The History of Understanding in Analytic Philosophy: Around Logical Empiricism 9781350159204, 9781350159235, 9781350159211

Citation preview

THE HISTORY OF UNDERSTANDING IN ANALYTIC PHILOSOPHY

Also available from Bloomsbury Contemporary Philosophy and Social Science, edited by Michiru Nagatsu and Attilia Ruzzene Skepticism: From Antiquity to the Present, edited by Diego Machuca and Baron Reed Spinoza in Twenty-First-Century American and French Philosophy, edited by Jack Stetter and Charles Ramond The Bloomsbury Companion to Analytic Philosophy, edited by Barry Dainton and Howard Robinson The Relevance of Hegel’s Concept of Philosophy, edited by Luca Illetterati and Giovanna Miolli

THE HISTORY OF UNDERSTANDING IN ANALYTIC PHILOSOPHY AROUND LOGICAL EMPIRICISM Edited by Adam Tamas Tuboly

BLOOMSBURY ACADEMIC Bloomsbury Publishing Plc 50 Bedford Square, London, WC1B 3DP, UK 1385 Broadway, New York, NY 10018, USA 29 Earlsfort Terrace, Dublin 2, Ireland BLOOMSBURY, BLOOMSBURY ACADEMIC and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2022 Copyright © Adam Tamas Tuboly and Contributors, 2022 Adam Tamas Tuboly has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as Editor of this work. For legal purposes the Acknowledgments on p. xiii constitute an extension of this copyright page. Cover design by Louise Dugdale Cover image © cjp/Getty Images. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party websites referred to or in this book. All internet addresses given in this book were correct at the time of going to press. The author and publisher regret any inconvenience caused if addresses have changed or sites have ceased to exist, but can accept no responsibility for any such changes. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Tuboly, Adam Tamas, editor. Title: The history of understanding in analytic philosophy : around logical empiricism / edited by Adam Tamas Tuboly. Description: London ; New York : Bloomsbury Academic, 2022. | Includes bibliographical references and index. Identifiers: LCCN 2021038612 (print) | LCCN 2021038613 (ebook) | ISBN 9781350159204 (hb) | ISBN 9781350159211 (epdf) | ISBN 9781350159228 (ebook) Subjects: LCSH: Logical positivism–History. | Analysis (Philosophy)–History. | Social sciences–Philosophy.–History. Classification: LCC B824.6 .H57 2022 (print) | LCC B824.6 (ebook) | DDC 146/.42–dc23 LC record available at https://lccn.loc.gov/2021038612 LC ebook record available at https://lccn.loc.gov/2021038613 ISBN: HB: 978-1-3501-5920-4 ePDF: 978-1-3501-5921-1 eBook: 978-1-3501-5922-8 Typeset by Newgen KnowledgeWorks Pvt. Ltd., Chennai, India To find out more about our authors and books visit www.bloomsbury.com and sign up for our newsletters.

To Gabó, Sára, Róza: thank you

vi

CONTENTS

L ist of C ontributors A cknowledgments

Introduction Adam Tamas Tuboly

ix xiii 1

Part I  Before Logical Empiricism 1 Georg Simmel on Historical Understanding Martin Kusch

13

2 Understanding, Psychology, and the Human Sciences: Dilthey and Völkerpsychologie 39 Lydia Patton 3 The Heuristic and Epistemic Account of Verstehen in Twentieth-century American Philosophy Fons Dewulf

63

Part II  Logical Empiricism 4 Is There a Hermeneutic Aspect in Carnap’s Aufbau? 87 Christian Damböck

viiiContents

5 More on Neurath on Verstehen: The Rejection of Weber’s Ideal Type Methodology Thomas Uebel

103

6 Neurath’s Debate with Horkheimer and the Critique of Verstehen 135 Andreas Vrahimis 7 Victor Kraft on Verstehen 161 Jan Radler Part III  After Logical Empiricism 8 The Leopard Does Not Change Its Spots: Naturalism and the Argument against Methodological Pluralism in the Sciences Jonas Ahlskog and Giuseppina D’Oro

185

9 Georg Henrik von Wright on Understanding (and Explanation) Henriikka Hannula

209

10 Psychological Understanding Tamás Demeter

229

11 Verstehen Redux: Understanding in Contemporary Epistemology 261 Aaron Preston I ndex

285

CONTRIBUTORS

Jonas Ahlskog is a postdoctoral researcher in Philosophy and History of Ideas at the Department of Philosophy at Åbo Akademi University, Finland. He works on the philosophy of history and the history of ideas. He has published The Primacy of Method in Historical Research: Philosophy of History and the Perspective of Meaning (2021), and he has articles in the Journal of the Philosophy of History, Rethinking History, and Acta Philosophica Fennica. Giuseppina D’Oro is Reader in Philosophy at Keele University, United Kingdom. She is interested in the metaphilosophical underpinnings of the reasons vs causes debate and is currently an executive editor for the Journal of the Philosophy of History. She is the author of Collingwood and the Metaphysics of Experience (2002) and of numerous papers on the philosophy and metaphilosophy of R.G. Collingwood. She co-edited Collingwood on Philosophical Methodology (2018), The Cambridge Companion to Philosophical Methodology (2017), Reasons and Causes (2013), and An Essay on Philosophical Method (2005). Christian Damböck is a Privatdozent at the Philosophical Institute, University of Vienna. He works on the philosophy of science and the humanities in the nineteenth and twentieth centuries, with a particular focus on logical empiricism. He is the author of (2017) and the editor of, among others, Influences on the Aufbau (2016), Young Carnap in

xContributors

an Historical Context: 1918–1935 (with Gereon Wolters, 2021), and Rudolf Carnap. Tagebücher 1908–1935 (2021). Tamás Demeter is Leader of the MTA Lendület Morals and Science Research Group and Professor of Philosophy at Corvinus University of Budapest, Hungary. His interests include the philosophy of mind, the philosophy of the social sciences, and the relation between eighteenth-century philosophy and sciences. He has published David Hume and the Culture of Scottish Newtonianism (2016) and edited Intellectuals, Inequalities and Transitions (2020), Conflicting Values of Inquiry (with Kathryn Murphy and Claus Zittel, 2015), and Essays on Wittgenstein and Austrian Philosophy (2004). He has published numerous papers in Synthese, Journal of Scottish Philosophy, Metascience, Early Science and Medicine, and Monist. His related publications include the forthcoming volume Mental Fictionalism: Philosophical Explorations (co-edited with Ted Parent and Adam Toon, Routledge, 2022), a special issue of The Monist (96, 2013/4) dedicated to “Mental Fictionalism,” and another of European Journal of Analytic Philosophy (5, 2009/2, co-edited with János Tőzsér) on “Challenges for Representational Theories of Mind.” Fons Dewulf is a postdoctoral researcher at the Department of Philosophy and Moral Sciences, Ghent University, Belgium. He works on the history of philosophy of science, the history of scientific explanations, and the twentieth-century philosophy of science. He has published articles in the British Journal for the History of Philosophy, HOPOS, Intellectual History Review, and the Journal of the History of Ideas. Henriikka Hannula is a doctoral student at the Department of Philosophy, University of Vienna, Austria. Her main interests are the philosophy of history and social sciences, hermeneutics, as well as the history of philosophy of late nineteenth century Germany. Martin Kusch is Professor of Philosophy at the University of Vienna, Austria. He works on general philosophy of science, philosophy and history of psychology and the social sciences, philosophy of language, and the sociology of scientific knowledge. His works include Psychologism (1995), Psychological Knowledge (1999), Knowledge by Agreement (2002), A Sceptical Guide to Meaning and Rules (2006) and Relativism in the Philosophy of Science (2021). Lydia Patton is Professor of Philosophy at Virginia Tech, USA, and Editor of HOPOS. She works on the history of philosophy of science. She has

Contributors

xi

edited Laws of Nature (with Walter Ott, 2018) and Philosophy, Science, and History: A Guide and Reader (2014), and she has published numerous articles in Synthese, The Monist, Kant-Studien, Studies in History and Philosophy of Science, and other journals. Aaron Preston is Professor of Philosophy at Valparaiso University, USA. His research interests include ancient philosophy, moral philosophy, philosophy of religion, and the history of analytic philosophy. In addition to a number of articles in these areas, he has published a monograph about analytic philosophy, Analytic Philosophy: History of an Illusion (2010), and edited Analytic Philosophy: An Interpretive History (2017) and co-edited The Disappearance of Moral Knowledge (2018). Jan Radler is an independent researcher interested in the history of philosophy of science, the cultural sciences, and logical empiricism. He has published papers in HOPOS, History of Psychology, and Conceptus. His main interest is the philosophy of Victor Kraft, about whom he published Victor Krafts konstruktiver Empirismus: Eine historische und philosophische Untersuchung (2006). Adam Tamas Tuboly is Leader of the MTA Lendület Values and Sciences Research Group and a research fellow at the Institute of Transdisciplinary Discoveries, Medical School, University of Pécs. He is interested in the history of philosophy of science and logical empiricism. He has edited numerous volumes and has published articles in Synthese, HOPOS, Perspectives on Science, and the British Journal of the History of Philosophy. Thomas Uebel is Emeritus Professor of Philosophy at the University of Manchester, England. He is the author of Overcoming Logical Positivism from Within (1992), Vernunftkritik und Wissenschaft (2000), and Empiricism at the Crossroads (2007), co-author of Otto Neurath: Philosophy Between Science and Politics (1996), editor of Rediscovering the Forgotten Vienna Circle (1991), and co-editor of Otto Neurath: Economic Writings. Selections 1904–1945 (2004). Andreas Vrahimis works at the Department of Classical Studies and Philosophy, University of Cyprus. His research focuses on the analyticcontinental divide from various perspectives. Besides numerous articles, he has authored Encounters Between Analytic and Continental Philosophy (2013), and he has just finished Bergsonism and the History of Analytic Philosophy (2022).

xii

ACKNOWLEDGMENTS

During the editorial work on this volume, I received support from the MTA Lendület Morals and Science Research Group, from the MTA Premium Postdoctoral Scholarship, and in the final stages by the MTA Lendület Values and Science Research Group. The kind help of all these programs is gratefully acknowledged. When this collection was first envisioned and the idea began to take shape on paper (or rather on our screens), it was a joint project with my former colleague, Ákos Sivadó. Unfortunately, Ákos decided to leave the academic sphere, but I am indebted to him for his friendship and for our numerous collegial discussions over the years. In the meantime, however, another challenge materialized: after the Covid-19 pandemic spread around the world in 2020, all of the authors (this editor included) experienced various difficulties in their personal and scholarly lives (which also became entangled in countless ways as we bent over our screens in living rooms with children running around us). I am more than honored and grateful that so many contributors nevertheless managed to write their pieces for this volume. I would like to express my special gratitude to Sammy Hirshland, who kindly went through most of these chapters to improve the language, while Christoph Gottstein, as always, provided indispensable proofreading support during the last two years.

xiv

Introduction ADAM TAMAS TUBOLY

0.1.  “UNDERSTANDING”: LOGICAL EMPIRICISM AND THE HISTORY OF ANALYTIC PHILOSOPHY In his contribution to The Cambridge Companion to Logical Empiricism, Thomas Uebel succinctly summarizes the received view of logical empiricism and its relation to the social sciences, stating that “philosophy of social science is unlikely to figure in many people’s judgment as a field in which logical empiricism effected great progress. If anything, impressions run to the contrary” (2007: 250). Thanks to the decades-long detailed work of Uebel, however, things are slowly changing.1 It recently emerged that logical empiricism was much more socially engaged with its own sociocultural and political milieu than was previously thought (Romizi 2012; Uebel 2012; Schliesser 2022), and the social and political activism of the logical empiricists (taken in the broadest sense) is now also seen in a new light; we know that for many of them, this was an issue with theoretical foundations (Reisch 2005; Howard 2019; Damböck and Tuboly 2022). Most importantly, however, recent scholarship (e.g., Nemeth 2007; Neuber 2022) has scrutinized the place and role of the social sciences in order to overcome some inherited misconceptions according to which the unity of science movement killed off the previously lively interest in the social sciences. There is still much to be said, however, about how the logical empiricists conceived of the social sciences and how the social sciences can be pursued from a logical empiricist perspective. “Verstehen,” “understanding,” “interpretation,” “empathy,” and similar notions are generally discussed in the context of the social sciences. And

2

The History of Understanding in Analytic Philosophy

what is more, they are usually conceived of as the characteristic features, or methodological elements, of the social sciences, in opposition to the natural sciences. Both the social and natural sciences have their own internal histories, and in relation to philosophy, there is much to be said about how early analytic philosophers—usually those in and around the logical empiricist camp—approached the matter. Recently, thanks to the thorough work of historians and philosophers of science, despite orthodox expectations rooted in various shadow histories, it emerged that (most) members of the Vienna Circle did not reject Verstehen altogether. For example, Otto Neurath, among others, worked out a quite sophisticated approach to the social sciences and their interpretative method that aimed to show how the idea of Verstehen could be purified from its metaphysical underpinnings and connotations to produce respectful knowledge about the social, historical, and economic world. Recent discussions point toward the following: There are still misconceptions about the history of analytic philosophy and of logical empiricism that need to be revised; various conceptions in the philosophy of the social sciences related to analytic philosophy (like non-reductive naturalism, interpretative conceptions of social phenomena, sociological approaches to metaphysics, and natural scientific theories) have their source, or better, their unknown and frequently neglected early forerunners, in 1920s and 1930s Vienna. As the often ambiguous, harshly and superficially formulated conceptions of Neurath and others (delivered usually in a misleading and unnecessarily militant language) were either missed or misunderstood by their opponents (Horkheimer), colleagues (Weber, Mises), and friends (Kaufmann), historical and philosophical questions of Verstehen in particular and understanding in general might provide some unexpected insights into the history and development of early analytic philosophy. The doctrine of Verstehen, as it is usually understood, could be mapped within the intellectual history of Europe in three successive phases. First, through the largely empathy-oriented ideas of the German historicist school of the nineteenth century, most famously championed and meticulously elaborated by Wilhelm Dilthey; second, through the shift of focus from empathetic understanding to motivational/rational reconstructions of the actions of agents found in the methodological writings of Max Weber (and given a different voice by R. G. Collingwood concerning historical understanding); and third, by the largely Wittgensteinian account of understanding rule-following as the basis of social action; for example, in the works of Peter Winch. While these three interpretive enterprises have their differences—identified in much of the vast literature they generated with the shift in targets, from an individual agent’s emotional life to that same agent’s

Introduction

3

subjective intentions and, finally, to the underlying principle that governs all individual actions regardless of their emotional or intentional components— what binds them together is their focus on those elements that positivistic or naturalistic accounts of social phenomena necessarily have to exclude from their scope: the qualitative differences between the subject matter of the natural and the human sciences. Accounts of various forms of Verstehen have therefore usually been treated as opposing attempts at unifying scientific methodology and its philosophical underpinnings, accentuating difference where overarching schemes would propose unity, and formulating a seemingly unbridgeable gap between traditionally understood analytic philosophy of science and enquiries into the social realm based on the largely German tradition of the Geisteswissenschaften. Instead of a dialog, the received view largely treats interpretive and positivistic discourses as talking past one another, since their relation to the reality they wish to illuminate is so remarkably different that any attempt at conversation would break down on the conceptual level.

0.2.  THE AIM AND STRUCTURE OF THIS VOLUME The aim of this volume is to enquire more deeply into the broadly conceived topics of Verstehen, understanding, and the philosophy of the social sciences and to contextualize the histories of the Vienna Circle and logical empiricism. In order to do that, the chapters of the volume are divided into three groups. The essays in Part I are devoted to the history of the social sciences before logical empiricism but already in the early phases of nineteenth- and twentieth-century positivism, with a hint toward some transitional stages. Part II follows the history of the social sciences in a chronological and logical manner. It contains papers devoted to logical empiricism that discuss the negative and positive critiques of various logical empiricists (mainly, of course, Neurath, as well as some of his colleagues, such as Felix Kaufmann and Victor Kraft) targeting the social sciences. The aim of these chapters is to show that the attacks launched by the logical empiricists were not due to misguided conceptions of what the social sciences are, nor to misdirected attempts at transferring the qualitative ideas of the social sciences to the solid quantitative foundation of the natural sciences, nor to a simplistic “positivist,” “eliminativist” conception of how the nature of science in general should be reformed. Finally, Part III traces the development of the methodological changes in the social sciences after the decline of logical empiricism, with a particular focus on how the idea of Verstehen became modified (or not) after the critique of logical empiricism.

4

The History of Understanding in Analytic Philosophy

In Chapter 1, Martin Kusch analyzes Georg Simmel’s theory of historical understanding. He highlights the theory’s debts to other authors and flags where Simmel’s proposal was original and provocative. The paper defends the following theses: First, Simmel’s theory of understanding was not initially part of a systematic attempt to demarcate the natural sciences from the social sciences or the humanities. Only from around 1905 onward did Simmel fall in line with the neo-Kantian demarcation efforts. Second, Simmel criticized some ingredients and forms of the “reenactment” theory of understanding. Third, Simmel’s theory was anti-realist in its likening of the historian to an artist and in thereby emphasizing the historian’s creativity in constructing “historical pictures.” This anti-realism was of a piece with Simmel’s neoKantian “critique of historical reason” with its emphasis on the “historical” or “relative a priori.” Fourth, Simmel’s reflections on understanding fitted with his general relativistic outlook. Still, fifth, Simmel was “anti-historicist” if by “historicism” one means the attempt to reduce philosophy to history. And sixth, the development of Simmel’s ideas on understanding culminated in his efforts to integrate his conception of (historical) understanding with his “philosophy of life.” Having reconstructed and analyzed Simmel’s position, Kusch briefly considers criticisms by his contemporaries. In Chapter 2, Lydia Patton examines the sociohistorical situation of two traditions in the Western analysis of culture and how they influenced each other. The two traditions are the hermeneutic approach of Wilhelm Dilthey and the Völkerpsychologie of Moses Lazarus and Chajim Steinthal. Patton’s account focuses on two elements. First, Lazarus and Steinthal attempted to motivate an account based on collective structures, or forms, of rationality made manifest by a people or Volk; the paper explores their account of how collective structures can be employed in sociohistorical analysis. Second, Dilthey rejected Lazarus’s and Steinthal’s argument that it was possible to identify the norms of action governing social phenomena. Dilthey rejected any account of psychology that took it to be law governed, even retrospectively, arguing that the “nexus of life” that is the ultimate basis of the human sciences cannot be reduced to any law-governed or explanatory relationship between the self, society, and nature. However, there is a deep tension in Dilthey’s position here which is evident in the development of his work over time. Patton explores this tension and its significance for the understanding of the subject and the role of psychology in the human sciences. Chapter 3 is a transitional chapter in the sense that it focuses on pre-logical empiricist figures against the background of logical empiricism. Fons Dewulf distinguishes two ways to approach Verstehen: a heuristic perspective and an epistemic one. Whereas the heuristic perspective conceives Verstehen merely

Introduction

5

as a subjective methodological aid to guide empirical research, the epistemic account conceives it as a way to transform source material into empirical evidence. He uses the contemporary investigation of Shakespeare’s sexuality to illustrate this distinction and discusses the heuristic account of Verstehen as originally defended by Carl Hempel and Theodore Abel to show that their account had its origin in an interpretation of Dilthey as a relativist, a view originally put forward by the American philosopher Maurice Mandelbaum. By revisiting Paul Oskar Kristeller’s and Otto Neurath’s earlier responses against Hempel’s heuristic account, Dewulf argues that the heuristic account lacked any credibility from the very beginning, precluding its application to real debates in the humanities, such as that about Shakespeare’s sexuality. In the opening chapter of the second part, Chapter 4, Christian Damböck elaborates on the now classical interpretation of Carnap’s Aufbau by Michael Friedman. His main idea is that while Friedman correctly explains the crucial formal notion of Carnap’s account, his conception does not tell the whole story of the role of objectivity in the Aufbau. The purely structural definite descriptions highlighted by Friedman are not the only condition for the intersubjective communicability of concepts; certain reducibility conditions also need to be added that include psychophysical parallelism and a somewhat Diltheyian notion of understanding via analogical conclusion. If we add these conditions, it turns out that objectivity in the Aufbau ultimately is not just a purely logical (or purely structural) issue but also involves the acceptance of certain rather traditional philosophical notions. This becomes possible because, in Carnap’s scientific framework, these notions are indeed not philosophical, given their character as nontrivial empirical claims. Following his influential 2019 article “Neurath on Verstehen,” Thomas Uebel argues in Chapter 5 that Neurath misunderstood Weber insofar as he overlooked his employment of causal reasoning as a control instance for the attribution of beliefs and desires or preferences and intentions. Neurath’s opposition to Weber’s interpretive sociology contributed much to his own mischaracterization as a reductive behaviorist. But that Neurath was wrong on this count does not mean that he was wrong in every one of his assessments of Weber’s conception of social science as categorically different from natural science. Two prominent examples are Weber’s ideal type methodology and his claim that all social science is interpretivist. The employment of so-called ideal types—concepts without exact complements in empirical reality—was regarded by Weber and others as a distinguishing criterion of social science that bars their integration into unified science. Neurath was skeptical about the use of Weberian ideal types for a number of reasons, all of them reflecting his concern regarding the lack of empirical control over such theorizing, and he therefore rejected the categorical

6

The History of Understanding in Analytic Philosophy

distinction proposed by Weber. Uebel thus reconstructs Weber’s ideal type theory and Neurath’s criticism and briefly outlines his alternative. In Chapter 6, Andreas Vrahimis presents another case study on Neurath, focusing on the failed attempt at collaboration between the Frankfurt School and the Vienna Circle that culminated in Horkheimer’s 1937 paper “The Latest Attack on Metaphysics.” Horkheimer relied on a caricature of positivism as espousing an uncritical myth of the given and drew far-reaching conclusions concerning positivism’s conservative prohibition of the radical questioning of appearances. However, Horkheimer later applied some of these criticisms to Dilthey’s conception of Verstehen, while presenting logical empiricism as dismissing Dilthey’s proposals as nothing more than poetry. By examining Neurath’s unpublished reply to Horkheimer and drawing on Thomas Uebel’s abovementioned account of logical empiricism’s stance towards Verstehen, Vrahimis attempts to dispel Horkheimer’s caricature. He highlights a parallel between Neurath’s attitude towards claims about Verstehen in social science and Horkheimer’s proposals for the use of Vernunft by a “supra-scientific” critical theory. Neurath argues that once Horkheimer’s objections to “positivism” are reformulated in non-metaphysical terms, they point towards a genuine problem concerning science’s relation to social praxis. For Neurath, the problem can be addressed from within (rather than, as Horkheimer would have it, from above) empirically minded investigations into the history and sociology of science. In Chapter 7, Jan Radler investigates an often overlooked figure of logical empiricism. He presents Victor Kraft’s Verstehen approach and important arguments for rejecting the nomothetic-ideographic dichotomy in Kraft’s work, which was the result of his occupation with contemporary geographical thought. In his discussion of the historical sciences, Kraft rebutted the widespread claim that history has its own methodology. Radler also compares Kraft’s position with the semantic ideas of scholars who developed (or proposed) another theory of meaning, such as Arne Naess’s empirical semantics. Naess was deeply influenced by Otto Neurath’s program of unified science. Kraft did not share the idea of empirical semantics or unified science. As a heuristic philosopher, however, he still believed that Verstehen could retain its position. The main difference between the two approaches that Radler presents at the end of his paper lies in their different conceptions of science. While Neurath stressed the social and collaborative aspect of science, Kraft, the more traditional thinker, stressed its foundational basis. After these chapters, with their focus on how to dispel certain entrenched myths about logical empiricism in the context of the social sciences, Part III moves forward both conceptually and chronologically. Following Dilthey and Weber, new conceptions of understanding emerged after World War II.

Introduction

7

In Chapter 8, Giuseppina D’Oro and Jonas Ahlskog challenge the view that articulating a defense of the sui generis character of humanistic explanations requires a commitment to the early modern conception of the mind as immortalized in Ryle’s metaphor of the (Cartesian) ghost in the machine and Quine’s metaphor of the (Lockean) myth of the museum. By caricaturizing the arguments for disunity in the sciences, these powerful metaphors have not only contributed to undermining the claim of methodological pluralism; they have also, and more worryingly, planted red herrings that have diverted attention away from the genuine issues at stake. D’Oro and Ahlskog’s paper is an exercise in removing these false clues to reveal what the claim of the autonomy of humanistic explanations really amounts to. In the following chapter, Henriikka Hannula takes as her point of departure a classic text in the tradition of the philosophy of the social sciences. Her chapter is a critical introduction to Georg Henrik von Wright’s Explanation and Understanding (1971). In his famous book, von Wright tackles the old explanation-understanding debate from an analytic perspective. He reconstructs the dichotomy between two models of explanation: the covering law model and the practical syllogism. He argues that while the former is a suitable model for the natural sciences, the human sciences need another explanation schema; namely, that of the practical syllogism. These models are also historically grounded. Von Wright distinguishes between two main traditions in the philosophy of science: the Galilean, which takes causal explanation as primary, and the Aristotelian, which emphasizes purpose and intentionality. Thus, von Wright also argues for a dualistic vision of science. Hannula summarizes, succinctly and enlighteningly, both the historical argument on the two traditions and the systematic arguments concerning causal and teleological explanations. She critically assesses how these different explanation schemas work in practice in the human sciences and, finally, whether von Wright’s overall argument supports the view that there is a genuine difference between the natural and human sciences. Apart from providing an accessible introduction to von Wright’s argumentation, Hannula also seeks to contextualize it in the broader tradition of the explanation-understanding debate. In Chapter 10, Tamás Demeter argues that psychological interpretation is not a fact-stating discourse and that its benefits are not of an epistemic nature, but of a social one. He states that by engaging in psychological discourse, we can do two things: resolve affective tensions that arise in the course of social interactions and provide orientation in the social world. Psychological interpretations can achieve this by a framing effect: they posit evaluative concepts for interpreting behavior, due to which psychological representations become affectively and motivationally significant. This

8

The History of Understanding in Analytic Philosophy

significance does not arise from any explanatory or predictive power, but from the specific light such interpretations shed on the behavior so interpreted; that is, from the way psychological concepts frame events by embedding them in a network of affectively significant concepts. Affective resolution and motivational significance are thus the chief non-epistemic benefits of our interpretive practices in psychology. In the final chapter, Aaron Preston takes a historical-critical look at the contemporary scene. He argues that for most of the twentieth century, analytic epistemology has been narrowly focused on the concept of knowledge and on closely related issues like justification. But decades of work on Gettier problems, lottery paradoxes, and the like have given mainstream analytic epistemology the appearance of a degenerating research program. In response, some analytic epistemologists have begun to investigate other cognitive states, such as wisdom and understanding, that many pre-analytic philosophers valued as much as or more than knowledge. In fact, Linda Zagzebski argues that Western philosophy exhibits the following historical pattern: in eras given to skeptical worries, philosophers have focused on certainty and justification; in eras less exercised by skepticism, philosophers have focused on understanding and the related issue of explanation. Preston’s aims are twofold. First, he adds to Zagzebski’s picture by showing that metaphysical and metaphilosophical issues have also contributed to this historical pattern. Second, he uses this expanded picture to explain the nineteenth-century advent of Verstehen as a term of art, the analytic tradition’s lack of interest in understanding, and the revival of interest in understanding in recent analytic epistemology. Preston concludes that the pro-Verstehen movement of the nineteenth century and the pro-understanding movement of the twenty-first century were motivated by similar concerns pertaining to the epistemic strictures of scientism, strictures which bear a complex relationship to skepticism. As these short summaries have shown, the present collection is indeed a stratified undertaking around logical empiricism. Its main motivation is to get a clearer picture of the logical empiricists’—but mainly Neurath’s— stance towards Verstehen (as documented in Part II), to understand where their arguments came from historically and culturally (as documented in Part I), and to examine how major figures of analytic philosophy, such as Quine, Winch, and von Wright, moved towards a more refined understanding of these issues, often based on a misleading notion of logical empiricism that pictured them falsely as dogmatist and reductionist. While Neurath, Kraft, Kaufmann, Carnap, and all the other figures in the positivist story—one might also add Edgar Zilsel and Heinrich Neider, though they are unfortunately not represented here—played a prominent role in the middle of the twentieth

Introduction

9

century (often through indirect channels and within unbalanced debates with people such as Apel, Mandelbaum, or Horkheimer), neither the philosophy of the social sciences nor the history of Verstehen would be exhausted by them. As the last two chapters of the volume show, contemporary discussions often focus on entirely different conceptions and goals in the context of understanding—even though many of them go back directly to the emphatic and interpretative versions of Verstehen. If the reader takes home the idea that the philosophy of the social sciences is a significantly underrated and underrepresented field among historians of analytic philosophy, then the volume will have achieved its main goal. For our aim is to show that prominent twentieth-century analytic philosophers, besides deliberating over the axiomatization of the natural sciences, the critique of metaphysics, and the problem of demarcation and induction, were also involved in significant traditional debates in philosophy concerning understanding, interpretation, and empathy, dating back to nineteenthcentury German traditions. In other words, the logical empiricists still have much to offer when it comes to sating our contemporary appetite for systematic concepts.

0.3. ACKNOWLEDGMENTS When we began working on this volume, I received support from the MTA Lendület Morals and Science Research Group and from the MTA Premium Postdoctoral Scholarship; in the final stages, I was supported by the MTA Lendület Values and Science Research Group.

NOTE 1 In fact, the whole idea of this volume goes back to the works of Uebel. Around 2017, his latest manuscript, “Neurath on Verstehen,” was circulated among colleagues, and many of them became quite interested (see Uebel 2019). The paper included some captivating and novel theses about Neurath and the history of the social sciences and shed new light on the philosophical projects of logical empiricism, especially regarding the key notions and method of understanding and interpretation (these are often taken to be the weakest elements in the philosophical project of logical empiricism). Thus a small, one-day workshop (Thomas Uebel on Verstehen and Logical Empiricism) was organized at the University of Pécs in September 2018. The main focus of the day was Uebel’s manuscript, but the speakers showed a much broader interest in the rehabilitation and contextualization of logical empiricism and the social sciences in the history of analytic philosophy. Quite happily and unexpectedly, this workshop held in Uebel’s honor coincided with his birthday; while at the time we only celebrated with a quick homemade cake, we now also present him with this volume.

10

The History of Understanding in Analytic Philosophy

REFERENCES Damböck, C., and A. T. Tuboly, eds. (2022), The Socio-ethical Dimension of Knowledge: The Mission of Logical Empiricism, Cham: Springer. Howard, D. (2019), “Otto Neurath: Philosopher in the Cave,” in J. Cat and A. T. Tuboly (eds.), Neurath Reconsidered: New Sources and Perspectives, 45–65, Cham: Springer. Nemeth, E. (2007), “Logical Empiricism and the History and Sociology of Science,” in A. Richardson and T. Uebel (eds.), The Cambridge Companion to Logical Empiricism, 278–302, Cambridge: Cambridge University Press. Neuber, M. (2022), “Nagel on the Methodology of the Social Sciences,” in M. Neuber and A. T. Tuboly (eds.), Ernest Nagel: Philosophy of Science and the Fight for Clarity, 215–32, Cham: Springer. Reisch, G. A. (2005), How the Cold War Transformed Philosophy of Science: To the Icy Slopes of Logic, New York: Cambridge University Press. Romizi, D. (2012), “ ‘The Vienna Circle’s “Scientific World-Conception’: Philosophy of Science in the Political Arena,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 2 (2): 205–42. Schliesser, E. (2022), “Philosophy of Science as First Philosophy: The Liberal Polemics of Ernest Nagel,” in M. Neuber and A. T. Tuboly (eds.), Ernest Nagel: Philosophy of Science and the Fight for Clarity, 233–53, Cham: Springer. Uebel, T. (2007), “Philosophy of Social Science in Early Logical Empiricism: The Case of Radical Physicalism,” in A. Richardson and T. Uebel (eds.), The Cambridge Companion to Logical Empiricism, 250–77, Cambridge: Cambridge University Press. Uebel, T. (2012), “Carnap, Philosophy, and ‘Politics in its Broadest Sense,’ ” in R. Creath (ed.), Rudolf Carnap and the Legacy of Logical Empiricism, 133–48, Dordrecht: Springer. Uebel, T. (2019), “Neurath on Verstehen,” European Journal of Philosophy, 27 (4): 912–38. von Wright, G. H. (1971), Explanation and Understanding, London: Routledge and Kegan Paul.

PART ONE

Before Logical Empiricism

12

CHAPTER ONE

Georg Simmel on Historical Understanding MARTIN KUSCH

1.1. INTRODUCTION Since the early nineteenth century, “understanding” (Verstehen) has been a central category in philosophical debates over the aims and methods of the sciences and the arts, over methodological differences between fields of study, and even over the human condition in general. The term first gained prominence in the German-speaking lands where historians like Johann Gustav Droysen, philosophers like Wilhelm Dilthey, or social theorists like Max Weber used it to demarcate history from philosophy, the humanities from the natural sciences, or the right from the wrong approaches in history, economics, and sociology. In this chapter, I shall analyze a contribution, of the same time and context, that has only rarely received detailed scholarly attention: Georg Simmel’s theory of historical understanding. Simmel’s account was closely related to the ideas of Droysen, Dilthey, and Weber; he critically responded to Droysen or Dilthey, and Weber reacted with both admiration and hostility to Simmel’s suggestions. At the same time, Simmel’s proposal was bold, original, and provocative. In what follows, I shall reconstruct Simmel’s reflections on understanding as he developed them from the 1890s until his death in 1918. Limitations of space do not, however, allow me to detail the various stages in the development of Simmel’s position.1

14

The History of Understanding in Analytic Philosophy

I shall try to make plausible the following theses. First, Simmel’s theory of understanding was not initially part of a systematic attempt to demarcate the natural sciences from the social sciences or the humanities. Only from around 1905 onward did Simmel fall in line the neo-Kantian demarcation efforts. Second, Simmel criticized some ingredients and forms of the “reenactment” (Nachbilden) theory of understanding. Third, Simmel’s theory was anti-realist in its likening of the historian to an artist and in thereby emphasizing the historian’s creativity in constructing “historical pictures.” This anti-realism was of a piece with Simmel’s neo-Kantian “critique of historical reason” with its emphasis on the “historical” or “relative a priori.” Fourth, Simmel’s reflections on understanding fitted with his general relativistic outlook. Still, fifth, Simmel was “anti-historicist” if by “historicism” one means the attempt to reduce philosophy to history. And sixth, the development of Simmel’s ideas on understanding culminated in his efforts to integrate his conception of (historical) understanding with his “philosophy of life” (Lebensphilosophie). Having reconstructed and analyzed Simmel’s position, I shall briefly consider criticisms by his contemporaries, especially Franz Eulenburg, Max Frischeisen-Köhler, Otto Hintze, and Max Weber. This will allow me to comment, in a context-sensitive way, on the strengths and weaknesses, as well as the originality, of Simmel’s position.

1.2.  VERSTEHEN AND THE NINETEENTH-CENTURY “DEMARCATION PROBLEMS” Nineteenth-century debates over the nature and role of understanding were tied to attempts to distinguish between philosophy, the natural sciences, the humanities, and the social sciences. Droysen ([1868] 2013: § 14) famously wrote: “In line with the objects and the nature of human thinking, there are three possible scientific methods: the (philosophically or theologically) speculative [method], the mathematical-physical [method], and the historical [method]. Their respective essence is: to come to know [erkennen], to explain, and to understand.” Dilthey used understanding in a similar way. His credo “nature we explain, mental life we understand” ([1894] 1990: 144) was frequently cited. The underlying thought was that in the study of “external nature, we postulate, below the appearances, a connection in the form of abstract concepts. But in the intellectual [geistig] world, the connection is experienced and reenacted [nachverstanden]” (Dilthey [1910] 2006: 49). At least in his early writings (e.g., Einleitung in die Moralwissenschaft), Simmel ([1891–2] 1989–91) emphasized the “unity of science.” Thus, the first edition of Probleme der Geschichtsphilosophie merely conceded that the natural sciences differ from historical research in regularly displaying

Georg Simmel on Historical Understanding

15

a stronger and broader consensus concerning “categories of ordering and evaluating” ([1892] 1989: 328). This was more a difference in degree than a difference in kind. Moreover, Simmel reasoned that since all historical research is ultimately concerned with human mental states, psychology must have a central position within it. He wrote that “if psychology were a science of laws of nature, then history would be applied psychology in the very sense in which astronomy is applied mathematics” ([1892] 1989: 304). Three things are important about this sentence. First, Simmel did not say that history is applied psychology. He formulated a counterfactual with a false antecedent; that is, he did not believe psychology to have (strict) laws (ibid.: 339–79), and hence he denied that historical research is simply applied psychology. Second, even if there were psychological laws of nature, history would still not reduce to psychology. After all, and to stick to Simmel’s comparison, many astronomical entities cannot be reduced to mathematics. Finally, while Simmel rejected the idea that history is applied psychology, he was comfortable with saying that “psychology is the a priori of the historical sciences.” And it was the descriptive—not the normative!—task of epistemology to find the rules “according to which, on the basis of mind-external documents and traditions, one infers not only mental processes but also the elements that suffice to create an ‘understandable’ connection between these processes” (ibid.: 338). In subsequent writings, Simmel’s attitude towards demarcation changed. The second (1905) and third (1907) editions of Probleme der Geschichtsphilosophie discussed the relationship between historical research and psychology in new ways. (I shall follow the 1907 edition throughout, since it differs only slightly from the 1905 edition.) To begin with, Simmel now aimed to explain how history’s focus on individuals could be reconciled with a (limited) role for psychological laws. Key here was an analogy with a major thesis in Simmel’s philosophy of natural science. According to this thesis, nature consists of two (kinds of) elements, laws of nature and a “primary matter” with a structure in and of itself. This primary matter is governed by laws of nature. It can be conceived of, Simmel [1907] 2015: 234) proposed, as a “total individual” (Gesamtindividuum). Concerning psychological laws in particular, Simmel reasoned that we can think of each human being as an individual in the sense introduced: each one of us is governed by the same psychological laws, and yet, in each case, these psychological laws govern a different primary structure, a different “real a priori.” In this way, general psychological laws and ultimately ineffable individuals both have to be reckoned with (ibid.: 235). Moreover, for Simmel historical research was situated between psychology and logic. Logic concerns itself with the contents of propositions

16

The History of Understanding in Analytic Philosophy

and the logical relations between them. Psychology investigates mental processes. History differs from both logic and psychology. Like psychology, history focuses on mental processes; and like logic, history is preoccupied with contents. In other words, history is interested in the “psychological development of contents”; it aims to understand how individuals experienced, shaped, and responded to such contents (ibid.: 237). Simmel also held that historical understanding is primarily preoccupied with “reenacting meaning” (Sinn). This sets history apart from the causal analysis of psychology and the inference-centered perspective of logic. In addition, historical understanding is distinctive also because it homes in on particular events rather than general patterns. Historical understanding draws contents and mental processes together into “unique constellations,” “syntheses of imagination,” where “there is no necessity of psychological events based on laws of nature, and there is no logical necessity connecting the contents with general validity” (ibid.: 274). Reaching this goal requires the peculiar talents and dispositions of outstanding individual historians. In the natural sciences— where “mathematics is the ideal”—the individuality of the researcher is of little significance in the constitution of the subject matter. But it is of great importance in the case of history (ibid.: 296). The demarcation efforts of the second and third editions of Probleme der Geschichtsphilosophie clearly resonated both with Droysen’s and Dilthey’s emphasis on understanding as unique to the humanities and with the Southwest School of neo-Kantians’ urging that history is essentially about particular events (Windelband, Rickert). In later writings, Simmel tied the concerns with demarcation to his “philosophy of life” (Lebensphilosophie). In a paper of 1918, he conjectured that the historian deals with a peculiar object: an object to which the historian “has to grant a being-for-itself ”; that is, an object in “the category of the you.” The “you” is the only element in the universe allowing for a “mutual understanding” (gegenseitiges Verstehen) and a feeling of unity. Simmel regarded it as obvious that the category of the “you” is, for the both the everyday and the historical worlds, as important as are the categories of substance or causality for the natural scientific world (1918: 161–2). With this claim, Simmel committed himself to a very strict divide between the natural and the historical sciences.

1.3.  THE CRITIQUE OF THE THEORY OF REENACTMENT One central thread running through all of Simmel’s discussions of understanding was the criticism of certain versions or ingredients of the

Georg Simmel on Historical Understanding

17

“reenactment” theory of understanding. An influential formulation of this version could be found in some passages of Droysen’s Historik: Understanding is made possible by the fact that […] every outward expression reflects inner processes. When it is perceived, the expression, by projecting itself into the inner of the perceiver, arouses the same inner process in him. When we hear the scream of fear we experience the fear of the person screaming. […] [T]‌he act of understanding […] occurs […] as an immediate intuition, as a creative act, like a spark of light between two approaching electrophoretic bodies, like conception in copulation. (Droysen [1868] 2013: §§ 9, 11) Similar notions can be found in many other nineteenth-century German historians, artists, theologians and philosophers (cf. Wach 1926, 1929, 1933; Beiser 2011). Dilthey claimed, for instance, that our attribution of mental states to others is based “on the achievement of a procedure that is equivalent to an inference of analogy.” We recognize other people’s behavior and then “transfer to them the mental states that we tend to express with such behaviour” (Dilthey [1883] 2017: 20). Elsewhere, Dilthey insisted on a central role for “the understanding that penetrates the alien expressions of life by way of a transposition of the abundance of one’s own inner experiences” ([1910] 2006: 48–9). Simmel raised several concerns regarding reenactment theory. First of all, he challenged the view that to understand an utterance U of another person P, one has to reproduce in one’s mind the very mental processes that caused P to produce U. Simmel believed that such “direct copying” is but a transient element of one kind of understanding: the understanding of “theoretical […] objective and logical” propositions. Assume P believes, and tells me, that (*) 57 + 68 = 125, and that I come to believe (*) on the basis of P’s telling. Simmel granted that in order to grasp (*), I have to momentarily copy P’s mental state of belief. But as soon as I have grasped (*), I can erase my belief about the origin of (*) in P’s mental life. My grasp of the content of (*) is then self-sufficient. Things are different when we are dealing with P’s expressions of emotions, moods, or volitions. In this case, some representation of P’s mental states is essential for our understanding of his expressions. But even here, such representation of P’s mental state is not “direct copy” (Simmel [1892] 1989: 317). Second, according to Simmel, it is wrong to think that the mental act M1, in and through which the historian H understands the mental act M2 of the historical actor A, can ever be identical to M2. Simmel pointed out that if M1 and M2 were identical, then the historian would willy-nilly turn into the

18

The History of Understanding in Analytic Philosophy

historical actor. M1 would no longer be about M2; it would be M2. It is no solution, Simmel added, to say that H’s M1 consists of two parts, M2 (of A) and the further “mental note” (of H), stating that M2 is primarily the mental state of A, and not H (ibid.: 319–20). Simmel rejected any suggestion that in knowing A’s M2, H needs to bring together separate elements—like the mental state and the mental note. Rather, when H “projects” beliefs and other mental states onto A, H performs a “unified act.” It is only retroactive reflection which inclines us to postulate two distinct elements. All this is not to deny, however, that H’s understanding of M2 is made easier when H, in the past, has had mental states similar to M2. But this does not mean that M1 has to be identical with, or similar to, M2 (ibid.: 320). Indeed, we often take ourselves to be able to correctly reconstruct others’ mental processes even when we ourselves have never had processes of this kind. Moreover, Simmel regarded as “cheap” the reply (by Droysen or Dilthey) that historians can understand unfamiliar mental states by “transforming their own experiences.” Simmel lamented that no one had explained how such transformation works and how historians would know whether the transformation has been successful (ibid.: 328). Third, Simmel reminded his readers that historians differ in their ability to understand various historical figures or periods. He believed that sometimes understanding historical actors even requires special innate talents. Relying on the idea of “organic memory”—influential with the tradition of Völkerpsychologie (see Otis 1994: 97–111)—Simmel asserted that “geniuses of historical research” owe their talents to biological inheritance. Geniuses have organically inherited mental dispositions, emotions, and beliefs of the distant past: “mental processes, remote from the genius’ own experience, occur within him since they have been deposited in his organism as memories of the human species” (Simmel [1892] 1989: 329). With this suggestion, Simmel even claimed to have recaptured the rational core of Plato’s doctrine of anamnesis (ibid.: 330). Still, however proud Simmel was of this idea in 1892, by the early 1900s, he was voicing caution. He now asked his reader to take his proposal in a heuristic way: our memory seems to work “as if ” there were such biological processes (Simmel [1907] 2015: 303–4). Fourth, Simmel maintained that the method of reenactment is of mixed value when it comes to understanding groups and social structures of the past. On the one hand, groups are easier to understand than individuals, since what binds group members together is the smallest common denominator of their cognitive and emotional states (Simmel [1892] 1989: 332). On the other hand, in the case of many social institutions, their functioning is not understood by the individuals involved. Reenacting the individuals’

Georg Simmel on Historical Understanding

19

conscious states therefore does not get at the institutions’ real causes or purposes. In support of this contention, Simmel ([1907] 2015: 253) referred to the Marxist analysis of history and ideology. Fifth, by 1905 Simmel had sharpened his criticism of the reenactment theory for the special case of understanding theoretical contents. Earlier, he had insisted that we can erase or forget the element of reenactment once we have grasped the proposition in question. Now he announced that such element of reenactment was not needed at all ([1907] 2015: 262). To understand an ordinary utterance of, say, the law of gravity, I need to grasp what was said and not who said it and why. This is so at least until further purposes come into play; maybe the speaker cited this law in order to provoke me or to make me feel stupid. Under these conditions, understanding the speaker’s mental states is important. Such a wider view is crucial also when the historian tries to understand how Newton discovered and formulated the law of gravity. The historian is aiming for an “image [Bild], an interpretation, a selection, a combination of psychological facts of individual, societal, [and] scientific […] kinds” ([1907] 2015: 263). This kind of “reproduction” is not a mere repetition of the same mental state (ibid.: 264). Thus, either reenactment is not needed at all, as in the understanding of propositions, or it is but a small element in a wider motley of methods. Sixth, Simmel alleged that previous accounts of reenactment had failed to explain how the reenacting historian can assess the success or failure of his action. Simmel proposed that what convinces us of the adequacy of a reenactment is a “feeling […] of typical validity.” Typical validity differs from logical validity in that it concerns the psychological connections between contents: it occurs when we find these connections to be of “über- [trans-, inter-, super-] personal character.” A paradigmatic case is the compelling interconnection between different artistic elements in an impressive and outstanding poem (ibid.: 267). Simmel also spoke of a “feeling of psychological probability” as the “criterion for when internally experienced psychological structure merits the status of objective validity” (ibid.: 268). Seventh, shortly before his death in 1918, Simmel attacked the reenactment theory as an expression of skepticism. He did so on the grounds that it assumed us having full access only to the physical outside of human beings and our attributing of mental states being impossible to confirm ([1918] 2015: 154–5). The reenactment theory also accepted that the greater the intellectual, emotional and cultural distance between the historian and the historical actors, the higher the probability that the historian’s hypotheses are false. In this spirit, it was sometimes alleged that the petit bourgeois historian could not comprehend Napoleon, Goethe, or

20

The History of Understanding in Analytic Philosophy

Nietzsche or that the average European would never understand “the soul of the Oriental.” Dilthey ([1883] 2017: 162) once speculated that they are unable to understand a culture without religious beliefs. Simmel himself wondered whether the historians of his time were able to understand the Athenians of the Persian Wars, or the mediaeval monks. At issue here was not a lack of sources; the point was rather that the historian’s mind is unable to find within itself relevant intellectual or emotional material (Simmel [1918] 2015: 155–6). Simmel readily acknowledged that the greater the intellectual, emotional and cultural differences between historian and historical actor, the lower the probability that the former will adequately understand the latter. Still, Simmel refused to accept the inverse: that strong similarity (alone) explains the possibility of understanding. For Simmel, the similarity between historian and historical actor is a background condition of understanding, not its immediate cause. To make his case, Simmel invoked our frequent difficulties in understanding some of our own recent actions. Here the historian and the historical actor are identical (ibid.: 158). Simmel also drew attention to a simple phenomenon that seems to refute the key idea underlying the reenactment theory. This is the idea that in order to understand another person’s, P’s, outward behavior B, as an expression of mental state M, I must have assembled the inductive knowledge that in my own case, occurrences of (type) B are expressions of (type) M. The simple phenomenon Simmel referred to was the human gaze. Although we frequently “read” a person’s mind by meeting their eyes, none of us has acquired—with the help of a mirror—inductive knowledge concerning the correlations between types of gaze and types of mental states (ibid.: 157). Eighth, Simmel insisted that the “you” is a “primitive phenomenon just like the I” and that we understand others as “immediately besouled.” Simmel drew a comparison with space. Space is not a container into which our perception projects objects (“like one moves with one’s furniture into an empty apartment”). Rather, “the property of being spatial [Räumlichkeit]” is “a primordial form of intuition [Anschauen].” The same goes for souls. Others’ souls are not constituted by our acts of projection; souls are immediately “apperceived” (or implicitly understood) in our encounters. Separating persons into body and soul is “the act of violence of an atomising form of thinking” (ibid.: 160–1). Again, this was not to deny that the full category of “you” and competent historical understanding are results of development and maturation, and thus have “complicated psychological preconditions.” Yet, Simmel thought it a mistake to assume that since the category of “you” is the result of a psychological development, it cannot be a primitive, irreducible phenomenon (ibid.: 1629).

Georg Simmel on Historical Understanding

21

1.4.  THE KANTIAN CRITIQUE OF HISTORICAL REALISM From around 1905 onward, Simmel ([1907] 2015: 229) couched his discussion of understanding as part and parcel of a “critique of historical realism” and as inseparable from the core neo-Kantian question: how is history [as a Wissenschaft] possible? Here Simmel followed in the footsteps of others. Droysen ([1843] 1925) had insisted that “what we need is a Kant … [or] a categorical imperative of history.”2 He also argued that historical understanding is perspectival and defended the view that history is invariably a “construction.” Accordingly, Droysen dismissed the correspondence theory of truth. Dilthey ([1903] 1990: 9) explicitly used the phrase “the critique of historical reason” to characterize his own work. Still, for all these influential predecessors, Simmel retained his own original take on these themes. One original idea was the claim that reenactment theory involves a “naïve” realist commitment to truth as correspondence. As Simmel ([1907] 2015: 275; [1918] 2015: 160) rendered reenactment theory, it involved the claim that the historical actor’s mental states are understood if, and only if, they correspond to, or are identical with, mental states of the historian. Simmel detected “naïve realism” also in Leopold von Ranke’s thought that in order to determine “how things actually have happened” (wie es eigentlich gewesen), the historian has to “extinguish himself.” As Simmel ([1892] 1989: 321) saw it, such extinguishing would destroy the very possibility of historical research itself (cf. Simmel [1907] 2015: 295). Simmel’s alternative was to emphasize “historical pictures” and to thereby propose similarities between the work of the historian and the work of the “artist.” All knowledge is due to “transformations of the immediately given into a new language with its own forms, categories and demands.” History is no exception. The historian understands historical actors and their actions and products by constructing pictures within which parts or aspects of these actors, actions and products are connected and made to cohere. Simmel took this to be obvious even in the case of autobiographical writing. When we look over our life as a whole, we often come to ask questions that did not arise in the course of life as it was originally experienced. We create a new “theoretical image” and a new “distance between experience and its image in our knowledge” (Simmel [1907] 2015: 277–9). The historian thus must not try to erase himself. On the contrary, only historians with a “strongly expressed intellectual character” can come to grips with certain historical individuals. As we have already seen, Simmel believed that here lies an important difference between history and natural sciences, which have their ideal model in mathematics. The difference also explains

22

The History of Understanding in Analytic Philosophy

why in the historical sciences the “ultimate results involve a more variable, more stretchable, and less final kind of objectivity than what we find in the natural sciences” (ibid.: 296–7). What is more, the constructed character of history is obvious, Simmel insisted, when we consider the question whether history as a serious of events is continuous or discontinuous. Historical research usually tries to narrow the causal distance between events, to fill in more details of causes and effects. But there invariably remain “gaps.” It is a question of “free […] choice” whether we highlight or neglect these gaps (Simmel [1909] 2001: 64). Simmel took his anti-realist position to be part and parcel of a neoKantian approach to history. His goal was not just to analyze our creativity in forming experience in the case of natural science—here Kant had shown the way—but to do so also in the case of history—here much of the work remained to be done. “Naturalism” places us in a deterministic causal nexus of physical and biological processes; “historicism,” in a deterministic causal sequence of economic and other sociopolitical forces. Kant rescued us from naturalistic determinism by showing that nature is itself “the achievement of the soul.” Simmel ([1907] 2015: 229) aimed to achieve a similar result concerning historicism by showing “that it is the mind itself that forms […] history according to its own categories.” Although the wording of these 1907 passages is more resolutely neo-Kantian than anything in the 1892 edition of Probleme der Geschichtsphilosophie, the theme of history as construction and the terms “historical” or “relative a priori” were present already in the earlier text. (I shall return to these concepts in Section 1.5.) The main shift was not from a non-Kantian to a neoKantian position; it is perhaps better described as a shift from a psychological neo-Kantianism to a neo-Kantian position that puts strict limits on the role of psychology in historiography (cf. Section 1.2). Particularly interesting for following Simmel’s neo-Kantian view of historiography is a lengthy paper from 1917–18 entitled “Die historische Formung”. It identified categories that Simmel took to be constitutive of historical research. To begin with, all historical material involves the category of “Erlebnis”; that is, pre-theoretical, lived experience. Erlebnis is “the most primordial way in which a conscious content somehow objectively stands opposite a conscious subject” (Simmel [1917–18] 2000, 321). Whereas other sciences aim to extract intellectual contents from the Erlebnis and to leave the Erlebnis behind, history instead “transforms the Erlebnis-character of contents.” Historical research gives the historical actors’ Erlebnisse “new forms” (ibid.: 322). One such form is concepts that organize the historical records into series of events. Simmel here was thinking of “special histories”: series of events

Georg Simmel on Historical Understanding

23

that are picked out by one guiding concept, like “foreign and domestic politics, … religion and fashions, medicine and the art, worldviews and technology” (ibid.: 324). History not only isolates these series; it also, and by the same token, constructs historical subjects as their correlates (ibid.: 330– 2). Another element of the historical a priori is concepts meant to achieve “unity” or “completeness.” They correct for the historian’s “fragmentary” knowledge of historical actors, rendering these actors as instances of “general concepts” or “types” (ibid.: 333). Simmel was also intrigued by historians’ construction of time. For example, depending on the kind of historical narrative, “the present” can comprise anything from a split second to several decades (ibid.: 350). Furthermore, whereas everyday life looks forward from the present, historical research looks forward only within stretches of time that as a whole are constituted as the past. Usually, these past stretches of time are built around “fixed points” like wars or religious conversions. Finally, historians frequently “compress” (verdichten) a longer time period, or set of contemporaneous elements, into one Zustand (condition, state) (ibid.: 357–67). Like all the other mentioned categories, the positing of a Zustand is “a deed of the observer and not predetermined in the objective unity of appearance” (ibid.: 361).

1.5. RELATIVISM German eighteenth- or nineteenth-century authors on historiography and understanding occasionally committed themselves to relativistic theses. In the eighteenth century, Johann Martin Chladenius argued that historical writing is inevitably perspectival and that there can be no impartial or absolute standards in historical writing (Beiser 2011: 45–51). Droysen wrote in 1857 that “the known historical truth is of course only relative truth; it is true as the narrator sees it, from his standpoint, from his point of view, from his stage of education; in a transformed time it can appear differently; each age has to rework, to rethink, the totality of history anew.” And Droysen also announced: “I do not want to appear to have more or less than the relative truth of my standpoint. […] One must have the courage to confess this limitation, and to console oneself that the limited and particular is richer than the universal” ([1857] 1977: 230, 287).3 Dilthey too had obvious relativistic leanings; for example, when he emphasized the plurality of radically different, self-contained worldviews. Still, Dilthey frequently declared that his overall project was to overcome relativism. Few were convinced at the time, and few are convinced now. Beiser (2011: 363) suggests that Dilthey—given his philosophical toolbox—could avoid the Scylla of relativism only by returning to the Charybdis of the philosophy of

24

The History of Understanding in Analytic Philosophy

history: declaring each worldview a necessary step in the unfolding of spirit (cf. Kinzel 2019). Simmel was outspoken in his advocacy of various forms of relativism (especially in the Philosophie des Geldes, 1900/1989). His historiography was no exception. Above, I quoted a passage implying that some historical periods might be incommensurable. A petit bourgeois historian cannot comprehend Napoleon, Goethe, or Nietzsche; and Europeans today are unable to understand “the soul of the Oriental,” the Athenians of the Persian Wars, or mediaeval monks (Simmel [1918] 2015: 155–6). Elsewhere Simmel ([1907] 2015: 287) spoke of different “specialised historical sciences” having different “concepts of truth.” He tried to make this idea palatable with pairwise comparisons between ways of writing history: thus the “history of philology” aims for detailed, exact, and narrow descriptions, whereas the social “history of mores” uses a broad brush; the “criteria of truth […] in both cases are different” (ibid.: 288). Or compare “history of technology” and “history of domestic politics”: the former has a clear and limited subject matter in the form of material objects, whereas the latter has to consider a wide range of different elements and strata (ibid.: 289). In the same context, Simmel proposed that even “the immediately experienced” is a historically variable “category” rather than something timeless or universal (ibid.: 291). Similar sentiments were already in play in 1892 when Simmel introduced the notion of a “relative a priori” as the condition of the possibility of historical research ([1892] 1989: 304). In so doing, he widened the category of the “a priori” beyond its original, Kantian, scope: In every moment, praxis as well as theory use combinatory forms for empirical material; praxis as well as theory use the peculiar plasticity of the mind that allows it to cast whatever content it encounters in the most varied of structures; and this is done by ordering, determining or emphasizing this content. (Simmel [1892] 1989: 304) Kant’s transcendental structures—space, time, and the categories—were joined in the realm of the generalized a priori by “combinatory forms” that apply more locally, to “a particular region of life, or a particular science.” The latter are the realm of the relative a priori (ibid.: 305). As far as history is concerned, a variety of relative a priori forms are important. A first kind works largely below the threshold of consciousness and concerns conditions for the attribution of mental states. We unreflectively assume the existence of mental processes and states in others just like we fill in the blind spot in our field of vision. In everyday life with its practical concerns, our relatively

Georg Simmel on Historical Understanding

25

simple inferences concerning others’ mental processes and behaviors are usually successful. More complex mental phenomena are less easy to understand and predict. To support this claim, Simmel offered two short case studies of historians (Heinrich von Sybel, Georg Friedrich Knapp) attributing mental states to historical actors on the basis of source material which, as a matter of fact, under-determined these attributions (ibid.: 306– 12). Simmel also highlighted historians’ practice of ascribing unconscious mental states to whole ethnic groups (“the dark longing of German people for Italy,” ibid.: 315). As Simmel had it, two kinds of a priori conditions are thus in play when historians do their work: the general or universal “powers and categories that naturally reside in their souls” and their individually and socially varying previous “actual experiences.” The latter are responsible for the petit bourgeois historian’s inability to understand Napoleon or Bismarck (ibid.: 325). Simmel also noted that sometimes historians’ background beliefs include prejudices which lead to distortions in the historical accounts. While it is often possible to correct for such prejudices, it would be a mistake, Simmel insisted, to assume that we could ever be free of all such subjective elements: “Just as no human being is a human being in general; just as no human being has only those properties which are shared by all human beings; so also there is no knowledge which is knowledge in general and thus is due only to the exercise of the most general a priori forms of thought” (ibid.: 327). There are thus both enabling and distorting relative a priori categories. The latter can and should be corrected since they lead the historian to ignore important sources or to misconstrue their message (ibid.: 333–4). Overall, Simmel’s relativism in historiography is Kantian insofar as it centers around the category of the a priori. Different historians in different historical periods and different specialties work with different conditions for the possibility of historical understanding, and these different specialties each that have their own concepts of truth.

1.6.  AGAINST RADICAL HISTORICISM It is inviting to conclude from the above that Simmel’s relativism was indicative of a commitment to historicism as well. Simmel himself thought otherwise. According to this rendering of the view, historicism reduces all forms of understanding to “historical derivation.” The paradigmatic culprits were Kuno Fischer and Friedrich Paulsen, whose famous battle cry in Kant scholarship famously had been: “To explain Kant is to derive him historically” (Paulsen 1875: iii). Simmel misquoted this as “to understand [sic!] Kant is

26

The History of Understanding in Analytic Philosophy

to derive him historically” ([1907] 2015: 261, [1918] 2015: 171). Simmel attacked this position as follows. Assume we read (at the time of Simmel’s writing) that a politician from the House of Hanover hated Bismarck. Hatred is a feeling familiar to us all, and we assume it to be at least to some extent universal or “transhistorical” (Simmel [1918] 2015: 163). But there was also something specific about the Hanoverian’s hatred of Bismarck; to get to this specificity, we have to look at the historical situation: to understand the Hanoverian’s hatred “historically,” we need to know about the war of 1866, the fact that Prussia annexed Hanover, and the passions triggered by these events in Hanover. Simmel’s focus was on the interplay of the transhistorical and historical elements. He was adamant that in order to understand the hatred historically, we need to first grasp hatred and its development in the transhistorical, timeless sense. Only subsequently can we reach historical understanding; this occurs when we situate discontinuous, timeless, transhistorical elements within a “continuous stream of life” (ibid.: 164). That is to say, to understand historically is to place the mental phenomena in a broader context, typically the context of the life of an individual or social group. Simmel rejected the alternative, according to which understanding a human deed or creation results from explaining it causally on the basis of earlier events. Life does not consist of mental atoms; to think that it does is to confuse logic and psychology. Historians need to grasp the “dynamic connection, fusion, unification of the manifold.” Understanding this dynamic process does not consist of moving from content to content; it is rather an appreciation of the “stream of life” with its “waves.” These waves are what seemingly independent contents turn out to be. Simmel thought that individual deeds and creations are best thought of as such waves (ibid.: 165–6). Put differently, for Simmel historicism fails to appreciate the crucial distinction between two types of understanding: one type of understanding is “completely free of history,” while the other is historical through and through (ibid.: 166). Simmel explains the distinction also in the case of art. Take a painting. We can study historically what the painter tried to express, or sought to tell her audience. Ultimately, only one answer to these historical questions can be true. However, as soon as the work of art is finished, it gains an independent “objective content” (Simmel [1918] 2015: 170) and thus the potential for a plethora of different interpretations based on the work’s fixed structure or permanent properties (ibid.: 167). Simmel used an analogy to press home this point. A human creation (e.g., a painting) is like a “riddle.” Its author may well have intended just one solution to work for it. And yet, later, other people may discover further solutions to the same riddle. When

Georg Simmel on Historical Understanding

27

this happens, Simmel insisted, the new solutions are just as correct as the solution envisaged by the author. Once the act of creation has “found the form of the objectivized spirit,” then all the different ways of understanding are equally to be accepted. Thus the “immanent understanding of the work of art” is “infinitely variable just like the feelings it triggers.” Essentially the same analysis also applies to the understanding of the purposes of machines, political constitutions, or individual laws (ibid.: 167–8). “Radical historicism,” as Simmel construed it, focuses exclusively on a work’s causal “conditions and stages of its temporal origins and development,” leaving no room for the “timeless qualities of being” (ibid.: 170). Historicists failed to see that their “seemingly pure historical understanding constantly uses the transhistorical, objective [sachlich] understanding.” Historicists therefore did not recognize their obligation to methodologically justify their use of transhistorical understanding (ibid.: 171). The complex interplay of historical and objective forms of understanding can also be shown when it comes to understanding Kant’s development—the very area in which Fischer and Paulsen had insisted on “historical derivation” (ibid.). They hypothesized that Kant’s thinking went through a series of stages. In so doing, they overlooked, however, that each of the stages can and must be grasped by the historian at first in the abstract, independently of its historical realization. Simmel compared this to someone’s thinking through the premises and inference of a syllogism. The premises and the inference have, according to Simmel, an objective character; in that sense they are fully timeless and independent of the temporal sequence. In order to understand the psychological processes of someone’s reasoning in syllogistic fashion, we again need to first understand the objective content (ibid.: 172). It is the same with the historical understanding of Kant as he moves from rationalism to empiricism to his own transcendental philosophy. If we did not understand the meaning of this sequence in and of itself, then we would not understand it within Kant’s thinking either. Thus, we can say both that to understand Kant we need to derive him historically and to derive Kant historically we need to understand him a-historically (ibid.: 173). There is still a further complication. Simmel wanted his readers to be puzzled by the claim that Kant “developed” across, or through, the mentioned three stages. Simmel asked how this could be since “is not each stage complete in itself?” He answered as follows. We bring development into the sequence insofar as we are implicitly committed to the “postulate of an ideal […] constructed subject beneath the series of standpoints.” It is the psychological continuity of this ideal subject, its continuity across the stages, that makes the stages steps of one and the same development (ibid.: 174). Something similar happens when we postulate a development in

28

The History of Understanding in Analytic Philosophy

the history of art. Here too we need the methodological step of postulating a subject, “an idea structure,” that constitutes the continuity below the many different discontinuous works of art. Even art as a whole requires an ideal “hypostatized” subject as an auxiliary term. And when we speak of love or hate quite generally, we again tacitly assume “an ideal bearer, a life as such.” These different feelings or emotions are once more like “waves” in the stream of life, each with their own specific height and “rhythm” (ibid.: 175). Simmel did not say much about the origin of the idea of an ideal subject. Still, he pointed out that the “vitality of such ideal subject is but a transformation or objectivization of a vitality that we feel in ourselves, albeit feel as something trans-individual, something of which we are merely an example.” After all, we feel in ourselves “at least a formal telos, a realisation of potentials, and unfolding of seeds that […] we are” (ibid.: 175–8). Above, we encountered Simmel’s view according to which to understand a content historically is to locate it within a totality of life. This theme resurfaces in a new form when we are concerned with understanding a development across objective contents. In the present case, Simmel thought that we need the totality of life, not of a real but of an idealized subject. We thus end up with not only two forms of understanding: objective and historical; we also are implicitly committed to two kinds of underlying subjects: one real; the other ideal.

1.7.  UNDERSTANDING AND THE PHILOSOPHY OF LIFE Finally, late in his life, Simmel tried to bring his account of historical understanding in line with his Lebensphilosophie (philosophy of life)— which, very roughly, marked a third phase of his work (the first two being the naturalism of Einleitung in die Moralwissenschaft and the neo-Kantian position of the 1905 and 1907 editions of Probleme der Geschichtswissenschaft). Here too Simmel was close to Dilthey ([1910] 2006). Simmel contrasted the philosophy of life—“the organicist or vitalist standpoint”—with “the mechanistic standpoint.” Mechanistic-atomistic thinking is unable to grasp “the unity and wholeness of living beings”; it splits life into separate pieces and then struggles to put them together again. This is why mechanistic-atomistic thinking is unable to appreciate “understanding as a primitive phenomenon” which involves at least two whole human lives, the life of the historian and the life of the historical actor. Mechanistic-atomistic thinking also misses the “creative aspect of the process of understanding” (Simmel [1918] 2015: 178). It inclines one

Georg Simmel on Historical Understanding

29

to think of understanding as a matching, or even identity, of two mental states of two different subjects. The historian understands the historical subject if the historian’s mental states somehow repeat the mental states of the historical subject. This explains why mechanistic-atomistic historical thinking renders historical understanding as a mere “repetition of ‘wie es eigentlich gewesen.’ ” Ranke thus turns out to be a mechanistic-atomistic historian. Interestingly enough, Simmel does not think that the difference between the two standpoints can be decided by argument; this is because the choice between them depends on one’s “Weltanschauung as a whole and in its depth” (ibid.: 179). One key feature of the “philosophy of life” perspective is to highlight interesting differences between understanding ourselves and understanding others, especially when the latter are historical actors. Our understanding of ourselves tends to be fragmentary: we know too many disparate things about ourselves for them to fit into one coherent picture. Since we know less about others, it is much easier to achieve a unified picture of them. Moreover, in our stream of experiences, important and unimportant, essential and accidental elements are mixed together and hard to separate. Again, this is different for (historical) others. Here principles of selection are at work from the very start. The historical person (qua character in the historian’s narrative) therefore “lives […] in a specially constructed, discontinuous time” (Simmel [1909] 2001: 63). Simmel’s philosophical focus on life also included reflections on the life of social groups and the role of historical understanding within them. In this vein, Simmel commented on the relationship between social “interaction” (Wechselwirkung) and “heritage” (Überlieferung). He compared social interaction with the way in which material masses influence each others’ movement. The difference between heritage and social interaction is that the former follows a single line from A to B, whereas the latter displays many feedback loops. Society as a whole comprises the sum of both processes. Society is a historical structure insofar as heritage is essential. In its heritage, the past of society still has an “effective reality” (ibid.: 65). Although Simmel did not put it in so many words, it is clear that for him interaction as well as heritage are mediated by processes of understanding. Heritage is the resource, medium and result of historical understanding. The most original feature of the life-philosophical perspective is that it urges us to see historical-scientific understanding as both growing out of and as sometimes contradicting everyday life and understanding. Life in its everyday mode evolves and unfolds from the present into the future. But when we engage in historical research “the direction of the stream of life” is turned around and attention flows from the present to the past: “History

30

The History of Understanding in Analytic Philosophy

[qua historical research] […] directs the gaze backwards (or to be more exact: the backwards-directed gaze creates history), against the stream of life that carries the gaze forward.” Put differently, in everyday life the past is a resource for the future, whereas when engaging in history, “life turns back in itself ” (Simmel [1917–18] 2000: 352–4). Historical understanding is a further development, or refinement, of everyday understanding. The latter always already works on the “remembered and inherited material of life”: it analyzes and synthesizes this material, orders it with concepts, or changes emphasis. Everyday understanding thereby uses, for practical purposes, categories which later prove of great theoretical significance in historical research. History as a science results when we detach the material of life and its structures from the original purposiveness of practice. In this way, we construct pictures of past life based on a “free-floating theoretical interest.” Thus, in everyday life, we are “as it were, embryonic historians of ourselves.” As scientific historians, we “perfect, and render absolute, attitudes and forms of structuring of prescientific life” (Simmel [1918] 2015: 153–4). Finally, some previously mentioned ideas also belong in the present context. Particularly significant is the thought that the “you” is a “primitive phenomenon just like the I” (ibid.: 160). Human life is essentially focused on human interlocutors intertwined in processes of understanding. Indeed, ultimately, we are all but “waves” in the “stream of life” (ibid.: 166).

1.8. CRITICISM Up to this point, I have sought to reconstruct and interpret Simmel’s theory of understanding. In this last section, I turn from reconstruction to criticism and evaluation. In this endeavor, I take as my guides and interlocutors, reviewers and other commentators who were Simmel’s contemporaries. I want to assess Simmel’s theory in light of standards and criteria prevalent in the early twentieth century; not in light of twenty-first-century historiography or philosophy of mind. Simmel’s work on historical understanding was, during his lifetime anyway, his most influential and valued contribution in philosophy and the social sciences. In part, this may have been because his writings on history and historiography were “conventional” in form, scope and style of argument— something that can hardly be said about his Einleitung in die Moralwissenschaft, the Philosophie des Geldes, or his books on Goethe or Rembrandt. It is also striking that the two main editions of Probleme der Geschichtsphilosophie were reviewed or otherwise discussed by leading philosophers, theologians, educationists, historians, and social scientists in the German-speaking world

Georg Simmel on Historical Understanding

31

and beyond: from Heinrich Rickert to Max Frischeisen-Köhler, from Franz Eulenburg to Max Weber, from Friedrich Meinecke and Kurt von Breysig to Ernst Bernheim or Otto Hintze, from Eduard Spranger to Otto Wach, and from Arvid Grotenfeld to David Morrison. Even Mind published separate review essays of the first and second editions of Probleme der Geschichtsphilosophie in the very year of their publication. As far as praise was concerned, Meinecke applauded Simmel’s “epistemological studies of the individual historian’s a priori” and his likening of history to art (1894: 71). Von Below (1905: 821), a fierce opponent of Karl Lamprecht, found nothing to object to in Simmel’s book (for Lamprecht’s attack on von Below, see Lamprecht 1899), and Wach (1926: 26) praised it for offering the first detailed analysis of historical understanding. Spranger (1906: 120) called Probleme “a systematic deed of fundamental significance.” Hintze (1906: 811) wrote that Simmel had been “entirely successful in demonstrating the subjective conditions of scientific history.” Bernheim (1908: 192) classified Simmel’s historiography as “convincing and delicate,” and Weber ([1903–6] 1988: 92) complimented him on the “logically most developed approach to a theory of ‘understanding.’ ” Eulenburg (1909: 171) went further by insisting that “Simmel’s epistemological stance taken as a whole seems to me the only acceptable one […;] it is the only one that makes history as a science ‘possible.’” Turning from praise to critical comments, first of all, several commentators objected to the essential role of psychology especially in Simmel’s early ([1892] 1989) historiography. This objection often culminated in the— at the time—all too familiar charge of “psychologism”; that is, granting psychology an inappropriate central position in a realm where other methods or sciences seem more appropriate.4 There is no doubt that the first edition of Probleme couched the analysis of the historians’ relative a priori very much in psychological terms. After all, it announced loud and clear that “psychology is the a priori of the historical sciences.” Heinrich Rickert (1902: 535) objected that this perspective “prevents a logical understanding of historical research.” For Weber, the problem was not confined to Simmel’s early work. In particular, Weber was unhappy with Simmel’s “psychologistic formulations” concerning the understanding of linguistic utterances. As Weber saw it, Simmel collapsed the normative question regarding what constitutes grasping a sentence meaning into the “causal question” of how an utterance of the proposition came about. Simmel’s account thereby missed the important feature that understanding a proposition is inseparable from “taking a stance” concerning its validity or truth (Weber [1903–6] 1988: 94–5). Weber granted however that Simmel came close to grasping the distinction in question when he distinguished between the “objective ‘understanding’ of the meaning of an

32

The History of Understanding in Analytic Philosophy

utterance” from “the subjective ‘interpretation’ of the motive of a (speaking or acting) human being” (ibid.: 93). Given the influence of neo-Kantian and phenomenological antipsychologism at the time, Simmel’s generous use of psychology was obviously salient (Husserl [1900] 1975; cf. Kusch 1995). And Weber is entirely right to observe that the three editions of Probleme did not recognize the normative elements of understanding. Of course, as we have seen, Simmel sought to separate history from both the causal analysis of psychology and the inference-centered perspective of logic. But he watered down the distinction immediately by claiming that historical understanding is primarily preoccupied with reenacting meaning (Sinn). That said, one might wonder though whether Simmel did not increasingly recognize Weber’s point himself. At least the 1918 attack on “radical historicism” is focused on distinguishing between “deriving Kant historically” and reconstructing Kant’s philosophy by our normative lights. Second, objections to Simmel’s disregard for normative questions also surfaced when critics took exception to his historiographical anti-realism. To be sure, almost every reviewer warmly agreed with Simmel’s Kantian project of defeating historicist determinism by emphasizing the historian’s alleged freedom in constituting the subject matter of history. After all, this move mirrored what Friedrich Albert Lange had done with respect to materialism in the 1870s (see Lange 1866; Frischeisen-Köhler 1920: 14). Still, many commentators felt that Simmel had gone too far by likening the historian to the artist and by insisting on the historian’s total and radical freedom in “painting […] historical pictures.” For example, the distinguished social historian Otto Hintze thought that Simmel was right to insist that the historian is no “copyist” of an independent historical reality. Nevertheless, Hintze rejected the contention that there is a strong similarity between historical writing and the artist’s unconstrained creativity. Instead, Hintze urged a parallel between historian and “translator.” Just as there are rules of method for adequate translations, so historians had long since developed a “genuine historical methodology.” Ranke’s talk of “eliminating himself ” and aiming to recover “how things really had happened” were expressions of such necessary methodological attitude. At the same time, Hintze accepted that different subfields of historical research were more or less constrained by methodological rules. In this vein, he distinguished between “epic” historical writing trying to recreate past life for its own sake and “didactic” historical research aiming to explain or change a presently existing institution. The latter field interacted closely with “the sciences of the state” and economics and was less constrained by traditional methodologies (Hintze 1906: 812). Other historians agreed with Hintze’s criticism.5

Georg Simmel on Historical Understanding

33

Philosophers, social scientists and theologians shared the historians’ unease. Frischeisen-Köhler (1920: 34–6) complained that Simmel overemphasized “the creative nature of the sovereign subject,” thereby destroying the “objectivity of historical knowledge” and its “connection with reality” (cf. Wach 1929: 86, 1933: 37). Frischeisen-Köhler was a student of Dilthey and perhaps he thought that Simmel had learnt too little from Dilthey’s account of the competing systems of rules of hermeneuticalhistorical understanding (Dilthey [1900] 1990). Unsurprisingly, Rickert (1896: 302) likewise opposed Simmel’s reluctance to identify epistemological norms and values involved in doing history. Weber agreed with Rickert: while the historian is free in the selection of values guiding his research, methodological rules govern the identification of causal links ([1903–6] 1988: 92, 124). Moreover, methodology and values did not just constrain; they also enabled understanding—they made it unnecessary to postulate a special biological “Platonist anamnesis” in the case of the historian-genius (ibid.: 100). One does not have to sympathize with the specific historiographical or epistemological position of the mentioned critics in order to agree with them that Simmel overstated the parallel between historian and artist and that he said too little about historical methodology. To be fair to Simmel, it is worth distinguishing two elements here: the project of a descriptive epistemology of historical research and the insistence on the historian’s radical freedom vis-à-vis the historical record. The latter is indeed a problematic exaggeration. The former, however, is not problematic as such, and it and fits with important strands in today’s philosophy of science. After all, there are not many philosophers in the twenty-first century who believe themselves able or entitled to prescribe what scientists or historians should do methodologically. This does not mean of course that philosophers must ignore scientific norms, but they are to be discussed as empirical phenomena like any others. Third, above we encountered two reasons for laying the psychologism charge at Simmel’s door: his emphasis on mental states in the theory of linguistic understanding and his psychological rendering of the historian’s “relative a priori.” Franz Eulenburg—a social scientist, economist, and student of Gustav Schmoller—pointed out a further (“psychologistic”) overemphasis on psychology in Simmel’s historiography; to wit, Simmel’s insistence that the mental states and processes of “historical personalities” are the primary or ultimate subject matter of all of history. As Eulenburg saw it, this assumption was problematic for a number of reasons. To begin with, it contradicted Simmel’s general Kantian anti-realism. Eulenburg fully endorsed the latter position: “It has always seemed

34

The History of Understanding in Analytic Philosophy

impossible [to me] to be a phenomenalist concerning the natural sciences, and a naïve realist concerning the humanities.” This “dualism” had become obsolete (Eulenburg 1909: 170–1). Although Simmel insisted on just this point himself, he backpedaled to the old dualism in treating the “historical personality” with their mental states as “the ens realissimum, the ‘thing in itself ’ of history” (ibid.: 180). Eulenburg suspected an “ethical moment” in Simmel’s reasoning: “the intention to preserve the freedom of the personality […] vis-à-vis nature and history.” Eulenburg rejected the intrusion of ethical considerations into metaphysical and historiographical reasoning, and thus he saw no grounds for Simmel’s “stress on the personality as such […] as the ultimate constituent of history, the ultimate secret for knowledge” (ibid.: 171). Eulenburg also rejected a Diltheyan element in Simmel’s thinking— to wit, the notion that in the humanities, the subject matter is somehow given more immediately than in the natural sciences: “I do not ‘experience’ unfamiliar forms of heroism more directly than I ‘experience’ the temperature of the sun, or the nesting of birds”—there might well be differences between these cases in degree and strength of sensations, but these did not amount to differences in kind (ibid.: 173). Moreover, the claim that the core of history are mental states and processes contradicted historians’ interest in un-personal, or transpersonal, powers, such as buildings, churches, monuments, technological artefacts, institutions, laws, language, expressions, corporations, and technologies. Although psychological processes are crucial for generating such entities, they later aquire “a meaning that prevails in good part independent of consciousness” (ibid.: 174). Eulenburg suggested calling them “psychological fossils” and called the “functional connection” between such fossils and human mental states and processes the “ultimate problem of history” (ibid.: 175). Going beyond psychological fossils, Eulenburg also highlighted the role of climate, geography, and biological inheritance (e.g., of illness) as important factors and determinants in history (ibid.: 175–8). Finally, Eulenburg did not wish to deny that one can write history by focusing on historical personalities. Biographies have their place and value. But Eulenburg accused Simmel of reducing all of history to biography— again implicitly situating Simmel very much in Dilthey’s school. At the same time, Eulenburg insisted that biographical writing must include an explicit answer to the question of how the historical personality is possible and why it is significant. Why does it matter that Napoleon’s various actions were carried out by one and the same personality rather than many different actors (ibid.: 182). Eulenburg’s critique seems to me to be spot-on.

