: PWN. Nowak, L. (1973). Typy idealne w koncepcji Maxa Webera [Ideal Types in the Concept of Max Weber], pp. 350–361. In: J. Kmita (ed.), Elementy marksistowskiej metodologii humanistyki. Poznań: Wydawnictwo Poznańskie. Nowak, L. (1977). Wstęp do idelizacyjnej teorii nauki [An Introduction to the Idealizational Theory of Science]. Warszawa: PWN. Nowak, L. (1978) Weber’s Ideal Types and Marx’s Abstraction. In: R. Bubner (ed.), Marx’ Methodologie (Neue Hefte für Philosophie, vol. 13), pp. 81–91. Goettingen: Ruprecht.

80

Godek

Nowak, L. (1980). The Structure of Idealization: Towards a Systematic Interpretation of the Marxian Idea of Science, Dordrecht: Reidel. Nowak, L. (1990). Abstracts Are Not Our Constructs. The Mental Construct Are Abstracts. In: J. Brzeziński, F. Coniglione, T.A.F. Kuipers and L. Nowak (eds.), Idealization I: General Problems. Poznań Studies in the Philosophy of the Sciences and the Humanities 16, pp. 193–206. Amsterdam-Atlanta: Rodopi. Nowak, L. (1992). The Idealizational Approach to Science: a Survey. In: J. Brzeziński and L. Nowak (eds.), Idealization III: Approximation and Truth. Poznań Studies in the Philosophy of the Sciences and the Humanities 25, pp. 9–63. Amsterdam-Atlanta: Rodopi. Nowakowa I. (2007). The Method of Ideal Type’s versus the Method of Idealization. In: J. Brzeziński, A. Klawiter, Th.A.F. Kupers, K. Łastowski, K. Paprzycka, P. Przybysz (eds.). The Courage of Doing Philosophy Essays Presented to Leszek Nowak, pp. 159– 166. Amsterdam: Rodopi. Rogers, R.E. (1969). Max Weber’s Ideal Type Theory. New York: Philosophical Library. Weber, M. ([1904] 1949). The Methodology of the Social Sciences, trans. By E.A. Schills, H.A. Finch. Glencoe, Illinois: The Free Press. Weber, M. ([1920] 1978). Economy and Society. An Outline of Interpretative Sociology, edited by G. Roth, C. Wittich. Berkeley, Los Angeles, London: Univeristy of California Press. Zielińska, R. (1989). Contribution to the Problem of Abstraction. In: J. Brzeziński, F. ­Coniglione, T.A.F. Kuipers and L. Nowak (eds.), Idealization II: Forms and Applications. Poznań Studies in the Philosophy of the Sciences and the Humanities 17, pp. 9–22. Amsterdam: Rodopi.

chapter 4

On Reduction in the Idealizational Theory of Science: A Case Study of the Relationship between the Concept of Rational Act and the Concept of a Habitual-Rational Action Mieszko Ciesielski Abstract The article discusses the issue of a reduction relationship between theories of action within the conceptual framework of the idealizational theory of science. The analysis of the relation between the concept of a rational act and the concept of a habitualrational action leads to the conclusion that not all necessary conditions for the occurrence of a reduction relationship are fulfilled in the discussed case. Although the two theories are conceptually compatible, they differ with respect to their hierarchization of essential factors, so there is essentiality disagreement between them. In the light of the idealizational theory of science, for a reduction relationship to occur between two theories, essentiality agreement between them is required. That conclusion becomes, in turn, the starting point for a critical analysis of essentiality agreement between theories as a necessary condition for a reduction in the field of humanities. The article ends with a postulate of weakening that condition with respect to theories formulated in the social sciences and the humanities, due to the methodological specificity of this domain of knowledge, for example, the qualitative and less precise language and the fundamental difficulty of conducting repeated experiments with particular factors, which influence the studied phenomenon, being effectively isolated.

1 Introduction The problem of the relationship of reduction (correspondence) between ­theories can be discussed in its three aspects. The explanatory aspect involves issues related to comparing the explanatory force of particular theories. The formal aspect concerns the inner dimension of a theory, for example, the ­similarities and differences of the concepts and structures of the compared theories, as well as the paraphrasability of the statements of those theories. The interdisciplinary aspect of the reduction relationship becomes manifest when one looks for connections among theories developed in various scientific © koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_006

82

Ciesielski

fields and for the possibilities (or the impossibility) of unifying p ­ rocesses within s­ cientific knowledge. The aim of this article is to present the formal aspect of the reduction ­relationship between theories of action, that is, the aspect concerning conceptual similarity and statement paraphrasability.1 I will use Katarzyna Paprzycka’s idealizational concept of reduction to define the relationship between the concept of a rational act explicated within the framework of Poznań School of Methodology2 and the concept of a habitual-rational action which results from critical reflection on the former concept.3 Both concepts are formulated in the same methodological language of the idealizational theory of science, which will make it possible to compare the terms and statements of those ­concepts in a more satisfactory manner than would be the case if two different methodological frameworks were used.4 We will, thus, apply the idealizational concept of reduction to humanities which, in principle, do not adhere by the rules of language precision, unambiguity of concepts, or experimental decidability in the way mathematized natural sciences do and, therefore, which are a frequent subject of analyses of inter-theoretical relationships.5 1 When I write about the relationship of reduction between the analyzed theories I mean a philosophical and not an ontological relationship. Therefore, we will be interested in a comparison of the theories and the possible explication of the concepts/statements of one of them with the use of the concepts/statements of the other one, and not in reflection on the characteristics of the objects described by those theories. The interrelationships between reduction from the theoretical and from the ontological point of view are noteworthy but that issue extends beyond the scope of this article. 2 When discussing to the concept of a rational act I will refer to Wojciech Patryas’s reflections (1979). Also see: Kmita (1971) and Nowak (1991; 1993; 2000). 3 The theory was presented in: Ciesielski (2012); a summary in English: Ciesielski (2011). 4 Methodological concepts alternative to idealization and concretization apply other, different criteria of the reduction relationship, which definitely has an influence on conclusions concerning particular cases, including the case, discussed here, of theories of action. The aim and scope of this article does not allow me to elaborate on the problem of various approaches to reduction and the issues and methodological disputes related to it, which is why I will limit my reflections to the concept of reduction in one, selected methodology – that of idealization and concretization. Reduction has been described in greater detail in the works of, among other, authors: Nagel (1961), Silberstein (2002), Hohwy, Kallestrup (2008), ­Niiniluoto (2013), and Riel (2014). 5 The issue of correspondence in sciences has been repeatedly discussed within the framework of the idealizational theory of science, for example, in physics: Nowakowa (1972; 1975b; 1982); in biology: Łastowski (1987); and in economy: Nowak (1974). Reflections about the inter-theoretical relationships were less frequent in humanities. One theoretician on that topic was Leszek Nowak (2000) who defined the relationships of the inter-human model of a human being with respect to the evangelical, liberal, and Marxist models.

On Reduction in the Idealizational Theory of Science

2

83

Idealizational Approach to the Concept of Reduction

The issue of inter-theoretical relationships has been raised in the idealizational theory of science (abbreviated as its) many times (see: Nowak 1971; 1977; 1980; Nowakowa 1972; 1975a; 1975b; 1991; 1994; Krajewski 1977; Egiert 2000). Already in 1970s systematic definitions of correspondence and dialectic refutation were made. The newest approach to the topic is Katarzyna Paprzycka’s (2005)6 “general concept of reduction”. In the author’s view, earlier approaches were fragmentary and did not exhaustively discuss all aspects of inter-theoretical relationships. Paprzycka’s approach was an attempt at filling in the blanks. The multi-layered essentiality structure of reality and the scientific laws in the form of a concretization chain, assumed in its, have a significant influence on the understanding of reduction. Paprzycka forms the following definition of reduction: (R) Theory t is reduced to theory T within the scope of D (…) always and only when: (a) D ⸦ Z(t) and D ⸦ Z(T ); (b) the concept of examined factor h in theory t can be explicated with the use of the concept of examined factor h’ in theory T; (c) with the use of a certain essential factor for factor h’ in theory T we can explicate a concept of each essential factor for factor h in theory t; (d) the essential structure of essential factors for factor h’ (in theory T) is an (explicatory) extension of the essential structure of essential factors for factor h (in theory t); (e) the (explicatory) counterpart of the idealizational law of theory t in theory T and the explicatory counterparts of all the concretizations of that law in theory T can be derived (by way of deduction or approximation, asymptotically, etc.) from theory T – given some additional assumptions (Paprzycka 2005, p. 128). In its the reduction relationship between theories can be analyzed in three dimensions: of explication, essentiality, and derivation (Paprzycka 2005, p. 121).7 For each of those dimensions Paprzycka formulates a necessary ­condition of the occurrence of reduction. In the explicatory dimension the necessary condition of the occurrence of the reduction relationship is the conceptual compatibility of both theories 6 Also see earlier works of that author: Paprzycka (1989; 1990). 7 When discussing Paprzycka’s approach, Witold Strawiński names the dimensions as follows: semantic, ontological, and logical. See more in: Strawiński (2008), p. 72.

84

Ciesielski

(points b and c of the definition). Such compatibility takes the form of either the identicality of the concepts of the reducing and reduced theories – in such a case, we speak of a homogenous reduction – or of the explication of the concepts of the reduced theory with appropriate concepts of the reducing theory8 – in such a case, we speak of a heterogenous reduction. Moreover, the condition includes the concepts of the examined factor and of all essential factors. In the essentiality dimension of reduction the reduced and reducing theories are compared and analyzed with respect to their essential structures. The necessary condition of the occurrence of the reduction relationship is the essentiality agreement of both theories which takes place when the essential structure of the reducing theory constitutes an extension of the essential structure of the reduced theory9 (see point d of the definition), that is, when: (a) all essential factors of the reduced theory are taken into account in the reducing theory, (b) the equally essential factors in the essential structure of the reduced theory are also equally essential in the reducing theory, and (c) for each two factors their relative order of essentiality in the essential structure of the reduced theory is replicated in the reducing theory (Paprzycka 2005, p. 161). When the condition is not fulfilled, we speak of essentiality disagreement. The essentiality disagreement can be fundamental – when it concerns the main factors of the theories in question – or non-fundamental (accidental) – when it only concerns the secondary factors.10 Using the concept of the extension of the essential structure we can distinguish the following three variants of reduction. Static reduction happens when the sets of essential factors of the reducing and reduced theories are the same. Corresponding reduction happens when at least one secondary factor is added to the reduction theory. There are two sub-variants of corresponding reduction: the supplementary one – when an additional secondary factor in the 8

9

10

In her approach to reduction Paprzycka uses the explicatory understanding of bridge rules: “bridge rules illustrate the explication of the concepts of the reduced theory with the use of the concepts of the reducing theory” (Paprzycka 2005, p. 116). Explication is, generally speaking, is: “the procedure by means of which it is possible to find the meaning of a term from a given language in another language” (Paprzycka 2005, p. 110). An extension of the essential structure takes place when “the reducing theory complements the existing essential structure without disturbing it either by removing some ­factors or by violating the relative order of essentiality of the existing factors” (Paprzycka 2005, p. 125). In the light of the concept of reduction, discussed here, because of the accidental essentiality disagreement (pertaining to secondary factors) only partial reduction is possible – within the scope of the original idealizational law and the factual statement (Paprzycka 2005, pp. 167–168).

On Reduction in the Idealizational Theory of Science

85

reducing theory is equally essential as some other, already recognized factor in the reduced theory – and the corrective one – when the additional secondary factor differs, in terms of essentiality, from the already recognized factors, so the essential structure of the reducing theory is extended by an ­additional essentiality level at which a new secondary factor is placed. ­Refutative reduction takes place when the reducing theory takes into account an additional main factor which has not been recognized in the reduced theory. In that case, as well – in analogy to corresponding reduction – two sub-variants can be ­distinguished: the supplementary one and the corrective one. The derivative dimension of reduction pertains to the statements of the reduced and reducing theories. A necessary condition of the occurrence of the reduction relationship is the possibility of deriving statements of the reduced theory (the most idealized law, particular concretizations and the factual statement) from the statements of the reducing theory (point e of the definition). With respect to that condition we distinguish a complete and partial reduction. The first type of reduction occurrs when the counterparts of all the statements of the reduced theory can be derived from the reducing theory. The second type of reduction occurrs when only some counterparts of all the statements of the reduced theory can be derived from the reducing theory and when among those counterparts there is a counterpart of the idealizational law.11 In Paprzycka’s approach, in order to be able to talk about the reduction relationship between theories, three conditions have to be fulfilled: conceptual compatibility (the explicatory dimension of reduction), essentiality agreement (the essentiality dimension of a reduction), and the possibility of deriving the statements of the reduced theory from the statements of the reducing theory (the derivative dimension of a reduction). Further in the article I will try to find out if Paprzycka’s necessary conditions of a reduction are fulfilled in the case of a relationship between two selected theories of action. 11

Apart from the types discussed extensively in this article Paprzycka also distinguishes global and local reductions. The first one pertains to situations in which the reduced statements can be derived within the whole scope of the subject to which the reduced theory refers. The second one does not fulfill that condition, that is, statements from a certain part of the reduced theory cannot be derived from the reducing theory. Local reduction can be open or closed. It is open when the reduced theory has a non-empty subset of the subject, which cannot be reduced to any other theory. Closed reduction ­pertains to situations in which statements from the whole scope of the subject are ­reduced. Also, branching reductions are mentioned, in the case of which the reduced theory is reduced to more than one reducing theory. Otherwise, we speak of a non-branching reduction.

86 3

Ciesielski

The Main Theses of the Concepts of a Rational Act and of a Habitual-Rational Action

Both the concept of a rational act (further abbreviated to cra) and the ­concept of a habitual-rational act (further abbreviated to ch-ra) describe practical human actions. Both define the scope of the subject applicability as a universum of actions of ‘normal’ subjects capable of intentional activity. 3.1 The Concept of a Rational Act The concept of a rational action has the form of model statements in which numerous assumptions idealizing an “action situation” are made. Below I will refer to the concept of rationality in Wojciech Patryas’s understanding (Patryas 1979). The author has systematically prepared an extensive list of idealizing assumptions on which the principle of rationality is based12 and concretized it with respect to all his assumptions. As Patryas uses the term “a rational act”, I will further name his concept the concept of a rational act. Patryas uses the concept of a “human act” (abbreviated to at) which is understood with respect to its outcomes (Patryas 1979, p. 5). A particular act, then, which is the object of an explanation, is defined by its actual outcome. at is not, as we can see, related to the behavioral (external) aspect of behavior or to the intentional aspect which is related to the subject’s intention to perform an act. The knowledge of the subject performing an action is an important aspect of each concept of the rationality of action; it is the knowledge of the subject performing an action that makes it possible to explain why particular actions are performed. Patryas characterizes the knowledge as that part of a subject’s convictions which concerns his or her acts. The knowledge includes the ability to use the means of performing an act (“the objective, non-intellectual and nonvolitional characteristics” (Patryas 1979, p. 8)) and the characteristic of a set of acts which can be performed and which are considered by the subject (Patryas 1979, pp. 10–11), that is, the convictions about whether the actions in the set of alternatives complement one another, whether the subject can ­perform all the alternative acts and only those acts from the set, whether particular acts from the set are mutually exclusive, and whether the only outcomes of the acts are the intended ones. 12

The non-idealizational nature of the assumption of rationality was also indicated by o­ ther authors who enumerated the most important idealizing assumptions. See, for ­example, Kmita (1972) and Nowak (1974).

On Reduction in the Idealizational Theory of Science

87

An analysis in the perspective of practical rationality is possible thanks to having a criterion for ordering particular acts with respect to the degree to which they fulfill that criterion. In Patryas’s approach the criterion is the ­function of the usefulness of individual outcomes, widely used by the researchers of the theory of practical rationality (Kmita 1971; Szaniawski [1971] 1994; Kleszcz 1998). The usefulness is defined with respect to the degree of the realization of the generally understood value/good of the subject performing an act. The usefulness criterion is connected with the relation of preference: the preference of one act to another is due to its greater individual usefulness for the subject performing an action (Patryas 1979, p. 11).13 An essential element of a theory formulated in the language of idealization and concretization is the hierarchy of factors influencing the examined factor. Within the framework of cra it is posited that there are three main factors influencing human acts: (1) the knowledge of the subject performing an act (Kat);14 (2) the relation of preferences on the basis of the usefulness of the outcome of an act (Pu); (3) the means of performing an act (Mat). Secondary factors are connected with the preference and knowledge of the subject performing an act. Patryas characterizes them as follows (Patryas 1979, pp. 9–11): (1) other people’s knowledge at the disposal of the subject performing an act (kop); (2) the subject performing an act taking into account other people’s or groups’ preferences (gp); (3) knowledge inadequacy (ki); (4) the variability of preferences and knowledge in time (vpk); (5) a lack of knowledge about whether the acts from the set of possible ­alternatives complement one another (lkc); 13

14

Here is a sample formula for the preference of one act to another, due to the individual usefulness of their outcomes: UO,X (ATj) > UO, X (ATk), that is: individual usefulness (U) for X of outcome (O) of act ATj is greater than the individual usefulness of the outcome of act ATk. For the purpose of this article I will introduce letter symbols for the essential factors in cra and ch-ra.

88

Ciesielski

(6) a lack of knowledge about whether there are other acts which can be performed by the subject, outside of the set of the considered alternative acts (lko); (7) a lack of knowledge about whether the acts from the set of possible ­alternatives are mutually exclusive (lke); (8) a lack of knowledge about how some acts are to be performed (lkat); (9) the knowledge about additional outcomes apart from the outcomes of the performed act (kao); (10) non-linear order of the outcomes of acts, on the basis of the individual usefulness for the subject performing an act (nlu). When he presents the essential structure of a rational act, Patryas makes an important remark about the hierarchy of secondary factors: There is also the problem of stratification principles according to which the assumptions are hierarchized (“assumptions” are understood as “­secondary factors canceled by the assumptions” and the hierarchization of the influence on the examined factor concerns factors and not assumptions [M.C.’s note]) 1–10. There are hardly any unambiguous stratification principles here. Rather, when I put the assumptions in order it was for pragmatic reasons. I assumed that the first assumptions to cancel would be those which lead to “simpler modifications” of the consequent of the principle of rationality. Nevertheless, it seems that assumptions 3 to 10 are more significant than assumptions 1 and 2 as the subject’s knowledge and preferences appear to be more important for the decision about what act is to be performed than other people’s knowledge or preferences, even if they are taken into account by the subject. From among assumptions 3 to 10 the most significant ones are, apparently, assumptions 5 to 10, as in order to account for the falsity of knowledge or its changing nature (and the changing nature of preferences) one must, first, somehow define that knowledge, which is determined by Patryas’s assumptions. patryas 1979, pp. 12–13

There was a reason why Patryas stressed the problem of the hierarchy of ­secondary factors, related to the determination of the degree of their influence on the examined factor. The methodology of idealization and concretization consists in forming statements in the form of a system of laws in which it is very important in what order of the factors influencing the examined factor are put with respect to the degree of that influence. The first and most idealized law is

On Reduction in the Idealizational Theory of Science

89

the regularity which shows the dependency between the examined factor and the main factor. Further statements are concretizations of the regularity with respect to the subsequently introduced factors considered to be secondary ones. The order of the essentiality of those factors, then, determines the order of the statements in the sequence of laws in a given theory: from regularities, through concretizations, to the factual statement. According to Patryas’s remarks, then, the essential structure of a human act (Sat) has the form: Sat : (0) Kat, Pu, Mat (1) Kat, Pu, Mat, nlu, kao, lkat, lke, lko, lkc (2) Kat, Pu, Mat, nlu, kao, lkat, lke, lko, lkc, vpk, ki (3) Kat, Pu, Mat, nlu, kao, lkat, lke, lko, lkc, vpk, ki, gp, kop In cra there is a regularity which says that in the antecedent there are realistic assumptions as well as ten idealizing assumptions canceling the influence of factors considered to be secondary ones.15 In the consequent there is a statement determining the act to be peformed by the subject. The regularity has the following form: (PR) IF X is in a situation in which he or she can do each of acts AT1,…, ATn and (A10) and (A9) and… and (A1), then: ATx = Max [U0, x(AT1),…, U0, x(ATn)].16 15

Particular secondary factors are canceled by the following idealizing assumptions: (A1) X in time period T only uses his or her own knowledge. (A2) X only takes into account his or her own preferences. (A3) X’s knowledge is true. (A4) X’s knowledge and preferences are constant. (A5) X knows that acts AT1,…, ATn complement one another. (A6) X knows that he or she can only do each of acts AT1,…, ATn. (A7) X knows that acts AT1,…, ATn exclude one another. (A8) X knows how to perform each of the doable acts. (A9) X knows that the only outcomes of the doable acts are their outcomes. (A10) Individual usefulnesses of the outcomes of the acts X can do are ordered in a linear way. Moreover, Patryas forms one realistic assumption: (Real. A.) X is in a situation in which he or she can do each of acts AT1,…, ATn. 16 Symbols: ATx – X’s act; Max – the maximal individual usefulness of the outcomes of an act; [U0, x(AT1),…, U0, x(ATn)] – a set of acts with a particular value of the individual ­usefulness of their outcomes for X.

90

Ciesielski

In other words, X will perform that act at the outcome of which will have the greatest individual usefulness (will be the most preferable) for him or her (Patryas 1979, pp. 13–14). Canceling the idealizing assumptions, from A10 to A1, leads to more modifications of the original formula. As the formal apparatus needed to present all concretizations of the regurality described above would be too complex, I will only present two concretizations. The first concretization of the cra regularity is constructed with respect to secondary factor nlu (non-linear order of the outcomes of acts, on the basis of the individual usefulness for the subject performing an act) (Patryas 1979, pp. 22–23). The A10 assumption which says that individual usefulnesses of the outcomes of acts are ordered in a linear manner is canceled and replaced with the A10’ assumption which assumes a partial order of the usefulnesses of outcomes. Therefore, the concretized principle of rationality is as follows: (PR1) If X is in a situation in which he or she can do each of acts AT1,…, ATn and (A10’) and (A9) and… and (A1), then: ATx = Max1 [U0, x(AT1),…, U0, x(ATn)] ∨… ∨ ATx = Maxj [U0, x(AT1),…, U0, x(ATn)]. Thus, the above concretization of the principle of rationality only says that X will do one of the acts in the set of acts with the highest, equal individual usefulnesses of their outcomes. For example, for j = 3 X will choose from among three acts: Max1 or Max2 or Max3. The conretization of cra, which makes allowances for the context of the social surrounding, is a statement which takes into account secondary factor gp (the subject performing an act taking into account other people’s or groups’ preferences). The idealizing assumption (A2): X only takes into account his or her own preferences is replaced with assumption/statement (A2’): X takes into account his or her own as well as Y1’s,…, Yi’s preferences. The principle of rationality concretized with respect to that factor has the form:17 (PR9’) IF X is in a situation where he or she can peform each of acts AT1,…, ATn and (A10) and (A9) and… and (A2’) and (A1), then: ATx = Max [U0, (x, y 1,…, yi)(AT1),…, U0, (x, y1,…, yi)(ATn)].

17

Here I present the concretization of the original principle of rationality, solely with respect to the secondary factor related to other people’s preferences. Patryas’s presentation of that concretization also takes into account the remaining secondary factors (Patryas 1979, pp. 42–43).

On Reduction in the Idealizational Theory of Science

91

Thus, the subject will perform the act the outcome of which is characterized by the greatest public usefulness. The public usefulness of an act is defined by the value of one of many possible functions which constitutes a certain combination of individual usefulnesses for particular people. 3.2 The Concept of a Habitual-Rational Action The concept of a habitual-rational action was created by rejecting the ­assumption which lay at the core of various approaches to a rational action and, in a way, restricted them (Hempel 1961/62; Watkins 1970; Kmita 1971; ­Szaniawski [1971] 1994; Kleszcz 1998; Nowak 2000). That assumption was called the ­principle of constant reflectivity (Ciesielski 2012, p. 66): (PCR): before undertaking any action an individual always reflects on alternative actions in order to determine which alternative will be the most useful with respect to the intended result. Thanks to the rejection of the assumption of a subject’s constant reflectivity, within the framework of ch-ra it was possible to conceptualize a model in which non-reflective actions, habitual and imitative, are included alongside the reflective ones. In the ch-ra language the original terms were defined which made it possible to conceptualize statements. “Human action” (abbreviated to an) is the realization of a particular state of things for the purpose of meeting a need, with the use of knowledge and the means of performing an action (Ciesielski 2012, p. 88). Human action is understood with respect to the subject’s intentions, that is, a given action is defined in terms of what the subject performing the action intended to do. The outcomes of an action may overlap with the intentions of the subject performing the action. When they do not, the action can be considered to be unsuccessful. The concept of the ‘need’ of the subject performing an action is defined as an individual’s characteristic consisting in the realization of a lack, an insufficiency (Ciesielski 2012, p. 84). There are three realms in which needs are met in the theory: the realm of want in which, although a state of things has been realized in which a given need is realized, an individual is still conscious of a lack with regard to that need; the optimal realm, in which a given state of things realizes a given need in such a way that the individual is not conscious of a lack with respect to that need; and the realm of abundance, in which the realized state of things realizes a given need at too high a degree, in such a manner that the individual is conscious of a certain excess with regard to that need. A subject’s “knowledge” is defined by pointing to its three realms (Ciesielski 2012, p. 88): objective ­(concerning the realized states of things), subjective (concerning the needs of the subject performing an action), and social (concerning other people’s actions).

92

Ciesielski

“The means of performing an action” are defined in two ways ­(Ciesielski 2012, pp. 88–91), as “tools” (states of things used in action) and as “preparatory actions” (the necessary actions which must be done by the subject in order to later perform the proper action). The criterion for the evaluation of particular actions is their “effectiveness”, that is the relationship between the degree to which a given need is met and the set of preparatory actions for a given action (Ciesielski 2012, p. 91). The fewer preparatory actions and/or the higher the degree to which a need is met, the more effective an action. Four main factors have been indicated within the framework of ch-ra ­(Ciesielski 2012, p. 101): (1) the knowledge of the subject performing an action (Kan); (2) the needs of the subject performing an action (Ns);18 (3) the relation of preferences described on the states of things with respect to the degree to which needs are met (Pn); (4) the means of performing an action (Man). The secondary factors are related to the characteristics of knowledge, needs, the preference relationship, and the means of performing an action (Ciesielski 2012, pp. 99–101): (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 18

the variability of the usefulness of the used tools (vuu); other individuals’ influence on one’s actions (oii); the size of the set of one’s needs (ssn); the type of the need met (tn); how high the costs of actions are (hc); the non-linear order of the states of things with respect to the degree to which a need is met (nls); knowledge concerning the unintended results of performed actions (kur); a lack of knowledge about how some actions are to be performed (lkan); the variability of the degree to which needs are met (vdn); knowledge inadequacy (ki). When enumerating the main factors the author of ch-ra states, in one place, that there are three such factors. He presents the needs of the subject performing an action together with the preference relationship described on the states of things, as one main factor (Ciesielski 2012, p. 101). However, the author’s further analyses show that the subject’s needs should rather be considered to be a separate main factor, for example, because the particular idealizing assumptions indicated by the author pertain, separately, to the subject’s needs and preference relationship.

On Reduction in the Idealizational Theory of Science

93

The author of ch-ra also notices the problem of the hierarchy of secondary factors: According to the idealizational concept of science, assumed here, the first idealizing assumption ought to cancel the secondary factor with the greatest influence on the examined factor; the second one – the factor with a smaller influence, the third one – the factor with still smaller influence, and so on. In very formalized sciences, such as physics, where the influence of particular determinants on the examined factors is defined quantitatively and precisely by way of experiments, that principle of ordering is strictly adhered to. In humanitites, it seems, the situation is slightly different. The influence of factors on the examined factor can hardly be presented in very strict, qualitative fashion, to say nothing about experiments. Hence, problems arise as regards ordering the space of essential factors on the basis of the degree of their influence. Such ordering in humanities is encumbered with a greater risk of inadequacy than in natural sciences. It seems, though, that when it comes to the first two idealizing assumptions, they cancel the secondary factors with the greatest influence on performing actions. ciesielski 2012, p. 101

The existing conceptualization of the the approach to action within the framework of ch-ra includes three models: the first and most idealized one omits all secondary factors, the second and third ones introduce factor vuu (the variability of the usefulness of used tools) and then factor oii (other people’s influence on the action). It seems, then, that at this stage of the conceptualization of the discussed approach it would be justifiable to assume that its essential structure consists of four levels and all the secondary factors which have not been introduced into the model yet are on the last level, as it would be hard to determine which of them is more essential when there is no possibility of determining the exact degree of their influence. The essential structure of human action (San), then, has the following form in ch-ra: San : (0) Kan, Ns, Pn, Man (1) Kan, Ns, Pn, Man, vuu (2) Kan, Ns, Pn, Man, vuu, oii (3) Kan, Ns, Pn, Man, vuu, oii, ssn, tn, hc, nls, kur, lkan, vdn, ki Within the framework of ch-ra a regularity is formed with regard to performing actions. Its antecedent contains a realistic assumption and ten idealizing

94

Ciesielski

assumptions which cancel the influence of factors considered to be secondary ones.19 Its consequent contains a statement determining what action will be performed. The regularity has the following form: (PH-R) If X is capable of performing actions and if performing action ANi allows X to best meet his or her needs, and if (AI) and (AII) and… and (AX), then: ANx = ANi. Thus, if the actions of X have so far allowed X to best meet his or her need, X will continue to satisfy the need with the same action, without reflecting on the effectiveness of alternative actions. The first concretization of the original model takes into account secondary factor vuu (the variability of the usefulness of used tools). It is assumed that: (AI’) the usefulness of the tools used by X constantly changes in time: it increases or decreases. That correction makes it possible to capture not only habitual but also rational actions. (PH-R1) If X is capable of performing actions and (AI’) and (AII) and… and (AX), then: ANx = MaxOp [E,x(AN1),…, E,x(ANn)]20 Therefore, X acts in a habitual fashion until the decreasing (increasing) ­usefulness of the tools causes insufficient (excessive) satisfaction of the needs, 19

Particular secondary factors are canceled by the following idealizing assumptions: (AI) X has at his or her disposal only ideal tools; (AII) X only performs individual actions in an isolated situation; (AIII) X only has one need, n and performs actions only to address that need; (AIV) need n is a permanent and necessary need for X; (AV) X’s choices are not motivated by the costs of alternative actions; (AVI) the states of things which meet X’s need n are ordered in a linear fashion, according to the degree to which they address that need; (AVII) X knows that the only results of his or her actions are his or her intended states of things; (AVIII) X knows how to perform the act he or she intends to perform; (AIX) the nature of need n is constant; (AX) X’s knowledge is true. Additionally, one realistic assumption is made in ch-ra: (real. a.) X is capable of performing actions in order to satisfy his or her needs. 20 Symbols: ANX – X’s action; MaxOp – the maximal effectiveness of the action within the scope of the optimal need satisfaction; [E,X(AN1),…, E,X(ANn)] – a set of actions with a particular value of their effectiveness for X.

On Reduction in the Idealizational Theory of Science

95

when X will reflect and change his or her action into a more effective one which will still fall within the scope of the optimal satisfaction of the need. The second concretization takes into acccount secondary factor oii (other individuals’ influence on the action). Idealizing assumption AII: X only performs individual actions in an isolated situation, is replaced with assumption AII’: X only performs individual actions in an interpersonal situation:21 apart from model X there is an acting individual Y which constitutes X’s social surrounding. The second concretization of the model makes it possible to form two statements differing in the presumed effectiveness of actions of individuals from the social surrounding. (A) actions performed by individuals from the social surrounding are also optimal for X or (B) actions performed by individuals from the social surrounding are not optimal for X. The second concretization, then, has the form of two statements: (PH-R2A) If X is capable of performing actions and if the actions ­ erformed by individuals from X’s social surrounding are also optimal p for X, and (AI’) and (AII’) and (AIII) and… and (AX), then: ANX = ANY , which means that X performs actions in a habitual fashion until the decreasing (increasing) usefulness of the tools causes insufficient (excessive) satisfaction of the need, when X will perform, imitatively, the action performed by Y, to restore the optimal satisfaction of the need. (PH-R2B) If X is capable of performing actions and if the actions ­performed by individuals from X’s social surrounding surrounding are not optimal for X, and (AI’) and (AII’) and (AIII) and… and (AX), then: ANX = MaxOp [E,X(AN1),…, E,X(ANn), E,X(ANY)], which means that X performs actions in a habitual fashion until the decreasing (increasing) usefulness of the tools causes insufficient (excessive) satisfaction of the need, when X will reflect and change his or her action into a more effective one which will still fall within the scope of the optimal satisfaction of the need.

21 The ch-ra conceptualization is based on the classification of human actions including actions: individual in an isolated situation, individual in an interpersonal situation, and social (of which there are three subtypes). See more in: Ciesielski (2012), pp. 78–82.

96 4

Ciesielski

The Relationship between cra and ch-ra in the Light of the Idealizational Concept of Reduction: An Attempt at an Interpretation

Let us, then, conduct an interpretation of the relationships between the concept of a rational act and the concept of a habitual-rational action, with respect to the three dimensions of reduction indicated by Paprzycka. In the case of the theories I compare it can be reasonably assumed that ch-ra is to be the reducing concept and cra – the reduced one, because in the first concept the spectrum of essential factors is broader – it includes factors related to human needs (Ns, ssn, tn, vdn) and to tools used for performing actions (vuu), which are not, in any way, taken into account in the concept of a rational act. 4.1 The Explicatory Dimension The necessary and introductory condition of the occurrence of the reduction relationship is the conceptual compatibility of the given theories. A reducing theory must express, in its language, all the principal concepts of the reduced theory. First of all, both theories must have the same object of study. An exact analysis of the conceptual compatibility of particular factors would require an in-depth comparison including a comparison of all concretizations of both concepts of action, which is impossible because in ch-ra ­concretizations with respect to secondary factors ssn, tn, hc, nls, kur, lkan, vdn, ki have not been made. Even so, we can roughly define the possibility of explicating the cra concepts through the ch-ra concepts. As regards the concept of a human act, Patryas views it from the perspective of its outcomes (Patryas 1979, p. 5). An act, then, is defined by its actual outcome. The intentions of the subject performing an act are not taken into account here. The concept of a human action, on the other hand, is presented with from the point of view of intentions (Ciesielski 2012, p. 88), that is, a given action is defined in terms of the intention of the subject performing the action. The two concepts, then, are different and require an explication. It seems that the concept of an act understood with respect to its outcomes can be explicated with the use of an action understood with respect to the subject’s intentions. The concept in which an action is understood with respect to the subject’s intentions allows two types of a situation: (i) an effective (succesful) action, when the subject’s intentions is fully realized and is identical with the actual outcome of the action; and (ii) an ineffective (unsuccesful) action, when the subject’s intention is not realized and is not identical with the actual outcome of the action.

On Reduction in the Idealizational Theory of Science

97

The first situation overlaps with an act understood with respect to its outcomes, where there is no difference between the subject’s intention and the outcome. The second situation cannot even be considered within the framework of a theory in which acts are viewed from the point of view of the outcomes because the differentiation between an effective act and an ineffective one can only be made when the intention of the subject is taken into account. Therefore, an action understood from the perspective of the subject’s intention is a broader concept from which one can explicate the concept of an act as related to its outcome: an act (in the outcome-related approach) is an effective action (in the intention-related approach).22 It seems that the main cra factors can be explicated with the use of ch-ra concepts. “The knowledge of the subject performing an act” (Kat), “the relation of preferences on the basis of the usefulness of the outcome of an act” (Pu) and “the means of performing an act” (Mat) are concepts which have their counterparts in appropriate concepts of the main ch-ra factors, that is, “the knowledge of the subject performing an action” (Kan), the “the relation of preferences described on the states of things with respect to the degree to which needs are met” (Pn), and “the means of performing an action” (Man). In the case of the concepts of knowledge and the means of performing an act/action it is possible to explicate in both directions. When it comes to the concept of ch-ra main factors including the term “the subject’s needs” (Ns and Pn), it is not possible to explicate those concepts within the framework of cra because they are not, in any way, included in the language of that theory. As regards the issue of the concepts of secondary factors, it can be assumed that the concepts of the following cra factors: nlu, kao, lkat and ki can be explicated with the use of the concepts of factors in ch-ra, respectively: nls, kur, lkan and ki. Moreover, concepts kop and gp (concerning the knowledge and preferences of individuals from the subject’s social surrounding) have their counterparts in the concept of factor oii which seems to be a general concept signifying several factors related to the influence of the individuals from the subject’s social surrounding on the performed action, for example, it signifies

22

Besides, in the outcome-oriented approach, in order to distinguish acts which express reflective decision-making (and such acts are the subject to be explained in the concept of rationality) it is necessary to presume the subject’s intentionality because otherwise it would be difficult to distinguish between instinctive behavior (such as breathing or closing one’s eyelids to protect the eyes from blinding light) and an act resulting from a conscious decision.

98

Ciesielski

the factor of the knowledge and preferences of individuals from the subject’s social surrounding.23 The concept of factor vpk (the variability of preferences and knowledge in time) is somewhat problematic as it is difficult to find a direct counterpart for it among the ch-ra factors due to its dual nature, that is, the variability of two different categories: preferences and knowledge. It seems that the influence of factor vpk in cra can be expressed with the influence of two ch-ra factors: vdn (the variability of the degree to which needs are met) – which makes it possible to explicate the concept of changes of preferences (when the degree to which needs are met thanks to a certain state of things is changed, the preferences regarding that state of things also change), and ki (knowledge inadequacy) – which makes it possible to explicate the concept of changes of knowledge (knowledge inadequacy may be gradual and change in time). As regards factors lke, lko and lkc (concerning the knowledge about the ­characteristics of a set of alternative acts), there are no directly explicatory counterparts in ch-ra. Most probably they could be found in the concept of factor ki which concerns the knowledge of the subject performing an action, including the knowledge about the set of alternative actions from among which the subject chooses the optimal ones (in particular, it includes the knowledge about complementing and excluding them and about which of them the s­ ubject is capable of performing). Therefore, the conceptual compatibility of the two theories can be presented as in Table 4.1. It seems, then, that all the basic concepts of the theory of a rational act, such as the concepts of the examined factor, main factors, and secondary ­factors, can be explicated with the use of appropriate concepts of the theory of a habitual-rational action. Therefore, the relationship between the theories fulfills the first condition of a reduction: the theories are conceptually compatible. Let us add that the compatibility involves an explication and not identicality of the concepts, which makes the possible reduction of the two theories heterogenous. 23

A certain non-specificity of concepts signifying particular essential factors is characteristic of humanities. Often, there is a whole set of factors related to one term. That is apparent in the light of the procedure of concretization used in its in which a cancelation of a particular idealizing assumption frequently entails the acceptance of another, ‘less idealizational’ one, which means that the original idealizing assumption did not cancel one secondary factor but a set of factors represented by one term. The problem is illustrated by the case of factors oii (other individuals’ influence on one’s actions) and ki (knowledge inadequacy) in ch-ra, which are sets of factors rather than single factors.

On Reduction in the Idealizational Theory of Science Table 4.1

99

A comparison of factors examined in the cra and ch-ra

The concepts of the examined factor and ­essential factors in cra

human act (AT) the knowledge of the subject performing an act (Kat) the relation of preferences on the basis of the usefulness of the outcome of an act (Pu) the means of performing an act (Mat)

The concepts of the examined factor and ­essential factors in ch-ra

human action (AN) the knowledge of the subject performing an action (Kan) the relation of preferences described on the states of things with respect to the degree to which needs are met (Pn) the means of performing an action (Man) the non-linear order of the states of non-linear order of the outcomes of acts, things with respect to the degree to on the basis of the individual usefulness which a need is met (nls) for the subject performing an act (nlu) the knowledge about additional ­results the knowledge concerning the apart from the outcomes of the ­unintended results of a performed ­performed act (kao) ­action (kur) a lack of knowledge about how some a lack of knowledge about how some acts are to be performed (lkat) ­actions are to be performed (lkan) knowledge inadequacy (ki) a lack of knowledge about whether the acts from the set of possible alternatives are mutually exclusive (lke) knowledge inadequacy (ki) a lack of knowledge about whether there are other acts which can be performed by the subject, outside of the set of the considered alternative acts (lko) knowledge inadequacy (ki) a lack of knowledge about whether the actions from the set of possible alternatives complement one another (lkc) the variability of preferences and the variability of the degree to which ­knowledge in time (vpk) needs are met (vdn) and knowledge inadequacy (ki) knowledge inadequacy (ki) knowledge inadequacy (ki) other individuals’ influence on one’s the subject performing an act taking actions (oii) into account other people’s or groups’ preferences (gp) other people’s knowledge at the disposal other individuals’ influence on one’s of the subject performing an act (kop) actions (oii)

100 Table 4.1

Ciesielski A comparison of factors examined in the cra and ch-ra (cont.)

The concepts of the examined factor and ­essential factors in cra

The concepts of the examined factor and ­essential factors in ch-ra

the needs of the subject performing an action (Ns) the variability of the usefulness of used tools (vuu) the size of the set of one’s needs (ssn) the type of the need met (tn) how high the costs of actions are (hc)

4.2 The Essentiality Dimension In the light of the approach discussed above, the necessary condition of the occurrence of the reduction relationship between two theories is their essentiality agreement. That happens when the reducing theory broadens the repertoire of essential factors with respect to the reduced theory, ­without ­disturbing their relative order. To be more precise, that happens when: (a) all essential factors of the reduced theory are taken into account in the r­ educing theory; (b) the equally essential factors in the essential structure are also equally essential in the other theory; (c) for each two factors their relative order of essentiality in the essential structure of one theory is replicated in the other theory. In the case of ch-ra the essential structure is extended with the following essential factors: – – – – –

main factor Ns: the needs of the subject performing an action; secondary factor vuu: the variability of the usefulness of used tools; secondary factor ssn: the size of the set of one’s needs; secondary factor tn: the type of the need met; secondary factor hc: how high the costs of actions are.

As a result of the extension of the essential structure by main factor Ns the ­possible reduction between cra and ch-ra in the essentiality dimension would assume the form of a refutational reduction. Because of the newly added secondary factor vuu, it would be the corrective sub-variant of a r­efutation, as that factor differs, with respect to essentiality, from other factors already ­recognized in cra.

On Reduction in the Idealizational Theory of Science the CRA essential structure

101

the CH-RA essential structure

AT

AN

(0) Kat, Pu, Mat

(0) Kan, Ns, Pn, Man

(1) Kat, Pu, Mat, nlu, kao, lkat, lke, lko, lkc

(1) Kan, Ns, Pn, Man, vuu

(2) Kat, Pu, Mat, nlu, kao, lkat, lke, lko, lkc, vpk, ki

(2) Kan, Ns, Pn, Man, vuu, oii

(3) Kat, Pu, Mat, nlu, kao, lkat, lke, lko, lkc, vpk, ki, gp, kop

(3) Kan, Ns, Pn, Man, vuu, oii, ssn, tn, hc, nls, kur, lkan, vdn, ki

Figure 4.1 A comparison of the cra and ch-ra essential structures, taking into account the conceptual compatibility; an arrow signifies the relationship of the explication of cra concepts in the language of ch-ra.

When the above solutions concerning the conceptual compatibility between the analyzed theories are assumed, the essentiality relationship can be illustrated as in Figure 4.1: A comparison of the essential structures of both theories of action shows that all essential kcr factors are taken into account in the ch-ra theory, ­however, the isomorphism of secondary cra factors and their explicatory counterparts in the ch-ra theory is not maintained. The essential ch-ra structure does not preserve the relative order of the secondary factors in the cra essential structure. The two least essential factors in cra: kop (other ­people’s knowledge which can be used by the subject performing an act) and gp (the subject ­performing an act taking into account other people’s or groups’ preferences) are at the third essentiality level, and their counterpart is factor oii which is at the second essentiality level in ch-ra. All other cra ­secondary factors have their ch-ra counterparts in the factors at the third essentiality level. There is, then, an accidental (as it concerns secondary ­factors) essentiality disagreement in this case. Therefore, the relationship between cra and ch-ra does not fulfill the second condition of a reduction because ch-ra extends and interferes in the essential structure of cra, by disturbing the relative order of the essentiality of factors: in cra factors kop and gp are considered to be less essential than

102

Ciesielski

other secondary factors whereas their counterpart, factor oii in ch-ra is more essential than the counterparts of the remaining secondary factors. 4.3 The Derivative Dimension The third necessary dimension of a reduction relationship in Paprzycka’s model is the possibility of deriving the statements of the reduced theory from the statements of the reducing theory. In the analyzed case the condition would be fulfilled if it were possible to derive, with certain additional assumptions, all counterparts of cra statements from ch-ra statements. A complete reconstruction of the relationship between cra and ch-ra in the derivative dimension is not as yet possible as no concretizations have been made in ch-ra with respect to secondary factors ssn, tn, hc, nls, kur, lkan, vdn, ki. One can only derive an explicatory counterpart of the most idealized cra law from an appropriate ch-ra statement. The initial cra law has the following form: (PR) If (Real. A.) and (A10) and (A9) and… and (A1), then ATX = Max [U0, X(AT1),…, U0, X(ATn)]. The above (PR) statement can be derived from the first ch-ra concretization: (PH-R1) IF (real. a.) and (AI’) and (AII) and… and (AX), then ANX = MaxOp [E,X(AN1),…, E,X(ANn)]. If we assume that there is no upper limit of optimal satisfaction of human needs (that people aim at the maximal and not optimal satisfaction of their needs), the concept of the maximal effectiveness of an action with regard to the optimal level of meeting the needs (MaxOp) becomes the concept of the maximal effectiveness of the outcome of an act (MaxU); additionally, when the concepts of the examined factor and the essential factors are compatible and with some factor values not included in cra (Ns = α, vuu = β, ssn = γ, tn = δ, hc = ε),24 we receive, within the framework of ch-ra, an explicatory counterpart of the (PR) regularity (let us denote it by: E(PH-R1)(PR)):

24

When deriving the counterparts of the statements of the reduced theory from the statements of the reducing theory it is necessary to assume particular values (in a boundary case it can be value zero) of factors not present in the structure of the reduced theory. See more in: Paprzycka (2005), p. 133.

On Reduction in the Idealizational Theory of Science

103

E(PH-R1)(PR) If (real. a.) and Ns = α and vuu = β and (AII) and ssn = γ and tn = δ and hc = ε and (AVI) and… and (AX), then ANX = Max [E,X(AN1),…, E,X(ANn)]. In other words, the first ch-ra concretization will point, in its consequent, if additional assumptions concerning the values of factors not included in cra (that is, factors Ns, vuu, ssn, tn, hc) are made, to the same X’s action as was indicated by the original cra law. Therefore, at that level of concretization, the necessary condition of a reduction in the derivative dimension is partially fulfilled, that is, we can derive a counterpart of the most idealized cra law from ch-ra statements. There is, then, only partial reduction. It would be difficult to determine if the remaining cra statements would have their counterparts in ch-ra because a further analysis of the derivative model is not possible, as not all ch-ra concretizations have been formed. At this point we can only have some intuitions which might shed some light on the derivative dimension of the relationship between cra and ch-ra. The aim of the conceptualization of the approach to a habitual-rational action was to overcome the limitation of the model of rationality related to omitting other types of action, especially habitual and imitative ones, while preserving all the statements which concern a rational action. In other words, the role of ch-ra is to overcome the one-sidedness of the two traditional approaches to human practice: the “reflective” one (which emphasizes individual and rational decision-making) and the “non-reflective” one (which emphasizes habituation, routine, and imitation) (Ciesielski 2012, p. 8). Moreover, within the framework of ch-ra it is possible to reveal what the limitations of the earlier theory consist in (the assumption of the constant reflectivity of the subject performing an action) and to identify the conditions in which cra correctly describes human practice (non-optimal satisfaction of needs and nonoptimal effectiveness of actions of individuals from the social surrounding). That can be illustrated as in Figure 4.2: It seems that further ch-ra concretizations ought to reveal the explicatory counterparts of particular cra statements, especially that the two theories are conceptually compatible and that the essentiality disagreement only concerns the two weakest secondary factors.25 25

In this text I only analyze the formal aspect of the reduction relationship between the cra and ch-ra statements. These reflections do not include the explanatory aspect – I do not compare the explanatory force of those theories in particular empirical cases. At

104

Ciesielski

statements concerning a rational act

the concept of a rational act (CRA)

statements concerning a rational (reflective) action;

statements concerning a habitual (non-reflective) action;

statements concerning an imitative (non-reflective) action;

action in the conditions of: -less than optimal satisfaction of needs -less than optimal effectiveness of actions of individuals from the social surrounding

action in the conditions of: -optimal satisfaction of needs (the effectiveness of actions of individuals from the social surrounding is irrelevant)

action in the conditions of: -less than optimal satisfaction of needs -optimal effectiveness of actions of individuals from the social surrounding

the concept of habitual-rational action (CH-RA)

Figure 4.2 A comparison of statements concerning particular types of action: in the concept of a rational act and in the concept of habitual-rational action.

5 Conclusions The interpretation of the relationship between cra and ch-ra presented in this article allows me to arrive at the following conclusions: – in the explicatory dimension: the condition of a reduction is fulfilled as the theories are conceptually compatible; – in the essentiality dimension: the condition of a reduction is not fulfilled as the theories are in essential disagreement; – in the derivative dimension: the condition of a reduction is partially fulfilled as not all ch-ra concretizations are formed and one can only derive the most idealized cra law in the ch-ra terms. According to Paprzycka if any of the three conditions is not satisfied, there is no occurred the reduction relationship. As there is no essentiality agreement between the cra and ch-ra essential structures we cannot define ch-ra as a this point let us only note that microeconomics theorists point to some limitations of the model of rationality, which manifest themselves in, for instance, consumer behavior based on habituation or imitation rather than rational decisions and maximizing the satisfaction of one’s needs. Irrational behavior, including unreasonable consumers’ activity, has been discussed by, among other authors: Schelling (1984, 2007), O’Shaughnessy (1987), and Edlin, Gelman, and Kaplan, (2007). It seems – although that is only my intuition which awaits verification – that we can explain behavior deviating from the standard model of rationality, used in microeconomics, on the grounds of ch-ra, thanks to the concept of the “optimization of needs” (which has substituted the concept of the “maximization of the aim”).

On Reduction in the Idealizational Theory of Science

105

theory which reduces the concept of rationality. It seems, however, that essentiality agreement, raised to the rank of a necessary condition, is too strict a condition when applied to the humanities. Paprzycka conceptualizes her approach to reduction with the idealizing assumption that a researcher can precisely define the strength of the influence of particular factors and build a hierarchy based on that knowledge. However, the cognitive specificity of the humanities, as opposed to natural sciences, is the inherent arbitrariness in hierarchization of factors. In the case of the humanities, with their primarily qualitative language, a stratification of the essential factors assumes the form of relative, intuitive solutions. Because of the lack of cognitively effective tools, for example repeateable experiments which would help to estimated the strength of influence of particular factors exerting the impact on the examined phenomenon. Moreover, the image of the essential structure reconstructed by a researchers is frequently influenced by non-cognitive reasons, for example, pragmatic considerations (such as the ease with which subsequent statements can be explicated).26 Therefore, the disturbed essentiality agreement of the discussed theories does not have to entail a lack of a reduction relationship between cra and chra because Patryas’s order of secondary factors is, as admitted by the author himself, arbitrary and could just as well be presented differently. For example, a factor related to the non-linear order of the outcomes of acts, on the basis of the individual usefulness for the subject performing an act (nlu), taken into account by Patryas in the first concretization of the original model, could just as well be taken into account at a later stage of the concretization (be ascribed lesser essentiality), especialy after having introduced factors related to the social surrounding (kop and gp – considered by Patryas to have the weakest influence). That would be the more desirable as factor nlu, in the itn understanding of essentiality as an exclusion of the possible occurrences of the examined factor, allows a greater number of possible acts than factors kop or gp which show that the subject will perform one particular 26

The difficulties of applying to humanities the same strict methodological rules as are applied to sciences are directly discussed by, among other researchers, Leszek Nowak who attempts to define the reduction relationship between his interpersonal model of a person and the three earlier theories: “The interpersonal model of a person with Gombrowicz’s correction can, then – if we set aside the imprecision of language inherent in social sciences, and, consequently, a certain non-specificity of all procedures realized in language, including the unifying ones – be viewed as a unification of the evangelical, liberal, and Marxian models of a person” (Nowak 2000, p. 72).

