Normative Species: How Naturalized Inferentialism Explains Us 1032484039, 9781032484037

Citation preview

Routledge Studies in Contemporary Philosophy

NORMATIVE SPECIES HOW NATURALIZED INFERENTIALISM EXPLAINS US Jaroslav Peregrin

Normative Species

This book is about rules, and especially about human capability to create, maintain and follow rules, as a root of what makes us humans different from other animals. The leading idea is that scrutinizing this capability is able to tell us who we humans are and what kinds of lives we live. It elaborates Wilfrid Sellars’ visionary observation that “to say that man is a rational animal, is to say that man is a creature not of habits, but of rules”; and it builds on the ideas of Sellars’ and Brandom’s inferentialism, in a novel naturalistic way. The main tenet of inferentialism is that our language games are essentially rule-governed and that meanings are inferential roles. Jaroslav Peregrin sees the task of reconciliation of inferentialism and naturalism as centered around the problem of naturalization of rules. He argues that the most primitive form of a rule is a cluster of normative attitudes. We humans are specific by our tendency to assume peculiar attitudes to what we do, and to do so in a specific way, which turns the attitudes into “normative” ones. This self-reflective structure characterizes our ability to build systems of interconnected rules, which have come to constitute our natural niche. Furthermore, Peregrin shows how our most important system of rules – that constitutive of our language – helped to lead us to our current position of rule-following, ultra-social, rational, and discursive creatures. Normative Species will be of interest to scholars and advanced students working in philosophy of language, philosophy of mind, social ontology, cultural evolution and cognitive science. Jaroslav Peregrin is a professor at the Faculty of Philosophy of the University of Hradec Králové, Czechia, and the research professor at the Department of Logic of the Institute of Philosophy of the Czech Academy of Sciences. He is the author of Doing Worlds with Words (1995), Meaning and Structure (2001), Inferentialism (2014), Reflective Equilibrium and the Principles of Logical Analysis (together with V. Svoboda, 2017) and Philosophy of Logical Systems (2020). His current research focuses on logical and philosophical aspects of inferentialism and on more general questions of normativity.

Routledge Studies in Contemporary Philosophy

Global Justice and Recognition Theory Dignifying the World’s Poor Monica Mookherjee The Mind-Body Problem and Metaphysics An Argument from Consciousness to Mental Substance Ralph Stefan Weir Empathy’s Role in Understanding Persons, Literature, and Art Edited by Thomas Petraschka and Christiana Werner Cosmopolitan Norms and European Values Ethical Perspectives on Europe’s Refugee Policy Edited by Marie Göbel and Andreas Niederberger The Politics of Recognition in the Age of Digital Spaces Appearing Together Benjamin JJ Carpenter Feminist Philosophy and Emerging Technologies Edited by Mary L. Edwards and S. Orestis Palermos Normative Species How Naturalized Inferentialism Explains Us Jaroslav Peregrin

For more information about this series, please visit: www.routledge.com/Routledge-Studiesin-Contemporary-Philosophy/book-series/SE0720

Normative Species How Naturalized Inferentialism Explains Us

Jaroslav Peregrin

First published 2024 by Routledge 605 Third Avenue, New York, NY 10158 and by Routledge 4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2024 Jaroslav Peregrin The right of Jaroslav Peregrin to be identified as author of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Peregrin, Jaroslav, author. Title: Normative species : how naturalized inferentialism explains us / Jaroslav Peregrin. Description: New York, NY : Routledge, 2024. | Series: Routledge studies in contemporary philosophy | Includes bibliographical references and index. Identifiers: LCCN 2023022726 (print) | LCCN 2023022727 (ebook) | ISBN 9781032484037 (hardback) | ISBN 9781032484044 (paperback) | ISBN 9781003388876 (ebook) Subjects: LCSH: Social norms. | Normativity (Ethics) | Logic. | Semantics (Philosophy) | Language and languages--Philosophy. Classification: LCC GN493.3 .P44 2024 (print) | LCC GN493.3 (ebook) | DDC 303.3/7--dc23/eng/20230724 LC record available at https://lccn.loc.gov/2023022726 LC ebook record available at https://lccn.loc.gov/2023022727 ISBN: 978-1-032-48403-7 (hbk) ISBN: 978-1-032-48404-4 (pbk) ISBN: 978-1-003-38887-6 (ebk) DOI: 10.4324/9781003388876 Typeset in Sabon by KnowledgeWorks Global Ltd.

Contents

Preface

vii

Introduction

1

1 Now I can go on!

6

2 Creature of rules

23

3 Preliminaries I: Rules and other human gear

38

4 Preliminaries II: Rules as part of nature

50

5 Preliminaries III: Kinds of rules

60

6 Normative attitudes

73

7 Rules in the natural world

91

8 The natural history of correctness

103

9 Systems of rules and institutions

114

10 Behavioral patterns

127

11 Practices

138

12 The space of meaningfulness

149

vi Contents

13 Logic167 14 Cooperation and morals178 15 Freedom190 16 The world198 17 Conclusion: We have become a normative species210 References220 Index234

Preface

The intellectual journey that led me to this book began in the 1980s, when I was engaged with formal semantics. I was excited about the possibilities offered by its abundant apparatus, but at the same time I was embarrassed by what I perceived as its naive philosophical grounding. It seemed to me that most of the formal semanticists took the theories as straightforward depictions of how words are fastened to the entities that are their meanings, the apparatus of possible worlds etc., expressing something hidden behind the facade of the natural world, something that served as some kind of metaphysical structure of the world. (Cresswell’s, 1973, book Logic and Languages can be seen as an example of this stance being expressed explicitly.) I felt that formal semantics needed a more solid philosophical background, which led me to write my book, Doing Worlds with Words (Peregrin, 1995). One of the crucial questions I kept asking myself during the work on the book, of course, was What is meaning? I was unsatisfied with theories which took language as a set of labels stuck to meanings – a picture that Quine dubbed the “museum myth”. I was looking for an alternative and I found the basis for one in the writings of Wittgenstein and Quine. This is what was embodied in my book, but I still was not fully satisfied. Shortly after its publication, I got hold of Bob Brandom’s (1994) book Making It Explicit. I was bewitched: this was the theory of meaning that I had always wanted to pursue if I had only been able to put it together like Brandom had! For a time I became a zealous soldier of the inferentialist army. In 2001, I published a book reconciling my ex-love, formal semantics, with my new one, inferentialism (Peregrin, 2001). This period culminated in my book Inferentialism (Peregrin, 2014a), where I tried to put together two partly independent strands of inferentialism: the philosophical and the logical. However, even as I was working on this book, I started to realize that I saw some of the topics differently from Brandom. The crucial issue was

viii Preface naturalism, to which I tended to subscribe (following in the tracks of my previous philosophical hero, Quine) and which Brandom did not seem to be particularly concerned about. I always took philosophy as being continuous with science, the two enterprises cross-fertilizing their ideas, thereby working toward new insights. I was disappointed to see that Brandom did not care about science very much and saw philosophy instead as being something orthogonal to it. Already in 2014, I published a paper summarizing some ideas about the evolutionary origin of rules that thereby suggested a naturalistic basis of inferentialism (Peregrin, 2014b). Some of the ideas also found a way into my 2014 book. Since that time, I have taken my mission to be to provide a basis for a naturalized version of inferentialism. During that time I also returned to my work on the philosophy of logic, explaining logic and its laws from a naturalistic perspective; I published two books – Peregrin and Svoboda (2017) and Peregrin (2020b). The present book, then, is the culmination of my naturalistic reconsideration of inferentialism. There are many people who have helped me, directly or indirectly, to shape the ideas laying the foundation of this book. Let me mention at least a few. First of all, my colleagues from the Department of Philosophy and Social Sciences of the Faculty of Philosophy of the University of Hradec Králové, where we have launched the great project of the naturalization of inferentialism: Ladislav Koreň, Preston Stovall and Matej Dobňák, as well as our oversea collaborators Ulf Hlobil and Mark Risjord. Certainly I must also give thanks to Bob Brandom, the long-term interaction with whom I owe for the development of many of the ideas presented in the book. I am also grateful to my friend and colleague from the Department of Logic of the Institute of Philosophy of the Czech Academy of Sciences, Vladimír Svoboda, who is a tireless and invaluable critic of all my new ideas. I am also grateful to Bartosz Kaluziński for detailed critical comments on the manuscript. This book presents the results of the research project Inferentialism Naturalized: Norms, Meanings and Reasons in the Natural World developed at the Philosophical Faculty of the University of Hradec Králové in Czechia, supported by the research grant No. G20-05180X of the Czech Science Foundation. It has swallowed up parts of my recent papers that addressed some of the topics integrated into this book, usually in a somewhat reworked form. In particular, Chapter 6 is based on the paper “Normative Attitudes” (in L. Townsend, P. Stovall and H. B. Schmid, eds.: The Social Institution of Discursive Norms, Routledge, New York, 2021, 121–137). Chapter 7 incorporates a small part of “Normativity between Philosophy and Science” (forthcoming in

Preface ix Philosophical Psychology), while Chapter 11 contains a small part of “Logic and Human Practices” (M. Blicha and I. Sedlár (ed.): The Logica Yearbook 2020, College Publications, London, 2021, 162–182). Chapter 12 is then based on “Inferentialism Naturalized” (Philosophical Topics 50, 2022, 33–54), and finally Chapter 15 follows “Human World” (Analítica No.1, 2021, 20–34).

Introduction

This book is about rules, and especially about the human capability to create, maintain and follow rules as a root of what makes us humans different from other animals. Indeed, it is meant as an elaboration of Wilfrid Sellars’ visionary observation that “to say that man is a rational animal, is to say that man is a creature not of habits, but of rules”. I am convinced that scrutinizing this capability will let us understand who we humans are and what kinds of lives we lead. The base camp of the intellectual journey that this book undertakes is inferentialism – the doctrine foreshadowed by Sellars and brought to full fruition by Robert Brandom. The core of the doctrine is the conviction that what we call the meaning of a linguistic expression is not something represented by the expression, but rather its inferential role. This presupposes that our language games (or at least some “central” ones) are essentially rule-governed, and hence that to understand language with its semantics is to understand (certain kinds of) rules and (a certain kind of) rule-governance. And it would seem that in so far as we humans can be characterized as linguistic creatures, we can also (and maybe more aptly) be characterized as a normative species. Moreover, it seems that it is not only language, but many other specifically human amenities that presuppose rules or are directly decomposable into various complexes of rules. It seems, indeed, that rules have managed to erect spaces within which we humans assumed our “unnatural” forms of life and which differentiate us so much from other animals. From this viewpoint, rules and complexes of rules start to look as something like the true key to our peculiar nature. The trouble is, alas, that rules and rule-following are a shadowy business. For a long time, philosophy did not pay enough attention to them, and now, when it does, there is a confusion of tongues. Many philosophers (and some scientists) talk about rules and normativity, but how exactly they understand these terms often remains obscure. Therefore, it would

DOI: 10.4324/9781003388876-1

2 Introduction seem that elucidating rules in their capacity of constituting the space of our peculiarly human form of life appears to me to be a very urgent enterprise. This book is an attempt to throw some new light on us humans from this peculiar angle. It starts out by giving consideration to two twentiethcentury philosophers who took the concept of rules seriously before it became a mainstream of philosophy – Wittgenstein and Sellars (Chapters 1 and 2). Their struggle with rules nicely reveals the complex problems that this concept, and the general concept of normativity, harbors. One important lesson from both these thinkers is that there is necessarily something like “implicit rules” – rules that are not a matter of any sentences or other kinds of symbolic articulation. At least some rules must thus be able to reside directly in behavior; though they can be made explicit later (or not). In my view, the conclusion that not every rule can be a linguistic affair is indicated, among others, by the fact that language itself is constituted by rules. In the introductory chapters, I review the discussions regarding rules as they stand, using the terminology in the – sometimes divergent – ways in which it is to be found. But, as I want to present a coherent theory of normativity, I need to put all of this on a more unified foundation. This brings about the necessity of analyzing the very concept of rule, together with other concepts related to it. This is what I undertake in the next three, “stage-setting”, Chapters 3–5. I argue that if rules are not always supported by a language, and if even human rationality, as Sellars suggested, is supported by rules (rather than the other way around), rules and the ability to follow them may have to do with the very differentiation of our species from the other ones. Therefore, the analysis of rules and rulefollowing may reveal to us a lot about our nature, i.e. the nature of the individuals of the genus Homo sapiens. It is also necessary to stress that the framework of our investigation is a broadly naturalistic one, which poses several restrictions on the means that we can use toward our purpose of analyzing the concepts of rules and rule-following. We cannot, for example, situate rules into a supernatural realm independent of the natural one. Our task is to find a place for rules in nature (where, however, nature is construed broadly enough to encompass humans and human communities). As rules cannot be generally identified with linguistic items (such as imperative sentences) and must be sought “within” behavior, the crucial question is what kinds of behavioral patterns amount to rule-following. And here I think it is crucial to pay attention to normative attitudes, the kinds of pro- and con- attitudes to human behavior, which are sensitive only to the kind of behavior (not to who is its source and target) and which we humans ubiquitously assume. I argue (in Chapter 6) that the most rudimentary kind of implicit rules are simply normative attitudes resonating across a society. Then I document (in Chapter 7) that such a

Introduction 3 view can be supported by some evidence concerning both human ontogeny and phylogeny. The concept of normative attitudes is heavily used by Brandom, but in his story, it belongs to the realm of the normative not reducible to the realm of the natural. In contrast to him, however, I argue that the attitudes are at bottom natural phenomena, which thus interconnect the two realms. The attitudes, I maintain, are capable of rendering something as right or wrong – thus grounding normativity – without themselves being right or wrong. In this way, I reach a naturalized version of Brandomian normativism and inferentialism, according to which the realm of the normative is embedded into the realm of the natural. Chapter 8 presents my conjecture concerning the development of the concept of correctness, especially the way in which correctness might have gained a certain independence from our actual attitudes, so that something may be correct despite it being generally held for incorrect (or vice versa). In my view, though correctness could hardly start otherwise than as the direct result of positive normative attitudes (hence, as what the bulk of the members of the relevant community hold for correct), it must have emancipated itself from this direct dependence. This is likely to have happened by means of a process in which normative attitudes retreated from determining an ultimate correctness to determining the criterion of the correctness. The common agreement thus no longer determines what is correct and settles for determining criteria – and thus, we may fail to know what is correct, despite there being a fact of the matter regarding this. In Chapter 9, I explain that while some of our rules are self-standing, more interesting rules are what I call integrative – rules which are operative only in concert with other rules. Such systems of rules, also known as institutions, have the peculiar property, which, I believe, can be best portrayed as creating “inner spaces”. Such “spaces” are remarkable in that they open, for us, the possibilities to carry out brand new kinds of actions. (Once you enter the “space of football”, built of its rules, you can score goals and do other things that are not available to you outside of the space.) And it is here where our specific human form of life originates: we have moved from the realm of nature into the complex of such normative spaces that we have built and that let us live our “unnatural” lives. All of this must be, of course, set into the framework of evolution theory – the ultimate framework of explanation of everything that happens in the animal realm, including the sub-realm of us humans. This is what I do in Chapter 10. A remarkable thing is that rules provide for the swift circulation of “cultural inheritance”: the thing is that, as we have an uncanny knack of not only submitting to the pressure of normative attitudes but taking them as what “ought to be” and thus join others in enforcing them (assuming corresponding normative attitudes), we get the

4 Introduction rules (and the institutions and practices that the rules founded) handed down from generation to generation, without any direct genetic support. Once the framework of evolution theory is in place, we can look at the specific human form of life also in terms of the concept of practice, which I do in Chapter 11. We humans not only do things but at the same time keep assessing our doings as right or wrong (often along more than one dimension). Thus, our practices have come to consist of (at least) two layers: on the lower layer we simply produce, on the upper one we evaluate; and this upper layer makes the behavior on the lower one into acting. The two levels are not really separable, they mingle into a complex whole. And our human predicament is that instead of just coping with the world and with each other, as other animals do, we engage in a labyrinthine collection of practices. In Chapter 12, I try to also shed some light on the nature of the “space of meaningfulness” in which our linguistic practices evolve and flourish. I argue that it is the structuring of our (proto-)assertions by means of the relationships of inferability and incompatibility that equips our utterances (and consequently their vehicles, sentences and their parts) by their semantics. I try to illustrate how an instinctive ejaculation of sounds could have been turned into rule-governed displays and then further into a system of linguistic utterances interlinked by inferential and other relationships, making their vehicles into meaningful expressions. In this context, I also address, in Chapter 13, the origins of logic. As I see it, the logical particles of natural language result from our tendency to make explicit the inferential rules that are originally only implicitly governing our language games. Then I turn my attention to several remaining topics that I think should be elucidated in the light of the conclusions reached up to this point. In Chapter 14, I address the concept of cooperation. I subscribe to the view that it was the need of intensified cooperation that has brought us humans on the trajectory that alienated us from other animal species so quickly that we became nature’s oddity. Our language and our reason, I maintain, arose from interpersonal interactions, leaving deep marks on them. Also our moral rules, in many respects the most important rules we have, have emerged out of the primordial soup of cooperation. The next concept I pay attention to, in Chapter 15, is that of freedom, which is, beyond doubt also something that is characteristic of us. In comparison to other animals, we are free in the sense that in every conceivable situation, we have a much vaster repertoire of behavior to display; but in the end, this is not what we call freedom. Freedom, in our human sense, has to do with rules. To be free in this sense, an individual must abide by rules and is free if he or she has a say in which rules to abide by. The last concept, addressed in Chapter 16, is that of an objective world. The question is how it is that individuals, at first communicating only in

Introduction 5 terms of unarticulated hoots, come to live in a shared objective world, talking about it and arguing what is true about it. In my view, the answer to this question can be derived from the process described in detail in Chapter 8 – the process where the concept of correctness (and also that of truth which is a species of correctness) parts ways with the concept of common agreement; to be correct (true) is no longer to be generally held for correct (true), but rather to be justified by methods that are generally held for correct. In this way, something may be the case despite its generally being taken to not be the case. The objective world defies intersubjective agreement. In the final chapter, Chapter 17, I summarize the fictive journey I have tried to depict in the book: the journey from our age of innocence, when we did not know what is right and what is wrong, to our life as a fully normative species.

1

Now I can go on!

1.1  Kripke on Wittgenstein on rules Not so long ago, the concept of rule was not considered to be crucially important in the context of the quest to reveal the nature of human beings. Traditionally, the leading candidates for the “specific factors” making us what we are were reason, language, morals etc. The concept of rule, to be sure, has never been alien to philosophy. Rules have always been studied by, for example, logicians and ethicists; and when our focus has been human communities, rules were then, of course, seen to help interlink them. However, the very existence of rules was not thought of as a fundamental factor of crucial importance with respect to the nature of us humans. Indeed, rules were usually considered a supplementary phenomenon arriving relatively late in the history of Homo sapiens: it seemed that their emergence must have been preceded by some long development of an already complex cognition (including perhaps a theory of mind and/or a language faculty), of language and of complex cooperation. Therefore, though rules were considered an important ingredient for human societies, they were not seen as something intrinsic to the identification of human nature. However, there were exceptional philosophers who insisted that we need to reflect on rules to understand not only logic or ethics but also mind and language. In the middle of the twentieth century, this was an exception; it was only in the recent decades that this view has received a broad attention. And one of the most important books that stimulated this was a book that appeared in 1982 and was about looking back at one of those clairvoyant philosophers who understood the importance of rules before this became commonplace – namely Ludwig Wittgenstein. Kripke’s (1982) book, Wittgenstein on Rules and Private Language, brought the attention of many philosophers to the question of why DOI: 10.4324/9781003388876-2

Now I can go on! 7 Wittgenstein, in his Philosophical Investigations (Wittgenstein, 1953; hereafter PI), paid so much attention to rules and rule-following. It, of course, was not the case that nobody would discuss it before Kripke, but Kripke’s book helped bring it to the very center-stage of philosophical discussions.1 The popular keyword of PI is “language game”, and on many pages of the book Wittgenstein strives to explain how this notion can help us understand the nature of language. In contrast to his earlier Tractatus (Wittgenstein, 1922), where he pursued the most basic principle of the relationship between language and the world (which he thought he had revealed to be “picturing”), in his later writings he realized that such a pursuit was in vain. There is no such single principle. He understood that there are myriads of ways of how language functions and how it is related to the world – myriads of different language games. However, Wittgenstein devotes even more space in the book to discussing the concept of rule and especially the problem of what it takes to follow a rule. Thus, though Wittgenstein is not quite explicit about this, it is clear that according to him, understanding this is crucial for understanding our language games, in particular that playing a language game typically involves following rules. True, language games are quite varied and rules may play a different – bigger or smaller – role in different kinds of games; but if they were not crucially important, Wittgenstein would not have dealt with them so extensively in PI. The meaning of an expression, according to Wittgenstein, is mostly a matter of the role or roles the expression has come to assume within our language games, of the ways it has come to be employed in them. Wittgenstein writes (PI, §43): For a large class of cases – though not for all – in which we employ the word “meaning” it can be defined thus: the meaning of a word is its use in the language. But already before this, he warns us against imagining the “use” or “function” of an expression in a language as something clear-cut and uniform (PI, §17): Think of the tools in a tool-box: there is a hammer, pliers, a saw, a screw – driver, a rule, a glue-pot, glue, nails and screws. – The functions of words are as diverse as the functions of these objects. (And in both cases there are similarities.)

8  Now I can go on! And, of course, insofar as the games are determined by rules, it is the rules that take part in determining the meaning. Hence, to grasp the meaning of an expression, I must fathom the rules according to which it is used. And what Wittgenstein, according to Kripke, indicated was that to pinpoint the very rules that govern a language game I witness is not really possible. Hence, according to Kripke, Wittgenstein’s argument is that as meaning rests on rules of usage, and as rules of usage turn out to be inscrutable, meaning cannot exist. Thus, according to this interpretation, Wittgenstein is a straightforward skeptic and his argument just propagates the skepsis regarding the possibility of meaning. This interpretation of Wittgenstein raised a wave of disagreement, which culminated in the mocking name “Kripkenstein”, which, according to Davidson (1994), is the straw man target of Kripke’s interpretation. Let us review, in a rough outline, why Kripke ascribes to Wittgenstein the argument he does. The famous example Kripke uses for this purpose is as follows. When one learns to add – i.e. learns the meaning of “+” – one is usually shown a lot of example cases like 1+1=2 2+3=5 etc. and after absorbing a sufficient number of such examples, one becomes able to add also numbers she has not encountered before. It is very probable that the numbers that figure in such school textbook examples are not too big, so we can take a number, n, such that no number bigger than n figures in them. (Kripke uses the number 57, but we can as well choose 1,000,000,000,000.) Now we consider the operation of quaddition, which yields us the same results as addition if both the summands are not bigger than n, and otherwise it yields something else. (Kripke’s example is that in such a case it always yields 5, but again we can imagine that it yields, e.g., the standard sum of the two numbers plus 2.) Now the question is: how come that which we learned at school was apparently that “+” is associated with addition, and not with quaddition (or some other operation indiscernible from addition for small numbers)? Kripke himself poses the question in an even sharpened way. How come, he asks, that when we say that 57 + 68 is 125, it is the correct answer given what I mean by “+”: the thing is that it is correct only if

Now I can go on! 9 my “+” expresses addition, rather than quaddition, or, in other words, only if what I mean by the “+” is plus, rather than quus – but how can this be the case? What would the act of meaning plus (addition), rather than quus (quaddition) be? Kripke insists that as Wittgenstein showed us that there is no such act, a skeptic who would argue that my answer 125 is merely an “unjustified leap in the dark” (p. 15) cannot be refuted. Kripke’s conclusion, then, is straightforward (p. 21): When I respond in one way rather than another to such a problem as “68 + 57”, I can have no justification for one response rather than another. Since the sceptic who supposes that I meant quus cannot be answered, there is no fact about me that distinguishes between my meaning plus and my meaning quus. Indeed, there is no fact about me that distinguishes between my meaning a definite function by “plus” (which determines my responses in new cases) and my meaning nothing at all. As a consequence (p. 21), It seems that the entire idea of meaning vanishes into thin air. As this example might seem to suggest that this is a problem not so much of semantics in general, but of mathematics, let us consider a different example already put forward (also famously) long before Kripke by Nelson Goodman (1955). Consider learning the meaning of “green”. Just like the case of “+”, we learn it by being confronted with a number of exemplars (and perhaps, for contrast, also with some nonexemplars). And just like the case of “+”, there are a plenty of concepts different from the ordinary concept of green (which speakers of English apparently associate with the word “green”), which are, however, indistinguishable from green in terms of all the exemplars with which we could have been confronted. Goodman himself chooses a distinctly weird example: the concept grue, where something is grue if it is either examined before some fixed future date (say 2050) and it is green, or it is examined after the date and it is blue. It is clear that so far everything has been green if and only if it has been grue; hence, there is no way of distinguishing between the two concepts by way of anything we can be confronted with. Again the question is how come we have associated “green” with the concept green, and not with grue (or any other similar monster concept).

10  Now I can go on! In both cases, then, the problem is that we would need to interconnect an expression with a definite meaning, while the information we have is not sufficient to let us do it. Suppose somebody tells us: “Bill is Anne’s son”, while we know that Anne has more than one son. This information, then, is not enough to let us link the name “Bill” with a definite person – to be able to do so, we would need to have some additional information (e.g., that Bill is Anne’s oldest son). Now the situation is similar with Kripke’s “+”, which can be, for all we know, associated with both the operation of addition and that of quaddition; it is the same with Goodman’s “green”, which may be associated both with the property of being green, and that of being grue. However, the situation with Anne’s son is simple in that additional information that would allow us to pick up a unique person may be readily available, while information that would pick up a unique operation (in case of “+”), or a unique property (in case of “green”), may not be available at all. Each such property has an infinity of alternatives, and the kind of information we can get cannot help us tell them all apart. Wittgenstein considers examples concerning finite series of numbers (such as we can encounter in IQ tests), from which we are to guess according to which rule it is created and hence how it is to continue. In the PI, Wittgenstein discusses this problem under the heading “knowing how to go on” (§151): A writes series of numbers down; B watches him and tries to find a law for the sequence of numbers. If he succeeds he exclaims: “Now I can go on!” – So this capacity, this understanding, is something that makes its appearance in a moment. So let us try and see what it is that makes its appearance here. – A has written down the numbers 1, 5, 11, 19, 29; at this point B says he knows how to go on. What happened here? But Wittgenstein points out that the task of determining the law is hopeless: there are always a number of rules compatible with any finite series. Indeed, consider the series 1, 2, 3, 4, 5, … where the rule may appear to be obvious. But aside from the obvious continuation …, 6, 7, 8, 9, 10, …

Now I can go on! 11 there are lots of others, all of which are also regular: …, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, … …, 7, 9, 11, 13, 15, 18, 21, 24, 27, 30, … …, 5, 5, 5, 5, 5, …. etc. It seems that the attempt to determine a definite rule from a finite set of exemplars is mission impossible. Hence, how have we managed to learn our language? Let us make a cautionary remark: it seems that the above considerations presuppose that learning a meaning of an expression is learning some rules, which does not seem to be an uncontroversial claim. But though later in the book we will indeed be arguing for this claim, here we need only its minimalistic (and not really controversial) version. It is hard to deny that part and parcel of learning the meaning of a word like “green” is learning to which objects it applies – and not to which it applies in the mouth of a randomly chosen speaker, but correctly. If we now return to our question, an answer may seem to be forthcoming: we can associate a lot of different concepts with “green”, but as a matter of fact, we all (or nearly all) pick up the same concept (or so it seems). Maybe this is a matter of our cognitive gear, in that we find some concepts (perhaps sub-consciously) more “natural” than others. And among all those concepts that might be associated with “green” given the evidence, the concept green is, in this sense, the most natural. Consider an analysis of how we may, in reality, come to “know how to go on” given by Hofstadter (1995). Consider the following series: 1, 4, 27, 256, 3125, … It is not so difficult to find out that it can be rewritten as 11, 22, 33, 44, 55, … and that hence its regular continuation can be …, 66, 77, 88, … .

12  Now I can go on! However, how do we really come to find this out? A typical process that may lead us to this conclusion, according to Hofstadter, is summarized in the following picture:

Figure 1.1 Analysis of 1, 4, 27, 256, 3125, …. From Fluid Concepts and Creative Analogies by Douglas R. Hofstadter, copyright © 1994. Reprinted by permission of Basic Books, an imprint of Hachette Book Group, Inc.

Considering the original series, we are, according to Hofstadter, first attracted by the number 27, which is, at first sight, characterized by the fact that it is the third power of three. (The first two numbers in the series are too ordinary to strike us as a unique characteristic feature; while the last two are too extraordinary.) This stimulates us to try to also imagine the other members of the series as powers, and while in the case of the first three we get the bases 1, 2 and 3, we try to express the fourth one as a power with the base 4. When we succeed and we see that we can have 1, 2, 3, 4 not only as bases, but also as exponents of the first four members of the series, we check whether the fifth one is also five to the fifth. In this way, we see that finding a regularity in a series is underlain by a lot of psychological mechanisms that may lead us to a specific regularity

Now I can go on! 13 at the expense of other ones. Thus, it is possible that our human cognitive gear, which we have acquired contingently in the course of evolution, favors one of the available regularities and does so for everybody (or almost everybody) of our kind. So why could we not say that addition is the kind of operation which we inevitably, given that our cognition is as it is, associate with “+”? Kripke observes that there is a problem with this answer. If we want to add 50 and 100, we do not only want to say that we, as a matter of fact, tend to give the result 150, but, moreover, that this is the correct answer. We want to say that somebody who would claim that it is, say, 152 is not (only) cognitively anomalous, but wrong. What is the source of this normativity, of this notion of rightness and wrongness, that we apply here? How do we get from the fact that we tend to do something to the fact (or “fact”?) that it is correct? A simple way may be to accept that to be correct is to be in agreement with the majority. This is a notion of normativity which Blackburn (1984, p. 85) aptly calls “democratic harmony”: according to this view, one is correct if and only if she is “in tune” with the others. The simplest case of such harmony is when everybody does the same: a rule is then just what the majority does. Of course, “the same” can be interpreted in various ways: when an orchestra plays a preconceived melody, then there is also a sense in which everybody is doing the same thing (viz. plays the same melody), despite individual musicians doing very different things. And we can consider also more complicated ways of being “in tune”. Imagine a jam session, with no direction given in advance: in such case a musician can also be “in tune” or “out of tune” with the others (though in such case it might be difficult to draw an explicit boundary). But, unfortunately, this pathway from facts to rules is not acceptable, or at least not generally so. Imagine, for example, that the majority of the members of an orchestra are overcome by a strange sort of confusion and start to play disharmonic sounds. Does it mean that this becomes correct and hence what the rest of the orchestra should now do? Or imagine that the members of a community are affected by a mental condition such that when they are to add large numbers, they tend to give wrong answers; and then suppose the majority will tend to give the same wrong answers. Would it follow that the correct results of addition change? It would be difficult to agree with this, especially if the folks could be made to recognize their errors. Of course, it might be that the concept of addition changes for a community – that it starts to associate “+” with a different procedure or function. If so, it might be correct to say that the result of “100+50”, understood in their way, is something else than 150. But this does not contradict the fact that the sum of 100 and 50 is 150. (As Abraham Lincoln

14  Now I can go on! is said to have remarked, calling the tail a “leg” does not change the fact that horses tend to have four legs.) However, this is not to say that correctness does not have anything to do with the agreement with the majority. At least in some cases, it would seem that something may be correct just because the majority agrees that it is correct: for example, the correct way to greet others in a community might be precisely the way the majority of the community holds to be correct.2 And though the sum of two numbers is not correct only because the majority would tend to agree with it but also because it is reached through the correct procedure, which particular procedure is correctly associated with “+” might be a matter of majority agreement. (Certainly the fact that we call addition “addition” and subtraction “subtraction” is a matter of nothing more than a historically contingent agreement among the speakers of English.) Kripke himself stresses that the Wittgensteinian problem, which he sees as distinctively skeptical, requires a skeptical solution. In particular, he maintains that (p. 108) we must give up the attempt to find any fact about me in virtue of which I mean “plus” rather than “quus”, and must then go on in a certain way. Instead we must consider how we actually use: (i) the categorical assertion that an individual is following a given rule (that he means addition by “plus”); (ii) the conditional assertion that “if an individual follows such-and-such a rule, he must do so-and-so on a given occasion” (e.g., “if he means addition by ‘+’, his answer to ‘68+57’ should be ‘125’”). That is to say, we must look at the circumstances under which these assertions are introduced into discourse, and their role and utility in our lives. 1.2  Wittgenstein on rules There are two especially important thoughts emerging from the discussion about rule-following that I want to emphasize. The first of them concerns the nature of rules. What is a rule? It may be that a first answer that comes to one’s mind is a certain kind of sentence or a proposition expressed by the sentence “You should stop at red traffic lights”, or “The bishop moves only diagonally”. Can rules be indeed equated with some kinds of sentences? Wittgenstein saw that this cannot be the case, at least not generally. It seems that there may be unwritten rules that are difficult if not impossible to express. Just like it may be difficult or impossible to articulate explicit instructions for riding a bike, the following of which would prevent one from falling down, it might be difficult or impossible to articulate the rules for maneuvering

Now I can go on! 15 one’s way through a foreign community without offending someone or without doing improper things. The reason is that both these things may be practical achievements, matters of know-how which we gain through long-term practice. But there are also deeper reasons why there are unwritten rules which cannot be written down, and that is that language itself is based on rules, and if all rules would have to be “written”, we would already need language to establish a language. The Wittgensteinian take on this problem, however, is more general. The question, for him, is what is the relation between a rule (whatever it is) and the application of the rule. Do we need some rules that tell us how to apply the rule? It is quite clear that if this were generally the case, an infinite regress would be forthcoming. Brandom (1994, p. 20), calling the view that all rules are explicit regulism, formulates this in the following way: If correctnesses of performance are determined by rules only against the background correctnesses of application of the rule, how are these latter correctnesses to be understood? If a regulist understanding of all norms as rules is right, then applications of a rule should themselves be understood as correct insofar as they accord with some further rule. Only if this is so, can the rule-conception play the explanatory role of being the model of understanding of all norms. A rule for applying a rule Wittgenstein calls an “interpretation” (Deutung). Brandom then refers to Wittgenstein (PI, §201): There is an inclination to say: every action according to the rule is an interpretation. But we ought to restrict the term ‘interpretation’ to the substitution of one expression of the rule for another. Hence, once there is generally a gap between a rule and its application, we need a further rule to link the original rule to its correct application. And as the same holds for this further rule, an infinite regress is inevitable. Now the gap is inescapable when rules are “symbolic objects”: we can follow a symbol only if we interpret it, and we must interpret it correctly, i.e. follow the appropriate rules of its interpretation. Suppose we have an explicit instruction, such as “You should stop at red traffic lights”. In order to act as the rule tells us we should, it must be interpreted – she who is to abide by it is to understand it. Moreover, it must be interpreted or understood properly: if somebody were to understand it as “You should accelerate at red lights”, it would not work. But to interpret or understand it properly is to use the right rules for its

16  Now I can go on! interpretation. Hence, any explicit rule already presupposes some other rules. And if the presupposed rules were to be explicit too, an infinite regress would be imminent. What we therefore appear to need is a notion of rule which is intrinsically connected with its application, which, hence, in a sense is its own application. Thus, to know a rule is to know how to apply it. This, I think, led Wittgenstein to the conclusion that the most basic form of rulefollowing is a practical “mastering a technique” (PI, §199): It is not possible that there should have been only one occasion on which someone obeyed a rule. It is not possible that there should have been only one occasion on which a report was made, an order given or understood; and so on. – To obey a rule, to make a report, to give an order, to play a game of chess, are customs (uses, institutions). To understand a sentence means to understand a language. To understand a language means to be master of a technique. Another way to make the same point is insisting that there must be rules which one follows “blindly” (PI, §219): When I obey a rule, I do not choose. I obey the rule blindly. But if following a rule is a matter of a technique the follower masters, rather than following an explicit instruction, the rule can be hardly identified with a sentence. There may be simply no sentence involved. Not all rules are explicit; some are merely implicit in certain human doings. Not all rules are written down – and some of them perhaps cannot be written down – so we seem to need a more encompassing notion of rule, a notion associating a rule with something as a “social configuration” rather than with a sentence. 1.3  Wittgenstein on rule-following The other important thought emerging from the post-Wittgensteinian rule-following discussion concerns the nature of rule-following. Probably the most frequently discussed paragraph of PI, among the participants in this discussion, is §202: And hence also “obeying a rule” is a practice. And to think one is obeying a rule is not to obey a rule. Hence it is not possible to obey a rule “privately”: otherwise thinking one was obeying a rule would be the same thing as obeying it.

Now I can go on! 17 The argument seems to be as follows: There is certainly a difference between the situation when one follows a rule and when she only thinks she follows the rule. As a rule is an objective (or at least intersubjective) matter, it is clear that there must be a possibility that somebody gets it wrong, that she only thinks she follows the rule but is mistaken. Hence, the existence of this difference is a condition sine qua non for anything that is adequately called a rule. Now to maintain this difference, a society is needed – only it can provide for the corrective needed for the division of the rules that I think I follow into those which I really follow and those which I do not. This is an interpretation that takes Wittgenstein’s pronouncement “it is not possible to obey a rule ‘privately’” at face value, as tantamount to “a rule can be obeyed only publicly” and hence “a rule can exist only in the public sphere”. This is an interpretation endorsed by Kripke and by some other authors (vehemently, for example, by Kusch, 2006). But a host of philosophers disagrees: according to them, this is too a strong reading. It seems that almost everybody agrees with Wittgenstein that rulefollowing is a practice (or a technique). The point of the quarrel is how public this practice is required to be. The question where the disagreement comes to the open is whether Robinson Crusoe, on his deserted island, could have followed rules. While the proponents of the strong reading of Wittgenstein conclude that Robinson, having been deprived of his community, cannot have followed any rules, their opponents see this as the reductio ad absurdum of the strong reading, for they take it for obvious that Robinson could have followed some rules. Kripke tends to the view that though it is only community that may establish an environment for genuine rule-following (and is thus not far from the strong reading), there is a sense in which we can say that Robinson does follow some rules – if only from our own perspective. “Does this mean”, he writes (p. 110), “that Robinson Crusoe, isolated on an island, cannot be said to follow any rules, no matter what he does? I do not see that this follows. What does follow is that if we think of Crusoe as following rules, we are taking him into our community and applying our criteria for rule following to him”. Blackburn (1984, p. 85), on the other hand, disagrees. He agrees that the notion of “democratic harmony” can cover even the Kripkean account of Robinson: “An orchestra coming across a solitary player, concentrating hard and making noises, might well say that if he were with them, he would be doing well or badly making these noises”. But at the same time he is convinced that this is not enough: “The problem Crusoe poses is that he does have a practice (follows a tune) regardless of how we or anybody else think of him”.