Georg Simmel on Historical Understanding

35

1.9. SUMMARY In this chapter, I have tried to reconstruct the central tenets of Simmel’s theory of historical understanding and to connect them to the thoughts of some predecessors and contemporaries, such as the neo-Kantians, Dilthey, Droysen, and Ranke. Clearly, Simmel’s account was very much a child of its time and part of larger conversations about the relationship between the natural sciences and the humanities, the role of the new psychology (i.e., “psychologism”) and evolutionary biology, the place of philosophical analysis, materialism and historicism, freedom and determinism, or the role of ethical considerations in history, epistemology, or metaphysics. Most commentators agreed that Simmel’s theory of understanding was an important and original step in historiography and the philosophy of mind more generally. The fact that his account served as a central foil for influential historians (e.g., Meinecke, von Below, Hintze), philosophers (e.g., Rickert, Frischeisen-Köhler) and social scientists (e.g., Eulenburg, Weber) attests to its significance and value. At the same time, it is clear that Simmel’s proposals were not (fully) satisfactory for conceptualizing work in the new forms of social and economic history, for distinguishing between good and bad historical writing, and for preventing the intrusion of dubious ethical principles. Still, a careful study of Simmel’s account is necessary, not only important, for understanding the debates of early-twentieth-century philosophers and historians. Simmel’s account is also intriguing today insofar as it constitutes something of a clear-cut version of what a relativist, constructivist, anti-realist historiography might be.

NOTES 1 It is striking that Simmel wrote extensively about understanding in historiography, but very little about its role in the social sciences; cf. Gerhardt (1998). Geßner (2003) situates Simmel’s philosophy of history in the context of his thinking about culture and its investigation. An invaluable source for understanding Simmel’s background and his relationship with Dilthey is Köhnke (1995). Beiser (2011) situates Simmel within the tradition of historicism, but he says too little about Simmel’s theory of understanding. Choi (2000) is a helpful though slim account of Simmel’s theory and its relation to Weber’s, Dilthey’s, and Droysen’s proposals. 2 I use the English translation of Beiser (2011: 289). 3 I use the English translation of Beiser (2011: 305, 306). 4 See, for example, Rickert (1902: 534–5), Gottl (1907: 269), and Weber ([1903– 6] 1988: 92–4). 5 See, for example, von Breysig (1893) and Bernheim (1908: 629), Meinecke (1894) being the exception that proves the rule.

36

The History of Understanding in Analytic Philosophy

REFERENCES Beiser, F. C. (2011), The German Historicist Tradition, Oxford: Oxford University Press. Bernheim, E. (1908), Lehrbuch der historischen Methode der Geschichtswissenschaft, 6th ed., Leipzig: Duncker & Humblot. Choi, H.-K. (2000), “Das historische Verstehen bei Simmel—ein Vergleich mit Droysen, Dilthey und Weber,” Simmel Studies, 10: 108–26. Dilthey, W. ([1883] 2017), Einleitung in die Geisteswissenschaften, ed. K. M. Guth, Berlin: Hofenberg Digital. Dilthey, W. ([1894] 1990), “Ideen über eine beschreibende und zergliedernde Psychologie,” in Gesammelte Schriften, Vol. 5, ed. G. Karlfried, 139–240, Göttingen: Vandenhoek & Ruprecht. Dilthey, W. ([1900] 1990), “Die Entstehung der Hermeneutik,” in Gesammelte Schiften, Vol. 5, ed. G. Karlfried, 317–31, Göttingen: Vandenhoek & Ruprecht. Dilthey, W. ([1903] 1990), “Rede zum 70. Geburtstag (1903),” in Gesammelte Schriften, Vol. 5, ed. G. Karlfried, 7–9, Göttingen: Vandenhoek & Ruprecht. Dilthey, W. ([1910] 2006), “Der Aufbau der geschichtlichen Welt in den Geisteswissenschaften,” in Gesammelte Schriften, Vol. 7, ed. B. Groethuysen, Göttingen: Vandenhoek & Ruprecht. Droysen, J. G. ([1843] 1925), “Theologie der Geschichte: Vorwort zu Geschichte de Hellenismus II,” in Historik, ed. E. Rothacker, 87–97, Halle: Niemeyer. Droysen, J. G. ([1857] 1977), Historik: Die Vorlesungen von 1857, ed. P. Leyh, Stuttgart-Bas Cannstatt: Fromann-Holzboog. Droysen, J. G. ([1868] 2013), Grundriss der Historik, in den Buchausgaben von 1868, 1875 und 1882, ed. H. Schleiff, Digitale Edition, Berlin: Diachron. Eulenburg, F. (1909), “Neuere Geschichtsphilosophie,” Archiv für Sozialwissenschaft und Sozialpolitik, 29: 168–97. Frischeisen-Köhler, M. (1920), “Georg Simmel,” Kant-Studien, 24: 1–51. Gerhardt, U. (1998), “Simmels Beitrag zur Methodologie des Verstehens in der Soziologie,” Simmel Newsletter, 8: 114–37. Geßner, W. (2003), Der Schatz im Acker: Georg Simmels Philosophie der Kultur, Weilerswist: Velbrück. Gottl, F. (1907), “Zur sozialwissenschaftlichen Begriffsbildung,” Archiv für Sozialwissenschaft und Sozialpolitik, 24: 265–326. Hintze, O. (1906), “Review of Simmel, Probleme der Geschichtsphilosophie, 2nd edition,” Jahrbuch für Gesetzgebung, Verwaltung und Volkswirtschaft im Deutschen Reich, 30: 809–14. Husserl, E. ([1900] 1975), “Logische Untersuchungen, Vol. 1, Prolegomena zur reinen Logik,” in Husserliana XVIII, ed. E. Holenstein, The Hague: Nijhoff. Kinzel, K. (2019), “The History of Philosophy and the Puzzles of Life: Windelband and Dilthey on the Ahistorical Core of Philosophical Thinking,” in M. Kusch, K. Kinzel, J. Steizinger, and N. Wildschut (eds.), The Emergence of Relativism, 26–42, London: Routledge. Köhnke, K. C. (1995), Der junge Simmel in Theoriebeziehungen und sozialen Bewegungen, Frankfurt am Main: Suhrkamp. Kusch, M. (1995), Psychologism: A Case Study in the Sociology of Philosophical Knowledge, London: Routledge.

Georg Simmel on Historical Understanding

37

Lamprecht, K. (1899), Die historische Methode des Herrn von Below, Berlin: Gaertner. Lange, F. A. (1866), Geschichte des Materialismus und Kritik seiner Bedeutung in der Gegenwart, Iserlohn: Baedecker. Meinecke, F. (1894), “Review of Simmel, Probleme der Geschichtsphilosophie, 1st edition,” Historische Zeitschrift, 72: 71–3. Otis, L. (1994), Organic Memory: History & the Body in the Late Nineteenth and Early Twentieth Centuries, Lincoln: University of Nebraska Press. Paulsen, F. (1875), Versuch einer Entwicklungsgeschichte der Kantischen Erkenntnistheorie, Leipzig: Fues. Rickert, H. (1896), Die Grenzen der naturwissenschaftlichen Begriffsbilding: Eine logische Einleitung in die historischen Wissenschaften, Tübingen: Mohr. Rickert, H. (1902), Die Grenzen der naturwissenschaftlichen Begriffsbildung: Eine logische Einleitung in die historischen Wissenschaften, Freiburg im Breisgau: Mohr. Simmel, G. ([1891–2] 1989–91), “Einleitung in die Moralwissenschaft. Eine Kritik der ethischen Grundbegriffe, Erster Band,” in Gesamtausgabe, Vol. 3, ed. K. C. Köhnke, Frankfurt am Main: Suhrkamp, 1989; “Zweiter Band” in Gesamtausgabe, Vol. 4, ed. K. C. Köhnke, Frankfurt am Main: Suhrkamp, 1991. Simmel, G. ([1892] 1989), “Die Probleme der Geschichtsphilosophie,” in Gesamtausgabe, Vol. 2, ed. H.-J. Dahme, 297–421, Frankfurt am Main: Suhrkamp. Simmel, G. ([1900] 1989), “Philosophie des Geldes,” in Gesamtausgabe, Vol. 6, ed. D. P. Frisby and K. C. Köhnke, Frankfurt am Main: Suhrkamp. Simmel, G. ([1907] 2015), “Die Probleme der Geschichtsphilosophie, 2nd and 3rd edition,” in Gesamtausgabe, Vol. 9, ed. G. Oakes and K. Röttgers, 227–419, Frankfurt am Main: Suhrkamp. Simmel, G. ([1909] 2001), “Beiträge zur Philosophie der Geschichte,” in Gesamtausgabe Vol. 12, 2nd ed., ed. R. Kramme and A. Rammstedt, 62–9, Frankfurt am Main: Suhrkamp. Simmel, G. ([1917–18] 2000), “Die historische Formung,” in Gesamtausgabe, Vol. 13, 2nd ed., ed. K. Latzel, 321–69, Frankfurt am Main: Suhrkamp. Simmel, G. ([1918] 2015), “Vom Wesen des historischen Verstehens,” in Gesamtausgabe, Vol. 16, 2nd ed., ed. G. Fitzi and O. Rammstedt, 141–80, Frankfurt am Main: Suhrkamp. Spranger, E. (1906), “Review of Simmel, Probleme der Geschichtsphilosophie, 2nd edition,” Philosophische Wochenschau, 1: 120–4. von Below, G. (1905), “Review of Simmel, Probleme der Geschichtsphilosophie, 2nd edition,” Zeitschrift für Sozialwissenschaft, 8: 819–21. von Breysig, K. (1893), “Review of Simmel, Probleme der Geschichtsphilosophie, 1st edition,” Literarisches Centralblatt, 18: 636–7. Wach, J. (1926), Das Verstehen: Grundzüge einer Geschichte der hermeneutischen Theorie im 19. Jahrhundert: Band I: Die grossen Systeme, Tübingen: Mohr. Wach, J. (1929), Das Verstehen: Grundzüge einer Geschichte der hermeneutischen Theorie im 19. Jahrhundert: Band II: Die theologische Hermeneutik von Schleiermacher bis Hofmann, Tübingen: Mohr. Wach, J. (1933), Das Verstehen: Grundzüge einer Geschichte der hermeneutischen Theorie im 19. Jahrhundert: Band III: Das Verstehen in der Historik von Ranke bis zum Positivismus, Tübingen: Mohr. Weber, M. ([1903–6] 1988), “Roscher und Knies und die logischen Probleme der Nationalökonomie,” in Gesammelte Aufsätze zur Wissenschaftslehre, 7th ed., ed. J. Winckelmann, 1–145, Tübingen: Mohr.

38

CHAPTER TWO

Understanding, Psychology, and the Human Sciences: Dilthey and Völkerpsychologie LYDIA PATTON

2.1.  INTRODUCTION: TRADITIONS At the end of the eighteenth century, in the Metaphysical Foundations of Natural Science, Kant argued that the only “true” sciences were those that could be given a mathematical foundation. He exempted psychology from the true sciences, arguing that no mathematical analysis could be given of the occurrence of thoughts and ideas. Johann Friedrich Herbart and others in the tradition of empirical psychology (later including Wundt, Fechner, and Weber) took Kant’s words as a challenge, coming up with ways to quantify sensation and perception and to test them experimentally. But Kant’s more fundamental distinction between mathematically based, a priori sciences and inductive, a posteriori sciences based on observation lingered. The schools of neo-Kantianism that followed took differing positions on Kant’s stance. The Southwest School, or Baden School, argued for a dualism between nature and value, taking the position that beyond the objects present to us, there are also meaningful and valid relations of valuation. The realm of value can be treated scientifically as well (Šuber 2009: § 14.2.2).

40

The History of Understanding in Analytic Philosophy

The Marburg School had it that the realms of culture and ethics as well as of science are characterized by a transcendental analysis of facts, which are revealed in history (Cassirer [1912] 2005; Matherne 2015).1 Along with the positions developed by the scions of the Southwest and Marburg schools, there were influential engagements with the question of how sciences based on human experience might be conducted and founded. The first was the approach of Wilhelm Dilthey and its connections with the Historical School (e.g., Droysen, Ranke, Schlosser, Niebuhr); and the second was the Völkerpsychologie movement founded by Moses Lazarus and Chajim Steinthal,2 responding to (and inspiring) currents in cultural anthropology, linguistics, and “the sciences of the state.”3 Both these approaches, of Dilthey and of Völkerpsychologie, were concerned with psychology and with subjectivity. Both rejected Kant’s conception of the distinction between subject and nature (while retaining elements of the Kantian approach). The rejection of Kant’s distinction allowed for a conception of the subject as a sociohistorical self who exists only in relation with others and with history, rather than as a subject characterized fundamentally by the ability to interpret representations and thus make judgments. During Dilthey’s early career, he was friends with Lazarus and Steinthal, visiting their houses and going on walks with them (Feest 2007: § 5.2; Kluback 1956: Chs. 1 and 2). However, they were not in agreement about how to approach the cultural or human sciences. The account that follows will focus on two elements of the thought of Dilthey on the one hand and Lazarus and Steinthal on the other. First, Lazarus and Steinthal attempted to motivate an account of Völkerpsychologie based on the Volksgeist, a collective structure, or form, of rationality made manifest by a people or Volk. Dilthey argued, especially in his early work, for individual “descriptive” psychology as the foundation of the human sciences (Lessing 2016: 84; Kinzel 2018). Second, Lazarus and Steinthal argued that it is possible to identify the norms of willing to action that govern social phenomena and that such phenomena, including psychology and language, are—at the same time— natural and anomic, and governed by laws and logic. Dilthey rejected any account of psychology that took it to be law governed, even retrospectively, arguing that the “nexus of life” that is the ultimate basis of the human sciences cannot be reduced to any law-governed or explanatory relationship between the self, society, and nature. However, one can find a deep tension in Dilthey’s position here, evident in the development of his work over time (Makkreel 2020: § 1.1). The account below will explore this tension and its significance for the understanding of the subject and the role of psychology in the human sciences.

Understanding, Psychology, Human Sciences

41

2.2.  COLLECTIVE STRUCTURES AND HISTORY Lazarus and Steinthal argued that neither Johann Friedrich Herbart’s individual psychology nor Georg W. F. Hegel’s idealism nor Wilhelm von Humboldt’s theory of culture can independently account for the reciprocal influence between individual and culture. The critics of the Volksgeist, including Dilthey, described it as a soul (Seele) or collective mind of a culture. For instance, the philosopher Adolf Lasson interpreted the Volksgeist as a “folk soul” (Volksseele), “without which there could not be a folk psychology” (Klautke 2013: 29). This is wrong as an interpretation of Lazarus and Steinthal in the following sense: Völkerpsychologie is the study of the Volksgeist, which develops within the “objective spirit” or “general spirit” (Gesamtgeist) of a culture (Klautke 2013: 19; Kluback 1956: 34). As Steinthal (1855: 388) put it, “the individual cannot be completely comprehended without regard to the mental whole (die geistige Gesamtheit) in which it has been created and in which it lives” (trans. Klautke 2013: 18). The aim of Völkerpsychologie was not to describe the development and faculties of the soul, but rather to investigate the mind or spirit of a group with respect to the collective structures and artefacts that make it manifest.4 In 1860 the first edition of Lazarus and Steinthal’s journal Zeitschrift für Völkerpsychologie und Sprachwissenschaft was published. The first article was a manifesto written by Lazarus and Steinthal bearing the title “Introductory Thoughts on Völkerpsychologie, as an Invitation to a Journal for Völkerpsychologie and Linguistic Science.”5 The text set out the program of the school and laid out a common set of assumptions and goals for its disparate researchers. The common assumption was that each cultural artifact or historical event is the result of prior thought, not of material causes alone, and thus cannot be regarded as mindindependent. Each article in the journal should show the relation between the phenomenon under investigation and the psychological or linguistic process of construction that lay behind it. The goal was “to investigate the historical life of peoples, in all its manifold aspects, in such a way as to account for the discovered facts from the innermost part of the mind, and thus to try to trace the facts back to their psychological roots” (Lazarus and Steinthal 1860: 1). Völkerpsychologie began with a critical analysis of Johann Friedrich Herbart’s psychological and epistemological views. Herbart relied on an empirical account of the occurrence of representations in the subject. But from Lazarus’s and Steinthal’s perspective, if you stay within the description of the subject, you cannot account for how an individual interacts with a culture to bring about artefacts, which can then be studied scientifically.

42

The History of Understanding in Analytic Philosophy

From the viewpoint of Völkerpsychologie, Herbart’s epistemology relied on an unfounded view that individual psychology could be associated with a universal rational psychology, which in turn is the sole ground of knowledge of culture. Lazarus and Steinthal read Herbart as arguing that the mind is a closed system in which the causes of all ideas or concepts are the brain’s responses to stimuli and those responses are conditioned entirely by the independent properties of the mind. Lazarus and Steinthal did not question Herbart’s argument that the principles of variation of sensation can be given a naturalist, empirical account. But they proposed that abandoning the narrow focus on physical science for an analysis of history yields a broader picture: the mind’s responses to stimuli are conditioned by the stimuli themselves and, in particular, by the material and intellectual environment in which an individual finds herself. Hence, Völkerpsychologie could not rest on the Herbartian view in isolation. Völkerpsychologie does not reject Herbart’s empirical psychology entirely, but argues that it must be broadened to take into account the thesis of reciprocal influence. In the articles Lazarus and Steinthal wrote together for the inaugural issues of the Zeitschrift für Völkerpsychologie und Sprachwissenschaft, they proposed a thesis of reciprocity between individual mind and communal structures; that is to say, historical and cultural facts and events. To use one of their examples, individuals create poetry, but poetry is also an expression of the collective imagination insofar as it expresses the borders of the collective intellectual potential of a people. Poetry stretches the borders of language, but it cannot do so beyond the recognition of the people meant to read it or it will not be effective. Lazarus and Steinthal wanted it to be possible for an idea or concept to be created in the mind of an individual that would nonetheless have a broad impact on social reality: “Wherever the power of great, general ideas spreads over many peoples, where a single thought seizes and dominates the genius of many nations, there psychological research will be directed not only to the behavior of the Volksgeist, but to the nature and the law of those societies, which goes beyond them” (Lazarus and Steinthal 1860: 6). Lazarus and Steinthal wanted to account for the fact that certain ideas, paradigm cases of which are mathematical or poetic ideas, have a broader application and effect than the determination of a representation. They wanted psychology to account not only for the phenomenon of individual representation, but also for (i) the fact that the individual’s ideas are a synthesis of facts about the world and (ii) the fact that there is an objectively determinable relationship between ideas and theories, and between ideas and societies, revealed in history. Psychological research should be directed

Understanding, Psychology, Human Sciences

43

not at individual responses to stimuli, but rather at the interaction between the “nature and the law of those societies” in which the influence of an idea is evaluated and the ideas themselves. To account for productive ideas, Lazarus and Steinthal turned Herbart’s analysis of representational psychology into a critique of psychological action as it is made manifest in history. Individual representational psychology is no longer the foundation of the description of knowledge. Instead, Lazarus and Steinthal developed a model of knowledge according to which the principles of knowledge are built from the reciprocal interaction between particular and general, between individual and collectivity, an example of which is the relation between a person and a “people” (Volk). This is the source of the name Völkerpsychologie. In 1863, Lazarus gave a Rektoratsrede (Über die Ideen in der Geschichte [On Ideas in History], cited as Lazarus 1865) in which he explained how his account of ideas as productive (schöpferisch) and effective (wirksam) in history does not succumb to Hegelian dialectic or to Humboldtian6 empiricism. One of Lazarus’s first moves was to apply Herbart’s empiricist criticisms to the Hegelian and Humboldtian philosophies of history. Lazarus wanted to identify two “violent” misuses of ideas.7 The first is Hegel’s theory that the individual person is only a vessel for the perfect, absolute idea.8 The second is Humboldt’s view that the activity of the person is a necessary and sufficient condition to achieve the being (Dasein) of the ideas. Lazarus argued that Hegel’s thought removes ideas too strictly from the material conditions of society, and thus it is unclear how we are to have access to them. Humboldt, on the other hand, argued that ideas have impact solely through the agency of single people, geniuses, who bring those ideas to fruition. But individuals do not control the impact of their ideas on society, because they do not have absolute control over the material conditions or other individuals in that society. Lazarus proposed to revise Herbart’s theory, avoiding the problems with Hegel and with Humboldt by giving an account of the relation between individual representational psychology and community structures. Lazarus located reason’s contribution to history in the influence of ideas on individual thought and action. Since the ideas are “partly realized” in “action and productivity,” the psychologist or philosopher can take the results of that action and productivity as the source material for analysis.9 The ideas themselves are not real things, independent of the person who has them (as they are for Hegel). However, ideas are made partly manifest in action: as cultural artefacts, as historical trends, and as scientific theories, for instance. The materials available to epistemology for analysis are the empirical data of experience and of recorded history. But Lazarus insisted

44

The History of Understanding in Analytic Philosophy

on the productive (schöpferisch) aspect of reason: ideas could be analyzed in their empirical manifestation in human actions and their results. Ideas in history are the ideal forms of human life: “The content of these ideas consists in all the norms of the will, in the criteria for action that keep the natural impulses of human life within certain bounds, describe goals and ends for it, and give form to individual and common human life […]. Thus structuring ideas [Ideen der Gestaltung] are the true ideas in history” (Lazarus 1865: 73–5). The word “Gestaltung” can mean “form,” “structure,” “configuration,” or “design.” Lazarus argued in On Ideas in History that recorded data and physical and cultural artifacts already possess ideal content as artifacts of human effort: the norms of the will impose a structure or form on the content of culture. That does not mean that the data of history and culture is already organized in such a way as to reveal ideal patterns or to answer, say, ethical questions. At first, the data is a bare assemblage of raw material— Lazarus’s example is of a mosaic. Once the historical mosaic is assembled, it must be analyzed to find the ideal content that is crystallized within it. The content of collective conceptions and Ideen der Gestaltung has to be brought to consciousness. According to Lazarus and Steinthal, this is done via three methods: “compression” (Verdichtung), “apperception of ideas,” and “representation” (Vertretung). When a person pursues a plan with “passion” and will, as Lazarus and Steinthal describe, that plan may succeed in its purpose. Someone may wish to build a steam train, found an educational institution, introduce a new recipe, or build a new theory. The process of achieving this purpose will require significant effort, thought, and planning. But once the plan is achieved, that thought is compressed (Kusch 2019: 254). As Wilhelm Jerusalem noted, Ernst Mach used ideas similar to compression in his analysis of the “economy of thought” in science (Uebel 2019: 27). Newton’s Principia built on the work of many before him and was a significant achievement. But a high schooler can learn the fundamental results of the Principia now without the need to learn the context and history: the material found in textbooks has been “compressed.” The ideal content is still there, though, and to Lazarus and Steinthal, that ideal form is the proper subject of study for Völkerpsychologie. Compression means, in part, that complex contents and processes can be simplified so that a subject can have access to the crucial elements without going through the entire process that it took to acquire that content. Through compression, ideas are made more widely available. The psychological phenomena that Lazarus and Steinthal identified as the target of Völkerpsychologie are “apperception” and “representation.”

Understanding, Psychology, Human Sciences

45

Apperception is “the conscious or unconscious interpretation of sensory or conceptual content in light of background beliefs” (Kusch 2019: 254). As Dilthey would argue later, Lazarus and Steinthal noted that “intuition” or “observation” are not strictly separate from conceptual framing and that experience comes already shaped by background beliefs, concepts, and analogies.10 The “interpretation of sensory or conceptual content” through apperception can be employed in “representation,” which, for both Lazarus and Steinthal, crucially involves language. Uebel presented Lazarus’s view in the following. The first, still non-linguistic protorepresentations select only biologically important feature[s]‌and processes and combine in intuitions (Anschauungen) of objects and states of feeling. Later on, linguistic ideas (Vorstellungen), which possess no intuitive content of their own, were formed to represent (vertreten) these intuitions by being associated in apperception with sounds that were invariably produced in the company of the intuitions. In this process the linguistic ideas “distil and condense (verdichten) as an extract” the intuitive content [[1857] 1885: 323). It was the interplay of synchronous Vertretung and diachronous Verdichtung that led to ever higher orders of representation: repeated and iterated innumerable times this process lead first to one-word sentences and after further differentiation … to the properly judgmental stage. (Uebel 2019: 29) Steinthal added to this account the view that the steps of this process can be traced using logic. In his Abriss der Sprachwissenschaft (Outline of Linguistic Science) (1871), Steinthal argued that while expression can be viewed as a natural, biological phenomenon—one that does not follow a priori laws or principles—it is nonetheless true that the use of language to express and to represent can be captured logically after the fact (see also Steinthal 1855). Revealing the content of apperception and compression is a way to show how the mind (Geist) is active in experience. And tracing the steps employed in judgment, expression, and representation becomes a logical enterprise for Völkerpsychologie, one that serves as a way to mediate between “ideas of form” (Ideen der Gestaltung) and the manifestations of culture: institutions, achievements, languages, theories. In this way, culture becomes the expression of a purpose but also of a “logic” in this sense. Here, ‘logic’ is understood as a formal process of thought, not as a subject matter with a fixed content.11 Thus culture is understandable as the product of a mind, but also explainable as a logical process that can be reconstructed as

46

The History of Understanding in Analytic Philosophy

following norms and principles of the will. It is this latter claim that would bring Lazarus and Steinthal into conflict with Dilthey (see Section 2.5).

2.3.  DILTHEY ON UNDERSTANDING Wilhelm Dilthey”s method of understanding (Verstehen), drawn from an earlier figure, Johann Gustav Droysen (Beiser 2012: 298), is based on the distinction between explanation, the characteristic method of the natural sciences, and understanding, characteristic of the human sciences— Erklärung and Verstehen (Feest 2009).12 Explanation is the central method of the natural sciences, and it involves the familiar methods of demonstrating causes via natural laws or reducing phenomena to elements that collectively explain them. Understanding (Verstehen), on the other hand, involves two features: the relation between the human and the physical world, and reflection on the “fact of humanity” as it is given in the ever-changing nexus of “lived experience” (Dilthey [1910] 2002). To Dilthey, inner experience as present in psychology can describe what is real: he sometimes refered to descriptive psychology as “real” psychology. Descriptive psychology captures the nexus of interaction: not a relation between a “subject” and “the world,” but the single thing that is life or lived experience (Erlebnis).13 We begin from a complex experience that involves not only inner experience, but also the engagement of others and the world, and frameworks that overlie that experience. The point of descriptive psychology is to uncover the immediately given certainties of inner experience, which are distinct from the more complex phenomena in which they are engaged and revealed.14 Descriptive psychology is thus a way of understanding the fundamental, given inner experiences that are the foundation of all experience. Because inner experience is not derived or constructed, but immediately given, it is a certainty that can serve as the foundation of our knowledge in the human sciences. According to this reading, the role of hermeneutics in Dilthey is to reveal, not solely to interpret, inner experience. For instance, when we use the categories of “subject” and “object,” these refer to elements of inner experience that are given to us immediately. It is true that they correspond to constructed categories, but those constructions are only the way we have chosen to present the subjective and objective features of experience. This reading of Dilthey may shield him from the charge, which he faced from neo-Kantians, of relying on unscientific psychological methods, including introspection. It is certainly true that bare, unallied introspection would be a poor foundation for the human sciences. But one would hope, on his behalf, that this is not Dilthey’s method. That would involve arguing

Understanding, Psychology, Human Sciences

47

that we commune with our own inner experience via introspection and then attempt to project what we learn—without any reasoning independent of it—onto the world of culture. Taking a more sympathetic reading of Dilthey, a world of culture emerges, and we learn through descriptive psychology, and reflective judgment in particular, how it has developed from our own subjectivity. Once that has been done, we can reconstruct the cultural nexus from the psychic nexus: “an account of how the objects of the sociohistorical world are themselves constituted by psychological forces” (Kinzel 2018: 358).

2.4.  DILTHEY AND THE VOLKSGEIST From this perspective, we can understand why Dilthey would have judged the Volksgeist of Lazarus and Steinthal to be “unscientific.” The founding idea of the Volksgeist is that culture depends on human action and that human activity and human experience in turn depend on the development of culture. Possibly influenced by Wilhelm Waitz, Dilthey argued that the Volksgeist has no scientific basis: only individuals are real agents, real subjects.15 The starting point of scientific reflection, for Lazarus and Steinthal, is the artefacts and achievements of a culture. They are undeniable as cultural facts, but the scientific question is: what made them possible? There are facts that can be understood and explained only as the result of collective action, including collective preparation of the means of action. For instance, a language might develop over time and might crystallize the ways a community has come to represent or express things. These means of expressing or representing the world come about in communal activity. But they can be derived logically after the fact. And this logical reasoning allows us to identify what is subjective and what is objective—but the difference, for Lazarus and Steinthal, is that “subjective” can include a collective subjectivity. Individuals do not achieve culture on their own: they require communal life for their achievements. Lazarus and Steinthal explained this through a subjectivity that cannot be reduced to the individual, but rather expresses the subjective conditions of cultural achievements, which require the participation of more than the individual. Dilthey would agree that cultural experience is not already given with subjective and objective neatly distinguished: his methods of reflection and articulation aim to distinguish the two. The difference between Dilthey and Völkerpsychologie is that Dilthey required articulation to ascribe subjectivity only to the inner experience of the individual, whereas Lazarus and Steinthal allowed for subjectivity to include cultural conditions for subjective actions that are communal or collective.

48

The History of Understanding in Analytic Philosophy

Those conditions arise naturally in the interactions between people— but, nonetheless, they can be reconstructed using logical reasoning. Lazarus and Steinthal proposed a science of subjectivity that broadens the scope of the “subject” to encompass not only individual experiences and actions but also collective structures. It appeals to cultural facts and the structural relationships they make manifest. Lazarus and Steinthal employed the transcendental method: beginning with a fact and tracing its conditions. The process of tracing those conditions distinguishes subjective from objective conditions of experience and knowledge. We understand subjectivity only through this transcendental process. For Lazarus and Steinthal, subjectivity is revealed in the common world of culture, and that world cannot be reduced to fundamental “inner experiences.” Rather, subjectivity is whatever functions of expression and representation, for instance, are found to be at the origin of culture. This approach is a point of continuity between Steinthal and Cassirer (see Patton 2021 and 2015a). Lazarus and Steinthal did not reject psychology as a way of identifying subjective action, but they argued that any such action should be understood as the origin of a process that results in objectification or in cultural achievement generally. The subject is known through its deeds, not through reflection on an individual person’s experience: and if subjective action requires collective action, then there are collective subjects. Or, at least, there are ways that subjects can be understood properly only in relation to collective facts or structures. From this perspective, while Völkerpsychologie focused on finding collective cultural ideal structures, Dilthey instead focused on finding the ground of culture in the individual experience of the sociohistorical subject. Both Dilthey and Völkerpsychologie require revealing or articulating the subject of knowledge and its role in the constitution of objects and culture. The Marburg neo-Kantian school would take up this question and, rejecting the psychological approach, argue that the ideal structures of Völkerpsychologie are structures of objectivity.16

2.5.  EXPLANATORY AND DESCRIPTIVE METHODS 2.5.1.  Völkerpsychologie Lazarus and Steinthal were able to parlay a reinterpretation of Herbart’s insight that individual psychology alone is not a good basis of epistemology or even of psychology itself into a theory that evaluates the reciprocal determination of history and psychology. For Völkerpsychologie, ideas, in

Understanding, Psychology, Human Sciences

49

their manifestation in individual psychology, are productive of history, of theories, and of human culture in general. The proper method in epistemology is to demonstrate how ideas are determined by the individual, but also how the individual determines ideas through her influence on history and even on theory. Lazarus and Steinthal developed a theory of intellectual history according to which cultural structures can be evaluated using rational criteria, since they are produced by reasoning. However, as we saw, Lazarus and Steinthal were concerned as well with the question of how to divorce historical analysis from the study of the psychology of the individual subject.17 Individual psychology constrains conceptual analysis to a description of psychological processes. Locating intellectual history in an analysis of collective cultural structures such as language allows for evaluating the impact of ideas in a broader context than individual psychological processes. For Lazarus and Steinthal, the goal of analyzing structures (Gestaltungen) is to identify the structuring ideas (Ideen der Gestaltung) that give rise to them. The structure itself, then, is the part of nature that we learn about through analysis. Assuming that we have given a sufficiently complete account of the structuring ideas, the resulting structure is the real thing described by the idea. For instance, if an architect plans a building, then the building itself is the partial realization of the architect’s idea. Insofar as the building instantiates the idea, by conforming to the architect’s plan, the building is the realization of the structuring idea. (Building could be seen as an example of compression: the building is a physical, easily observable manifestation of a complex plan.) Being a realization of a structuring idea becomes part of the determination of the building as an object. The tradition of Völkerspychologie developed a method of: 1. recording facts about culture (myth, language, art, science); 2. tracing the origin of those facts in an interaction between humans, culture, and the material world; 3. finding the logic embodied by the process of deriving facts from experience by means of inferences and processes, material and ideal— this logic is made manifest in structuring ideas or ideas of structure (Ideen der Gestaltung). Understanding culture requires all three of these elements. Importantly, basic psychological and linguistic phenomena—for both Lazarus and Steinthal— are independent of the logic used to explain them. Lazarus and Steinthal saw human actions and achievements as evidence for the development of a particular kind of collective process, describable by logic after the fact.

50

The History of Understanding in Analytic Philosophy

It’s crucial, here, to understand the role of artefacts and achievements for Lazarus and Steinthal. We can have knowledge of a play, poem, theory, or philosophical argument by recording its occurrence and its properties, as a fact and as an artefact of culture. That is the first step of their method of analysis. The second step shows—using facts as evidence—that this achievement could not have come about absent the participation of others in the culture. Moreover, many achievements depend on previous achievements. New scientific theories build on standing results. Understanding the contribution a scientist has made with a new theory requires understanding the previous ones on which it builds to be able to determine what is new. This second step is the one that depends on law-governed reasoning about what can be done based on what. Thus, the third step—tracing the logic that gives rise to the achievement— requires showing how an individual builds on cultural structures and achievements. For instance, Edmund Hillary’s summit of Everest would not have taken place without the access he had to Indigenous Sherpa knowledge of the mountain, through his climbing partner Tenzing Norgay. Norgay himself called on hundreds and even thousands of years of Indigenous knowledge of Everest, not merely his own experience of the mountain. It is still true that Norgay and Hillary are the first on record to reach the summit of Everest. But understanding their achievement requires knowing how they employed Indigenous and Western techniques and knowledge as well as what they may have done differently from others. One may say: “But then their particular achievement—once we’ve understood how it goes beyond what they took from their cultures—is individual, not an element of a Volksgeist.” That disregards the fact that climbers to this day study Norgay’s and Hillary’s techniques. They involve themselves in learning how the summit of Everest was reached and in incorporating those methods into their own. And these techniques have been built into subsequent approaches to mountaineering. Even as individuals transcend the Volksgeist, their particular way of transcending becomes part of the Volksgeist. The Volksgeist of Lazarus and Steinthal is not independent of the material progress of culture: its nature, not just its manifestation, depends on the historical and material conditions that obtain. The Volksgeist is also not universal. It develops from the cultural conditions in a particular society and time. Understanding the Volksgeist does not depend on grasping an ideal entity, expressing a universal logic independent of the world of matter. Revealing the Volksgeist requires understanding the interaction between collective, cultural processes, material facts, and individual psychology in understanding how an artefact or event came about. Doing that requires

Understanding, Psychology, Human Sciences

51

tracing the logic of a particular achievement in order to identify the structure of the ideas made manifest in that achievement. Dilthey’s criticism of the scientific basis of Völkerpsychologie comes more into focus in this context.18 We might ask: What are the constraints on the analysis of the Volksgeist? What determines its location in a particular society or region, and what natural or psychological laws govern its development? What is the evidence for these laws?19 Dilthey, Lazarus, and Steinthal all tried to relate psychological processes in the individual to broader social phenomena and structures. Dilthey appealed to individual psychology as a foundation for the human sciences and for understanding as a method. Lazarus and Steinthal argued that “collective” structures can make manifest the relations between individuals and the Gesamtgeist or collective mind.20 They encounter tensions between the irrational elements introduced into their systems by locating Erlebnis or the Volksgeist in individual psychology and the attempt to give a “scientific” explanation using natural law. The “norms” or laws of society, which are key to the methods of Lazarus and Steinthal, are revealed only in action that is the result of human will. In this way, Völkerpsychologie relies on a psychological account, but one that includes appeal to collective structures or forms that make it possible to act based on a plan, principle, or idea. Those structures or forms (Ideen der Gestaltung) are embedded in artefacts of culture, and we can find their meaning by engaging in reflection that reveals the compression of ideas in them as well as the ways in which the artefacts are manifestations of human plans and purposes. 2.5.2. Dilthey on Descriptive Psychology One of the most significant elements of Dilthey’s approach developed in the late nineteenth and early twentieth centuries is his distinction between descriptive and explanatory psychology (Feest 2007; Damböck 2016, 2020). In the 1880s and 1890s, Dilthey was steadfast in arguing that the psychology that is the foundation of the human sciences must be a descriptive or “real” psychology, rather than one based on explanation. The differences between Dilthey, Lazarus, and Steinthal partly hinge on their view of whether psychology should involve methods of explanation as well as description. However, there are tensions on this score. Recognition of these tensions may have contributed to the evolution of Dilthey’s approach over time. There is some disagreement among recent interpreters of Dilthey’s work over whether he is a methodological holist (Patton 2015b; Šuber 2009) or whether his project is epistemological and foundational (Kinzel 2018;

52

The History of Understanding in Analytic Philosophy

Lessing 2016; Feest 2007). The supporters of the methodological holist reading emphasize Dilthey’s account of the cooperation between the human sciences and his remarks to the effect that the unity of the nexus of life makes it impossible to reduce knowledge to any one perspective. Most poignantly, Dilthey argued, “it is clear that the human sciences and natural sciences cannot be logically divided into two classes by means of two spheres of facts formed by them” ([1910] 2002: 103). Defenders of the reading that Dilthey was engaged in a foundational, epistemological project emphasize his remarks—which are certainly there!—that psychology (and anthropology) are the basis of all the human sciences; that “inner experience” provides “immediate certainty” about its contents; and that the methods of articulation and reflection reveal the basis for the constitution of objects of experience and knowledge. As Kinzel noted: “Dilthey’s remarks about the privileged status of psychology are best understood not as expressing a methodological thesis. Rather, they express a thesis about psychology being central for explicating the conditions of valid knowledge—of valid knowledge in general and of valid knowledge in the human sciences in particular” (2018: 350). Kinzel (2018: 348) emphasized Dilthey’s descriptive psychology as developed in his earlier works from the 1880s and 1890s. That psychology emphasizes the “articulation” of the functions of the “psychic nexus,” which is the totality of inner and outer experience (2018: 352). Dilthey defended the following claims in his early works (as elaborated here by Kinzel 2018): (a) The psychic nexus articulates its functions. (b) The psychic nexus articulates representations of the contents of experience. (c) The psychic nexus articulates representations of itself (as a content of inner experience). (d) These representations are valid because psychic structure is universal. In Dilthey’s words, Psychological thought articulates and distinguishes by starting with the overall given nexus. […] Psychic life will be conceived as a nexus of functions connecting the constituent parts, which in turn consists of specific systems, each of which presents new tasks for psychology. Since these tasks can only be accomplished by means of articulation, descriptive psychology must at the same time be analytical psychology. By analysis we always understand the articulation of a given complex reality. ([1894] 2010: 148)

Understanding, Psychology, Human Sciences

53

Descriptive psychology does not begin from the assumption of a “subject” who has experience of “objects.” Rather, the psychic nexus involves experience from within—“inner experience”—and experience of outer phenomena, which are given as a whole or totality, not in separation from each other. Achieving knowledge of the “subject” requires working to articulate the representation of that nexus itself (Kinzel’s step c above). The key is that articulation involves the drawing of connections between elements of systems, which are part of “the overall given nexus.” Those connections do not result in law-governed descriptions of how historical events result from psychological phenomena. Rather, sociohistorical phenomena are epistemically accessible from the perspective of the articulation of inner experience (Kinzel 2018: § 2). The aim of that articulation is not to arrive at causal explanations of sociohistorical phenomena, but rather to understand the connections between them; that is, to articulate the relationships within the psychic nexus. Feest (2007) and Kusch (2019: 261) noted that Dilthey’s criticisms of explanatory psychology in the 1890s may be aimed in part at Völkerpsychologie (as well as at his well-known target, English historian William Henry Buckle). As we saw above, the Volksgeist or Volksseele involves positing collective forms or structures that are intended to demonstrate how sociohistorical phenomena are made manifest as products of human willing and action. Dilthey likely saw this as of a piece with explanation in the natural sciences: taking “artefacts” of culture, as Lazarus and Steinthal did, as evidence of how events were made possible and how they happened. In his 1894 essay criticizing “explanative” or “explanatory” psychology as the foundation of the human sciences, Dilthey remarked that a science based on explanation “is one that subsumes a phenomenal domain to a causal system by means of a limited number of univocally determined elements that are the constituents of the system” (in [1924] 2010: 113; see Damböck 2020: §2). Descriptive psychology, on the other hand, always sees the phenomena as part of the nexus of life itself and does not provide explanations that reduce a system to its constituent elements. As a method of understanding, it is appropriate to the human sciences but not the natural sciences. But if Dilthey intends the method of “articulation,” taken as an epistemic project, to be fundamental to the human sciences, then he has a problem similar to the one he diagnoses in Lazarus and Steinthal. If Dilthey were able to make a program of drawing all the elements of culture from individual psychology work, then he would show, at the same time, that we can demonstrate how culture is built from a process based on a description of more fundamental elements. No matter how much Dilthey may have insisted

54

The History of Understanding in Analytic Philosophy

that because the “psychic nexus” cannot be reduced to anything more fundamental, this is not a form of explanation—it would be, nonetheless. These methods would be, properly speaking and in Dilthey’s own terms, methods of explanation, where the “univocally determined elements” of inner experience are taken as the explanans of more complex phenomena. If psychology is to be the ground of the human sciences and if the fundamental method of the human sciences is understanding (Verstehen), then psychology cannot be the independent epistemic ground of all the other results of the human sciences. That would make the fundamental method of the human sciences explanation, not understanding—despite Dilthey’s best efforts to argue that they are not. To be clear, I am not saying that Dilthey argued that his foundation of the human sciences in psychology is explanatory. I am saying that if Dilthey really was claiming that valid knowledge is based on articulation of inner experience, then his methods were those of explanation: specifically, the constitution of a complex domain from immediately given parts. Dilthey’s psychology would resemble Carnap’s constitution system, not Droysen’s history. Nonetheless, it is true that Dilthey saw psychology (and by extension the account of inner experience) as a foundation of the human sciences (Lessing 2016; Kinzel 2018). Lessing put it this way: “As is well known, psychology (and anthropology) take on a special function in Dilthey’s philosophy of the human sciences: it is the fundamental science of the human sciences” (Lessing 2016: 84, my translation). In my view, this is not a matter to be decided by choosing rival readings of Dilthey. Rather, it is crucial to note that Dilthey seems to have recognized this tension in his own approach and to have changed his mind over time. Kinzel and Feest focused on the early writings of the 1880s and 1890s. Those who emphasized the methodological holism of Dilthey’s approach (Patton 2015b; Šuber 2009) emphasized later works, including The Formation of the Historical World in the Human Sciences of 1910 (Dilthey [1910] 2002). In this work, Dilthey explicitly argued that the human sciences cannot be reduced to any one domain and that there is no fundamental distinction between the human and the natural sciences in terms of the facts dealt with by each. In my view, then—and this is consistent with the account in Makkreel (2020: § 1.1)—Dilthey’s position evolved over time. Dilthey seems to have recognized the tension between his commitment to the view that individual psychology is the fundamental science of the Geisteswissenschaften and his argument that individual experience is itself given only as part of a more fundamental psychic nexus, which is part of an even broader “nexus of life.” There are two problems here. First, individual psychology cannot be meaningful in isolation from the psychic nexus—on

Understanding, Psychology, Human Sciences

55

Dilthey’s own terms. If we require methods other than individual psychology to even identify the “individual” within the nexus, in what sense is individual psychology “fundamental”? Second is the problem identified above. If the results of individual psychology are intended as the basis of the constitution of knowledge, how is that not a form of explanation? The methods of Verstehen (understanding) involve the entire process of tracing back a sociohistorical nexus to its ground in experience. But they also involve emphasizing the connections between elements of that nexus, not just the constitution of one set of elements based on another.21 Verstehen requires much more than a ground, or foundation: Verstehen requires understanding how and why articulation and reflection, for instance, fit into a larger framework, which is not reducible to individual psychology. Saying that one science is the most fundamental science of the Geisteswissenschaften is not to say that individual psychology can function as the epistemic ground of all the results of the human sciences, independently of all the other human sciences. Even recognizing an “individual” who is capable of having “inner experience” requires analysis of immediately given experience: the distinction between subjective and objective is not immediately given, as Dilthey himself noted. It is key to notice that all the other human sciences are significant and that they are necessary in drawing out the implications of what we learn from descriptive psychology—and even in identifying the “subject” of psychology itself. It is significant, however, that when faced with the tensions described above, Dilthey did not abandon his position that psychology is the psychology of the individual. Instead, he retained his preference for individual psychology and simply emphasized anew the cooperation of the human sciences and the irreducible holism of the nexus of life. Dilthey still rejected the idea, familiar to him from Völkerpsychologie, that sociohistorical categories broader than the individual might be meaningful in psychological reasoning.

2.6.  CONCLUSION: POSTERITY The differences between the Marburg School (exemplified by Cassirer and Cohen) and the Southwest School (Windelband, Rickert) trace two paths for the philosophy that came afterward (Friedman 2000). Cohen, Cassirer, Windelband, and Rickert were, in some ways, responding to the nineteenthcentury exchanges between Dilthey, Lazarus, and Steinthal. The difference between Völkerpsychologie and Dilthey was consequential to the development of nineteenth- and even twentieth-century philosophy. Hermann Cohen was affiliated with Lazarus and Steinthal at first, but then converted their analysis—beginning with the facts of culture and tracing

56

The History of Understanding in Analytic Philosophy

their logical conditions—to a Kantian analysis of the transcendental subject (Edgar 2020; Köhnke 2001). Recent and classic scholarship on Rudolf Carnap’s Aufbau has interpreted the work as responding to Cohen’s methods, via Bruno Bauch (Richardson 1998, 2003; Friedman 2006), and Dilthey’s analysis, via Hans Freyer and his significant works including the Theorie des objektiven Geistes (Tuboly, forthcoming; Gabriel 2004; Dewulf 2017; Damböck 2012). The reception of Lazarus’s and Steinthal’s work at times turned their sociohistorical views on Völker into racist theory. Kusch (2019) has shown in devastating detail how Richard Thurnwald turned the analysis of Völkerpsychologie into a focus on “races and personalities” rather than social institutions and histories and how “Willy Hellpach in his 1938 book Introduction to Völkerpsychologie sought to make the field compatible with Nazi ideology” (Kusch 2019: 263; see Klautke 2013: Ch. 3). Steizinger (2020) has detailed the influence of both Dilthey and Völkerpsychologie on the National Socialist sympathizer Erich Rothacker.22 On the positive side, the greatest exponents of something like Lazarus’s and Steinthal’s methods are Ernst Cassirer, Georg Simmel, and Gilbert Simondon.23 Cassirer’s symbolic forms and his analysis in terms of “functions” do not require a Kantian transcendental subject. Rather, Cassirer looked at the conditions for objectivization, but within a context of a broader analysis of the facts of culture. This is superficially similar to Dilthey, except that Cassirer rejected Dilthey’s foundation in inner experience. Cassirer’s argument that there is a “logic” of the cultural sciences has clear similarities to positions in Steinthal (Cassirer [1942] 2000; Patton 2015a), and Cassirer’s analysis of language is indebted to Steinthal as well (Cassirer 1942; Patton 2021).24 Kusch (2019: 251) identified a “strong” program of Völkerpsychologie, identified by “methodological neutrality and symmetry; causal explanation of beliefs based on causal laws; a focus on groups, interests, tradition, culture, or materiality; determinism; and a self-referential model of social institutions.” Kusch cited as adherents to the strong program Georg Simmel, Emil Wohlwill, Hermann Cohen, and Wilhelm Windelband—though Kusch (2019: § 5) showed that Simmel did not succeed in his attempts to build on the strong program, and he was often critical of Lazarus and, especially, Steinthal.25 Gilbert Simondon cited Lazarus and Steinthal in his histories of psychology (2015: 111–12). While he is often considered a philosopher of technology, Simondon’s works L’individu et sa genèse physico-biologique (1964) and L’individuation psychique et collective (1989) focus on the process of individuation and on how the individual depends on the collective—both central concerns of Lazarus and Steinthal. Moreover, Simondon’s genetic,

Understanding, Psychology, Human Sciences

57

biological account of the origins of individuality and conscious experience find echoes in Steinthal’s and Lazarus’s work. While other, more direct influences on Simondon were likely more decisive, he should certainly be considered among those whose work deals with questions central to Völkerpsychologie and illuminates those questions. Similarly, the work of Wilfrid Sellars on collective intentionality and collective, communal structures might be read in connection with the Völkerpsychologie movement.26 Overall, however, the differences between Dilthey, Lazarus, and Steinthal are less consequential than their similarities. While Dilthey rejected the causal or law-governed elements in Lazarus and Steinthal, and while Völkerpsychologie rejected the individual psychology of Dilthey, they each constructed sociohistorical accounts of subjectivity. Each saw the self as indistinguishable from the sociohistorical nexus in which a person lives, and each saw history and anthropology as fundamental to capturing the facts of culture. The fundamental framework for the analysis of how culture is related to understanding owes much to their work and influence.

2.7. ACKNOWLEDGMENTS Many interlocutors have enriched this chapter, and I am very grateful for their contributions. First among them must be Adam Tamas Tuboly, who provided insightful comments on a draft and commissioned the work. Katherina Kinzel, Martin Kusch, and Thomas Uebel made invaluable contributions to the revision of the chapter, making suggestions that hit on key issues and resulted in significant improvements, for which I am very grateful. An anonymous referee for the volume made valuable comments. An early version of the research on Völkerpsychologie was done for my dissertation (Patton 2004), although it is now in a very different form.

NOTES 1 For Cassirer’s distinction between the sciences, see Capeillères (2007). 2 Sometimes Germanized as Moritz and Heymann. 3 I am very grateful to Martin Kusch for providing the historical context in this paragraph (references to the Historical School and to the context of Völkerpsychologie). See chapter 2 of Kluback (1956) for more detail on the relationship between Dilthey and the Historical School. 4 I am grateful to Martin Kusch for clarification of this passage (remaining errors are mine). 5 “Einleitende Gedanken über Völkerpsychologie, als Einladung zu einer Zeitschrift für Völkerpsychologie und Sprachwissenschaft” in Lazarus and Steinthal (1860: 1–73). All translations from this work are my own except where noted.

58

The History of Understanding in Analytic Philosophy

6 Wilhelm von Humboldt (1767–1835) was a humanist scholar known for, among other subjects, his contributions to linguistics and to aesthetic theory. He founded the Humboldt University in Berlin (Uecker 1990). The foundation of Steinthal’s linguistic study was in Humboldt’s linguistic theory (Lassahn 1995). 7 “It should arrest the critic’s attention compellingly that the great force of ideas is equally strongly emphasized in two such fundamentally different points of view as Hegel’s and Humboldt’s. Certainly one of the most important ways that ideas are determined is in relationship to acting and productive people, to the individuals who appear to have them. However, whereas in Hegel conscious or unconscious generality comes into the foreground, with Humboldt [it is] personal individuality. For the former [Hegel], the individual is only a medium […;] for the latter [Humboldt] the individual is the higher expression, the true life of the idea; for the former the expression: ‘we do not have ideas, but they have us’ is common; for the latter the doctrine is that only in the productive personality do ideas attain a productive existence” (Lazarus 1865: n. 41). 8 Thus, readings of Lazarus and Steinthal that emphasize the Hegelian elements are, in their view, one-sided. While the term “Volksgeist” is certainly suggestive of Hegel, and while Lazarus and Steinthal do incorporate some of Hegel’s notions, they are not, ultimately, Hegelian idealists. Their view is intended, instead, to be based on the human sciences (Geisteswissenschaften) and on secure scientific methods. I am grateful to Thomas Uebel for emphasizing this key point. 9 “Ideas in history are the ideas that are effective in the lives and activities of men, that is, of individuals and peoples, and thus in the life of humanity. They are not transcendental powers found outside the human soul, which somehow affect it from outside, but are actual [wirkliche] ideas, that is, ideas that appear within people as acts of their mental agency. They are produced, shaped, and developed within the human soul, and are partly realized in action and productivity” (Lazarus 1865: 73). 10 See Damböck (2016) for a discussion of Dilthey and Steinthal as exponents of German empiricism. 11 See Patton (2004: Ch. 1) for a discussion of Erkenntnislogik and Erkenntnistheorie. Thanks are due to Katherina Kinzel for encouraging clarification of this passage. 12 During the 1860s, Dilthey wrote several drafts of an article called “Contra Lazarism et Lazaristas, Millium, etc.” explicitly against Lazarus’s (and Steinthal’s) methods. According to Kluback (1956: 36), “this article that was never published furnished the foundation for his first published systematic work,” Introduction to the Human Sciences. 13 Dilthey’s Erlebnis has been compared with similar ideas from phenomenology. I am grateful to Adam Tuboly for emphasizing this point. 14 “Descriptive psychology .… seeks to recapture the intrinsic structure of inner experience and preserves its immediate givenness” (Kinzel 2018: 353). 15 “At the core of Waitz’s history of civilization stands the individual person: opposing romantic ideas about Volksgeist, he insists that ‘there is no agent, real and substantive, which can be considered as the spirit of a people or of humanity; individuals alone are real’ (Waitz 1863: 324). Dilthey would later agree with this plea for the crucial role of individuality” (Martinelli 2018: § 5).

Understanding, Psychology, Human Sciences

59

16 Edgar (2020: § 6) explored the move from Cohen’s early work with Lazarus and Steinthal to his later focus on objectivity. 17 Steinthal focused, in much of his work, on the interaction between humans as natural beings, their language, and how language grounds culture. Linguistic expression, for Steinthal, is fundamental to grasping the world in logical categories. But the expressive use of language is outside logic: it is an entirely natural phenomenon (see Patton 2021 and 2015a for discussion). 18 Thanks are due to Martin Kusch for raising this important point. 19 Adam Tuboly’s suggestions motivated several of these questions. 20 In this sense, it is wrong to see in Lazarus and Steinthal a quasi-idealist method, as Thomas Uebel has emphasized. 21 As Katherina Kinzel has emphasized to me (personal communication), this is already true of descriptive psychology. 22 Unfortunately, Steinthal himself engaged in discourse about “primitive” cultures and ethnic inferiority. And the introduction that Lazarus and Steinthal composed for their first journal issue engages in triumphalist thinking about the historical priority of the Germanic peoples (1865: 7). 23 Kusch (2019) identified most of the posterity of Völkerpsychologie and drew clear connections to the impact of their work. I was reminded of Simondon’s importance by a public remark by Grant Maxwell. 24 Ikonen (2011) cited the influence of Völkerpsychologie on Cassirer’s analysis of culture, and Kalmar (1987) argued that their work was influential on the modern concept of ‘culture’ generally. Luft (2015) analyzed Cassirer’s notion of “culture.” 25 Klautke (2013) analyzed more of the critical and sympathetic reaction to Lazarus and Steinthal. 26 I am grateful to Adam Tuboly for raising this latter point.

REFERENCES Beiser, F. (2012), The German Historicist Tradition, Oxford: Oxford University Press. Capeillères, F. (2007), “Sciences de la nature et sciences de l’esprit dans l”architectonique de la Philosophie des formes symboliques,” Revue Germanique Internationale, 6: 217–35. Cassirer, E. ([1912] 2005), “Hermann Cohen and the Renewal of Kantian Philosophy,” Angelaki, 10 (1): 113–26. Cassirer, E. (1942), “The Influence of Language upon the Development of Scientific Thought,” Journal of Philosophy, 39 (12): 309–27. Cassirer, E. ([1942] 2000), The Logic of the Cultural Sciences, trans. S. Lofts, New Haven: Yale University Press. Damböck, C. (2012), “Rudolf Carnap and Wilhelm Dilthey: ‘German’ Empiricism in the Aufbau,” in R. Creath (ed.), Rudolf Carnap and the Legacy of Logical Empiricism, 67–88, Dordrecht: Springer. Damböck, C. (2016), : Studien zur Philosophie im deutschsprachigen Raum 1830–1930, Dordrecht: Springer.

60

The History of Understanding in Analytic Philosophy

Damböck, C. (2020), “What Is Descriptive Psychology? Ebbinghaus’s 1896 Criticism of Dilthey Revisited,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 10 (2): 274–89. Dewulf, F. (2017), “Rudolf Carnap’s Incorporation of the Geisteswissenschaften in the Aufbau,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 7 (2): 199–225. Dilthey, W. ([1910] 2002), Wilhelm Dilthey. Selected Works, Vol. 3: The Formation of the Historical World in the Human Sciences, ed. and trans. R. Makreel and F. Rodi, Princeton: Princeton University Press. Dilthey, W. ([1894] 2010), “Ideas for a Descriptive and Analytic Psychology,” trans. R. Makkreel and D. Moore, in Wilhelm Dilthey: Selected Works, Vol. 2: Understanding the Human World, ed. and trans. R. Makkreel and F. Rodi, 115–210. Princeton: Princeton University Press. Dilthey, W. ([1924] 2010), Wilhelm Dilthey. Selected Works, Vol. 2: Understanding the Human World, ed. and trans. R. Makreel and F. Rodi, Princeton: Princeton University Press. Edgar, S. (2020), “Völkerpsychologie and the Origins of Hermann Cohen’s Antipsychologism,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 10 (1): 254–73. Feest, U. (2007), “‘Hypotheses, Everywhere Only Hypotheses!’: On Some Contexts of Dilthey’s Critique of Explanatory Psychology,” Studies in History and Philosophy of Biological and Biomedical Sciences, 38: 43–62. Feest, U., ed. (2009), Historical Perspectives on Erklären and Verstehen, Dordrecht: Springer. Friedman, M. (2000), A Parting of the Ways, Chicago: Open Court. Friedman, M. (2006), “Carnap and Quine: Twentieth-century Echoes of Kant and Hume,” Philosophical Topics, 34 (1/2): 35–58. Gabriel, G. (2004), “Introduction,” in S. Awodey and C. Klein (eds.), Carnap Brought Home, 3–23, Chicago: Open Court. Ikonen, S. (2011), “Cassirer’s Critique of Culture,” Synthese, 179 (1): 187–202. Kalmar, I. (1987), “The Völkerpsychologie of Lazarus and Steinthal and the Modern Concept of Culture,” Journal of the History of Ideas, 48 (4): 671–90. Kinzel, K. (2018), “Inner Experience and Articulation: Wilhelm Dilthey’s Foundational Project and the Charge of Psychologism,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 8 (2): 347–75. Klautke, E. (2013), The Mind of the Nation, New York: Berghahn Books. Kluback, W. (1956), Wilhelm Dilthey’s Philosophy of History, New York: Columbia University Press. Köhnke, K. (2001), “‘Unser Junger Freund.’ Hermann Cohen und die Völkerpsychologie,” in W. Marx and E. W. Orth (eds.), Hermann Cohen und die Erkenntnistheorie, 62–77, Würzburg: Königshausen & Neumann. Kusch, M. (2019), “From Völkerpsychologie to the Sociology of Knowledge,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 9 (2): 250–74. Lassahn, R. (1995), “Heymann Steinthal,” in F. W. Bautz and T. Bautz (eds.), Bautz Biographisch-Bibliographischen Kirchenlexicon, Vol. X, 1335–8, Spalten: Verlag Traugott Bautz.

Understanding, Psychology, Human Sciences

61

Lazarus, M. (1865), Über die Ideen in der Geschichte, Berlin: Ferdinand Dümmlers Verlagsbuchhandlung. Lazarus, M. ([1857] 1885), Das Leben der Seele in Monographien über seine Erscheinungen und Gesetze, 3rd ed., Berlin: Schindler. Lazarus, M., and H. Steinthal, eds. (1860), Zeitschrift für Völkerpsychologie und Sprachwissenschaft: Vol. 1, Berlin: Ferdinand Dümmlers Verlagsbuchhandlung. Lessing, H.-U. (2016), Die Autonomie der Geisteswissenschaften, Nordhausen: Traugott Bautz. Luft, S. (2015), The Space of Culture: Towards a Neo-Kantian Philosophy of Culture (Cohen, Natorp, and Cassirer), Oxford: Oxford University Press. Makkreel, R. (2020), “Wilhelm Dilthey,” in E. N. Zalta (ed.), Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/win2020/entries/dilthey/. Martinelli, R. (2018), “Defining Human Sciences: Theodor Waitz’s influence on Dilthey,” British Journal for the History of Philosophy, 26 (3): 498–518. Matherne, S. (2015), “Marburg Neo-Kantianism as Philosophy of Culture,” in S. Luft and T. Friedman (eds.), The Philosophy of Ernst Cassirer, 201–31, Berlin: De Gruyter. Patton, L. (2004), Hermann Cohen’s History and Philosophy of Science, PhD Dissertation, McGill University. Patton, L. (2015a), “Cassirer and Steinthal on Expression and the Science of Language,” Cassirer Studies, 7: 99–117. Patton, L. (2015b), “Methods of the Sciences,” in M. Forster and K. Gjesdal (eds.), The Oxford Handbook of German Philosophy in the Nineteenth Century, 594– 606, Oxford: Oxford University Press. Patton, L. (2021), “Symbolic Forms and the Logic of the Cultural Sciences,” in A. Pollok and L. Filieri (eds.), Ernst Cassirer’s Philosophy of Symbolic Forms: The Method of Culture, 261–78, Pisa: Edizioni ETS. Richardson, A. (1998), Carnap’s Construction of the World, Cambridge: Cambridge University Press. Richardson, A. (2003), “Conceiving, Experiencing, and Conceiving Experiencing,” Topoi 22: 55–67. Simondon, G. (1964), L’individu et sa genèse physico-biologique; l’individuation à la lumière des notions de forme et d’information, Paris: Presses universitaires de France. Simondon, G. (1989), L’individuation psychique et collective, Paris: Aubier. Simondon, G. (2015), Sur la psychologie (1956–1967), Paris: Presses universitaires de France. Steinthal, H. (1855), Grammatik, Logik, und Psychologie. Ihre Prinzipen und ihre Verhaltnis zueinander, Berlin: Dümmler. Steinthal, H. (1871), Abriss der Sprachwissenschaft, Berlin: Ferd. Dümmler’s Verlagsbuchhandlung. Steizinger, J. (2020), “From Völkerpsychologie to Cultural Anthropology: Erich Rothacker’s Philosophy of Culture,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 10 (1): 308–28. Šuber, D. (2009), “Social Science between Neo-Kantianism and Philosophy of Life,” in U. Feest (ed.), Historical Perspectives on Erklären and Verstehen, 267–90, Dordrecht: Springer.

62

The History of Understanding in Analytic Philosophy

Tuboly, A. T. (forthcoming), “The Constitution of Geistige Gegenstände in Carnap’s Aufbau and the Importance of Hans Freyer,” in C. Damböck, G. Sandner, and M. Werner (eds.), Logical Empiricism, Life Reform, and the German Youth Movement/ Logischer Empirismus, Lebensreform und die deutsche Jugendbewegung, Cham: Springer. Uebel, T. (2019), “Wilhelm Jerusalem, the Social Element in his Pragmatism, and its Antecedent in Völkerpsychologie,” European Journal of Pragmatism and American Philosophy, 11 (1): 24–38. Uecker, T. (1990), “Wilhelm von Humboldt,” in F. W. Bautz and T. Bautz (eds.), Bautz Biographisch-Bibliographischen Kirchenlexicon, Vol. II, 1168–73, Spalten: Verlag Traugott Bautz. Waitz, T. (1863), Introduction to Anthropology, trans. J. F. Collingwood, London: Longman, Green, Longman, and Robert.

CHAPTER THREE

The Heuristic and Epistemic Account of Verstehen in Twentiethcentury American Philosophy FONS DEWULF

3.1.  TO BI OR NOT TO BI? When discussing a philosophical concept, it can be easy to forget the reason why the introduction of that concept mattered in the first place. Without care for the rationale of their investigations, philosophers often find themselves lost among the possible positions or theories concerning their concepts. To retrieve the original intent of the introduction of a philosophical concept is one purpose which the history of philosophy can pursue. This historical essay serves such a purpose. The object of its investigation is the philosophical reason for theorizing about Verstehen. I aim to show the importance of distinguishing between two different and noncomplementary ways to approach Verstehen: from a heuristic perspective and from an epistemic perspective. Whereas the heuristic perspective conceives Verstehen merely as a subjective methodological aid to guide empirical research, the epistemic

64

The History of Understanding in Analytic Philosophy

account conceives it as an essential part of empirical research itself because it transforms source material into empirical evidence. Instead of choosing a specific philosopher’s point of view to begin my investigation, I start from a paradigmatic question which the concept of Verstehen, I think, should illuminate. Did William Shakespeare have an erotic attraction to males, and did he actually express this attraction in physical acts of affection? This is a factual question about the sexuality of a man who has been dead for over four hundred years. Since this question is about Shakespeare’s inner desires and sexual intention, there is no way of ascertaining an answer by some form of direct observation, even if we would have a time machine at our disposal. In order to answer the question, like most questions in literary or intellectual history, we have to use available physical manifestations, like texts or behavior, as evidence of Shakespeare’s sexual orientation. In the case of Shakespeare, the remaining possible manifestations of his sexual orientation are scant.1 There are no reliable autobiographical documents or contemporary witnesses and no preserved correspondences. The only manifestations of Shakespeare’s inner life are his published works, and of those, only the published collection of sonnets are likely candidates for an autobiographical interpretation. Although it might seem ridiculous to use only these sonnets as evidence to answer the question of Shakespeare’s sexual orientation, every interpretation of the sonnets begs this very question. In the first and largest sequence of sonnets (1–126), the author persona addresses a “fair youth” who is clearly identified by several pronouns and descriptions as male. Only in the second and much shorter sequence (127–54) does a “dark lady” appear in the narrative as an antagonist for the love between the author character and the fair youth. One reading of the poems thus suggests some kind of love triangle where the author of the sonnets is attracted to both male protagonist and female antagonist. Ever since the reprinting of the sonnets in the 1640 edition by John Benson, the sexuality of Shakespeare has been a point of contention for readers (Matz 2010; Smith 2007). Benson decided to rearrange some sonnets, placing all poems directly addressed to a male in the beginning of the sequence, suggesting that all others, especially those that express more overt sexual desires, are addressed to a woman (Matz 2010: 484). He even changed some gendered words to cohere with his rearranged sequence. After Benson’s publication, multiple editors of the sonnets have time and again attempted to rearrange the sequence and gendered words. Sometimes they wanted to dissuade a potential bisexual reading, or they tried to ensure that the affection of Shakespeare for a man was not interpreted sexually, but as a form of friendship.

THE Heuristic and Epistemic Account

65

In contemporary Shakespeare scholarship, there is no consensus on the question of Shakespeare’s sexuality. Some scholars still defend that the sequence addressed to the fair youth describes a passionate but nonsexual friendship between two men. Other scholars argue that the sonnets are fictional and should not be used to assess Shakespeare’s sexuality, which subsequently must remain hidden to us (Bevington 2020). Yet, a growing number argue in favor of both the autobiographical interpretation of the sonnets and the overtly bisexual nature of Shakespeare’s erotic feelings that are laid bare in the collection of poems (Wells 2004; Pequigney 1985; Hamill 2005). Despite a lack of consensus, the state of the debate has reached a fascinating evidentiary complexity. The nature of sonnet writing and homoerotic desires in the Elizabethan period are crucial in the various accounts. Based on this research, at least one interpretation has been discarded on evidentiary grounds: Shakespeare’s description of sexual desires towards the dark lady, if they are read as autobiographical, cannot be taken as evidence of his heterosexual identity, since sexual orientation at this time was never conceived within mutually excluding identity terms; that is, that a person is either homosexual or heterosexual (de Grazia 1993). I take the discussion of Shakespeare’s sexuality to be a paradigm example of the evidentiary difficulty that scholars in the humanities face: they have to justify why they understand a physical symbol, like a text, as evidence of some aspect of the inner life of a person. Other paradigm cases of this problem would be the debate over whether there were atheists in the sixteenth century (Wootton 1988) or whether Hawaiians worshipped Captain James Cook as the god Lono (Obeyesekere 1997).2 The German philosopher Wilhelm Dilthey identified this evidentiary problem as the problem of Verstehen (understanding).3 He introduced it as a distinct epistemological problem for the humanities (Geisteswissenschaften) in opposition to the natural sciences. Ever since the introduction of Verstehen as central to the epistemic identity of the Geisteswissenschaften (human sciences), philosophers have disputed its relevance in an evaluation of these sciences.

3.2.  THE HEURISTIC ACCOUNT OF VERSTEHEN Philosophical disputes over Verstehen have seen many faces. However, within the Anglo-American intellectual world, two mid-twentieth-century classic texts have greatly shaped later discussions on Verstehen in the philosophy of history: Carl Hempel’s “The Function of General Laws in History” and Theodore Abel’s “The Operation Called Verstehen.” Hempel and Abel similarly defended that Verstehen should be conceived as a mere heuristic method in humanities research, with no epistemological significance—I

66

The History of Understanding in Analytic Philosophy

will call this the “heuristic” account. On this account, Verstehen refers to the use of one’s commonsense familiarity with human behavior to come up with a lawful generalization concerning human behavior, like “a person ‘feeling cold’ … will ‘seek warmth’ ” (Abel 1948: 213). Such a generalization expresses the regular connection between two identifiable instances of some kind of behavior, in this case between feeling cold and looking for warmth. As an illustration of his ideas, Hempel (1942: 40) used the toy example of a migrating population. Although one could imagine many reasons for people to move from one location to another, the most likely connection according to (Hempel’s) common sense is the following: people who live in poor conditions will migrate to regions that they recognize as offering them better prospects. This generalization connects two kinds of events; namely, being in a region with poor living conditions and migrating to another region with better prospects. Such a generalization can operate in an inference: 1. Johanna lives in poor conditions and thinks that living in California would be better. 2. People who live in poor conditions will migrate to regions that they recognize as offering them better prospects. 3. Johanna migrates to California. The generalization expressed in (2) connects Johanna’s individual behavior/ situation as expressed in (1) and (3). Discovering a generalization like (2) guides an investigator of migrating populations in two ways. First, if the conditions of the population before and after migration are known, the researcher can use the generalization to state why the migration was to be expected given the initial conditions. If the migration has not yet occurred, the researcher can even predict its occurrence given information of the initial conditions (Hempel 1942: 38). Second, if the condition of the population before migration is not known, the investigator can use the theoretical generalization to guide research on their living conditions before the migration (Hempel 1942: 48). How does the researcher come up with generalizations like (2)? Abel calls such invention “internalizing.” “We imagine how we would have been affected by such an impact” and “we find, then, that in all its essential features the operation of Verstehen is based upon the application of personal experience to observed behavior” (Abel 1948: 215, 216). Hempel (1942: 44) described the method in a similar way: “the historian … tries to realize as completely as possible the circumstances under which they acted, and the motives which influenced their actions; and by this imaginary self-identification with his heroes, he arrives at an understanding.” Crucial to the heuristic reading of Verstehen is the relativity of this procedure to a historian’s personal

THE Heuristic and Epistemic Account

67

imagination.4 The heuristic account allows the possibility that the intuited generalizations of two historians might differ radically if they have had different life experiences. It essentially relativizes the validity of the intuited generalization to subjective credibility only (Abel 1948: 217). For both Abel and Hempel, Verstehen, relativized in this way, could only be a research heuristic. Because the commonsense generalization was given to the investigator through his subjective experience, it had to be confirmed by appropriate evidence independently of its validity based on the researcher’s imagination (Hempel 1942: 44; Abel 1948: 216).5 Thus, it becomes a matter of luck whether the use of one’s imagination is actually helpful in research: some intuitive generalizations connecting observable behavior will be vindicated by further evidence; others will not. However, if the researcher uses Verstehen without the necessary, evidentiary control, it also becomes a dangerous method. Hempel (1942: 45) explicitly mentioned the possibility that investigators could produce a correct prediction of some event, based on the generalizations originating in their imagination, even though they have no empirical confirmation for the generalizations. Such a prediction would not be warranted from an epistemological point of view. Abel and Hempel each concluded that Verstehen, at best, is an intuitive shortcut in the scientific investigation of the human world. At worst, it is used as a fictional procedure of metaphysics which illegitimately replaces the justificatory role of empirical research. Thus, for both Hempel and Abel, the reason to investigate Verstehen lay precisely in their cautionary warning not to confuse Verstehen with proper reasoning in the human sciences, which they understood as continuous with reasoning in the natural sciences (Abel 1948: 218; Hempel 1942: 48). Hempel’s and Abel’s accounts of Verstehen and their rationalizations of historical practice in general have been received extremely poorly by professional historians ever since they were published (Novick 1988: 399). If one applies this account to research on Shakespeare’s sexuality, it is not hard to see why so much scorn was directed toward the heuristic account. In that case, intellectual historians would have to use their imagination and their own familiarity with erotic feelings to intuit some kind of generalization expressing a regular connection between “writing sonnets like Shakespeare’s” and “having a bisexual orientation.” Afterwards, the historian would have to find empirical evidence to support the intuited generalization. However, no Shakespeare scholar ever discussed any such generalization. Moreover, it is unclear what kind of evidence could be used to support it. One possibility would be to find all sonnet writers in the early modern English world and then attempt to find out their sexual orientation independently from an interpretation of their sonnets. However, this only moves the problem to

68

The History of Understanding in Analytic Philosophy

another area. When can you use witness reports or correspondence from the sixteenth century as evidence of sexual orientation? Even if one excludes this problem, the fictional scenario in favor of a heuristic account has its own issues. The heuristic account presupposes that the behavior or the historical event under investigation can be subsumed under some general concepts. This presupposition enables the supposedly intuited universal generalization to inferentially connect two instances of these concepts. In the case of Shakespeare’s sexuality, the presupposition is hard to sustain. It’s not by virtue of Shakespeare’s writing of sonnets in general that scholars infer anything about his sexuality, but by virtue of the sonnets Shakespeare actually wrote, with the specific, gendered characters and the specific affectionate wordings that he used. There is no comparable sequence of sonnets. There is no meaningful kind of sonnet-writing behavior that the literary historian can connect to having some kind of sexual orientation. Even in the ideal scenario that there is independent evidence that every other writer of sonnets in the Elizabethan era had sexual passions for men, the possibility could remain that Shakespeare, even though he was a writer of sonnets, had no sexual passions for men. Such ideal scenario would yield extremely weak evidence because it offers no theoretical reason to relate both kinds (writing of sonnets and having a sexual passion for men). The current debate surrounding Shakespeare’s sexuality is a debate over what the sonnets, in their historical and literary uniqueness (specific characters, emotional phrases, etc.), are evidence of. Within a heuristic account, there is no room for such a debate. By design, any question concerning nongeneralizable events or behavior cannot be the object of evidentiary dispute, because all evidence is only meaningful within a theoretical framework that uses generalizations over kinds of events or kinds of behavior. From the perspective of the heuristic account, the contemporary investigation into Shakespeare’s sexuality is nonsensical, a mere play of words. Yet, such an investigation exists and has resulted in a vivid scholarly— might one say scientific?—debate. Given the inapplicability of the heuristic account to this paradigm case of Verstehen, there might be something wrong with the way Hempel and Abel framed their theories of Verstehen as a cautionary tale against scholars in the humanities. Two early responses to Hempel’s 1942 paper—one from Paul Oskar Kristeller and another from Otto Neurath—already attempted to diagnose the problem of this framing.

3.3.  RESPONSE I: PAUL OSKAR KRISTELLER When Hempel’s 1942 paper was published in The Journal of Philosophy, Kristeller was a lecturer in the history of philosophy at Columbia University.