106

Ciesielski

act. Patryas notes that lack of a limitation to only one act and explains that it is most probable that, ultimately, there are other, “complementary determinants of acts”, not taken into account in the model and playing a role in the final choice of a particular act (Patryas 1979, pp. 22–23). Referring to additional determinants which were not introduced into the concept in a systematic fashion, and doing so at already on the occasion of the first concretization of the model, confirms Patryas’s claim that the hierarchy of secondary factors is highly arbitrary and governed more by formal reasons (the ease with which it is possible to conceptualize and form statements, etc.) than by subject-matter reasons (the real degree of the influence of particular secondary factors on the examined factor). The above characteristic of research limitations in the field of h ­ umanities – should we accept it – also colors our understanding of the reduction relationship between theories in human sciences. If we asume that the relationship between the theories of action analyzed in this text are representative for the remaining cases in the field of humanities, then the specificity of that field speaks for a modification of Paprzycka’s approach to reduction. Namely, the essentiality agreement as a necessary condition of a reduction ought to be weakened and replaced by a condition grounded in the following alternative: the necessary condition of a reduction relationship between two theories from the field of humanities is either essentiality agreement or accidental essentiality disagreement (concerning solely secondary factors). It is only fundamental essentiality disagreement (concerning main factors) that prevents a reduction relationship. The meaning of the discussed condition of a reduction can also be expressed as follows. We can talk of a reduction in humanities when: (i) the reduced and reducing theories are conceptually compatible (the explicatory dimension of a reduction); (ii) the reducing theory mirrors the relative order of the main factors with respect to the secondary factors in the reduced theory; such order with regard to secondary factors is not necessary (the essentiality dimension of a reduction); (iii) it is possible to derive, from the reduced theory (with a certain ­accepted degree of approximation and with additional provisions), the original law of the reduced theory and all its concretizations (the derivative ­dimension of a reduction). The assumption of such a ‘weakened’ approach to reduction would make it possible to interpret the relationship between the concept of a rational act and

On Reduction in the Idealizational Theory of Science

107

the concept of a habitual-rational action as a partial,27 heterogeneous, refutational reduction in the corrective sub-variant. If the intuitions expressed above are accurate, the concept of a reduction makes it possible to define the further differences between the humanities (as well as the social sciences) and natural sciences.28 One of the specificity of the humanities and social sciences manifests itself in less stringent criterion has to be fulfilled for a relationship of the reduction between two theories. Translated by Agnieszka Wróblewicz References Brzechczyn, K. (1998). Odrębność historyczna Europy Środkowej. Studium metodologiczne [Historical Distinctiveness of Central Europe. A Methodological Study]. Poznań: Wydawnictwo Fundacji Humaniora. Brzechczyn, K. (2009). Methodological Peculiarities of History in Light of Idealizational Theory of Science. In: idem (ed.) Idealization XII: Modeling in History. Poznań Studies in the Philosophy of the Sciences and the Humanities 97, pp. 137–157. Amsterdam – New York: Rodopi Ciesielski, M. (2011). Problem of Limits of Rational Model of Human Actions. An Attempt to Capture Habitual-Rational Action. Studia Europaea Gnesnensia 4, 393–398. Ciesielski, M. (2012). Zagadnienie ograniczeń racjonalnego modelu działań ludzkich. Próba ujęcia działania nawykowo-racjonalnego [Problem of Limits of Rational Model of Human Actions. An Attempt to Capture Habitual-Rational Action]. Poznań: Wydawnictwo Poznańskie. Edlin, A., A. Gelman, N. Kaplan (2007). Voting as a Rational Choice: Why and How People Vote to Improve the Well-Being of Others. National Bureau of Economic Research, Cambridge, MA.

27

28

At this stage of the ch-ra concretization the nature of the reduction is partial but one cannot exclude the possibility that the conceptualization of the remaining ch-ra concretizations would make it possible to derive all cra statements and then call the reduction complete. Brzechczyn (1998; 2009) analyzes the difference between the two fields of study at the level of essential structure. The author identifies two types of essential structures: an essential structure dominated by the main factor (its influence is greater than the sum of influence exerted by secondary factors) and an essential structure dominated by a class of secondary factors (their total influence is greater than the influence exerted by the main factor) (Brzechczyn 2009, pp. 141–142).

108

Ciesielski

Egiert, R. (2000). Parafrazy idealizacyjne [Idealizational Paraphrases]. Poznań: Wydawnictwo Fundacji Humaniora. Hempel, C.G. (1961/62). Rational Action. Proceedings and Addresses of the American Philosophical Association 35, 5–23. Hohwy, J., J. Kallestrup (eds.) (2008). Being Reduced: New Essays on Reduction, Explanation, and Causation. Oxford: Oxford University Press. Kleszcz, R. (1998). O racjonalności. Studium epistemologiczno-metodologiczne [On Rationality. An Epistemological-Methodological Study]. Łódź: Wydawnictwo Uniwersytetu Łódzkiego. Kmita, J. (1971). Z metodologicznych problemów interpretacji humanistycznej [Some Methodological Problems of Interpretation in the Humanities]. Warszawa: PWN. Kmita, J. (1972). Kilka uwag o idealizacji w badaniach logicznych nad językiem nauki. Studia Semiotyczne 3, 57–74. Krajewski, W. (1977). The Principle of Correspondence and the Growth of Knowledge. Dodrecht – Boston: Reidel. Łastowski, K. (1987). Rozwój teorii ewolucji. Studium metodologiczne [The Develop­ment of the Theory of Evolution. A Methodological Study]. Poznań: Wydawnictwo UAM. Nagel, E. (1961). The Structure of Science: Problems in the Logic of Scientific Explanation. Harcourt, Brace & World. Niiniluoto, I. (2013). Is Science Progressive? Dordrecht: Springer. Nowak, L. (1971). U podstaw Marksowskiej metodologii nauki [Foundations of the Marxian Methodology of Science]. Warszawa: PWN. Nowak, L. (1974). Zasady marksistowskiej filozofii nauki. Próba systematycznej rekonstrukcji [The Principles of the Marxist Philosophy of Science. An Attempt at a Systematic Reconstruction]. Warszawa: PWN. Nowak, L. (1977). Wstęp do idealizacyjnej teorii nauki [An Introduction to the Idealizational Theory of Science]. Warszawa: PWN. Nowak, L. (1980). The Structure of Idealization. Towards a Systematic Interpretation of the Marxian Idea of Science. Synthese Library, vol. 139. Dordrecht/Boston/London: Reidel. Nowak, L. (1991). U podstaw teorii socjalizmu, tom 3: Dynamika władzy. O strukturze i konieczności zaniku socjalizmu [The Foundations of the Theory of Socialism, vol. 3: The Dynamics of Power. On the Structure and Necessity of the Decline of Socialism]. Poznań: Nakom. Nowak, L. (1993). Two Inter-Human Limits to the Rationality of Man. In: L. Nowak, M. Paprzycki (eds.). Social System, Rationality and Revolution. Poznań Studies in the Philosophy of the Sciences and the Humanities 33, pp. 197–204. Amsterdam – A ­ tlanta: Rodopi. Nowak, L. (2000). Człowiek i ludzie. Modele z Gombrowicza [Man and People. Models from Gombrowicz]. Warszawa: Prószyński i S-ka.

On Reduction in the Idealizational Theory of Science

109

Nowakowa, I (1972). Zasada korespondencji w fizyce [The Principle of Correspondence in Physics]. Poznań: Ph.D. thesis. Dept. of Philosophy, Adam Mickiewicz University in Poznań. Nowakowa, I (1975a). Dialektyczna korespondencja a rozwój nauki [The Dialectical Correspondence and the Development of Science]. Warszawa – Poznań: PWN. Nowakowa, I (1975b). Idealization and the Problem of Correspondence. Poznań Studies in the Philosophy of the Sciences and the Humanities 1, 65–70. Nowakowa, I (1982). Dialectical Correspondence and Essential Truth. In: W. Krajewski (ed.), Polish Essays in the Philosophy of the Natural Sciences, pp. 135–146. Dordrecht: Reidel. Nowakowa, I (1991). Zmienność i stałość w nauce. [Change and Stability in Science]. Poznań: Nakom. Nowakowa, I. (1994). The Dynamics of Idealizations. Poznań Studies in the Philosophy of the Sciences and the Humanities 34. Amsterdam – Atlanta: Rodopi. O’Shaughnessy, J. (1987). Why People Buy? New York – Oxford: University Press. Paprzycka, K. (1989). Reduction and Explanation in the Light of the Idealizational Approach to Science. Honors Thesis: Harvard University. Paprzycka, K. (1990). Reduction and Correspondence in the Idealizational Approach to Science. In: J. Brzeziński, F. Coniglione, T.A.F. Kuipers and L. Nowak (eds.), Idealization I: General Problems. Poznań Studies in the Philosophy of the Sciences and the Humanities 16, pp. 277–286. Amsterdam – Atlanta: Rodopi. Paprzycka, K. (2005). O możliwości antyredukcjonizmu [On the Possibility of Anti-reductionism]. Warszawa: Semper. Patryas, W. (1979). Idealizacyjny charakter interpretacji humanistycznej [Idealizational Nature of Humanistic Interpretation]. Poznań: Wydawnictwo Naukowe UAM. Riel, R. van (2014). The Concept of Reduction. International Publishing Switzerland: Springer. Schelling, T.C. (1984). Choice and Consequence. Cambridge: Harvard University Press. Schelling, T.C. (2007). Strategies of Commitment and Other Essays. Cambridge: Harvard University Press. Silberstein, M. (2002). Reduction, emergence and explanation. In: P.K. Machamer, M. Silberstein (eds.), The Blackwell Guide to the Philosophy of Science, pp. 80–107. Cambridge: Blackwell. Strawiński, W. (2008). O możliwości antyredukcjonizmu: uwagi polemiczne [On the Possibility of Anti-reductionism. Polemical Remarks]. Diametros 17, 70–81. Szaniawski, K. ([1971] 1994). Kryteria podejmowania decyzji. In: K. Szaniawski, O nauce, rozumowaniu i wartościach [On Science, Reasoning, and Values] (selection and edition: J. Woleński), pp. 286–304. Warszawa: PWN. Watkins, J. (1970). Imperfect Rationality. In: R. Borger, F. Cioffi (eds.), Explanation in the Behavioural Sciences, pp. 167–217. Cambridge: Cambridge University Press.

part 2 Idealization in the Social Sciences

⸪

chapter 5

Idealization in Economics: A Structuralist View1 Adolfo García de la Sienra Abstract The aim of the present paper is to deal, within a structuralist perspective, with the problem of the nature and types of idealization in positive economics. Even though the topic of idealization in normative economics is very important, and some results presented here might become also useful in dealing with such topic, a proper treatment of the same would require more space than provided here. The most complete discussion of idealization from a structuralist perspective is due to De Donato (2011), but this work is concerned mainly with idealization in physics. Are the claims made about idealization in physics valid mutatis mutandis in economics? This is one of the leading questions of the paper.

Motivation According to Balzer and Zoubek, The basic picture of concretization and idealization seems to be that there is ‘the’ real world which is described by human theories to more or less satisfactory degree. As the real world is very complex the theories describing it in the beginning provide very rough pictures, many real features are left out of account. Theories therefore are idealized. In the historical process more and more of the features originally ignored are incorporated into the theories which in this way become more accurate pictures of the world. This is the process of concretization. balzer and zoubek 1994, p. 57

Yet, even though Balzer and Zoubek recognize that “‘the’ real world seems to be necessary as a guide for concretization” for, without having it in view, “concretization has no ‘direction’ and becomes arbitrary, or so it seems,” they nevertheless find problematic “the existence of a fully determined real world 1 The present paper was produced with support of conacyt Project 127380, Filosofía de la Economía.

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_007

114

de la Sienra

independent of human description” and so they conclude that “idealization and concretization should be regarded as ordinary intertheoretical relations between theories” (Balzer and Zoubek 1994, p. 57). It is true that some philosophers have thought that some way of referring to concrete actual beings is necessary in order to develop a complete semantics of theories. A recent defense of this view is due to Muller (2011), who introduces the concept of concrete actual being in order to perform this task. Indeed, according to Marx’s method of political economy, the subject (das Subjekt) of any science is real and concrete (real und konkret); it is an already given concrete organic whole (ein gegebnes konkretes, lebendiges Ganzes) that remains outside the head (außerhalb des Kopfes) independent of it (in seiner Selbständigkeit bestehen), and must always be kept in mind as a presupposition of the representation (als Voraussetzung stets der Vorstellung vorschweben), when the theoretical method (theoretische Methode) is employed.2 Even though Marx’s Subjekt (in economics), as described according to his method, is far from being “a fully determined real world independent of ­human description,” some extremely empiricist and instrumentalist philosophers would find it objectionable. Yet, some kind of conceptualization is necessary in order to fix the target system of any intended theorization, if only to know which empirical data or measurements might turn out to be relevant to its understanding, and indeed as a pre-condition to produce any idealized representation of it. Some philosophers have thought that nothing less than a full-fledged metaphysical or ontological theory is required for this endeavor.3 Marx’s scientific practice seems to presuppose that only a certain minimal description of das Subjekt, the target system, is necessary in order to determine the real and concrete referent. This description is given by Marx, in the case of economics, in a language which contains general but not idealized terms. He says that, starting with a small number of determinant, abstract, general and simple concepts (einfachste Bestimmungen) of aspects of the target system, a richer concept of the same is built by way of thought (im Weg des Denkens), as a concentration of many determinations (Zussamenfassung vieler Bestimmungen), as a unity of the diverse (Einheit des Mannigfaltigen). Clearly, this concept is not the real Subjekt, but only a “spiritual” (geistig) reproduction of

2 Cf. „Die Methode der politischen Ökonomie” in Marx (1983); I follow the Penguin Edition (Marx 1973). 3 I myself believed this when I wrote that a philosophia de ente was necessary for that (cf. ­García de la Sienra 1990). Stigum (2003, pp. 37ff.) also seems to believe it when he introduces his “world of facts.”

Idealization In Economics: A Structuralist View

115

the same, just a concrete totality in thinking (Gedankentotalität, als ein Gedankenkonkretum, in fact ein Produkt des Denkens). Marx claimed that in economics the real Subjekt always had to be seen as an organisches Ganzes and conceptualized, no matter which concrete real given economy was going to be studied, through four general categories: production, distribution, exchange and consumption. These categories have to be further specified, seen in their mutual “interaction”, i.e. in their inner connections, and applied to given properties or relations of the economy under scrutiny. Clearly, the resulting Gedankenkonkretum, no matter to what extent the categories are specified, is still a general concept, but it is nevertheless far from being a theoretical model. Rather, it fixes a target system and opens the possibility of leading inquiries, probing into it, collecting empirical information (empirical data), making some measurements, and eventually the possibility of building idealized theoretical models of some of its aspects. Hence, the construction of the Gedankenkonkretum by way of thought should not be confused with the process of concretization as understood by Leszek Nowak and some structuralists (as Balzer and Zoubek, quoted above, or De Donato 2011). Rather, idealization and concretization in this latter sense must be represented exclusively, in effect, as a relationship between theoretical models. But the reference to the Gedankenkonkretum is the required guide for concretization. 1

Empirical Structures

By ‘empirical structure’ I mean a set-theoretic structure that records, in a ­systematic way, empirical observations about some aspect or aspects of a previously determined Subjekt. These can be data related to the aggregate production of a concrete economy during a certain period of time (e.g., the American economy from 1919 to 1929); tables reporting positions of “sparkling points” in the sky during a “long series of nights,” or the data resulting from the comparison of several bodies, as to their weight, by means of an equal-arms balance. Other examples are given by tables recording the observation of business cycles or, more recently, by Patrick Suppes’ record of electroencephalographic data collected in order to establish brain representations of words.4 It is not the case that the data collected by means of systematic observations are always recorded in order to test or apply a previously given theory. For instance, in spite of their rich algebraic and computational tradition, the 4 For a survey of this last project, cf. Crangle (2014).

116

de la Sienra

Babylonian astronomers did not produce anything that might be deemed as a “theory” of planetary motion similar to that of Ptolemy. This situation is also typical in econometrics, where it is common the production of statistical correlations without the use of much theory. These correlations, normally obtained through the method of ordinary least squares, constitute an elementary form of idealization for which there is nothing resembling a concretization process. But this method is only one of the many methodologies used in order to interpret or organize quantitative observations: the wealth of data in the different sciences is so overwhelming, and the range of methodologies used in quantitative observations is so vast, “that in their full detail [they] are theoretically immeasurable” (cf. Suppes 2011, p. 119). Esentially, empirical structures are sets of n-tuples ⟨x1, …, xn⟩, whose coordinates are taken as belonging to certain empirical categories X1, …, Xn. For instance, X1 may be required to be a set of consecutive years, X2 the recorded annual gnp of a country during these years, and so on. These data are usually presented in tables and stored in databases. Some empirical structures are built in order to serve the purposes of a given scientific theory. These purposes are varied: testing of a particular theoretical systematization, application, or computation of a particular magnitude. At a micro-logical level, sometimes a particular form of a magnitude is posited in order to explain some phenomenon, in which case empirical data are used to check the accuracy of that form. Sometimes, the point is not to check whether the theory is good, but just to apply it in order to obtain realizations useful for some particular purpose (for instance, information in order to put a satellite in a required orbit). Or just to compute the value of some magnitude (like the mass of the sun). A most typical scientific endeavor is that of trying to imbed a data structure into a partial potential model of a certain theory T. For the sake of the example, let us suppose that, out of the observation of the behavior of a consumer, we get a structure of data of the following form: ​D̂ ​ = 〈X, F, ​η̂ ​〉,​ where X is a set of vectors representing the consumption vectors available to the agent, F is a finite subset of P × W, P is the family of all possible price systems (vectors), W is the interval of all possible levels of wealth for the agent, and ​η̂ ​ is a function defined over F recording the observed choice of the agent at each level of prices and wealth (p, w). Function ​​η̂ ​​is obviously finite and ­discrete, and can be presented as in Table 5.1.

Idealization In Economics: A Structuralist View Table 5.1

117

Empirical structure ​D̂ ​represented as a table

Argument

Value of ​η̂ ​

(p1, w1) (p2, w1) ⋮ (pk, wk) ⋮ (pn, wn)

​x ​̂ 1​ = ​η̂ ​(​p​ 1​, ​w​ 1​)​ ​x ​̂ 2​ = ​η̂ ​(​p​ 2​, ​w​ 2​)​ ⋮ ​x ​̂ n​ = ​η ̂ (​ ​p​ k​, ​w​ k)​ ​ ⋮ ​x ​̂ n​ = ​η̂ ​(​p​ n​, ​w​ n​)​

If ≳; is the preference relation of the agent (usually represented by some ­utility function u: X → R) the models of consumer choice theory T are of the form ​M = 〈X, P, W, η, ≿〉,​ and are defined as those potential models that satisfy the law of utility maximization: η(p, w) is actually optimal within the set {x ϵ X│px ≤ w}. Since ≳ (and hence u) indeed smells like T-theoretical under some circumstances, the partial potential models of T have the following aspect: B ​  = 〈X, P, W, η〉 = ρ​(M)​.​ (P is the functor that eliminates the theoretical terms of the potential models of T). Thus, the problem is to show that ​D̂ ​ is “approximately” imbedded in B ​  = 〈X, P, W, η〉​; that is to say, there exists a structure ​D̂ ​ = 〈X, F, ​η̂ ​〉,​ such that, for every ​(​ p, w)​ ϵ F, ∥ ​η̂ ​(p, w)​ − ​η̂ ​(p, w)​ ∥ r consumers’ demand equals zero (see, proposition 1 in Varian’s paper).

142

Hardt

f­ollowing questions are worth asking. First, how such a model explains? ­Second, how it enables us to learn about the actual world? Three, what criteria one should apply in choosing the right model of price dispersion in case of having multiple models fulfilling the MECH conditions? The next section deals with each of these questions. 3

Learning from Models

Historically, in economics, the nomological deductive model of explanation was a dominant one. It equated explanations with deductions of descriptions of economic phenomena from premises including scientific laws. Although there were (and still are) some controversies about the nature of scientific laws, philosophers generally agree that such laws are “nonaccidental generalization that made no reference to particulars and supported counterfactual claims” (Kincaid 2012, p. 137). The deductive nomological model of explanation was finally rejected due to fundamental problems in distinguishing scientific from nonscientific laws as well as the general issue of many examples showing that derivation from a law was neither necessary nor sufficient to explain. Therefore, many alternative theories of explanation emerged (for an interesting overview, see, e.g., Salmon 1989). It is for sure beyond the scope of this paper to present all of them, however, a common trait among them is that they do not explain by invocations to universal laws. As far as economics is concerned, I should comment now on a typical way economists defend the centrality of laws in economic theory. It is interesting that they do it by referring to modelling practices of economics. In other words, they ask how laws produced by models with false assumptions can e­ xplain. They answer that these laws are not false but qualified as ceteris ­paribus. As Kincaid (2012, p. 145) rightly notices, that view is supported by Cartwright’s analysis of science, especially physics, since its fundamental laws are just ­ceteris paribus laws, e.g., the force on a body due to gravity is equal to mass times acceleration only assuming no other physical forces are present. The same holds for economics where a typical way of reasoning is to have models as producers of theories that are always true in models but only true in ceteris paribus sense if referred to the actual world. For instance, in the neoclassical model of consumer choice, the higher the price, the less a given consumer is willing to buy (the law of demand). This law is always true in the model, but not vis-à-vis the actual world where its ceteris paribus interpretation is claimed to be true, namely all other things being equal or held constant (ceteris paribus), the higher the price, the less a given consumer is

Between Isolations and Constructions

143

to buy. ­However, ­philosophers of ­economics have several objections against such a way of defending the centrality of laws in model-based explanations. First, claims qualified as ceteris paribus laws seem to be unfalsifiable (Earman et al. 2002). Second, if one attempts to spell out ceteris paribus laws, then one arrives at the nomological deductive model of explanation which is however claimed not to be the adequate one. Thus, postulating the existence of ceteris paribus laws does not offer an acceptable answer to the question on how models with false assumptions explain. Here the semantic view can help us in solving such a puzzle, since it claims that this is a separate empirical question whether there is anything in the world corresponding to the abstract entity (e.g., a mechanism) described by the theory and embedded in a theoretical model. Therefore, the question on how we may have good empirical evidence for models is distinct from the question on how models explain. I am to return to this former question later, but now let me continue my reflections on the latter. The question on how literally false models explain has gained a considerable attention in philosophy of science with such answers as the following ones (according to Kincaid 2012): 1. Models provide “insights”. This is a common informal rational given by ­social scientists in defense of particular models. 2. Models unify, i.e., they show how different phenomena might be captured by the same model (Morgan and Morrison 1999). 3. They serve as instruments – we do things with models (ibid.). 4. Models are isomorphic to the phenomena of interest (Giere 1988). 5. Models are nomological machines and theoretical models are just blueprints for such machines (Cartwright 1983). Now, how about the kind of explanation the above-described Varian’s model offers us? It is for sure that his model does not unify (Ad. 2) and it is not used as a typical instrument (Ad. 3). Also, it is not isomorphic to the phenomena of interest (Ad. 4). Thus, the very first (“insights”) and the last (“machines”) of the five above mentioned options seem to be worth investigating. Some insights are given by Varian himself in his 1978 paper where the emphasis is put on the causal fit between the model and the real world: The goal of causal application is to explain aspects of the world that can be noticed or conjectured without explicit techniques of measurement. In some cases, an aspect of the world (such as price dispersal […]) is noticed, and certain aspects of the micro-situation are thought perhaps to explain it; a model is then constructed to provide the explanation. In

144

Hardt

other cases, an aspect of the micro-world is noticed, and a model is used to investigate the kinds of effects such a factor could be expected to have (emphasis added). gibbard and varian 1978, p. 672

An important virtue of explanation is its simplicity: “[…] it is important that one be able to grasp the explanation. Simplicity, then, will be a highly desirable feature of such models. Complication to get as close as possible a fit to reality will be undesirable if they make the model less possible to grasp” (Gibbard and Varian 1978, p. 672). Sugden (2000, p. 13) interprets this passage from Varian’s paper in terms of simplicity as a model’s characteristic which makes communication with the audience easier. Also, for him simplicity serves as device legitimizing the presence of highly unrealistic assumptions in models. I disagree with such an interpretation of Varian’s ideas. In my reading of his paper, I treat simplicity as a suggestion for searching essential mechanisms explaining the real. Or, even more, a good model tells a story (one story) (Gibbard and Varian 1978 p. 666) which must give an answer that is right in its essentials (p. 669) and unrealistic assumptions “are chosen not to approximate reality, but to exaggerate or isolate some feature of reality” (Gibbard and Varian 1978, p. 673). So, the mechanism (treated in the sense of MECH) is such an answer, since it refers to the essentials of a given systems. Also, my analysis of his Model of Sales supports the above interpretation of simplicity from his 1978 paper. Here enters my conception of a believable world under which a model is an entity containing mechanisms that are believed to be similar to the ones operating in the real world. They are similar because a model of a mechanism demonstrates the reality of mechanism by isolating it. According to MECH what we find in mechanisms are not the universal regularities but rather a set of Millian tendencies or capacities (natures) as Cartwright would name them. They are crucial for the mechanism’s ability to explain: “Our most wideranging scientific knowledge is not knowledge of laws but knowledge of the natures of things” (Cartwright 1999, p. 4) and later she adds that “Idealizations and the inference to natures form a familiar two-tiered process that lies at the heart of modern scientific inquiry” (Cartwright 1999, p. 83). If “capacities are real” (Cartwright 1989, p. 1), then mechanisms are real too. Also, the reality of capacities leads us to the acceptance of the idea of the world as a world of powers (capacities) (Mumford 2013, p. 17). This is a typical realist claim and thus anti-Humean about powers. In such a conceptual schema instead of referring to capacities in the form in-the-nature-of-sth.-is-to-produce, one may say it-is-believable-that-sth.-is-to-produce.

Between Isolations and Constructions

145

If models are understood as believable worlds, then the theoretical insights they produce are beliefs.11 The very category of belief is present in contemporary epistemology where the quality of understanding that the models offer is taken as a measure of their goodness (cf. Suarez 2010). It is important to note that understanding means having true beliefs about the world, or, more precisely, believing truths and not believing falsehoods. So, understanding is defined in terms of belief and not in terms of universal knowledge. Let me explain. Invocations to knowledge are much absent from contemporary epistemology (David 2001, p. 152). What contemporary epistemologists value most is connecting justification to the non-epistemic concept of truth. As Alston states it: Epistemic evaluation [justification] is undertaken from we might call ‘the epistemic point of view’. That point of view is defined by the aim at maximizing truth and minimizing falsity in a large body of belief […]. For a belief to be justified is for it, somehow, to be awarded high marks ­relative to that aim (emphasis added).12 alston 1985, pp. 83–84

The aim is usually defined in terms of searching for truth, however, with emphasis on searching, since arriving at universal knowledge that is free from doubts is hardly possible (cf. Lehrer 1990, pp. 20–38). While searching for the 11

12

Here the idea of models as believable worlds differs from the hypothetical one of models as possible worlds (cf. Nowak 1992, pp. 9–10). In the case of the latter, the theoretical claims built upon a given possible world “[do] not intend to speak about reality. A pure theory is just a picture of a possible world which does not actually exist” (Händler 1982, pp. 74–75, making this observation in his discussion of the empirical applicability of the general equilibrium theory in economics). In the case of the former, from the very beginning the model is constructed in such a way as to mimic its real target, so it intends to speak about reality. However, on the other hand, the Nowak’s claim that “[T]he smallest is the distance between the intended possible world of the kind and the actual world, the truer the counterfactual is” (1992, pp. 9–10) underlines some similarities between these ideas. But still more research is needed in comparing these ideas, since, for instance, the counterfactuals that are made in models of believable worlds are different from the ones of models of possible worlds (cf. theory of counterfactuals by D. Lewis, e.g., 1973). I would like to thank an anonymous referee for bringing the idea of possible world to my attention. This claim can be treated as a symptom of Alston’s general denial of deflationism in the theory of truth. However, an in-depth study on this issue is beyond the scope of my paper.

146

Hardt

truth we successively produce some often vague descriptions of the reality ­giving us insights into the way the world works (max truth) and does not work (min falsity). In the same vain we may say that the goodness of a model’s fit to the reality can be conceptualized in virtue of the extent to which a given model offers us a justified belief about the real world, namely the belief aiming at maximizing truth and minimizing falsity about the model’s target. If our model meets the MECH requirements and refers to the essentials of the ­modeled, then it is a believable world, and thus: ​DEF1 : Any model which meets the MECH requirement and refers to the ­essentials of the modeled is a believable world.​ It is worth noticing that meeting the MECH requirements only is not enough for a model to give us a good understanding of the way the world works (max truth and min falsity). The model must correctly make distinctions between essential explaining items (including mechanisms) and the ones of secondary importance. This is particularly important, since we may have a set of models (in MECH sense) that offer different explanations of the phenomena of interest. This is precisely the third question from the reminder of the previous section, namely what criteria one should apply in choosing the right model of price dispersion in case of having multiple models fulfilling the MECH conditions? But before focusing on this question, let me conclude my analysis on the ways economic models explain. Here I subscribe to the view that explaining means identifying the cause(s) of a given phenomenon. However, as Steel (2011, p. 122) claims: (M) X is a cause of Y if and only if there is a mechanism from X to Y. (M) is not intended as a universally true principle regarding causality, since there is presumably some “rock bottom” level of physical causation below which no mechanism lie. steel 2011, p. 122

So, models explain by depicting structures which enable the working of mechanisms, or models are just mechanisms’ descriptions. And such models of mechanisms produce beliefs about the real world and thus these beliefs are always true in models producing them (cf. “semantic view” on models). Here my answer to the question on how models explain refers closely to the Cartwright’s idea of models as blueprints of nomological machines that produce “insights” about regularities present in the actual world. Although the identification of a mechanism between X and Y is a necessary condition for explaining Y, it is not a sufficient condition for having a believable world (in sense of DEF1) of X–Y causal interplay. What is needed is high level of essesimilitude of such a model

Between Isolations and Constructions

147

(cf. essentials in DEF1) which is understood as likeness or closeness of model’s mechanism to its real counterpart (cf. Mäki 1991).13 This is what the method of isolation aims at in economics. However, economists often isolate so strongly that a phenomenon is isolated in its “pure” form without disturbing factors. As Niiniluoto (2002) rightly claims such situations are not “parts” of the real world and thus theoretical claims describing the model world are not true in its target. My solution of this problem is that they are partially true, namely they are beliefs about the target. Here, an interesting insight emerges from my analysis. It seems that models being constructed in such a way as to comply with DEF1 are inherently ­isolations. Although more in-depth research is needed to confirm such a proposition, the basics arguments in favor of it are based on the following: 1/a mechanism must have subparts and hence a researcher is obliged to identify a set of forces making the overall effect of the mechanism and consequently the risk of concentrating on the wrongly chosen single force is minimized; 2/the requirement that each component of mechanism is described by a generalization that is invariant under interventions guarantees that only influential forces are to be taken into account what immediately excludes false generalizations such as the one that manipulating the barometer is to cause the storm; 3/the independently changeable character of mechanism’s components ­assures the modeler not to include in the description (model) the forces that change their rules of work in the presence of other forces and also enables the modeler to assess the strength of each force, and thus the risk of taking into account some pseudo forces is minimalized; 4/the (iv) condition of MECH assures the model to be similar to the real in that sense that it enables summing up the effects of the workings of all forces and hence arriving at the overall effect of the mechanism. On a more philosophical level the justification for the MECH models ability to capture the essentials is due to the fact that in the real world mechanisms have the status of unobservables and thus their existence is only manifested on the empirical and factual levels of reality.14 Consequently, a given mechanism’s manifestation (in the real world) and a given mechanism’s 13

14

Although in philosophical literature the notion of essesimilitude refers usually to theories, I think that one may use it (at least analogically) in reference to models, if models are understood as entities producing theoretical claims (cf. Mäki’s (2012) insights on models and truth). Here I assume that models capturing the “essence” of the target are to give rise to theoretical claims about this very “essence” (cf. Niiniluoto 2002, p. 218). However, an in-depth study on this issue is beyond the scope of this paper. I refer here to the idea of three levels reality by T. Lawson (1997, p. 21), namely the empirical (experience and impression), the actual (actual events in addition to the empirical), and the real (structures, powers, mechanisms, and tendencies).

148

Hardt

description (in the model world) refer to the very same mechanism. However, the fit is perfect in the case of the actual world (manifestations are products of the true and acting mechanism) and imperfect in the case of the model (e.g., a model of mechanism does not capture all the ingredients of the mechanism operating in the real). But still the goal of explanation “is to posit a mechanism (typically at a different level to the phenomenon being explained) which, if it existed and acted in the postulated manner, could account for the phenomenon singled out for explanation” (Lawson 1997, p. 212). So, both in the real world and in the model the mechanism is somehow hidden – in the case of the former behind its manifestations and in the case of the latter behind its description.15 Now, the second question should be asked: how the mechanism described by the model relates to the one operating in the real world? Or, what is the nature of model-world inference? First, I agree with Hausman that the point of models in empirical science is to assist scientists in making claims about the world. In doing so, however, we do not test the model as such vis-à-vis the real world, but just “an application of a model, a hypothesis stating that certain elements of a model are approximately accurate or good enough representations of what goes on in a given empirical situation” (Guala 2005, p. 219). Next, he adds: “The fact that a model turns out not to work under certain circumstances does not count as a refutation of the model but only as a failed test of its applicability in a given domain” (Guala 2005, p. 220). The above is relatively intuitive: the closer a given empirical domain to the model’s structure is, the higher probability that the model’s insights are to correctly explain the workings of such a domain.16 However, the fit is never perfect and thus the idea of models’ insights as beliefs about the targets. But how the very act of making inferences from models in MECH sense to the real world looks like? Let me start by naming the mechanism responsible for price distortion (F, henceforth) and described in Varian’s 1980 paper as MECH1. Also, the set of regularities between variables of his model is depicted here as R (take, for instance, relations between model’s parameters as presented in Table 1 of Varian’s paper, e.g., the claim that more uninformed consumers 15

16

Cf. Steel’s (2011, p. 130) thesis that mechanisms are of central importance for learning about cause and effect in social sciences. This follows from his assertion that if there is a mechanism from X to Y, then X is a cause of Y. I put the term ‘probability’ in italics, since by using it I refer to the idea of ­verisimilitude – we are interested in theories (“models insight’s”) with high degree of verisimilitude (closeness to truth) (cf. Popper’s insights on the origin of the idea of verisimilitude (in contradistinction to probability) and its closeness to the idea of belief (as opposed to truth), e.g., Popper 1963/2002, p. 540).

Between Isolations and Constructions

149

cause the a­ verage price paid by the uninformed consumer to rise). Therefore, the structure of model-world inference is as follows: E1 – in the Varian’s 1980 model, F is caused by MECH1, and this model produces R; E2 – F occurs in the real world (“[…] the form of the resulting price strategy […] does not seem out of line with commonly observed retailing behavior” observes Varian 1980, p. 658); E3 – MECH1 operates in the real world (“[…] some aspect of economic life is noticed” says Gibbard and Varian 1978, p. 673); Therefore, there is reason to believe that: E4 – in the real world, F is caused by MECH1, and thus statements taken from R are believed to adequately describe the real world. In other words, they are beliefs about the target. Now, the interesting question is why knowledge about a mechanism for a causal relationship between shops’ behavior and price distortion can be inferred more reliably than the causal relationship as such between these two factors (cf. Reiss 2008b, p. 113)? It is particularly important, because economists quite often explain and make theory-world inferences on statistically inspecting patterns in empirical data-sets. However, such an approach is possible only if one has data, but also knowledge about possible confounders as well as the relationship between causality and probability. Practicing economists know well that such an econometrically informed way of doing economics is often hardly possible. This is also the case of Varian’s model – we do not have a statistical data analysis here but rather a conceptual exploration into various causes of price dispersion. This way of making model-world inferences not only diminishes the risk of doing economics without theoretical underpinnings, but can help econometricians by telling them, for instance, what kind of data is necessary for explaining given phenomena (cf. critique of theoretical emptiness of econometrics by McCloskey 1984). In this context, a must to be cited piece from economic literature is an interesting passage from Friedman and Schwartz monumental work on The Monetary History of the us: However consistent may be the relation between monetary and economic change, and however strong the evidence for the autonomy of monetary changes, we shall not be persuaded, unless we can specify in some detail the mechanism that connects the one with the other (emphasis added). friedman and schwartz 1963, p. 229

150

Hardt

What I should add is that an identified mechanism, e.g., the one Friedman refers to, does not enable us to learn about the real world itself. It does its job by giving us insights (beliefs) about the target. As it was stated earlier, the closer the model to the target is, the more accurate beliefs it produces are. So, the remaining question is how we can compare different MECH models of a given phenomenon. As it was just mentioned, models are used as producers of theoretical insights about their targets. So, in assessing models one in fact is to check to what extent theories brought upon by models survive transition from the world of the model to the real world. After all that have been said above, it should be clear that theory which is true inside the model is never (totally) true vis-àvis the real world. Let me refer again to the final part of the Varian’s paper, namely the conclusion that “Although this causal empiricism can hardly be conclusive, it suggests that the features of the model described here may have some relevance in explaining real worlds behavior” (p. 658). Before offering such a statement, Varian describes his own observations on how shops in his neighborhood area use sales. So, for him, his model should at least offer some theoretical insights explaining these initial empirical observations. However, what is needed is a systematic empirical investigation into the applicability of the model’s theoretical claims to a particular domain. It should be noted, however, that from the outset these claims are not to be suitable for every domain. So, a given model offers a specification of the conditions that make its insights potentially applicable in real situations. For instance, since in Varian’s model shops can freely set prices thus this model’s insights will be more appropriate in free markets (e.g., in the us) than in highly regulated market environments – it is for sure unreasonable to expect North Korean shops to behave in a manner described by Varian. So, the above mentioned statistical empirical investigations of the validity of a given theory should be conducted in domains at least slightly similar to the conditions of the model that was used to produce such a theory. It is thus unreasonable to test models’ insights in environments far beyond the ones specified by models’ structures. Now, let us look at the idea of the believable world by comparing it to the concept of a credible world by Sugden (2000). According to him such a world is significantly similar to the real one (Sugden 2000, p. 23). Also, a credible world can be understood as a description of how the world could be (Sugden 2000, p. 24) and credibility in models is like credibility in ‘realistic’ novels (Sugden 2000, p. 25). Next, credibility in economic models means that they are coherent (assumptions of models are not arbitrary chosen) as well as “they cohere with what is known about causal processes in the real world” (Sugden 2000, p. 26). However, in summing up his arguments Sugden concludes that “[he] cannot give anything remotely like a complete answer” (ibid.) to the question

Between Isolations and Constructions

151

of how a credible world should be defined. In his 2009 paper on Credible Worlds, Capacities, and Mechanisms he clarifies his ideas and claims the following: Credibility is not the same thing as truth; it is closer to verisimilitude or truthlikeness. We perceive a model world as credible by being able to think of it as a world that could be real […]. One crucial difference between a credible world and an isolation is that a credible world may be constructed around general empirical regularities – one might say, empirical laws – that are merely postulated. For all we know, these regularities may not be part of how the world really works. All that is required is that, in the current state of knowledge, they are credible candidates or truth. sugden 2009, p. 18

In the subsequent parts of his 2009 paper Sugden offers two sections – no 6 on models as isolating tools and no 7 on social mechanisms. Although he is quite ambivalent whether to accept Cartwright’s approach to isolation and Schelling’s ideas of social mechanisms, he somehow sees a need to conceptually link the ideas of credibility, isolation, capacity, and mechanism. Finally he concludes that we do not have such an overwhelming framework and hence “there is still a gap to be crossed and that requires inductive inference” (Sugden 2009, p. 26). If I am to compare now my idea of believable world to the one of credible world, I should start by saying that these two are not identical. Mine is more sharp and thus less general that the Sugden’s concept of credible words. Also, it is more applicable, since it is based on the MECH definition. Moreover, it attempts to resolve the problem of inductive leap in learning from models by claiming that gaining knowledge about the real world rests on assuming the similarity between the model and its target what guarantees the applicability of model’s insights to such a domain. But still more research on MECH models is needed in order to check the validity of this approach in accounting for the ways economists model the real world. In this context it is worth checking how the above developed ideas can help us in solving the Reiss’s (2012) paradox. This is done in the following section and serves as an assessment of the validity of my approach to economic modelling. 4

The Explanatory Paradox by Julian Reiss and an Attempt at its Resolution

Reiss’s 2012 paper brings together many issues philosophers of economics are interested in. It is particularly intriguing for those dealing with economic

152

Hardt

­models. Its value rests in showing that the great many contemporary philosophical approaches to economic models are themselves contradictory. Also, it raises the issues so fundamental that they should be of interest not only for philosophers of economics. The Reiss’s message is nicely given in the form of the following trilemma: (1) Economic models are false. (2) Economic models are nevertheless explanatory. (3) Only true account explain. My main point of disagreement is with the trilemma’s first point not because I claim the contrary, but because I think that economic models are neither true, nor false, but rather they aim at maximizing truth and minimizing falsity in a large body of belief about the real world. They are producers of such beliefs. So, truth and falsity mix in insights (beliefs) models give us about the real world and thus the idea of models as believable world. This leads also to my disagreement with the trilemma’s third point. I begin with the first claim of the paradox. Reiss saying that economic models are false does not mean that models have truth values, since according to him models are not sentences and only sentences are true or false, so when he speaks about truth and falsity of models he “speaks elliptically” (Reiss 2012, p. 49). Further he clarifies his views in the following way: “[…] the slogan ‘all models are false’ […] draw[s] attention to the undisputed fact that all models also misrepresent their targets in a myriad of respects” (Reiss 2012, p. 49). In what follows Reiss argues against Mäki’s way of resolving the problem, precisely his claim that a model can be true despite containing many falsehoods. So, for Mäki a model is always ‘false’ in many unimportant respects, but ‘true’ in what captures the causal factor of interest, e.g., the Earth’s gravitational pull in Galileo’s case. Reiss’s cites here the following passage from Mäki’s accounts of von Thünen’s model of the isolated state: If there is a natural truth bearer here, it is neither this model as a whole nor just any arbitrary parts of it. It is rather a special component of the model, namely the causal power or mechanism that drives this simple model world: the Thünen mechanism. This truth bearer has a fair chance of being made true by its truth maker, the respective prominent causal ‘force’ or mechanism in the real system. It is the mechanism that contributes to the transformation of distance into land use patterns through transportation costs and land values. mäki 2011, p. 60

Between Isolations and Constructions

153

So, in Mäki’s approach, models are not true as such but may contain truths about its various parts and most notably causal powers of mechanisms. However, Reiss denies the validity of ‘isolation by idealization’ tradition by claiming that economic models do not isolate in the Galilean sense for three reasons: 1/Galilean idealizations are absent in Galilean thought experiments;17 2/Galilean idealizations are quantitative, not categorical;18 3/Galilean idealizations have natural zero19 (Alexandrova and Northcott 2013, p. 263). Therefore, Reiss develops his final and fundamental criticism of Mäki’s approach first by claiming that “the models of economics […] are by and large very much unlike Galilean thought experiments [‘isolations by idealizations’]” (Reiss 2012, p. 51), and consequently “we do not know where to look for ‘truth in the model’” (Reiss 2012, p. 52). Mäki (2013) rejects Reiss’s denial of the possibility of isolating by idealizing in economics. For instance, he does not claim that having a natural zero is a necessary condition for an assumption to be of a Galilean kind. More fundamentally, in his research Mäki presents many cases of economic models that include many idealizing assumptions. The only thing Mäki acknowledges in regard to the Reiss’s criticism is a need for checking the role a given assumption plays in the model. If, for instance, one assumes a balanced state budget, then it may be interpreted as an assumption defining the cases in which this model can be applied – if you have a balanced budget, then you can use that model. We have a domain assumption here. On the other hand, saying that a budget is balanced may be interpreted that the state of the budget does not 17

18

19

To be honest I share with U. Mäki (2013, p. 274) difficulties in fully understanding the Reiss’s distinction between Galilean and non-Galilean assumptions, so in here presented clarifications I just offer some insights taken rather straightforwardly from Reiss (2012). While claiming that Galilean idealizations are absent in Galilean thought experiments, Reiss states that the factor assumed away does not normally appear in the latter: “The assumption of no air resistance cannot be read off the model […]. In contrast to, the non-Galilean assumptions Hotelling uses are all explicit part of the model, and they are assumptions without which no result could be calculated at all” (p. 51). According to him, the majority of assumptions economists are making in their models are similar in kind to the ones of Hotelling. Reiss (2012, p. 51) explains it as follows: “Galilean assumptions usually concern quantitative causal factors. Different media produce different degrees of resistance. Hotelling’s assumptions are categorical. Different geographies are different kinds of thing and not the same kind of thing to a different degree”. Here he states that “Galilean assumptions usually concern a causal factor that has a natural zero. No air resistance is such a natural zero” (Reiss 2012, p. 51), but the ones in economics, e.g., the geographical shape of the market, do not have natural zero.

154

Hardt

have any importance for the process we try to explain. So, in this case we have a negligibility assumption (Musgrave 1981). I share with U. Mäki his strong conviction that the method of isolation is not only possible, but that is the central method of science, including economics (Mäki 1992; 1994; 2009; 2011; cf. Cartwright 1989). The possibility of isolating in its broadest sense is a necessary condition also for my idea of a believable world, since without being able to isolate, it would be hardly possible to distinguish between mechanisms’ parts and identifying mechanisms as such would not be feasible. Now, let us come back to the first point of the Reiss’s trilemma but now in contrast with the idea of a model as a believable world (in sense of DEF1). Such models give raise to beliefs containing falsehoods as well as truths about their targets. So, the question is the following: do models as believable worlds have truth values? Does Mäki’s (2011, p. 60) claim that a mechanism inside a given model can be qualified as a truth bearer imply that models of mechanisms (in MECH sense) have truth values? I would not say that for at least two reasons. First, I do not believe in the ability of models to perfectly represent the mechanisms operating in the real world. Even I doubt whether we can have purely true and purely false elements in such models. What I claim is that we can only assess the extent to which a given model offers us a justified belief (not knowledge) about the real. This is a virtue of maximizing truth and minimizing falsity about the model’s target. So, each element of a given model is in some distance from truth. The aim of the modeler is to reduce this gap. Or, in other words, to include in mechanisms only these forces that are approximately correct and thus reducing the gap between the model’s mechanism and its real counterpart (cf. Hausman 2013, p. 253). For instance, in the case of Woodward’s (2002) block sliding down an inclined plane it means to include two forces only (­gravitational and the one due to frictions) and not some extra ones, e.g., the force exercised on the block by the wind.20 Such an approach guarantees also the stability of the explanation – the behavior it ­describes is stable across different environments, so robustness testing is possible in case of believable worlds (cf. Grüne-Yanoff 2013, p. 255). Thus, for believable worlds, the Reiss’s trilemma can be restated as follow:

20

Adding of such a force would break the condition of the independently changeable c­ haracter of generalizations governing each component of mechanism, since wind is ­definitively to impact the workings of the kinetic friction. So, MECH requirements ease the process of idealization.

Between Isolations and Constructions

155

(1′) Believable worlds offer insights that maximizes truth and minimizes falsity in a large body of belief about the real world. (2′) Believable worlds are nevertheless explanatory. (3′) Only beliefs that are candidates for true explain. The first above statement recapitulates what have been said earlier. The second follows from DEF1 (cf. discussion in Section 2). So, the third only requires now some more comments. Here, instead of Reiss’s ‘account(s)’ I refer to beliefs and ‘candidates for true’ replaces the simple true from the original trilemma. What remains is the invocation to explain, since even on briefly inspecting the most important economic models their authors usually put explanation as the main research goal while crafting and using models, e.g., the very last sentence from the Varian’s model is the following: “[…] it suggests that the features of the model described here may have some relevance in explaining real world retailing behavior” (1980, p. 658). Although the very notion of ‘accounts’ from Reiss’s trilemma is imprecise, the detailed reading of his paper suggests that by true accounts he understands first credible worlds and next he checks whether the (3) holds for models as unifying entities (in the sense of Kitcher’s (1981) accounts of unification). He concludes that neither credible worlds, nor the letter approach can account for models as being explanatory, since they are not true and only true account explains. Thus, according to him, the paradox remains. In my formulation of Reiss’s (3) the emphasis is put on saying that it is the justification of the belief that makes it a candidate for true and the highly justified beliefs are the ones that maximizes truth and minimizes falsity about the model’s target. Importantly then is to conclude that we cannot have a purely true belief, or, a model perfectly representing the real, but what we should aim at is the model whose structure and its parts (together giving raise of mechanism) are as close as possible to the characteristics of its target. For instance, in case of the Varian’s model of sales what explains is a set of regularities between variables of his model and these regularities are beliefs about what is going on in the real world. The model-world inference is thus due to the identification and subsequent description of the mechanism responsible for price dispersion and it proceeds from E1 via E2 and E3 and then via inductive leap to E4 (see, Section 3). As it was explained in Section 2, beliefs based on regularities produced by models do not have a status of purely true descriptions of mechanisms operating in the real, but rather these descriptions are as close as possible to their empirical ideals. Thus, (1′), (2′), and (3′) are all true and hence mutually consistent and thus the paradox is resolved.

156

Hardt

5 Conclusions The goal of this paper was to shed some light on the dichotomical treatment of models as isolations and constructions. In doing so I was able to show that what characterizes the great many of economic models is that they are not (only) idealizations, nor (only) constructions, but believable worlds, precisely models of mechanisms giving us justifiable beliefs about the way the world works. The mechanisms are defined here using the ideas of Woodward (2002). The example of Varian’s model of sales was presented as a typical case of such a model. Also, the question on how models explain was answered. It was shown that models explain by producing theoretical insights that are always true within models but they are just beliefs if claimed to accurately describe the real world. Thus, such beliefs are more credible if the target is close enough to the model’s structure. This enabled me also to shed some light on the issue of model-world inferences. The validity of my approach was tested by using it in an attempt at resolving the Reiss’s paradox of explanation. It was done in by reformulating it in the following way: 1/models are not true or false but rather they maximize truth and minimize falsity in a large body of belief about the real world; 2/such beliefs can never function as perfectly true but rather as candidates for true; 3/finally, these models are explanatory. The ability of believable worlds to explain lies in the fact that they refer to mechanisms that operate in the real world. However, more research is needed in order to further justify the claim that constructing models of mechanisms (in Woodward’s sense) raises the probability of identifying candidates for crucial mechanisms and hence explaining economic phenomena well. This paper offers same basic ideas supporting such an assertion. Also, the concept of economic models as believable worlds better suites with the claim that in economics we do not have universal laws but rather Millian tendency laws (capacities) or empirical regularities of no-law status. In this context, a lot more in terms of philosophical refection need to be done in order to refer the here proposed ideas to the one of models as blueprints of nomological machines (Cartwright 1999). I think that the research endeavors sketched above can enhance our knowledge about the status of models in economics. This paper is an invitation to take up this challenge. Acknowledgements The research here presented was financed by a grant from the National Science Centre of Poland (grant no 2011/01/D/HS4/03829). I would like to thank

Between Isolations and Constructions

157

anonymous referees for their helpful comments. Also, I thank prof. F. Guala for stimulating discussions that helped me to shape the final version of the paper. References Alexandrova, A. (2006). Connecting Economic Models to the Real World: Game Theory and the FCC Spectrum Auctions. Philosophy of the Social Sciences 36,173–192. Alexandrova A. and R. Northcott (2013). It’s Just a Feeling: Why Economic Models do not Explain? Journal of Economic Methodology 20 (3), 262–267. Alston, W.P. (1985). Concepts of Epistemic Justification. Monist 68 (1), 57–89. Boumans, M. (2004). Models in Economics. In: B. Davis, A. Marciano, J. Runde (eds.), The Elgar Companion to Economics and Philosophy, pp. 260–282. Cheltenham: Edward Elgar. Boumans, M. (2005). How Economists Model the World to Numbers. London: Routledge. Cartwright, N. (1983). How the Laws of Physics Lie. Oxford: Oxford University Press. Cartwright, N. (1989). Nature’s Capacities and their Measurement. Oxford: Clarendon Press. Cartwright, N. (1998). Capacities. In: J.B. Davis, D.W. Hands, U. Mäki (eds.), The Handbook of Economic Methodology, pp. 45–48. Cheltenham: Edgar Elgar. Cartwright, N. (1999). The Dappled World: A Study of the Boundaries of Science. ­Cambridge: Cambridge University Press. Cartwright, N. (2009), If No Capacities Then No Credible Worlds. But Can Models ­Reveal Capacities? Erkenntnis 70 (1), 45–58. Chakravartty, A. (2007). A Metaphysics for Scientific Realism: Knowing the Unobservable. Cambridge: Cambridge University Press. David, M. (2001). Truth as the Epistemic Goal. In: M. Steup (ed.), Knowledge, Truth, and Duty, pp. 151–170. New York: Oxford University Press. Earman, J., Glymour, C., Mitchell, S. (eds.) (2002). Ceteris Paribus Laws. Dordrecht: Springer Netherlands. Friedman, M. and A. Schwartz (1963). A Monetary History of the United States, 1867– 1960. Princeton: Princeton University Press. Gibbard, A. and H.R. Varian (1978). Economic Models. The Journal of Philosophy 75 (11), 664–677. Giere, R. (1988). Explaining Science: A Cognitive Approach. Chicago: University of ­Chicago Press. Glennan, S. (1996). Mechanisms and the Nature of Causation. Erkenntnis 44, 49–71. Grüne-Yanoff, T. (2013). Genuineness Resolved: A Reply to Reiss’ Purported Paradox. Journal of Economic Methodology 20 (3), 255–261.