18  Now I can go on! The most straightforward way a community can provide me with the feedback necessary to differentiate the situations when I really follow the rule and those when I merely think I follow a rule is that they correct my errors – the cases when I violate the rule, especially those which I do not perceive as violations. This comes naturally when people around me also follow the same rule, for correcting the violations is part and parcel of its following. If, however, this were the only possibility, it would mean that one cannot follow a rule save as a member of a community that follows the rule as a whole. This is the strong reading of Wittgenstein. Kusch (2002) presents a very fine-grained classification of the views according to which the strong reading and its complete rejection are only two extremes, in between which we can find several intermediary standpoints. The strong reading corresponds, according to Kusch, to what he calls the Strongest Present-Tense Community Thesis (p. 181): “An individual is able to follow a rule only if the individual is currently a participating member of a group in which the very same rule is followed by other members”. It seems clear how, if this is the case, the community can help the individual to maintain the gap between following the rule and merely thinking that one is following a rule: her peers are ready to correct her mistakes and slips. According to this view, one can follow a rule only if she is a participant of the communal practice of following the very rule. Consequently, there is no room at all for seeing Robinson as a rule-follower. There are several ways of relaxing this strong requirement. One is that it is not the following of a specific rule which necessitates the community, that it is rule-following more generally. Hence, according to this view, to be able to follow a rule, one must be a participant of the communal practice of rule-following, not of the practice of following the very rule. This is what Kusch terms the Strong Present-Tense Community Thesis (ibid.): “An individual is able to follow a rule only if the individual is currently a participating member of a group in which some rule or other – but not necessarily the same rule – is followed by other members”. In this view, once one is following rules under the communal auspices, she can start to follow a new rule all by herself. Another relaxation is based on the idea that following a rule is a skill, which one has to acquire under the communal auspices, but then she is in its possession and can go on with the rule-following even outside of the community. This yields us what Kusch calls the Past-Tense Community Thesis (p. 182): “An individual is able to follow a rule only if the individual has been, at some point in the past, a participating member of a group in which the very same rule, or some other rule, was followed by other members”. Note that if we accept this version, we will be able to accept that we may see Robinson as a rule-follower.

Now I can go on! 19 Then we can go on relaxing the condition further and require that to follow a rule one need not be a member of a rule-following community but only behave in the same way as a member of the community. Thus, we have Kusch’s Moderate Community Thesis (ibid.): “An individual follows a rule if and only if the participating members of an existing group judge parts of the individual’s behaviour to be similar to their behaviour in following the very same rule. The individual in question need not be a member of the group in question”. This, in effect, is the reading maintained by Kripke. Relaxing this still further, we can have the Weak Community Thesis (ibid.): “An individual follows a rule if and only if it is possible to imagine a community in which that rule is followed by participating members”. The weakest reading, Kusch’s Weakest Community Thesis, then amounts to the following position (ibid.): “An individual follows a rule if and only if the participating members of an existing group judge aspects of the individual’s behaviour to be similar to aspects of their behaviour in following some rule or other. The individual in question need not be a member of the group in question”. It follows that saying that rule-following is a “public business” is saying something not very specific. It covers a variety of different views. Hence, it is important to specify what exactly is the role of the public within the individual cases of rule-following and how rules that are followed emerge out of the social interaction. 1.4  Wittgenstein on language We have already noted that probably the most popular term coined by Wittgenstein was “language game”. It alludes to the fact that we do a great variety of things with language and that it is in vain that we try to find their common denominator (PI, §65): Here we come up against the great question that lies behind all these considerations. – For someone might object against me: “You take the easy way out! You talk about all sorts of language-games, but have nowhere said what the essence of a language-game, and hence of language, is: what is common to all these activities, and what makes them into language or parts of language. So you let yourself off the very part of the investigation that once gave you yourself most headache, the part about the general form of proposition and of language”. And this is true. – Instead of producing something common to all that we call language, I am saying that these phenomena have no one thing in common which makes us use the same word for all, – but

20  Now I can go on! that they are related to one another in many different ways. And it is because of this relationship, or these relationships, that we call them all “language”. Thus, while Wittgenstein’s earlier project was to reduce everything we do with language to its picturing capacity, his later philosophy was led by the recognition of the fact that neither can picturing play this role, nor can we find anything else in its stead. Our language games are irreducibly diverse. This led Wittgenstein to a view of language very different from the one prevailing among his contemporaries, according to which expressions are representations of something extralinguistic and to elucidate their meanings we need to find out what exactly they represent and how exactly they manage to represent it. In contrast to this, Wittgenstein came to see expressions as tools, as vehicles of various kinds of activities we carry out. Hence, instead of explaining our linguistic activities by explaining language and explaining language by means of explaining meaning, he strove to explain language and meaning by explaining our linguistic activities. The most basic question is therefore no longer What does an expression represent? or What does it mean? but rather What does language, or a particular expression, allow us to do? This led him to a view of language very different from the ordinary one, based on the assumption that expressions are representations (or pictures, as he himself used to put it). Wittgenstein saw how nontrivial problems surround the concept of representation, which may seem prima facie straightforward (PI, §26): One thinks that learning language consists in giving names to objects. Viz, to human beings, to shapes, to colors, to pains, to moods to numbers etc. To repeat – naming is something like attaching a label to a thing. However, imagine that you literally attach a label, perhaps a wooden table with an engraved word, to a thing, perhaps to your house. Do you thereby give the house a name, do you let the word on the table stand for the house? Perhaps. Maybe in your neighborhood it is usual to give names to houses in this way. But in a different neighborhood such a table may be taken as displaying a magic word intended to keep evil spirits away from your house. Or it may be taken as something to entertain the passers-by. All of this depends on the infrastructure of habits, rules and customs of the neighborhood. Therefore, Wittgenstein concluded, the ultimate level of explanation of language is not the level where we talk about symbols, representing etc., but an underlying level, on which we talk about customs, conventions or rules. Ultimately, he thinks that the whole

Now I can go on! 21 level of meanings as entities is an artifact of the ways we talk about rules (Wittgenstein, 2016, p. 195): [T]he mere fact that we have the expression “the meaning” of a word is bound to lead us wrong: we are led to think that the rules are responsible to something not a rule, whereas they are only responsible to rules. Therefore, he proposes an utterly different route to the explanation of meaningfulness, leading him to a “use theory of meaning”. The meaning of an expression is determined by the way in which the expression is used within our language games; and as the language games are – more or less – rule governed, it is determined by the rules. From this vantage point, we could perhaps understand why Wittgenstein’s musings about rules in PI are so intertwined with his reflections of the plurality of language games. If our question is how our language games work and which new “ways of words” they open up for us, then the analysis of the rules of games, insofar as they are rule governed, is of the utmost importance. Understanding chess, for example, certainly presupposes understanding its rules. To be sure, certainly not all language games are governed by rules in the way chess is. And Wittgenstein might be reluctant to even admit that all language games are rule governed, for then “rule-governedness” would be essential to them, whereas language games, according to him, have no essence. But be that as it may, it is clear that most of our language games do involve rules, and to understand them is to get a clear command of them (PI, §125): The fundamental fact here is that we lay down rules, a technique, for a game, and that then when we follow the rules, things do not turn out as we had assumed. That we are therefore as it were entangled in our own rules. This entanglement in our rules is what we want to understand (i.e. get a clear view of). Of course, this is not too surprising; that there are correct and incorrect ways of using expressions, and that it is the former one we learn as novices, is quite clear. However, it was Wittgenstein who pointed out how intricate the concept of a rule of (not only) language is. 1.5 Summary The first philosopher to stir up the modern discussion of rules as something fundamental, rather than only marginal, for philosophy was the

22  Now I can go on! later Wittgenstein. On many pages of his Philosophical Investigations, he discusses the concept of rule and especially the problem of what does it take to follow a rule. Such attention paid to rules in the context of philosophy of language and of mind was unusual – Wittgenstein was the first to understand the importance of normativity even in these areas. Based on this, Wittgenstein put forward a picture of language and of our linguistic activities very different from the usual one: according to him, the key to understanding language is understanding our language games; and to understand our language games, we need to understand the rules involved in them. However, many things regarding the role of rules remain slightly enigmatic, so it is no wonder that Wittgenstein’s exposition sometimes produces more questions than it answers. In any case, Wittgenstein excavated some important facts regarding rules and normativity in general. One of Wittgenstein’s crucial observations was that rules cannot be generally identified with anything like sentences – the most fundamental rules must be implicit. (The reason is that a sentence, to be usable as a rule, must be interpreted and it must be interpreted correctly, which means that here there must be some rules of its interpretation.) Another of the crucial observations is that rule-following, and hence rules, must be a public matter. (Here the reason is that there must be a distinction between following a rule and merely thinking one is following a rule, which can exist only in the public sphere.) Wittgenstein thus, among other things, introduced a brand new way of looking at language: a way of looking at it as a basically rule-governed enterprise. Notes 1 See, for example, Baker & Hacker (1984), McDowell (1984), Goldfarb (1985), Boghossian (1989), Haugeland (2000) or Kusch (2006). 2 Of course, we can imagine a scenario where things are more complex: in a community there might be, for example, a consensus that correct greetings are those that accord with the tradition, where everybody might be mistaken with respect to the tradition (for example, because of systematic errors in the chronicles). But this possibility does not disturb the fact that normally in such cases appropriateness reduces to what is taken to be appropriate.

2

Creature of rules

2.1  Sellars on rules Wittgenstein, we saw, paid a lot of attention to rules, but he was never quite explicit with respect to what roles the rules play concerning language and the whole human form of life. At approximately the same time, another philosopher (residing on the other side of the globe), Wilfrid Sellars, was more explicit: “To say that man is a rational animal”, he wrote (Sellars, 1949, p. 311), “is to say that man is a creature not of habits, but of rules”. In this way, he put forward the explicit claim that rules might be something like the differentia specifica of us humans – the idea that fueled the writing of my book. Why are rules, according to Sellars, so important for us? It is not easy to say; Sellars is a complicated thinker and his thought developed via quite tortuous trajectories. At some points of his development, his views about rules and normativity may seem to be quite sharp and clear; at other times, they are far more intricate and obscure. What I take from him are mostly the sharp and clear views, expressed most explicitly in his paper quoted above (Sellars, 1949).1 And what I think is the crucial articulation of Sellars’ view of the role of normativity in human life is the passage immediately preceding the quote above (pp. 297–298): Certainly, we learn habits of response to our environment in a way which is essentially identical with that in which the dog learns to sit up when I snap my fingers. And certainly these learned habits of response – though modifiable by rule-regulated symbol activity – remain the basic tie between all the complex rule-regulated symbol behavior which is the human mind in action, and the environment in which the individual lives and acts. Yet above the foundation of man’s learned responses to environmental stimuli – let us call this his tied behavior – there towers a superstructure of more or less developed systems of rule-regulated symbol activity which constitutes DOI: 10.4324/9781003388876-3

24  Creature of rules man’s intellectual vision. It is in terms of such systems of rule-regulated symbol activity that we are to understand an Einstein’s grasp of alternative structures of natural law, a Leibnitz’ vision of the totality of all possible worlds, a logician’s exploration of the most diversified postulate systems, a Cantor’s march into the trans-finite. Such symbol activity may well be characterized as free – by which, of course, I do not mean uncaused – in contrast to the behavior that is learned as a dog learns to sit up, or a white rat to run a maze. On the other hand, a structure of rule-regulated symbol activity, which as such is free, constitutes a man’s understanding of this world, the world in which he lives, its history and future, the laws according to which it operates, by meshing in with his tied behavior, his learned habits of response to his environment. I read this so that what is characteristic of “man” is her ability to get over the merely “tied behavior” and become “free”. At first sight, this may seem to be a matter of acquiring a new kind of mind (or perhaps soul?); a mind which is not only reactive, but rather proactive, not merely able to respond to external stimuli, but also to freely choose new courses of action. But Sellars does not want to characterize the human condition in terms of such a characterization of the human mind – he is too well aware how elusive such characterizations tend to be, and so he is inclined to behaviorism (though, ultimately, to behaviorism of his specific kind, which he calls “verbal”2). Hence the most appropriate characterization he finds is in terms of human normativity: we humans are free in that we are not tied to the environment – and to each other – not only via conditional reflexes and habits, but rather via the peculiar attractive and repulsive forces of the “superstructure” of rules we have erected atop of our habitual background and “which constitutes our intellectual vision”. Sellars tends to see our progress from habits to rules as entering a new kind of realm. In this sense, his view is reminiscent of Kant, who accounted for human exclusivity in term of the distinction between the realm of the concept of nature and the realm of the concept of freedom:3 unlike other entities in our world, we live not only in the former, but also in the latter. The distinction put forward by Sellars (1962a) is similar to Kant’s: it is the distinction between two ways of envisaging the world which he calls the scientific image and the manifest image. The scientific image is something we have achieved via our painstaking effort to develop rigorous and reliable methods of exploring the world; it is inhabited by entities interacting with each other, within space-time, according to the causal laws discovered by natural science. Of course, we people belong to this image: as organisms which behave in certain ways. And there is a sense in which it is this picture that discloses the “true

Creature of rules 25 nature” of our world. This is revealed by Sellars’ version of Protagoras’ Homo mensura claim, which has been dubbed Scientia mensura: “In the dimension of describing and explaining the world, science is the measure of all things, of what is that it is, and of what is not that it is not” (Sellars, 1956, §41). However, there is something missing from the scientific image; something we can find only in the manifest image. We are in the scientific image, but merely as organisms, not as persons, acting and being responsible for their actions, and handling meanings. In this guise, we are present only within the manifest image, which is our primordial image of the world, predating the scientific one.4 And though there is a sense in which the scientific image is an upgrade on the manifest image, it cannot simply replace it, because we cannot wholly abandon – to use Max Weber’s vivid term – the “enchanted” version of the world. Sellars claims that the most central entities of the manifest image are persons, and he claims that the irreducibility of persons to organism, and consequently the irreducibility of the whole manifest image to the scientific image, “is the irreducibility of the ‘ought’ to the ‘is’” (Sellars, 1962a, p. 39). This means that the scientific image lacks a normative dimension that is (crucially) present in the manifest image, and it is the absence of this dimension on the scientific image that constitutes the crucial difference between the two images. And we must be “creatures of rules” to be able to maintain also the manifest image. Of course, characterizing humans as creatures of rules requires a scrutiny of the concept of rule. Again, I think that Sellars is not always optimally clear, but in general, he suggests that “a rule is an embodied generalization which to speak loosely, but suggestively tends to make itself true” (Sellars, 1949, p. 299). This indicates that according to him, a rule (at least insofar as it is explicit) has something in common with a (general) declarative sentence that, however, it has also the inverse “direction of fit”, like an imperative. (If a sentence, say Everybody greets everybody else in the morning, does not fit the world, hence if everybody does not greet everybody else in the morning, then if it is meant as a description, the failure is on its part, for it is false, while if it is meant as an imperative, then the failure is on part of the world, for it is not obeyed). Now explicit rules are, according to Sellars, somewhere between declarative and imperative sentences – they may have both directions of fit.5 Typically, such a sentence makes explicit what is already implicit in the behavior of the relevant community, viz. recapitulates a “tradition”, while aiming at making the members of the community further comply with it. (This appears to ignore the fact that in developed communities, there are rules which do not continue any existing tradition and hence act as pure imperatives.)

26  Creature of rules Hence, a rule has the features of an imperative, which is typically a tool of one’s effort to make other people fit their vision of the world. But rules are usually not tools of a single person. They must be backed up by an authority, which is often distributed – often we all see to it that we all follow a rule. Therefore, more than simply fitting others into the vision of a single person, they have to do with fitting everybody into the common vision of everybody. I think – though Sellars does not say this in so many words – that one of the important reasons why we need the manifest image over and above the scientific one is that we humans have become future-oriented creatures who live in terms of plans and ventures which we project into the future.6 (This, of course, is closely connected with our freedom mentioned above.) And rules, I am convinced, help us with such projections. Rules aim at how we imagine things to be, and though, of course, they allow us to assess past and present things, their main scope of activity is the future: they spell out how we plan things to go in the times to come.7 They are tools, if you want, of our “wishful thinking” – and if we want it to be “wishful thinking” not in the usual pejorative sense of the word, we must keep in mind that they concern not how things are, but rather how we want them to be. They do not concern reality, but our vision of it. Thus, the manifest image, unlike the scientific one, is not a matter of description of the real world, but rather envisaging of its ideal version. Hence the entities present on the manifest image (like persons and meanings), from the viewpoint of the scientific image, do not exist. But they are not just some regrettable delusions, but they are also components of ideal states-of-affairs that we need to orient our lives, to get a direction to move into the future. 2.2  Rules as regularities? Like Wittgenstein, Sellars points out that the most primitive form of the existence of a rule is not a linguistic one: “The mode of existence of a rule is as a generalization written in flesh and blood, or nerve and sinew, rather than in pen and ink” (Sellars, 1949, p. 299). It is only when we come to acquire language that we can make explicit the rules that were only implicit in behavior before (and we can also establish new rules by means of language). The “making explicit” of the rules was raised to an emblematic human activity by Brandom (1994). However, if rules are not linguistic objects that would be coined by somebody’s declaration, what are they? What are we talking about when we say that members of a society accept a rule, or that the rule is in force in the society, or that somebody follows a rule? It might seem that to explain this is not so difficult: To accept a rule is to perform as the rule instructs

Creature of rules 27 us. To accept You should stop at red traffic lights amounts to stopping at red traffic lights, whereas to accept The bishop is moved only diagonally amounts to moving bishops only diagonally. But it is obvious that this at least cannot be the whole story. It seems plausible to assume that in the typical case, in a society which accepts a rule, we will encounter many more cases of behavior that complies with the rule than that violates it. But it does not seem plausible to assume that there will be no violation whatsoever. (Also there may be rules that are violated more often than not.) Moreover, and this is more of an obstacle, it does not seem that everything that people do not violate is a rule accepted by the society. Consider Running faster than light is prohibited, One ought to sleep at night, or You should not use your bishop to pick your teeth. Presumably, all these instructions are complied with in our communities (certainly more often than not), and yet we would hardly want to say that they are rules accepted by us. One moral is that it makes little sense to call a rule something that cannot be – as a matter of principle – violated. This is what disqualifies Running faster than light is prohibited as our rule. And this is a point worth dwelling on a little, for there are things which are sometimes called rules and which do not fulfill this requirement and hence are not rules in our sense. Natural laws, for example, are not rules in this sense. Nor is anything that follows from them, such as principles governing the functioning of the brain. This is a crucial point that cannot be overemphasized. A lot of theories agree that language is a matter – partly or wholly – of rules. However, what they understand by rule is sometimes quite divergent. In particular, not everything such theories take to be a rule is a rule in the sense of this book. What we claim is that the kind of rules that underlie language are “social” rules: the kind of rules that people impose on people and that are usually not enacted in a non-exceptional way. Take, for example, the “rules” of Chomskyan universal grammar (Chomsky, 1955; 1986; 2000). They are sub-personal – in so far as we follow them, it is because of the setup of our language faculty, which we cannot influence.8 Hence they are not rules in our sense of the word. They are closer to principles of the functioning of a machine – though the machine can malfunction, we do not see this as disobeying the principles – and in so far as the rules cannot be disobeyed, they cannot be obeyed for obeying presupposes the possibility of disobeying. Unlike laws of nature that determine what can and what cannot happen, rules determine what should happen (though it may not happen) and what should not happen (though it may).9 However, both One ought to sleep at night and You should not use your bishop to pick your teeth may be violated, though they are presumably

28  Creature of rules not, or not very often. In the case of the first of them, sleeping at night seems to be a habit (with a very clear rationale) rather than a rule. (Though we can easily imagine a society in which this habit starts to be something that one should do – viz. a rule.) In the case of the second, we would not even talk about a habit – picking one’s teeth with a bishop would be a behavioral pattern that could be called a (bad) habit, but not doing so is not what we would call a pattern, let alone a habit. Hence, that something is not violated (though it could be) is not a sufficient condition for it to be a rule. At a very general level, we can say that the above proposal amounts to saying that a behavior is rule-governed once it is simply regular, and that this vastly over-generates. In addition to what we have already seen, it would also render a billiard ball moving in a straight line as following a rule (namely Descartes’ first law of motion) in the same sense in which a driver keeping her car on the right (or left) side of a highway follows a rule (of traffic). This would be neglecting a huge difference – one which Kant proposed to render as the distinction between “working according to laws” and “acting in accordance with the representation of laws” (Kant, 1785, §4:412). However, Kant’s term “representation” (“Vorstellung”, sometimes also translated as “conception”) appears to suggest that what makes the difference is something sealed within the mind of the rule follower, which we want to avoid here: we want to capture the difference on the behavioral level. This goal is based on the assumption that rule-governed behavior differs from the merely regular one not merely by the presence, in the minds of the protagonists of the former one, of something that is absent in those of the protagonists of the latter one. There must be a difference in behavior – on the penalty of the two forms of behavior being effectively indistinguishable. The idea is that those who “act in accordance with the representation of laws” differ from those who merely behave “according to laws” in that they manifest their comprehension of the rules (“laws”) by some practical attitudes going beyond complying with the rules: attitudes that “render” the actions correct by the rule as correct and those which are incorrect as incorrect. And one of the tasks of this book will be to characterize these attitudes in greater detail. Hence, the identification of rule-governed behavior with behavior that is only regular over-generates; and as we already saw that the identification of rules with sentences under-generates (for it is not able to accommodate unwritten rules), it seems that the genuine concept of rule must be somewhere in between these two extremes. It seems, then, that to give an adequate account of rules we must, as Brandom (1994, §1.III) puts it, steer clear of both the Scylla of regulism (= the assumption that rules are always explicit and we follow them by grasping them as instructions) and

Creature of rules 29 the Charybdis of regularism (= the assumption that rules are mere regularities). It seems that rules are more than mere regularities, though they are not always explicit instructions. What, then, could they be? 2.3  Pattern-governed behavior Can somebody behave in a specific way because of some rule without following the rule in the sense of being led by it? A possibility seems to be that she is forced to behave in this way by another person, who is led by the rule. This might be the case of behavior that is not merely regular (for it takes place because of a rule), while at the same time not being rulefollowing in the sense that she who displays the behavior would display it because she is being led by the rule. Imagine a toddler who starts to be taught her mother tongue. She emits various kinds of sounds and gets a feedback from her parents: she is encouraged to emit some of the sounds in certain situations, while being diverted from emitting some sounds in some circumstances.10 Her emissions start to reflect the patterns that, according to the parents, are prescribed by the rules of language. Her emissions are thus not rule-followings in the sense that she would be led directly by the rules, and yet they are here because of the rules (for they have been brought about by the parents who do follow the rules). We can call this kind of behavior rule-governed; and in view of the fact that a rule typically institutes a behavioral pattern we can speak, more generally, about pattern-governed behavior. This is a term introduced by Sellars (1954), who considered not only patterns formed by rules, but also, e.g., patterns emerging out of evolution. Bees, for example, display their specific kind of dance, and Sellars insists that there is a sense in which we can say that the “turnings and wigglings” of an individual bee “occurs because they are part of a complex dance”, despite it being clear that it is not the case that “the bee envisages the dance and acts as it does by virtue of intending to realize the dance” (p. 208). This suggests that there is a kind of regularity of behavior, which is “purposeful” in that it emerges out of the process of natural selection, and the acts of individual bees are building blocks of the regularity. Sellars (1954) thus sees a middle way between regularism (as he puts it, the conviction that rule-following is “merely conforming to rules”) and regulism (the conviction that it is the “obeying of rules” in the sense of comprehending them and acting as they instruct us). (According to Sellars, the former amount to “doing A in C, A’ in C’ etc. where these doings ‘just happen’ to contribute to the realization of a complex pattern”, while the latter to “doing A in C, A’ in C’ etc., with the intention of fulfilling the demands of an envisaged system of rules”.) The middle way is a species of

30  Creature of rules his pattern-governed behavior: the species in which the relevant pattern is enforced by rules. Because the terminology starts to be a little bit labyrinthine, let us appropriate the refined terminology introduced by Stovall (2021). Let us call the behavior that is merely regular pattern-conforming behavior. Pattern-governed behavior, then, is the behavior that is here because of the pattern (like the bee dance). Rule-governed behavior will be a species of pattern-governed behavior, where the relevant pattern is constituted by rules (like the linguistic behavior of infants who still do not understand language). And what I will call rule-following11 behavior is the kind of behavior that is displayed by those who know the rules and abide by them. However, rule-following behavior is still of two kinds. One is the case when the follower can abide by the rule as a matter of know-how: she knows how to do it like one knows how to ride a bicycle, without necessarily being able to articulate the rule explicitly. Let me call this behavior rule-enacting. (I am afraid this term may be misleading, but I can think of none better.) And finally, we have the case when one follows the rule on the basis of comprehending an explicit instruction. Hence we have:

• pattern-conforming behavior, which is a behavior that is merely regular, conforms to a pattern

• pattern-governed behavior, which occurs because of a pattern • rule-governed behavior, which occurs because of some rules • rule-enacting behavior (also called rule-following), which occurs be-

cause the individual who displays it has a know-how amounting to following the rule • rule-obeying behavior (also called rule-following), which occurs because the individual who displays it comprehends an explicit rule and is instructed by it Note that rule-governed behavior arises out of a certain mixture of pattern-conforming and rule-following. There is an “enforcer” who follows rules, she knows them (be it merely a know-how) and strives to do what they require; and there is an “enforcee”, who comes to display a pattern, because she is enforced into it – she does not know the rules and merely behaves in a regular fashion. Hence, on one side of the interaction, there is a fully fledged rule-following behavior, whereas, on the other side, there is a mere pattern-conforming behavior, i.e. mere regularity. It is the interplay of these two that produces the rule-governed behavior: the merely pattern-conforming behavior becomes more than that because it is brought into being by rule-following tutelage.

Creature of rules 31 Given all these kinds of behavior, the rule-governed behavior is a subspecies of the pattern-governed one. The only difference is that the pattern to be instantiated by the behavior is delimited by rules, like when teaching and learning language. However, from the phylogenetic perspective, the rule-governed behavior is dependent on the pattern-governed one, because rule-following, on which rule-governed behavior depends, is itself a (very complex) behavioral pattern. This overarching pattern comes into being because some of our behavioral patterns emerging out of evolution as being adaptive come to be supported by a sophisticated behavioral superstructure, helping to hold the patterns in place, and possibly “adjusting” them. (We will try to analyze, in the rest of the book, the nature of such a superstructure.) In this way, the behavior becomes rule-governed, instead of merely pattern-governed, because we call the crucial components of the behavior regulated in such a way rules. Note also that not everybody can obey a rule. To do so, one has to be able to understand a representation (usually linguistic) of the rule. Typically, to do so one must be already a discursive, concept-using creature. This is the problem which, we saw, led Wittgenstein to conclude that not every rule may be explicit. And Sellars’ pattern-governed behavior is meant to be a part of the solution of this conundrum on the ontogenetic level. Those who enter human society become subject to the pressure of the elders who already know the rules and feel the urge to instill the corresponding pattern into them. But they come to not only instantiate the pattern, but also take it as something that ought to be, and what is thus to be required. In this way, they themselves become potential teachers, ready to educate the next generation. The proliferation of the pattern, by exclusively paragenetic means, is up and running. The idea of pattern-governed behavior and of the enforcees naturally mutating into enforcers nicely explains the ontogenesis of rule-following: the enforcers (who are already rule-followers) make the enforcees (who are not) instantiate the appropriate patterns, while the latter slowly adopt the patterns as something that ought to be and join the enforcers. They internalize the external directions their teachers give them – but not only in the sense that they start to behave accordingly, but also in the sense that they start to make others behave accordingly. In this way, their behavior becomes rule-governed and slowly “mutates” into rule-following behavior. What it does not explain so clearly is the phylogenesis of rule-following. How did the human-specific features that have made it possible that we became rule-followers and let the rules (together with everything that rests on them) proliferate from generation to generation? Two things are needed for this proliferation to take place. First, rule-governed behavior must have the tendency of mutating into a

32  Creature of rules rule-following one. This, I maintain, is the peculiar human tendency towards rule-following, which I want to focus on in some of the upcoming chapters. Second, teachers can produce teachers, but at the beginning, it would seem, there must be some teachers who are not produced by previous teachers – some ur-teachers, as it were. This is likely to be explainable by a bootstrapping – at the beginning, there were individuals who came to appreciate some rudiments of rules, which they implanted into their youths, who later appreciated rules in a slightly less rudimentary shape, etc. However, even the increased tendency of the elders to teach the youths as well as the tendency of the youths to become taught, which must underlie such a bootstrapping, mark the specific evolutionary trajectory of our species (Laland, 2017). 2.4  Ought-to-do’s, ought-to-be’s and the rules of language Consider the first of the two problems mentioned in the last paragraph of the previous section. The problem is that the infants entering human communities are not yet agents who could follow rules; indeed, an important outcome of their education is that they become agents, rulefollowers and speakers of the language of the community. Sellars (1969) solves this by assuming that aside from ordinary rules, which regulate human actions and which he dubs ought-to-do’s, there is a different kind of rules, ought-to-be’s, which are not directly concerned with what one should do, but rather with in which state something should be. The rule that can be articulated as Karl, you should drive on the right side of the road is an example of an ought-to-do, while Cars should go on the right side of the road is an example of an ought-to-be. The interplay of these two kinds of rules is important for the proliferation of rules addressed in the previous section. The ought-to-do’s are just instructions and can be followed only by those who grasp them – they can only be the source of rule obeying. (Karl can follow the instruction contained in the sentence You should drive on the right side of the road only if he understands it.) The ought-to-be’s, in contrast to them, are not construable as instructions, as they are not explicitly directed at an agent. Rather, they mark a state as desirable. They may lead to actions, because they bear ought-to-do’s, via a specific kind of generic “practical syllogism”: If something ought to be, and doing A is likely to bring it into being, then do A! (If Karl endorses the statement Cars should go on the right side of the road, then as he knows that his driving at the right side will contribute to cars going on the right side, he will derive Karl, you should drive on the right side of the road!). Again, one must comprehend the relevant concepts to use the ought-to-be to carry out this syllogism.

Creature of rules 33 Aside of being an agent following ought-to-do’s and endorsing oughtto-be’s, a person may also be a subject of an ought-to-be. And, according to Sellars, there is an important difference between X, you should do A, which requires X to be an agent comprehending A, and X should be in state φ, which does not involve any such requirement. The latter is rather a “freefloating” rule, which is up for grabs for any agent and comprehender (including, possibly, X herself). (The difference is that between Karl, you should use “This is a dog” only if you point at a dog, which instructs Karl what to do, and Karl should use “This is a dog” only if he points at a dog, which, especially in the situation when Karl is a toddler, can be picked up by anybody and used to educate Karl.) Sellars concentrates especially on rules of language, for in their case it is clear they cannot be propagated only by explicit instructions. And Sellars’ (1969, p. 512) picture of language learning is that of moving from the position of a subject of certain ought-to-be’s to the position of their endorser: [T]he members of a linguistic community are first language learners and only potentially ‘people,’ but subsequently language teachers, possessed of the rich conceptual framework this implies. They start out by being the subject-matter subjects of the ought-tobe’s and graduate to the status of agent subjects of the ought-to-do’s. Linguistic ought-to-be’s are translated into uniformities by training. This indicates that if the teachers of Karl endorse the ought-to-be of the form One should be in a state φ (e.g. One should say “This is a dog!” only when pointing at a dog), their tutelage brings about, in the long run, not only Karl’s being in state φ (saying This is a dog! only when pointing at a dog), but also Karl’s comprehension and endorsement of the ought-to-be (and consequently deriving commands of the form I should do so as make X say “This is a dog!” only when pointing at a dog). In short, when educating humans (or candidates of humanity), forcing behavioral patterns results not only into the patterns’ coming into being, but also into the patterns being endorsed as ought-to-be’s. Hence, what is crucial here is the diachronic aspect of the interplay between rule-following and rule-conforming. Rule-following, on part of the enforcers, brings about the rule-conforming of the enforcees, and then rule-conforming of the enforcees brings about, in the long run, their

34  Creature of rules rule-following. It is this last move that closes the circle of language reproduction and which is the heart of the matter. An important point regarding rules of language that Sellars (1953a) brings forth is the identification of the kind of rules that are crucial for language – they are, he suggests, the rules of inference. This may be somewhat surprising – for it appears to bring language into too a close connection with logic. But Sellars, I think, has good reasons for this suggestion. In contrast to Wittgenstein, he does not believe that all language games are on the same level, that there is no game that would be central, hence that any kind of speech act is as important as any other. Sellars gives pride of place to what deVries (2005, p. 135) termed the “knowledge game” (Brandom, as we will see later, introduced the apt term “game of giving and asking for reasons”): “In the course of a young human being’s development it becomes increasingly appropriate – indeed, increasingly necessary – to characterize his or her behaviour in terms of moves in the knowledge game made with regard for and because of the rules constituting that practice”. And as claims to knowledge are most directly manifested by assertions, it is assertions that are accentuated along with the accentuation of this kind of game. It follows that the crucial rules of language are those that characterize the speech act of assertion. A straightforward case of such a rule would be one stating under which conditions we may assert a given sentence. The conditions may be extralinguistic circumstances: The sentence “This is a dog” may be asserted when pointing at a dog. But the conditions may also concern the assertability of other sentences: The sentence “This is a dog” may be asserted when both “This is a dog or a cat” and “This is not a cat” may be asserted. Also there may be consequences of the assertability of a sentence: If “This is a dog” may be asserted, then both “This is not a cat” and “This is an animal” may be asserted. Note that in specifying these conditions, we often specify what counts as a reason for the sentence in question. The point is that a necessary (not sufficient) condition for A being the reason for B is that B is inferable from A. This only reflects the close connection among assertion, reason, justification and knowledge. Now a lot – though not all – of the above conditions can be rewritten as inferences:

This is a dog or a cat      This is not a cat This is a dog



This is a dog This is not a cat



This is a dog This is an animal

Creature of rules 35 If you ask a logician, then she will be likely to select the first of these as a prototype of an inference, for it appears as an instance of a general schema known as disjunctive syllogism:



A ∨ B ¬B A

But Sellars insists that aside from such logical rules, we must consider rules which he calls material and which are exemplified by the other two cases above. Indeed, he urges that it is these kind of inferential rules that are fundamental – it is they who establish the roles making ordinary sentences (and, as a consequence, words that are parts of the sentences) meaningful. Sellars is thus, like Wittgenstein, convinced that the meaningfulness of linguistic expressions is a matter of their involvement in our language games, of their roles within the games, but, moreover, he is convinced that the roles have essentially to do with the rules of inference. This may be seen as connected with the fact that we humans are distinctively rational creatures, viz. we recognize the force of reasons, where a reason for a claim is something the claim is inferable from. In general, Sellars sees the roles of expressions within our language games as determined by three kinds of “transitions” (Sellars, 1974). The language entry transitions lead us from the world to language: they are the transitions of the kind that go from seeing a dog to uttering This is a dog. The intralinguistic transitions remain inside of language: they are the inferences proper. And the language exit transitions lead from the language to the world: they may be exemplified by the one leading us from We will go this way to going this way. However, we should not forget that it is not the individual transitions that determine the meanings of expressions, it is the rules which govern them. And Sellars holds that it is the rules of inference that are central, they correspond to the rules of chess according to which the pieces are moved. (While the entry rules may instead be compared to the regulation according to which the pieces are arranged on the chessboard in the initial position.) However, despite this prominent place reason and reasoning plays within Sellars’ theory of our linguistic transactions, he would not see it as something achieved independently of language and underlying our linguistic abilities. On the contrary, reason in a sense is a product of reasoning, and hence a product of language that must underpin reasoning. And, as language is a product of rules, rules come to appear as the ultimate foundation. We will consider this order of explanation in greater detail in the next chapter.

36  Creature of rules 2.5 Summary Wittgenstein paid a lot of attention to rules, but he was never quite explicit with respect to what roles the rules play concerning language and the whole human form of life. Wilfrid Sellars was more explicit: “To say that man is a rational animal”, he wrote, “is to say that man is a creature not of habits, but of rules”. In this way, he put forward the explicit claim that rules might be something as the differentia specifica of us humans. Like Wittgenstein, Sellars thought that rules, in general, cannot be anything like sentences, but, on the other hand, he stressed that they cannot be mere regularities. Hence, he sought a middle way between the Scylla and Charybdis of these two views, and he saw a solution in his “pattern-governed behavior”. This kind of behavior exists because of a rule (or, more generally, a pattern), but it is not the result of its protagonists following the rule in that they are instructed by it. An example is the rudimentary language of toddlers: it is here because of certain rules (namely the rules their parents and teachers wanted the behavior of the toddlers to conform to), but not because the toddlers would themselves follow the rules (they are too young to be able to comprehend them). And since the toddlers do grow up to comprehend the rules and its protagonists become teachers of new toddlers, this can also explain, according to Sellars, the proliferation of language from generation to generation. Sellars was also more specific than Wittgenstein with respect to those rules of language that have to do with meanings – they are, according to him, inferential rules. Meaning thus comes out as an inferential role of an expression. Notes 1 Further development of these views can be traced through his subsequent papers, especially Sellars (1953a, 1954, 1969). For a thorough analysis, see deVries (2005) and O’Shea (2007). 2 Sellars’ doctrine of verbal behaviorism (VB) does not go so far as to claim that thinking is generally talking to oneself, but it maintains that we understand thought on the model of linguistic utterances. As Sellars (1974, pp. 418–419) himself puts it: “According to VB, thinking ‘that-p,’ where this means ‘having the thought occur to one that-p,’ has as its primary sense saying ‘p’; and a secondary sense in which it stands for a short term proximate propensity to say ‘p’”. 3 Kant introduces these terms explicitly only in his third Critique (Kant, 1790), though the underlying distinction is clearly important for the whole of his practical philosophy. See, e.g., Risser (2009) for a discussion. 4 More precisely, Sellars reckons that both the images descended from an urimage he calls original, on which everything was a person. But while the manifest image is a direct continuation of the original one, only with everything save us stripped of the status of person, the scientific image is completely reformed. 5 Millikan (1995) speaks, in this context, about “pushmi-pullyu representations”.