THE Heuristic and Epistemic Account

69

Hempel and Kristeller had both arrived in New York in 1939 as intellectual migrants, escaping Nazi-occupied Germany. They were both acquainted with Ernest Nagel and Sidney Hook, and most likely both visited an informal philosophy discussion group (The New York Circle) organized by Nagel and Hook (Dewulf 2018). Philosophically, Hempel and Kristeller were direct opposites. Kristeller was a historically trained philosopherphilologist who before 1933 had been working with Martin Heidegger. When he left Germany, he had letters of recommendation not only from first-rate German philosophers, like Heidegger and Ernst Cassirer, but also from first-rate German philologists, like Werner Jaeger and Eduard Fraenkel (Obermayer 2014). In contrast, Hempel was a ahistorical philosopher who had been trained by two of the leading German scientific philosophers, Hans Reichenbach and Rudolf Carnap. Hempel’s interests were mainly situated in the formal analysis of metascientific concepts, like confirmation and later also explanation. During the Second World War, Kristeller became assistant editor of the Journal of the History of Ideas, one of the first English journals specifically focused on intellectual history. Because of his background in German historical philosophy and his own historical work on the Renaissance, Kristeller was well placed to discuss Hempel’s ideas on the use of imagination in historical practice. Kristeller responded in the paper “Some Remarks on the Method of History” (with Lincoln Reis as coauthor), which was published in The Journal of Philosophy in 1943.6 Reis and Kristeller argued that the historian should not attempt to replicate all scientific standards from the natural sciences. “If he wishes to be scientific, he will reach his goal not so much by talk about science or by blind imitation of the specific methods of another science, but rather by applying to his particular method and subject-matter its own demands of scientific accuracy” (Reis and Kristeller 1943: 234). Their reason for emphasizing methodological autonomy for the historian lay in what they took to be the misapplication of epistemic norms to domains where they don’t apply. For Reis and Kristeller, the scientific historian aims to infer and reconstruct past events (including ideas and feelings of historical actors) from the available source material. In contrast to Hempel’s heuristic account, they claimed that generalizations are not a necessary part of this inferential procedure. In order to use source material as evidence for a specific claim about the past, two basic justificatory procedures must be applied: factfinding and interpretation (Reis and Kristeller 1943: 235). As they admit themselves, they only repeat well-known earlier methodological principles of the historian’s practice from traditional historiographical manuals of Bernheim, Droysen, and Langois and Seignobos. The historian must collect

70

The History of Understanding in Analytic Philosophy

possible sources, select the relevant ones for the investigation, and then perform a critical analysis (investigating whether the sources are from the relevant period and were authored by the relevant person)—all processes that differ from methodological problems in the natural sciences. In the case of Shakespeare’s sexuality, this translates to key choices. Should one adjust the arrangement and wording of the sonnets as they were first published in 1609? Was the author of the sonnets also the author of the other works in the Shakespearean corpus? These choices have to be justified by the historian, and only then an attempt can be made to reconstruct a past intention or a past event based on the available source material. Reis and Kristeller identify this reconstruction as the evidentiary problem particular to the humanities—Verstehen. “This is where scope for a properly disciplined historical imagination enters and in a sense is the special mark of the historian” (Reis and Kristeller 1943: 237). According to Reis and Kristeller, a historical imagination is necessary to reconstruct the past from available traces. This is most obvious when only fragments of a work remain and the historian is forced to reconstruct an author’s thoughts or intentions from those fragments. In their view, an analogous situation prevails everywhere in literary history, history of art, intellectual history, etc. However, they did not think that the historian should “indulge in wild hypotheses or in phantasies” in order to arrive at some generalization (Reis and Kristeller 1943: 237). The historical imagination is disciplined and is also not aimed at the production of generalizations. Every reconstruction of the historian must be supported by further evidence from source materials. In the case of the sonnets, most scholars in favor of the bisexuality thesis maintain that sonnets 18–126 describe the increasing infatuation of the author character with the male fair youth. As they point out, these sonnets use phrases to express the homoerotic affection of the author character for the fair youth. Whether phrases such as “he was but one hour mine” (sonnet 34) are verbal data that detail physical intimacies between the author and the fair youth requires an interpretation of the meaning of these phrases—something that in itself requires familiarity with the idioms of sixteenth-century English poetic language, a broader interpretation of the phrase within the narrative of the poem, which in turn is supported by the narrative of the sonnet sequence. Nothing in this case suggests the use of a vivid imagination on the part of the historian or the literary critic which would only be valid relative to their subjective experience. According to Reis and Kristeller (1943: 238), the chief types of error in history are “disregards of available evidence, lack of criticism, a failure to verify the sources and a misinterpretation of the available source materials.” Verstehen, understood in this way, delineates the

THE Heuristic and Epistemic Account

71

particular way in which historians of the human past have to infer intentions and beliefs of actors from available source material, just like physicists must support their theoretical models with available experimental data. However, unlike physicists, historians of the human past often cannot conceptualize the objects of their investigation as members of some general kind. Unlike the experimental sciences, the historian mostly studies objects which are not recurrent. The historian should allow the possibility that his object of study can only be conceived in its concrete uniqueness. “Such concreteness does by no means forbid description or even understanding, but it can not fit any neat scheme of inductive generalization” (Reis and Kristeller 1943: 242). In historical studies (of human culture), experimental repetition is excluded from the potential methodology. In the case of Shakespeare’s sonnets, this problem is evident: there is no way to repeat Shakespeare’s life, alter his sexual orientation, and find out whether the content of the sonnets also changes. Although Kristeller was a student of Heinrich Rickert, he never mentioned the idiographic-nomothetic distinction between the cultural sciences and the natural sciences as introduced by Wilhelm Windelband and later developed by Rickert. Kristeller’s point is not that historiography should only be concerned with the specificity or uniqueness of historical events (idiographic approach). His point is (a) that some historical events (like, e.g., Shakespeare’s sexual orientation) have to be investigated scientifically in absence of a lawful, theoretical framework covering those events and (b) that they also can be investigated in such a way.

3.4.  RESPONSE II: OTTO NEURATH In 1943, Hempel also received very similar criticism for his ideas on Verstehen from someone who was intellectually much closer to Hempel’s own tradition of scientific philosophy. Otto Neurath, trained as a historical economist, became an important member of the Vienna Circle in the 1920s. From 1934 onwards, he was the central organizer of the Unity of Science movement, a broad collective of intellectuals, including Hempel, who intended to renew both philosophy and science (Reisch 2005). From 1934 onwards, Hempel and Neurath had an extensive correspondence about the future of the Unity of Science movement, especially the relevance of historiography and historical perspectives on science for that movement (Dewulf 2020). During the Second World War, Neurath found refuge in Oxford. Unlike his colleagues in the movement, he decided to remain in Europe, but became increasingly concerned with intellectual developments in the United States.7 In 1935 Neurath had already complained to Hempel that the logic of science

72

The History of Understanding in Analytic Philosophy

should not be conceived as a “cookbook” containing abstract recipes for rationality applicable in all domains.8 After reading Hempel’s 1942 paper, he was disappointed in Hempel’s treatment of the historical sciences as if they were entirely similar to mathematical physics. Specifically, he did not agree that historical events can always be subsumed under generalizations, either of a deterministic kind (all As are Bs) or of a probabilistic kind (95% of all As are Bs). In his letters of 1943 and 1944, Neurath emphasized the “unpredictability in principle” of most events in the social domain.9 He wrote that “the unpredictability of chance is already there before, in so far as items are concerned, which are not sufficiently repeated.”10 In another letter, he wrote: You see I think the so-called probability standpoint in history would imply, that we get 6000 Hitlers and then tell something how they behave “on an average”. Usually people think that the behaviour of Hitler and the Nazis may be based on probability statements dealing with masses, as if that were in discussion here. IT IS NOT. Could you show me, very roughly, what you intended in your article on p. 41, with illuminating examples, please. (Neurath to Hempel, November 25, 1944, Nr. 246 VCA) On that page, Hempel had suggested that historical generalizations could be of a statistical nature. He gave the example of a child, Tommy, who comes down with the measles (Hempel 1942: 42). After some days, Tommy’s brother also comes down with the measles. According to Hempel, the infection of Tommy’s brother could have been expected with a high probability given the condition that Tommy and his brother spent time together in close proximity. Hempel presumed that a probabilistic generalization was in play, something like “it is very probable that someone who has not yet had the measles and has been in close contact with a person who has the measles will also get the measles.” Hempel believed that generalizations in history are of the same character: their probability values would not be numerical, only “at best known quite roughly.” Neurath was not convinced by this epidemiological example of the application of probabilistic generalizations. In the case of the measles, at least, it is possible to support some kind of probability for the contagion of the measles, through repeated observations or experiments. The contagion of Tommy’s brother can be subsumed under a general kind of event: a contagion in the context of close contact with an infected individual. Given the right kind of retrospective observational study of multiple infectious cases, it is possible to infer the probability of contagion in general given these initial conditions (which were assumed similar in all investigated cases).

THE Heuristic and Epistemic Account

73

However, the social, political, and cultural preconditions of historical events differ in varying degrees. As a consequence, in many cases an assessment of probability cannot be made. Neurath gave the example of Hitler’s rise to power: unlike the contagion of Tommy’s brother, Hitler’s rise to power cannot be grouped under some general kind of event which is open to investigation through observations or experiments. In the case of Tommy’s brother, it is possible to isolate the relevant initial conditions, similar in all cases, such that the contagion could be expected; namely, the close contact with another infected person. It is more difficult to discern what the relevant initial conditions for a historical event are. Neurath emphasized this problem to Hempel: “You speak of ‘determining conditions’ without even touching the point that the whole aggregate may appear as ‘determining conditions.’ ”11 As Kristeller also pointed out to Hempel, the historian cannot isolate which aspect of his object is relevant beforehand. What is relevant about the sonnets in terms of their ability to serve as evidence for Shakespeare’s sexuality, is not guaranteed by a general theory which conceives them under a general kind of behavior. Is it Shakespeare’s writing of sonnets, writing of sonnets to a man, writing of sonnets to a man within a particular narrative, etc.? When Neurath wrote his comments to Hempel in 1943 and 1944, he was frustrated by Hempel’s neglect of actual historical research. Almost no historical work discusses historical events as if they can be conceived as members of a general kind. Presuming that historical studies are like the epidemiological study of the measles without any real example from historical literature was a mockable idea according to Neurath—as if we could generate 6,000 Hitlers under various conditions and investigate the outcome! He was especially upset that Hempel quoted Maurice Mandelbaum’s book The Problem of Historical Knowledge as a positive inspiration for his paper: “That you call M. Mandelbaum’s analysis clarifying hurt me deeply, I just wanted to quote this book as an example of overwhelming metaphysical confusion.” “Why start from this Mandelbaum business and not from some historian or anthropologist?”12

3.5.  THE SOURCE OF HEMPEL’S FRAMEWORK: MAURICE MANDELBAUM Mandelbaum’s The Problem of Historical Knowledge was one of the first English monographs which discussed German and Italian philosophers of history, including Croce, Dilthey, Mannheim, Simmel, Rickert, and Troeltsch. In the book, Mandelbaum argued against a relativist epistemology for the historical sciences and defended a realist interpretation of historical

74

The History of Understanding in Analytic Philosophy

narratives: events in historical narratives represent real events, and the historical narrative represents causal relations between the events (Mandelbaum [1936] 1967: 188). Mandelbaum did not conceive the causal relations as lawful connections between events, but as singular products and producers of change ([1936] 1967: 264). Hempel disagreed with this interpretation of historical causation and argued that every causal relation should be understood as a regularity relation between events, expressible as a lawful generalization (Hempel 1942: 48). Although Hempel and Mandelbaum had a difference of opinion in this respect, they each framed their account in opposition to a relativist understanding of historical science. Mandelbaum identified Croce, Dilthey, and Mannheim as the source of such a relativist understanding: all three threatened to reduce the validity of historical accounts to the personal values of the historians or their imagination (Mandelbaum [1936] 1967: 83). Mandelbaum’s reading of Dilthey’s method of Verstehen relied purely on the use of the historian’s imagination. According to Mandelbaum, Verstehen entailed an argument from analogy; that is, that only those experiences of historical actors which were also experienced personally by the historian could be inferred from past manifestations. This limits the validity of Verstehen to one’s personal knowledge of inner experiences. Mandelbaum understood this subjective foundation of Verstehen as the main problem of Dilthey’s work: it introduced a fundamentally subjective element in the writing of history. Mandelbaum’s conclusion on Dilthey was harsh: “In the end, there is a final subjective feeling of certainty which cannot be transcended or rendered intelligible by logical means” ([1936] 1967: 66). On this reading of Dilthey, contemporary historians with a sexual passion for men could recognize Shakespeare’s sonnets as manifestations of their own sexual affection, whereas contemporary heterosexuals would be barred from such an inference. If this is how one conceives Verstehen, it is epistemologically defunct from the beginning. By using personal imagination as confirmation for one’s claim, the validity of the historical conclusions is necessarily relative to the subjective experience of the historian defending them. In agreement with Mandelbaum, Hempel and later Abel emphasized that the use of “empathetic understanding” should not confuse an appeal to our personal imagination with empirical soundness. Consequently, Hempel and Abel both solved Mandelbaum’s subjective interpretation of Verstehen by relegating it to mere research heuristic (with potential dangers). Both Mandelbaum and Hempel believed that Verstehen had no role to play in the acquisition of evidence for historical claims. However, such heuristic account of Verstehen entails a problem: how can the historian use manifestations from the past, like the sonnets, as evidence of past events

THE Heuristic and Epistemic Account

75

in the first place, given the absence of any general theoretical framework about those events? Given that historians have no reliable knowledge of how men in general manifest their sexual affection for other men, nothing about Shakespeare’s sexuality can be inferred from the availability of his sonnets. Thus, these poems are destined to remain mute to the extent that they are conceptualized as singular expressions of Shakespeare and not as instances of some general kind. If it were possible to create a general theory of human expressions applicable across times and cultures, the epistemological problem of Verstehen would not exist. However, in the absence of such a theory, there are three possibilities. This first option is skepticism about knowledge of the past, especially of human intentions: Shakespeare’s sexuality, and most other topics like it, cannot be objects of our knowledge. At best, they would be objects of speculation only. The heuristic accounts of Hempel and Abel take this option: to the extent that the sonnets are not subsumed under some well-confirmed generalization, they cannot be evidence of anything at all. Any inference that is based on their contents can only be speculation. As mentioned earlier, Hempel’s and Abel’s reason to have a theory about Verstehen lay exactly in the warning that such a theory entails; that is, not to confuse the use of one’s imagination with proper empirical standards of justification. The second option is a form of acquaintance epistemology. This would imply that the historian can, in absence of any mental mediation through a general theory, infer the correct conclusion from his source material. Mandelbaum believed that “the historian need merely look at his material in order to find ties of existential dependence” (Mandelbaum [1936] 1967: 261). In the case of Shakespeare’s sonnets, this position entails that a historian of English literature would merely have to look at the sonnets to understand that the first 126 poems originated from Shakespeare’s sexual affection for men. This option introduces a miracle into the epistemological process: it replaces the problem of evidence with an instant acquaintance of the causal relation between historical events. It is not surprising that Neurath classified Mandelbaum’s position as overly metaphysical. It was precisely because it could not serve a purpose given his acquaintance epistemology that Mandelbaum did not have a theory of Verstehen at all. A third option is to conceive Verstehen as a necessary element in an epistemic practice where the objects under investigation cannot be captured within a general theory—I will call this option the “epistemic” account of Verstehen. The reason to have a theory about Verstehen lay in the recognition that there is an epistemological problem typical for the human sciences: the use of sources as evidence in absence of a theory which conceives those

76

The History of Understanding in Analytic Philosophy

sources under general concepts. Kristeller and Neurath, although they worked within very different intellectual traditions, each urged Hempel to abandon the heuristic account and take up an epistemic account. Such an account entails an intellectual challenge: to reconceive what is a rational inference and what is not, given the absence of a general theory about the object under investigation. If you have a general theory, then an instance which was predicted by the conjunction of theory and initial conditions can count as confirmation of that theory. Or, you can use a well-confirmed theory to infer one object given another one; for example, you can infer sexual passion for men given specific poems. In absence of such a general theory, this is not an option. Consequently, how you think about confirmation and evidence must change as well. In his exchange with Hempel, Neurath pointed out that theories in the cultural/social domain should not be evaluated with the same epistemic values as theories in the nonhuman domain, specifically concerning prediction. Some theories in physics allow precise predictions for the state of a system, given conditions of its past. It is not rational to desire such predictions from historiographical theories. Neurath wrote Hempel that the difference between physics and historiography “is not a question of exactness, but of non-predictability.”13 In his book Empirische Soziologie (1931), Neurath had already argued that prediction should not be expected in the social domain, because regularities between social events are unstable. According to Neurath (1931: 130), each individual in a social system can react differently to new events in his social environment: some individuals will perform abnormal behavior compared to the others in the group, and this abnormal behavior can then expand among groups and between groups. Contrary to the parts of complex machines, individuals in a society can change their regular behavior. Moreover, sociological theories can also become part of the social system that they are describing and have the potential to change that very system by describing it (Neurath 1931: 132). Unlike machine parts or the movement of planets, the regularities displayed by human behavior are not robust under constantly changing conditions. The problem is not that human behavior is very complex, but that it is variable, especially in relation to the (scientific) representations of that behavior. This limits the projection of social regularities into the future and consequently the possibility of predictions. Reis and Kristeller also discussed the difference between rational norms for theories in historiography and the natural sciences. Because historians interpret their sources in the absence of a general theory, every object of historical interpretation can be conceived in a great variety of ways (Reis and Kristeller 1943: 242). Both the selection of the sources and the aspects

THE Heuristic and Epistemic Account

77

of the sources that historians construe as relevant will determine their interpretation. Consequently, historical interpretations will vary depending on these choices. Reis and Kristeller believed that the best way to describe the product of historiography is as a narrative. Historical narratives have to be confirmed by evidence—they have to remain in accord with the sources. However, the choice of the narrative structure will play an important role, both in selecting the sources and deciding which aspects to highlight in the absence of any lawful generalization (Reis and Kristeller 1943: 243).14

3.6.  DILTHEY’S COGNITIVE AND LOGICAL PERSPECTIVE ON VERSTEHEN Given that most actors involved with historiography/sociology never endorsed a heuristic account of Verstehen, how did both Hempel and Mandelbaum come to frame Verstehen as a subjective procedure relying purely on the personal imagination of the historian? Hempel most likely never read Dilthey, but Mandelbaum, who was an important source for Hempel’s 1942 paper, did. And Mandelbaum’s ideas hinge on a specific reading of the last chapter of one of Dilthey’s unfinished manuscripts, The Formation of the Historical World in the Human Sciences. Specifically, Mandelbaum refers to the following passage from the chapter, titled “Exegesis or Interpretation,” to emphasize that there always remains a final subjective feeling of certainty in the process of understanding: “There is something irrational in all understanding, just as life itself is irrational; it cannot be represented in a logical formula. The ultimate, although quite subjective, sureness residing in this re-experiencing cannot be replaced by any cognitively tested inferences that explicate the process of understanding” (Dilthey 2002: 239). Here, Dilthey defended that the reconfiguration of physical objects as the manifestation of an inner object necessarily relies on the cognitive process of re-experiencing what was at stake for the historical actors involved. Mandelbaum concluded from this passage that Dilthey considered Verstehen ultimately as a subjective and irrational procedure. What Mandelbaum did not mention is Dilthey’s characterization of Verstehen in the paragraphs directly following this one. There, Dilthey conceived Verstehen not only from a cognitive point of view but also from a logical point of view; that is, as an inference. Dilthey contrasted inferences in the natural sciences with those in the human sciences because the latter do not use a general theory to infer an inner thought from an external manifestation: “The process of understanding based on this is to be conceived as inductive. This induction is of the kind in which an incomplete series of cases is used not to infer a

78

The History of Understanding in Analytic Philosophy

universal law but a structure or ordering system that gathers the cases as parts of a whole” (Dilthey 2002: 240). Because the induction does not rely on general laws connecting the inner thought and the external manifestation, it would be mistaken to evaluate the success of inferences in the human sciences in the same way as one evaluates the success of inferences in the natural sciences. For Dilthey, the task of a logic of the human sciences is to find the rules to assess this type of inference (Dilthey 2002: 240). Mandelbaum never mentioned this task as an epistemological challenge, nor Dilthey’s specific epistemic characterization of Verstehen as a form of inference. Mandelbaum chose to focus purely on Dilthey’s cognitive assessment of Verstehen as the subjective re-experiencing of an event in the historian’s imagination. The logical characterization of Verstehen was abandoned. Hempel followed this reading. Neurath’s and Kristeller’s early responses to Hempel’s 1942 paper were attempts to disambiguate the cognitive and the logical aspects of Verstehen which Dilthey had mixed together. Neurath did not conceive the re-experiencing of an event in one’s imagination as an important task for the historian, and Kristeller equally did not attribute any importance to the historian’s personal imagination. Instead, each attempted to characterize epistemic norms for inferences in the human sciences—norms that differ from those in the natural science, given the absence of general laws covering the objects contained in the inferences. Despite these early responses, the heuristic account of Verstehen remained in place as a potential position on the problem of Verstehen. Hempel’s reputation within American philosophy steadily rose in the 1950s and 1960s. Throughout his life, he never abandoned the heuristic account: he never replied to Neurath’s long letters about the 1942 paper, and he never engaged Kristeller’s narrative perspective.15 As a consequence, many scholars set themselves the task to respond against Hempel. The amount of papers written between 1945 and 1975 against the heuristic accounts of Hempel and Abel is immense (Danto 1966; Dray 1957; Mink 1966; Skinner 1969; Weingartner 1961; Taylor 1971). One of the most cited papers within this set of critics is “Interpretation and the Sciences of Man” by Charles Taylor (1971: 49), who gave three reasons why general hypotheses have no role to play in the sciences dealing with the cultural domain. First, there is the problem that historical and social events cannot be shielded from possible external interference: experiments cannot be set up to isolate specific preconditions to an event. Second, and related to the first reason, is the problem that differing interpretations of human expressions cannot be judged by their different predictions. Third, and most fundamentally, humans are self-defining animals: their self-understanding

THE Heuristic and Epistemic Account

79

can shift radically through time. In close similarity to Neurath, Taylor also argued that human behavior is variable in relation to the representations of that behavior in a given society. Any representation of behavior has the potential to alter that very behavior. “The success of prediction in the natural sciences is bound up with the fact that all states of the system, past and future, can be described in the same range of concepts, as values, say, of the same variables” (Taylor 1971: 49). However, the states of cultural systems are dynamical to the extent that the range of concepts used to describe these states differ throughout time. “Human science looks backward. It is inescapably historical” (Taylor 1971: 51). Taylor’s position resembles arguments made by Kristeller and Neurath against Hempel’s heuristic account: by virtue of the fact that we have no means to find a warranted general theory capable of covering all historical events as kinds, we should not use the epistemic norms available in the natural sciences when evaluating theories in the humanities.

3.7. CONCLUSION Was Shakespeare sexually attracted to a man? And did he express his affection in physical intimacies with this man? There is no available evidence to settle these questions beyond rational dispute. However, given a background of interpretative assumptions, the collection of Shakespearean sonnets can be used to support several mutually exclusive answers. Our theories about Shakespeare’s sexuality are underdetermined by the available evidence. This epistemic situation is not unique to the human sciences in any way; it abounds in the natural sciences as well. Given that new evidence about Shakespeare himself is not likely to turn up, progress in the debate must rely on progress in the support for a given set of interpretative assumptions. In fact, research on the nature of sexuality in the sixteenth century has achieved such success: we can now assume that sexual attraction to people of the same sex was a recognized category of behavior in the sixteenth century, as a natural but sinful affection in all men. This assumption excludes all interpretations of the sonnets that portray a bisexual Shakespeare as anachronistic. The absence of a general theory concerning sexuality and its relation to poetic expression did not make the debate impossible or speculative. The sonnets play an evidentiary role: an interpretation of their content limits possible positions concerning Shakespeare’s sexuality. From the perspective of the heuristic account, as defended by Hempel and Abel, the absence of any general theory about the relation between sonnet writing and sexuality nullifies the evidentiary importance of the sonnets. From the perspective of the epistemic account, the debate on Shakespeare’s sexuality proves the

80

The History of Understanding in Analytic Philosophy

possibility of rational argumentation in absence of generalizations covering the objects of investigation. Defenders of the epistemic account have recognized the need to evaluate the relation between theory and evidence in the human sciences differently from that relation in the natural sciences exactly because the objects of investigation often cannot be conceived as members of some general kind. That one can argue about Shakespeare’s sexuality is the reason why Verstehen is an epistemic problem. This inferential characterization of the problem of Verstehen has been available ever since Dilthey. Unfortunately, in Anglo-American philosophy, Verstehen has often been framed as a subjective procedure which relies entirely on the use of one’s personal imagination. This is how Maurice Mandelbaum introduced Dilthey’s ideas to the American philosophical community in 1938. Once an interpretation of a philosophical concept has been introduced within a community, it is hard to remove. Hempel’s framing of Verstehen along the lines of Mandelbaum is a good example: it ensured that generations of philosophers either accepted his framing of Verstehen as the subjective use of imagination or were forced to argue exactly against it. Ultimately, any position on Verstehen comes down to the attitude one has towards cases like the debate on Shakespeare’s sexuality: should we conceive it as a form of investigation that is part of what we take rationality to be, or not? A positive answer leads to an epistemic account of Verstehen; a negative one to the heuristic account and its related skepticism concerning our knowledge of the past.

NOTES 1 In this essay I use Shakespeare as a name to refer to the author of the Shakespeare corpus. Some scholars see an important link between the question of the author’s sexual orientation and the question of who that author was (Hamill 2005). In this essay, I am neutral as to who actually was the author of that corpus. 2 For an illuminating discussion of this case as an epistemological problem, see Roth (2008). 3 In contemporary philosophy of science, there has been a revival in the study of scientific understanding. Within this revival, understanding refers to an epistemic aim of scientific theories, in contrast to explanation and description (de Regt 2017). Because it is necessary to distinguish Dilthey’s problem of understanding in the human sciences with this contemporary debate in philosophy of science, I chose to use Dilthey’s term of Verstehen. If understanding is an aim of science, one might question whether Dilthey’s Verstehen (e.g., next to explanation) also gives understanding. However, I want to stay neutral to (the meaningfulness of) this question and will not discuss it. 4 Both Hempel and Abel, in the context of their discussion of sociology/ historiography, equate “understanding” to einfühlen, the process of imaging

THE Heuristic and Epistemic Account

81

oneself to be someone else (Abel 1948: 216). The intellectual origin of this equation is discussed in Section 8.5. 5 It is important to note that in his later career, Hempel revisited the notion of “understanding,” in the context of science in general, as the intellectual aim of scientific explanations. Explanations enable us to understand ourselves and our world (Hempel 1965: 333). 6 At the time, Lincoln Reis was a student of history at Columbia University. Because Reis did not have an academic career after he left Columbia, it is difficult to assess what his contribution to the paper was. 7 Neurath’s increasing worries about the intellectual developments in the United States during the war are exemplified in his war correspondence with Rudolf Carnap (Cat and Tuboly 2019: 512–685). 8 Neurath to Hempel, February 2, 1935, Nr. 244 Vienna Circle Archives (henceforth VCA), Noord-Hollands Archief, Haarlem, the Netherlands. Quoted by permission of Wiener Kreis Stichting, Amsterdam. All rights reserved. 9 For a discussion of Neurath’s ideas about unpredictability in principle, see Reisch (2001). For Neurath’s positive reception of some hermeneutic ideas about Verstehen, see Uebel (2019). 10 Neurath to Hempel, September 21, 1943, Nr. 246 VCA. 11 Neurath to Hempel, November 25, 1944, Nr. 246 VCA. 12 Neurath to Hempel, November 25, 1944, Nr. 246 VCA. 13 For Neurath, this distinction does not imply that the sciences are disunified. Sciences which lack the capacity of prediction are still sciences in the sense that one can confirm their statements through observation and experimentation expressed in a physicalist language. Neurath was merely pointing out that predictive power is not an essential feature of scientific theories. Reis and Kristeller made a similar point: historiography is a part of science even though it lacks predictive power. 14 Today, the narrative tradition in philosophy of history has greatly expanded this very same idea. For a recent contribution to this tradition, see Roth (2020). 15 Hempel wrote Kristeller in 1943 that despite Kristeller’s interesting discussion of historical criticism, he did not agree with Kristeller’s rejection of the role of generalizations in history (Hempel to Kristeller, November 25, 1943, Paul Oskar Kristeller Papers, box 42, folder 24, Rare Book and Manuscript Library, Columbia University Library). This letter seems to be the only exchange Kristeller and Hempel had about their papers.

REFERENCES Abel, T. (1948), “The Operation Called Verstehen,” American Journal of Sociology, 54 (3): 211–18. Bevington, D. (2020), “William Shakespeare—Sexuality,” Encyclopedia Britannica. https://www.britannica.com/biography/William-Shakespeare/Sexuality. Cat, J., and A. T. Tuboly, eds. (2019), Neurath Reconsidered: New Sources and Perspectives, Cham: Springer.

82

The History of Understanding in Analytic Philosophy

Danto, A. C. (1966), “The Problem of Other Periods,” Journal of Philosophy, 63 (19): 566–77. de Grazia, M. (1993), “The Scandal of Shakespeare’s Sonnet,” in S. Wells (ed.), Shakespeare Survey: Volume 46: Shakespeare and Sexuality, 35–50, Cambridge: Cambridge University Press. de Regt, H. W. (2017), Understanding Scientific Understanding, Oxford: Oxford University Press. Dewulf, F. (2018), “Revisiting Hempel’s 1942 Contribution to Philosophy of History,” Journal of the History of Ideas, 79 (3): 385–406. Dewulf, F. (2020), “The Place of Historiography in the Network of Logical Empiricism,” Intellectual History Review, 30 (2): 321–45. Dilthey, W. (2002), The Formation of the Historical World in the Human Sciences, Princeton: Princeton University Press. Dray, W. H. (1957), Laws and Explanation in History, Oxford: Oxford University Press. Hamill, J. (2005), “Shakespeare’s Sexuality: And How It Affects the Authorship Issue,” The Oxfordian, 8: 1–35. Hempel, C. G. (1942), “The Function of General Laws in History,” Journal of Philosophy, 39 (2): 35–48. Hempel, C. G. (1965), Aspects of Scientific Explanation, New York: Free Press. Mandelbaum, M. ([1936] 1967), The Problem of Historical Knowledge, New York: Harper & Row. Matz, R. (2010), “The Scandals of Shakespeare’s Sonnets,” ELH, 77 (2): 477–508. Mink, L. O. (1966), “The Autonomy of Historical Understanding,” History and Theory, 5 (1): 24–47. Neurath, O. (1931), Empirische Soziologie, Vienna: Springer. Novick, P. (1988), That Noble Dream: The “Objectivity Question” and the American Historical Profession, Cambridge: Cambridge University Press. Obermayer, H. P. (2014), Deutsche Altertumswissenschaftler im amerikanischen Exil: Eine Rekonstruktion, Berlin: De Gruyter. Obeyesekere, G. (1997), The Apotheosis of Captain Cook: European Mythmaking in the Pacific, Princeton: Princeton University Press. Pequigney, J. (1985), Such Is My Love: A Study of Shakespeare’s Sonnets, Chicago: University of Chicago Press. Reis, L., and P. O. Kristeller (1943), “Some Remarks on the Method of History,” Journal of Philosophy, 40 (9): 225–45. Reisch, G. A. (2001), “Against a Third Dogma of Logical Empiricism: Otto Neurath and ‘Unpredictability in Principle,’ ” International Studies in the Philosophy of Science, 15 (2): 199–209. Reisch, G. A. (2005), How the Cold War Transformed Philosophy of Science: To the Icy Slopes of Logic, New York: Cambridge University Press. Roth, P. A. (2008), “Beyond Understanding: The Career of the Concept of Understanding in the Human Sciences,” in S. P. Turner and P. A. Roth (eds.), The Blackwell Guide to the Philosophy of the Social Sciences, 311–33, Oxford: Blackwell Publishing. Roth, P. A. (2020), The Philosophical Structure of Historical Explanation, Evanst: Northwestern University Press. Skinner, Q. (1969), “Meaning and Understanding in the History of Ideas,” History and Theory, 8 (1): 3–53.

THE Heuristic and Epistemic Account

83

Smith, B. R. (2007), “Shakespeare’s Sonnets and the History of Sexuality: A Reception History,” in R. Dutton and J. E. Howard (eds.), A Companion to Shakespeare’s Works, 4–26, Oxford: Blackwell Publishing. Taylor, C. (1971), “Interpretation and the Sciences of Man,” The Review of Metaphysics, 25 (1): 3–51. Uebel, T. (2019), “Neurath on Verstehen,” European Journal of Philosophy, 27 (4): 912–38. Weingartner, R. H. (1961), “The Quarrel about Historical Explanation,” Journal of Philosophy, 58 (2): 29–45. Wells, S. (2004), Looking for Sex in Shakespeare, Cambridge: Cambridge University Press. Wootton, D. (1988), “Lucien Febvre and the Problem of Unbelief in the Early Modern Period,” Journal of Modern History, 60 (4): 695–730.

84

PART TWO

Logical Empiricism

86

CHAPTER FOUR

Is There a Hermeneutic Aspect in Carnap’s Aufbau? CHRISTIAN DAMBÖCK

4.1. INTRODUCTION Michael Friedman (1999: 89–113) argues that the main aim of the constructional system of Carnap’s Der logische Aufbau der Welt (The Logical Structure of the World, [1928] 1967), hereafter called Aufbau, is to provide intersubjectively communicable and therefore objective definitions for concepts. This intersubjective communicability, in turn, becomes granted via what Friedman calls purely structural definite descriptions of concepts (hereafter PSDDs): Carnap argues that only the logical form or structure of a relation is objective or scientifically communicable: any excess “content” going beyond logical structure must rest ultimately on ostensive definitions, and these, according to Carnap, provide no intersubjective meaning. For truly objective communication, then, we must require that all relations are given only through descriptions of their structure—through what Carnap calls “purely structural definite descriptions” (§§ 12–15). (Friedman 1999: 97) A famous example of a PSDD in the Aufbau is the definition of the “Eurasian railroad network” (§ 14). One of the stations in the network might

88

The History of Understanding in Analytic Philosophy

be characterized by the structural characteristic that emerges if we draw lines between all stations being directly connected in the railroad network. If the resulting graph is unique for Station A, then it constitutes a PSDD for A. In this chapter, I recommend separating Friedman’s discussion of PSDDs from a more general discussion of how to make concepts intersubjectively communicable. While the main innovative feature of the Aufbau is certainly the idea of PSDDs, it is also important to see that there is another aspect of intersubjective communicability involved in the Aufbau which does not boil down to structure, but instead directly connects with more traditionally minded philosophical notions such as psychophysical parallelism and understanding via analogical conclusion. Because all these notions are characteristic of Dilthey’s and other nineteenth-century hermeneuticists’ views generally, including their conception of hermeneutics and objectivity; for example, this story also adds an additional task to highlighting Dilthey’s work as relevant background reading to the Aufbau. Whereas previous readings (Damböck 2012; Dewulf 2017) were more focused on the status of mental objects (geistige Gegenstände) in the Aufbau, the present chapter has a different focus; namely, how crucial psychophysical parallelism and understanding via analogical conclusion as Diltheyian notions are for the general scientific and empirical side of objectivity. But how does this work? How can Carnap, who unequivocally claims in the Aufbau and elsewhere to be neutral regarding philosophical questions such as realism versus phenomenalism (§§ 176–8), at the same time defend such seemingly super-philosophical nineteenth-century claims? The answer is that there is a clear difference between, for example, the realism topic and the notions of psychophysical parallelism and understanding via ontological conclusion being addressed here. Though the latter were defended mainly by philosophers and insofar are philosophical because they belong to the nineteenth-century philosophical discourse culture, they are still different from philosophical conceptions in the sense of realism and phenomenalism, because they involve nontrivial empirical claims. If psychophysical parallelism holds, then for every mental fact we need to have a corresponding empirical fact, and this first needs to be established at the level of facts and natural laws. Moreover, understanding via analogical conclusion only works if for every particular case of a mental state, we also can find a strategy to connect this with empirical facts (at the level of behavior or neurons); again, we need to have empirical facts and laws of nature here that make this conception tenable. Thus, it is also quite problematic to call these conceptions “philosophical”; one might better call them general scientific notions, because they are bound to empirical facts and natural laws that first need to be established in sciences such as psychology, sociology, history, physiology,

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

89

and the neurosciences. In order to do justice to this, we henceforth call these crucial notions philosophical/scientific notions because they stem from nineteenth-century philosophical discourse but still depend on concrete empirical facts and natural laws. In Section 4.2, I briefly characterize the particular brand of hermeneutics that was represented by Wilhelm Dilthey and highlight those crucial notions of this conception that also might have been relevant for Carnap. In Section 4.3, I review several reduction procedures as crucial elements of the Aufbau which obviously parallel Dilthey’s conception in ways that at least partially can be traced back to Dilthey. On this basis, I claim in Section 4.4 that objectivity in the Aufbau cannot entirely be boiled down to PSDDs.

4.2.  DILTHEY (AND OTHER NINETEENTHCENTURY PHILOSOPHERS) ON HERMENEUTIC OBJECTIVITY Note that the views being attributed to Dilthey here are characteristic also for work of several other German philosophers, including the Völkerpsychologie of Moritz Lazarus and Chaijm H. Steinthal, the empiriocriticism of Richard Avenarius, and the neo-Kantianism of Hermann Cohen and Ernst Cassirer. These views were commonplace for the mainstream of German philosophy in the second half of the nineteenth century, but they were increasingly attacked at the beginning of the twentieth century with the effect that in the 1920s, facing the new age of philosophy as represented by Edmund Husserl, Martin Heidegger, and others, they were widely rejected in the German philosophy community: twentieth-century continental philosophy henceforth took a stance widely at odds with the more rational and objectivist views of these nineteenth-century philosophers (Damböck 2020, forthcoming). Given that overall picture, it is indeed remarkable that Carnap still decided to defend these nineteenth-century pro-scientific views against his somewhat postmodern contemporaries in the 1920s. I take Dilthey as my chief witness here only because he and some of his followers first influenced Carnap regarding his general views on objectivity and understanding. So, this is the story that could be told in similar ways using Steinthal, Lazarus, Avenarius, Cohen, or even Cassirer. As a counterpart to the more well-known notion of objectivity as a means of observing nature in an entirely nonsubjective way, the German philosopher Wilhelm Dilthey specified a second form of objectivity as a crucial notion of hermeneutics—viz., objective understanding.1 Whereas the former was a matter of the natural sciences, the latter was intended as its counterpart in the field of the human sciences. However, even

90

The History of Understanding in Analytic Philosophy

Dilthey made some effort to point out that the first variety of objectivity that comes from the natural sciences is crucial for all kinds of science. Following the lead of Helmholtz, Dilthey (2011: 690) was convinced “that we got a grip on the proofs that it is really the case and the world is built by objects being utterly identical with our representations.” However, objectivity of that kind had to be complemented, for Dilthey, by a second variety being tailored for the humanities: “The objectivity of knowledge that is sought here has a different sense; the methods for approaching the ideal of objectivity of knowledge here display essential differences from those by which we approach the conceptual cognition of nature” (Dilthey 2002: 92). The “laws” that are relevant here, for Dilthey, are grounded in the nature of “the experienceable, conceivable, and understandable” (ibid.). The key problem for Dilthey, however, in the task of gaining objectivity in the field of the human sciences, is that while the physical world is, in his view, directly accessible through our senses, the cognitive world of others is accessible indirectly, at best. He says, for example, “We can understand other people only by means of analogy to our own inward” (Dilthey 1990: 315). Therefore, in Dilthey’s view, the hermeneutic method has to be developed on the basis of everyday communication, because only there do we come close to the “inward” of others; analogical understanding, therefore, has to be based on a study of everyday communication. According to Dilthey, the “elementary forms of understanding” as covered by a study of that kind provided the “letters” of our alphabet of understanding which we put together in order to develop “higher forms of understanding” (Dilthey 2002: 228– 34). Thus, understanding as a whole is entirely based, for Dilthey, in the development of what we today might rather call a semiotic practice that allows us to decipher the talk and behavior of others. Intuition plays an important role here, of course, since it is intuition alone that allows us to see the just mentioned “analogies to our own inward.” However, intuition of that kind is not anything irrational. Dilthey follows here the scientific attitude of his teacher August Boeckh, according to whom hermeneutics has been based on scientific analyses in the fields of (what we now would call) linguistics, history, psychology, and sociology. This variety of hermeneutics converges with twentieth-century semiotics, although it has little to do with the varieties of hermeneutics in the style of Heidegger, Habermas, and Gadamer. Though, for Boeckh, hermeneutics was based on intuition or “feeling” (Gefühl), this side of hermeneutics was just the last step in a corpus of investigations that could become scientific only if they were developed in an entirely empirical way (Boeckh, Bratuscheck, and Klußmann [1886] 1877: i, 86).

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

91

The key features of this empirically minded variety of hermeneutics— which were quite characteristic for nineteenth-century German philosophy, although they became somewhat twisted in twentieth-century so-called continental philosophy—are the following: 1. A variety of psychophysical parallelism, here seen as the thesis that there is no mental phenomenon without a clearly demarcated physical representation, or as Dilthey (1989: 67) puts it: “Only by means of abstraction is mental life separable from the psychophysical life-unit.”2 2. Understanding becomes possible by means of the fundamental thesis that mental events as well as meanings of utterances are related to behavior in similar ways among human beings. 3. Mental objects become empirically accessible only via their physical representations—Hegel’s “objective spirit,” therefore, becomes turned on its head; the Hegelian notion of a mental being who is objective already at the mental level is replaced with a mental being whose objectivity exists only by means of the physical representations of mental objects (Dilthey 2002: 170–4). In the following, I will argue that the constitutional system of the Aufbau is based on notions of intuition and understanding being quite similar to those developed by Dilthey. More precisely, it will be my crucial claim that objectivity in the Aufbau at the end must be seen as surprisingly close to Dilthey’s second, hermeneutic notion of objectivity, while PSDDs are the tools that allow us to achieve that goal in a scientific way.

4.3.  REDUCIBILITY CONDITIONS The Aufbau offers several fundamental reduction procedures which all rest on philosophical/scientific assumptions related to Dilthey’s (and other nineteenth-century philosophers’) conception of hermeneutics. This does not change the fact, of course, that the way in which these procedures become realized in the Aufbau is entirely different from Dilthey’s view, which clearly does not offer anything like PSDDs. These are the Ditlheyian procedures as they appear in the Aufbau. (1) Reduction of the physical to the autopsychological is the most fundamental reduction procedure of the Aufbau. I will not discuss here whether Carnap’s seemingly failed attempts to provide such a reduction procedure might finally come to a happy end in the context of a new and improved system (Chalmers 2012; Leitgeb 2011). Rather, I simply want

92

The History of Understanding in Analytic Philosophy

to point out that to provide such a procedure is fairly crucial; failure of a reduction procedure of that kind would imply an entire breakdown of science and rationality. In the Aufbau, Carnap defends, several times (and quite explicitly), psychophysical parallelism (§§ 19, 57) as the basis of this step. Carnap identifies “the central problem of metaphysics” independently from the “psychophysical problem” (§ 22) and therefore locates it outside the purview of the Aufbau, where he clearly defends psychophysical parallelism and rejects substance dualism (§ 162). If a physical object were irreducible to sensory qualities and thus to psychological objects, this would mean that there are no perceptible indicators for it. Statements about it would be suspended in the void; in science at least, there would be no room for it. Thus, all physical objects are reducible to psychological ones. (§ 57, original emphasis) To be able to talk about a physical object, one must have a mental grasp of it. Otherwise, the object will be a myth or something for which, “in science at least, there would be no room.” There are no isolated spheres of physical and psychological objects for Carnap. The strategy to reduce psychological objects to their parallel physical objects can be twofold: it can be based on “parallel processes” in the brain or on (behavioral) “expression relations” (ibid.). Without reduction procedures of that kind, the external spatiotemporal world would be inaccessible to the human mind. Therefore, psychophysical parallelism and the respective reduction procedure, for Carnap, are fundamental conditions for the possibility of a constructional system and science as a whole. (2) Similarly, the reducibility of heteropsychological objects to autopsychological ones (by means of physical objects) is an equally fundamental condition for Carnap. This can be illustrated by the following episode which occurred when Carnap started his work on the Aufbau in summer 1920 with a brief manuscript entitled “Skelett der Erkenntnistheorie” (Skeleton of Epistemology).3 This manuscript reflects the outcome of a meeting Carnap had organized in summer 1920 in order to do philosophy of science with a number of friends (who, by the way, all belonged to the Dilthey school; see Dahms 2016; Damböck 2021). It consists of a sketch of the interrelation of the first three object types of the Aufbau and considerations on possible reduction procedures. The starting points are “experiences” or “facts of consciousness” (Erlebnisse (Bewusstseinstatsachen)). Certain experiences that Carnap calls “representations” (Vorstellungen) allow us to identify their objects, which either can be “representations” again or “physical objects.” Distinguished physical objects are “my body” and “other bodies.” “Several

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

93

sequences of physical events of my body are often simultaneous with certain sequences of experiences; when I find similar sequences of events in other bodies then I will produce the fiction that even their ‘experiences of other minds’ take place” (RC 081-05-04). This shows that in 1920, Carnap already had a system in mind whose major aim would have been the development of reduction procedures between autopsychological, physical, and heteropsychological objects (whereas the idea of PSDDs was developed at a later stage). At the end of the “skeleton,” Carnap provides a reference to Eduard Sokal’s paper “The salto mortale of thought” (“Das Salto-Mortale des Gedankens”) from the 1904 issue of Ostwald’s Annalen der Naturphilosophie (Sokal 1904). Since this paper, which (at least to my knowledge) was never cited in Carnap’s subsequent work, seems to have been important for the formulation of the very first draft of the Aufbau, it should be worthwhile having a brief look at it. After a number of poems and romantic considerations on the small man and the big stars, Sokal turns to a rather clear and concise discussion of the psychophysical problem, with special focus on the problem of accessibility of other minds. Sokal’s discussion mentions Gustav Theodor Fechner as its major witness and points out that we are able to know from other minds only by means of “analogical conclusion” (Analogieschluss). It is not perception but rather (logical) deduction that allows us to gain knowledge of others. [This] knowledge of others that we only can deduce and conjecture but never can perceive [undertakes] a steady and essential correction to our own content of consciousness and [serves] as integral part of our fundamental and important conceptual differentiations. Without this eternally enigmatic knowledge of the other that we achieve through a salto mortale of thought we obviously would never have been able to supply to our own perceptions the examination and objectivation that appears to be so important, both in a practical and in a thought economical, i.e., scientific sense. (Sokal 1904: 108) Carnap seems to have agreed with the conclusion that there is no communication and, as a consequence of this, there is also no objectivity if science fails to offer this salto mortale and connects different minds via the behavior of the respective people and via the fundamental thesis that behavior represents mental events in all human beings in similar ways. This is, in fact, what Carnap describes in those paragraphs of the Aufbau that deal with the problem of intersubjectivity (see Section 4.4). (3) For Carnap, mental objects (geistige Gegenstände) have to be reducible to psychological and physical representations. He rejects the idea of mental

94

The History of Understanding in Analytic Philosophy

objects as being located in a transcendent realm of platonic ideas. Carnap’s respective conception is closely related to Hans Freyer’s account in Theory of Objective Mind (Theorie des objektiven Geistes). Freyer’s book, in turn, is based on the conceptions of Wilhelm Dilthey, especially those in Dilthey’s late book The Formation of the Historical World in the Human Sciences (Der Aufbau der geschichtlichen Welt in den Geisteswissenschaften; cf. Dilthey 2002; Freyer 1923; Tuboly, forthcoming). The upshot of these writings of Freyer and Dilthey is that mental objects exist only by means of their mental representations (in the head of the persons who represent them) or the pieces of literature, art, and artifact that represent them. This psychological and semiotic approach to mental objects is crucial for Carnap because only in the context of such a non-Platonist treatment do suitable reduction procedures become possible. These three examples demonstrate that Carnap’s constructional system, though neutral with regard to metaphysical conceptions such as realism and phenomenalism, and also based on the purely logical account of PSDDs, is not neutral with regard to fundamental philosophically/scientific claims. Psychophysical parallelism, the accessibility of other minds, and the nontranscendent status of mental objects and values are quite concrete and empirically meaningful philosophical/scientific positions. For Carnap, to affirm these positions and to reject their counterparts is a fundamental condition for the possibility of rationality, for intersubjective communication, and for the sciences as a whole. The constructional system of the Aufbau is inevitably linked to these philosophical/scientific positions, and it is not a purely logical undertaking, nor is it neutral. Those philosophical conceptions, in turn, that Carnap regards as a matter of convention—viz., realism versus phenomenalism—are compatible with the same empirical reality and are indeed empirically neutral and somewhat irrelevant for science. Psychophysical parallelism and hermeneutics via ontological conclusions, by contrast, are by no means neutral in this sense, because they lead to a wealth of empirical implications. Therefore, only because they are not philosophical at all in the sense in which Carnap takes philosophy as a mere matter of convention can these notions be adopted by Carnap on a level that goes beyond matters of convention. The idea that Carnap has a deep commitment to philosophical neutrality (Carnap [1928] 1967, §§ 157–78; Schilpp 1963, 44–6) is misleading insofar as he only rejects philosophical claims (or tries to be neutral at least—i.e., accepting both negative and affirmative readings of these claims) that have no positive empirical meaning at all. However, the reducibility claims we attribute to the Aufbau are certainly empirically meaningful. Psychophysical parallelism (in the sense Carnap takes it) has the concrete empirical

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

95

implication that every psychological phenomenon must have a physical representation (the existence of psychological phenomena that fail to have such representations would refute that claim); the claim of accessibility of other minds implies the possibility of universal understanding (the existence of entirely inaccessible, un-understandable persons would refute that claim). The claim that mental objects only exist by means of their psychological and physical representations demands an entirely empirical approach that rules out any aprioristic talk about the mental. In other words, the just-described philosophical/scientific claims of the Aufbau are empirically meaningful because their failure would in each case imply the failure of the respective reducibility condition, and as a result of this failure, the empirical world would look entirely different.

4.4.  HOW PURELY STRUCTURAL IS THE NOTION OF OBJECTIVITY IN THE AUFBAU? The aim of the Aufbau is to achieve an entirely objective constructional system. Whereas in other systems of concepts (such as Husserlian phenomenology or sense data empiricism), the meaning of concepts becomes specified only by means of some irreducibly subjective aspects (phenomena in the sense of Husserl are irreducibly intentional, sense data in the sense of Hume, Mill, Russell, are irreducibly ostensive4), Carnap’s approach in the Aufbau is based on the representations of the logical values of concepts via PSDDs, which do not contain any subjective element and are therefore perfectly objective. However, this (purely logical) sense of objectivity is not the only sense of objectivity that can be found in the Aufbau. Rather, Carnap also wants to make the full meaning of each concept—the epistemic value rather than the logical value—objectively available. The logical value of a concept is purely structural, but it may cover not even an aspect of the epistemic value. Carnap notes that “a constructional transformation may, for example, turn a true, epistemically valuable statement into a triviality” (§ 50). It certainly was not the end goal of the Aufbau to produce objective definitions at the price of turning meaningful concepts into (cognitively useless) trivialities. In other words, it is not quite true that, as Alan Richardson puts it (closely following Michael Friedman), “the objective and the objectively communicable are exhausted by the structural properties of the objects of science” (Richardson 1990: 12, emphasis added). Though purely structural properties are of crucial importance for the Aufbau’s story about the route to objectivity, it is certainly not only the structural property of a concept that Carnap wants to make intersubjectively communicable and therefore objective. Carnap’s aim is instead to make the whole epistemic meaning of a

96

The History of Understanding in Analytic Philosophy

concept intersubjectively communicable and therefore objective. This implies that my reading of the following passage of the Aufbau seems to somewhat diverge from the readings of Friedman and Richardson: “Even though the subjective origin of all knowledge lies in the contents of experiences and their connections, it is still possible, as the constructional system will show, to advance to an intersubjective, objective world, which can be conceptually comprehended and which is identical for all observers” (§ 2). To achieve that goal, we first (but not exclusively) must develop purely structural definitions for all concepts (here I am in complete agreement with Friedman and Richardson). These definitions are (in themselves) both intersubjectively communicable and objective, of course; but they are by no means identical to the “intersubjective, objective world” which, according to Carnap, the constructional system allows us to make sense of. For Carnap, what makes a concept objective in the full sense that also includes its epistemic value is not merely the development of PSDDs (this is merely a first step), but also the procedure of “intersubjective correspondence” (“intersubjektive Zuordnung,” § 146), which allows us to map each concept c of each type to its intersubjective sphere. Intersubjective correspondence is a procedure that allows us to specify counterparts cM to c, for every instance of person M in the world of heteropsychological objects. This procedure relocates a given concept c in the autopsychological realm of a person M in such a way that it covers exactly the same meaning (epistemic value) that it initially covered in the autopsychological realm of M. In order to shift the concept from person M to another person N, we simply use the autopsychological basis as an intermediary step and consider both cM and cN. Intersubjective correspondence therefore allows us to distribute any concept c to the intersubjective realm; the epistemic value of c becomes available to each person M (provided that M is capable of dealing with our system at all). The whole constructional system appears to be a device for understanding, which is a hermeneutic tool. The account of “intersubjective correspondence” as given in Richardson (1998: 82–6) is perfectly compatible with my interpretation. There is no disagreement here with respect to the question of how intersubjective correspondence works. Disagreement may arise only with respect to the question of how the function of intersubjective correspondence works for the overall conception of a constructional system. It is true that in § 16 Carnap points out “that, for science, it is possible and at the same time necessary to restrict itself to structure statements.” But this does not imply that, for Carnap, science is exhausted by its structuralist side. According to Carnap, science must be purely structural as long as it is also something purely linguistic, such as a class of statements or deductions,

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

97

or the kind of thing that we might lay down in the form of a computer program. However, this is not the only side of science for Carnap. For him, science is also a cognitive undertaking which is carried out by human beings. Thus, for Carnap, even “in science” “cognition” takes place (“for the most part”) “intuitively and not in the rational form of logical deductions” (§ 100). Carnap at one point says that the “constructional system is a rational reconstruction of the entire formation of reality, which, in cognition, is carried out for the most part intuitively” (ibid.). The qualification that Carnap makes in this passage is not that he wants us to overcome intuition and replace it with logic. Referring again to the section just quoted, in § 143, Carnap unequivocally says that “the construction [i.e., the constructional system] does not represent the actual process of cognition in its concrete manifestations.” This leads to a “deviation of the construction from the actual process of cognition” (ibid.). In other words, the constructional system provides a rational reconstruction which is something quite different from “the actual process of cognition.” PSDDs do not represent actual concepts but do represent something else. PSDDs are not even similar to concepts. They only uncover their structural locus and their functional role in the whole process of cognition in order to allow us to build bridges between different realms of cognition, the most important of which are the bridges between the concepts in different minds. PSDDs are a tool that allows us to support understanding, but they are located in a realm that appears to be rather disjointed from cognition in itself. Carnap says that the “course of the process of cognition is not misrepresented by the construction” (ibid.), because construction ultimately has little to do with cognition. It instead represents an entirely different realm—viz., the realm of structural/analytic objects—whereas cognition is an empirical matter. “Construction” “reconstructs” the actual “course of the process of cognition,” and as Carnap points out at the end of § 143, these entities are “not even fictitious” (ibid.). And this is a qualification which seems to be important for the whole of the constructional system. The remark that immediately follows this passage unequivocally points out that what Carnap tells us here is to be taken as a fairly general observation with serious consequences for the whole of the constructional system: “These remarks hold, beyond the present problem [of production of signs], quite generally for all construction” (ibid.). In other words, we have good reasons to believe that this is a crucial passage of the Aufbau. Recall, on the other hand, Carnap’s statement in § 16 “that, for science, it is possible and at the same time necessary to restrict itself to structure statements.” This is an important observation, of course. But concepts belong to the constructional system in two ways. First, they belong to it as

98

The History of Understanding in Analytic Philosophy

something produced by the constructional system—viz., something purely formal or analytic—and not as something produced by the scientist who produces the constructional system—viz., something cognitive or synthetic. In this sense, concepts are PSDD. But these purely structural objects must be distinguished sharply from Carnap’s understanding of concepts as the sources of cognition which provide the external sources of the constructional system. These external sources are not produced by the constructional system but by human beings, and they are not analytical but cognitive! The constructional system is intended to help us to stabilize concepts in the latter fully-fledged sense of cognitive products of the human mind. To achieve this goal, we need PSDDs, to be sure. Second, there is also a variety of this process of stabilization which becomes visible inside of the constructional system and functions independent from the formal story of PSDDs. The end goal of the constructional system is the construction of “the intersubjective world” (§ 148). The fundamental qualification of the Aufbau is that “there exists a general one-to-one correspondence between all such systems, that is, between the worlds of all persons (i.e., normal persons known to me), including myself ” (ibid.). This intersubjective world “forms the actual object domain of science” (§ 149). Intersubjectivity in that sense is entirely detached from the structural story, because it is realized as a concrete empirical process of distribution of concepts among human beings. This qualification also becomes quite necessary if we survey the different places in the Aufbau where Carnap uses the term “objectivity.” This term is used together with “intersubjectivity,” which appears already in that passage where the former term first occurs (in the already cited § 2). The next time Carnap uses the term “objectivity” is in the famous (and already cited) § 16, this time with an explicit reference to “the problem of intersubjective reality,” to which he announces he will return “later.” Finally, in § 66, Carnap points out that “the precise method for achieving objectivity in the sense of intersubjectivity” will be demonstrated in §§ 146–9; that is, in the paragraphs already cited where Carnap introduces the notion of intersubjective correspondence that we identified as crucial for the constructional system as a whole. It is only the technique of intersubjective correspondence that allows us to establish objectivity in its fully-fledged form, whereas PSDDs function as auxiliary devices here. The richer form of objectivity as just characterized above points to an important (and somewhat neglected) side of Carnap’s thought, which is connected to the role that psychology plays in the Aufbau. As pointed out by Uljana Feest (2007), Carnap’s understanding of psychology in the Aufbau is neither based on Gestalt psychology nor on Brentano’s notion of intentionality. Rather, Carnap deals with elementary experiences in

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

99

the Aufbau which have strong similarities with conceptions of psychology developed by authors such as Hans Cornelius and Wilhelm Dilthey. Already the very first draft to the Aufbau—the abovementioned “Skeleton of Epistemology” from 1920—was based on “facts of consciousness,” a term that has been almost exclusively used by Dilthey and his followers. Given this and the fact that Carnap developed the 1920 draft in collaboration with intellectuals who all belonged to the Dilthey school, there is good evidence to suspect that the psychological basis of the Aufbau was informed by the Dilthey tradition. The crucial aspect of “facts of consciousness” as understood in this tradition is that they are taken as objective representations of subjective states (“The given does not have a subject,” § 65). They are taken to be “facts”—that is, there is no irreducibly subjective aspect, no “intentionality,” as in Brentano’s scholastic treatment of psychological states—but at the same time they are facts “of consciousness”—that is, something owed only by a certain subject. A certain “fact of consciousness” or, as Carnap calls it in the Aufbau, a certain “elementary experience” e is something that (like any other fact) occurs just once; e cannot be repeated, either in the same person’s or in another person’s stream of consciousness. Thus, a question comes up (and this was, of course, the crucial question of hermeneutics that involved Dilthey’s hermeneutic conception of objectivity) about how to gain access to such singular events by means of other events of consciousness. How can we map e to another mental event (“fact of consciousness,” “elementary experience”) e' in such a way that e and e', though not the same event, nevertheless might be identified as being identical in some way? Carnap’s answer to that question and his solution to Dilthey’s problem of hermeneutic objectivity is encapsulated in the notion of intersubjective correspondence as introduced in § 146 of the Aufbau. In conclusion, though the Aufbau’s most innovative contribution is certainly the idea of intersubjective communicability of concepts via PSDDs, it nevertheless serves another aim that is related to objectivity and intersubjective communicability; namely, the aim of gaining access to other minds on the basis of notions such as psychophysical parallelism and understanding via analogical conclusion. The reduction procedures of the Aufbau almost all rest on such hermeneutic notions. It also needs to be pointed out that the presence of these material notions of objectivity may allow us to uncover another quite interesting and innovative feature of Carnap’s structuralism. The strength of the formal story of the Aufbau increases as soon as we realize that this formal story serves a certain external, material purpose—viz., to gain objectivity for real concepts and not just their structural representations. Moreover, psychophysical parallelism and understanding via analogical conclusion are neither metaphysical nor

100

The History of Understanding in Analytic Philosophy

analytic, but empirical claims, because there must exist certain facts and laws of nature which allow us to uphold our fallible claims about other minds. This nontrivial, non-reductionist, and non-behaviorist variety of empiricism was continuously defended by Carnap in all later periods of his intellectual life. Taken from a more general and broader perspective, the story that is here told on the Aufbau would lead to an examination of “intersubjective accountability” in the philosophy of the left wing of the Vienna Circle (Uebel 2020), because the hermeneutic aspect of psychophysical parallelism and understanding via analogical conclusion is not only a key element of the scientific world conception, in general; it also becomes crucial for the left Vienna Circle’s attempt to develop a political philosophy of science. Finally, it is also worth noting that not only Carnap but also Otto Neurath committed to key notions of nineteenth-century hermeneutics (Uebel 2019; Damböck forthcoming). This demonstrates that Carnap’s respective commitment in the Aufbau was by no means only an early aberration, but became characteristic of all varieties of Carnap’s and the left logical empiricists’ philosophy.

4.5. ACKNOWLEDGMENTS This chapter was written for the most part already in spring 2015 during a stay at The Center for Philosophy of Science, University of Pittsburgh. For comments in the center’s reading group, I am grateful to William Bechtel, Sara Green, Wayne Myrvold, John Norton, Hamid Sayedsayamdost, and Raphael Scholl. The paper elaborates on those sections of my habilitation thesis that deal with the theoretical aspects of the Aufbau and their relationship with the Dilthey school (Damböck 2017: 164–90). It also is related to some extent to Damböck (forthcoming) and to a talk with the same title that was presented at a workshop in Pécs in 2018. For comments on this talk, I am grateful to Akos Sivado, Adam Tuboly, Thomas Uebel, and Andreas Vrahimis. Thanks, finally, to an anonymous reviewer of this chapter for extremely helpful comments and to Sammy Hirshland for language edits that highly improved the readability of this chapter.

NOTES 1 It is worth mentioning that Daston and Galison (2007), though correctly describing objectivity as an “epistemic virtue” that became crucial only in the second half of the nineteenth century, by no means deal with this second form of objectivity. For them, objectivity is exclusively a business of the natural sciences, in spite of the fact that for Dilthey and other main representatives of the humanities, the second, hermeneutic aspect of objectivity has been no less crucial for their scientific worldview than the first one.

A HERMENEUTIC ASPECT IN CARNAP’S AUFBAU?

101

2 This variety of psychophysical parallelism is weaker than the classical notion in Fechner, because unlike Fechner, Dilthey does not base his notion on a rejection of causal relationships between the mental and the physical. Carnap, as I will point out, also uses the notion in this weaker and more general form as it is used by Dilthey, not mentioning at all the mind-body-causality topic. 3 Archives for Scientific Philosophy, Hillman Library, University of Pittsburgh, Rudolf Carnap Papers, RC 081-05-04. 4 The sense data empiricism of Russell is not identical with the sense data empiricism of Hume and Mill; cf. Richardson (1990). While Hume and Mill take sense data as mental phenomena, Russell takes them to be physical entities. However, the different varieties of sense data empiricism as mentioned here seem to converge at least with respect to their “ostensive” nature.

REFERENCES Boeckh, A., E. Bratuscheck, and R. Klußmann ([1886] 1877), Encyklopädie und Methodologie der philologischen Wissenschaften, Leipzig: Druck und Verlag von B. G. Teubner. Carnap, R. ([1928] 1967), The Logical Structure of the World/Pseudoproblems in Philosophy, London: Routledge. Chalmers, D. (2012), Constructing the World, Oxford: Oxford University Press. Dahms, H.-J. (2016), “Carnap’s Early Conception of a ‘System of the Sciences’: The Importance of Wilhelm Ostwald,” in C. Damböck (ed.), Influences on the Aufbau, 163–85, Dordrecht: Springer. Damböck, C. (2012), “Rudolf Carnap and Wilhelm Dilthey: ‘German’ Empiricism in the Aufbau,” in R. Creath (ed.), Carnap and the Legacy of Logical Empiricism, 75–96, Dordrecht: Springer. Damböck, C. (2017), : Studien zur Philosophie im deutschsprachigen Raum 1830–1930, Dordrecht: Springer. Damböck, C. (2020), “(Dis-)similarities: Remarks on ‘Austrian’ and ‘German’ Philosophy in the Nineteenth Century,” in D. Fisette, G. Fréchette, and F. Stadler (eds.), Franz Brentano and Austrian Philosophy, 169–80, Cham: Springer. Damböck, C. (2021), “Die Entwicklung von Carnaps Aufbau 1920–1928,” in C. Damböck and G. Wolters (eds.), Der junge Carnap in historischem Kontext 1918– 1935 / Young Carnap in a Historical Context 1918–1935, 19–53, Cham: Springer. Damböck, C. (forthcoming), “Dilthey and Historicism,” in T. Uebel and C. LimbeckLilienau (eds.), Routledge Handbook of Logical Empiricism, London: Routledge. Daston, L., and P. Galison (2007), Objectivity, New York: Zone Books. Dewulf, F. (2017), “Rudolf Carnap’s Incorporation of the Geisteswissenschaften in the Aufbau,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 7 (2): 199–225. Dilthey, W. (1989), Introduction to the Human Sciences. Selected Works, Volume I, ed. with an introduction by R. A. Makkreel and F. Rodi, Princeton: Princeton University Press. Dilthey, W. (1990), Logik und System der philosophischen Wissenschaften. Vorlesungen zur erkenntnistheoretischen Logik und Methodologie (1864–1903). Gesammelte Schriften, Band XX. Göttingen: Vandenhoeck & Ruprecht.

102

The History of Understanding in Analytic Philosophy

Dilthey, W. (2002), The Formation of the Historical World in the Human Sciences. Selected Works Volume III, ed. with an introduction by R. A. Makkreel and F. Rodi, Princeton: Princeton University Press. Dilthey, W. (2011), Briefwechsel Band I: 1852–1882, Göttingen: Vandenhoeck & Ruprecht. Feest, U. (2007), “Science and Experience/Science of Experience: Gestalt Psychology and the Anti-Metaphysical Project of the Aufbau,” Perspectives on Science, 15: 1–25. Freyer, H. (1923), Theorie des objektiven Geistes. Eine Einleitung in die Kulturphilosophie, Leipzig: Teubner. Friedman, M. (1999), Reconsidering Logical Positivism, Cambridge: Cambridge University Press. Leitgeb, H. (2011), “New Life for Carnap’s Aufbau?” Synthese, 180: 265–99. Richardson, A. (1990), “How not to Russell Carnap’s Aufbau,” in A. Fine, M. Forbes, and L. Wessels (eds.), Proceedings of the Philosophy of Science Association Meetings 1990, 1: 3–14. Richardson, A. (1998), Carnap’s Construction of the World. The Aufbau and the Emergence of Logical Empiricism, Cambridge: Cambridge University Press. Schilpp, P. A., ed. (1963), The Philosophy of Rudolf Carnap, Chicago: Open Court. Sokal, E. (1904), “Das Salto-Mortale des Gedankens,” Annalen der Naturphilosophie, 3: 96–110. Tuboly, A. T. (forthcoming), “The Constitution of geistige Gegenstände in Carnap’s Aufbau and the Importance of Hans Freyer,” in C. Damböck, G. Sandner, and M. Werner (eds.), Logical Empiricism, Life Reform, and the German Youth Movement / Logischer Empirismus, Lebensreform und die deutsche Jugendbewegung, Cham: Springer. Uebel, T. (2019), “Neurath on Verstehen,” European Journal of Philosophy, 27: 912–38. Uebel, T. (2020), “Intersubjective Accountability: Politics and Philosophy in the Left Vienna Circle,” Perspectives on Science, 28: 35–62.

CHAPTER FIVE

More on Neurath on Verstehen: The Rejection of Weber’s Ideal Type Methodology THOMAS UEBEL

5.1. INTRODUCTION In “Neurath on Verstehen” (2019a) I argued for a broadened understanding of the social scientific thought of a leading logical empiricist and advocate of the unity of science. Often portrayed as a reductive physicalist on account of his temperamental interventions against geisteswissenschaftliche obscurantists at the time of the embattled first republics in both Germany and Austria, Otto Neurath in fact was a defender and practitioner of interpretive methods who warned against their abuse inside and outside of academic social science and history. Having placed Neurath’s efforts in the context of the Vienna Circle, in particular in the context of the ideas of his fellow physicalist Rudolf Carnap, I clarified his talk of “social behaviorism” and his dismissal of the methodology of empathetic understanding (Verstehen) as grounds for the sharp separation of social from natural science under the guise of Geisteswissenschaft. Both of these points represent prominent stumbling blocks for a proper appreciation of his views. I argued that what, in the case of

104

The History of Understanding in Analytic Philosophy

Carnap, are widely taken to be reductive definitions of types of mental states in purely behavioral terms were in fact inductive generalizations intended to provide defeasible criteria for third-person mental state attributions, and that this reading is readily extended to Neurath’s “physicalism” and “social behaviorism.”1 Rejection of intentionality or of the use of the intentional idiom was neither intended nor entailed; what Neurath rejected was an antiphysicalist, dualist understanding of mind.2 Yet Neurath also claimed that Max Weber’s “interpretive sociology” (verstehende Soziologie) abandoned the empiricist strictures on scientific theorizing.3 This criticism focused on seemingly metaphysical turns of Weber’s methodology (encouraged by his claim to be incorporating ideas by Heinrich Rickert) and on the mystifying use made of empathetic understanding by advocates, like Werner Sombart (and others like Max Scheler), who linked it to an altogether different ontological sphere of “spirit.” This context makes Neurath’s rejection of interpretive sociology understandable, but it does not undo his failure to appreciate the empirical ground that Weber’s own interpretive method strived for throughout.4 (Weber’s interpretivism was not a WittgensteinWinchean anti-causalism.) Importantly, however, this failure did not entail a wholesale rejection of interpretive methods in social science or history on the part of Neurath. To establish this, I reconstructed Neurath’s view on the true role of empathetic understanding from remarks that all along can be found in his Empirical Sociology but which are often overshadowed by misreadings of his arguments against separatist Geisteswissenschaft. Further support was then adduced from various instances of Neurath’s pre-Vienna Circle work, including his critique of Spengler and his still earlier work in economic history and the history of ideas. Overall these excavations strongly support the thesis that having been trained as a historian, Neurath never abandoned his recognition of the hermeneutic insights and practices that are basic to engaging in historical research, but sought to integrate these in his philosophy of social science— which, in turn, he regarded as an integral part of science as a whole, of “unified science.” Yet to be clear: while Neurath’s philosophy of science made room for hermeneutic practices in the social sciences, it did not designate the latter as exclusively hermeneutic, as some advocates of interpretive sociology often had it. His was a pluralistic version of the methodology of the social sciences. Neurath’s notion of unified science, too, was a non-dichotomous conception: it countenanced no categorical distinction of Naturwissenschaft and Geisteswissenschaft as was commonly pronounced to be the case in German-language philosophy since the nineteenth century, prominent exponents being Dilthey and the neo-Kantians Windelband and Rickert. In

MORE ON NEURATH ON VERSTEHEN

105

philosophy of science terms, this denial of a categorical distinction makes Neurath’s position a naturalistic one: the human sciences are held to be no less empirical and able to seek causal explanations than the natural sciences. Often, what is understood by the term “naturalism” is a position that ascribes dominance to the natural sciences as providing the example of scientific propriety which the social sciences have to follow if they are to count as sciences at all. (This is also what is sometimes condemned as “scientism.”) As noted, what distinguishes Neurath’s early naturalism is that it did not involve—despite what some rather careless formulations of his may suggest5—the reduction of the terms and statements of the languages of the different individual sciences to those of the language of one basic science, physics. Importantly, Neurath’s unified science respected the autonomy of the individual sciences.6 This is an important aspect of what is meant when his conception is called a version of “non-reductive naturalism.” It is clear then that the integration of everyday hermeneutics and the elaboration of their scientific application is a central component of Neurath’s non-reductive naturalism. Now it might be granted that Neurath’s intentions did indeed lie in this non-reductive direction, as argued, but the suspicion may remain that he was altogether mistaken in thinking that something like his empirical sociology was required in place of Weber’s conception. It is this skepticism that I wish to address here; namely, that Neurath misunderstood Weber even more badly than conceded so far and that there was no need for an alternative conception. To answer this concern, I investigate his criticisms of Weber’s methodology for the social sciences more generally. Weber’s separatist claim was that the social sciences employ so-called ideal types and that this grounds their claim to autonomy from the natural sciences.7 To show that Neurath did have reasonable grounds to reject Weber’s arguments and seek an alternative, Weber’s theory of ideal types must therefore be reconstructed first. Before I begin, let me note that already in an early work, Neurath judged that the essay in which Weber introduced ideal types contained “much insight besides some mistakes” ([1910] 2004: 291, n. 33). Whether or not this remark bespeaks youthful arrogance, his was not an across-the-board rejection of Weber’s ideas.8 What is crucial, however, is that already then Neurath rejected the separation of natural from social science.9 It is clear, therefore, that Weber’s ideal type methodology must have come in for criticism that unfortunately remained unspecified at the time and received hardly any elaboration in later years. A fair amount of excavation is required, therefore, to see what his reasons were, especially given the complexity of Weber’s theory, which demands a very careful reconstruction in its own right.10

106

The History of Understanding in Analytic Philosophy

5.2.  IDEAL TYPES AND IDEAL TYPE THEORY Weber introduced the notion of ideal types as characteristic for social scientific theorizing in his programmatic 1904 paper “The ‘Objectivity’ of Knowledge in Social Science and Social Policy.” Its opening remarks on the editorial policy of the journal in which it appeared were followed by a detailed discussion of social scientific methodology.11 After a long discussion of the value-ladenness of the very concept of culture (more on this in Section 5.3), Weber turned to “the question that is of interest to us from a methodological point of view when we are dealing with the ‘objectivity’ of cultural knowledge: […] what is the importance of theory and of theoretical concept formation for the knowledge of cultural reality?” ([1904] 2014: 121). Weber’s answer was to roundly reject “the naturalistic prejudice, according to which one must, by establishing those concepts, create something akin to the exact natural sciences,” a prejudice which “had simply led to a misconception of the nature of those theoretical concepts” (ibid.: 123). The employment of ideal type concepts set history and the social or, as Weber preferred, the “cultural” sciences apart from the natural sciences. Weber’s example was economics. Interestingly, not only the British empiricist John Stuart Mill and his followers but also those of Carl Menger, the founder of Austrian economics and foremost opponent of the Historical School—whose inductivism Weber also castigated sharply—fell under this censure.12 The reason was that their “ ‘abstract’-theoretical method” held “the supreme goal of science” to be the pursuit of “laws (in the strict sense)” and so presupposed “that we ourselves have constant and immediate experience of the actual interrelations of human action and therefore […] are able, with axiomatic evidentness, to render the course of such action directly understandable, and by this means to uncover the ‘laws’ that govern it” (ibid.). Mengerian aprioricism as regards the fundamental concepts used, instead of empiricist abstractionism, made no difference. The fatal mistake was to hold that “the only way to gain intellectual mastery of the multiplicity of social life, at least as far as the fundamental phenomena of social life are concerned, must be to draw up a system of abstract and—therefore—purely formal propositions, analogous to those of the exact natural sciences” (ibid.). Weber’s bête noire was the idea that social life was to be captured by universal and abstract theoretical laws that allowed the deduction of events.13 It also was of no help at all, according to him, to indulge in the fantasy of breaking down social reality into a play of different factors with a separate theory for each of the myriad motives of human action, arriving at a mosaic conception of the interrelation of the sciences such that “the totality of these theories would contain the true reality of

MORE ON NEURATH ON VERSTEHEN

107

things”; for instance, that “exact economic theory would determine the consequences of one psychological motive” and “it would be the task of other theories to develop all the other motives in a similar way” (ibid.). For Weber, the key lay in appreciating the difference of “abstract theory” in social science. “The propositions of abstract theory are only seemingly ‘deductions’ from fundamental psychological motives; in actual fact, they are rather a special case of a kind of conceptual construction that is specific, and to a certain extent indispensable, to the sciences of human culture” (ibid.: 124). Weber offered an original version of the by this time conventional distinction of Geisteswissenschaften and Naturwissenschaften. Instead of following Dilthey in postulating a distinct type of psychology to underwrite his contrast of empathetic understanding (Verstehen) and explanation (Erklären), Weber extended the views of the neo-Kantians Windelband and Rickert, who not only spoke of Kulturwissenschaften in place of Geisteswissenschaften but also located the contrast in the type of concept formation characteristic of each. Windelband had contrasted the generalizing concepts of the nomothetic sciences, which allow individual events to be subsumed as tokens of a general type and thereby fall under laws, and the individualizing concepts of the idiographic sciences which sought to capture the very individuality of single events or agents. Rickert deepened the analysis of these individualizing concepts as value-laden and elaborated that it was by virtue of this value relation, which the occurrences at issue bear to investigators who focus on them, that it was the very particulars at issue that mattered, not their genus. Weber accepted much of this but saw the need to go further to avoid having to distinguish between generalizing and individualizing Kulturwissenschaften as Rickert had to, depending on whether the token or the type was primarily valued on the occasion.14 Of what distinctive nature then are the concepts of social science? According to Weber, they are “ideal types,” the use of which gives rise to a distinct type of theory. The abstract economic theory offers us an example of those syntheses that are usually called “ideas” of historical phenomena. It presents us with an ideal image of what goes on in a market for goods when society is organized as an exchange economy, competition is free, and action is strictly rational. This mental image brings together certain relationships and events of historical life to form an internally consistent cosmos of imagined interrelations. The substance of this construct has the character of a utopia obtained by the theoretical accentuation of certain elements of reality. Its only relation to the empirically given facts of [real] life is the following: if it is established or assumed that interrelations of the same

108

The History of Understanding in Analytic Philosophy

kind as those represented in abstract form in the construct […] to some extent operate in reality, then we can pragmatically clarify the distinctive character of that interrelation and make it understandable, by means of an ideal type. This possibility can be important, and even essential, both for heuristic purposes and as an aid to exposition. (Ibid.: 124–5) For Weber, ideal types are building blocks for theoretical models of a highly specific sort. Ideal types and ideal type theories are characterized by their accentuation of only certain features of aspects of social reality. Due to the high degree of idealization involved, these theories cannot be regarded as realist depictions, but only as instruments for a partial and “pragmatic” understanding of that reality.15 Here a first worry may surface. Why should this accentuation of certain aspects be detrimental to a realistic understanding? Is that not what all sciences do? And is this not just an instance of the phenomenon of scientific specialization that the mosaic conception set out to address? Weber evidently did not think so. Why is not yet clear, but a possible reason will emerge presently. Weber quickly added that ideal types are not to be understood as ideal in a normative sense, as expressing a standard to be emulated.16 Theirs is a broadly descriptive office—only “broadly” because due to their selective accentuation of features, they are not identical to the “dominant ideas” of an epoch (however much we may use them to explicate the latter).17 In research, the ideal type seeks to render the scholar’s judgement concerning causal imputation more acute: it is not a “hypothesis”, but it seeks to guide the formulation of hypotheses. It is not a depiction of reality, but it seeks to provide [the scientific] account with unambiguous means of expression. […] This is done not by forming the concept “city economy” as, say, the average of the economic principles actually to be found in all the cities one examines, but rather, again, as an ideal type. It is obtained by means of one-sided accentuation of one or a number of viewpoints and through the synthesis of a great many diffuse and discrete individual phenomena (more present in one place, fewer in another, and occasionally completely absent), which are in conformity with those onesided, accentuated viewpoints, into an internally consistent mental image. In its conceptual purity, this mental image cannot be found empirically anywhere in reality. It is a utopia, and the task of the historian becomes that of establishing, in each individual case, how close reality is to, or how distant it is from, that ideal image. […] Provided it is employed with care, the concept [of the ideal type] renders specific services for the purposes of research and exposition. (Ibid.: 124–5)

MORE ON NEURATH ON VERSTEHEN

109

Social phenomena are human phenomena in history and come in endlessly graded varieties. The range of their variability may not differ from that of natural phenomena; the difference lies, rather, in how these phenomena are conceptualized: they are typed differently. The basic concepts of theories of social life are not formed, like generic concepts, by “generalising abstraction” (ibid.: 122) or even by averaging certain salient values.18 Rather, they are formed by “one-sided accentuation” and “synthesis of a great many diffuse and discrete individual phenomena” (ibid.: 125) The one-sidedness of ideal types is a very much a condition of serving their theoretical function: they unify and connect what to the untutored eye are unrelated occurrences and so bring to attention what previously was allowed to escape it. The distinction between ideal type concepts and the “generic” concepts employed in the natural sciences turns on this functional difference. The ideal type is chiefly an attempt to comprehend historical individuals, or their component parts, in genetic concepts. Let us take the concepts “church” and “sect”. By way of [generic] classification, they may be broken down into complexes of characteristics; but in that case, not only the boundary between the two concepts but also their substantive content will necessarily remain fluid. However, if I want to comprehend the concept of “sect” genetically, for instance with regard to certain ways in which the “sectarian spirit” has been particularly important for modern culture, then certain characteristics of both concepts become essential because they stand in an adequate causal relationship to those effects. At the same time, however, the concepts become ideal-typical—that is to say: [empirically,] they cannot or can only rarely be found in [their] completely pure conceptual form. Here, as everywhere, any concept that goes beyond mere classification leads away from reality. (Ibid.: 127) Ideal types are theoretical in a strong sense: not only do they designate what cannot be simply observed, but they also incorporate the theory of which they are part into their definition of phenomena by presupposing, when they are applied, the obtainment of certain causal relationships between related phenomena under investigation. It is this that distinguishes ideal types and renders them less than straightforwardly empirical even though they are composed entirely of straightforwardly empirical elements: “every individual ideal type is composed of conceptual elements that are generic and take the form of ideal types” (ibid.: 131). Ideal types are not wholly categorically different then from the empirical concepts of science. Even so, for Weber, “an ‘objective’ treatment of cultural occurrences, in the sense that the ideal aim of scientific work would be to reduce the

110

The History of Understanding in Analytic Philosophy

empirical [reality] to ‘laws’, is absurd” (ibid.: 119). Taken together with his remarks about the mistaken expectation to be able to “deduce” real phenomena, this suggests that it is nomological explanation—in particular, strict nomological explanation—that Weber thought inappropriate for social science. This would certainly distinguish it from the natural sciences. To the characteristic of ideal type theory being merely instrumental, we may thus add that of being not strictly nomological. In one way, this feature of ideal type theory is puzzling, but ultimately it is deeply revealing. We recall that Weber was not adverse to causal explanation in the social sciences. Neurath was evidently aware of Weber’s objection to nomological explanation in social science, but what he overlooked was that this was not the end of the matter. Weber took the statistical plausibility of the obtainment of the causal connections about to be imputed—clearly an empirical matter—to serve as a controlling instance for whether interpretations that appear adequate on the level of meaning should also be considered causally adequate and therefore fully acceptable (see Weber [1906] 2014: 175–81 and [1921] 1978: 11–12). Importantly then, for Weber, causal explanation in the social sciences did not rely on having established strict laws between types of events, but on the employment of the probabilistic notion of “objective possibility,” according to which the likelihood of an event type was regarded as a function of objective conditions being fulfilled.19 For Weber, it seems, the social sciences nevertheless could not help but be less realist in their ambition than the natural sciences, because in their case, mere probabilities were the only ground of causal imputation. Again one may wonder whether this is reason enough to declare these sciences type distinct. It is important therefore that this lack of strict nomologicality on the part of the social sciences is considered in the context of the distinctive relation they bear to their subject matter. Thus Weber noted, “the ‘one-sidedness’ and unreality of a purely economic interpretation of historical material is only a special case of a principle that is generally valid for the scientific knowledge of cultural reality” ([1904] 2014: 113). What is this principle and what bearing does it have on the validity of ideal type theory and the separation of the natural from the social sciences?