158

Hardt

Guala, F. (2005). The Methodology of Experimental Economics. Cambridge: Cambridge University Press. Händler, E.W. (1982). The Evolution of Economic Theories. A Formal Approach. Erkenntnis 18 (1), 65–96. Hardt, Ł. (2011). An Inquiry into the Explanatory Virtues of Transaction Cost Economics. The Journal of Philosophical Economics 5 (1), 120–147. Hausman, D. (2012). Mäki’s MISS. In: A. Lehtinen, J. Kuorikoski, P. Ylikoski (eds.), Economics for Real: Uskali Mäki and the Place of Truth in Economics, pp. 81–95. New York: Routledge. Hausman, D. (2013). Paradox Postponed. Journal of Economic Methodology 20 (3), 250–254. Kincaid, H. (2012). Some Issues Concerning the Nature of Economic Explanation. In: U. Mäki (ed.), Philosophy of Economics, pp. 137–158. Amsterdam: North-Holland. Kitcher, P.(1981). Explanatory Unification. Philosophy of Science 48 (4), 507–531. Lewis, D. (1973), Counterfactuals. Cambridge: Harvard University Press. Lawson, T. (1997). Economics and Reality. London: Routledge. Lehrer, K. (1990). Theory of Knowledge. London: Routledge. Machamer, P.K., Darden, L., Craver, C.F. (2000). Thinking about Mechanisms. Philosophy of Science 67(1) 1–25. Mäki, U. (1991). Comment on Hands. In: N. de Marchi and M. Blaug (eds.), Appraising Economic Theories: Studies in the Methodology of Research Programmes, pp. 85–90. Aldershot: Edward Elgar. Mäki, U. (1992). On the Method of Isolation in Economics. In: C. Dilworth (ed.), Idealization IV: Intelligibility in Science, Poznan Studies in the Philosophy of the Sciences and the Humanities vol. 26, pp. 317–351. Amsterdam-Atlanta: Rodopi Mäki, U. (1994). Reorienting the Assumptions Issue. In: R. Backhouse (ed.), New Directions in Economic Methodology, pp. 236–256. London: Routledge. Mäki, U. (2001). Models. In: N. Smelser and N. Baltes (eds.), International Encyclopedia of the Social and Behavioral Sciences 15, 9931–9937. Mäki, U. (2006). Remark on Models and their Truth. Storia del Pensiero Economico 3, 7–19. Mäki, U. (2009). MISSing the World. Models as Isolations and Credible Surrogate Systems. Erkenntnis 70 (1), 29–43. Mäki, U. (2011). Models and the Locus of Their Truth. Synthese 180, 47–63. Mäki, U. (2012). The Truth of False Idealizations in Modeling. In: P. Humphreys and C. Imbert (eds.), Models, Simulations, and Representations, pp. 216–233. New York: Routledge. Mäki, U. (2013). On a Paradox of Truth, or How not to Obscure the Issue of whether Explanatory Models can be True. Journal of Economic Methodology 20 (3), 268–279.

Between Isolations and Constructions

159

McCloskey, D. (1984). The Loss Function has been Mislaid: the Rhetoric of Significance Tests. American Economic Review 75 (2), 201–205. Mill, J.S. (1836). On the Definition and Method of Political Economy. In: D. Hausman (ed.) (1994). The Philosophy of Economics. Cambridge: Cambridge University Press. Morgan, M. (2012). The World in the Model: How Economists Work and Think. Cambridge: Cambridge University Press. Morgan, M. and M. Morrison (1999). Models as Mediators: Perspectives on Natural and Social Science. Cambridge: Cambridge University Press. Morgan, M. and T. Knuuttila (2012). Models and Modeling in Economics. In: U. Mäki (ed.), Philosophy of Economics. Handbook of the philosophy of science 13, 49–87. Mumford, S. (2013). The Power of Power. In: R. Groff and J. Greco (eds.), Powers and Capacities in Philosophy. The New Aristotelianism, pp. 9–24. New York: Routledge. Musgrave, A. (1981). Unreal Assumptions in Economic Theory: the F-Twist Untwisted. Kyklos 34, 377–387. Niiniluoto, I. (2002). Truthlikeness and Economic Theories. In: U. Mäki (ed.), Fact and Fiction in Economics, pp. 214-228. Cambridge: Cambridge University Press. Niiniluoto, I. (2012). The Verisimilitude of Economic Models. In: A. Lehtinen, J. Kuorikoski, P. Ylikoski (eds.), Economics for Real: Uskali Mäki and the Place of Truth in Economics, pp. 65–80. New York: Routledge. Nowak, L. (1992). The Idealizational Approach to Science: A Survey. In: J. Brzeziński and L. Nowak (eds.), Idealization III: Approximation and Truth. Poznań Studies in the Philosophy of the Sciences and the Humanities 25, pp. 9–63. Amsterdam: Rodopi. Nowak, L. (1994). The Idealization Methodology and Economics. In: B. Hamminga and N. de Marchi (eds.), Idealization in Economics. Poznań Studies in the Philosophy of Sciences and the Humanities 38, pp. 303–336. Amsterdam: Rodopi. Pissarides, C.A. (2000). Equilibrium Unemployment Theory. Cambridge: MIT Press. Popper, K. (1963/2002). Conjectures and Refutations: The Growth of Scientific Knowledge. New York: Routledge. Reiss, J. (2008a). Social Capacities. In: S. Hartmann, C. Hoefer, L. Bovens (eds.), Nancy Cartwright’s Philosophy of Science, pp. 265–288. New York: Routledge. Reiss, J. (2008b). Error in Economics: The Methodology of Evidence-Based Economics. London: Routledge. Reiss, J. (2012). The Explanation Paradox. Journal of Economic Methodology 19 (1), 43–62. Salmon, W. (1989). Four Decades of Scientific Explanation. Minneapolis: University of Minnesota Press. Salop, S. and J. Stiglitz (1977). Bargains and Ripoffs: A Model of Monopolistically ­Competitive Price Dispersion. Review of Economic Studies 44, 493–510. Schelling, T.C. (1978). Micromotives and Macrobehaviour. New York: Norton.

160

Hardt

Schmidt-Petri, Ch. (2008). Cartwright and Mill on Tendencies and Capacities. In: S. Hartmann, C. Hoefer, L. Bovens (eds.), Nancy Cartwright’s Philosophy of Science, pp. 291–302. New York: Routledge. Shilony, Y. (1977). Mixed Pricing in Oligopoly. Journal of Economic Theory 14, 373–388. Steel, D. (2011). Social Mechanisms and Causal Inference. In: F. Guala and D. Steel (eds.), The Philosophy of Social Sciences Reader, pp. 119–131. New York: Routledge. Suarez, M. (2010). Fictions, Inference, and Realism. In: J. Woods (ed.), Fictions and Models: New Essays, pp. 225–246. Munich: Philosophia Verlag. Sugden, R. (2000). Credible Worlds: the Status of Theoretical Models in Economics. Journal of Economic Methodology 7 (1), 1–31. Sugden, R. (2009). Credible Worlds, Capacities and Mechanisms. Erkenntnis 70 (1), 3–27. Suppe, F. (1989). The Semantic View of Theories and Scientific Realism. Urbana and Chicago: University of Illinois Press. Tversky, A. and D. Kahneman (1981). The Framing of Decisions and the Psychology of Choice. Science 211 (4481), 453–458. van Fraassen, B.C. (1980). The Scientific Image. Oxford: Oxford University Press. Varian, H.R. (1980). A Model of Sales. The American Economic Review 70 (4), 651–659. Varian, H.R. (1999). How to Build an Economic Model in your Spare Time. In: M. Szenberg (ed.), Passion and Craft: Economists at Work, pp. 256–271. Ann Arbor: University of Michigan Press. Woodward, J (1999). Causal Interpretation in Systems of Equations. Synthese 121, 199–257. Woodward, J. (2000). Explanation and Invariance in the Special Sciences. The British Journal for the Philosophy of Science 51, 197–254. Woodward, J. (2002). What Is a Mechanism? A Counterfactual Account. Philosophy of Science 69, 366–377. Woodward, J. (2005). Making Things Happen. Oxford: Oxford University Press. Ylikoski, P. and J Kuorikoski (2010). Explanatory Power. Philosophical Studies 148 (2), 201–219. Zeidler, P. (2013). Models and Metaphors as Research Tools in Science. Berlin: Lit Verlag.

chapter 7

The Fallacy of the Reification of Idealization in Economic Research Adam Czerniak Abstract The reification of idealization, i.e. using sophisticated formalized models without any proper concretization, is an immanent feature of economic as a science. When its existence is ignored it can cause large damage to economic research as well as to real economy through production of bad guidelines for policy makers. In this paper I argue that this flaw of economic methodology contributed to the recent global financial crisis. I show the consequences of applying econometric models such as Value-at-Risk (VaR) to account for systematic risk and market meltdowns. Furthermore I claim that large scale idealized econometric models like DSGE are usually not suitable to identify economic imbalances and analyze the effects of monetary policy on macroeconomic stability.

1

Financial Models and Market Crashes

Before the 1950’s the dominant approach in financial economics was old-­ fashioned institutional analysis (MacKenzie 2006). At that time the use of mathematical models was limited mainly to classical macroeconomics (Mirowski 1988; Birner 1994). In finance, the methodological revolution started in the late 1950’s and transformed this science into a theoretical, analytical and increasingly mathe­matical branch of economics. The shift had three main strands: (1) the investigations of Markovitz on the asset portfolio returns and risk; (2) the research of Modigliani and Miller on the dependence of capital structure and companies values; (3) the propositions formulated by Eugene Fama on stock-price movements. Those pioneers of modern finance introduced basic mathematical models and lay down methodological fundaments for the modern theory of finance. For the purpose of this article I will skip a precise description of their research and concentrate purely on the description of the methodology that they used. 1.1 The Methodological Foundations of Modern Financial Models Markovitz (1952) took up the problem of optimal asset portfolio choice. He stressed that, for an investor, it is more important to know the expected value © koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_009

162

Czerniak

of an asset portfolio then the expected value of a given stock, since the probability distributions of different stocks are usually interdependent. Furthermore he argued that the optimal portfolio choice is not only about expected value but also about risk, which Markovitz defined as the variability of expected portfolio returns (MacKenzie 2006). Hence he came up with a formal model of asset portfolio choice in which he combined the risk and return of an asset portfolio with the investors attitude towards risk. The two basic implicit assumptions of this approach were: (1) there exist a finite and constant variance of asset price fluctuations; (2) there exists a constant (in a given period of time) relationship between the distributions of asset returns. Both assumptions can be seen as intuitively true or possibly true and hence they have not been widely discussed in the works of Markovitz and his followers. However, as statistical evidence suggest (see below), both assumptions are not met in financial markets. So – from a methodological point of view (Nowak 2000) – these assumptions are pure idealizing conditions and are not merely used for technical idealization purposes. The works of Markovitz are also important for the development of financial research for a second reason. He was one of the first to use pure matrix algebra to solve problems of economic origin. He used it for the computation of correlations between changes in asset prices, introducing a whole new method of analyzing interdependence of asset returns. As Milton Friedman – a member of Markovitz PhD thesis defense commission – stated: “What he did was a mathematical exercise, not an exercise in economics” (MacKenzie 2006, p. 50). Modigliani and Miller (1958) analyzed company-value determinants as if they were functioning in a “perfect capital market”. This was a methodological breakthrough. Years before, a vast majority of financial academics concentrated on real-life case studies, in which they described the dependence of valuation, returns, dividends and other financial measures of companies’ efficiency on institutional and economic factors. Modigliani and Miller proposed to use the theoretical and highly idealized concept of a “perfect capital market”, i.e. a situation in which there are no transaction costs sensu largo – taxes, provisions, liquidity premia, bankruptcy costs etc. (Cools et al. 1994). Such an approach was directly borrowed from mainstream economics, where idealizedmarket concepts and models were widely used in sensitivity analysis since Alfred Marshall published his textbook “Principle of Economics”. Presumably, the most influential concept in the theory of finance, that also influenced mainstream macroeconomics, was the efficient market hypothesis formulated by Eugene Fama (1970). It was based on research that could be traced back as far as to the works of Bachelier from 1900 and Wiener from 1923 (MacKenzie 2006). Both of them stated that stock prices move in line with

The Fallacy of the Reification of Idealization

163

a stochastic process similar to Brownian motion from physics. Hence, information on past price changes cannot influence future changes and cannot be used for profitable speculation. A second supposition that Fama built on was the old conventional wisdom that “If one could be sure that a price will rise, it would have already risen” (Samuelson 1965, p. 41). Binding those two suppositions together Fama proposed the efficient market hypothesis. Putting it simply, according to this hypothesis, financial markets are “efficient”, i.e. asset prices fully reflect all the information that is available to investors and, hence, price movements that are not driven by new information publications, are of a purely random nature. Fama opened-up a totally new field of research in which asset prices could be modeled with the use of sophisticated tools developed by theoretical statisticians. The assumption that had to be made to introduce formal modeling into financial economics was very simple – all changes in asset prices that are not driven by changes in fundamental factors (i.e. new information) can be described by a stochastic random walk process, usually of a Brownian nature. Eventually new models of asset prices, market equilibria, optimal portfolio choice etc. started to dominate the theory of finance replacing descriptive methods of institutional analysis which was a methodological U-turn for this field of economics. The shift in methodology attracted a large number of mathematicians, statisticians and physicists to analyze financial phenomena as they were of an algebraic nature. The problems began when theoretical financial models started to be applied in practice. In many cases, idealized assumptions that allowed formal modeling in the theory of finance were considered as theorems. This lead to the fallacy of the reification of idealization in economic finance, as many market participants used theoretical models to construct their optimal asset portfolios and estimate prices of assets and derivatives without proper concretization methods. As research shows, this was one of the reasons for the severity of the market meltdowns that have happened since the mid 80’s in the world economy (i.a. MacKenzie 2006; Knorr Cetina, Preda 2005). In this paper I focus only on one example – a model that is used, usually without any concretization procedure, by financial institutions, the so called Value-at-Risk (VaR) model. 1.2 Value-at-Risk Models and Their Assumptions VaR methods are now used universally by banks and hedge funds to measure market risk (imf 2007). They are also used for setting upper limits on risk ­exposure by financial institutions, for internal risk control purposes and for financial supervision tests (Bałamut 2002). Furthermore, the VaR model is an

164

Czerniak

archetype for constructing more advanced risk measurement techniques that have been developed in recent years. The basic VaR model was developed in the 1980’s and became a mainstream risk measurement practice after JP Morgan released its RiskMetricsTM system in 1994. It has also been included into the Basel ii Accord as a standard technique to control risk exposure. Technically, VaR models are used to compute a firms asset portfolio loss that will not be exceeded within a given time horizon at a certain confidence level (Alexander 2008). Hence a VaR model has two basic parameters: α – the significance level; and h – the modeling horizon. The core of the VaR model is the profit and loss function (P&L), which can be defined as follows: ​L​(​Pt​ ​ , ​Xt​ ​)​ = v​(​Pt​ ​)​ − v​(​Pt+h ​  ​)​ = v​(​Pt​ ​)​ − v​(​Pt​ ​*  exp​(​Xt​ ​)​)​​

(1)

where P is the vector of specific values that can be transformed into real asset prices through the valuation function v(P), and X is a vector of possible returns from the asset portfolio. The distribution of the L function is usually called the P&L distribution. Accordingly, the value at risk is the α-quantile of the P&L distribution (see Figure 7.1.). For instance, if a firm’s one-day estimated VaR is $9.3 million, at a confidence level of 99% (i.e. significance level of 1%) this implies that a loss of $9.3 million or greater is expected on one trading day out of 100. The above described concept of VaR modeling is in fact very simple. ­However problems arise when it comes to using it in practice, because of five implicit assumptions that have to be made to compute the value-at-risk figure. I list them below: 1. The P&L density function gets a finite variance, i.e. the tails of the P&L density function are relatively thin. 2. The P&L density function is constant over time and independent between periods. 3. Correlations between changes in asset prices in a given portfolio are constant. 4. All assets are traded in a perfect market. 5. Financial markets are efficient, which means that all uncertainty can be treated as risk and described by a P&L distribution. Some of these assumptions might seem technical, however, in the light of empirical research on the nature of financial markets, they have to be seen as idealization assumptions. VaR modeling only describes the underlying relations between risks and returns and not a real market situation. The real dependence

The Fallacy of the Reification of Idealization

165

Figure 7.1  Illustration of the VaR metric Source: Alexander 2008

between portfolio choice and possible loss are thus much more complex than the VaR approach implies. Below I discuss each of the above iterated assumptions and show what effect on risk assessment they have. Ad 1. P&L density functions are usually estimated with the use of a finite sample of recent observations of asset price changes. Using such data, practitioners compute a sample mean and variance and assume that they are the first and second moment of a given probability distribution. Usually it is assumed that the density function is normal (Blommestein et al. 2010) – which can be seen as a legacy of the works of Bachelier, Wiener and Fama. However, as Mandelbrot (1963) proved, such assumptions can be misleading, because in real markets the population moments are very often infinite. In such a world, extreme events like market crashes happen much more often than is implied by the normal distribution function as used in standard financial risk measurement models. Anecdotal evidence supports this point of view. The 19th October 1987 futures market crash was an event whose probability was estimated at 10−160 (MacKenzie 2006), and during the 2007 credit market meltdown, Goldman Sachs employees “observed” an event that under a n ­ ormal

166

Czerniak

­distribution assumption occurs approximately once every 14*1010 years, i.e. fourteen times the age of our universe (Danielsson 2008). Furthermore, in the past 100 years stock prices have fallen, in one day, by 7% 48 times, even though, under the normal distribution assumption, such an event happens, on average, once every 300,000 years (Bervas 2006). Using standard VaR models without proper concretization will lead to an underestimation of market meltdown probability. Ad 2. To forecast asset portfolio risk, one has to assume the nature of the relationship between past observations and future price changes. In the VaR approach, as in standard econometric models, it is assumed that density functions do not change with time – i.e. past observations on asset prices can deliver statistically robust inferences relating to the probability of future price changes (Blommestein et al. 2010). This is rarely true in financial markets, especially during financial shocks. Market volatility tends to vary in time substantially – e.g. in periods of booms and busts, the stock index volatility is more than three times higher than in periods of moderation (Danielsson 2002). The importance of this assumption increases with higher h and lower α parameters in the VaR model. The more distant and less frequent events the model has to predict, the more data is needed to estimate the P&L distribution. For example, if data from a one year interval is used to estimate a 95% value-at-risk figure, ten years of data are needed to estimate a 10 day 95% value-at-risk figure with the same statistical significance, and for a 10 day 99% value-at risk figure the data requirements increase to fifty years of observation. Hence, the assumption that over such a long period of time the P&L distribution has remained constant is unrealistic. Such properties of VaR models lead to a severe procyclicality of risk indicators – in times of moderation the VaR model underestimates the probability of future losses and in times of market tensions the risk is overestimated. Ad 3. In VaR modeling it is assumed that past correlations between price changes are an unbiased indicator for future dependence between the values of assets in a given portfolio. This is a very strong assumption. As empirical research shows, correlations between asset prices differ substantially between booms and bust periods. For example, the average correlation between German and us stock prices in times of bull markets equals 8.6% and in times of bear markets is as high as 52% (Erb et al. 1994). These changes in correlation are mainly caused by financial market imperfection and inefficiency – both factors are discussed in the next two paragraphs. However, even in a perfect and efficient financial market, the correlations between asset prices can vary due to macroeconomic factors. Let’s take the example of mortgage default risk. When computing the VaR figure, the default probabilities of different mortgage loans are taken as independent events, conditional only on the individual

The Fallacy of the Reification of Idealization

167

characteristics of the borrowers. This is true in times of an economic boom, but not necessarily when the economy is contracting. In such a case mortgage defaults become highly correlated even among loans granted to different economic groups in different geographical regions (Danielsson 2008). The same applies to stock prices or the yields of emerging market bonds – the risks for all assets can be independent on a micro level but are usually highly correlated when macroeconomic factors are taken into account. Furthermore, in some cases, credit contagion is also possible. Due to the macroeconomic interdependence of different market participants an increase in defaults can cause a snowball effect and lead to a massive decrease in asset valuations (Egloff et al. 2003). Such systemic risks are usually not included in typical VaR models, which leads to an underestimation of the value-at-risk estimation in times of economic crisis. Ad 4. The assumption that goods are traded in a perfect financial market is a typical idealization assumption that has been used in theoretical models since the works of Modigliani and Miller. In VaR models the perfect market assumption implies that: (1) there are no costs for selling assets – i.e. no taxes, no provisions, no spreads between bid and ask prices; (2) there are no liquidity problems – buying or selling an indefinitely large asset portfolios does not affect the market price of these assets; (3) the first Goodhart law1 does not apply. All these three properties of financial markets have to been taken into account in risk management. Firstly, taxes, provisions and spreads can increase the cost of selling assets, especially during periods of financial tensions when the bid/ ask difference can rise substantially. As a result the expected real loss is higher than under perfect market assumption conditions. Secondly, in periods of increased risk aversion, financial markets can dry up causing liquidity problems for portfolio managers. In such a case, the amount of assets an investor is willing to sell or buy can substantially affect the market price of those assets (Jarrow 2005). For example, if liquidity is low, to sell a large basket of stocks one has to offer them at below-the-market price. Hence, the possible losses during a financial crisis will be higher than implied by VaR models that are computed on the basis of market prices. Thirdly, the most complex violation of the perfect market assumption is the relationship between models and markets. Usually economists assume that the models they use do not affect the phenomena they are trying to explain. However, according to Goodhart law: if models are based on statistical relationships (e.g. Philips curve), using them for policy purposes affects the 1 Using stylized facts for policy purposes does not affect the observed statistical regularities (Chrystal et al. 2003).

168

Czerniak

analyzed phenomena and breaks down the observed relationship. Also, as Danielsson (2002) argues, if a risk model, such as VaR, is used for risk management purposes, then it breaks down. This happens because different institutions base their decisions on the same VaR figures, which increases the correlation between otherwise independent asset prices and influences the future P&L distribution function – in times of market meltdown, investors which use VaR models tend to sell the same assets, driving down their prices more than historical volatility distribution would imply. One could of course imagine a situation in which different institutions use different risk management ­models – however such a situation is not Pareto-optimal. According to the terminal isomorphism theorem (Pozner et al. 2010), there is only one risk management strategy that will be used by market participants in equilibrium. If every investors’ goal is to maximize his profits and he has an upper limit of permissible risk – most probably set by a regulatory institution – then in the long-term all market participants will use only one risk model that allows them to hold the most profitable asset portfolio. As a result, endogenous risk emerge. Hence, VaR models, which are based on perfect market assumptions, tend to underestimate the real risk of an asset portfolio. Ad 5. At a very fundamental level, in almost every asset price model it is assumed that all uncertainty can be measured in a reliable fashion using a probability distribution (Blommestein et al. 2010). This is a legacy of the efficient market hypothesis, which implies that all the known up-to-date information is included in asset valuation. As a result, the remaining moves in asset prices are pure white noise, which of course can be approximated by a distribution function. This is an idealizing assumption, because as Shiller (1981) proved, asset price volatility is higher than the volatility of the underlying firm’s dividends, which under the efficient market hypothesis is impossible (Fama 1970). The correlation between past and future price changes can be attributed to the existence of animal spirits or to use of availability heuristics by investors. Empirical research (Peterson 2007) shows, that in times of booms, investors tend to underestimate the impact of new negative information and overestimate the impact of new positive information on fundamental asset prices. As a result, asset prices can become highly correlated during turning points in financial markets when at the same time a large amount of investors realize that their former convictions are improper and thus change their investment strategy. Furthermore, uncertainty can influence market participants’ behavior, not only on the micro, but also on the macro level – institutional and systemic uncertainty, such as the relationship between new financial derivatives and linkages between different financial markets, can also influence asset prices and cause the underestimation of VaR figures.

The Fallacy of the Reification of Idealization

169

1.3 The Fallacy of Reification and Concretization Attempts As shown above, the idealizing of assumptions of the VaR model causes that the forecasted value-at-risk figure of an asset portfolio is usually lower than in reality. Hence, using the VaR approach without proper concretization methods can lead to a fallacy of the reification of idealization, i.e. to an underestimation of the real risk of a given portfolio. However, risk models do have a valuable function so long as their limitations are recognized – for example, they are useful in managing the risk of small, uniform portfolios, but not for computing VaR figures for large divisions, financial institutions or for whole financial systems (Danielsson 2008). Their usefulness is also limited in times of market meltdowns, when the underestimation of potential loss tends to increase. The VaR models and their idealizing assumptions can be compared to the Galilean law of falling bodies, which states that any object will fall with uniform acceleration under the idealizing assumption that the resistance of the medium through which it was falling does not exist. In such a case the Galilean law could be applied to compute the real velocities of falling bodies unless medium frictions were negligible – this is the so-called Approximate Inference Assumption (Pemberton 2005). The same assumption is also used when VaR models predictions are used for forecasting real financial events. However the Approximate Inference Assumption is sound only when VaR models are used to measure day-by-day risks of frequent small events – in such a case the implications of violating the idealizing assumptions are negligible. Since the invention of VaR models, a large number of scientists have tried to overcome the problems connected to the limited use of this risk measurement technique. These VaR concretization attempts can be divided into four groups, each corresponding to one of the assumptions mentioned in Section 3.2. Ad 1. Two important ideas have been developed on how to overcome the problem of an infinite variance of the P&L distribution. One of these, proposed by Mandelbrot et al. (1997), is the multifractal model of asset returns, which can be computed regardless of the shape of the P&L distribution ­function – even in cases of Lévy-stable distributions (i.e. those with an infinite variance). Also other similar attempts to take into account “fat” and “super fat” tails of P&L distributions in the VaR modeling have been undertaken (imf 2007; Bałamut 2002). The second idea for concretization of the VaR model is the expected shortfall (es) approach (Artzner et al. 1999). The es is the conditional expectation of loss given that the VaR number is exceeded. After the VaR number is computed with standard methods, the distribution of losses beyond the VaR threshold is estimated. In this way the thickness of the tail of the initial P&L distribution can be better approximated. Both ideas, however, have the disadvantage that more data is needed to properly estimate the shape

170

Czerniak

of the tail of the distribution function. As Bams et al. (2005) show, models with more sophisticated tail behavior lead to more parameter uncertainties, which in turn lead to greater uncertainty in the predicted VaR numbers. Ad 2. The most developed part of VaR research is on how to allow the P&L distribution function to change with time (Bałamut 2002; Danielsson 2008). Methods such as the exponentially moving average, normal or fat-tailed garch allow for smooth changes to the forecasted volatility levels. However, if the parameterization and estimation horizon of the VaR models after a volatility shock remains the same, the model has a very similar steady state volatility, which can be updated only if the high volatility period is sufficiently long (Danielsson 2002). Thus, such sophisticated models do not eliminate – especially in times of financial market meltdowns – the value-at-risk bias connected to the second idealizing assumption of VaR models. Ad 4. There have also been several attempts to incorporate the risk of financial market liquidity dry-up into risk management models (Bervas 2006). Usually in this approach a statistical relationship between market liquidity and the transaction costs of selling asset prices is estimated. Subsequently some aposteriori assumptions on the probability distribution of market liquidity levels are made. Compiling those two together, one can approximate the probable cost of selling asset prices which in turn can be implemented into the standard VaR modeling. This statistical method helps to overcome the bias related to transaction costs and liquidity problems, however it does not help to account for the Goodhart law problem. Because using simple backward induction one can prove that taking into account transaction costs, liquidity levels and possible decisions taken by other market participants, increases global fire sales during financial crises, driving up asset price correlations. In an extreme case, VaR models that account for the relationship between risk management strategies and market liquidity would assign high loss probability to all possible assets and hence would become useless for risk management. Ad 5. Only the conditions from 1 to 5 can be replaced by realistic negation in the form of mathematical formulas. The last idealizing assumption is more of a qualitative kind, because uncertainty cannot be quantified ex definitione. To account for unpredictable risks one has to use institutional analysis and sociological methods which can point to unprecedented events that cannot be forecasted by econometrical methods. Regarding the VaR models such qualitative methods had already been used to show how value-at-risk figures have been underestimated (e.g. Danielsson 2002). All the above described concretization methods are used quite rarely by market participants – financial institutions and supervisory commissions. This is because such procedures usually require larger data sets that are rarely

The Fallacy of the Reification of Idealization

171

­available. Furthermore the results of proper concretization are often imprecise, i.e. instead of one VaR number one gets a large set of possible VaR levels with lower accuracy. Last but not least, such procedures are more complex, hence often too sophisticated to be understood by executives and thus not preferred. As a result, the VaR models are used improperly. This leads to a severe risk of underestimation in the whole financial sector, making it vulnerable to negative economic shocks like the last housing market breakdown. 2

dsge Models and Their Ability to Predict Financial Shocks

Dynamic Stochastic General Equilibrium (dsge) models are the dominant approach taught in u.s. graduate schools and are frequently used by us macro economists (Colander 2009b). They are also often one of a group of models used by leading central banks (Du Plessis 2010). Hence, general equilibrium models have quite often been the subject of research in the field of economic methodology (i.a. Hoover 2001, Pemberton 2005, Colander 2009b). Hence, in this paper I focus only on the relationship between the use of dsge models by monetary policy decision makers and their ability to adequately predict and counteract economic shocks stemming from financial market instability. For this purpose I enumerate the idealizing assumptions that eliminate the impact of financial tensions on the economy from the model, and make the dsge inadequate to predict effects of financial shocks. 2.1 dsge Models Used for Monetary Policy before the Financial Crisis dsge models are consistent with the Walrasian approach, which assumes that under certain conditions (Blaug 1994 [1985]) a general stable equilibrium of the economy does exist. This means that the state of the economic system at any point of time is the solution of a system of simultaneous equations representing the demand for goods by consumers, the supply of goods by producers, and the equilibrium condition that supply equals demand on each market in the system (Arrow, Debreu 1954). In dsge models consumers and producers are usually modeled as rational representative agents that Pareto-optimally allocate their resources each period – this approach is based on the micro-­ foundation theory (Hoover 2010). In more advanced models some heterogeneities in preferences, income, or the type of constraints that agents face are allowed. Usually Bellman equations are employed to compute the equilibrium level of prices, the amount of traded goods and policy decisions (Canova 2005). dsge modeling requires a posteriori calibration of the model parameters, which has to be based on assumptions on the relationships between economic

172

Czerniak

categories (Hoover 1995). The most important relate to the shape of the utility function, to the set of available production technologies, and the characteristics of the policy impulse response functions (Rizvi 1994; Colander 2009b; Du Plessis 2010). As dsge modeling supporters argue: “If one doesn’t develop and stay true to such a carefully specified theoretical model, one will likely be fooled by spurious empirical relationships” (Colander 2009b, p. 3). dsge models are idealizations along many dimensions (Hoover 2005), of which the most important is that these models deal in aggregates while the economy is composed of individuals. As Pemberton (2005) notices: the ArrowDebreu proof for the existence of general equilibrium rests on modeling the economy as “a multi-dimensional vector space in which certain sets of points are assumed to be compact – a quality which, only in the loosest sense, could be deemed to be descriptive of reality” (p. 7). Furthermore, other important idealization dimensions are related to the notion of perfect rationality, the role of money in the economy, and nominal frictions. In practice the dsge model usually used by central banks is more of a hybrid model, its core is the solution to the inter-temporal optimization problem of representative agents – usually with rational expectations or real analysts’ consensus expectations (Du Plessis 2010). In some aspects they use vector auto-regression estimations and policy response functions, like the Taylor rule. Despite this, the general equilibrium models used by central banks are still pure idealized models that have to be calibrated and carefully evaluated, because their usefulness for policy making is highly limited by many idealizing assumptions. 2.2 Selected Assumptions of dsge Models There are two important idealizing assumptions of typical dsge models that do not allow them to be used – without proper concretization – for the purpose of assessing the risk of financial market tensions and breakdowns. Firstly, as Canova (2005, p. 38) points out, in classical general equilibrium models “with a full set of Arrow-Debreu claims, money is a redundant asset”. Therefore some frictions have to be introduced into dsge modeling which make money a desired asset. The two typical ways of introducing money into those models is the cash-in-advance (cia) and money-in-the-utility-function (miu) approach (Bhattacharjee, Thoenissen 2007). The cia approach is based on the assumption that first representative agents have to acquire money so that they in turn can buy goods. The miu approach makes money implicitly a desired asset – the more money agents have the higher their utility. These two approaches allow for the introduction of different types of price-setting frictions – transaction costs, liquidity constraints, sticky prices or sticky wages. Other

The Fallacy of the Reification of Idealization

173

typical non-price setting frictions are excluded from dsge models – i.e. information asymmetry between banks and borrowers, principal-agent problems, the depth of the credit market, bank lending and bank capital constraints, and firm and household balance sheet constraints, etc. (Du Plessis 2010). This is because of an implicit idealizing assumption of dsge models: financial institutions are neutral intermediaries in the monetary transmission mechanism of turning savings into investments (Assumption 1), i.e. financial markets work perfectly. In particular, agents can save and borrow in any quantity at the risk free rate fully controled by the central bank (Brzoza-Brzezina et al. 2013). Secondly, dsge models deal in aggregates of representative firms and households, which implies that each of them is identical in terms of income and preferences (Assumption 2). Hence, each agent in the model is assumed to act in the same way, i.e. makes the same decision. In more sophisticated models some heterogeneity is allowed – usually a couple of types of agents that are different in wealth, preferences, or rationality are introduced (Canova 2005, p. 35). However, there are no in-group differences and between-group mobility is impossible or highly limited. As Hoover (1995) points out “The analogue for physics would be to model the behavior of gases at the macrophysical level, not as derived from the aggregation of molecules of randomly distributed ­momenta, but as a single molecule scaled up to observable volume” (Hoover 1995, p. 40). These are crucial idealizing assumptions for modeling the effects of financial frictions. Without the heterogeneity of firms, households, and banks (once they are introduced into the model) in terms of their budget constraint, balance sheets, information sets and information costs, the modeling of financial non-price setting frictions is impossible. Due to the two above mentioned idealizing assumptions, macroeconomic general equilibrium models used by central banks and other major research centers could not forecast and account for the rising imbalances in asset markets and for the instability of financial institutions on the eve of the financial crisis. This inability brought heavy criticism on the heads of macroeconomists all over the world after the global recession started in 2008 (Krugman 2009). There also exists a large amount of secondary idealizing and technical assumptions of dsge models that are made during the calibration process. Calibration is usually preferred when the number of parameters is too large to permit econometric estimation. In such cases stylized facts and theoretical relationships are used to determine the value of parameters (Hoover 1995). Using this methodology, one implicitly assumes that relationships observed in the past will not break down in the future (see also assumption 2. and 4. in Section 1.2.). Furthermore, idealizing and technical assumptions made while

174

Czerniak

formulating economic theories that are used in the process of calibration become secondary assumptions of the dsge model itself. In some cases they can also be an important factor that has to be taken into account when dsge models are used to deal with financial market frictions. For example restrictions imposed on the utility function parameters defining the intertemporal substitution of leisure can substantially influence the level of savings, leading to an underestimation of household propensity to borrow in times of economic booms (see also Section 2.3.). 2.3 Concretization Attempts and Some Technical Problems After the crisis a large number of economists realized that financial markets should be taken into account when constructing and using general equilibrium models. As a result models with financial frictions – previously at the margin of professional interest – got into the mainstream. Eventually imperfect financial markets have been added to models used by central banks for policy purposes (Brzoza-Brzezina et al. 2013). These concretization attempts can be divided into two groups: in the first group, financial frictions are introduced into a classical neo-keynesian dsge framework with representative agents; in the second group heterogeneity of banks and financial frictions are modeled with the use of non-ge models and their results are compared to the results of dsge models without financial frictions to correct for the eventual shortcomings of the later. It is interesting that the two basic concepts of how to introduce financial institutions into the dsge models were proposed long before the last crisis. The first one was proposed by Kiyotaki and Moore (1997). This line of research introduces financial frictions via collateral constraints, i.e. borrowers have to provide a collateral, usually housing (Iacoviello 2005), for their loans. This directly affects the quantity of loans in the economy (Brzoza-Brzezina et al. 2013) and introduces financial frictions stemming from shocks to the required loanto-value ratio. Such an approach requires two groups of agents that have different time preferences for consumption – lenders and borrowers (so-called entrepreneurs). Furthermore, in this approach it is assumed that the interest on loans is higher than the risk free rate charged on deposits. The spread between these two can change over time due to technological reasons and s­ tochastic shocks. The second concept that allows for financial frictions is based on a paper by Bernanke and Gertler (1989). In this line of research, frictions arise due to transaction costs connected to the monitoring of the loan applicant and enforcing the credit contracts. Technically these transaction costs are introduced through information asymmetry. On the one hand, entrepreneurs (borrowing

The Fallacy of the Reification of Idealization

175

agents) are prone to idiosyncratic shocks, which are observed by them for free. On the other hand, lenders have to pay monitoring costs to learn about the realization of the shock. These transaction costs translate into a difference between the lending rate and the risk free rate – this spread is called the external finance premium. Hence, in contrast to the first line of research, in this approach financial frictions affect the economy via the prices of loans (BrzozaBrzezina et al. 2013). The second group of researchers proposes to abandon the dsge models with representative agents, because this methodology is not suitable for investigating financial frictions. They argue that heterogeneity is at the heart of financial market imbalances and cannot be modeled only through adding some risk premium and/or idiosyncratic shocks (Du Plessis 2010). The most representative model of this approach is the Risk Assessment Model for Systemic Institutions (ramsi) developed at the Bank of England. At the core of ramsi are the detailed balance sheets of the largest uk banks, including approximately 400 asset classes and 250 liability classes (Aikman et al. 2009). The link between the economy and the various risks on the banks’ balance sheets is modeled with the use of a large-scale Bayesian var – this part of the model is the only source of shocks in ramsi. This framework allows for systemic risks to be taken into account, such as macro-credit risk, income risk, network interactions, and feedback effects. As a result, ramsi is very useful for investigating the domino effect of bank defaults and modeling systemic risk of high leverage ratios. These results can, in turn, be used to correct the dsge model outputs. Both lines of research have their advantages and disadvantages. The dsge models with financial frictions are a step forward in concretizing the outputs of general equilibrium models, however they are still based on a large number of unrealistic assumptions as discussed in Section 2.1. Furthermore, the introduction of the banking system forced dsge designers to make new idealizing assumptions about attitudes towards risk, monetary transmission mechanisms and the types of financial market competition. For example it is assumed that: borrowers are risk neutral, households are risk averse, the banking sector is perfectly competitive, and banks break even every period, the risk-free interest rate exists and is set by the central bank. As a result, the outputs of these tweaked dsge models are still in some points counterintuitive and at odds with empirical evidence (Brzoza-Brzezina et al. 2013). On the other hand, abandoning the representative agent concept helps to model financial market frictions and interdependencies more realistically, but is in fact a step backwards in economic modeling. The real economy, with all its complexity, is described in the ramsi model by a set of equations estimated with a black box method – the Bayesian var. It lacks the theoretical backbone

176

Czerniak

of dsge models, which make them useful in explaining economic phenomena and in investigating effects of policy decisions. As in the VaR models the use of Bayesian var estimation for forecasting and policy modeling requires a very strong idealizing assumption – that past observed relationships between economic variables will not breakdown in the future, especially in times of economic crisis. From a theoretical point of view, the best solution would be to merge these two different approaches into one advanced dsge model without representative agents. But this theoretically simple way of concretization is impossible at the moment due to technical constraints. State-of-the-art economic modeling techniques do not allow for an infinitely large number of different agents that simultaneously optimize their decisions in an infinite number of periods (Canova 2005). Even models with simple heterogeneity with two-three types of representative agents are “daunting to compute” (Fernández-Villaverde et al. 2010, p. 6). 3

The Immanency of the Ricardian Vice

Improper use of dsge and VaR models made it difficult – or even ­impossible – for decision makers to forecast and prevent the last financial crisis at the macro, as well as at the micro level. The idealizing assumption of these models made them useless for assessing the systemic risk in financial markets, despite this, dsge and VaR models were widely used for this purpose. Such a misuse of idealized models can be seen as a classic example of the Ricardian vice (Schumpeter 1954) or as Nowak (2000) would call it – the fallacy of the reification of idealization in economics. However, this is not the first and most likely not the last example of this methodological error in economics. Why is economics so prone to the Ricardian vice? There are three complementary answers to this question, each relating to another level of methodological analysis. 3.1 Ricardian Vice and the Features of Economics as a Science There are three deeply rooted characteristics of economics that ­encourage ­researchers and policy makers to use economic models without proper ­concretization. Firstly, economics is a science that is closely connected to policy making. Far more than in any other discipline, the results of economic ­research are used by politicians and other policy makers to choose, defend, and evaluate financial, fiscal and monetary policy decisions. They are used, for example, to find the right way to counteract unemployment, decrease inflation, solve ­liquidity problems, or to boost entrepreneurship, foster investment, and

The Fallacy of the Reification of Idealization

177

mitigate financial risk. As a result there is a large demand for models that can promptly provide policy makers with straightforward answers to any arising questions. Hence sophisticated, formal idealized mathematical models are preferred over non-idealized inductive or descriptive institutional analysis, they are hard to falsify and the results can seldom be generalized. Furthermore, such mathematical models are easier to apply for supervisory reasons – for example to set limits on public expenditures or financial institutions capital requirements. Secondly, unlike physics, economics can be based mostly on natural experiments. The use of experimental economics is limited to consumption and investment decision making – the domain of microeconomics and f­inance. Macroeconomic phenomena can be analyzed only without the ­ceteris ­paribus clause, which makes it harder to distinguish between better and worse ­economic models and theories. Furthermore, some models can forecast and explain economic phenomena accurately in normal times, when their idealizing assumptions do not seem to matter, but are useless without proper concretization in times of financial tensions and economic busts (see the VaR model described in Section  2.). As a result, idealizing assumptions and a model’s shortcomings are often neglected as long as the model’s forecasts are accurate. Thirdly, macroeconomics is quite a young science and does not have a firm theoretical foundation like physics or chemistry. As Colander points out “the current field of macro [is] an engineering field in which the best we can hope for are loose laws to guide policy that follow from common sense and empirical studies” (Colander 2009a, p. 3). This might be seen as an oversimplification of the problem. One might argue that other social sciences like sociology or psychology have the same problem of a lack of theoretical foundations and at the same time are more immune to the fallacy of the reification of idealization. However, if we combine a lack of proper concretization tools and methods with high demand for unambiguous policy guidelines and difficulties in empirical falsification of theories, then one can see why economics is more prone to the Ricardian vice than other sciences. The choice that economists very often have to face is: whether to provide no answer to a research question or to provide the least wrong answer, which often means the application of an idealized model for policy purposes without proper concretization. Ricardian Vice and the Philosophical Foundations of Economic Methodology A large number of heterodox researches argue that the problems of economic methodology stem from the adoption of instrumentalism by mainstream researches. As articulated in 1953 by Milton Friedman in his famous essay “The

3.2

178

Czerniak

Methodology of Positive Economics”, economic models are only man-made tools and what matters is only their predictive success. In his opinion a model can be judged as good, not if it is true or if its assumptions are realistic, but if its predictive performance is better than the performance of alternative models. Such a methodological approach neglects the problem of idealization and concretization, because it puts aside the problem of the realism of any given model or theory. Instrumentalism treats inquires about the truthfulness of a given model as fruitless and redundant (Hoover 2001). As a result this methodology favors the use of idealized models without proper concretization for policy purposes. Another methodological stance widely adopted by mainstream researchers since the 1980’s is formalism. According to historian Mark Blaug (2003), economists have converted the subject into a sort of social mathematics in which analytical rigor is everything and the solution of an economic model is the final answer to the question that prompted the investigation in the first place. Hodgson goes even a step further and argues that “the victory of technique over substance is a chronic problem within modern economics” (Hodgson 2006, p. 1). Empirical research (Klamer, Colander 1990) shows that only 3% of graduate students on top us economics programs perceived ‘having a thorough knowledge of the economy’ to be ‘very important’ for professional success. On the other hand, 57% believed that ‘excellence in mathematics’ is what matters. In my opinion formalism can be seen as the methodological justification of the wide use of sophisticated mathematical models which is driven by high demand for policy guidelines. 3.3 Ricardian Vice and the Institutional Setting of Economic Research In recent years the development of economics, along with other social sciences, has been largely influenced by the development of citation based indices of research success. This has in turn led to a proliferation of “producers” of macroeconomic models. According to Colander (2009a) a producer of a economic model does not need to know the limitations of a theory, but must know every technical aspect of modeling. This type of scientists brings the latest analytical and computational technologies to the existing model, tweaking and upgrading it. Such research methods are preferred in the current academic institutional setting, since they drive up citations and make their authors more successful – in academic and financial terms. This pushes other researchers, mostly heterodox economists, outside the mainstream preventing their results from being recognized as important and published to a wider audience. Furthermore, a drive for producer economists has pushed out the political and

The Fallacy of the Reification of Idealization

179

institutional economy and lead to a change in the training of undergraduate economic students. Courses of economics became more technical and mathematicized in line with the postulates of formalism. Although producers do not design their models for policy purposes, their ­results are very often used to prepare guidelines for policy decision makers. The problem comes, as I have shown in Sections  2. and 3., when idealized models are used for policy purposes without taking into account that their assumptions were chosen on the basis of analytical tractability, not policy applicability. In such cases, idealizing assumptions are very often neglected and models are applied directly to analyze real economic phenomena, as long as their predictions are roughly correct. For example VaR and dsge models were used as long as financial frictions and systemic risks could be neglected. After the financial crisis they have been upgraded or withdrawn from usage. 4 Conclusions The Ricardian vice is an immanent feature of economic as a science. When its existence is ignored it can cause large damage to economic research as well as to the real economy through the production of bad guidelines for policy makers. As I have shown, the improper usage of the financial VaR model and the macroeconomic dsge model deludes market participants – central bankers supervisory institutions and banking risk managers – making them believe that a financial crisis was remote and that the economy was widely balanced. Furthermore, this is most likely not the last time in economic history that idealized models, which are used for policy purposes, will misguide leading decision makers on the current and future condition of the economy. The adverse effects of this propensity of economics can however be mitigated by the introduction of some measures that would minimize the probability of the misuse of idealized models. First and foremost, economic researchers should promote a culture of enumerating, in research papers, the primary idealizing assumptions of the models they produce, especially, when those models can be used for policy purposes or for answering questions which are important to decision makers. Furthermore, commonly used point forecasts produced with the use of economic models should be substituted by interval forecasts. This technical change would promote, among public opinion, a larger recognition of the uncertainty connected to economic forecasts. I also agree with the propositions formulated by Colander (2009a), to stimulate – at an institutional level – cooperation between mainstream and heterodox

180

Czerniak

economists in the application of economic theory for policy issues. This list of possible measures is of course incomplete and further research – empirical and theoretical – on how to successfully mitigate the adverse effects of the Ricardian vice is highly desirable. References Aikman D., P. Alessandri, B. Eklund, P. Gai, S. Kapadia, E. Martin, N. Mora, G. Sterne and M. Willison (2009). Funding Liquidity Risk in a Quantitative Model of Systemic Stability. Bank of England Working Paper no. 372. Alexander, C. (2008). Value-at-Risk Models. Chichester: John Wiley & Sons Ltd. Arrow, K.J., and G. Debreu (1954). Existence of an Equilibrium for a Competetive ­Economy. Econometrica 22 (3), 265–290. Artzner, P., F. Delbaen, J.-M. Eber, and D. Heath (1999). Coherent Measures of Risk. Mathematical Finance 9 (3), 203–228. Bałamut, T. (2002). Metody estymacji Value at Risk. Materiały i Studia, zeszyt no. 174. Warszawa: Narodowy Bank Polski. Bams, D., T. Lehnert, C.C.P. Wolff (2005). An Evaluation Framework for Alternative VaR-Models. Journal of International Money and Finance 24 (6), 944–958. Bernanke, B., and M. Gertler (1989). Agency Costs, Net Worth, and Business Fluctuations. American Economic Review 79 (1), 14–31. Bervas, A. (2006). Market Liquidity and its Incorporation into Risk Management. ­Financial Stability Review No. 8, pp. 63–79. Paris: Banque de France. Bhattacharjee, A., and C. Thoenissen (2007). Money and Monetary Policy in Dynamic Stochastic General Equilibrium Models. The Manchester School 75 (1), 88–122. Birner, J. (1994). Idealizations and the Theory Development in Economics. Some ­History and Logic of the Logic of Discovery. In: B. Hamminga, N.B. De Marchi (eds.) Idealization VI: Idealization in Economics. Poznań Studies in the Philosophy of the ­Sciences and the Humanities 38, pp. 277–301. Amsterdam: Rodopi. Blaug, M. ([1985] 1994) Teoria ekonomii. Ujęcie retrospektywne [Economic Theory in Retrospect]. Warszawa: Wydawnictwo Naukowe PWN. Blaug, M. (2003). The Formalist Revolution of the 1950s. In: W.J. Samuels, J.E. Biddle and J.B. Davis (eds.). A Companion to the History of Economic Thought, pp. 395–410. Oxford: Blackwell Publishing. Blommestein, Hans J., L.H. Hoogduin and J. Peeters. (2010). Uncertainty and Risk ­Management after the Great Moderation: the Role of Risk (Mis)Management by Financial Institutions. SUERF Studies: The Quest for Stability: The View of Financial Institutions, pp. 7–29.

The Fallacy of the Reification of Idealization

181

Brzoza-Brzezina, M., M. Kolasa and K. Makarski (2013). The Anatomy of Standard DSGE Models with Financial Frictions. Journal of Economic Dynamics and Control 37 (1), 32–51. Canova, F. (2005). Methods for Applied Macroeconomic Research. Princeton: Princeton University Press. Chrystal, K.A., P.D. Mizen and F.C. Artis (2003). Goodhart’s Law: its Origins, Meaning and Implications for Monetary Policy. In: Mizen, P. (ed.) Central Banking, Monetary Theory and Practice: Essays in Honour of Charles Goodhart. Volume One, pp. 221–243. Cheltenham: Edward Elgar Publishing Limited. Colander, D. (2009a). How Did Macro Theory Get So Far off Track, and what Can Heterodox Macroeconomists Do to Get it Back On Track? Middlebury College Economics Discussion Paper, no. 09–11. Colander, D. (2009b). Economists, Incentives, Judgment, and the European CVAR ­Approach to Macroeconometrics. Middlebury College Economics Discussion Paper, no. 09–12. Cools, K., B. Hamminga, T.A.F. Kuipers (1994). Truth Approximation by Concretization in Capital Structure Theory. In: B. Hamminga, N.B. De Marchi (eds.) Idealization VI: Idealization in Economics. Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 38, pp. 205–228. Amsterdam: Rodopi. Danielsson, J. (2002). The Emperor Has no Clothes: Limits to Risk Modelling. Journal of Banking & Finance 26 (7), 1273–1296. Danielsson, J. (2008) Blame the Models. Journal of Financial Stability 4 (4), 321–328. Du Plessis, S. (2010). Implications for Models in Monetary Policy. Journal of Economic Methodology 17 (4), 429–444. Egloff, D., M. Leippold and P. Vanini (2003). A Simple Model of Credit Contagion. Online version: http://www.nccr-finrisk.uzh.ch/media/pdf/wp/WP141_2.pdf [downloaded 9.02.2014]. Erb, C.B., C.R. Harvey and T.E. Viskanta (1994). Forecasting International Equity ­Correlations. Financial Analysts Journal 50 (6), 32–45. Fama E. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. Journal of Finance 25 (2), 383–417. Fernández-Villaverde J., P.A. Guerrón-Quintana and J. Rubio-Ramírez (2010). Reading the Recent Monetary History of the US, 1959–2007. National Bureau of Economic Research Working Paper no. 15929. Hodgson, G. (2006). On the Problem of Formalism in Economics. Voprosy Economiki 3, 112–124. Hoover, K.D. (1995). Facts and Artifacts: Calibration and the Empirical Assessment of Real-Business-Cycle Models. Oxford Economic Papers 47 (1), 24–44. Hoover, K.D. (2001). The Methodology of Empirical Macroeconomics. Cambridge: ­Cambridge University Press.