Creature of rules 37 6 Though Sellars’ stress on intentions that are turning into volitions to guide our behavior (see, e.g., Sellars, 1968, Chapter VII) clearly does betray the orientation towards the future. 7 Cf. Stovall (2022), who elaborates on this “planning” aspect of the Sellarsian view of us humans. 8 As Partee (2018, p. 173) characterizes the Chomskyan notion of rules: “There is no such thing as ‘not knowing’ the syntactic rules of your language—what you know defines what your language is”. 9 It is worth noting that Chomskyan nativism, which used to be the leading paradigm in linguistics, is slowly giving way (Fernald & Marchman, 2006). Though we are certainly pre-disposed to acquire certain kinds of rules and languages, the picture that the rules are already “wired in” in our “language faculty” has come to be disputed (Christiansen & Chater, 2016). 10 From the current perspective, it is not quite clear how active the parents (and other adults) must in this respect be. According to recent research, it does not seem that much. This is connected with the extraordinary ability of toddlers to identify rules, which we deal with in Section 7.1. 11 Here I deviate from Stovall, who uses the term rule-following for what I am going to call rule-obeying. Also, I replace his rule-representing by my ruleenacting, which is not a purely terminological replacement.

3

Preliminaries I Rules and other human gear

3.1  Rules and language The writings of both Wittgenstein and Sellars render rules as something crucially characteristic of us humans that, nevertheless, are somewhat enigmatic. Rules, we saw, cannot be generally identified with symbolic items (such as imperative sentences) – because there are no symbols without interpretation and there is no interpretation without rules. So the task of the further chapters will be to elucidate the concept of rule, the way we humans follow rules and the question how important are rules for our human nature. Are they more than an epiphenomenon of what is sometimes called our “ultrasociality” (Richerson & Boyd, 1998)? We humans, no doubt, are peculiar animals. During the last eight-odd million years, we have alienated ourselves from the rest of the animal kingdom in an unprecedented way – having developed culture, science and technology in such a vehement way that we have come to totally dominate our planet, leaving most of nature at our mercy. What is the feature or features that have provided for the fundamental difference between us and other animals? There are a lot of features that are considered to make us humans special. The most obvious ones are perhaps reason, language or morals. A very naive version of this story might go as follows: we humans came to grow bigger brains which made us rational. Then, it turned out that it would be a useful thing to exchange our rational thoughts among ourselves, and we therefore developed language. And once we had language, it turned out to be useful to solidify our communal bonds with rules, moral and others, which we now could stipulate.

DOI: 10.4324/9781003388876-4

Preliminaries I 39 This, most probably, is not a story anyone would want to defend in such a caricatured form. But there are more sophisticated versions of the story, which preserve the order reason    ⇓ language    ⇓ rules. And, indeed, it would seem that this diagram, in general, captures the most typical view of the relationship between these three human features. This would imply that the most basic difference between us and other animals, the difference that is responsible for our differentiation, is our reason. In this way, we would turn out to be the rational animals of traditional philosophy. Here, however, we want to explore an alternative view (not denying that we are rational, but that it is rationality that underlies all other our specifica) connected with the inversion of the above diagram. We have already seen that rules, in general, are not underlain by language. Though a rule may come into being via its linguistic articulation (and in a developed human society with a developed language, this may be the prototypical way), this is not the only possibility; as we saw, rules that do not presuppose language are not only possible, but in a sense necessary. This opens up the possibility that language, the other way around, is underlain by rules. This fact is interesting and important because, as we already saw, it opens up a possibility of explanation of how language functions in a way alternative to the usual representationalism. The usual explanation is that linguistic expressions are meaningful in that they stand for (sophisticated) mental representations, which represent something else, typically something extralinguistic. Thus Sinha (2014, p. 41) notes: Symbols […] are conventional, depending upon shared understanding that the symbol is a token representing some referential class, and that the particular token represents a particular (aspect of) a shared universe of reference and, ultimately, of discourse. Conventional symbol systems are grounded in an intersubjective meaning-field in which speakers represent, through symbolic action, some segment or aspect of reality for hearers. This representational function is unique to symbolization, and is ultimately what distinguishes a symbol from a signal. A signal can be regarded as a (possibly coded) instruction to behave in a certain way. A symbol, on the other hand, directs and guides not the behaviour of the organism(s) receiving the signal, but

40  Preliminaries I their understanding (construal) or (minimally) their attention, with respect to a shared referential situation. And, of course, individual representational theories may differ quite a lot in their views on the nature of the representations and of their relation to the items represented. Let us consider the influential theory of Fodor (1975, 2008). Fodor argues that as the human mind is essentially a matter of manipulation of representations, these representations must form a comprehensive system forming a language – a language of thought – in its own right. A natural language is then just an externalization of the language of thought. The meaning of the expressions of the former is the corresponding “expressions” of the latter – representations we are born with. Fodor (1975, p. 86) writes: What, then, is being denied? Roughly, that one can learn a language whose expressive power is greater than that of a language that one already knows. Less roughly, that one can learn a language whose predicates express extensions not expressible by those of a previously available representational system. Still less roughly, that one can learn a language whose predicates express extensions not expressible by predicates of the representational system whose employment mediates the learning. Here Fodor insists that in order to learn a natural language, an individual must already possess a language of at least the same expressive power. (Let us leave aside the fact that Fodor reduces the expressive power of a language to that of its predicates.) The latter language, obviously, cannot have been learned; hence, it must be inborn. We cannot learn to express anything that we are not already able to express – any semantics of a natural language we learn cannot be but a rehash of a semantics we already have. Thus, language, and especially its semantics, is essentially linked to the innate representational capacity of the human mind. In Fodor’s view, the link is quite straightforward: expressions of the alleged language of thought directly are representations, whereas natural language expressions represent these representations, thereby gaining the ability to refer to appropriate items in the environment. There are many theories that see representations as laying the foundation of semantics, perhaps differing from Fodor’s in explaining the details of how they do it. However, such theories are not unproblematic: if only for the fact that representations are subjective, while language is essentially intersubjective, its constitutive feature is that it can be grasped by different persons, thus providing for an intersubjective understanding. If there are such things as meanings (and, for the record, I do not think that meanings are things), then they must be an intersubjective property. It is not that representations would not play a role within human cognition; it is that the road from

Preliminaries I 41 cognition to meanings is less straightforward than such theories suggest. In particular, meaning is not achieved by an individual cognition, it originates, as it were, in the intersection of many cognitions. Therefore, a quite different approach to the explication of the nature of meaning may be desirable. And once we explain rules as – in their ur-form – independent of language and hence capable of laying the foundation of language, we have a wholly different model of functioning of language and of meaningfulness. Of course, this view of language is also not unprecedented: it started to emerge in the works of some twentieth-century philosophers; already, in fact, Wittgenstein, as we saw in the first chapter of this book, stressed that our “language games” have a lot to do with rules. Inferentialism, which took this normative picture of language as its cornerstone, was then developed by Sellars, Brandom and their followers. The most basic idea of the inferentialist approach to language and meaning consists in the role an expression comes to acquire as it is gradually entangled within the inferential structure of language. This structure constitutes what can be called the space of meaningfulness, a virtual space delimited by the rules of language. Thus, as we already noted, it is not the case that to explain language we need to explain meanings and the links that keep them attached to expressions; it is rather the case that we need to explain the inferential rules of language and the edifice they make up. I have discussed this view in detail in many previous writings – Peregrin (2008a, 2012, 2014a) – and hence will not do it again here (but see also Chapter 12). What interests us here is the fact that this view of the nature of language also opens a new kind of view of possible language origins, which differs from those associated with the representational paradigm and, in my view, is more plausible than they. And this is something we are going to discuss in detail in upcoming chapters. But now we return to the above diagram of the relationship among reason, language and rules. 3.2  Rules and reason We have explained why we want to see the arrow marking the dependence of rules on language inverted. What then, about reason? Should it stay above, as underlying rules which then underlie language, or should it too be underpinned by rules? (What, exactly, do we mean by reason here? In a nutshell, it is supposed to be our human propositional thinking, our belief-desire psychology, prominently including theoretical and practical reasoning, allowing us to cope with the environment – in addition to the reactive and practical way – also in a proactive and theoretical way.) We will argue that we do not need reason to underpin (the ur-form of) rules, for rules bootstrap themselves into existence without the help of anything like reason – out of certain practical attitudes, which start out

42  Preliminaries I not as propositional attitudes, but rather as more or less reflexive reactions to the behavior of others. And in fact neither do we need it to underpin language, for language, as we will see, is fully underpinned by rules. This allows us to move reason to the very bottom of the diagram, so that we have rules    ⇓ language    ⇓ reason. Is the assumption that language underlies reason, rather than the other way around, plausible? Again, though it is probably not the mainstream view, it is, of course, far from being unprecedented. What I call the “mainstream view” is, I think, very well illustrated by the following quote from a paper of Steven Pinker (2010, p. 8995): The end product of learning survival skills is information stored in one’s brain. Language is a means of transmitting that information to another brain. The ability to share information via language leverages the value of acquiring new knowledge and skills. Language, it would seem from this, is a means of the externalization of information. Thanks to it, some information, originally stored internally in one’s brain, becomes external and thus accessible to individuals other than the owner of the brain. As Hurford (2007, p. x) puts it, we gained “the specific conceptual abilities that accrued to humans as a direct result of going public with their thoughts” and being “willing to open up their private concepts to others”. An extreme form of this view is that we consider human reason – including propositional thinking – as having evolved independently of language, which then lets us explain language quite easily: sentences are just labels for the propositions already existing within human minds and sub-sentential expressions are labels of their components. In this way, we could even see the whole belief-desire psychology and the corresponding mechanisms of reasoning as evolved before language, our explicit game of giving and asking for reasons then being merely a public presentation of what already goes on in human heads. But I think this picture is suspect. I think that the understanding of language as merely a means of externalization of thought is untenable, for the way we humans think is closely connected with the ways of our social interactions, which, in our human case, are mostly linguistic interactions. Thus, unlike Pinker, Tomasello (2014, pp. 122–123) argues that genuine

Preliminaries I 43 human thinking and reasoning (leading to “information”) resulted from the internalization of the viewpoints of others: Whereas early humans internalized and referenced the perspective of what Mead (1934) calls the “significant other”, modern humans internalized and referenced the perspective of the group as a whole, or any group member, Mead’s “generalized other”. Human thinking at this point is no longer a solely individual process, or even a second-personal social process; rather, it is an internalized dialogue between “what I do think” and “what anyone ought to think” (Sellars, 1963b). Human thinking has now become collective, objective, reflective, and normative; that is to say, it has now become full-blown human reasoning. One of the offshoots of Pinker’s view that reason developed independently of language and of social interaction is his “cognitive niche” hypothesis.1 Pinker (2010) argues that our specifically human way of life is primarily the result of the improvement of human intelligence. According to him, our “cognitive niche” is characterized by “the idea that in any ecosystem, the possibility exists for an organism to overtake other organisms’ fixed defenses by cause-and-effect reasoning and cooperative action – to deploy information and inference, rather than particular features of physics and chemistry, to extract resources from other organisms in opposition to their adaptations to protect those resources” (pp. 8993–8994). The most fundamental achievement is a matter of the upgraded cognition of an individual human organism; its spread through the human communities then being bolstered by the invention of language. This conception is criticized by many of those who think that it is not the improvement of cognition that came first and underlies our peculiar kind of human “cultural” sociality, but that rather it is the sociality that boosted the cognition. Thus, the evolutionary success of us humans is not (or is not only) a matter of the boost of individuals’ cognitive capacities, but of our tendency to form social “networks of knowledge”, in which nobody knows everything, but everybody can make use of the knowledge of the whole network.2 Boyd et al. (2011) write (p. 10919): “The cognitive niche hypothesis overestimates the extent to which individual human cognitive abilities allow people to succeed in diverse environments and misunderstands the role that culture plays in a number of important ways. We suggest, instead, that our uniquely developed ability to learn from others is absolutely crucial for human ecological success. This capacity enables humans to gradually accumulate information across generations and develop well-adapted tools, beliefs, and practices that no individual could invent on their own. We have entered the ‘cultural niche,’ and our

44  Preliminaries I exploitation of this niche has had a profound impact on the trajectory of human evolution”. Thus, while the “cognitive niche” hypothesis is based on the assumption that our reason developed independently of communication, as an individual “problem solving device”, the “cultural niche” hypothesis entertains the thesis that a great deal of the “problem solving” was already social. Thus, the emergence of our reason was inseparably connected with social interaction, and, I would add, with the emergence of language. There is one more issue that should be clarified, and this is the relationship between reason and cognition. Fodor, we saw, stressed that to learn a language we already need a language. I do not see why this should be so,3 but I do not dispute that to learn a language we already need some cognitive capacities. (For example, to learn a meaning based on some kind of distinctions, we need to be able to make the distinctions.) Hence, if reason depends on language, there must be some cognition that is not yet reason, some “pre-rational” cognition. And, indeed, it is especially important to stress that reason should not be equated with cognition in general, and the thesis that rules underlie reason is not the thesis that rules underlie any kind of cognition. Reason, we can say, is a specific kind of cognition, our specifically human kind. But reason also should not amount to our human cognition in its entirety. Reason is better seen as a layer of our human cognition, as its upper layer that rests on some lower layers which we may share with other animals. Perhaps reason in this sense can be identified with System 2 of the “dual-process theories of reasoning” (Evans, 2003; Evans & Over, 2013). According to this theory, there are two distinct cognitive systems underlying reasoning. While System 1 is evolutionarily ancient and shared with some other animals, System 2 is more recent and restricted to humans. The former system is swift, but not quite precise (thus it is suited for “online” reasoning where speed is more important than precision); the latter, in contrast to this, is not only more precise, but also more time-consuming (thus being suited for “offline” deliberation, possibly correcting the errors and imprecisions of System 1). We therefore have a theory of two relatively independent systems of reasoning in a single brain characterizing our human condition. It is plausible to assume that starting the bootstrapping of rules, language and reason presupposes some nontrivial cognition, and unless we identify cognition with reason, this does not interfere with our proposal. The cognition that is here before the outset of the bootstrapping process can be perhaps seen as something like System 1, and anyway “as a form of universal cognition shared between humans and animals” (Evans,

Preliminaries I 45 2003, p. 454). Cognition is not a direct subject matter of this book, but it may be of interest to sketch what such a cognition predating reason could look like.4 In a classic paper, Premack (1983) distinguished between what he calls the “codes of man and beast” and what he also characterizes as “propositional” and “imaginal” codes, respectively. While the former one is closely connected with a language (and is thus restricted to humans, plus, according to Premack, some apes with the proclivity to learn the rudiments of language), the latter can be imagined, with an amount of oversimplification, as “thinking in pictures”. Note that though we indeed can do a lot of things using the imaginal code that may perhaps also be called “reasoning” (such as imagining consequences or impacts of alternative courses of action and choosing one of them accordingly), it is not reasoning in our sense. How far we can get with the purely “imagistic” cognition (as he calls it) has been explored by Gauker (2011). He concludes that it may be further than we would think at first sight; but even so Gauker appreciates the surplus level of “conceptual” thinking that is inseparably interconnected with language. (Gauker argues that it is natural language which plays the role of bringing about the “conceptual” or “propositional” level of thinking, rather than a language of thought. I concur – but this is not the place to discuss this in detail.) Thus, if the main argument of Fodor (1975) against the thesis that we think in terms of a public language is that even animals without such a language think (“the obvious […] refutation of the claim that natural languages are the medium of thought is that there are nonverbal organisms that think” – p. 56), then this is not an argument against our view. Aside from the fact that the view does not necessarily involve the claim that a natural language is directly the medium of thought, the important thing is that natural languages influence our human kind of thinking, not that they make any kind of thinking or cognition possible.5 It is also possible to question the “layer cake” picture of human cognition that the proponents of System 2 as laid over System 1 propose. Boyle (2016) urges that “reason” is not something to be added to an underlying layer of cognition, leaving its underlayment intact. The addition, according to him, essentially affects the underlying layer, so that the result is a wholesale transformation of the whole cognition. It is not our task here to adjudicate between the two hypotheses – the “layer-cake” thesis and the “wholesale transformation” thesis, respectively. Our point is that rules lay the foundation not of cognition in general, but rather of our specific human kind of cognition – aka reason – which is added to (or, following Boyle, transformative of) a previous, simpler kind of cognition.

46  Preliminaries I 3.3  Rules and emotions We should also mention the intricate relationship among rules, reason and emotions. In ethics, rules are often seen as in opposition to emotions. Morality is seen as rooted in one or the other, and in this sense, they are played against each other. Ethical theories of the kind represented by Kantian deontology insist that to become moral, man has to become normative first, for genuine human morality is strongly rule-based. The theories based on the Humean appreciation of the “passions of the soul”, on the other hand, see morality as fully rooted in the emotions and see any moral rules as merely instrumental to the goals the emotions direct us at. Some recent results in the field of experimental philosophy (Haidt, 2012) appear to confirm that it is the Humean view that is more adequate. There is a sense in which it is the Kantian view that is closer to the position defended here, for according to this view, it is rules and spaces constituted by rules that crucially constitute our form of life and thus our nature. However, from our perspective, this does not denigrate the role of emotions. On the contrary, rules and emotions are complementary. Everything an animal (and especially a human) does is underlain by drives and emotions; this holds also for human rule-following, viz. for treating something as correct and something else as incorrect. From this viewpoint, emotions look like part of the implementation of rule-following; indeed, treating something as correct/incorrect seems to be inseparably connected with some emotional reaction to things of this kind. And as emotions certainly pre-exist rule-following, we may conjecture that we are able to follow only rules that accord, in a certain way, with our emotions; hence, that it is emotions that take part in determining which rules we can have. And indeed in literature we can find arguments that the role of emotion, especially w.r.t. ethical rules, is deeply underestimated (Haidt, 2001; Greene, 2008) and that rules that accord with our emotions are much more likely to flourish than those that do not (Nichols, 2002). Is it so, then, that rules and rule-following are only our distorted picture of what in reality is an interplay of emotions? No; rules are a genuine superstructure of emotions, which has a significant feedback on the emotions that lie in its foundations. As Kumar and Campbell (2022, p. 84) put it (concentrating on moral norms and emotions): Moral emotions and moral norms co-evolved. The result was a new, adaptively complex system in which emotions and norms were knitted together. This affective and cognitive system is a central feature of the moral mind. It’s also the key to understanding the internal structure of moral norms and what philosophers and scientists call “moral intuition”.

Preliminaries I 47 Anyway, emotions, and the way they support (and influence) norms, are not the subject matter of this book. We do not doubt that there would be no rule-following without emotions – just like there would be no football without feet. But to study rules from the social perspective, we need not study emotions – just like to study football as a social phenomenon, we need not study the feet of its players. This does not mean that a psychology of rule-following would not be an interesting subject – just like the study of the legs and feet of footballers. 3.4  Why concentrate on rules? There are, no doubt, many features other than language, reason and rules that are characteristic of us humans. We are, for example, the notorious “featherless bipeds” of Plato. Why give the special pride of place to rules and sideline all the other features? One reason might be that the other features can be shown to presuppose the ability to follow rules. I am trying to show this for language and reason, but of course I am not able to do it for all the other features specific to us humans. I am certainly not going to claim that we are featherless bipeds because we follow rules. (And I think that trying to show that, vice versa, we are rule-followers because we are the featherless bipeds is equally unpromising. This is not to say that there cannot be some links between the two features – see, e.g., Planer & Sterelny 2021, §2.4 – but searching for them is not a subject matter of this book.) Let us consider some more proposals regarding the distinctive features of us humans. In a thoughtful paper, Carruthers (2013) gives a survey of the most significant proposals of the essential features that might have provided for the deep difference between us humans and other animal species, and especially which provide for the kind of cumulative culture we have. They are as follows:

• • • • • • •

Imitation, shared intentionality and mind-reading Language Creativity Fine motor coordination Naïve physics Inference to the best explanation Cooperation and norms

His view is that it is in vain that we seek a single feature which has brought all the others in its wake. But be this as it may, it is certainly so that some of the features may depend on other features, or presuppose them. There is no doubt that

48  Preliminaries I fine motor coordination is presupposed for many human achievements; in the same way that it is clear that our language presupposes a very specific development of our speech organs. The same holds for the ability of imitation or for physics as a module of our brain. But these appear to be traits nearer to human biology than language or rule-following; hence, they seem to operate on a “lower” level than those that interest us here. What we urge is that the emergence of rules could have enabled the development of language, which could have enabled the development of our kind of reason (including such feats as the inference to the best explanation). As we will also see later (Chapter 15), becoming a “creature of rules” is interdependent with becoming “free” in our specific, human sense, and it also opens up the space for our creativity. Hence, the reason why we concentrate on rules, and why we accept Sellars’ proposal that we humans are essentially creatures of rules is not that we would be able to trace all our specific features back to rules or the ability to follow rules. The reason is that an analysis of this ability may tell us something very nontrivial about our species and that other such features, like language or reason, seem to presuppose it rather than the other way around. It is, I will try to show, precisely this ability that is behind our specific “form of life” that differs, dramatically, from the forms led by all our animal cousins. The point is that the form can be, I believe, nicely envisaged as a form taking place in new kinds of spaces, and that these spaces can, in turn, be plausibly characterized as constructed out of rules. It is also important to clarify that the development of rules has made rules in their fully fledged form, as we know them today, into something very different from the rudimentary form of rules as they first emerged out of evolution. It is clear that once language is on board, even rules get upgraded to a new level. In a society with language, the most usual way in which a rule comes into being will be by way of an announcement by somebody with the requisite authority. It may be that a new law is printed in a bulletin and declared to be in force; or that somebody invents a new game and presents its rules on the internet; or a change in the rules of orthography of a language is presented in a professional journal. This presupposes not only that language is up and running, but also that the society in question is normatively structured, so that various people have – or may acquire – various kinds of authorities that let them bring various rules into force, or put them forward for others to accept them. Following some such rules must be backed up by an elaborate system of establishments. Thus, certain kinds of rules may be announced by a government, complying with them may be looked after by the police, and violators may be sentenced by courts and put in jails, and so on. This creates a vast range of problems (addressed by philosophers of law, social scientists, etc., such as Hart, 1961, or Bicchieri, 2005) that I will not address – or

Preliminaries I 49 will address only cursorily – in this book. I will instead concentrate on the problem of how rules could have come into being without being backed up by reason or language, how implicit rules can exist and how rules can lay the foundation to language and reason, and, as a consequence, to our human form of life as we know it. Hence, we will concentrate on the phylogenetically early stages of the occurrence of rules. 3.5 Summary According to a common parable, rules are underlain by language (for rules are sentences), which, in turn, is underlain by reason (for language can come into being only when there are thoughts for it to express). The considerations of Wittgenstein and Sellars, however, indicate that it may be otherwise: rules do not presuppose language, for language, the other way around, presupposes rules. And similarly language does not presuppose reason, for reason presupposes language. (It is not that any kind of cognition would presuppose language, but the specifically human kind of propositional cognition, aka reason, does presuppose it.) This opens up a new way of thinking about us humans: once we move rules to the bottom of the hierarchy of our specific features, we may start to think about Homo sapiens as about a species that is distinctively normative. Notes 1 See Peregrin (2020a) for a more detailed discussion. 2 Sloman and Fernbach (2018) show that we tend to see a lot of knowledge accessible to us through the network as residing in our own minds. But this, the authors argue, is a similar illusion as seeing a lot of what is retrievable through one’s computer from the internet as residing inside the computer. 3 From the viewpoint presented here, this sounds a bit like To be able to learn to ride a bike, we already need to know how to ride a bike-of-thought. 4 Dual-process theories usually present themselves as theories of reasoning; hence, even System 1 is supposed to be a system of reasoning. Is it, then, not a contradiction to consider it a “pre-reason reasoning”? Not if you understand reason and reasoning, as we do here, as a matter of manipulating propositions – then the cognition affected by System 1 is at most a pre-reasoning. 5 Cf. Sellars’ verbal behaviorism, mentioned in endnote 2 of chapter 2.

4

Preliminaries II Rules as part of nature

4.1  Philosophy and science In this book, we want to portray humans as a species that evolved rules as the most important element of our social organization and consequently of the environment that shapes us. This, by itself, may not seem to be very difficult: we see rules all around us, we can describe them and, by analyzing the ways they interact and interlock, we may arrive at something we call their “grammar” (Bicchieri, 2005). Not that assembling such a grammar would be easy, it is certainly a deep achievement, but the enterprise does not seem to pose any deep conceptual or philosophical problems that would force us to go beyond the framework of ordinary social science. Where we do encounter deep conceptual problems, as we have already seen in the previous chapter, is accommodating rules to the description of the rest of the world. And again, it would not be problematic if we were to put no restrictions on the description – if we could say that rules are simply strange kinds of intangible entities, perhaps made of some esoteric stuff. But of course this is not what we want: we want the rules to fit in the kind of descriptions that are produced by our most reliable and most widely accepted tools, viz. by science. Hence we want to approach the phenomenon of rules – and of us humans as normative creatures – from a standpoint that is compatible with science, informed by science and reflecting the findings of science. Not that the book would itself be scientific; it draws on results of science but arranges them together in a way that is too speculative and too bold to be scientific. This arrangement is not something that could be substantiated doing justice to the standards of science; it is (or hopes to be) instead a vision that might perhaps help some scientists find a new orientation for their further research. Its main contribution (if any) is, however, philosophical. Not all philosophers whom this book takes as being essential for understanding rules are similarly obliging towards science. Wittgenstein, for one, thought that philosophy has very little to do with science. Unlike science, DOI: 10.4324/9781003388876-5

Preliminaries II 51 philosophy, according to him, is not a theory, but rather something like “therapy”. Whereas science treats its problems exclusively by struggling to solve them, philosophy should treat its problems “as illnesses”: its primary aim is to help those who rack their brains over philosophical problems escape from being harassed by the problems, be it by solving them, by showing that they are not real problems, or in whatever other way. In one of the most colorful similes of PI, Wittgenstein writes (§119): The results of philosophy are the uncovering of one or another piece of plain nonsense and bumps that the understanding has got by running its head up against the limits of language. These bumps make us see the value of the discovery. The situation is more complicated with Sellars. He had a great respect for science (which is expressed, among other things, in his Scientia mensura catchphrase cited in Section 2.1), and yet he would not interlock philosophy with science too closely. The reason is that though science rules, sovereignly, over our scientific image of the world, we cannot get rid of our manifest image, which escapes its jurisdiction. And as we saw, what crucially differentiates the latter from the former is the presence of rules. This ambivalence of Sellars has given birth to a schism between his followers, who can be divided into the “right” Sellarsians, who put stress on Sellars’ allegiance to science and try to squeeze rules into the scientific image, and “left” Sellarsians, who insists on the autonomy of the manifest image and of the normative (Peregrin, 2016a). Robert Brandom, a prominent left Sellarsian and the most important contemporary philosopher from the viewpoint of this book, maintains that philosophy is not supposed to tell us something about the world, but rather “to figure out ways to increase semantic and discursive self-consciousness” (Brandom, 2009, p. 128). Brandom thus thinks that the basic task of philosophy is to help equip us with concepts that foster our understanding of who we are and how we should live – not tell us facts about us, our societies and the rest of the world. Thus, according to him, philosophy is not predestined to closely cooperate with science. However, despite the fact that Brandom does not have the ambition to reveal novel facts about language or the world, he and his fellow-travelers do tell us lots of things: about our language, about the games we play with it, about the rules that govern them and our social practices more generally. Should we not take them seriously? Should we see them merely as non-committal tales, providing a fertile ground for the concepts that are the important resource? And what if we do take them more seriously than this?

52  Preliminaries II True, Brandom does not pay much attention to the results of empirical science, even if they concern language, social rules or other things that appear to be in the center of his attention. And it is true, by far, that not everything he writes about such things is useful or accurate when construed as reports. But I believe that his view on language and social practices suggests a useful orientation of empirical research that may bring us not only new and general insights, but also new and specific findings. What would such an orientation require? It would require, first and foremost, that we see the essential importance of rules for our human world. Brandom maintains that living within human communities means steering through webs of normative relationships, thereby acquiring and losing all kinds of commitments and entitlements and consequently gaining various kinds of social statuses. And to see this clearly may be difficult if we are viewing humans through a magnifying glass ground to scientific standards. What I take to be especially important is Brandom’s view of language. It too is a matter of rules, and as such it is to be seen as a “toolbox” (Wittgenstein, 1953, §11) rather than a set of symbols or representations. But as our language games are – more or less – rule-governed, it is most fruitful to look at language as a game like chess, or maybe even better like football (Lance, 1998). Just like the rules of chess turn pieces of wood into pawns, kings or bishops, so the rules of language turn types of sounds into reports that the sun is shining, questions about whether the train has already left or requests to shut a window. Viewing language as a “toolbox”, viz. as a set of utensils with which one can achieve various things, of course, is far from unprecedented. It is a view of language owed to, besides Wittgenstein, some of the classical pragmatists (Dewey1), as well as to a number of “neo-pragmatists” (Quine, Rorty, Davidson; see, e.g., Hildebrand, 2018), not to mention the ordinary language philosophers (Austin, Grice.; see, e.g., Hanfling, 2013) and the like. Seeing language as a basically rule-governed enterprise does not have so many precedents; but Wittgenstein and Sellars, again, are among them. However, if we want this to serve as an orientation for science, we must know how to embed rules into the causal structure of the world, and this is notoriously difficult. And Brandom makes it no easier for us: for him the “realm of the normative” is self-encapsulated, with no back-end connecting it seamlessly to the “realm of the causal”. However, as a matter of fact, Brandom’s views have already come to resonate with some recent developments in science. A prominent case is the theory of Michael Tomasello and his followers (Tomasello, 1999, 2014; Rakoczy & Schmidt, 2012, 2013). The reason for this resonance, as far as I can see, consists in the fact that Tomasello and company see the

Preliminaries II 53 human kind of intelligence as largely an imprint of human social interaction and human culture on an individual. Hence, like Brandom, they see culture not as the result of the interconnection of individuals whose intelligence was bolstered by evolution and they thereby additionally acquired new potentials for sophisticated interaction, but rather as an evolutionarily emergent phenomenon which, derivatively, formed the cognitions of its individual partakers. Tomasello himself acknowledges being influenced by Brandom’s mentor Wilfrid Sellars as well as by Brandom himself (Tomasello, 2014), whereas his teachings in turn are utilized by a host of current philosophers developing the ideas of Sellars and Brandom (Gauker, 2011; Rouse, 2015; Koreň, 2021; Stovall, 2022, etc.). Tomasello stresses the social origin of human thought and the importance of rules for human societies, while his followers have already done a lot of work anatomizing the role of rules within human ontogeny. In this way, the ideas of the later Wittgenstein, of Sellars and of Brandom get into a fruitful interaction with science and help to oil the wheels of scientific research. 4.2 Naturalism The philosophers who insist on the inseparability of philosophy and science nowadays mostly rally under the banner of naturalism. The term was originally used by American philosophers around John Dewey and Roy Wood Sellars (Wilfrid Sellars’ father), who argued that there is nothing save nature (hence nothing “supernatural”; see Krikorian, 1944) – a position, needless to say, that admits many divergent interpretations and elaborations. Later, Quine gave this thought a concise and general articulation: “knowledge, mind, and meaning are part of the same world that they have to do with, and (…) they are to be studied in the same empirical spirit that animates natural science” (Quine, 1969, p. 26). And it is this kind of naturalism that I want to subscribe to here. Quine’s concern is not so much to exclude the “supernatural” in some esoteric sense (this, he maintains, should not find its way into philosophy in the first place) but to exclude an investigation of a subject (the seat of “knowledge, mind, and meaning”) by methods that are alien to the methods we use to investigate the rest of the world (such as a phenomenological analysis, a priori speculations and introspection). Quinean naturalists are convinced that the subject (as contrasted with the object, or the world) is not to be investigated by any specific methods not known to the sciences; insofar as it is to be investigated at all, it must subject itself to the standard scrutiny of those sciences. There is, from the Quinean viewpoint, no first-person science or philosophy, only the usual third-person investigation yielding findings that are publicly accessible and subject to an

54  Preliminaries II interpersonal check (cf. Dennett’s, 2018a, manifesto to the same effect). This gives us the following picture of cooperation between science and philosophy (Quine, 1969, pp. 126–7): I see philosophy not as an a priori propaedeutic or groundwork for science, but as continuous with science. I see philosophy and science as in the same boat – a boat which, to revert to Neurath’s figure as I so often do, we can rebuild only at sea while staying afloat in it. There is no external vantage point, no first philosophy. All scientific findings, all scientific conjectures that are at present plausible, are therefore in my view as welcome for use in philosophy as elsewhere. According to Quine, then, and according to many other naturalists, philosophy is at least continuous with science. The two enterprises are not orthogonal, as scientific question-answering passes smoothly into philosophical question-answering when scientifically respectable evidence comes to be found lagging. Note that this says that both science and philosophy are parts of one and the same endeavor aimed at revealing the nature of our (natural) world, yet not very much about a division of labor between them. This contrasts with some later, more outspoken naturalists who sometimes do not shy away from claiming that any meaningful question is to be answered by natural science. Thus, if Alex Rosenberg (2013, p. 17) claims that “naturalism is the label for the thesis that the tools we should use in answering philosophical problems are the methods and findings of the mature sciences – from physics across to biology and increasingly neuroscience”, it is easy to read it so that there is no philosophy beyond natural science. (Of course, if we interpret Rosenberg’s claim in this way, it verges on being self-stultifying – for the thesis itself is hardly one that can be justified by “the methods and findings of the mature sciences”.)2 Naturalism can be seen as a metaphysical or ontological doctrine, telling us what there is, or, more precisely what there is not: any entities that are not recognized by science. Then, perhaps, if we want to see all of science as reducible (in principle, of course) to natural science and to see all natural science as ultimately reducible to physics, then as physics sees – with a certain amount of oversimplification – all objects as conglomerates of atoms, this kind of naturalism, or maybe better physicalism, tells us that there are merely atoms in the void. This version of naturalism claims that there are no thoughts, or, say, numbers – unless they are conglomerates of atoms or some aspects thereof. But I think it is important to see that there is also a more modest version of naturalism, especially a version that does not say what there is

Preliminaries II 55 but rather what kind of language we need to describe the world.3 We can admit an irreducible plurality of languages, each being possibly useful for some purpose (as Rorty, 1989, does), and many such languages do not lead us to see the world as consisting only of atoms. But such languages may be “local”, and we may be interested in seeking a “global” language – a language for the description of the whole world. True, it is not so easy to explain what “whole” means here; hence, there is no hope for yes/no verdicts. But certainly there is a great difference between languages that have the ambition to describe the entire world (like the language of physics or those of some religious doctrines) and those that concentrate on some limited sector or aspect (like the language of arithmetic or that of folk psychology). Seen from this visual angle, naturalism may be seen as an attempt to see how far we can get, in the project of describing and elucidating the world, with the language of science – or as a defense of the view that we can get far enough with it to be able to argue that it is the “whole world”. Note that from the Rortyan standpoint, the promotion of the naturalistic position does not conflict with the existence of other useful languages not reducible to it and coexisting with it. It does not deny that there may be languages useful for some special – or even general – purposes that would be not translatable into the language of science. Yet the language of science is in a sense special: it has been produced by the people that we have delegated to work on the answers to the hardest or the most pressing questions concerning our world and our place within it. Naturalism thus can mean entrusting oneself to science as being the enterprise of finding out as much as possible about the world as carried out by our most seasoned researchers. 4.3  Two (or more) images of the world? We have seen that Sellars argued that we people are endowed with two “images” of the world, which are irreducible to each other (though they can be perhaps joined in a “stereoscopic vision”4) and yet have the ambition to encompass the “whole world”. Does this not compromise naturalism?5 And how does it square with the fact that Sellars is a self-avowed naturalist? (deVries, 2005, p. 15, goes so far as to claim that “Sellars’s deepest philosophical commitment is to naturalism”.) Risking the allegation of misinterpretation, I will explain this prima facie discrepancy in a way about which I am convinced that if it was not Sellars’ own way, then it should have been. I think that though the manifest image is at least as important for us as the scientific image, the purpose of the former is slightly different from that of the latter. This, I am convinced, is the reason of their irreducibility to each other – of the fact

56  Preliminaries II that we need both of them and, optimally, their coordination in the “stereoscopic vision”. How does the manifest image differ from the scientific one and what makes it irreducible to it? The manifest image, I maintain, has to do not (only) with what there is, but also with what there should be, or what there will be in the optimal case. As I have already indicated, we need the manifest image because we are future-oriented creatures who live in terms of plans and ventures projected into the future. And just as imperative or interrogative sentences are not translatable (without a remainder) into declarative ones, so also the normative sentences are not. Here we should pause and say a few words about normative sentences – or normatives as I am going to call them for short – viz. sentences to the effect that something is (in)correct or (im)proper, or that it should (not) be done. From the viewpoint of syntax, normatives are just declarative sentences; but they differ from other declarative sentences in that they do not simply indicate (or picture, using the terminology of the early Wittgenstein) facts of the world. (That the syntactically declarative sentences form a heteronymous bundle, so that only some of them can be seen as declarative in the proper sense of the word was pointed out by several authors.)6 Normatives express what we will later call our normative attitudes and they can be construed, if we insist on seeing them as pictures, as pictures of an ideal world. In a paper we already discussed, Sellars (1949, p. 315) urges that “a rule, properly speaking, isn’t a rule unless it lives in behavior, ruleregulated behavior, even rule-violating behavior”. He claims that “to describe rules is to describe the skeletons of rules”. Using this terminology, we can say that normatives belong to the tools that help make some rules live; they are tools of the endorsement of rules that make them into more than “skeletons”. We can describe the whole process of keeping the rules alive: of their coordinated endorsement and of the attitudes that render – practically – the acts of complying with the rules “acceptable” (“right”), while making those of not complying with them “unacceptable” (“wrong”). This can be seen as a very complex and very sophisticated behavioral pattern, in which normatives (and, more fundamentally, the attitudes they make explicit) play a central role. However, if rules are, according to Sellars, what distinguishes the manifest image from the scientific one, is it not so that they are absent from the scientific image? And when a naturalist restricts herself to the scientific image, is it not so that she must completely leave out rules from her purview? Is it not so that normatives have no place in it? Well, it is clear that questions or commands are not translatable into descriptions – and yet we do not say that questions and commands are beyond the natural order. They are part of the order because we can describe, in a naturalistic language, the nature

Preliminaries II 57 and the role of the corresponding speech acts. And in a similar way, we can embed normatives into the natural order. This is to say that although we cannot translate normatives into declaratives, and hence we cannot reduce the manifest image to the scientific one, we can account for the normatives, and hence for the scientific image, in a naturalistic way. And just as in case of questions and commands, this is all we can reasonably require. Moreover, it is precisely the possibility to view rules as specific behavioral patterns that makes them part of the scientific image – hence even if we restrict ourselves to the scientific image, it is not true that rules vanish from our sight. We do see them, though as that which Sellars dubs their “skeletons”. We can tell the behavioral pattern characteristic of rule-following from other behavioral patterns. True, we do not see the rules “live”, as we renounce endorsing them – but we can detect their presence even within the scientific image. Hence, there is a way of embedding the manifest image into the scientific image. Without reducing the former to the latter, we may use the causal language of the scientific image to explain how “the normative” arose out of a rib of “the causal”. We can tell a scientifically respectable story of how rules came to be a part of our world, or how we humans came to move from the “realm of the causal” into the “realm of the normative”. However, the story still does not exist in a commonly accepted form. Sellars (1953b) himself introduces a useful terminological distinction: he distinguishes between “logical” and “causal” reducibility and he claims that though the talk about minds that we avail ourselves of within the manifest image (and hence about persons that are central to the image) is not logically reducible to the scientific idiom (viz. is not translatable into it), it is causally reducible to it. This means that we can, with the help of nothing more than the scientific vocabulary, tell the story about how and why the mentalistic and normativistic idiom came into being and how they function within our world. Hence there is the scientific image of the world that is self-contained and in a sense complete. It contains also rules (though devoid of their “lives”). Naturalism is the urge to remain, “in the dimension of describing and explaining the world”, within its bounds. But the fact is that in general we not only describe and explain the world – we also make plans, try to predict what will happen, and depending on it, what we are to do. It is because of this that we need the manifest image and not because we need to get hold of some occult, “supernatural” stuff that science prevents us from perceiving. 4.4 Summary The aim of this book is not only to portray us humans as a species that evolved (social) rules as the most important element of their form of life

58  Preliminaries II and consequently of the environment that shapes them, but also to do so in the naturalistic spirit, viz. without recruiting any specific conceptual resources going beyond what can be accommodated by sciences. Quine characterized naturalism by the claim that “knowledge, mind, and meaning are part of the same world that they have to do with, and they are to be studied in the same empirical spirit that animates natural science”, thus making philosophy continuous with science and renouncing any methods alien to science. However, Wittgenstein and Brandom are not Quinean naturalists; they think the goals of science and philosophy diverge. (The situation with Sellars is more complicated: he is a self-avowed naturalist, though he grants normativity a certain amount of autonomy.) Hence, going naturalistic means turning off the route marked out by them. It means developing a naturalistic theory of rules, drawing on not only Brandom, but – indirectly – also on Quine; but precisely this is the project of this book. It is, however, important to realize that the naturalistic approach to rules need not involve a seamless joining of the realm of the normative with that of the natural into one all-encompassing realm. The normative attitudes that we humans developed are unprecedented and introduce a new form of life, which is impossible to assimilate to anything that was here before. There is then a sense in which the normative keeps a certain kind of autonomy – relating to the world in this novel, future-oriented way, from “inside” of systems of rules and normative practices. However, the systems of rules and practices, though capable of mediating us the new kind of experience yielded by this new kind of relation to the word, are not something that would escape a naturalistic description. Notes 1 Quine (1969, p. 27) writes: “Further along [(Dewey, 1925)] expanded the point thus: ‘Language is specifically a mode of interaction of at least two beings, a speaker and a hearer; it presupposes an organized group to which these creatures belong, and from whom they have acquired their habits of speech. It is therefore a relationship’ (185). ... When Dewey was writing in this naturalistic vein, Wittgenstein still held his copy theory of language”. 2 Readers conversant with the philosophy of logical empiricism of the Vienna Circle may read this as an echo of their manifesto (Hahn et al., 1929): “The scientific world-conception knows no unconditionally valid knowledge derived from pure reason, no ‘synthetic judgments a priori’ of the kind that lie at the basis of Kantian epistemology and even more of all pre- and post-Kantian ontology and metaphysics. ... It is precisely in the rejection of the possibility of synthetic knowledge a priori that the basic thesis of modern empiricism lies. The scientific world-conception knows only empirical statements about things of all kinds, and analytic statements of logic and mathematics”. 3 I think that this possibility became acute after the linguistic turn in philosophy (Rorty, 1967).