5.3.  THE VALUE-RELATEDNESS OF IDEAL TYPE THEORY For the principle governing all the sciences of cultural reality, we have to turn to the earlier part of Weber’s essay that elaborates the idea of their value-relatedness. Weber stated bluntly that “there is no absolutely ‘objective’ scientific analysis of cultural life … independent of special and ‘one-sided’

MORE ON NEURATH ON VERSTEHEN

111

points of view, according to which [the relevant phenomena] are … selected as an object of inquiry, analysed and presented in an orderly fashion” (ibid.: 113, original emphases). Here, Weber clearly equated absolute objectivity with absence of perspective: this equates to what in more recent times has been called the “absolute conception of reality,” “the view from nowhere,” or “the viewpoint of no-one in particular.”20 Needless to say, it is questionable whether this idea of absolute objectivity has application anywhere, but let’s hold this objection for the moment as well. The point to focus on is that it is having a perspective on things, perspectivality as such, that was deemed by Weber to make for the distinguishing characteristic of social or cultural science. We want to understand the distinctive character of the life in which we are placed and which surrounds us—on the one hand: the interrelation and the cultural significance and importance of its individual elements as they manifest themselves today; and, on the other: the reasons why these elements historically developed as they did and not otherwise. (Ibid.: 114) Here we must ask: what precisely is “cultural significance and importance”? It is clearly related to perspectivality, but given that causal explanation too is interest relative, this does not yet exclude causal inquiry unless the “understand” were weighted in the anti-causalist Diltheyan fashion (which, however, is precluded by Weber’s probabilist causalism). So Weber may have merely misunderstood natural science, not realizing that the interest relativity of causal explanation in social science is not unique to it. However likely, this does not yet touch on the core of the matter. More is involved in the distinctive perspectivalism of the cultural sciences than interest relativity: “what is crucially distinctive about the approach adopted by the cultural sciences” is that their “aim is to acquire knowledge about the culturally significant aspects of the manifestations of life” (ibid.: 115). Weber did not elaborate, but it would appear that what makes for cultural significance and importance and allows its investigation the kind of non-absolute objectivity it does still possess (let’s call it “perspectival objectivity”) is its nonarbitrariness. The social or cultural sciences bear a value relation to their objects; moreover, something about this value relation makes this interest nonarbitrary and sets it apart from causal contingency. Neither the significance of the configuration of cultural phenomena nor the basis [of that significance] can be deduced, explained or made comprehensible by any system of law concepts, however perfect, since that significance presupposes the relation of cultural phenomena to

112

The History of Understanding in Analytic Philosophy

value ideas. The concept of culture is a value concept. Empirical reality is “culture” to us because, and to the extent that, we relate it to value ideas; it comprises those, and only those, elements of reality that acquire significance to us because of that relation. Only a tiny part of the individual reality that we observe at a given time is coloured by our interest, which is conditioned by those value ideas, and that part alone has significance for us; it has significance because certain of its relations are important to us by virtue of their relation to value ideas. (Ibid.: 115–16) So what are these specific human interests or value perspectives that are at issue? Weber did not directly address the first question, but he explored what he regarded as the universality of the value relation.21 To begin with, Weber elaborated the presuppositional nature of this value relation, what he called its “transcendental” status: The transcendental precondition of every cultural science is not that we find a particular, or indeed any, “culture” valuable, but that we are cultural beings, endowed with the capacity and the will to adopt a deliberate position with respect to the world, and to bestow meaning upon it. Whatever this meaning may be, it will become the basis on which we are, in our life, led to judge certain phenomena of human existence in common and to adopt a (positive or negative) position with respect to them because we regard them as significant. No matter what this position may be, these phenomena possess cultural significance for us, and it is upon this significance alone that their scientific interest rests. (Ibid.: 119) Here, this cultural value orientation is affirmed for all humanity. Again no specific values were named, and it was stated instead that even if in a particular case it were determined what these basic value ideas are—irrespective of whether they hold for just one or all cultures—it does not follow that the individuals under investigation or the investigators themselves subscribe to them.22 One cannot help but wonder whether this freedom was a function of the merely ideal-typical nature of these values or whether the universality in question was that of a social norm from which deviation was possible. Had Weber left it there, he would only have provided, in his own terms, a “formal” characterization that does not measure up to the substantive role these values are supposed to play. But he added: Consequently, all knowledge of cultural reality is always knowledge from specific and particular points of view. When we demand of historians and social scientists, as an elementary prerequisite, that they must be able

MORE ON NEURATH ON VERSTEHEN

113

to distinguish between what is important and what is unimportant, and that they must possess the “points of view” necessary for making this distinction, this simply means that they must be able—consciously or unconsciously—to relate what happens in [empirical] reality to universal “cultural values”, and, on this basis, to select those relationships which are significant for us. (Ibid.) Here for once Weber spoke of “universal ‘cultural values’” (emphasis added). But again we must ask: what are they? Was Weber taken in by the values peculiar to Germany’s academic “mandarins,” that his philosopher friend Rickert designated as universal,23 or by the values of state-building and of the “great men” directing it, that the historical profession at the time was much concerned to defend against materialist upstarts? Yet having designated them “universal,” Weber left these values unspecified.24 Still, what Weber did specify is enough to diagnose an unresolved tension in his view: the tension between his belief in the universal validity of cultural values and the realization that this universal validity lies beyond anything empirical science can establish. It is this very tension, of course, that animates Weber’s accommodation of the objectivity of social science amid the value pluralism of modernity via the twin concepts of “value freedom” and “value relevance.” Scientists can choose research programs that possess relevance to the cultural values they adhere to (including political ones), but their scientific results cannot include the unconditional endorsement of any such values.25 Weber’s views on the value freedom of science have long been controversial but are not at issue here. Suffice it to say that given the impossibility of establishing unconditional norms empirically, Weber’s conception not only accommodates the tension involved, but also allows ample room for socially engaged and even politically relevant research while respecting the demand to keep scientific pronouncements free of value statements that exceed an empirical base. (Without the Rickertian value metaphysics that this conception of scientific value freedom is, for better or worse, associated with in Weber, this is the position that was also adopted by Neurath [1913] 2004 and Carnap 1934.) The question that remains is whether the tension between Weber’s belief in the universal validity of cultural values and his realization of the impossibility of establishing this validity scientifically can be resolved equally well with regard to his ideal type theory.

5.4.  THE OBJECTIVITY OF SOCIAL SCIENCE There is one reading that would solve Weber’s difficulty. The import of the value relation for cultural science lies in how it delimits its field of investigation

114

The History of Understanding in Analytic Philosophy

and how it plays out in the life of the subjects investigated, in their valuations (and their consequences); it does not lie in what the nature and standing of the values may be that the subjects attach to the objects of their attention. It is this shift of attention that allowed Weber to claim perspectival objectivity for social science. What made this relocation of objectivity possible was the dual brief he had drawn up for the social sciences: to investigate “the interrelation and the cultural significance and importance” of the elements of contemporary social life and “the reasons why these elements historically developed as they did and not otherwise” ([1904] 2014: 114). Having drawn attention to the constitutive role that value relations play for social science, Weber declared the status of any particular value thus brought into play irrelevant to the validity of investigative findings. The search for knowledge in the field of the cultural sciences (in our sense) is therefore tied to “subjective” preconditions insofar as it is only concerned with those parts of reality that are—however indirectly— connected with occurrences to which we attach cultural significance. Nevertheless, the knowledge [that we are seeking] is of course purely causal, in exactly the same sense as knowledge of significant individual natural occurrences that have qualitative character. (Ibid.: 120) The “knowledge about the culturally significant aspects of the manifestations of life” (ibid.: 115) that Weber sought did not concern the nature of this significance. The role of cultural significance is different: it does not serve as the object of inquiry but as what frames it. Cultural significance or judgments thereof select what is to be causally explained; namely, how the phenomenon came about and what its consequences are. According to Weber, here, it is this causal explanation—as we shall see presently, the genetic account—of social phenomena that constitutes the aim of social scientific investigation. To be sure, significance in a more limited, strictly subjective sense also plays a role in the provision of the causal explanations of social phenomena. Thus Weber’s famous definition of sociology as a science which attempts the interpretive understanding of social action in order to thereby arrive at a causal explanation of its course and effects. In “action” is included all human behavior when and in so far as the acting individual attaches a subjective meaning to it. Action in this sense may be either overt or purely inward […] it may consist of positive intervention in a situation, or of deliberately refraining from such intervention or passively acquiescing in the situation. Action is social in so far as, by virtue of the subjective meaning attached to it by the acting individual

MORE ON NEURATH ON VERSTEHEN

115

(or individuals), it takes account of the behavior of others and is thereby oriented in its course. (Weber [1921] 1978: 4) But the explanatory role of subjective meaning—in other words, the recognition of the indispensability of intentional descriptions of agents in explanations of their actions—is very different from the role that judgments of cultural significance played in setting up what is to be explained in social science in the first place. On the present reading, then, the role of cultural values in interpretive social science is fundamental, but social scientists are entirely in order to leave unanswered the questions of what they are and what precisely their standing is. To be sure, metaphysicians demand to go further and investigate the nature of the constitution afforded by value relations and the resultant domain, but a theorist and methodologist of science like Weber can stop right there. So it looks like the tension between Weber’s belief in the universal validity of cultural values and his realization of the impossibility of establishing any one such value or their validity in general by scientific means can be resolved equally well with regard to his ideal type theory. The validity of these cultural values—be they universal or parochial—is not an issue for the social or cultural sciences after all, whatever their motivating power may be.26 Weber concluded: Undoubtedly, the value ideas are “subjective”. Between the “historical” interest in a family chronicle and the “historical” interest in the evolution of the greatest conceivable cultural phenomena that over long periods have been, and still are, common to a nation or mankind, there is an infinite scale of “significance”. And in the same way, they will naturally be subject to historical change, with changes in the character of the culture and the ideas governing human beings. But obviously, it does not follow from this that research in the cultural sciences can only have results that are “subjective” in the sense of being valid for one person but not for another. Rather, what varies is the degree to which such results interest one person and not another. In other words: what becomes the object of investigation, and how far this investigation extends into the infinity of causal relationships—that is determined by the value ideas that govern the scholar and are dominant in his age. As far as the method of investigation is concerned—how it proceeds—the guiding “point of view” (as will become apparent) determines the formation of the conceptual tools employed by the scholar; but in applying these tools, he is obviously, here as everywhere else, bound to the norms governing our thought: scientific truth is only that which claims validity for all who seek truth. (Ibid.: 121)27

116

The History of Understanding in Analytic Philosophy

Irrespective of how subjective the constitutive value relation may be, the inquiry itself can claim truth-valuable validity for its results (what earlier we called “perspectival objectivity”). Yet while there is much to be said for the empiricist reading provided so far, it must be noted that not everything Weber wrote on the matter fits well with it. He offered not only formulations that are highly ambiguous—for example, “the aim of ideal-typical concept formation is always to bring out clearly what is distinctive, and not what is generic, in cultural phenomena” (ibid.)—but also inconsistent ones like the following assertion in the final paragraph of his “Objectivity” paper: “Nothing should be emphasized more strongly than the following proposition: helping to acquire knowledge of the cultural significance and importance of concrete historical relationships is the single, exclusive, ultimate goal [of social science], for the attainment of which the formation and critique of concepts is one of the tools” (ibid.: 137– 8). Here, Weber clearly reaches beyond the boundaries of empirical social science: it is the determination of what the cultural significance of concrete historical phenomena is, what it is for culture and society at large, that here falls to social science as one of its tasks. Now what is a reader to conclude concerning this evident contravention of empiricist strictures? For those who take Weber’s references to Rickert at face value, it is surely this: the determination of cultural value reoccupies the central place it held in neoKantian Kulturwissenschaft from which it seemed to have been banished for the sake of retaining perspectival objectivity by Weber. Here, the tension between his (elsewhere suppressed) substantive value beliefs and his claim to empirical probity snapped in contradiction of the latter.

5.5.  NEURATH’S CRITICISMS It is precisely this tension resulting in a clear contravention of empiricist strictures that Neurath’s criticism focused on. Concerned as ever to establish social science as empirical and part of physicalist unified science and to differentiate it clearly from competitor conceptions that either offer or appear to offer what are compromising accommodations of idealist tendencies, Neurath focused on the very points that a sympathetic reading of Weber, less invested in the methodological battles of the day, would gloss over. As ever, he did not easily forgive anti-empiricist sins, however much they may show more carelessness than metaphysical intent. Neurath’s basic tenor throughout the years was that the “dualism ‘natural sciences’-‘mental sciences’, the dualism ‘natural philosophy’-‘cultural philosophy’ is, in the last resort, a residue of theology” ([1931b] 1981: 69). This persistent criticism was epistemologically grounded. The same type

MORE ON NEURATH ON VERSTEHEN

117

of point applies to the norms of categorization in scientific practice. Of all scientific categories, we must ask what their legitimacy, their normalizing force for social scientific theory construction, depends on. For Neurath, the goal of scientific inquiry into any subject matter most broadly is to increase our ability to predict the phenomena involved, however imperfectly; and this ability was regarded as dependent on the empirical grounding of the systems of concepts used to comprehend the subject matter. Unlike for Aristotelians or neo-Kantians, for Neurath their cognitive authority cannot derive from some supposed deeper insight into the nature of the phenomena themselves, either as intuited essences or as synthetic a priori ideas that must be presupposed. This, of course, is simply to affirm empiricist sensibilities: how are they offended by ideal type methodology? One potential conflict readily comes to mind. “We cannot start from a ‘human being as such’, but only from historically formed human beings who stand in specific historically given connections” ([1931a] 1973: 399, trans. amended). Neurath rejected a presupposition that found expression in the ideal type of agent employed throughout neoclassical and Austrian economics. The ideal type of the homo economicus, the self-interested individual endowed with the economic motive, could not be appealed to without further ado: at a minimum, its use will have to be justified on a caseby-case basis, and there is no guarantee at all that it can be so justified. To assume otherwise, that “a certain determinate way of acting is presupposed,” ultimately leads to theoretical “paralysis” ([1910] 2004: 273–4).28 Weber would not disagree with Neurath here, but Neurath also did not have Weber in mind, for his opponent in this passage was Wilhelm Wundt, whose use of the homo economicus concept was described in terms that fit very well with what Weber characterized as the mistaken psychologizing of ideal type methodology.29 Moreover, Neurath himself did not oppose all theoretical idealizations, or “auxiliary constructions” as he called them, but only certain types of use of ideal-typical concepts; for instance, Carl Menger’s purely theoretical, abstract types. Since Menger did not require his abstract types to have any complements in empirical reality, he was unable to ensure the possibility of empirical verification: his economics were an a priori construction whose applicability was in question (see Menger [1883] 1963: 62). To be sure, Neurath too reserved the right to theories with the help of abstract constructs that did not have a historical realization, especially in his investigations of alternative economic orders. The important difference here is that Menger’s homo economicus is a primitive concept, not built up from empirical elements, whereas Neurath’s imaginary systems of alternative economic institutions are composed of elements and relations that themselves are “as a matter of principle empirical” ([1911] 1998: 493).30 In other words, whereas the former

118

The History of Understanding in Analytic Philosophy

represents an intuited essence (of either an Aristotelian or a Kantian type), the latter are constructed from empirical elements. Some types of ideal types, then, Neurath could reject as plainly unempirical, but Weber’s escape this criticism. Neurath’s rejection cannot rest on the fact that they represent, by Weber’s own admission, “utopias,” abstractions that are not present in empirical reality, for his too are built from empirical generic concepts, as we have seen. Nor could his criticism rest on the claim that Weber’s ideal-typical theories resist testing in principle for, as noted, Weber insisted on being able to test the causal adequacy of his hypotheses. Neurath’s complaints, finally, that “with Weber, again and again, the empathetic immersion appears in place of science” and that “over this poetic activity there is no test, no control, nor does it belong in a scientific account” ([1931a] 1973: 357) also seem less than fully justified. It is true, of course, that Weber claimed that “the social sciences are concerned with the influence of processes in the human mind; and the re-experiencing ‘understanding’ of such processes is of course a task that is specifically different from that which the formulas building on the exact knowledge of the natural world are at all able, or designed, to solve” (Weber [1904] 2014: 115). But it could be argued that it is precisely the use of ideal types that distinguishes Weber’s social scientific methodology from that of Dilthey: it introduces conceptual and potentially multidisciplinary theoretical mediation in place of empathetic directness and its grounding in a new psychology. The mention of “re-experiencing ‘understanding’ ” on Weber’s part need not be read as an endorsement of Dilthey’s methods, but is better understood as a mere tip of the hat to a major predecessor, indeed to a large part the originator of the philosophical tradition very broadly conceived that Weber signed up to. Yet when we dig deeper, we find that Weber’s ideal-typical methodology is not free of fault. We saw that his appeal to the role of value relatedness of social science did not remain wholly innocent. Here is another instance: The relation of reality to value ideas which lend it significance, and the selection and ordering, according to their cultural significance, of the parts of reality coloured by this [value relation] form an approach that is completely heterogeneous and disparate from [that of] analysing reality with a view to finding laws, and ordering it in general concepts. No necessary logical relations exist between those two kinds of intellectual ordering of reality. (Ibid.: 117) Passages like this problematize readings of Weber’s defense of the objectivity of social science that relegate the importance and relevance of what he

MORE ON NEURATH ON VERSTEHEN

119

deemed its constitutive value-relatedness, metatheoretically, to the selection of problems to be worked on and, on the object level, to the causal nexus under investigation itself. The assignment to social science to provide an “ordering, according to their cultural significance, of the parts of reality coloured by this [value relation]” goes well beyond this. What this requires is a deeper, more substantial engagement with these values, which, after all, Weber deemed “universal.” This demand is problematic in more than one sense. Quite obviously it undermines the carefully constructed criterion of objectivity for social science, for it entails transgression of the, in other respects, carefully constructed border between intersubjectively testable science and philosophy. It also calls into question the possibility that social science should be pursued by empirical means, for more is required than these can provide. Social phenomena cannot be ordered according to their cultural significance unless that significance is known, and the resultant ordering cannot be considered valid unless it is intersubjectively validatable. Just that is, of course, extremely questionable—and evidently was questioned by Weber himself.31 Inevitably, all this puts enormous pressure on the particular conception of the value-relatedness of the cultural sciences that Weber borrowed from Rickert. Neurath was aware of this, stating clearly: “From Dilthey the path leads through Windelband and Rickert to Weber, who is clearly conscious of his own outlook: ‘It is a purpose of this study [Roscher and Knies] to test the applicability of Rickert’s thoughts to the methodology of our science’ ” ([1931a] 1973: 357; quote from Weber [1903–6] 2014: 6, fn. 7). Moreover, Neurath’s own position on value freedom in science reflected this directly. As noted, deflationarily construing the notion of value involved as turning on the cognitive interest on the part of researchers all along, Neurath was happy to accept Weber’s concept of value relevance in choice of research topics alongside the demand for freedom from unconditional value statements in the presentation of scientific results.32 What this points to is that it is the metaphysical weight that Rickert put on the value relation and which Weber apparently shed only partially that Neurath’s criticism correctly focused on. Accordingly, the charge reads as follows: construed as something more substantial than the interest relativity of research topics that the social sciences share with the natural sciences, the value relation that the cultural sciences are said to bear to their objects undermines their status as empirical sciences by demanding they go beyond what intersubjective validity can be established for. Since Weber did not stick to limiting the role of social science’s value-relatedness to mere guidance in topic selection, as we saw, Neurath’s charge is fully justified.

120

The History of Understanding in Analytic Philosophy

Importantly, Neurath’s criticism of ideal type theory was not merely of the abstract, global verificationist sort; it also had a concrete, local dimension concerned with the application of ideal types in sociological analysis. According to Weber, the ideal type “has the status of a purely ideal limiting concept against which reality is measured—with which it is compared—in order to bring out certain significant component parts of the empirical substance of [that reality]” ([1904] 2014: 127). Against this, Neurath inveighed in later years as part of his campaign against the use of what he called “absolute terms,” terms without any indefiniteness in their application. This conception of language he deemed to be, except in cases pertaining to islands of full formal regimentation, about as metaphysical or pseudo-rational as the idea of a pure protocol language composed of atomic sentences (see Neurath [1932b] 1983 and [1936b] 2004). The “absolutism” of terms appears particularly in terms which speak of certain “standards”—Max Weber, e.g., wants to estimate the “irrationality” of a certain behaviour by comparing it with the behaviour which would appear, if a man would know completely all elements of a situation and all possible ways and therefore the most effective way. This tendency to create an “ideal type” is connected with the old strain of rationalism. The “homo economicus” and similar beings have been created in this way. Laplace imagined a superhuman being which should be able to predict all future dislocations of all bodies by knowing their present positions and the formulae which describe their actions. (Neurath [1941] 1983: 225) Now precisely what is so objectionable to Neurath here is not fully spelled out, but his stock reference to Laplace makes clear that he rejected the underlying assumption of perfect insight, not only as an operational possibility but even as a limit concept. Just such a type of limit concept was invoked by Weber. Objectively correct rationality can be used by [interpretive sociology] as an ideal type in relation to empirical action; purposive rationality [can be used] as an ideal type in relation to what is psychologically understandable in terms of its meaning; and what is understandable in terms of its meaning [can be used] as an ideal type in relation to action whose motivation is unintelligible. The comparison with the ideal type makes it possible to determine the causally relevant elements of irrationality (this term being defined differently at each level) for the purposes of causal imputation. (Weber [1913] 2014: 279) What Weber said here seems harmless enough, but it wasn’t for Neurath.

MORE ON NEURATH ON VERSTEHEN

121

Let’s reconstruct what moved him. For Neurath, Weber gave the game away when he continued: It is true that we in our work also deal with important problems where the relationship between empirical behaviour and the correctness type actually also becomes a causal factor [determining the] development of empirical phenomena. But to demonstrate this state of affairs is not to divest the object [of investigation] of its empirical character: it is [to realize] the goal orientation of the investigation, determined by value relations and determining what kinds of ideal type are to be employed and what their function is to be. (Ibid.: 279–80) While defending his criterion of objectivity in so many words, Weber also conceded that the employment of the “correctness type” was of central importance to the attainment of empirical results; namely, the imputation of causal relations between individual events, or singular causal connections.33 What precisely then, Neurath will have asked, is it that this correctness type comprises: perhaps not just a correct assessment of the circumstantial conditions that make some courses of action more likely to succeed than others, but also an assessment of whether the ultimate goal was a reasonable one? Surely, at the most general level of inquiry— “objectively correct rationality”—the relevance of such considerations of the goal directness of action cannot be denied (it may be so perhaps at lower levels). And this means, of course, that the very question of the rationality of goals, bracketed on the grand scale by Weber as pertaining to philosophical orientations beyond empirical tests, haunts the modest practitioner in her everyday work. In other words, the very function of ideal types of serving as limiting concepts for understanding rational action is bound to involve ideal type theorists in substantive value considerations. They have to ask themselves whether an agent’s chosen goal is indeed a proper goal of action. And with this, they move beyond science into ethics or politics. Neurath’s was not a challenge from outside, as it were, but was founded on immanent criticism.34 It shows the perspectival objectivity of Weber’s interpretive social science to be compromised by the very feature which was claimed to account for its separatist identity: its value relation. As long as the assessment of the propriety of substantial non-epistemic values is part of social science, its empirical standing cannot but be compromised. This challenge is deep but not overbearing: replace the metaphysical value relation with that of interest relativity and much of Weber’s edifice remains in rude health—albeit in a non-separatist vein.

122

The History of Understanding in Analytic Philosophy

It might be argued that all I have shown above is that Weber’s assertions about gaining “knowledge of the cultural significance and importance of concrete historical relationships” and “ordering” them accordingly, about the “correctness type” and “objectively correct rationality,” allow for an interpretation that renders his conception problematic from an empirical standpoint, but not that these are essential features of interpretive social science as such. This can be conceded, but it does nothing to lessen the justification and force of Neurath’s criticism. Certainly from what evidence was available to Neurath—he was unable to access the criticisms Weber made of Rickert in private—the conclusion that Weber was committed to substantial assessments of the value ideas does follow. Moreover, he rightly reasoned that if Weber were not so committed, then he had little reason to propagate social scientific separatism. De-metaphysicalized or re-empiricized, the value relation becomes interest relativity and is shared as such even with the natural sciences. Then ideal types continue to be employed—albeit as regular theoretical terms or concepts bearing the imprint of their distinctive discipline.

5.6.  OTHER EMPIRICIST CRITICISMS OF WEBER It is instructive to compare Felix Kaufmann’s and C. G. Hempel’s stances on ideal types here. While far more sympathetic to the project of interpretive sociology than Neurath, Kaufmann nevertheless considered mistaken the claim that the employment of ideal types distinguished the natural from the social sciences. To start with, he rejected the idea of a separate realm of validity with independent laws (1936: 219), but this may well be more of a criticism of Rickert than of Weber himself. Still, Kaufmann held Weber himself to be “not entirely innocent” of “overdrawing” the difference between natural and social science (ibid.: 227). Kaufmann preferred to think of ideal types, rather, as representing “interpretive schemata,” which as such constituted the complement of the laws of natural science for the science of the social world (ibid.). Moreover, Kaufmann stressed that ideal types, far from representing Platonic universals, partake of “historicity”: certain “assumptions appropriate to certain historical epochs” enter into their formation (ibid.: 235). Since Weber agreed with this point as well, the basis of Kaufmann’s objection to his separatism is not easily determined. Direct criticism of ideal types was issued by Hempel, who claimed outright that “the alleged differences between the explanatory use of ideal types and the method of explanation in natural science are spurious” ([1952] 1965: 162).35 Hempel argued that since ideal types are “meant to serve as an interpretive or explanatory schema[ta] embodying a set of ‘general empirical principles’

MORE ON NEURATH ON VERSTEHEN

123

which establish ‘subjectively meaningful’ connections between different aspects of some kind of phenomenon” (ibid.), they “must be construed as theoretical systems embodying testable general hypotheses” (ibid.: 166). The subject matter of these schemata are, of course, “certain aspects of human behaviour” (ibid.: 169); their introduction thus requires “the specification of a set of characteristics … and of a set of general hypotheses connecting those characteristics” (ibid.: 168). In the absence of such specifications, ideal type explanations presumably would count, at the very best, only as explanation sketches that remain to be filled in. We can conclude therefore that as far as ideal types are concerned, Kaufmann and Hempel in effect joined Neurath’s protests against their supposedly distinctive and distinguishing role in social science (whether they were aware of its grounds or not). In addition, Hempel gave explicit arguments as to why ideal types—suitably reconceived, of course—could be integrated into unified science. Neurath remained alone in criticizing the overgeneralizations and absolutizations that ideal types tended to encourage, but also, more importantly, in criticizing the seemingly unempirical commitment to substantive values that their use in Weber’s hands seemed to demand. Characteristic for Neurath was that he saw in ideal type methodology a standing invitation to metaphysics. This point is cast into relief by Neurath’s resistance to Weber’s thesis about the rise of capitalism from the mindset of Protestantism. Needless to say, Weber’s thesis was controversial from the start in very many quarters, and Neurath was fond of citing the objections by J. B. Kraus, S. J., in particular.36 Neurath himself objected to Weber’s supposed attempt to invert the relations postulated by historical materialism as much as to the perceived lack of evidence.37 This criticism was partly misplaced, since arguably Weber objected to the idea of one-way determinism as such, yet Neurath also had straightforward empirical objections: “Theological doctrines are so far from having unambiguous consequences, that one cannot control whether a life still corresponds to a given doctrine or not” ([1931a] 1973: 358). Too many ways of life could be held in agreement with certain religious doctrines for us to invest such doctrines with causal power of changes in the ways of life of a population. However, Neurath’s main objection reverts to his criticism of Weber’s ideal type theory. Max Weber more than anyone else has endeavoured to treat sociology empirically but nevertheless he also has always striven for an antibehaviorist attitude in principle, he who time and again has ascribed a moving power to the spirit of the age. […] [Weber’s] turn away from strictly scientific formulations is not just an external matter. It is not as

124

The History of Understanding in Analytic Philosophy

though one could simply transform his statements into scientific ones. This is impossible when he speaks of the “rational economic ethos” as a force alongside behavior […] and when in any case he treats Protestantism like a reality which acts on men. What can this mean in a physicalist sense? ([1931a] 1973: 356–8) Neurath’s criticism here went beyond Weber’s use of correctness types in assessing the rationality of historical actors and focused on Weber turning his own idealizing abstractions—recall that ideal types are not realistic depictions of dominant ideas—into causal agents. Thus he wrote: “There exist Protestants, but there is no Protestantism” ([1931a] 1973: 358). Even if the particularly malevolent influence of Rickert’s metaphysics were discounted, there remained other grounds for Neurath’s worries about Weber’s (here, historical) sociology; namely, what he regarded as undue reification of abstractions. But it was not only Weber’s philosophical associations or the prominence he gave to ideal types that prevented Neurath’s acceptance of his interpretive sociology. There was also Weber’s monopoly interpretivism, which found expression in his famous definition of sociology (given in Section 5.4) in which two things are notable: first, sociology is limited to the investigation of meaningful action; second, at least in the eyes of many readers at the time, causal efficacy is accorded to what were widely understood to be nonphysical mental entities (or even supra-mental ones). Neurath objected to both.38 In postwar publications, Hempel joined Neurath’s reasoning on the first point, stating: “Weber’s limitation of the explanatory principles of sociology to ‘meaningful’ rules of intelligible behavior […] is untenable: many, if not all, occurrences of interest to the social scientist require for their explanation reference to factors which are ‘devoid of subjective meaning’ and thus ‘non-understandable uniformities’ ” (Hempel [1952] 1965: 163). Hempel rejected Weber’s delimitation of sociology as “interpretation of action in terms of its subjective meaning,” for it is “either an arbitrary restriction of the concept of sociology … or else it amounts to an a priori judgement as to the character of the system of concepts that can possibly yield an explanatory sociological theory” (ibid.: 164). At that point, Hempel still placed his faith in “explanatory principles for purposive action in purely behavioristic terms” (ibid.) and so appears to have been more restrictive than Neurath. It is clear, however, that Hempel too was worried by the dualist undertone of Weber’s definition. Such anti-metaphysical qualms certainly also motivated Neurath when he invoked against loose invocation of “motives” as springs of action. More generally, as we saw, he was worried that since a theorist might designate just about anything as an ideal type, interpretive sociology opened

MORE ON NEURATH ON VERSTEHEN

125

the door to metaphysical abuse. In a related vein, Ernest Nagel reminded proponents of interpretive sociology in the 1960s that imputations of subjective states required evidence “obtained by observation of ‘objective’ occurrences” (1961: 485). To object to the limitation of sociology to the study of meaningful social action—as did Neurath, Hempel, and Nagel39—may seem unduly picky given that Weber ([1921] 1978: 4) himself conceded that the dividing line between intentional action and merely conditioned behavior is not sharp. However, for Neurath, sociology also encompassed the study of social institutions and their consequences for social life, which demanded a broader vision. It is the task of sociology to focus on groupings with certain customs which show a definite structure in spite of changing individuals, to characterise typical groups of customs and to establish their relations. If the sociologist wants to master the multitudes of changes he must try to formulate laws. At the same time he has to keep close to actual events and their peculiarities. He can speak of “social laws” which are valid for distinct social formations. […] But of general sociological laws there are certainly only a few. In general one has to make use of laws of individual behaviour and other laws. (Neurath [1931a] 1973: 370–1)40 To study the regularities and interconnections of the behaviors of groups that remain stable despite changing memberships means to attend to structural relations and their consequences (and formulate non-strict laws of limited ranges). The question of the “coherence” of customs plays a decisive role in development of sociology. Though many sociologists pay attention to the coherence of stimuli, they neglect the socio-technical coherence, i.e. investigations of how certain institutions produce crises, mass misery, and the ways of behaviour which follow. The analysis of the relations by which these processes can be explained has less to do with stimuli than with the consequences of behaviour. Here schematic economic analysis will lead the way for sociological investigation. Just as the theory of extrapolation owes much to the behaviourists […] so the theory of coherence owes much to the economists. (Ibid.: 380) Clearly, Neurath’s sociology found lots of room for theoretical schema—ideal types, if you like, that are radically de-metaphysicalized. There was no need, however, for all of them to recur to meanings that agents attached to their actions. It was this broad conception of sociological inquiry that Neurath

126

The History of Understanding in Analytic Philosophy

juxtaposed to Weber’s in his 1931 monograph Empirical Sociology: The Scientific Content of History and Political Economy.

5.7. CONCLUSION Did Neurath wholly misunderstand Weber? Given the reading outlined here, by no means. While he overlooked Weber’s attempts to provide empirical validation for his ideal-typical theories—and cannot be absolved of this— Neurath rightly detected the danger inherent in Weber’s adoption of Rickert’s concept of the essential value-relatedness of the cultural sciences in transcending the bounds of empirical science. He also contrasted the limitation of interpretive sociology with his own more broadly conceived conception of sociology. Neurath need not be read as rejecting the inquiries pursued by interpretive sociology as wholly inappropriate, however: suitably recast, they found a place in his unified science as well. Neurath’s claim to argue for an empirical sociology that contrasts with Weber’s was well grounded and legitimate.

NOTES 1 Sociology treats men in the same way as other empirical sciences treat animals, plants, and stones; it is “social behaviorism.” Without having to adopt the particular results of behaviourism in Watson’s sense, it shares in principle the standpoint of the consistent behaviourists which so far could take root only in the USA and the USSR. In Germany the aversion to behaviourism is especially strong; here any ‘naturalistic,’ i.e. scientific sociology is rejected. … Scientific sociology is in opposition to the usual psychology which operates with “lived experience” and empathetic activities; it relates everything back to physicalistic data; it must also oppose the spread of those biological views which are the fashion today, so-called “neovitalism.” Sociology on a materialistic basis knows no effective structures which are not spatial and/or temporal. (Neurath [1931a] 1973: 361, trans. altered) 2 Physicalism does not hold the thesis that ‘mind’ is a product of ‘matter’ but that everything we can sensibly talk about is spatially and temporally ordered. What appears in statements as ‘mental’, ‘personality’, ‘soul’, must be expressible as something spatio-temporal or else vanish from science” (Neurath [1931a] 1973: 325). “Spatio-temporal processes, including spoken and written words—even words spoken to oneself (in the sense of behaviourism), at least if they can be incorporated as hypotheses—are objects for scientific sociology. (Neurath [1931a] 1973: 362–3)

MORE ON NEURATH ON VERSTEHEN

127

3 Max Weber more than anyone else has endeavoured to treat sociology empirically but nevertheless he has also striven for an anti-behaviorist attitude in principle. … He is unwilling to admit that “other minds” enter into science only insofar as they reveal themselves in spatio-temporal structures. … We see how one of the most important non-Marxist sociologists of Germany struggles for an “interpretive sociology,” from which, nevertheless, ever again he goes over to a purely scientific sociology. He tries, like Rickert, to carry out a division in the pursuit of science, which makes impossible any universal connection of all lawlike features that are required by the empirical sciences in their forecasts. With Weber, again and again, the empathetic immersion appears in place of science. Over this poetic activity there is no test, no control, nor does it belong in a scientific account (Neurath [1931a] 1973: 356–7) 4 “Understanding—in the sense of evident ‘interpretation’—and ‘empirical experiencing’ are not opposites, since all ‘understanding’, psychologically speaking, presupposes ‘empirical experience’ and can only, logically speaking, be demonstrated as valid by reference to ‘empirical experience’ ” (Weber [1903– 6] 2014: 74, cf. [1913] 2014: 273–4). The view is most clearly expressed here: We apply the term “adequacy on the level of meaning” to the subjective interpretation of a coherent course of conduct when and in so far as, according to our habitual modes of thought and feeling, its component parts taken in their mutual relation are recognized to constitute a “typical” complex of meaning. … The interpretation of a sequence of events will on the other hand be called causally adequate insofar as, according to established generalisations from experience, there is a probability that it will always actually occur in the same way. […] A correct causal interpretation of a concrete course of action is arrived at when the overt action and the motives both have been correctly apprehended and … their relation has become meaningfully comprehensible. A correct causal interpretation of typical action means that the process which is claimed typical is shown to be both adequately grasped on the level of meaning and at the same time is to some degree causally adequate. (Weber [1921] 1978: 11–12) 5 Unfortunately (see Uebel 2019a) many of these can be found in Neurath’s monograph Empirical Sociology ([1931a] 1973), but his intended meaning can still be ferreted out of passages like those quoted here in Notes 1–3. 6 See the quotations in Note 2 above. Still clearer affirmations of Neurath’s intertheoretical anti-reductionism are cited in Uebel (2007). For the record, Neurath, like Carnap, also endorsed intra-theoretical anti-reductionism. His frequent use of “zurückführen” with regard to the relation of the terms and claims of the theoretical to the observation language insists on testability qua mere confirmability, not verifiability. It is misleadingly translated as “reduce.” 7 The relation of reality to value ideas which lend it significance, and the selection and ordering, according to their cultural significance, of the parts of reality coloured by this [value relation] form an approach that is completely heterogeneous and disparate from [that of] analysing reality with a view to

128

The History of Understanding in Analytic Philosophy

finding laws, and ordering it in general concepts. No necessary logical relations exist between those two kinds of intellectual ordering of reality. (Weber [1904] 2014: 117, and see further, 124–7)

For discussion, see below.

8 In later years, Neurath noted that “Max Weber more than anyone else has endeavored to treat sociology empirically” and that even after remarkable failings on this count “ever again he goes over to a purely scientific sociology” ([1931a] 1973: 356–7). 9 The division taken over by Wundt, namely that between the natural sciences and the human sciences [Geisteswissenschaften,] has been attacked repeatedly and, indeed, does not seem expedient. It would be well worth to consider disregarding these large divisions entirely for the time being, for instance, the group ‘social sciences’, and to content oneself with a more precise separation among the individual areas. It seems very doubtful whether more appropriate principles of division are already available today. (Neurath [1910] 2004: 267) 10 The vast literature on Weber’s ideal types that has accumulated mostly in the last seventy years will not be discussed here for the simple reason that Neurath could not have taken account of it. Having debated with Weber at the Verein für Sozialpolitik in 1909 and 1913, Neurath regarded himself as his junior colleague and did not comment on his reception by contemporaries except in the case of his Protestantism thesis (see Section 5.5). Instead a close reading of the texts by Weber with which Neurath was familiar will be offered and related to relevant remarks by Neurath. For an overview of the interactions between Neurath and Weber and his circle, see Hübinger (forthcoming). 11 This paper marked Weber’s assumption of the co-editorship, with Werner Sombart and Alfred Jaffe, of the journal Archiv für Sozialwissenschaft und Sozialpolitik. The translation of Weber’s works used here is the one by Hans Henrik Bruun in Weber (2014), which has much to recommend it (as does the collection from which it stems). All emphases in quotations from Weber are as in the translated original unless otherwise noted; square brackets in quotations indicate explications by the translator or the present author. 12 Weber partly exempted Carl Menger himself from the criticism ([1904] 2014: 123). 13 That it is theoretical laws, not phenomenal-observational laws, that are at issue is indicated by the preceding phrase: “the only exact form of knowledge, i.e. the formulation of laws that are immediately and intuitively evident, is also the only one that makes it possible to draw conclusions concerning the occurrences which have not been directly observed” (Weber [1904] 2004: 123). Bruun translated “anschaulich” as “clearly” instead, presumably correctly to avoid the association with sensory intuition, but loses the nod towards intellectual intuition. 14 See Weber’s critical remark about Rickert’s concept of “systematic cultural sciences” in his letter to Rickert of November 3, 1907, in Weber (2014: 393). 15 Weber grounded his anti-realist instrumentalism in Kantianism:

MORE ON NEURATH ON VERSTEHEN

129

The basic idea of modern epistemology, which goes back to Kant, is that concepts are, and can only be, theoretical means for the purpose of intellectual mastery of the empirically given; and for anyone who carries this idea to its logical conclusion, that fact that precisely defined genetic concepts are by necessity ideal types will not constitute an argument against the formation of such concepts. ([1904] 2014: 134–5) 16 Weber expressly stated that “the idea of what ought to be, of an ‘ideal’, must be carefully distinguished from the theoretical constructs that we are discussing and that are ‘ideal’ in the strictly logical sense of the term” ([1904] 2014: 126). 17 Whenever the “ideas” that govern (i.e. are diffusely active in) the human beings of a certain epoch are mental constructs of a somewhat more complicated nature, we can only group those ideas themselves with conceptual precision in the form of an ideal type, as they are of course empirically present in the minds of a large, indeterminate and varying number of individuals, and can be found there in a multitude of variant with regard to form and content, clarity and meaning. ([1904] 2014: 128) 18 Thus for Weber: Obviously the difference between ordinary generic concepts, which simply comprise what is common empirical phenomena, and genetic ideal types—as for instance an ideal-typical concept of the “essence” of craft—may vary from case to case. But no generic concept as such has a “typical” character, and there are no purely generic ‘average’ types. Whenever we speak of “typical” magnitudes—for instance in statistics—this involves more than a simple average. ([1904] 2014: 131) 19 For the significance and influence of Johannes von Kries’ theory of probability from which Weber took over the idea of objective possibility, see Heidelberger (2010). 20 See Williams (1978), Nagel (1989), and Fine (1998), respectively. 21 We know from his so-called Nervi fragment (2014: 413–14) that Weber had severe reservations about Rickert’s claim of universal validity for the values he deemed basic to the cultural sciences, but his own resolution of this challenge to this neo-Kantian conception remained utterly unclear. 22 “ ‘Culture’ is a finite section of the meaningless infinity of events in the world, endowed with meaning and significance from a human perspective. This is true even when a human being is opposed to a particular culture as its mortal enemy and demands a ‘return to nature’. For he can only arrive at this position by relating that particular culture to his value ideas and finding it ‘wanting’ ” (Weber [1904] 2014: 119). 23 Religion, Church, law, the state, customs and morals, science, language, literature, the arts, the economy and the technical means necessary for it

130

The History of Understanding in Analytic Philosophy

are, at least from a certain level of development onwards, cultural objects or commodities in just the sense that the value which attaches to them is recognised as valid by all members of the community or that its recognition is forced on them. (Rickert [1899] 1986: 40) 24 Certainly, without the investigator’s value ideas, there would be no principle according to which the subject matter could be selected, and no meaningful knowledge of individual reality. … And the values to which the scientific genius relates the objects of his inquiry may be able to determine—that is: be decisive for—the ‘conception’ of a whole epoch, concerning not only what is considered “valuable,” but also what is considered significant or insignificant, “important” or “unimportant” about the phenomena. (Weber [1904] 2014: 120) 25 See the first outline of Weber’s doctrine in Part I of the paper at issue here ([1903–6] 2014: 101–8) as well as the later discussion in his ([1917] 2014). The original version of the latter appeared in the same internal discussion document of the Verein für Sozialpolitik as Neurath ([1913] 2004). 26 In a letter to Rickert of June 14, 1904, Weber declared: “I feel that a category of this kind is indeed necessary to distinguish between ‘valuation’ and ‘value relation’ ” (2014: 375). For Weber, the function of “the idea of the ‘ideal type’ ” was to allow value ideas—uniquely characteristic of human affairs—to become objects of scientific investigation without these investigations loosing their empirical character and becoming normative. 27 Weber’s forward reference (“as will become apparent”) is to his discussion of ideal types (reviewed in Section 5.2): their accentuation of certain features of a phenomenon by their ideal type conceptualization. 28 Since Neurath did not reject neoclassical or Austrian economics as altogether unscientific, this rejection on its own only amounted to a plea for methodological pluralism in economics; see Neurath ([1917] 2004). 29 Many years later ([1939] 1983: 212, n. 4) Neurath criticized a very similar claim by Frank Knight (1931: 67) that likewise falls under Weber’s strictures: “In human behavior we have a kind of direct knowledge of motives, whereas we only infer the existence of physical forces from observations of changes specific to each.” 30 For an overview of Neurath’s socialization proposals, see Cartwright, Cat, Fleck, and Uebel (1996: 22–43), and for Neurath’s reply to Weber’s criticsms of them, see Uebel (2019b). 31 In the field of the empirical social sciences of culture, the possibility of gaining meaningful knowledge of what is important to us in the infinite multitude of occurrences is tied to the unremitting application of viewpoints that have a specifically particular character and that are all in the last resort oriented towards value ideas. These value ideas may be established and experienced empirically as elements of all meaningful human action, but their validity cannot be proved on the basis of the empirical material. ([1904] 2014: 137)

MORE ON NEURATH ON VERSTEHEN

131

32 See Neurath ([1913] 2004). Incidentally, this was Neurath’s contribution to the privately published discussion document (for members only) of the Verein für Sozialpolitik on the place of value statements in science; Weber ([1917] 2014) represented an edited version of his submission to the privately printed discussion document which elaborated the position he had broadly outlined already in ([1904] 2014). Neurath was fully cognizant of Weber’s recognition—alongside his banishment of unconditional value judgments from all science—of the role values play, as cognitive interests, in the selection of social scientific data and of the problems to work on, and he put this to use for his own purposes. Thus, for Neurath, this value relevance provided “no occasion to diagnose a special ‘value-relatedness’ of the domain of history, which would lift it above the profane spheres of the physical” ([1936a] 1983: 773). Instead, he held that “the data which are used in sociology to construct the theoretical schema to make concrete predictions depend, on the one hand, on the interest one has in certain aspects, and on the other, on the chance to detect lawful connections” ([1931a] 1973: 388). 33 For the purposes of causal imputation of empirical events, we do in fact need rational constructions—empirical-technical or logical, as the case may be—that furnish an answer to the following question: what would a particular pattern of external action or, say, a theoretical construct (for instance a philosophical system) look like (or would have looked like) under the assumption of total rational, empirical and logical “correctness” and “consistence”? (Weber [1917] 2014: 330). 34 In later years, yet another consideration came into play in Neurath’s opposition. Ideal types are dangerous insofar as they can lead to hasty absolutizations that neglect the limits of our inquiries into what are the “aggregational complexes” of social life ([1941] 1983: 225). For Neurath, the phenomena we encounter are so complex and dependent on the satisfaction of so many conditions that the basis for generalization is much smaller than is typically realized. This point applies insofar as ideal types claim to be the result of a so-called isolating abstraction, where certain other variables often associated with them are simply neglected. This, however, is not a criticism of ideal type methodology as such but only of a use of ideal types that uncritically overlooks the limitations of what Weber called their “one-sided accentuations.” 35 Hempel ([1952] 1965: n. 7) referred to Kaufmann’s “stimulating critical discussion.” 36 See Neurath’s ([1931b] 1981) review of Kraus (1930). 37 See also Neurath ([1931b] 1981: 401–3 and [1932a] 1983: 83). This is not to say that Neurath thought the doctrine of historical materialism was unproblematical. Having entertained an, in all appearances, “functional” interpretation in Viennese discussions with Carnap in 1930, Neurath later on explored versions of a predictive asymmetry model—in which knowledge of the substructure alone allows for deductions and predictions better than knowledge only of the superstructure—to account for the cognitive claims made by historical materialism, but he never solved the issue to his satisfaction. For the functionalist interpretation, see “Diskussion, Boltzmanngasse, 5. V. 1930” (RC 029-18-01 ASP) and compare with Cohen

132

The History of Understanding in Analytic Philosophy

(1978); for the predictive asymmetry model, see Neurath ([1936c] 1981: 714–15, [1941] 1983: 225, and 1944: 22); in between, Neurath toyed with interpreting historical materialism as specifying dependencies of types of social phenomena on others: see Neurath ([1931a] 1973: 431, and [1932a] 1983: 82–3). 38 The second objection was elaborated by way of reconstruction of Neurath’s own acceptance of everyday hermeneutics in Uebel (2019). 39 Nagel argued in a Neurathian vein: It is beyond dispute that human behavior is frequently purposive; and it is likewise beyond question that when such behaviour is described or explained, whether by social scientists or by laymen, various kinds of “subjective” (or psychological) states are commonly assumed to underlie its manifestations. Nevertheless, … even when the behaviors studied by the social scientist are indisputably directed toward some consciously entertained ends, the social sciences do not confine themselves to using only distinctions that refer to psychological states exclusively; nor is it clear why these disciplines should place such restrictions upon themselves. For example, in order to account for the adoption of certain rules of conduct by a given community, it may be relevant to inquire into the ways in which members of the community cultivate the soil, construct shelters, or preserve food for future use, and the overt behaviors these individuals exhibit in pursing these tasks cannot be described in purely “subjective” means (1961: 475). 40 The first sentence in this quotation echoes the definition of sociology by Georg Simmel (see Simmel [1908] 1992), whose lectures Neurath attended as a student in Berlin.

REFERENCES Carnap, R. ([1934] 2013), “Theoretische Fragen und praktische Entscheidungen,” in C. Damböck (ed.), Der Wiener Kreis, 175–9, Stuttgart: Reclam. Cartwright, N., J. Cat, L. Fleck, and T. Uebel (1996), Otto Neurath: Philosophy between Science and Politics, Cambridge: Cambridge University Press. Cohen, G. A. (1978), Karl Marx’s Theory of History: A Defense, Oxford: Oxford University Press. Fine, A. (1998), “The Viewpoint of No-one in Particular,” Proceedings and Addresses of the American Philosophical Association, 72: 7–20. Heidelberger, M. (2010), “From Mill via von Kries to Max Weber: Causality, Explanation and Understanding,” in U. Feest (ed.), Historical Perspectives on Erklären und Verstehen, 241–65, Dordrecht: Springer. Hempel, C. G. ([1952] 1965), “Typological Methods and in the Natural and the Social Sciences,” in Aspects of Scientific Explanation and Other Essays, 155–71, New York: Free Press. Hübinger, G. (forthcoming), “Otto Neurath, Emil Lederer und der Max-WeberKreis,” in C. Damböck, G. Sandner, and M. G. Werner (eds.), Logischer Empirismus, Lebensreform un die deutsche Jugendbewegung/Logical Empiricism, Life Reform and the German Youth Movement, Cham: Springer.

MORE ON NEURATH ON VERSTEHEN

133

Kaufmann, F. (1936), Methodenlehre der Sozialwissenschaften, Vienna: Springer. Knight, F. (1931), “The Relation of Utility Theory to Economic Method in the Work of William Stanley Jevons and Others,” in S. Rice (ed.), Methods in Social Science, 59–69, Chicago: University of Chicago Press. Kraus, J. B. (1930), Scholastik, Puritanismus und Kapitalismus. Eine vergleichende dogmengeschichtliche Uebergangsstudie, Leipzig: Dunker & Humblot. Menger, C. ([1883] 1963), Problems of Economics and Sociology, Urbana: University of Illinois Press. Nagel, E. (1961), The Structure of Science: Problems in the Logic of Scientific Explanation, London: Routledge & Kegan Paul. Nagel, T. (1989), The View From Nowhere, Oxford: Oxford University Press. Neurath, O. ([1910] 2004), “Zur Theorie der Sozialwissenschaften,” in T. Uebel and R. S. Cohen (eds.), Economic Writings: Selections 1904–1945, 265–91, Dordrecht: Kluwer. Neurath, O. ([1911] 1998), “Nationalökonomie und Wertlehre, eine systematische Untersuchung,” in R. Haller and U. Höfer (eds.), Gesammelte ökonomische, soziologische und sozialpolitische Schriften Band 1, 470–519. Vienna: Hölder-Pichler-Tempsky. Neurath, O. ([1913] 2004), “On the Role of Moral Value Judgments in Economic Science,” in T. Uebel and R. S. Cohen (eds.), Economic Writings: Selections 1904– 1945, 297–8, Dordrecht: Kluwer. Neurath, O. ([1917] 2004), “The Conceptual Structure of Economic Theory and Its Foundation,” in T. Uebel and R. S. Cohen (eds.), Economic Writings: Selections 1904–1945, 312–41, Dordrecht: Kluwer. Neurath, O. ([1931a] 1973), “Empirical Sociology,” in M. Neurath and R. S. Cohen (eds.), Empiricsm and Sociology, 319–421, Dordrecht: Reidel. Neurath, O. ([1931b] 1981), “Marxismus eines Jesuiten,” in R. Haller and H. Rutte (eds.), Gesammelte philosophische und methodologische Schriften, 401–6, Vienna: Hölder-Pichler-Tempsky. Neurath, O. ([1932a] 1983), “Sociology in the Framework of Physicalism,” in R. S. Cohen and M. Neurath (eds.), Philosophical Papers 1913–1946, 58–90, Dordrecht: Reidel. Neurath, O. ([1932b] 1983), “Protocol Statements,” in R. S. Cohen and M. Neurath (eds.), Philosophical Papers 1913–1946, 91–8, Dordrecht: Reidel. Neurath, O. ([1936a] 1983), “Encyclopedia as ‘Model,’” in R. S. Cohen and M. Neurath (eds.), Philosophical Papers 1913–1946, 145–58, Dordrecht: Reidel. Neurath, O. ([1936b] 2004), “Sociological Predictions,” in T. Uebel and R. S. Cohen (eds.), Economic Writings. Selections 1904–1945, 506–12, Dordrecht: Kluwer. Neurath, O. ([1936c] 1981), “Mensch und Gesellschaft in der Wissenschaft,” in R. Haller and H. Rutte (eds.), Gesammelte philosophische und methodologische Schriften, 11–17, Vienna: Hölder-Pichler-Tempsky. Neurath, O. ([1939] 1983), “The Social Science and Unified Science,” in R. S. Cohen and M. Neurath (eds.), Philosophical Papers 1913–1946, 209–12, Dordrecht: Reidel. Neurath, O. ([1941] 1983), “Universal Jargon and Terminology,” in R. S. Cohen and M. Neurath (eds.), Philosophical Papers 1913–1946, 213–29, Dordrecht: Reidel. Neurath, O. (1944). Foundations of Social Science, Chicago: University of Chicago Press.

134

The History of Understanding in Analytic Philosophy

Rickert, H. ([1899] 1986), Kulturwissenschaft und Naturwissenschaft, Stuttgart: Reclam. Simmel, G. ([1908] 1992), Soziologie. Untersuchungen über die Formen der Vergesellschaftung, Frankfurt: Suhrkamp. Uebel, T. (2007), “Philosophy of Social Science in Early Logical Empiricism: The Case of Radical Physicalism,” in A. Richardson and T. Uebel (eds.), The Cambridge Companion to Logical Empiricism, 250–77, Cambridge: Cambridge University Press. Uebel, T. (2019a), “Neurath on Verstehen,” European Journal of Philosophy, 27: 912–38. Uebel, T. (2019b), “Rationality and Pseudo-rationality in Political Economy: Neurath, Mises, Weber,” in J. Cat and A. T. Tuboly (eds.), Neurath Reconsidered: New Sources and Perspectives, 197–216, Cham: Springer. Weber, M. ([1903–6] 2014), “Roscher und Knies and the Logical Problems of Historical Economics,” in H. H. Bruun and S. Whimster (eds.), Max Weber: Collected Methodological Writings, 3–99, London: Routledge. Weber, M. ([1904] 2014). “The ‘Objectivity’of Knowledge in Social Science and Social Policy,” in H. H. Bruun and S. Whimster (eds.), Max Weber: Collected Methodological Writings, 100–38, London: Routledge. Weber, M. ([1906] 2014), “Critical Studies in the Logic of the Cultural Sciences,” in H. H. Bruun and S. Whimster (eds.), Max Weber: Collected Methodological Writings, 139–84, London: Routledge. Weber, M. ([1913] 2014), “On Some Categories of Interpretive Sociology,” in H. H. Bruun and S. Whimster (eds.), Max Weber: Collected Methodological Writings, 273–301, London: Routledge. Weber, M. ([1917] 2014), “The Meaning of ‘Value Freedom’ in the Sociological and Economic Sciences,” in H. H. Bruun and S. Whimster (eds.), Max Weber: Collected Methodological Writings, 304–34, London: Routledge. Weber, M. ([1921] 1978), Economy and Society, Berkeley: University of California Press. Williams, B. (1978), Descartes: The Project of Pure Enquiry, London: Routledge.

CHAPTER SIX

Neurath’s Debate with Horkheimer and the Critique of Verstehen ANDREAS VRAHIMIS

6.1. INTRODUCTION Though the Nazis in Germany and the fascists in Austria had persecuted a number of individual philosophers, there were only two philosophical institutions that they targeted directly: the Frankfurt Institute for Social Research and the Ernst Mach Society. As a result, during the late 1930s, almost all members of both the Vienna Circle and the Frankfurt School became refugees fleeing into exile. In 1936–7, in large measure due to the political situation they found themselves in, both institutions would negotiate an attempt at collaboration.1 The attempt, led by Max Horkheimer and Otto Neurath, would fail, resulting in Horkheimer’s well-known critique of positivism in his 1937 article “The Latest Attack on Metaphysics.” At its outset, this chapter will look at the details of Horkheimer’s 1937 article, analyzing some of the arguments Horkheimer develops against what he presents as logical empiricism’s newest variety of “scientism” and “positivism.” Horkheimer claims that logical empiricism was committed to the view that all knowledge is derivable from value-free facts immediately given to the senses. Opposing this view, Horkheimer proposes a critical tribunal of reason that would further probe behind the givenness of facts and into the social activity that conditions them. Horkheimer later applies

136

The History of Understanding in Analytic Philosophy

some of his 1937 arguments against positivism and its myth of the given in developing, in 1940, a critique of Dilthey’s psychological conception of Verstehen. Despite accusing Dilthey of being a positivist, Horkheimer nonetheless misrepresents logical empiricism as providing the main argument against his conception of Verstehen. In fact, as this chapter will show, Horkheimer’s portrayal of logical empiricism, both in 1937 and in 1940, is in large measure a caricature.2 In the final parts of this chapter, I will question Horkheimer’s portrayal of the logical empiricist response to Verstehen in light of Uebel’s (2019) recent reconsideration of Neurath. I will examine Neurath’s responses to Dilthey and Sombart and compare them with Horkheimer’s critique of Verstehen. As Uebel (2019) shows, Neurath’s strategy for addressing claims about Verstehen is to rephrase them using non-metaphysical terminology. Neurath parallels this strategy towards Verstehen when he argues that once Horkheimer’s objections to logical empiricism are reformulated in empirically minded terms, they point towards a genuine problem concerning science’s relation to social praxis. As I will show, for Neurath, the problem can be addressed within empirically minded investigations in the history and sociology of science (rather than, as Horkheimer would have it, from a standpoint “above” science). I will conclude by showing how, in Neurath’s responses both to metaphysical conceptions of Verstehen and to Horkheimer’s conception of critical Vernunft, Neurath defends Unified Science against metaphysical “separatist” tendencies.

6.2.  HORKHEIMER AGAINST POSITIVISM Horkheimer’s 1937 article portrays his contemporary philosophy as divided between two poles: metaphysics and scientism.3 In attempting to define the dichotomy, Horkheimer draws up a genealogy tracing these two elements back to Descartes’ dualism. In Horkheimer’s view, Descartes attempted to disenchant scientific enquiry into the res extensa while at the same time opening a metaphysical path towards retaining a mental excess which cannot be disenchanted (Horkheimer [1937] 1972: 134–5). This Cartesian enterprise would give rise to the tensions that eventually evolved into the polarization Horkheimer diagnoses within his contemporary philosophy. Horkheimer describes this polarization as follows: Although the formulations of science now offer the most advanced knowledge of nature, men continue to use obsolete forms of thought long discarded by scientific theory. … It turns out, however, that many of these ideas are not merely superfluous, but also meaningless. Notions

Neurath’s Debate with Horkheimer

137

of absolute space, absolute time, and other metaphysical categories have been proved untenable. In addition, the doctrines of substance, causality, the soul, the mind-body relation, at least in their traditional form, have come into conflict with modern scientific methods. Yet, for all that, the pattern of ordinary thinking has not changed. This fact is really the projection of a contradiction that has persisted throughout the modern era. (Horkheimer [1937] 1972: 133–4) Horkheimer thus sets out by plainly stating that some of the basic areas within traditional metaphysics have been superseded by the advances of science. In this way, the lineage which derives from Descartes is seen by Horkheimer as reaching an impasse in his contemporary philosophical scene’s division between metaphysics and scientism. Horkheimer clarifies that what he takes to be regressive metaphysics is, since the First World War, represented by “romantic spiritualism, Lebensphilosophie, and material and existential phenomenology” ([1937] 1972: 136). What unifies all these strands of philosophy is, in Horkheimer’s view, their shared insistence on retaining the various categories which science has rendered redundant. Horkheimer sees his contemporary metaphysicians as radicalized by comparison to their earlier modern predecessors. In Horkheimer’s view, the “Cartesian solution” ([1937] 1972: 135) of substance dualism enabled earlier metaphysicians to draw a sharp distinction between their field and scientific endeavors. Nonetheless, as science rendered metaphysical categories more clearly obsolete, more and more strain would be put onto the conflict between science and metaphysics. According to Horkheimer, during and after the First World War this would lead to the following bifurcation in philosophy: The various attempts at harmonization fall into two extremes. One is the statement that science is the only possible form of knowledge and that the last traces of metaphysical thought must give way before it. The other is the deprecation of science as a mere intellectual technique answering to subordinate considerations of human existence. True knowledge, it is urged, must emancipate itself from science. (Horkheimer [1937] 1972: 136) In other words, Horkheimer sees the latest rendition of metaphysics as radicalized insofar as it has become overly skeptical about the value of science. Horkheimer connects this skepticism about science to a reactionary political attitude, contending that “postwar metaphysics paved the way intellectually for the authoritarian system of government in Germany”

138

The History of Understanding in Analytic Philosophy

([1937] 1972: 139). On the opposite pole, according to Horkheimer, stands a reaction to such skepticism which seeks to eliminate metaphysics. Horkheimer calls this reaction “scientism.” Scientism, which is the main target of Horkheimer’s critique in 1937, is exemplified by positivism. Its latest manifestation in the postwar philosophical scene to which Horkheimer is responding is logical empiricism, which Horkheimer refers to as a “neopositivist mode of thought” ([1937] 1972: 139).4 In brief, for Horkheimer, positivism is equivalent to the view that there is no possible object of knowledge outside empirically acquired knowledge. As Horkheimer sees it, due to its philosophical opposition to metaphysics, positivism thereby also happened to misguidedly attract “wide circles opposed to fascism” ([1937] 1972: 139). Horkheimer’s typology is an account of manifold divisions that fold into a dichotomy between the Frankfurt School and a number of other philosophical schools. But where is the Frankfurt School positioned in relation to its rivals? At first glance, Horkheimer seems to suggest that he is paving a path between two poles which the Frankfurt School is to navigate. Their path is one that will avoid falling back into traditional metaphysics, without thereby committing the scientistic error of limiting the knowable to supposedly value-free empirical knowledge of a given. In other words, Horkheimer’s presentation implies that the Frankfurt School was set to programmatically avoid the “parting of the ways” that Friedman (2000) sees as enacted between Heidegger and Carnap ([1931] 1959).5 Horkheimer’s ([1937] 1972) main philosophical objections to logical empiricism (from which a series of political criticisms follow) are addressed against what he construes as its positivistic reliance on a conception of givenness to the senses.6 In Horkheimer’s critique, positivism (as exemplified by logical empiricism) must subscribe to a myth of the given, according to which value-free facts are somehow immediately grasped by the senses. There are two aspects of his version of the myth of the given that Horkheimer focuses on in his objections. First of all, according to Horkheimer, an essential characteristic of positivism is that it prohibits any further probing concerning the givenness of facts. In Horkheimer’s view, positivism thus sets too strict a limit on critical inquiry. In his conception of positivism, any scientific attempt at studying what “lay before or behind individual facts and their interrelations” ([1937] 1972: 154) would go beyond the bounds of sense. Horkheimer’s strategy for resisting this limitation is, like various subsequent post-positivist critics of the Vienna Circle (see Dahms 1994: 58), to relativize givenness by pointing to the continuously changing historical and social forces which shape our senses, as well as our practices of “the selection, description, acceptance,

Neurath’s Debate with Horkheimer

139

and synthesis of facts” ([1937] 1972: 145). Thus, what the positivist takes to be observational facts, constitutive of scientific knowledge, are already conditioned by what Horkheimer calls “social praxis” ([1937] 1972: 154). Horkheimer insists that the positivists’ restrictions on meaningfulness necessitate this blind spot concerning the sociohistorical conditioning of facts. Horkheimer insists that in order to discover this, critical inquiry must necessarily investigate beyond mere facts as they are given to the senses. Of course, apart from facts given to the senses, there is another element that is crucial to logical empiricism. In Horkheimer’s account, logical empiricist epistemology allows for the systematic logical “arrangement and rearrangement” ([1937] 1972: 145) of given facts. Horkheimer even criticizes the positivist account of science for overestimating the significance of such rearrangements. He notes that for the positivist, it matters not what facts are selected from the infinite number that present themselves. He proceeds as if the selection, description, acceptance, and synthesis of facts in this society have neither emphasis nor direction. Science is thus treated like a set of containers which are continually filled higher and kept in good condition by constant repair. This process, which was previously identified with the activity of the understanding, is unconnected with any activity which could react on it and thereby invest it with direction and meaning. Everything designated … by materialism, as social practice and conscious historical activity, is related to science essentially as objects of observations and not as constitutive interests and directive forces, insofar as empiricism concedes them to be conditions of knowledge at all. (Horkheimer [1937] 1972: 145) Thus, in the Kantian terminology Horkheimer ([1937] 1972: 145, 156, 158) employs, the logical empiricists conceive of science as consisting of facts discovered by sensation, on the one hand, and their conceptual arrangement by the understanding, on the other. What is missing from logical empiricism, in Horkheimer’s Kantian picture, is the faculty of reason ([1937] 1972: 152, 156–7, 186). Horkheimer appeals to reason (Vernunft), in the Kantian sense, as that faculty which allows for a further probing into the social practices which, qua “constitutive interests and directive forces” ([1937] 1972: 145), condition sensation. Horkheimer’s proposal for such further probing involves taking an interdisciplinary critical standpoint. In Horkheimer’s view ([1937] 1972: 186), this standpoint is neither metaphysical, in the sense of being autonomous from scientific inquiry, nor scientistic, in the sense of being limited to given facts and their logical ordering. Critical reason, in Horkheimer’s view, must be allowed to rise above the individual sciences

140

The History of Understanding in Analytic Philosophy

in order to question the social praxis which underlies what has uncritically been taken as given to the senses. Second, according to Horkheimer, the positivist myth of the given is tied to a conception of facts as value-free, which he considers problematic.7 If facts are conditioned by social praxis, as Horkheimer thinks, this further implies that they are value-laden. In Horkheimer’s account, the very idea of value-free science is itself value-laden. The conception of science as valuefree unavoidably leads to a relativistic outlook, as exemplified in the work of Neurath (Horkheimer [1937] 1972: 165).8 According to Horkheimer ([1937] 1972: 165), relativism has various dangerous political implications. He argues that the common correlation between relativism and liberalism is misconceived. He illustrates this with cases in which relativism had also been instrumentalized by brands of authoritarianism. Horkheimer ([1937] 1972: 165) cites Mussolini’s championing of relativism as a case in point and even draws comparisons between Mussolini and Neurath in this regard. Horkheimer also draws various other far-reaching political conclusions from his portrayal of positivism as bound to the myth of the given. According to Horkheimer, positivism’s upholding of the myth of the given entails that it is bound to a type of conservatism prohibiting the radical questioning of appearances.9 From there, he leaps to an identification of positivism with the silencing of critical reason and even to the declaration of the compatibility of positivism with authoritarianism. Partly due to his stated avoidance of dwelling “on the shades of differences among its adherents” ([1937] 1972: 141), Horkheimer’s critique involves a number of stereotypes about logical empiricism that are now easily demonstrable as false (see also Vrahimis 2020a). For example, Horkheimer’s central argument concerning the commitment of positivism to the myth of the given seems to simply ignore the disagreements between the Vienna Circle’s members on precisely this topic in the context of the protocol-sentence debate (see Uebel 1992: 295–301, 2007). Perhaps more pertinent to our discussion in this chapter is Horkheimer’s misidentification of positivism with a kind of very strict “physicalism” ([1937] 1972: 146) and “behaviourism” ([1937] 1972: 149–50, 152–3), to which I shall return in what follows. Based on this straw man, Horkheimer proceeds to claim that despite being seen at the time as one of the main “intellectual weapons against the totalitarian frenzy” (Horkheimer [1937] 1972: 140; see also Jay 1996: 17), positivism is somehow “as securely bound as metaphysics to the established order” ([1937] 1972: 140).10 Having conceived of the positivist strictures on meaningfulness in the manner outlined above, Horkheimer can thus

Neurath’s Debate with Horkheimer

141

conclude that a description and critique of social praxis lies beyond such strictures. As Horkheimer sees it, the positivist is thus confined to the position of being unable to distinguish between the nonsensical metaphysics that support authoritarianism and its non-positivist critics: Harmony and significant existence, which metaphysics wrongly designates as true reality as against the contradictions of the phenomenal world, are not meaningless. Powerful economic forces welcome a philosophy that professes not to know what to make of these conceptions and for that reason prefers to stick to facts; a philosophy that resolves not to make any essential distinction between the conspiracy of brutal despots against all human aspiration to happiness and freedom, on the one hand, and the struggles to defeat these tyrants on the other; a philosophy that reduces the two to the abstract concept of the “given” and even glorifies such conduct as objectivity. (Horkheimer [1937] 1972: 178) It is in this way that Horkheimer is able to contrast his caricature of positivism with his own position, which becomes attentive to social praxis by breaking away from the positivist strictures while nonetheless avoiding falling back into the errors of traditional metaphysics. As already noted, the politically conservative nature of positivism is contrasted by Horkheimer with his own conception of critical reason, which was to guide the Frankfurt School’s endeavors under his leadership. Horkheimer calls for a kind of metascientific theoretical endeavor that is free from the limitation of being derivable from the given and, thus, is able to somehow approach the social praxis that conditions givenness. In “The Latest Attack on Metaphysics” (Horkheimer [1937] 1972), this is conceived as being able both to subvert authoritarian metaphysics and to critically react to empirical science. The crucial question at stake here for Horkheimer is whether it is possible to meaningfully develop such a critical metascientific standpoint without reverting to traditional metaphysics and its deprecation of science.11 Is it not the case that any account of social praxis will inevitably rely on some empirical observations? If this is admitted, then would not a less strict version of positivism (i.e., something other than the caricature painted by Horkheimer) be able to recognize such an account as meaningful? If not, then how can Horkheimer even know that his position differs from traditional metaphysics? It seems, however, as I have shown above, that Horkheimer’s conception of this metascientific enterprise is reliant on an argument against a positivist straw man, caricatured in a way which none of its adherents, and especially not his interlocutor Neurath, actually uphold.

142

The History of Understanding in Analytic Philosophy

Before returning to this issue, I will now examine how Horkheimer’s critique of positivism informs his later reception of Dilthey.

6.3.  HORKHEIMER ON DILTHEY, VERSTEHEN, AND LOGICAL EMPIRICISM In 1940, three years after the attempted collaboration between the Frankfurt School and the Vienna Circle had failed, Horkheimer published one of the clearest articulations of his views on Verstehen in “The Relation between Psychology and Sociology in the Work of Wilhelm Dilthey.” Dilthey is among the various philosophers of his time who had insisted on drawing a strict separation between the Naturwissenschaften and the Geisteswissenschaften. This was, as we shall see in more detail later, a thesis which would come under attack by various philosophers associated with the Vienna Circle, including perhaps most prominently Otto Neurath. Given this critical attitude by the Vienna Circle, it is interesting to see that Horkheimer’s 1940 paper presents Dilthey’s conception of Verstehen as part of an overall alignment with a positivistic conception of scientific methodology. Throughout his article, Horkheimer (1940: 430, 432–4) portrays Dilthey as committed not only to a positivist methodology for science, but also specifically to a “denunciation of metaphysics” (1940: 433). According to Horkheimer, Dilthey understands the cultural sciences (of which history was the most central for his goals) as reducible to the science of psychology.12 In Horkheimer’s account, Dilthey conceives of the split between the Naturwissenschaften and Geisteswissenschaften in psychological terms; that is, as pertaining to the difference between what Carnap would later call autopsychological and heteropsychological observations.13 The natural sciences are engaged in a systematic determination and classification of the facts given by sense perception in their space-time relationships. Cultural sciences have to deal with the same objects. Reality is not divided into nature on the one hand, and mind on the other. Cultural sciences have to deal with the same reality under another specific aspect. Certain objects in nature compel us to regard them as the expression of life past or present, and we are able to know their true character from an understanding of what we ourselves are. We must go back from them to our own life. (Horkheimer 1940: 433–4) Horkheimer thus sees Dilthey as arguing that both the Naturwissenschaften and the Geisteswissenschaften make claims about different aspects of the same reality. Horkheimer makes it clear that he sees the division Dilthey

Neurath’s Debate with Horkheimer

143

draws as purely methodological, and he rejects the thesis that it may involve a metaphysical distinction between mind and matter. The Diltheyan methodological division, in Horkheimer’s account, is not incompatible with the positivist strictures which he presents Dilthey as upholding; that is, the limitation of what is known to what is given to the senses (which, as we have seen, Horkheimer also attributes to the logical empiricists). Both the natural and the cultural sciences have empirical content, though the types of experiences involved in each are different. Only the latter involve the types of experiences that Dilthey categorizes under the heading of Verstehen. Horkheimer (1940: 434) explains that Diltheyan Verstehen is not to be thought of as some mystical irreducible intuition (of the type that we find, e.g., in Bergson). Rather, as Horkheimer (1940: 434) puts it, “Dilthey insisted that a basic scientific analysis of all individual data and relationships must precede understanding.” Horkheimer thus does not, at this point in his analysis, see a clash between Dilthey’s methodological account of the Geisteswissenschaften and his overall commitment to positivism. Horkheimer appears to accept that Dilthey’s account of Verstehen makes a valuable contribution to the methodology of the cultural sciences, but only insofar as it is acknowledged that it faces certain limitations. Thus Horkheimer’s criticism does not mainly focus on rejecting Dilthey’s psychological methodology for the cultural sciences, but on highlighting what is omitted by its application. Horkheimer even goes on to chastise Dilthey for precisely the same reasons he attacks the logical empiricists; namely, for subscribing to the myth of the given. Horkheimer accuses Dilthey of insisting in the “unfounded belief that valid insight must confine itself to the realm of the immediately given” (1940: 437) and, as a result, setting too strict a limit on what can be known through the science of psychology and by extension through the cultural sciences.14 Horkheimer directs against Dilthey one of the main criticisms that he had also directed against the Vienna Circle: by buying into the myth of the given, Dilthey is bound to a kind of individualism which inevitably fails to account for social phenomena and for the supra-individual “worldly character of the human spirit” (Horkheimer 1940: 437). Horkheimer thinks that Dilthey’s individualist psychologism thus fails to account for the sphere of the “objective spirit” (1940: 442) at the level of theory. Horkheimer argues, however, that Dilthey brings objective spirit into the picture in the guise of specific phenomena which he falsely presents as if they were purely subjective. Horkheimer sees this problem in Dilthey’s position as the result of a contradiction between Dilthey’s positivistic theoretical commitments and his concrete analyses of specific historical phenomena. Horkheimer sees the latter as bringing in social phenomena through the back door, so to speak. For example, Horkheimer points out that Dilthey’s psychological-historical

144

The History of Understanding in Analytic Philosophy

analyses of specific artists’ lives and work involve extensive discussions of tendencies in their contemporary art world, or that when Dilthey “speaks of a Roman he draws in Roman law and Roman imperialism as real forces in the soul of the Roman citizen” (Horkheimer 1940: 443). Thus his psychological analyses of purportedly “subjective” phenomena contain discussions of social phenomena disguised as lived experiences. According to Horkheimer, due to Dilthey’s unwarranted assumption of positivism (and thereby individualism), what is missing from his project is an explicit targeting of such social phenomena, which are indispensable from any analysis of history. Confusingly, though Horkheimer admits that Dilthey had a positivistic orientation in his conception of Verstehen, he also mentions logical empiricism as giving rise to the main form of objection to such a conception: If Dilthey is right, a large part of our knowledge depends upon the inner richness of the individual, who strives for such knowledge. The act of knowing, in Dilthey’s own words, comes close to the artistic process. In this case not a few modern methodologists would strike this whole section of knowledge from the realm of science and assign it to poetry, somewhat after the fashion of the logical empiricists. In economics they would like us to limit ourselves to mathematics, in human psychology to experiments with the tachystoscope or similar apparatus. But if these critics are right, a sphere of decisive experience would thus fall outside the range of scientific activity. (Horkheimer 1940: 434–5) In other words, Horkheimer thinks that Dilthey sets too strict a limit on what is knowable by espousing the positivistic myth of the given and also that his positivist critics set too strict a limit on what is knowable by espousing a positivistic myth of the given that undercuts even Dilthey’s notion of Verstehen. Horkheimer seems, at least indirectly, to further append to his earlier caricature of logical empiricism the above claims concerning the methodology of economics and psychology, for which he adduces no evidence.15 He is nonetheless perhaps more circumspect than in his earlier work insofar as he does not talk of the logical empiricists themselves but of “a few modern methodologists” who hold such beliefs “somewhat after the fashion of logical empiricists” (Horkheimer 1940: 434).