182

Czerniak

Hoover, K.D. (2005). Quantitative Evaluation of Idealized Models in the New Classical Macroeconomics. In: M.R. Jones, and N. Cartwright (eds.). Idealization XII: Correcting the model, idealization and abstraction in the sciences. Poznan Studies in the Philosophy of the Sciences and the Humanities, vol. 86, pp. 15–34. Amsterdam: Rodopi. Hoover, K.D. (2010). Microfoundational Programs. HOPE Center Working Paper No. 2010–01. Iacoviello, M. (2005). House Prices, Borrowing Constraints, and Monetary Policy in the Business Cycle. American Economic Review 95 (3), 739–764. IMF (2007). Do Market Risk Management Techniques Amplify Systemic Risks? In: Global Financial Stability Report: Financial Market Turbulence. Causes, Consequences, and Policies. Washington: International Monetary Fund, pp. 52–76. Jarrow, R.A. (2005). Liquidity Risk and Classical Option Pricing Theory. Online version: http://forum.johnson.cornell.edu/faculty/jarrow/111%20wp.pdf [downloaded 9.02.2014]. Kiyotaki, N., and J. Moore (1997). Credit Cycles. Journal of Political Economy 105 (2), 211–248. Klamer, A., and D.C. Colander (1990). The Making of an Economist. Boulder: Westview Press. Knorr Cetina, K. and A. Preda (2005). The Sociology of Financial Markets. Oxford: ­Oxford University Press. Krugman, P. (2009). How Did Economists Get It So Wrong? The New York Times. September 6. Online version: http://www.nytimes.com/2009/09/06/ magazine/06Economic-t.html [Downloaded 26.10.2009]. MacKenzie, D. (2006). An Engine, Not a Camera. How Financial Models Shape Markets. Cambridge: The MIT Press. Mandelbrot, B.B. (1963). The Variation of Certain Speculative Prices. The Journal of Business 36 (4), 394–419. Mandelbrot B.B., A.J. Fisher and L.E. Calvet (1997). A Multifractal Model of Asset ­Returns. Cowles Foundation Discussion Paper No. 1164. Markovitz, H. (1952). Portfolio Selection. The Journal of Finance 7 (1), 77–91. Mirowski, P. (1988). Against Mechanism: Protecting Economics from Science. Lanham: Rowman & Littlefield Publishers, inc. Modigliani, F. and M.H. Miller (1958). The Cost of Capital, Corporation Finance and the Theory of Investment. The American Economic Review 48 (3), 261–297. Nowak, L. (2000). The Idealizational Approach to Science: A New Survey. In: L. Nowak, and I. Nowakowa, Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities 66, pp. 109–184. Amsterdam-Atlanta: Rodopi. Pemberton, J. (2005). Why Idealized Models in Economics Have Limited Use. In: M.R. Jones, and N. Cartwright (eds.). Idealization XII: Correcting the Model, Idealization

The Fallacy of the Reification of Idealization

183

and Abstraction in the Sciences. Poznan Studies in the Philosophy of the Sciences and the Humanities, vol. 86, pp. 35–46. Amsterdam: Rodopi. Peterson, R.L. (2007). Inside the Investor’s Brain. The Power of Mind over Money. Hoboken: John Wiley & Sons, Inc. Pozner J.-E., M.K. Stimmler, P.M. Hirsch. (2010). Terminal Isomorphism and the SelfDestructive Potential of Success: Lessons from Subprime Mortgage Origination and Securitization. In: M. Lounsburry, P.M. Hirsch (eds.) Markets on Trial: The Economics Sociology of the U.S. Financial Crisis. Research in the Sociology of Organizations, vol. 30, pp. 181–214. Bingley: Emerlad Group Publishing Limited. Samuelson, P.A. (1965). Proof that Properly Anticipated Prices Fluctuate Randomly. Industrial Management Review 6 (2), 41–49. Rizvi, S.A.T. (1994). The Microfoundations Project in General Equilibrium Theory. Cambridge Journal of Economics 18 (4), 357–377. Schumpeter, J.A. (1954). History of Economic Analysis. New York: Oxford University Press. Shiller, R.J. (1981). Do Stock Prices Move Too Much to be Justified by Subsequent Changes in Dividends? American Economic Review 71 (3), 421–436.

chapter 8

Strategies of Comparative Analysis in Historical Comparative Sociology: An Attempt at an Explication within the Conceptual Framework of the Idealizational Theory of Science Krzysztof Brzechczyn Abstract The aim of this article is to explicate the basic types of comparative analysis used in historical-comparative sociology, within the conceptual framework of the idealizational theory of science (its), and to determine the usefulness and place of the ­comparative method in the structure of the idealizational theory. In the second ­section some classifications of comparative analysis are presented. An analysis of comparative methods used in States and Social Revolutions and what methodological discussion it stimulated is outlined in the third section. Section 4 presents the key tenets of the idealizational theory of science. Section 5 explicates the basic types of comparative analysis within the conceptual framework of the theory and defines the place of that method in the idealizational theory. In the final section, I consider the status of ­historical-comparative sociology in light of the view of the development of science provided by the idealizational theory.

1 Introduction Historical-comparative sociology was born in the 1960s in the usa and Western Europe but it only gained methodological and theoretical traction in the 1980s and 1990s. That consciousness has manifested itself in some theoretical (Abrams 1982, Skocpol 1984b, Smith 1991) and methodological (Bonnel 1980, Tilly 1981, 1984, Skocpol 1984a, Braembussche 1989, Mahoney, Rueschemeyer 2003, Lange 2013) works presenting its theoretical achievements. Theda Skocpol’s book State and Social Revolutions is considered to be a classic in this scientific sub-discipline. That work laid out the further development of the theory of revolution and became the subject of an animated theoretical and methodological discussion (Goldstone 2003, p. 62, p. 65).

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_010

Strategies of Comparative Analysis

185

The aim of this article is to explicate the basic types of comparative analysis used in historical-comparative sociology, within the conceptual framework of the idealizational theory of science (its), and to determine the usefulness and place of the comparative method in the structure of the idealizational theory. 2

Classifications of Comparative Analysis. Selected Examples

Victoria E. Bonnell (1980) distinguishes two basic comparative strategies: an “illustrative” one and an “analytical” one. In the case of the illustrative strategy: The main point of comparison is between equivalent units, on the one hand, and the theory or concept on the other. This variant evaluates individual units on the one hand and a theory or concept on the other. This variant evaluates individual units not in relation to each other but in relation to a basic theory or concept applicable to all of them. bonnell, 1980, p. 165

And in the case of the analytical strategy: The main point of comparison is between or among equivalent units. The comparison involves an identification of independent variables that serve to explain common or contrasting patterns of occurrences. The investigator juxtaposes equivalent units with each other in order to discern regularities that might provide explanatory generalizations. bonnell 1980, pp. 164–165

The criterion of division here is the place in a comparison where theoretical assumptions are made. In the illustrative type of comparison the researcher has a ready theory and determines the range of its applications with the use of the compared cases. In the analytical type of a comparison the comparison is expected to give rise to a theory of the compared cases. Theda Skocpol and Margaret Somers (Skocpol, Somers 1980) proposed a different classification of comparative analysis, distinguishing a parallel, ­contrast-oriented, and a macro-causal type. According to the authors, the ­parallel comparative analysis: serves as an ancillary mode of theoretical demonstration. Historical instances are juxtaposed to demonstrate that the theoretical arguments

186

Brzechczyn

apply convincingly to multiple cases that ought to fit if the theory in question is indeed valid. Cases are selected to cover all possibilities, or to represent a range of sub-types or points on continua. The point of the comparison is to assert a similarity among the cases – similarity, that is, in terms of the common applicability of the overall theoretical arguments. skocpol, somers 1980, pp. 176–177

They also point out the following important feature of that method: “it is characteristic of all works of parallel comparative history to elaborate theoretical model and hypotheses before turning to historical case illustrations” (Skocpol, Somers 1980, p. 177). In this respect the parallel method is reminiscent of the illustrative method from Bonnell’s classification. The second type of comparative analysis involves looking for contrasts among the compared cases. Scholars who use this comparative strategy “bring out the unique features of each particular case included in their discussions, and to show how these unique features affect the working-out of putatively general social processes” (Skocpol, Somers 1980, p. 178). Unlike in the parallel method, researchers who use the contrast-oriented comparative method “aim to place historical limits on overly generalized theories, but they do not aspire to generate new explanatory generalization through comparative historical analysis” (Skocpol, Somers 1980, p. 178). The macro-causal comparative method combines the parallel and the ­contrast-oriented ones. On the one hand, [m]acro-analysts can try to establish that several cases having in common the phenomenon to be explained also have in common the hypothesized causal factors, although the cases vary in other ways that might have seemed causally relevant. skocpol, somers 1980, p. 183

On the other hand, [m]acro-analysts can contrast cases in which the phenomenon to be explained and the hypothesized causes are present to other (“negative”) cases in which the phenomenon and the causes are both absent, although they are as similar as possible to the “positive” cases in other respects. skocpol, somers 1980, p. 183

Strategies of Comparative Analysis

187

In their meta-methodological commentary Skocpol and Somers claim that the macro-causal comparative history is an application of Mill’s methods of agreement and difference (Mill [1843] 1974). 3

On the Methodology of States and Social Revolutions

Let us consider how comparative analysis is used in Theda Skocpol’s States and Social Revolutions. A research process can be divided into three stages: (i) forming a research problem and selecting case studies for analysis; (ii) identifying the main factors influencing the selected case studies; (iii) testing the hypothesis about the essential influence of the selected main factors on the studied cases, through looking for ‘negative’ examples, in the case of which some of the main factors have occurred but did not lead to the same results as in the case of ‘positive’ studied cases. Ad (i). Skocpol’s research problem is how to explain the outbreak of social revolutions defined as a “rapid, basic transformation of a society’s state and class structures; […] accompanied and in part carried through by class-based revolts from below” (Skocpol 1979, p. 4). Skocpol studies the causes of the outbreaks of revolutions in France, Russia, and China. The societies are said to have been ruled with a system of agrarian bureaucracy consisting of four main classes or social groups: bureaucrats, landowners, bourgeoisie, and peasantry. Offices were only partially specialized in these cases. Officials were subject to hierarchical control which did not, however, include all their functions. State authorities did not have sufficient resources to finance wars, the court, and the officials. Therefore it was necessary to employ officials from aristocratic landowning families and to bestow them with administrative power over peasants. Those landowners abused their political authority to increase the economic exploitation of peasants. Ad (ii). A victorious revolution resulted from the joint impact of two factors: (A) a political crisis of the state and (B) mass peasant revolts. Geopolitical pressure weakened the international position of states based on agrarian bureaucracy and triggered political crises. Elites in those countries tried to stand up to the challenges of the international situation by conducting reforms which loosened the relationship between political power and the landowning class. That led to internal political crises which, in turn, paved the way for the

188

Brzechczyn

o­utbreaks of peasant rebellions (C). An alliance of the political elites from the opposition (D) with grassroots peasant movements helped coordinate the actions of that social class on the level of the whole society and to articulate a separate vision of social order. As a result, state authority was abolished and a new state order was introduced. The model of social revolution is presented in the Figure 8.1. Ad (iii). Skocpol verifies her hypothesis by considering negative cases, i.e. social situations in which international pressure did not lead to an inner political crisis and in which peasant revolts did not evolve into social revolutions. The first type of such negative cases includes countries (Germany 1807–14 and 1848–50, Japan 1868–73 and Great Britain 1640–1689) which did not experience a political crisis although they suffered from external pressure. Skocpol explains this with reference to the peculiarities of the social-political systems of Japan and Germany and with the differences of the intensity of geopolitical pressure. In her view Japan and Germany were highly bureaucratized, absolutist countries in which the landowning class was subject to the state. Therefore when the political powers introduced political reforms, it did not cause a political crisis. Moreover, according to the author, England in the 17th century and Germany during the Spring of Nations only experienced “mild” international pressure. The second type of negative cases is represented by societies in which local peasant revolts did not evolve into a mass-scale social revolution despite conducive social conditions. This was the case in Germany (1848–1850), Japan (1868–1873), and England (1640–1689). Skocpol analyzes the impact of various

Successful revolution

Politcial crisis

Peasants revolts

Geopolitical pressure

Oppositional political elites

Figure 8.1  The model of successful revolution

189

Strategies of Comparative Analysis

factors (e.g. economic development and the structure of peasants’ property ownership) on peasants’ ability to mobilize. In the first two countries peasant communities were controlled so efficiently, be it by the junkers allied with state bureaucracy (Germany) or by the state itself (Japan) that it was impossible to transform local revolts into a national revolution. In England the early capitalist transformations destroyed rural communities, which prevented a social outbreak, even though the quality of life of some strata of peasantry deteriorated. The results of Skocpol’s studies (1979, pp. 155–157) are summed up in the table 8.1. Skocpol’s methodology and the results she obtained caused a vigorous discussion lasting over 30 years. In this article I only present the methodological aspects of the dispute. The most important critical voices were those of Elizabeth Nichols (1986), Michael Hechter and Edgar Kiser (1991), Michael Burawoy (1989), Rebecca Jean Emigh (1997), and Igor Hanzel (2010, pp. 328–333). The following issues were examined: 1. The role of theoretical assumptions in a comparative analysis (Hechter, ­Kiser 1991, pp. 23–24, Hanzel 2010, pp. 331–332). 2. The validity of treating the French, Russian, and Chinese revolutions as representing the same type of a revolution, as such a decision involves certain silent theoretical assumptions: the French revolution is considered to be a bourgeois movement, and the Russian and Chinese ones – proletarian (Burawoy 1989, p. 769; Nichols 1986, p. 177). Table 8.1 Explanation: sbm – a semi-bureaucratic absolute monarchy, hbas – a highlybureaucratic absolutist state, sba – a semi-bureaucratic absolutist state, hbs – a highly-bureaucratic state, nba – a non-bureaucratic state. State

Type of Geopolitical pressure leading to a government political crisis

Peasant A victorious revolts ­social revolution

1. France 2. Russia 3. China 4. Prussia 5. Japan 6. England

sbam hbas sbas hbas hbs nbs

Yes Yes Yes Yes Yes No

Moderate Extreme Strong Strong (1807–1814) Mild (1848) Strong: imperialistic intrusion Mild

Yes Yes Yes No No No

190

Brzechczyn

3. The lack of a clear theory of a revolution – it is not apparent why the two co-existent factors: a political crisis and peasant revolts are deemed to be, jointly, a necessary prerequisite for a revolution (Nichols 1986, p. 177). 4. In spite of her inductive declarations, some factors present in the analysis are “more privileged than other ones” (Nicholas 1986, p. 178); for example, peasant revolts and a political crisis were not the only factors present in all three of the analyzed cases of revolutions.1 5. The incongruence of Skocpol’s methodological declarations with her ­actual research approach (Hechter, Kiser, 1991, pp. 14–15; Burawoy 1989, p. 763, 769; Emigh 1997, pp. 652–653): although the author claims to have used inductive reasoning according to Mill’s methods of agreement and difference (Skocpol 1979, pp. 36–37, 1984, pp. 378–379), critics note that there are some deductive theoretical assumptions and lines of reasoning in her writing. Skocpol’s answer to the above critique was not particularly convincing as it did not deepen the knowledge about the position of comparative analysis in the construction of a theory in historical-comparative sociology. On the one hand, the author states that: finally, it needs to be stressed that comparative historical analysis is not substitute for theory. Indeed, it can be applied only with the indispensable aid of theoretical concepts and hypotheses. For the comparative method alone cannot define the phenomenon to be studied. It cannot select appropriate units of analysis or say which historical cases should be studied nor can it provide the causal hypotheses to be explored. All of these must be done from the macro-sociological imagination, informed by the theoretical debates of the day, and sensitive to the patterns of evidence for such historical cases. skocpol 1979, p. 39

On the other hand, however, she argues: the causal arguments positively invoked in States and Social Revolutions are not derived from a pre-existing theory of revolution, this is very true – and the great advantage of the book lies precisely in this fact! How are we 1 For instance, Rosemary H.T. O’Kane mentions civil war as an important factor in the ­construction of a post-revolutionary state in the three societies studied by Skocpol (O’Kane 1995, p. 20).

Strategies of Comparative Analysis

191

ever to arrive at new theoretical insights if we do not let historical patterns speak to us, rather than always viewing them through the blinders, or the heavily tinted lenses, of pre-existing theories? skocpol 1986, p. 190

However, the author also admits that some elements of theoretical reasoning are, indeed, necessary and that comparative study involves a creative synthesis of deduction and induction. She confesses: during the years that I worked on the arguments of States and Social Revolutions, I moved from history to hypotheses and back to history. First, I used comparative historical reading to discover surprising ­regularities – between the supposedly “bourgeois” French Revolution and the supposedly “Communist” Chinese Revolution. Then, I formulated hunches about the structural and historical conditions that accounted for the regularities. Finally, I turned my hunches into more precise causal ­hypotheses, forced myself to operationalize these hypotheses in consistent ways across cases, and extended my comparative historical analysis to include another “positive” case, Russia, and a series of negative cases sharing many background conditions with France, Russia and China. skocpol, 1986, s. 190

We know, then, what the interaction between theory and empirical observations was in the process of writing the book. What we still do not know, however, is how we could re-construct those relationships in historical-comparative sociology and delineate the place of comparative analysis in the process of constructing a scientific theory. 4

The Basic Concepts of the Idealizational Theory of Science

It would be interesting to ascertain whether comparative analysis could somehow be incorporated in the process of building the idealizational theory of ­science. In order to verify that, I will present the basic concepts of the idealizational theory of science (Nowak 1970, 1971, 1974, 1977a, 1980, 1992, Nowak, Nowakowa 2000). A crucial notion of its is the notion of influence. Factor H influences magnitude F when the adoption of a certain value by H excludes the adoption of any value by magnitude F. Metaphorically speaking, magnitude F which is under the influence of factor H is limited in the ‘choice’ of intensity. The set of

192

Brzechczyn

restricted values φ1, φ2, …, φi of magnitude F is called the power of the influence of factor H on magnitude F and it is symbolically noted as WF(H). Let us assume that nF is a set of all values adopted by magnitude F. It is possible to distinguish three kinds of influence of factor H on the magnitude factor F: – maximal – when factor H excludes NF-1 of the values assumed by magnitude F, what means that the value adopted by magnitude F is strictly determined by that factor H; – predominant – if factor H, assuming value h, excludes the majority of values adopted by magnitude F from set nF; – minimal – if factor H, assuming value h, excludes the minority of values adopted by magnitude F from set nF . The factors recognized as main ones in the idealizational theory of science exert a maximal and predominant influence; the secondary ones have a minimal impact. A hierarchy of the influence of factors is based on their strength of influence. Let us say that factor H is more important for magnitude F than factor p when the set of restrictions of the values adopted by magnitude F due to the influence of factor H is composed of more elements than the set of restrictions of the values adopted by magnitude F due to the influence of factor p.2 ​WF​  ​(H)​ > ​WF​  ​(p)​.​ A researcher constructs a scientific theory by taking a particular ontological and theoretical perspective. The process should comprise the stages described below. 1. On the basis of the chosen ontological perspective the researcher selects factors, defining what type of factors has an impact on magnitude F and 2 The explication of the influence, presented for illustrative purposes, involves many subtle assumptions which, if taken into account, would further complicate our discussion. For example, it is presumed that the considered factors are homogenous which means that the power of the influence of factors p and H on magnitude F is independent from the intensities of the determining factors (Brzeziński, et al. 1976, pp. 116–118). Moreover, it is presumed that the ranges of the influences of factors H and p on magnitude F are divergent, i.e. that the sets of the restricted intensities of magnitude F are divergent (Paprzycka, Paprzycki 1992). The issue of the essentiality of factors in the idealizational theory of science is presented in Paprzycka 2005, pp. 88–104, Zajęcki, 2013.

Strategies of Comparative Analysis

193

what type of factors has no influence on it; in this way the space of factors essential for the studied magnitude F is created. 2. On the basis of the selected theoretical perspective the researcher determines the type of factors which can be the main ones and the type of factors which can be secondary for the studied magnitude; in this way the factors which influence a given magnitude are divided into main and secondary ones. 3. The power of the influence of particular factors becomes the basis for the creation of a hierarchy of factors which allows the construction of an image of the essential structure of magnitude F. 4. Idealizing assumptions which eliminate the influence of factors considered to be secondary ones are introduced, and idealizational statements which show the dependence between magnitude F and the main factor acting on it are formed. 5. The accepted idealizing assumptions are gradually canceled out and the idealizational statement is gradually modified to show how magnitude F depends on the main and secondary factors which modify the basic dependency. 6. The reason why magnitude F has a particular value on object O is explained by revealing how that magnitude is related to its main and secondary factors. The research procedure of making a theory adequate is independent from forming an idealizational statement and concretizing it (Nowak 1979). A researcher builds an image of space – of range U – of the factors which are ­essential for magnitude F in a given scope. Range U can be an adequate reconstruction of the scope of magnitude F but it can also happen that a researcher will not accurately determine the range of the influence of factors significant for magnitude F. Let us illustrate this with an example. Set U of all objects with magnitude F with some intensity is a universum of that variable. Let us assume that we can distinguish certain subsets, Z1 …, Zn, in the universum. Those sets are F-species and the sum of F-species with the same main factor forms a type of a given magnitude in the given scope. Particular F-kinds differ from one another with respect to the main factor (for simplicity’s sake we assume that there is only one main factor in an essential structure of a given variable) whereas F-species which belong to the same F-kind and have the same main factor ­differ from one another with respect to secondary factors (Nowak 1977b, p. 83). The ­differentiation can be illustrated as follows:

194

​  ​S​ F ​ ​ : H  ​SZ3 ​ F ​ ​ : G ​ ​ ​SZ1​ F ​ : H ​  ​  ​    Z2 ​    ​  ​ ​              H, p               H, q               G, q

Brzechczyn

The Z1, Z2, Z3 sets is are F-species. Besides, factor F has two F-kinds. Kind R1 is identical with the sum of classes Z1 and Z2, whereas kind R2 is identical with class Z3. The range of factor G will be inaccurately determined if a researcher presumes that it includes two classes, Z2 and Z3 constituting one kind R2. Factor G is the main factor only in class Z3 and it is not present at all (or it may be a secondary factor) in class Z2 which belongs to kind R1. Although a researcher has a theoretical perspective at their disposal which delineates the type of factors (e.g. materialistic, idealistic, etc.) that ought to be considered as the main ones, they are not able to determine the range of the influence of particular main factors on magnitude F in particular scopes of its universum. An identification of the type of factors which are recognized as the main ones does not, however, mean that all those factors are the main ones for magnitude F in its whole universum. They can be main factors only in some scopes of magnitude F. Comparative analysis allows the researcher to select, from a pool of main factors indicated with the use of a theoretical perspective, a group of those main factors which influence magnitude F in the given scope. I will elaborate on this in the next section. 5

Strategies of Comparative Analysis: An Attempt at an Explication within the Theoretical Framework of Idealizational Theory of Science

Let us, then, present an explication of three kinds of comparative analysis: the parallel comparative history, the contrast-oriented comparative history, and the macro-causal comparative method in the conceptual framework of the idealizational theory of science and determine their place in the construction of a scientific theory. Suppose that a researcher wants to explain why magnitude variable F defined on two objects, O1 and O2, has value n. Before the researcher chooses a comparative strategy he or she preliminarily delineates the object of the search. Based on the assumed ontological perspective, in the first stage of constructing a theory the researcher determines what factors influence the studied magnitude F and what factors have no impact on it. In the second stage of the construction of a theory the researcher delineates, based on the assumed theoretical perspective, what types of factors can be the main factors for the

Strategies of Comparative Analysis

195

studied phenomenon. After a pool of the main factors has been determined, the researchers states a hypothesis about whether magnitude F defined on two objects belongs to one or to two F-kinds, i.e. if it is subject to the influence of one or of two different main factors. If the researcher concludes that variable F defined two objects, O1 and O2, belongs to two separate F-kinds, then the ­contrast-oriented comparative method will be used. If the researcher concludes that it belongs to one F-kind, then the parallel comparative method will be used. Let us take a closer look at this. The researcher has identified a pool of main factors for magnitude F on two objects, O1 and O2. They are: A, B, C, D, E, G H, I, and J. Next, the researcher concluded that magnitude F on objects O1 and O2 belongs to two separate F-kinds and uses the contrast-oriented comparative method. The set of main factors in the essential structure of magnitude F on object O1 includes A, B, C, D, E, and G, whereas the set of main factors of the essential structure of magnitude F defined on object O2 includes A, B, C, H, I, and J. O1: A, B, C, D, E, G O2: A, B, C, H, I, J We will call factors A, B, and C common factors for magnitude F defined on both objects and we will call factors D, E, G and H, I, J particular factors for variable F. Based on that decision each factor belonging to common factors is a secondary factor for magnitude F (although in the other range of the universum of magnitude F the factors recognized as secondary ones are the main factors and, therefore, they belong to the selected set of main factors). The accepted research hypothesis makes the researcher look for the main factors for magnitude F among particular factors. In the case presented above the main factors for magnitude F on object O1 can be factors D, E, or G, and for magnitude F on object O2 – factors H, I, or J. Nevertheless, it ought to be emphasized that not every particular factor can be the main factor for magnitude F defined on a given object. Such factors may include both the secondary factors for magnitude F (which are the main factors for a different scope of that magnitude) and main factors for magnitude F in the studied range. All the compared determining factors, however, belong to the set of main factors determined by the researcher’s theoretical assumptions. This is, then, the cognitive efficiency of the contrast-oriented comparative analysis in light of the proposed explication. It allows us to narrow down the range of factors among which we can look for the main factors. The method turns out to be a useful heuristic tool in the second stage of constructing the

196

Brzechczyn

idealizational theory, after the initial division of factors into essential and unessential ones has been made and a theoretical perspective delineating what type of factors are the main ones has been adopted. The comparative method allows us to select those factors from a pool of main factors which are the main factors in a given scope of magnitude F. If the researcher concludes that magnitude F defined on two objects, O1 and O2, belongs to two F-species which form one F-kind, then only one and not two separate main factors are looked for. For that purpose the parallel comparative strategy is used. In such a case we have two sets of factors influencing magnitude F on two objects: O1: A, B, C, D, E, G O2: A, B, C, H, I, J On the basis of the adopted theoretical assumptions the researcher then searches for the main factor among the factors common for magnitude F on two objects. In the analyzed example these are factors A, B, and C. Particular factors can be the secondary determinants for magnitude F in the considered range (although they are main factors in other ranges and, therefore, belong to the set of the main factors selected on the basis of the theoretical assumptions). These are: D, E, and G influencing variable F on object O1, and H, I, and J influencing variable F on object O2. Once again, the parallel comparative method only narrows down the range of factors among which there can be ‘hidden’ main factors in a given range of magnitude F. Further hierarchization of factors and formulation of dependencies between magnitude F and those factors which influence the variable is directed by the theory used by the researcher. In order to further narrow down the range of factors among which there can be the main factors the researcher applies the macro-causal comparative method which combines the parallel and contrast-oriented methods. In the first stage the researcher finds that magnitude F defined on two objects, O1 and O2, belongs to the same F-kind. In order to discover the main factor the researcher uses the parallel comparative method. O1: A, B, C, D, E, G O2: A, B, C, H, I, J It follows that the main factor ought to be looked for among factors A, B, and C. The researcher then uses the contrast-oriented comparative method in ­order to further restrict the range of factors. The essential structure of ­variable

Strategies of Comparative Analysis

197

F ­defined on objects O1 and O2 is compared with the essential structure of ­variable F on object O3. The comparison must fulfill two conditions: – magnitude F on object O3 must take a value other than value n which is taken by variable F on two objects, O1 and O2, and – the essential structure of magnitude F on object O3 must be sufficiently ­similar to magnitude F defined on two objects, O1 and O2. This can be presented in the following way: O1: A, B, C, D, E, G O2: A, B, C, H, I, J O3: C, D, E, G, I, J The common factors for magnitude F on two objects, O1 and O2, are factors A, B, and C. This is where, then, the main factor(s) ought to be sought. The researcher conducts a case analysis in which magnitude F on object O3 assumes a different value than n (“negative case”) despite the “great similarity” of essential structure F defined on object O3 compared to essential structure F defined on objects O1 and O2. In the essential structure of the “negative case” there is factor C, preliminarily considered to be the main factor, and there are some secondary factors (D, E, G, I, J) but there are not the two remaining factors considered to be main ones, i.e. A and B. Since magnitude F assumed a value different than n we may presume that the absence of factors A and B was decisive in this case. In order to define which of the two factors, A or B, is the main one, the researcher may continue the comparison. Another control case has to be found: magnitude F defined on object O4 has a different value than n but its essential structure is sufficiently similar to magnitude F on the objects analyzed before. Let us create a symbolical formula of that process. O1: O2: O3: O4:

A, B, C, D, E, G A, B, C, H, I, J C, D, E, G, I, J B, D, E, G, H, J

In the essential structure of the second “negative case” factor B considered to be a main factor. The two remaining two main factors – A and C – are not there. Since, in the first control case, factors A and B were missing in the essential structure and now the missing factors are A and C, we may presume that the factor with the greatest influence on the fact that magnitude F on objects O1

198

Brzechczyn

and O2 and assumes value n is factor A. Where factor A occurred in the essential structure of magnitude F it assumed value n and where factor A was missing magnitude F assumed a value other than n. It appears, then, that the macro-causal comparative method is more effective than the two other methods in that it allows us to restrict more precisely the range of factors among which we can look for the main factors of a given phenomenon. However, its effectiveness, as well as the effectiveness of the other methods, depends on certain hidden theoretical assumptions which are very rarely, if ever, fulfilled in social and historical sciences. First, it is difficult to find in historical science such study cases which would only differ in one factor within the researcher’s area of interests.3 Second, the assumption that the influence of factors on a given magnitude is independent from the intensity of the factors with an influence on the variable is a simplification.4 Third, it is presumed that the factors do not interact.5 Fourth, the studied cases are expected to be independent from one another.6 Such conditions are rarely fulfilled in historical science, which limits the applications of comparative analysis or influences the degree to which it can be reliable. 6 Conclusions In the light of the explication presented above, comparative analysis, or at least the version of used in historical-comparative sociology and illustrated 3 Mill himself was conscious of this fact: he believed his canons would find applications ­primarily in experimental and not social sciences. 4 The condition is not fulfilled in Skocpol’s work because in her narrative she changes the status of factors understood as constant into nominal and ordinal factors. For example, she introduces four kinds of societies based on agrarian bureaucracy: a semi-bureaucratic absolute monarchy, a highly-bureaucratic absolutist state, a semi-bureaucratic absolutist state, and a highly-bureaucratic state. These are treated as separate nominal factors. The author classifies the states of external pressure as moderate, mild, strong, and extreme. The transformations are done intuitively so the influence of those sub-factors on the studied social phenomenon (an outbreak of a revolution) is unknown. A methodological analysis of that aspect of Skocpol’s theory can be found in: Mahoney 2000, pp. 389–390, 404–405. 5 For an explication of the interactions of factors in the conceptual framework of the idealizational theory of science see: Brzeziński 1975, Brzechczyn 1998, pp. 101–103, 2009, pp. 147–148. 6 Such an assumption is introduced by Skocpol (1979, p. 39) in reference to the French, Russian, and Chinese revolutions. The assumption is obviously counterfactual as the 1789 revolution in France influenced the later, 19th-century revolutions in Europe, at least with respect to ideology, and the Chinese revolution was modeled on the Bolshevik one.

Strategies of Comparative Analysis

199

with the example of States and Social Revolutions, is not a theoretically neutral, purely inductive method. Theoretical assumptions determine the choice of a research problem and of study cases (Ragin 1981, p. 117), the identification of essential factors, and the selection of the main factors from the pool of essential factors. Comparative analysis is, however, useful for determining the main factors in a given range of magnitude F. If a researcher assumes that the analyzed research cases belong to different kinds he or she uses the contrast-oriented method and looks for the main factors within the set of particular factors occurring in the essential structures of the compared cases. If a researcher assumes that the analyzed research cases belong to the same kind he or she uses the parallel method and looks for the main factors within the set of common factors occurring in the essential structures of the compared cases. To sum up, comparative analysis is useful when the researcher already has some ontological and theoretical assumptions, and knows what elements are to be compared and for what purposes. The explication allows us to demonstrate why “some variables are more privileged than others” (Nichols 1986, p. 177) and to interpret the stages of Skocpol’s work more accurately than has been done by the author herself. Contrary to the author’s methodological comments, we can distinguish three and not two stages in her work. In the first stage, before the author started work on the book, she assumed a theoretical perspective which combined the Marxist concept of class relationships and the theory of political conflict (Skocpol 1979, p. 13). In the second, empirical stage of her work, based on the literature on the subject, the author examined the influence of factors which she presumed to be the main ones, within her theoretical perspective. Using comparative analysis she selected from the set of main factors those which exerted an influence on the outcome in the cases she studied. In the third stage of her work she analyzed negative cases, further restricting the set of main factors and testing the hypotheses put forth in her work. If we view States and Social Revolutions as a representative case, then in the case of theories developed within the historical-comparative sociology we do not generally see consciously introduced idealizational assumptions, on the basis of which idealized models of historical development could be built and later concretized. It would be difficult, then, to view those theories as a ­manifestation of a consciously applied idealizational method. However, the methodology of that scientific sub-discipline can be, as we can see, e­ asily ­reconstructed in the conceptual framework of the idealizational theory of ­science. How, then, can we explain such a paradox?

200

Brzechczyn

Francesco Coniglione has distinguished four types of scientists’ attitudes: 1. Not using idealization in scientific practice and not being methodologically conscious of the possibility of using such a method; 2. Using idealization in scientific practice but not being methodologically conscious of doing it (or interpreting it in the wrong way); 3. Using idealization in scientific practice and being methodologically conscious of it; 4. Not using idealization in scientific practice but being methodologically ­conscious of the possibility of using such a method, in non-scientific fields of knowledge (Coniglione 2004, p. 63). It seems that Skocpol’s case represents type (ii) of a scientific approach, as States and Social Revolutions are not, in any respect, a typical factographic history but we can find a clear model of a revolution in the book. According to Skocpol’s methodological declarations, her theory of revolution is an example of the application of Mill’s methods of agreement and difference. Some commentaries on Skocpol’s work (Nicholas 1986, Hanzel 2010, p. 332) emphasize the fact that those declarations are not in line with the author’s actual research practice, which has been indirectly admitted by Skocpol herself (1986, p. 188). Leszek Nowak distinguished two stages of the development of science: the empirical-gathering stage and theoretical-explanatory stage. In the first stage the method of systematic observation of facts and generalization from those facts is used; the basic cognitive method at that stage is induction. In the ­second stage the gathered empirical material is explained; the basic cognitive method here is idealization which leads to the construction of sophisticated theoretical models. In different sciences that methodological breakthrough occurred at different times. In physics it was at the beginning of the 17th century, thanks to Galileo, in biology and economy in the 19th century – thanks to, respectively, Darwin and Marx, in linguistics it was in the 20th century, thanks to Chomsky. According to Nowak, though: “the transition has not occurred in all empirical sciences yet. Numerous disciplines in the humanities are still in the first stage of development, gathering the empirical data necessary for proper explanation of their subjects of study, which can, however, only be done with a theory of a different nature than a set of empirical generalizations” (Nowak 1977a, p. 61). Nevertheless, it seems that the division of science into two stages is too crude and it ought to be supplemented with a transitional, pre-theoretical-­ explanatory stage at which science already sets theoretical-explanatory objectives and avails itself of certain rudimentary forms of idealization which are interpreted inductively.

Strategies of Comparative Analysis

201

When we characterize the stages of science mentioned above with the use of Coniglione’s classification of approaches to science we can say that in the ­empirical-gathering stage the first type dominates (researchers do not apply idealization to scientific practice and do not have methodological consciousness of the possibility of using that method), in the theoretical-explanatory stage it is the third approach (researchers apply idealizations to scientific practice and have methodological consciousness of the possibility of using that method), and in the transitional pre-theoretical-explanatory stage – the second approach (researchers apply idealizations to scientific practice but do not have methodological consciousness of the possibility of using that method). It seems, then, that it is this transitional pre-theoretical-explanatory stage in the development of science that is currently represented by historical-­comparative sociology, or, more precisely, its theory of revolution. References Abrams, Ph. (1982). Historical Sociology, Ithaca, NY: Cornell University Press. Bonnell, V.E. (1980). The Uses of Theory, Concepts and Comparison in Historical Sociology. Comparative Studies in Society and History 22, 156–175. Braembussche, A.A. van den, (1989) Historical Explanation and Comparative Method: Towards a Theory of the History of Society. History and Theory 28, no. 1, 1–24. Brzechczyn, K. (1998). Odrębność historyczna Europy Środkowej. Studium metodologiczne [Historical Distinctiveness of Central Europe. A Methodological Study]. Poznań: Humaniora. Brzechczyn, K. (2009), Methodological Peculiarities of History in the light of Idealizational Theory of Science. In: idem (ed.) Idealization XIII: Modeling in History. Poznań Studies in the Philosophy of the Sciences and the Humanities 97, pp. 137–158. Amsterdam-New York: Rodopi. Brzeziński, J. (1975). Interaction, Essential Structure, Experiment. Poznań Studies in the Philosophy of the Sciences and the Humanities 1, pp. 43–58. Brzeziński, J., I. Burbelka, A. Klawiter, K. Łastowski, S. Magala. L. Nowak, W. Patryas (1976), Prawo, teoria, sprawdzanie [Law, Theory Testing]. Poznańskie Studia z Filozofii Nauki 1, pp. 107–133. Burawoy, M. (1989). Two Methods in Search of Science. Skocpol versus Trotsky. Theory and Society 18, 759–805. Coniglione, F. (2004). Between Abstraction and Idealization: Scientific Practice and Philosophical Awareness. In: F. Coniglione, R. Poli, R. Rollinger (eds.) Idealization XI: Historical Studies on Abstraction and Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities 82, pp. 59–110. Amsterdam-New York: Rodopi.

202

Brzechczyn

Emigh R.J. (1997). The Power of Negative Thinking: The Use of Negative Case Methodology in the Development of Sociological Theory. Theory and Society 26, 649–684. Goldstone, J. (2003) Comparative Historical Analysis and Knowledge Accumulation in the Study of Revolution. In: J. Mahoney, D. Rueschemeyer (ed.) Comparative Historical Analysis in the Social Sciences, pp. 41–90. Cambridge: Cambridge University Press. Hanzel, I. (2010). Studies in the Methodology of Natural and Social Sciences. Frankfurt am Main: Peter Lang. Hechter, M., E. Kiser (1991). The Role of General Theory in Comparative-Historical Sociology. American Journal of Sociology 97, 1–30. O’Kane, R.H.T. (1995). The National Causes of State Construction in France, Russia and China. Political Studies 43 (2), 2–21. Lange, M. (2013). Comparative-Historical Methods. Thousands Oaks, CA: Sage. Mahoney, J. (2000). Strategies of Causal Inference in Small-N Analysis. Sociological Methods and Research 28, pp. 387–424. Mahoney, J., D. Rueschemeyer (2003) (ed.) Comparative Historical Analysis in the Social Sciences. Cambridge: Cambridge University Press. Mill, J.S. ([1843] 1974). A System of Logic. Toronto: University of Toronto Press. Nichols, E. (1986). Skocpol on Revolution: Comparative Analysis vs. Historical Conjuncture. Comparative Social Research 9, 163–186. Nowak, L. (1970). O zasadzie abstrakcji i stopniowej konkretyzacji [On the Principle of Abstraction and Gradual Concretization]. In: J. Topolski (ed.) Założenia metodologiczne “Kapitału” Marksa, pp. 123–218. Warszawa: KiW. Nowak, L. (1971). U podstaw Marksowskiej metodologii nauki [Foundations of the Marxian Methodology of Science]. Warszawa: PWN. Nowak, L. (1974). Zasady marksistowskiej filozofii nauki. Próba systematycznej rekonstrukcji [The Principles of Marxist Philosophy of Science. An Attempt at a Systematic Reconstruction]. Warszawa: PWN. Nowak, L. (1977a). Wstęp do idealizacyjnej teorii nauki [An Introduction to the Idealizational Theory of Science]. Warszawa: PWN. Nowak, L. (1977b). U podstaw dialekty ki Marksowskiej. Próba interpretacji kategorialnej [Foundations of Marxian Dialectics. Towards a Categorial Interpretation]. Warszawa: PWN. Nowak, L. (1979). Problem adekwatności teorii w idealizacyjnej teorii nauki. In: K.  Łastowski, L. Nowak (red.), Konfrontacje i parafrazy, pp. 29–38. WarszawaPoznań: PWN. Nowak, L. (1980). The Structure of Idealization. Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht: Reidel. Nowak, L. (1992). The Idealizational Approach to Science: A Survey. In: J. Brzeziński, L.  Nowak (eds.) Idealization III: Approximation and Truth. Poznań Studies in the

Strategies of Comparative Analysis

203

Philosophy of the Sciences and the Humanities 25, pp. 9–67. Amsterdam-New York: Rodopi. Nowak, L., I. Nowakowa (2000). The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities 69. Amsterdam-New York: Rodopi. Paprzycka, K. (2005). O możliwości antyredukcjonizmu. Warszawa: Semper. Paprzycka, K., M. Paprzycki (1992), Accuracy, Essentiality and Idealization. In: J. Brzeziński, L. Nowak (eds.) Idealization III: Approximation and Truth. Poznań Studies in the Philosophy of the Sciences and the Humanities 25, pp. 9–67. ­ ­Amsterdam-New York: Rodopi. Ragin, Ch. (1981). Comparative Sociology and the Comparative Method. International Journal of Comparative Sociology 22, 102–120. Skocpol, T. (1979). States and Social Revolution. Cambridge: Cambridge University Press. Skocpol, T. (1986). Analysing Causal Configuration in History: a Rejoinder to Nichols. Comparative Social Research 9, 187–194. Skocpol, T., M. Somers (1980). The Uses of Comparative History in Macro-Social Inquiry. Comparative Studies in Society and History 22, 156–173. Skocpol, T., (1984a). Emerging Agendas and Recurrent Strategies in Historical Sociology. In: Th. Skocpol (ed.), Vision and Method in Historical Sociology, pp. 356–391. Cambridge: Cambridge University Press. Skocpol, T., (ed.) (1984b). Vision and Method in Historical Sociology. Cambridge: ­Cambridge University Press. Smith, D. (1991). The Rise of Historical Sociology. Cambrigde: Polity Press. Tilly, Ch. (1981). As Sociology Meets History. New York: Academic Press. Tilly, Ch. (1984). Big Structures, Large Processes, Huge Comparisons New York: Russell Sage Foundation. Zajęcki M. (2013). Założenia esencjalistyczne idealizacyjnej teorii nauk [Essentialist Assumptions of Idealizational Theory of Science]. Studia z Filozofii Polskiej 8, 1–22.

chapter 9

The Structure of Idealization in Noam Chomsky’s Generativist Theory Barbara Konat Abstract The early work of Noam Chomsky has been described by Leszek Nowak as the “­idealization turn in linguistics” (Nowak 1971; 1976; 2000). Nowak claims that only ­after Chomsky has linguistics become a mature science. Prior to Chomsky, structuralist ­distributivism had focussed mainly on linguistic data gathering and on the properties of particular languages. Along with the introduction of the concept of linguistic competence, Noam Chomsky set forth idealized models to linguistics. This paper ­further develops Nowak’s analysis on Chomsky’s work, proposing a description of the structure of idealization in Chomsky’s Generativist Theory. The paper reconstructs – in terms of the Idealizational Theory of Science – the essential structures underlying Chomsky’s theory. I also put forward the claim that the hierarchy of factors described by Chomsky determines the process of concretization in the Generativist Theory. This hierarchy also shapes current research methods in generative linguistics

1 Introduction The aim of this article is to propose a new reconstruction of Noam Chomsky’s linguistic work based on reconstruction formulated by Leszek Nowak. I a­ ssume the Idealizational Theory of Science (henceforth: its), allowing for expanded analysis of the internal structure of Chomskyan theory. As a result, the essential structure of Generative Grammar (henceforth: gg) is reconstructed. Laws and theories in science do not aim at reflecting the world’s phenomena with the highest accuracy. They rather express a certain system of a researcher’s beliefs and observations resulting in a complex hierarchy of observed factors. This claim is the source of the Idealizational Theory of Science (Nowak, 1970; 1971; 1974; 1977; 1980). At the center of Nowak’s approach lies the idea of the gradual processes of idealization and concretization. In Nancy Cartwright’s words, Nowak’s idealization is about “how to get from an abstract law to the concrete laws that make up its phenomenal content” (Cartwright 1989, p. 197). Cartwright (1983) claimed that the laws of physics lie. Nevertheless, physicists are not the only scientists who can be suspected of lying. The laws of © koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_011

Idealization in Noam Chomsky’s Generativist Theory

205

Chomskyan linguistics also “lie”, in Cartwright’s sense. By “lying” we mean here idealization – an attempt to simplify the picture of the phenomenon under investigation, by removing the influence of disturbing factors. This method has been known in science since Galileo. Idealization in Cartwright’s sense is a process which precedes the formulation of a scientific law: [I]n idealization we start with a concrete object and we mentally rearrange some of its inconvenient features – some of its specific properties – before we try to write down a law for it. cartwright 1989, p. 187

This article presents how the method of idealization and gradual concretization, as described by Leszek Nowak, can be used in an analysis of Chomskyan linguistics, and can provide new explanations for the changes which generativist theory has undergone. Even though some elements of Chomskyan linguistics have changed over the years, the essential structure of the theory seem to have survived in its original form. What constitutes this essential structure? What is the general methodological nature of Noam Chomsky’s linguistic t­heory? Which idealization and concretization procedures were used by him? These questions have not been answered yet. The history of linguistics shows us, that the issue of gradual idealization as an internal structure of Chomskyan linguistics seems to be of great importance as Chomsky was the first to challenge the vision of purely empirical linguistics, with no assumptions or idealizations. 2

Leszek Nowak’s Work on Chomskyan Linguistics

Noam Chomsky is a linguist and a philosopher who founded the linguistic ­orientation named Generativism. The generativist movement has been supported by numerous scholars, and has gone through several significant changes over the years. Yet the basic ideas formulated by Noam Chomsky in the early 1960’s thrive and constitute the methodological and philosophical identity of the current movement. The Chomskyan chapter in linguistics started with his publication of “A Review of B.F. Skinner’s Verbal Behavior” (1959). Shortly after, Chomsky started publishing a series of works, which currently mark stages in the development of generativist theory (Chomsky, 1965; 1966; 1981; 1995). For the purpose of this analysis, we list five key stages in the development of generativism: 1. Early Transformational Grammar, 2. Standard Theory, 3. Government and Binding, 4. The Minimalist Program and 5. Phase Derivation. The analysis presented here focuses on theoretical basis formulated during the two initial phases.

206

Konat

Leszek Nowak mentioned Chomskyan linguistics as an example of idealization in the social sciences in most of his books and articles. However, in two articles Nowak analyzed Chomsky in more detail, and these works are summarized below. In the article “Uwagi o stosunku logiki do językoznawstwa [On the relation between logic and linguistics]” (1973), Nowak discusses Kazimierz Ajdukiewicz’s thesis, concerning the methodological relation between linguistics and logic. According to Ajdukiewicz, the logical theory of language differs from linguistics with regards to the research subject. Subjects of the investigation of logic are artificial languages, whereas subjects of the investigation of linguistics are natural languages. Ajdukiewicz claims that artificial languages are, in fact, idealizations of natural languages. He further concludes that between logic and linguistics there is a relation of correspondence. Leszek Nowak argues with Ajdukiewicz from the perspective of the Idealizational Theory of Science, and proposes his own interpretation of the relation between logic and Chomskyan linguistics. In the interpretation, Nowak refers to the Chomskyan ideal speaker-hearer construction, in which Chomsky assumes an ideal, homogenous speech community and an ideal speaker that makes no errors (Chomsky 1965, p. 3). In response to Ajdukiewicz, Nowak concludes that in both linguistics and the logical theory of natural language, idealizing assumptions are present (Nowak 1973, p. 83). What is different between logic and linguistics, according to Nowak, is the fact that logical theory of language relies on many more idealizing assumptions than Chomskyan linguistics. This is why linguistics is less abstract (that is – more concrete) than the logical theory of natural language. Nowak suggests that it is the relation of concretization (and not correspondence, as in Ajdukiewicz’s thesis) that occurs between the logical theory of language and Chomskyan linguistics (Nowak 1973, p. 84). A second analysis is provided by Nowak in the study: “Ajdukiewicz, Chomsky and the Status of the Theory of Natural Language” (2000). Nowak argues there, in more detail, with Ajdukiewicz’s claim that linguistics empirically describes natural languages without any idealizations. Nowak disagrees with that view, and writes: In fact, the linguistic theory of Noam Chomsky seems to be a typical example of idealizational concepts. First and foremost, it is constructed with methodological self-consciousness which is rare among scientists. nowak 2000, p. 326

This rare methodological self-consciousness of Chomsky is visible in his ­criticisms of the “purely inductive” structures proposed by structuralistic distributivism. Nowak reconstructs Chomsky’s argument against ­distributivism

Idealization in Noam Chomsky’s Generativist Theory

207

in the following way: “Observation, classification and induction are ­unable to separate the factors significant for the linguist, and therefore they cannot serve properly as a mean to build a real linguistic theory” (Nowak 2000, p. 327). In his reconstruction of Chomsky’s theory, Nowak describes a series of grammars which determine the steps in idealization: (1) generative grammar (2) generative-transformative grammar (3) generative-transformative-­phonetic grammar (Nowak 2000, p. 329). According to Nowak, these grammars form a series of models, in which the first one (generative) is the most abstract, and further models are more concrete. It is worth mentioning here, that Nowak focused on the relation between linguistics and the logical theory of natural languages, rather than on the internal structure of Chomskyan linguistics. Nowak’s ideas, presented in the two above studies are also mentioned briefly in his other works. For example, the three models of Chomskyan linguistics are described in the article “On the Hidden Unity of Social and Natural S­ ciences” (Nowak [1998] 2012). Analysis of the internal structure of Chomskyan theories is presented in (Nowak 1974). He analyses the essential structure of Generativism and compares it to the essential structures found in other social sciences. He starts with the origin of Chomsky’s theory, which was the refusal of traditional linguistics: We can see now, what Chomsky reproaches taxonomists for: that they simply describe what can be observed in language behavior, instead of presuming principal factors and building a theory of language that will rely on those principal factors.1 nowak 1974, p. 86

Nowak further identifies and analyses the essential structure of Chomskyan theory, by pointing out the linguistic competence as the principal factor influencing all acts of producing and understanding the language (Nowak 1974, p. 87). Nowak acknowledges there the existence of possible secondary factors, but does not go further into analysis. To sum up the description of Nowak’s work on Chomskyan linguistics, we may claim that, even though Nowak believed generative grammar to be a great example of idealization, he was interested in the relation between linguistics and logic, rather than the internal structure of linguistics. In the context of a general methodology of science, Nowak presented his view on the importance 1 Original quote in Polish: „Widzimy więc, co Chomsky zarzuca taksonomistom: że po prostu opisują to, co daje się w zachowaniach językowych zaobserwować miast wyróżnić w nich czynniki główne i budować teorię uzależniającą rozumienie języka od owych czynników głównych” (Nowak 1974, p. 86). Translation: b.k.

208

Konat

Galileo

Physics

Economy

Marx

Biology

Darwin

Chomsky

Linguistics XVII c.

XIX c.

XX c.

Figure 9.1 Idealization thresholds in science (Nowak 1977, p. 60)

of Chomsky’s work in the diagram, showing the thresholds of idealization in various scientific disciplines (see Figure 9.1.). Nowak’s remarks on Chomsky, usually serve as a means to other aims.2 Nowak’s reconstruction seems to be incomplete, and raises at least two questions: 1. What is the exact formal reconstruction of the ideal speaker-hearer assumption in terms of its? 1. Has this basic assumption survived in the next generations of gg? The application of the its to selected works of Chomsky, presented in further sections of this article, aims at answering those questions. 3

Idealizational Theory of Science

The Idealization Theory of Science was developed by Leszek Nowak. In his work he adopted an essentialist perspective on science. The main source of 2 It may seem odd, that such a great methodologist as Leszek Nowak has not analyzed in more detail the work of another great thinker – Noam Chomsky – and has not prepared a larger study on him. The explanation of this unusual situation was provided to this paper’s author by Prof. Krzysztof Łastowski – Leszek Nowak’s former co-worker. In personal communication he claimed that, in fact, Nowak prepared a book on Chomsky, but the manuscript has been lost.

Idealization in Noam Chomsky’s Generativist Theory

209

i­nspiration was the Marxian method – the method of idealization. In the words of Nancy Cartwright: “The most detailed work on abstraction has been done by Leszek Nowak, as part of his reconstruction of Marx’s thought in Das Kapital” (Cartwright 1989, p. 202). In the first pages of “The Structure of Idealization” Nowak presents a case study of the Marxian law of value. Let us now look closer at this presentation, as it will introduce basic notions in the Idealizational Theory of Science. ­According to Nowak, the law of value is a “typical theorem of Marx in C ­ apital” (Nowak 1980, p. 3). Nowak claims that the Marxian law of value is, in fact, based on an assumption that in an economy supply and demand are balanced. Nowak further claims that Marx was aware of the counterfactual nature of this proposal, and yet he still decided to call the law of value a “law”. Nowak claims that this counterfactual statement was not a mistake or accident in Marx’s work, but rather was a part of his methodological strategy: a strategy of stepby-step abstraction. In fact, according to Marxian methodology, a mistake in science would be to examine the world in its particular form. Nowak claims that the economic system, in Marxian description “­resembles ideal gases, perfectly rigid bodies, and other constructs of the type” (Nowak 1980, p. 9). Following the idealization procedure present in the natural sciences, Marx wanted to base his economy on the principle of abstraction. This principle includes numerous unrealistic assumptions, put forth to describe what is essential, not phenomenal and particular in an economy. Marx accused other economists of the time of lacking the “force of abstraction” (Nowak 1980, p. 9). When conducting a detailed analysis of the Marxian method, Nowak discovered that the method of idealization, although used in economics for the first time by Marx, is also used in other social sciences and humanities. Based on procedures observed in a Marxian economy, Nowak has built his own concept of the scientific theory which is a series of models. The first model is the most idealizing one. This model contains numerous unrealistic assumptions. Let us consider the situation where the observed factor F is under the influence of three factors: H, p1 and p2 (in the set Z, which is the set of objects under investigation). We label the principal factor (that is the most influential factor) H, whereas all the secondary factors, ordered by the strength of their influence, will be labelled p1 and p2. The essential structure of the factor F will thus look as follows: ​S​ ZF​ ​ : (2) H (1) H, p1 (0) H, p1, p2.