Preliminaries II 59 4 See O’Shea (2012). See Peregrin (2018) for a discussion of the extent to which this duality is parallel with the duality stemming from the anomalous monism of Davidson (1970). 5 Sellars’ (1962a, p. 35) often quoted delimitation of philosophy runs as follows: “The aim of philosophy, abstractly formulated, is to understand how things in the broadest possible sense of the term hang together in the broadest possible sense of the term”. It follows that for him, what philosophy is after is a bird’s eye view of the whole world; presumably a view that science alone is not able to achieve. 6 See, e.g., Mandelkern (2020) for modal sentences or Punčochář and Gauker (2020) for conditional ones.

5

Preliminaries III Kinds of rules

5.1  Important distinctions Let us glance at the variety of classifications of rules and norms that are due to various philosophers (and social scientists). There are so many such classifications that it makes no sense to try to take all of them into account. Let us, however, at least take a look at a few interesting cases. One issue we have already tackled is the distinction between rules, as prescriptions that can be, as a matter of principle, violated and laws, which are immune from violation. Unfortunately, the terminology is not quite settled here. While the term “law” is often used for cases of what I call rules (like “criminal law”), the term “rule” is, conversely, sometimes used for the things I would like to call laws (like the “rules” of Chomskyan universal grammar). A very important distinction was pointed out long ago by Rawls (1955). He distinguished between what he called the “summary conception” and the “practice conception” of rules. In the former case, “the decisions made on particular cases are logically prior to rules” (p. 21). This means that in this case the rule aims at a summarization of a practice which is already up and running; in the latter case, the rules make the practice possible in the first place (“the rules of practices are logically prior to particular cases”, p. 24). This makes a lot of sense if we understand “rules” as sentences: the former case corresponds to the situation where the sentence is used to make explicit an already existing implicit rule, whereas the latter corresponds to the case where it is used to bring the rule into being. In the hands of John Searle (2010, Chapter 1), Rawls’ distinction has mutated into a distinction between “regulative” and “constitutive” rules, which has become quite influential. The former rules “regulate antecedently existing forms of behavior” (p. 10), like traffic; the latter, in contrast to this, “create the possibility of the very behavior that they regulate”, like the rules of chess. Searle also suggests that a typical regulative rule has the form Do X, while a typical constitutive rule is of the form X counts as Y in context C. DOI: 10.4324/9781003388876-6

Preliminaries III 61 I have some problems with this way of developing Rawls’ ideas. First, the term “regulative rule” seems to me ambiguously between the case where it, like Rawls’ summary rules, denotes regulation of a behavior that is already normative and the case where it introduces a normative regulation into a behavior which had, up to that point, no normative dimension at all, which is closer to Rawls’ practice conception than to his summary conception. Second, I do not understand that a typical constitutive rule is of the form X counts as Y in context C. Take chess: are the rules crucial to it of the sort this piece of wood counts as a rook in the context of the upcoming game? It does not seem so. One thing is that no physical pieces are necessary for a game of chess (as blind games attest); another thing is that the assignment of a role to a concrete item does not even belong to the definition of the game, it is the rules constitutive of the roles (The bishop moves only diagonally) that do. An influential classification of norms and rules was also put forward by von Wright (1963). He discusses what he calls the three most basic kinds of norms, two of them being not very far from the Searlian constitutive and regulative kinds of rules. The first kind are rules of games, the only norms which von Wright wants to call rules. They are close to Searlian constitutive rules, for they make the practices they govern possible at all. von Wright writes that aside from rules of genuine games, this category also comprises rules of the grammar of languages and rules of calculi. The second category of norms von Wright puts forward are prescriptions or regulations. These are norms that are explicitly stipulated, so there is typically a clear authority that is behind them. Prototypical examples of norms of this kind are military commands, orders and permissions given by parents to children, and the rules of traffic. The third type of norms, according to von Wright, are directives or technical norms; they are a matter of specifying adequate means for attaining a given end. A prototypical example is directions for use. Note that von Wright’s technical norms are instrumental rules, and so their status as rules may be seen as dubious. The technical norm If you want to be rich, you should steal would seem to be synonymous with If you do not steal, you will not be rich. The latter is simply a non-normative sentence (a conditional one, for that matter); hence, as it has the same meaning as the former one, it would seem that even the former one cannot mean anything genuinely normative. So perhaps the should in technical norms is used in a “non-normative” sense. For Brennan et al. (2013), there are also three categories of norms that are most basic; however, they are rather different from von Wright’s. Their formal norms are close to von Wright’s prescriptions: they are explicit rules issued by an authority. However, the second kind of norms they put forward, moral norms, have no counterpart in von Wright’s classification.

62  Preliminaries III (According to him, moral norms do not represent a homogenous group.) According to Brennan et al., they are characterized by being non-formal and being “purported approximations of objectively valid principles of morality” (p. 6). The third and last group of this classification is social norms, norms that are non-formal like the moral norms, and that, however, are much more arbitrary. We can perhaps say that formal norms are those which are typically brought into being by a sentence or sentences having been issued by an appropriate authority; and the rest of the norms, those which lack such canonical expression, then can be divided into central, viz. moral, ones and peripheral, social, ones. Roughley and Bayertz (2019), in their book about humans as normative animals, present still another triad. As the first and most basic kind, they have social norms; norms that are valid, because they are “accepted” by a society. Next, they discuss moral norms, which some authors of the contributions to the volume consider as simply the most central of social norms, others as archetypes of social rules, and still others as sui generis. The third kind of rules discussed in the book are linguistic norms, the nature of which is again subject to disputes. Some of the contributors see them as constitutive of our linguistic practices, others deny that rules play any important role regarding language and hence that there is almost or literally nothing such as specifically linguistic rules. Now putting all these proposals aside, I see a few crucial distinctions that we must observe when we want to analyze the concept of rule. We have seen that, in general, rules cannot be identified with sentences or propositions – for aside from explicit rules, the vehicles of which are such linguistic entities, there are also implicit (“unwritten”) rules, which need not be associated with anything like this. This brings us to the first of our conceptual differences: Svoboda (2018) proposes to articulate it using the following terminological convention (which, in his view, distinguishes two basic ways in which the term “rule” is commonly used): use the term L-rule (“linguistic rule”) for a rule in the sense of a sentence and use the term F-rule (“rule as a social fact”) for the rules that are in force in a certain community, perhaps merely implicit in a social praxis. While the concept of L-rule (or rule-sentence) is relatively clear, what exactly an F-rule can be considered to be is not so clear. As a first approximation, we can perhaps say that F-rules are specific (yet quite varied) social facts. We then use phrases like “the community accepts a rule” or “its members tend to follow a rule” to point out a way in which the members of the community behave – a complex and permanent behavioral pattern they display (crucially involving, as we will see, what we call normative attitudes). The relationship between L- and F-rules is complex. An F-rule can be brought into being by means of an L-rule; or the F-rule exists before the

Preliminaries III 63 corresponding L-rule and the L-rule is invoked to make it explicit. Or the F-rule may exist merely in the implicit form, without any explicitating L-rule. And conversely, there may be an L-rule without any corresponding F-rule: a rule-sentence that is not accepted by any society. Admittedly, the concept of F-rule, which, as both Wittgenstein and Sellars indicated, refers to rules in their primordial form, remains slightly enigmatic. It will take us many of our upcoming chapters to clarify it completely. However, what is usually possible is to represent an F-rule by an L-rule that would make it explicit; and this is what we will do. (It follows from the considerations of Wittgenstein that there may be no unique L-rule explicitating a given F-rule; sometimes there might even be none, or it might be too complicated to articulate it – but in most of the cases we will be dealing with the use of such representations of F-rules will be unproblematic.) Consider an L-rule, such as No shirt, no service. An owner of a bar can hang it out in front of her enterprise, which may prevent the guests from entering the bar shirtless. The (social) fact that people should not enter the place without shirts (and if they do, they will likely be thrown out) then amounts to an F-rule. Alternatively, the F-rule may pre-date the L-rule: the fact that people are not allowed into the bar without a shirt may be a matter of an unwritten tradition, which an owner may only later decide to make explicit in terms of the sign. Also the tradition may be so complicated (for example it may involve some unorthodox view of what counts as a shirt) that it might not be easy to articulate it in terms of an L-rule. The second conceptual difference we point out was vividly articulated by Dennett (2013, p. 51), as the difference between what he calls “Pittsburgh normativity” and “Consumer Reports normativity”: The former is concerned with the social norms that arise within the practice of communication and collaboration. […] The latter, in contrast, is concerned with quality control or efficiency, the norms of engineering, you could say, as revealed by market forces or just by natural failures. This is nicely highlighted by the distinction between a good deed and a good tool, or, negatively, between naughty and stupid. People may punish you for being naughty, by their lights, but nature itself may mindlessly punish you for being stupid. Hence, we have instrumental rules (“rules of engineering”) and rules that are non-instrumental (i.e. are not straightforwardly derived from a goal). Here we will concentrate on Dennett’s second kind of rules, on the social rules (in the broadest sense), which are products of human societies. They are in force in the society insofar as they are “accepted” – in such or another way – by members of the society.

64  Preliminaries III These rules are stipulated by us (though in many cases not willingly), while those of second kind are independent of our conducts. As they are “norms of engineering” (von Wright’s “technical directives”), the correctness they institute is an instrumental one: they sanction something as correct iff it leads to a pre-given goal. Obviously, some of these rules will not be very interesting: If you want to make an omelet, you should break some eggs. As we already noted, this sentence does not appear to say more than If you do not break some eggs, you will not make an omelet, so it seems to be only a paraphrase of a non-normative declarative sentence, and thus the question is whether the “should” it contains is more than a stylistic tool, i.e. that it marks a genuine normativity. But the fact that the “norms of engineering” exist and are “in force” independently of what we humans do may appear to grant them (or some of them) the quality of being “absolute” in the sense of “superhuman”. Is not, after all, this kind of “absoluteness” the hallmark of a genuine rule? Is not a rule that is not “absolute” in the sense of being irreversible only a makeshift rule? My answer is negative. Note that a rule, as we understand the term, cannot be absolute in the sense of being unbreakable, of wielding its normative power with an irresistible force. We have already stressed that rules that cannot be violated by mortal beings are not rules in our sense of the word – they are laws (in the sense of laws of nature). A “norm of engineering” can certainly be violated – on pain, of course, of not achieving the goal to which the rule is instrumental. You can violate the rule You should break some eggs – but you will hardly make an omelet. However, the goals to which various rules are instrumental may be of a different kind. If there are goals that we cannot avoid having, then the corresponding rules will be, after all, close to absolute (for us). Are there such rules? There appear to be two kinds of candidates. First, we may consider rules that are absolute in the sense that they are posed by a god with an absolute authority. As there is no room for gods in our naturalistic picture of the world, we disregard this possibility. The other kind of candidates appears to be rules that are instrumental to achieving something that is intrinsically valuable. If we, for example, consider life as intrinsically valuable, then any rule that leads us to save a life is absolutely in force. But the trouble is determining the intrinsic values (independently, do not forget, of any contingent matters of human societies). Life, for example, is a good candidate, but we could hardly find a human society where it would be valued so absolutely that there would not be situations where killing somebody is tolerable (or even celebrated). However, there seems to be one more interesting possibility: any animal seems to be predestined to value its own life. In a sense, this is what it takes to be an animal, or more generally an organism. This opens up a

Preliminaries III 65 remarkable way of thinking about normativity – but we will leave it for the next section. The final conceptual difference regarding rules I want to point out is that between the rules that are prescriptive in the sense that they tell us what we should do and those that are restrictive in the sense that they tell us what we should not do. Consider the prohibition Do not spit on the floor. It does not restrict the possibilities of my behavior in any significant way: it does not prescribe me to do any specific thing; even if I follow it and refrain from spitting on the floor, I may do whole lots of other things. And consider, by contrast, the command Spit on the floor: it tells me to do something quite specific. Thus, while prescriptions order us to do something, to be active in a specific way, restrictions do not require a specific activity (they can often be obeyed by “doing nothing”). To be sure, the distinction between prescriptions and restrictions is not sharp. If we reformulate Do not spit on the floor as, say, Avoid spitting on the floor, it may come to look as a prescription rather than as a restriction. The difference is not in the grammatical form, but rather, to repeat, in the spectrum of actions that it leaves open: Spit on the floor requires one quite specific action, while Do not spit on the floor (or Avoid spitting on the floor) permits all kinds of actions with the exception of spitting on the floor. But of course there are a lot of rules that are in between prescriptions and restrictions and no sharp dividing line can be drawn. The reason why I think it is crucial to keep this distinction in mind is that the background of the whole rule following discussion inspired by Wittgenstein and taking place in the second half of the twentieth century is constituted by the Wittgensteinian examples which concern straightforward prescriptions. Rules that tell us how to continue a number series are prescriptive in the most direct sense: they tell us which number to display next. And if this is taken to be the prototype of a rule, then the whole discussion appears to be about prescriptions. I think this is dangerously misleading, because, as concerns the role of rules in human societies, restrictions are much more important than prescriptions. We use rules to mark out the spaces in which our societies function, and in which we, consequently, live our lives. And it is restrictions that are suitable to mark out playgrounds on which we can play new kinds of “games”; and of course they are often not games in the literal sense of the word, but rather much more substantial social activities. In this way, they can – and do – alter our human forms of life. Thus, far from only restricting us, they are constitutive of the arenas in which our social – and consequently also personal – lives are taking place. Another important thing is that if we concentrate on prescriptions, then it seems that rules compromise our freedom, strangle our creativity and

66  Preliminaries III suppress our spontaneity. In this way, they are foes of many features that we feel make us human and that we therefore cherish. But nothing could be further from the truth – we can see this if we pay due attention to restrictions, rather than prescriptions: they do not route us into a single direction, they only erect barriers of a playing field where our spontaneity and creativity may come to a full fruition. In any case, as we will try to show later in the book, they are the basic building block of our social order (more about this in Chapter 15). 5.2  Absolute rules? In the previous section, we posed the question about goals that we cannot avoid having. If there are such goals, then there is a sense in which the rules that are instrumental to them are absolute (for us). And there are philosophers who insist that there are such goals, and that this follows from the fact that we are organisms. Thus, Okrent (2017, p. 28) notes: It is necessary that if life continues, if the ‘organism’ survives as what it is, as a living thing, then it engages with its environment in ways that are conducive to its survival, ways that thereby become the ‘right’ ways for it to interact with the environment. If the organism interacts with its environment in the ‘wrong’ way, it simply ceases to exist; it dies. That is, since interacting with an environment in certain ways is a necessary condition on the life of any organism, from the perspective of the living organism, any living organism, its own continued life supplies a standard that is to be achieved, relative to which how it actually interacts with its environment is to be evaluated. Life, as such, carries with it its own norms of appropriate interaction with its environment. This view is usually (though perhaps not inevitably) connected with evolution theory.1 Organisms move through the bottleneck of natural selection so that only those who are able to take care of themselves survive and produce offspring. Therefore, to be an organism is to take care of oneself in the manner that lets you survive and reproduce. Hence, everything that an organism does and that is instrumental to its fitness is correct, in the sense that it accomplishes the mission of the organism as such. Thus, there are rules inherent to the biological world, and as we humans are part of this world, these rules apply to us as well as to other organisms. What applies to the whole organisms applies also to their parts, components or organs; and it may apply also to their behavior or their tools. Our heart, for example, works correctly if it contributes to the fitness and the survival of the organism, which it does (oversimplifying) if it pumps blood

Preliminaries III 67 in an appropriate way; so we may be tempted to say that we have hearts in order to have our blood pumped. Evolution theory tells us that this outright teleology can be explained away (the statement We have hearts in order that blood can be pumped can be reduced to Our hearts, in the course of evolution, were selected to pump blood), but there is a sense in which evolution does produce goals, which generate norms. These alleged rules are “superhuman” in that they hold independently of what we people do, think or wish, which makes them attractive, and it may seem that it may be the root of any other kind of normativity. This line of thought has been pursued by various authors, especially Millikan (1989; 2005, 2013). She suggests that it is this very kind of rules that is the ultimate source of any normativity in the biological world, which, of course, also comprises us humans. This bestows on all kinds of rules that we humans may have come to follow the seal of absolute validity, a validity that is not dependent on our contingent vagaries. Thus, for example, rules like Hearts should pump blood or, perhaps, We should not eat poisonous plants are seen as absolutely valid. However, I think that though it is a fact that we are here thanks to evolution and that our features embody that which helped us to survive (or are by-products of something such), our evolution has not been as straightforward as that of other species. In particular, our evolution has led us to a point where we require a repertoire of behavior so vast that it can start to be very “opportunistic”: not predictably dependent on external stimuli, but dramatically variable. This opened-up space for developing more complicated behavioral structures has been determined by a mechanism inscribed in us by evolution in only an indirect way. We have come to have the possibility of “experimenting” – trying new ways to behave on a much larger scale than other animals. We have gained, that is to say, the ability to also explore the space of possible evolutionary trajectories in terms other than via mutations, at least as far as behavioral patterns are concerned. We can realize unprecedented and unprecedentedly complex ways of behavior, sometimes even just for “the hell of it”.2 Of course, it is evolution that has the final say (we propose, it disposes), so that the patterns that do not enhance fitness are wiped out. But while mutations can only push evolution in small steps (we know that an eye is fitness-enhancing, but we must find a chain of small steps which lead up to it, each of which is fitness-enhancing), our creative activities can take much larger steps. Part and parcel of this development was that we became able to set up not only rules that were directly profitable for us as individuals, but also rules that we only fancied, for such or another reason. (We may put together strange kinds of games, instituting various rituals, etc.) That does not mean, of course, that we extricated ourselves from the rails of

68  Preliminaries III evolution, we continue to be accountable to it; but our accountability no longer directs all of our steps. We have extricated ourselves, as it were, from the online monitoring – from the modus in which every tiny achievement must be in accordance with evolution, lest it be wiped out. We have started to be capable of producing things offline, as it were, that are then sanctioned, with a delay, by evolution (or not). As Laland (2017, p. 234) puts it: “Our culture hasn’t stopped biological evolution – that would be impossible – but it has left it trailing in its wake”. What kind of rules that are not directly instrumental to our survival, and yet eventually come to be sanctioned by natural selection, can there be? One important example are those rules which have to do with the structure of our societies, with our institutions and practices (see later for details). It is a notorious fact that rules of cooperation are not directly fitness-enhancing, for they require some investment (which, in an optimal case, return later with interest, but which can be also misused by freeriders – see Section 9.3). If these rules are to be directly X-rayed by natural selection, we may have a problem (witness the Prisoner’s Dilemma – see Section 14.2), but once we are able to also establish some noninstrumental rules, they are a prominent candidate. Okrent (2017) argues that it is not only us humans that have such noninstrumental rules, determining social hierarchy, etc., but also, at the least, other primates. However, he claims that while in the case of other primates they are genetically wired in and hence rigid, we humans have developed them in a flexible, “cultural” shape, which makes it possible for us to adjust them to circumstances. As he puts it (p. 105): “The human world is organized in terms of non-instrumental norms that specify different classes of persons and different classes of objects, and these non-instrumental rules have to do ultimately with how different human beings ought differentially to act in relation to other humans and nonhuman objects”. While the rules (“rules”?) in the focus of Millikan characterized what is correct from the viewpoint of evolution, we gained the ability of establishing our own rules so as to render what is correct for us – and though what thus becomes correct for us by our own lights is not likely to contradict what is correct for us by the lights of evolution, the former correctness is also no mere shadow of the latter one. Now, while the Millikanian “superhuman” rules are not characteristic especially of us humans (but rather of the whole biological world),3 the rules we ourselves pose (not only those that we consciously articulate, but rather also those which come to govern our communities implicitly) are such. It is these rules, and our very ability to create them, which makes us into the peculiar normative animals we are. This is not to say that we do not (or should not) treat as absolute some of the rules we accept. It is part and parcel of the idea of this book that

Preliminaries III 69 we humans cannot live without a coordinate system of rules, and we must treat the most basic rules of the coordinate system as absolute. Thus, we have what may be called the “inner” and the “outer” perspectives: from the outer perspective, every rule is social and contingent, whereas from the inner one the rules appear absolute. As I will argue later in the book, systems of human rules display the peculiar ability to build “inner spaces” where those who follow them can dwell. The dichotomy of the “inner” and the “outer” aspect of rules is a consequence of this fact; it is what makes rules so crucial for our human form of life. But more about this is detailed later in the book. As we are going to talk about the emergence of rules in human societies, we will have to consider various forms of rules, from the most rudimentary forms (these will be the implicit ones) to the fully fledged ones as we know them from our current societies. We will have much more to say about the former than about the latter – our focus will be on the process during which rules established themselves as the crucial determinant of human societies. We will also concentrate on the rules that may be considered “primordial”: those that lay down the primary “normative infrastructure” of a society and provide for the possibility of adding further layers of normative relationships. The rules of this primary infrastructure may, for example, establish a hierarchy granting some individuals the authority to impose rules on others, which opens room for creating the filigree of multi-layer normative relationships as we know them from contemporary human societies. To summarize from the perspective adopted here, it is misleading to talk about rules established directly by evolution. Everybody is, of course, free to use the term “rule” as she wishes, but here I prefer the perspective according to which rules are strictly social phenomena. 5.3  Rules, norms, correctness and normativity Now it is time to start fixing our terminology. In the previous two sections, we tried to put the concept of rule on a firm footing. We point out several crucial conceptual differences, without which it is difficult to grasp the concept of rule in all its complexity. Then there is the term norm, which is sometimes construed as synonymous with rule, while sometimes it is used in various different senses. In view of the problem discussed above, it might be tempting to use one of the terms for merely the explicit rules (sentences or propositions), while using the other one for the implicit ones (or perhaps for both the explicit and implicit ones). This is not, by far, the usage that would be generally accepted. “Rule” and “norm” are often not used interchangeably; thus, e.g., von Wright (1963) employs the terms so

70  Preliminaries III that rules are specific kinds of norms, whereas Svoboda (2018) proposes to see, vice versa, norms as specific kinds of rules. However, as each of the terms has some unwanted connotations which would complicate such terminological regulations, we will refrain from them and use “rule” and “norm” interchangeably. The next concept we must clarify is the concept of correctness: some actions, like adding 3 and 4 with a result of 7, or calling a green thing green, are correct, whereas other actions, like adding 3 and 4 with a result of 8, or calling a blue thing green, are incorrect. Of course, there are a vast number of other kinds of items which are correct/incorrect or may be considered as such. In a typical case, correctness is closely connected with a rule. When I formulate a rule (say a rule of a game that I invent), then what is in accordance with the rule becomes correct (in the context of the game), while what is not becomes incorrect. An L-rule like Bow to the chief or Bowing to the chief is correct classifies potential ways to behave – namely bowing to the chief – as correct. But the same can be achieved by an F-rule, namely by the fact that members of a community approve of bowing to the chief and disapprove of not doing so. Thus, every rule generates a correctness. Is, vice versa, every correctness generated by a rule? It depends on what exactly we call a rule. If what we call so is an L-rule, then no, not every correctness is generated by a rule. There are, we saw, “unwritten rules”, correctnesses that are not generated by any explicit instruction. And the “unwritten rules” need not be always capable of being written down, i.e. expressed by sentences. However, we could also construe the term rule so that “unwritten rules” will also be rules, viz. F-rules. And in such a case, it is just the occurrence of correctness (generated, as we will see, by normative attitudes) that shows the presence of a rule. Hence, on this broad notion of rule, every correctness is a matter of a rule. And we will use the term in this broad sense. It is important to stress that correctness is not an absolute notion (as discussed in the previous section). Many things may be correct from the viewpoint of some rules or normative attitudes, while incorrect from that of other ones. This is a consequence of the fact that the rules we consider are our, human creations and nothing can prevent us from creating conflicting rules. Sometimes the viewpoints will differ in import. An utterance may be correct from the viewpoint of English grammar, while incorrect in that it offends somebody (and on top of that, it may be correct in that the offence saves somebody’s life.) However, it is often difficult to decide which of some conflicting correctnesses we should prefer; indeed, human life is partly about such decisions. Then we have ought’s and should’s (we use these terms interchangeably). According to our construal, what is correct is what one should

Preliminaries III 71 (or ought to) do. Again, this is a broad construal of should – it counts on the fact that a should is not absolute, that it does not oblige us come what may. A should may be trumped by another, more pertinent should. The fact that according to the rules of chess, I should move a bishop only diagonally, does not contradict the fact that there may occur a situation where I should move it other than diagonally, and in which the latter case trumps the former one. I might, for example, perform a role in a theatrical play that prescribes me to violate the rule of chess. The usage of “correct” can be pulled apart from that of “should”, and thereby from “rule” and “norm”. Thus, for example, Glüer and Wikforss (2009, p. 36) insist that “there are non-normative uses of ‘correct’”, i.e., cases of correctness which do not entail anything that should (or ought to be) done. A case in point, according to Glüer and Wikforss, are rules of language: to say, for example, that the term dog correctly applies to dogs, in their view, does not instruct us to say dog whenever we see a dog or anything else. I do not think that this is a reasonable view. True, rules of language usually do not dictate that we should use a particular expression on a particular occasion. Just like the rules of chess do not tell us (with some marginal exceptions) that we should do a particular move. But the rules (both of chess and of language) tell us what we should not do: move a bishop otherwise than diagonally, say This is a dog! when pointing at something other than a dog, etc. And true, the correctnesses instituted by these rules can be trumped by other correctnesses: thus, I can move a bishop horizontally or say This is a dog! when pointing at a certain person to make a joke, for example. This is connected with the difference between prescriptive and restrictive rules discussed in Section 5.1: we have the correlative difference between the prescriptive should and the restrictive should not (again, with no sharp boundary between them). And the should’s relevant for chess or language tend to be of the latter kind. We are not told what we should do because we are only told what we should not do, which is needed for the delimitation of the virtual arenas in which we can play the games of chess or of communication. And, finally, then there is the term normativity, which has come to be used by some authors as a kind of magic formula with no clear content. It is clear that the term signals the presence of some norms (rules); it is, however, not always clear, in which capacity they are supposed to be present. Most often something is deemed normative if it involves actions that can be carried out correctly or incorrectly – language, for example, is usually called normative by those who are convinced that speech acts have such or another correctness conditions.

72  Preliminaries III Hence, our terminology is such that all the terms correct, rule, should, norm and normativity, as we construe them, are closely interlinked. Correctness and rules are two sides of the same coin and so is correctness and should, rules and norms are the same thing and normativity amounts to something as a presence of norms. 5.4 Summary The terminology surrounding rules and normativity is complex and intransparent. Different authors stress various features and use the same terms in different ways. Here we urge a few crucial distinctions. We must distinguish between laws, which are unbreakable, and rules, which can be violated. Also we must distinguish between rules in the sense of symbolic objects and rules in the sense of aspects of social reality. And we must also distinguish social rules (which are more or less arbitrary) from “rules of engineering” (which reflect how to reach a given goal). Moreover, it is good to realize that rules interesting from the viewpoint of language, reason, etc., usually have the character of restrictions rather than prescriptions. In general, we treat other basic terms surrounding the concept of rule as closely linked to it. In particular, we treat the term norm as synonymous with rule, we treat the concept of correctness as what is instituted by a rule (of course not necessarily an explicit one) and we take it that what is correct should be done. And we take something to be normative if it involves actions that can be carried out correctly or incorrectly. Notes 1 Okrent himself associates it with Aristotle. 2 Cf. Dor and Jablonka (2014) on “innovative behavior” and “interactive exploration”. 3 As Kusch (2006, p. 71) remarks: “It even seems to follow that only evolutionary biologists can ultimately tell us what we mean by our words. After all, according to Millikan’s proposal, what we mean by our words is ultimately derived from our unexpressed biological purposes. And these purposes are studied and investigated by evolutionary biologists”.

6

Normative attitudes

6.1 Sanctions Having clarified all the conceptual issues that we dealt with in the previous chapters, we can finally turn our attention to the nature of rules and of their roles with respect to human nature. We have seen that we cannot identify rules with anything like sentences (L-rules); that rules must have started as implicit. Now it is time to clarify what exactly an “implicit rule” is. And the answer is based on the thesis that the basic building blocks of an implicit rule are normative attitudes. We have seen that, according to Sellars, the middle way between the Scylla of regulism and the Charybdis of regularism is secured by an interaction between “enforcers” and “enforcees” (where members of the society may play both the roles). How do the enforcers make the enforcees display the pattern the former make the latter instantiate? Well, by preventing them from doing “improper” things and instigating them to do the “proper” ones. If the enforcees are already linguistic creatures, this may proceed via explicit instructions, but as language is also a system of rules that must be learned, this possibility is not always open. In such a case, the prevention and instigation must acquire the shape of much more demonstrative actions – perhaps positive and negative sanctions, or, in extreme cases, making it physically impossible for the enforcee to do the improper things and haul them in to do the proper ones. Thus, what is crucial from the viewpoint of the proliferation of rules as sketched by Sellars are the sanctions (especially “corrective” or “censoring” acts) of the enforcers, which squeeze the behavior of the enforcees into the required pattern. The term which has been employed for these kind of attitudes is “normative attitudes”. The first author who I know employed the term (and thus probably coined it) was Hart (1961).1 Then the concept assumed an important place in the foundations of the conceptual edifice of Brandom’s (1994) inferentialism. Nowadays, the term penetrates even into the theories of empirical scientists (Schmidt & Rakoczy, 2019). DOI: 10.4324/9781003388876-7

74  Normative attitudes Brandom (1994, p. 33), surveying the proposal of Haugeland (1982), writes: The approach being considered distinguishes us as norm-governed creatures from merely regular natural creatures by the normative attitudes we evince – attitudes that express our grasp or practical conception of our behavior as governed by norms. These normative attitudes are understood in turn as assessments, assignments to performances of normative significance or status as correct or incorrect according to some norm. The assessing attitudes are then understood as dispositions to sanction, positively or negatively. Finally, sanctioning is understood in terms of reinforcement, which is a matter of the actual effect of the sanctioning or reinforcing responses on the responsive dispositions of the one whose performances are being reinforced, that is sanctioned, that is assessed. We have seen that Sellars held normative attitudes for something that plays a central role in the process of constitution of rule-governed behavior and thus in the proliferation of rules, but their involvement with rules may be even more substantial. We have also seen that the primordial form of rules is the implicit one – the rules being implicit in what people do. And the presence of normative attitudes may be precisely what makes some human practices rule-governed. In this sense, rules themselves can be seen as constituted by the attitudes, as “clusters of normative attitudes” (a view ascribed by Brennan et al., 2013, to Hart, 1961). But we should say, more precisely, what exactly the normative attitudes are and how they constitute the implicit rules. The first thing to stress is that the attitudes are not a matter of mere psychology, they are not merely in one’s mind. They should be seen rather as behavioral syndromes: as, in Brandom’s terms, “dispositions to sanction”, typically as the “corrective behavior” in terms of which we try to divert people from doing what they should not do. The presence of such behavior is thus the hallmark of a genuine rule. As Kelly and Davis (2018, p. 59) put it: “[F]rom a psychological point of view, a distinctive, defining feature of norms concerns approval and disapproval, and the use of punishments and rewards to influence behavior. […] In short, norms imply ‘oughts’, and ‘oughts’ imply punishment and reward”. The second thing to stress is that as normative attitudes lay the foundations of rules, they themselves cannot presuppose rules or meanings. They are nothing like propositional attitudes; we do not assume a normative attitude because we comprehend a rule. The normativity of the attitudes is “primitive” in the sense of Ginsborg (2011). (More about this in Chapter 11.)

Normative attitudes 75 The third thing is that, of course, not every reaction to what somebody else does counts as a normative attitude. Such an attitude should not just be an expression of one’s inviting pleasure and avoiding pain. Even very simple organisms display such behavior, and the simpler they are, the more uniformly they would do so. But we would not want to call the fact that a cuttlefish squirts ink when attacked its normative attitude to the attacker. A necessary condition for calling an attitude normative is that it is an attitude to a kind of behavior, independently of whether she who assumes the attitude is the same person as the one who is its source or target. Thus, to hinder somebody from beating me is not a normative attitude until it also extends to other cases of people being beaten: if I hinder somebody not only from beating me, but also from beating anybody else.2 Finally, we can ask what the road is from normative attitudes to an implicit rule. It is, we can answer, the “collective adoption” of the rule by the community in question. The simplest version of such a “collective adoption” is a mere resonance across the community – the fact that the attitudes of most of its members tend to align with each other. However, both Sellars himself and some of his followers urge something stronger, some kind of attitude that becomes “collective” in a more robust sense. Sellars himself talks about “we-intentions” that develop out of the individual intentions, but are not reducible to them (Koons, 2021). Recent investigations of various forms of “collective intentionality” follow suit (Bratman, 2013; Tuomela, 2013; or Koreň et al., 2021). Thus, a (rudimentary) rule originates from the attitudes of individual members of the community that come to resonate with one another, but of course the resonation is likely to lead to various additional forms of “reinforcements” of the rule, which may lead to such a superstructure as we find underpinning some rules within our developed societies, such as police, law courts, jails, etc. There is a terminological issue here: as Wittgenstein taught us, and we have already discussed, any genuine normativity is a public, supraindividual matter. Therefore, it may be misleading to call an attitude of a single person “normative” – it is normative only in the sense that it is a potential component of a rule. But talking about, say, pro- and conattitudes to kinds of behavior would be too cumbersome. Anyway, now we are able to see the basic “anatomy” of implicit rules: their most basic constituents are normative attitudes. Also, we know something about their “physiology”: the most basic way of putting normative attitudes together to aim at an implicit rule is mere alignment, whereas a more complicated “physiology” is possible and is likely to develop out of the primitive one.