6.4.  AGAINST THE CARICATURE: UEBEL ON NEURATH’S RESPONSE TO VERSTEHEN For the purposes of this study, what is most interesting about Horkheimer’s (1940) article is its take on the logical empiricist critique of Verstehen.

Neurath’s Debate with Horkheimer

145

Such common misconceptions of the logical empiricists have often resulted from inadequate attention paid to the detail of their work. Similar views to those presented by Horkheimer (1940) have often been falsely supposed to follow from some of the other features of Horkheimer’s ([1937] 1972) earlier caricature concerning the Vienna Circle’s commitment to crudely reductionist forms of behaviorism and physicalism. I will not here attempt to address this topic in connection to logical empiricism in general, but will focus on Neurath, who had been in debate with Horkheimer. Uebel’s recent work (2010, 2019) reconsiders Neurath’s critique of Verstehen by questioning whether he upholds reductionist behaviorist and physicalist views.16 Uebel (2019) concedes that at first glance the writings not only of Neurath but also of Carnap and other members of the Vienna Circle may deceptively appear to justify Horkheimer’s view that they are committed to a reductive physicalism and behaviorism.17 This is insofar as “behaviorism” and “physicalism” make frequent appearances as emic terms to describe positions defended by, and even identified with, logical empiricism. Nonetheless, further scrutiny allows us to dispel the caricature. The logical empiricists’ apparent championing of behaviorism and physicalism often appeared in the context of their polemicizing against various rival philosophers who had championed a demarcation between Naturwissenschaften and Geisteswissenschaften. Dilthey is among the names Neurath discusses in criticizing this demarcation in his 1931 book Empirical Sociology. According to Neurath, although Dilthey himself wished to avoid metaphysics, he inspired it greatly. Even those people who tend to be empiricist, bow before the “humanistic sciences” and praise Dilthey’s significant achievement which, however, lies in his concrete analysis and not in his theoretical investigations. His inclination towards classification has given ideological support to those who would like to avoid causal investigations in social matters. German sociology above all is shot through with ideas of Dilthey’s, some grasped and some misunderstood. Just as in certain circles the scientific world-view is spreading ever more widely, so the leading academic sociologists in Germany are strongly metaphysical in outlook, or at any rate, they are far removed from any endeavor intended to further a unified science on a materialist basis. (Neurath [1931] 1973: 356) Neurath’s account of Dilthey here displays some interesting parallels with Horkheimer’s (1940) later analysis.18 Neurath values Dilthey’s “significant achievement” insofar as, in agreement with Horkheimer, he acknowledges Dilthey’s declared intention to avoid metaphysical tendencies in the cultural

146

The History of Understanding in Analytic Philosophy

sciences. As shown above, Horkheimer criticizes Dilthey’s theoretical framework for being unduly restrictive due to his espousal of positivism. By contrast, he thinks that Dilthey’s concrete analyses overcome the strictures of positivist individualism by bringing in elements of social praxis disguised as psychology. Neurath reverses Horkheimer’s critique of Dilthey when he favors the content of Dilthey’s concrete analyses over the theoretical framework in which they are undertaken.19 Neurath criticizes Dilthey’s theoretical framework for allowing philosophers in his wake to leap to a metaphysical segregation of the “humanistic” from the “natural” sciences. Neurath (e.g., [1930] 1983: 44–5, 2020a: 245) attacks the metaphysical division between Naturwissenschaften and Geisteswissenschaften as a remnant of the Christian religious division between the earthly and the heavenly realms. Neurath ([1930] 1983: 44–5) sees that in fields such as history and sociology, scholars have given refuge to religious and theological notions of the heavenly, attempting to rescue them from being brought back to earth by naturalistic explanations.20 The Weberian conception of Verstehen as a prerequisite for explanation even threatens to make such metaphysically phrased religious notions indispensable for science (see Uebel 2019: 6–8). Such misuses of metaphysical nonsense would rely on Dilthey’s notion of Verstehen, nonetheless divorcing it from the empirically minded manner in which Dilthey draws its connection to natural science (which both Horkheimer and Neurath emphasize). Uebel (2019) demonstrates that Neurath’s main criticisms of Verstehen are thus directed not against Dilthey himself, but against various other figures in Dilthey’s line of influence, such as Sombart or Spengler.21 Recent studies (e.g., Gabriel 2004; Damböck 2012, 2017; Dewulf 2017; Nelson 2018) highlight that Dilthey’s work was at least partly influential for various Vienna Circle figures, most prominently Carnap. This influence can best be understood in light of some fundamental differences between Dilthey and some of his followers. There are substantial philosophical differences between what Horkheimer sees as a positivistic bias in Dilthey and the metaphysical tendencies of some of the philosophers in his line of influence. There is also a notable political difference between the former and the latter. While Dilthey himself was committed to a brand of democratic Enlightenment liberalism,22 many of the figures influenced by his work were politically reactionary.23 This political difference at least partly informed the logical empiricist reception of Dilthey’s work (see Nelson 2018: 322). Indeed, in 1931, Neurath offers an analysis of how later metaphysical uses of Dilthey were connected to the state’s power over academic appointments:

Neurath’s Debate with Horkheimer

147

This advance of metaphysics becomes understandable if we take a look at those who hold such views. On the whole the representatives of metaphysically directed sociology are at the same time representatives of the ruling order. Most governments and other centres of power favor metaphysically inclined scholars, even theologizing ones, whereas they are mostly suspicious of anti-metaphysically inclined scholars, especially those who wish to establish this outlook in the field of sociology. Since the ruling bourgeois classes are more or less closely linked with clerical circles, and Christian priests generally defend the ruling order against revolution, a certain protection of the theological and metaphysical outlook seems probable from the outset. (Neurath [1931] 1973: 356) Despite their fundamental disagreements concerning the political dimensions of logical empiricism, Neurath seems here to be largely in agreement with the connections Horkheimer ([1937] 1972) draws between metaphysics and the authoritarian state. After all, by the 1930s the Vienna Circle had already fought various battles against academics (e.g., Othmar Spann) whose “metaphysically inclined” and “theologizing” views had sought to give respectability to various political positions (e.g., clericalism, Austrofascism, pan-Germanism, corporatism) that both the Frankfurt School and the Vienna Circle opposed (see, e.g., Stadler 2015: 254; Romizi 2012; Scott 2020: 230–1). Neurath’s political opposition to the metaphysical interpreters of Dilthey informs his critical responses to the debate over Verstehen. Neurath did in fact explicitly target the idea that there is such a thing as an irreducible type of intersubjective “empathetic understanding, Einfühlung or Verstehen” (Uebel 2019: 1) which could provide a methodological backdrop exclusive to the Geisteswissenschaften. Horkheimer’s diagnosis was thus at least partly correct: in response to such cases, Neurath polemicizes against metaphysical claims about Verstehen by characterizing them as being mere poetry (e.g., Neurath [1931] 1973: 357; see Uebel 2019: 7). Nevertheless, Horkheimer, like many others, is wrong in thinking that Neurath’s critique of metaphysical statements cover all and any claims about Verstehen. According to Uebel (2019: 5), Neurath does not think that the concept of Verstehen itself was altogether meaningless. Neurath’s objections are directed against a subset of claims employing this concept; namely, the specific range of views that hold Verstehen to be somehow unanalyzable and irreducible.24 It is precisely from such dubious premises that his contemporaries had leapt to the metaphysical separation between two types of science (“human” and “natural”) which Neurath targets. A consequence of such metaphysical views about Verstehen is that due primarily to the language in which they are phrased, they mistakenly purport

148

The History of Understanding in Analytic Philosophy

to preclude any type of empirical testing. By contrast, as Uebel (2019) shows, Neurath’s proposal is to probe deeper into such statements, clearing away their metaphysical components by showing how they can, upon further scrutiny, be rendered empirically testable. According to Neurath, “when the concept of Verstehen is analysed scientifically, then everything about it that can be communicated in words to somebody else turns out to be a statement about structure, just as in the natural sciences” (quoted in Uebel 2019: 6). In this way, Verstehen is reinstated in a non-metaphysical form. Neurath’s goal is not to eliminate the statements of the human sciences, but rather to correct their mode of expression. A certain range of statements about Verstehen tends to systematically commit the error of employing metaphysical language. Neurath’s aim is to show that the error can be avoided by carefully rephrasing statements in a way that renders them empirically testable (e.g., by rendering first-person introspective reports about conscious states into third-person protocols about such first-person reports; Uebel 2019: 922–3). According to Uebel (2019), Neurath’s position on Verstehen can be detected in the details of his various critical responses to a number of contemporary philosophers and social and cultural theorists. Neurath does often engage in excessive polemics, and he misunderstands some of his targets’ views.25 It is only once one looks past his polemical excesses and misunderstandings are cleared away that Neurath’s intention in these debates comes into view. Neurath shows, in each case, how the relevant metaphysical claims about the autonomy of sciences based on Verstehen from other sciences can be cleared away, rendering non-metaphysical, empirically testable claims about Verstehen part of his vision of a materialist Unified Science (see Uebel 2019: 7). In connection to the debate with Horkheimer, an interesting case among Neurath’s targets is that of Sombart. Neurath criticizes Sombart for moving away from the kind of individualistic psychological conception of Verstehen that we find in Dilthey towards what he calls a “noological,” nonpsychological conception. Sombart also provides us with a particularly good example of the aforementioned political dimension of the dispute, since he conceives of this distinction in nationalistic terms. He categorizes the psychological conception of Verstehen as part of “Western” sociology, which he contrasts with “German” sociology and its noological conception.26 The latter involves an appeal to an objective “spirit [which] cannot be reduced to mind and [to] human culture [which] cannot be derived from elementary psychological drives” (quoted in Uebel 2019: 6). Sombart’s is a clear case of how claims about Verstehen are transposed from empirically testable psychological claims (as we find, e.g., in Dilthey) into what the Vienna Circle would condemn as metaphysical nonsense. Furthermore, the nationalistic

Neurath’s Debate with Horkheimer

149

usage to which Sombart puts his metaphysical nonsense exemplifies precisely the kind of political attitude that Neurath wishes to oppose.27 Sombart’s criticism of the individualistic psychological conception of Verstehen is to some extent similar to Horkheimer’s critique of the limitations of Dilthey’s individualism. Sombart, like Dilthey himself makes use of the Hegelian terminology of “objective Spirit” and, as we have already seen, Horkheimer (1940: 442–3) makes explicit reference to this element in objecting to Dilthey. Horkheimer notes that the “postulate, that the subjective mind can be rightly understood only in connection with the objective mind (using the Hegelian terminology), has not been fully worked out in his [i.e., Dilthey’s] historical writings” (1940: 443). Though Sombart and Horkheimer were on very different sides of the political spectrum, their proposed criticisms of individualism and their connected appeals to social phenomena (via the Hegelian notion of objective spirit) may appear here to lie philosophically quite close. In Sombart’s case, it was easy for Neurath to detect his straightforwardly metaphysical attitude, denying that any type of empirical test would be possible for his proclamations about spirit. But what about Horkheimer’s claims about the possibility of grasping “social praxis” through the use of critical reason while at the same time avoiding the pitfalls of metaphysics?

6.5.  NEURATH’S REPLY TO HORKHEIMER Though at least obliquely relevant, the methods of Verstehen are not what is directly at stake in the case of Horkheimer’s proposal for a critical approach to social praxis. Instead, Horkheimer’s main disagreement with Neurath concerns the viability of pursuing some other type of nonempirical metascience. In his reply to Horkheimer’s 1937 article, Neurath summarizes Horkheimer’s claims as follows:28 there is an extra-scientific method, which is based on everything that is scientifically established, and which can criticise the sciences in particular by exhibiting their historical position in a way that is foreign to the sciences. This standpoint, named by Horkheimer “dialectical” or “critical”, goes beyond both the metaphysical, as well as the scientific; it was precisely the error of the “Scientivists” that they are “opposed to thought, whether it tend forward with reason, or backward with metaphysics” [Horkheimer [1937] 1972: 186]. He takes himself to be defending Reason [Vernunft] against Empiricism, whereby, in the sense of traditional German philosophy, he distinguishes between Understanding [Verstand]—which he will not deny to Empiricism—and Reason [Vernunft]. (Neurath 1937a: 4)29

150

The History of Understanding in Analytic Philosophy

Thus, it is reason (in the Kantian sense outlined above) that is, according to Horkheimer, the vehicle of that critical metascientific activity which may stand over and above empirical science. The main question which thereby arises for Neurath is whether Horkheimer’s proposal concerning a critical perspective on the relation between scientific and social activity requires taking some irreducibly metaphysical a priori standpoint above science. In Neurath’s view, the correct answer to this question is: no. Neurath concedes that Horkheimer raises a question that is crucial for an empiricist understanding of science’s social impact and social activity’s impact on science. Nonetheless, to reach the kind of valuable insight involved in Horkheimer’s criticisms, his objections need to be reformulated in empirically testable terms. Neurath opens his attempt to undertake such a reformulation by pointing out that scientific activity is unavoidably a kind of social activity, the consequences of which are often not determinable by its agents. In Neurath’s example (1937a: 3), a conservatively minded scientist may unwittingly contribute to tomorrow’s progressive social struggles. One possible nonempiricist response to this uncertainty is to attempt to fix the possible relations between scientific and social activity through some aprioristic form of metaphysics. Indeed, Neurath sees Horkheimer as being in the line of influence of such metaphysical approaches stemming from German idealism. For the empiricist, however, this is not a viable option. It prematurely flees from the reality of empirical uncertainty into some realm of false a priori certainty (what Neurath elsewhere [e.g., [1921] 1973] calls “pseudorationalism”). Once a priori metaphysical speculation has been ruled out as a possible solution to the problem, it becomes clear how difficult it is to predict the future social consequences of scientific activity and vice versa. This, for Neurath, is a genuine problem which arises in a purely empirical manner, and which is of the utmost concern to the empiricist who realizes that there is no possible appeal to metaphysics that could resolve it. In other words, the problem that Horkheimer cites turns out to be much more serious than he himself takes it to be! Horkheimer’s preference for metaphysically charged terminology seems largely to be the reason preventing him from giving an adequate formulation of the problem he raises. Neurath detects metaphysical elements both in the language which Horkheimer uses to pose his question and in the resulting attempt to answer it in a “supra-empiricist” (Neurath 1937a: 4) manner. While attempting to address a question which, Neurath argues, should be understood as arising within empirical science, Horkheimer’s proposed solution appeals to some supra-empirical critical tribunal by reason. Horkheimer’s appeal to reason, which is reached through the application of what Neurath sees as a poorly defined method of dialectics, promises to

Neurath’s Debate with Horkheimer

151

determine the manifold of relations that hold between scientific discovery and social praxis. Neurath thinks that Horkheimer is held back by traditional German philosophy’s ways of posing and answering such questions in a priori terms. Neurath briefly states that Horkheimer’s views are traceable back to metaphysical terminology employed by the German idealists (1937a: 4, 13) and points to Horkheimer’s reservedly sympathetic references to “Hegel, Kant, the Neo-Kantians and Husserl” (1937a: 13) as evidence that this is the case. Horkheimer retains such metaphysical terminology not only in his conception of “reason” but also, for example, when he appeals to the “essence” that is hidden beneath the mere “appearances” that are given to the senses (Neurath 1937a: 16; see also O’Neill and Uebel 2004: 80–3). In place of metaphysical appeals to an a priori metascience, Neurath proposes what he sees as a viable empirical method for investigating the relation between scientific discovery and social praxis. This empirical method may be, for Neurath, pursued in examining the history and sociology of science. Horkheimer makes a number of claims concerning the history of empiricism, which seem to be, as Neurath sees it, proclaimed ex cathedra without empirical evidence. Instead, Neurath claims that: If one interprets Horkheimer’s emotional turns of phrase, his metaphorical and allegorical statements, and his metaphysical expressions, in an empirical manner as much as it is possible, one comes to serious problems. […] It would certainly be valuable to analyse these problems, which are also related to the status of empiricism, using concrete material from the last 100 years. A comprehensive historico-sociological account of the history of scientific Empiricism is a scientific desideratum; had it existed, one could more easily find out whether one or another of Horkheimer’s metaphysically draped theses point to certain connections, which have so far escaped the notice of many empiricists—but even if such examination were to prove this, it would still be a long way from proving that Horkheimer’s method is useful and that his metaphysical style is necessary. (Neurath 1937a: 24) Thus, according to Neurath, determining the relation between scientific and social activity constitutes a project that is vital to an empiricist and materialist conception of Unified Science. Nonetheless, in Neurath’s view, the supraempiricist method that Horkheimer approaches it with is questionable, if not simply redundant. Instead, according to Neurath, an empirical social scientific and historical approach to the role empiricism has played within the history of science provides a method of testing Horkheimer’s claims

152

The History of Understanding in Analytic Philosophy

(once they are correctly reformulated, avoiding metaphysical terminology). Neurath calls for a critical and reflexive standpoint that would be able to scientifically address the problems addressed by Horkheimer. Neurath refers to this standpoint as being within the “comprehensive scientific outlook” (1937a: 8) which the Unity of Science movement advocates. The proposed critique would thus come from within empirically minded science, rather than from some impossible position above it. In this way it is possible to test whether Horkheimer’s claims concerning the social effects of empiricism are true or false.

6.6. CONCLUSION In conclusion, there is an interesting parallel to be drawn here between Neurath’s criticism of Horkheimer’s critical conception of Vernunft and his views about Verstehen. As we have already seen, Neurath responds to statements about Verstehen by attempting to salvage their empirical content while avoiding being misled into metaphysical errors through the misuse of language. Neurath’s strategy in defense against Horkheimer’s critique of positivism (or at least a straw man thereof) runs parallel to this empirical rehabilitation of Verstehen insofar as it is an exhortation to separate the metaphysical chaff from the empirical grain. Some of the views defended both by Dilthey’s followers and by Horkheimer challenge, in different ways, Neurath’s view of a Unified Science. Both institute different types of separation between different types of endeavors that Neurath would like to unify. In the case of metaphysical conceptions of Verstehen, the threat comes in the guise of a division within science; namely, its separation into Naturwissenschaft and Geisteswissenschaft. Neurath’s proposed rehabilitation of the metaphysical vocabulary engaged in the envisaged separation intends to pave a path towards meaningful communication between Naturwissenschaften and Geisteswissenschaften within the framework of Unified Science. Horkheimer’s Frankfurt School, on the other hand, threatens to create another type of division; namely, that between scientific activity and the critical theory that purportedly stands above it (e.g., Horkheimer [1937] 1972: 182–3; see Uebel 1992: 295). Neurath’s clearing away of the metaphysical elements in Horkheimer’s objections looks towards an integration of the Frankfurt School’s critical project into the encyclopedic project of Unified Science. In both cases, Neurath resists the “separatist” tendencies by proposing manners of empirically rehabilitating them, thus making room for them within the wide scope of Unified Science. Neurath’s strategy of rehabilitation involves showing how to avoid the employment of metaphysical language which had prompted such separatist tendencies in the first place.

Neurath’s Debate with Horkheimer

153

The critical aspect of such a project—what was often misconstrued as the goal of “eliminating” metaphysical language—allowed for various caricatures to emerge; for example, the one which we find in Horkheimer. Neurath (1937a: 8–9) is conscious of this when he notes in his reply that, though of course highly significant, the overcoming of metaphysics through the use of logical analysis is but a small aspect of the logical empiricist project. While Horkheimer’s caricature focused almost exclusively on limitations imposed by anti-metaphysical strictures, Neurath’s vision for Unified Science was inclusive and encyclopedic, and it included reflexive and critical elements (see Neurath 1937a: 4–6). Contrary to the apparent implications of his infamous use of an index verborum prohibitorum for metaphysical terms, Neurath did not seek to ban either claims about Verstehen or critical reflection on scientific activity and its relation to social praxis. Of course, though Neurath’s strategies against Verstehen and against Horkheimer’s Vernunft may be similar, their targets are not the same. The method of Verstehen, for Dilthey’s metaphysically inclined followers, involves an explicit conception of a type of knowledge that purports not to be empirically testable. As we have seen, Neurath responds to this by showing how to rephrase the terminology involved so as to render the claims in question relevant to empirically testable statements. What Horkheimer attempts to develop, on the other hand, is not a special type of knowing (like that attempted by what he calls “metaphysics”), but rather a type of reflection on the results of empirical science.30 This is how he claims to avoid both metaphysics and scientism. Thus Neurath’s response, in addition to showing which parts of Horkheimer’s critique may be reformulated so as to be rendered compatible with logical empiricism, also needs to show why his own view of an immanent critique of science is superior to Horkheimer’s critical theory qua supra-empirical metascience. The debate is one about the bounds within which the critique of scientific practice may take place. For Neurath and other logical empiricists, it is clear that there can be no standpoint outside or above scientific practice from which a metascientific critique is possible. In other words, critique is only possible “immanently” (though the metaphysically loaded word would not be in agreement with Neurath’s point of view). On the other hand, Horkheimer and other critical theorists insist on supra-empirical reason as constitutive of the critical standpoint. Horkheimer and Neurath are, however, as we have seen, in agreement concerning a number of other topics. They both agree on the need to criticize postwar tendencies in Germanophone philosophy, which they each see as politically allied with authoritarianism. Furthermore, they each partly agree that the method of Verstehen has limited applications to social

154

The History of Understanding in Analytic Philosophy

research, though each has different reasons and agendas in diagnosing these limitations. Finally, despite appearances to the contrary (as I have shown above), both Neurath and Horkheimer in fact agree about the crucial importance of instituting a kind of reflexive critique of scientific activity in light of its social repercussions. Their main disagreement is focused on the positioning of this critical activity in relation to science. Though the breakdown of the Vienna Circle’s attempt to collaborate with the Frankfurt School may have come about largely for extra-philosophical reasons (see Dahms 1994), such methodological disagreements between Horkheimer and Neurath were a contributing factor—at least insofar as they provided an excuse for the breakdown. The opportunity to address such differences through further debate during the late 1930s was missed, alongside the failure of the attempt of the Vienna Circle to collaborate with the Frankfurt School. Instead, the increasingly polemical Frankfurt School criticisms of what its members construed as positivism between the 1940s and the 1960s received no replies from logical empiricists.31 Since both movements would soon be led into serious philosophical trouble due to upholding the positions in dispute, subsequent debates over the autonomy or immanence of the critique of scientific practice were bound to take very different shapes and forms.

NOTES 1 For more detailed studies of the relation between the Vienna Circle and the Frankfurt School, see: Dahms (1994); Bowie (2000); O’Neill and Uebel (2004, 2018); Reisch (2005: 121–5); Wheatland (2009: 115–21); Chase and Reynolds (2011: 31–4); Samios (2012); Turner and Factor (2014: 224–5); Dewulf (2017); Vrahimis (2020a). 2 Dilthey’s conception of Verstehen is discussed in detail by, for example, Makkreel (1992: 247–72), while a defense against common misportrayals by Dilthey’s critics has been undertaken by, for example, Harrington (2001). Similarly, both Horkheimer’s and Neurath’s portrayals of Dilthey could be said to be somewhat reductive. The attempt to correct their respective misreadings of Dilthey exceeds the limits of this inquiry. 3 Horkheimer uses the term “Szientivismus,” which has been translated into English as “scientivism”; I will avoid this translation, using in its place the more common term “scientism.” 4 Horkheimer uses the term “logical empiricism” (and only once “logical positivism” [1937] 1972: 185] and “neopositivis[m]‌ ” [1937] 1972: 139]) to describe the work of the Vienna Circle (as well as, misguidedly, that of Russell and Wittgenstein); in Horkheimer’s view, logical empiricism is a brand of positivism and a representative example of scientism. 5 Though he does not explicitly state this, Horkheimer’s article at least indirectly responds to Carnap’s critique of Heidegger. In fact, following Adorno’s

Neurath’s Debate with Horkheimer

155

recommendations, Horkheimer even deleted his mention of Heidegger by name from an earlier draft of his paper; see Dahms (1994: 122–3) and Vrahimis (2020a). 6 A more extensive analysis of Horkheimer’s arguments is offered by Vrahimis (2020a). 7 A year before Horkheimer’s 1937 article, and based on a very selective reading of the early Schlick, Marcuse ([1936] 1968: 47) had already briefly sketched this criticism of the “one-dimensional” positivist emphasis on value-free facts. Marcuse appeals to the Marxist critique of Erscheinungsform in objecting to Schlick’s rejection of the metaphysical distinction between essence and appearance. 8 But see Uebel’s (1992: 295–301) answer to Horkheimer’s charges; Uebel explains how normative questions could be addressed from within Neurath’s reflexive framework for Unified Science. See also Uebel (2020). 9 O’Neill and Uebel (2018: 152) contend that the divorcing of reason from action which Horkheimer began to defend during the 1930s would eventually lead to “the conservative political quietism of Horkheimer’s postwar work.” 10 Marcuse ([1936] 1968: 66), who also sees positivism as serving “the interest of small and powerful economic groups,” nonetheless concedes a “critical tendency” which distinguishes positivism from metaphysical dogmatism and the “regressive social interests” it serves to bolster. 11 In his correspondence with Neurath, Horkheimer (2007: 125) admits that he is unable to produce a criterion which would justify claims made from this metascientific enterprise. Horkheimer nevertheless goes on to note that this is consistent with the main thrust of his work, which had been “trying to demonstrate the extent to which the required criterion can’t be specified, particularly for the most decisive problems” (2007: 125). See also Vrahimis (2020a). 12 On Dilthey’s conception of psychology, see Feest (2007). 13 See Uebel (2019: 3–4). On Carnap’s dialogue with Dilthey, see, for example, Damböck (2012, 2017) and Dewulf (2017). 14 The other case of subscribing to the myth of the given Horkheimer (1940: 437, see also 435) mentions here is not that of logical empiricism, but Bergson. Elsewhere, both Horkheimer (e.g., [1933] 1972: 40) and Adorno ([1970] 2013: 46) argue that Bergson’s position concerning the immediate givenness of data to the senses amounts to a type of positivism. However, Bergson’s appeal to the immediate givenness of intuitive knowledge is heavily criticized by both Schlick and Carnap; see, for example, Uebel (2007: 74–9). 15 Such caricatures of positivism become more and more prominent in later works of the Frankfurt School, escalating during the 1960s into the Positivismusstreit; see Dahms (1994). 16 In earlier works, Uebel (1992: 295–301) and O’Neill and Uebel (2004) addressed the inapplicability of Horkheimer’s caricature to Neurath more generally. 17 On Carnap’s relation to Verstehen, see, for example, Uebel (2019) and Dewulf (2017). Dewulf (2017) argues that critics like Horkheimer and Cassirer did not, in the 1930s, consider Carnap’s earlier Aufbau and its way of accommodating the method of Verstehen within its overall understanding of the Geisteswissenschaften.

156

The History of Understanding in Analytic Philosophy

18 Horkheimer ([1937] 1972: 145, 152–3) was clearly aware of, and had commented on Neurath’s Empiricism and Sociology. It is possible that a superficial reading of the more polemical claims made by Neurath in that book may have informed Horkheimer’s caricature of the logical empiricist critique of Verstehen. 19 Neurath expresses a very similar attitude in his reply to Horkheimer. There, Neurath (1937a: 14) clearly states his appreciation of the concrete sociological analyses undertaken by the Frankfurt School, which he sees as compatible with the logical empiricist conception of Unified Science. By contrast, Neurath criticizes various aspects of Horkheimer’s theoretical framework. Horkheimer in turn reciprocates by expressing, in private correspondence, his admiration for Neurath’s concrete analyses (e.g., his paper on the standard of living [Neurath 1937b] or his work on ISOTYPE [International System Of TYpographic Picture Education]) while remaining critical of the logical empiricist theoretical framework (Dahms 1994: 178–9). 20 Neurath (2020a: 246) points out that even the common view of the social sciences as Geisteswissenschaften is problematic, since they study processes to which the term “Geist” would be admitted as inapplicable even by its proponents. 21 Neurath also criticizes Max Weber and neo-Kantian philosophers such as Rickert and Windelband. It should here be noted that in responding to Spengler, Neurath does criticize another specific aspect of the late Dilthey’s legacy; namely, his metaphysics of worldviews. Even here, however, Neurath should be understood as calling for a kind of empirical rehabilitation of Dilthey’s conception of worldviews, rather than rejecting it wholesale as metaphysical. In two recently translated articles, Neurath tackles the concept of worldviews in relation to his own brand of Marxism (Neurath 2020a) and in his critical response to Mannheim (Neurath 2020b). On the complex engagements of the Vienna Circle with Lebensphilosophie, see Vrahimis (2021). 22 On the middle-class political and cultural history of nineteenth-century conceptions of Verstehen; see, for example, Phillips (2010) and Arens (2010). 23 Cooper (1996) argues that despite Dilthey’s own liberal politics, his holistic conception of Verstehen was compatible with, and would even enable, later irrationalist fascist appeals to some purported organic unity of “life.” 24 Cooper (1996) claims that Dilthey shares with later fascist tendencies the view that claims about Verstehen are irreducible. 25 For example, Uebel (2019: 6–8) shows that in his analysis of Max Weber, Neurath exaggerates the influence of Rickert’s metaphysics on him, missing the fact that Weber had in fact expressed his doubts about Rickert. 26 See Uebel (2019: 6). In Sombart’s account, Germany and its Kultur were differentiated from the merely technological civilizations of the “West”; see also Vrahimis (2015: 89–90). 27 A similar conflation between metaphysics and nationalist politics occurs in Spengler, who is elsewhere targeted by Neurath ([1921] 1973). 28 Neurath’s reply remained unpublished from 1937 to 2011, when it was first made available in English translation (Neurath 2011).

Neurath’s Debate with Horkheimer

157

29 To avoid confusion it should be noted that, as Neurath clarifies, the sense of Verstand here is the Kantian one involved in Horkheimer ([1937] 1972), rather than the Diltheyan notion discussed in Horkheimer (1940). 30 It should be noted here that Neurath’s conception of the Unity of Science is also centrally preoccupied with the possibility of reflecting on the results of individual sciences from within an empiricist framework; see, for example, Zolo (1989) and Uebel (1992: 294–301). 31 See also Vrahimis (2020a). Interestingly, Horkheimer’s critique of Verstehen as overly subjective is more or less retained by Habermas in the context of the Positivismusstreit in the 1960s (Adorno et al. 1976: 140). One of the theses Popper defends during that debate is that “we cannot reduce to psychology what has often been termed ‘verstehende Soziologie’ (the sociology of [objective] understanding)” (cited in Adorno et al. 1976: 63).

REFERENCES Adorno, T. W. ([1970] 2013), Against Epistemology: A Metacritique: Studies in Husserl and the Phenomenological Antinomies, trans. W. Domingo, Cambridge: Polity. Adorno, T. W., H. Albert, R. Dahrendorf, J. Habermas, H. Pilot, and K. R. Popper (1976), The Positivist Dispute in German Sociology, trans. G. Addey and D. Frisby, London: Heinemann. Arens, K. (2010), “Erklären, Verstehen, and Embodied Rationalities: Scientific Praxis as Regional Ontology,” in U. Feest (ed.), Historical Perspectives on Erklären and Verstehen, 141–60, Dordrecht: Springer. Bowie, A. (2000), “The Romantic Connection: Neurath, The Frankfurt School, and Heidegger, Part Two,” British Journal for the History of Philosophy, 8 (3): 459–83. Carnap, R. ([1931] 1959), “The Elimination Of Metaphysics Through Logical Analysis of Language,” in A. J. Ayer (ed.), Logical Positivism, 60–81, Glencoe, IL: Free Press. Chase, J. R. and J. Reynolds (2011), Analytic versus Continental: Arguments on the Methods and Value of Philosophy, Durham: Acumen. Cooper, D. E. (1996), “Verstehen, Holism and Fascism,” in A. O’Hear (ed.), Verstehen and Humane Understanding, Royal Institute of Philosophy Supplement, Vol. 41, 95–107, Cambridge: Cambridge University Press. Dahms, H.-J. (1994), Positivismusstreit, Die Auseinandersetzung der Frankfurter Schule mit dem logischen Positivismus, dem amerikanischen Pragmatismus und dem kritischen Rationalismus, Frankfurt: Suhrkamp. Damböck, C. (2012), “Rudolf Carnap and Wilhelm Dilthey: “German” Empiricism in the Aufbau,” in R. Creath (ed.), Rudolf Carnap and the Legacy of Logical Empiricism, 67–88, Dordrecht: Springer. Damböck, C. (2017), «Deutscher Empirismus»: Studien zur Philosophie im deutschsprachigen Raum 1830–1930, Dordrecht: Springer. Dewulf, F. (2017), “Rudolf Carnap’s Incorporation of the Geisteswissenschaften in the Aufbau,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 7 (2): 199–225.

158

The History of Understanding in Analytic Philosophy

Feest, U. (2007), “‘Hypotheses, Everywhere Only Hypotheses!’: On Some Contexts of Dilthey’s Critique of Explanatory Psychology,” Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences, 38 (1): 43–62. Friedman, M. (2000), A Parting of the Ways: Carnap, Cassirer, and Heidegger, Chicago: Open Court. Gabriel, G. (2004), “Introduction: Carnap Brought Home,” in C. Klein and S. Awodey (eds.), Carnap Brought Home: The View from Jena, 3–24, Chicago: Open Court. Harrington, A. (2001), “Dilthey, Empathy and Verstehen: A Contemporary Reappraisal,” European Journal of Social Theory, 4 (3): 311–29. Horkheimer, M. ([1933] 1972), “Materialism and Metaphysics,” in Critical Theory: Selected Essays, 10–46, New York: Continuum. Horkheimer, M. ([1937] 1972), “The Latest Attack on Metaphysics,” in Critical Theory: Selected Essays, 132–87, New York: Continuum. Horkheimer, M. (1940), “The Relation between Psychology and Sociology in the Work of Wilhelm Dilthey,” Zeitschrift für Sozialforschung, 8: 430–43. Horkheimer, M. (2007), A Life in Letters: Selected Correspondence, Lincoln: University of Nebraska Press. Jay, M. (1996), The Dialectical Imagination: A History of the Frankfurt School and the Institute of Social Research, 1923–1950, Berkeley: University of California Press. Makkreel, R. (1992). Dilthey: Philosopher of the Human Studies, 2nd ed. with afterword, Princeton, NJ: Princeton University Press. Marcuse, H. ([1936] 1968), “The Concept of Essence,” in Negations: Essays in Critical Theory, 31–64, Boston: Beacon. Nelson, E. (2018), “Dilthey and Carnap: The Feeling of Life, the Scientific Worldview, and the Elimination of Metaphysics,” in J. Feichtinger, F. Fillafer, and J. Surman (eds.), The Worlds of Positivism, 321–46, Cham: Palgrave Macmillan. Neurath, O. ([1921] 1973), “Anti-Spengler,” in M. Neurath and R. S. Cohen (eds.), Otto Neurath: Empiricism and Sociology, 159–213, Dordrecht: D. Reidel. Neurath, O. ([1930] 1983), “Ways of the Scientific World-Conception,” in M. Neurath and R. S. Cohen (eds.), Philosophical Papers: 1913–1946, 32–47, Dordrecht: D. Reidel. Neurath, O. ([1931] 1973), “Empirical Sociology,” in M. Neurath and R. S. Cohen (eds.), Otto Neurath: Empiricism and Sociology, 319–421, Dordrecht: D. Reidel. Neurath, O. (1937a), “Einheitswissenschaft und Logischer Empirismus: Eine Erwiderung,” Unpublished manuscript, Wiener Kreis Archive (203/K.63), NoordHollands Archief, Haarlem. Neurath, O. (1937b), “Inventory of the Standard of Living,” Zeitschrift für Sozialforschung, 6 (1): 140–51. Neurath, O. (2011), “Unity of Science and Logical Empiricism: A Reply,” in O. Pombo, J. Symons, and J. M. Torres (eds.), Otto Neurath and the Unity of Science, 15–30, Dordrecht: Springer. Neurath, O. (2020a), “Worldview and Marxism (original edition in 1931, translated from German by Alan Scott),” Sociologica, 14 (1): 243–8. Neurath, O. (2020b), “Bourgeois Marxism. A Review Essay on Karl Mannheim, Ideologie und Utopie (original edition in 1930, translated from German by Alan Scott),” Sociologica, 14 (1): 235–41.

Neurath’s Debate with Horkheimer

159

O’Neill, J., and T. Uebel (2004), “Horkheimer and Neurath: Restarting a Disrupted Debate,” European Journal of Philosophy, 12 (1): 75–105. O’Neill, J., and T. Uebel (2018), “Between Frankfurt and Vienna: Two Traditions of Political Ecology,” in S. Smith and K. Forrester (eds.), Nature, Action and the Future: Political Thought and the Environment, 133–56, Cambridge: Cambridge University Press. Phillips, D. (2010), “Epistemological Distinctions and Cultural Politics: Educational Reform and the Naturwissenschaft/Geisteswissenschaft Distinction in NineteenthCentury Germany,” in U. Feest (ed.), Historical Perspectives on Erklären and Verstehen, 15–35, Dordrecht: Springer. Reisch, G. A. (2005), How the Cold War Transformed Philosophy of Science, Cambridge: Cambridge University Press. Romizi, D. (2012), “The Vienna Circle’s “Scientific World-Conception”: Philosophy of Science in the Political Arena,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 2 (2): 205–42. Samios, P. M. (2012), Η κριτική του Max Horkheimer απέναντι στο λογικό θετικισμό μέχρι το 1940, με επίκεντρο τις φυσικές επιστήμες, PhD Dissertation, National and Kapodistrian University of Athens. Scott, A. (2020), “Introduction to Otto Neurath’s “Bourgeois Marxism” (1930) and “Worldview and Marxism” (1931),” Sociologica, 14 (1): 227–34. Stadler, F. (2015), The Vienna Circle. Studies in the Origins, Development, and Influence of Logical Empiricism, Switzerland: Springer. Turner, S. P., and R. A. Factor (2014), Max Weber and the Dispute Over Reason and Value: A Study of Philosophy, Ethics and Politics, London: Routledge. Uebel, T. (1992), Overcoming Logical Positivism from Within: The Emergence of Neurath’s Naturalism in the Vienna Circle’s Protocol Sentence Debate, Amsterdam: Rodopi. Uebel, T. (2007), Empiricism at the Crossroads: The Vienna Circle’s Protocol-Sentence Debate Revisited, Chicago: Open Court. Uebel, T. (2010), “Opposition to Verstehen in Orthodox Logical Empiricism,” in U. Feest (ed.), Historical Perspectives on Erklären and Verstehen, 291–308, Dordrecht: Springer. Uebel, T. (2019), “Neurath on Verstehen,” European Journal of Philosophy, 27 (4): 912–38. Uebel, T. (2020), “Intersubjective Accountability: Politics and Philosophy in the Left Vienna Circle,” Perspectives on Science, 28 (1): 35–62. Vrahimis, A. (2015), “Legacies of German Idealism: From the Great War to the Analytic/Continental Divide,” Parrhesia, 1 (24): 83–106. Vrahimis, A. (2020a), “Scientism, Social Praxis, and Overcoming Metaphysics: A Debate between Logical Empiricism and the Frankfurt School,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 10 (2): 562–97. Vrahimis, A. (2021), “The Vienna Circle’s Responses to Lebensphilosophie,” Logique et Analyse, 253: 43–66. Wheatland, T. (2009), The Frankfurt School In Exile, Minneapolis: University of Minnesota Press. Zolo, D. (1989), Reflexive Epistemology: The Philosophical Legacy of Otto Neurath, Dordrecht: Kluwer.

160

CHAPTER SEVEN

Victor Kraft on Verstehen JAN RADLER

7.1. INTRODUCTION Throughout this text, the German expression “Verstehen” is used instead of its English-language equivalent “understand.” Although this paper is not an exercise in linguistic philosophy, but rather an essay in the history of philosophy (of science), the question may arise of what the different situations and contexts in which Verstehen is used all have in common. It is not difficult to provide the answer: Verstehen refers to teleological explanations of human actions and of institutions and animals—past, present, and future. (This, of course, implies the interpretation of written texts and thus hermeneutics). Von Wright coined the term “Aristotelian tradition” for this type of explanation. By using the German term throughout this essay, I would like to allude to the obviously well-known tradition in German1 philosophy contrasting the (scientific) explanation with (human) Verstehen (cf. Abel 1948: 211, n. 1, for this terminological convention). The former holds for the inanimate nature, and the latter is reserved especially for humans. This is famously summarized in Wilhelm Dilthey’s classical words “Die Natur erklären wir, das Seelenleben verstehen wir” (Nature we explain, mental life we understand) (quoted after Acham 1974: 121). For several decades, logical empiricism has been an object of intensive historical research. Many historical influences on the main proponents of logical empiricism have been illuminated by scholars in Austria, Germany, and elsewhere. In addition to personal contacts and teacher–student relationships, the intellectual and cultural climate of Austria (Topitsch 2005: 85–103), which differed remarkably from that of Germany (i.e., mainly Prussia),

162

The History of Understanding in Analytic Philosophy

has been identified as an object of historical scholarship. Scholars began studying the main proponents of logical empiricism (e.g., Rudolf Carnap, Moritz Schlick, and Otto Neurath), and recently the philosophers in the second line have been rediscovered, Victor Kraft among them. Many logical empiricists had an intellectual background in the natural sciences (especially physics) or mathematics. Because of their interest in the natural sciences, it is not surprising that the “Aristotelian tradition” was only of marginal interest to them. In this, however, Kraft was an exception.2 He had a profound knowledge of humanities and history and even studied for one semester in Berlin under Dilthey and other humanities scholars (Topitsch 1960: iii). Therefore, it might be asked whether Kraft developed a theory of Verstehen; or to put it in other words: Did he develop a theory of hermeneutics or a theory of (human) action? The answer to this question is a strict and plain “no.” Kraft’s main interest in the field of the philosophy of science was to develop a hypothetico-deductive methodological framework for all branches of human scientific reasoning. This effort was supplemented with remarks on how the curious human mind constructed (new) hypotheses. Accordingly, Kraft’s philosophy has a strong constructive inclination, which found its expression in one of his last papers, “Constructive Realism” (Kraft 1973a). How did Kraft reach this result? In an analysis of scientific research practice (initially in the area of geography), he showed how deductive reasoning in science proceeds. Central to his philosophical research program is not only the all-encompassing constructive-deductive methodology but also a normative inclination regarding the most fundamental (metaphilosophical) questions. He scrutinized and evaluated philosophical solutions to determine whether they furnished the desired consequences (Kraft 1960: Ch. 1). In the following, I mainly highlight the historical influences on Kraft. Instead of developing a theory of human action and understanding, Kraft showed how far history can be subordinated under the deductive explanatory scheme. By doing so, he argued against a philosophical tradition which claimed that history is a literal activity and not a scientific one. Kraft took an explicitly rational stance in the contemporary philosophical discourse. Decades later, analytic philosophy tried to mediate between Verstehen and Erklären (cf. for an early3 overview, Stegmüller 1983), but Kraft did not participate in this debate, which started when he was in his eighties. This chapter summarizes his sparse remarks about the Verstehen problem and shows that it was not one of Kraft’s main interests. To be sure, by analyzing humanities, Kraft presupposed hermeneutic assumptions, but he did not spell these assumptions out. Therefore, he can be regarded as a representative of the logical empiricists’ standard view, but his “constructive inclinations” put him in opposition to physicalistic thinkers such as Neurath.

VICTOR KRAFT ON VERSTEHEN

163

7.2.  LOGICAL EMPIRICISM’S STANDARD VIEW ON VERSTEHEN AND KRAFT’S PHILOSOPHY During the meetings of the Vienna Circle, Kraft played only a marginal role. He insisted that he was never an “orthodox” adherent of logical empiricism. The main differences between Kraft and other members of the Vienna Circle will be mentioned shortly. His main interests were the philosophy of science and the theory of knowledge as well as the theory of values and moral philosophy. Kraft always adhered to a decidedly realistic philosophy,4 meaning that he assumed a world independent from the subject could explain regularities in perception. His argument for this has the form of an abductive inference. After the establishment of a realistic world concept, the explanation of regularities in perception is possible. Thus, the core of Kraft’s argument has a normative flavor: one has to accept a world concept if one aims to explain the regularities in perception. Alternatives are excluded because they give rise to misleading results, which means that they cannot guarantee the explanation one is looking for. From a historical point of view, it is worth mentioning that later Paul Feyerabend and Hans Albert picked up this normative element and incorporated it in their own philosophical positions (for details, see Radler 2006 and Kuby 2020). Using the same type of rational argumentative strategy, Kraft tried to establish the foundations of morale. Kraft never became very explicit about this standard of rationality— and this might be the first hint that he missed the first step of developing of a theory of (rational) understanding. After establishing a realistic “world concept”, an explanation is possible. Kraft’s argument represents a hypothetico-deductive position, which Popper later became famous for. (Popper knew Kraft very well, but Kraft was never a falsificationist like Popper; Kraft is—as we shall see later— for other, immanent reasons, a “hypotheticist.”) On the one hand, Kraft’s approach was, as already mentioned, basically normative. On the other hand, Kraft developed his philosophical position hand in hand with a neat examination of many scientific disciplines. In addition to this, he tried to show that the deductive approach can be found in those very disciplines. Logical analysis, which was (or became) very central to the more prominent members of the Vienna Circle, cannot be found in Kraft’s writing. (To a lesser degree, this changed after the Second World War, but Kraft was never a logician.) Kraft’s orientation towards scientific practice was intended as an indirect critique of other positions, such as the classical positivism represented by Ernst Mach or Richard Avenarius. Kraft developed a kind of bottom-up theory of knowledge. This approach profited from his broad interests in

164

The History of Understanding in Analytic Philosophy

cultural science. Did this have an impact on Kraft’s handling of the Verstehen approach? Before answering this question, it is useful to present the standard view of the Verstehen approach in logical empiricism. Danneberg and Müller (1984: 183) summarized the standard view as follows: 1. The Verstehen approach is only relevant for history and the social sciences. The problem of hermeneutics (understanding of texts) is either completely neglected or merely a special case of the Verstehen approach in a broader sense, which does not give rise to any additional problems. 2. (2) Verstehen is identified with activities such as ‘empathizing’ or ‘reliving’ (nacherleben). 3. (3) Verstehen concerns only the realm of heuristics, and for ‘unified methodology’ (Einheitsmethodologie) or ‘methodological naturalism’, Verstehen cannot be implemented, or can only be implemented by relinquishing its specific traits. I will show that Kraft’s position can be regarded as representative of the standard view; however, the third point applies only to a lesser degree.

7.3.  THE HISTORICAL CONTEXT OF KRAFT’S CONSIDERATIONS OF THE VERSTEHEN APPROACH Kraft developed his position regarding the Verstehen approach mainly in engagement with two references. That is, he engaged on the one hand with neo-Kantianism and on the other hand with Alfred Hettner’s writings on geography. 7.3.1.  Kraft’s Interpretation of Southwest Neo-Kantianism All schools of neo-Kantianism have in common that they ask for the validity of knowledge claims; that is, they ask for the validity of propositional knowledge. How it is possible to justify knowledge, especially mathematical and theoretical knowledge? Any approach to justifying knowledge by referring to psychological states of mind is regarded by Kraft as futile. During his seven-decades-long period of publication, Kraft always asked for the validity of knowledge claims, and it is not difficult to discover the root of this philosophical division of labor: Wilhelm Windelband provided Kraft the separation between “critical” and “genetical” methods. “Genetical” is, broadly speaking, an umbrella term for all scientific investigations of the human capacity for reasoning (cf. Radler 2006: 116, n. 7). However, most

VICTOR KRAFT ON VERSTEHEN

165

importantly, it would not be correct to describe Kraft as a neo-Kantian philosopher. First of all, his academic teachers (e.g., Friedrich Jodl) adhered to a position which might be rejected by neo-Kantians as psychologistic. Nevertheless, Kraft discussed neo-Kantian philosophers broadly and with a certain distance in his early monographs (1912, 1925). Heinrich Rickert, the most prominent exponent, besides Windelband, of the Southwest School of neo-Kantianism, claimed that acts of valuation mediate between subjective thinking and objective thoughts. According to Rickert, judgments can be formulated as questions, and by answering these questions, the subject takes a stance affirming or rejecting the judgment. By taking this position, Rickert tried to overcome the subjectivity of thinking and to establish the objective validity of knowledge claims. Rickert spoke from a “transcendent ought,” which has nothing to do with practical norms of morality. In a modern manner of speaking, this would mean the following: insofar as a proposition is formed, the transcendent ought is acknowledged. This is, for Rickert, the only way to guarantee the objectivity of knowledge. In addition, values also play an important role by separating sciences from humanities (for details, see Staiti 2018). The cultural sciences (humanities) regard the individual as important, whereas the sciences are general. In addition to this, the objects of cultural sciences are value related. This differentiation between sciences and humanities provides enough space for the opinion that individual processes can also be analyzed in natural science. Consequently, Rickert dissociated himself from Windelband, who introduced in philosophy the differentiation between “nomothetic” and “idiographic” sciences, a differentiation Kraft criticized as well. Kraft was perfectly clear that Rickert’s position is a kind of idealism. Regarding Rickert, Kraft stated that “knowledge, as a product of mind, is a methodological concept formation, a processing of the sensually given after the requirements of logical thinking” (Kraft 1912: 155). This processing depends on the acknowledgment of “transcendental norms” (ibid.). Although Kraft criticized Rickert for his idealism, his own realistic epistemology was grounded in a normative postulation of the foundations of knowledge. He stated, “[i]‌f one tries to avoid a [kind of subjective idealism] then the principles of knowledge can only be regarded as normative axioms” (Kraft 1912: 196, emphasis added). However, after reading this one might ask the following: normative axioms for whom? Rickert postulated that these normative axioms are for an “ideal subject”, but as a critical contemporary commentator remarked, the acceptance of theoretical values must be considered as a “real psychological act” (Messer 1923: 373). Thus, if one takes Rickert’s normative epistemology and strips it of his transcendentalism, one can transfer Rickert’s intuition into a philosophical framework that is

166

The History of Understanding in Analytic Philosophy

open and connectable for empirical (psychological) findings. This is how Kraft proceeded, as he combined a normative stance with empirical findings. Nevertheless, in his late writings, Kraft mentioned that southwest neoKantianism is quite right in emphasizing the role of normative conventions (Festsetzungen) as the basis of epistemology. According to Kraft, the concept of knowledge (Erkenntnisbegriff) is a norm. This implies that such a concept of knowledge has only hypothetical validity and, as a consequence, the quest for an absolute fundament of epistemology is to no avail. Here, the difference between Kraft’s and Rickert’s philosophies becomes very clear. This normative argumentation strategy can be found in three areas of Kraft’s philosophy: (a) in his argument for a realistic world concept, (b) in his discussion of the concept of knowledge, and (c) in his discussion of the foundations of morale.5 In any case, Kraft argued that alternatives to his proposals cannot furnish the desired results. This requires a consensus about theoretical and normative aims. (In particular, it is far from obvious whether consensual agreement on the foundations of morale could be established.) Is the statement that expresses the consensus a synthetic or an analytic proposition? (cf. Stegmüller 1983: 490–5). To repeat, Kraft was clear that he proposed a hypothetical world concept (Kraft 1912: 183); thus, his answer would be that he had formulated a synthetic proposition. The concept of knowledge, the realistic world concept, and the foundations of morale are stipulated based on practical consideration.6 These claims are important in the context of the Verstehen problem since the philosophical analysis no longer rests on a transcendental ideal subject but on human beings. Their actions or reasoning have to be understood. From these claims, it is clear that Kraft presupposed that something is intended by these normative stipulations. How does Kraft proceed in detail? In his first monograph (Kraft 1912), he argued for a realistic world concept; that is, there is a world that is independent from the epistemological subject. By stipulating the “outer world” (Außenwelt), the regularities in perception can be explained. This explanation is an abductive inference to objective objects whose existence causes perception (cf. Schurz 2008). The explanandum is the psychological fact of object permanence. This illustrates Kraft’s tight connection to empirical sciences. Given the logical possibility that an extravagant idealistic world concept can be established, Kraft introduced a normative premise into his argumentation. Any realistic world concept is more parsimonious and is based on commonsense realism. In the end, however, it remains an (ultimate) decision. For example, Kraft said, “[a]‌reality beyond the consciousness can only be stipulated, it persists for us, because the reality is theoretically and practically acknowledged” (1912: 209, original emphasis, cf. 218–19).

VICTOR KRAFT ON VERSTEHEN

167

This normative approach is for Kraft exemplified by the Southwest School of neo-Kantianism. However, there are also important differences between his normative approach and neo-Kantianism. Kraft’s ought is not a “transcendent” one but one that posits practical aims. This practical syllogism has nothing to with transcendental philosophy (cf. Bast 1999: xi) but bears similarities with pragmatism. The assumption of an external world is used as a proposition that explains the perceptual regularities deductively. Kraft noted that this assumption of an external world “proves its validity due to its concrete benefits” (1912: 198, emphasis added). Kraft received important ideas for his philosophy from the Southwest School of neo-Kantianism but reframed its basic assumptions in order to fit them in his realistic and, to a certain degree, naturalistic epistemology. Even in 1960, he said that he had shocked his readership with his “relativistic” foundation of epistemology but claims that southwest Kantianism with its normative stance has made a point (Kraft 1960: 32). Any alternatives to his pragmatic foundation are bound to fail. Kraft’s reference to the Southwest School differs from other logical empiricists’ connections to neo-Kantianism. Carnap, just to name the most prominent of them, refers heavily to the Marburg School (e.g., Cassirer, Cohen, Bauch). Typical for the Southwest School is “an explicitly dualistic conception according to which the realm of pure thought stands over and against a not yet synthesized manifold of sensation (a not yet formed ‘matter’), whereas the Marburg School strives, above all, to avoid this dualism” (Friedman 2000: 31). It was due to this dualism that Kraft was able to argue normatively and to include empirical findings in his philosophical argumentation. Let us interrupt our historical reconstruction of Kraft’s philosophy and dwell on its implication for the Verstehen problem. Kraft argued for realism. His aim was to convince others to reconsider his position. This means that everyone who aims to explain the perceptual regularities has to accept an external world as a common cause for those regularities. This is a hypothetical assumption. Imagine a person whose perceptual impressions bear no regularities; such a person is in no need of a realistic worldview. (In addition to this, it is inconceivable how such a person could learn anything.) Since we live in an ordered world, Kraft’s argument becomes reasonable. His argument rests on general empirical assumptions, called Ducasse sentences by Wolfgang Stegmüller (1983) in his reconstruction of Georg Henrik von Wright’s classical position (1971/2004).7 Without going into the details, it suffices to notice that a Ducasse sentence is an empiricalhypothetical formulation of a law. Thus, it is possible to speak from a “double relativization”. On the one hand, Kraft clearly acknowledged the

168

The History of Understanding in Analytic Philosophy

plurality of different concepts of knowledge, while on the other hand, his argumentation refered to a general (empirical) statement, and it was thus fallible. It was a synthetic proposition. Or to put it in other words: There is a plurality of aims and a general fallibility concerning the propositions expressing the means (including the ancillary conditions). 7.3.2.  Kraft’s Methodology of Geography Kraft made his first steps in the philosophy of science in the realm of geographical methodology, and this was obviously due to his academic education in this field (for the following, see Radler 2008 and Ostermeier 1987: 214–20). Geography is characterized by a “double dualism”: on the one hand, the general geography, and on the other hand, the regional geography (Länderkunde). Conversely, one can also distinguish between physical geography and human geography. Initially, for Kraft, geography was both a nomothetic and an idiographic science, and he mentioned the famous distinction made by Windelband. Already in his first publications, he adhered to a deductive concept of explanation: “Explanation means, to understand the singular by referring to its conditions and not only to plainly state it, but deduce” (Kraft 1914: 4; see also 1912: 175, 1973b: 21–2). He distinguished the regional geography and the general geography by referring to their different epistemic goals. However, for Kraft, geography is neither strictly nomothetic nor strictly idiographic. The methodology of geography includes both facets: scientific observation like that in other sciences as well as “conceptual processing” (begriffliche Verarbeitung). On the one hand, travel reports must be the topic of a critical assessment,8 and on the other hand, idiographic description uses universal type concepts, and these concepts are theoretical ones (Kraft 1914: 8). The procedure is both inductive9 and deductive. An inductive procedure needs (conceptual) constructions “in order to reach general laws” (Kraft 1914: 8). Again, after the construction of those laws, it is possible to deduce a test statement that can be “compared with the reality whether it bears those deductive appearances” (Kraft 1914: 7; cf. Davis 1912: 334–51; Radler 2006: 77–81). Referring to the historical sources of his position, it is worth mentioning that he was inspired by William M. Davis’ geomorphological theory, which he got to know through his academic teacher Albrecht Penck. Davis described how an uplifted landscape is formed through erosive forces. Davis’s theory clearly has a deductive structure, although he emphasized that deductive reasoning goes hand in hand with inductive reasoning and observation (e.g., Davis 1912: x, 29 et passim).10 In a remarkable letter to the German geographer Alfred Hettner, Kraft explained that his occupation with Davis’s theory

VICTOR KRAFT ON VERSTEHEN

169

had led him to a more critical position towards the nomothetic-idiographic dichotomy (see Radler 2008 for references). Fifteen years later, Kraft repeated his methodological remarks about geography. He emphasized that he argues as an epistemologist, not as a (geographical) practitioner. Now, he concluded that his analysis of geographical research practice has made clear to him that there is neither a common epistemological goal for geography nor a uniform method. Of course he had already made these points in 1914, but interestingly, in 1929 he did not use the neo-Kantian predicates idiographic and nomothetic. This might mean that Kraft adjusted his intellectual compass, and probably here the influence of the Vienna Circle played a role. 7.3.3.  Kraft and Verstehen in Historical Science What are Kraft’s views about the methodology of the historical sciences? He presented his position mainly in two essays (Kraft 1929 and Kraft 1963). In his paper “Intuitive Understanding in Historical Science,” he emphasized primarily that history is concerned with laws and regularities (Gesetzeswissenschaft) and not with “individualized historical knowledge” (individualisierende Geschichtskenntnis, Kraft 1929: 2). He distanced himself from Rickert, who primarily focused on value embedded in cultural appearances, and he stressed that it is not value but sense (meaning) that is the starting point for analysis, so one could therefore speak of (to use Gomperz’s, 1929, telling expression) “sense structures” (Sinngebilde) to denote the objects of history and cultural sciences. Because “sense” was now the starting point of his analysis, Kraft explained why the problem of understanding becomes important. Thus, Kraft cut the cord in two steps. First, he became more and more critical about the nomothetic-idiographic distinction due to his occupation with geography; second, he explicitly admitted that “value” is not the central concept that should guide the analysis of cultural and historical sciences, but that the comprehension of sense is the central concept that should guide this type of analysis. In critical discussion with Eduard Spranger, Kraft sketched his position concerning Verstehen in detail: Thus, verstehen means, in contrast to explain, to think something in a teleological sense-context, not in a causal one. In this way also for Spranger verstehen is primarily the grasping of an objective sense and, secondly, reliving (nacherleben) of subjective mental states the specific methodological procedure of the humanities, thus also for history. […] Explaining means deducing logically from laws. […] Here, the validity

170

The History of Understanding in Analytic Philosophy

rests on logic and natural laws and the empirically given special values. Verstehen is a direct presentation. Intuitively, the sense is grasped, we have to able to directly plunge into the foreign mental life (fremde Innenleben). Verstehen intuitively takes place (sich vollziehen). (Kraft 1929: 5–6)11 Several aspects of Verstehen are mentioned in this remarkable quotation. First, Verstehen has a semantic dimension, which has nothing to do with the understanding of causal relations. Second, Verstehen proceeds intuitively and—this was not mentioned by Kraft—at a remarkable speed. (This is called “parsing.”) Third, Verstehen is a metarepresentational concept (e.g., “I know what you know”). Here, Kraft had formulated three aspects of a modern theory of Verstehen (cf. Detel 2014), but he did not unfold them further. Instead, he concentrated on the concept of sense (Sinnbegriff). Kraft differentiated the concept of sense in three ways. Primarily, “sense” means the symbolic content of a sign. In other words, “sense” means meaning. In addition to this, one can distinguish “structures of sense” (Sinngefüge), such as sentences, logical conclusions, or narratives. This presupposes that the use of the sign is consistent within these structures of sense. Finally, one can attribute “sense” to artworks, cultural objects, machines, or actions. Because of the last fact, he uses a teleological concept of sense (Kraft 1929: 10). Verstehen in historical science means that a “complex of facts” (Tatsachenkomplex) appears in a new light if sense is attached to it. To be sure, the same complex of facts can be understood differently. This applies to actions because several layers of intentions can be identified (cf. von Wright [1971]2004: 26; Acham 1974: 69; Anscombe 2020: §§ 23, 26). I will discuss this point later. Kraft emphasized that Verstehen depends heavily on the interpreter. On many occasions, Verstehen takes place intuitively (Kraft 1929: 25), but on many other occasions, it depends on the interpreter’s acquired knowledge. In any case, Verstehen always remains subjective and therefore lacks an objective validity. On its own and from an epistemological point of view, becomeunderstandable (verständlich werden) means only a heuristic idea. If this idea is correct it still has to be tested. The conviction, linked with verstehen is a subjective feeling, provides no ground for legitimacy (Geltungsgrund). Intuitive understanding only denotes the psychological act of its formation, but does not provide the epistemological fundament, on which knowledge rests. From a psychological point of view everything that was said about the intuitive character of verstehen is correct; a leading role of intuition can and ought not be denied. But epistemologically this is not enough. (Kraft 1929: 17–18)

VICTOR KRAFT ON VERSTEHEN

171

Consider, for example, the understanding of an action. The action can be understood differently depending on the sense attached to it, but it can be understood in all cases with the same intuitive feeling. Kraft said, for example, “Intuition cannot provide any basis for legitimation (Geltungsbasis)” (Kraft 1929: 19). This is not only true for history but also more generally. However, how is it possible to erect history as a science on a firm fundament? If intuition is ineligible, then only a methodological foundation remains. These were, for Kraft, the “processes of logical reasoning” (Kraft 1929: 20).12 Essentially, Kraft had an inference to the best explanation in mind, using the juristic term “circumstantial evidence” (Indizienbeweis). According to Kraft, the meaning of a sign is normally given. In other cases, only an investigation might show whether a “complex of facts” has a meaning. This procedure is inductive (Kraft 1929: 21). The sense of text is provided by an interpretation and is based on a deduction based on “general conditions,” which are formed through inductive13 generalization (Kraft 1929: 21). To give an example of how this would work: the text is compared with other texts of the same type or from the same author; then one formulates hypothetical statements about the text; and finally one generalizes them and deduces in a second step a sentence that can be tested against the original text. The same procedure (inductive generalization and then deduction) can be used for teleological understanding. This means that Kraft embedded Verstehen in his general methodological hypothetical-deductive framework, which was presented four years earlier (Kraft 1925). An explanation of a historical person’s action can only be worked out by referring to the historical sources and by referring to generalized assumptions about humans’ disposition to act. Therefore, a modern theorist of Verstehen refers not only to empirical assumptions (Ducasse sentence, see Section 7.3) but also to a basic rationality (e.g., Detel 2011: 348). According to Kraft, the result of this analysis is the understanding of the inner life of historical persons. It provides “a model to relive the historical [person]” (Kraft 1929: 24) and has the heuristic function of paving the way for new (historical) knowledge. That means that progress in historical science does not only occur through the discovery of new sources. Moreover, the more brilliant the historian, the greater the chance for a new interpretation. As already mentioned, this is only a heuristic function, and history gains legitimacy only in the same way as other knowledge: through deduction. Kraft has merely highlighted the historical synthesis and described it as a psychological-logical process. How does the historian proceed? Her or his primary task is the critique of historical sources. Kraft emphasized that this is logical reasoning from laws or rules and known singular facts. First, according to Kraft, one has

172

The History of Understanding in Analytic Philosophy

to assume that the sense for all the different historical sources has to be the same. Second, one has to reassess the coherence of the sources. Here, circumstantial evidence has to be taken into account as well. Third, the “historical synthesis” is constructed. This “hypothetical construction” is tested by deducing sentences and comparing them against the facts (Kraft 1929: 26–7). Given this rather schematic procedure, how is a “new” interpretation of history possible? How is falsification possible? Clearly, the sense of the central terms has to be the same, and a falsifying counterexample destroys the interpretation’s coherence. It becomes clear that on a more basic level, Verstehen, even in Kraft’s standard account, plays a central role. The hypothetico-deductive reconstruction presupposes the correct understanding of the central terms. However, this understanding was for him rather a psychological phenomenon, not a semantic one, which can be transformed into a logical analysis (cf. von Wright [1971] 2004: 30). Kraft did not revise his position. In 1963, he published a short paper with the title “Geschichtsforschung als strenge Wissenschaft” (Historical Research as Strict Science).14 Additionally, in this late publication, Kraft spoke about “immediate understanding” and referred to intuition (Kraft 1963: 73). He separated this understanding from the comprehension of a new language; rather, this intuitive process is a rediscovery of sense. In the same way that in natural science a new15 theory is postulated, intuition plays this role in historical research. Obviously, Kraft refered here to the so-called “eureka moment”. Once again intuition alone is not sufficient to provide knowledge (Kraft 1963: 76), and after constructing a synthesis one has to deduce test statements in order to verify (or falsify)16 these hypothetical stipulations. Even at the end of his life he did not modify his position. He contested that Verstehen might provide some validity in addition to the usual scientific procedures, such as logical deduction and testing. By doing so, he distanced himself from the positions of Dilthey and Karl Vossler (Kraft 1973b: 24). The main ideas are repeated nearly literally. Verstehen is a heuristic idea, and intuition provides the synthesis of propositions about singular facts but cannot provide any foundation for knowledge (Kraft 1973b: 26). Kraft’s comments are a striking example of the classical logical empiricist standard (dis)solution of the Verstehen problem. It is worth mentioning that Kraft did not develop his position in a discussion with logical empiricists or their publications. Instead, his standpoint has other, earlier, roots. Insofar as Kraft is in accordance with other logical empiricists, it might be interesting to dwell on whether he and other logical empiricist philosophers referred to the same sources. I think that here the hint regarding Neo-Kantianism might

VICTOR KRAFT ON VERSTEHEN

173

be appropriate (cf. Staiti 2018). Kraft always distinguished between factual and validity claims (Faktenfragen and Geltungsfragen). The latter are secondorder questions. The neo-Kantianism starting point was always the notion of separating psychological from transcendental discussions.17 This separation was central for Kraft as well, since he spoke from the psychological process of synthesis. Intuition is only the first step for new knowledge. Surveying Kraft’s remarks on the Verstehen problem, it becomes clear that they lack originality. However, what is remarkable is the historical context. Kraft and his contemporaries observed philosophical developments that are regarded as irrational or at least in strict opposition to “scientific philosophy” (cf. Bühler [1927] 1978); hence Kraft’s permanent insistence that there is no unique or special method of historical sciences that is connected with a philosophy, which highlights “intersubjective verifiability.” This argumentative strategy embraces the central insight that science (including humanities and history) is a collective endeavor.

7.4.  AN ALTERNATIVE: VERSTEHEN AND “EMPIRICAL SEMANTICS” Kraft’s standard view can be contrasted with another theoretical development, that goes back to Otto Neurath. In the following, I will sketch these ideas as briefly as possible (for details, see Radler 2015). In doing so, I will illuminate Kraft’s position on the Verstehen problem from another perspective. Kraft always looked at Neurath’s philosophy from a certain distance. This distance becomes clear if we analyze the following lengthy quotation in which Kraft is critical of Karl Bühler’s solution to the Verstehen problem. The intention behind Kraft’s critical appraisal is clear: empiricism is in need of assumptions that are argued for in a second-order analysis. This epistemological-critical bound [i.e., the impossibility to perceive other minds directly] holds also for Bühler’s “contact theory.” According to this theory other minds are perceived in the same manner as other perceptual traits of bodies. […] Foreign mental states are given with their physical expressive symptoms. […] “[If this denoting function is given,] then it makes no difference whether the redness of the cherry is a sign for their sweetness, or whether the redness of the human cheek serve as a sign for shameful and shy behavior” (Bühler 1927/1978, 97). But the difference remains indissoluble that data, which functions as perceptual signs, are in the case of physical properties perceivable on their own, but in the case of mental states, they cannot be perceived directly but have to be added in thought. (Kraft 1929: 19, n. 1)

174

The History of Understanding in Analytic Philosophy

To put it simply: the sweetness of a cherry can be perceived directly by tasting it, but we lack a sense organ that allows us to perceive mental states of other persons directly. Only by supplementing the perceptual given (e.g., the perception of a red cheek) can the mental state be inferred. However, what is the correct supplement? Kraft was not very explicit, but his notion “to add something in thought” (hinzudenken) alludes to an important point. The redness might be a natural sign (Anzeichen) of hypertension, or the cheeks might be red after a walk on a frosty winter morning. The redness in both of these and in similar cases stands in a causal relationship to some other phenomenon. Shame has a semantic content since it is a feeling (I feel shame for something). This content implies the understanding of actions and verbal statements, and then the feeling of shame is attributed. Kraft demanded correctly that an emphatic relive has a heuristic value, but he did not spell out what it means to understand a person. The metarepresentational assumption that my vis-à-vis has mental states, that she thinks and that she is bounded to a basic rationality, is an assumption that has to be made in order to understand actions (cf. Detel 2011: 334, 2014: 56). This assumption makes behavior understandable and explainable in the same way the assumption of the external world makes perceptual regularities understandable and explainable (Kraft 1960: 273, 280). Kraft was not very precise on this, but a basic rationality has to be assumed (Kraft 1960: 278–9). In addition to this, an act of comprehension and to relive is necessary. Verstehen is not only to know which perceivable behavior precedes or follows an action or reaction, neither it is the ability to react appropriately on perceived behavior, including verbal behavior. Rather, verstehen consists in the ability to relive (nacherleben) what one feels by acting or reacting. Thus I understand the behavior of another person, for example such a simple one as a child’s desire for sweets, since I know that it tastes good for the child since I had made the experience and had had the desire of tasting candy. On the other hand, I cannot understand why a person smokes because I am a non-smoker and it does not matter how good I know the smoker’s behavior. (Kraft 1960: 281) Thereby, Kraft tied Verstehen (of human behavior) to the metarepresentational ability of reenactment. As we have seen, the explanation of historical phenomena as an appeal to the intellectual horizon of the historian is indispensable. By doing this, Kraft introduced very high standards and unduly narrowed the concept of Verstehen. Researchers can be confronted with a completely strange and unknown human behavior which is completely new, but such a behavior is nevertheless understandable.18 Modern theories

VICTOR KRAFT ON VERSTEHEN

175

of Verstehen refer at this point to, for example, Donald Davidson’s theory of radical interpretation (e.g., Detel 2011: 379–91). The historical context of Kraft’s position is his distance from the logical empiricists’ physicalism.19 Let me begin with Bühler, who was the starting point for our discussion. Bühler started his career by investigating thought processes. In his experiments, difficult or even cryptic formulations (aphorisms, philosophical phrases, sayings, and so on) were presented to test subjects. Bühler observed at the end of the thought process that the test subjects suddenly grasped the meaning of the given formulations. Bühler described this moment, this “latching” of meaning, as an “Aha experience” (Aha-Erlebnis). He said, “Not seldom the spontaneity of the task-solution emerged [… ;] the test persons did not know better as to describe it that with an inner Aha! the solution fell into place” (Bühler 1922: 21). The interjection “Aha!” is a mere announcement (Kundgabe), a sign of the “sudden emerging of meaning” (plötzliche Eintreten des Sinns, Bühler 1909: 117).20 Is it possible to erect a research program on Bühler’s basic methodological assumptions? This would be a research program21 that does not insist on subjective reenacting but interprets the “latching” of meaning as a sign that a cognitive process has come to an end. This task was taken up by, among others, Bühler’s student Egon Brunswik. In his studies about perception, he argued that test subjects should estimate the volume of bodies in comparison to reference objects. In our context, the following is important: The interjection “aha!” is not a sign that the cognitive process has come to an end, but the verbal report that the two compared objects seem to be of the same volume. Brunswik (1934: 40, cf. 21, 28) referred to the notion of sameness; thus, Bühler’s and Brunswik’s methodological considerations were based on the desire to eliminate the subjective elements as far as possible. (Remember that Kraft argued that we ought to incorporate subjective elements in our methodological considerations!) Brunswik started his research with the perception of physical properties, but mentioned that other psychological phenomena could be investigated in the same manner (Brunswik 1934: 210–25; cf. Radler 2013, 2015). In Brunswik’s research program, subjective elements are eliminated as far as possible. By doing so, he was able to guarantee the intersubjective status of psychological research. Arne Naess used the same methodology in his “empirical semantics” (Naess 1953), and it was in his work that Verstehen gained its central place. Bühler started with rather complex and cryptic formulations, presupposing that a test subject was acquainted with the formulation (Bühler 1909: 117). Brunswik investigated facets of human perception, and Naess asked test subjects to decide whether or not two formulations were synonymous. The analogy to Brunswik’s approach is obvious. The report of sameness (or

176

The History of Understanding in Analytic Philosophy

synonymity) was systematized in both approaches. An important result was the discovery of the “depth of intention” (Brunswik 1934: 101–5; Naess 1953: 77), a point that was later used by Anscombe (2020) and von Wright. Let me state the basic idea as simply as possible. Brunswik discovered that test subjects see “through” property p1 another property p2. (One sees a coin’s value through the coin’s size.) Thus, mediation plays a very important role. Naess’s use of the term ‘depth of intention’ was slightly different: some people are able to understand or to produce a formulation in a more specified manner. This is quite normal: the layman uses language in a sloppy way, whereas the expert is able to differentiate linguistically the tiniest details. Here, different knowledge might be a first explanation for this fact. The next move is the step from understanding formulations to understanding human action. Different knowledge furnishes different interpretations of behavior. In other words, the same observable behavior can be interpreted differently depending on the knowledge of the observer (interpreter) and the knowledge of the observed person. It is worth mentioning that it is not the causal relationship that constitutes the unity of different observations, but the same intention (von Wright [1971]2004: 89, cf. 26). Naess made interesting use of his results. Formulations with a low degree of preciseness (a low depth of intention) can be understood (though superficially) by many different people and may foster communication. They are a kind of argumentative platform, and only in preceding discussions is further differentiation possible. He thought that those vague and ambivalent formulations were important tools for practical-ethical issues, since by providing an argumentative platform, they foster collaboration and problem solutions. From a historical point of view, it is remarkable that Naess took up ideas that Neurath also had in mind. Brunswik and Naess joined his collaborative program of unified science (though after the Second World War, Naess dropped out). From a methodological point of view, Neurath’s congestions (Ballungen) played an important role in his program of unified science. These vague and ambivalent cluster concepts are exactly the same as formulations with a low level of intention (for details, see Radler 2015 and Rutte 2004: 364–65, and see Cartwright, Cat, Fleck, and Uebel 1996 for a detailed discussion of Ballungen’s historical background). Those congestions might lead to the process of collaborative research and new results. Despite all of their differences, Bühler, Brunswik, Neurath, and Naess are all philosophers who rejected any subjectivistic notion of Verstehen. Kraft took a different stance.22 For him, Verstehen is a subjective act of comprehension and reliving and reenacting, mainly for the reason that in “synthesis,” the axioms of a deductive explanation are constructed. For the

VICTOR KRAFT ON VERSTEHEN

177

other thinkers mentioned above, this is irrelevant. The mind is considered a black box, and only a “statement of sameness” has to be reported. (This, of course, presupposes intentions. See Radler 2015 and 2013 for details.) Kraft rejected such radical ideas. He was far away from reaching the current state of research, so (e.g., Detel 2014) Verstehen always had a subjective component for him. (His critical and short discussion of Gilbert Ryle’s The Concept of Mind illustrates this; for Kraft, Ryle’s position amounted to materialism [Kraft 1960: 281, cf. 263]). In Neurath’s program, there was no place for subjective notions, and there was no reference to the researcher’s personality. This amounted to Neurath’s famous and casual dictum that Verstehen or introspection might foster research, but no more than a good cup of coffee (Neurath 1931: 56).

7.5.  KRAFT ON VERSTEHEN: A SUMMARY According to Kraft’s analysis, Verstehen is connected with reliving and introspection. In his discussion of the Verstehen of historical persons, actions, and language, Kraft did not exclude this subjective moment as Neurath and others did. To be sure, Kraft’s statements are at best no more than sketchy remarks that represent the logical empiricists’ standard view on Verstehen. In addition, Kraft did not dwell on hermeneutics, as he was too attached to the deductive concept of scientific explanation. Therefore, he did not develop a fully-fledged theory of mind. Kraft alluded only to “synthesis,” the creative power of intuition, and its heuristic value. However, how thoughts are connected with the real world was not spelled out. Remarkably, by referring to “synthesis” and “intuition,” he opened philosophy to empirical science, thus diminishing the separation between questions of fact and questions of validity. However, nearly 100 years ago, the state of psychological research was still in its infancy. Because of this, his philosophical results became obsolete and outdated. However, Kraft probably would have endorsed the view that philosophy is open to progress.