210

Konat

In the essential structure we can see that the influence of the principal factor H on the factor F is stronger than the influence of the side factors p1 and p2. On the phenomenal level (0), the influence of all factors is taken into account, however on the essential level (2) only the principal factor H plays a role. According to the Idealizational Theory of Science, the determination of the essential structure of factors, as presented above, is the first step in the construction of a scientific theory (Nowak 1974, p. 37). The next step is the formulation of an idealizational statement. In its first, most idealized form, the scientific law claims that only the primary factor H has an influence on the observed factor F, and the influence of all the secondary factors is minimal (and in particular, equals zero). The most idealized model, based on the essential structure will therefore be presented as follows: (​ ​T ​ 1)​ ​ if Z​(x)​ and ​p​ ​(x)​ = 0 and ​p​  ​(x)​ = 0 then F​(x)​ = f​(H​(x)​)​.​ 1

2

In the example (T 1), the two formulas in the antecedent: p1(x) = 0 and p2(x) = 0 represent two idealizing assumptions. The expression Z(x) represents ­realistic description (assumption), which determines the scope of predication of a given law (for example “x is a rigid body”). The function f describes the dependence between the observed factor F and the principal factor H. Even though the proposed law counterfactually assumes the minimal influence of secondary factors, it properly describes the essential relation between the o­ bserved factor F and its principal factor H. The second step in formulating the theory is the removal of one of the idealizing assumptions, together with the introduction of the concretization assumption. According to the essential structure presented above, the second most influential factor after factor H is factor p1. Therefore, the idealizing ­assumption p1(x) = 0 should be removed. In the second step we have an idealizational statement: (​ ​T ​ 2)​ ​ if Z​(x)​ and ​p​  ​(x)​ ≠ 0 and ​p​  ​(x)​ = 0, then F​(x)​ = f’​(H​(x)​, ​p​  ​(x)​)​.​ 1

2

1

The theorem (T 2) contains one concretizational statement: p1(x) ≠ 0 in the place of the idealizational one. Since the influence of the factor p1 is no ­longer ­assumed to be minimal, this influence is included in the consequent of (T 2). The corrective function f ’ is added, which symbolizes the influence of the ­factor p1. The last step in formulating a scientific theory is the removal of all idealizing assumptions. In the analyzed example, this will mean the removal of the second and last idealizing assumption: p2(x) = 0. The final concretization leads

Idealization in Noam Chomsky’s Generativist Theory

211

to the formulation of a factual statement in which the influence of all principal and secondary factors are taken into account: ​(​T ​ 3​)​ if Z​(x)​ and ​p​ 1​(x)​ ≠ 0 and ​p​ 2​(x)​ ≠ 0, then F​(x)​ = f’’​(H​(x)​, ​p​ 1​(x)​, ​p​ 2​(x)​)​.​ According to Nowak, a scientific theory consists of a series of statements of the type (T 1), (T 2), (T 3). The series starts with an idealizational statement (T 1), then includes the first concretization (T 2), leading to the factual statement (T 3). The steps in this process are determined by the essential structure of ­factors. The scientific theory in its is then, a series of models – the first model is the most idealized (or deformed), and the further models, via concretizations, move closer and closer to reality. To sum up, we should point to the fact that the Marxian method, together with the Idealization Theory that follows, differ from other methodological schools. As Nowak recognizes – The Marxian law of value is not a law in the positivist view. It has something new to say about science. In Marx’s theory, science is somehow paradoxical from the viewpoint of common sense (Nowak 1980, p. 9). As we shall see, Noam Chomsky’s concept of language is also based on counterfactual assumptions, and it stands in opposition to common intuitions about language. 4

The Steps of Idealization in Chomskyan Linguistics

Basic philosophical assumptions of generative linguistics were presented by Noam Chomsky in his Aspects of the Theory of Syntax (Chomsky 1965). Representing The Standard Theory stage in generativist theory (1962–1966) this book presents the crucial moment of the formulation of the idealizing assumption. Barbara Stanosz, claimed that “Aspects of the Theory of Syntax presents the core of generativist theory, and whatever comes later will be based on that core” (Stanosz 1977, p. 18). Twenty years later, Lyons (Lyons, 1998) confirmed her expectations, declaring that all the critics of Chomsky still refer to the basic theoretical core of Aspects of the Theory of Syntax (Chomsky 1965). As Chomsky himself declares, his aim in Aspects was to present the “main background assumptions” of Generativism (Chomsky, 1965, p. 3). Of those ­assumptions, the first, and most famous, placed on the first page, is the concept of the ideal speaker-hearer: Linguistic theory is concerned primarily with an ideal speaker-listener, in a completely homogeneous speech-community, who knows its language

212

Konat

perfectly and is unaffected by such grammatically irrelevant conditions as memory limitations, distractions, shifts of attention and interest, and errors (random or characteristic) in applying his knowledge of this language in actual performance. chomsky 1965, p. 3

The idealization proposed by Chomsky in this paragraph resembles the ­Marxian counterfactual constructions in economy, as described by Nowak. A completely homogenous speech community cannot be found anywhere in the world. Also, no language user is unaffected by memory limitations, distractions or shifts of attention, as those are part of the human cognitive system. Noam Chomsky was completely aware of the idealizational character of his work and he knew that language production is influenced by many other factors: To study actual linguistic performance, we must consider the interaction of a variety of factors, of which this underlying competence of the ­speaker-hearer is only one. In this respect, the study of language is no different from the empirical investigation of other complex phenomena. chomsky 1965, p. 4

Chomsky introduced the concepts of “competence” and “performance” to show the idealizational character of his own work: We thus make a fundamental distinction between competence (the speaker-hearer’s knowledge of his language) and performance (the actual use of language in concrete situations). Only under the idealization set forth in the preceding paragraph is performance a direct reflection of competence. chomsky, 1965, p. 4

Definition of “competence” as a “speaker’s knowledge” highlights the mentalistic character of Chomsky’s theory: he was not interested in language production itself, but in the mental processes underlying this process. The main ­assumptions of generative linguistics presented by Noam Chomsky in “­Aspects” can be summarized as follows: 1. The main purpose of linguistic research is an explanation of mental processes taking part in the mind of the language user, that is the description of her linguistic competence.

Idealization in Noam Chomsky’s Generativist Theory

213

2. Acceptability belongs to the level of linguistic performance and is gradable, whereas grammaticality belongs to the level of linguistic competence. 3. Linguistic competence of a language user is not directly observable and it can be analyzed only through research into the performance level. 4. Performance is influenced by – at least – two factors: a. influence of grammar, b. cognitive factors. 5. The influence of factors other than grammatical ones is accidental and can be ignored. 6. Competence would equal performance only in the case of an ideal speaker-hearer. 7. Generative grammar consists of three main elements (phonological, syntactic and semantic). The grammaticality level is thus a level of adjusting the utterance to those three factors. 8. Non-linguistic, cognitive factors also influence linguistic performance. This influence is gradable. Based on the reconstruction of the above-presented claims, let us propose here a new reconstruction of the Chomskyan theory in terms of the its. The labels used in the formulas are: U(x) – a predicate “x is an utterance” (realistic assumption); G(x) – grammaticality of x; c(x) – cognitive factors; A(x) – acceptability of x; f – function. The essential structure of factor A (acceptability of the utterance) in the ­generative approach presents as follows:

​S​ UA​ ​ : (1) G (0) G, c

In the essential structure we can see that the influence of the principal factor G on factor A is stronger than the influence of the secondary factor c. On the phenomenal level (0), the influence of all factors is taken into account. ­However, on the essential level (1) only the principal factor G plays a role. That is, in Chomskyan linguistics, on the phenomenal level – the level of ­performance – the acceptability of an utterance depends on two factors: grammaticality and the cognitive factor. However on the essential level – the level of

214

Konat

competence – the influence of secondary factors is ignored and acceptability depends solely on grammaticality. In its first, most idealized form, Chomskyan theorem (T 1) claims that only the primary factor G has an influence on the observed factor A, and the influence of all the secondary factors is minimal. The most idealized model, based on the essential structure presents as follows: ​(​T ​ 1​)​ If U​(x)​ and c​(x)​ = 0 then A​(x)​ = f​(G​(x)​)​.​ The Chomskyan theorem (T 1) reads: if “x” is an utterance, and the influence of cognitive factors on that utterance production is minimal (and particularly equals zero), then the acceptability of this utterance depends solely on its grammaticality (in the way described by the function f). The formula in the ascendant and c(x)  =  0 represents the idealizing assumption. The expression U(x) represents a realistic description which restricts the law only to utterances. The function f describes the connection between the observed factor A and the principal factor G. Even though the proposed law counterfactually assumes the minimal influence of secondary factors, it properly describes the essential relation between the observed factor A and its principal factor G. The step by step concretization determined by the structure leads to the ­formulation of the factual statement (T 2), in which the influence of all principal and secondary factors are taken into account: (​ ​T ​ 2)​ ​: If U​(x)​ and c​(x)​ ≠ 0 then A​(x)​ = f’​(G​(x)​, c​(x)​)​.​

The factual statement (T 2) reads as follows: if “x” is an utterance, and the influence of cognitive factors on that utterance production is greater than zero, then the acceptability of that utterance depends on the combined influence of its grammaticality and cognitive factors (in the way described by the function f’). As we can see, during the process of step by step idealization and concretization, the idealizing assumption c(x) = 0 has been removed. 5

Conclusion – Concretization Procedures in Chomskyan Linguistics

The analysis proposed above confirms Nowak’s intuitions about Chomskyan linguistics. Generativism is, indeed, essentialist in nature. Chomsky considered grammaticality to be a principal factor in the acceptability of utterances. The analysis reveals, that Chomsky was not only aware of the idealizational

Idealization in Noam Chomsky’s Generativist Theory

215

­ rocedures, but he also included procedures of step by step concretization. p These results shed new light on further developments of Chomskyan linguistics. Confirmation of the presence of (T 2) in generative linguistics can be found in his later work. Chomsky (2002) clearly shows the structure of the process of concretization in linguistics by introducing the notions of ­Faculty of Language in the Narrow Sense (fln) and Faculty of Language in the Broad Sense (flb) (Hauser, Chomsky, & Fitch 2002). The first mechanism – fln, consists of the purely grammatical, recursive mechanisms required for language. The second mechanism – flb, includes cognitive and sensory-motor systems. In flb, secondary factors can be taken into account. In the context of the results presented above, we can postulate that the fln system will be described by (T 1), whereas the flb system – by (T 2). Application of the Idealizational Theory of Science to Chomskyan linguistics reveals the steps in idealization and concretization that influenced the development of generativism. Leszek Nowak called Marx “the Galileo of the social sciences” (Nowak 1980, p. 34). Noam Chomsky, thanks to his introduction of idealization and concretization to language studies, should be named the Galileo of linguistics. References Cartwright, N. (1983). How the Laws of Physics Lie. New York: Oxford University Press. Cartwright, N. (1989). Nature’s Capacities and their Measurement. Oxford/New York: Clarendon Press/Oxford University Press. Chomsky, N. (1959). A Review of B.F. Skinner’s Verbal Behavior. Language 35 (1), 26–58. Chomsky, N. (1965). Aspects of the Theory of Syntax. Cambridge: MIT Press. Chomsky, N. (1966). Cartesian Linguistics: A Chapter in the History of Rationalist Thought. New York/London: Harper and Row. Chomsky, N. (1981). Lectures on Government and Binding. Dordrecht/Cinnaminson: ­Foris Publications. Chomsky, N. (1995). The Minimalist Program. Cambridge: MIT Press. Hauser, M.D., N. Chomsky, & W.T. Fitch, (2002). The Faculty of Language: What Is It, Who Has It, and How Did It Evolve? Science, 298 (5598), 1569–1579. Lyons, J. (1998). Chomsky. Warszawa: Prószyński i S-ka. Nowak, L. (1970). O zasadzie abstrakcji i stopniowej konkretyzacji [On the Principle of Abstraction and Gradual Concretization]. In: J. Topolski (ed.) Założenia metodologiczne “Kapitału” Marksa, pp. 123–218. Warszawa: KiW. Nowak, L. (1971). U podstaw Marksowskiej metodologii nauki [Foundations of the ­Marxian Methodology of Science]. Warszawa: PWN.

216

Konat

Nowak, L. (1973). Uwagi o stosunku logiki do językoznawstwa [On the Relation ­between Logic and Linguistics]. Studia Semiotyczne 3, 75–84. Nowak, L. (1974). Zasady marksistowskiej filozofii nauki. Próba systematycznej rekonstrukcji. Warszawa: PWN. Nowak, L. (1976). Wykłady z Filozofii Marksistowskiej [Lectures on Marxist Philosophy]. Poznań: Wydawnictwo Naukowe UAM. Nowak, L. (1977). Wstęp do Idealizacyjnej Teorii Nauki [Introduction to the Idealizational Theory of Science]. Warszawa: PWN. Nowak, L. (1980). The Structure of Idealization: Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht/Boston/London: Reidel. Nowak, L. ([1998] 2012). On the Hidden Unity of Social and Natural Sciences. In: K. Brzechczyn, K. Paprzycka (eds.). Thinking about Provincialism in Thinking. Poznań Studies in the Philosophy of the Sciences and the Humanities, 100, pp. 15–50. Amsterdam-New York: Rodopi. Nowak, L. (2000). Ajdukiewicz, Chomsky and the Status of the Theory of Natural Language. In: I. Nowakowa and L. Nowak, Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and in the Humanities, vol. 69, pp. 325–337. Amsterdam-Atlanta: Rodopi. Stanosz, B. (1977). Próba rewolucji w lingwistyce [An Attempt at Revolution in ­Linguistics]. In: B. Stanosz (Ed.) Lingwistyka a filozofia [Linguistics and ­Philosophy]. Warszawa: PWN.

Part 3 Idealization in the Humanities

∵

chapter 10

The Method of Idealization and Concretization on the Ground of Negativistic Unitarian Metaphysics Krzysztof Kiedrowski Abstract The basic purpose of this paper is to present a reconstruction, criticism and reinterpretation of one of the constructional methods used by Leszek Nowak in building negativistic unitarian metaphysics. Nowak’s statement, that the basic method used by him is the method of idealization and concretization is unsustainable for the following reasons: postulated by Nowak structure of metaphysical doctrines fails to comply with the procedure of concretization on the grounds of the unitarian metaphysics, the sequence of models is in several cases arbitrary, the order of the notions introduced to the conception is to a certain extent arbitrary, and the method used does not lead to decreasing of the degree of the speculativeness of the system. In this article I propose to consider the above-mentioned method as a method of abstraction and disabstraction, and to characterize the different way of understanding the method of idealization and concretization. As a result of these procedures the content of ‘Being and Thought’ is regarded as the speculative first model of the negativistic metaphysics.

1

Introductory Comments

1.1 The Method of Idealization and Concretization The philosophical works of Leszek Nowak are distinguished by a methodological unity (Nowak 2012 [1998]). That is, the method of idealization and concretization as a basic method for both the developed sciences and ­humanistic disciplines, including philosophy. This method, Nowak says, was, for the first time, ­applied by Galileo in physics, in biology by Charles Darwin, and in economics (­humanities) by Karl Marx in Das Capital. From the 1960s onwards, the scientific ­community gathered around Nowak conducted extensive research on the method of i­dealization and concretization. An extensive comparison of the results of these studies was published in the work Idealization X: The Richness of Idealization (Nowakowa and Nowak 2000), the findings of which are the basis of the p ­ resent paper.

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_012

220

Kiedrowski

The method of idealization and concretization is based on two procedures: of idealization, understood as a directed deformation of a fact that is being investigated, and, of concretization which relies on the successive elimination of that deformation. The philosophical horizon for the above-mentioned ­procedures is the thesis of essentialism, that every fact embraces the essence (the set of principal factors for the investigated factor) and the ­phenomenon (the set of secondary factors for the investigated one).1 The procedure of ­idealization tends therefore to reveal principal factors and to show, along with the simultaneous lack of the secondary factors’ influence, how they ­affect the factor which is being investigated. This way that idealizational law is ­being ­proposed, as a law which is a counterfactual conditional as follows (­Nowakowa and Nowak 2000, p. 117): ​T​ k​ : if (G​(x)​ & ​p​ 1​(x)​ = 0 & ​p​ 2​(x)​ = 0 & ⋯ & ​pk−1 ​  ​(x)​ = 0 & ​p​ k​(x)​ = 0,  then F​(x)​ = ​f​ k(​ ​H​ 1(​ x)​, ⋯, ​H​ n(​ x))​ ​ where G(x) is a realistic assumption presenting a set of objects, to which the theory refers, variables p1, …, pn represent secondary factors, the influence of which is in the idealizational law suspended (expression ‘pk(x) = 0’ is called an idealizing condition), F is the factor that is being investigated in a given theory, and function fk represents the quantitative character of the influence of principal factors H1, …, Hn on the investigated factor F. Distinguishing the principal and secondary factors, and establishing their degree of importance, i.e. the image of the essential structure, constitutes the preliminary stage of the researcher’s work. After proposing the idealizational law, the procedure of concretization is then applied, i.e. the procedure of successive elimination of consecutive idealizing conditions, in accordance with a previously established hierarchy of the factors’ importance. Successive idealizational statements (concretizations) therefore are of the form (Nowakowa and Nowak 2000, p. 118): ​T​ k−1​ : if (G​(x)​ & ​p1​ ​(x)​ = 0 & ​p2​  ​(x)​ = 0 &⋯& ​pk−1 ​  ​(x)​ = 0 & ​pk​  ​(x)​ ≠ 0,  ( ) ( ) ( ) ( ) then F​ x ​ = ​fk−1 ​  (​ ​H1​ ​ x ​, ⋯, ​Hn​  ​ x ​, ​pk​  ​ x )​ ​ ​T​ k−2​ : if (G​(x)​ & ​p1​ (​ x)​ = 0 & ​p2​  (​ x)​ = 0 &⋯& ​pk−1 ​  (​ x)​ ≠ 0 & ​pk​  (​ x)​ ≠ 0,  then F​(x)​ = fi​(​H1​ (​ x)​, ⋯, ​Hn​  (​ x)​, ​pk​  (​ x)​, ​pk−1 ​  (​ x)​)​,​ ........................................................................ 1 Nowak did not consider it necessary to adopt the thesis of essentialism. The effectiveness of the method of idealization and concretization is also reliable when we treat it entirely instrumentally.

The Method of Idealization and Concretization

221

​T ​ 1​ : if (G​(x)​ & ​p1​ ​(x)​ = 0 & ​p2​  ​(x)​ ≠ 0 &⋯& ​pk−1 ​  ​(x)​ ≠ 0 & ​pk​  ​(x)​ ≠ 0,  then F​(x)​ = ​f1​ ​(​H1​ (​ x)​, ⋯, ​Hn​  (​ x)​, ​pk​  (​ x)​, ​pk−1 ​  (​ x)​⋯, ​p2​  ​(x)​)​,​ ​T ​ 0​ : if (G​(x)​ & ​p1​ (​ x)​ ≠ 0 & ​p2​  (​ x)​ ≠ 0 &⋯,  & ​pk−1 ​  (​ x)​ ≠ 0 & ​pk​  (​ x)​ ≠ 0,  then F​(x)​ = ​f0​  (​ ​H1​ (​ x)​, ⋯, ​Hn​  (​ x)​, ​pk​  (​ x)​, ​pk−1 ​  (​ x)​⋯, ​p2​  (​ x)​, ​p1​ (​ x))​ ​,​ where: the expression pn  ≠  0 represents the elimination of a given idealizational condition, that is, consideration of the influence of the factor pn on the investigated factor F, function fk-1, …, f0 quantitatively characterizes the influence which is considered at a certain level as the secondary factor affecting the investigated factor as well as factors which are located higher in the essential structure, and T0 is a factual statement, which characterizes the fullness of relationships between all the factors that are crucial for the investigated one. Thus, idealization leads to the most unreal statement (idealizational law) revealing the essence of the fact which is being investigated, while concretization is a gradual making real of the theory in order to, as a result, attain the statement which fully characterizes the relationship which is being investigated. The simplified core of the method of idealization and concretization, above, is applied to the natural sciences within qualitative-quantitative language. Using this method in the humanities (and metaphysics in particular) necessitates the carrying out of a correction due to the exclusively qualitative character of the language which is applied here. However, the fundamental idea that underpins the method of idealization and concretization remains invariable. Formulating a theoretical recognition of reality (or part of it) relies on the proposition of its simplified image (model) depicting its undisturbed essence. Moreover, this image is successively enriched by the whole diversity of the secondary determinants. 1.2 Negativistic Unitarian Metaphysics The last and unfinished philosophical concept of Nowak is the idea of negativistic unitarian metaphysics (Nowak 1998; 2004; 2007), which was to be, as the author intended, a metaphysical horizon for the method of idealization and concretization (Nowakowa and Nowak 2000, p. 431).2 It does not seem to be an interpretative error to state that all the fundamental metaphors developed within this metaphysical system – of attributivism, negativism and pluralityof-worlds – derive from reflection on the idealizational concept of science. 2 A shortened report on the postulative field of the unitarian metaphysics are presented by Kiedrowski (2010, 2013). Various aspects of the negativistic unitarian metaphysics were also discussed by Falkiewicz (2002), Jadacki (2010, pp. 161–174), Paprzycka (1996; 1999), Perzanowski (2002), Poli (2007), Przybysz (2007) i Woleński (2007) and Zajęcki (2012).

222

Kiedrowski

In an extensive way Nowak developed only the objective part of u ­ nitarian metaphysics. The metaphysics of subject comprising, among others, metaphysical elaboration on cognitive deformation procedures, is available only in its tentative and residual form (Nowakowa and Nowak 2000, pp. 431–438; Nowak 1995). It seems that the basis for the metaphor of attributivism is the recognition that the subject of studies conducted, when applying the method of idealization and concretization, are the factors and relationships between them (Nowakowa and Nowak 2000, p. 130). Their theoretical images (or properties, Tuchańska 1979, pp. 11–16), constitute the conceptual basis for any idealizational theory. This split is considered to be inconsistent with the metaphilosophic thesis of unitarism assumed by Nowak, which states the necessity of the indication of a one, and ultimate, reason (arche) of any metaphysical division. The lack of it, which is characteristic of distinctivism, leads to a plurality of principles (archai), that are self-sustained and not derived from their components of being. Therefore these are the only attributes constituting the metaphysical basis for a division: factor (objective metaphysics) – property (subjective metaphysics). Thereby attributivism becomes another stage of Parmenidesian statement on the identity of being and thought. Reflection on the nature of idealizing conditions leads, however, to a questioning of another of Parmenides’ theses, that of the positiveness of being. If a given factor adopts a minimal value (pn = 0), it means, according to Nowak, a lack (inessentiality) of that factor at a given level of idealization. Subsequently, this level relates to such a world, whose component is a negative situation (negative), which is a reference to the idealizing condition in the form of pn = 0. Being includes not only what is positive (positive realizations of attributes), but also that which is negative (negative realizations of attributes), and which we get to know precisely because of idealization. The negative sphere of the unitarian universe (beingness) is considered to be, metaphysically, the most significant – it constitutes the right field for the category of essentiality (­influence) and existence. This creates the basis for the thesis of suprarealism, i.e. the statement that the subject of scientific research exists objectively, and the right way of arriving at it is by applying the method of idealization and concretization. Every chain of idealizational statements corresponds to an equinumerous set of worlds. In this sense, the idealizational statement, emptyfulfilled, in regard to ‘our’ (‘current’) world, has its reference in a certain other world. Therefore, plurality-of-worlds in a more radical version than in Lewis’ philosophy (Nowak 1998, pp. 117–121) constitutes a reason for the monism of the current world, and one that we try ultimately to explain through scientific theories.

The Method of Idealization and Concretization

223

The method of idealization and concretization was, according to Nowak, applied in the construction of negativistic unitarian metphysics, which develops the metaphor of attributivism, negativism and plurality-of-worlds. With regard to it, this system includes nine more and more realistic models, in which the following categories are developed: 1. Model i (mi): beingness; 2. M ii: existence, essentiality; 3. M iii: causality, modal categories; 4. M iv: change; 5. M v: time; 6. M vi: eternity; 7. M vii: the absolute, logical resulting, enigma of beingness; 8. M viii: Reality; 9. M ix: simple body. The basis statement of the present paper is to question whether, when constructing the objective part of unitarian metaphysics, the Author of Being and Thought used the method of idealization and concretization. I will endeavor to demonstrate that the method which was used was in fact a method of abstraction and disabstraction. 1.3 Abstraction and Idealization Nowak characterizes the relation of idealization to abstraction as follows: The crucial point which follows from the proper understanding of the notion of idealization is that idealization is not abstraction (…). Roughly, abstraction consists in a passage from properties AB to A, idealization consists in a passage from AB to A-B. nowakowa and nowak 2000, p. 116

Abstraction is therefore understood as an impoverishment of a set of properties by a certain property, while idealization consists of the transformation of a given set of properties as such, which embraces a lack of properties. An extensive discussion on this problem was developed by Renata Zielińska (1979; 1981; 1989). The author distinguishes three different procedures of abstracting: 1. abstraction – abstracting from the existing of specific factors (factors’ elimination);

224

Kiedrowski

2. idealization – abstracting from essentiality (influence) of certain factors on other ones (elimination of the factors’ essentiality); 3. stabilization – abstracting from a variability of specific factors (elimination of the factors’ variability) (Zielińska 1979, pp. 89–90). For the qualitative language only the two first procedures are important. The abstracting conditions take the form x ∉ Z(Ci), where i = 2, 3, …, k, which means here ‘that the objects from the universe that is being considered are not within the range of the relevant factor’ (Zielińska 1981, p. 38). Thus, the disabstraction procedure consists of successively making more and more realistic the domain of deliberation by taking into account the existence of a given class of factors. The method of abstraction and disabstraction leads to a separation of the following structure: (1) ​ ​C​ 1​ (2) ​​C​ 1​ , ​C​ 2​ .................. (​ k)​ ​C​  ​ , ​C​  ​ , ⋯, ​C​  ​ .​ 1 2 k Thus, abstraction consists of a transition from the sequence of factors C1, C2, …, Ck to the factor C1. This seems to be the right counterpart of the procedure of abstraction in Nowak’s terms. The idealization procedure which abstracts from the influence of a given factor (Ci = 0) does not annul the factor’s range for a given field of deliberation, it only annuls its essentiality. Therefore, the structure which arises through ­applying the method of idealization and concretization takes on the form: (1) ​ ​C​ 1​ , ​C​ 2​ = 0, ⋯, ​C​ k​ = 0​ (2) ​ ​C​ 1​ , ​C​ 2​ , ⋯, ​C​ k​ = 0​ ......................... (​ k)​ ​C​  ​ , ​C​  ​ , ⋯, ​C​  ​ .​ 1 2 k The above scheme seems to be, in turn, an appropriate way to reflect the way of understanding the procedure of idealization by L. Nowak. For here we are dealing with a transition from the sequence C1, C2, …, Ck to the sequence C1, C2 = 0, …, Ck = 0. The author of Being and Thought considered, however, the abstraction and idealization in the perspective of Zielińska as procedures, which he described as – respectively – reduction and ideation: ‘Zielińska labels “abstraction” which is termed below reduction and “idealization” which is termed here ideation’ (Nowakowa and Nowak, p. 116). Thus, the idealization understood by Nowak

The Method of Idealization and Concretization

225

as a combination of the procedures of reduction and ideation is not identical to the idealization in Zielinska’s perspective. To bring a way of understanding these procedures closer on the negativistic unitarian metaphysics basis, let us assume the set of properties X that embraces positive properties W1, W2, W3 and negative properties: −W4, −W5, −W6. Ideation is a weak counterfactual procedure which transforms the initial set X = {W1, W2, W3, −W4, −W5, −W6} for instance, into a set X’ = {W1, W2, W3, −W’4, −W’5, −W6}. Thence, it relies only on a quantitative change of factors, and not a qualitative one (the set X’ embraces as many positive and negative properties as the set X). However, reduction which is a strong counterfactual procedure leads to a qualitative change. The initial set X is transformed as a result of the application of this procedure, for example into the set X” = {−W1, W2, −W3, −W4, −W5, −W6} (Nowakowa and Nowak 2000, pp. 434–435). Considering abstraction as an equivalent of reduction, and idealization as an equivalent of ideation, does not seem, however, to be correct on the grounds of the unitarian metaphysics. Both reduction and ideation are ‘accessible’ procedures on the grounds of negativism. Abstracting from the factor’s existence cannot be equivalent to the transformation of a positive property into a correspondingly negative one, since the latter (on the grounds of negativistic unitarian metaphysics) exists. Thus, reduction does not lead to a factor which one deprives of existence, but rather to a factor which is entitled to its existence in the strongest possible terms. Furthermore, idealization cannot be understood as ideation, for the minimal value of a given factor (negative one) does not change its negative nature. In accordance with the definition of essentiality adopted in unitarian metaphysics, every negative takes part in an essentiality relation, which is at variance with the way of understanding idealization by Zielińska (Nowakowa and Nowak 2000, p. 164). From the perspective of unitarian metaphysics the criticism of identification abstraction and idealization in the perspective of Zielińska with reduction and ideation in the meaning of Nowak, is reasonable only in so far as this concept can be considered as an appropriate horizon for understanding the method of idealization and concretization. This is exactly what the author of Being and Thought postulated (Nowakowa and Nowak 2000, p. 431). 2

The Reconstruction of the Methodology of the Unitarian Metaphysics from Nowak’s Perspective

In unitarian negativistic metaphysics four methods are applied: the method of idealization and concretization, the method of deduction, the method of ­criteria of adequacy, and the method of paraphrase, the choice of which is

226

Kiedrowski

Table 10.1 The methods of unitarian metaphysics and their role in the system

Method

Function

Metaphilosophic assumption

idealization and concretization deduction criteria of adequacy

constructional

thesis of the methodological unism thesis of metaphysics as a system thesis of metaphysical radicalism

paraphrase

constructional constructional-­ argumentative argumentative

thesis of incommensurableness of philosophical conceptions, thesis of philosophical pluralism

closely connected with the metaphilosophic assumptions of the author of ­Being and Thought. The functions which these methods perform in the system are presented in Table 10.1. The Method of Idealization and Concretization and the Method of Deduction The basic method that enables the building of the system of unitarian metaphysics is the method of idealization and concretization. The reason for the presence of this method in metaphysics constitutes the thesis of methodological monism, which states that in every mature research discipline (in both the areas of science and social science or humanities) one applies the method of idealization and concretization, understood methodically as a deforming of the facts which are being investigated, and it is this deforming that enables their explanation (Nowak 1998, p. 20). The aforementioned metaphilosophic thesis can be considered as an expression of methodological naturalism, since the method of idealization and concretization is used, in Nowak’s opinion, universally, on the grounds of natural science. The humanities, says Nowak, have only recently realized the necessity of using this method. The qualitative character of the language of philosophy allows, however, for only the general frames of the idealizational science concept to be preserved, while a number of technical terms and specific solutions must here undergo modifications, which mainly consist of demathematizing the idealizational statements. Thus, the predecessors of such statements embrace qualitative idealizing conditions (for example that attributes do not influence each other), whereas their apodoses – embrace the network of categories relevant to a given model (metaphysical categories are the counterparts of factors 2.1

227

The Method of Idealization and Concretization Table 10.2 The categorial and idealizational characteristics of the unitarian metaphysics’ models

Model i Model ii

Model iii Model iv

Model v

Model vi Model vii

Model viii

Model ix

Idealizing condition

Main category(ies) Categorial content (symbol) of the model

situations do not influence each other attributes are not ­connected by full determination relationship situations do not change there are no temporal ­connections between situations there are no timeless ­connections between situations peculiar beings are not taken into account dynamics of beingness (as a whole) is not taken into account the occurence of simple bodies in beingness is not taken into account –

beingness (B)

MI = 

existence (E!), ­essentiality (Ess)

MII = 

causality (C) change (Z)

MIII =  MIV = 

time (T)

MV = 

timeless (ΘT)

MVI =  MVII = 

enigma (E), the absolute (A) Reality (R)

MVIII = 

simple bodies (Cł)

MIX = 

from the idealizational concept of science). The Table 10.2 shows the correlations between the idealizing conditions that are valid for a given model, the ­metaphysical categories which are adopted on its ground, and the categorial content of the model at a given level of idealization. The objective part of Nowak’s negativistic unitarian metaphysics embraces a sequence of nine models that become increasingly realistic. Each model ­(except the last one) is constructed while the validity of the set of idealizing conditions, the number of which decreases along with the degree of the model being made real. For this reason Model ix is no longer based on any of ­idealizing conditions postulated by Nowak. The idealization procedure in metaphysics

228

Kiedrowski

is therefore understood here as the adoption of a set of ­qualitative idealizing conditions that are characteristic for a given model, and that exclude from its categorial content the metaphysical notions which are introduced to the system in the following models. Elimination of the idealizing condition, as well as the introduction to the system of a metaphysical category, to which this condition refers, takes place, however, within the procedure of concretization. It leads to an enriching of the categorial content of each model, that is, to an increase of its being made real. The direction of concretization is, again, not arbitrary, but rather it reflects the essentiality level of metaphysical categories (Nowak 2004, p. 272). The full development of the model takes place with the implementation of a traditionally understood method of deduction. It allows a formulation of metaphysical statements based on definitions and statements introduced to the concept in a given model or in prior ones. Philosophy as a system, postulated by Nowak, is realized through a deductive deriving of metaphysically significant consequences by the author of the concept, consequences which create, along with definitions, a non-contradictory set of sentences. Metaphysical categories, as far as they are not primitive, are characterized through natural definitions, on which following conditions are imposed: 1. the natural definition is based solely on the beings of the assumed universe (here: the unitarian one); 2. in the definiens, only the notions that are being introduced into the system in the prior models in respect of which an adequate relation is being determined are used; 3. as a result, we receive a metaphysical category, the range of which is in line with the previously adopted criteria of adequacy (Nowak 2004, p. 85). Metaphysical categories are still characterized by meaning postulates. The constructive function of the methods discussed above comes down to the methodical construction of unitarian metaphysics in a vertical and horizontal dimension. The method of idealization and concretization leads to making more realistic of the concept through successively expanding the categorial content of consecutive models, whereas the method of deduction makes it possible to fill these models with statements which constitute the consequences of the application of the procedure of concretization. 2.2 The Method of Criteria of Adequacy Negativistic unitarian metaphysics is a radical concept in the sense that it breaks, on postulative ground, away from common-sense theses (for instance with a statement on the uniqueness of the world or on the positive nature of

The Method of Idealization and Concretization

229

reality and others). The radicalism that is being postulated, enabled by speculation based on the procedure of idealization, is not, however, aimed at questioning the significance or veracity of statements of standard (common-sense) metaphysics. What is primarily at stake here is the indication of the metaphysical relations that would be appropriate for them, that is a demonstration of the secondariness of those statements on the grounds of unitarian metaphysics. Constructing the metaphysical categories, which, as a result of the adopted assumption, breaks with common sense could lead to a formulation of entirely arbitrary definitions of notions. The method of criteria of adequacy constitutes a form of protection against the abovementioned situation. On the one hand, it enables the construction of a notion that is, in terms of content, radical. On the other hand, it fulfills a certain array of intuitions, which are usually connected with a given notion. As an example, from a temporal consequence relationship, we should expect that, independently of how its elements will be characterized, it will be a relation that is unreflexive, asymmetrical and transitive. Constructing the metaphysical category while adopting a method of adequacy criteria begins with the establishment of the notion which is to be defined (for instance the above-mentioned ‘temporal consequence’). The author of the concept arbitrarily picks the set of adequacy criteria on the basis of preexisting concepts (for example the Aristotelian concept of time) or universally prevailing intuitions that are associated with a given notion. These criteria, in so far as it is necessary, are subsequently paraphrased into a language of a parent concept and then the natural definition of a given metaphysical category is formulated. A notion which is constructed in such a way should correspond with the range of the initially adopted adequacy criteria. The method of criteria of adequacy is, therefore, on the one hand a method supporting construction methods and in a way it constitutes an argument in favor of the adequacy of the metaphysical categories that are proposed on the ground of a given concept. On the other hand, it protects from the arbitrary distinguishing and denominating of the construct that has been sought by the author of the radical concept. 2.3 The Method of Paraphrase Supplying as many alternative models of the Whole (being) as possible is considered by Nowak the cultural function of philosophy. None of the models are, at the same time, objectively cognitively distinguished, since one cannot take the neutral perspective of an observer, which would single out some of the concepts. This perspective is always burdened with metaphysics. Yet, it is not cognitively disqualifying the disadvantage of philosophy (metaphysics). On

230

Kiedrowski

the contrary, the multiplicity of philosophical concepts provides various horizons of understanding reality. Which of them will be used by science, religion, art or other cultural fields for philosophizing is a secondary issue. Various metaphysical systems are incomparable. Criticism of a philosophical concept from the perspective of a parent concept, i.e. on the grounds of its settlements, constitutes an abuse, since each metaphysical category is located in a specific-for-a-given-metaphysics categorial net, and it is the net that legitimates the category. Moreover, it is impossible to formulate any conclusive metaphysics that are superordinate to another one. The only way to refer to a different philosophical concept is the method of paraphrase. It enables an expression of theses of a different concept in a conceptual apparatus of the parent one as well as verification whether that paraphrased statement constitutes the thesis of the parent concept. Whatever the result of the paraphrase would be, it does not constitute a basis to state the truthfulness or falseness of the concept that is being paraphrased. Application of the method of paraphrase proceeds according to the following scheme: 1. one selects a thesis from the system which is being paraphrased; 2. one tries to add sense to that thesis on the ground of the parent conceptual apparatus, that is, to find the notional counterparts of the terms which appear in the thesis that is being paraphrased; 3. if the second step cannot come into being, it means that either the parent concept is defined by lesser metaphysical power than the concept which is being paraphrased (i.e. it does not have a sufficient conceptual apparatus), or that the stage of the theory’s construction is inappropriate for paraphrasing; 4. if sense can be added to the thesis, then one next verifies whether the ­assumptions and definitions of the parent concept stem from this very ­concept (positive paraphrase) or the negation of it (negative paraphrase). Thus, the method of paraphrase enables verification and, as a consequence, criticism of the parent concept, and not the paraphrased one. In the first ­volume of Being and Thought Nowak treats positive and negative paraphrases equally; in the second volume he assumes that the more positive paraphrases that can be carried out within one, the better the concept is.3 It is not a 3 The criterion applied by Nowak in second volume of Being and Thought seems to be doubtful. Ideational and logical non-contradiction of the metaphysical system results in some of the philosophic concepts being, in some measure, paraphrased solely negatively in advance.

The Method of Idealization and Concretization

231

b­ reaking of different metaphysical ideas but an attempt at the comprehension of them from our own perspective that should constitute the proper philosophic attitude towards the multiplicity of metaphysical horizons. The argumentative function of the method of paraphrase comes down to demonstration: that conceptual apparatus which is being proposed along with the resource of metaphors, and that which is expressed within that apparatus, are distinguished by a higher heuristic power. According to Nowak, the metaphysics being proposed, should be as radical as possible, but, simultaneously, it cannot remain entirely beyond any alternative metaphysical proposals. Furthermore, the basis for paraphrases can be both a philosophic treatise and an excerpt stemming from literature or a poem. The inability of a first concept to add sense to other philosophic theses is evidence against it and constitutes the basic incentive to its further development, transformation or discarding. The methods discussed above, used by the author of Being and Thought, used for construction or validation in negativistic unitarian metaphysics, constitute an essential component of this system. In the following part I will demonstrate that the method of idealization and concretization was recognized improperly by Nowak. The characteristics of this method, which was reconstructed in point ii.1 refers thence to a different method (of abstraction and disabstraction), and the method of idealization and concretization should be re-characterized and applied on the grounds of unitarian metaphysics. The Figure 10.1 summarises all methods distinguished by the author of Being and Thought, maintaining the terminology proposed by Nowak (Kiedrowski 2013, p. 42). 3

Reinterpretation of the Method of Idealization and Concretization on the Grounds of Unitarian Metaphysics

Difficulties Related to Nowak’s Approach of the Method of Idealization and Concretization If the method applied by Nowak to construct the objective part of unitarnian metaphysics is the method of idealization and concretization, then at least four difficulties emerge: 3.1

It seems that the metaphysical power of the system (that is, the ability of the conceptual apparatus to paraphrase the alternative statements) should be measured with the ability to add sense to theses of alternative metaphysical concepts. Positive paraphrases prove only that the statements of other philosophic concepts, which are (positively) paraphrased, constitute a common ideational line.

232

Kiedrowski criteria of adequacy for model I

MODEL I criteria of adequacy for model II

MODEL II criteria of adequacy for model III

MODEL III criteria of adequacy for model IV

T H E W H O L E

MODEL IV criteria of adequacy for model V

MODEL V criteria of adequacy for model VI

MODEL VI criteria of adequacy for model VII

MODEL VII criteria of adequacy for model VIII

MODEL VIII criteria of adequacy for model IX

MODEL IX

positive negative paraphrases

positive negative paraphrases

positive negative paraphrases

positive negative paraphrases

positive negative paraphrases

positive negative paraphrases

positive negative paraphrases

positive negative paraphrases

A L T E R N A T I V E M E T A P H Y S I C A L S Y S T E M S

positive negative paraphrases

-deduction -concretization -paraphrase

Figure 10.1

Structure of Nowak’s negativistic unitarian metaphysics

1. the concept of the structure of metaphysical doctrines, postulated by the author of Being and Thought, is incompatible with the procedure of concretization on the grounds of unitarian metaphysics; 2. the possibility of various sequences of models in the system of unitarian metaphysics is incompatible with the thesis of essentialism that is being adopted;

The Method of Idealization and Concretization

233

3. the possibility of introducing metaphysical categories at an earlier stage of construction than takes place in Nowak’s system, excludes ­consideration of the simplifying conditions for given models as idealizing conditions; 4. applying the procedure of concretization should lead to a limiting of the speculativeness of the metaphysical concept, whereas metaphysical categories proposed by Nowak (up to and including model ix) are characterized by too high a degree of abstraction. The structure of metaphysical doctrines is characterized by the author of Being and Thought as follows: (…) the recursive form: (1) the creations of B0 exist (they are beings), (2) if x is a being and x remains in relation R to y, then y is a being; the set of beings B consists of B0 (base beings), and also R (B0), R(R(B0)) beings etc. Every metaphysical doctrine includes, therefore, theses of existential qualification determining prima facie the class (classes) of beings as well as theses of existential qualification determining relation (relations), in which being to being must remain, and in the last resort to the base being (Nowak 1998, p. 24). (…) The set of metaphysical structures of a given type will be called a metaphysical paradigm. Individual paradigms will, however, differ either objectively or structurally. 1. Considering only objective differences between metaphysical structures, one cannot state anything in general terms. What is possible to indicate, is that it is only the reasons that induce the philosophers to begin the deliberations from particular primary beings within systems they build and to try to put them in order within more or less capacious classes (…). (…) rules of existential identification can belong to two essentially different types. The procedure of identification R is defined factually in the case, where there is a following relationship: if y exists and yRx, then exists x (sc. in the same meaning of the word). In other words, factualistic rules of identification enable enlargement of the set of primary beings by other elements of the same universe. (…) rules of identification are called counterfactual, if they ascribe to the objects that bear an appropriate relation to primary beings a different modus of existence than one to which the first ones are entitled. In other words, counterfactual rules of identification do not enlarge the set of primary beings, but they add to it other sets that consist of beings in a way from another level.

234

Kiedrowski

(…) So, certain objects do exist actually, while the rest of them are ‘only-possible’. Also, sometimes, to emphasize given difference, the distinction between existence and subsistence is introduced (Nowak 1996, pp. 294–295). Every metaphysical concept in this perspective thus indicates, on the one hand, a set of basic beings, and, on the other hand a relation which allows the existence of the remaining beings. Analysis of the first two models of the objective part of the unitarian metaphysics allows the conclusion that the above-mentioned characteristics are applied only to model ii, where negative situations (negatives) are considered as basic beings and the factualistic rule is introduced (exists only if at least one element is negative) as well as the counterfactual one (any x subsists when it is a part of the unitarian universe). The postulatory content of model i, which comes down to separation within the unitarian universe (the beingness) of various possible classes of beings, is not contained in the postulated concept of the structure of the metaphysical system. In this situation it is difficult to understand in what way model ii is a concretization of model i. This gives weight to the argument that the notions which are proposed in it are of another categorial order. In addition, model ii includes characteristics of the main metaphysical relation of the system which is being discussed, that is, the essentiality relation. Again, it does not seem possible that any metaphysical system could be deprived of such a relation. Let us recall, by way of illustration, the relation between Platonic physical objects and ideas, between the prime mover and the Universe in the concept of Aristotle, or the pre-established harmony concept of Leibniz. Thus, I argue for extension of the concept of the structure of the metaphysical system of Nowak by certain additional postulates. Thus, the metaphysical doctrine embraces: 1. ontological theses – they characterize a potential content of the assumed universe; 2. theses of qualification and existential identification – they organize the content of the universe from the existential perspective; 3. relational theses – describing the main relation – (characteristic) for a given system, based on the classes of beings distinguished by ontological theses. The above modification allows us to comprehend the postulatory content of the first two models of the unitarian metaphysics. It does not explain, though, in what relation to each other these models remain. The core of a given metaphysics should be the basis of a possible concretization. Therefore, in the case

The Method of Idealization and Concretization

235

of negativistic unitarian metaphysics it would be the content of the first two models. Consecutive difficulty is connected with the consequence of the thesis of essentialism, as assumed by Nowak. It extorts a specific and invariable order of models that constitute a more and more realistic sequencing of them. The order of models is, however, in two cases, arbitrary (the author of Being and Thought informs the reader of these simplifications explicite), that is, it is the technical aspects that determine their place in the system of unitarian metaphysics. This includes model v, which, metaphysically is established in model vi as well as model vii, which constitutes a combination of definitions and statements concerning particular components of beingness. This issue is meaningful from the side of idealizing conditions, which in idealizational methodology are decisive for the possibility to formulate a model at a given stage of construction. For a given model, the elimination of the idealizing condition X, and holding of the idealizing condition Y, means that the condition X refers to a factor that is more essential than the condition Y. Hence, from the perspective of concretization it is unacceptable that model P which constitutes a metaphysical basis for model R appears in the system later, i.e. at the level which has been made more realistic. The third issue concerns the categorial order of the objective component of negativistic unitarian metaphysics. Essentialism extorts not only a certain definite order of models, but also a definite sequence of metaphysical categories that are developed within respective models. Of metaphysical categories, which are counterparts of the factors in natural sciences, we can state the characteristic for a given metaphysics level of essentiality: ‘notion x is less essential than notion y then and only then, when, in the definiens of notion x, notion y appears’ (Kiedrowski 2013, p. 50). Analysis of the categorial content of the nine models of unitarian metaphysics leads to the conclusion that the categories of necessity, transformation, temporal consequence, real sequence and simple body, could be defined at the earlier stages of construction (see Kiedrowski 2013, pp. 50–53). This means that the order in which they occur in the system is not essential and the simplifying conditions which enable their definition are not idealizing conditions. The last of the indicated problems is associated with the main difficulty of every metaphysical system, that is, with the transition from highly speculative notions and statements to ones which, in an adequate way, can be related to the reality surrounding us. In light of the distinction of metaphysics into radical and subservient ones regarding the common sense, this means that model ix should be far more “commonsense” than model i. The analysis of definitions and statements of the objective part of unitarian metaphysics, meanwhile,

236

Kiedrowski

does not allow ascertainment that this concretization has actually taken place. The definition of a simple body, formulated by Nowak in model ix will serve as proof here: [Cł] As a simple body (sensu stricto) of type r we will call an object C, the constitutive of which embraces the sequence R = 〈RG, σ, r, S°, Ν 〉, such that 1. the deep structure RG of the body C is a sum of the self-existent sequence σ and the set of unilateral causes and results of the situations from σ; 2. S° is a set of hidden negatives of a body: no n ∈ S° ⊆ RG has an epiphenomenon in RP; 3. the distinguished (final) element is a situation r ∈ RG, which is the result of at least one component of a selfexistent sequence σ; 4. the superstructure N of a body C is a sum of a surface structure RP and a set of accidents A° of the body C; 5. the surface structure RP of the body C is a sum of the set ^σ of the epiphenomena of selfexistent sequence σ  ⊆  RG as well as of the set of beyondselfexistent (unilateral bound causal components) epiphenomena of the structure RG; 6. the components of the superstructure Ν = RP ∪ A° of the body C are such, that: 6.1. if p^ ∈ RP, to p^ is an epiphenomenon of the negative p which belongs to the deep structure RG of an object C, but not necessarily the other way round (the set S° can be non-empty); 6.2. every f ∈ A° is a positive from N, which a. is not an epiphenomenon of whichever of the situations that belong to the deep structure RG of an object C, b. is neither necessary nor sufficient condition for any epiphenomenon from the surface structure RP of an object C; 7. sets RG, RP are non-empty, while sets S°, A° can be empty (it is so for regular bodies: p ∈ RG wtw p^ ∈ RP) or not (for overloaded irregular bodies S° = ∅, A° ≠ ∅, for irregular unfulfilled bodies S° ≠ ∅, A° = ∅). (Nowak 2007, pp. 336–337). As we see above, on the one hand a metaphysical category, which constitutes one of the fundamental categories for common sense thinking, is being introduced, yet, on the other hand this category is characterized in such a way that it has to be considered as a speculative one. This ascertainment, concerns, I believe, all the notions introduced into the system by the author of Being and Thought, and thereby it is out of the question that the procedure used to introduce them would be the concretization.

The Method of Idealization and Concretization

237

3.2 The Method of Abstraction and Disabstraction The difficulties which were presented in the previous point can be overcome only if we consider the method of abstraction and disabstraction as a construction method, as applied by Nowak. Still, the characteristics of this method proposed by the author of Being and Thought do not change if, when distinguishing the procedure of abstraction from the procedure of idealization, one uses the distinction which was discussed in point 1.3 of the present paper and proposed by Zielińska (1979; 1981; 1989). The simplifying procedure, applied by Nowak, relies on non-consideration of a metaphysical category at a given stage of construction, that is, in the lack of definition of a given notion. If we treat this measure as a counterpart (in science) of abstracting from the existence of a factor, then the procedure that has been used should be considered as abstraction. The construction of model i relies on the validity of the simplifying condition (that from now on I will call by the term ‘abstracting condition’) about the non-influencing of the attributes on each other, as a result of which beingness and its content are described. Elimination of the mentioned assumption – disabstraction – leads to a formulation, on the grounds of model ii, a definition of existence and essentiality, as well as further specifying the content of the unitarian universe within these two categories. Obviously, for model ii the following abstracting condition is applicable, which undergoes elimination in model iii and so on. The recognition of the method used by Nowak for the method of abstraction and disabstraction forces us into making certain terminological modifications that are necessary for preserving the traditional, for the idealizational concept of science, terminology. Since applying the method of idealization and concretization leads to the construction of a sequence of models, in the case of the method of abstraction and disabstraction, we will talk about modules. Thus, the objective part of the negativistic unitarian metaphysics of Nowak, i.e. its postulatory plane, constitutes the most idealized model i, embracing nine modules that create a series of disabstractions in ascending order. This model is still, though, located at the same level of idealization. The method of abstraction and disabstraction does not lead to the difficulties enumerated and described above. And so: 1. the relation between the first two modules is the relation of disabstraction, which leads to a completion of the set of ontological theses, which is embraced by the module i, the completion is via theses of qualification and existential identification as well as the relational ones; since both modules occur at the same level of idealization, the core of metaphysics as a whole

238

Kiedrowski

will undergo, on the one hand, further disabstraction (in modules iii – ix), and on the other hand – concretization (model ii). 2. The order of models can be designated by both logical dependences and technical reasons, for the method of abstraction and disabstraction is not based on the thesis of essentialism. 3. In the case of metaphysical categories, the matter can be seen in a similar light; their place in the system is not related to the degree of essentiality, but it is mainly determined by logical reasons; the abstracting condition excludes from the deliberation such a category, which logically does not precede the category that is central for a given module and which constitutes a category that refers to a different range of deliberation. 4. What is also not surprising is the degree of speculativeness of the notions which are introduced in subsequent modules, since they are located at the same level of idealization. Thus, further specifying the metaphysical universe through the procedure of disabstraction constitutes a preliminary step to applying the procedure of concretization. The author of Being and Thought, therefore used the method of abstraction and disabstraction to build the objective part of negativistic unitarian metaphysics. It fulfills the following functions in the system: 1. constructive function – together with the method of deduction, it enables the construction of a sequence of modules of an increasing degree of disabstraction and which jointly create a model; 2. logical function – it enables the separation of the metaphysical categories that take different places in a definitional order; 3. technical function – through the characteristics of distinct metaphysical problems in different modules, it leads to a clear exposition. 3.2 Reinterpretation of the Method of Idealization and Concretization With regard to that discussed in point iii.2 (modification of the unitarian methodology), it becomes necessary to determine what the procedures of the metaphysical idealization and concretization relies on in a different way to that proposed by Nowak. For their adequate characteristics I assume that: 1. it is necessary to indicate the idealizing condition, the obligation of which enabled the construction of model i; 2. the procedure of idealization is supported by the idealizing condition which concerns the one of three groups of statements comprising the structure of a metaphysical system;

The Method of Idealization and Concretization

239

3. the procedure of concretization relying on the elimination of the idealizing condition, leads to the distinguishing of a more realistic metaphysical level as well as a homogeneous modification (concretization) of the conceptual apparatus of model i. From the metaphysical perspective, for the most legitimate, I consider that the idealizing assumption should concern the main relation: essentiality. Its modification will entail, on the one hand, the natural distinguishing of a new class of beings on the basis of the content of the universe, presented in module i of model i (abbreviated to: mIMI). On the other hand, it should enable the concretization of the metaphysical categories (including the category of existence), introduced in model i. From a methodological perspective, by concretization I will understand the elimination of the idealizing condition considering the main relation, that is the introduction to the system of an additional thesis, the negation of which, that is, an idealizing condition, is applied in model i. Thus, the procedure of idealization relies on distinguishing the idealizing condition and, while it being in force, constructing a given model. The justness of the above way of understanding both procedures is strengthened by the postulate of metaphysical holism, which states that it is not a single statement but only a certain set of them which constitutes an appropriate (comprehensive) metaphysical argument (Nowak 1998, p. 142). On this basis, the heuristic power of the idealizing condition is proven by the content of the model, which was constructed while the condition was in force, or as a result of its elimination. The formal and general way of presenting the metaphysical idealizing statements for the first two models of negativistic unitarian metaphysics is as follows: • i​ f U & R & I & ¬ Z,​ ​then ​M​ I​ = ​{, }​,​ * • ​if U & ​R​  ​ & I,​ ​then ​M​ II​ = {, ,  n MII MII MII MII },​ where U is a set of meaning postulates that characterize attributes, R – the set of definitions and statements characterizing the essentiality relation, I – theses of existential qualification and identification, ¬Z – idealizing condition (about the lack of the constancy of the essentiality relationships between attributes), MI – model i, CnL – the set of logical consequences, K1MI, …, KnMI – the set of definitions and statements that characterize the metaphysical categories

240

Kiedrowski

of the model i, R* – the set R complemented with the sentence Z, MII – the model ii, K1*MI, …, Kn*MI – the set of definitions and statements that characterize metaphysical categories of model i taking account of the modification of relation R, K1MII, …, KnMII – the set of definitions and statements that ­characterize metaphysical categories of model ii (Kiedrowski 2013, p. 60). Should the above methodological reinterpretation be accurate, in the construction of the system of the negativistic unitarian metaphysics five methods are used: 1. the method of idealization and concretization; 2. the method of abstraction and disabstraction; 3. the method of deduction; 4. the method of criteria of adequacy; 5. the method of paraphrase. The adopted way of understanding the method of idealization and concretization entails the necessity of a correction in respect of the method of paraphrase. Now, every model shall correspond to a different field of paraphrases due to the different categorial content of the models. It means that the ability of the conceptual apparatus of the negativistic metaphysics to paraphrase the statements of different concepts is different according to the model, which we will take as the basis of a paraphrase. Hence, we evaluate the metaphysical power of a system due to the global field of paraphrases embracing the field of paraphrases of model i and model ii. A graphic depiction of the structure of the negativistic metaphysics from the methodological perspective is shown in Figure 10.2. 4

Concluding Comments

The present methodological analysis is limited to the system of negativistic unitarian metaphysics. Whether the methodology proposed by Nowak and reinterpreted above can be a valuable tool for the analysis and reconstruction of other metaphysical systems should be studied in further detail. Such an interpretation would be, as a matter of fact, of an adaptational character (and not a historical one), yet it would be an additional argument in favor of the universality of the method of idealization and concretization in metaphysics. So far, the first two modules of model ii have been developed, which ­constitute the negativistic concept of matter (Kiedrowski 2013, pp. 71–101). An initial understanding of the concept of the phenomenon (Kiedrowski 2013,

241

The Method of Idealization and Concretization

CRITERIA OF ADEQUACY MODULES

T H E W H O L E

MODEL I

II

I

III

IV

V

VI

VII

VIII

IX FIELD OF PARAPHRASES ppMI

CRITERIA OF ADEQUACY MODULES I

MODEL II

II

III

Figure 10.2

n

FIELD OF PARAPHRASES PPMII

S Y S T E M S

GLOBAL FIELD OF PARAPHRASES = PPMI + PPMII

– DISABSTRACTION – DEDUCTION

...