76  Normative attitudes 6.2  Brandom on normative attitudes Brandom himself eventually rejects the picture of rules as “clusters of normative attitudes” as not satisfactory. The reason why he introduced normative attitudes was that he did not want to reduce rule-governedness to mere regularity; and yet it seems that (though at the end of the day, rather than at the outset) we do arrive at such a reduction, only at a higher level: we reduce rule-governedness to regularities not of “ground-level” behavior, but of the “higher-level”, namely of the assessments of the “groundlevel” behavior. Thus, with an oversimplification, we do not reduce the rule One should bow to the chief to people usually bowing to the chief, but we do reduce it to people usually pestering those who do not bow to the chief. And this is what Brandom does not want to accept. Hattiangadi (2003, pp. 425–426), trying to convince us that “Brandom’s picture is still largely, or perhaps even entirely, founded on dispositions”, formulates the objection quite clearly: [I]t appears as though Brandom is offering a dispositionalist account of the determination of correctness – since the starting point includes nothing more than behavioral dispositions. Moreover, when we come to his positive account of the structure of the social practices necessary for conceptual content, nothing is added that would distinguish the account from dispositionalism. The structure of practices productive of content are certainly specified in deontic vocabulary, but the practices that institute normative status presuppose only the propensity to perform various forms of sanction. It seems that even for Brandom, sanctioning – whether it is beating with sticks or social exclusion – constitutes or amounts to the attitude of taking someone as committed or entitled. We have already seen that the most basic problem with the reduction of rule-governedness to regularity is that it erases some distinctions that appear vital – both the behavior of a person following the rule of traffic and of a stone falling down “following” the law of gravity will come out as the same kind (see Section 2.2). This, however, does not directly concern the reduction to the “second-level” regularity – the regularity of normative attitudes. But Brandom also cites subtler problems. One of them is the underdetermination of rules by regularities, as pointed out already by Wittgenstein. As we saw, any finite set of regularities underdetermines a rule – we can always find an infinite number of regular continuations. (And Brandom thinks that the problem recurs if we think about regularities of normative attitudes.) Another problem is that the normative attitudes themselves, according to Brandom, can be assessed for correctness.

Normative attitudes 77 Let me start from the last problem. Brandom, in effect, argues that a rule is induced not by a cluster of normative attitudes, but rather by a cluster of correct normative attitudes. And this stands in the way of straightforward naturalization – at least until we are able to naturalize the relevant notion of correctness. Therefore, he maintains, “defining normative attitudes in terms of dispositions to apply sanctions does not by itself reduce the normative to the non-normative – it just trades off one sort of norm for another” (Brandom, 1994, p. 42). I agree that insofar as normative attitudes can be sorted out into correct and incorrect, it is only the correct ones that matter. In such a case, however, we need some “second-level” normative attitudes that would render the original ones correct/incorrect. Are these “second-level” attitudes also correct/incorrect? And, more generally, is any kind of normative attitudes correct/incorrect, i.e. are there the corresponding “higher-level” attitudes for them? Clearly, some normative attitudes can themselves be taken to be correct/incorrect. For example, we can ask if it is correct to scorn a person who does not want to bow to the chief, and we can ask if it is correct to deride someone exclaiming This is a dog! when pointing at a kangaroo. The question, however, is, whether without exception all possible kinds of normative attitudes can be considered correct or incorrect. More generally, there may be rules for assessing rules. For example, we have theories of democratic social order, according to which some rules holding in democratic societies are correct, while others may be incorrect. Of course that there may be also rules for assessing the rules that are themselves used to assess rules. The question, however, is whether this can continue to infinity. My conviction is that the hierarchy of metarules does not run to infinity, that there are rules for which there are no metarules. These rules are then places where the non-normative gives birth to the normative: where a configuration of something from the realm of the natural produces something rudimentarily, but genuinely normative. This, however, is something Brandom does not seem to accept; for him, any rule is either correct or incorrect. If we assume that a rule must be present before that which it renders correct/incorrect (consider chess: the rules must be given already before I can make moves as such), then the assumption that for every rule there must be a corresponding metarule appears to lead to a straightforward infinite regress. To establish any rule, we would already have to have a rule for evaluating it, and hence a rule for evaluating this second rule etc. Thus, insofar as the rules we are talking about are social rules, viz. our creations, a metarule would have to be established before any rule. Can we avoid the regress via a bootstrapping? Can we say that a rule comes into being

78  Normative attitudes together with a rule for its evaluation? An obstacle to this seems to be that we would have to assume not only that a rule brings about another rule, but that the other rule brings about still another rule, etc. However, we should not forget that the primordial rules are implicit to our practices. They consist in the processes of evaluation; hence they cannot exist before the phenomena they target – they exist in the form of attitudes to the phenomena. Again, I think that we end up with some normative attitudes that are no longer the target of normative attitudes. However, this is what Brandom rejects; so let me consider the possibility that for all normative attitudes there exist normative attitudes that render them correct or incorrect. Now, no straightforward infinite regress is necessarily forthcoming – if rules are constituted by normative attitudes, then it would be enough to assume that every such attitude potentially evokes normative attitudes targeting it. However, there is still the air of a puzzle: Am I really able to readily assume normative attitudes to something I have never before encountered? The only plausible answer I can see is that I can acquire a general tendency to assume normative attitudes to any normative attitudes with which I am confronted (plus, of course, to many items other than normative attitudes). Thus, it is possible to imagine that once people gained their ability to assume normative attitudes, they became able (and maybe were instigated) to assume normative attitudes to anything (or nearly anything) people do, including any new normative attitudes they assume. Hence, the infinite regress here is tamed by the assumption that normative attitudes do not presuppose further, actual normative attitudes targeting them. It is possible to assume that we do not necessarily already adopt the normative attitudes toward any other normative attitudes, but only that we are ready to forge them when it comes to it. To every normative attitude, there potentially exists a normative attitude targeting it. 6.3  Norms as irreducible To avoid any misunderstanding, let me stress that there is a notion of the irreducibility of “the normative” to “the natural” that is different from the one discussed in the previous section and in the adoption of which I concur with Brandom. This irreducibility is most clearly manifested by the fact that there is no way of translating many normative claims to nonnormative (declarative) ones. True, to see this problem quite clearly, we must move forward in the process of instituting rules, from implicit rules carried by merely practical normative attitudes to explicit rules articulated in language. I agree that normative claims are not translatable into non-normative declarative

Normative attitudes 79 sentences, at least not without a remainder. In this sense, I think, Brandom is right to see the “normative world” as something new and unprecedented that we enter on our way to becoming human. But this irreducibility of the “normative” to the “natural” is nothing mysterious. It is, I think, the consequence of the fact that we have developed a new kind of attitude that cannot be transformed into those already existing. This new kind of attitude (the “normative” one) has given rise to a new kind of speech act, and it, in turn, has given rise to new kinds of linguistic vehicles. And these vehicles of the acts – we have dubbed them normatives – are not translatable into declarative sentences. Why are normatives not genuine declaratives and not translatable into them? Because, to put it briefly, they do not state what is the case. We have already indicated this in the chapter about Sellars: normatives spell out how things should be, viz. how they are in an ideal world at which we should aim. In this sense, normatives are like proposals: they envisage the kind of future their proponents vote for, and if other members of the community concur, this kind of future becomes their common goal (cf. Gauker, 2007). Normatives, then, may look like declaratives in that they appear to report something, that is, various proprieties. Such proprieties, however, do not exist in the same way as things do: they are correlates of our future-oriented projects. Speaking about the evolution of moral rules, Schloss (2014) urges a distinction between the explanation of their content (“what is the evolutionary cause or adaptive significance – if any – underlying central tendencies or cross-cultural universals in what humans feel, think, or do in morally salient situation x?”) and that of the underlying capacity (“to feel, think, or behave in certain ways”). Using this terminology, we can say that rules are irreducible with respect to their content (the content is not reducible to the content of non-normative idioms), but the capacity to produce them is explainable, without a remainder, in non-normative terms: we can explain why and how norms came into being, via the establishment of the peculiar speech acts (with their peculiar – and irreducible – contents), in purely naturalistic terms. We have already seen (Section 4.3) that Sellars (1953b) introduced a similar distinction, namely the distinction between his logical and causal reducibility of the normative idiom, to the non-normative one. The former (“the definability of Ought in descriptive terms”) concerns the reduction of Schloss’ content, while the latter concerns the explanation of his capacity. The latter case, according to Sellars, occurs when the explanandum “occurs in the antecedent of a properly constructed causal explanation only as a subordinate element in a descriptive mentalistic context”. Thus, what is explained is not “X ought to …”, but rather something like “X believes she ought to …”. As Christias (2015, p. 151) duly points out in

80  Normative attitudes his enlightening reconstruction of Sellars’ exposition: “Pace non-normativism, normative facts can indeed be irreducible to non-normative facts if this irreducibility is understood as being logical or conceptual in character (i.e., if it concerns relations of logical or conceptual entailment between normative and non-normative facts), but at the same time, pace normativism, this does not exclude the possibility that normative facts can be, in another (causal-explanatory) sense, reducible to non-normative facts”. It is interesting to consider Brandom’s views on the matter. I think he has no reason to disagree; indeed, Brandom (2008, p. 12) mentions this idea at least as an open possibility (“although normative vocabulary is not reducible to naturalistic vocabulary, it might still be possible to say in wholly naturalistic vocabulary what one must do in order to be using normative vocabulary”). On the other hand, he does not subscribe to it explicitly. But I think he has no reason to oppose this. Recall Hattiangadi’s objection that Brandom’s picture is founded on dispositions. It is clear that from what I have said up to now it follows that, though this is perhaps not the case with Brandom, I am willing to accept this (and the project of this book, to a significant degree, depends on it.) The birth of the normative out of the non-normative can be described in exclusively non-normative terms when viewed as a causal process that is a part of the evolution of us humans, as a story about how we came to assume a new kind of attitude to each other and to the world, and how this attitude launched an unprecedented development of our species. And a recapitulation of the story may yield us reductions of the above kind. But then why does Brandom keep opposing the idea that the “normative” is reducible to the “non-normative”, and why do I say that I agree with him? Why do I still reject the “logical” reduction, and why do I claim that the normative idiom cannot be translated into the non-normative one? The reason is that the above reduction treats the normative One should bow to the chief as just a declarative sentence, which it is not. Or more precisely, it can be used as a declarative (and then the above reduction is exhaustive), but its unique function, the function for which normatives were conceived, is different. By uttering a normative, we do not (or do not only) state a fact, but also issue an endorsement, we express our support to a proposal or a project. So uttering One should bow to the chief we do not (or do not only) state that this is – as a matter of fact – a rule that holds for our tribe but rather endorse it, support the proposal that we go on doing it in this way. We have already seen that the work of rules can be envisaged by speaking about “inner space”, which they open up and which we humans can enter, and here it may be helpful to invoke this metaphor. From “outside” of the space, we can describe what it takes to “play the normative game”, but only from its “inside”, can we become “players” of the game.

Normative attitudes 81 To unpack the metaphor, being “inside” the space means not only acknowledging, but endorsing the rules that delimit the space, and thereby being able to relate to the social environment in ways that are unprecedented and not reducible to those available “outside”; this concerns not only normative linguistic utterances, but also the pre-linguistic attitudes which the linguistic ones make explicit (while presupposing certain normative attitudes, namely those which establish the rules of our language games and thus make such utterances possible in the first place). The normative dimension of human life is – viewed from “inside” of the practices – something truly new and not reducible to what was here before. It is a matter of experiencing the world in a new way and unveiling the possibility of living a brand new form of life. From the outside, people living “within” a normative world (usually it is an assembly of interconnected normative spaces) look like displaying a complex behavioral pattern, characterized by a “second-level” behavior targeting some “first-level” one (“sanctions”). From this perspective, there is nothing that would escape a naturalistic description. So I agree with Brandom that the view from inside a normative space is not reducible to the view from outside – and normatives, as you utter them only from inside, are not reducible to declaratives, available to you if you are outside. However, insofar as I understand him, it is, according to him, not only this that is not available from the outside. From the outside, he maintains, you cannot get hold of the notion of correctness and any of the interlocked normative notions that mark the crucial dimension of our normative world. Thus, there is a battery of concepts that are accessible only from within. Hence, an outside observer not only cannot become a “player” of the game, but he cannot even become its faithful recorder. She cannot say: The natives hold as being correct bowing to the chief for she is not in possession of the requisite concept of correctness. But is this correctness really a concept? We said that its native use is within the normatives, sentences that are not (or not purely) descriptive. On the other hand, normatives are so close to declaratives that it is natural to see them as endowed with truth values. (This is why the normatives can also be used in the declarative mode and why even when used in the normative mode they are often felt to have a descriptive component.) Hence it is correct (true) to apply this concept of correctness to some things and not to other things. But this is not, according to Brandom, what an outside observer can do, or at least do properly – she is not in possession of the concept.3 This is a consequence of the Brandomian self-encapsulation of the normative. According to it, anything, especially a bundle of normative attitudes, can be a source normative force only if it is correct and hence also a target of normative force; and so the correctness cannot be naturalized. Hence the corresponding space consists of irreducibly normative

82  Normative attitudes components; it cannot be seen as a sophisticated reconfiguration of the components of the natural world. And this is where I come to disagree: as I see it, what is peculiar to the view “from within” and what is not available to the outside observers is not a kind of knowledge. It is not so that the insiders were able to know something – perhaps that something is correct – that cannot be known by the outsiders. Their peculiar experience, which is mediated by the normative space and which is not shared by the outsiders, is not an epistemic matter – it is a matter of specific ways of experiencing the world. As I already said, from outside I can give a faithful record of how the game evolves, but it is only from inside that I can play the game, become an actor in the evolution of the game. Consider chess: the difference between a player and a spectator is not that the former knows something that the latter does not. Both of them know the rules, both see the moves and both of them can record how the game develops. But there is clearly a deep difference between them: the player can influence the development, can win the game or lose it; and indeed her whole attitude to the game is tinged with her perspective of winning or losing. Though the spectator may know everything about checking an opponent’s king, it is only the player who can check the king of her opponent. Nevertheless, I am convinced that there are normative attitudes that are the “bedrock” source of normative force without being its target – attitudes that are, in themselves, neither correct, nor incorrect, and yet they institute correctness. This, I am convinced, makes it possible to naturalize even the correctness of the attitudes – and thus, we can approach the “normative world” in a scientific way. Hence, while normative attitudes certainly lend themselves to scientific scrutiny, if what I say is correct, the same holds for correct normative attitudes. In this way, the anchoring of the normative world within the natural one becomes explicit. But one aspect of the “normative world” that is not accessible to science survives. In the end of the previous section, I concluded that while the birth of the “normative” out of the natural can be described in exclusively non-normative terms, this does not lead to a “logical” reduction. Now we can express this in more colorful terms: while the establishment of rules can be described in exclusively non-normative terms when we view them from the outside, this does not capture the inside, which the rules have the peculiar ability of building up. 6.4  Rules all the way down? Return to chess. The ways I may move the pieces over the chessboard may be either correct, or incorrect – viz. in accordance with the rules of chess

Normative attitudes 83 or not. Are the rules themselves correct or incorrect? Well, the rules may perhaps be called “correct” in the sense that they make up an enjoyable game. But this is not well expressed by the term “correct”, better would be to call them “successful” or “sophisticated”. In any case, they are not correct/incorrect in the sense in which the individual moves are. Now suppose nobody has bothered to formulate the rules of chess explicitly, hence they exist only via the normative attitudes of the players and other supporters. Are the attitudes correct/incorrect? Perhaps we should call them correct if they enact the (genuine) rules of chess, and incorrect if not? But we must not forget that the rules exist only through the attitudes, hence the rules are nothing else than what the attitudes enact. There is no significant room for deviation. But perhaps some of the attitudes can be incorrect in that they are “out of tune” with the majority, and these attitudes can be deemed incorrect? If so, then we have a theory, according to which correctness is easily reducible to non-normative facts. Hence, if there is significant room for deviation, then the deviation, and hence the corresponding correctness, can be easily naturalized. Or consider a language, such as English. It is clear that it is correct to say This is a dog when pointing at a dog, and not pointing at other things. This is what new speakers learn when they are initiated into English – we say that it is a matter of the meaning of the word dog. (Let us leave aside, for the time being, whether the rules hold because of the meaning, or whether the rules are directly (co-)constitutive of the meaning.) Is the rule itself correct? Is it correct that we use the word dog for dogs and not for, say, storks? We could perhaps make up a very specific context in which this question makes sense, but normally we would certainly find it senseless. Or consider rules of logic, such as that of the disjunctive syllogism (DS), which would take us from A or B and not A to B. This rule is (taken to be) correct. Why? From the contemporary viewpoint, it is because it can be shown to be correct, it can be proved. For example, within the Gentzenian system of natural deduction, it can be proved using the elimination rule for negation (NOT-E, taking us from A and not A to B) and the elimination rule for disjunction (OR-E, taking us to C from A or B and the respective proofs of C from A and B).4 Are the elimination rules correct? Insofar as this question is understood in its descriptive sense, it is empirical, for they target the English words not and or, and it is an empirical question whether English speakers approve of these usage rules. Alternatively, we can construe the question in a normative sense: Do the rules constitute negation and disjunction in a “right” way (and thus perhaps show how the English words should be used)?

84  Normative attitudes The descriptive version of the question, obviously, is not pertinent for a logician or a philosopher – answering it is a matter of empirical linguistics. But what about the normative version? What can the “right” in it amount to? It is clear that (NOT-E) or (OR-E), in contrast to (DS), cannot be proved (in a nontrivial way), because they are used as “axioms” to characterize negation and disjunction. Can we say that they characterize negation or disjunction correctly (or, as the case may be, incorrectly)? Prima facie it might seem that indeed we can. It may seem that it is obvious to characterize both the English or and disjunction in general in terms of (OR-E), rather than, say, modus ponens, which would take us from A and A or B to B. However, it is not difficult to see that this is really only a trivial sense: we feel that this is correct, for this is what it takes to be disjunction, and because we are convinced that the English or is an instance of disjunction. This does not therefore amount to more than saying that (OR-E) is correct because or is a disjunction and disjunction is, ex definitione, something that is governed by (OR-E). Thus (OR-E) is correct for disjunction in general because disjunction is what obeys (OR-E), and it is correct for or in particular since or has been found, empirically, to obey it (to be a disjunction) (cf. Peregrin, 2010b). Hence, like the rules of chess or those of language, the rules of reasoning are neither correct, nor incorrect – though they are certainly eminently useful. They help us produce disjunction, negation, implication etc. that are superb tools of our reasoning. Also, the corresponding normative attitudes are not correct/incorrect, save for a trivial sense that does not exclude them from the natural world. And I think that this holds about rules in general. There may be rules for the assessment of rules, but sooner or later we reach rules that are not rule-evaluable, they are not correct or incorrect. We reach the Wittgensteinian “bedrock”, where our “spade is turned” (PI, §199). They may be useful (and this is probably why they flourish, like the rules of logic which govern our argumentation and, as a matter of fact, make it possible at all). Thus, in contrast to Brandom, I think that there are some basic normative attitudes that are themselves neither correct nor incorrect. Notice that this also alleviates the Kripkean worries (which, as mentioned earlier, Brandom cites as the other obstacle to the naturalization of normativity). If correctness rests – ultimately – on normative attitudes that are no longer themselves norm-evaluable, then all we need to take into account is their de facto occurrence. We concur in assessing certain results of addition as correct – and hence they are correct. Why is it therefore correct to use “+” as addition rather than “quaddition”? Well, because addition is what our normative attitudes – de facto – are found to converge upon. We have considered the argument that no de facto regularity is able to determine a unique rule. According to this view, a rule is like an infinite

Normative attitudes 85 regularity, and any finite part admits many alternative continuations. No matter how many concrete results of cases of addition are fixed, this underdetermines the cases not yet encountered. This is the picture sketched by Wittgenstein (1953): 218. Whence comes the idea that the beginning of a series is a visible section of rails invisibly laid to infinity? Well, we might imagine rails instead of a rule. And infinitely long rails correspond to the unlimited application of a rule. Wittgenstein does not subscribe to this picture: 219. “All the steps are really already taken” means: I no longer have any choice. The rule, once stamped with a particular meaning, traces the lines along which it is to be followed through the whole of space. – But if something of this sort really were the case, how would it help? No; my description only made sense if it was to be understood symbolically. – I should have said: This is how it strikes me. Hence according to Wittgenstein, we should not imagine a rule as a given infinite regularity (of which we can be always confronted only with a finite part): the illusion that this is so comes from the fact that when I do follow the rule, I do not feel an uncertainty and a need to choose. I just do what strikes me as correct. It is clear that being rule-governed is something more than being simply regular. But what I proposed is that the “more” consists in something else than a mysterious presence of an infinite continuation of what I do, but rather the presence of normative attitudes: being rule-governed is being determined by normative attitudes. Now, however, as some critics pointed out, there reappears the above objection: no de facto normative attitudes are able to determine a unique rule. No matter how many times the results of addition are approved or disapproved, this will not determine whether the “correct” kind of addition is the standard one or its alternatives (like quaddition). A “dispositional” answer to this question, which is available to the naturalist, is that while the practices of addition have produced (or have approved of) only a finite number of cases as of yet, the practitioners have the dispositions to produce all the needed remainder. The results are thus present, as it were, in the dispositions. But while I think that we cannot wholly make do without the concept of disposition, I do not think this answer, as it stands, is more than a sleight of hand. It is no less void than

86  Normative attitudes the claims of the non-naturalists that the rule – the infinite continuation of its part that has already come to the open – lies somewhere in the Platonist heaven or in some mysterious depths of human minds. I think that the only way out of the impasse is the rejection of the assumption that a rule is an infinite regularity. According to a naturalist, any form of the infinite is only our projection of our finite experience. We are, to be sure, very well equipped for such projections: our survival in the world depends on our ability to transform our past experience into a reasonable expectation of the future. It is hard to deny that we are disposed to give specific results to arithmetical queries we are made to face. I, for one, am disposed to give certain results and I expect that others will tend to give the same ones. And it is this structure, of past resonances and of expectations of future resonances yielding the normative attitudes, that amounts to the existence of a rule. If we experience people giving, and approving of, results for various cases of addition, we naturally expect which results they will give in the future, possibly in cases we have never experienced yet. We also presume that there will be an agreement with respect to these results as it has been before. (We, for example, assume which results of addition will be generally expected even for big summands never encountered yet. And I reckon that as a matter of brute fact, we expect that they will be the result of addition, not quaddition. That this is so is a matter of our being cognitively equipped as we are.) Thus, it may be reasonably expected that we will continue to add, rather than quad. This, however, is not guaranteed: a rule is not a rail leading us into the infinity in an inescapable way, a rule is always still in the making. Hence, a rule is based on the resonance of our normative attitudes; there is no guarantee that our dispositions to evince the attitudes will ensure resonance in the future. However, it is reasonable to expect this – just as it is reasonable to expect that the sun will keep rising each morning in the future. There is thus a rule insofar as there is a resonance of normative attitudes, and expectation that the resonance will continue and there will be no substantial frustration of the expectation (yet?). Consider the picture sketched by Hattiangadi (2003, p. 426): Consider a face-to-face interaction between two members of the kind of community described by Brandom (call them John and Emma). John says to Emma, pointing, ‘that’s red’. We are supposed to imagine that John makes these sounds and gestures, and Emma, taking all of this in, attributes certain commitments and entitlements to John. This just means that Emma becomes disposed to sanction John-disposed, that is, to punish John under some circumstances but

Normative attitudes 87 not under others. Imagine, further, that at some later time poor John is punished. The question is what has John been punished for? Has Emma attributed the commitment to say ‘that’s not blue’, or has she attributed the commitment to say ‘that’s not grue’? Which of these commitments has John violated? I think that saying that “Emma […] attributes certain commitments and entitlements to John” just means “that Emma becomes disposed to sanction John” is wrong. Being disposed to sanction is not yet ascribing a specific commitment: specific commitments may emerge from the general pattern of sanctioning when they tend to support types of behavior in the “agent-neutral” way. And in no case can they be more specific than what is implicit to the attitudes (and expectations of the participants). If John is punished for saying “that’s red” in some particular case, and if he and others are usually punished by saying this when pointing at things that are not red, then there is a rule prohibiting saying “that’s red” when pointing at things that are not red. Why is it not the rule that prohibits saying “that’s red” when pointing at things that are not red when not at the same time wearing a bucket on one’s head? Well, as far as our present concerns go, there is no need to distinguish between the two rules; if the need for this arises, then this will be resolved. Can we then say that one who “quadds” instead of adds is not only weird, but also wrong? And here, I am convinced, we must say that this is a place where the non-normative gives rise to the normative. Here what is right is what is accepted by us: it is right because our attitudes – as a matter of fact – are what they are. We tend to take additions as correct, and “quadditions” as incorrect. Let me stress that this is not to generally embrace the “democratic harmony” theory and equate correctness with being in tune with the majority. First, it is not only doing what others do that makes me correct, but rather accepting what others accept – hence aligning my normative attitudes with others. But even so, there are, of course, a lot of correctnesses that are not a matter of a majority vote. (In case of sentences, for example, while it may be the case that their meaning is a matter of such a vote, their correct assertability – which, according to Sellars, amounts to their truth – is typically not. We will have more to say about this in Section 12.4.) However, there are some primitive (“bedrock”) cases where it is impossible for correctness to be anything over and above an endorsement by the majority. This holds for the most elementary ways of using words – what could their correctness consist in save the endorsement of the members of the relevant linguistic community? (More about this in Chapter 8.) Let me also stress that the fact that there must be normative attitudes that are not themselves norm-evaluable does not mean that these attitudes

88  Normative attitudes must be genetically hardwired. In fact, we will see (in Chapter 10) that the attitudes can proliferate by way of pure pedagogy, admittedly resting on a genetic infrastructure, but not being themselves necessarily genetically encoded. 6.5  Hume’s thesis and the Frege-Geach problem The famous Hume’s thesis states that we cannot infer an ought from an is, i.e. that from purely non-normative premises we cannot infer a normative conclusion (Hudson, 1969). It seems, however, that the project of naturalizing normativity does bring about a certain version of the reduction of ought to is: we argue that what individuals ought to do (in our social sense of ought) is determined by the attitudes of members of the communities to which the individuals belong. If we think that we can see normative attitudes as purely natural phenomena (pace Brandom), does it mean that we reject this thesis? In particular, if we say that being correct is being an object of positive normative attitudes, does it follow that we can infer, e.g., (H) People of our community support bowing to the chief and condemn not doing so           We should bow to the chief The answer to this question is negative, and in this sense, I maintain that Hume’s thesis does hold. But to explain why, we must return to the distinction between normatives and other declarative sentences tabled in the previous section. First we consider a similar inference (H*) People of our community support bowing to the chief and condemn not doing so             We have a rule to bow to the chief Unlike (H), (H*) does hold. However, its conclusion is not really normative, it is a mere statement of a fact. If it is a fact that the normative attitudes of the people of our community are as the premise tells us they are, then it is a fact that we have the rule. What is the difference between (H) and (H*)? The most basic difference is that the conclusion of (H) is not declarative, but rather normative. (More precisely, it can perhaps also be read as a purely declarative sentence, perhaps as a synonym of the conclusion of (H*), but its more standard meaning is different.) By claiming We should bow to the chief I do not – or do not only – state that there is this kind of rule, but rather also that I subscribe to it.

Normative attitudes 89 We saw that despite appearances, normatives are not declarative sentences: they express the brand new kind of attitudes which we humans have come to acquire and which we dub normative. They may be seen as declarative sentences insofar as they are seen as spelling out that the corresponding rule, as a matter of fact, exists (or perhaps as stating what holds in an ideal state-of-affairs). Thus, we can see the meaning of a normative as consisting of two parts, a declarative part and an endorsement part; and it is the endorsement part that makes it different from a declarative. The problem of (in)correctness of (H) is the same as the problem of (in)correctness of (H**) People of our community support bowing to the chief and condemn not doing so               Do we bow to the chief? Just like a question, a normative is not something that would fit as a conclusion (nor, for that matter, as a premise) into ordinary inferences. That a normative is, roughly, a declarative plus an endorsement also explains how this view overcomes the so-called Frege-Geach problem (Schroeder, 2008): the problem that a theory of meaning must take into account that a sentence can not only function as a self-standing utterance, but also as a component of more complex sentences. Its meaning must therefore not only explain its former role, but also the contribution it brings, in its latter role, to the meaning of the complex sentences of which it is a part. Suppose that I have theory that normatives are just expressions of delight or of disgust (sometimes called a “hoorah – boo theory”). Now consider a sentence like If we should bow to the chief, then … . We expect that its meaning is a function of those of its parts which work well if the meanings are truth values or classes of possible words, but the theory we consider does not tell us anything about a meaning that could serve as a contribution to meanings of complex sentences. Does not our theory fall prey to this problem? Not really. As the meaning of a normative consists of the declarative and the endorsement parts, when a component of a complex sentence, the normative is stripped off its endorsement part and retains only its declarative part. (The declarative part of its meaning can be modeled in any of the well-known standard ways, like a class of possible worlds.) In this way, normatives can contribute even to the meanings of the complex sentences in which they occur.

90  Normative attitudes 6.6 Summary What is crucial from the viewpoint of the proliferation of rules as sketched by Sellars are the sanctions of the enforcers that squeeze the behavior of the enforcees into the required pattern. Brandom calls these sanctions “normative attitudes” and puts them into the foundations of the conceptual edifice of his inferentialism; they are, according to him, “dispositions to sanction, positively or negatively”. The most elementary kind of (implicit) rule consists in the normative attitudes resonating across the community. However, this, according to Brandom, does not mean a reduction of rules to dispositions, for Brandom insists that it is only correct normative attitudes that are relevant; hence a rule is made up by correct normative attitudes, which are made correct by the correct normative attitudes we assume to them, which are made correct by the correct normative attitudes etc. This infinite regress prevents the naturalistic reduction of rules to dispositions to evince the attitudes. Hence, according to Brandom, “defining normative attitudes in terms of dispositions to apply sanctions does not by itself reduce the normative to the non-normative – it just trades off one sort of norm for another”. But here I part ways with Brandom: I think that though some rules may be evaluated as correct or incorrect, at bottom there must be rules that render something correct without themselves being either correct or incorrect. This, pace Brandom, leads to the naturalistic reduction. Notes 1 See Shapiro (2006) for an explanation of the role of the concept of normative attitude for Hart and especially for his own conception of legal practices as having an “inside”. 2 Giromini (forthcoming) calls this constitutive property of normative attitudes projectivity. 3 Note that an outside observer will be in possession of a concept of correctness, though one different from the one available only to the insiders. The point is that the outside observer, insofar as she is able to use language, is also within a normative space, in particular in the space of meaningfulness of the language. (Remember Sellars’ point that we humans cannot be outside of all normative spaces.) So the observer, though outside of the normative space of the natives (and deprived of their concept of correctness), is inside of her own normative space (and in possession of her own concept of correctness). 4 The proof can be depicted as follows: [A]             not A                 [B]                   B                         B                           A or B                                                                            B

7

Rules in the natural world

7.1  Ontogenesis of rules Now let us look at rules from the “outside”, i.e. the perspective of ordinary science. If rule-following behavior is not simply a behavior that is regular, and if the surplus that makes it such is not something esoteric, it should be scrutable even from this perspective. Hence, we are warranted in expecting an answer to the question as to what is it that distinguishes human normative practices – to repeat, viewed from “outside” – from behavior that is merely regular. And my answer is that it is the presence of normative attitudes that is revealed and characterized in normal scientific – viz. naturalistic – terms. What has been said about normative attitudes up to now are mostly philosophical speculations, so the question that comes to mind is: Do we have any empirical evidence that normative attitudes exist and are ubiquitous in human activities? And it turns out that during the recent two decades normative attitudes and phenomena close to them have become the subject matter of numerous empirical studies, which not only confirm their existence but also tell us a lot about them. In 2008, when the role of rules within human ontogeny started to move into the focus of empirical scientists, Rakoczy et al. (2008) published the results of an experiment, in which children (three-year-old) first learned how to play a simple game, and then they were exposed to a puppet violating the rules of the game. It turned out that the children displayed “normative responses” (i.e. tried to correct the puppet), but they did not display them in the context where the puppet did the same thing outside of the context of the game. The authors concluded that “even very young children have some grasp of the normative structure of conventional activities” (p. 875). A year later, a similar experiment was reported by Casler et al. (2009). Young children (two- and three-year-old) familiarized themselves with the functions of some artifacts. A puppet subsequently used the artifacts in DOI: 10.4324/9781003388876-8

92  Rules in the natural world atypical ways, which also elicited “normative responses” on the part of the children. The authors summarized their results by claiming that they “depict toddlers as already sensitive to the uniquely human, normative nature of tool use” (p. 240), but an important thing this experiment indicated was that young children were not only quick in grasping genuine rules, but that their tendency to extract rules from the environment was so intense that it led them to over-generate – to interpret as normative what they felt as normal. Many similar studies followed. In 2011, for example, Clément et al. (2011) published the results of a similar experiment, in which children (three- and four-year-olds) underwent versions of the false-belief task, some of which were modified by the introduction of a rule or a regularity. It turned out that when the task included a rule, the performance of three-year-olds – who fail traditional false-belief tasks – significantly improved. What was also remarkable about this experiment was that in one of its versions, the presence of the rule which played the crucial role in it was not indicated verbally, but exclusively in terms of the normative attitudes. (“Instead of giving a verbal rule, like in Study 1, Study 2 tested the capacity of 3-year-old children to infer the presence of a rule from the censoring behaviour of a significant figure (teacher) and to use this inferred rule in order to predict the behaviour of the protagonists”, p. 913.) In this way, it vindicated the claim that a rule can exist solely in terms of attitudes. The results of a similar experiment were published, in the same year, by Schmidt et al. (2011). It concentrated on the question of the basis on which young children detect the presence of a rule, in the absence of its explicit linguistic articulation. The result was that the only cue the children used was adult social pragmatic marking of the action as familiar, as if it were a token of a well-known type (as opposed to performing it, as if inventing it on the spot). “These results”, the authors claim (p. 530), “suggest that—in the absence of explicit normative language— young children interpret adult actions as normatively governed based mainly on the intentionality (perhaps signaling conventionality) with which they are performed”. More studies followed. What the studies reveal – and what is important for us here (it is not always what the studies concentrate on primarily) – is the following: children are extraordinarily sensitive to the presence of rules in their social environment, they tend to interpret their findings of how things are done as how things should be done, thus over-generating. The ultimate indicator of rules, however, is the normative attitudes – which, in the most straightforward case, are expressed explicitly by linguistic utterances, which can be, however, also expressed merely implicitly, via

Rules in the natural world  93 “censoring behavior”. This further indicates that rules, in their primordial form, can consist merely in nonverbal normative attitudes. What experimental evidence has shown us is that we humans have a specific sensitivity for rules – we have, we can say, an uncanny knack for identifying rules. This concerns not only adult humans, but especially human toddlers. When an infant starts to explore the world, she inevitably encounters boundaries to the “space” which she can inhabit – via finding out that some of her activities get regularly frustrated; they are, as it were “bounced off” by the environment. We might perhaps conjecture that cognitively she starts to represent her world in terms of what is possible1; but relatively soon, as her caregivers switch to more subtle ways of posing the normative boundaries, she recognizes the normative boundaries for what they are – something essentially different from natural ones (Kalish, 1998). What turns out to be the case is that the infant is already prepared to meet the rules halfway. She is sensitive to their occurrence in her environment (which, of course, means that she is sensitive to the normative attitudes that are the vehicles of the rules), readily understanding that some regularities within the environment (especially those displayed by the caregivers and later by a growing range of “persons”) are more than just regularities, that they manifest rules, viz. something that ought to be followed. The proclivity of children to pick out rules from the environment is documented especially by the fact that in this respect they often over-generate, expecting or seeing rules where there are mere regularities. The experiments described in the previous section clearly documented that children tend to interpret what adults do (including instructions that are not verbalized, but mediated merely by normative attitudes) as the correct way it is done, even in cases where such an expectation is not substantiated. Rakoczy and Schmidt (2013) introduce the term “promiscuous normativity”, which expresses the fact that children sometimes draw normative conclusions too quickly.2 Thus, empirical research indicates that children detect and take on rules in a surprisingly swift way.3 7.2  Phylogenesis of rules The empirical results concerning normative attitudes presented in the previous section concern human ontogeny, the role of the attitudes within the integration of newbies into human societies. This is a matter that can be easily documented. The situation, of course, is much more complicated with respect to phylogeny, which cannot be studied in such a direct manner. Here any theories cannot but be largely speculative.

94  Rules in the natural world Some scientists researching the development of mankind, however, do not doubt the basic importance of norms in this process. Thus, Henrich (2015, p. 188): Over our evolutionary history, the sanctions for norm violations and the rewards for norm compliance have driven a process of selfdomestication that has endowed our species with a norm psychology that has several components. First, to more effectively acquire the local norms, humans intuitively assume that the social world is rule governed, even if they don’t yet know the rules. … Second, when we learn norms we, at least partially, internalize them as goals in themselves. This internalization helps us navigate the social world more effectively and avoid temptations to break the rules to obtain immediate benefits. Similarly Laland (2017, p. 267): [A]t some juncture in our history, our ancestors began systematically to correct the behavior of the individuals they taught; in the process, they shifted their society away from reliance on mere conventions and toward governance through norms. People stopped illustrating a way to behave and began insisting on the way to behave. Eventually, each society was characterized by a particular set of norms that dictated how individuals should behave (e.g., how they should build a fire, how they should catch a turtle, how they should till the soil), each of which was propagated through verbal instruction. Norms specify rules of social interaction too, including specification of how people should respond to norm violation. With the advent of norms, hominin social life effectively became transformed from simply living in groups to identifying with the group, abiding by its rules, and privileging in-group members. Norms facilitated group coordination and thereby substantially enhanced the society’s capacity for cooperative endeavor. In order to resolve conflicts or prevent social problems from arising in the future, institutionalizing the norm explicitly as a “rule of law” that all members of the society must abide by, and agreeing on sanctions for violators, would sometimes have been necessary. The question, of course, is which kind of behavior became the target of normative attitudes – and why and how did this happen. But there is no easy answer to such questions as rules do not leave any direct archeological record. But it is not difficult to imagine plausible scenarios of how rules, considered consonant normative attitudes, could have appeared.