NOTES 1 Schnädelbach (1999: 138) referred to the post-Hegelian period in Germany. After the breakdown of Hegel’s all-embracing philosophical system, philosophy had to face its historicization. Therefore, secular hermeneutics provided the starting point of a long-lasting historical development. 2 Neurath was also an exception and familiar with Dilthey’s ideas (Uebel 2019: 925). Thanks to Adam Tuboly for providing me with this information. 3 The first edition was published in 1969.

178

The History of Understanding in Analytic Philosophy

4 To be clear, other members of the Vienna Circle adhered to similar positions as Kraft; for example, those of early Moritz Schlick and Herbert Feigl. 5 I am not able to discuss this topic here but would like to refer to Radler (2010) and to Siegetsleitner (2014: 332–86). 6 “Pragmatic” because he considers the consequences of his stipulation. If the result is not as it should be, the stipulation has to be modified. 7 Through the assumption of a general law, the practical syllogism is transferred into a special type of the Hemple-Oppenheim-schema (Stegmüller 1983: 492), a form of a deductive-nomological explanation. The Ducasse sentence can be formulated as follows: whenever one intends p and believes that q is causal necessary for p, she realizes q. Stegmüller’s reconstruction is about actions, but if a methodological decision is the basis of Kraft’s epistemology, it is reasonable to “naturalize” this further and to argue that a decision is an action as well. One might intend to posit a philosophical statement. 8 Here, Kraft missed the opportunity to spell out in detail what a critical assessment of sources implies. He did not discuss hermeneutical suggestions for text interpretation. 9 “Inductive” was Kraft’s wording. Nowadays, one would use the term “abductive,” thus highlighting the important epistemological function those constructed laws possess. 10 Davis’s use of the word “deduction” is quite different from the use in the philosophy of science. His aim was to provide an explanation of the how the actual landscape became existent. Thus, this involves general natural laws. By advising his students to “Look and think!” (Davis 1912: xvii), it became clear he had deductive reasoning in mind, not, for example, a syllogism like modus ponens. He spoke as a scientific practitioner and not as a logician. 11 It is not completely clear what “empirically given values” means. I understand this as a kind of mindset of humans which guides their research and can be discovered by empirical research. 12 It remains open whether this means logic in its proper sense or logical thinking, which is a cognitive human ability. 13 Above, I do not use Kraft’s term “inductive” but “abductive,” which seems to me more appropriate. 14 The paper had already been published in 1955 under the title “Geschichtsforschung als exakte Wissenschaft,” and in 1963 Kraft added a short afterword. 15 Clearly, the invention of a new theory implies the invariance of meaning of the old and new theory’s central terms. There is no semantic incommensurability. 16 As already mentioned, in contrast to Popper, Kraft was never a falsificationist. 17 “The critical method never has to interpret the facts of psychology and history as grounds for validity for the norms, but they are needed as objects of philosophical analysis and contemplation” (Windelband [1883] 1924: 129). 18 Anthropologists are able to understand people with a completely different worldview (and language system). See, for example, Everett’s popular Das glücklichste Volk (2010).

VICTOR KRAFT ON VERSTEHEN

179

19 Physicalism means—to put it simply—that observational sentences are formulated in the third person perspective, thus not containing any reference to one’s own subjective feelings; this holds even for self-knowledge (cf. Rutte 2004: 351). From the lines written above, it should be clear that Kraft did not adhere to such a position (cf. Kraft 1960: 275), which is what, for example, Otto Neurath is famous for. For the problem of the understanding of other persons, it is interesting to note that Neurath did not reject the Verstehen approach altogether. “[H]‌e accepted de-metaphysicalized Verstehen into the de facto methodology of social science that his unified science encompassed” (Uebel 2019: 929). Already Cartwright, Cat, Fleck, and Uebel (1996: 140–1) had shown that Neurath adhered to a basic form of intentional understanding, without which the intersubjectivity of science simply does not work. 20 Strictly speaking, that has nothing to do with the eureka! experience but with the understanding of verbal statements. 21 For a broad overview of the historical development of psychological research in Vienna, see Benetka (1995). 22 In a recent publication, Uebel (2019) emphasized that Neurath did not reject the Verstehen approach altogether. Important to note is that “Neurath’s complaint was rather that talk of empathy misrepresents because of the metaphysical allegiances of the term” (Uebel 2019: 921–2). For Neurath, “intentions and intentional states were basic descriptive building blocks” (Uebel 2019: 923) and thus played an important role in his later philosophy of science.

REFERENCES Abel, T. (1948), “The Operation Called Verstehen,” American Journal of Sociology, 54 (3): 211–18. Acham, K. (1974), Analytische Geschichtsphilosophie, Freiburg: Verlag Karl Alber. Anscombe, G. E. M. (2020), Absicht, Berlin: Suhrkamp. [Originally, Intention, 1957]. Bast, R. A. (1999), “Vorwort des Herausgebers,” in H. Rickert, Philosophische Aufsätze, vii–xi, Tübingen: Mohr Siebeck. Benetka, G. (1995), Psychologie in Wien. Sozial- und Theoriegeschichte des Wiener Psychologischen Instituts 1922–1938, Vienna: WUV-Universitätsverlag. Brunswik, E. (1934), Wahrnehmung und Gegenstandswelt. Grundlegung einer Psychologie vom Gegenstand her, Leipzig: Deuticke. Bühler, K. (1909), “Über das Sprachverständnis vom Standpunkt der Normalpsychologie aus,” in F. Schumann (ed.), Bericht über den III. Kongreß für experimentelle Psychologie (Frankfurt 1908), 94–130, Leipzig: Johann Ambrosius Barth. Bühler, K. ([1927] 1978), Die Krise der Psychologie, Frankfurt am Main: Ullstein. Bühler, K. (1922), Die geistige Entwicklung des Kindes, Jena: Fischer. Cartwright, N., J. Cat, L. Fleck, and T. Uebel (1996), Otto Neurath: Philosophy between Science and Politics, Cambridge: Cambridge University Press. Danneberg, L., and H.-H. Müller. (1984), “Wissenschaftstheorie, Hermeneutik, Literaturwissenschaft. Anmerkungen zu einer unterbliebenen und Beiträge zu einem künftigen Dialog über die Methodologie des Verstehens,” Deutsche

180

The History of Understanding in Analytic Philosophy

Vierteljahresschrift für Literaturwissenschaft und Geistesgeschichte, 58 (2): 177–237. Davis, W. M. (1912), Die erklärende Beschreibung der Landformen, Leipzig: Teubner. Detel, W. (2011), Geist und Verstehen, Frankfurt am Main: Klostermann. Detel, W. (2014), Kognition, Parsen und rationale Erklärung, Frankfurt am Main: Klostermann. Everett, D. (2010), Das glücklichste Volk, Munich: DVA. Friedman, M. (2000), A Parting of the Ways, Chicago: Open Court. Gomperz, H. (1929), Sinn und Sinngebilde, Tübingen: Mohr. Kraft, V. (1912), Weltbegriff und Erkenntnisbegriff, Leipzig: Johann Ambrosius Barth. Kraft, V. (1914), “Gegenstand, Aufgaben und Methoden der Geographie als Wissenschaft,” in Handbuch der geographischen Wissenschaften, Part 1, ed. O. Kende, 1–8, Berlin: Vossische Buchhandlung. Kraft, V. (1925), Die Grundformen der wissenschaftlichen Methoden, Vienna: HölderPichler-Tempsky A.G. Kraft, V. (1929), “Intuitives Verstehen in der Geschichtswissenschaft,” in Mitteilungen des Österreichischen Instituts für Geschichtsforschung, Ergänzungsband XI, ed. W. Bauer, 1–30, Innsbruck: Universitäts-Verlag Wagner. Kraft, V. (1960), Erkenntnislehre, Wien: Springer. Kraft, V. (1963), “Geschichtsforschung als strenge Wissenschaft,” in Logik der Sozialwissenschaften, ed. E. Topitsch, 72–82, Cologne: Kiepenheuer & Witsch. Kraft, V. (1973a), “Konstruktiver Realismus,” Zeitschrift für allgemeine Wissenschaftstheorie, 4: 313–22. Kraft, V. (1973b), Die Grundformen der wissenschaftlichen Methoden, 2nd, completely revised ed., Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Kuby, D. (2020), “Decision-based Epistemology: Sketching a Systematic Framework of Feyerabend’s Metaphilosophy,” Synthese, https://doi.org/10.1007/ s11229-020-02934-3. Messer, A. (1923), “Rickert und der kritische Realismus,” Kant-Studien, 28: 364–75. Naess, A. (1953), Interpretation and Preciseness, Oslo: Jacob Dybwad. Neurath, O. (1931), Empirische Soziologie, Vienna: Springer. Ostermeier, J. F. (1987), De opvattingen van Alfred Hettner (1859–1941) over de plaats van de geografie in het system van de wetenschappen. Een bijdrag tot zijn intellectuele biografie, PhD thesis, Katholieke Universiteit te Nijmegen. Radler, I. (2008), “Victor Kraft und die Geographie,” in M. Fürst, W. Gombocz, and C. Hiebaum (eds.), Analysen, Argumente, Ansätze. Beiträge zum 8. Internationalen Kongress der Österreichischen Gesellschaft für Philosophie in Graz, Vol. 2 , 65–75, Franfurt am Main: Ontos. Radler, J. (2006), Victor Krafts konstruktiver Empirismus, Berlin: Logos. Radler, J. (2010), ‘Abduktive Argumentationsformen in Krafts Moraltheorie,” in Logischer Empirismus, Werte und Moral, ed. A. Sigetsleitner, 157–77, Vienna: Springer. Radler, J. (2013), ‘Neurath’s Congestations, Depth of Intention and Precization: Arne Naess and His Viennese Heritage,” HOPOS: The Journal of the International Society for the History of Philosophy of Science, 3 (1), 59–90. Radler, J. (2015), ‘Bringing the Environment in: Early Central European Contributions to an Ecologically Oriented Psychology of Perception,” History of Psychology, 18 (4), 401–13.

VICTOR KRAFT ON VERSTEHEN

181

Rutte, H. (2004), ‘Moritz Schlick und Otto Neurath – die intellektuelle Spannweite des Wiener Kreises,” in Philosophie und Religion: Erleben, Wissen, Erkennen: Vol. 6.1 Geschichte der österreichischen Humanwissenschaften, ed. K. Acham, 335–83, Vienna: Passagen Verlag. Schnädelbach, H. (1999), Philosophie in Deutschland 1831–1933, 6th ed., Frankfurt am Main: Suhrkamp. Schurz, G. (2008), ‘Patterns of Abduction,” Synthese, 164: 201–34. Siegetsleitner, A. (2014), Ethik und Moral im Wiener Kreis, Vienna: Böhlau. Staiti, A. (2018), “Heinrich Rickert,” in Stanford Encyclopedia of Philosophy, ed. E. N. Zalta. https://plato.stanford.edu/entries/heinrich-rickert/. Stegmüller, W. (1983), Probleme und Resultate der Wissenschaftstheorie und Analytischen Philosophie: Vol. I, Erklärung – Begründung – Kausalität, Studienausgabe, Teil C, Berlin: Springer. Topitsch, E. (1960), “Zum Geleit,” in Probleme der Wissenschaftstheorie, iii–vi, Vienna: Springer. Topitsch, E. (2005), Überprüfbarkeit und Beliebigkeit, ed. K. Acham, Vienna: Böhlau. Uebel, T. (2019), ‘Neurath on Verstehen,” European Journal of Philosophy, 27: 912–38. von Wright, G. H. ([1971]2004), Explanation and Understanding, Ithaca: Cornell University Press. Windelband, W. ([1883]1924), “Kritische oder genetische Methode?” in Präludien: Aufsätze und Reden zur Philosophie und ihrer Geschichte, Vol. 2, 99– 135, Tübingen: Mohr.

182

PART THREE

After Logical Empiricism

184

CHAPTER EIGHT

The Leopard Does Not Change Its Spots: Naturalism and the Argument against Methodological Pluralism in the Sciences JONAS AHLSKOG AND GIUSEPPINA D’ORO

Vulpes pilum mutat, non mores

8.1.  INTRODUCTION: THE RED HERRING OF THE MIND AS AN INNER THEATER The connection between meaning and inner inscrutable psychological items was immortalized by Locke, who claimed that a parrot does not speak a language stricto sensu, even though it can utter a word, because there is no idea in the parrot’s mind for which the word is the outward expression. Without the underlying idea in the mind the word remains a mere sound and does not qualify as a linguistic item. The parrot does not

186

The History of Understanding in Analytic Philosophy

mean anything when it utters a word. This close connection between the meaning of an utterance and the underlying psychological item underpins Locke’s well-known account of what it takes to understand others (cf. Locke [1690] 1964: Book III). On Locke’s account, when a person speaks, he or she attaches a word (which, like a label, is public and observable) to an idea (which is inner and to that extent inscrutable). To understand the meaning of the speaker’s utterance, therefore, the listener must “translate” the public observable sign, the written or spoken word, back into an idea in their own mind, thereby recovering the speaker’s original meaning. In chapter 2 of Word and Object, Quine ([1960] 2001) puts forward a positive proposal for how to “understand” others that dispenses with appeal to mental entities. The task of understanding others, for Quine, can (and should) be undertaken without invoking the mental notions which, on the Lockean picture of the relation between thought and language, are what allegedly bestow meaning onto the speaker’s linguistic behavior. Quine ([1960] 2001: 29) considers how a field linguist who only has access to the publicly observable linguistic behavior of “natives,” without any prior familiarity with their language, can translate sentences from the natives’ language into her own. Translation from alien languages, Quine argues, is “radical,” in the sense that the linguist can only appeal to observable behavioral aspects. What enables the linguist to render a term such as “gavagai” as “rabbit” is the native speaker’s behavioral disposition to assent to the term “gavagai” when in the presence of a rabbit, rather than some supposed access to the concept or idea which, in the Lockean view, the word signifies. Radical translation, to be clear, is not just the method of the field linguist; it is a proposal for an alternative science of understanding which dispenses with the notion of meaning as no more than a convenient way of speaking (cf. Kirk 2004: 152). Quine’s argument for radical translation is therefore often seen as providing indirect support for his earlier (and more direct) attack on intensional notions in “Two Dogmas of Empiricism” (Quine 1951; for a discussion of this, see Kemp 2006: 35). In the last half of the twentieth century, Quine’s proposal for how to articulate a science of understanding which requires nothing beyond a purely extensional context of explanation has become something of an orthodoxy. What seems to have made Quine’s account of radical translation compelling to many is the assumption that—given the close connections between meaning and mental items as depicted in Locke’s account of language—rejecting the view that to understand others requires recovering inner, inscrutable psychological items requires abandoning the notion of meaning itself. In other words, if one rejects the Lockean picture of how one understands others (the speaker has an idea, which she translates into a

The Leopard Does Not Change its Spots

187

word; the listener hears the word and translates the word back into an idea in her own mind), one must also reject the Lockean claim that there is more to a linguistic utterance than mere linguistic behavior. In this chapter, we wish to challenge the view that to provide an account of how to understand others which dispenses with inscrutable inner mental items one must extrude all intensional notions as inherently suspicious and give up altogether on the notion of meaning. We argue that it is based on a red herring which has (quite effectively) diverted attention away from the genuine issues at stake between those who defend the autonomy and possibility of humanistic explanations and those who endorse a form of scientism which denies that such explanations are sui generis. The red herring is that in order to develop a humanistic account of how we understand others that does not fall prey to what Quine (1969: 27) refers to as the Lockean “myth of the museum in which the exhibits are meanings and the words are labels,” and is earlier referred to by Ryle ([1949] 1990: 17) as the “myth of the ghost in the machine,” it is necessary to endorse Quine’s positive proposal for a science of understanding—the view that all translation is radical—and its corollary— that there are no fixed meanings (the indeterminacy of translation). We argue that the bone of contention between those who defend the autonomy of humanistic explanations and those who endorse a form of scientism which denies such explanations are sui generis is not that the former commit to the existence (and accessibility) of inner mental items that are inscrutable from a purely behavioral perspective, while the latter reject them. What is at stake, rather, is whether explanations which invoke normative considerations are different in kind from causal explanations. Once the criticism that the defense of a distinctive form of humanistic understanding necessarily rests on a commitment to suspicious mental entities is exposed as a red herring, the motivation for endorsing Quine’s positive proposal concerning how to understand others (radical translation) and its corollary (indeterminacy of translation) is substantively weakened. We trace the history of this red herring to the way in which an earlier attempt to defend the possibility of distinctive humanistic explanations—viz., Collingwood’s—was dismissed by falsely accusing it of relying on the early modern conception of the mind as an “inner theater.” Then we consider how this objection raised its head again and was mobilized by the Quinean naturalist against Peter Winch’s defense of humanistic explanations. Section 8.2 introduces Collingwood’s argument for a humanistically oriented historiography, an argument that was redeployed by W. H. Dray to rebut Hempel’s neopositivist defense of the unity of science. We defend this argument against the epistemological objection that humanistic explanations, as depicted by Collingwood and Dray, ascribe historians

188

The History of Understanding in Analytic Philosophy

implausible telepathic powers of access to the minds of historical agents and their contents. We argue that, as Dray makes clear, Collingwood’s defense of a humanistically oriented historiography neither assumes that historians have special epistemological powers of access to the minds of historical agents, nor rests on a commitment to ontologically suspicious entities; it consists rather in the claim that explanations which invoke normative considerations are different in kind from causal explanations. Section 8.3 considers a similar objection raised by the Quinean naturalist against Peter Winch’s argument for methodological diversity in the sciences. According to Quinean naturalists, Winch’s philosophy subscribes to an ontology of stable and determinate mental entities—designated as “the meaning” of human actions—which serve as the objects of understanding in social science research. Since naturalism explicitly denies the existence or knowability of such mental entities, naturalists dismiss Winch’s philosophy of the social sciences as a latter-day idealist ontology about ghosts in the machine. We claim, once again, that this objection draws attention away from the true bone of contention; that is, whether or not the logical form of humanistic explanation is irreducible to that of causal explanation. Tracing the history of this particular red herring back to the methodological debate for and against methodological unity in the sciences as it was played out between Hempel and Dray shows that naturalism, like the leopard, has not changed its spots: the same straw man objection is being remobilized here to motivate support for a solution to what is actually a false problem. Naturalism, of course, comes in many shades of grey that are not examined in this chapter. It could therefore be argued that we are making our lives considerably easier by focusing on the renewal of the argument for methodological unity as it was articulated by naturalism’s chief villain rather than focusing on softer forms of naturalism, such as those found in Donald Davidson’s anomalous monism and John McDowell’s liberal naturalism. They, unlike Quine, do not pursue an eliminativist agenda, but rather a reconciliatory one. Although it cannot be argued here, such attempts to do justice to the normative sui generis character of the mental within a naturalistic framework cannot succeed; such reconciliatory efforts are doomed to fail just like the attempt to square the circle, but to explain why this is the case would take a long excursus into deeply buried metaphilosophical assumptions that cannot be fully explored here (see D’Oro 2012 and D’Oro, Giladi, and Papazoglou 2019). We focus on Quine because instead of pursuing a reconciliatory agenda, he seeks to undercut the need to do justice to the distinctive character of normative/rationalizing explanations by dispensing with the very entities (inner mental items) which, he believes, generate the need for such sui generis explanations in the first instance. We argue that

The Leopard Does Not Change its Spots

189

the need to acknowledge the sui generis nature of rationalizing explanations was never born out of a commitment to the early modern conception of the mind as an inner theater and that Quine’s attempt to undercut the need for sui generis rationalizing explanations by eliminating the “beetle” in the mind is directed at a straw man. Irreducibly normative/rationalizing explanations are needed not to acknowledge the existence of the beetle, but to do justice to the subtleties of language, to capture the distinction between “actions” understood as bodily movements (and thus as mere species of events which can be explained causally) and “actions” understood as responses to norms (i.e., as a sui generis concept, which do not stand to the concept of “event” as a species to its genus in the way in which, e.g., the species “cow” stands to the genus “mammal”). Abolishing the distinction between these two senses of the term “action” is to fail to see the difference between abiding by a norm and following a (natural) law.

8.2.  THE HEMPELIAN CHALLENGE The possibility of a distinctive type of humanistic explanation that differs in kind from explanations in the natural sciences was a hotly disputed issue in mid-twentieth-century philosophy. The debate for and against methodological unity in the sciences was reignited by the publication of Hempel’s influential 1942 paper “The Function of General Laws in History,” where he argues that humanistic explanations are not different in kind from scientific explanations, because, appearances notwithstanding, they share the same logical form or structure. The battle for and against methodological unity in the sciences was by and large fought out on the turf of the philosophy of history, not least because the example used by Hempel to undermine the irreducibility of humanistic explanations was lifted from the pages of a historical text. Hempel’s famous example is that of dust bowl farmers who were said to have migrated because of deteriorating living conditions. Hempel’s claim is that such a historical explanation is not different in kind from the nomological explanations used in the natural sciences, because it covertly relies on the general law that populations will tend to migrate when living conditions deteriorate. Historical explanations are incomplete nomological explanations or “explanation sketches” that when filled in or duly completed, are revealed to be nomological explanations which are not different in kind from explanations used in the natural sciences to account for natural events such as volcanic explosions or solar eclipses. Hempel’s challenge is picked up by W. H. Dray (1957, 1958, 1963, 1964, 1980), who mobilizes Collingwood’s argument for the autonomy of historical explanation against Hempel’s claim for methodological unity. Historical

190

The History of Understanding in Analytic Philosophy

explanations, Dray argues, are not incomplete nomological explanations, as Hempel states, but complete explanations of a different kind. In a nomological explanation, the explanandum is the conclusion of an inductive argument. The fact of water freezing, for example, is explained by pointing out that this is what normally happens when the temperature drops below 0°C. The explanandum is inferred from a general law (if the temperature drops below 0°C, water freezes) and certain antecedent conditions (the dropping of the temperature). In historical explanations, by contrast, the explanandum is the conclusion of a practical argument, not an inductive one. When explained historically, the migration of the dust bowl farmers is presented as the result of a reasoning process in which the conclusion (the migration) is the logical or rational thing to do. Just as in Aristotle’s practical syllogism, the conclusion that eating chicken is good is rationally entailed by the premises that chicken is light meat and that eating light meat is good for you, here the migration of the dust bowl farmers is presented (in the context of a historical explanation) as the reasonable conclusion of the argument that it is better to migrate where living conditions are more favorable to the sustenance of life. Humanistic explanations differ from scientific explanations not because they are incomplete nomological explanations or explanation sketches, as Hempel argues, but because they are complete explanations of a different kind; they are normative explanations which account for what agents do by invoking what they ought to do if they act reasonably. Dray’s criticism of Hempel exposes the claim for methodological unity in the sciences as resting on a failure to distinguish between what Collingwood takes to be the categories of actions and events (cf. Collingwood [1946] 1993: Part V, Epilegomena). Actions (which he takes to be the distinctive subject matter of history) are the correlative of normative/rationalizing explanations; humanistic explanations have a distinctive explanandum because they have a sui generis method that differs from the nomological explanations in natural science, where the explanandum is Events (including human bodily movements understood as mere behavior). From Collingwood’s perspective, the claim for methodological unity in the sciences is based on a category mistake (the conflation of the category of action and that of event) which arises when one fails to discern the different senses that the term cause/because possesses in different explanatory contexts; cf. Collingwood [1940] 1998: 285–327). The Hempelian challenge to the claim that there is a sui generis form of action explanation does not deny the legitimacy of notions such as beliefs and desires; it rather co-opts them in the context of nomological explanations: just as the dropping of the temperature, together with the general law that water freezes at 0°C, can be used to explain why water in a

The Leopard Does Not Change its Spots

191

bucket froze overnight, so (for Hempel) beliefs and desires can be deployed in the context of nomological explanations to explain, for example, why the dust bowl farmer migrated when living conditions deteriorated. This attempt to co-opt propositional attitudes such as beliefs and desires in the context of nomological explanations is however premised on the mistaken assumption that the relation between the explanans and the explanandum in an action explanation is an empirical connection that is established by observation and inductive generalization, rather than a conceptual connection. But reasons for actions, for Collingwood and Dray, are not Humean causes that are temporally prior to, and logically independent of, their effect: how one rationalizes the action determines what kind of action it is. Whether a person opening a window is “letting air in” or “letting a fly out” is not logically independent of how the action is rationalized. Because action explanations are rationalizations, they resist assimilation in a naturalistic picture in which they are presented as essentially types of events which are caused by antecedent conditions of a particular (internal) kind. As a contemporary causalist such as Albert Mele (2000: 279) would put it, “our actions are, essentially, events … that are suitably caused by mental items, or neural realizations of those items.” From the perspective of Collingwood and Dray, actions are not events, because they are the correlative of different kinds of explanations; the claim that actions are events, or a species of events (events which are the effects of internal rather than external causes) rests on a conflation of two distinct meanings of the term “because.”1 Hempel’s challenge is therefore met (and undermined) by the consideration that rationalizing explanations (of actions) are not causal/nomological explanations (of events), because the former establish conceptual rather than empirical connections and have a normative dimension that the latter lack: even when they are used predictively, rather than retrospectively, rational explanations do not foresee future behavior by subsuming it under a general psychological law, but anticipate how one would act if one acted rationally; that is, as one ought to act in the circumstances. The historical understanding of action is therefore a matter of rendering past behavior intelligible by ascribing agents a practical argument that rationalizes (in an in instrumental rather than a moral sense) the action. Rationalization, to be clear, is not a historical method only insofar as it applies to the understanding of past agents. One might say, in an exact reversal of Quine’s claim that radical translation applies not only to the field linguist but also to the translator of living languages, that rationalization begins at home: it is in the same way that we understand our contemporaries, and indeed ourselves, insofar as we explain what is done qua action (cf. Collingwood [1946] 1993: 219). The divide between humanistic and historical explanations is therefore not

192

The History of Understanding in Analytic Philosophy

a divide between the future-directed explanations of science and the pastdirected explanations of history, but between radically different ways of homing in on the explanandum (cf. D’Oro 2018). One important implication of the distinction between rationalizing explanations and causal explanations as defended by Collingwood and Dray is that actions, unlike events, are identified by invoking the thoughts which they express. Insofar as the description of an action requires appeal to a thought process or rationalization, understanding action requires going beyond a purely extensional context of explanation, or the description of mere bodily movements. For, after all, the same bodily movements could be the expression of different rationalizations, as in Anscombe’s (1957: 41) example of the man who operates a water pump: is he replenishing the water supply or poisoning the inhabitants of the house? It is this claim—viz., the claim that the understanding of action requires going beyond a purely extensional context of explanation to expose the thought of which the action is the expression—that gives rise to the criticism that Collingwood’s defense of the autonomy of historical explanation rests on the dubious ontological commitment to the existence of hidden mental processes (the kind that Quine is keen to eliminate) and the equally suspicious epistemological claim that historians enjoy special epistemic powers of insight. Collingwood came to be seen as the target of Ryle’s criticism of the myth of the ghost in the machine, and Collingwood’s doctrine of reenactment, which claims that to understand an action historically is to rethink the thoughts of historical agents, was mocked for bestowing on historians unlikely telepathic powers of access to other minds (cf. Gardiner 1952a, 1952b: 213). Objections of this nature may have been prompted by Collingwood’s unguarded use of language—by the claim, for example, that actions have an inside which events lack (Collingwood [1946] 1993: 213). But one only needs to scratch beneath the veneer of what he writes to reveal that talk of an “inside” and an “outside” is a highly metaphorical way of expressing the claim that we are dealing here with different kinds of explanations, with a different logical form in each case, which suit the distinctive explanatory goals of scientific and humanistic explanations. The historical meaning or significance of an action, for Collingwood, is not a psychological process of which the action is an outward manifestation in the way in which, for Locke, a word is an observable sign that stands for an idea in the mind that is inscrutable from a third-person perspective. The thought context which historians should take into account are not agents’ inner states. They are the epistemic, moral, legal, aesthetic norms which form the background of their actions. Just as understanding a character in a period novel requires suspending the rules that govern the historian’s own life and imaginatively reconstructing what

The Leopard Does Not Change its Spots

193

it means to live under a different set of expectations and demands, so too understanding agents in history is seeing how they responded to the rules and expectations of their world. In other words, it is to enter into what Wittgenstein calls their “forms of life.” As Collingwood puts it: The Historian, investigating any event in the past, makes a distinction between what may be called the outside and the inside of an event. By the outside of the event I mean everything belonging to it which can be described in terms of bodies and their movements: the passage of Caesar, accompanied by certain men, across a river called the Rubicon at one date, or the spilling of his blood on the floor of the senate house at another. By the inside of the event I mean that in it which can only be described in terms of thought: Caesar’s defiance of Republican law, or the clash of constitutional policy between himself and his assassins. The historian is never concerned with either of these to the exclusion of the other. He is investigating not mere events (where by a mere event I mean one which has only an outside and no inside) but actions, and an action is the unity of the outside and inside of the event. (Collingwood [1946] 1993: 213) As this passage makes (abundantly) clear, the meaning of Caesar’s action, which is what historical investigation aims to uncover, is not a thought inside Caesar’s head, but what that action signifies in relation to the norms which governed the Roman Republic. To uncover what Caesar’s action means requires understanding republican law, just as to understand the meaning of a word is to understand the grammar of the language to which it belongs, not trying to enter the speaker’s head. To distinguish the way in which explanations in the human and the natural sciences operate and the nature of their explanandum, there is no need to commit either to the ontological claim that there are mental processes or to the epistemological claim that the historian has a hotline to these thoughts. On Collingwood’s account of Caesar’s action, the crossing of the Rubicon has a determinate meaning which is defined by the laws of the republic along with what counts as an infringement of them, just as the gesture of raising one’s hand at an academic conference, in this specific context, signifies the request to ask a question. Collingwood would have accepted that to understand the meaning of Caesar’s action, one needs insider knowledge of the rules and regulations that applied to Roman generals under the republic, but he would have denied that to have such “insider” knowledge, one needs somehow to peek into the heads of Romans or indeed assume that the meaning of Caesar’s action is captured by an internal monologue that he recited to himself while crossing the Rubicon. Meaning or significance, so

194

The History of Understanding in Analytic Philosophy

understood, is not located in the head: to grasp the significance of Caesar’s action, one needs “insider” knowledge of the Roman world, not knowledge of what goes on “inside” Caesar’s head, just as to understand the meaning of raising one’s hand at an academic conference requires understanding the rules of engagement that govern that context.2 Quine’s argument for the indeterminacy of translation arguably rests on an equivocation of these two different meanings of “inside.” He denies the radical translator (whose task is to render a word into the home language without any prior knowledge of the target language) “insider knowledge” in the sense of prior familiarity with the norms (linguistic and otherwise) that govern the agent’s behavior on the grounds that presupposing familiarity with the cultural context of the native speaker would be tantamount to being able to “see” “inside” the head of the speaker, just as a bilingual speaker could “see” the idea that stands behind the words of each language.3 As we see in the next section, this objection has been raised by the Quineaninspired naturalist against Peter Winch’s defense of sui generis social science explanations. But once again, the objection that the notion of meaning is inevitably bound up with an earlier conception of the mind as an inner theater misses the target: radical translation is an implausible solution to an entirely fabricated problem.

8.3.  THE QUINEAN CHALLENGE Along with Ludwig Wittgenstein, Collingwood is the other main source of inspiration for Peter Winch’s philosophy of the social sciences. Winch’s seminal work, The Idea of a Social Science (ISS; Winch [1958] 1990), makes frequent reference to Collingwood, and the very title of the book consciously alludes to The Idea of History. However, while Winch often explicitly acknowledges the Wittgensteinian influence on his work, the inspiration he derives from Collingwood remains far less explicit. One reason for this may be that Winch himself is not fully aware of all the fundamental similarities between his own position and Collingwood’s. In addition, Winch is not always the most astute reader of Collingwood, and in ISS he repeats the common misinterpretation of reenactment as psychological Einfühlung rather than seeing it, as Collingwood would have insisted, as a critical reconstruction of past reasoning along with the historically contingent conditions of its meaningfulness. Given these circumstances, it is only quite recently that scholars have articulated fundamental connections between Collingwood’s and Winch’s work, pointing to ways in which Winch’s philosophy of the social science can be read as an implicit development of themes found in Collingwood’s

The Leopard Does Not Change its Spots

195

philosophy of history (Ahlskog and Lagerspetz, forthcoming). Collingwood’s and Winch’s seminal books on “the idea” share views on what the human sciences are about. Both thinkers argue that the human sciences involves forms of understanding and explanation that are logically distinct from those employed in the natural sciences. In the following section, we present the challenge to Winch’s philosophy of the social sciences that has been mounted from the point of view of Quinean naturalism, a challenge that, if valid, applies ipso facto to Collingwood’s defense of a humanistically oriented historiography. The naturalist critique is inseparably connected with an ontological reading of central arguments in Winch’s philosophy of the social sciences. In a key passage in ISS, Winch argues that there is, in humanistic forms of explanation and understanding, a logical priority for the participant’s unreflective understanding of social phenomena. Winch’s crucial argument is that the description of a social or intellectual phenomenon cannot even get started unless the researcher can identify what belongs to the phenomenon (and thus counts as “the same thing”) according to classifications that the participants in the target society themselves would potentially use— specifically identified by Winch as “the unreflective understanding” of the participants. In that sense, the social scientist is not like an engineer investigating physical processes, but more like an apprentice engineer who wants to understand the activities of his colleagues in terms of their concepts, reasons, and norms of belief (Winch [1958] 1990: 88). However, Winch is keen to point out that this argument is about logical requirements for understanding social phenomena and not an ad hoc stipulation about the supposed impossibility of critical engagement with the participant’s own understanding; the latter interpretation has been the source of many misrepresentations of Winch’s philosophy (cf. Ahlskog and Lagerspetz 2015). Winch writes: I do not wish to maintain that we must stop at the unreflective kind of understanding of which I gave as an instance the [apprentice] engineer’s understanding of the activities of his colleagues. But I do want to say that any more reflective understanding must necessarily presuppose, if it is to count as genuine understanding at all, the participant’s unreflective understanding. … [A]‌lthough the reflective student of society, or of a particular mode of social life, may find it necessary to use concepts which are not taken from the forms of activity which he is investigating, … still these technical concepts of his will imply a previous understanding of those other concepts which belong to the activities under investigation. (Winch [1958] 1990: 89)

196

The History of Understanding in Analytic Philosophy

The decisive question raised by this quote is what Winch’s notion of the participants “unreflective understanding” means. Throughout ISS, Winch explicates this notion as the ideas, concepts, and rules that belong to the social phenomena themselves, and he argues that it is by viewing the behavior of the agents in relation to such ideas and concepts that the social scientist makes sense of actions by historical agents. Winch emphasizes that ideas and concepts are not something apart from the action itself and indeed advances something similar to Collingwood’s argument for the internal connection between action and thought. His point is that actions are what they are only in relation to the concepts and ideas embodied in the social relations and institutions of the target society itself. Without an understanding of the relevant ideas and concepts shared by the agents of the target society, one would not be able to explain how the actions and events in question follow from the motives and reasons of the agents. The reason for this, of course, is that it is only in relation to ideas and concepts—expressed in practices and ways of thinking—that actions appear as reasonable responses to some situation the agent is facing. Winch’s main argument for showing how the meaning of actions derives from ideas and concepts is based on Wittgenstein’s idea of rule-following. Winch argues that actions are cases of rule-following and that this feature distinguishes the category of action from brute biological or psychological dispositions and habits. According to Winch ([1958] 1990: 52), “all behaviour which is meaningful is ipso facto rule-governed.” Consequently, the social scientist who endeavors to understand social phenomena from the perspective of the agents must do so in terms of the historically specific rules and norms in relation to which reasons and motives carry normative force. According to the naturalist critique, Winch’s notion of rules stipulates an ontology of Platonic meanings that constitutes and supplies the social world with a rigid and univocal structure (cf. Bohman 1993; Pleasants 1999, 2000; Pettit 2000; Roth 1987, 2003, 2011; Turner and Roth 2003; Turner 1994). As Mark Theunissen (2020: 259) has recently summarized this interpretation of Winch: “According to this reading, ISS argues that the social world can only have the necessary stability if built on a web of shared rules independent of any one individual, which magically constrains each one of us from destabilizing the social reality we are members of.” According to Paul Roth (1987: 134), Winch’s philosophy “hypostasizes social rules and talks of them … as if they were an independent object of study.” This reading is found also among scholars sympathetic to Winch. For instance, Nigel Pleasants (1999, 2000) writes that Winch ontologizes Wittgenstein, while Philip Pettit (2000) complains that Winch’s ontology of rules imposes overly rigid constraints on the freedom of agency.

The Leopard Does Not Change its Spots

197

There are also those who view Winch’s supposed ontology in positive terms. Alice Crary (2018: 31) construes Winch as proposing “a distinctive social ontology. At issue is an ontology on which objective features of the social world are irreducibly ethical.” Winch, on Crary’s (2018: 31) reading, tells us that values are just as real, or almost as real, as physical objects are— for Winch’s account “is giving us an image of a region of objective reality as an intrinsically ethical realm.” Still, Winch’s supposed ontology of rules is more often read as a misguided endorsement of both cultural relativism and social conservatism (Roth 2003, 2011; Turner 1994; Turner and Roth 2003). But it is equally common, especially among naturalists, to construe Winch’s supposed ontology as dependent on a hypothesis about mental entities. For example, Roth (1987: 138) derides, with implicit reference to Winch, philosophers of social science “who believe that there exist conceptual models lurking in mental space awaiting discovery.” The main line of attack in the naturalist critique is to contend that Winch’s ontology steers social science in the wrong direction. For if translation of meaning is, as Quine argues, essentially indeterminate, then the task of discovering supposedly univocal meaning entities turns out to be an entirely misguided endeavor. According to Roth, this dismissal follows directly from Quine’s thesis about the indeterminacy of translation and the uprooting of the analytic/synthetic distinction on which it is based. As Roth writes (1987: 139), the key issue is that Quine’s thesis undermines the entire idea of believing that there exists a “mental model that a propitious translation might mirror.” The point is, as Quine is supposed to have shown, that in contrast to the “objectively determinable stimulus conditions for observation sentences,” there can be no “similar warrant for an assumption about a shared semantic model” (Roth 1987: 143). There are, Roth claims, no observable meaning facts parallel to natural facts. Quine’s thesis shows, therefore, that “there is no behavioral fact of the matter [to separate between impositions and discoveries;] moreover … there are no introspective facts of the matter to settle the question either” (Roth 1987: 234). On the basis of such considerations, Roth (1987: 143–4) concludes that Quine’s thesis directly undercuts what he takes to be Winch’s central proposal for the social sciences: “If we have no reason for assuming that there exists some unique set of rules by which individuals jointly make sense of their social environment, then it can hardly be maintained that the sole purpose of social analysis is to discover such rules.” There are two distinctly different ways in which one could defend Winch against the naturalist critique. One alternative is to tone down the rigidity of Winch’s supposed ontological claims about the social world. Such a defense can be mounted by stressing that Winch makes many qualifying remarks

198

The History of Understanding in Analytic Philosophy

in this connection. Social rules, for Winch, are neither static nor univocal building blocks of the human realm (cf. Winch [1958] 1990: § 1.8). Winch also shows the diversity of social rules, some of which are implicit and others explicit; some are rigid while others are more permissive (Winch [1958] 1990: §§ 2.2–2.5).4 Theunissen (2020: 265) uses this line of defense and writes that for Winch, “rules aren’t a homogeneous class of ontological entities that fix social life as ready for interpretation, rather they are varied, multifaceted, contingently or necessarily indeterminate, subject to change and thus as pluriform and dynamic as we expect social life to be.” However, the problem with this line of defense is that it does not pull Winch firmly enough out of the ontological trench. Emphasizing the diversity of rules simply risks transforming Winch from a rigorous ontologist who subscribes to determinate Platonic meanings into an indecisive ontologist who believes that social rules are so varied that one cannot say much about them at all while, at the same time, lamenting that meaning entities are as complex as life itself. Furthermore, this kind of defense carries no weight against naturalism. For whether meaning entities are univocal or multifaceted makes no difference for the thesis of indeterminacy of translation—the argument is that no specifiable meaning can be discovered “behind” behavioral facts at all (cf. Roth 1987: 143).5 An alternative line of defense is to dispute that Winch was engaged with ontology in the first place, which means denying that ISS is in the business of determining what kinds of things generally exist in the world. In that case, one must explicate what kind of philosophical work Winch himself assigns to the much-debated concept of rule-following. It is important to note that Winch does not understand rule-following as a concept for making ontological claims, but rather as a concept used for delineating logical distinctions between forms of understanding and explanation in social and natural science, respectively. When Winch reflects on the main argument of ISS in his new introduction for the second edition of the book, he makes the following clarification: The discussion of the distinction between the natural and the social sciences in the book revolves round the concept of generality and the different ways in which this characterizes our understanding of natural and social phenomena respectively. I expressed this difference by saying that our understanding of natural phenomena is in terms of the notion of cause, while our understanding of social phenomena involves the categories of motives and reasons for actions. Furthermore, I argued, whereas the category of cause involves generality by way of empirical generalizations, that of a reason for action involves generality by way

The Leopard Does Not Change its Spots

199

of rules. And these notions—of generalization and of rule—differ from each other in important logical respects. (Winch [1958] 1990: xi, original emphases) In the paragraphs which follow these remarks, Winch furthermore clarifies that he is not merely making a verbal distinction here, as if he was legislating about the way we should use our words. Winch emphasizes that we can, and do, speak about actions as caused—in phrases such as “what was the cause of his doing that?”—which implies that causal terms do apply to human actions also. Nonetheless, Winch ([1958] 1990: xii) points out that this does not matter for the logical distinction that he is investigating: “it would be a great mistake to think that, in saying [that causal terms apply to action], we are saying anything substantial about the form of explanation and understanding of his behaviour that is in question.” For Winch, the concept of rule serves a logical function in the form of explanation that distinguishes social science from natural science. The social scientist does not start by hypostasizing that such and such social rules exist as ready-made meaning entities. Rather, rules, norms, and concepts are what social scientists must (logically) look for in order to render action intelligible by invoking motives and reasons on the part of the agents. For insofar as what counts as legitimate reasons and motives for action are not determined by universal human nature, the possibility of giving explanations in terms of reasons and motives will, necessarily, depend on understanding historically specific rules, norms, ideas, and concepts—expressed in practices and ways of thinking—by virtue of which reasons and motives have normative force. In fact, in a very similar way, for the natural scientist, using the concept of cause does not entail the supposition of causes as ready-made entities in the natural world. Rather, “causality” characterizes the form of explanation by way of which the natural scientist investigates empirical generalizations. Consequently, the question of whether social rules are univocal or multifaceted—or whether causes are simple or complex—is not one to be solved by ontological theory, but rather an empirical question that finds its answer in the actual work of the social and natural scientists, respectively. Again it is important to note that Winch stresses these points by quoting passages in which Wittgenstein speaks of “language-games” and “rules” not as entities independent of our forms of understanding, but as concepts of comparison through which we render experience intelligible (cf. Winch [1958] 1990: xiv). The naturalist critique of Winch rests on confusing his delineations of logical commitments, inherent in different forms of explanation, with ontological claims about what kinds of thing exist in the world. Importantly,

200

The History of Understanding in Analytic Philosophy

Winch explicitly rejects the assumption that naturalists believe to be the most problematic feature of his philosophy of the social science; namely, that explanation by way of social rules, norms, and concepts presupposes an idea of hidden mental entities. Winch ([1958] 1990: 119) writes: “It must be said very firmly here that the case for saying that the understanding of society is logically different from the understanding of nature does not rest on the hypothesis of an ‘inner sense.’ ” Furthermore, Winch’s own account of “social rules” is directly opposed to the idea that such rules are to be understood as “inner” mental, conceptual structures: “In fact it follows from my argument in Chapter II that the concepts in terms of which we understand our own mental processes and behaviour have to be learned, and must, therefore, be socially established, just as much as the concepts in terms of which we come to understand the behaviour of other people” (Winch [1958] 1990: 119, original emphases). Hence, the core mentalist presupposition which the naturalist critique ascribes to Winch is in fact explicitly undermined by the main arguments of ISS. However, does this mean that Winch is actually closer to naturalist accounts? For if Winch does not think that social rules are ontological entities residing in a hidden “mental space,” does this not mean that rules, norms, and concepts must—as several naturalists argue (cf. Roth 2011)— be construed as mere intellectual tools, devised by the social scientist and imposed for purposes of explanation? Winch addresses this question head on in his critique of Karl Popper’s methodological individualism. Popper’s position (1957) implies that there is no significant distinction to be made between, on the one hand, intentional descriptions of the historical agent’s action and, on the other hand, the theoretical models used by the researcher to explain the agent’s behavior. As Winch points out, Popper’s neglect of this distinction is closely related to his project of presenting the social scientist’s descriptions as analogous to the construction of theoretical models used in the natural sciences, an approach which is very similar to that of present-day naturalists (cf. Roth 1987; 2011; Turner and Roth 2003). Popper was fighting what he called “methodological essentialism,” which, he believed, would involve the assumption that behind the explanatory models of the scientist there would be some kind of “observable ghost or essence” (Popper, quoted in Winch [1958] 1990: 127). Popper claims therefore, again in a way that is reminiscent of present-day naturalists, that our knowledge of social action and events is no different from other kinds of empirical knowledge—for all such knowledge simply consists in the explanation of experiential data by way of the researcher’s imposition of scientific categories and theoretical models (cf. Roth 2012: 315, 2020: 48, 61).

The Leopard Does Not Change its Spots

201

One well-known consequence of this view is Popper’s description of social institutions as mere models that are applied for explanatory purposes. Winch ([1958] 1990: 127) calls that idea “palpably untrue.” He argues, instead, that the actions of historical agents embody concepts that belong to the perspectives of meaning inherent in the ways of life and thought of the agents themselves. In other words, the explananda of social science embody concepts. Winch argues, furthermore, that this is the chief feature distinguishing the social scientist’s relation to their subject matter from that of the natural scientist. As Winch writes: The ways of thinking embodied in institutions govern the way the members of the societies studied by the social scientist behave. The idea of war, for instance, which is one of Popper’s examples, was not simply invented by people who wanted to explain what happens when societies come into armed conflict. It is an idea which provides the criteria of what is appropriate in the behaviour of members of the conflicting societies. Because my country is at war there are certain things which I must and certain things which I must not do. My behaviour is governed, one could say, by my concept of myself as a member of a belligerent country. The concept of war belongs essentially to my behaviour. But the concept of gravity does not belong essentially to the behaviour of a falling apple in the same way: it belongs rather to the physicist’s explanation of the apple’s behaviour. To recognize this has, pace Popper, nothing to do with a belief in ghosts behind the phenomena. Further, it is impossible to go far in specifying the attitudes, expectations and relations of individuals without referring to concepts which enter into those attitudes. (Winch [1958] 1990: 127–8, original emphases) Again, Winch’s argument here is not ontological, but concerns a logical distinction between the different presuppositions for describing phenomena within social and natural sciences, respectively. As Winch emphasizes, the very possibility of describing human action as an intelligible response to a historically specific context is logically dependent on understanding the norms and concepts that belong to the agents’ understanding of their situation. Furthermore, as Winch’s point about the concept of war aims to show, the normative dimension is not something added to actions by social scientists in and through their explanations. Quite the contrary, the normative dimension is already part of the original context of the agents, which is clear from the fact that appealing to the normative dimension of “being at war” would also be essential if the agents themselves were to make sense of what they were doing. In addition, the concept of being at war would shape

202

The History of Understanding in Analytic Philosophy

explanations of their actions independently of the fact that individual agents may have cheered the war on, while others opposed the very idea of war as conscientious objectors. Thus rationalization is not a feature added by social scientists, but an aspect that logically belongs to every form of description of behavior as action—irrespective of whether such descriptions are made by historical agents or social scientists. The internal relations between ideas, social relations, and the identity of action is further explored in Winch’s critique of Max Weber’s “scientific” attempt to pry apart those very internal relations by describing social phenomena in purely extensional terms. Winch writes: Weber ceases to use the notions that would be appropriate to an interpretative understanding of the situation. Instead of speaking of the workers in his factory being paid and spending money, he speaks of their being handed pieces of metal, handing those pieces of metal to other people and receiving other objects from them; he does not speak of policemen protecting the workers’ property, but of “people with helmets” coming and giving back the workers the pieces of metal which other people have taken from them; and so on. In short, he adopts the external point of view and forgets to take account of the “subjectively intended sense” of the behaviour he is talking about: and this, I want to say, is a natural result of his attempt to divorce the social relations linking those workers from the ideas which their actions embody: ideas such as those of “money”, “property”, “police”, “buying and selling”, and so on. Their relations to each other exist only through those ideas and similarly those ideas exist only in their relations to each other. (Winch [1958] 1990: 117–18) Winch does not deny that Weber’s “externalization” may have some beneficial results, such as a Verfremdungseffekt that offers researchers a possibility for observing details that may have remained unnoticed due to their familiar and obvious status. However, what Winch stresses is that the benefits of externalized descriptions presuppose that the situation has already been rendered intelligible by way of interpretive understanding. The external standpoint is not a perspective that is, somehow, more real or true, but one that is parasitic on the social scientist’s ability to make sense of the agent’s behavior as action which is expressive of certain concepts and ideas. Winch’s point about logical requirements for describing social phenomena takes us to the heart of the naturalist challenge. For what Winch shows is that without interpretive understanding, there cannot even exist any relevant candidate descriptions to choose between for the social scientist’s investigation. That there are such candidate descriptions is something

The Leopard Does Not Change its Spots

203

naturalists take for granted. For instance, Roth argues that the translation of meaning is a process of empirical investigation, one that parallels the way theories are tested in the natural sciences (Roth 1987: 240). He concludes that deciding between competing translations is like a “scientific test” in which the researcher does not aim for the right answer: “the search, as always, is for the best—the most empirically adequate, simplest—account we can give at the time” (Roth 1987: 244). However, later on in the same section, Roth (1987: 244) concedes that our study may sometimes “[lead] us to unusual translations of what others are about.” The same unwarranted presupposition (that candidate descriptions are available on naturalist premises) is also betrayed by Quine himself. It is embedded in the very argument for the indeterminacy thesis when he speculatively writes about what the “likely” reference of the native speaker’s “gavagai” expression may be (Quine 1969: 34). If Quine’s empiricist approach is endorsed, with the result that all intensional terms are to be eschewed and only empirical descriptions of behavioral data are permitted, it follows that there can be no grounds for saying that one translation is more “unusual,” “likely,” or “better” than the other. The reason is that the very first attempt to make out what the native speaker’s expression means already requires that the expression in question is grasped from within the context of utterance to which it belongs. This contextualization requires the very understanding of rules, norms, and concepts that Winch argued for. It is in light of this consideration that Sören Stenlund (1990) detects an incoherence at the very heart of Quine’s influential thesis. Stenlund argues that if one takes Quine’s empiricism seriously, then speaking of an “indeterminacy of translation” is already saying too much, for such a position requires candidate descriptions for translation. What one ends up with is, rather, an “indeterminacy of forms of expressions” in which the social scientist is permitted to say nothing more than that the native speakers are uttering expressions that the social scientist cannot understand. As Stenlund writes: It is presupposed in Quine’s argument that the field linguist, engaged in radical translation, is able to identify phonetically the expressions of the foreign language. He is supposed to be able to recognize two different utterances of the same sentence as such. And this is an important point because it is on this basis that he is able to formulate hypotheses of translation and test them empirically. He must be able to identify the sentences in order to make hypotheses about their correct translation. But how can he even do that? By what criteria is he judging two different utterances as utterances of the same expression? That the linguist thinks

204

The History of Understanding in Analytic Philosophy

that two utterances sound alike may not be a good criterion if it is a radically foreign language. Can he determine the “phonetic norms” of the foreign language prior to the investigation of the meaning and the uses of expressions? Quine is obviously presupposing that he can. (Stenlund 1990: 94, original emphasis) The core of Stenlund’s argument is that Quine neglects the fact that a normative context is also needed for identifying the relevant expressions to be used in the field linguist’s empirical test. Phonetic norms that govern tone, pitch, rhythm, etc. make a great difference to the meaning of what is said, and without this kind of understanding “how is the linguist to know what features of two utterances are essential to them qua utterances of the same sentence?” (1990: 95). According to Stenlund, this stronger form of indeterminacy undermines the entire conceptual and methodological structure of Quine’s indeterminacy thesis. What Quine erroneously assumes is that norms governing expressions can be identified in purely naturalistic terms; that is, independently of an investigation of the uses of such expressions in actual contexts of utterance—uses of expressions that must be fleshed out in relation to concepts, rules, and norms of the target society. According to Stenlund, the fact that Quine assumes the possibility of a strictly naturalist identification of expressions shows that Quine assumes actual human language to have a formal structure, like a scientific theory. A structure that can be identified and specified in isolation from the forms of use of its expressions. However, actual language derives its meaning from its application in different contexts of utterance, which entails that understanding a language is not like understanding a scientific theory. As Stenlund (1990: 96) writes: “If a human language is conceived of as a system of rules for using expressions, then what constitutes following the rules correctly is determined ultimately within the forms of life where the language belongs” (original emphases). The naturalist alternative to Winch’s account of the social sciences fails for three related reasons. First, naturalists assume that the only way to account for the normative dimension of action explanation is to hypostasize a hidden mental structure behind the overt behavior of historical agents. Winch makes no such assumption. He argues instead that norms, rules, and concepts are not hidden ontological entities, but rather features of the forms of understanding that logically distinguish social science from natural science. Second, naturalists erroneously assume that the only legitimate way that normativity—in the form of rules and concepts—enters social science research is by way of imposition on the part of the social scientist. Against this idea, Winch argues that concepts and rules belong to the very

The Leopard Does Not Change its Spots

205

identification of behavior as action, which means that normativity is also intrinsic to the ways in which historical agents themselves understood what they were doing. Third, Winch points out that descriptions of social phenomena could not even get off the ground without understanding the background of rules, norms, and concepts in terms of which actions and expressions can be identified for what they are. Furthermore, as Stenlund’s elaboration of the argument shows, the naturalist fails to appreciate that actual human expressions are always already rooted in contexts of utterance that reflect the rules, norms, and concepts of the agents. This entails that there simply is no such thing as a purely naturalist identification of the sense or reference for an expression. Without invoking intensional terms, the naturalist would only be able to speculate about, say, the way in which the tone of voice or the way signs are jotted down on the paper show signs of biologically or psychologically disposed behavior patterns. In a trivial sense, purely extensional descriptions of human behavior are not impossible. However, such descriptions of social phenomena cannot (logically) communicate anything of interest to historians and social scientists. In conclusion, Winch is not putting forward an argument about what kinds of things exist in the world; on the contrary, he wants to outline the kinds of logical commitment that shape our descriptions and explanations of behavior and events qua actions and social phenomena.

8.4. CONCLUSION The Quinean challenge to the possibility of distinctive humanistic explanations is more radical than the earlier Hempelian argument for methodological unity in the sciences. Hempel’s challenge treats beliefs and desires as internal causes of external behavior, as antecedent conditions which lie in the mind rather than the external world. Quine, by contrast, dismisses them as belonging to the Lockean “myth of the museum.” He is not so much an epistemological skeptic about the possibility of recovering the meaning behind actions or words as he is a kind of nihilist (Kirk 2004: 153). These differences notwithstanding, the Hempelian and the Quinean challenges to the possibility of a sui generis humanistic science of interpretation both rest on a failure to grasp that humanistic explanations have a normative dimension that is irreducible to causal explanations. In both cases, attention is diverted away from the genuine issues at stake by the accusation that any attempt to defend the possibility of humanistic explanations cannot escape a commitment to the early modern conception of the mind as an inner theater. Collingwood’s account of reenactment is mocked by Gardiner for endowing historians with telepathic powers of

206

The History of Understanding in Analytic Philosophy

access to the minds of historical agents, and Quine’s positive proposal for a science of understanding acquires its plausibility from the consideration that any defense of humanistic explanations must be committed to the existence of ontologically dubious entities. Once this misrepresentation is corrected, it becomes clear that there is no need to choose between the Lockean account of the relationship between words and ideas, with its associated model of how we understand others, and Quine’s eliminativism about meaning and his associated account of radical translation. The myth which needs debunking is no longer that of the ghost in the machine. It is the myth that any attempt to defend the autonomy of humanistic explanations must involve buying into the Lockean “myth of the museum.”

NOTES 1 The debate as to whether humanistic explanations are or are not causal has undergone a recent revival in the philosophy of mind. For more recent discussion and survey, see D’Oro and Sandis (2013) and Schumann (2019). 2 Hermenuticians such as Gadamer also deny the need to invoke psychological processes such as authorial intentions to reconstruct the meaning of a text. What is distinctive about Collingwood’s account of historical understanding is that the notion of “context” that is relevant to the historical understanding of action in history is the agent’s own, not that of the interpreter. He was not an advocate of the presentist view that the past must be understood from the perspective of the interpreter and, therefore, necessarily rewritten from different “presents.” He would have been as unsympathetic to Gadamer’s notion of the “fusion of horizons” as to the narrativists’ claim that the past is a construction from the perspective of the present. 3 We have not discussed here Davidson’s (2001) more nuanced attempt to avoid recourse to an intensional context of explanation without eliminating altogether the notion of meaning. On this, see D’Oro (2004). 4 Important here is also the fact that in the new preface to ISS ([1958] 1990: xiv– xvi), Winch himself admits that his account emphasized too strongly the role of rules in meaningful behavior. 5 In relation to this issue, it should be pointed out that there are readings of Winch which claim that ISS is not really concerned with foundations or methodological questions in relation to the social sciences, but is essentially concerned with an “ethical orientation” and, thus, tries to show how social inquiry is connected with philosophical and cultural questions about self-knowledge (cf. Theunissen 2020). However, it is unclear how this ethical orientation reading could escape the force of the naturalist critique, regardless of Winch’s intentions.

REFERENCES Ahlskog, J., and O. Lagerspetz (2015), “Language-games and Relativism: On Cora Diamond’s Reading of Peter Winch,” Philosophical Investigations, 38 (4): 293–315.

The Leopard Does Not Change its Spots

207

Ahlskog, J., and O. Lagerspetz (forthcoming), “The Pincer Movement: Winch and Collingwood on the Metaphilosophical Dimension of the Human Sciences,” History of the Human Sciences. Anscombe, G. E. M. (1957), Intentions, Oxford: Basil Blackwell. Bohman, J. (1993), New Philosophy of Social Science: Problems of Indeterminacy, Cambridge, MA: MIT Press. Collingwood, R. G. ([1940] 1998), An Essay on Metaphysics, rev. ed., with an introduction by R. Martin, Oxford: Oxford University Press. Collingwood, R. G. ([1946] 1993), The Idea of History, rev. ed. with an introduction by J. Van der Dussen, Oxford: Oxford University Press. Crary, A. (2018), “Wittgenstein Goes to Frankfurt (and Finds Something Useful to Say),” Nordic Wittgenstein Review, 7: 7–41. Davidson, D. (2001), “Radical Interpretation,” in Inquiries into Truth and Interpretation, 125–39, Oxford: Clarendon Press. D’Oro, G. (2004), “Re-enactment and Radical Interpretation,” History and Theory, 43: 198–208. D’Oro, G. (2012), “Reasons and Causes: the Philosophical Battle and the Metaphilosophical War,” Australasian Journal of Philosophy, 90 (2): 207–21. D’Oro, G. and C. Sandis (2013), “From Causalism to Anti-causalism and Back,” in G. D’Oro and C. Sandis (eds.), Reasons and Causes: Causalism and Anti-causalism in the Philosophy of Action, 7–48, Basingstoke: Palgrave. D’Oro, G. (2018), “The Touch of King Midas: Collingwood on Why Actions Are Not Events,” Philosophical Explorations, 21 (1): 1–10. D’Oro, G., P. Giladi, and A. Papazoglu (2019), “Non-reductivism and the Metaphilosophy of Mind,” Inquiry, 62 (5): 477–503. Dray, W. H. (1957). Laws and Explanation in History, London: Oxford University Press. Dray, W. H. (1958), “Historical Understanding as Rethinking,” University of Toronto Quarterly, 27: 200–15. Dray, W. H. (1963), “The Historical Explanation of Actions Reconsidered,” in S. Hook (ed.), Philosophy and History, 105–35, New York: New York University Press. Dray, W. H. (1964), Philosophy of History, London: Prentice-Hall Inc. Dray, W. H. (1980), “R. G. Collingwood and the Understanding of Actions in History,” in Perspectives on History, 9–26, London: Routledge and Kegan Paul. Gardiner, P. (1952a), The Nature of Historical Explanation, London: Oxford University Press. Gardiner, P. (1952b), “The Objects of Historical Knowledge,” Philosophy, 27: 211–20. Hempel, C. G. (1942), “The Function of General Laws in History,” Journal of Philosophy, 39: 35–48. Kemp, G. (2006), Quine: A Guide for the Perplexed, London: Continuum. Kirk, R. (2004), “Indeterminacy of Translation,” in R. F. Gibson, Jr. (ed.), The Cambridge Companion to Quine, 151–80, New York: Cambridge University Press. Locke, J. ([1690] 1964), An Essay Concerning Human Understanding, Glasgow: Collins Fount Paperbacks, William Collins Sons & Co Ltd. Mele, A. (2000), “Goal-directed Action: Teleological Explanations, Causal Theories and Deviance,” Philosophical Perspectives, 14: 279–300. Pettit, P. (2000) “Winch’s Double-Edged Idea of a Social Science,” History of the Human Sciences, 13 (1): 63–77.

208

The History of Understanding in Analytic Philosophy

Pleasants, N. (1999), Wittgenstein and the Idea of a Critical Social Theory: A Critique of Giddens, Habermas and Bhaskar, London: Routledge. Pleasants, N. (2000), “Winch, Wittgenstein and the Idea of a Critical Social Theory,” History of the Human Sciences, 13 (1): 78–91. Popper, K. (1957), The Poverty of Historicism, London: Routledge. Quine, W. V. O. (1951), “Two Dogmas of Empiricism,” Philosophical Review, 60: 20–43. Quine, W. V. O. ([1960] 2001), Word and Object, Cambridge, MA: MIT Press. Quine, W. V. O. (1969), Ontological Relativity and Other Essays, New York: Columbia University Press. Roth, P. (1987), Meaning and Method in the Social Sciences: A Case for Methodological Pluralism, Ithaca: Cornell University Press. Roth, P. (2003), “Mistakes,” Synthese, 136 (3): 389–408. Roth, P. (2011), “The Philosophy of Social Science in the Twentieth Century,” in I. C. Jarvie and J. Zamora-Bonilla (eds.), The SAGE Handbook of the Philosophy of Social Sciences, 103–17, London: SAGE Publications. Roth, P. (2012), “The Pasts,” History and Theory, 51 (3): 313–39. Roth, P. (2020), The Philosophical Structure of Historical Explanation, Evanston, IL: Northwestern University Press. Ryle, G. ([1949] 1990), The Concept of Mind, London: Penguin Books. Schumann, G. (2019), “Introduction,” in G. Schumann (ed.), Explanation in Action Theory and Historiography, 1–42, New York: Routledge. Stenlund, S. (1990), Language and Philosophical Problems, London: Routledge. Theunissen, M. (2020), “‘Don’t Mention the Idea of a Social Science’: The Legacy of an Idea,” in M. Campbell and L. Reid (eds.), Ethics, Society and Politics: Themes from the Philosophy of Peter Winch, 251–71, Cham: Springer. Turner, S. P. (1994), The Social Theory of Practices: Tradition, Tacit Knowledge, and Presuppositions, Chicago: Chicago University Press. Turner, S. P., and P. Roth (2003), “Introduction. Ghosts and the Machine: Issues of Agency, Rationality and Scientific Methodology in Contemporary Philosophy of Social Science,” in S. P. Turner and P. Roth (eds.), The Blackwell Guide to the Philosophy of the Social Sciences, 1–17, Oxford: Blackwell. Winch, P. ([1958] 1990), The Idea of a Social Science and Its Relation to Philosophy, 2nd ed., London: Routledge.

CHAPTER NINE

Georg Henrik von Wright on Understanding (and Explanation) HENRIIKKA HANNULA

9.1. INTRODUCTION Georg Henrik von Wright (1916–2003) was a Finnish philosopher best known for his work in logic, values, human action, and analytic hermeneutics. This chapter will focus on the last aspect. Von Wright’s Wittgenstein scholarship is also noteworthy, as are his general essays (written mostly in Swedish) about his contemporary culture, humanism, and ecology. Therefore, this chapter will be concerned with only a fragment of von Wright’s overall work, concentrating on the central themes and arguments in his 1971 Explanation and Understanding. Special attention will be given to the historical continuity of von Wright’s arguments within the long tradition of the question of “understanding.” In his book, von Wright argues for a fundamental difference between the explanation models in the natural and human sciences. He animates the old question with analytic rigor and clarity while also bridging some gaps between the analytic and continental divide. Explanation and Understanding deals with causal explanation, intentionality and teleological explanation, and explanation in history and the social sciences. Von Wright backs his systematic argumentation up by presenting a historical argument. He distinguishes between two main traditions in the history of science and philosophy—the Aristotelian and

210

The History of Understanding in Analytic Philosophy

the Galilean. These traditions are associated with the opposition between teleological and causal thinking, respectively. Through this historical survey, von Wright defends the Aristotelian tradition, seeking to support his thesis that the teleological explanatory model is characteristic to the human sciences, separating it from the natural sciences that are in turn characterized by the causal model. Von Wright’s central systematic argument is that human actions cannot be explained by reducing them to simple causality. They have to be understood as intentional and as motivated, usually, by different desires and beliefs. He discusses the implications of this conclusion for the disciplines of history and sociology, arriving at a dualist vision of scientific explanation. While the causal and nomological explanation model is suitable for the natural sciences, the social sciences need a new, but just as logically sound and rigorous, explanatory model. Von Wright puts forward the “practical syllogism” as a candidate for this basic model. A practical syllogism explains actions in light of the intentions motivating it. A simple example might be that the premise “sleep cures tiredness” and the premise “I am tired” lead to the action of me going to sleep. The central argument of the book is twofold: first, that the practical syllogism is the characteristic explanation model in the social and human sciences and, second, that this implies that there is a fundamental difference between the human and the natural sciences.

9.2.  EXPLANATION AND UNDERSTANDING IN CONTEXT The understanding versus explanation debate surfaced for the first time in Germany from the mid to late nineteenth century. There it was connected with the debate over the independence of the Geisteswissenschaften (human sciences) from the natural sciences. The most prominent advocates of this distinction were Johann Gustav Droysen and Wilhelm Dilthey. The second phase of the debate (see Apel [1979] 1984 and Väyrynen 2015: 324) can be associated with the debates around positivism and anti-positivism in the social sciences in the 1950s and 1960s, with contributions from analytic hermeneutics on the one hand (people like Winch, Dray, and Anscombe) and people drawing from the German tradition of sociological methodology on the other (Popper and Adorno).1 Von Wright (1971: 29) recognizes some affinities his project shares with his contemporary continental hermeneutic or hermeneutic-dialectic philosophy, (Gadamer and Habermas). However, von Wright approaches the problem through the more analytic strain of the philosophy of action. He draws

Georg Henrik von Wright on Understanding

211

from his contemporary neo-Wittgensteinians (e.g., Anscombe, Melden), who emphasize the importance of intentionalist action explanations and draw inspiration from Aristotelian logic as to practical reasoning (Tuomela 1976: 183). Here, von Wright sees a potential bridge between the continental and analytic traditions. He proclaims that both continental hermeneutic philosophy and analytic philosophy of action orient themselves after the thought of the later Wittgenstein, thus sharing an interest in “the idea of language” and “language-oriented notions such as meaning, intentionality, interpretation, and understanding” (von Wright 1971: 29–30). As any interesting philosophical thesis should, Explanation and Understanding provoked many critical responses from both analytic and continental philosophers. Following a symposium in Helsinki focusing on von Wright’s work, many of these were collected in the 1976 anthology Essays on Explanation and Understanding—Studies in the Foundations of Humanities and Social Sciences (Manninen and Tuomela 1976). Von Wright later revised some of his views.2 However, this chapter will concentrate mainly on elucidating the arguments in Explanation and Understanding and leave aside the critical reception. An extensive discussion on von Wright’s later views is beyond the scope of this chapter.

9.3.  VON WRIGHT’S HISTORICAL ARGUMENT: THE ARISTOTELIAN AND THE GALILEAN TRADITIONS As stated above, Von Wright has two independent lines of argumentation, historical and systematic (Riedel 1976: 6). In the first chapter of the book, von Wright distinguishes between the Aristotelian and the Galilean traditions in the philosophy of scientific explanations. The Aristotelian tradition seeks to understand things teleologically, and the Galilean tradition is about explaining them causally, or rather, taking the “causal-mechanistic point of view” when explaining phenomena. This distinction should be seen as a typology rather than a definite historical theory (von Wright 1971: 1), but even so, it motivates von Wright’s systematic argumentation. The historical argumentation is there to bring weight to the principal claim of Explanation and Understanding: that there is an important difference between the models of explanation characteristic of the social scientific and natural scientific disciplines. In the section about history, von Wright positions himself in relation to the Verstehen tradition and hermeneutics and his contemporary analytic trends in the philosophy of action. Even though there had been proponents of this causal-mechanistic strain of scientific explanations already in ancient philosophy, the Galilean

212

The History of Understanding in Analytic Philosophy

tradition got its most definitive and influential expression in the “platonizing spirit of Renaissance and Baroque science” (von Wright 1971: 8). As the name suggests, “its most impressive champion was Galileo” (von Wright 1971: ibid.). In turn, the Aristotelian tradition gained its modern form and relevance through the development of the modern human and social sciences throughout the nineteenth century. According to von Wright, Hegel developed the first modern iteration of the Aristotelian philosophy of method. For both Aristotle and Hegel, “explanation consists in making phenomena teleologically intelligible rather than predictable from knowledge of their efficient causes” (von Wright 1971: 8). Another line of development in the teleological tradition von Wright discusses is the anti-positivist philosophy of the human sciences or Geisteswissenschaften that developed primarily in Germany in the late nineteenth century. He mentions Johan Gustav Droysen, Wilhelm Dilthey, Georg Simmel, and Max Weber as important proponents of this tradition. He also sees British philosopher of history and art R. G. Collingwood as a part of this same tradition of thinkers. Interestingly, von Wright does not engage further with these well-known thinkers of the understanding tradition, but outright dismisses their theories as outdated and psychologistic (1971: 6). Of all the continental philosophers, it is Hegel (and to a lesser extent, Marx) whose philosophy von Wright actually engages with and discusses in some length. According to von Wright, in the nineteenth and twentieth centuries, the distinction between the Aristotelian and Galilean traditions developed into two opposing positions in the philosophy of the scientific method: dialecticalhermeneutical and analytical. This distinction determined the relationship between modern historical Geisteswissenschaften (i.e., human sciences) and natural sciences. This historical argument supports von Wright’s own dualist vision of scientific explanations. But what about modern social and “behavioral” sciences? Von Wright (1971: 6) argues they were born under the influence of a “cross pressure of positivist and anti-positivist tendencies” and, thus, became a battleground for the two traditions in the philosophy of science. Both the Aristotelian and Galilean traditions tried to claim the social sciences as their own. Von Wright (1971: 9) defines positivism as “a philosophy advocating methodological monism, mathematical ideals of perfection, and a subsumption-theoretic view of scientific explanation.” The philosophies of Auguste Comte, Emile Durkheim, and John Stuart Mill fit this description— indeed, it is Comte who coined the term “sociology,” which testifies to the positivistic roots of the discipline. For von Wright (1971: 7), Max Weber is at a focal point in this battle between the traditions: the “positivist colouring”

Georg Henrik von Wright on Understanding

213

of his interpretative sociology is constrained by an emphasis on teleology and “empathic understanding.” Von Wright states that as a tradition, analytic philosophy has usually taken the positivists’ side in questions on the philosophy of science. However, this alliance is only accidental and historically contingent, owing to the heritage of Russell’s and the early Wittgenstein’s logical atomism and the neopositivism of the Vienna Circle. There is nothing in analytic philosophy itself that forces this to be the case. Also, because of this heritage, analytic philosophers of science were long concerned almost exclusively with the methodology of the exact natural sciences, only much later getting interested in the philosophy of the social and behavioral sciences. In von Wright’s view, the discussion of problems of explanation within the tradition of analytic philosophy got its decisive impetus from Carl Gustav Hempel’s 1942 paper “The Function of General Laws in History.”3 In the following decades, this controversial paper defines much of the subsequent discussions around the analytic philosophy of the social sciences. Hempel argues that history and society are governed by general laws just as the natural world is governed by natural laws. The “Hempelian model” is also the principal adversary or interlocutor for von Wright’s vision in Explanation and Understanding. Besides the historical reconstruction of the development of the two traditions, the other function of the first chapter’s argumentation is methodological (Riedel 1976: 7). Von Wright claims that “practical reasoning” is of great importance to the explanation and understanding of action. Thus, it makes sense to evoke the Aristotelian tradition, which had been revived in the (mostly British) philosophy of action in the previous decades. Von Wright mentions Elizabeth Anscombe and Peter Winch as important inspirations for his work. Von Wright counts both as fellow fighters in the anti-positivist camp in the theory of explanation. Both of them criticize philosophical theories that see motives and beliefs as causes of action (Stoutland 1982; Tuomela 1976). Winch’s 1958 The Idea of a Social Science and its Relation to Philosophy is “an attack on positivism and a defence of an understanding of social phenomena by methods different in principle from those of the natural sciences” (von Wright 1971: 28). Von Wright (1971: 26–7) especially emphasizes his intellectual debt to Elizabeth Anscombe’s work. Her 1957 book Intention makes the notion of intentionality central to the subsequent discussion of the philosophy of action among analytic philosophers. This is important for von Wright: only if actions are understood in terms of intentions, not in terms of causes, the distinction between explanation and understanding becomes conceptually significant. Anscombe also revives the Aristotelian practical inference, which then becomes von Wright’s main candidate for the paradigmatic case of

214

The History of Understanding in Analytic Philosophy

social scientific explanation against the Hempelian subsumption theory of explanation. Von Wright reconstructs Anscombe’s key idea in the following: The starting point or major premise of the syllogism mentions some wanted thing or end of action; the minor premise relates some action to this thing, roughly as a means to the end; the conclusion, finally, consists in use of this means to secure that end. Thus, as in a theoretical inference the affirmation of the premises leads of necessity to the affirmation of the conclusion, in a practical inference assent to the premises entails action in accordance with them. (von Wright 1971: 26)4 Von Wright ends the first chapter on a programmatic note. He argues that the practical syllogism should become the characteristic model for a social scientific explanation. In short, he sets himself the task of formulating an analytic, logically solid model for the social sciences, thus modernizing the Aristotelian tradition and bridging gaps between the analytic and continental hermeneutic traditions. It is a tenet of the present work that the practical syllogism provides the sciences of man with something long missing from their methodology: an explanation model in its own right which is a definite alternative to the subsumption-theoretic covering law model. Broadly speaking, what the subsumption-theoretic model is to causal explanation and explanation in the natural sciences, the practical syllogism is to teleological explanation and explanation in history and the social sciences. (von Wright 1971: 27) The analytic and positivistic Galilean tradition is associated with the idea of the unity of science: that the social and human sciences ultimately share the same principles and methods with the natural sciences. In von Wright’s view, the Aristotelian tradition can defend a dualistic vision of sciences— that there is a real difference between the sciences of human and the sciences of nature. Interestingly, von Wright’s exegesis of the two traditions ends in a brief concession that the intentionalist and causalist views are incommensurable with each other. Between these explanations, there is “a basic opposition, removed from the possibility both of reconciliation and of refutation” (1971: 32) built into the basic concepts of whole argumentation. The choice between the two is existential, not logical. For the rest of Explanation and Understanding to be convincing, the reader has to choose to accept the basic premises of the dualistic vision of scientific explanation in the book. This choice is what the historical argumentation aims to encourage and justify.

Georg Henrik von Wright on Understanding

215

9.4.  CAUSAL VERSUS TELEOLOGICAL EXPLANATION While this chapter will mainly be concerned with the hermeneutic “understanding” side of von Wright’s theory, as the book’s title suggests, it is as much about explanation as it is about understanding. Indeed, von Wright dedicates one-fourth of the book to the problems of causality. In the interest of space, the view will be summarized only insofar as it is relevant for his overall argument about the difference between the explanatory schemas in the natural and social sciences. For von Wright, the distinguishing feature of an actual causal relation is that cause and effect are logically independent of one another. He calls this kind of causal relationship Humean without, however, committing himself to the rest of Hume’s views about the nature of causation (von Wright 1971: 93). An example would be: when a billiard ball hits another and the second one moves, there is no logical but merely a factual connection between the two events. These two events are said to be causally related, but all we see is that they instantiate certain regularities of succession that, if repeated, seem to happen over and over again. There is no logical or conceptual connection that explains this relationship, only states of affairs that may obtain or not on given occasions. These states are logically independent, meaning that it is logically possible for them, on any given occasion, to obtain or not obtain in any combination (von Wright 1971: 42–3). Causal explanation is understood in terms of the covering law model. Von Wright’s primary target here is the Hempelian theory of explanation, later named the covering law model by its critic William Dray (1957: 1) or, the subsumption theory of explanation (cf. Mandelbaum 1961). According to this model, explaining an event by reference to another event (as it is with causation) necessarily presupposes a law or a general proposition underlying these events. Simply put, if I infer that when a billiard ball hits another, the other always starts moving, I postulate a covering law. Alternatively, I explain the relationship between the two billiard balls by referring to a covering law. A causal explanation appeals to laws or general propositions correlating events of the type to be explained (explananda) with events of the type cited as its causes or conditions (explanans/explanation). According to Hempel, scientific explanations always involve at least one law. In “The Function of General Laws in History,” Hempel applies this model to history and historical explanation, arguing for methodological monism and the unity of natural and social sciences. For him, all scientific explanations share the same logical structure. Explaining a historical occurrence is a deductive or probabilistic derivation of the explanandum

216

The History of Understanding in Analytic Philosophy

from general laws and facts. In Hempel’s view, an explanation is thus merely about subsuming an occurrence or event under general laws. He tries to show that in history, like in any other branch of empirical inquiry, a scientific explanation can be achieved only by means of suitable general hypotheses (Hempel 1942: 44). Von Wright accepts that the covering law model works for the natural sciences. Hempel goes astray when applying it to historical explanation (and, by extension, to other social sciences). Indeed, von Wright (1971: 10–11) notes that “it seems almost an irony of fate that the fullest and most lucid formulation of the positivist theory of explanation should have been stated in connection with the subject matter for which, obviously, the theory is least suited, viz. history.” While von Wright does not deny that causal explanations can be present in the social sciences or even that they can be very important, they are still not the characteristic model of social scientific inquiry. A social scientist or historian is interested in things that have to do with human agency. They seek to uncover the intentions and motivations that lead to action. Could then a motivation or intention be considered a cause of action? Or the cause of the outcome of the action? Von Wright asks us to consider the action of the “ringing of a doorbell,” resulting in the bell ringing. Can the intention or will to ring the bell cause this result? Von Wright argues that an intention can never be a Humean cause. For this purpose, von Wright updates the so-called “logical connection argument,” initially developed by Melden in (1961) and Davidson in (1963).5 For them, causal theories of action are not only incorrect but also radically confused. The argument goes as follows: The relation between a genuine cause and effect must be contingent, or it is not (Humean) causation. Since the relationship between a want and the resultant action is logical, the relationship cannot be deemed causal. In other words, action cannot be explained causally and wants, desires, and motives cannot be considered as the causes of action. In light of the logical connection argument, we need another schema for a teleological explanation: the practical inference. This explains an action in terms of the actor’s goals or intentions. An intentional explanation is there to explain why “someone does something.” When we observe an action, such as pushing a button, we are not usually interested in the cause of the said action. Instead, the question we ask is “Why?” And the answer often is: “In order to bring about p” (von Wright 1971: 96). The most basic schema that von Wright gives for a teleological explanation is: A intends to bring about p. A considers that he cannot bring about p unless he does a.