A L T E R N A T I V E M E T A P H Y S I C A L

– CONCRETIZATION

The structure of negativistic metaphysics’ system

pp. 157–176) and subject (Kiedrowski 2013, pp. 177–192) have been outlined. The concretization of model I leads to the division of a set of attributes, which in model i is homogeneous. However, the appearance of categories of matter, phenomena and subject proves, as a consequence, that the second model is more realistic, i.e., it is a result of applying the procedure of concretization which constitutes an argument in favor of the reasonableness of the reinterpretation proposed above. References Falkiewicz, A. (2002). Wobec nowożytnego rozumu. O Bycie i myśli Leszka Nowaka ­[Towards Modern Reason. On Being and Thought by Leszek Nowak]. In: J. Brzeziński, A. Klawiter, T.A.F. Kuipers, K. Łastowski, K. Paprzycka, P. Przybysz (eds.) Odwaga filozofowania. Leszkowi Nowakowi w darze, pp. 439–465. Poznań: Humaniora. Jadacki, J.J. (2010). Metodologia i semiotyka [Methodology and Semiotics]. Warszawa: Wydawnictwo Naukowe Semper. Kiedrowski, K. (2010). Zarys negatywistycznej metafizyki unitarnej [The Outline of ­Negativistic Unitarian Metaphysics]. Poznań: Wydawnictwo Poznańskie.

242

Kiedrowski

Kiedrowski, K. (2013). Metodologiczne podstawy negatywistycznej metafizyki unitarnej Leszka Nowaka [Methodological Foundations of Leszek Nowak’s Negativistic Unitarian Metaphysics]. Poznań: Wydawnictwo Naukowe Instytutu Filozofii UAM. Nowak, L. (1995). Anty-cogito, magia, unitarna koncepcja metafizyki [Anti-cogito, Magic, the Unitarian Concept of Metaphysics]. Kultura współczesna 1–2, pp. 5–34. Nowak, L. (1996). Kilka uwag na temat miejsca empiryzmu logicznego w dwudziestowiecznej filozofii [Some Remarks on the Place of Logical Empiricism in 20th Century Philosophy] In: R. Egiert, A. Klawiter, P. Przybysz (eds.) Oblicza idealizacji. Poznańskie studia z filozofii humanistyki 2(15), pp. 293–310. Poznań: Wydawnictwo Naukowe UAM. Nowak, L. (1998). Byt i myśl. U podstaw negatywistycznej metafizyki unitarnej, t. I Nicość i istnienie [Being and Thought. Foundations of Negativistic Unitarian Metaphysics, vol. I: Nothingness and Existence]. Poznań: Wydawnictwo Zysk i S-ka. Nowak, L. (2012 [1998]) On the Hidden Unity of Social and Natural Sciences. In: K. Brzechczyn, K. Paprzycka (eds,) Thinking about Provincialism in Thinking, Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 69, pp. 15–50. ­Amsterdam/Atlanta: Rodopi. Nowak, L. (2004). Byt i myśl. U podstaw negatywistycznej metafizyki unitarnej, tom II: Wieczność i zmiana [Being and Thought. Foundations of Negativistic Unitarian Metaphysics, vol. II: Eternity and Change]. Poznań: Zysk i S-ka. Nowak, L. (2007). Byt i myśl. U podstaw negatywistycznej metafizyki unitarnej, tom III: Enigma i rzeczywistości [Being and Thought. Foundations of Negativistic Unitarian Metaphysics, vol. III: Enigma and Realities]. Poznań: Zysk i S-ka. Nowakowa, I., and L. Nowak (2000). The Richness of Idealization, Poznań Studies in the Philosophy of the Sciences and the Humanities 69. Amsterdam/Atlanta: Rodopi. Paprzycka, K. (1996). Dziury metafizyczne. O ontologicznych podstawach koncepcji związku w metafizyce unitarnej [Metaphysical Gaps. On the Ontological Basis of the Concept of Relationship in Unitarian Metaphysics]. In: R. Egiert, A. Klawiter, P. Przybysz (eds.) Oblicza idealizacji. Poznańskie studia z filozofii humanistyki 2(15), pp. 311–320. Poznań: Wydawnictwo Naukowe UAM. Paprzycka, K. (1999). Idealizacja w metafizyce unitarnej [Idealization In Unitarian Metaphysics]. In: A. Klawiter, L. Nowak, P. Przybysz (eds.) Umysł a rzeczywistość, Poznańskie Studia z Filozofii Humanistyki 5(18), pp. 393–404. Poznań: Zysk i S-ka. Perzanowski, J. (2002). Pozytywizm a negatywizm ontologiczny [Ontological Positivism and Ontological]. In: J. Brzeziński, A. Klawiter, T.A.F. Kuipers, K. Łastowski, K. Paprzycka, P. Przybysz (red.) Odwaga filozofowania. Leszkowi Nowakowi w darze, pp. 477–495. Poznań: Humaniora. Poli, R. (2007). Formal and Ontological Roundabouts. In: J. Brzeziński, A. Klawiter, T.A.F. Kuipers, K. Łastowski, K. Paprzycka, P. Przybysz (eds.) The Courage of Doing Philosophy. Essays Presented to Leszek Nowak, pp. 325–335. Amsterdam-New York: Rodopi.

The Method of Idealization and Concretization

243

Przybysz, P. (2007). What does to Be Mean In Leszek Nowak’s Conception of Unitarian Metaphysics? In: J. Brzeziński, A. Klawiter, T.A.F. Kuipers, K. Łastowski, K. Paprzycka, P. Przybysz (eds.) The Courage of Doing Philosophy. Essays Presented to Leszek Nowak, pp. 315–324. Amsterdam-New York: Rodopi. Tuchańska, B. (1979). Czynnik a wielkość [Factor versus Magnitude]. In: A. Klawiter, L. Nowak (Eds.) Odkrycie, abstrakcja, prawda, empiria, historia a idealizacja [Discovery, Abstraction, Truth, Experience and History vs. Idealization], pp. 11–21. Warszawa-Poznań: PWN. Woleński, J. (2007). Metalogic and Ontology. In: J. Brzeziński, A. Klawiter, T.A.F. Kuipers, K. Łastowski, K. Paprzycka, P. Przybysz (eds.) The Courage of Doing Philosophy. Essays Presented to Leszek Nowak, pp. 337–351. Amsterdam-New York: Rodopi. Zajęcki, M. (2012) O potrzebie (recepcji) negatywistycznej metafizyki unitarnej Leszka Nowaka [On the Need (of Reception) of Leszek Nowak’s Negativistic Unitarian Metaphysics]. Filozofia Nauki 20 no. 2, 143–154. Zielińska, R. (1979). Przyczynek do zagadnienia abstrakcji [A Contribution to the ­Issue of Abstraction]. In: A. Klawiter, L. Nowak (eds.) Odkrycie, abstrakcja, prawda, ­empiria, historia a idealizacja, pp. 87–97. Warszawa-Poznań: PWN. Zielińska, R. (1981). Abstrakcja, idealizacja, generalizacja. Próba analizy metodologicznej [Abstraction, Idealization, and Generalization]. Poznań: Wydawnictwo N ­ aukowe UAM. Zielińska, R. (1989). Contribution to the Problem of Abstraction. In: J. Brzeziński, F. Coniglione, T.A.F. Kuipers and L. Nowak (eds.), Idealization II: Forms and A ­ pplications. Poznań Studies in the Philosophy of the Sciences and the Humanities 17, pp. 9–22. Amsterdam: Rodopi.

chapter 11

The Choice of the Principles of Justice in the Political Philosophy of John Rawls: An Idealizational Interpretation Piotr Przybysz Abstract In this paper I propose a methodological interpretation of John Rawls’ theory of ­justice. According to this proposal Rawls’ theory can be better understood if we view it in terms of the method of idealization and concretization proposed by Leszek Nowak. My aim is to interpret Rawlsian original position, i.e. situation of choice of the principles of justice, as idealizational model of social situation construed on the basic of a limited number of idealizing and realistic assumptions. I present a discussion of some methodological and theoretical consequences of this theory building method.

1 Introduction The aim of this article is to propose a methodological interpretation of John Rawls’ theory of justice presented in A Theory of Justice (1971; 1999). Scholars usually consider Rawls’ theory as an interpretation of the ethical and anthropological foundations of a modern democratic liberalism. Firstly, Rawls introduced into modern political philosophy a vast array of new concepts such as the original position, the veil of ignorance, reflective equilibrium, justice ­between generations and the like. Secondly, he revived in liberalism a mode of thinking that refers to notions of the state of nature, social contract and the person’s rights. As a result, the liberal philosophy which “under the reign of” utilitarianism in a form of legal positivism represented by Jeremy Bentham, John Austin and Wesley Hohfeld was an institutional and legalistic doctrine received moral justification. Thirdly, thanks to Rawls, the role of theoretical “constructivism” as a valuable research method in political philosophy was ­appreciated (e.g. Daniels ed. 1975; Pogge 2007; Young ed. 2009). The main goal of this reconstruction is not, however, to discuss Rawls’ ­theses. It rather refers to the methodological interpretation of Rawls’ theory of justice so as to demonstrate that it was decisively affected by the research

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_013

The Choice of the Principles of Justice

245

method used to establish this theory.1 I will argue that Rawls’ theory described in A Theory of Justice can be understood better if viewed in terms of the method of idealization and concretization as proposed by Leszek Nowak (Nowak 1980; Nowak 1992; Nowakowa, Nowak 2000). According to this proposal, idealization is one of the crucial procedures used to construct scientific theories. The idealizational theory of science (henceforth its) involves an assertion that numerous notions and theoretical constructs are, as a matter of fact, idealizational constructs that result from the adoption of relevant simplifying, abstracting and idealizing assumptions. For example, it refers to notions of a material point or perfect gas (natural sciences) and notions of a free market or an ideal employer (social sciences) (Nowak 1992, pp. 19–29). Many of those idealizational constructs are applied for theoretical purposes through the use of them in scientific theories. According to the its a scientific theory comprises a sequence of models (theorems) arranged according to the number of adopted idealizing assumptions: ​M​ k​ , ​M​ k−1​ , … ​M​ i​ , A ​M​ i​ where: k, k−1, … i is the number of adopted idealizing assumptions. A sequence of models begins with the most general and most idealized model Mk that describes dependencies between the main factors, i.e. those that affect a given phenomenon to the greatest extent, and omits secondary factors (listed as k – assumptions). According to the its – the idealization procedure needs to be supplemented by a procedure of concretization that explains how the fundamental law formulated in the initial idealizational theory Mk is affected by secondary factors. This is done through the introduction of some corrections into the initial formula of the idealizational statement: the previously adopted idealizing assumptions are removed step by step, whereas the consequent is corrected accordingly. Therefore, the fundamental idealizational model is gradually concretized: The Mk−1 model is the concretization of the Mk model and describes how a given phenomenon is affected by secondary factors omitted in. The concretization is continued until the model Mi is reached (together with i – idealizing assumptions) that can be approximated into reality. The last model of this sequence (AMi) delineates how dependencies described in the model Mi are approximated to the empirical reality (Nowak 1992, p. 13). 1 Methodological, theoretical and ideological layers almost never occur in Rawls’ work ­separately. It is rather dominated by mixing theoretical and methodological issues with ­ideological ones.

246

Przybysz

The method of idealization, understood in that way, has been used for constructing new theories and for methodologically re-constructing the existing ones, both in the sciences and the humanities. Unfortunately, the peculiarities of the method have not been sufficiently analyzed with respect to their suitability for building or interpreting theories in different social sciences, such as theories of politics or social philosophy. Some analyses have been made to interpret the theories of historical materialism and Marxian social thought by means of the conceptual apparatus of the its (Nowak 1983). However, it is possible that the method of idealization and concretization has wider applications and can also used within other ideological traditions, e.g. liberalism. In line with this presumption I am going to demonstrate that elements of idealizational approach are also found in contemporary theories of liberalism. 2

Preliminary Characteristics of Rawls’ Theory

The aim of this reconstruction is to show that Rawls’ theory is constructed on the basis of the scheme recommended by the its. First of all, I will make an attempt to prove that Rawls’ model of the person who chooses the fundamental principles of justice is an idealizational construct. Secondly, I show that the theory of justice recommended by Rawls is constructed in accordance with the principles of idealization and concretization. Before reconstructing Rawls’ choice of the principles of justice theory, it is worth describing briefly the most essential elements – theoretical and ­methodological – included in this conception. Hence, let us start from the ­initial assumption that Rawls’ theory is: (*) theory of justice. The point of departure of the theory of justice is a thesis that the aim of the political philosophy that addresses the issues of justice is to go beyond the ordinary register and description of people’s opinions. Rawls’ goal is to demonstrate the deep structure of justice; this requires the development of a complex theoretical construction. He writes: “I wish to stress that a theory of justice is precisely that, namely, a theory …. A theory of justice is subject to the same rules of methodology as other theories.” (Rawls 1971, pp. 50–51). (**) contractarian theory of justice.2 On the basis of this theory, a mode of justifying fundamental principles plays a special role. Rawls claims that they 2 Kukathas and Pettit (1990, pp. 17–35) classified the formulation of contract issues in liberalism in the context of their relation to Rawls’ theory. About the contract theory in the political philosophy, compare: Hampton (1986).

The Choice of the Principles of Justice

247

will be necessarily and sufficiently justified, if they are chosen as part of the social contract. A crucial subject of the author’s interest includes procedures and rules that may lead to such a contract and ensure its impartiality. Rawls writes: “My aim is to present a conception of justice which generalizes and carries to a higher level of abstraction the familiar theory of the social contract as found, say, in Locke, Rousseau, and Kant” (Rawls 1971, p. 11). (***) hypothetical theory of justice. For the considered theory, however, it is essential to disregard both the historical interpretation of the social contract and actual opportunities to consent and agreement in existing social arrangements. The social contract is rather “a thought experiment” in which the contract is concluded only in a simulated situation in compliance with strictly specified conditions. Thus, the argumentation structure is a conditional clause, e.g. if specific conditions X, Y and Z were fulfilled, the contract Q would be agreed. (****) decisionist theory of justice. It is also characteristic that the issue of the social contract is reduced by Rawls to the choice between alternative principles of justice made in a hypothetical situation by persons or, actually, by a person who is a typical representative of all the others.3 The characteristics of this situation show that, to employ the language of decision theory, the choice is made by a person in conditions of uncertainty. Therefore, the only criterion of legitimacy and legality of the principles of justice is that they would be chosen in this hypothetical situation. (*****) idealizational theory of justice. The fundamental decision pertaining to the choice of the relevant principles of justice is possible only if some strict conditions described by idealizing assumptions are observed. These assumptions, however, exemplify an evident discrimination of normally existing social and economic relations and psychological conditions. As a result, the fundamental theorems of a theory concerning the choice of the principles of justice should have a status of the idealized theorems. In the following section of the paper I shall focus on the methodological aspects of the theory mentioned above. I wish to demonstrate that the idealizing status should be granted to Rawls’ models of man and society, i.e. the original position in which the principles of justice are selected. 3 E.g. Kukathas and Pettit (1990, p. 21). According to some scholars who associate the contract issue with haggling and reaching a compromise, the decisive representation of the contract issue proposed by Rawls is fundamentally erroneous. Rawls’ contract is not a result of the agreement and compromise between individuals as any diversification of interests is nipped in the bud beforehand to such an extent that it is pointless discussing a compromise between the aspirations and aims of various individuals.

248 3

Przybysz

Idealization as a Method of Constructing the Original Position

The fundamental aim of the author of A Theory of Justice was to construct a specific model of the social situation referred to as “the original position” in which the fundamental principles of justice are chosen. It is the equivalent of the “state of nature” similar to those known, for example, from papers of such classical representatives of liberalism as Hobbes and Locke. However, the model of choosing fundamental principles of justice proposed by Rawls was developed on the basis of a vast array of simplifying and idealizing assumptions. The idealizational characteristic is applied by Rawls to the representation of people who make such choice. Such persons are subject to numerous idealizing ­assumptions which, according to Rawls, make them conscious and responsible persons to make decisions. The adoption of idealizing assumptions means – as it seems – that there is a will to disregard any coincidental, and hence unreliable or irrelevant features of their nature and to keep only those that may permanently and fundamentally determine the moral nature of the persons. Therefore, we receive the idealizational construct of “the original society” that is beyond history and normal public life. And such idealized persons are competent to choose the principles of justice describing the form of the basic political and economic structure of the society. In this way, the idealized persons choose universal and general principles of justice in the idealized society.4 Realistic and Idealizing Assumptions that Characterize the Original Position Amongst assumptions used by Rawls to construct his model I would like to distinguish between two different types of them: (1) realistic assumptions, and (2) idealizing ones. In general, a realistic assumption on the basis of the its is an assumption that characterizes features of a realistic domain. It is said more precisely that a realistic assumption based on knowledge K is an assumption that capture properties of the object, if the knowledge K does not show that the object having this properties does not exist (Nowak 1980, p. 28). I maintain that in A Theory of Justice Rawls characterizes a model of person by means of the following realistic assumptions: 3.1

4 The best known dispute with Rawls’ theses providing that “general” individuals choose “­general” principles of justice was carried out by Robert Nozick (1974).

The Choice of the Principles of Justice

249

R.1: Persons coexist together on a definite territory. R.2: Each person is similar in physical and mental powers to other persons at any rate inasmuch as that no one among them can dominate the rest. R.3: Persons are vulnerable to hostile attacks of other persons, but they are also able to protect themselves against them. R.4: They have their own plans for life. R.5: They desire primary social goods. R.6: The area where persons coexist together is characterized by a moderate scarcity of goods. These assumptions preliminarily characterize the original position in which the principles of justice are chosen. Therefore, it is assumed that the person who makes a choice lives with others in a specific territory considered as the place of the making of the contract. Rawls writes: “many individuals coexist together at the same time on a definite geographical territory” (Rawls 1999, p. 109). All people living in this territory are similar to one another (mentally and physically) and the level of this resemblance decides that none of them is able to dominate the others. They are “roughly similar in physical and mental powers; or at any rate, their capacities are comparable in that no one among them can dominate the rest” (Rawls 1999, pp. 109–110). Further, it is realistically assumed that the analyzed persons are exposed to hostile actions by other people and their own plans can be thwarted by a coalition of the remaining persons: “They are vulnerable to attack, and all are subject to having their plans blocked by the united force of others” (Rawls 1999, p. 110). The persons have their own different plans and goals in life: “These plans, or conceptions of the good, lead them to have different ends and purposes, and to make conflicting claims on the natural and social resources available” (Rawls 1999, p. 110). It is also essential to assume that the natural purpose of the persons’ actions and aspirations is specific goods whose distribution rules this person is interested in. One of the kinds of goods for which the persons strive are the so-called “social primary goods” comprising: “rights, liberties, and opportunities, and income and wealth” (Rawls 1999, p. 54).5 The goods at the disposal of people are limited.6 5 Rawls mentions that primary social goods do not exhaust all goods within reach of the individual: “Other primary foods such as health and vigor, intelligence and imagination, are natural goods; although their possession is influenced by the basic structure, they are not so directly under its control” (Rawls 1999, p. 89). 6 “Finally, there is the condition of moderate scarcity understood to cover a wide range of ­situations. Natural and other resources are not so abundant that schemes of cooperation

250

Przybysz

On the other hand, let us analyze idealizing assumptions used by Rawls to construct the original position model. According to the its the idealizing assumption, in a broad sense, is an assumption according to which the modeled object (i) has a certain feature which a real object does not have or (ii) does not have a certain feature which a real object does or (iii) has a certain feature in a greater or lesser value than in a real object (particularly to a minimum and maximum volume) (cf Nowak 1980, pp. 28–29; Nowakowa, Nowak 2000, pp. 431–433). Employing idealizational characteristics Rawls is able to omit an array of features that characterize real and historical persons such as the failure to comply with contracts or envy. The idealizing assumptions cause that the model person becomes a rational and responsible decisive person, in short, an impartial person competent to choose the principles of justice governing the social order. I assume that Rawls’ model of the person is characterized by the following idealizing assumptions: Z.1: Persons have a sense of justice and they can rely on one another with respect to the compliance of the agreed principles. Z.2: Persons are rational. Z.3: Persons are disinterested in one another, e.g. they are not moved by envy. Z.4: They take care of the welfare of future generations. Z.5: Persons do not have knowledge of: (a) their own position in the society; (b) any economic and generational divisions in the structure of the ­society, to which they belong; (c) the civilizational level of their society, to which they belong, its ­political culture and natural resources. In the original position the persons are considered to have a sense of justice and “to understand and to act in accordance with whatever principles are ­finally agreed to” (Rawls 1999, p. 125).7 Whereas, it is known that people do not always follow their choices and they are frequently suspicious as to whether others will comply with the principles and rules of conduct or not.8 Hence, become superfluous, nor are conditions so harsh that fruitful ventures must inevitably break down. While mutually advantageous arrangements are feasible, the benefits they yield fall short of the demands men put forward” (Rawls 1999, p. 110). 7 “The parties are presumed to be capable of a sense of justice and this fact is public knowledge among them. This condition is to insure the integrity of the agreement made in the original position” (Rawls 1999, p. 125). 8 “The assumption only says that the parties have a capacity for justice in a purely formal sense: taking everything relevant into account, including the general facts of moral psychology,

The Choice of the Principles of Justice

251

this assumption is of an idealizational nature. Moreover, it is assumed that persons who are in the original position behave rationally.9 At the same time, it is assumed in a counter-factual way that the persons are not interested in the others10 as they do not have any negative affective feelings to each other such as envy (or even shame and humiliation, Rawls 1999, p. 144) that would be meaningful in the event of the choosing of the principles of justice.11 Another assumption provides that when choosing the principles of justice they take into consideration their future descendants’ interests, which means that they should consider them as representatives of the future descendants’ interests. However, the most crucial role in Rawls’ theory is played by the next assumption that limits the persons’ knowledge (Rawls 1971, pp. 136–141). Rawls applies the veil of ignorance and maintains that it deprives the persons who are in the original position of their knowledge. It means that the persons are subject to information restrictions, and hence their decisions about the choice of the principles of justice are not affected by any detailed and circumstantial information and attitudes resulting from individual preferences or social conditions. Rawls states that the persons do not know “any details about their own plans in life, the strength of their own intelligence, inclinations, and psychical predispositions”, nor their place in the social structure. The persons are deprived of information on details concerning their own society. They do not know “their political and economic condition”, nor the level of its cultural and civilizational development: It is assumed, then, that the parties do not know certain kinds of particular facts. First of all, no one knows his place in society, his class position or social status; nor does he know his fortune in the distribution of natural assets and abilities, his intelligence and strength, and the like. Nor, again,

9 10

11

the parties will adhere to the principles eventually chosen” (Rawls 1999, p. 126). Further, he also assumes that when “principles are acknowledged the parties can depend on one another to conform to them” (Rawls 1999, p. 125). “I have assumed throughout that the persons in the original position are rational” (Rawls 1999, p. 123). One feature of justice as fairness is to think of the parties in the initial situation as rational and mutually disinterested. This does not mean that the parties are egoists, that is, individuals with only certain kinds of interests, say in wealth, prestige, and domination. But they are conceived as not taking an interest in one another’s interests (Rawls 1999, p. 12). “The special assumption I make is that a rational individual does not suffer from envy. He is not ready to accept a loss for himself if only others have less as well. He is not downcast by the knowledge or perception that others have a larger index of primary social goods” (Rawls 1971, p. 143).

252

Przybysz

does anyone know his conception of the good, the particulars of his rational plan for life, or even the special features of his psychology such as his aversion to risk or liability to optimism or pessimism. More than this, I assume that the parties do not know the particular circumstances of their own society. That is, they do not know its economic or political situation, or the level of civilization and culture it has been able to achieve. The persons in the original position have no information as to which generation they belong. rawls 1999, p. 118

I suggest singling out three areas of knowledge subject to the veil of ignorance: (a) the person’s knowledge of his own, including the knowledge of “his own position in the society, knowledge of his own dispositions and skills, details on his own plans for life or even specific features of his own psychology such as a propensity to risk” and the like, (b) knowledge of economic and generation divisions in the society, and (c) knowledge of natural resources, the level of economic development and political culture achieved by this society. In spite of the “veil of ignorance” idealized Rawls’ persons still have nomological knowledge. Hence, they understand the general sense of politics and know the general rights of economics. They have a general knowledge of the basics of social organization and of the general rights of psychology or economics.12 I assume that the above assumptions from R.1 to R.6 and from Z.1 to Z.5 ­characterize Rawls’ conditions of the original position in which the persons make the impartial choice of the principles of justice. 3.2 Selection of Principles of Justice According to interpretation, the fundamental aim of the author of A Theory of Justice is to prove that persons characterized in that way choose two principles of justice in the idealized original position. Therefore: If the persons are in the position characterized by assumptions R.1– R.6 and Z.1–Z.5(a)–(c), they will choose the following two principles of justice: 12

“It is taken for granted, however, that they know the general facts about human society. They understand political affairs and the principles of economic theory; they know the basis of social organization and the laws of human psychology. Indeed, the parties are presumed to know whatever general facts affect the choice of the principles of justice” (Rawls 1999, pp. 119).

The Choice of the Principles of Justice

253

The first principle of justice Each person is to have an equal right to the most extensive basic liberty compatible with a similar liberty for others rawls 1971, p. 60

The second principle of justice Social and economic inequalities are to be arranged so that they are both: (a) to the greatest benefit of the least advantaged and (b) attached to offices and positions open to all under conditions of fair equality of opportunity. rawls 1999, p. 83

Rawls endeavors to show that the fundamental sense of these principles is expressed by the idea of equal liberties (first principle) and by economic ­inequalities that are to be arranged only if they are advantageous for the least advantaged (second principle). He holds that the persons will be more willing to agree on equal than unequal liberties and on safe inequality more than ­other risky solutions. The author of A Theory of Justice maintains that the choice of the above principles results from the fact that the rule for choice in the decisive situation is the maximin principle. This is caused by the fact that the persons who are affected by, among others, the veil of ignorance and are not aware of their own political and economic position in society will be inclined to choose those principles that maximally minimize the risk of making decisions in this regard. Rawls constructed the original position in such a way that the maximin rule was a natural decision rule (Rawls 1999, pp. 132–136). In view of this rule, the first principle of justice asserting that all the persons are to have an equal right to basic liberties dominates over and above any other principles laying down the unequal distribution of liberties because, regardless of the real position of the person in the society (when the person needs to make a choice, he or she does not have any knowledge of it), it guarantees the satisfactory minimum; in the worst case, a given person will have the same liberties as the others. The same applies to the second principle of justice. In the case of a different choice, it could emerge that the person who is actually economically handicapped and is not aware of this fact (veil of ignorance) would choose a disadvantageous solution, e.g. by providing better salaried persons with additional benefits to the detriment of himself. Rawls strives to prove that risky rules of conduct, e.g. a promise of large profits and, simultaneously, a high risk of failure, will not be appreciated by the

254

Przybysz

idealized person due to other reasons, as well. The assumption laying down that the person is uninterested in others and is deprived of envy decides that he or she is not affected by arguments such as another person has more than I do, and hence there is no reason to suppose that he or she will strive for more shares in the pool of social goods than all the others. On the other hand, due to this person’s rationality, he or she will not agree to reduce his or her shares, ­either. Therefore, the proposed principles are to guarantee the minimum safety: “the minimum assured by the two principles in lexical order is not one that the parties wish to jeopardize for the sake of greater economic and social ­advantages.” (Rawls 1999, p. 135) 4

Concretization of the Principles of Justice

The author of A Theory of Justice maintains that the theory proposed in his book comprises two parts that supplement each other. The first “ideal” part consists of the original position model reconstructed above, where in the idealized characterized situation two principles of justice are chosen, and the second “non-ideal” part includes assumptions on the operation of the principles of justice in slightly more realistic conditions: The intuitive idea is to split the theory of justice into two parts. The first or ideal part assumes strict compliance and works out the principles that characterize a well-ordered society under favorable circumstances. It develops the conception of a perfectly just basic structure and the corresponding duties and obligations of persons under the fixed constraints of human life. My main concern is with this part of the theory. Non-ideal theory, the second part, is worked out after an ideal conception of justice has been chosen; only then do the parties ask which principles to adopt under less happy conditions. This division of the theory has, as I have indicated, two rather different subparts. One consists of the principles for governing adjustments to natural limitations and historical contingencies, and the other of principles for meeting injustice. rawls 1971, pp. 244–245

Unfortunately, Rawls’ theory does not have a distinct procedure for the transition between the ideal and non-ideal parts of the theory of justice and a lack of clear-cut guidelines in this regard constitutes a fundamental difficulty for this reconstruction. Declaratively, Rawls suggests using the so-called reflective

The Choice of the Principles of Justice

255

equilibrium method. This method requires a start with any commonly shared and potentially weak assumptions and the use of them to reach “a significant group of rules.” However, if it appears to be impossible, it is recommended to revise the initial assumptions and – on the basis of them – once again make an attempt to reach general rules. This procedure, which is only outlined here, is sometimes characterized by Rawls as a method to discover general and fundamental rules and ideas, or a method to reconcile them with any commonly shared convictions and judgements (Rawls 1999, pp. 18, 48–50). It may be ­intuitively envisioned that this method is used to concretize general rules in order to tailor them to our beliefs and to change our beliefs in order to adjust them to the rules (compare e.g. Dworkin 1977, pp. 292–323, and Sadurski 1988, pp. 20–22). I will claim that although Rawls’s method of reflective equilibrium is not identical with the method of concretization as it is understood within the framework of the its, they both lead to the same result. The method of reflective equilibrium, just like the method of concretization, makes it possible to transition from statements and principles at a higher level of abstraction and idealization to less idealized and more concrete ones. The difference between them lies in that the method of reflective equilibrium points to a particular, mental and reflective, mechanism by which people change their ideas and principles, while the method of concretization, understood in terms of the its, only points to the formal procedure – i.e. the removal of the idealizing ­assumptions – leading from more idealized statements and principles to less idealized ones. Nevertheless, the two methods are not contradictory to each other. Both can facilitate building up of increasingly realistic, concretized theories. That is why I will be referring to Rawls’s method of reflective equilibrium as a method of concretization of the theoretical model of the original position. Let us begin with what we know on the basis of A Theory of Justice. First of all, it is known that the idealizational model of the person is concretized: the veil of ignorance is lifted (Rawls 1999, pp. 195–201), hence the image of the person must change. Secondly, it is known that the principles of justices themselves are concretized, as well: their contents fundamentally change as they are provided with the so-called priority rules (cf. Rawls 1999, pp. 250, 302–303). These two things are known for sure. Nevertheless, the applied concretization procedure is not sufficiently clear. This problem is quite significant ­because – according to the assumptions of the its whose method of the idealizational reconstruction of scientific theories is applied here – the researcher is required to distinctly concretize his or her idealizational theorem and make an attempt to make it more realistic. The concretization of the idealizational

256

Przybysz

theorem should be as follows: having removed a specific idealizing assumption from the antecedent of this theorem, the succession of this theorem is modified. Literally speaking, Rawls’ theory includes two threads of its concretization. The first thread describes how the veil of ignorance is removed and leads to deliberations on more specific – and hence closer to reality – models of the person who decides about more concrete rights that govern the society. The second thread in A Theory of Justice includes very vast and dispersed deliberations on the necessity to concretize the principles of justice which are to apply to more and more specific circumstances and this thread ends with formulating the concretized version of the principles of justice. Now, I am going to present an interpretation that aims at joining these two threads of concretization. Due to the fact that Rawls abandoned any deliberations on this issue, this proposal may seem to be slightly artificial in view of the loyalty to the character of Rawls’ deliberations, but I believe that it is completely natural with respect to the inner logic of the reconstructed idealizational theory. Let us repeat again: it is true that Rawls shows the concretized form of the principles of justice that are to function in more realistic conditions, thus he should have earlier removed those conditions that are very idealized in the original position model. And de facto he does it as he gradually “removes the veil of ignorance”. But the thing is, Rawls’ work does not explicitly show the relationship between these two threads of concretization. I make an attempt to show the nature of this relationship that may be found in Rawls’ work. In view of the interpretation adopted herein, the ideal and non-ideal parts of Rawls’ theory are developed as part of the so-called four-stage sequence. This sequence is a basic scheme for developing Rawls’ theory that characterizes its subsequent stages that follow one after another (Rawls 1999, pp. 195–201). The sequence of the subsequent stages is – according to me – determined by the concretization rule. Intuitively speaking, it means that consequent stages of the development of the theory of justice (i.e. its subsequent models) are ordered linearly with respect to its level of idealization, i.e. force and type of systematic simplifications. A characteristic feature of each level is its decreasing idealization. We start with the description of the persons who are in the original position. In this highly idealized situation the persons make impartial, unanimous and rational choice of basic principles of social justice. At the next, less “abstractive” stage the persons are to find a relevant constitution and evaluate it on the basis of the previously chosen principles of justice, but this time – according to my ­interpretation – in a concretized version. The next stage is legislative where

The Choice of the Principles of Justice

257

the persons choose relevant fair rights that particularly pertain to the operation of the economic and social infrastructure of the country. These rights must also be assessed as to their compliance with the principle of justice, but in a more concretized form. What condition needs to be satisfied in order to go from the strongly idealized model of the original position to the constitutional stage model and to any further models? It seems that it is necessary to revise the assumptions of Rawls’ initial model. While the unanimous and rational choice of basic principles of justice was only possible in the cases of the persons characterized by a complete set of these assumptions, i.e. R.1–R.6 and Z.1–Z.5, the performance of the remaining roles by the persons – constitution creators, legislators and citizens – is only possible thanks to the systematic removal of the assumptions that characterize them. Due to the fact that Rawls’ man model is brought closer to reality, two principles of justice must also be subject to concretization. The need to concretize two principles of justice is affected by the changing situation of the person that results from the transition from the original position to the constitutional order and then to the legislative one. At the subsequent stages of social reality, the social perception of the world changes as a result of the removal of the veil of ignorance. The person who is at the constitutional or legislative stage will understand justice slightly differently as his or her knowledge of various circumstances is greater. Changes make it necessary to again review the formula of principles of justice with respect to their possible adjustment to new conditions. Any corrections made in this way form the basis for formulating more realistic versions of the justice concept which the citizens should take into consideration when choosing the constitution and proposing more detailed legal acts pertaining to any social and economic policy. In the light of this interpretation any amendments to the subsequent versions of the principles of justice do not seem to be coincidental supplements, but they are more systematic and form the basis for changing the formula of principles of justice. So, at the constitutional stage the persons seek a relevant constitution evaluated on the basis of the principles of justice (ibidem, p. 198). On the one hand, it must comply with the previously adopted principles, but on the other hand – as with every constitution – it must take into consideration the political culture of a given country, cultural and civilizational realities. Therefore, the assumption reconstructed as Z.5(c) must be removed. Through the partial removal of the veil of ignorance the persons regain their partial knowledge of the society in which they live and this knowledge is necessary to establish the constitution. From this moment, they have access to any previously unknown

258

Przybysz

information on the civilizational and cultural level of their society, including the economic development and natural resources. Rawls writes: Since the appropriate conception of justice has been agreed upon, the veil of ignorance is partially lifted … they now know the relevant general facts about their society, that is, its natural circumstances and resources, its level of economic advance and political culture, and so on. rawls 1971, p. 197

At the stage of the constitutional convention, the persons still do not have any detailed knowledge of themselves and they do not have, e.g. any access to ­information about any existing economic divisions in the society, e.g.: “The persons in the convention have, of course, no information about particular persons: they do not know their own social position, their place in the distribution of natural attributes, or their conception of the good” (Rawls 1971, p. 197). According to my interpretation, due to the partial removal of the veil of ­ignorance, i.e. the abolition of assumption Z.5(c), the principles chosen in the original position must be reviewed, again. Since when persons became aware of the civilizational and cultural conditions existing in the society, the most urgent business has been to concretize the principles of justice. Hence, the constitutional order should be evaluated on the basis of such concretized ­principles of justice. According to Rawls the first principle named the principle of equal rights is essential for the constitutional order: an equal right for everyone to the most extensive scheme of basic liberties for all people. However, after the partial removal of the veil of ignorance, when the person has knowledge of the civilizational and cultural position of the society in which he or she lives (e.g. of the existing political culture, natural conditions and the like), the situation changes. The persons who are currently taken into consideration are able to notice the difference among various types of liberties, e.g. among various types of political liberties. They are not presently considered as qualitatively homogeneous, and hence relationships among them are much more complex. Therefore, instead of simply referring to the basic liberty, then it is necessary to mention a system of basic liberties. Maybe, Rawls bears it in mind when he writes: Thus the delegates to a constitutional convention, or the members of the legislature, must decide how the various liberties are to be specified so as to yield the best total system of equal liberty. They have to balance one liberty against another. The best arrangement of the several liberties

The Choice of the Principles of Justice

259

depends upon the totality of limitations to which they are subject, upon how they hang together in the whole scheme by which they are defined. rawls 1971, p. 203

Any amendments to this principle pertain to two closely related issues specified by this principle. Firstly, in the light of cultural and civilizational conditions which the person is already aware of, it is necessary to review the idea of the broadest fundamental liberty, i.e. the broadest scope of liberty. The range of liberty for all the people must sometimes be narrowed due to liberty itself: a smaller share of liberty for all the people means strengthening the entire system of liberties. For example, the liberty of conscience which in the ideal part of the theory is not restricted, presently – in the unideal part – it needs to be adjusted to the political liberty that constitutes a guarantee of the public order: “Liberty of conscience is limited, everyone agrees, by the common interest in public order and security” (Rawls 1971, p. 212). This happens when we place, in advance, liberties of all the people within a certain scope in order to prevent a situation when – accidentally – intolerant sects would communicate to their believers the idea of fight against the unbelievers or when we restrain various groups that fight each other from using violence. The restriction of liberties and the prohibition of all people against taking some actions will be advantageous to the whole system of liberties of all (Rawls 1999, pp. 246–250). Secondly, situations that arise in the social reality also indicate a need to modify the idea of equal liberty (resp. equal rights as formulated by the first principle). The attention is drawn to the fact that situations in which we rigidly keep equal liberties for all are frequently questioned by natural restrictions, accidents of human life, accidental historical or civilizational circumstances which the persons are already aware of. For example, in certain situations it is necessary to approve some political inequalities and any related inconveniences in order to avoid much greater injustices. Even the slavish system affected the idea of liberty less than some stricter solutions (e.g. those which preferred killing soldiers who were prisoners of war). If we could choose between them, it would be rational to select the first one despite the explicit injustice of enslavement. However, importantly, Rawls maintains that the possible acceptance of this state of affairs by the most injured parties, i.e. those who, in this case, received a smaller share of liberties than others, always constitutes the criterion of inequality admissibility. The result of this argumentation is that it is only admissible, as shown, in the conditions in which the persons regain at least their part of the detailed knowledge of the society in which they live, e.g. the activity of intolerant sects or various inconveniences concerning inequalities in the political milieu, is to

260

Przybysz

specify the first principle of justice to the extent that it would cover the threads in a form of a special clause called: “the first priority rule”.13 The specific fundamental theorem of the theory of justice will be as follows: If the persons are in a situation characterized by assumptions R.1–R.6 and Z.1–Z.5(a)–(b), they will need a specific version of the first principle of justice: The first principle of justice Each person is to have an equal right to the most extensive total system of equal basic liberties compatible with a similar system of Liberty for all. rawls 1971, p. 250

Priority rule: The principles of justice are to be ranked in lexical order and therefore liberty can be restricted only for the sake of liberty. There are two cases: (a) a less extensive liberty must strengthen the total system of liberty shared by all, and (b) a less than equal liberty must be acceptable to those citizens with the lesser liberty. rawls 1971, p. 250

In turn, at the legislative stage the persons decide on choosing the principles that govern the social and economic order. At this point, it is necessary to dismiss the assumption reconstructed by myself as Z.5(b), thanks to which the persons would regain their knowledge of economic and social inequalities that arise in the society: “At this point the full range of general economic and social facts is brought to bear. The second part of the basic structure contains the distinctions and hierarchy of political, economic, and social forms which are necessary for efficient and mutually beneficial social cooperation” (Rawls 1971, p. 191). This new knowledge is essential in establishing the most universal rights concerning the function of economic institutions. At the stage of legislature, the veil of ignorance assumption was not, however, completely dismissed and the persons are still deprived of particular knowledge of themselves, ­private preferences, own biographies and knowledge of their own plans for life.

13

The thesis that remains effective at a constitutional level in whole pertains to the superiority of liberty over any other goods. Therefore, liberty is an unexchangeable value and may be restricted only due to itself, in other words, the restriction of civil liberties cannot be compensated, e.g. through increasing economic benefits.

The Choice of the Principles of Justice

261

At this stage, the second of the selected principles must be reviewed. It comprises – let me remind – principle 2(a) that postulates a specific method of the most just distribution of economic inequalities, and principle 2(b) that considers access to positions and public functions as one of the conditions that may legitimize the existence of economic inequalities. As it is with the concretization of the first principle, two points would presently be amended: the distribution of inequalities to make it beneficial for the least advantaged and fair opportunities to access public positions. And similarly to the concretization of the first principle, changes would need to be made in the formula of the second principle and take the form of another priority rule. Firstly, at this stage of the theory development it would be necessary to consider the unfair social opportunities with respect to access to status and positions in public institutions. The persons who are aware of the economic and social inequalities in society must realize that such inequalities exist. Similarly to the restriction of equal liberty, the admissibility of unfair opportunities, e.g. due to specific historical or natural circumstances, may take place only when it strengthens and ensures the positions of the least privileged. Rawls claims that it is necessary to accept some specific inequalities, if they are advantageous to the least advantaged in comparison with each alternative determined for them: We must also claim that the attempt to eliminate these inequalities would so interfere with the social system and the operations of the economy that in the long run anyway, the opportunities of the disadvantaged would be even more limited. The priority of fair opportunity, as in the parallel case of the priority of liberty, means that we must appeal to the chances given to those with the lesser opportunity. We must hold that a wide range of more desirable alternatives is open to them than otherwise would be the case. rawls 1971, p. 301

Secondly, the analogous line of thought recognizes the exception to the principle laying down that inequalities are distributed in such a way that they are advantageous to the least advantaged. Currently, Rawls expands it with a clause providing that taking care of the benefits of the least advantaged should be consistent with a fair saving principle for future generations. This issue may be illustrated by Keynes’ line of thought. He holds that the excessive (and advantageous to the future) capital accumulation in the 19th century was realized by the capitalists: Society in the nineteenth century, he says, was arranged so as to place the increased income in the hands of those least likely to consume it.

262

Przybysz

The new rich were not brought up to large expenditures and preferred to the enjoyments of immediate consumption the power which investment gave. rawls 1971, pp. 298–299

but its burden was carried by the workers. Although their circumstances appear harsh, Keynes presumably maintains that while there were many ostensible injustices in the system, there was no real possibility that these could have been removed and the conditions of the less advantaged made better. Under other arrangements the position of the laboring man would have been even worse (Rawls 1971, p. 299). Rawls seems to accept this last element of Keynes’ line of thought, and hence there occurs his clause laying down that inequalities that are not arranged to the distinct benefit of the least advantaged are admissible, if they may be treated as saving costs for the future generations, considering that any burdens carried by those to whose detriment they are established would be greater in a different social configuration. “Whenever the constraints of justice in the matter of savings are infringed, it must be shown that circumstances are such that not to trespass upon them could lead to an even greater injury to those on whom the injustice falls” (Rawls 1971, p. 299). It should be noted that this line of thought became possible following the regaining by the idealized persons of the piece of knowledge of generation divisions in the society, thanks to which they are presently able to distinguish the interest of their own generation and the interest of future generations. This knowledge imposes a relevant clause that restricts any current economic actions on the second principle through using a proper rate of saving for the benefit of any future generations. Therefore, at this stage Rawls’ theorem reads as follows: If the persons are in a situation characterized by assumptions R.1–R.6 and Z.1–Z.5(a), they will need a specific version of the second principle of justice: The second principle of justice Social and economic inequalities are to be arranged so that they are both: (a) to the greatest benefit of the least advantaged, consistent with the just savings principle, and (b) attached to offices and positions open to all under conditions of fair equality of opportunity. rawls 1971, p. 302

The Choice of the Principles of Justice

263

The second priority principle The second principle of justice is lexically prior to the principle ofefficiency and to that of maximizing the sum of advantages; and fair opportunity is prior to the difference principle. There are two cases:(a) an inequality of opportunity must enhance the opportunities of those with the lesser opportunity; (b) an excessive rate of saving must on balance mitigate the burden of those bearing this hardship. rawls 1971, pp. 302–303

The concretization procedure, during which assumptions Z.5(b)–Z.5(c) are ­removed, proves that at each of the subsequent more specific and realistic stages of the political reality the persons are characterized by the lesser number of idealizational assumptions (that pertain to the persons’ knowledge). 5

The Idealizational Character of Rawls’s Theory: Rawls and Chomsky

The aim of the methodological reconstruction of Rawls’s theory of the choice of the principles of justice, presented above, was to demonstrate the role idealizational constructs play in that theory. As has been shown, in Rawls’s theory, on the individual level, the status of the idealizational construct applies to a person who chooses the basic principles of justice. On the social level the idealizational construct is the so-called original position defined as a model of the circumstances in which individuals make such a basic choice and which is, at the same time, the idealizational prototype of a society consisting of rational people who have a sense of justice. We have seen that the models do not have real-life counterparts in historical and contemporary societies. They are of a hypothetical nature. The construction of such a model is a kind of thought experiment in which people’s behavior is analyzed in an idealized and isolated situation. According to the idealizational theory of science, the researcher who employed the idealizational method in contemporary humanities was Noam Chomsky (Nowakowa, Nowak 2000, pp. 325–337). In Chomsky’s opinion, language is subject to significant distortions in everyday use. Those distortions hinder access to the latent language structures. As he claims, what is observed in the surface layer of language is neither important nor significant, while that which is important and significant is often very hard to observe (Chomsky 1965). Chomsky’s main postulate is to focus on the deep structure of language, which we can access by idealization from natural language. Chomsky juxtaposed his own linguistic theory to the paradigm still prevalent in linguistics in 1960s, that is, to structuralist distributivism (Harris 1963).

264

Przybysz

Chomsky questioned the thesis that one can build an adequate theory of ­language with the use of systematic description, classification, and inductive generalizations only. He discarded descriptive linguistics which remained on the level of the surface of language, for the sake of a theory which looked for deep rules governing language: Any interesting generative grammar will be dealing, for the most part, with mental processes that are far beyond the level of actual or even potential consciousness; furthermore, it is quite apparent that a speaker’s reports and viewpoints about his behavior and his competence may be in error. Thus a generative grammar attempts to specify what the speaker actually knows, not what he may report about his knowledge. chomsky 1965, p. 9

An analogous attitude one can find in Rawls’s book. He claims that an accurate explanation of moral dispositions, such as a sense of justice, must be based on something more than observation and classification of common convictions: There is no reason to assume that our sense of justice can be adequately characterized by familiar common sense precepts, or derived from the more obvious learning principles. A correct account of moral capacities will certainly involve principles and theoretical constructions which go much beyond the norms and standards cited in everyday life; it may eventually require fairly sophisticated mathematics as well. rawls 1971, p. 47

Both Chomsky and Rawls, then, reject, in their respective disciplines, the method based on collecting data. Instead, they wish to build theories which would go beyond standard reporting and classification of what appears on the surface of language and social life. Another similarity concerns the research program offered by the two scientists. At the core of Chomsky’s theory there is the model of an ideal language user. However, in order to build such a model it is necessary to make a number of idealizing assumptions, or, more precisely, assumptions which idealize ­actual language users. According to Chomsky, an ideal language user is an ideal speaker-listener, in a completely homogeneous speech-­ community, who knows its language perfectly and is unaffected by such grammatically irrelevant conditions as memory limitations, distractions, shifts of attention and interest, and errors (random or characteristic)

The Choice of the Principles of Justice

265

in applying his knowledge of the language in actual performance. This seems to me to have been the position of the founders of modem general linguistics, and no cogent reason for modifying it has been offered. chomsky 1965, p. 3–4

Here, the question arises whether we can find an analogy to such an approach in Rawls’s theory. This appears to be the case. Rawls’s theory also involves an idealized subject – not a language user but a person choosing the principles of justice. When constructing the model of such a person, Rawls, too, availed himself of idealizing assumptions, for example, his subjects do not know the details concerning their own life plans, the political and economical situation in their society, or the cultural and civilizational level of the development of that society. At this point we can already glimpse another similarity between Chomsky’s and Rawls’s theories. The aim of Chomsky’s idealized construction of an ideal language user was to create a theory of language competence: “A grammar of a language purports to be a description of the ideal speaker-hearer’s intrinsic competence” (Chomsky 1965, p. 4). The ideal speaker-hearer can, for example, apply grammar rules without error to form correct sentences. Potentially, language competence is ascribed de facto equally to all senders and receivers of language but the actual use of it is hindered by numerous confounding factors. Thus, the functioning of language competence can only be described within the framework of an idealized model. It seems that in Rawls’s theory the application of idealizational methodology has a similar effect. Rawls’s ideal subject is ascribed – per analogiam to language competence – the competence of choosing principles of justice. Although Rawls assumes that all individuals have the competence, in real life it is blunted due to various unfavorable conditions. We can reveal how the competence functions thanks to the construction of an idealizational model in which unimportant factors are omitted. The comparison of Chomsky’s and Rawls’s theories, presented above, allow us to put forth the following hypothesis. It seems that Rawls’s theory played the same role in political philosophy as Galileo’s hypotheses and theories did in theoretical physics (Nowakowa, Nowak 2000, pp. 17–27; Such 2004, pp. 12–36), Darwin’s – in biology (Łastowski 1991; Nowakowa, Nowak 2000, pp. 63–94), Marx’s – in economy (Nowak 1980), and Chomsky’s – in linguistics (Nowakowa, Nowak, pp. 325–337). Rawls rejected the old methods of conceptualizing the phenomenon of justice, which consisted in describing and c­ lassifying common language uses and common beliefs about justice and just actions. Instead, he proposed a theory of justice of a clearly idealizational ­nature. ­Figure  11.1

266 Physics

Przybysz Galileo

Biology

Darwin

Economy

Marx

Linguistics

Chomsky

Political philosophy

Rawls 17th c.

Figure 11.1

19th c.

20th c.