Rules in the natural world  95 It is a platitude that every organism that is able to modify its environment so that it becomes better suited to its needs is likely to do so. In the case of inanimate parts of the environment, there may be straightforward ways of bettering it (like removing hindrances). However, to influence those components of the environment which consist of other organisms in a useful way may be much more difficult. There is, nevertheless, no reason to think that an organism capable of doing so would shun influencing even this part. What is to be expected is that the organism would try to defend itself from the endangering behaviors of other organisms, especially its conspecifics. A next step may be to proactively discourage others from producing these kinds of behavior (and perhaps encourage them to produce beneficial ones). But this is still in line with fostering one’s own flourishing, albeit in a sophisticated form; there is still nothing normative in play. The rules come into being at the point when the pro- and con-attitudes come to be applied to kinds of behavior, irrespectively of whether the behavior in question directly concerns the individual who assumes the attitudes, and when these come to resonate across the community. It is one of the cornerstones of the complex behavioral syndrome that produces what we call human normativity. It is the step from a purely egocentric “evaluation” in terms of harmful for me vs. helpful for me to the more abstract right vs. wrong, which applies to myself as well as to everybody else. But why would one take the step? It is readily imaginable that some individuals behave in a way that is beneficial for the whole community, which comes to be supported by others; or that they behave in a harmful way and come to be hampered. Let us consider an example (inspired by Beisecker, 2013). It is likely that our predecessors accompanied their behavior by vocal displays, at first perhaps involuntarily. Anyway, some of these displays may be provoked by certain events in the (social or natural world), and as such they may be utilized by others as indications of the events, in specific cases perhaps as warnings. The association of a certain kind of vocal display may indicate an approaching danger in a similar way in which associating clouds are taken to indicate an approaching rainstorm. This is as yet nothing over and above the way in which regularities in their environment are exploited even by simpler organisms (at least by those able to learn through trial and error). However, just as clouds may sometimes fail to be followed by a rainstorm, such displays may fail to be followed by a real danger. But the latter case, unlike the former, may lead to consequences different from those invoked by clouds or other natural phenomena. While in their case the only sound reaction is the modification of one’s expectations to fit reality, in case of displays of one’s peers we can also try to modify reality to fit the expectations – we may try to force those who display the behavior to

96  Rules in the natural world modify it so that the expectations will not get frustrated. This creates an opportunity for the emergence of normative attitudes: it is understandable that if there are sounds usually indicating danger, there may be a tendency to make the very sounds predict danger in a reliable way. I base my behavioral decisions on expectations I acquired inductively on the basis of observed regularities – natural and social. And my decision – viz. if I evince the normative attitudes trying to influence the behavior of my peers – may turn out to be “good” in the sense that it will help me in the future, or “bad” in the sense that it will not. Of course, I do not know – for sure – whether my choice will turn out to be “good” or “bad” (whether, that is, it will lead to less or more frustration of my expectations). I can only estimate this. And here is where something quite specific for us humans may step in: we may take into account not only the behavior in question, but also the reactions of others to it. The interesting thing is that the decision whether to react to the frustration of one’s expectation by changing the expectation or by trying to modify the behavior of the individual who gave rise to the expectation may be based on the reactions of our peers. If most of them try to make the individual change her behavior, then I join them, if not, I modify my expectations.4 Note that from the viewpoint of evolution, there is nothing mysterious about this: the behavior I thus come to display is wholly in my own interest because I strive to maximize the chances that the choice of my behavior will be “good”. This strategy, however, has profound consequences. First, it tends to align the reactions to behavior across the community so that the encouragement of the displays (or, as the case may be, their suppression in improper cases) becomes what could be called a “collective endeavor”. Second, it tends to make this endeavor independent of any individual’s parochial perspective. Moreover, by consulting the reaction of others, I may gain a feeling for which displays should be suppressed (and which should be encouraged), hence the former can be termed correct, and the suppressions of the incorrect ones may be termed sanctions. In this way, a rudimentary normativity is forthcoming. Of course, such rudimentary normative attitudes do not yet provide for fully fledged rules. But supplementary steps are readily describable. The normative attitudes keep aligning with each other so that they form a compact virtual “barrier”. (Perhaps they interconnect so as to form a “collective intention” or a “we-intention”.) Normative attitudes may give rise to social mechanisms of prosecution, punishments and rewards. But of course the greatest of these supplements is the development of language, itself based on implicit rules, but also presenting us with the possibility of making our rules explicit, and this essentially upgrades the nature of the whole game (Peregrin, 2014a).

Rules in the natural world  97 7.3  Rules and coordination problems Vocal displays usable as indications resp. warnings, of course, are not the only kind of behavior that is deemed so useful as to be cemented by normative attitudes and hence rules – thus, certainly not all kinds of rules are likely to have originated in the way described in the previous section. In general, rules come into being via some kinds of behavior being promoted as correct and others as incorrect. “Rules of engineering” or “technical norms” are thus promoted for a clear reason: they make us able to deal with the environment in an effective way. But the rules we are interested in, non-instrumental social rules, are not of this kind, they facilitate our social coexistence and are not determined that uniquely. It seems that a lot of different rules could do the work of the actual ones as well as they do. (Instead of driving on the right side of the road, we could just as well drive on the left side, and perhaps we could also effectively regulate traffic in other ways.) So here we can ask: Why have we promoted precisely these ways of behavior to rules? We have seen that the promotion proceeds via the establishment of corresponding normative attitudes; so the question is why have we started to assume the normative attitudes toward some, but not other, ways of behavior? And a simple, general (though not very informative) answer is that we targeted those ways of behavior that were turning out to be useful (in such or another way) – sometimes probably just by chance. To make this answer more specific (and at least a little bit more informative), we must return to the situation where a (perhaps spontaneous) reaction to a circumstance turns out to be useful (e.g., as a warning, but there may be many other reasons) for the members of a tribe. Then it seems to be natural that the members slowly come to compel each other to react always in this very way and perhaps not display this particular reaction in any other case. This might acquire the form of a brute force on those who do not comply, or some rather subtle means. And gradually it is likely to develop into a system of far more subtle incentives and rebukes, which may have to do merely with the participation in the profitable praxis of “early-warning” and later, more generally, in that of linguistic communication. It is clear that the case we have just discussed, the vocal (or other) displays which turn out to be useful as they can be taken as indications of (natural and social) events, is only one among many. One of the other cases, often seen as the most important one, is behavior that establishes itself spontaneously to solve various “coordination problems”, which is then naturally cemented by rules. Imagine a sidewalk on which people (or, for that matter, other animals) walk in both directions. If the traffic is light, then two individuals approaching in opposite directions simply sidestep

98  Rules in the natural world each other. But if it is heavy, such sidestepping would be so frequent that it may wreak havoc. Therefore, it is very probable that the traffic develops to the state where the individuals going in one direction travel on one side of the sidewalk, while those going in the opposite direction travel on the other. (And, it is important to note, this may happen without the individuals realizing that this is a reasonable way to behave. It may simply be easier to walk behind somebody than to clear the way for oneself, and this may eventually lead to a reasonable arrangement.) It hardly needs to be added, then, that the result is so satisfactory that it evokes normative attitudes so as to hold it in place. Hence, it would seem that normative attitudes come to support ways of behavior which are beneficial for the community in question and into which the community happens to fall or perhaps sometimes falls in a systematic manner. The idea of spontaneous solutions to coordination problems was elaborated by Lewis (1969); according to him, in a typical case, rules have to do with conventions which have to do with such problems. In his influential book, Lewis states that conventions arise as implicit solutions of coordination problems. In particular, he claims (p. 42): A regularity R in the behavior of members of a population P when they are agents in a recurrent situation S is a convention if and only if, in any instance of S among members of P, 1 everyone conforms to R; 2 everyone expects everyone else to conform to R; 3 everyone prefers to conform to R on condition that the others do, since S is a coordination problem and uniform conformity to R is a proper coordination equilibrium in S. One of the examples Lewis gives concerns telephone calls. Suppose that the calls are frequently terminated abruptly for some reason such as the signal dropping-out. Then the following convention may develop (and it did develop in a documented case): the one who is to call to renew the interrupted call is the one who was the original caller before the interruption. This is a useful convention, because it prevents both participants from calling each other (and finding the other’s line busy) or both waiting for the other’s call. Lewis also addresses the relationship of convention to rules and norms. According to him, many conventions (though not all) are already norms, whereas many norms (though not all) are conventions. For example, the above convention concerning telephone calls, according to Lewis, is normative because given that it is in place, one should abide by it, in the sense that it is advantageous to the person in question.

Rules in the natural world  99 However, we have already seen a lot of rules that have little to do with any coordination problems. Take the rule The bishop only moves diagonally. Not only does it not seem to help solve any pre-existing coordination problem; as it stands, it does not seem to be beneficial for anybody in any way. The reason is that this rule comes as part and parcel of a larger collection of rules, the rules of chess. It is only this whole collection that may be considered beneficial (though it is not easy to say what the benefit consists in). Also, the case of warning that we considered earlier in detail does not seem to have a lot to do with coordination (unless we consider “coordination” excessively broadly). There may be, to be sure, still other ways in which rules could have come into being. Take the rule that prohibits killing other people – it is not probable that its origin has to do with coordination (or, for that matter, prediction). It is certainly vital for me that I do not get killed, and being a pre-historic hunter or gatherer, one of the most important projects of my life must be avoiding this. However, it may turn out that a viable way to prevent myself from being killed is to take part in a collective endeavor of preventing everybody from killing anybody. Or take the rule that prohibits the stealing of private property. Its origins might lie in the power of some alpha males to usurp things that up until that time were up for grabs and then to beat up anybody who would touch these things. This privilege of the powerful may have then given rise to the rule establishing the institution of private property, making it into something that is observed publicly. Hence, the roads to rules may be diverse, but there are some general constants. Rules target what is done but not who does it to whom. Rules emerge out of normative attitudes having the tendency – for such or another reason – to align across communities; they are underpinned by the emerging normative attitudes to the extent that the rules can be identified with clusters of the attitudes. The emergence of the attitudes is driven by something useful for their partisans – in the most straightforward case something enhancing their fitness (e.g., the increased ability to better predict further events, or to decrease the danger of being killed). 7.4  Life under the eye of evaluation The sensitivity to rules is not restricted to children, it accompanies us throughout our whole lifespan, maturing into our specific way of inhabiting our world. The anthropologists Castro et al. (2010, p. 349) write: During ontogeny, the assessor communication between parents and offspring is extended by other evaluative interactions where the

100  Rules in the natural world approval or disapproval of behavior is provided by other unrelated individuals. Throughout their lifespan, a person establishes a social reference group with individuals that interact closely during a particular stage of life (parents, partner, friends, and colleagues). It appears to be an obvious fact that people assess the behavior of their conspecifics as right or wrong, and that this assessment plays a crucial role in human ontogeny. We enter our societies by being continuously curtailed by our elders, and we learn to inhabit social spaces, the “walls” of which are of nothing but curtailments. We have come to extricate our evaluations of actions from the egocentric perspective and have come to evaluate them independently of how they relate to the evaluator. Tomasello (2014) describes the situation in the following way (p. 120): [I]n their collaborative interactions modern humans conform to the collectively accepted ways of doing things, based on norms of cooperation, and in their communicative interactions they conform to the collectively accepted ways of using language and also linguistically formulated arguments, based on the group’s norms of reason. This leads to the situation in which many human doings acquire a characteristic shape, in that they are carried out with an eye on correctness. Kern and Moll (2017, p. 324) describe the transformation in the following way: “when a human walks or talks, her walking or talking is guided by an understanding of what it means to walk or talk, including an understanding of how it is done correctly”. The maturing of the infants of our species is guided by massive education (unlike that of the infants of other species, where purposeful education is an exception and the only learning is a matter of imitation – see Gergely & Csibra, 2006; Laland, 2017). During this education, the ways we behave are curtailed to fit into socially delimited channels – we learn the “correct” ways to behave, to solve problems, to deal with others. As a result, our mere doing of many things is transformed into doing them under the surveillance of correctness. The difference is like the one between merely kicking a ball around the playground and playing football. Though we can imagine that in both cases the people on the playground do similar things, the two cases are very different; this is because it is only in the second case, but not in the first, that things happen within a certain framework of rules (Peregrin, 2016b).

Rules in the natural world  101 Of course, the presence of the framework is not an esoteric business – it is present in that the participants have it, in some sense, “in their minds”. On the other hand, it is not enough for a person’s doings to be governed by rules that the person thinks they are governed by rules (remember that, as Wittgenstein, PI, §202, stressed and as we discussed it in the first chapter, “to think one is obeying a rule is not to obey a rule”). The reason is that the rule is a social institution: though there is a sense in which it is minddependent, it cannot depend on a single mind. It always exists, as it were, in the intersection of more minds. 7.5 Summary Do we have any empirical evidence that normative attitudes exist and are ubiquitous in human activities? It turns out that during recent decades normative attitudes and phenomena close to them have become the subject matter of numerous empirical studies, which not only confirm their existence but also tell us a lot about them. Many studies document a palpable presence of normative attitudes within human ontogeny. Infants are sensitive to the occurrence of rules in their environment, readily understanding that some regularities within the environment are more than just regularities, that they manifest rules, viz. something that ought to be followed. The situation, of course, is much more complicated with respect to phylogeny, which cannot be studied in such a direct manner. But even here there is evidence that rules played an important role. The emergence of normative attitudes aligning across communities is driven by something useful for their partisans – in the most straightforward case something enhancing their fitness. Rules may emerge from not only spontaneous solutions to coordination problems, but also as part and parcel of a larger collection of rules. The sensitivity to rules is not restricted to children, it accompanies us throughout our whole lifespan, maturing into our specific way of inhabiting our world. We enter our societies by being continuously curtailed by our elders, and we learn to inhabit the social spaces, the “walls” of which are of nothing but curtailments. Notes 1 Phillips and Knobe (2018) argue that in the first stages of cognitive development children tend to represent norms on a par with natural boundaries, simply as something the violation of which is “impossible”, and that even in the later stages of cognitive development a similar kind of representation stays prominent. 2 For other results regarding the role of rules within human ontogenesis, see Kenward et al. (2011) or Schmidt et al. (2016).

102  Rules in the natural world 3 Of course, this somewhat uncanny sensitivity to rules must be underlain by a cognitive (and consequently neurological) mechanism (it might be related to what some authors see as inherited cognitive specialization of “deontic reasoning” – see Cosmides & Tooby, 2005), but the mechanism is not as yet entirely clear. 4 Of course not everybody can wait for everybody else’s decision. The point, however, is that going with the majority is one of the factors that influences the decision process.

8

The natural history of correctness

8.1  Protagorean correctness In the previous chapter, we discussed the suggestion that the origins of language may be sought in the unification of sounds emitted by prelinguistic humans under the pressure of their peers who do it because emitting certain sounds in certain circumstances may be useful, for example as warnings. If there is, for instance, a sound which many members of a pre-linguistic tribe spontaneously emit in the presence of danger, it is understandable that it may be beneficial when the members of the tribe come to force each other to emit this very sound always and only in this situation. In this way, this and many other sounds may come to be associated with specific circumstances, extralinguistic or linguistic, and held in place in terms of incipient social rules. I argued that the most primitive form of such rules is constituted by normative attitudes resonating across the society. The normative attitudes are practical attitudes (behavioral syndromes) encouraging some behaviors and discouraging others, where the behavior is identified by its kind, not by who is its source or target. Let me repeat that, perhaps in contrast to Brandom, I hold that despite the name, there is yet nothing normative about such an individual attitude – it is normative only in the sense that it has the potential to take part in producing rules or norms when it comes to resonate with attitudes of other people. Anyway, the behavior encouraged by such resonating attitudes then can be seen as not only beneficial or advantageous for a particular person, but as beneficial in some “higher” or “supraindividual” sense having to do with the welfare of the whole community – so that it may start to invite the terms like right or correct. As this requires a certain kind of resonance across the society to form the intersubjective stance, which is largely independent of any one’s own attitude, we may want to consider Blackburn’s “democratic harmony” theory of normativity (see Section 1.1), according to which one is correct if and only if she is “in tune” with the others. DOI: 10.4324/9781003388876-9

104  The natural history of correctness To be sure, Blackburn rejects this theory as unacceptable, and I, of course, concur. It seems to be clear that many things can be correct even though the majority (or, for that matter, everybody) does not do them. However, though being clearly implausible in general, could it not work in some special cases? Consider the rule (D)   It is correct to say This is a dog only when pointing at a dog This, as a matter of fact, is correct1 – but why is this? Can we say that it is because the bulk of speakers of English do say This is a dog only when pointing at a dog? Could the most primitive form of normativity be a matter of simply doing what others do? The answer to this question is “no” – regularity alone does not give birth to normativity. No matter how regularly people will emit the sound This is a dog only when pointing at a dog, this will not make it correct. What is able to make it correct is that people hold this, practically, for correct: that they assume positive normative attitudes towards it. Hence, the resonance that is able to bear the (rudimentary form of) correctness is the resonance of normative attitudes – it is the attitudes that are the vehicle of any normativity. Normativity does not result, e.g., from the mere fact that people are uniformly reacting to danger; they must also assume normative attitudes to each other, supporting some kinds of the reactions and discouraging others. Thus, for there to be correctness, people must assume a peculiar kind of attitude to forms of behavior. This explains the most rudimentary form of correctness and also the emergence of an “intersubjective stance”. As incipient rules tend to transcend individual, selfish perspectives in favor of a social, intersubjective one, they require a resonance of the attitudes across the society to form the intersubjective attitude, which is largely independent of anyone’s own attitude. The thesis is hence that correctness is borne by the resonance of the normative attitudes: here is where the normative arises out of the non-normative. Things come to be correct not because somebody does them or wants them to be done but because they are generally accepted as worth being supported (plus/minus everybody displays positive normative attitudes towards them). Also, there comes to appear the distinction between what one does and what one (and everybody) should do. This intersubjectivity is nevertheless still a far cry from objectivity. While intersubjectivity involves independence from anybody, but not everybody, objectivity involves independence from everybody.

The natural history of correctness  105 The situation is even clearer when we concentrate on the rules of language, especially on the rules of correct assertability, which, as we will see later (Section 12.4), amounts to truth. As long as there is no objectivity beyond intersubjectivity, there is no truth beyond common opinion; hence, no truth worth the name. The history of the establishment of the concept of objectivity is thus the history of the establishment of the concept of truth.2 Returning to the case of a warning – why would people assume normative attitudes towards emitting the particular sound in the particular circumstance? Well, of course, precisely because it is useful as a warning: there is a customary regularity that is worth being reinforced by the attitudes. We want everybody to do this and we direct our normative attitudes so as to bring this about. In this particular case, we promote something normal to now be correct: at least in some cases, we tend to see what is normally done as something that should be done. To be sure, the regularity need not already exist to become correct – what is correct need not have been normal before. The only regularity that must be present is that of the normative attitudes, for there is a rule and hence a normativity only if the relevant normative attitudes resonate across the community. Hence, emitting This is a dog only when pointing at a dog is correct not because people regularly do so, but because people regularly assume positive normative attitudes towards it. Thus, I am convinced that there are some primitive (“bedrock”) cases in which this version of correctness holds, where it is impossible for correctness to be anything over and above endorsement by a (vast) majority. This holds for the most elementary ways of using words – what could their correctness consist in save the endorsement of the members of the relevant linguistic community? We have thus reached a view of correctness that I have dubbed Protagorean elsewhere (Peregrin, 2010a): something is correct because a great majority holds it for correct. A property is generally called Protagorean when something has it iff a vast majority of people take it to have it.3 I argue that the most primitive form of correctness is Protagorean in this sense: in these most primitive cases, something is correct iff it is held for correct by a vast majority. I certainly agree that no kind of correctness follows the “democratic harmony” model, but I claim that there are correctnesses, such as (D), that are Protagorean, which amounts to “democratic harmony” on a higher level, on the level of normative attitudes. (To explain why one can only emit This is a dog when pointing at a dog, it is enough to state that this is what the bulk of speakers of English holds to be correct.) And, moreover, I am convinced that any normativity ultimately derives from this primitive, Protagorean form.

106  The natural history of correctness 8.2  Non-Protagorean correctness Of course, not all cases of correctness are by any means Protagorean. Consider, for example, (D′)   It is correct to say This is a dog when pointing at Fido Here we know that this may not be correct even if everybody were to agree that it is (or the other way around). Perhaps Fido is an animal extraordinarily similar to a dog, despite not being a dog. We know that in some cases, it takes some research to reach an ultimate verdict, and we are prepared to change our prima facie attitudes if we learn the results of such a research. This, however, presupposes an additional dynamics of normative attitudes. Perhaps we assume some “prior”, prima facie normative attitudes, but these can later change into “posterior”, ultimate attitudes. The emergence of the mechanisms that make people go from the prior attitudes to the posterior ones marks an essential breakthrough in the grasping of normativity. Consider the sound that is emitted by members of a pre-historic tribe and that is bolstered by normative attitudes to a warning. At first, it might be that emitting the sound in a certain situation is correct if it is outright held as correct by the members of the tribe. But this may quickly change: the members of the tribe may start to endorse deferred, considered verdicts – be willing to change their prima facie attitude and to assume an attitude different from it. Why change the attitude? Well, of course, because something may be discovered to be worth supporting though it may prima facie seem not to be such, or vice versa. She who assumes the attitude may come to look closer and change it, but especially she may be convinced by her peers, via the incipient “game of giving and asking for reasons”. Maybe it seemed that there is nothing dangerous around, but then someone shows us a hidden tiger in a nearby bush. So we may see it so that there are prior attitudes that are completely erased by posterior ones, formed on the basis of some research or reflection. It is important to realize that the “posterior” attitudes are something like a “limit” notion. It is not so that after some time for research and argumentation there will come the modified attitudes which operate in a Protagorean way. The situation is rather such that what is decisive would be the normative attitudes that would result from an exhausting research and faultless argumentation, which is not something to be encountered in the real world. It is, then, obvious that any de facto posterior attitude may be mistaken – there is always the possibility that it will still be altered.

The natural history of correctness  107 Hence, it seems that not only the prior attitudes but also the posterior ones cease to have any practical significance; while the former ones merely start the way to the posterior ones, the latter ones become the carrot on the stick which we keep pursuing but never really achieve. It may seem therefore that we must abandon the thesis presented above, viz. that it is Protagorean correctness that is the cradle of all normativity – for this correctness appears to lose any important role now. But I think this need not be the case: the Protagorean attitudes can only shift from directly determining correctness to determining the criterion of the correctness. Take then the property of being a dog (which may be seen as expressed by the term dog). What happens is that the attitudes soon cease to determine which particular animal is a dog (something may be a dog even when the majority of speakers do not think it is) but determine the criterion of doghood (i.e. what does it take to be a dog, especially which observable features identify a dog?). More generally, if a property P is Protagorean, then to have P is simply to be taken to have P. If it is not, then to have P is something else, let us say to display a feature F. In a particular case, people may be mistaken about something displaying F and it may take some research to find this out; it may be that after the research is carried out, the opinion of the community changes so that the thing in question then has P just in case it is taken to have P. Thus, what is decisive are not the prior attitudes (before the research), but the posterior ones (after the research). But in order for the research to effect this, it is necessary that there is an agreement that to have P is to display F and that hence displaying F is the criterion of having P. In the simplest case, this is Protagorean. Let us return to calling an animal dog. The prior correctness results from the resonance of immediate normative attitudes: the utterance is in this sense correct iff a vast majority of speakers would immediately hold it for correct. It is correct in the posterior sense if a vast majority of speakers would hold it for correct after the animal is checked for doghood in an exhaustive way and if its doghood is irrefutably demonstrated by means of argumentation (i.e. something that can happen only in an ideal world). The relevant Protagorean attitudes determine how we can move from the prior to the posterior attitudes: by determining the relevant methods of research and argumentation, e.g., the (correct) criterion of doghood. Let us refine our terminology. Let us call a property that we have called Protagorean up to now zero-order Protagorean, because then we can define the notion of a first-order Protagorean property: a property is firstorder Protagorean iff it is not zero-order Protagorean, but the property of being the criterion of having the property (i.e. the way to find out whether it has the property or not) is zero-order Protagorean. Then we can define a property as being second-order Protagorean, etc. Thus, having Covid, for example, is not a zero-order Protagorean property: one does not have

108  The natural history of correctness Covid just because people take her to have it. But it is probably not even first-order Protagorean, as the criterion of having Covid is not necessarily what the general public may take it to be (weariness, loss of smell or what have you). What may be ultimately Protagorean is that to be the criterion is a matter of being proclaimed as such by people that have the right kind of schooling and appropriate expertise (or perhaps they only decide what is the criterion of being the criterion?). Hence, what is Protagorean, in this case, is being a criterion of a criterion. Thus, the above example illustrated how the correctness of calling an animal dog can move from a zero-order Protagorean property to a firstorder Protagorean one. Protagorean attitudes thus continue to be crucial, only the correctness they determine is not so interesting for its own sake, but rather because it determines a criterion of some more interesting correctness. It may determine what it takes to be a dog, which is not so interesting by itself, but rather because it determines, indirectly, which particular animals are dogs. From this viewpoint, it is an auxiliary correctness; it points out how we may move from the prior to the posterior correctness. If what is in question is the move from Fido is a dog to Fido is a wolf, then a correct way may be showing that Fido displays some features that are peculiar to wolves, rather than to dogs; an incorrect way may be claiming I feel that inside of Fido, there is the soul of a wolf! Thus, what counts are no longer the prior attitudes towards the utterance, but rather the normative attitudes to the ways of working on the establishment of the posterior attitudes. These may include the rules of the “game of giving and asking for reasons” or of corresponding research. (How it is correct to probe an animal for doghood?) In short, some zero-order Protagorean properties become first-order Protagorean. Consider correct assertability of sentences (which, as we already noted, amounts to their truth). It is quite clear that in nontrivial cases, it is not the majority voice that decides what is correctly assertable, there may be complicated justifications showing that given some basic rules, a given sentence is correctly assertable or not. The speakers may have some prior attitudes to the sentence, but they are ready to change them on the basis of ongoing interaction to posterior ones, which override the prior ones and become decisive. And it may be the case that nobody knows whether a sentence is correctly assertable – despite there being a fact of the matter regarding this. Anyway, the simplest cases of correctness are the Protagorean ones, and hence this kind of correctness came to realize the birth of the normative out of the non-normative. What is important is that once we recognize the difference between prior and posterior normative attitudes, and the fact that it is the latter (in their ideal limit form) that are decisive, we recognize that in the cases where this distinction is in play (viz. in the case

The natural history of correctness  109 of non-Protagorean correctness), our attitudes (or at least many of them) come to be fallible. Especially in the case of truth, we recognize that we all can be mistaken, for we can never be sure that there is no potential way to show us that what we claim is not the case. 8.3  Negotiating correctness Let us now look at the same process from a different angle. We humans have come to assume the normative attitudes. What we should do is not merely what others do, but rather what the normative attitudes of others point at. (This may be, in some cases, just what the others do, but this is far from universal.) The emergence of rules presupposes the resonance of normative attitudes across the community in question. Yet it is hardly imaginable that the resonance would be exceptionless – there will always likely be some dissidents, or even some clashes between significant parts of the community. How will the community treat them? I think that here there are two possibilities. Some people’s deviant attitudes may be such that they amount to the ignorance or rejection of a rule. Someone would simply ignore the rule that one should say This is a dog only when pointing at a dog or that one should not kill other people. (He will keep saying This is a dog when pointing to things other than dogs and will not be bothered when other people do the same, or he will kill other people and approve of others doing so.) Such a person is likely to be excluded from the relevant game (which may be meaningful English communication or the very life of the community). But there may be some more subtle deviations. Perhaps a person’s attitudes towards the usage of This is a dog or towards killing do generally resonate with the majority, but in some particular cases, they come to dissent. What happens then? Of course, a possible resolution of the situation is that the dissident is cut down to size: she becomes the target of negative normative attitudes, which make her give up her dissenting stance (or leave the game). In such a case, the dissenter is taken to be mistaken. However, surprisingly, it turns out that this is not what always happens. It may sometimes happen that the dissident makes the majority change their attitudes – to recognize they were mistaken. This may happen if the establishment of the correctness moves beyond simple resonance towards some more complicated methods of determining agreement – towards correctness that is not a matter of initial resonance, but rather of some less trivial process of reaching agreement. Thus, doing something comes to be correct not if it directly elicits positive normative attitudes, but if the utterer is able, as a matter of principle, to make her peers elicit the positive attitudes (even when originally their attitudes were negative). For example, an utterance of This is a dog, in a particular

110  The natural history of correctness circumstance, may be correct not if it incites positive reactions (right off), but if the reactions can be made positive (for example by showing that the animal possesses some characteristic features that distinguish it from a wolf). Of course, in such a situation, the dissident cannot do whatever to convince the opponents. She may do a research that is deemed correct in the present situation, and she may try to change the attitudes of the majority only in an acceptable way. In this way, the target of the normative attitudes shifts from the final verdict to the method of reaching the verdict. It is no longer so that the correctness of the utterance of This is a dog is irreversibly correct iff the majority’s attitudes to it are positive, the attitudes are taken as pro tanto, they may be reversed by presenting the results of a research carried out by correct methods. As a result, something is correct not iff the majority of normative attitudes to it are positive, but iff there is an application of correct methods capable of making the majority assume the positive attitudes. Hence, something may be correct despite the majority’s negative attitudes, viz. despite the majority’s conviction that it is not correct. Here we therefore need rudimentary forms of persuasion, negotiation or argumentation; eventually, a rudimentary form of what Brandom calls the “game of giving and asking for reasons” (GOGAR). Hence, there must emerge the rudiments of the rules of such a process, one that makes others change their normative attitudes. These will include not only general rules of the game but also rules specific for the case of correctness in question. Thus, my effort to persuade others that my utterance of This is a dog was correct will be governed not only by general rules of persuasion and argumentation but also by some specific rules concerning properly showing that something is a dog. So here there emerges a novel mechanism for determining what is correct. Note that originally we considered a single mechanism – the approval of the majority (which rendered every correctness that which we called zero-order Protagorean). To say This is a dog when pointing at a dog was correct when the normative attitudes of the majority of speakers of English approved of it, and not to kill anybody is correct because this is approved of by the majority of the relevant community. (Note that we can easily imagine alternatives: there may be a community in which it is correct to utter This is a dog when pointing at a cat; and we can imagine one in which killing is, under certain conditions, admirable.) But now we come to have properties that are not zero-order, but rather first-order (or possibly second-order …) Protagorean. Hence, we may have different criteria: saying This is a dog when pointing at a particular animal in a particular situation may be correct even if the majority were to disapprove of it. It is correct when the entity pointed at is really a dog, which may be determined by

The natural history of correctness  111 empirical research and/or argumentation. The majority of the community may be mistaken and may be led to recognize its mistake. Protagorean attitudes still form the “bedrock” – only now they do not concern the final verdict, but rather the ways to reach it. 8.4  From resonance to agreement Let us summarize the route which leads from the (untenable) view of correctness as mere resonance with others to the rich and multifaceted notion of correctness we have today. (Let me stress that this is not a hypothetical rendering of what really happened in the pre-history of mankind, it is an attempt to illustrate the steps that lead from what we conjectured must have been the birth of normativity and the current role of rules within human conduct.) We have stated that the most straightforward way from facts to norms that may come to mind is the “democratic harmony” theory: what one does is correct when it is “in tune” with what the majority does. This would yield us a principle that could be called agreement in doing: Doing something is correct if it resonates with what others do. After we concluded that correctness is yielded by normative attitudes, we have to improve this as concerning the attitudes: what one does is correct when it is a target of positive normative attitudes (where the attitudes are normative only when they are “in tune” with the majority). Hence, we have a principle that could be called agreement in assessing of doing: Doing something is correct if it is approved by others – if it is assessed as correct by resonating normative attitudes But even so, we must conclude that this cannot underlie our rich notions of correctness and normativity: there are certain things that are correct despite the majority not thinking they are correct (not assuming positive normative attitudes towards them). Now we finally see what must be still added to reach a viable concept of normativity as issuing from social facts: it is not always “being in tune” with the others regarding directly the normative attitudes determining correctness, but sometimes regarding the methods of how to reach the resonance of the normative attitudes. So, finally, we have a principle called agreement about methods of assessing of doing: Doing something is correct if others can be made to approve of it – if it is rendered correct by methods assessed as correct by resonating normative attitudes

112  The natural history of correctness We see here then how correctness, which could not have started otherwise than as “being in tune” with the majority (though what must be “in tune” are the normative attitudes), could have broken loose from this and gone on more complicated routes. The thing is that some correctnesses are not direct projections of normative attitudes; they result from the application of some correct method of determination. A property is Protagorean (for a community) iff it is possessed by an entity only in the case that the majority of the community holds it to possess it. For example, a sound possesses the property of meaning dog iff the majority of the speakers of the relevant language holds it for such. But an entity is a dog not iff the majority of members of a community holds it for such – here the majority may be mistaken. Hence, an animal does not possess the property of being a dog iff the majority of the members of the speakers of the relevant community holds it for such. It is a dog when the correct methods of checking its doghood can lead to a result that is convincing for the majority. The agreement of the majority still plays a key role, but it is no longer an agreement with respect to what one does: the agreement concerns, much more specifically, normative attitudes. Moreover, what is substantial are not normative attitudes that are immediate, it may be normative attitudes resulting from a certain process that is endorsed by (immediate) normative attitudes. Hence, in the end, the fundamental agreement of normative attitudes concerns not what is correct, but rather how to find out what is correct – not the property itself, but rather its criterion. We agree that something is a dog iff it displays such and such features. This is the vote of the majority. But whether some particular animal is a dog is no longer a matter of the vote, it is a matter of the animal displaying the decisive features. Again, if there is a quarrel with respect to whether the features are really present that cannot be resolved by negotiation or argumentation, then the ultimate decision may again be the “vote” of the majority, but this would be an exceptional situation. 8.5 Summary What is it that makes something correct? In previous chapters, we have given an unambiguous answer: it is the normative attitudes. In the simplest case, we have the “Protagorean” form of correctness, viz. doing something is correct if it is approved by others (if it is assessed as correct by resonating normative attitudes). However, this gives rise to more complex forms of correctness, by allowing for a negotiation of normative attitudes, i.e. their informed change by methods that again are liable to normative attitudes. In some cases, then, doing something is correct if others can be

The natural history of correctness  113 made to approve of it (if it is rendered correct by methods assessed as correct by resonating normative attitudes). This is important, because it introduces the possibility of everybody being incorrect. Notes 1 To avoid misunderstanding, let me repeat that this is a kind of correctness, possibly one among many other kinds. Developed societies will be replete with such kinds, possibly competing with each other and trumping each other (see Section 5.3). We can say therefore that (D) holds from the viewpoint of English, while it can be easily trumped by another kind of correctness: you may be an actor in a theater play that prescribes that you violate this rule of English. 2 Price (2011) argues that the role of truth in human communication consists in the determination that a conflict in opinions should not remain unresolved. This, needless to say, is an important step on the road to objectivity: reality is only one way; it does not tolerate conflicting facts. 3 Thus, the Protagorean properties are those “the measure” of which is we humans alone.

9

Systems of rules and institutions

9.1  Integrative rules The rationale behind some rules is obvious: it is clear why a rule such as You should always wash your hands before eating turned out to be useful and promoted by evolution. However, we have already seen that, by far, not all rules are like this. The rationale behind rules like The bishop moves only diagonally or You should receive Communion under one kind but not under both – and indeed You may say “This is a dog” when pointing at a dog – is far from obvious. And the reason is that such rules do not work alone, but rather in interdependence with many other rules. I call these kinds of rules integrative: they are not self-standing but are integrated with a greater system of rules. It is rules of this kind that are especially characteristic of our species. There is a lot of discussion about whether animals other than us humans can follow rules; there seems to be strong evidence that, for example, primates or wolves do respect certain rules (or proto-rules) regarding pack hierarchy, etc. (Fitzpatrick, 2020; Andrews, 2020). This may be partly a terminological problem, but be that as it may, I believe that what, in any case, is an exclusively human matter are the integrative rules. As Okrent (2017, p. 124) puts it: [W]hile baboons live in communities that are differentiated according to social roles and social positions, and a single baboon can occupy different positions synchronically and diachronically, all baboons live in communities with the same kind of system of social positions, defined by the same norms for the various positions, and the same rules for social position occupancy. We humans, on the other hand, live in communities with a vast variety of systems of kinds of social positions, each with different norms for occupants of those different kinds of social positions, and different rules for the position occupancy of those different kinds of social positions. DOI: 10.4324/9781003388876-10

Systems of rules and institutions  115 While self-standing rules can often be imagined as being unproblematically carried along by normative attitudes, the more tightly they become integrated into a system, the more complicated it is to sustain them – to sustain, that is, the whole system – without the help of language; some of the systems then simply cannot exist otherwise than in terms of an explicit and complex linguistic infrastructure. How could systems of integrative rules, such as those behind chess, football or language, come into being? A rule in general is “holistic” in the sense that it is in force in a society only if it is accepted as such by a significant number of the members of the society. Hence, it usually comes into being by means of bootstrapping (at least before the establishment of language, which makes it possible to establish a rule by means of a stipulation, which, however, itself rests on already accepted rules). An integrative rule displays an additional kind of “holism”: it is in force in a society only if a number of other rules are accepted by the society. So another kind of bootstrapping may be needed. It is tempting to see systems of rules as inventions, as something that came into being by somebody clever formulating them (or some of their ur-versions) so that no bootstrapping was needed. This may be plausible in the case of chess (who knows?), but it is not so plausible in the case of football, not to mention language. It seems much more plausible to me that football evolved, through the emergence of implicit rules (those who did something that it was felt obstructed the game would be pestered about it), from rule-less frolicking to the form governed by fully explicit rules that we know today. And to see language as such invented systems of rules is not only implausible, but simply impossible. In any case, an integrative rule forms a whole together with the other rules with which they form the system; if there is an evolutionary rationale behind this, then it is related to the whole system. It is, of course, very difficult to specify such a rationale, and I do not pretend that I can do it entirely clearly. What I can offer is a closer examination of the systems of integrative rules in which I point out some of the crucial features that can easily escape attention, especially to point out the achievements to which they led and which may have to do with their rationale. Why do we implement restrictions imposed by rules? Does it make sense to rid ourselves of some of the possibilities of behavior we have? We have already seen that the point of some self-standing rules is to lead us to the most efficient ways of behavior and rid us of those that lead into blind alleys. And this is certainly something understandable. But the point of the integrative rules is often different. We must see that just like real walls, rules not only limit us, but if they are put together in a proper way, they are also able to delimit new spaces for us to inhabit.