Georg Henrik von Wright on Understanding

217

Therefore A sets himself to do a. (ibid.) This, however, is not yet a logically conclusive argument. Von Wright goes through several scenarios: the agent might fail to carry out her intention by, for example, forgetting it. She might be prevented from acting by an external force. Sometimes time is essential to performing actions successfully: if an agent intends to go to a concert or cinema, she has to intend to buy the tickets in advance. Von Wright then tries to make the schema for practical inference as watertight as possible, as he wants to assign it a truly binding force. The final formulation of the inference schema is: From now on A intends to bring about p at time t. From now on A considers that, unless he does a no later than at time t’, he cannot bring about p at time t. Therefore, no later than when he thinks time t’ has arrived, A sets himself to do a, unless he forgets about the time or is prevented. (von Wright 1971: 107) Even though von Wright principally accepts the truth of the logical connection argument, he admits that the premises of a practical inference do not as such entail certain behavior with logical necessity. “The syllogism when leading up to action is ‘practical’ and not a piece of logical demonstration” (von Wright 1971: 117). It could be possible that someone intends to bring about something and knows the means necessary for this end, but still fails to act according to the intention. Von Wright gives an example of a person who has resolved to shoot an authoritarian tyrant. He is standing in front of the tyrant, did not forget any preparations and is not prevented from acting, and now it is time for him to act. Still, von Wright (1971: 116–7) argues, it can be that he does not set himself to act in accordance with the premises. The anarchist revolutionary might not carry out the act of shooting the tyrant. This, because there is no logical compulsion behind the practical syllogism. This admission is somewhat puzzling given that von Wright’s argument builds so much on the logical connection argument. It also seems to undermine the general argument of Explanation and Understanding (cf. Dray 1989: 484). His argument, however, seems to be that while there is a logical connection between the actions, there is no logical necessity. The necessity of the practical inference schema is only conceived ex post actu (1971: 117), explaining an action that has already happened. Then the premises and the conclusion are conceptually dependent on each other.6 Von Wright recognizes that both human causes and human intentions can be present in an action. If I raise my arm and push a button, one can

218

The History of Understanding in Analytic Philosophy

presumably give a causal explanation as well as an intentional one. One could probably also track a causal chain of neural events when I raise my arm and push the button. This, says von Wright, is not because the event would have two different “causes”—the neural event and me. His argument is that the interpretation of behavior as an intentional act can be compatible with the behavior having a Humean cause. While in both causal and teleological explanations the explanandum is an occurrence or an event in the world, in a causal explanation the explanans is an underlying general law and in a teleological explanation the explanans is the intention or motivation behind the action. In the case of an intentional explanation, however, it is better to say that we understand a behavior rather than explaining it (1971: 123). For von Wright, understanding here thus stands for having a proper teleological explanation for an act.

9.5.  EXPLANATION IN THE SOCIAL SCIENCES AND HISTORY In the last chapter of the book, von Wright discusses explanation in history and the social sciences. He puts his systematic argumentation to practice and evokes different examples, although all of his chosen examples have to do with historical explanations. It is not entirely clear how the practical syllogism is supposed to be a relevant explanation model in the other social and human sciences. Von Wright readily admits that practical syllogism is not sufficient to explain social phenomena or crowd behavior. Not all of those have a purpose or a meaning that could be extracted from the purposes or intentions of individual men and women. Thus, a historian or a social scientist needs to interpret the event. Von Wright suggests we call this act of interpretation “explicative.” Before the explicative interpretation can begin, however, the explanandum—the event or phenomenon that is to be explained—must be described. In other words, it needs to be grasped as something, in “the sense of means or signifies” (1971: 135, original emphasis). Von Wright thus uses the word “understanding” in two different ways (cf. Kusch 2003: 329). As we saw in the previous section, it can refer to the understanding of intentional acts of individuals. But it also refers more broadly to our ability to grasp events and phenomena as something that is imbued with meanings and significance. This is the preliminary of all teleological explanations. When we see a crowd shouting and holding placards that say “the government must resign,” we recognize this as a phenomenon with an intended meaning. That is, we understand it as a demonstration. Only now can further explicative interpretation begin, either by evaluating the

Georg Henrik von Wright on Understanding

219

historical and cultural reasons leading to it or by analyzing the material and economic reasons for people’s dissatisfaction. Von Wright also argues that causal explanation includes a certain kind of preliminary understanding; that is, grasping something as something. For example, I understand a flower as a flower. This kind of understanding is relatively trivial, but it would be, says von Wright, misleading to say that understanding versus explanation marks the difference between two types of scientific intelligibility. One can nonetheless say that the intentional or non-intentional character of their objects marks the difference between understanding and explanation (1971: 135). In other words, it is intentionality in general that characterizes the subject matter of history and the social sciences and thus demarcates them from the natural sciences. There is a place for genuinely causal explanations in the social sciences and historical explanation, but they are usually subordinate to other forms of explanation and not the primary interest of historians. How and why Vesuvius erupted and how it led to the destruction of Pompeii is only interesting for the historian insofar as it is linked to the subsequent human affairs. As a Humean cause of the destruction of the houses and the death of the inhabitants, it is not as interesting. It is usually not relevant to know the biological and physical mechanisms that led to people’s gruesome deaths in Pompeii. However, it is sometimes historically significant to determine whether or how something was possible. For example, it is significant to know whether and how it was causally possible for the people in prehistoric Britain to transport the stones needed to construct Stonehenge. But even if this is determined, it is still an insufficient answer if we wanted to understand why Stonehenge was built; that is, what its significance and meaning was. Von Wright uses the outbreak of the First World War as an example to illustrate the difference between a causal and intentionalist explanation in history. We can and do speak about “the causes of the First World War.” However, von Wright explores whether and in what way we can say that the murder of Archduke Franz Ferdinand in Sarajevo in July 1914 was the “cause”—or at least one of many causes—of the outbreak of the war. What is the conceptual nature of the mechanism connecting the explanans with the explanandum here? The requirement of the logical connection argument is thus met. Clearly, the killing of the archduke and the outbreak of the war are not logically connected. It would also be strange to claim that a covering law explains the connection between the events. While Franz Ferdinand’s assassination has often been figuratively called the “spark which made the powder barrel explode”—that is, it started the war—surely the relationship between the two events is not causal in the same way as when a spark literally makes a powder barrel explode.

220

The History of Understanding in Analytic Philosophy

Von Wright’s suggestion is to construct the connection between the events not as a general law but as a set of singular occurrences that constitute the premises of practical inferences. The explanandum, the historical event that is to be explained, should be understood as a chain of situations that all include practical inferences. When the archduke was killed, this triggered a deliberation process in Austria-Hungary. This deliberation process was based on the motivational background of Austrian power politics, which then resulted in the issuing of an ultimatum. This created another situation, actuating a new practical inference (the Russian cabinet defending its interests and issuing another ultimatum) which again created a new situation (the mobilization), actuating further practical inferences. The final conclusion of this chain of situations was the outbreak of the war. Thus, the explanandum is reached through a number of steps of intermediate conclusions to practical inferences. Von Wright (1971: 142) calls these explanations quasi-causal. As established, they do not fit the covering law model, so they are not genuinely causal. They cannot really be called teleological either, as the outbreak of the war is not an intentional act in the same sense as pushing a button is, even though teleology, of course, enters into the practical inferences that link the various explanans to the explanandum. As some critics note,7 von Wright does not consider the use of counterfactuals in assessing the conceptual nature of historical explanation. It is a tool now widely used in historiography: counterfactual considerations are considerations about “what if things had been different”; for example, what if Franz Ferdinand would not have been assassinated. Theoretically, a historian could at least consider this question. They might even come to the unlikely conclusion that the First World War would not have started had the archduke not been killed. In this case, the assassination would look much more like a genuine cause, and the conceptual distinction is not as clear as von Wright’s would have us believe. Also, because von Wright defines causality so much through the idea of a covering law, he does not entertain the possibility that genuine causes could also be singular and isolated events. As a contrast, let us briefly consider the Hempelian model of covering laws in history. Hempel’s model has fallen out of favor in the social sciences. In von Wright’s time, however, it was the received view, and this is the position von Wright is arguing against (Raatikainen 2017). What Hempel understands by general laws in history remains rather vaguely defined in his 1942 paper. He appeals to regularities in individual and social psychology and collective behavior. We can observe that groups tend to migrate to regions that offer better living conditions or that growing discontent tends to lead to an outbreak of a revolution. These observations seem to be accurate, and it

Georg Henrik von Wright on Understanding

221

is intuitively appealing to assume that they could be some sort of regulative laws. Understood like this, they also seem causal and yet still include human intentions and motivations in this causal connection. It is certainly true that “growing discontent caused the revolution” has a causal ring to it and can be a perfectly valid historical explanation. To bring it back to the example about the First World War, one could say (and many historians do) that militarism, imperialism, and nationalism caused the war. It seems plausible, or at least it cannot be outright refuted, that there could be a universal regularity at play here: growing militarism and nationalism tend to cause wars. If there are general laws in history, this seems like a perfectly plausible one. However, for genuine Humean causality, one would have to postulate a real lawlike regularity connecting the events. Here the connection between motivation (e.g., growing discontent in a population) and the outcome (revolution) is not nomic; it is not necessary and sufficient as we take genuinely causal relationships to be. People who are discontent might also not start a revolution, they might fail to start it, or the situation could morph into something else. It is certainly possible that the outbreak of the First World War could have been avoided even after Franz Ferdinand was shot in Sarajevo. It also would seem strangely dogmatic to insist that the First World War was an inevitable causal result of the militarism and nationalism of the nineteenth century and that things could not have turned out differently. Closely related to this is von Wright’s interesting discussion on determinism. Given the argument of the book, he deems many beliefs and ideas concerning determinism in history to be simply confused, stemming from the mistaken view that the human world works according to causal laws, analogous to those that govern the natural realm. However, one can meaningfully talk about determinism if one works with a more precise definition. One can distinguish between two types of determinism, which von Wright calls “predetermination” and “post-determination.” The former is connected with predictability, and the latter with ideas of intelligibility of the historical and social process. As was the case with the practical inference schema, the necessity of which is only conceived ex post actu, the intelligibility of history is a determinism ex post facto (von Wright 1971: 161). Von Wright briefly considers the possibility for genuine teleology in history—in the sense that history as a whole would have a direction or people, a destiny—but understandably denies such a possibility outright. He does admit that philosophical theories postulating a direction to history may “nevertheless be of great interest and value” (1971: 167). Hegelian or Marxist dialectics come to mind—even if their teleological theories are rejected as credible descriptions of how things are and will be, once these

222

The History of Understanding in Analytic Philosophy

theories are out in the open, they might still have significant effects in the real world. Von Wright, however, recognizes that there can be developments in history that look rational—that they had to happen so that something else could happen. The suffering of a people might lead to rational, even generally beneficial outcomes. Von Wright’s example is Poland’s economic recovery under Casimir the Great (1333–1370), which was largely due to the fact that Jews, who had been expelled from German territory, were invited to settle in Poland. They did not leave Germany in order to contribute to Poland’s economic and cultural flourishing, yet that was the outcome. He calls these kinds of developments “quasi-teleological” (1971: 156). There is no fate or reason in history. Such developments were not “meant to be” in any strong sense.8

9.6.  DIFFERENCE BETWEEN NATURAL AND HUMAN SCIENCES Many critics of Explanation and Understanding point out that the practical syllogism—at least in the form that von Wright would have it—is too narrow to work as the paradigmatic model of explanation in the human and social sciences.9 As noted above, it does not really provide adequate tools for explanations of complex historical processes and developments that are not reducible to individuals’ intended actions. Later, von Wright agrees with much of this criticism and revised many of his views. He abandons the idea of the centrality of the practical syllogism in the human and social sciences (von Wright 1989: 814) and even questions whether action explanations themselves could be a key to historical and social scientific explanations (von Wright 1981: 128). Von Wright, having his roots in the Wittgensteinian tradition, echoes his teacher’s logical atomism.10 For him, the “states of affairs” are the sole ontological building blocks of the world. Those are what a scientific explanation tries to describe; that is, a scientific investigation is interested in some fact or set of facts. This has important implications for von Wright’s view on the social and human sciences at the outset. His view of historical explanation is concentrated on important events: their causes and effects. Von Wright’s theory seems to be tailored to an idea of history as a succession of important events and less suited for other interests. Historians, or social sciences more generally, are not usually only interested in causes and effects. One could argue quite plausibly that, for example, the interests of the history of ideas or art history are not reducible to just events.

Georg Henrik von Wright on Understanding

223

Von Wright’s practical syllogism is also concentrated on individual action (Dray 1989: 479). The practical syllogism takes individuals to be rational agents who act according to considerations on necessary and sufficient reasons. It only seems to take seriously purposeful or goal-oriented rational action, action that Max Weber would call zweckrational. However, in Weber’s theory of social action, people may also engage in value-oriented rational action (wertrational). They may act from emotional or affective motivations or engage in traditional action. Von Wright’s practical syllogism is not well suited to explaining these kinds of actions, yet they are of paramount importance to social scientific explanations (see also Apel [1979] 1984: 112–20). The later von Wright agrees that the action explanation theory should be developed to take a more collectivistic viewpoint. He proposes that we should focus on customs, practices, rules, institutions, offices, roles, and how they regulate individuals’ actions in a community (von Wright 1989: 842). For later von Wright, external customs and rules can be reasons for action and, as such, significant for action explanations. For example: if I have to understand why people do not generally jaywalk, I do not have to assign them a specific intention or motivation “not to jaywalk”—it is perfectly adequate to explain the phenomenon by referring to social customs, norms, and rules around jaywalking. Action explanation is about widening the context of the explicative principles; that is, seeing it in the context of cultural and social norms and the agent’s life history (von Wright 1981: 142). Despite these concessions to critics and several papers in which von Wright further refines his ideas, he does not develop a new collectivist theory of action.11 Neither does he put forward a new proposition for a philosophical toolkit for the human and social sciences that would replace his theory about the practical syllogism. If, as von Wright’s concedes, the practical syllogism cannot function as the paradigmatic explanatory model for the social and human sciences, what is left of the argument about the fundamental distinction between the natural and the social sciences? Sami Pihlström (2019) puts forward a suggestion that perhaps von Wright should be read as a “pragmatic pluralist” when it comes to scientific explanation. When we explain or understand an event taking place in the world, “we must always explain it under some description, first conceptualizing and thereby understanding it either as a natural event to be causally explained or as a human action with meaning, to be intentionally, rationally, and hence teleologically explained” (ibid.). So, even if one were to grant that social and natural sciences both mix explanation and understanding, applying the model that always best suits the subject matter, the result would be methodological pluralism. Going further, one could still

224

The History of Understanding in Analytic Philosophy

insist, with von Wright’s argumentation in mind, that social/human sciences and natural sciences characteristically use different kinds of explanatory models. They tend to conceptualize their objects differently: the natural sciences as a collection of causal events to be explained, the human sciences as a result of human action that needs to be understood intentionally. In this view, the difference between natural and human sciences would still hold, although it would be grounded on pragmatic considerations. One might still want to insist that there is something yet more fundamental to be said about the difference between the natural and human sciences. In contrast to the older theories in the tradition of “understanding,” von Wright does not ground the theory on psychology or a theory of mind. He is decidedly anti-foundationalist. However, one can ask whether the epistemological puzzle of understanding versus explanation can be solved without an epistemological foundation. Understanding for von Wright is a semantic rather than a psychological act. This is no wonder, as he argues that understanding should be strictly distinguished from empathy or Einfühlung. He associates Einfühlung with people like Droysen, Dilthey, and Collingwood and considers them to be hopelessly old-fashioned and “psychologistic.” That said, von Wright is not entirely justified in dismissing the older tradition of hermeneutics and the idea of Einfühlung or “reenactive empathy” from the outset. For people like Dilthey and Droysen, “empathy” does not mean “sympathy” or “being empathetic.” It means recognizing that when we try to explain why humans do or say something, we understand that something because it is done in the context of our shared world. It demarcates the epistemic distinctiveness of the human sciences by recognizing the epistemic centrality of reenactive empathy for our understanding of rational agency.12 In fact, this idea does not seem to be too different from von Wright’s considerations on the teleological or motivational background of actions. He also explicitly suggests that intentions, wants, and beliefs are developed in a social and cultural context, probably also drawing from later Wittgenstein’s concept of “forms of life” (Wittgenstein [1953] 1967: 8, 11, 88, 174). In order to understand, elucidate, and explicate actions, we have to consider them in this cultural and social nexus. Again, coming from a Wittgensteinian background, von Wright uses language as a metaphor for social and cultural reality: Intentional behaviour, one could say, resembles the use of language. It is a gesture whereby I mean something. Just as the use and the understanding of language presuppose a language community, the understanding of action presupposes a community of institutions and practices and

Georg Henrik von Wright on Understanding

225

technological equipment into which one has been introduced by learning and training. One could perhaps call it a life community. We cannot understand or teleologically explain behaviour which is completely alien to us. (von Wright 1971: 114–15) Here, von Wright seems to consider “sharing a life community” a necessary epistemological foundation for the social and human sciences. Of course, it is unclear what counts as a life community. How well can I understand the intentional acts of someone who comes from a different culture? Or, important for a historical explanation, how can we attribute feelings and motivations to someone in the past? In any case, it seems that von Wright’s concept of a life community is not so far removed from the concept of “reenactive empathy” of the older hermeneutics and, perhaps more importantly, Wittgenstein’s “form of life.”13 It is what makes possible the understanding of things as meaningful and significant. Unfortunately, von Wright does not pursue this strain of argumentation further in Explanation and Understanding, and his theory remains primarily a theory of understanding individual action. His brief remarks about larger social communities and institutions, however, open a door for a more collectivist theory of action explanations. Thus, if developed further, the concept of life community could function as an epistemological foundation grounding the human and social sciences and thus justifying the explanation versus understanding distinction.

NOTES 1 For a more extensive historical reconstruction of the explanation/understanding controversy, see Apel ([1979] 1984). 2 For a good overview of the criticisms and von Wright’s revisions of the arguments of Explanation and Understanding, see Kusch (2003). 3 On Hempel’s role in the history, see Fons Dewul’s chapter in this volume. 4 It should maybe be noted that Anscombe (1989) disagrees with von Wright’s use of practical syllogism, claiming that for von Wright “practical inference” functions as a theoretical inference put to practical use, not genuinely practical. 5 Other versions of the logical connection argument are also presented by Anscombe (1957), Winch (1958), and Taylor (1964). See also Otten (1977) and Schumann (2016). 6 The point about ‘conceptual’ relation and relation of ‘conceptual necessity’ becomes more prominent and explicit in von Wright’s later argumentation when he abandons the logical connection argument (von Wright 1976a, 1976b, 1980). 7 See, for example, Väyrynen (2015: 327) and Raatikainen (2017). Väyrynen notes that the use of counterfactuals in history was first proposed by David Lewis in his

226

The History of Understanding in Analytic Philosophy

book Counterfactuals, which was published in 1973, two years after Explanation and Understanding. 8 Von Wright returns to these questions in his later essayistic writings; for example, in “The Myth of Progress” (von Wright 1993). 9 See, for example, Martin (1976), Wellmer (1979), Apel (1979/1984), Markovic (1989), and Dray (1989). 10 He explicitly evokes the concept of the “Tractatus world” (the world is everything that is the case) in the context of his discussion on causality. There, he takes it to be a good model of the world (von Wright 1971: 43–4). It is, however, not clear to what extent he thinks that the atomistic conception would work for the human world. 11 Despite this, von Wright’s views have inspired developments in the philosophy of social institutions (e.g., Tuomela 1995). Kusch (2003) develops von Wright’s theory towards a collectivistic action explanation theory through insights from the sociology of knowledge. 12 For a modern take on this matter, see Stueber (2012). 13 Wittgenstein’s “forms of life” has been discussed extensively in Wittgenstein research, even though the concept is explicitly mentioned only five times in the Philosophical Investigation. For an introduction, see Tonner (2017).

REFERENCES Anscombe, G. E. M. (1957), Intention, Oxford: Basil Blackwell. Anscombe, G. E. M. (1989), “Von Wright on Practical Inference,” in P. A. Schilpp and L. E. Hahn (eds.), The Philosophy of Georg Henrik von Wright, 177–404, La Salle, IL: Open Court. Apel, K.-O. ([1979] 1984), Understanding and Explanation. A TranscendentalPragmatic Perspective, trans. G. Warnke, Cambridge, MA: MIT Press. Davidson, D. (1963), “Actions, Reasons, and Causes,” Journal of Philosophy, 60: 685–700. Dray, W. H. (1957), Laws and Explanation in History, Oxford: Oxford University Press. Dray, W. H. (1989), “Von Wright on Explanation in History,” in P. A. Schilpp and L. E. Hahn (eds.), The Philosophy of Georg Henrik von Wright, 470–87, La Salle, IL: Open Court. Hempel, C. G. (1942), “The Function of General Laws in History,” Journal of Philosophy, 39 (2): 35–48. Kusch, M. (2003), “Explanation and Understanding: The debate over von Wright’s philosophy of action revisited,” in L. Haaparanta and I. Niiniluoto (eds.), Analytic Philosophy in Finland, 327–53, Amsterdam: Rodopi. Lewis, D. (1973), Counterfactuals, Oxford: Blackwell. Mandelbaum, M. (1961), “Historical Explanation: The Problem of ‘Covering Laws,’” History and Theory, 1 (3): 229–42. Manninen, J., and R. Tuomela, eds. (1976), Essays on Explanation and Understanding—Studies in the Foundations of Humanities and Social Sciences, Dordrecht: Reidel.

Georg Henrik von Wright on Understanding

227

Markovic, M. (1989), “Von Wright on Explanation Versus Understanding: The Relation of the Sciences of Nature and the Sciences of Man,” in P. A. Schilpp and L. E. Hahn (eds.), The Philosophy of Georg Henrik von Wright, 445–70, La Salle, IL: Open Court. Martin, R. (1976). “Explanation and Understanding in History,” in J. Manninen and R. Tuomela (eds.), Essays on Explanation and Understanding: Studies in the Foundations of Humanities and Social Sciences, 305–34, Dordrecht: Reidel. Melden, A. I. (1961), Free Action, London: Routledge. Otten, J. (1977), “Reviving the Logical Connection Argument,” Canadian Journal of Philosophy, 7 (4): 725–43. Pihlström, S. (2019), “Finnish Versions of Pragmatist Humanism,” European Journal of Pragmatism and American Philosophy, 11 (1). https://journals.openedition.org/ ejpap/1543. Raatikainen, P. (2017), “Explanation and Understanding Revisited,” in I. Niiniluoto and T. H. Wallgren (eds.), On the Human Condition: Philosophical Essays in Honour of the Centennial Anniversary of Georg Henrik von Wright, 339–53, Helsinki: The Philosophical Society of Finland. Riedel, M. (1976), “Causal and Historical Explanation,” in J. Manninen and R. Tuomela (eds.), Essays on Explanation and Understanding: Studies in the Foundations of Humanities and Social Sciences, 3–26, Dordrecht: Reidel. Schumann. G. (2016), “Von Wright’s Theory of Action and the ‘Logical Connection Argument’,” in I. Niiniluoto and T. H. Wallgren (eds.), On the Human Condition: Philosophical Essays in Honour of the Centennial Anniversary of Georg Henrik von Wright, 325–38, Helsinki: Philosophical Society of Finland. Stoutland F. (1982), “Philosophy of Action: Davidson, von Wright, and the Debate over Causation,” in G. Fløistad (ed.), Contemporary Philosophy, A New Survey: Volume 3, Philosophy of Action, 45–72, Dordrecht: Springer. Stueber, K. R. (2012), “Understanding versus Explanation? How to Think about the Distinction between the Human and the Natural Sciences,” Inquiry, 55 (1): 17–32. Taylor, C. (1964), The Explanation of Behaviour, London: Routledge and Kegan Paul. Tonner, P. (2017), “Wittgenstein on Forms of Life: A Short Introduction,” E-LOGOS – Electronic Journal for Philosophy, 24 (1): 13–18. Tuomela, R. (1976), “Explanation and Understanding of Human Behaviour,” in J. Manninen and R. Tuomela (eds.), Essays on Explanation and Understanding: Studies in the Foundations of Humanities and Social Sciences, 183–205, Dordrecht: Reidel. Tuomela, R. (1995), The Importance of Us: A Philosophical Study of Basic Social Notions, Stanford, CA: Stanford University Press. Väyrynen, K. (2015), “Historialliset selitykset suomalaisessa 1900-luvun historianfilosofiassa: Grotenfelt, Salomaa ja von Wright,” in K. Väyrynen and J. Pulkkinen (eds.), Historianfilosofia, 307–30, Tampere: Vastapaino. Wellmer, A. (1979). “G. H. v. Wright über ‘Erklären’ und ‘Verstehen,” Philosophische Rundschau, 26 (1): 1–27. Winch, P. (1958), The Idea of a Social Science and its Relation to Philosophy, London: Routledge and Kegan Paul. Wittgenstein, L. ([1953] 1967), Philosophical Investigations, Oxford: Basil Blackwell. von Wright, G. H. (1971), Explanation and Understanding, London: Routledge and Kegan Paul.

228

The History of Understanding in Analytic Philosophy

von Wright, G. H. (1976a), “Determinism and the Study of Man,” in J. Manninen and R. Tuomela (eds.), Essays on Explanation and Understanding: Studies in the Foundations of Humanities and Social Sciences, 415–35, Dordrecht: Reidel. von Wright, G. H. (1976b), “Replies,” in J. Manninen and R. Tuomela (eds.), Essays on Explanation and Understanding: Studies in the Foundations of Humanities and Social Sciences, 371–413, Dordrecht: Reidel. von Wright, G. H. (1980), Freedom and Determination, Amsterdam: North-Holland. von Wright, G. H. (1981), “Explanation and Understanding of Action,” Revue Internationale de Philosophie, 135: 127–42. von Wright, G. H. (1989), “A Reply to My Critics,” in P. A. Schilpp and L. E. Hahn (eds.), The Philosophy of Georg Henrik von Wright, 731–887, La Salle, IL: Open Court. von Wright, G. H. (1993), “The Myth of Progress,” in The Tree of Knowledge and Other Essays, 202–28, Leiden: Brill.

CHAPTER TEN

Psychological Understanding TAMÁS DEMETER

To the memory of Adam Morton

10.1. INTRODUCTION Psychology is widely thought to have epistemic benefits. It is taken to be a field of inferences about a putative domain of facts pertaining to the causal processes that produce behavior. Knowledge of these facts is thought to be deployed in explanations and predictions about ourselves and others and, thereby, facilitates navigation in the social world.1 We might call this the widely shared “metarepresentational presupposition”: Psychology facilitates social interaction by providing knowledge of the internal mechanisms responsible for behavior. In this chapter, I will offer an alternative account of psychological discourse. To remove the spell of this presupposition, we might derive inspiration from Jaegwon Kim’s ([1989] 1993: 240, n. 4, 263, n. 46) passing remarks: he suggests, without spelling this out in any detail, that psychological propositions can be seen as facilitating noncausal understanding and evaluation of behavior. I will develop my account in this spirit. The core idea is this: psychological understanding is an affective process whose benefits are not of an epistemic but of a social nature. I will argue that by engaging in psychological discourse, we can do two things: resolve affective tensions that arise in the course of social interactions and provide

230

The History of Understanding in Analytic Philosophy

orientation in the social world. Psychological interpretations can achieve this by a framing effect: They posit evaluative concepts for interpreting behavior due to which psychological representations become affectively and motivationally significant. This significance does not arise from any explanatory or predictive power, but from the specific light such interpretations shed on the behavior so interpreted; that is, from the way psychological concepts frame events by embedding them in a network of affectively significant concepts. Affective resolution and motivational significance are thus the chief non-epistemic benefits of our interpretive practices in psychology. By characterizing our psychology as an evaluative discourse, I distinguish my proposal from others that treat psychology as a normative, prescriptive, regulative discourse (e.g., Kusch 1997, 1999: Part II; Morton 2003; McGeer 2007). In this sense, psychological platitudes—which some take to be empirical generalizations and others view as principles of rationality— prescribe the behavior to be followed in a given community, and insofar as they are observed, they play a causal role in the production of behavior. This is the property that allows us to rely on them in epistemic contexts, in explaining and predicting behavior—a property that mere evaluations do not possess. If it were possible to attach unequivocal evaluation to a given piece of behavior—that is, if there were any facts of the matter determining its evaluation—then it would be plausible to read evaluations as norms for behavior. But, as it will transpire, no such facts are available (see also Demeter 2009a, 2013). Therefore, evaluations cannot be norms for behavior, as there is no way to tell which behavior accords to a psychological “norm” and which one does not. However, to anticipate what follows, it can be argued that they can become norms for subsequent interpretations and evaluations. What I offer is the thus following: The main function of psychology is interpretative and evaluative. It does not directly serve the purpose of social navigation, which is enabled by more fundamental mechanisms (see, e.g., Hutto 2008; Dennett 2000; Bermúdez 2003: 27, 46–7, 2006). But it is through psychology alone that we can represent agents as persons—it allows us to relate to agents as persons and not as causes of behavioral patterns, physical objects, or animals, etc. The concept of a person is organically and intricately connected to the possibility of mental state ascription: if mental states cannot be ascribed to a specific agent, then that agent is not a person. Representing agents as persons requires coherent psychological narratives that bestow on them a relatively stable character and personality. Without coherence, the agent is seen as insane and as falling outside both the psychological discourse and the community. In this sense, a coherent psychological narrative is also a criterion of belonging to a community,

Psychological Understanding

231

and no moral evaluation is possible unless an agent is counted as a person within it.

10.2. INTERPRETATION Let us start from a simple case borrowed from a different context (taken from Kennett 2001: 3–4). Every day on his way home from work, John enters a nearby shop, buys cream cakes, and eats them. He does this regularly despite knowing that they do harm to his health: he is overweight, his doctor advises him to lose weight, and heart disease and diabetes are not unknown in his family, etc. There is thus a tension between what John knows and what he does; given what he knows, he should behave differently. Why then is he doing this? There are at least three explanations. We could say that John is an irresponsible person, who simply does not care about the risks involved in his behavior. Or we could say that he has a compulsive personality: while being aware of the risks involved, he just cannot fight his desire. Or we could also say that he is akratic, someone whose desire is not irresistible and who could therefore fight it, but who for some reason does not exercise this power. How can we decide which interpretation of John’s behavior to accept? We cannot rely on the behavior itself, as it is equally compatible with all three interpretations. John eats the cakes while agreeing to the proposition that they do harm to his health. We could ask him to pick the correct interpretation, but this would not settle the question either. His interests may influence his answer, and he might also be deceiving himself. The second interpretation exempts him from responsibility and, what is more, might even induce sympathy for him. The first one, while putting the blame on him, may also cast him in a favorable light—perhaps not in the case of irresponsible cake-eating, but, for example, if he were deliberately exceeding the speed limit: he may seem to be brave, at least to some people. Even John himself cannot settle the question, no matter how honest and truly committed he may be. Let us consider a bit more closely why this is so. Although there are criteria that can help us in deciding which of several conflicting interpretations to favor, they cannot help us in telling which one is true. The best they can offer are pragmatic considerations for acceptance. One of these criteria may be simplicity. Some of the possible systems of mental states by which an agent may be interpreted will be implausibly complex and can, therefore, be excluded in favor of more simple ones. But a simple system is not necessarily a true one, only one that is easier to handle. Besides, more often than not, we find that any type of behavior matches more than one equally simple

232

The History of Understanding in Analytic Philosophy

and plausible interpretation (Dennett 1991: 49). A better criterion might be to consider the degree to which a given interpretation is coherent with an overall interpretation of the agent derived from previously accepted interpretations of his behavior. But this does not go hand in hand with truth either. It only shows that we have arrived at the current interpretation against the background of previously accepted interpretations and thus understand new behavioral evidence in the light of previous interpretations: we simply use previous interpretations as criteria for the current one. But again, this is surely not the way to arrive at true representations of the agent’s internal states. John’s choice for interpreting his behavior depends on a variety of factors, including what he thinks and feels about himself; what kind of personality he thinks he has; and whether a narrative portrays his behavior as coherent (rational) and he can be content with it. This is the background against which John interprets his own behavior; his interpretation may rarely be explicit, but it can reveal itself from the character traits, beliefs, desires, moods, emotions, etc. that he ascribes to himself. John’s honest answer to the question about his motivation depends on how it coheres with the overall story he can accept about himself. These narratives are crucial in psychological practice. We tell them, typically in fragments, not only to ourselves but also to others. A version of this story could be called “the public image of ourselves” (Velleman 2006: 53), which may not be identical with the story we would tell us about ourselves, as the private image may have parts that we would not like to share with others, perhaps because they are intimate or compromising in nature. We also tell psychological narratives about others, with the aim of representing their behavior coherently. The success of this representational exercise gives us a kind of cognitive security in the social world, which thus appears to be intelligible. Beyond that, our narratives are useful in social interactions as they express how we relate to others and are instructive as to how we would behave if cooperative or competitive situations were to arise. The case is thus not, as, for example, Goldie (2000: 142) and Velleman (2006: 231) think, that the narratives render the person about whom they are told predictable; instead, they are expressive of the interpreter’s stance. Psychological narratives help us in navigating the social world by revealing a great deal about those who put them forward and, if they are taken to be reliable, also help us relate to those about whom the stories are told. At their core, such narratives are not true or false representations of the agents’ internal states—notwithstanding the declarative semantics of the sentences making up these narratives, which prima facie are about those very states. But despite how they seem, psychological propositions are not

Psychological Understanding

233

representations of the agents’ internal states. And this is reflected in the fact that there can be disagreements in interpretation even if the interpreters are epistemic peers, meaning their disagreement cannot be explained by their relative epistemic advantage or disadvantage. As Dennett points out: I see that there could be two different systems of belief attribution to an individual which differed substantially in what they attributed—even in yielding substantially different predictions of the individual’s future behavior—and yet where no deeper fact of the matter could establish that one was a description of the individual’s real beliefs and the other not. In other words, there could be two different, but equally real patterns discernible in the noisy world. The rival theorists would not even agree on which parts of the world were pattern and which noise, and yet nothing deeper would settle the issue. The choice of a pattern would indeed be up to the observer, a matter to be decided on idiosyncratic pragmatic grounds. (Dennett 1991: 49) In other words, there are no deeper facts that can make an interpretation true, there is no evidence independent of the psychological concepts deployed for their interpretations, and evidence emerges or disappears as we choose the concepts for interpretation. There is no way of stepping outside the conceptual framework of psychology and considering behavioral evidence from within it and still being true to the discourse. We can only obtain evidence that is alien to, and cannot be treated by, the conceptual resources of psychology. Evidence put forward in terms of bodily movements and neural mechanisms can contribute to understanding only if interpreted by psychological concepts; that is, if they find their place among psychological phenomena. The concepts chosen for interpretation deliver the evidence: they exclude some aspects of behavior as irrelevant “noise” while making visible other aspects by arranging them into “patterns.” But this suggests that there is no independent evidence that could single out a particular interpretation as true. It is the psychological concepts themselves that show some aspects of behavior as evidence—and not vice versa, as there is no independent evidence to support interpretations. The evidence supporting one interpretation may be entirely irrelevant if viewed from an altogether different angle, simply because it exploits different concepts in making sense of behavior and, therefore, creates different evidence to support itself. This does not mean, however, that there are no constraints on weaving psychological narratives. They facilitate understanding not by explaining internal mechanisms for producing behavior, but by conforming to the

234

The History of Understanding in Analytic Philosophy

conventions of psychological storytelling (see Demeter 2013). We understand these narratives because we are used to talking about persons in this way. Conventions set the limits of and specify the possibilities for interpretation. We acquire these conventions while taking part in the interpretative practices of the community as listeners, and subsequently producers, of psychological narratives for interpreting agents. It is not only everyday interactions that should be counted as relevant here. In this context, literature too has a specifically social function: it elaborates and spreads ideas about how people are working in various circumstances (see, e.g., Morton 1980: 7–8, and the classic essay by T. S. Eliot 1975). Taking part in these practices teaches us the paradigmatic uses of psychological terms and how to apply them in our own narratives (see Hutto 2007: 56–7; Goldie 2000: 33, 102–6). We grow into competent interpreters by learning to use paradigmatic narrative structures, thereby acquiring the conventions that govern their application. While learning the conventions of psychological storytelling is a precondition for interpretations, these conventions also induce some fundamental biases, as illustrated by the widely discussed phenomenon of fundamental attribution error (Ross 1977; see also Morton 2003: 151–3; Doris 2002: 93). There is a systematic distortion in social understanding: while putting great emphasis on the contributions of the agents involved, we underestimate the influence of those features that are independent of them. Our understanding of social situations is focused on persons, and these situations are understood as resulting from the latter’s motivations, actions, and character. The mastery of psychological concepts makes us sensitive to the aspects of social situations that can be interpreted by these conceptual resources and thus opens up specific ways of understanding while, at the same time, excluding others. Arguably, character traits are attributed by overemphasizing the relevance and frequency of past behaviors on the basis of earlier interpretations. Doris (2002: Ch. 3) and Alfano (2013) report several experiments in which the person-centered and character trait-centered perspective of our interpretations is revealed. For example, it is hard to account for the systematic correlation between helping behavior and smells: People are more likely to help in an environment that smells good. And from a psychological point of view, it is far from immediately transparent why people who have found a coin in a public telephone kiosk are more likely to help than those who did not. These experiments suggest that some features of social situations are motivationally relevant, but they cannot be easily placed in the psychological discourse. Although it is possible, it still sounds all too strange to say that good smells put people in a better mood and that this is why they are more

Psychological Understanding

235

inclined to help. It may be true in some rare cases, but it is far from a plausible explanation of the systematic correlation demonstrated by the experiments. This is an unconvincing fudge, because it brings a stable personality trait (helpfulness) under the influence of temporary mood fluctuations. Traits belong to a psychological category quite distinct from that of moods, and the transactions between them are not straightforward. This is how the conventions of psychological narratives set the limits of our interpretations: they put constraints on the situations in which they can facilitate understanding. At the same time, without these concepts, agents cannot be represented as persons; they thus open up, to use Hacking’s (1995: 239) happy phrase, ways of being a person. But the price to be paid for this is to sacrifice sensitivity to aspects of situations that cannot be integrated coherently into psychological narratives. Another bias encoded in psychological discourse is rationality understood as coherence. Psychological interpretation, as Davidson ([1974] 1980: 237) aptly emphasizes, is a business that commits us to finding significant consistency in the agent’s behavior. This produces both a constraint and a bias: Psychological interpretations cannot do their job without it, but given that we, as a matter of course, understand one another through psychological terms, we have a natural tendency to look for rationality in behavior. Coherence can be achieved if the internal logic of narratives is in concert with the conceptual connections that are characteristic of psychological discourse (see Demeter 2009a). This does not mean that the concepts initially applied to an interpretation of an agent lay down an a priori path to follow in the course of future interpretations. Psychological concepts are linked in diverse ways: they have connections in several directions, and interpretations can thus be developed in divergent ways while preserving consistency. But as long as we accept any given interpretation, it is also clear that any further interpretations must align with it. Therefore, if I derive a specific action of an agent as based on goodwill, then the future negative consequences of that action, if any, cannot be interpreted as intentional without revising the initial interpretation. Our current interpretation depends on previous interpretations, due to rationality, conceptual connections, the conventions of psychological storytelling, paradigmatic narrative structures, etc. And this produces an essentially personal bias in our interpretations. Peter Goldie’s specific claims about emotions are generally true of psychological interpretations: there are many ways in which the narrative of an emotional experience or of an emotion can get pieced together or filled in from a starting-point of limited information. Interpretation of the elements of the narrative may

236

The History of Understanding in Analytic Philosophy

require us to look far back into the aetiology of the person’s character, mood, upbringing, and historical and cultural situation. Furthermore, if we are adequately to interpret another’s emotion, that person’s selfinterpretation may also be relevant. […] [U]‌nderstanding and explaining is personal, normative, and holistic. (Goldie 2000: 187, 189) The practice of interpretation consists in weighing up narrative elements. This is not an arbitrary process in the sense that the interpreter chooses the narrative elements as he wishes. Any interpretation is composed against the background of one’s personal interpretative history, which also configures a psychological sensitivity: it makes us responsive to certain aspects of situations and agents, thereby inclining us, in specific cases, towards accepting some interpretations over others. Thus our interpretations are always and irredeemably personal, never objective. To some degree, it is, of course, possible to withdraw from this personal perspective and adopt that of someone else, but it is not possible to bracket it entirely: our respective responses are, at bottom, gut reactions whose cognitive control is limited. Besides configuring psychological sensitivity, personal interpretative history also defines the stock of narrative elements and improves and refines the conceptual resources at hand, thereby configuring an interpretative perspective. This process yields the stock of concepts with which the events of the social world—that is, events that are marked as such by our psychological sensitivity—are interpreted. Our interpretative perspective supplies the concepts available to the interpreter for representing an agent’s behavior as coherent (rational) in line with our psychological sensitivity. This coherence, just like Dennett’s patterns, is relative to the interpreter’s perspective. Talk of consistency or inconsistency simpliciter is meaningless, and inconsistency relative to one standard may be consistency relative to another. Accordingly, each person may exhibit a plethora of consistencies and inconsistencies, inasmuch as we may observe her behavior from a multitude of perspectives, and any run of behavior will be consistent to some perspective. (Doris 2002: 80) As there is no objective perspective from which personal perspectives, and thus the truth of the interpretations based on them, could be judged, there is no escape from personal bias either. The acceptance of a psychological interpretation depends not on some matters of fact, but on personal interpretative histories, on the experiences and the socialization of the interpreter.

Psychological Understanding

237

Peter Goldie (2000: 186) aptly illustrates, with an example from War and Peace, the fact that we interpret on the basis of previous interpretations. For the attentive reader, Prince Andrew’s behavior at Schöngrabern will make sense: the context makes clear why he attends personally to the withdrawal of the artillery in spite of the obvious risks involved and of the alternative options at his disposal. The narrative Tolstoy gives us about Prince Andrew configures the reader’s psychological sensitivity so that it is almost natural for her to expect something like this. This pre-history influences the sensitivity of the reader towards Prince Andrew’s behavior—and so configured, this sensitivity would have been violated if Prince Andrew had not done as he did. Tolstoy’s narration is tendentious in the sense that it portrays Prince Andrew and other characters by directing the reader’s attention to a particular pattern that has no alternatives. That is to say, novels in general submit ready-made interpretations of agents, in the context of which there is not too much room for interpretative maneuvering on the reader’s part, precisely because they configure psychological sensitivity in a very specific way. It is only half of the story to say that psychological narratives influence our psychological sensitivity; the other half is that they also express it. The current interpretation expresses the interpreter’s psychological sensitivity, configured by his personal interpretative history; that is, it expresses the relevant affective reactions induced by a particular situation. Psychological interpretation is, at its core, part of an expressive discourse, not a descriptive one. Adam Morton (1980: 17) hints at a possible construal of interpretative practice similar to the one proposed here when he points out that the practice of interpretation starts from imagining complexes of thoughts, emotions, potential actions, etc. that we cannot properly capture with words. We imagine the internal states of the agent we are interpreting and try to find words for them. In place of this approach, I propose to replace metarepresentational imagination with the expression of our affective reactions: while engaging in psychological discourse, instead of describing internal states of agents as we imagine them, we try to express affective states arising in us during the relevant social interactions, and we try to induce others to relate to the object of interpretation in some similar way. Our interpretations articulate, in a conventional way, how we feel about the interpreted agent or situation. Affective reactions lack conceptual structure. They can be taken to be embodied reactions or “somatic markers” of bodily processes.2 They have motivationally relevant properties: affective appraisals guide our navigation in the social world by alerting us, as Jenefer Robinson (2005: 183) says, to what is going on—but before cognition comes into the process. Their function

238

The History of Understanding in Analytic Philosophy

is to alert the agent very fast and automatically to whatever in the (external or internal) environment is of significance to the agent’s wellbeing, and to produce instantly a physiological state that readies the agent for appropriate action and signals to the others the state of the agent. […] Almost always—even in non-human animals—there is subsequent cognitive monitoring of the situation. Indeed, in human beings, these affective appraisals are usually both preceded and succeeded by cognitive appraisals. Human beings not only respond automatically to whatever seems to be important to their well-being; they also have the cognitive ability to monitor and modify these responses. (Robinson 2005: 197) One way of doing this is through psychological narratives, as they express affective reactions and thereby configure future reactions. This is a continuous process during social interactions: Physiological reactions arise from psychological narratives and social situations, and their affectively appraised aspects tune our psychological sensitivity (see Robinson 2005: 157). It follows that not even the general acceptance of an interpretation means that it is at least a potentially true representation of an agent’s mental world, only that it adequately expresses our relevant affective reactions. In these cases, we acknowledge that it is appropriate to apply psychological concepts to the agents involved, which in turn means that the conventional standards of psychological interpretation are duly followed. Besides, by accepting an interpretation, we agree that we would all tell a similar story about the situation and the agents involved; that is, that the narrative is acceptable from our interpretative perspective and that we feel similarly about the situation. This consensus is obviously vulnerable even among competent interpreters. It is always an open question if an interpretation is correct, if an interpretative perspective (i.e., the concepts and narrative structures applied) is appropriate, if a psychological sensitivity is well configured (if it is sensitive to what it should be), or, at a more specific level, if everything has been taken into account or if any major omissions have been made in the interpretation, etc. But these doubts cannot be put forward from an objective point of view, only from another interpretative perspective. Psychology is therefore different from scientific (and also folk) theories, which operate with descriptions of phenomena that are at least adequate for predictive and explanatory purposes. Psychological discourse operates in a radically different way—it is much more like an art than a science. As Quine (1960: 219, 222, 248) aptly puts it, the discourse on attitudes is essentially a “dramatic idiom” with which we make “dramatic projections.” This insight is especially revealing because psychological discourse, similarly to drama, focuses our attention on agents, their relations, and their actions.

Psychological Understanding

239

The analogy to literary fiction is further strengthened when Quine talks about the psychological idiom as a “myth” whose ontological commitments do not pose a problem as long as this dramatic element is kept in mind and provided that this idiom is not used “seriously”; for example, in scientific or fact-stating contexts. Following Quine, our psychological idiom can thus be treated as a fictional discourse whose production more closely resembles literary composition than theory construction (see also Demeter 2013). But this should be constrained: psychology is also a discourse about a putative domain of investigation (i.e., the mental). And in this respect it differs from literary fiction: in literature, there is no way to refine or revise the author’s claims, and neither do we ask for reasons why specific statements are accepted. Given the fictional character of the discourse, psychological platitudes cannot be empirical generalizations, lawlike propositions, etc. that would support explanations and predictions. But they can be conventions stabilizing the conceptual connections in the network of psychological concepts. They can be rules governing the correct application of concepts, revealing which attributions can be combined with each other, and which ones can entail— ceteris paribus, or in this case without further attributions—which other ones. Psychological platitudes are conventions that stabilize conceptual connections; they thus set the tracks on which psychological narratives can be advanced without violating the requirements of coherence. Due to their conceptual connections, attitude ascriptions are holistic, and therefore agents can only be interpreted by systems of attitudes whose number is in principle infinite: we can combine any attitude or other mental state with any other one provided that we are ready to ascribe further mental states that would lend coherence to the interpretation or to sacrifice its plausibility. The current possibilities of interpretation are to some extent limited practically by any of the earlier interpretations that we accept at a given time: coherence may require extreme complexity in narratives if contradictory elements are to be combined in them. In such cases, we are inclined to withdraw assent from some interpretations accepted earlier and replace them with new ones. And this indeed happens sometimes, given that increasing complexity places constraints on our ability to deal effectively with situations; and psychological practice does not easily tolerate high complexity. While it is true that any type of behavior can be rationalized, this never happens without a context: we acquire the psychological idiom as a narrative practice (see Hutto 2008), and behavior is always interpreted against the background of our interpretative perspective as configured by our communities’ interpretative standards, the conventions of the psychological idiom, and our personal interpretative history.

240

The History of Understanding in Analytic Philosophy

This is the sense in which psychological narratives set limits—not for behavior, which does not depend directly on psychological interpretations (though these narratives play a causal role by triggering affective reactions), but for further interpretations of agents and situations. By ascribing a specific mental state to someone, we do not offer for acceptance a true or false proposition about her; rather, we commit ourselves to an interpretation and thereby circumscribe what further interpretations we are willing to make about her future behavior. We outline a narrative space in which her behavior will be positioned by our standards. Similarly, in first-person cases, an interpretation commits us to a range of possible further interpretations whose coherence can be maintained or violated in exactly the same way as in third- and second-person cases and whose affective content may motivate behavior. Self-interpretations also have their roots in affective states: Their acceptance depends on a self-image we can live with and on the interpretative perspective that generates the least dissonance in our self-images and lets us live in peace with ourselves. While this perspective is normative, it is not directly so for behavior—it does not prescribe how we ought to behave, but how to interpret our own behavior in order to be in concert with our previous selves as represented by our self-images. This is not always possible. Simon Blackburn reminds us of a not unusual situation: I decide that one course of action is the one to take. But when the time comes, I find myself doing another. The natural response is to suppose that, perhaps as a surprise to me and others, I actually cared more about some aspect of the situation than I realized. If I behaved better than I intended, then it gives me an opportunity to learn something about myself. My sense of honour, pride, altruism, prudence, or whatever is stronger than I had realized. If I behave worse, I can similarly learn that my greed, fears, lusts, imprudence, or whatever are stronger than I had realized. To make the decision intelligible is to look back and recognize that my emotions and dispositions were not quite as I had taken them to be. (Blackburn 1998: 191) In this respect, I put special emphasis on affective states: if we feel that in a given case we did not do what we should have done, and if this feeling is strong enough, then it may induce us to revise our self-image, which can happen without our confessing it publicly. But it should be noted that “what has been done” is again a question of interpretation and that the need for revision arises from interpretations as expressions of affective reactions to the situation, not from facts of the matter—exactly the way it is in the interpretation of others.

Psychological Understanding

241

10.3. UNDERSTANDING Let us take a closer look at what it means to understand psychological narratives and, through them, agents. To begin with, speaking the psychological idiom facilitates the understanding of behavior through the application of familiar concepts embedded in interpretative practice and, thereby, connecting it to phenomena previously interpreted in a similar manner. By these concepts, we reveal or, more precisely, create patterns in behavior (recall Dennett 1991: 49). The application of concepts is stabilized by community-wide conventions, where conformity sheds familiar light on the behavior interpreted. Children grow into a community that uses, as a matter of course, the psychological idiom while talking about a specific class of agents—namely, those taken to be persons—and the situations in which they find themselves. Narratives told in terms of these concepts allow for a synoptic view of agents and their behavior that might otherwise seem unconnected. They enable an understanding by creating a unity that is not present in behavior if it is described as a product of causal processes—as the examples in Alfano (2013), Doris (2002: Ch. 3), and Velleman (2006) show. Interpretations arrange pieces of behavior into an order which highlights their relevance in a wider context. The success of an interpretation depends partly on its ability to represent, as a unified whole, the interconnections of a person’s character, mental states, and behavior and the situation in which they are relevant. Viewed from this angle, pieces of behavior acquire meaning through their teleological significance in the network of these interconnections. This significance can be specified with reference to the conclusions towards which our narratives gravitate. This could be an image of a person’s character, an outline of future interpretations, the possible significance of a specific situation for other events, etc. These are typically virtual conclusions that are suggested, but not necessitated or made unequivocal, by our narratives; they can be drawn in various ways, are revisable, and bear the mark of the interpreter’s personality. The point towards which our narratives gravitate is thus elusive. This is a feature of psychological narratives that has no parallel in causal histories: causes and effects can follow one another in an endless chain without ever coming to a conclusion (see Velleman 2003: 10–11), while causal histories cannot lend teleological significance to events. This is the framework in which psychological interpretations make sense of behavior: with an attention to purpose and a commitment to finding a high degree of coherence. This is the background against which the meaning and rationality of behavior can be shown and life can be seen as meaningful.

242

The History of Understanding in Analytic Philosophy

This way, virtual conclusions can be present without being explicit. They can be explained by “implicature”: psychological propositions that are not in themselves informative or relevant can be charged with teleological significance by recognizing their narrative implications through paradigmatic narrative structures. Partial interpretations imply virtual conclusions due to conceptual connections and the biases discussed above. Implications can be drawn in several directions from a partial interpretation, depending on which affects one takes to be expressed in it. This network of implications specifies the position of the partial interpretation in the overall interpretation of a situation or a person, and its acceptance depends on the affective reactions it induces. Narratives so produced are not useful devices in everyday social interaction, and they need not be. In everyday interactions, we rarely need explanations of isolated pieces of behavior (see, e.g., Bermúdez 2003), because social behavior typically follows habitual patterns whose understanding does not pose a special problem. Consequently, our psychological practice should not be construed as building up from singular pieces of explanation and prediction of behavior—despite philosophical interest being traditionally focused on them. Contemporary debates typically start from a picture of motivation (frequently called Humean) in which explanations emerge from an ascription of a belief-desire pair. Behind this analytical practice seems to lie the conviction that our psychological practice can be understood from these allegedly basic cases. In real life, however, we are much more interested in overall images of personality and character as well as in revealing patterns in agents’ behavior and explaining situations and agents that seem to be somehow problematic to us. Our psychological practice should not be understood as bottom-up—that is, as unfolding from simple belief-desire explanations—but as top-down—that is, descending from complex narrative structures (see Hutto 2007, 2008). If the task of hermeneutics rests, as Gadamer (1993: 295) likes to say, “on a polarity of familiarity and strangeness,” then psychology is a practical hermeneutic device of social interactions: it is a tool for weaving narratives that make the social world familiar to us, which is why coherent narratives are valuable to us in themselves. In this sense, psychological practice is a narrative practice, and the basic cases of psychological understanding are not individual predictions and explanations, but more complex narrative structures. These provide the raw material for interpreting behavior that does not fit smoothly with the habitual ways in which things are done. We need them where we feel uneasy with the situation, or feel some tension, or where we do not see clearly, etc. These affective reactions create a need for and motivate psychological interpretation. What we need in these cases is an

Psychological Understanding

243

affective resolution of the situation:3 psychological interpretation replaces “strangeness” with “familiarity.” In this sense, the value of psychological understanding lies in its collective therapeutic capacity. Habermas (1971: 262–4) considers psychoanalysis a cooperative effort of the patient and the psychotherapist, aimed at the elaboration of a narrative whose success does not depend on empirical adequacy but on therapeutic efficacy. This is congruent with everyday psychological practice. The success of psychological narratives is not measured by the adequacy with which they represent the internal mechanisms underlying behavior, but by how effectively they can treat these situations. An interpretation is successful if it makes the social situation familiar by resolving affective tensions. Offering an interpretation for acceptance to others is an attempt at inducing affective states. This communication is successful if the listener understands how the interpreter feels about the situation, agent, etc. that is being interpreted. An example of orientation, this process is successful if the listener accepts the narrative as expressing his own feelings about the situation; that is, if his affective reactions to the situation become similar. Disagreements as to the correct interpretation are about how one should feel about the objects of interpretation; the process of communication refines our affective reactions and our psychological sensitivity. Communicating affective states is a vulnerable process, as understanding can easily be imperfect and partial. Therefore, psychological understanding is a matter of degree: it can be deepened by refining the interpretation, but by its very nature, there are no criteria for correct understanding, and it is not clear what would count as a correct recognition of the interpreter’s feelings. It is easier to judge, however, what counts as unsuccessful communication: These are the cases in which we experience imaginative resistance, which can come in several possible varieties (see Gendler 2006; Walton 2006). One can dispute the admissibility of a given narrative element on the grounds that it does not conform to the conventions of psychological storytelling, or violates the interconnections of narrative elements that we acknowledge, thereby making the story incomprehensible. Or we may acknowledge the possibility of a given interpretation while at the same time considering the feelings that they express to be inappropriate in the given case, thereby deeming the interpretation to be implausible. This kind of resistance can be explained by reference to the difference of interpretative perspectives: typically, we do not accept interpretations that portray the social world in ways radically different from how we understand it. Psychological understanding does not entail knowledge and is not an epistemic state.4 On the one hand, psychological understanding is transparent (see Zagzebski 2001: 246–7): while I can know something without knowing

244

The History of Understanding in Analytic Philosophy

that I know it (e.g., some logical consequence of my knowledge that I have not thought of yet, but if I was asked, I would give the correct answer in a straightforward manner; or something that I have not tried yet, but if I was to try it, I would know it), I cannot understand something without understanding that I understand it. On the other hand, understanding, as opposed to knowledge, does not presuppose external criteria; more precisely, it does not presuppose truth. I cannot know that bad weather was caused by a witch’s curse, as this is not true. But I can understand weather conditions as being so determined if my interest is satisfied by this explanation; for example, because I am used to “explanations” offered in these terms. The truth of psychological interpretations is similarly indifferent with respect to their understanding. Understanding does not require truth in the same sense that knowledge does, so there is no need for truthmakers here. It only requires some affective state, a somatic marker of a process that we express by saying something like “I see.” Psychological understanding is in itself of value to us, but this value is not epistemic. This is not the case with the explanation of weather conditions, as this is an epistemic, rather than a hermeneutic, activity. Therefore, understanding weather conditions as resulting from the intervention of witches is problematic because the discourse about weather conditions is a fact-stating one that is supposed to have epistemic virtues. Psychology is not a discourse of this kind (see Demeter 2009a, 2009b). The process of psychological understanding is thus as follows: a narrative put forward in the psychological idiom expresses, by conventional means, the interpreter’s affective reactions to the object of interpretation. The interpreter thereby communicates her feelings, configuring the listeners’ psychological sensitivity to orient them in navigating the social world by framing the events being interpreted. This framing is made possible by the conventional conceptual resources and the paradigmatic narrative structures of psychology. By acquiring the conventions of the psychological idiom, we learn, on the one hand, how to express the affective reactions that are appropriate in social situations so as to be effective in our social interactions and, on the other, how to understand others’ expressions. Paradigmatic narrative structures entail affective reactions that we learn during the process of socialization: we learn which characters in tales to sympathize with or have feelings against, and later, from everyday interaction and reading literature, we learn which comments express the socially relevant reactions of the interpreter, etc. We thus acquire an expressive code, step by step, with which to encode our own affective reactions and decode those of others. This knowledge can be exploited for the purposes of manipulation: sometimes, psychological narratives may not express the interpreter’s affective reactions, but only serve the purpose of dishonest manipulation. The know-how of

Psychological Understanding

245

psychological narration allows us to influence others’ sensibilities, and it can be used in various ways. This can be illustrated by several examples, such as John’s case, jealousy, problems of self-interpretation, personal conflicts, or our unsettled interpersonal relations. For elaborate cases, it is worth turning to literary examples. It is here, and not in the simplistic action explanations favored by philosophers, that the complexity of everyday interactions, the proper context of psychology, can be found. But one must be careful in this regard. Plays are typically a more appropriate source of illustrations than novels, as the former portrays interpersonal situations without any manifest psychological interpretation. Novels frequently present a picture of social relations that is already interpreted to a significant extent. This is the case with Goldie’s example taken from War and Peace. It is not entirely up to the reader to interpret Prince Andrew on the basis of the reader’s sensitivities and perception of his behavior. The reader only has access to Prince Andrew through Tolstoy’s narration, which suggests a ready-made interpretation. By the time the reader arrives at this point, her sensitivity is fundamentally influenced by Tolstoy’s previous interpretations. And even though some freedom of interpretation still remains, it cannot be compared with that offered by a play whose rules of composition leave much less room for configuring sensibilities—in other words, in drama there is no room for detailed descriptions of character. There are exceptions, of course. David Lodge (1992: 26) is probably right in saying that Henry James is a master of manipulating points of view, and this may explain why he is one of the favorite novelists of many a philosopher. Let us therefore take a closer look at The Golden Bowl.5 The plot is not too complex. Adam Verver, a rich American businessman, travels around Europe with his daughter, Maggie, collecting works of art for the museum he wants to build back home. A friend, Fanny Assingham, introduces Maggie to an impoverished Italian prince, Amerigo, and eventually they get married. Before their engagement, the prince had a liaison with the beautiful but also poor Charlotte Stant, from whom he parted because of their financial circumstances. Although Charlotte is an old friend of Fanny, and also a school friend of Maggie, neither of them informs Maggie about this relationship. Following the wedding, Maggie urges her father to marry again. He then heeds her advice and marries Charlotte. Nevertheless, Maggie and her father spend most of their time together and with Maggie’s recently born child, thereby gradually bringing Amerigo and Charlotte closer, causing them to renew their liaison at last. Although Maggie discovers this, she tries to sort things out while keeping everything under the surface. She succeeds; Adam and Charlotte return to America, while Maggie and Amerigo begin a new life.

246

The History of Understanding in Analytic Philosophy

The story has no unexpected turns and is not untrue to life. Its sophistication springs from how the characters, their relationships, and the indeterminacy of their psychology are portrayed. For example, the most obvious question considered by Fanny in the novel is about the significance of Charlotte’s coming to London, where the wedding is about to take place. The problem is not that one must choose the true interpretation from among rival ones in light of the evidence at hand, but that the meaning to be revealed thus is simply not there—not even for Charlotte, despite her belief that her intentions explain everything. The meaning of her arrival depends on her and others’ reactions, expectations, intentions, notions of appropriate behavior, etc.; that is, on several previous and future interpretations. The inconsistencies of Fanny’s interpretations and behavior can be traced back to her ambivalent feelings, as a result of which she does not commit herself to any possible interpretation. The general features of this characteristically Jamesian setting are summarized by Robert Pippin: Of course, to say that many James characters do not seem to know what they think does not mean that they are somehow unaware of their own mental states. It is to note that what I am in fact thinking consciously about my own attitudes, or what I think about others, might but need not have much relevance to what opinions and beliefs may rightly be (eventually) attributed to me about such matters. I may never have “actually thought” the opinions I do have, and what I consciously tell myself I believe may not at all be what I believe. This might still seem to suggest that the problem is some sort of epistemological opacity (there is such an opinion and I am blocked somehow from acknowledging it), or am in self-denial, or it is unconscious, and so on. I might be said to reveal what I really believe by what I do, not by what I say or consciously think, for example. But … James’s position seems much more radical than such “opacity” positions. The question of what belief I actually hold, or what my true motivation was, and so on, is simply not the sort of thing that could be said to be “there” at all, however epistemologically refined our insight might become. It is available retrospectively in some sense only because “there,” in the future, is the only place it could “be.” (Pippin 2000: 64–5) The metaphysical opacity of mental states, lacking the possibility of true interpretation, is clearly perceived in a key scene of the novel. Charlotte asks for Amerigo’s help in choosing a suitable present for Maggie without letting her know. It is not entirely clear, not even for Charlotte herself, what her purpose is in doing so, apart from having the prince in her company and

Psychological Understanding

247

having something said between them—something utterly vague. What does this scene mean? Charlotte may want to compromise Amerigo; or she may want to get closer to the prince by this common secret; or to initiate the renewal of the relationship; to manipulate him by pricking his conscience, as he is about to get married for financial reasons; or to remind him who he is actually in love with. Alternatively, the situation can also be interpreted as a heroic attempt to gain some kind of recognition for their former liaison or even as a symbolic act of closing the affair. And it could also be argued that this scene has no special significance and simply constitutes an empty conversation in an uneasy situation that may have arisen from a bad idea. Which one to choose then? The answer depends on how one understands this event, which in turn depends on previous interpretations, none of which corresponds to facts better than the others. The choice will be made depending on which interpretation provides the reader with a synoptic view of the events from her personal vantage point; which one corresponds best with the purposes the interpreter ascribes to the characters; which one seems the most appropriate given the interpreter’s affective reactions to the situations encountered; which one makes her feel that she understands the situation and the behavior of the characters in them; which one fits best with the overarching interpretation of the events portrayed in the novel, etc. In order to see the incompatibility of various interpretative possibilities, we do not need to turn to the intricacies of Jamesian psychology. Morton (1980: 47–50) introduces the interpretation of Aida’s behavior in Verdi’s opera in five varied styles, some of which are compatible and some incompatible with one another. For example, Aida can be understood in a “romantic” style, with reference to passions and character traits; in a “rational” style, with reference to the calculation of propositional attitudes; in a “social” style, with reference to interpersonal relations and the conflict or congruence of the agents’ plans; in a “symbolic” style, with reference to the symbolic content with which the agents interpret the situation; or in a “primitive” style, with reference to her purposes and internal processes as actions. These styles can be used as general strategies in interpreting agents while they emphasize different aspects of the situations and frame what is to be understood in each of these different cases. Their partial incompatibility arises from the properties of the categories of interpretation. For example, “symbolic” and “primitive” interpretations are incompatible for Morton, as they represent mental states in a radically different manner, while “romantic” and “rational” interpretations ascribe different roles to character traits in understanding behavior. These conflicts can be explained by pointing out that the core concepts used in these divergent styles have divergent conceptual connections. The concepts used in the initial interpretation specify different

248

The History of Understanding in Analytic Philosophy

ranges of concepts and ways in which they can be used. What is common to them, however, is that they define, as Morton says, the “moral universe” in which social situations can be represented.

10.4. EVALUATION The possibility of moral evaluation is based on psychological interpretation. Hume rightly says that ’Tis evident, that when we praise any actions, we regard only the motives that produc’d them, and consider the actions as signs or indications of certain principles in the mind and temper. The external performance has no merit. We must look within to find the moral quality. This we cannot do directly; and therefore fix our attention on actions, as on external signs. But these actions are still consider’d as signs; and the ultimate object of our praise and approbation is the motive, that produc’d them. (Hume 2002: § 3.2.1.2) It is only through psychological concepts that agents can be represented as persons moved by evaluable motives. This is not possible in the physical— or any scientific—description of agents. The description of the physical/ physiological causal chain producing behavior is an impersonal one. It represents the agent as if events were happening to him rather than him doing something—it does not represent his behavior as his actions. When behavior is represented as resulting from mental states, its real causes are not actually represented. This does not prevent us, however, from representing agents as persons, as suitable objects of psychological understanding and moral evaluation: these representations express affective reactions that are relevant in both senses. For example, character traits or reasons for action are elements in representations connected to motivationally relevant feelings that influence how we behave towards those to whom we ascribe them. Psychological narratives are suitable means of moral orientation: they convey implicit moral evaluations and thus configure our moral sense. In the process of socialization, we acquire paradigmatic narrative structures that induce affective responses that are stabilized by repeated encounters. Some sources of motivation, if ascribed, shed a favorable and others an unfavorable light on the agent or behavior. Due to these interpretations, we understand an interpreter’s reactions and the object of interpretation. Affective reactions as conveyed by psychological narratives are partly responsible for how we navigate in social situations, and normative ethics is precisely about how to regulate

Psychological Understanding

249

moral feelings and interpersonal behavior appropriately (see Frankfurt 1988: 80). Representing an agent (or anything else, for that matter) as a person entails the acknowledgement that it is appropriate to apply to her the categories of freedom and responsibility—categories that have no counterparts in scientific dictionaries. Agents can be seen as free if they can be interpreted psychologically: otherwise, their behavior is at most indeterminate. Without psychological ascriptions, behavior can be explained physiologically in causal terms, but from this, the contribution a person makes to the situation cannot determined—on the contrary, causal explanations reveal determining factors and thereby exempt people from responsibility. This is thus incompatible with the idea of personal decisions that follow from freedom and are presupposed by responsibility. Responsibility is judged based on the psychological narrative and on actual behavior. John’s and Charlotte’s responsibility in the above examples can be judged differently depending on how one understands their psychology. Alongside their hermeneutic purpose, psychological narratives thus also serve evaluative purposes. Due to the seemingly causal connections between reasons and actions (see Demeter 2009a), persons can be treated as having deliberative capacities and as the sources of their actions—precisely because conceptual connections take some philosophical effort to be revealed. But such “causal” connections are of a different kind from those of physiological explanations. In psychological narratives, we cite special “causes”—character traits, motivations, etc.—as pertaining to persons who have authority over them: they can decide on them, influence them, and are therefore responsible for them. Because of this authority over (at least some of) their internal states, these persons are responsible also for actions, which are understood as springing from these “causes.” These “causal connections,” as opposed to physiological ones, do not obliterate responsibility—on the contrary, they are presupposed by responsibility because they belong to the realm of a person’s special authority. Beyond that, moral evaluation does not presuppose freedom in any robust metaphysical sense. Our behavior may be subject to natural laws without our knowledge, but the narratives we tell to and about ourselves must be based on the illusion of free will—without it, there is no way of understanding behavior as autonomous action. Will, as Daniel Wegner (2002: 325–7) argues, can be seen as an affective shadow of some physiological processes (a somatic marker), which reminds us that some events are attributable to us. If we interpret a piece of behavior as an intentional action, we ascribe this to the agent so interpreted. This ascription plays a central role in judging moral responsibility and in deciding who deserves what; that is, in the context of praise and punishment, but not in an exclusively social context—bad

250

The History of Understanding in Analytic Philosophy

conscience is a typical example of moral self-punishment. By growing into a competent user of the psychological idiom, we learn which narratives to interpret as expressions of bad conscience and which ones can incite this feeling—we learn the relevant paradigmatic narrative structures. One’s bad feelings about one’s past behavior can be interpreted by psychological concepts and understood as bad conscience. And vice versa, narratives about our behavior told by others, and our own self-interpretation, can also incite similar bad feelings. But we cannot have bad conscience about events that we do not view as attributable to us. Since understanding behavior and the ascription of responsibility go hand in hand, psychological interpretation and moral evaluation spring from the same sources. Kathleen Wilkes (1998: 155) rightly points out that an episodic life—whose events are not connected by the concept of responsibility and are not understood as appropriate objects of praise or blame, and in which emotions are mere feelings without a proper history, etc.—cannot be moral. Moral evaluation presupposes that we treat ourselves as fairly stable intentional systems in order to represent ourselves as such by psychological narratives. These narratives bestow moral features on agents through the attribution of reasons, motivation, character, mood, etc.—that is, essentially evaluable properties. As Morton (2003: 43–5) points out, in understanding actions, we frequently rely not on propositional attitudes but on character traits, virtues or imperfections that have moral overtones. Concepts deployed in psychological understanding are typically evaluative concepts or at least have evaluative implications. Our psychological sensitivity is not merely the basis of our moral sensitivity, but is moral in itself. Psychological narratives endow agents with evaluative properties that we ascribe because we are sensitive to certain aspects of the surrounding world; namely, to the contributions of agents similar to us (see Mackie 1977: 31–4). These agents may not even have a grounding property, or set of properties, in common that warrants the ascription of a given evaluative property. What is common to them is that they are ascribed a given property that cannot be identified independently of the act of evaluation; for example, by means of science. Goldie (2000: 30) adduces the property of being “dangerous” as an example: “bulls with long horns, dogs with rabies, exposed electric fires, icy roads, strangers with sweets, certain ideas, Lord Byron” have nothing in common that would ground this evaluative property that they share. The possession conditions of evaluative concepts cannot be based on our ability to recognize common features in things subsumed under them—instead, they may be grounded in some sort of response; for example, in affective reactions. And arguably, this is what we do in our psychological interpretations and moral evaluations: we subsume

Psychological Understanding

251

fuzzy affective reactions under concepts, thereby structuring them and making them communicable. The intricate connections of interpretation and evaluation reveal why it is problematic to consider certain kinds of circumstances in moral evaluation and how the problem of moral luck arises. The fundamental attribution bias, which I have derived from the natural tendency to apply psychological concepts in our attempts to understand social situations, is especially important here. As most of the work of moral evaluation is done by psychological interpretations, circumstances that are not interpretable psychologically cannot be taken into account effectively. From this angle, Bernard Williams ([1976] 1982: 39) is right in suggesting that “luck” could be integrated into our morality only if it is thoroughly revised. This is a natural corollary of the agent-centered focus of psychological interpretation, and it also explains why consequentialist ethics entail a revision of everyday moral discourse (see Doris 2002: 129): The practice of moral evaluation attends more to the psychological roots of action than to its consequences. The essentially evaluative character of psychological ascriptions clarifies one aspect of why “mental terms stand or fall together” (Lewis [1972] 1991: 209). Evaluative concepts are defined contrastively, and the contrasts specifying their content are not binary, but obtain in several different directions and to various degrees. As Charles Taylor puts it, No one can have the idea what courage is unless he knows what cowardice is, just as no one can have a notion of “red”, say, without some other colour terms with which it contrasts. It is essential to both “red” and “courage” that we understand with what they are contrasted. And of course with evaluative terms, as with colour terms, the contrast may not just be with one other, but with several. And indeed, refining an evaluative vocabulary by introducing new terms would alter the sense of the existing terms, even as it would with our colour vocabulary. (Taylor [1977] 1985: 19) If the network of the various contrasts that are constitutive in the meaning of psychological terms changes, their meaning changes too. By introducing new psychological concepts, the concept of a person also changes, and so do our moral evaluations. Just think of how reference to the unconscious (in Freud’s sense) can transform the concept of an autonomous, deliberating, responsible, etc. person; or of how the concepts of desire, motivation, reason, etc. can thus also be reshaped. The time for conceptual transformations comes when we feel our established concepts are incapable of expressing our relevant affects or when we find that arbitrarily introduced concepts can serve well the purposes of expression and understanding. This is a two-way

252

The History of Understanding in Analytic Philosophy

process: transformations of psychological concepts both imply and arise from changes in morality and social imagery. Seen from the angle of contrastive concepts, several mysterious features of psychology disappear. For example, this explains how psychological interpretations bring coherence to behavior through contrasts and implications; it also explains why action explanations are always contrastive: they explain why an agent did what he did as opposed to doing something else (see Lewis 1986: 229). While actions understood as free events can be explained only contrastively, other indeterminate, but not free, events (like radioactive decay) cannot be contrastively explained—only the causal chain resulting in the given event can be described. Contrastivity is also in close harmony with the need for psychological interpretation that arises in unfamiliar situations—where we need to know why someone did what he did—as opposed to pursuing a more familiar action. The evaluative aspect of psychology is often overlooked in theories of action and motivation, frequently called Humean (see Smith 1987) for their focus on rational calculation with propositional attitudes. If it is admitted that evaluative aspects cannot be separated from psychological concepts, then it is doubtful whether they can be used in purely descriptive contexts if no further provisos are added. Such a proviso could be that psychological terms refer to natural kinds that we do not have sound enough knowledge of and whose nature is supposed to be revealed by further empirical research (see Lycan 1988: 31–2). Our psychology may provide initially contingent, and thus possibly false, descriptions of entities based on which we can start an inquiry, but it has no special importance beyond that. In this case, the conceptual resources of psychology are not privileged in the explanation of human behavior and cognitive functioning. Philip Pettit ([1995] 2002: 229–8) points out the problems arising from such a technical interpretation of mental terms in relation to economics. He sees a tension between the rich moral and quasi-moral idiom of our psychology and the way psychological concepts are used in economic theories. The anthropology of economics is different from that of our psychological idiom, given that they portray motivations differently: while we understand ourselves in the evaluative psychological idiom, economists work with theories, frequently successfully, in which this evaluative aspect has no role to play, as they engage in exclusively rational calculation without moral overtones. This tension, however, is only superficial: in the context of social science, psychological concepts do not play the role they play in the natural interactions of everyday life—they are used as technical terms. And this is precisely the proviso we need to add. We have to realize that in technical contexts, the use of psychological terms is restricted to a few aspects of their

Psychological Understanding

253

complex meaning and that those aspects depend on the theory that exploits them. Consequently, the terms of our evaluative discourse are turned into technical terms with purely descriptive meaning—and we have to bear in mind that their meaning thus changes.