Thresholds of idealization in science

provides a graphic which illustrates, in a simplified form, the transition from the ­empirical approach based on collecting data to the theoretical, explanatory approach, in various disciplines of empirical science. Such a style of research, characteristic for the scholars mentioned above, is sometimes called “Galileo’s style”. It seems that the theories have been successful for two reasons: their content and the method applied for their construction. In my interpretation the method of idealization is a crucial argument for the importance of Rawls’s theory. The method began to be applied in modern science at the time of Galileo, Rawls introduced in a systematic and conscious fashion, into political philosophy in 1960s. That was the most significant aspect of Rawls’s breakthrough in political philosophy. The role Rawls’s theory played in political philosophy, then, is similar to that of Galileo’s, Darwin’s, Marks’s, and Chomsky’s in their respective fields of science. In that sense we can even call Rawls “the Galileo of political philosophy”. At this point it ought to be emphasized that using idealizational constructs is a distinctive feature not only of Rawls’s theory but of liberal theories in ­general. A reconstruction of liberal theories will show that they contain many such idealizational models of an individual (e.g. an entrepreneur, a moral ­individual, or a decision-making subject), as well as idealizational models of society (e.g. anarcho-capitalist, contractual, utilitarian, or egalitarian). Just like Rawls’s conception of a person and of the original position, all those models are theoretically idealized constructs because they only take into account a limited number of properties characterizing an actual human being (O’Neill 1987; Hamminga, De Marchi 1994). Thus, when theoreticians build a model of

The Choice of the Principles of Justice

267

a moral individual, they usually ascribe to that individual certain natural moral rights; in the case of the model of a decision-making subject the ascribed original characteristic is the ability to make autonomous choices; in the case of the model of a utilitarian individual the emphasis is put on maximizing happiness and individual usefulness. What accounts for the differences among the particular idealizational constructs in liberal political theories are the theoretical procedures used for building them. The most salient example are the dissimilarities between Rawls’s model of a decision-making person and the model of an ideal legislator proposed in classical utilitarianism, e.g. in Jeremy Bentham’s works (cf. his 1789). The difference between these models refers to the level of knowledge attributed to such a person by the authors. While Rawls minimizes the knowledge held by the persons, the utilitarian construct presumes its maximization in a clearly idealizing way. The utilitarian legislator must, in principle, possess wide knowledge on facts and possible scenarios, as well as what profits and losses can result from them, and must be able to make, on the basis of that data, proper calculations concerning the utilitarian balance of pleasure and pain. In this case, then, the Rawls and the utilitarians differ in their symmetrical, so to speak, approach to model individuals’ knowledge: Rawls presupposes, counterfactually, that it is very limited, while utilitarians, also counterfactually, presume the opposite, that it is as extensive as possible. Other liberal political theories make just as wide a use of idealizational models of society. Hobbes’s and Locke’s models of the state of nature are classical examples of such use. Those models were designed by their authors to reconstruct the original form of social relationships and, as such, they omit a number of characteristics of human society which manifest themselves in more complex, historical and real, circumstances. The same applied to Rawls’s idea of the original position in which the historical-civilizational background of the social contract is idealizationaly omitted. Robert Nozick’s model of a state of protective associations is based on similar idealizing assumptions. From the methodological point of view, in that model there is a certain analogon to the original position. The theoretically idealized social structure is filled with individuals invested with an absolute and equal right to self-preservation, protection of private property, and a right to directly punish intruders and criminals, which is, admittedly, a highly idealized situation (Nozick 1974, pp. 10–15). Still, even though liberalism commonly relies on idealizational constructs of individuals and society, Rawls’s postulates in the field of political and moral philosophy are still innovative and of major importance. The merit of his proposal is that it shows how to make the transition from the ideal to the nonideal level of description. It seems that when Rawls combined, in A Theory of

268

Przybysz

Justice, the idealizational procedure with the concretizing one, he managed to put forth a credible transition from the level of abstraction to the level of concretization, which has always been a difficulty for theoreticians of society (cf. e.g. Simmons 2010; Hamlin, Stemplowska 2012). The tactic of removing ­subsequent idealizing assumptions and of concretizing the basic statements of a theory constitutes an alternative to many theories which do not go beyond the level of idealization. Such theories are very often criticized for excessive abstractiveness and being too far removed from the realities of life. That kind of criticism is not devoid of substance. Therefore, indicating systematic concretization as a method of moving from general to particular principles and from abstract to more realistic constructs seems worthy of our attention. References Bentham, J. (1789). Introduction to Principles of Morals and Legislation. Chomsky, N. (1965). Aspects of the Theory of Syntax. Boston: The MIT Press. Daniels, N. (ed.) (1975). Reading Rawls. Critical Studies on Rawls’ A Theory of Justice. ­Oxford: Oxford University Press. Dworkin, R. (1977). Taking Right Seriously. Cambridge, Mass.: Harvard University Press. Hamminga, B., De Marchi, N. (1994). Idealization and the Defence of Economics: Notes Toward a History. In: B. Hamminga, N. De Marchi (eds.) Idealization VI: Idealization in Economics. Poznań Studies in the Philosophy of the Sciences and the Humanities 38, pp. 11–40. Atlanta: Rodopi. Hamlin, A., Z. Stemplowska (2012). Theory, Ideal Theory and the Theory of Ideals. ­Political Studies Review 10, 48–62. Hampton, J. (1986). Hobbes and the Social Contract Tradition. Cambridge: Cambridge University Press. Harris, Z. (1963). Structural Linguistics. Chicago: Chicago University Press. Kukathas, Ch., P. Pettit (1990). Rawls. A Theory of Justice and its Critics. Cambridge: Cambridge University Press. Łastowski, K. (1991). Two Models of Evolution in Darwin’s Theory. Variability and E­ volution 1, 25–37. Nowak, L. (1980). The Structure of Idealization. Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht/Boston/London: Reidel. Nowak, L. (1983). Property and Power. Toward a Non-Marxian Historical Materialism. Dordrecht: Reidel Publishing Company. Nowak, L. (1992). The Idealization Approach to Science: A Survey. In: Brzeziński, J. Nowak, L. (eds.) Idealization III: Approximation and Truth. Poznań Studies in the Philosophy of the Sciences and the Humanities 25, pp. 9–63. Atlanta: Rodopi.

The Choice of the Principles of Justice

269

Nowakowa, I., Nowak, L. (2000). Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities 69. Atlanta: Rodopi. Nozick, R. (1974). Anarchy, State, and Utopia. New York: Basic Books. O’Neill, O. (1987). Abstraction, Idealization, and Ideology in Ethics. Royal Institute of Philosophy Lecture Series 22, 55–69. Pogge, Th. (2007). John Rawls: His Life and Theory of Justice. Oxford: Oxford University Press. Rawls, J. (1971). A Theory of Justice. Harvard: Belknap Press. Rawls, J. (1999). A Theory of Justice. Revised Edition. Harvard: Belknap Press. Sadurski, W. (1988). Teoria sprawiedliwości. Podstawowe zagadnienia. Warszawa: PWN. Simmons, A.J. (2010). Ideal and Non-Ideal Theory. Philosophy & Public Affairs 38(1), 5–36. Such, J. (2004). Multiformity of Science. Poznań Studies in the Philosophy of the Sciences and the Humanities 79. Atlanta: Rodopi. Young, S. (ed.) (2009). Reflctions on Rawls: An Assessment of His Legacy. Burlington: Ashgate Publishing Company.

chapter 12

Inductive Modeling and Discursive Idealization in Scenario Planning1 Zenonas Norkus Abstract This paper discusses the distinctive features of modeling and idealization in ­scenario planning. Scenarios are analytical narratives or complex causal stories about the ­distant future which are constructed in future studies and strategic management (business studies). They are different from conventional scientific predictions that are expected to be true or most probable, because scenario planning focuses on the uncertain and open aspects of the future which are the sources of threats and opportunities for stakeholders. Scenarios also differ from science fiction and social utopias, because scenario planners accept more stringent restrictions on creative imagination. Scenarios are inductive models grounded in the discursive multi-model idealizations, because they are created in group discussions with the participation of the experts and stakeholders, who alternate “brainstormings” and critical discussions, and do not expect to generate the best model but to learn from the future.

1 Introduction The aim of the paper is to discuss distinctive features of idealization and ­modeling in future studies and strategic management, widely taught at business schools. It is beyond my competence or the space of this paper to p ­ rovide ­encompassing analysis of the „philosophical foundations“ of the strategic management and future studies. Instead, I will focus on one epistemic practice 1 This research is/was funded by European Social Fund under the Global Grant measure (VP-1 3.1-ŠMM-07-K-01-010). A part of the paper was presented at the European Sociological Association Research Network midterm workshop “Rethinking and Theorizing the Experience of Transformations in Eastern and Central European Societies” in Tartu, Estonia, 22–24 March, 2013. I thank workshop participants for stimulating questions and suggestions. I want also to thank the reviewers for their careful reading and their constructive criticisms, and Vincentas Giedraitis (Ph.D. in Sociology at University of California at Riverside) for improving my English.

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_014

Inductive Modeling and Discursive Idealization

271

of the future representation which is very important or maybe even central in these fields. This practice is alternatively designated as „scenario planning“, „scenario construction“ or „scenario analysis“.2 Although these designations refer to the same referent, they highlight its different features. „Scenario planning“ spotlights the application of the scenarios for the practical goal of planning: the representation is practically useful, helping to guide the action. „Scenario construction“ emphasizes its affinity to what authors writing „science fiction“ are doing: representation of future should be imaginative or creative. At the same time (this is the point of naming these practices of the future representation „scenario analysis“), it should remain „realistic“, differing from mere „science fiction“ or „social utopia“. Both futurists and strategic management analysts do scenario planning. However, future studies with their interest in global trends are more „macro“-focused, while strategic management analysis can be described as „­microfuturists“ or experts in the applied future studies, because they do the work commissioned by specific business corporations, not-for-profit organizations or public agencies. In practice, it is useful for an „applied micro-futurist“ to know futurological literature on mega- and macro- issues, and it is difficult for an adviser in strategic management analysis to establish her reputation in the field without publishing a bestseller on „macrofuturism“, i.e. global trends. In my contribution I will provide an exploratory methodological characteristic of scenario planning using the vocabulary which was suggested by the editors of this volume. Firstly, I will argue that scenario planners are modeling, but do this differently from the „hard science“ modellers. This is the point for my use of the adjective „inductive“ in my desciption of the representations of future in the strategic management and future studies literature as „­inductive models“. However, a reader should be warned in advance do not misread „inductive models“ as the synonym of „statistical models“. Secondly, in building their models, scenario constructors idealize in a distinctive way. I will designate this idealization „discursive idealization“, although I am not insisting that this designation is the best. I will finish my exploratory analysis with the tentative conclusion that discursive idealization may be a case of the multiple-models idealization according to the influential classification by Michael Weisberg (2007). The reader will find these observations on modeling and idealization in scenario planning in the final, fourth section of my paper. However, as far as I cannot assume even the casual knowledge of the scenario planning among 2 See e.g. Chermack (2011), Godet ([1997] 2006), Heijden ([1996] 2005), Lindgren and Bandhold (2009), Ogilvy (2002), Ralston and Wilson (2006), Ringland (2006), Wright and Cairns (2011).

272

Norkus

philosophers, I am starting with a very short historical overview of the emergence and development of scenario planning in the second section. It is continued with the description of the prevailing practices on how scenario planning is done, illustrating them with examples. The reader is welcome to use the references provided for a deeper acquaintance with the field. 2

What Scenario Planning is and Where Did It Come from

Unlike economics, where most of the neoclassical mainstream still endorse its famous definition by Lionel Robbins as „the science which studies human behavior as a relationship between ends and scarce means which have alternative uses“ (Robbins ([1932] 1934, p. 15), there is no universally agreed definition of scenario planning. According to perhaps the most frequently quoted source, scenario planning is “a disciplined methodology for imagining possible futures in which organizational decisions may be played out” (Schoemaker 1995, p. 25). Herman Kahn (1922–1983), the author of the futurological bestsellers „On Thermonuclear War“ (1960), „Thinking about the Unthinkable“ (1962), „On Escalation: Metaphors and Scenarios“ (1965), „The Next 200 Years: a Scenario for America and the World“ (1976; co-authored with William Brown and Leon Kahn), has the reputation to be both the founding father of future studies as the intellectual field in u.s. and selected the word „scenario“ among its several alternatives (e.g. „screenplay“, „script“, „story“) to designate the output of the futurists. Obviously, this is not the same as what filmmakers mean, when they speak about scenarios, because no futurist or decision analyst has the power to make actors to act or play according to his scenario. According to one definition, “scenarios are literally stories about the future that are plausible and based on analysis of the interaction of a number of environmental variables” (Kloss 1999, p. 73). Scenarios look like predictions, but as a matter of fact, they are different from predictions. ‘Scenario building is a narrative forecast that describes potential courses of events and actions’ (Anheier and Katz, 2009, p. 244). A prediction is about finding which scenario is true (in a deterministic s­ etting), or about calculating the relative probabilities of several different scenarios. Scenario constructing is about imagining the future – producing a set of visionary but tenable stories or scripts of the future. To assess the probability of a certain type of event, one must firstly take it into consideration or imagine it. It is repeated experience that events happen or discoveries are made which were beyond the realm of the imagination of even the most well-informed contemporaries. One is invited to remember the failure of thousands of

Inductive Modeling and Discursive Idealization

273

professional Sovietologists in 1986 just to imagine the dissolution of the ussr in only 5 years or the failure of of technology forecasters to predict the emergence of the internet in the early 1990s. On the other hand, people did repeatedly pray for utopian hopes and expectations. A part of Soviet youth, about 1962, believed that they will live under communism in only 20 years, according to the „scientific predictions“ printed in the textbooks of „social science“ they had to study and learn. „Scenario planning attempts to compensate for two common errors in decision making – underprediction and overprediction of change“ (Schoemaker 1995, p. 27). It would be wrong to describe scenario planning as only just the preparatory phase or assistance for the prediction. Conventional decision theory (also known as rational choice theory), which provides the theoretical foundation for the neoclassical microeconomics, conceives decision making as utility maximization. In the decision theoretical models, two fundamental components are distinguished: the predicted outcomes of the actions, and the values assigned to these outcomes. These models tell which actions will maximize benefits and minimize cost given the range of probable outcomes and preferences of the actor. These models may work well in the context of operational management which is concerned with internal issues of an organization such as controlling costs and improving efficiency. In this context, all ­causally ­relevant conditions may be under control, constituting the situation of ­decisions under certainty, or involve only calculable risk, constituting the situation of the decision under risk. The very reason for the emergence of strategic management as a separate branch of organization theory and management studies is the distinction between operational decisions and strategic decisions, involving a long-term perspective. In such a perspective, there may be no possibility for rational calculation of the risk related to the external issues because no statistical data are available. No less importantly, the ultimate goals or criteria for the assessment of the outcomes, i.e. the very identity of the organization and the definition of what „utility“ and „success“ mean may become a problem (Mitchell 2009, pp. 85–104). In the short-term perspective, the extrapolation of the present state or trends may be the most error-safe technique of the prediction, grounding the expected-utility maximizing choices of the organization or the individual decision-maker. However, it may turn out to be just myopic in the long term perspective. Scenario analysis as the strategic management technique helps to maintain open-minded organizational culture which makes a corporation alert to the changes in its environment and to sensitize it to the unpredicted challenges. “Scenario planning is inherently a learning process that ­challenges

274

Norkus

the comfortable conventional wisdoms of the organization by focusing ­attention on how the future may be different from the present” (Thomas 1994, p. 6). This view of scenario planning implies that scenarios themselves are “tools for foresight-discussions and documents whose purpose is not a prediction or a plan, but a change in the mind-set of the people who use them” (De Geus 1997, p. 46). Such scenarios should include new thoughts, which may be painful to contemplate, because they contradict current common assumptions. „When used effectively, scenario planning functions as an organizational ‚radar‘ scanning the environment for signals of potential discontinuities“ (Chermack 2011, p. 10). Even if H. Kahn was the inventor of the concept of scenario planning and helped (as the founder of the famous Hudson institute) to make scenario planning a financially rewarding activity (in the commercial business and public policy consulting), one can argue that he was not the inventor of the scenario planning practices. Arguably, the Prussian general staff of the mid-XIXth century can be credited for the invention of this method. Following the Prussian success in 1866 against Austria, the French, British, Italians, Japanese, Russians and the Austrians themselves all began to make use of wargaming as a training tool. However, the wargaming needs scenarios as much as film making does, although filmmakers allow much less improvisation for actors than organizers of the war games do for generals. Actually, war games are played to „test“ war plans. Since the late XIXth century, the planning for wars and „testing“ these plans by war games (models of real war) were the two main activities of ­general staffs. A scenario provides the framework for a war plan and for its modeling realization in a war game. These plans and scenarios are continuously updated to take into account the experiences of new wars (like the very recent „hybrid war“ in the Ukraine) as well as the lessons of war games, usually played out with real troops. Each full scale military exercise enacts some specific military operation plans, based on a scenario. Of course, general staffs used to concentrate on planning the war against the most probable adversaries, with the „Schlieffen plan“ of the German general staff and its French counterpart (Plan xvii) providing two very famous examples (Hull 2005). However, the obligation of well-working general staffs is „to think unthinkable“ – to work out the operation plans also against the states which may be considered as improbable adversaries at the time when plans are made. To introduce such a plan, one must consider the course of events leading to such event beyond the common imagination. For example, in 1890–1939 the u.s. military considered more than 10 ­scenarios involving their country into war (see Carlson 1998, Gole 2003, Ross 2002). The list of these scenarios included not only war with Germany (plan

Inductive Modeling and Discursive Idealization

275

„Black“) and Japan („Orange“), but also with the United Kingdom („Red“) and Canada („Crimson“), France („Gold“), Mexico („Green“) and other Latin American countries („Purple“, „Violet“). American military strategists choose the names of colours for their military plans to colour-code the scenarios for the multilateral wars, involving the u.s. into the conflicts with coalitions. These were so-called „rainbow scenarios“, with the Black-Orange scenario describing the war of the u.s. with a German-Japan coalition, Red-Crimson-Gold – the war against United Kingdom and Canada allied with France and so on. As a matter of fact, by 1939 u.s. military planners had elaborated only 5 such „rainbow plans“, based on the scenarios of multilateral conflicts. This example should help to grasp how and why scenario planning is different from prediction. If the contemporary u.s. military occasionally has a plan for a war with Mexico, this does not mean that there are people in the contemporary u.s. government who predict such a war. On the other side, how many members of the Russian establishment considered a war with Ukraine as serious possibility some 10 years ago? But if some time in the future historians will find out in the archives that the Russian general staff started to work on the scenarios of such a war immediately after the dissolution of ussr, this will only be the evidence of the high professionalism of this dreadful institution. So scenario planning as (deadly) serious intellectual activity was already practicised by the military already for decades, when it was applied for the more peaceful purposes in the new field of the strategic management of organizations which has emerged in the 1960s. The novelty, which was introduced by H. Kahn in the application of scenario analysis was to refocus from courses of events, leading to some critical uncertainties (e.g. thermonuclear war), to the systematic exploration of their uncertain consequences, producing reports as though people might write them in the future. Thermonuclear war is such a critical uncertainty, because before the actual exchange of nuclear strikes, it is nearly impossible to predict the loss of human life of a specific country. Will American missiles hit the stationing positions of Soviet missiles first or vice versa? How much of the Soviet nuclear force will survive a preemptive American attack or retaliatory strike? This was the reason why so many intellectuals just decried thermonuclear war as terrible and „unthinkable“. H. Kahn did object with the argument that there is huge difference between the thermonuclear war with the total human loss for the u.s. of some 10 million, and another one with the loss of 100 million. Given this difference, the obligation of the responsible government is not just to deter the ussr, but to plan for victory (meaning survival with a minimal loss) and for the postwar reconstruction. For those goals, one needs well-articulated scenarios of alternative futures after a „nuclear Armageddon“, even if one cannot assign to

276

Norkus

them exact numeric relative probabilities.3 Obviously, one faces similar „critical uncertainties“ (even if not so dramatic ones) pondering over the construction of a nuclear plant, liberalizing immigration laws or allowing genetically modified organisms into agriculture. No reasonable discussion about them is possible as far as the alternative scenarios of the possible consequences are not articulated as explicitly as possible. Along with other simulation and forecasting techniques, scenario planning is used in the future studies to explore long-term societal consequences of technological, social, and political change. The famous report „Limits to Growth“ in 1972 (which sold 12 million copies), commissioned by the Club of Rome, remains the most broadly known single contribution to the newly established field of „futurology“ (Meadows et al. 1972). It was grounded in the computer modeling of exponential Malthusian economic and population growth with finite resource supplies, and presented three scenarios. Two of them described the collapse of the global system in the XXIst century, while a third scenario presented a stabilized world. At the same time, scenario planning made its way into the commercial world after it was successfully used in the early 1970s by the Royal Dutch Shell company for generating and evaluating its strategic options. Allegedly, it made this corporation more fit than its competitors to cope with challenges in the aftermath of the „Oil Shock“ in 1973. „Because the oil shock was so devastating to views of a stable future, by the late 1970s the majority of the Fortune 100 corporations had adopted scenario planning in one form or another“ (Chermack 2011, p. 12). 3

Thinking Multiple Futures in the Scenario Planning

By now, future studies and strategic management are huge fields which even the insiders overlook with difficulty. They are differentiated into many subareas, schools and approaches, using many different techniques. Importantly, scenario planning is only one of them, used in conjunction with many ­others. These complementary or alternative techniques with a more or less broad range of applications include: (1) swot analysis, which examines both internal elements of the organization – Strengths and Weaknesses – and external ­elements – Opportunities and Threats; (2) pest (Political, Economic, Social and Technological) analysis, which uses a framework of macro-­environmental 3 This does not mean the assumption that one can imagine the situation where nucleat war can be rational choice, as famous dictum of Carl von Clausewitz “War is the continuation of politics by other means” says about the conventional war.

Inductive Modeling and Discursive Idealization

277

factors to scan the environment of an organization for opportunities and threats; (3) Michael Porter’s „National Diamond“ framework for the analysis of patterns of comparative advantage among industrialized nations; (4) his “Five Forces” framework for industry analysis and business strategy development, and many others.4 After H. Kahn, Peter Drucker and Michael Porter were the most important thinkers in the field under review. Similarly to Kahn, Michael Porter’s work on business strategy took a cue from some of the military planning concepts and applied them to business organizations. His work concentrated on the idea that there can be both unique solutions to strategic problems and general solutions that may be examined for relevance to any strategic situation. chermack 2011, p. 7

A recent list of the approaches to scenario planning includes more than 10 positions, and makes the picture even more confusing.. One finds on this list the Global Business Network, The French School, The Futures Group, Lindgren and Bandhold, Soft creative methods, procedural scenarios, reference scenarios and many more approaches to scenario planning (Amer et al. 2013). However, some differences between these approaches are not substantial, and others may be exaggerated. One can see that some names of „approaches“ refer to commercial „think tanks“ busy to advertise their own trademark version of scenario planning. Their target group of potential customers are commercial corporations looking for how to improve their performance and to enhance their competitive position on the markets of their own. Under tight market competition, business consulting corporations are eager to position their ­services, exaggerating the alleged advantages of their home products, not much differently as pharmacy firms do, selling medicines with identical active substances under different brand names (cp. Giedraitis 2003). I would like to invite the reader who may be interested in more detail about the “schools” or „approaches“ in scenario planning to consult available introductions, textbooks or surveys.5 In the remaining part of this section, I will focus on the shared precepts of how to conduct scenario planning, with special consideration given to those practices where idealization and modeling as features of scenario planning transpire in the most straightforward way. One 4 See Brews and Hunt (1999), Burgelman et al. (2009), Hitt et al. (2011), Dess et al. (2010), Freeman ([1984] 2010), Gamble and Thompson (2009), Kahane (2004), Mintzberg (1994) M ­ orden (2007), Porter 1980; 1985; 1990. 5 See footnotes and references. For recent survey of the field, see Amer et al. 2013.

278

Norkus

of them is the scenario building technique called „future compass“ or „2×2 matrix”. It was famously used by the Global Business Network in its work for the Royal Dutch Shell company in the early 1970s, and therefore can be considered as „classical“. The whole process includes the following steps/operations: 1. Decide on the key questions to be answered by the analysis (outcomes of interest) and the major stakeholders who will be affected and have an interest in the possible outcomes. 2. Set the time frame and the scope of the analysis. Some authors recommend a time frame not less than 5 years into the future, as far as in the short time frame most reasonable inference about the future is the extrapolation of the present. The time frame of 10±5 years may be optimal for scenario planning. The price for the extension of the time frame into the future is the decreasing ratio of the predictable and unpredictable elements of the situation. 3. Identify major forces (factors) which have an impact on the outcomes of interest. Forces (factors) that are considered unimportant should be discarded. 4. Assess each force (factor) on two scales: importance/unimportance and ­certainty/uncertainty (relative unpredictability). While in the search for relevant factors (step (3)), the factors can be designated as „variables“ in accordance with the established usage in the science, specific terminology is used in the next step. Importance means just high causal relevance or a significant contribution to the explanation of the outcomes. Important variables with low uncertainty are called „givens“, while those with high uncertainty are designated as „drivers“. Importantly, one should not confuse „givens“ with „parameters“ in the sense of conventional modeling and prediction. „Parameters“ are important variables which are assumed or expected and do not change their values. „Givens“ in the scenario planning are both parametric variables and „true“ variables, provided the modelers can forecast their values for the time frame of interest. For example, if the age structure of the population and the party affiliation of the u.s. president are both judged as important factors for the changes in the American healthcare system during the next decades, the age structure but not party affiliation of the next two u.s presidents is a „given“. It is not very difficult to deduce what the age structure of the u.s. population will be in 2025 from the information about the present birth rates, mortality rates and other data. However, there is much more uncertainty whether a Democrat or Republican president will be elected in 2016 and 2020. 5. Among all drivers, identify those which have the most impact on the outcomes of interest and which are most uncertain. Consider all remaining ­factors as given.

Inductive Modeling and Discursive Idealization

6.

279

Explore the range of possible variation in the most powerful and most uncertain drivers, and identify their extreme states. This exploration includes the inquiry, whether all states of drivers are compatible with factors considered as givens, limiting the range of their variation. 7. Explore mutual compatibility and interactions between the drivers. 8. Produce initial mini-scenarios which describe the most probable outcomes of the different combinations of drivers. Depending on the number of drivers and their states, this step may lead to seven, nine or more scenarios. 9. Reduce the number of initial scenarios. The minimal target number for reduction are two scenarios, because ending with only one scenario erases the difference between the scenario planning and the conventional single-track forecasting. Besides that, such reduction would annihilate most of the benefits which are claimed for scenario analysis as the powerful means to change the minds, making them open for innovation and change. So the acceptable minimal number of scenarios are three s­cenarios, usually including an „optimistic“, a „pessimistic“, and „­balanced“ or „realistic“ scenario. Customers and stakeholders are the source of pressure for scenario analysts to conclude scenario planning in this way. However, such praxis is not acclaimed by the experts themselves, because stakeholders tend to identify „realistic“ with „most probable“. This interpretation is wrong: “scenario planning aims at plausibility, not probability” (Ruser 2015, p. 278) because scenario planning aims at plausibility, not probability. Scenarios are about the developments produced by the interaction of drivers, and the drivers are unpredictable forces according their definition. In the context of scenario planning, „realistic“ means rather „business as usual“ (now), which may have only a low probability after some 10 years. This misinterpretation can be neutralized using the framework of the „future compass“ technique, which produces four final scenarios, and will be shortly described below. 10. Draft scenarios, narrating a compelling story about what and why will happen because of the interaction of the assumed states of drivers. ­Ideally, the elaboration of the mini-scenario into the full-blooded scenario should be similar to the report written by the people in the future observing the scenario coming true. However, unnecessary details should be avoided, focusing on the mechanisms and the „logic of the situation“ forcing changes to end with the outcomes expected for assumed states of drivers. The story should be internally consistent and represent relatively stable outcome situations in the end. This story should be qualitative, although it may include diagrams and numerical data. However, it must remain followable and persuasive for the readers without mathematical

280

11.

Norkus

or statistical background. A good practice is to designate a scenario with a short name or catchword which is easy to remember and communicates the key insight of the story. A Difficulty to find such a „catchy“ name may indicate the poor quality of a scenario. Identify issues arising from the scenarios for the customer organization. What do the scenarios mean for the future of the organization (or nation state), what are critical threats and opportunities in each of the alternative futures, represented by the final scenarios? These answers are critical for the derivation of the strategies of an organization. In the literature, scenario planning is associated with the „adaptive“ approach in the choice of strategy, which prefers „robust“ strategies, rather than aiming for performance (profit) maximization by gambling on one outcome. Robust „strategies might not have the best possible option available as any one outcome, but their satisfactory outcomes occur in the largest range of future contingencies“ (Mitchell 2009, p. 93). If for some reason an organization prefers optimizing or utility maximizing approach, scenario analysis helps to identify issues which need further research. Such research may help to reduce the uncertainty about the future and to assess different scenarios in terms of their comparative probability. In this case, scenario planning helps the organization know what it needs to know.

The use of the „future compass“ technique for the reduction of mini-scenarios involves identification of two forces which share two features: they are most important and the most uncertain (unpredictable). Ideally, variables describing these forces should have extreme opposite values. In this case, uncertainty means the uncertainty about the direction of the change after some critical conjuncture. Then it is possible to plot two drivers on the grid, shaped by the intersection of the X and Y axis. In this grid, each of the quadrants on the ­coordinate plane represents a specific scenario. This is visualized by Fig. 12.1 which represents „energy descent“ scenarios constructed by the Australian environmental designer, ecological educator and writer David Holmgren, who jointly with Bill Mollison has developed the idea of permanent agriculture (permaculature; see Mollison and Holmgren 1978; Holmgren 2002). Permaculture is the philosophy of ecological engineering and environmental design which seeks to exploit and imitate naturally occurring patterns, and provides a body of insights relevant for agriculture, landscape architecture and ecology. D. Holmgren promotes permaculture as the best response (at the very least, in his native Australia) to coming challenges of peak oil and climate change, which according to broad consensus may be the two most important trends for the economic and political future of the world. I have

Inductive Modeling and Discursive Idealization

281

chosen D. Holmgren’s scenarios as the illustration of the „future compass“ technique, because they are concerned with the issues of general interest and are easy to access on his website.6 While there is scientific consensus since the early 1990s about global warming, there is no agreement concerning the speed of climate change. Opinions are dividing the forecasters who predict rapid climate warming during next decades, and those who consider climate change as a slower secular process (with the possibility of short stops or even backward movements), providing for humanity ample time to innovate and to adapt. In the Fig. 12.1, this critical uncertainty is represented by the Y axis, its upper part standing for the rapid climate change and the lower part representing slow change. X axis stands for another important and uncertain factor – Peak Oil (see e.g. Hubbert 1956, Deffeyes 2001, Aleklett 2012). This is the point of maximum oil production, after which the rate of production will enter terminal decline. As a matter of common knowledge, the world’s natural oil supply is fixed. So the present trend of the growing oil production cannot go on indefinitely. There are two uncertainties about Peak Oil. One of them is about the exact timing, when the annual CLIMATE CHANGE rapid

BROWN TECH (Top Down Constriction)

LIFEBOATS (CivilisationTriage)

OIL PRODUCTION DECLINE slow

rapid

GREEN TECH (Distributed Powerdown)

EARTH STEWARD (Bottom Up Rebuild) slow

Figure 12.1

David Holmgren’s global climate change and energy descent scenarios Source: [http://www.futurescenarios.org/content/view/ 27/46/index.html; 01.30.2015]

6 See . 30.01.2015.

282

Norkus

production of oil will reach its all-time maximum. On this question, pessimists maintain that will be the case during the coming years, while optimists predict its growth for the coming 15–30 years. Another and even more important uncertainty is about the rate of the decline of oil production. One extreme is represented by the forecast of its rapid decline, while the opposite prediction describes slow a decline, providing sufficient time to substitute oil by other energery resources and to develop energy saving technologies. The left part of the X axis represents the possibility of the slow decline of oil production, while the right side stands for its rapid contraction. All four scenarios describe the alternative futures of the energy descent after Peak Oil. This is a transitional phase, when humankind goes from the ascending use of energy that has occurred since the industrial revolution to a descending use of energy. We do not know when (or if) the phase of energy descent will come, and it may happen that you the reader are already living at this time. Despite this uncertainty, there are no obstacles to explore different scenarios of the post-peak oil phase and maybe to prepare global, national and local energy descent action plans, which are advocated by environmentalists. The scenarios should describe the impact of the energy descent on energy and agriculture, settlement and mobility forms, economy and finance, politics, gender relations and culture. This is what David Holmgren actually does in his four energy descent scenarios. Two of them may be qualified as extreme. These are Life Boats or Civilization Triage scenarios. The first scenario is pessimistic, because it paints a social and economic catastrophe on a larger scale than the Black Death in Medieval Europe was. It would be caused by an unhappy coincidence of the rapid decline of high quality fossil fuels supply and the melting of the arctic glaciers during only a few years. This catastrophe includes local wars with the use nuclear weapons (but not World War iii), a halving of the global population in a few decades, abandonment of large urban areas, etc., with a few „oases“ or „life boats“ successfully fencing off themselves against the masses of the displaced people (new nomads). In the long run, the drastic reduction of fossil fuels consumption will stop or even reverse global climate warming. However, because of the lag effects and positive feedbacks in the climate system this cannot be expected to happen within a few years after the collapse. After a longer time, one can expect a rebirth of the civilization, but it will differ from the present civilization no less than medieval Western Christian civilization was different from the Ancient Roman civilization after the „triage“ of the Dark Ages in vi–x centuries. „Triage“ refers to the sorting out of the legacy of the former civilization what is useful for the survival, and discarding what is not.

Inductive Modeling and Discursive Idealization

283

The Green Tech or Distributed Powerdown scenario is the benign extreme case, hoped by optimists. In this scenario, adverse climate and oil supply changes a take long time, allowing the economy and society to successfully adapt to them. This adaptation means the substitution of the renewable energy sources of wind, biomass, solar, hydro, tidal, wave etc. for fossil fuels energy. Nevertheless, this is no happy-go-lucky future, because the scenario forecasts the conflict in the allocation of the production resources between biofuels and food. Farmers, agribusiness corporations, and countries with the underused capacities for agricultural production will emerge as the net winners. The transition to a low carbon, renewable resources economy will reduce greenhouse gas emissions, slowing down or stopping global warming and climate change. Slow energy descent will allow smooth adaptive changes in culture, including curbing the excesses of consumption capitalism, the maturation of feminism and environmentalism etc. The two remaining scenarios describe how economies, societies and cultures are shaped by the interaction of the contrary processes of rapid climate change under slow oil production decline (Brown Tech or Top Down Constriction scenario) and slow climate change under rapid (10% and more annually) oil production decline (Earth Steward or Bottom Up Rebuild scenario). A slow decline of oil production will provide time to develop technologies to get more out of lower grade non-renewable resources (e.g. tar sands, coal and uranium), thereby accelerating climate warming. Strong state action to address mounting natural disasters and other threats resulting from rapid climatic changes will be necessary, including massive centralised investments by the governments into construction of dams to protect coast settlements from the rising sea waters and the relocation of populations. One should expect the rise of resource nationalism, short but intense international conflicts for resources, fragility of the financial system and the increasing control of the state over the allocations of the resources, which will roll back present trends of globalization and liberalization. This is what the alternative designation of this scenario (Top Down Constriction) communicates, referring to the „renaissance“ of the national state burdened with the new emergency responsibilities in the face of contracting energy and food supplies. Quite differently, under slow climate change but rapid contraction of the production of oil there will be few natural disasters and emergency situations, prompting the massive centralized action of the national state and establishing something like a „war economy“ like that in times of wwi and wwii. Under no overwhelming natural disasters, one can expect that the adaptation to the new energy supply conditions will be achieved by the market forces and local

284

Norkus

communities. This is the message of the alternate designation of this scenario (Bottom Up Rebuild). However, a transition to the new equilibrium under this scenario will involve even more dramatic changes in the everyday life and settlement patterns than under Brown Tech, which would be simply a less affluent and more regulated version of high tech urban life. In the Earth Steward scenario, one should expect radical reduction in mass mobility of both people and goods, migration of people out of big cities to smaller towns and villages, ruralization, simplification and localisation of society and life patterns. 4

Scenarios as Models and Idealizations with Adjectives

I will not present D. Holmgren’s scenarios of energy decline in more detail, it is possible to read them at the author’s website.7 Neither will I discuss their substance, because they were presented as a samples or illustrations of what the scenarios are. Now I can proceed to the main questions: (1) are the scenarios models? (2) Do scenario planners idealize? There is broad literature on the epistemological foundations of economics, including significant contributions by the researchers working in association with the Poznań school in the philosophy of science and humanities (e.g. Jones and Cartwright 2005, Hamminga and DeMarchi 1994). Importantly, its seminal ideas go back to the Leszek Nowak’s analysis of the applications of idealization in Das Kapital of Karl Marx, which was a contribution to political economy (Nowak 1980, Borbone 2011). However, one can find in the huge literature on the philosophy of economics and, more broadly, social science, only a few contributions on strategic management and future studies.8 However, in these few publications the reader searches in vain for a discussion of the questions, informed by the contemporary philosophy of science. Instead, arcane questions about the relations of different „paradigms“ or schools of strategic management are discussed, drawing on post-modernist philosophers like Jacques Derrida or Michel Foucault (see e.g. Rasche 2008). A partial explanation for this omission may be an elitist prejudice that future studies and strategic management is too „soft“ or even „parascientific“ as a subject for discussion in mainstream analytical philosophy of science. Most probably, mainstream philosophers are infected by this prejudice because of the credulity toward what is the opinion only of the part of the people whose profession is to speak and write about economy and business. 7 See . 30.01.2015. 8 See Ackoff (1981), Bracker (1980), Bourgeois (1984), Rasche (2008), Tsoukas (2000).

Inductive Modeling and Discursive Idealization

285

Among economicsexperts, one can presently observe the division or even „culture war“ between two sub-communities with different epistemic cultures. One of them has its fortifications in the study programs in economics, finance, and econometrics, with neoclassical economics and mathematical modeling providing the main articles of trade. In business schools which offer education in business management, there is a different intellectual environment which is mainly sceptical about the practical relevance of the mathematical modeling and econometrics for the education of the successful chief executive officers (ceo) of business corporations. „If you’re so smart, why ain’t you rich?“ – this is a famous American question. There is no evidence that graduating in economics or finance makes a significant difference for the chances in becoming rich. Therefore, many programs in business management focus on the case studies of „real business“ practices, which are analysed using the ideas from psychology, sociology, and anthropology in addition to standard microeconomics. Most importantly, they provide practical training of the skills in the creative decision making in high uncertainty situations, using simulations and business games. Strategic management is part of the education of future ceo’s, teaching them the skills of disciplined and organized thinking about the distant future of business corporations and not-for-profit organizations. In practical terms, this means thinking some 5–10 or even more years ahead. Although the ­practical relevance of the education in such thinking is beyond dispute, such a distant future is not accessible for „hard“ science (or social science mimicking such „hard“ science) as a topic for „serious“ prediction. So the field is free for the practitioners in future studies and strategic management, who are busy teaching the skills to represent the future in imaginative and stimulating ways. In terms of publishing output and study programs, these are big industries, as one can get the impression visiting „business“ or „management“ departments in a campus bookstore.9 The situation is not unsimilar to that in philosophical aesthetics, which focuses on the elitist high brow art, neglecting commercially successful popular culture as just „trash“. It is difficult to find in the academic aesthetics ­treatises the discussion of representation practices in popular culture. Future studies and strategic management may be neglected in the analytical philosophy of science for similar reasons, as a marginal and „unserious“ subject. I do not know whether entertainment business workers are „real“ artists. But I would argue that futurists and strategic management analysts are scholars, 9 See e.g. Burgelman et al. (2009), Hitt et al. (2011), Dess et al. (2010), Freeman ([1984] 2010), Gamble and Thompson (2009), Morden (2007).

286

Norkus

even if they are not scientists. As far as the philosophy of science does not limit its interest to the „hard“ (100%) science, but benevolently extends its sphere to include the “humanities”, it should not neglect future studies and strategic management, even if there is difficulty to identify and classify in terms of the established division between hard science and humanities. However, this is only part of the explanation why modeling and idealization practices used in the strategic management and future studies are neglected in the science of philosophy. Futurists and business strategy analysts should blame themselves. While neoclassical economists care a lot about their philosophical clout inserting references to Karl Popper, Thomas Kuhn or Milton Friedman into their work (McCloskey 1990, p. 10), in the broadly read futurological books one will not find references to contemporary analytical philosophy of science (and sometimes even to academic sources). Although practitioners in the future studies and strategic management are divided into many schools engaging each other in the sometimes acrimoniuos debates, they do not ­display much interest in the philosophical analysis or justification of what they are doing. Good book sales and high enrollments seem to provide all the justification they may need. So my answers to the questions inspired by the purpose to contribute to the present collection on the idealization and modeling in science can be only tentative and preliminary, moving in the still unexplored terrain. So are there scenario models? I would give to this question a qualified affirmative answer. Yes, scenarios are just verbal models of possible processes or developments, providing their representations for special purposes. I already explained these purposes in the first part. To repeat, they are “tools for foresight-discussions and documents whose purpose is not a prediction or a plan, but a change in the mind-set of the people who use them” (De Geus 1997, p. 46). The qualification is necessary because there is no universally agreed definition of the „model“ in the philosophy of science. Too strict a definition of a “model,” opposing ­models as abstract structures described by sets of mathematical equations, would disqualify scenarios from the running for a „model“ title. However, influential contemporary authors do not advise to identify „models“ with „mathematical models“. Verbal, graphical, and even material structures can qualify as „­models“ too.10 Inverse conceptual extremity is too permissive use of „models“, which leaves this concept without a fruitful contrasting concept. On this side, there are important reasons to make a distinction between „just so stories“ and „models“. In history writings, there were schools which made their mark, 10

See recent authoritative statements in: Morgan 2012, Morgan and Knuuttila 2012.

Inductive Modeling and Discursive Idealization

287

criticisizing traditional or narrative historiography for the „story telling“ and pleading for the self-conscious model building (see e.g. Brzechczyn 2009; Topolski 1976). However, both my description of the scenario planning practice and its e­ xamples should be sufficient to prove that scenarios are not „just so stories“, which only report a temporal sequence of events which did actually happen. Properly constructed scenarios are grounded in the analysis of the causal interaction of at least two factors. So they are hypothetical complex causal stories or analytical narratives about the future (cp. Bates et al. 1998). In this intermediate sense, a distinctive feature of a model is the representation of a causal structure, where some specific outcome is produced under given initial conditions or inputs(cp. Craver 2006). Scenarios display this feature, because they describe how interacting drivers under specific givens produce a range of outcomes. Many years ago Donald (now Deirdre) McCloskey made the point „that economists are like other human beings in that they both use metaphors and tell stories“ (McCloskey 1990, p. 7). Mary Morgan did elaborate this point in more detail, criticisizing the tendency to treat a narrative like a hypothesis in science. Unlike a hypothesis (including predictions), a „narrative cannot be confirmed or disconfirmed, it can be plausible or implausible, satisfying or not, meaningful or not, insightful or not“ (Morgan 2001, p. 382). This is a wonderful summary of the distinctive features of a good scenario. However, while in mainstream economics, telling stories is a casual or peripheric activity, serving to formulate a question or provide the setting to introduce deductive (top down) models, in the strategic planning production of compelling narratives is a core activity, producing inductive models sui generis, which should not be conflated with the statistical models in econometrics. However, before proceeding to describe their distinctive features, I owe the answer to the second question: do scenario planners idealize? My answer is affirmative – but again with qualifications. Scenario planning involves idealization, but when their construction is completely true to its purpose described above („a change in the mind-set of the people who use them“, so De Geus 1997, p. 46), a specific – discursive – idealization procedure is used. I will not enter into a further question, whether this idealization procedure is unique to scenario planning, or rather one that can find its applications in other intellectual contexts. No matter how broad the scope of the application of this procedure is, the analysis of its uses in strategic management and future studies may contribute to the agenda recently proposed by the Poznan school in the philosophy of science and humanities: to explore the uses of idealization not only in the natural and social science, but also in metaphysics, theology, political theory, history and other humanities (Brzechczyn 2009, Przybysz 2009).

288

Norkus

For this goal, one should expand the influential definition of idealization as an „intentional introduction of distortion into scientific theories“ (Weisberg 2007, p. 639) by assuming the truth of the statements which are known to be false. Given the commitment of the historical profession to Leopold Ranke’s famous description of its mission – to represent the past as it actually was (wie es eigentlich gewesen), one can worry about the intentional introduction of distortions into historical narratives. However, alternative models of the future in the narratives produced by the scenario planners cannot be produced without a „distortion of facts“. Moreover, such distortion is a conscious practice. More specifically, facts are distorted in the scenario planning making in two crucial steps. Firstly, doing (4) and (5), results in the differentiation between the givens and drivers (see former section). Considering selected causally powerful factors as „givens“, strategic planners make an idealizing assumption about the state of their knowledge, assuming they know the values of the relevant variable during all time frame of the scenario. This is strictly speaking not the case, because the difference between the causal factors considered as „givens“ and those explored as „drivers“ is that of degree and not of kind. Without this (first) idealization, scenario planning cannot start, while without the next (second) idealization it cannot be accomplished. This idealization is involved in the step (9), where the number of initial mini-scenarios is reduced to 2–4 scenarios. To make the point using D. Holmgren’s scenarios, many more uncertain and important factors (drivers) in addition to the rate of Global Warming and the speed of the oil production decline will influence the shape of the economy, politics and culture after Peak Oil. However, real causal complexity is self-consciously disregarded by assuming that neglected drivers will not have „considerable“ impact, although as a matter of fact they can preempt the working of the selected drivers. As a matter of principle, there is no logical difference between the reduction of the initial variety of scenarios to the final minimal set and the idealization in the natural scientific models, where conclusions are derived, self-­consciuosly making false assumptions about the irrelevance of other ­factors except those taken into consideration. Galileo Galilei’s famous model of freely falling ­bodies provides a classical example. Most probably, Galilei did arrive to his conclusions not by generalizing from the experiments, but by way of a thought experiment, involving conscious „deformation of reality“, which is distinctive future of i­dealization in comparison with abstraction (cp. Nowak 2000a, Borbone 2011, pp. 241–242). Quite similarly, each scenario can be considered as a „thought experiment“ (Gedankenexperiment). While in real experiments the conclusion is validated by the control of „irrelevant factors“, the plausibility

Inductive Modeling and Discursive Idealization

289

of the scenarios in the scenario planning is grounded in the „control“ of „irrelevant drivers“ by means of idealization. So Alexander Ruser makes a good point, describing scenario planning as the work in the “sociological quasi-labs”, enabling researchers “to make plausible statements about future developments without making prophecies” (Ruser 2015, p. 179). Modeling and idealization in scenario planning are methodologically interesting because of distinctive procedures (modus procedendi) which are used in most of the steps of scenario planning. When they are not used, scenario planning may fail to achieve the main goal – “to change in the mind-set of the people who use them” (De Geus 1997 p. 46).11 As a matter of principle, the scenarios which may claim the status of the „best practices“, are no creation of a single analyst or small team of them. Critically, the participation of the stakeholders with the insider knowledge is crucial for the success of the whole enterprise. In the scenario analysis commissioned by the business corporation, these insiders and stakeholders are the ceos of the corporation. During most steps ((3)-(9)) of the scenario planning, the role of the professional strategic planner (scenario analyst) is limited to that of a discussion group moderator, organizing and directing the group discussions, where experts and stakeholders participate. Minimally, three rounds of group discussion are needed for scenario planning to achieve its goals. First of them is dedicated to find out the factors which are causally relevant for the outcomes in question. This round is organized as a group ideation or brainstroming session, where the participants should feel free to generate unusual ideas (no matter how „wild“ they are). During the brainstorming session, criticism of ideas generated is not allowed. Instead, participants should focus on extending, combining or improving their ideas. Criticism is reserved for the later stage of the process, which may take several sessions. During these group discussions, factors compiled during the „brainstorming“ session are arranged in two orders: from the least important to the most important, and from the most uncertain to the certain. These two sequences may emerge out of the consensus during a discussion, but they may be also decided by voting. In this case, the most important and most uncertain are considered the factors which collect most votes. Voting may be not so useful at the next stage of the work, where drivers are arranged into meaningful patterns. At this stage, „brainstorming“ is needed 11

This may be the flaw in the D. Holmgren’s scenarios of the pending energy descent, because obviously neither I nor you, reader, did participate in the scenario planning which may have generates them.

290

Norkus

again, where participants are asked to „intuit“ these patterns in the complex data. The addition of new factors are allowed, provided they help to organize already selected drivers into meaningful clusters. Each of these clusters provides a mini-scenario, which should be „tested“ by the attempt to elaborate it into a compelling story. The reason why such an attempt may fail is the incompatibility between the drivers or between the drivers and the factors which already are considered as givens. After some 9–10 compelling scenarios are produced, a „critical“ phase follows where the decision should be made about the number of final scenarios. Here counting the votes can help to select the most promising mini-scenarios. Usually, these final mini-scenarios are developed into full-blooded scenarios by the professional scenario planners, who are also orchestrating the whole enterprise. For this goal, they may collect additional information, doing or ordering supplementary research. The scenarios are presented and discussed at the final session where the participant stakeholders should assess the relevance of the multiple “futures” disclosed by scenarios for the organization. This discussion involves identification of the vulnerabilities and opportunities which remained unnoticed before the work on scenarios, discussion and assessment of the possible adaptive actions. The participation of the key stakeholders in the scenario planning makes it a meaningful „learning from the future“ process for ceos of organizations. The whole series of brainstormings and ensuing critical discussions may have the effect on the mindset of the corporation’s ceo, which is similar to the impact which successful Freudian psychoanalysis is reputed to have on the personality of an individual patient. A critical difference is that the „brain“ of the corporation learns from the future, while a Freudian patient learns from her suppressed past. Because idealizations (intentional distortions of the facts) during the scenario planning are the outcome of collective discussions and participatory decision-making, the name of „discursive idealizations“ highlights their specificity. Alternatively, one could use a different adjective, calling them “discursive idealizations”. Although the work on scenarios includes deductive elements on some stages, the whole procedure corresponds almost perfectly to the description of grounded theory method provided in sociological methodology textbooks (Glaser and Strauss 1967, Glaser 1994; 1998, Urquhart 2013). In the standard social research procedure, a researcher selects a theoretical framework, derives hypotheses, and tests them with the data, applying a theoretical model to the phenomenon she studies. Grounded theory method attempts to discover theory through analysis of data. The researcher starts with the collection of mainly qualitative data. For data collection, semi-structured or unstructured interviews and focus group discussions with the participation of

Inductive Modeling and Discursive Idealization

291

key informants or representative stakeholders are recommended as the best methods. In textual data, a series of codes are extracted and used to mark key points. These codes or concepts are used as raw materials to develop categories which are the building blocks of a model or theory.12 One can recognize the procedures involved in the „discovery of theory“ by means of the grounded theory method in scenario planning. Therefore, the scenarios which emerge out of the discursive process may be described as the inductive models of the future. My purpose in using the adjective „inductive“ to specify the particularity of scenario planning models is to point out the similarities between scenario planning in strategic management and grounded sociology, which may unclear to the experts in the philosophy of science who are familiar only with theory and model construction in economics and econometrics. In this environment, „inductive models“ are synonymous with statistical models. The linear regression equation is the elementary example of such statistical or econometric models, with various kinds of „curves“ or „curvilinear“ surface fitting in the n-dimensional data space (where n is the number of independent variables) providing the most sophisticated versions of this epistemic practice. Unlike the inductive models of this kind, inductive models in strategic planning are not models of data, but those of the phenomena themselves, which draw upon the tacit and explicit knowledge of the well-informed insiders who have stakes in these phenomena which are modeled by the analytical narratives produced by the discursive idealization. There are only two remarkable differences between „grounded theories“ and scenarios. (1) In scenario planning, informants and stakeholders not only provide the data, but also participate in processing of these data, negotiating intermediate mini-scenarios and final full scenarios.13 (2) In scenario planning, the final product is not the unique theory or model, but the range of the models representing multiple futures. These models may be assessed in terms of their relative probability. However, this topic is peripheral to scenario planning, because it is fore-mostly the way to organize the thinking about the distant and therefore open future. The distinction mark of the good prediction is truth or high probability, while good scenarios are praised for imaginative and plausible thinking about the future, alerting stakeholders for the threats and 12 13

Grounded theory methodologists do not care much about the distinction between „­model“ and „theory“. However, such participation is explicitly recommended in the variety of the grounded theory methodology, known as „Constructivist Grounded Theory“. See Bryant and Charmaz (2007), Charmaz (2006).

292

Norkus

opportunities which otherwise would have remained unperceived behind the veil of the „official future“. According to Peter Godfrey-Smith, “one actual and reasonable aim of science is to give us accurate descriptions (and other representations) of what reality is like. This project includes giving us accurate representations of aspects of reality that are unobservable“ (Godfrey-Smith 2003, p. 176). This is a good description of the „realist“ view of science, which usually provides the explicit or implicit background for the philosophical discussions about the idealization. If the final goal of science is to present an accurate description of reality, how can one explain and justify its intentional distortion in the idealizations? The most common strategy for this legitimization is to describe the idealizations as a necessary means to isolate key or essential factors or main causes to be explained. After the main causes are described, the idealization should be followed by the concretization or de-idealization, which takes into account secondary factors and special circumstances, modifying the effects of the main causes. In scenario planning, the concretization or de-idealization is not the issue, because the goal of scenario planning is not the prediction of the future. Rather, scenarios describe multiple futures of unknown probability, but with important lessons for the actors in the present. These lessons may motivate stakeholders to act in ways, which may preempt some scenarios from becoming true predictions, or may move the stakeholders to change themselves internally by modifying their individual or collective selves. Because idealizations involved in scenario planning do not contribute to the realist „project giving us accurate representations of aspects of reality that are unobservable“ (Weisberg 2007, p. 639), they can be classified as “multimodel idealizations”. Such idealizations are „not justified by the possibility of de-idealization back to the full representation“ and differ from realist Galilean idealizations „in not expecting a single best model to be generated“ (Weisberg 2007, pp. 645–646). I would like to leave for discussion and further research, whether idealizations used in scenario planning may be explicated and justified using available philosophical approaches to idealization,14 or does their philosophical account provide the challenge to enlarge their supply. Further research, drawing upon more case studies of scenario planning, may find out that different schools or paradigms in strategic management apply many different forms of idealization. However, the exploratory analysis provided in this 14

According to recent inventorization, their list includes neo-Duhemian, neo-Weberian, neo-Leibnizian, neo-Millian, and neo-Hegelian approaches. See Nowak 2000b, pp. 109–110.