116  Systems of rules and institutions We have already seen that the working of rules can be sometimes portrayed as opening up an “inside space” which we can “enter”. And it is especially quite complex systems of rules that work in this way. That the systems may resemble virtual edifices is based precisely on the ability of its component rules to act as walls, in that they restrict our behavior to block, as it were, our movements in certain directions. And if such walls are combined in an appropriate way, what we get may be a dwelling. And inside the dwelling, we may become able to do things that cannot be done outside of it. And rules, indeed, can be like walls. Imagine you are in a supermarket and have no money. Perhaps you are hungry, and there is a plenty of food around. But you feel you simply cannot take it – it is impossible, because it contradicts the rules you accept. The food is separated from you by a virtual wall of the rules just as effectively as it would be separated by a real, solid wall. (True, you may break the rules and steal the food – but it is also true that in the case of a solid wall, there may be ways to get over it.) Consider playing chess. What does this amount to, how do I start playing chess? I must accept the rules of the game (as governing the specific chessboard and the pieces that are in front of me). By doing this, I enter the “space of chess”, whereby the pieces become pawns, bishops, knights, etc., and what I do with them has to be within the boundaries of the space, the space in which chess games develop. However, the delimitation of this space does not merely restrict me because, within the space it constitutes, I acquire the potential to produce brand new kinds of actions. It is only within the framework of the rules of chess that I am able to check an opponent’s king, that I have the freedom to plot how to checkmate my opponent, and that I can organize defenses against her attacks. The most important space is the space of our language, the “space of meaningfulness”. Just as in the space of chess, pieces of wood become chess pieces that can form an “army” with the potential to fight an opponent’s “army”, so in the space of language, certain kinds of sounds become specific kinds of reports, orders or questions, with the potential of carrying out a fruitful linguistic intercourse. They become vehicles for playing our “language games”, which transform us into the kind of “discursive creatures” that we are. We have already mentioned Sellars’ (1949, p. 315) distinction between a rule that “lives in behavior” and its mere “skeleton”, which we get when we “attempt to grasp it from without”, when we “describe” it (Section 4.3). Let us now quote the passage in full: We saw that a rule, properly speaking, isn’t a rule unless it lives in behavior, rule-regulated behavior, even rule-violating behavior. Linguistically we always operate within a framework of living rules.

Systems of rules and institutions  117 To talk about rules is to move outside the talked-about rules into another framework of living rules. (The snake which sheds one skin lives within another.) In attempting to grasp rules as rules from without, we are trying to have our cake and eat it. To describe rules is to describe the skeletons of rules. A rule is lived, not described. Thus, what we justify is never a rule, but behavior and dispositions to behave. The “ought” eludes us and we are left with “is”. The metaphor of being within/without a framework of rules underscores the distinction between a rule that “lives in behavior” and one that is a mere “skeleton”: a rule “lives” if we view it from inside; once we move outside, its life fades away and what remains is the skeleton. It is only from within that a rule yields a full-blooded “ought” (The bishop ought to be moved only diagonally), from without the “ought” is wrapped into an “is”: It is a rule of chess that the bishop ought to be moved only diagonally. There is one more important message in Sellars’ paragraph quoted above. We can step out of some systems of rules, perhaps out of any system of rules, but only “into another framework of living rules”. Thus, we seem to be unable to be out of every system of rules at once. Why is that? The reason, obviously, is that we humans cannot exist outside of all systems of rules – at least not without ceasing to be the kinds of creatures we are, quitting our specifically human “form of life”. Systems of rules constitute our natural niche, which we need to be able to live the kinds of lives we do. This is just another articulation of the fact that we are normative creatures. Thus, we can be outside of any system of rules, but not of every system of rules (at once). Sellars also indicates a certain ineffability of the ingredient that makes a rule into a living one. Once we grasp a rule, we possess only its skeleton. Following a rule is not necessarily grasping it and then acting accordingly; it is a kind of action that is sui generis. We saw that being inside does not mean knowing something more than one can know from outside, it is not a matter of surplus knowledge, but rather of assuming a different attitude. 9.2 Institutions The upshot of the previous section is that we humans have established multifarious normative spaces in which we can perform unprecedented kinds of actions, similar to the kinds of actions we become prone to perform when submitting to the rules of a game. They are, however, not mere moves in a game, but often socially much more “serious” actions. At this

118  Systems of rules and institutions point, it is helpful to realize that what we call institutions are usually such systems of rules. Kumar and Campbell (2022, p. 138) write: “The core of any given institution is a set of interlocking norms—that is, a set of shared and mutually reinforcing social rules about what actions are obligatory or forbidden”. And similarly Bowles (2009, p. 47): “Institutions are the laws, informal rules, and conventions that give a durable structure of social interactions among the members of a population”.1 A little clarification may be in place here. Take a prototypical institution like a university. Is it really merely a system of rules? What about the buildings, the students, the teachers, etc.? It is quite clear that an institution cannot exist without an “embodiment”, without certain real-world objects and individuals. But this is because the rules institute a system of roles that need to be instantiated. A professor, for example, is part of the university not qua a concrete person, but qua the professor, which is a role consisting of certain – possibly dynamically changing – bundles of commitments and entitlements (see the next section). The institution as such rises and falls with the rules constituting the roles, commitments and entitlements, not with the persons and things which bear them. In this spirit, Turner (1997, p. 6) claims that an institution is “a complex of positions, roles, norms and values lodged in particular types of social structures and organizing relatively stable patterns of human activity with respect to fundamental problems in producing life-sustaining resources, in reproducing individuals, and in sustaining viable societal structures within a given environment”. Hence, institutions are systems of rules (in the broad sense of “rule” which we entertain here) establishing a structure of positions and roles; I think that we do nothing terminologically too devious when we say that any systems of integrative rules can be generally called an institution. True, we usually do not apply this term to games such as chess, which we use as paradigmatic examples of a system of rules (though we certainly do call the International Chess Federation an institution), but we do not seem to do any harm if we stretch the use in this direction. But do we want to claim that language is also an institution rather than a psychological reality of the Chomskyans or a “biological category” of Millikan (1984), etc.? Is this not an over-intellectualized account of language? Not if we keep in mind that a rule, and hence an institution, is not necessarily the result of an explicit agreement – that it may emerge out of what we do, without any explicit intention, once we become the “creatures of rules”. And of course that language has to do with our psychology and our biology – but just as it would be futile to try to understand football by studying the psychology and biology of its practitioners, I hold it as being misguided to try to understand language in this way.

Systems of rules and institutions  119 The important point is that institutions are ubiquitous. They have pervaded our natural world and changed it to the extent that allowed us to assume a brand new form of life. As Packer and Cole (2019, p. 175) put it: We argue that humans construct an environmental niche that is unique in being composed of institutions, which function to coordinate activity over multiple time scales. Institutions involve not simply customs or conventions but a deontology of future-binding rights, responsibilities, duties, and obligations. 9.3  Roles, commitments and entitlements The metaphor of inhabitable spaces built out of rules extends to the point that the same space can be inhabited in different ways by different people. In the world of football, you can be a goalkeeper, a defender or a forward; and you can be also a referee or a newspaper reporter. This is made possible by the fact that spaces created by complex systems of rules provide for multiplicities of different roles which can be assumed by different participants. But to explain this in greater detail, we must first extend our anatomization of the concept of the role. Rules may involve objects of the real world. We may have a rule If, and only if, the Sun shines, you may do A or If, and only if, the Moon shines, you may do B. The objects involved thereby obtain certain roles with respect to the rules. (The role obtained by the Sun vis-à-vis the abovementioned rule can be called A-enabler, that by the Moon B-enabler.) It is, however, more common that the objects involved with rules are not quite specific, that they are any kinds of objects complying with certain given standards. Take the ball involved in the rules of football. It can be any kind of object which complies with the requirements specified by the rules: it must have certain dimensions, be inflated to a certain value, etc. Let me call the objects required by rules in this way the equipment. An extreme kind of equipment are chess pieces, as there are almost no requirements placed on them. The knight does not have to look like a knight (and in fact it usually does not), it does not have to be made of any specific material and have any specific size. Of course, we presuppose that the pieces are such that they can be easily moved by a single hand, but even this is not a strict requirement: it is conceivable that players play chess using cranes moving pieces heavier than a human could handle. It is even possible to play chess without any physical pieces; hence, we can see chess pieces as pure roles, which are assumed by physical objects not out of necessity, but only for convenience. Anyway, a system of rules may specify a number of roles which can be assumed with various kinds of real objects (pawns, rooks, bishops, etc.,

120  Systems of rules and institutions of chess), possibly with some kinds of prescribed properties. Also, people entering the system – let me call them participants – may assume various kinds of roles. Again, some requirements may be connected with some roles (next to anybody can become a goalkeeper in a yard game, but to be a goalkeeper of a professional team you need much more). Sellars (1963a, p. 637) invites us to consider a game called “Tess”, which happens to be “isomorphic” with chess: Thus, if Texans had independently developed a game played with automobiles and counties called “Tess” with its own distinctive vocabulary for its pieces and moves, we might have come first to appreciate isolated similarities between Tess and chess, and then to see that they could (along the above lines) be regarded as different ways of playing a game which chess would be another way of playing. At this stage, instead of coining a new vocabulary for the “same game” we would probably raise the criteria for being a “pawn,” a “king,” a “board,” and consequently for being a game of chess, to a higher degree of abstraction, and begin to contrast “Texas chess” with “conventional chess” as (materially) different varieties of chess. Before this move, we could speak of two similar games, and, even, of two games “so similar that they could be regarded as different ways of playing one and the same game.” Only after this step could we speak, without qualification, of two forms of the same game. Thus, he observes that a system of rules involving nontrivial equipment may institute an abstract structure that may come to be instantiated by different kinds of equipment and/or participants plus relationships among them. Consequently, it may give birth to a new, abstract reality. When the abstract roles, which may be instantiated by various concrete items, come to be seen as objects (when we see, as Shapiro, 1997, puts it, “places of objects”), the roles may appear as kinds of object alongside the items instantiating them and linked to them. According to Sellars, this usually happens with our language, where we do not see meanings as roles of linguistic items and see them as things associated with expressions. It is also important to realize that the system of roles brought about by a system of rules is not just static; the rules may provide for its change or development. Take chess: at the beginning of the game, both the kings are not just kings, but kings with the entitlement to castle. This entitlement exists until the point the king moves. Or consider the football player: if she gets a yellow card, her status changes – her next yellow card changes into a red one and she is out of the match. These examples may look rather marginal to the games, but many practices are based on the constant change of the statuses of the equipment and especially the participants.

Systems of rules and institutions  121 Take, for example, a university. There are a number of roles – students, professors, deans, etc. – which are assumed by various people complying with corresponding requirements. Assuming a role in a practice involves accepting the rules constitutive of the practice. In other words, it involves certain obligations which become the commitments of the player. The other side of the same kind are permissions of the players, which become her entitlements. Moreover, the commitments and entitlements are highly dynamic: they change in the course of the practice. Thus, a student starting at a university commits herself to passing a specific collection of exams and will become entitled to various things, such as attending the lectures or working in the library. In the course of time, as she passes the exams, she may get clear herself of some of the commitments but will gain various new commitments and entitlements, such as the entitlement to enroll in certain new courses (especially those which require some of the exams she has passed as prerequisites) and the commitment to pass the corresponding exams. For Brandom, the concepts of commitment and entitlement appear to be the foundation of his normative theory of language and society. I do not find these concepts primitive, I see them as constituted in terms of more primitive concepts, especially the concept of normative attitude. It is normative attitudes that constitute (implicit) rules, it is systems of rules that constitute systems of roles; and it is the existence of the system of roles that opens up the possibility of committing oneself and becoming entitled. It is, admittedly, not easy to pin down the evolutionary rationale of our ability and tendency to build the normative spaces constructed by means of integrative rules. But this ability grants us something that is likely to be connected with the evolutionary rationale and that completely reorganizes the way we humans live our life as a social species. Let me return to Sellars’ distinction between the manifest image and the scientific image (Section 4.3). The difference, as Sellars puts it, is that the former has us as persons (while the latter has us only as organisms). What is the difference between a person and an organism? We may say that a person is an organism which is woven in a web of normative relationships, capable of accepting commitments and entitlements. A person usually has a couple of social roles, so maybe we can say that a person in general is an organism capable of assuming social roles (and commitments and entitlements). In general, a person, unlike an organism, is somebody who is responsible for what she does; hence, her behavior is more than that, it is acting. The thing is that responsibility (and persons, and actions …) are brought about by our normative attitudes, by our holding the behavior of each other for correct or incorrect. If you ought to do something, you are obliged to do it and you are responsible for doing it. If it is not the case

122  Systems of rules and institutions that you ought not to do something, you are permitted to do it. And it is thus that our behavior mutates into actions, and we mutate into persons. This set up lets us establish a kind of cooperation that is unprecedented in the animal realm – cooperation based on normative relationships and on responsibility. The discussion about the origins of cooperation in the animal realm, which has been taking place for a couple of decades, reveals that the core of the problem consists of breaking the barrier of what, in game theory, acquired the shape of the Prisoner’s Dilemma (Poundstone, 1992; McElreath & Boyd, 2008; see also Section 14.2). The point, as we already saw, is that true cooperation requires an investment – which, in the optimal case, returns itself with surplus interests. The interests result from the fact that the value of goods is context-dependent: its value for me depends on how I need it in the moment in question. When I am hungry, a piece of meat is much more valuable for me than when I am satiated. (And in the extreme case that I am dying from hunger, its value can be immense). Thus, when I am not hungry and have some meat, it is not a great loss for me to give it to my peer who is hungry and for whom it thus is a great gain. Cooperation therefore lets me transmute a small loss into a big gain, and in the optimal case a small loss for me into a great gain for me – this is if my peer reciprocates and gives me some meat when I am hungry (and she is not). However, there is always the temptation on the part of the one who is to reciprocate to turn the investor down and refrain from reciprocating. If you share your meal with me when I am hungry, you expect that I will do the same in the future, but if I do not do it, I may profit. And it seems that the Prisoner’s Dilemma shows that evolution must support this selfish behavior. It is clear that as we (and lot of other animal species) obviously do cooperate, the barrier has been broken. And it seems that in our human case, it is our ability to commit ourselves and to accept the responsibility that has been part of the solution. Of course, the mere possibility of committing oneself solves nothing – any commitment can be broken. But the whole organization of community based on commitments and entitlements makes it difficult to be a permanent free-rider. Maybe we cannot say with certainty how this normative scaffolding of our societies came into being; we can however say a lot of about how it works now, because we see it in front of our eyes. The actual normative statuses – commitments and entitlements – of individual people have become, in our eyes, the features that characterize persons in substantial ways, and a great deal of what we observe as social movement consists in changes in these statuses. Our orientation within our social landscape consists, to a large part, in what Brandom called, borrowing from Lewis (1979), deontic scorekeeping.

Systems of rules and institutions  123 9.4  Building a shared future We have already seen that the fact that normativity rests on the attitudes of individual people allows us to use it for the purposes of what I call “building a shared future”. The point is that I evince attitudes that have to do with the way the world should be in the future, and others are welcome to join me, to evince similar attitudes, which, if they become the “collective will”, direct our common efforts (as a “regulative ideal”) to bring the desired future into being. Within the studies of evolution, increasing attention is being paid to the concept of “niche construction” (Odling-Smee et al., 2003). We will deal with this in the next chapter; here we point out that for us humans, niche construction is something in which we not only excel, but which we have appropriated to make it into the most determining feature of our evolution.2 We have acquired the ability to rebuild our environment in an unprecedented way. But we not only change our physical environment (unfortunately in such a way that we have become a menace for the whole planet), we have also introduced various “virtual” extensions and complements for it. We have filled up our environment with specific sounds that we emit and that provide for specific ways of our coexistence, coordination and cooperation; we have especially intertwined it with rules that not only govern the employment of the sounds but also govern a lot of other things we do and employ. The fact that our environment is so replete with rules makes me think that the appropriate name for it would be a “normative niche” (Peregrin, 2020a). The novices of our societies come into an environment in which they not only put to work their genetic predispositions or learn how to behave by trial and error but also readily directed to certain ways of behavior. They are actively taught how to convey their behavior to certain “standard” channels and thereby to enter their human community, assume their roles and act as responsible persons.3 When we return to niche construction, it is not surprising that we try to eliminate or diminish dangers and hindrances and we try to support and cultivate such aspects of the environment that are helpful. However, not all features of the environment are equally yielding to our interventions. We can move a stone or a tree trunk and it stays where it is. We can mow grass, but it will not stay mown; we have to mow it again after some time. And the most complicated issue is to influence the part of the environment that is constituted by our human peers: here we must wield continuous pressure to keep it in the shape we think it should be in. And the kind of pressure we have developed, in the form of normative attitudes, is a unique achievement of our species.

124  Systems of rules and institutions Suppose a bunch of prehistoric humans travelling along a path are impeded by a huge tree trunk blocking their way. If they do not possess developed means of communication, it might be reasonable for a member of the tribe to go and start lifting the trunk, even if he alone is obviously unable to move it. This action may be interpreted as an invitation or an appeal and other members of the tribe may soon join the pioneer and help him remove the obstruction. (Of course, if the tribe does possess a developed language, it would be better to make use of an explicit appeal: Let us move this trunk out of our way!) This particular situation concerns a transient problem and only minimally involves any shared future. A shared future amounts to building virtual walls to delimit our future (normative) dwellings. Imagine, once more, a university: we have already seen that it usually employs some tangible buildings, but the buildings alone are insufficient to constitute the university (and in an extreme case, they need not be even necessary). It is, in fact, much more a matter of the virtual spaces that are built out of rules, built from all the things a student or a teacher of the university should and should not do. Just as in the case of the hindering tree trunk, we can invite others to participate in building a virtual, normative “wall” by simply starting to assume the corresponding normative attitudes; others may then concur (or not). The difference from the previous case is that now the action is not transient, but persistent, and aims to administer an ongoing normative pressure. The situation, of course, accelerates dramatically if a rich language is up and running in the community. (Remember, however, that language itself is already a matter of a normative space – therefore, it cannot help us open the normative spaces from the beginning.) If this is the case, then I can not only assume the normative attitudes moving others to do X and preventing them from doing Y: I can say that X should be done (or that it is correct), while Y should not be done (or it is incorrect). In addition to perhaps glowering at someone who does not bow to the chief, I can also say things such as One should bow to the chief or It is correct to bow to the chief. But we have already seen that such pronouncements, which I proposed to call normatives, can be ambiguous. In one sense, they can be read as stating facts concerning the set-up of the society in question: in this case, its members tend to assume normative attitudes towards bowing to the chief, in particular they approve of the bowing. However, if a speaker says such things as One should bow to the chief or It is correct to bow to the chief, it is also possible, and usual, to interpret the speaker as not (only) stating such facts, but assuming the corresponding attitude. Thus, one is interpreted as not only stating a fact in a disengaged way, but as approving of bowing to the chief.

Systems of rules and institutions  125 Normatives, then, are successors to “implicit” normative attitudes, making them explicit. At the same time, these acts are construed as close enough to declarative utterances to be truth evaluable (though their status as successors to normative attitudes might suggest that they should be assimilated rather to a kind of imperative, prescribing people what to do and what not to do). The reason, I think, is that these speech acts are apt tools for building a common future. The thing is that the normative worlds in which we humans mostly dwell presuppose a robust resonance of our normative attitudes, perhaps, as we have already seen, some “we-intentions” or “collective intentions”. Such a resonance can be achieved by long-term calibration (which is what happens with pre-linguistic normative attitudes), or in a much swifter way using the normatives. It is as if the normative describes a desired state, for which the utterer votes and other members of the society may concur – or, as the case may be, refuse to concur. If they do concur, the “wishful thinking” behind the normative is turned into reality (Peregrin, 2016a). 9.5 Summary Some rules, integrative rules, do not work alone, but rather in interdependence with other rules; they are not self-standing but are integrated with a greater system of rules. It is rules of this kind that are especially characteristic of our species. The working of integrative rules can be portrayed as opening up an “inside space” which we can “enter”. Systems of rules, sometimes also known as institutions, establish structures of positions and roles, which can be assumed by various kinds of real objects and concrete people. The people gain and lose various commitments and entitlements, often in a very dynamic way. It is thus that our behavior mutates into actions, and we mutate into persons. This setup lets us establish a kind of sociality that is unprecedented in the animal realm – sociality based on sharing normative spaces and on responsibility, which is inseparable from normativity. The fact that normativity rests on the attitudes of individual people allows us to use it for the purposes of “building a shared future”. I evince attitudes that have to do with the way the world should be in the future, and others are welcome to join me, to evince similar attitudes, which, if they become the “collective will”, direct our common efforts (as a “regulative ideal”) to bring the desired future into being. Notes 1 For a general discussion of the relationship between conventions and institutions, see Zachník (2020). 2 Applying this concept to human evolution has produced some interesting theories (Kendal, 2011; Laland & O’Brien, 2011).

126  Systems of rules and institutions 3 Packer and Cole (2019, p. 203), who use the term “deontic niche” instead of our “normative niche”, write: “[R]ecognizing the deontological character of human culture and of the human environmental niche allows re-location of the sources of the process that dual-inheritance theorists attribute to a genetically evolved psychological ability and motivation to learn from others and to teach others. Capabilities that seem to be within the individual are in fact distributed through the community and the environment. Capacities that theories of cultural learning view as inherited and biological are in fact located in the community’s persisting deontic niche”.

10 Behavioral patterns

10.1  The very idea of evolution Organisms display various behavioral patterns. Some birds from Europe fly to Africa for winter. We continental Europeans tend to drive our cars on the right side of the roads. Spiders make their webs. And so on. How come? The wisdom of the recent decades is that everything in the biological world is explained by evolution, and behavioral patterns are no exception. So a short answer may be that the patterns are here because they have been selected by natural selection. Let us summarize the most basic facts regarding evolution. In the simplest case, evolution is a matter of some entities called replicators, which produce their copies.

Figure 10.1 Replication

To survive, the replicators need some resources, which are in short supply; hence, the replicators must compete for them and only some of them survive (and produce further copies). The ability to survive and produce copies is called fitness. Hence, those kinds of replicators that do produce more copies are – statistically speaking – more fit than those which do not. This process is called selection – it is the sieve which lets the more fit through while discharging the less fit.

DOI: 10.4324/9781003388876-11

128  Behavioral patterns

Figure 10.2 Selection

In the above picture, the three kinds of replicators display about the same level of fitness. But the situation may be different. Suppose, for example, that the kind ⊗ is slightly more fit than ∅ and the latter is in turn slightly more fit than ⊕. Selection may then work in the following way:

Figure 10.3  Selection over two generations

Behavioral patterns  129 In this way, the fittest replicators tend to survive and do survive, in the long run, eliminating those that are not so fit. Replicators may also undergo random mutations, which may make them more or less fit for the struggle for resources (and which are permanent in the sense that they are present in all the copies of the mutant). Given this, the replicators, obviously, evolve in the direction of greater fitness. The point is that if there is a way to become more fit, it will occur (sooner or later): due to the random nature of mutations, in the long run, all possible changes occur, and those which are helpful persist and solidify. Hence, evolution, if it works in this way, is guaranteed to lead to ever greater fitness. Darwin pointed out that evolution is what is behind the development of our organic world, for individual organisms are replicators of the above kind (though the situation is more complicated than in the above sketch in the case of those organisms which reproduce sexually). Natural selection causes individual organisms to increase their fitness, but the ways in which different kinds of organisms come to do this are very diverse (some developing wings, others gills, and still others rules, language and reason). And according to the current version, the replicators of evolution theory are genes, while organisms are the vehicles the genes build so as to replicate. Which properties of an organism determine its fitness? Certainly, it is in part its body build, organs, extremities etc. But it is also its dispositions to behavior, the behavioral patterns it displays in response to events in its environment. In the case of some organisms, those patterns are pretty rigid – they can be seen as directly “hardwired”. The behavior of other organisms is more flexible – the patterns they display are not uniquely determined, and hence, such organisms are, for example, able to learn by trial and error. 10.2  Evolution and culture The mechanism of natural selection, as described in the previous section, constitutes the backbone of evolution; at present, however, we know much more about the specific nature of the evolution of our organic world, including many other factors which intervene and make the process much more intricate than in the above simple model. Jablonka and Lamb (2014), for example, present evolution as a “four-dimensional process”, only one of the dimensions being the genetic system as sketched out above. The second dimension, according to the author, is “epigenetic inheritance”, in which cells pass information to their daughter cells in other ways than via genes. The third dimension concerns issues discussed in detail here: social learning and cultural inheritance. The fourth dimension then, according to Jablonka and Lamb, concerns “how information is transmitted through

130  Behavioral patterns language and other forms of symbolic communication”. Let us say a few words about the third and fourth dimensions. It is clear that in the case of such complex organisms as us humans not everything they do can be traced back to their genes. However, not everything that we do and that is not a matter of genes is merely random; a lot of such behavior can be traced back to what we tend to call, very loosely, culture. Some of the behaviors we display are due to having been taught them, because we saw that they are useful, or simply because this is how things are done in our society. Hence, it seems that our behavior patterns can have one of two sources: genetics (or “nature”, as it is sometimes called) and culture (or “nurture”) (Goldhaber, 2012; Ceci and Williams, 1999). Thus, it would seem that there is not only the genetic inheritance we sketched in the previous section but also something like a “cultural” inheritance. Therefore, Richard Dawkins, one of the founding fathers of the modern version of evolution theory, had already in his path-breaking book The Selfish Gene (Dawkins, 1976) considered there to be a direct parallel between genetics and culture: he thinks that just like genes undergo natural selection, there are units of culture, which Dawkins calls memes and which undergo “cultural selection”. They live in the brains of people and, if they are not useful or appealing, they fade away, whereas in the opposite case, they flourish and spread. However, the theory of memes (sometimes called memetics – see, e.g., Blackmore, 1999) has not really turned out to be viable. It is useful as a metaphor, but attempts to change it into a true theory have foundered on the impossibility of clearly delimiting the concept of meme so that there would be a parallel with genetics. What, however, is even more challenging is that recently the relationship between genetics and culture (“nature” or “nurture”) has turned out to be much more complex than it would have seemed at first sight. There is even a question as to whether genetics and culture can be considered separately at all. One of the phenomena that shows how complicated such a relationship can be is the niche construction, already mentioned in Section 9.4. We have already stated that animals, naturally, try to make their environment as hospitable as possible. Thus, the environment they leave for future generations may differ from the one they inherited from previous generations – and the difference may be, from the viewpoint of evolution, significant. If the changes in the environment aim at greater hospitality, they may blunt some of the pressures that fuel the evolution of the species in question. Imagine, for example, that some feature of the environment represents a danger for a species – then the species may be expected to work toward doing away with it, or at least moderating it. However, the same feature is likely to represent pressure providing for the selection of those individuals of the

Behavioral patterns  131 species that are able to evade it. Hence, doing away with it, or diminishing its effects, may change the specific evolutionary pressure on the species. It may not be a great change, but the change may further accelerate the niche construction, which may further accelerate the change of the evolutionary trajectory, thus starting a self-propelling spiral. The theory of niche construction is closely connected with that of geneculture coevolution, which can be seen as its special case. The point is again that we can route the genetic evolution by developing our culture. Culture builds a “new niche” for us to inhabit (as we have seen, I maintain that the niche is in fact a condominium of normative spaces), and this new niche moderates the evolutionary pressures that form us. In this way, culture gets imprinted into our genes. Gintis (2011, p. 880) discusses an important example. According to him, “the physiology of speech and facial communication” evolved precisely in this way – as a genetic adaptation to our “linguistic niche”. He writes: The increased social importance of communication in human society rewarded genetic changes that facilitate speech. Regions in the motor cortex expanded in early humans to facilitate speech production. Concurrently, nerves and muscles to the mouth, larynx and tongue became more numerous to handle the complexities of speech […]. Parts of the cerebral cortex, Broca’s and Wernicke’s areas, which do not exist or are relatively small in other primates, are large in humans and permit grammatical speech and comprehension […]. This is important, because, for a long time, language was considered to be a matter of human “biology” (Chomsky, 1986; Pinker, 1995; Berwick and Chomsky, 2016). Gintis and other scholars have recently come to think otherwise. According to them, language first started to be developed as a “cultural tool”, which was then, via the gene-culture coevolution, projected into our genes. As Dor and Jablonka (2014, p. 16) put it succinctly, “First we invented language, then language changed us” (see also Christiansen and Chater, 2016). It follows that, according to Gintis and others, the whole complex physiology underlying our ability to produce and comprehend the nuanced sounds that are the vehicles of our languages evolved because there were already some proto-linguistic sounds around and the ability of mastering them enhanced an individual’s fitness. Many such human features, which are produced by our genes, found their way into the genome because of culture (because, that is, the niche in which we evolved was largely cultural). Thus, the question of “nature” vs. “nurture”, i.e. genes or culture, may not always have a clear sense. True, if we consider an individual, or a group of individuals, then the question of whether a trait of hers has some

132  Behavioral patterns genetic support may, indeed, have a clear sense; but the question whether the trait is ultimately a matter of “nature” rather than “nurture”, i.e. whether it got into our genome independently of any cultural facts, may not (Lewens, 2017). 10.3  Ways of establishing a behavioral pattern The point of speaking about behavioral patterns here is that rule-following can be seen as a matter of such patterns. Patterns that are characteristic of rule-following display a certain “reflectivity” – they consist of a (“groundlevel”) behavior and additional (“first-level”) behavior targeting it. (They may involve also “second-level” behavior targeting “first-level” behavior, etc.) This is of course only a very rough characterization; however, the change of visual angle that this embedding of rules into evolution brings about may help us see what might escape us otherwise. It may seem that organisms display behavioral patterns because this is somehow hardwired in their brains (or whatever they have instead of them) by evolution, for it turned out, in the past, to enhance their fitness. But this would be too hasty. It seems that at least in the case of us humans (but also some other “higher” – viz. very complex – organisms), the situation is such that a lot of what we do seems to be too random or too “opportunistic” to be directly hardwired. Patterns may also originate from our interaction with the environment: we learn that some courses of action, rather than others, get us what we want or let us avoid pitfalls. (Of course, even this can only operate with hardwired support, at least the infrastructure for trial and error, but it is not itself hardwired.) But the environment we interact with may also be social: we may submit to the pressure of our peers. And normative attitudes are precisely an effective tool for forcing us into behavioral patterns; the pattern may come into being because individuals are compelled into it because they are rewarded when behaving in accordance with the pattern and penalized when not. Of course, then the question that arises is where do the normative attitudes come from? Are they hardwired in the minds of the individuals assuming these attitudes? This would seem odd. Why would evolution produce a pattern in such a detoured way, producing “enforcers” to force it upon “enforcees” instead of just making the enforcees display it right away? And even if this were to happen, would it not constitute an evolutionary advantage for those who happened to have the pattern already wired in, so that the pattern would soon be propagated in the genetic way after all? Is there the possibility that normative attitudes can also be something that people were forced into, rather than having it hardwired? Perhaps, but

Behavioral patterns  133 then we would need some “second-level” normative attitudes that would enforce the “first-level” ones; we would need not only the normative attitudes that would force people to bow to the chief, but we would need also attitudes that would force people to force people to bow to the chief, but then also attitudes that would force people to force people to force people to bow to the chief, etc. An infinite regress would be imminent. If we have a behavioral pattern (like bowing to the chief), let us call the pattern that enforces this one (like forcing others to bow to the chief) an associated higher-level pattern. Now the trouble is that every pattern must be either transmitted genetically, or there must be the associated higher-level pattern. Thus, whichever pattern we take, there must be an infinite hierarchy of associated higher-level patterns (the pattern associated with it, one associated with the associated one etc.) or there must be a pattern, somewhere in the hierarchy, which would be innate. In the former case, we have an infinite regress; in the latter, we do not get rid of the innateness, only move it somewhere up the hierarchy. Could this dilemma be bypassed? Imagine that at some point in the hierarchy, the enforcers would be able to enforce not only the relevant pattern but also, in one sweep, its associated higher-level pattern. Imagine that this happens immediately at the bottom of the hierarchy. The enforcers would be able to not only make the enforcees bow to the chief but also force others to bow to the chief. There would no longer be any need for further levels of the hierarchy – there would be no infinite regress and no need for innateness. Call a normative attitude supernormative if it has the property that it enforces not only its target pattern but also the associated higher-level pattern. What we have just indicated is that once a species was able to wield such supernormative attitudes, it would be ready for a purely non-genetic spreading of behavioral patterns. What I want to claim here is that we humans are such a species – we have developed a peculiar ability of grasping normative attitudes targeting us as not only something to which we should yield, but rather as something that ought to be – something that should also be required of others. As we already pointed out, it seems that human infants, unlike the young of other animal species, display a specific kind of response to facing normative attitudes. Not only do they yield to the pressure of the enforcers, adjusting their behavior to display the requisite pattern, but they also take it to include the associated higher-level pattern. The fact that in our specific case, this works in this way is significant; indeed, I think this is the most basic key to our human specificity. Why would human children display this kind of idiosyncrasy? Why would they be so sensitive to normative attitudes that they make them into

134  Behavioral patterns supernormative ones? The answer to such questions may, of course, only be speculative and it may only rest on considerations about the struggle for fitness in evolution. But if we look at the situation from this viewpoint, a candidate answer is forthcoming: it is precisely this kind of idiosyncrasy that provides for the massive paragenetic proliferation of behavioral patterns that is at most meager in other species (but see, e.g., Whiten, 2021). In other words, it is what lets us have, aside from genetic heredity, also an additional “cultural” one. Hence, the idea of this book is then that we humans are special (mind me: I do not mean better!) – we are a normative species and part and parcel of being such a normative species is precisely being able to establish behavioral patterns in this “cultural”, rather than the “genetic”, way. Note that it follows that normative attitudes need not be matters of genetically encoded dispositions. True, there may be innate normative attitudes, but the supernormative nature of the attitudes makes innateness dispensable. The only thing that must be a matter of genetics is the infrastructure that provides for the supernormativity of the attitudes. It can be seen as a mechanism that tends to turn any normative pattern – that is learned – into a supernormative one, supplementing it with the associated higher-level pattern. 10.4  Why bypass genes? We have seen that, aside from genetic hardwiring, a behavioral pattern can come into being by means of normative attitudes; now we see that there is a way for normative attitudes to bootstrap themselves into existence without being hardwired. This, then, establishes a way of proliferating patterns that is completely “paragenetic”. As a result, genetic evolution gets supplemented by this paragenetic level. Now, however, the question is why we need such behavioral patterns, handed down from generation to generation, instead of merely those that are engraved into our genes directly by natural selection. In a typical case, a behavioral pattern is a response to something that goes on in the environment. Birds go south because winters are not so accommodating in Europe. We drive on the right side to coordinate traffic flow and to minimalize traffic accidents. Spiders make webs so they can catch enough prey to survive. It is very good to have such rough and ready responses to the environment. But this only holds so long as the environment is stable. What is characteristic of us humans is that our behavior has come to be unprecedentedly flexible; that the relationship between the behavioral “outputs” and the “inputs” that stimulate them, which is relatively tight in case of other species, has been largely loosened. We humans have achieved

Behavioral patterns  135 a state where our behavior is so versatile and flexible that the “inputs” largely underdetermine the “outputs” and give our behavior, as Dennett (2018b, p. 7) puts it, “another degree of freedom”. Such versatility of behavior – that is, the fact that our reactions to various stimuli are nothing close to hardwired – is useful, especially within changing environments. If an organism reacts to a stimulus in a fixed way, i.e. in a way that turned out to be useful for its ancestors in the given environment, then there may be a problem when the environment changes. This is not the case when the organism has a larger spectrum of responses that it can try – then perhaps it can find one useful in the new environment. However, until the environment changes, it is useful to have some “standard” reactions to recurrent situations. Thus, on the one hand, it is useful to have a rich repertoire of reactions; on the other hand, however, it is useful to tame this richness so that in “standard” circumstances we can readily display “standard” (which, in this case, means: tried and true) reactions. The solution to this predicament is usually described in terms of the concept of “conformity”. Or, more specifically, “informational conformity”: “Informational influence is based on the desire to be accurate; others’ responses are used as a source of information about reality, and people conform because they believe that the others may be correct” (Campbell and Fairey, 1989, p. 458). People follow those who are successful, so the story goes, since copying their ways may be profitable.1 But copying and conforming is at best only a part of what is really happening, at least in our human case.2 It is not only that people acquire a set of “standard” reactions by copying the behavior of some model individuals, but the standard reactions are also actively instilled into the apprentices of human societies. And the most basic tools of such instilment are precisely normative attitudes. The truly important role for normative attitudes, however, emerges within the context of a different kind of “standardization” of reaction. When we deal with nature, then “standard” amounts to plus/minus successful. However, when we deal with each other, then “standard” often means something else: it means in agreement with deliberately established convention. We not only need to be “standard” to make use of the “tried and true” methods of achieving our goals, we also need to be standard to be predictable for each other. The reason is that we are also essentially social creatures, and our success depends on how we cope with each other – and here again, flexibility might be a hindrance. Cooperation and prosperous coexistence depend, to a large extent, on predictability, and the more flexibility of behavior you have, the less predictable you tend to be.