10.5. COPENHAGEN Let me conclude with a discussion of Michael Frayn’s Copenhagen as an ideal illustration for the present account of our psychological idiom. The situation it portrays cries out for psychological interpretation. The meaning of the protagonist’s behavior is opaque and thus unsettling for the other characters, and the moral evaluation of the situation depends on its psychological interpretation. In Frayn’s (2009: 136) own interpretation, the play is about “the epistemology of intention,” with the following main lesson: What people say about their own motives and intentions … is always subject to question—as subject to question as what anybody else says about them. Thoughts and intentions, even one’s own—perhaps one’s own most of all—remain shifting and elusive. There is not one single thought or intention of any sort that can ever be precisely established. (Frayn 2009: 99) Picking this play as an illustration is, I admit, a bit tendentious, as its author’s intentions are somewhat close to the view suggested here. But it will be enlightening to see how Frayn, as I believe, misreads the philosophical lesson of his play precisely because he implicitly endorses the metarepresentational presupposition. It is only with this in the background that he can raise the question of whether we have epistemic access to the relevant facts that could substantiate psychological knowledge claims. The event presented in the play is the infamous meeting of Heisenberg and Bohr in Copenhagen in September 1941. Later, they would remember and interpret the meeting and its significance differently. The play stages an otherworldly conversation between them by reconstructing what had actually been said. The primary aim is to reveal Heisenberg’s intentions in visiting Bohr and to review the prospects of moral evaluation of Heisenberg’s behavior. The characters’ “spirits” reconstruct the meeting in several inconclusive versions, thereby demonstrating the indeterminacy of interpretation.6 The central question of the play is about Heisenberg’s motivation: why did he, the head of the German atomic bomb project, go to Nazi-occupied Denmark to visit Bohr? The spectrum of possible and incompatible

254

The History of Understanding in Analytic Philosophy

interpretations is wide enough. Did he want to get approval, or even help, from Bohr in building an atomic bomb; or did he want to gain inspiration from him for solving a related problem? Or was it a visit undertaken to demonstrate his sympathy for the Jewish Bohr? Or did he want to send a message to the allied powers via Bohr that he would not build a bomb and that they thus should not do so either? Or did he want to find out if there was an allied atomic bomb project in order to be able to decide, in light of this information, how to proceed with the German project? Did he want to convince Bohr that the German cause was the right one and that it would be victorious; or did he want to persuade Bohr that it would be in his own interest to build better connections with the German authorities in Denmark? The line of possible interpretations could be continued and made more complicated by taking into account questions such as whether Heisenberg intended to build a bomb or why he did not try to calculate the critical mass of fissionable uranium needed for a bomb. It may seem as if the available facts do not settle which interpretation is true (see “Postscript” and “Post-Postscript” in Frayn 2009). This brings us back to Dennett (1991: 49): taking the visit and its significance as the starting point, Heisenberg’s behavior, past and future, exhibits several patterns depending on how it is interpreted. But its significance cannot be specified on the basis of facts. This could be done if some interpretation of his past or future behavior had already been accepted. In this case, the significance of the visit must be interpreted coherently with what is accepted—once it has been accepted. Heisenberg’s intentions in visiting Bohr can only be specified against the background of further intentions that are not matters of fact either—just as we have seen with James’s character above. For example, revealing whether or not Heisenberg wanted to build a bomb might shed some light on why he went to Copenhagen. But again, this is not settled by facts. On the one hand, he was the head of the German atomic bomb project, and he made some progress in this regard. On the other hand, the results he delivered, the available documents, and his moderate enthusiasm for the project, which is known from the testimony of others (i.e., from interpretations), suggest that he was not eager to build a bomb—but if this was the case, why not? Perhaps he thought that it was impossible (i.e., he was ignorant of important parts of the science) or perhaps he did not want to build it for moral, political, or other considerations. The answer depends on the issue of the calculation of critical mass. Heisenberg did not even try to calculate it, in spite of his habit of calculating everything he could—as Bohr tells us in the play. This poses the question: Why not? Because, as Heisenberg notes in the play, he did not want to build a bomb, meaning this knowledge was not at all interesting to him? Or because

Psychological Understanding

255

he judged the whole enterprise to be hopeless and would thus not even try? Or because he could not see the very problems that had to be solved? Whatever answer we choose, there will be facts supporting it. But it will not be conclusive; it will only reveal a pattern, a possible arrangement of behavioral evidence, thereby filling it with meaning and evaluation. There are no facts to access that could settle the question. Frayn suggests that it may have been unclear even to Heisenberg himself what his actual motives were (and this is a feature he shares with many of James’s characters— and ourselves, of course). But it is clear from the above that whatever interpretation we choose could be opposed by another interpretation revealing an alternative pattern of behavior. Even if we presuppose that Heisenberg’s self-interpretation was honest and that he was not deceiving himself, it cannot be independent of his previous self-interpretations, as they configured his psychological sensitivity and interpretative perspective towards himself. The same is true about Bohr’s interpretation and that of his wife, Margarethe. They relied on their previous interpretations of Heisenberg and their feelings about him and the broader situation (i.e., about their common past, the occupation, Germany, anti-Semitism, the war, etc.) which had configured their psychological sensitivity. Before this background, they (especially Margarethe) could not distance themselves from the thought that Heisenberg wanted to talk about fission or some related problem about its practical use. This initial stance made them respond to the situation in a certain way, which led Bohr to cut off the conversation—and thus their friendship. There seems to be no impersonal perspective from which a true psychological interpretation of Heisenberg’s behavior can be advanced. The moral evaluation of the situation depends on the motivational narrative that we accept about the background of Heisenberg’s behavior. In some interpretations, his behavior could be seen as positive; for example, if he is interpreted as worrying about the safety of his old friend or as wanting to ensure that the bomb was not being built by anyone. Other interpretations would portray him neutrally; that is, if he wanted to find out whether there was an allied atomic bomb project in order to decide what to do about it. The case becomes morally problematic if he is interpreted as attempting to convince Bohr about Germany’s just cause. And he could be treated with moral contempt if he sought to exploit Bohr in some way for his atomic bomb project. The evaluation depends on how we react affectively to various reconstructions and interpretative possibilities, and it is conditioned by our judgment of the given historical period and Heisenberg’s less than unequivocal role in it, etc. And the case is similar with Bohr. Depending on how Heisenberg’s role is understood, one could condemn Bohr for

256

The History of Understanding in Analytic Philosophy

expressing impatience or unjustified moral indignation or interpretative prejudices. To sum up, the acceptance of a psychological narrative hangs on how we feel about Heisenberg’s and Bohr’s behavior and their perspective on historic events, etc., but mostly on the interpretative alternatives offered in the play. Our affective reactions are configured by our interpretative perspective and psychological sensitivity, which make certain aspects of the characters’ behavior salient and offer us conceptual resources for representing these by means of a coherent narrative. Both depend on our personal interpretative history: they supply us with the relevant concepts and narrative structures for expressing our affective reactions. By expressing these reactions, we represent the object of interpretation as a person, as a bearer of properties relevant for psychological understanding cum moral evaluation. And that is how psychological narratives play a role in our social life. They do not communicate knowledge about the internal world of the agents about whom they are told, but convey how the interpreter feels about those being interpreted. Interpretations proposed for acceptance aim at configuring affective sensibilities of others so as to enable us to feel similarly about the object of interpretation. Affective sensitivity is both the basis and the target of psychological narratives: they are expressed, accepted, or rejected on this basis, and they tune how we feel about the components of the social world. Our narratives reflect the way we feel about others and ourselves, and via our psychological sensitivity, they influence our navigation in the social world. Interpretation and evaluation concur in this process.

10.6. ACKNOWLEDGMENTS At various stages of its preparation, this chapter has benefited from comments by Gábor Forrai, Jane Heal, Martin Kusch, Miklós Márton, Adam Morton, Csaba Pléh, and János Tőzsér. Of course, they are not responsible for the outcome, but their advice is much appreciated. This chapter contributes to the research program of the MTA Lendület Morals and Science Resarch Group.

NOTES 1 I take this to be the common core of the various approaces called “theory theory,” “simluation theory,” and “interpretivism” as well as of personality psychology and several forms of pschyotherapy. 2 I borrow the term from Damasio (1994), albeit with a slightly different content. For him, emotions are somatic markers, while I side with Goldie (2000) in claiming that emotions are at least partly conceptual and have narrative structure.

Psychological Understanding

257

Therefore, they cannot be purely bodily processes in the sense in which ‘somatic marker’ is used in the present context. 3 In the case of narratives, I have in mind a kind of resolution similar to Velleman’s (2003: 13) “emotional resolution,” though I am inclined to agree with Goldie (2000: 13–14, 103) that emotions have a narrative structure, which is not characteristic of affects. In this case, I am thus talking about affective reactions and not about emotions. 4 As such, it can be neatly contrasted with accounts of “explanatory understanding” as an epistemic state; see de Regt (2017). 5 The Wings of the Dove or The Portrait of a Lady would also qualify. Note, however, that philosophical interest in James is primarily focused on his moral contents (see, e.g., Pippin 2000; Nussbaum 1983; Jones 1983) and that psychological lessons are treated merely as secondary. 6 At first sight, the material may not seem a good fit for a play. The plot needs important background information to be shared with the audience, which pushes the exposition in a more novel-like direction and away from the natural dynamics of dialog. But a closer look reveals that the material can indeed be conveyed in the form of a play, as a novel would easily tend towards more detailed psychological (self-)interpretations and a more intensive configuration of our psychological sensitivity. As I suggested above, novels typically offer ready-made interpretations and leave less room for everyday indeterminacies of interpretation. From the participants’ points of view, the structure of social interaction resembles that of a play more than that of a novel: We rarely describe circumstances or evaluate others in great detail, and seldom offer elaborate internal monologues to ourselves. The social world spins more quickly than the world of the novel.

REFERENCES Alfano, M. (2013), Character as Moral Fiction, Cambridge: Cambridge University Press. Bermúdez, J. L. (2003), “The Domain of Folk Psychology,” in A. O’Hear (ed.), Minds and Persons, 25–48, Cambridge: Cambridge University Press. Bermúdez, J. L. (2006), Philosophy of Psychology: A Contemporary Introduction, London: Routledge. Blackburn, S. (1998), Ruling Passions, Oxford: Clarendon. Damasio, A. (1994), Descartes’ Error: Emotion, Reason and the Human Brain, New York: Putnam. Davidson, D. ([1974] 1980), “Psychology as Philosophy,” in Essays on Actions and Events, 229–44, Oxford: Clarendon. Demeter, T. (2009a), “Folk Psychology Is Not a Metarepresentational Device,” European Journal of Analytic Philosophy, 5: 19–38. Demeter, T. (2009b), “Two Kinds of Mental Realism,” Journal for General Philosophy of Science, 40: 59–71. Demeter, T. (2013), “Mental Fictionalism: The Very Idea,” Monist, 96: 483–504. Dennett, D. (1991), “Real Patterns,” Journal of Philosophy, 88: 27–51.

258

The History of Understanding in Analytic Philosophy

Dennett, D. (2000), “Making Tools for Thinking,” in D. Sperber (ed.), Metarepresentations: A Multidisciplinary Perspective, 17–29, New York: Oxford University Press. de Regt, H. W. (2017), Understanding Scientific Understanding, New York: Oxford University Press. Doris, J. M. (2002), Lack of Character, Cambridge: Cambridge University Press. Eliot, T. S. (1975), “The Social Function of Poetry,” in On Poetry and Poets, 15–25, New York: Farrar, Straus & Giroux. Frankfurt, H. G. (1988), The Importance of What We Care About, Cambridge: Cambridge University Press. Frayn, M. (2009), Copenhagen, London: Methuen. Gadamer, H. G. (1993), Truth and Method, 2nd ed., New York: Continuum. Gendler, T. S. (2006), “Imaginative Resistance Revisited,” in S. Nichols (ed.), The Architecture of Imagination: New Essays on Pretence, Possibility, and Fiction, 147–73, Oxford: Clarendon. Goldie, P. (2000), Emotions: A Philosophical Exploration, Oxford: Clarendon. Habermas, J. (1971), Knowledge and Human Interest, Boston: Beacon. Hacking, I. (1995), Rewriting the Soul, Princeton: Princeton University Press. Hume, D. (2002), A Treatise of Human Nature, Oxford: Oxford University Press. Hutto, D. D. (2007), “The Narrative Practice Hypothesis: Origins and Applications of Folk Psychology,” in D. D. Hutto (ed.), Narrative and Understanding Persons, 43–68, Cambridge: Cambridge University Press. Hutto, D. D. (2008), Folk Psychological Narratives, Cambridge, MA: MIT Press. Jones, P. (1983), “Philosophy, Interpretation, and The Golden Bowl,” in A. P. Griffiths (ed.), Philosophy and Literature, 211–28, Cambridge: Cambridge University Press. Kennett, J. (2001), Agency and Responsibility: A Common-sense Moral Psychology, Oxford: Clarendon. Kim, J. ([1989] 1993), “Mechanism, Purpose, and Explanatory Exclusion,” in Supervenience and Mind: Selected Philosophical Essays, 237–64, Cambridge: Cambridge University Press. Kusch, M. (1997), “The Sociophilosophy of Folk Psychology,” Studies in History and Philosophy of Science, 28: 1–25. Kusch, M. (1999), Psychological Knowledge: A Social History and Philosophy, London: Routledge. Lewis, D. ([1972] 1991), “Psychophysical and Theoretical Identifications,” in D. M. Rosenthal (ed.), The Nature of Mind, 204–10. New York: Oxford University Press. Lewis, D. (1986), “Causal Explanation,” in Philosophical Papers, Vol. 2., 214–40, Oxford: Oxford University Press. Lodge, D. (1992), The Art of Fiction, London: Penguin. Lycan, W. (1988), Judgement and Justification, Cambridge: Cambridge University Press. McGeer, V. (2007), “The Regulative Dimension of Folk Psychology,” in D. D. Hutto and M. Radcliffe (eds.), 137–56, Folk Psychology Re-assessed, Dordrecht: Springer. Mackie, J. L. (1977), Ethics: Inventing Right and Wrong, London: Penguin. Morton, A. (1980), Frames of Mind: Constraints on the Common-sense Conception of the Mental, Oxford: Clarendon. Morton, A. (2003), The Importance of Being Understood: Folk Psychology as Ethics, London: Routledge.

Psychological Understanding

259

Nussbaum, M. (1983), “Flawed Crystals: James’s The Golden Bowl and Literature as Moral Philosophy,” New Literary History, 15 (1): 25–50. Pettit, P. ([1995] 2002), “The Virtual Reality of homo economicus,” in Rules, Reasons, and Norms: Selected Essays, 222–44, Oxford: Clarendon. Pippin, R. B. (2000), Henry James and Modern Moral Life, Cambridge: Cambridge University Press. Quine, W. V. O. (1960), Word and Object, Cambridge, MA: MIT Press. Robinson, J. (2005), Deeper than Reason: Emotion and Its Role in Literature, Music, and Art, Oxford: Clarendon. Ross, L. (1977), “The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process,” Advances in Experimental Social Psychology, 10: 173–220. Smith, M. (1987), “The Humean Theory of Motivation,” Mind, 96: 36–61. Taylor, C. ([1977] 1985), “What Is Human Agency?” in Human Agency and Language: Philosophical Papers 1, 15–44, Cambridge: Cambridge University Press. Velleman, J. D. (2003), “Narrative Explanation,” Philosophical Review, 112: 1–25. Velleman, J. D. (2006), Self to Self: Selected Essays, Cambridge: Cambridge University Press. Walton, K. (2006), “On the (So-called) Puzzle of Imaginative Resistance,” in S. Nichols (ed.), The Architecture of Imagination: New Essays on Pretence, Possibility, and Fiction, 137–48, Oxford: Clarendon. Wegner, D. M. (2002), The Illusion of Conscious Will, Cambridge, MA: MIT Press. Wilkes, K. (1998), “ΓΝΩΘΙ ΣΕΑΥΤΟΝ (Know Thyself),” Journal of Consciousness Studies, 5: 153–65. Williams, B. ([1976] 1982), “Moral Luck,” in Moral Luck: Philosophical Papers 1973–1980, 20–39, Cambridge: Cambridge University Press. Zagzebski, L. (2001), “Recovering Understanding,” in M. Steup (ed.), Knowledge, Truth, and Duty: Essays on Epistemic Justification, Responsibility, and Virtue, 235–52, Oxford: Oxford University Press.

260

CHAPTER ELEVEN

Verstehen Redux: Understanding in Contemporary Epistemology AARON PRESTON

11.1. INTRODUCTION The past several decades have seen a new interest in “understanding” among a small group of analytic epistemologists. This is a significant and, in my view, promising departure from long-standing norms of analytic epistemology, which for most of the twentieth century focused attention narrowly on the concept of knowledge and closely related issues like justification. The new interest in understanding grew out of dissatisfaction with those norms. On the one hand, decades of work on Gettier problems, lottery paradoxes, and the like seemed to be going nowhere, giving mainstream analytic epistemology the appearance of a degenerating research program. On the other, there was a dawning realization among some epistemologists that, historically, philosophers had valued other cognitive states, like understanding and wisdom, as much as or more than knowledge and that they had done so for good reasons. In fact, Linda Zagzebski argues that since the time of Plato, philosophical thought has cycled between the two epistemic poles of certainty and understanding. In eras given to skeptical worries, philosophers

262

The History of Understanding in Analytic Philosophy

have focused on certainty and justification. But in eras less exercised by skepticism, philosophers have focused on understanding and the related issue of explanation. “Understanding and certainty,” she says, are two epistemic values each of which has enjoyed pride of place for long periods in the history of epistemology, but rarely, if ever, at the same time. […] Usually whichever one of the two concepts dominated was the one connected with the concept of knowledge, so Plato comes very close to identifying knowledge with understanding, while Descartes comes very close to identifying knowledge with certainty. (Zagebski [2001] 2020: 59) In this essay, I aim first to add to this picture by showing that not only epistemic issues like skepticism but also metaphysical and metaphilosophical issues have contributed to this historical pattern. Second, I use the expanded picture to explain (i) the nineteenth-century advent of Verstehen as a term of art, (ii) the analytic tradition’s lack of interest in understanding, and (iii) the revival of interest in understanding in recent analytic epistemology. I conclude that the pro-Verstehen movement of the nineteenth century and the pro-understanding movement of the twenty-first century have been motivated by similar concerns pertaining to the epistemic strictures of scientism, strictures which bear a complex relationship to skepticism. But first, what is understanding?

11.2.  THE NATURE OF UNDERSTANDING As with knowledge, the exact nature of understanding is a disputed matter.1 However, there seems to be broad agreement among those now working to rehabilitate the notion that understanding is a cognitive grasp of the interrelationships of the parts of complex systems, a grasp which, ceteris paribus, enables one to explain the system and/or to act competently if not skillfully with regard to it. This grasp is holistic: understanding is not reducible to a conjunction of otherwise discrete states of knowledge. As John Greco puts it: “one can know individual or isolated facts about a subject matter, but understanding seems to come in larger packages. Understanding ‘hangs together’ in ways that knowledge need not” (2014: 287). What’s more, although propositional knowledge may play a role in (some forms of) understanding, understanding is not reducible to propositional knowledge as the latter is usually understood. As one leading advocate of understanding, Jonathan Kvanvig, observes, there is a “difference in focus when understanding is before our minds rather than knowledge”:

VERSTEHEN REDUX

263

When we think about the nature of understanding, what is foremost in our minds are the ways in which pieces of information are connected with each other. To understand is to grasp the variety of such connections. It involves seeing explanatory connections, being aware of the probabilistic interrelationships, and apprehending the logical implications of the information in question. There is, of course, an element of factivity to the notion of understanding, just as there is with the notion of knowledge. But when we move past the alethic aspect of both notions, our attention turns to diverse paths. When the question is whether one knows, the issues that are foremost in our minds are issues about evidence, reliability, reasons for belief, and, perhaps most importantly, non-accidentality regarding the connection between our grounds for belief and the truth of the belief. When the question is whether one has understanding, the issues that are foremost in our minds are issues about the extent of our grasp of the structural relationships (e.g., logical, probabilistic, and explanatory relationships) between the central items of information regarding which the question of understanding arises. (Kvanvig 2017: 96–7) Zagzebski also characterizes understanding as a cognitive grasp of structural features of reality. However, she emphasizes that unlike knowledge on the prevailing “information model” ([2001] 2020: 72), the objects of understanding are often nonpropositional in nature: “understanding is a state of grasping nonpropositional structures of reality such as the structure of an automobile, the structure of a work of art or literature, or the structure of a human mind” (Zagzebski 2020: 3). Zagzebski acknowledges that understanding can be directed on propositions or “items of information,” but for her this is not essential to understanding. The general notion of understanding as a cognitive grasp of structural features of reality which (ceteris paribus) facilitates explanation of and competent interaction with its objects guides us in the remainder of the essay.

11.3.  THE CERTAINTY VERSUS UNDERSTANDING CYCLE As noted, Zagzebski ([2001] 2020) argues that the history of philosophy is characterized by a pattern of oscillation between the two epistemic poles of certainty and understanding, apparently determined by the extent to which the philosophy of a given era was preoccupied with the threat of skepticism. This strikes me as basically correct, but there is more to the story. Specifically, skepticism does not rise and fall in a vacuum,2 but is explained

264

The History of Understanding in Analytic Philosophy

by other, correlated historical factors. Some of these are social or cultural in nature, as when the Reformation’s challenge to the authority of the Catholic Church contributed to the rise of skepticism in the early modern period.3 But other relevant factors are ideational in nature, and they often include other philosophical positions whose logical implications point toward or away from skepticism. For example, the theology of the Reformation was deeply influenced by the via moderna of late medieval nominalism, which (as we shall see) provided independent ideational grounds for a shift toward skepticism relative to prevailing ideational norms under the via antiqua. The full matrix of relevant explanatory factors for any significant historical shift will usually be quite complex, so any manageable account will have to be selective. In what follows, I focus on interconnected networks of views in epistemology, metaphysics, and metaphilosophy as providing a more complete explanation of the changing attitudes toward understanding at key junctures in the history of philosophy, especially in connection with the origin of Verstehen in the “technical” sense and the origin of the analytic tradition. Traditionally, philosophy has embraced what I have called “meta­ philosophical eudaimonism” (ME; see Preston 2019b and forthcoming). On ME, philosophy is seen as a cognitive-behavioral project of moral formation whose overarching aim is both to inform us of and to form us in the good life, the life of human flourishing, or eudaimonia. As Nussbaum (1994), Hadot (1995, 2002), Sellars (2003), and others have shown, from very early on, philosophy was understood as a way of life involving practices aimed not only at the production of theory but also the shaping of one’s character in light of the theory produced.4 Thus, philosophy in “the great tradition” beginning with Socrates aimed to provide wisdom for living, grounded in a deep understanding of human nature in relation to the larger orders (social, cosmic, etc.) that humans inhabit. It is primarily in this sense that the tradition responds to the Delphic injunction to “know thyself.” In Plato’s Philebus (48c–e), Socrates implies that “knowing oneself ” is a matter of knowing the true nature of the psyche and its virtues; and as we see throughout the Platonic corpus, but perhaps most clearly and completely in the Republic, this is a task which ultimately requires a well-developed picture not only of our mental capacities along with their proper functions and virtues but also of the relations among the virtuous person, the human social world, and the cosmos, understood at the most fundamental level. Plato thus established as the overarching cognitive task of philosophy the use of reason to arrive at a comprehensive “worldview” which (i) locates human nature in the context of the larger orders it inhabits and (ii) orients it toward the good.

VERSTEHEN REDUX

265

Clearly, the type of cognitive state required to grasp a “worldview” of this sort, and to make use of it to order one’s life, fits better with Kvanvig’s and Zagzebski’s characterizations of understanding than with any standard characterization of knowledge. Little wonder, then, that Plato was more interested in understanding than in mere knowledge. As many scholars of ancient Greek philosophy have observed—and as contemporary advocates of understanding have been keen to point out—the word “epistēmē” as used by Plato and Aristotle, which is usually translated as “knowledge,” is better translated as “understanding.”5 Understanding is therefore an essential cognitive aim for ME; and although an embrace of ME is far from necessary to appreciate the value of understanding, there is a suggestive tendency among those contemporary epistemologists who have taken an interest in understanding to do so out of considerations that fit naturally with ME. ME was the mainstream position in ancient and medieval philosophy, but in the modern period, one sees commitment to ME beginning to wane. One reason for this, I believe, concerns the relation between late medieval nominalism and skepticism. Put simply, nominalism’s rejection of universals was motivated in large part by anti-essentialism in service of a voluntarist conception of God. Voluntarism insists that divine power and freedom are unrestricted by most of what we would normally regard as necessary facts or truths, such as the truths of logic or of morality. The voluntarist answers Euthyphro’s dilemma by saying that the holy is so only because God loves it; it is not the case that God loves it because it is holy in itself. By extension, the good, the right, and the rational are determined by the divine will, which is elevated over other elements of the divine nature, including reason and goodness. On this view, there is no compelling reason to suppose that God could not or would not perform actions contrary to reason and to morality. This voluntarist conception of God is the most likely source of inspiration for Descartes’s skepticism-inducing omnipotent deceiver, his genium malignum (see Gillespie 1995: Ch. 1). It was certainly one potent source of the skeptical worries which turned early modern philosophers’ attention to questions about certainty and justification in so pronounced a manner that the period is often characterized as making epistemology “first philosophy.” Additionally, nominalism and its corollaries undermined traditional views about knowledge and reason. Since the time of Plato, knowledge had been regarded as a matter of grasping universals, and reason was seen as the faculty which made this possible, in part by tracing purportedly necessary connections among universals. Denying the existence of universals required new theories of reason and knowledge grounded in our apprehension of particulars, and such theories were slow to gain traction, thereby adding to the skeptical worries generated by voluntarism.

266

The History of Understanding in Analytic Philosophy

Of course, these are two sides of the same nominalist coin. Through them, philosophy entered a centuries-long period in which questions of certainty and justification, as well as various answers to those questions, would dominate the mainstream of philosophical thought, to the neglect of understanding. In what follows, I focus on the impact of one such answer: scientism. Scientism is the view that the methods of the natural sciences are the only sources of justification sufficient to generate knowledge—that, as Bertrand Russell ([1935] 1997, 243) once put it, “whatever knowledge is attainable must be attained by scientific methods.” Scientism’s main attraction, historically, was the promise of achieving certainty, or something near enough. Of course, this is now widely understood to be a naïvely overoptimistic view of what the sciences can achieve. Nonetheless, this naïve view played a powerful role in the ascent of the natural sciences to positions of supreme epistemic authority in Western culture. As I see it, the respective origins of Verstehen and of the analytic tradition embody opposite reactions to the rise of scientism, with the former rejecting and resisting and the latter accepting and capitulating to the scientistic tendencies of the late modern period. In the next two sections, I explain how this is so.

11.4.  VERSTEHEN As a term of art, Verstehen owes its existence to a sense that the modes of explanation characteristic of the natural sciences (Erklären) are incapable of providing an adequate account of the psychosocial dimensions of human life. This task requires a form of cognition which grasps the motives and meanings of human actions in terms of agency, purpose, and value. Verstehen, the ordinary German word for “understanding,” was adopted as a technical name for this form of cognition by historian Johann Droysen in the late 1850s.6 It was further developed by Wilhelm Dilthey, who refined and elaborated the concept over the following half century in his quest to make Verstehen the defining mark of the Geisteswissenschaften. Thanks to Dilthey’s advocacy, the concept was kept alive into the twentieth century, when it was revised, extended, and restricted in a variety of ways by a number of thinkers.7 Consistent with its neglect of understanding in the general sense, the technical concept of Verstehen has garnered little attention from the analytic mainstream. Apart from some critical engagement from the logical positivists in their pursuit of unified science, it has for the most part been relegated to peripheral areas of the tradition, such as the philosophy of history (e.g., Dray 1957) and the philosophy of the social sciences (e.g., Winch 1958). This lack of engagement may be due in part to the fact that Dilthey’s Verstehen

VERSTEHEN REDUX

267

has been widely misunderstood as merely or mainly a form of “empathetic intuition.” This type of misreading is often attributed to the positivists,8 but it was shared by Gadamer and Habermas (see Harrington 2001). Some of Dilthey’s contemporaries (e.g., Windelband and Rickert) developed a similar misunderstanding, thinking that Dilthey’s Verstehen had only to do with insight into the “inner experience” of others, empathetic or otherwise. Such an array of distinguished misinterpreters suggests that Dilthey was less than sufficiently clear in expressing his views on Verstehen. But it seems that an additional source of the misunderstanding is a failure to properly contextualize Dilthey’s thought about Verstehen against the backdrop of Droysen’s.9 Consequently, I focus here on the original conception of Verstehen developed by Droysen. Verstehen in the technical sense makes its initial appearance as the centerpiece of Droysen’s epistemology of history. Droysen proposed Verstehen as an alternative to the nomothetic model of explanation favored in the natural sciences of his day and, for several centuries after Newton, widely idealized as a general model for knowledge as such. The sociohistorical result of this idealization is what I have elsewhere called the “drive toward scientism” in modern thought (see Preston 2007: Ch. 6). Perhaps the most famous advocate of applying the nomothetic model in history was the Englishman Henry Buckle, whose 1857 History of Civilization in England ([1857] 2011) is remembered as an exemplar of this “positivistic” approach. However, as Fredrick Beiser (2011: n. 7) shows,10 the nomothetic approach had already taken root in German historiography in the generations before Buckle. Although Droysen famously deployed the concept of Verstehen in a critical review of Buckle’s book, it was primarily in opposition to the scientism of other German historians that Droysen developed it in the first place. As Droysen saw it, history was neither like the philosophy of his day, which sought to know by deriving truths from first principles, nor like natural science, which sought to explain by subsuming individual cases under general laws. Rather, history seeks to understand contingent individuals in all their particularity through a form of cognitive apprehension akin to understanding a text or a speaker. For this reason, Verstehen is described as an interpretive or hermeneutic phenomenon, but for Droysen it is less an inferential process or stepwise method than a form of immediate cognition. Droysen recognizes four hermeneutic dimensions individually necessary and jointly sufficient for historical Verstehen, modeled on Aristotle’s four causes. First, “pragmatic interpretation” focuses on the efficient causes of historical events. It is here that the nomothetic model of explanation can contribute to historical understanding. But understanding requires more. It requires, second, the

268

The History of Understanding in Analytic Philosophy

“interpretation of conditions,” which concerns the physical, social, and moral conditions that made an historical event possible. Droysen saw these conditions as analogous to Aristotle’s material causes. Third, “psychological interpretation” seeks to understand the final causes of historical events in terms of the aims of historical actors. Finally, the “interpretation of ideas” seeks to understand the beliefs, ideas, and ideals that guided historical actors and served as the formal causes of historical events. Only when we take account of all of these factors do we have Verstehen of a historical event (cf. Beiser 2011: 301).11 Already we can see that Verstehen fits the general characterization of understanding given in Section 11.2. Additionally, this way of characterizing Verstehen comports well with certain contemporary accounts of understanding which make it out to be not merely structural knowledge, but knowledge focused on causal relations (Grimm 2014) or dependence relations (Greco 2014) among the elements of structural systems. Droysen’s Verstehen also has a normative focus that marks it as a close relative of the kind of understanding required for ME, although Droysen and Verstehen theorists generally shrink from the kind of evaluative comparisons needed to enact ME. For Droysen, the hermeneutic endeavor begins with something empirically given, which stands as a sign of something hidden from empirical view. An encounter with language begins with a sensible sign in speech or writing, but to be understood it must first be recognized as an outward expression of the inner life of the speaker or writer. Likewise, the historian begins with artifacts of various sorts, from monuments to texts, taken as expressions of the inner lives of the individuals who produced them. The inner life will include thoughts, beliefs, desires, valuations, volitions, and the like, understood as parts of an organic whole: the self. Holism about experience and the self is a central commitment of early Verstehen theorists, and it is here that we find the greatest genuine disagreement with those, including some positivists, who embraced a Humean/atomistic view of experience. Although the inner lives of others are not presented to consciousness in the same manner as one’s own, Droysen insists that they are somehow given in and through their outward expressions, that the interpretive process is not a matter of reasoning by analogy or of inference to the best explanation.12 At the same time, Droysen insists that the elements of the inner life can be properly understood only by locating them in a larger social context of shared values and meanings, of linguistic and moral norms. Following Hegel, he calls this “the ethical world.” The ethical world “comprises all those institutions, values, practices and traditions in which the individual participates,” and is, for Droysen, “the basic unit of history, the framework within which all intersubjective understanding takes place” (Beiser

VERSTEHEN REDUX

269

2011: 302). Ultimately, then, Verstehen is the cognitive result of locating human actions and artifacts within their ethical worlds. Hence, “the fundamental difference between the natural sciences and history … is that between causal and normative explanation” (Beiser 2011: 315). ME requires a similar form of understanding insofar as its project is to evaluate different “ethical worlds,” actual or merely possible, and their suitability for directing human life toward eudaimonia. One should not be thrown by Beiser’s suggestion that Verstehen is connected to normative “explanation.” The familiar opposition between “understanding” and “explanation” in discussions of Verstehen becomes misleading if we forget that the “explanation” in question is a very specific type of explanation, modeled ultimately on Newtonian physics. As most contemporary discussions of understanding reveal, understanding in the general sense is conceptually connected to explanation in the general sense. It is therefore not surprising that Droysen’s narrower species of understanding is associated with certain forms of explanation. In fact, as we have seen, Droysen’s Verstehen accepts the nomothetic explanations of the sciences as part, but not the whole, of a complete, Verstehen-generating or -generated explanation of human behavior. Consistent with contemporary statements characterizing understanding as a grasp of complex structure, Droysen’s Verstehen is a matter of seeing how the broad range of phenomena constituting the ethical world, the inner lives of persons, and their outward expressions fit together so as to explain human actions, interactions, institutions, artifacts, etc. As Rudolph Makkreel puts it in a description of Dilthey’s Verstehen: Verstehen … stands for a deliberate process that finds the proper context to relate others and their objectifications [i.e., expressions of thought, purpose, value, etc., in forms such as language, art, cultural institutions, etc.] to what is already familiar to us. It is a reflective mode of inquiry that provides the framework for more specific explanations, whether causal or rational. (Makkreel 2020, my emphasis) Dilthey himself acquired the concept of Verstehen directly from Droysen’s 1857 Berlin lectures (Maclean 1982: 348, n. 6). Although he spent the next half century refining the concept in various ways, it is not clear that he intended to fundamentally alter the character of Verstehen as established by Droysen. To the contrary, most of the differences between Droysen and Dilthey seem to arise from the latter’s attempts to further develop and clarify certain features of the Droysenian concept to meet new challenges pertaining to the nature and knowability of “inner” or “lived” experience

270

The History of Understanding in Analytic Philosophy

and related matters pertaining to the nature of psychology as a science. But, in the end, Verstehen remains for Dilthey “the procedure from which we know something psychic from signs given externally to the senses,” and in this sense “the interpretation of the state of the soul from the context of its entire life under the condition of its milieu” (quoted in Beiser 2011: 347–8).

11.5.  THE DEMISE OF VERSTEHEN AND THE RISE OF ANALYTIC PHILOSOPHY13 As a form of cognitive illumination focused on the interior dimensions of human agency in relation to the domain of value, Droysen’s Verstehen harkens back to Socrates’ primordial clarification of the nature of mentalistic explanation, as contrasted with the materialistic explanations favored by certain pre-Socratic naturalists: human behavior, at least, requires explanation in terms of mind or intellect (nous), and the intellect always considers options in terms of value, of “what is best” (Phaedo: 96a–99c). This is precisely what Droysen was insisting on under the heading of Verstehen; namely, that the human world requires explanation in terms of the noetic and normative features of human experience, motivation, and action. Of course, what initially gripped Socrates’ imagination was the possibility that the whole of reality, and not just the human world, might be subject to mentalistic explanation. This possibility has captivated the Western tradition for most of its history, and it has proven more contentious and less plausible than Droysen’s more restricted claim.14 Consequently, although Socrates and Plato established the mentalistic framework as the default position of the Western intellectual tradition for over two millennia, views contesting it are also a fixture of the tradition. As Plato observed in the Sophist (246a–b), Western thought was marked from the beginning by “a battle like that of the gods and the giants,” with the latter contending that “that alone exists which can be touched and handled” and the former contending that “real existence consists of certain Ideas which are conceived by the mind and have no body”—although in point of historical fact, it was the Mind or minds containing the Ideas, and not the Ideas themselves, that were more often viewed as the fundamental realities by those on this side of the debate. The rise of modern science was a turning point in this battle of gods and giants: for the first time since Plato’s era, the “giants” had at their disposal a tool with which they would, in the span of several centuries, make prima facie credible attempts to “drag down everything from the heavens and the invisible realm to earth” (Sophist: 246a). In many cases, these attempts succeeded and the success proved salutary: they advanced human knowledge and improved human life. But in a few crucial areas pertaining to the personal

VERSTEHEN REDUX

271

and moral dimensions of human life, they have had the opposite effect.15 Droysen’s advocacy for Verstehen was an attempt to hold some ground for the mentalistic framework in these crucial areas in the face of scientism’s advance. Clearly, his gambit failed. Why? There are two key factors which worked in tandem to sideline Verstehen in Droysen’s sense. The first has to do with the tendencies toward subjectivism and historicist relativism in Droysen and especially Dilthey. Droysen was inconsistent on a number of crucial points, including the possibility of objective metaphysical, moral, and trans-historical knowledge. He emphasized the perspectival nature of all human experience, including the historian’s, without doing enough to explain how trans-perspectival knowledge was possible. In the end, this “left the philosopher and historian alike with troubling tensions. … [A]‌nd there is no defense against relativism” (Beiser 2011: 320, cf. 305–7). This was to prove a liability in in era increasingly given to the idealization of the universality and objectivity of scientific knowledge. These problems were exacerbated by Dilthey, who was widely perceived to have extended the perspectivalist elements in Droysen’s thought in the direction of a full-blown relativism (cf. Baghramian 2010: 44). This is a perception shared by Dilthey’s contemporaries, like Rickert and Windelband, as well as by later thinkers both Continental (Gadamer, Habermas) and analytic (Carnap, Abel). Like most figures accused of relativism, Dilthey resisted the charge, and there are who defend him on this point (see Makkreel 2020). However, he hung his aspirations to objectivity on the emerging science of psychology, pursued along the lines of Brentano’s “descriptive psychology,” or those of the early Husserl’s phenomenology. But this strategy was bound to fail in an era increasingly committed to a model knowledge and objectivity drawn from Newtonian physics. And this—the forward march of scientism—is the second key contributor to the demise of Verstehen as originally conceived. The nineteenth century was marked by debate about the nature of science itself. Key points of contention included the nature of scientific laws (necessary or contingent? fixed or mutable?), the nature of scientific observation (theory laden or theory neutral?), and the place of “subjective” methods and their objects, like introspection and the inner, mental life (should they count as “scientific” methods or not?). This last topic was of critical significance not only for Verstehen and its place in the social sciences, but also for the discipline of philosophy. Given the forward march of scientism in modern thought, a discipline’s status as a genuine field of knowledge came increasingly to depend on its ability to represent itself as a “science” of some sort. But since there was not a clear consensus about

272

The History of Understanding in Analytic Philosophy

the nature of science in the nineteenth century, there were many avenues to claiming scientific status. One prominent avenue was to link a discipline to psychology in a way that (supposedly) facilitated the transfer of scientific status. Dilthey’s attempt to ground the objectivity of Verstehen in descriptive psychology is one example of this strategy, enacted on behalf of all the Geisteswissenschaften. This strategy was also adopted by the tradition of British empiricism, where, since the time of Locke, the conception of philosophy as a mental science adjacent to psychology had been cultivated by its leading thinkers. Thus George Croom Robertson, the first editor of Mind,16 places the associationist psychologist David Hartley alongside David Hume as a major figure in British philosophical thought, noting that “psychology … cannot be neglected in a history of philosophic thought in England, where it has been so steadily cultivated without being too carefully discriminated from philosophy proper” (Robertson 1877: 358). For Robertson, “philosophy proper” differs from psychology mainly in two respects: its willingness to engage the “subjective” or first-personal dimensions of the mental, and its willingness to engage questions of a more general nature, upon which one could not reasonably hope to achieve the level of consensus characteristic of the hard sciences. Like Droysen and Dilthey, Robertson believed that psychological phenomena are fundamentally subjective or first-personal and that our subjective awareness of them is fundamental in studying them. “Philosophy proper” embraced this: “the fundamental consideration of mind is and must be subjective”; hence “Psychology may be shown to pass inevitably into Philosophy.” But, again like Droysen and Dilthey, Robertson thought this transition could be made “without abandoning the firm ground of the positive sciences” (1876: 4). However, mainstream thought in the sciences, including psychology, was already moving against such a view, and by 1900 it was largely untenable. By then, experimental psychology had begun to establish itself as the best approximation to a “Newtonian” science of mind, and it had done so by rejecting “subjective” methods like introspection (cf. Kukla and Walmsley 2006: Ch. 2). Other disciplines quickly followed suit. A pivotal episode occurred between 1910 and 1930, when the social sciences attempted to rigorize so as to more closely approximate the “hard sciences” and thereby secure their own status as sciences. By the 1920s, a strong consensus was emerging in the global disciplinary communities of the social sciences that scientific objectivity required (i) the elimination of the private or first-personal, (ii) a commitment to operational definition, and (iii) elimination of all value from both sides of inquiry, that of the researcher and that of the researched (Purcell 1973: 21 ff). Obviously, this was a death blow to the scientific aspirations

VERSTEHEN REDUX

273

of any field which based its work on the data of first-person experience. Likewise for any field which aspired to moral knowledge (cf. Reuben 1996; Willard et al. 2018). Vershehen in its original form was out, as was ME. It is no coincidence that the Vienna and Berlin circles were both founded in this same decade—both were effects and also perpetuating causes of this growing consensus. By 1930 this new consensus about the nature of science was having profound effects beyond scientific disciplines themselves, altering curricular, pedagogical, and other institutional norms in leading American universities in ways that forced changes in other disciplines, including philosophy, the arts, and the humanities (Reuben 1996). The publication of Ayer’s Language, Truth and Logic (1936) inaugurated a similar sea change in Britain (Preston 2021). As it became normalized in the social sciences and the broader academic sphere, this rigorized vision of science excluded Verstehen as a legitimate source of knowledge and set the stage for the emergence of analytic philosophy. It is also no coincidence that the term “analytic philosophy” emerged in the early 1930s, just as the standpoint of the rigorized social sciences began to have broad effects on the academic world. Although early uses of the term display some variety as to its intended intension and extension, there is nonetheless a fairly clear pattern of usage that took shape in the early 1930s and persisted through about 1970, representing what I have called the “traditional conception” of analytic philosophy.17 As traditionally conceived, analytic philosophy originated in Moore’s and Russell’s revolutionary break with the British Idealist tradition around the turn of the twentieth century, which curiously was also seen as an epoch-marking development in the Western philosophical tradition. This was predicated on the view that they had innovated a novel philosophical method—linguistic analysis—and along with it a new perspective on the nature of philosophy itself; namely, that philosophy is nothing more than the analysis of language. Thus, analytic philosophy was seen as replacing the “new way of ideas” characteristic of early modern philosophy with its own “new way of words” (Sellars 1948). The way thought and talk about “analytic philosophy” initially took shape bears a non-coincidental relationship to the scientism of the 1930s in several respects. First, the notion that Moore and Russell had taken the “linguistic turn” along with Wittgenstein and the logical positivists requires one to ignore their own highly metaphysical self-understandings of what they were up to in offering “analyses” and to focus only on superficial similarities grounded in the third-person observable features of the language in which those analyses were proffered (see Preti 2017). Second, the metaphilosophical position at the heart of the traditional conception was tailor-made to comply with the new restrictions on knowledge and scientific status imposed by the

274

The History of Understanding in Analytic Philosophy

rigorization of the social sciences via its broader cultural and institutional effects. To retain relevance in the academy, each discipline had to show either that it met the requirements of the rigorized social sciences or that it performed some useful noncognitive task. Many of the arts and humanities took the latter approach, aiming to provide university students with what was essentially an emotivist form of moral formation (cf. Reuben 1996). Analytic philosophy split the difference by presenting itself as dealing in clarificatory and usually deflationary “analyses” of linguistic meanings, thereby offering improved formulations of things already known, rather than generating new knowledge itself. In at least these ways, scientism bears a foundational and systemic relation to the analytic tradition. To be clear: I am not suggesting that scientism was a core tenet or “defining doctrine” of analytic philosophy or even that large numbers of analysts explicitly defended it. To the contrary, many prominent analysts have been explicitly anti-scientistic, and the whole of the “ordinary language” side of the early analytic tradition, associated with Moore and the later Wittgenstein, is usually thought of as anti-scientistic. But even here we find a restricted (Moore) and deflated (Wittgenstein) role for philosophy as a purported field of knowledge that, in my view, is best explained in terms of the culturally pervasive scientism of the early twentieth century (cf. Preston 2007: Ch. 6). Capitulation to this scientism was implicit in the founding conceptualization of analytic philosophy, without which there would have been no “analytic philosophy” to speak of. That conceptualization established norms of practice, including norms delimiting what counted as acceptable questions and acceptable answers among philosophers. This is why so many of the major trends within the analytic tradition have been suggestive of and amenable to scientism. While there is no such thing as “the analytic view” on any topic, there are certain views which, due to their prevalence or prominence, count as characteristic of analytic philosophy, especially at particular stages in its history. There are also some persistent tendencies of thought reasonably seen as characterizing the analytic tradition as a whole. Scientism is implicated in many of these characteristic views and overarching tendencies. As David Cooper observes: Analytical philosophy has generally proved … friendly and sympathetic to science. Where the traditional ambitions of philosophy have appeared to conflict with the claims of science, the response has typically been sanguine acceptance of a reduced role for philosophy—as conceptual analysis, the logic of scientific theory, or whatever. Beyond this, it has often been the explicit aim to provide analyses which will accommodate apparently recalcitrant phenomena to the scientific understanding. Tarski’s “scientific

VERSTEHEN REDUX

275

semantics” is scientific not only in the sense of “exact”, but in that of bringing semantics, through its elimination of intensional notions, into “harmony with the postulates of the unity of science and of physicalism.” Behaviorism and materialism in the philosophy of mind, and the various breeds of “anti-cognitivism” in ethics, have been motivated by a similar quest for harmony. (Cooper 1993: 10)18 It is in this sense that the analytic tradition embodies an acceptance of late modern scientism, as I claimed at the end of Section 11.3.

11.6.  VERSTEHEN REDUX: THE RETURN OF UNDERSTANDING IN ANALYTIC EPISTEMOLOGY Having made the case that the Verstehen tradition and the analytic tradition originated in opposite reactions to the rise of scientism, I turn to the claim that the nineteenth century’s pro-Verstehen movement and the twentyfirst century’s pro-understanding movement grew out of similar concerns pertaining to the epistemic strictures of scientism. To see how this is so, we must consider more closely the basis of the analytic tradition’s neglect of understanding. Zagebski’s insight into the understanding versus certainty cycle might be taken to suggest that whenever understanding is neglected, one should be able to trace a straight explanatory path back to a preoccupation with certainty grounded in skeptical worries. But, while this works in some instances (as with Descartes), it does not work very well with the analytic tradition. Apart from G. E. Moore, the problem of skepticism was simply not a central concern for any leading analysts of the first half of the twentieth century. This is not to say that no one other than Moore discussed skepticism, of course; only that it wasn’t a dominant focus for them or for the emerging analytic tradition. Rather, the dominant focus was on meaning and clarity. And while an interest in clarity does have some interesting connections to concern with certainty (as in Descartes), what drove analysts to focus on meaning and clarity was less a preoccupation with certainty and more a desire to emulate science, combined with a willingness to defer to it in all matters epistemic—in other words, scientism. Insofar as scientism equates to a strong affirmation of the possibility of knowledge, it seems to be the very opposite of skepticism and, hence, a potential counterexample to Zagzebski’s thesis that neglect of understanding correlates with skepticism. However, such a conclusion misses the important fact that the flip side of scientism’s exclusive affirmation of scientific knowledge is skepticism about other purported forms of knowledge and other purported

276

The History of Understanding in Analytic Philosophy

forms of cognitive achievement, like wisdom or understanding. Thus Russell follows his scientistic declaration that “whatever knowledge is attainable must be attained by scientific methods” by rephrasing it in the negative: “what science cannot discover, mankind cannot know” (Russell [1935] 1997: 243, emphasis added). This is why R. G. Collingwood (1933: 137–50), in one of the earliest discussions of the newly dubbed “analytic philosophy,” identifies it as a form of philosophical skepticism. As Collingwood observes, the analytic philosophy of the early 1930s was skeptical not about knowledge in general, but about philosophical knowledge as traditionally conceived. According to the analysts, “philosophy cannot establish positive or constructive positions,” but is limited to “taking the propositions which we are given by science and common sense, and revealing their logical structure or ‘showing exactly what we mean when we say’, for example, that there is a material world” (Collingwood 1933: 141–2). So skepticism of a sort was still part of the picture in early analytic philosophy, but its effects were channeled through a scientism-driven deference to science. This was what set the analytic tradition on the path of neglecting understanding and, along with it, ME. In his historical work on the analytic tradition, Scott Soames has identified several “underlying themes or tendencies that characterize,” but do not define, analytic philosophy (2003: xiii). One is that “philosophy done in the analytic tradition aims at truth and knowledge, as opposed to moral or spiritual improvement … the goal of analytic philosophy is to discover what is true, not to provide a useful recipe for living one’s life” (2003: xiv). This clearly puts it at odds with the aims of ME. Additionally, analytic philosophy is characterized by “a widespread presumption … that it is often possible to make philosophical progress by intensively investigating a small, circumscribed range of philosophical issues while holding broader, systematic issues in abeyance” (2003: xv). Although true, this is an understatement. The analytic tradition has been characterized by a positive aversion to addressing “broader, systematic issues,” especially on anything like the scale required by ME. It is hard to see how one could make progress on specific philosophical problems without focusing principally on the phenomena immediately involved. But to hold broader, systematic issues in abeyance indefinitely, as has been the norm in the analytic tradition, is to obstruct the attainment of understanding. This aversion to doing the kind of systematic work needed for understanding was in large part due to methodological norms established by early analysts like Moore and Russell, whose principal focus was on the analysis of propositions taken singly. This atomistic orientation initially emerged as an almost accidental byproduct of their rejection of British

VERSTEHEN REDUX

277

Idealism and its holism, combined with Moore’s philosophical style (not a self-conscious “method,” as Moore himself later explained; 1942: 676). But it soon became a matter of methodological principle in connection with the logical atomism of Russell and the early Wittgenstein. From that point onward, it persisted, in no small part because it was a way to approximate the exactness of the natural sciences19 while also deferring to them. Even before the rise of logical positivism and the new consensus of the rigorized social sciences, Russell had come to reject the systematic and ethical aspects of the traditional philosophical project. In his 1914 Herbert Spencer lecture, titled “On Scientific Method in Philosophy,” Russell dismissed the idea that there are facts pertaining to the universe as a whole and that some of these are ethical facts as mere anthropomorphic projections. An early adopter of ethical emotivism, Russell rejected “ethical metaphysics” as “fundamentally an attempt […] to give legislative force to our own wishes” (1914: 107–8). His attitude toward philosophical system-building was not entirely negative (see Carey 2017), but what the analytic tradition took from him as it coalesced over the next two decades was the idea that “the essence of philosophy … is analysis, not synthesis” (Russell 1914: 113). This had far-reaching effects within the discipline of philosophy, not only in terms of maintaining the atomistic orientation toward propositions considered singly, but also in the proliferation of specialized subfields in philosophy. The atomistic orientation established by the early analysts remained a norm of practice in analytic philosophy long after any principled reason for it had faded from memory. It is one of the things that recent understanding theorists have found objectionable in mainstream analytic epistemology’s exclusive focus on knowledge and its elements (see Kvanvig 1992: 181–2, 2017: 15; Zagzebski [2001] 2020: 59). This is one reason why it makes sense to see parity between nineteenth-century Verstehen theorists and twenty-first-century understanding theorists: in both cases, advocacy for understanding was a reaction to strictures grounded, ultimately, in scientism. Additionally, in contrast with the rigorized social sciences’ demand that scientific inquiry be value-free, theorists both of Verstehen and of understanding have been overwhelmingly motivated by concern with value. For Droysen, it was concern that the explanatory role of value in human agency be given its due. For contemporary understanding theorists, the concern with value arises in the form of several “value problems” pertaining to knowledge, beginning with doubts about the value of knowledge as compared with true belief (or other subsets of the elements of knowledge) and extending to doubts about the value of knowledge as compared with other ostensibly valuable cognitive states, like wisdom and understanding

278

The History of Understanding in Analytic Philosophy

(see Kvanvig 1992, 2003, 2009, 2017; Zagzebski [2001] 2020; Pritchard, Turri, and Carter, 2018). By the 1990s, when contemporary understanding-theorists first began to appear, the “fact-value distinction” had long since been rejected and ethical noncognitivism was in decline. But, given the fragmentation of the analytic tradition into numerous specialized subdisciplines, aptly characterized by Michael Kremer (2005) as “balkanized fiefdoms,” the recovery of value was not immediately felt throughout the entire analytic kingdom. Analytic epistemology labored on in a relatively value-free zone until the 1980s with the advent of virtue epistemology (Turri, Alfano, and Greco, 2019). Virtue epistemologists construe valuable cognitive states, like knowledge, wisdom, or understanding, as virtues that are not only instrumentally useful but which are also valuable in themselves and confer an admirable or praiseworthy status on their possessors. Not surprisingly, contemporary interest in understanding as a valuable cognitive state has been driven by a group of virtue epistemologists. Only as virtue epistemology began to grow in popularity in the 1980s and 1990s did analytic epistemology undergo what Wayne Riggs (2006) has dubbed its “value turn,” effecting “a rapprochement between epistemology and ethics” (Pritchard 2007: 85). The advent of contemporary understanding theory was part of this value turn, and only with that turn did analytic epistemology escape the strictures of norms established over half a century earlier under the dominion of scientism. It is in this sense that the twenty-first-century pro-understanding movement, like the nineteenth-century’s pro-Verstehen movement, grew out of concerns pertaining to the epistemic strictures of scientism.

11.7.  CODA: ME REDUX? Most of my own work on the history of analytic philosophy has been motivated by the sense that its neglect of philosophy’s traditional eudaimonistic task is highly problematic. Not only does this render philosophy less useful to ordinary people in crafting good and meaningful lives, but it also seems likely that there is a non-coincidental relationship between the abandonment of ME over a century ago and many of the social ills of our age.20 In After Virtue, Alasdair MacIntyre draws attention to the relationship between the incoherence of our moral vocabularies and the “shrillness and interminability” that had become characteristic of moral and political disagreement in the twentieth century (1981: 8). Now, four decades later, that shrillness has only become worse and has with increasing frequency spilled over into violence perpetrated by extremists on both the left and the right. (In fact, as I write this paragraph, an armed mob has just breached the U.S. Capitol building in

VERSTEHEN REDUX

279

an effort to disrupt the certification of the recent presidential election.) On both sides, moral and political ideologies grounded in ill-formed worldviews are a large part of the problem. These worldviews are sad counterfeits of those that philosophy under ME once aimed to produce, for they are not developed or tested by reason in the way ME requires. Consequently, they are usually ill-founded, underdeveloped, and even internally inconsistent. Plausibly, one reason so many people are now guided by such defective worldviews is that the practice of intellectually responsible worldview formation and critique no longer has any institutional home as it once did in our colleges and universities, especially within the discipline of philosophy (cf. Reuben 1996; MacIntyre 2009; Willard et al. 2018; Soames 2019: 371– 2). If so, then one prospect—perhaps the only prospect—for adequately addressing our present ills is to move back in the direction of ME as a guiding vision for philosophy. Since understanding is essential to the task of ME, the recovery of understanding by contemporary virtue epistemologists is an important step in the right direction. As we have seen, this important step was facilitated by a rapprochement between epistemology and ethics. The recovery of ME will require a further rapprochement: between epistemology, ethics, and metaphilosophy. Twenty years ago, Zagzebski ([2001] 2020: 58) urged that “it is time we cease the obsession with justification and recover the investigation of understanding.” Today, it is time that we implement at the level of metaphilosophy what two decades of investigation into the nature and value of understanding has unearthed and that we do so with the aim of understanding what humans need most to understand: our nature, our place in the world, and, in light of that, how to live good lives.

NOTES 1 For a good overview of the main positions, see Gordon (n.d.). 2 Not that Zagzebski suggests that it does; but approaching the matter from the standpoint of epistemology as she does, her focus is understandably on the epistemic factors at work in this historical pattern. 3 See Popkin (1979). Cf. Gregory (2012), but note that Gregory locates the source of modern skepticism in Scotus’s voluntarist view of the divine will rather than in nominalism. He has a point, but as I see it, Scotus’s view of will became problematic only when combined with the Occamist rejection of universals as grounds of necessary truth. 4 These authors emphasize different aspects of ancient philosophical practice, and they disagree over a number of important details, but they also exhibit a clear “overlapping consensus” on the fact that philosophy originally had practical, moral aims that were intrinsically linked to its theoretical aims.

280

The History of Understanding in Analytic Philosophy

5 See Moravcsik (1979), Burnyeat (1981), Lear (1988), Everson (1990), Fine (1990), and Benson (2000). Cf. Zagzebski (2001/2020), Greco (2014), and Kvanvig (2017). 6 Droysen first published on his concept of Verstehen in his 1858 Historik: Vorlesungen über Enzyklopädie und Methodologie der Geschichte, but he was already lecturing on Verstehen at Jena and Berlin in 1857 (cf. Beiser 2011: Ch. 7). 7 Brian Fay (2017) gives a useful taxonomy of the main positions. 8 See Makkreel (2020). However, as Uebel (2010, 2019), as well as a number of essays in the present volume, seek to show, the positivist reception of Verstehen was more sophisticated and more diverse than this received view allows. 9 See Beiser (2011: Ch. 8) and Maakkreel (2020). This is ironic given that this sort of contextualization is an essential component of Verstehen—thus the received view of Verstehen was in part determined by a failure to achieve Verstehen about Verstehen. 10 The account of Droysen’s Verstehen given here draws heavily on Beiser. 11 Note that these causes and conditions do not, for Droysen, amount to a complete causal explanation of any historical event involving human action. For Droysen, human action proceeds from free will, understood along what we now call “libertarian” lines. Awareness of the full range of conditions within which human decisions are made enables us to understand those decisions, but the conditions do not fully explain (causally) the decisions. Understanding the inner workings of what we now call “agent causation” is beyond the historian’s scope of expertise and, for Droysen, may be beyond human comprehension. But he accepts it, in Kantian fashion, as a condition of the possibility of moral experience. As we shall see, the moral domain is of central importance for history on Droysen’s account. 12 Droysen’s view has strong affinities with what are now called “perceptual” theories of our knowledge of other minds. See Avramides (2020: § 1.4). 13 Most of what I say about analytic philosophy in this section has been developed at greater length in Preston (2007). 14 This is only partly due to the fact that it would seem to require some version of theism. It is also partly due to the fact that many facets of the natural world— from the values of the cosmic constants to indeterministic quantum events—seem to be if not entirely arbitrary then at least contingencies that are not determined by any facts about “what is best.” Even if one is a theist, the contingencies of the natural world require a restriction and qualification of Socrates’ hope. Although Droysen had theological views, including views relating the doctrine of providence to human history, in his mature work he was careful to separate his theological beliefs from what he thought the historian could achieve using proper historical method, the method of Verstehen. 15 I cannot argue for this here, but see Reuben (1996), Willard et al. (2018), Hunter and Nedelisky (2018), and Preston (2019a, 2021). 16 Mind was founded in 1876 by Alexander Bain, then the leading representative of the empirical tradition in Britain. His great achievement was to bring the associationist psychology of Hartley, Mill, etc. up to date by relating it to what were then the most modern theories of physiology and, thus, to lay the

VERSTEHEN REDUX

281

groundwork for much progress in experimental psychology. Dubbing it “a quarterly review of psychology and philosophy,” Bain founded Mind for the express purpose of continuing research in the line of Bacon, Hobbes, Locke, Berkeley, Hume, Hartley, and the Mills. 17 See Preston (2007, §§ 2.3, 3.4), Glock (2008, §3.1), Beaney (2013: 44), and FrostArnold (2017). Most of what follows in my discussion of analytic philosophy and scientism has been defended at greater length in Preston (2007). 18 Similar observations have been made by Putnam (1992: ix), Sorell (1991), Clarke (1997), Willard (1998: 28), Capaldi (2000), Margolis (2003), and Haack (2017). A 2009 survey of over 3,000 philosophers working in analytic contexts found that 49.8 percent of respondents were naturalists, while only 25.9 percent were nonnaturalists. Similar statistics hold for other perspectives amenable to scientism, including compatibilism about free will (59.1%) and physicalism about the mind (56.5%) (see Bourget and Chalmers 2014). 19 As Geoffery Warnock observed in 1958, “the search for philosophical analyses always took the form of an attempt to formulate a sort of linguistic equation” (1958: 59). 20 I cannot argue for this here, but see Preston (2019b, 2021).

REFERENCES Avramides, A. (2020), “Other Minds,” in Stanford Encyclopedia of Philosophy, ed. E. N. Zalta. https://plato.stanford.edu/archives/win2020/entries/other-minds/. Ayer, A. J. (1936), Language, Truth and Logic. London: Victor Gollancz. Baghramian, M. (2010), “A Brief History of Relativism,” in M. Krausz (ed.), Relativism: A Contemporary Anthology, 31–50, New York: Columbia University Press. Beaney, M. (2013), “What Is Analytic Philosophy?” in M. Beaney (ed.), The Oxford Handbook of the History of Analytic Philosophy, 3–29, Oxford: Oxford University Press. Beiser, F. (2011), The German Historicist Tradition, Oxford: Oxford University Press. Benson, H. (2000), Socratic Wisdom: The Model of Knowledge in Plato’s Early Dialogues, New York: Oxford University Press. Bourget, D., and D. J. Chalmers (2014), “What do Philosophers Believe?” Philosophical Studies, 170 (3): 465–500. Buckle, H. T. ([1857] 2011), History of Civilization in England, Cambridge: Cambridge University Press. Burnyeat, M. (1981), “Aristotle on Understanding Knowledge,” in E. Berti (ed.), Aristotle on Science: The Posterior Analytics, 97–139, Padua: Editrice Antenoire. Capaldi, N. (2000), The Enlightenment Project in the Analytic Conversation, Dordrecht: Kluwer. Carey, R. (2017), “Russell’s Philosophical Method: How Analytic Philosophy Is Shaped By and Perpetuates Its Misinterpretation,” in A. Preston (ed.), Analytic Philosophy: An Interpretive History, 85–102, New York: Routledge. Clarke, D. S. (1997), Philosophy’s Second Revolution, La Salle, IL: Open Court. Collingwood, R. G. (1933), An Essay on Philosophical Method, Oxford: Clarendon.

282

The History of Understanding in Analytic Philosophy

Cooper, D. (1993), “Analytical and Continental Philosophy,” Proceedings of the Aristotelian Society, 94: 1–18. Dray, W. (1957), Laws and Explanation in History, Oxford: Oxford University Press. Everson, S. (1990), “Introduction,” in S. Everson (ed.), Epistemology, Volume 1 of Companions to Ancient Thought, 1–10, Cambridge: Cambridge University Press. Fay, B. (2017), “Verstehen and the Reaction against Positivism,” in L. McIntyre and A. Rosenberg (eds.), The Routledge Companion to Philosophy of Social Science, 29–40, London: Routledge. Fine, G. (1990), “Knowledge and Belief in Republic v– vii,” in Epistemology, Volume 1 of Companions to Ancient Thought, 85–115, Cambridge: Cambridge University Press. Frost-Arnold, G. (2017), “The Rise of ‘Analytic Philosophy’: When and How did People Begin Calling themselves ‘Analytic Philosophers’?” in S. Lapointe and C. Pincock (eds.), Innovations in the History of Analytical Philosophy, 27–67, London: Palgrave Macmillan. Gillespie, M. A. (1995), Nihilism before Nietzsche, Chicago: University of Chicago Press. Glock, H.-J. (2008), What Is Analytic Philosophy? Cambridge: Cambridge University Press. Gordon, E. C. (n.d.), “Understanding in Epistemology,” in Internet Encyclopedia of Philosophy. https://iep.utm.edu/understa/ Greco, J. (2014), “Episteme: Knowledge and Understanding,” in K. Timpe and C. A. Boyd (eds.), Virtues and Their Vices, 285–301, Oxford: Oxford University Press. Gregory, B. (2012), The Unintended Reformation, Cambridge, MA: Harvard University Press. Grimm, S. (2014), “Understanding as Knowledge of Causes,” in A. Fairweather (ed.), Virtue Epistemology Naturalized: Bridges Between Virtue Epistemology and Philosophy of Science, 329–45, Dordrecht: Springer. Hadot, P. (1995), Philosophy as a Way of Life, trans. M. Chase, Oxford: Blackwell. Hadot, P. (2002), What is Ancient Philosophy?, trans. M. Chase, Cambridge, MA: Harvard University Press. Harrington, A. (2001), “Dilthey, Empathy and Verstehen: A Contemporary Reappraisal,” European Journal of Social Theory, 4 (3): 311–29. Kremer, M. (2005), Review of Philosophical Analysis in the Twentieth Century, 2 vols., Notre Dame Philosophical Reviews. https://ndpr.nd.edu/reviews/book-1philosophical-analysis-in-the-twentieth-century-vol-1-the-dawn-of-analysis-book2-philosophical-analysis-in-the-twentieth-century-vol-2-the-age-of-meaning/. Kvanvig, J. (1992), The Intellectual Virtues and the Life of the Mind, Lanham, MD: Rowman and Littlefield. Kvanvig, J. (2003), The Value of Knowledge and the Pursuit of Understanding, New York: Cambridge University Press. Kvanvig, J. (2009), “The Value of Understanding,” in A. Haddock, A. Millar, and D. Pritchard (eds.), Epistemic Value, 95–112, Oxford: Oxford University Press. Kvanvig, J. (2017), “Understanding,” In F. D. Abraham and W. J. Aquino (eds.), Oxford Handbook on the Epistemology of Theology, 175–90, Oxford: Oxford University. Haack, S. (2017), “The Real Question: Can Philosophy be Saved?” Free Inquiry, 37 (6): 40–3.

VERSTEHEN REDUX

283

Hunter, J. D. and P. Nedelisky (2018), Science and the Good: The Tragic Quest for the Foundations of Morality, New Haven: Yale University Press. Kukla, A. and J. Walmsley (2006), Mind: A Historical and Philosophical Introduction, Indianapolis, IN: Hackett. Lear, J. (1988), Aristotle: The Desire to Understand, New York: Cambridge University Press. MacIntyre, A. (1981), After Virtue, Notre Dame: University of Notre Dame Press. MacIntyre, A. (2009), God, Philosophy, Universities, Lanham, MD: Sheed and Ward Book/Rowman & Littlefield Publishers. Maclean, M. J. (1982), “Johann Gustav Droysen and the Development of Historical Hermeneutics,” History and Theory, 21 (3): 347–65. Makkreel, R. (2020), “Wilhelm Dilthey,” in Stanford Encyclopedia of Philosophy, ed. E. N. Zalta. https://plato.stanford.edu/archives/win2020/entries/dilthey/. Margolis, J. (2003), The Unraveling of Scientism, Ithaca: Cornell University Press. Moore, G. E. (1942), “A Reply to My Critics,” in P. A. Schilpp (ed.), The Philosophy of G. E. Moore, 535–677, La Salle: Open Court. Moravcsik, J. (1979), “Understanding and Knowledge in Plato’s Philosophy,” Neue Hefte für Philosophe, 15: 53–69. Nussbaum, M. (1994), The Therapy of Desire: Theory and Practice in Hellenistic Ethics, Princeton: Princeton University Press. Popkin, R. H. (1979), The History of Scepticism from Erasmus to Spinoza, Berkeley: University of California Press. Preston, A. (2007), Analytic Philosophy: The History of an Illusion, London: Continuum. Preston, A. (2019a), “Personhood in Twentieth- and Twenty-first-century Anglophone Philosophy,” in A. LoLordo (ed.), Persons: A History, 263–300, Oxford: Oxford University Press. Preston, A. (2019b), “Soames on Ethics: A New Vision for the Future of Analytic Philosophy?” Philosophical Studies, 176 (5): 1347–55. Preston, A. (2021), “Ayer’s Book of Errors and the Crises of Contemporary Western Culture,” in A. T. Tuboly (ed.), The Historical and Philosophical Significance of Ayer’s Language, Truth and Logic, 333–64, Cham: Palgrave Macmillan. Preston, A. (forthcoming), “Philosophy and Its Past: A Eudaimonistic Perspective,” in S. Lapointe and E. Reck (eds.), Historiography and the Formation of Philosophical Canons, New York: Routledge. Preti, C. (2017), “Some Main Problems of Moore Interpretation,” in A. Preston (ed.), Analytic Philosophy: An Interpretive History, 70–84, New York: Routledge. Pritchard, D. (2007), “Recent Work on Epistemic Value,” American Philosophical Quarterly, 44 (2): 85–110. Pritchard, D., J. Turri, and A. Carter (2018), “The Value of Knowledge,” Stanford Encyclopedia of Philosophy, ed. E. N. Zalta. https://plato.stanford.edu/archives/ spr2018/entries/knowledge-value/. Purcell, E. A., Jr. (1973), The Crisis of Democratic Theory: Scientific Naturalism and the Problem of Value, Lexington: University Press of Kentucky. Putnam, H. (1992), Renewing Philosophy, Cambridge and London: Harvard University Press. Reuben, J. (1996), The Making of the Modern University: Intellectual Transformation and the Marginalization of Morality, Chicago: University of Chicago Press.

284

The History of Understanding in Analytic Philosophy

Riggs, W. (2006), “The Value Turn in Epistemology,” in V. Hendricks and D. Pritchard (eds.), New Waves in Epistemology, 300–32, Aldershot: Ashgate. Robertson, G. C. (1876), “Prefatory Words,” Mind, 1 (1): 1–6. Robertson, G. C. (1877), “English Thought in the Eighteenth Century,” Mind, 2 (7): 352–66. Russell, B. (1914), “On Scientific Method in Philosophy,” in Mysticism and Logic, 97–124, New York: Longmans, Green and Co. Russell, B. ([1935] 1997), Religion and Science, Oxford: Oxford University Press. Sellars, J. (2003), The Art of Living: The Stoics on the Nature and Function of Philosophy, Burlington: Ashgate. Sellars, W. (1948), “Realism and the New Way of Words,” Philosophy and Phenomenological Research, 8 (4): 601–34. Soames, S. (2003), Philosophical Analysis in the Twentieth Century, 2 vols., Princeton: Princeton University Press. Soames, S. (2019), The World Philosophy Made, Princeton: Princeton University Press. Sorell, T. (1991), Scientism: Philosophy and the Infatuation with Science, New York: Routledge. Turri, J., M. Alfano, and J. Greco (2019), “Virtue Epistemology,” Stanford Encyclopedia of Philosophy, ed. E. N. Zalta. https://plato.stanford.edu/archives/ fall2019/entries/epistemology-virtue/. Uebel, T. (2010), “Opposition to Verstehen in Orthodox Logical Empiricism,” in U. Feest (ed.), Historical Perspectives on Erklären and Verstehen, 291–309, Dordrecht: Springer Nature. Uebel, T. (2019), “Neurath on Verstehen,” European Journal of Philosophy, 27 (4): 912–38. Warnock, G. J. (1958), English Philosophy Since 1900, London: Oxford University Press. Willard, D. (1998), “Who Needs Brentano? The Wasteland of Philosophy Without Its Past,” in R. Poli (ed.), The Brentano Puzzle, 15–43, Aldershot: Ashgate. Willard, D. et al. (2018), The Disappearance of Moral Knowledge, New York: Routledge. Winch, P. (1958), The Idea of a Social Science and its Relation to Philosophy, London: Routledge and Kean Paul. Zagzebski, L. T. ([2001] 2020), “Recovering Understanding,” in Epistemic Values: Collected Papers in Epistemology, 57–77, Oxford: Oxford University Press. Zagzebski, L. T. (2020), “Introduction,” in Epistemic Values: Collected Papers in Epistemology, 1–8, Oxford: Oxford University Press.

INDEX

Abel, T. 5, 8, 65, 67, 74–5, 78, 79, 80 n.4, 271 Adorno, T. W. 210 Albert, H. 163 Anscombe, G. E. M. 192, 210, 211, 213–14, 225 n.4 anti-psychologism 32 Aristotle 190, 211–12, 213–14, 265, 270 Avenarius, R. 89, 163 Ayer, A. J. 273 Bacon, F. 281 n.16 Bain, A. 280 n.16 Bauch, B. 56, 167 Beiser, F. 269 Bergson, H. 155 n.14 Berkeley, G. 281 n.16 Bernheim, E. 31, 69 Blackburn, S. 240 Boeckh, A. 90 Bohr, N. 253–6 Brentano, F. 99, 271 Breysig, K.v. 31 Brunswik, E. 175–6 Bruun, H. H. 128 n.11, 128 n.13 Buckle, W. H. 53, 267 Bühler, K. 173, 175, 176

Carnap, R. 5, 8, 56, 69, 81 n.7, 87–8, 91–7, 98–9 107–8, 127 n.6, 131 n.37, 138, 142, 146, 154 n.5, 155 n.13, 155 n.14, 155 n.17, 162, 167, 271 Cassirer, E. 48, 55, 56, 69, 89, 155 n.17, 167 Chladenius, J.M. 23 Cohen, H. 55–6, 59 n.16, 89, 167 Collingwood, R.G. 2, 187–8, 189, 190–4, 195, 196, 205, 212, 224, 276 Comte, A. 212 Cooper, D. 275–6 Cornelius, H. 99 Croce, B. 73, 74 Davidson, D. 175, 206 n.3, 216, 235 Davis, W. M. 168, 178 n.10 Dennet, D. 233 Descartes, R. 136, 265, 275 Dilthey, W. 2, 3, 5, 6, 13, 14, 18, 21, 23, 33, 35 n.1, 40, 41, 45, 46–7, 48, 51–3, 54–5, 56–7, 58 n.13, 65, 73, 74, 77–8, 80, 80 n.3, 88, 89–90, 91, 99, 100 n.1, 104, 107, 111, 118, 136, 142–4, 145–6, 148–9, 152, 155 n.13, 156 n.21, 161, 172, 210, 212, 224, 266–7, 269, 270, 271, 272

286Index

Dray, W. H. 187–8, 189–90, 191–2, 210, 215 Droysen, J. G. 13, 14, 17, 18, 21, 23, 35, 40, 46, 54, 69, 210, 212, 224, 266–9, 270–2, 277, 280 n.6, 280 n.11, 280 n.14 Durkheim, E. 212 Ernst Mach Society 135 Eulenburg, F. 14, 31, 33–4, 35 Fechner, G. T. 39, 93, 101 n.2 Feest, U. 98 Feigl, H. 178 n.4 Feyerabend, P. 163 Fischer, K. 25 Fraenkel, E. 69 Frankfurt Institute for Social Research 135 Frayn, M. 253 Freud, S. 251 Freyer, H. 56, 94 Friedman, M. 5, 87–8, 95–6 Frischeisen-Köhler, M. 14, 31, 33, 35 Gadamer 90, 206 n.2, 210, 242, 267, 271 Galileo, G. 211–12, 214 German idealism 150–1 Goethe 19, 24 Goldie, P. 235, 237 Grotenfeld, A. 31 Habermas, J. 90, 210, 271 Hacking, I. 235 Hartley, D. 272, 280 n.16 Hegel, G. W. F. 41, 43, 58 n.8, 91, 177 n.1, 212, 221, 268 Heidegger, M. 69, 89, 90, 138, 154 n.5 Heisenberg, W. 253–6 Helmholtz, H. v. 90 Hempel, C. G. 5, 65–7, 69, 71–3, 74–7, 78–9, 80 n.4, 81 n.5, 81 n.15, 122, 124–5, 187–91, 205, 213, 214, 215, 220 Herbart, J. F. 39, 41–2, 48 Hettner, A. 164, 168

Hintze, O. 14, 31, 32, 35 Hobbes, T. 281 n.16 Hook, S. 69, Horkheimer, M. 2, 6, 8 Humboldt, W. v. 41, 43, 58 n.6 Hume, D. 95, 101 n.4, 215–16, 219, 221, 248, 272, 281 n.16 Husserl, E. 89, 95, 271 Jaeger, W. 69 Jaffe, A. 128 Jerusalem, W. 44 Jodl, F. 165 Kant, I. 24, 25, 27, 39, 40, 56, 116, 139, 150, 167, 280 n.11 Kaufmann, F. 2, 3, 8, 122, 131 n.35 Kim, J. 229 Kinzel, K. 52–3 Knapp, F. 25 Knight, F. 130 n.29 Kraft, V. 3, 6 Kraus, J. B. 123 Kremer, M. 278 Kries, J. v. 129 Kristeller, P. O. 5, 68–9, 70–1, 73, 76–9, 81 n.13, 81 n.15 Kusch, M. 56 Kvanvig, J. 262–3, 265 Lamprecht, K. 31 Lange, F. A. 32 Langlois, C.-V. 69 Laplace, P.-S. 120 Lasson, A. 41 Lazarus, M. 4, 40, 41–5, 47–51, 53, 56–7, 58 n.8, 59 n.16, 89 Lewis, D. 225 n.7 Locke, J. 186–7, 192, 281 n.16 Lodge, D. 245 Mach, E. 44, 163 MacIntyre, A. 278 Makkreel, R. 269 Mandelbaum, M. 5, 8, 73, 74–5, 77– 8, 80 Mannheim, K. 73, 74, 156 n.21 Marcuse, H. 155 n.7

Index

Marx, K. 221 McDowell, J. 188 Meinecke, F. 31, 35 Melden, A. I. 211, 216 Menger, C. 106, 117, 128 Mill, J. S. 95, 101 n.4, 106, 212, 280 n.16 Moore, G. E. 273–4, 275, 276–7 Morrison, D. 31 Morton, A. 229, 237, 247 Naess, A. 6, 175–6 Nagel, E. 69, 125, 132 n.39 Napoleon, B., 19, 24 Neider, H. 8 Neurath, O. 2, 5, 6, 71–3, 76, 79, 81 n.13, 100, 162, 176, 177, 179 n.19 Newton, I. 44, 267, 272 Niebuhr, R. 40 Nietzsche, F. 20, 24 Occam, W. 279 n.3 Paulsen, F. 25 Penck, A. 168 Pettit, P. 252 Pihlström, S. 223 Pippin, R. 246 Plato 18, 261, 264–5, 270 Popper, K. 157 n.31, 163, 178 n.16, 200–1, 210 Quine, W. V. O. 7, 8, 186–8, 191, 194–5, 197, 203–6, 238–9 Ranke, L. 29, 32, 35, 40 Reichenbach, H. 69 Reis, L. 69, 70, 76–7, 81 n.6, 81 n.13 Richardson, A. 95, 96 Rickert, H. 16, 31, 33, 35, 55, 71, 73, 104, 107, 113, 116, 119, 122, 124, 129 n.21, 130 n.26, 156 n.25, 165– 6, 267 Riggs, W. 278 Robertson, G. C. 272 Robinson, J. 237 Roth, P. 196–7, 203 Rothacker, E. 56

287

Russell, B. 95, 101 n.4, 213, 266, 273, 276–7 Ryle, G. 7, 177, 187, 192 Scheler, M. 104 Schlick, M. 155 n.7, 155 n.14, 162, 178 n.4 Schmoller, G. 33 Scotus, D. 279 n.3 Seignobos, C. 69 Sellars, W. 57 Shakespeare, W. 64–5, 70–1, 73–5, 79, 80, 80 n.1 Simmel, G. 4, 14–16, 56, 73, 132 n.40, 212 Simondon, G. 56 Soames, S. 276 Socrates 264, 280 n.14 Sokal, E. 93 Sombart, W. 104, 128, 136, 148–9, 156 n.26 Spann, O. 147 Spengler 104, 156 n.21, 156 n.27 Spranger, E. 31, 169 Stegmüller, W. 167 Steinthal, C. H. 4, 40, 41–5, 47–51, 53, 56–7, 58 n.6, 58 n.7, 59 n.16, 59 n.17, 59 n.22, 89 Taylor, C. 78, 79, 251 Thurnwald, R. 56 Troeltsch, E. 73 Uebel, T. 1, 45, 145, 147–8, 155 n.8, 179 n.22 Vienna Circle 100, 103–4, 138, 140, 142, 145, 147, 148, 154, 163, 178 n.4, 213 von Below, G. 31, 35 von Mises, R. 2 von Sybel, H. 25 von Wright, G. H. 7, 8, 161, 167 Vossler, K. 172 Wach, O. 31 Waitz, W. 47 Weber, M. 2, 5, 6, 13, 14, 31, 32, 33, 35, 39, 156 n.21, 156 n.25, 202, 212, 223

288Index

Williams, B. 251 Winch, P. 2, 8, 188, 194–9, 200–2, 204, 206 n.4, 206 n.5, 210, 213 Windelband, W. 16, 55, 71, 104, 107, 164, 165, 267 Wittgenstein, L. 193, 194, 196, 199, 209, 213, 222, 224–5, 226 n.13, 273–4, 277

Wundt, W. 39, 117 Zagzebski, L. 8, 261, 263, 265, 275, 279, 279 n.2 Zeitschrift für Völkerpsychologie und Sprachwissenschaft 41 Zilsel, E. 8



289

290