Inductive Modeling and Discursive Idealization

293

paper validates the conclusion that it is not Aristotelian abstraction, but the idealization (of whatever kind) which makes the scenario analysis tick. 5 Conclusion u.s. Secretary of Defense Donald Rumsfeld made headlines, following an answer to a question at a u.s. Department of Defense news briefing on February 12, 2002 about the lack of evidence that Saddam Hussein did supply weapons of mass destruction to terrorist groups: „Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know.“ (us Department of Defense, February 12, 2002). Commenting on this citation, Slavoj Žižek made Rumsfeld’s typology complete, pointing out that besides known knowns, known unknowns and unknown unknowns there are also unknown knowns: facts that we intentionally refuse to acknowledge to know (Žižek 2004). Differently from Unlike the past, which is singular (although can be presented in different ways) and determined (the Law of the excluded middle applies to the propositions about the past), futures are multiple and underdetermined. Prediction is about known unknowns. In terms of the Rumsfeldian typology, scenario planning is an intellectual activity in the grey or interface area between known unknowns and unknown unknowns. I did use the workhorse concepts of the Poznań school in the philosophy of social science and humanities „modeling“ and „idealization“ to provide an exploratory epistemological discussion of the scenario planning, where scenario planning is a core activity. This discussion was exploratory, because mainstream analytical ­philosophy of social science focuses on economics and econometrics, which among all social sciences were most successful in establishing their reputation as „­serious“ or „hard“ science. The situation is not unsimilar to that in philosophical ­aesthetics, which focuses on the elitist high brow art, neglecting commercially successful popular culture as just „trash“. While this may be the correct attitude in aesthetics, the neglect of the epistemic practices of scenario ­planning in the philosophy of social science cannot be justified because of their pedigree in military science and the paramount role in contemporary corporation management.

294

Norkus

After tracing the genealogy of scenario planning, I described its modus procedendi, providing the case study of the „future compass“ scenario planning techniques and its application in the construction of D. Holmgren’s global climate change and energy descent scenarios. While the analysis of other approaches to scenario planning may lead to different findings, my case study validates the description of future scenarios as sui generis inductive models, produced by applying idealization procedures of a specific kind. Scenarios as inductive models display important similarities to grounded theories in sociology, emerging from the bottom, from the quantitative and qualitative data, which are the inputs provided by the stakeholders. Differently from the statistical models of the quantitative data in econometrics, which may involve abstraction but not idealization, scenarios are constructed by intentionally distorting reality, displaying this characteristic feature of idealization. The idealizations involved in the scenario planning can be tentatively characterised as discursive multi-model idealizations, because they are created in group discussions with the participation of experts and stakeholders, who alternate „brainstormings“ and critical discussions, and do not expect to generate one best model but to learn from the future. References Ackoff, R.L. (1981). On the Use of Models in Corporate Planning. Strategic Management Journal 2, 353–359. Aleklett, K. (2012). Peeking at Peak Oil. New York: Springer. Amer, M., T.U. Daim, A. Jetter. (2013). A Review of Scenario Planning. Futures 46(1), 23–40. Anheier, H, H. Katz (2009). Introducing Futures Research: Forecasting and Scenarios. In: A. Kumar, J.A. Scholte, M. Kaldor, M. Glasius, H. Seckinelgin, H.K. Anheirer (eds.). Global Civil Society 2009: Poverty and Activism, pp. 238–251. London: Sage. Bates, R.H., A. Greif, M. Levi, J.-L. Rosenthal, and B.R. Weingast (1998). Analytic N ­ arratives. Princeton: Princeton UP. Borbone, G. (2011). A New Idea of Science: The Epistemological Legacy of Leszek Nowak. Epistemologia. An Italian Journal for the Philosophy of Science 34 (2), 233–258. Bourgeois, L.J. (1984). Strategic Management and Determinism. Academy of Management Review, 9, 586–596. Bracker, J. (1980). The Historical Development of the Strategic Management Concept. Academy of Management Review 5, 219–224. Brews, P.J., M.R. Hunt (1999). Planning to Learn and Learning to Plan: Resolving the Planning School/Learning School Debate. Strategic Management Journal 20, 889–913.

Inductive Modeling and Discursive Idealization

295

Bryant, A., K.C. Charmaz, eds. (2007). The SAGE Handbook of Grounded Theory. ­London: Sage. Brzechczyn, K., (ed.) (2009). Idealization XIII: Modeling in History. Poznań Studies in the Philosophy of the Sciences and the Humanities 97. Amsterdam: Rodopi. Burgelman, R.A., C.M. Christensen, S.C. Wheelwright (2009). Strategic Management of Technology and Innovation. 5th ed. Boston: McGraw-Hill Irwin. Carlson, A. (1998). Joint U.S. Army-Navy War Planning on the Eve of the First World War: its Origins and Its Legacy. Carlisle Barracks, Pa.: Strategic Studies Institute. Charmaz, K. (2006). Constructing Grounded Theory: a Practical Guide Through Qualitative Analysis. London: Sage. Chermack, Th.J. (2011). Scenario Planning in Organizations. How to Create, Use, and A ­ ssess Scenarios. San Francisco: Berrett-Koehler Publishers. Craver, C.F. (2006). When Mechanistic Models Explain. Synthese 153 (3), 355–376. Deffeyes, K.S. (2001). Hubbert’s Peak: The Impending World Oil Shortage. Princeton, N.J.: Princeton University. De Geus, A. (1997). The Living Company. Boston: Harvard Business School Press. Dess, G.G., G.T. Lumpkin, A.B. Eisner (2010). Strategic Management: Creating Competitive Advantages. New York: McGraw-Hill Irwin. Freeman, R.E. ([1984] 2010). Strategic Management: a Stakeholder Approach. ­Cambridge: Cambridge UP. Gamble, J.E., A.A. Thompson (2009). Essentials of Strategic Management: the Quest for Competitive Advantage. Boston: McGraw-Hill Irwin. Giedraitis, V. (2003) Selling Madness: Psychopharmaceutical Companies in Lithuania, 1990–2000. Lituanus. Lithuanian Quaterly of Arts and Sciences 49 (1), 6–27. Glaser, B.G., ed. (1994). More Grounded Theory Methodology: a Reader. Mill Valley, CA: Sociology Press. Glaser, B.G. (1998). Doing Grounded Theory: Issues and Discussions. Mill Valley, CA: ­Sociology Press. Glaser, B.G., A. Strauss (1967). The Discovery of Grounded Theory: Strategies for Q ­ ualitative Research. Chicago: Aldine Pub. Co. Godet, M. ([1997] 2006). Creating Futures: Scenario Planning as a Strategic Management Tool. London: Economica. Godfrey-Smith, P. (2003). Theory and Reality. Chicago: Chicago UP. Gole, H.G. (2003). The Road to Rainbow: Army Planning for Global War, 1934–1940. ­Annapolis: Naval Institute Press. Hamminga, B., N.B. De Marchi, eds. (1994). Idealization VI: Idealization in Economics. Poznań Studies in the Philosophy of the Sciences and the Humanities 38. Amsterdam: Rodopi. Heijden van der, K. ([1996] 2005). Scenarios: The Art of Strategic Conversation. ­Hoboken, NJ: John Wiley & Sons.

296

Norkus

Hitt, M.A., R.D. Ireland, R.E. Hoskisson. 2011. Strategic Management: Competitiveness and Globalization: Competitiveness and Globalization: Concepts & Cases. 9th ed. ­Mason, Ohio: South-Western Cengage Learning. Holmgren, D. (2002). Permaculture: Principles & Pathways Beyond Sustainability. ­Hepburn, Vic.: Holmgren Design Services. Hubbert, M.K. (1956) Nuclear Energy and the Fossil Fuels Drilling and Production ­Practice. Shell Development Company, Publication no. 95. Hull, L. (2005). Absolute Destruction. Military Culture and the Practices of War in I­ mperial Germany. Ithaca, NY: Cornell University Press. Jones, M.R., N. Cartwright, eds. (2005). Idealization XII: Correcting the Model. Idealization and Abstraction in the Sciences (Poznań Studies in the Philosophy of the Sciences and the Humanities, vol. 86). Amsterdam: Rodopi. Kahane, A. (2004). Solving Tough Problems. San Francisco: Berrett-Koehler. Kahn, H. (1960). On Thermonuclear War. Princeton, N.J.: Princeton UP. Kahn, H. (1962). Thinking About the Unthinkable. New York: Horizon Press. Kahn, H. (1965). On Escalation: Metaphors and Scenarios. New York, Praeger. Kahn, H., W. Brown, L. Kahn. (1976). The Next 200 Years: a Scenario for America and the World. New York: Morrow. Kloss, L. (1999). The Suitability and Application of Scenario Planning for National ­Professional Associations. Nonprofit Management & Leadership 10 (1), 71–83. Lingren M., H. Bandhold (2003). Scenario Planning: the Link between Future and Strategy. Basingstoke, Hampshire: Palgrave Macmillan. McCloskey, D. (1990). Storytelling in Economics. In: Ch. Nash (ed.) The Uses of the ­Storytelling in the Sciences, Philosophy and Literature, pp. 5–22. London: Routledge. Meadows, D.H., Dennis L. Meadows, J. Randers, W.H. Behrens III. (1972). The Limits to Growth: a Report for the Club of Rome’s Project on the Predicament of Mankind. New York: Universe Books. Mintzberg, H. (1994). The Rise and Fall of Strategic Planning. New York: Prentice Hall. Mitchell, S.D. (2009). Unsimple Truths. Science, Complexity, and Policy. Chicago: The University of Chicago Press. Mollison B., D. Holmgren (1978). Permaculture One: A Perennial Agriculture for Human Settlements. Melbourne: Transworld. Morden, T. (2007). Principles of Strategic Management. Aldershot: Ashgate. Morgan, M. (2001). Models, Stories and the Economic World. Journal of the Economic Methodology 8 (3), 361–384. Morgan, M. (2012). The World in the Model. How Economists Work and Think. Cambridge: Cambridge UP. Morgan M.S., T. Knuttila (2012). Models and Modelling in Economics. In: U. Mäki, D.M. Gabbay, P. Thagard and J. Woods, eds. Philosophy of Economics 13. In: H ­ andbook of the Philosophy of Science, pp. 49–87. Amsterdam: Elsevier.

Inductive Modeling and Discursive Idealization

297

Nowak, L. (1980). The Structure of Idealization: Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht: Reidel. Nowak L. (2000a). Galileo-Newton’s Model of Free Fall. In: I. Nowakowa and L. Nowak, Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities 69, pp. 17–62. Amsterdam: Rodopi. Nowak L. (2000b). The Idealizational Approach to Science: a New Survey. In: I. Nowakowa and L. Nowak (eds.), Idealization X: The Richness of Idealization. Poznań Studies in the Philosophy of the Sciences and the Humanities 69, pp. 109–184. Amsterdam: Rodopi. Ogilvy, J.A. (2002). Creating Better Futures: Scenario Planning as a Tool for a Better T ­ omorrow. New York: Oxford UP. Porter, M.E. (1980). Competitive Strategy. New York: Free Press. Porter, M.E. (1985). Competitive Advantage. New York: Free Press. Porter, M.E. ([1990] 1998). The Competitive Advantage of Nations. New York: Free Press. Przybysz, P. (2009). Modele teoretyczne we współczesnej filozofii politycznej liberalizmu [Theoretical Models in Current Political Philosophy of Liberalism]. Poznań: WN WNS UAM. Rallston, B., I. Wilson, eds. (2006). The Scenario-planning Handbook: A Practitioner’s Guide to Developing and Using Scenarios to Direct Strategy in Today’s Uncertain Times. Mason, Ohio: Thomson South-Western. Rasche, A. (2008). The Paradoxical Foundation of Strategic Management. Heidelberg: Physica-Verlag. Ringland, G. (2006). Scenario Planning: Managing for the Future. Chichester, England: Wiley. Robbins, L. ([1932] 1934). An Essay on the Nature and Significance of Economic Science. London: Macmillan and Co. Ross, S.T. (2002). American War Plans, 1890–1939. London: Frank Cass. Ruser, A. (2015). Sociological Quasi-labs: The Case for Deductive Scenario Development. Current Sociology Monograph 63(2), 170–181. Schoemaker P.J.H. (1995). Scenario Planning: A Tool for Strategic Thinking. Management Review 36 (2), 25–40. Thomas, C.W. (1994). Learning from Imagining the Years Ahead. Planning Review 22 (3), 6–10. Topolski, J. (1976). Methodology of History. Dordrecht: D. Reidel. Tsoukas, H. (2000). Knowledge as Action, Organization as Theory – Reflections on the Organizational Knowledge. Emergence 2, 104–112. Urquhart C. (2013). Grounded Theory for Qualitative Research: a Practical Guide. ­London: Sage.

298

Norkus

US Department of Defense (2002). News Briefing – Secretary Rumsfeld and Gen. M ­ yers, February 2, 2002 11:30 AM [http://www.defense.gov/transcripts/transcript .aspx?transcriptid=2636. accessed 30.01.2015]. Weisberg, M. (2007). Three Kinds of Idealization. The Journal of Philosophy 104 (12), 639–659. Wright G., G. Cairns. (2011). Scenario Thinking: Practical Approaches to the Future. New York: Palgrave Macmillan. Žižek, S. (2004). What Rumsfeld Doesn’t Know That He Knows About Abu Ghraib. , accessed 30.01.2015].

chapter 13

Historical Narration in the Light of the Idealizational Theory of Science: A Recapitulation and Expansion Piotr Szwochert Abstract The purpose of this article is to analyze the historical narration in the terms of ­idealizational theory of science. The author presents hitherto formulated concepts of narration in idealizational approach to science and proposes its own extensions: ­supplementation of the idealizational concept of narration by persuasive and axiological components.

1

The Idealizational Theory of Science: Basic Premises

An analysis of historical narration within the framework of the idealizational theory of science (its) requires the description of some basic premises of this theory (Nowak 1970, 1971, 1980). The author of the its, Leszek Nowak, emphasizes first the difference between essentialism and phenomenalism: Essentialism attaches an objective sense to the differentiation between the essence and the appearance. In the same reality, certain aspects of a given phenomenon are main or principal, while others are secondary or apparent. Phenomenalism stands against this differentiation and considers it an unauthorized imposition of all which is related to man onto the outside world. In the phenomenalist approach, only to us, i.e. people, may certain aspects of phenomena seem significant (essential), while some other may appear insignificant (secondary). nowak 1977, pp. 50–51

The conviction of the essentialist hierarchy of reality (acceptance of the ontological thesis of essentialism) determines the choice of a specific research procedure, which may be structured as follows:

© koninklijke brill nv, leiden, ���6 | doi 10.1163/9789004318847_015

300

Szwochert

a) determination of essential factors for the investigated phenomenon; b) determination of principal and secondary factors within the essential factors; c) introduction of the so-called idealizing assumptions (neglecting the secondary factors) and proposal of the initial hypothesis (idealizational law) determining the dependence of the phenomenon on the principal factor(s) considering the idealizing assumptions; d) waiving the idealizing assumptions (in consideration of particular ­secondary factors) and modification of the initial hypothesis, i.e. approximation of the complete presentation of relationships between the investigated phenomenon and factors essential for it (Nowak 1977, pp. 52–53). The establishment of an essential structure (the identification of the principal and secondary factors for a given phenomenon) and the formulation of the idealizational law is referred to as idealization. As well as the above procedure Nowak also stresses the role of concretization consisting of the gradual removal of particular idealizing assumptions (in consideration of secondary factors) (Nowak 1977, pp. 82–83). It should also be emphasized that researchers rarely possess sufficient knowledge: About the influence of each considered secondary factor […] – it can be assumed that a researcher has information about the effects of certain factors he or she considers secondary. The possessed theoretical knowledge as well as empirical material permit the concretization of only some selected secondary factors. nowak 1977, pp. 134–135

Leszek Nowak indicates, therefore, the need of approximation, i.e. the attribution of certain values to secondary factors, which are only approximate to the actual values adopted by secondary factors.1 In the its, explanation depends upon the presentation of the dependence between the investigated phenomenon and the principal factor (formulated idealizational law) and gradual concretization (or approximation) of the antecedent condition until a factual statement is drawn (Nowak 1977, p. 93). 1 According to Nowak: “An idealizational statement is approximated when it may not be concretized, i.e. when the influence of other factors on the investigated factor can be estimated, but there is not sufficient data about the way they influence it […]. Approximation is, therefore, a substitute for concretization” (Nowak 1977, p. 137).

Historical Narration

301

In his further research Nowak showed that the its is one of several types of deformation procedures used in culture. In his metaphysical concept he distinguishes between soft and hard deformation procedures. The former include negative potentialization and positive potentialization, the latter, reduction and transcendentalization (Nowak 1989, 1990). Following Nowak, let us assume we have an initial object A provided with a certain set of properties characterized by a certain intensity. As a result of the use of positive potentialization, the properties of object A’ are characterized by a greater intensity than the properties of the antecendent object A. In the case of the use of negative potentialization the intensity of properties in object A’ is lower than the intensity in the antecedent object A.2 Transcendentalization provides object A’ with a certain property, which the initial object A does not have. In the case of reduction object A’ is deprived of a certain property possessed by the initial object A (Nowak 1990, p. 197). According to Leszek Nowak: These four deformation procedures are combined into complex procedures characteristic of human culture. For example, science is characterized by idealization, i.e. a combination of reduction and ideation [­extreme negative potentialization – P.S.]. In essence, a material object is both a reduct and an ideal type of a material body. It possesses only some properties of bodies (e.g. it does not have chemical or biological properties), and there are certainly those that can be sensibly attributed to it (e.g. spatial dimensions) are not attributed to it at all. Fictionalization, i.e. a combination of reduction and mythization [extreme positive potentialization – P.S.] seems to be characteristic of art. A caricature of a figure is a reduct of this figure, since certain properties of the antecedent figure are neglected. But the caricature is also a mythization of the figure as the presented properties are extremely accentuated. Mythology, in turn, assumes absolutization, i.e. a combined use of transcendentalization and mythization. Objects in mythology are to be ontologically richer than the surrounding objects, and they also feature “superlative” attributes (the best, the strongest, etc.). nowak 1989, pp. 6–7

2 Nowak also observes that: “an extreme case of negative potentialization is ideation, when the postulated object A’ is attributed a property of object A of zero intensity. Mythization is an extreme case of positive potentialization in which the postulated object A” is attributed a property of object A of the highest intensity (infinite intensity).” (Nowak 1989, p. 6).

302

Szwochert

Assuming the idealizational character of science is tantamount to the treatment of science as a cultural practice (a deformation procedure used in culture). The aforementioned observations can be considered in an analysis of ­historical narration in the light of selected criteria of the idealizational theory of science. 2

Idealizational Theory of Historical Narration

Structure of Historical Narration in the Light of Idealizational Theory of Science Izabela Nowakowa begins her description of historical narration within the its framework by drawing a distinction between two main lines in the theory of truth: Aristotelian and Hegelian (Nowakowa 1990, p. 31). In the former, truth consists of an “exact reproduction of a given state of affairs” (Nowakowa 1990, p. 31).3 In the latter, truth is “a deformation of the state of affairs, as one which allows for the separation of what is essential in it from what is only secondary to it” (Nowakowa 1990, p. 31). The Hegelian line is based on the assumption of an essential structure of reality (aforementioned ontological thesis of essentialism). The Hegelian line of interpretation of truth deals with idealizational statements “whose aim is to neglect what is considered to be secondary and to focus on what is treated as principal for a given phenomenon” (Nowakowa 1990, p. 31). According to Nowakowa, the idealizational formula:

2.1

​U​(x)​ and ​p​ 1​(x)​ = 0 and ​p​ 2​(x)​ = 0 and, …, and  ​p​ k​(x)​ = 0 → F​(x)​ = ​f​ k​(H​(x)​)​ is essentially true when H is a principal factor for phenomenon F, influencing F in an fk-way. It is essentially false when H is only a secondary factor for F, or influences F in a way different than fk on F, or is lacking any influence whatsoever upon F (Nowakowa 1990, p. 32). Furthermore: Depending on the fact of whether all the principal and all the secondary factors are referred to in a given (idealizational) statement, and, in the latter case, of what significance, etc. – one obtains some variants of the essential truth or falsity. nowakowa 1990, p. 32

3 “What is truth? The classic answer to this question is that truth is conformity with reality” (Ajdukiewicz [1949] 2003, p. 19).

Historical Narration

303

In her construct of historical narration Nowakowa assumes the essential interpretation of truth. She distinguishes between two layers of historical narration: the phenomenal and the deep. The phenomenal layer registers the states of the described phenomena as they occur in time. The deep part refers to facts that are treated as essential determinants of these phenomena (Nowakowa 1990, p. 33). In accordance with the its premises one can distinguish between principal and secondary factors. Therefore, according to Nowakowa, the deep layer of narration may and should be correspondingly hierarchized upon the so-called narrative levels. The first level of historical narration describes the state of an investigated phenomenon accounting only for the influence of the principal factor. The second level contains more developed interpretations as it accounts for the principal factor as well as secondary factors of the first order. The subsequent levels of historical narration include even more developed interpretations of a given phenomenon as they also include other secondary factors identified in the deep layer (Nowakowa 1990, pp. 33–34). Thus, in Nowakowa’s view, historical narration is not a set of singular statements but a structure of a higher order (Nowakowa 1990, p. 35). According to her: What seems to be most important for narration is not what it contains but rather what it omits, that is, the criterion of selection of the historical material. And different styles of narration differ mainly as to the selective criteria of which they make use. The Marxist narration of a given period of history of a given society focuses on the transformations of the technology and the class structure […]. In contrast to this, the idealist narration focuses the state and/or the evolution of human consciousness and usually neglects the economic dimension of human history. nowakowa 1990, p. 32

Thus epistemological assessment of concrete historical narrations is not so much concerned with factual statements in the phenomenal layer,4 but rather with the essential truthfulness of the deep layer (reflecting the structure of the assumed idealizational theory) (Nowakowa 1990, p. 36). 2.2 Cascadeness and the Structure of Historical Narration Nowakowa’s views were used in an analysis of historical narration by Krzysztof Brzechczyn, who distinguishes between two types of essential structure of 4 A factual statement, in Nowakowa’s view is true or false (in the classic meaning of the concept of truth) only on its own – it does not determine the epistemological value of entire narrations (Nowakowa 1990, p. 36).

304

Szwochert

investigated phenomena: those dominated by the principal factor and those dominated by a set of secondary factors (Brzechczyn 1998, pp. 95–96; 2009b). In Brzechczyn’s view: In an essential structure dominated by the principal factor, its influence on the phenomenon is greater than the total of influences of secondary factors. In an essential structure dominated by secondary factors their combined influence is greater than the influence exerted by the principal factor, although the influence of the latter is – by definition of essential structure – greater than the influence of each individual secondary factor. brzechczyn 1998, p. 96

Brzechczyn also notices that the essential structure dominated by secondary factors is characteristic of social phenomena (Brzechczyn 1998, p. 97). Moreover, in Brzechczyn’s view, there may occurred be an effect of cascadeness in this type of essential structure. It is often that: an initial essential structure dominated by the principal factor may – due to the presence of an increasing number of new secondary factors – be transformed into an essential structure dominated by secondary factors. A reverse cascadeness effect, i.e. a decline of influence of certain secondary factors, may restore the dominance of the principal factor. When cascadeness occurs the first type of essential structure is transformed into the second type of essential structure in which the total influence of secondary factors is greater than the influence of the principal factor. brzechczyn 1998, p. 99

The secondary factors in the process of cascadeness may influence the investigated phenomenon separately or jointly by interacting with one another. ­Following Russell L. Ackoff,5 Brzechczyn notes that an interaction between factors A and B in the essential structure of phenomenon F takes place when the total influence of A and B on F is not equal to the total of influences exerted

5 Following Russell L. Ackoff, two variables interact with each other if the influence of one of them on a phenomenon depends on the value the other variable attains (Brzechczyn 1998, p. 101).

Historical Narration

305

separately by each factor on F (Brzechczyn 1998, p. 102). For example, a weakening interaction between A and B on F occurs when factor A influenced by B (or vice versa) attains a value at which the influence of A (or B) on F decreases. A strengthening interaction between A and B on F takes place when factor A influenced by B (or vice versa) attains a value at which the influence of A (or B) on F increases (Brzechczyn 1998, p. 102). The effect of cascadeness also involves a different construction of the essential structure (indicating the principal and secondary factors). According to Brzechczyn, a modeling sequence characteristic of the its (initial consideration of the principal factor only, followed by a gradual accumulation of secondary factors) becomes reversed when cascadeness takes place: The primary model describes the cascading effects of secondary factors. The effect of the principal factor is described by the secondary model. The researcher includes in the primary model all the factors of the cascade because their total influence is greater than the influence of the principal factor on the investigated phenomenon. Thus the cascadeness model is more realistic than the standard essential structure model as it contains more factors. brzechczyn 1998, p. 103

Also historical narration describing phenomena influenced by a cascade of secondary factors involves, according to Brzechczyn, a certain peculiarity. Due to the effect of cascadeness the first level of narration describes the state of a given phenomenon in consideration of the influence of secondary factors. It is the consecutive level of narration which accounts for the influence of the principal factor (Brzechczyn 1998, p. 105). Moreover, Brzechczyn observes that: When factors interact with one another in a cascade, the structure of narration becomes more complex. Each strip of narration consists of two levels. The first level describes the influence of interacting factors on the investigated phenomenon. The second level involves interactions between factors in terms of their range and type (weakening or strengthening interaction). brzechczyn 1998, p. 105

Within the its framework, historical narration appears to be a complex, ­multi-level structure, which reflects – more or less accurately – a complex (­essentially hierarchical) social reality.

306

Szwochert

Narrativism in the Light of the Idealizational Theory of Science Suplemented by the Selected Premises of Unitarian Metaphysics According to Brzechczyn, the its framework broadened with selected categories of Unitarian Metaphysics6 also allows for the paraphrasing of the controversy between positivism and narrativism regarding the place of history in culture.7 By referring to the concepts of negative potentialization, positive potentialization, reduction and transcendentalization, the fundamental question of whether history is closer to science or literature can be posed as: “What deformation procedures find primary application in history: idealization or fictionalization?” (Brzechczyn 2009a, p. 16). Brzechczyn’s attribution of primary importance to idealization (combination of reduction and negative potentialization) indicates the scientific character of history. In this view the “fictionality” of historical narration as seen by narrativists – and mistaken by them for literary “fictionalization” – is of idealizational character, i.e. it results from historians’ use of the scientific idealizational procedure. According to Brzechczyn:

2.3

Idealization resembles White’s act of prefiguration. In order to neglect certain dimensions of the phenomenon, one needs courage and imagination in escaping the load of details, which weigh heavily on the theory’s construction. Nobody knows why White attributes this courage and imagination to poets, and denies them to scientists. brzechczyn 2009a, p. 16

The application of a broadened its framework also allows for the paraphrasing of H. White’s historical narration paradoxes of the historical record as 6 See the aforementioned two types of deformation procedures used in culture: soft (negative and positive potentializations) and hard (reduction and transcedentalization); (Brzechczyn 2009a, pp. 14–15). 7 In the positivist approach, whose main proponent is Carl G. Hempel, history is a science like all other sciences, including natural sciences. This means it uses the same models of explanation (deductive-nomological model), formulation and confirmation of scientific laws as, e.g., physics. In the narrativist view, chiefly represented by Hayden White (1973) history is closer to literature than to science. According to White, the historian – in the deepest level of his a­ ctivity – performs a “poetic” prefiguration (creation) of history research using one of four p ­ oetic tropes: metaphor, metonymy, synecdoche or irony. This pre-figured c­ reation is then subject to other strategies aimed at organizing the past. They include strategies of emplotment, argument and ideology. Each strategy falls into four characteristic types: “­emplotment into romance, tragedy, comedy, satire; argumentation into formism, mechanicism, ­organicism and contextualism; and […] ideology into anarchism, conservatism, ­liberalism, radicalism” (Domańska 1992, pp. 32–34).

Historical Narration

307

­being both too full and too sparse, which purportedly proves the close relationship between history and literature (Brzechczyn 2009a, p. 19). Following the paradox of the historical record being too full: there are always more facts in the record than the historian can possibly include in his narrative representation of a given segment of the historical process. And so the historian must “interpret” his data by excluding certain facts from his account as irrelevant to his narrative purpose. white 1985, p. 51

According to White, the historian’s interpretation, i.e. exclusion of irrelevant facts, is made through the use of poetic tropes (metaphor, metonymy, synecdoche, irony). In the its it is made by creating an essential structure, i.e. identifying principal and secondary factors for a given phenomenon and by excluding insignificant data (Brzechczyn 2009a, p. 19). Following the paradox of the historical record being too sparse: In his efforts to reconstruct “what happened” in any given period of history, the historian inevitably must include in his narrative an account of some events or a complex of events for which the facts that would permit a plausible explanation of its occurrence are lacking. And this means that the historian must “interpret” his materials by filling in the gaps in his information on inferential or speculative grounds. white 1985, p. 51

The “filling in of the gap” is the idealizational theory of science adopted by the historian. “The scientific theory in historical research plays not just an explanatory but also a heuristic role” (Brzechczyn 2009a, p. 25). Both paradoxes of the historical record are evidence of not so much the affinity between history and literature, but they also confirm the assertion that history is of a scientific character (in the its framework). It should be emphasized that indicating the scientific character of historiography is not synonymous with the positivist approach to historiography.8 8 The its and positivism are based on different ontological assumptions, i.e. on essentialism (its) and phenomenalism (positivism). The positivist approach is closer to the classic concept of truth, while the representatives of the Poznań methodological school adhere rather to the Hegelian interpretation of truth. It should be stressed, however, that positivism and its are both naturalist approaches. For a comparison between its and positivism see: (Nowak 1973, pp. 277–302).

308

Szwochert

Idealization in science, in the framework of Leszek Nowak’s Unitarian Metaphysics, is regarded as a deformation procedure used in culture. Following Brzechczyn, history attains scientific status, when science itself is regarded as an area of culture. The positivist approach (as noted by Wojciech Wrzosek): “highlights differences and blurs similarities between science and culture. It is because the scientific reflection in its modern form was developed in opposition to those areas of culture it had previously been associated and identified with (religion, theology, art, ideology)” (Wrzosek 2010, p. 19).9 3

An Attempt at Extension of the Idealizational Theory of Historical Narration

Both Nowakowa and Brzechczyn stress that the presented theory of historical narration is based on a number of idealized assumptions. It stipulates that, e.g., “If a factor A influences factor B, then the factor B does not influence the factor A” (the so-called one-sidedness of influence) (Brzechczyn 2009a, p. 19).10 This results in a passive understanding of the investigated factor. The researchers regards the factor in the theory as an object not a subject of influence. Jan Pomorski notices, however, that in empirical sciences, beside explaining how a given factor depends on factors essential for it, one tries to reveal how a distinguished factor affects sets of factors and is essential for them (Pomorski 1981, pp. 62–63). Thus Pomorski replaces the assumption of one-sidedness of influence with an assumption that: “every distinguished factor F ontically generates a set of factors essential for F and a set of factors for which F is an essential factor. These sets of factors are essentially ordered sequences in terms of their strength of influence on F and by F” (Pomorski 1981, p. 63). According to Pomorski, between a set of factors essential for F and a set of factors upon which F exerts its influence, relations of identity, subordination, superordination, intersection and exclusion may occur (Pomorski 1981, pp. 62–63). The case in which sets of factors essential for F and for which F is essential are mutually exclusive is regarded by Pomorski as characteristic of historiographic practice. This situation allows Pomorski to waive another its assumption which states that: “factor F is subordinate to the essential impact of solely directly essential factors” Pomorski 9 10

The science/culture opposition is also close to the narrativist standpoint, see, e.g., Iggers (1990; 1997), Domańska (1994, pp. 58–59) and K. Brzechczyn (2009a, p. 45; 2014). I. Nowakowa defines the one-sidedness of influence: “the considered properties are, at most, unilaterally, essential one for another, that is, it is presupposed that if A is essential for B, then B is not essential for A” (Nowakowa 1990, p. 37).

309

Historical Narration

1981, p. 76). Waiving this assumption allows for the introduction of the notion of an indirectly essential factor. An indirectly essential factor for the F factor is a factor which influences F through the mediation of another factor. Depending upon the number of factors-mediators present in this chain of influences, we can distinguish the first, second, etc, levels of indirectness. A zero level of indirectness is the level of direct essential influence (Pomorski 1981, p. 76). After imposing the relation of temporal ordering upon the relation of influence, an exemplary chain of influences between factors A, F and B can be: ​A​ ​t​  ​​ → ​F​ ​t​  ​​ → ​B​ ​t​  ​​ m

n

o

where: → – relation of influence; t – temporal variable; n, m, o – natural ­numbers m ≤ n ≤ o (Pomorski 1981, p. 67). In this formula factor F is a principal factor for factor B, and A is a principal factor for F. Factor A is then a factor indirectly essential for B, and factor F is the directly essential factor. Pomorski’s assertions were used by K. Brzechczyn, who provided the idealizational concept of historical narration with a more realistic form. Following his proposals, a simplified structure of historical narration for the chain of ­influence of factors A and F on factor B would be as follows: Strips of narration

Phenomenal layer

Deep layer

1st strip

B (a, to) = b F (a, tn) = f

F (a, tn) = f A (a, tm) = a

After waiving the assumption about the one-sidedness of influence, the statement which describes the state of factor F in time tn appears both in the phenomenal and deep layers of narration. The statement which describes the state of factor B in time to is present only in the phenomenal layer; the statement describing the state of factor A in time tm is only in the deep layer of narration. The waiving of the one-sidedness assumption leads to a partial cancellation of the division of narration into the phenomenal and deep layers. As Brzechczyn observes, “certain statements belong both to the phenomenal and deep layers of narration, some belong only to the phenomenal layer, and some only to the deep layer” (Brzechczyn 2009a, pp. 21–23). The idealizational theory of historical narration also involves simplifications which are not idealizing assumptions introduced consciously by the theory’s

310

Szwochert

authors. In our opinion, they result from the excessive removal of Nowakowa’s and Brzechczyn’s considerations from historians’ research practice. As Zenonas Norkus notes: neither Brzechczyn nor Nowakowa provides even a single example of the application of this method [interpretation within the its framework – P.S.] to real historical work. Without such “proof by doing”, Brzechczyn’s claim about the ability of its to mediate in the debate between the positivist and narrativist approaches to history remain unsubstantiated. norkus 2012, p. 303

An analysis of historical works shows that one must differentiate between at least two components of the phenomenal layer of historical narration: factographic and persuasive. In a given narration both components are intermingled and they can be separated from the text more or less successfully. The factographic layer refers directly to statements reporting the state of affairs of examined phenomena. The persuasive one deals with the ways in which these statements are presented. One can therefore assert that if the content of narration, i.e. what is considered and what is omitted, depends on a theory, the way the selected material is presented depends – to some extent – on the used artistic means (poetic tropes, figures of speech, rhetorical devices, etc.). According to Jerzy Topolski, the narrative practice of historians involves several types of rhetorical frameworks which determine “the general atmosphere of narration”. They are divided into approving, disapproving, apologetic, ironic and quasi-impartial rhetorical frameworks (Topolski 2005, p. 120). According to him the approving or disapproving rhetorical frameworks: require the historian to take sides in the drama of the past […]. This could be, for example, a description of revolutionary events from the standpoint of revolutionaries, or from the standpoint of those (e.g. a royal court) against whom the revolutionary events were directed. topolski 2005, pp. 120–121

The apologetic framework features rarefied nomenclature, e.g. the name Great October Socialist Revolution for the Russian Revolution of 1917. The opposite of the apologetic framework is the ironic framework. It can take the form of heavy irony but also a form of “patting on the shoulder” of described historical figures (Topolski 2005, p. 121). In general, however, historians: want to be perceived as impartial and they are inclined to choose a rhetoric framework I call quasi-impartial […]. Complete impartiality is

Historical Narration

311

c­ ertainly unattainable […] one can, however, avoid apologetic, ironic or clearly approving or disapproving rhetorical frameworks by pursuing a “placid” kind of narration. topolski 2005, pp. 121–122

In our opinion, one may also add to Topolski’s set of rhetorical strategies the so-called judging narration framework which places the historian in the role of a fair judge assessing the actors of past events “in the name of History”.11 Obviously, a given historical narration may combine the aforementioned frameworks. For example, in his book “Agresja 17 września 1939” (Aggression: September 17, 1939) Jerzy Łojek makes extensive use of a judging rhetorical framework by criticizing the desistance of the Polish authorities (President, government, military high command) in the areas of domestic and foreign policy, following the Soviet invasion of Poland on September 17, 1939, as well as before the outbreak of World War ii (Łojek 1990). In his emphasis of the fatal outcomes of the inactivity of the Polish military high command in the face of the Soviet invasion (best illustrated by the issuing of the general directive of the Polish Commander-in-Chief), Łojek describes the following situation: At 12.45 another message was received at Zbaraż. A local military commander of the Polish forces, General Mieczysław Smorawiński, forbade his troops to offer resistance to the Soviet army. We know today he did it on his own responsibility before receiving any respective orders from the Polish High Command. Very soon History will add a punchline to this story: in 1943 General Smorawiński’s body will be exhumed from the mass grave in Katyń. łojek 1990, p. 101

In another part of his book Łojek is partial to irony in commenting the transcript of a phone message from September 17, 1939 (8.30 a.m.) found in the military records of the Polish High Command: The transcript says: “Skała is taken. The barracks in Podwołoczyska under attack. The bridge on the Zbrucz blown up. Riding in their tanks through Borszczów they [Soviet troops] proclaim to the locals: We come to defend your homes. They open fire only when they meet resistance from our troops…” Even taking into account the tragic chain of circumstances and 11

The inclusion of the judging narration framework was inspired by Karol Modzelewski’s division of models of historiography into: traditional (judging), positivist (scientific) and inspired by anthropology; Modzelewski (2009).

312

Szwochert

misunderstandings on September 17 this report deserves a highly critical comment. Were they to open fire if there was not any resistance? If on September 1st the Polish troops had not fired upon the German troops crossing the western border, the Wehrmacht tanks would have been soon in Warsaw without firing a single shot, and their crews would be sitting comfortably in the tanks with their hatches open. The attacker is always fired upon by the attacked. Attaining the objective without bloodshed is always in the aggressor’s interest. It is astonishing they did not understand it in Kołomyja on September 17th. łojek 1990, p. 82

Another oversimplification, which in our opinion should be corrected, is the assertion that: “In the case of narration of historical phenomena, the criterion of selecting historical material is supplied by a scientific theory which provides a hierarchy of factors influencing a given phenomenon” (Brzechczyn 2009a, p. 19).12 In our opinion, the form of narration (what it includes and what it omits) is also determined by axiology, i.e. values respected by the historian as a member of a given society or nation.13 This modified narration can be ­presented as follows: The figure 13.1 shows that the theoretical and axiological parts of the deep layer of historical narration interact with each other. Depending on

Axiology

Source material

Historical narration

Theory

Figure 13.1 12

13

→ one-sided influence, ↔ mutual influence

According to Nowakowa this assumption is: “the deep part [of historical narration – P.S.] refers to the facts that are treated as essential determinants of [investigated – P.S.] ­phenomena. The deep part of narration is a consequence of the assumed – openly or tacitly – idealizational theory of the described phenomena” (Nowakowa 1990, p. 33). This axiology is different from the axiology of values accepted by the historian as a member of the community of professional scholars. We are also fully aware that each theory (model) assumes a certain ontology, ideals of science, vision of the world and man, etc.

Historical Narration

313

the dominance of one of the two parts, one can distinguish between theoretical and axiological narration. Theoretical narration is characterized by ­materialistic holism, i.e. considering a historical reality as a “whole” (societies, socio-economic systems, etc.), and accounts for principal material factors, e.g. technological level of economy, natural environment or biological needs of individuals. Axiological narration features idealism accounting for principal ideal factors individually (individual convictions, traits of character, etc.) and holistically (ideology, religious beliefs, etc.). Examples of theoretical narration can be found in works within the paradigm of New Economic History. According to Jan Pomorski, the proponents of New Economic History (e.g. Robert W. Fogel, Douglass C. North): represent a high level of methodological self-awareness, which is a ­distinctive characteristic of this “school of thought” in comparison to other modern approaches to historiography. The majority of the works of cliometrists – as they call themselves – are founded on explicit methodological assumptions. They also deal with issues of general methodology in works presenting the novel character of their own paradigm. pomorski 1995, pp. 12–13

Jan Pomorski enumerates five basic components of methodological aware­ ness of new economic historians: (1) conceptualization (“specifications”) of subject matter; (2) explicite reference to the theory; (3) precise formulation of hypothesis allowing for its verification; (4) quantification of historical data; and (5) refutation of commonly accepted arguments (Pomorski 1995, p. 67).14 14

Robert W. Fogel, author of the classic work in cliometrics “Railroads and American Economic Growth: Essays in Econometric History” lists six characteristics that make “scientific history” (cliometrics) distinct from “traditional history”: 1. subject matter – “scientific” historians focus on the study of collectivities and recurring events, while the “traditional” historians deal with concrete events and individuals; 2. preferred types of evidence – “­traditional” historians prefer literary sources, i.e. qualitative sources, while “scientific” historians favor quantitative sources; 3. standards of proof and verification – “traditional” historians seek reliable witnesses and try to prove they knew what happened, while “­scientific” historians formulate all explicite empirical assumptions in the initial hypothesis and, on the basis of source material, attempt to accept or disprove these assumptions; 4. the role of controversy – in “scientific history” discussions and polemics are considered to be significant factors of knowledge development, while in “traditional history” all criticism is understood as criticism ad hominem (undermining the researcher’s credibility); 5. attitudes toward collaboration – “traditional” historians avoid team research, in the case of “scientific” historians the types of sources and the diversity of applied research ­methods

314

Szwochert

Most cliometric works contain these components.15 For example, a paper “Changes in American and British Stature since the Mid-Eighteenth Century” focuses on the impact of long-term changes in the levels of health and nutrition of a population on economic growth (Floud et al. 1982).16 The proposed index of health and nutrition is the stature of individual members of the investigated population. Therefore, the authors of the paper, as well as their application of economic theories also refer to works in the areas of anthropology and psychology (Floud et al. 1982, pp. 4–8). They give credence to the hypothesis that health and nutrition improvement was of primary importance to the economic growth of the American society from the mid-18th century to the beginning of the 20th century. They dismiss the common practice of economists who treated health and nutrition merely as background influences, but not as principal determinants of long-term economic growth (Floud et al. 1982, p. 1). Also characteristic is the authors’ approach to historical sources. They do not cite reports from individual witnesses but focus on quantitative sources, e.g. military recruitment records from the American War of Independence (1775– 1783), and use them in statistical analysis (Floud et al. 1982, pp. 10–11 and 14). An example of axiological historical narration can be the aforementioned “Agresja 17 września 1939” by J. Łojek. This book, originally published underground in communist Poland, deals with events “passed over and consciously omitted from the national awareness” related to the Soviet invasion of Poland in September 1939. Łojek regarded his description of these events as a “scholarly, moral and patriotic duty” (Łojek 1990, pp. 7 and 10). In Łojek’s view the causes of the 1939 political disaster17 can be found in the weaknesses of character of the Polish authorities (he therefore represents an

15 16 17

make team research a necessity; 6. communication with the history-reading public – “­traditional” historians write for the general public, whereas the works of “scientific” ­historians are aimed at a small group of professionals; (Fogel 1982, pp. 34–49). As confirmed by Pomorski’s analysis of 80 historical works and 48 theoretical and methodological works by new economic historians; see (Pomorski 1995, pp. 102–128). This work was not studied by Jan Pomorski in his book. The political disaster of 1939 should not be identified with the Polish military failure resulting from the “objective” differences in the military potentials of Poland, the Third Reich and the Soviet Union. As J. Łojek observes, “Poland could not have been saved in September 1939, but the Polish government elite could have taken a proper political course of action and designed political declarations, propaganda actions, appeals to the public opinion worldwide, prepared a contingency plan of armed resistance on the eastern border and issued instructions to the general public about their civic duties in a situation of concerted aggression and occupation by both Nazi Germany and the Soviet Union” (Łojek 1990, p. 18).

315

Historical Narration

idealistic-individualist approach). In Łojek’s opinion, the Polish Commanderin-Chief Marshal Edward Rydz-Śmigły, had no sufficient authority (nor character traits) to make any decisions regarding the fate of the Polish state: “Rydz was no Józef Piłsudski. He lacked the political charisma of the First Marshal and was no match for the Commandant in moral authority or political talents” (Łojek 1990, p. 54). Rydz’s dominant position resulted from the impotence of the other centers of state authority, in particular, of the President of the Republic of Poland “the crummy Ignacy Mościcki who was taken aback by the unfolding national tragedy a few months before his anticipated retirement” (Łojek 1990, p. 54). Such a description of representatives of the Polish authorities allows Łojek to question their decisions and criticize their negligence. ­Furthermore, Łojek also proposed his own solutions (e.g. a draft statement of the President and the Government of Poland on the Soviet aggression) alternative to those by the Polish government of the day (Łojek 1990, pp. 97–98). The role of sources in the structure of historical narration must be also ­taken into consideration. On the one hand, a theory or a set of axiological ­values accepted by historians determine their selection of data from source material for a given historical narration. On the other hand, source data may falsify ­explanations proposed by the historian or force the historian to modify them.18 Therefore, our final model of historical narration should be presented in the figure 13.2.

Axiology

Source material

Historical narration

Theory

Figure 13.2 18

Where → one-sided influence; ↔ two-sided influence

It should be noted that the character of historical sources often determines the way of their analysis (or organization). Francois Furet (1971) distinguishes, for example, between quantitative and qualitative historical sources. He divides the former into structurally numerical sources grouped together as such, structurally numerical sources used by the historian substitutively to find the answers to questions completely outside the original field of investigation, and non-structurally numerical sources that the historian can use quantitatively, by a process involving two substitutions.

316

Szwochert

In conclusion, the idealizational theory of historical narration developed by I. Nowakowa and K. Brzechczyn should be supplemented with the persuasive component of the phenomenal layer of narration (and retain – to some extent – the literary dimension of history) and with the axiological component of the deep layer of historical narration thus bringing the its closer to the actual ­narrative practice of historians. References Ajdukiewicz, K. ([1949] 2003). Zagadnienia i kierunki filozofii. Teoria poznania. M ­ etafizyka [Problems and Theories of Philosophy. Epistemology. Metaphysics]. Kęty-Warszawa: Wydawnictwo Antyk- Fundacja Aletheia. Brzechczyn, K. (1998). Odrębność historyczna Europy Środkowej. Studium metodologiczne [Historical Distinctiveness of Central Europe. A Methodological Study]. Poznań: Wydawnictwo Fundacji Humaniora. Brzechczyn, K. (2009a). Between Science and Literature: The Debate on The Status of History. In: idem (ed.), Idealization XIII: Modeling in History. Poznań Studies in the Philosophy of the Sciences and the Humanities 97, pp. 7–30. Amsterdam/New York, NY: Rodopi. Brzechczyn, K. (2009b). Methodological Peculiarities of History in the light of Idealizational Theory of Science. In: idem (ed.), Idealization XIII: Modeling in History. Poznań Studies in the Philosophy of the Sciences and the Humanities 97, pp. 137–157. Amsterdam/New York, NY: Rodopi. Brzechczyn, K. (2014). Between Positivism, Narrativism and Idealisation in Polish Methodology of History, Historein. A Review of the Past and Other Stories 14, no. 1, 75–83. Domańska, E. (1992). Metafora–Mit–Mimesis (Refleksje wokół koncepcji narracji historycznej Haydena White’a) [Metaphor – Myth – Mimesis: Reflection on Hayden White’s Concept of Historical Narration]. Historyka. Studia Metodogiczne 22, 29–44. Domańska, E. (1994). Kryzys tradycyjnego rozumienia historii w filozofii anglosaskiej [The Crisis in the Traditional Comprehension of History in Anglo-Saxon Philosophy]. Historyka. Studia Metodogiczne 24, 57–65. Floud, R., R. Fogel, G. Friedman, S. Engerman, K. Sokoloff, R. Steckel, T. Trussell, G. Villaflor and K. Wachter (1982). Changes in American and British Stature Since the Mid-Eighteenth Century: A Preliminary Report on the Usefulness of Data on Height for the Analysis Of Secular Trends in Nutrition, Labor Productivity, and ­Labor ­Welfare. National Bureau of Economic Research Working Paper Series Working Paper No. 890, pp. 1–64.

Historical Narration

317

Fogel, R. (1982). “Scientific” History and Traditional History. In: L.J. Cohen, J. Łoś, H. Pfeiffer and K.P. Podewski (eds.), Logic, Methodology and Philosophy of Science, vol. VI, pp. 15–61. Warszawa–Amsterdam–New York–Oxford: PWN. Furet, F. (1971) Quantative History, Daedalus 100(1), 151–167. Iggers, G. (1990). Historical Studies in the 1980’s. Some Observations. In: J. Topolski (ed.), Narration and Explanation. Contributions to the Methodology of the Historical Research. Poznań Studies in the Philosophy of the Science and the Humanities, vol. 19, pp. 11–30. Amsterdam/Atlanta: Rodopi. Iggers, G. (1997). Historiography in the Twentieth Century. From Scientific Objectivity to the Postmodern Challenge. Hanover & London: Wesleyan University Press. Łojek, J. (1990). Agresja 17 września 1939. Studium aspektów politycznych [Aggression of September 17, 1939. The Study of The Political Aspects]. Warszawa: Instytut Wydawniczy PAX. Modzelewski, K. (2009). Trzy modele historiografii [Three Models of Historiography]. Nauka 2, 15–21. Norkus, Z. (2012). Modeling in Historical Research Practice and Methodology: Contributions from Poland. History and Theory 52, 292–304. Nowak, L. (1970). O zasadzie abstrakcji i stopniowej konkretyzacji [On the Principle of Abstraction and Gradual Concretization]. In: J. Topolski (ed.), Zalożenia metodologiczne “Kapitału” Marksa, pp. 123–218. Warszawa: KiW. Nowak, L. (1971). U podstaw Marksowskiej metodologii nauki [Foundations of Marxian Methodology of Science]. Warszaw: PWN. Nowak, L. (1973). Pozytywistyczna koncepcja praw i wyjaśniania [Positivist Concept of Laws and Explanation]. In: J. Kmita (ed.), Elementy marksistowskiej metodologii humanistyki, pp. 277–302. Poznań: Wydawnictwo Poznańskie. Nowak, L. (1977). Wstęp do idealizacyjnej teorii nauki [An Introduction to the Idealizational Theory of Science].Warszawa: PWN. Nowak, L. (1980). The Structure of Idealization. Towards a Systematic Interpretation of the Marxian Idea of Science. Dordrecht: Reidel. Nowak, L. (1989). Byt i myśl. Przyczynek do metafizyki unitarnej [Being and Mind. Contribution to Unitarian Metaphysics]. Studia Filozoficzne 1, 1–18. Nowak, L. (1990). Abstracts Are Not Our Constructs. The Mental Constructs Are Abstracts. In: J. Brzeziński, F. Coniglione, T.A. Kuipers, L. Nowak (eds.) Idealization I: General Problems. Poznań Studies in the Philosophy of the Sciences and the Humanities 16, pp. 193–206. Amsterdam/Atlanta, GA: Rodopi. Nowakowa, I. (1990). Historical Narration and Idealization. In: J. Topolski (ed.), Narration and Explanation. Contributions to the Methodology of the Historical Research. Poznań Studies in the Philosophy of the Science and the Humanities 19, pp. 31–40. Amsterdam/Atlanta: Rodopi.

318

Szwochert

Pomorski, J. (1981). Próba idealizacyjnej teorii nauk historycznych [An Attempt at an Idealizational Theory of Historical Sciences]. Studia Metodologiczne 20, 61–93. Pomorski, J. (1995). Paradygmat „New Economic History.” Studium z teorii rozwoju nauki historycznej [Paradigm of New Economic History: A Study in the Theory of Development of Historical Science]. Lublin: Wydawnictwo Uniwersytetu Marii Curie-Skłodowskiej. Topolski, J. (2005). Wprowadzenie do historii [Introduction to History]. Poznań: Wydawnictwo Poznańskie. Hayden White (1973). Metahistory. The Historical Imagination in Nineteenth-Century Europe. Baltimore/London: Johns Hopkins University Press. White, H. (1985). Tropics of Discourse. Essays in Cultural Criticism. Baltimore/London: The Johns Hopkins University Press. Wrzosek, W. (2010). Historia. Kultura. Metafora. Powstanie nieklasycznej historiografii [History – Culture – Metaphor. Facets of non-Classical Historiography]. Wrocław: Wydawnictwo Uniwersytetu Wrocławskiego.