136  Behavioral patterns Campbell and Fairey (1989, p. 458) write about this kind of conformity, calling it “normative conformity”, but they characterize it in a way which I find misleading: “Normative influence is based on the desire to maximize social outcomes. Even when people believe the others are wrong, they may conform in order to gain the rewards or avoid the punishments that such agreement and disagreement mediate”. This sounds as if the problem were to seek an optimal way through the “social landscape” in a similar way in which we seek it throughout the “natural landscape”. But here we must first create the “social landscape”, and the normative attitudes and the conformity they induce do, at the most basic level, just this. While nature wields us resistance which we must get around by the clever ways we invent (and which are then subject to imitation or conformity), the resistance wielded by society is, by and large, constituted by the normative attitudes to which we conform. Here, obviously, the role of the normative attitudes is much more substantial than in the previous case. There, they were employed to accelerate the ability to acquire useful reactions, but here they determine which reactions are useful and thus constitute our very forms of social life: they decide which actions are “successful” and need to be reinforced, and which are not. (We must not only learn to do things in the right way according to our community – we must institute what is right in the first place.) 10.5 Summary Rule-following can be seen as a (very complex) behavioral pattern, and the wisdom of the recent decades is that everything in the biological world, including such patterns, is explained by evolution. Patterns that are characteristic of rule-following crucially involve normative attitudes – a “higher-level” behavior targeting a “ground-level” one. In general, a behavioral pattern may be genetically “hardwired”, but it can be also enforced by normative attitudes. But then, it would seem, either the attitudes must be “hardwired” or they must be enforced by further normative attitudes, etc. There is, however, one more possibility, which is really instantiated by us humans: our normative attitudes work as “supernormative”. They not only enforce the required pattern but also enforce the enforcement of the pattern. (Toddlers learning language not only learn what to say when, but they also come to understand that this is something to require from others, something that ought to be.) In this way, its proliferation ceases to require any genetic support: the enforcers not only produce enforcees, but the enforcees mutate into new enforcers. This enables our “cultural inheritance”, which is the carrier of our culture. Behavioral patterns handed down from generation to generation allow us to convey our unprecedentedly

Behavioral patterns  137 flexible behavior into some standard routes; where “standard” means either effective (in case of pursuing goals), or respecting social conventions (when interacting with other people). Notes 1 This is connected with the phenomenon of prestige – an attribute possessed by those who are worth being copied. See Henrich and Gil-White (2001). 2 Conformity is argued to be far from specific to us. See Claidière and Whiten (2012).

11 Practices

11.1  Behavior and “metabehavior” In the case of the Sellarsian rule-governed behavior, we have the “enforcers” who strive to form the behavior of the “enforcees” into a pattern. Hence, the normative attitudes are displayed by individuals different from those whose behavior is its target. And the whole thing works because the enforcees not only acquire the requisite patterns, but as a part and parcel of this they become the enforcers themselves – the normative attitudes work as supernormative. However, the enforcers do not aim their normative attitudes only at those who are not yet level with them (at the learners of the rules in question), they aim them at all followers (and would-be followers) of the rules indiscriminately. This means that they aim the normative attitudes also on those who aim their normative attitudes on them, and indeed they aim the attitude also on themselves. In this way, the activities involving rulefollowing come to target themselves in this peculiar manner. To elucidate this specific feature of human activities, let us look at them from a different angle. Imagine a behavioral pattern displayed by some animals; a flock of hens, for example, rushing out of a henhouse looking for food. From our viewpoint (though, presumably, not from the hens’), there is a “metalevel” to this behavior. On the metalevel, we (though not they) can describe their behavior, we can take it for “correct” (measured by our aims) or “incorrect”, and we can attempt to regulate it. (We can open the doors of the henhouse at certain hours, prepare food for the hens at certain places etc.) The existence of the metalevel is given by the fact that we humans can assume certain attitudes toward the behavior of the hens; we can report it or try to influence it. Of course, the same happens if the animals displaying the basic level pattern are us, humans. Imagine a school: there is a host of children whose behavior is molded by a (much smaller) host of teachers. We can look at this enterprise so that its point is to instill appropriate behavioral patterns DOI: 10.4324/9781003388876-12

Practices  139 in the former (reading, writing, counting, …) by means of the corresponding “metapatterns” of the latter (teaching). Hence again: there is a behavior and its metalevel, because we humans can also assume certain attitudes to the behavior of other humans. However, in this case, it can happen that those who display the attitudes on the “metalevel”, who display a “first-level” behavior targeting the “ground-level” one, are the same humans as those who display the “ground-level” behavior. In this case, the whole pattern, consisting of the two levels, becomes what we can call self-reflective. We may call such activities practices. And the thesis which I defend here is that human activities are mostly characterized by being practices in this sense of the word. In other words, while any behavior of animals (or, for that matter, “behavior” of inanimate things) can be described on a metalevel and regulated from without, what we call human practices already incorporates the metalevel, they are regulated from within (hence: self-regulated). To become a competent practitioner of the human language games, viz. a speaker of language, for instance, the speaker, apart from becoming able to produce appropriate “languagings”, must also, as Sellars (1974, p. 424), puts it, “acquire the ability to language about languagings, to criticize languagings, including his own”. (This does not mean that knowing language involves being able to articulate the rules explicitly; but it involves being able to verbalize the normative attitudes.) Similarly, Brandom (2000, p. 20ff) stresses that it is the self-reflective quality of our human conceptual activities that enables us to put the rules that regulate them into words and thereby become “semantically self-conscious”. Brandom (2000, p. 20) explains the rise of self-consciousness in terms of his “making explicit”: Here the original inferential-propositional model of awareness (in the sense of sapience) is applied at a higher level. In the first application, we get an account of consciousness – for example that Leo is a lion. In the second application we get an account of a kind of semantic self-consciousness. For in this way we begin to say what we are doing in saying that Leo is a lion. For instance, we make explicit (in the form of a claimable, and so propositional content) that we are committing ourselves thereby to his being a mammal by saying that if something is a lion, then it is a mammal. Note that this two-tier organization of human practices and human awareness is what a rule-following behavior becomes when seen under a different visual angle. We impose rules on each other; and we do it in terms of our normative attitudes that aim at regulating our behavior. Hence, there

140 Practices is the “ordinary”, “ground-level” behavior, and then there are the “metalevel” normative attitudes that are aimed at it. 11.2  What is a practice? In his famous paper, Rawls (1955, p. 33) uses the term practice as “any form of activity specified by a system of rules which defines offices, roles, moves, penalties, defenses, and so on, and which gives the activity its structure”. According to Rouse (2007, p. 48), “a performance belongs to a practice if it is appropriate to hold it accountable as a correct or incorrect performance of that practice”. The essence of practices of this kind thus consists in what Rouse calls “the mutual accountability of their constitutive performances”. I think both these delimitations are on the right track, we only need to anatomize the concept a little more closely. We have already touched upon Ginsborg’s (2011) concept of “primitive normativity”. She employs it to meet the challenge posed by Kripke (1982). It amounts to the fact that we hold some acts for correct as a matter of a primitive attitude, and not as a result of any conscious application of a rule. “Thus”, she writes (pp. 244–245), “your disposition is not just to say ‘125’ in answer to ‘68 plus 57,’ ‘126’ in answer to ‘68 plus 58,’ and so on; it is also, in each case, to take what you are saying to be the appropriate response to the query. You are disposed not only to respond with a number which is in fact the sum, but to consider that particular response appropriate”. I think that the “primitive normativity” Ginsborg is urging here is the kind manifested by our normative attitudes. When we carry out the additions, there are two things in play: not only the disposition to produce the results, but also the disposition to take the results as adequate or correct. It is crucial to stress that the normative attitudes’ being “primitive” involves their being nothing like propositional attitudes, and not being based on an appreciation of rules or meanings. On the contrary, these attitudes underlie all rules in general, and the linguistic rules that underlie meanings in particular. It is a feature of us humans that we have developed these selfreflective patterns involving these pro- and con- attitudes to the behavior. Hence in this sense, Ginsborg’s proposal is wholly in the spirit of our approach. But there is an important difference: it renders the responses of a person correct or incorrect because of the existence of the corresponding normative attitudes of the very same person. This, I think, stems from the conviction of the author that counting is primarily an individual, mental activity. The same would hold, according to many authors, about reasoning more generally. But this assumption, I am convinced, is mistaken, and it blocks us from arriving at an adequate understanding of human practices.

Practices  141 It is not that I would not apply the normative attitudes also to myself. The point is that practices such as counting or reasoning cannot evolve as purely private, because a public dimension is in their very essence. This is not to say that an individual, independently of a society, cannot have evolved some technique of classifying groups of objects according to their numerosity, or a technique to estimate what will be the case if something else is the case; but it is to say that to make this into fully fledged counting or reasoning the individual needs a society, because it is only within the context of a society that the practices can acquire the normative dimension which qualifies them as being distinctively human (“self-conscious”) practices.1 It may seem strange that something as essentially mental as reasoning would have evolved not in the mind but in the arena of the intersection of many minds; however, current research is bearing this out. Most forcefully it is put forward by Mercier and Sperber (2011, 2017), who argue, I think rightly, that private reasoning is secondary to public argumentation rather than the other way around. This is of a piece with the recent trend to see the human mind as much more a social product than used to be typical (see, e.g., Tomasello, 2014). All in all, I think that to understand the practice of reasoning, as well as the other practices, we must accept that it involves normative attitudes, but not merely the normative attitudes of a subject to her own inferences (as Ginsborg insists). Rather, we must accept that the very practice, along with so many other distinctively human practices, presupposes assuming such normative attitudes to each other. It is this kind of mutual assessment that constitutes the practices as such and gives them their essence. When this is up and running, an individual may tend to internalize the attitudes and evaluate her own thinking; this, however, is possible only after the public practice is in place. 11.3 From practices to normative virtual spaces to uniquely human thinking We saw that we humans mostly live “under the eye of evaluation”, that most of what we do is evaluated as right or wrong (in various senses of the word) by our peers; and we are permanently and acutely aware of this fact. This has become an ever present background of our doing things; it is, we saw, what makes our doing things into acting. In our words, we live most parts of our lives within our normative spaces (which thus may appear like cages into which we are born and which it is hard to escape, see Maryanski & Turner, 1992). In the first chapter, I indicated that for our words and expressions to serve as the useful kinds of tools they serve as, we must create virtual

142 Practices spaces in which they become tools useful just like hammers, saws or pliers are useful in the natural world. What makes a hammer a useful tool? Certainly the fact that the world contains objects that can be crushed by it and contains objects into which it is easy to drive nails by means of it. Hence, certain parts of the world “respond” to using hammers in a useful way. Some sounds may also be useful in the natural world. A sound may perhaps be used to chase some dangerous animals away, or to attract attention. However, over and above this, the usefulness of sounds is quite limited. The situation, however, changes when we develop linguistic practices that establish the normative spaces in which sounds are responded to not by the effects of the natural world, but by countermoves within the rule-governed space. And the responses to The house is very nice are very different from those to The house is on fire – which makes each of the two sentences into a very specific tool of the language of English language games. It is important to realize that the “space of meaningfulness”, as the most important case of the normative space, is virtual not only in the sense that it is a man-made extension of the natural world – it is virtual also in the sense that it allows us to achieve things without a significant investment of energy, what is usually a sine qua non of an engagement with the natural world. While doing anything significant in the natural world requires the exertion of much energy (as Zahavi & Zahavi, 1997, pointed out, a signal in the animal world can be considered reliable only if it is costly, for it is only when it costs the signaller something that could not be faked by mere pretenders that it can be “trusted”), the world constituted by the rules of language provides for the possibility of achieving (a peculiar kind of) significance at virtually zero energetic cost.2 It is also important to see that the normative spaces that we build are not only something that shape us from without – they are already imprinted in our thinking. The assessments guiding human enculturation have come to equip us with what some psychologists have come to call norm psychology (Sripada & Stich, 2007), or what Tomasello (2014, p. 38) calls “self-governance”: The general process is thus that the young child imagines how some social interactant is comprehending or evaluating her, and then she uses this to socially self-regulate. Scaling up the sociality involved, children from about three years of age (but, needless to say, not apes) socially self-regulate on the basis of cultural structures – such as, prototypically, conventional and moral norms – that are based in cognitive processes of collective intentionality, what we may call normative self-governance. This process operates in much the same

Practices  143 way as social self-regulation in a dyadic context, but the social other in this case is more generalized and authoritative. Thus, from sometime during the late preschool period, young children self-regulate both their thinking and actions not just by how efficacious they will be in the current context (as do apes), and not just by how they will affect a particular person’s thoughts or evaluations (as do younger children), but also by the perspective of how these will fit with the normative expectations of the social group. This process essentially constitutes the construction of a normative point of view as a selfregulating mechanism, arguably the capstone of the ontogeny of uniquely human cognition (normative rationality) and sociality (normative morality). And, again, engaging in social self-regulation or normative self-governance can lead to developmental change without any additional inputs from either maturation or learning. What happens, then, is that the normative spaces, which we construct out of rules to structure our social life, are as if absorbed into individual minds and become operative in structuring our mental life. The “eye of evaluation” under which we have learned to live our lives is drawn inside and becomes part of us. (Freud’s idea of superego is not only incidentally similar to this). The result of this process of self-governance is a true self-consciousness, which is characteristic of us humans. It is the outcome of a process by which we turn our normative attitudes, primarily aimed at others, on ourselves. There are indications that it is precisely this kind of self-targeting structure that is characteristic of a number of processes which have to do with how we humans have become what we are. In psychology, we can find this self-reflective pattern in the so-called higher-order theory of consciousness (Lau & Rosenthal, 2011). According to this theory, consciousness is the result of the mind turning on itself. The conscious state of a subject, according to this theory, is a state that the subject is aware of (Lycan, 2001). To put it in an oversimplified way, it is the thinking that is subject to thinking. Thinking, as such, first evolved to help us deal with the outer world, but once we were able to make it into the self-targeting matter, we got the bonus of consciousness, and hence our distinctively human form of mental life. Carruthers (2011) presents an even more epic thesis about the self-targeting of the mind: he claims that the very notion of our mind is the result of our interpretive capabilities, developed for the purposes of making sense of others, being turned on ourselves.3 From the viewpoint of theories of behavior, a similar story is told about domestication. Domestication appeared as a skill to turn some animals into species which serve us and live in a certain symbiosis with us.

144 Practices But domestication, according to some theories, has also come to be selftargeted: it seems that we have domesticated also ourselves (Wrangham, 2019). This seems to be a story parallel to the previous one, only concerning behavior rather than thinking. All in all, the normative dimension we have added to our world has made it possible (and in fact inevitable) for us to live very specific kinds of lives. The self-reflection which we have imputed to our practices by way of supplying the ever present evaluability of what we do, so that we do everything as if under the eye of a normative beholder, makes the practices and our lives “self-conscious” in a sense not too dissimilar to Hegel’s4 (though, as we saw, we can account for this also in some very different, perhaps naturalistic terms). In any case, our life is the life of understanding the reflective assessment of our actions. 11.4 Affordances We have indicated that we want to embrace the picture of language according to which expressions are tools. However, we have seen that expressions are quite peculiar tools, very different from hammers, pliers or saws. A hammer is a tool because its form and its constitution make it suitable for accomplishing certain tasks, like pounding nails. But the form or constitution of a sound like It is raining outside in no way predisposes it as a tool for accomplishing a task such as reporting the weather outside. So how can expressions be tools? In fact, we know that the sentence It is raining outside, in countries where people speak English, can be used as a tool of reporting weather, so how is it possible? An ordinary response to this is that the sound does not serve as this kind of tool qua sound, but rather qua being a symbol. This might be an acceptable answer, but now everything hangs on the explanation of how we make a sound into a symbol. One possible explanation is the representational one: we let the sound stand for something, making it into a signifier and the thing into a signified. (Indeed, a celebrated theory of symbols due to Peirce, 1998, p. 478, explains symbols as a subspecies of signs, which are characterized as “anything which is so determined by something else, called its Object, and so determines an effect upon a person, which effect I call its interpretant, that the later is thereby mediately determined by the former”.) However, we have noted that Wittgenstein saw that the seemingly transparent notion of “letting something stand for something” is not transparent at all, and that the attempts to clarify it may lead us into various blind alleys. Given this, I offer a quite different answer to the question of how can expressions serve as tools. The answer I give is that we are able to create a specific kind of “virtual reality” – viz. a practice in which words can

Practices  145 serve in a similar way as hammers or pliers do in the natural world. Just as the hammer has the features needed to accomplish a task we face in the natural world, given the natural laws of the world, so does the sound It is raining outside possess the features to accomplish a task in its “virtual world” – because the world is put together in such a way that it is acutely sensitive to subtle differences between the sounds. Where does such a “virtual world” come from and how can it be put together in such a way that it makes our sounds into such useful tools that make up languages as we know them today? A short answer I want to give to this question in this book is that it is a matter of our practice, which we have built out of our rules. The answer is not so odd as it may look at first sight. Compare: we have built a “virtual world of chess” out of certain rules for how to handle the pieces, which has made the pieces into tools for competing with an opponent in the memorable game. The idea is that language works in a similar way: it is the rules of our language games that have built the virtual world in which words and sentences serve as tools of our meaningful talk. And, more generally, we have built a couple of other, more or less important “virtual worlds” in which we play our “social games” and in which various items of our natural world become tools of unseen significances. Thus, we have completely altered our “form of life”. The effectivity of the tools in a world is yielded by the nature of the world and the way the tools are able to incite its “response”. If we are able to construct new worlds (or “worlds”), we enable new kinds of tools (or “tools” – in the case of the space of meaningfulness, they are usually called “symbols”) that are effective in the new worlds. And as the worlds are not quite of the same kind as the natural one (they are “virtual”), the tools also are not of the kind that works in the natural one. It is important to realize that tools and environments are complementary – they do or do not fit together. And while the natural world appears to be simply given to us so that we must invent our tools to fit it, in the case of normative spaces, it is also the environment that is our product, so we may shape it to fit our would-be tools. (Also the natural world is being upgraded, by us, to fit our tools – for example, we produce various constructions using the screws and nuts to fit with the screwdrivers we may want to use on them sometimes in the future.) Almost half a century ago, J. J. Gibson introduced the term affordance (Gibson, 1979) that has become immensely popular in psychology and philosophy by stipulating that “the affordances of the environment are what it offers the animal, what it provides or furnishes, either for good or ill” (p. 127). The main idea is that an animal does not perceive its environment as just what there is, going on to figure out what these things may be good for, but it directly perceives what they are good for, in particular

146 Practices what actions on its part they invite. (The parallel between the concepts of Zuhandensein and Vorhandensein of Heidegger, 1927, is obvious.) Gibson’s original approach focused on perception and on the links between perception and action. Gibson opposed the picture according to which perception gave rise to mental representations, which then formed the input of an information-processing system, inciting the actions.5 According to him, there is a sense in which perception directly elicits actions. Thus, an animal always already perceives a “meaningful” world – it does not perceive things devoid of any significance until its information-processing system determines how they can be perused, it already perceives things as aids or hindrances, as tools to be used in various ways. This way of thinking about perception and action then gave rise to what has become known as ecological psychology (Chemero & Turvey, 2007). The concept of affordance has since been employed in multifarious ways, often not really in the sense in which Gibson coined it (Chong & Proctor, 2020), and here also I do not intend to do full justice directly to Gibson’s ideas. What is important here is that affordances, clearly, need not be only the “features of the environment that have species-specific or transcultural significance”, but also “features that have significance only within a particular sociocultural context” (Heft, 1989, p. 1). If you see a chess bishop, then perhaps any human (or any primate?) will see it as something that is easy to grasp, but most members of our culture (more specifically the players of chess) will see it as something that moves diagonally over the chessboard. While the existence of these “cultural” affordances are sometimes paid some attention (Ramstead et al., 2016), what, to my knowledge, entirely escapes the attention of philosophers is the ways in which affordances can be produced by culture. And it is precisely this that can serve us as a way of explaining the functioning of the normative spaces we have been talking about. How do we imbue the environment with the kind of significance that links it directly to actions? While there are “natural” affordances the perception of which can be inborn, the perception of the “cultural” ones must be learned. Hence they must be, on the one hand, imprinted into the environment so that it, on the other hand, will be intelligible to humans perceiving it. We have already seen that establishing the normative spaces we discussed in the previous chapters can be seen especially as making available new kinds of actions. Opening up the space of chess is letting us play chess with all the actions that make up the game. Opening up the space of meaningfulness lets us play our language games, viz. utter meaningful sentences that have complex consequences both for the further development of the game and for the world outside of the game. In other words, it

Practices  147 is all about new, “cultural” affordances, sometimes overlaying, sometimes complementing the original “natural” ones. To look at the situation via the prism of affordances is instructive. While we talk about rules, it might seem that our human becoming the kind of normative species we are is about restrictions, prohibitions or discipline. However, once we see that the rules, combined into normative systems, produce new affordances, its point is much more immediately seen. 11.5 Summary We can describe the behavior of animals and we can (try to) regulate it. Our behavior to this effect can be seen as constituting a “metalevel” to the described or regulated behavior of the animals, our “metalevel” interventions come to the “ground-level” behavior “from without”. However, the animals displaying the “ground-level” behavior can also be humans, and, moreover, they may be the same humans as those who target the behavior from the “metalevel”. In this case, we have a complex behavioral pattern involving its own “metalevel”. And what we call human practices are just such patterns already regulating themselves “from within” (hence: “self-regulating” themselves). Brandom stresses that it is the self-reflective quality of our human conceptual activities that enables us to put the rules that regulate them into words and thereby become “semantically selfconscious”. As a practice involves normative attitudes of its participants to each other, it is imbued with the mutual assessment of the individual moves of the participants. When such pervasive reciprocal evaluation is up and running, an individual may tend to internalize the attitudes and evaluate her own thinking; this, however, is possible only after the public practice is in place. Thus, the essence of practices of this kind consists in what Rouse calls “the mutual accountability of their constitutive performances”. Practices can be also seen as constituting new affordances: a specific kind of “virtual niche”, in which our cultural tools, like words, can act as hammers or pliers in our natural niche. Just as the hammer has the features needed to accomplish a task we face in the natural world, given the natural laws of the world, so does the sound It is raining outside possess the features to accomplish a task in its “virtual world” – because the world is put together in such a way that it is acutely sensitive to subtle differences between the sounds. Notes 1 Cf. the discussion of the public nature of rules in the first chapter of this book. 2 As Knight (2008, p. 124) points out: “Each animal can make a difference only physically, only with its body—with signals inseparable from the body [...]. By contrast, a human linguistic utterance—a ‘speech act’—is an intervention in a

148 Practices different kind of reality [...]. A speech act, like a move in a game of ‘let’s pretend’, is internal to reality of this kind [...]. When human life became subject to the rule of law, participation in this kind of reality became possible for the first time. Because signals internal to this novel domain were no longer evolving in a Darwinian world, the familiar laws of signal evolution [...] no longer applied. Intrinsic reliability was no longer a requirement, allowing zero-cost signaling to emerge. Among other consequences, messages could now be encoded as digital shorthands”. 3 Carruthers (2011, p. xii) writes: “[O]ur only mode of access to our own thinking is through the same sensory channels that we use when figuring out the mental states of others. Moreover, knowledge of most kinds of thinking (and hence by extension knowledge of our own standing attitudes) is just as interpretive in character as other-knowledge. Our common-sense conception of the transparency of our own minds is illusory, I shall argue. On the contrary, for the most part our own thoughts and thought processes are (in a sense) opaque to us. For they can only be discerned through an intervening sensory medium whose contents need to be interpreted”. 4 Especially if we read Hegel in the “sociological” terms of some of his interpreters, such as Pinkard (1996) or Brandom (2019). 5 This amounts to what Hurley (2008, p. 2) ridiculed as the “sandwich conception of the mind”, which “regards perception as input from world to mind, action as output from mind to world, and cognition as sandwiched between”.

12 The space of meaningfulness

12.1  The problem of circularity We have already discussed, on a very general level, a possible route from incipient normative attitudes to rudimentary rules as sketched by Beisecker (2013). More specifically, Beisecker is interested in the rudimentary rules of language and he conjectures that our pre-linguistic predecessors registered that certain “grunts” or “hoots” of their peers could be associated with certain circumstances and started to make each other emit the specific sounds in these specific situations by evincing rudimentary normative attitudes. In particular, they aligned the attitudes with those of their peers in the sense that concurrent positive/negative attitudes tend to reinforce. In this way, I suggest, we can imagine the very origins of language. True, some more details should be added, but before I do this, let me mention one important desideratum of theories of this kind. When we consider the theories of the emergence of meaning, we often face something which can be called the problem of circularity. The point is that some of these theories already presuppose meanings (usually only covertly, of course), or it is difficult to make sense of them without such a presupposition. Consider perhaps the most widespread tale about the emergence of meanings due to Grice and his followers. Here, meanings are derived from the intentions of potential speakers and their relatively complex expectations regarding their audience. Yet the most straightforward way to make sense of this complex state of the speaker’s mind is in terms of propositional attitudes. (The speaker intends that the audience believes that ….) Of course, more sophisticated elaborations of the Gricean framework are possible, but it is not easy to construe Gricean intentions in a non-circular way. Likewise, some naturalistic theories of human cognition and language can easily lead us to a circular explanation of the emergence of meaning. Take the effort of Michael Tomasello and his group to base our human specificity on “the ability to participate with others in DOI: 10.4324/9781003388876-13

150  The space of meaningfulness collaborative activities with shared goals and intentions: shared intentionality” (Tomasello et al., 2005, p. 675). Such shared intentionality is usually considered to require an interpersonal reflective structure such as the one proposed by Bratman (1992). Tomasello et al. (2005, p. 680) write: According to Bratman (1992), joint cooperative activities, as he calls them, have three essential characteristics that distinguish them from social interaction in general (here modified slightly): (1) the interactants are mutually responsive to one another, (2) there is a shared goal in the sense that each participant has the goal that we (in mutual knowledge) do X together, and (3) the participants coordinate their plans of action and intentions some way down the hierarchy – which requires that both participants understand both roles of the interaction (role reversal) and so can at least potentially help the other with his role if needed. Again, it is not easy to build such a structure without already involving things such as beliefs or other propositional attitudes. One more example: approaches based on meta-representations, like Mercier and Sperber (2017). As much as I like the overall approach of the authors to reason and to reasoning, it is once again hard to avoid a circularity here. The trouble is that in representing an argument (for the purposes of “seeing” its validity), one seems to need to represent it as a series of propositions, or at least meaningful sentences. Thus, while I want to argue that meanings are a product of inferential practices, here it seems that these practices already presuppose meanings.1 The truth is that our “inferentialistic” exposition of the genesis of meaning is also frequently accused of circularity: the usual objection is that inferences cannot constitute meanings, because they presuppose meanings. As Boghossian (2014, p. 17) puts it, “reasoning is an operation on thought contents and not on symbols”. Elsewhere (Peregrin, 2017), I have shown that this objection does not work against Brandomian inferentialism because it disregards the crucial difference between inferences and inferential rules (normative attitudes). But, of course, no explanation of the birth of meaning out of either inferences or inferential rules can start from inferences that already presuppose meanings. And indeed, the idea I am presenting, as we have seen, is that at the outset inferences were just habitual emittings of sounds in reaction to other sounds. No content, no meaning is present at the outset. What then happens is that the habitual links between kinds of sounds are turned normative by means of the emergence of normative attitudes. And with

The space of meaningfulness 151 the growing complexity of such normative entangledness of the various sounds (potentially imbued with empirical content because some of the rules that govern them incorporate the world2), the roles of the kinds of sounds become what we now call meanings. The picture presented in the following sections is meant to indicate that indeed this could have happened. By no means is it meant as a depiction of what really happened. It is meant merely to show that this kind of development was possible. The framework of evolution poses severe limits to any such possibility, and it is not so easy to come up with a story that can be comprised within its limits. Dennett (2013, p. 59) urges: Human beings devote a lot of time and energy to their game of reason giving, and however stable and satisfying the view appears from inside the space of reasons, the existence of this elaborate set of human behaviors is just as much in need of a biological account as the distraction displays of the birds or the dam-building enterprises of the beavers. Sellars, unlike his more recent Pittsburgh followers, took this question seriously, and while he didn’t – apparently – develop anything beyond a sketch of an evolutionary account of the origins of the practice of asking and giving reasons, he recognized the need for such an account and also pointed to the deep similarity between the law of effect (or Skinnerian operant conditioning) and evolution by natural selection […]. Hence, in what follows, I want to illustrate how the reflexive emitting of sounds could have been transformed, with the help of normative attitudes, into an inferentially structured system of proto-sentences, which further developed into a language as we know it today. Before going into detail, let me return, once more, to chess. The game of chess is made possible by the space framed by the skillfully interconnected rules of chess. By subjecting our pieces of wood to the relevant rules, the pieces become rooks, pawns, bishops, etc., which we use to contest our opponent. And analogously, our language games are made possible by the space framed by their own skillfully interconnected rules. Subjecting sounds to rules enables sounds to become, say, indications that it is raining outside, questions as to whether a train has already left, or an invitation to a party, which we may use to interact with our fellow people on a wholly new level. What is the general nature of the rules of language? How could they have come into being? And how could they have given rise to meanings as we now know them?

152  The space of meaningfulness 12.2 From the game of indication to the game of giving and asking for reasons Inferentialism, which is the main source of inspiration for this book, claims that meaning is a matter of rules, especially inferential rules. Yet, in the previous section, we saw that as inference cannot presuppose meanings, it must have started as some rule-based links among meaningless sounds, only later making the sounds meaningful by weaving them into the growing web of inferential relations. So how could this process have started? Imagine a situation when members of a pre-historic tribe react to some external stimulus by a similar kind of sound. (I think it is not too farfetched to suppose that they could react, e.g., to danger in a similar way. And besides such natural reactions, it is not too difficult to imagine other reasons for producing similar sounds in response to other kinds of stimuli.) If this is the case and if such sounds turn out to be useful (e.g., as warnings, but more generally as sources of information), it may happen that they may come to be supported and directed by incipient normative attitudes. Hence, when individuals come to spontaneously emit a similar sound when they detect a danger, the members of the tribe may slowly come to compel each other, by means of normative attitudes, to always emit that very kind of sound in those cases where they detect a danger, and only in those cases. And in such cases, we have a rudimentary rule: it becomes correct to emit the sound in the relevant situation and incorrect to emit it otherwise. We may call such a correct reaction to an external stimulus a quasiinference – something similar to inference, but leading us to a conclusion not from linguistic premises, but rather from a “situation”. There may be many more conditions framing the correct emission of sounds, but let us concentrate on this one, fixing the emissions as sources of information, as an indication of various (social and natural) events. Let us call the simple language game the game of indication (GI).3 The only kind of rule governing this game is (1) One should emit A only in circumstances C. Now imagining a “proto-language” consisting of the sounds Tiger, Danger and Calm, all being linked to the situations indicated by their current English senses, we have (1a) One should emit Danger only when there is a danger. (1b) One should emit Calm only when there is no danger. (1c) One should emit Tiger only when there is a tiger around.

The space of meaningfulness 153 Let me stress that though I have borrowed the English words functioning, in current English, as terms, this does not mean that they function as such in the hypothetical proto-language. They are just proto-sentences, reactions to circumstances, not terms or classifiers of things. To borrow the vivid phrase that Quine (1995, p. 23) used in a slightly different context,4 they are “innocent of any thought of reification and reference, they are on par with ‘It’s cold’ or ‘It’s raining: just things to say in distinctive circumstances”. Moreover, the function of such sounds would most probably not be exhausted by indicating; they would function in a holistic way as put forward by Wray (1998). There is, clearly, a long way to go from this austere proto-language to a language proper. A lot of hurdles must be cleared to get from it to something like present-day English, including complex syntax and displaced reference. There are, in fact, many speculative accounts of this development (Laland, 2017; Okrent, 2017; Planer & Sterelny, 2021). We are not going to speculate about this, we will concentrate on the inferential structure of language. (The reason is that we follow Sellars in conjecturing that it is the inferential structure of language that is beyond the whole of semantics; that it is, therefore, this structure that we must throw light on.) And the important thing is that the inferential structure is likely to emerge already on the level of our proto-language. Notice also that that there is a straightforward parallel between chess and such incipient language. We can see the various sounds comprising a language on a par with chess pieces, we need to assume that our language games are governed by rules which tell us what will be classified as correct “moves” with those sounds. A rule of chess has the form, e.g., A bishop may move diagonally. Analogously, a rule of a language game could be The sound X may be displayed in such and such circumstances; e.g., The sound “Tiger” may be displayed when there is a tiger approaching. Thus, within our speculative scenario, sounds become tied, by normative attitudes, to various outer circumstances; it becomes correct to emit a particular sound in some circumstances, and incorrect to emit it in other ones. Given (1), A is a correct indication of C. Given a couple of sounds governed by such kinds of rules, we have a “language” consisting of something reminiscent of rudimentary versions of what Quine (1960) called “occasion sentences”5; and we have GI. Now it is easy to imagine how this game might obtain a more complex superstructure. The circumstances that determine the propriety of emitting certain sounds may become constituted by not (only) the extralinguistic situations (an approaching tiger), but also by other moves of the game. In particular, when you have a sound A indicating the circumstance C, and

154  The space of meaningfulness a sound B indication the circumstance C′, where C and C′ cannot occur simultaneously, we have (2) When it is correct to emit A, it is not correct to emit B. For example, one should never emit a sound indicating an approaching tiger (Tiger) and, at the same time, a sound indicating the absence of danger (Calm). Similarly, if C′ is bound to occur whenever C does, we have (3) When it is correct to emit A, it is correct to emit B. If you emit a sound indicating an approaching tiger (Tiger), you may also emit a sound indicating danger (Danger). (Or perhaps more realistically: if you emit a sound indicating an approaching tiger, you already thereby count as also emitting a sound indicating danger.) Thus, the proto-language becomes equipped with a rudimentary semantic structure. A rule of the form (2) renders the sound B incompatible with A (Tiger is incompatible with Calm); while (3) renders B inferable from A (Danger is inferable from Tiger). And note that this is already not a quasiinference, but an inference proper – its premise and conclusions are linguistic items. Note that the normative attitudes instituting the most rudimentary features of the inferential structure of the proto-language need not be very different from those that fasten sounds to circumstances. Just like uttering Calm when there is a tiger approaching evokes negative attitudes, so will, for example, negative attitudes be evoked by uttering Calm right after uttering Tiger. Note also that the structure is not likely to keep merely mirroring the facts concerning the necessary or impossible concomitance of events in the world. The incompatibility of Tiger and Calm (or the inferability of Danger from Tiger) derives from the experience that any occurrences of tigers so far meant danger; but once we embed the incompatibility (resp. the inferability) into the language, it acquires a life of its own and starts to regulate, rather than mirror, our future experiences. If we discover some animals that are like tigers, but are not dangerous, we are bound to refrain from calling them tigers. (Of course we can, in principle, also adjust our language and retract the incompatibility, but linguistic change tends to be a slow and tedious process.) Thus, it is here that the rules of language start to be underpinned by an arbitrariness that impedes their construal as direct products of evolution. Here we go beyond what is immediately useful (in the sense of enhancing fitness), for we go beyond what is directly grounded in the empirically given world. Of course, this is not to say that this would not be reasonable, nor

The space of meaningfulness 155 is it to say that we would extricate ourselves from the accountability to evolution – but we have extricated ourselves from being closely monitored by it. We have a limited possibility to go the way we choose (though in the long run it may turn out to go against evolution, in which case it will be hampered, or else accord with evolution, in which case it will be bolstered). In this way, the semantic structure of language comes to be, ever more, no mere shadow of our experience, it starts to take part in the organization of experience and starts to resemble a prism through which we observe the world and discern its structures. More precisely, there is a “dialectical” process of adjustment of the structure of language to that of the world and vice versa, including the ongoing negotiations among the speakers of the language regarding this twofold structure. Given this, we can speculate about other rules that may come into being: (4) If somebody emits an incorrect indication of the present situation, you should emit a correct one. For example, if you see that there is a danger around and somebody emits a sound indicating the absence of danger (Calm), you should emit a sound indicating danger (Danger). (5) If you emit a correct indication of the present situation, and somebody reacts with an incorrect one, you should emit another correct indication from which the original is inferable. Imagine that you see an approaching tiger, emit the sound indicating danger (Danger), and somebody emits a sound indicating the absence of danger (Calm). Then you should emit a sound indicating tiger (Tiger). The upgrades on the GI we have just considered have promoted it halfway to Brandom’s the game of giving and asking for reasons (GOGAR). Rule (4) may lead to the obligation to react to what we think is an incorrect indication: it prompts us to emit a sound that is incompatible with it (in the current situation) and thus to challenge it. Rule (5) can then lead to the obligation to give a reason for a claim that is challenged: to emit sounds that are more likely to go without a challenge and from which the original claim is inferable. Given the above example of proto-linguistic vocabulary (where Danger is inferable from Tiger, while Calm is incompatible with Danger), we can start to play a rudimentary version of GOGAR: X: Danger! Y: Calm! X: Tiger!

156  The space of meaningfulness The first move amounts to X’s indication of a danger. But then Y, as a response to this indication, indicates something that is incompatible with X’s indication, namely the absence of danger. This will not leave X calm, for she has learnt to react to an indication incompatible with her own indication, and she would put forward a more specific indication of the situation she indicates: Y must see the tiger, and if she sees it, she must accept that it is correct to indicate it, and hence that it is correct to indicate “danger!”. Like GI, a rudimentary form of GOGAR can still have only one kind of “multipurpose” move. Displaying one of the available sounds may count as a challenge (a request for reasons, if it is incompatible with a previous display) or as a support (a presentation of reason, if a previous display is inferable from it). And once giving a reason becomes what one should do when one’s claim gets challenged, this move starts to acquire the character of assertion (given that assertion is a speech act characterized by a specific kind of responsibility, the obligation of giving reasons for it if it is challenged). It is to be expected that the set of sentences will grow ever bigger and more intricately interconnected. Moreover, we may expect an emergence of “logical” words. The fact that Danger follows from Tiger may come to be expressed in the language: If tiger, then danger. And similarly the fact that Danger is incompatible with Calm: If danger, then not calm. In this way, the language becomes enriched with what we can call “logical vocabulary” and thereby also with sentences that are not occasional.6 We thus now have rules like (6) It is correct to emit A and B if it is correct to emit both A and B. (7) It is correct to emit if A then B if B is inferable from A. etc. 12.3  Meanings, propositions and concepts It is important to appreciate the difference between the rudimentary forms of GOGAR and its present form. We now play GOGAR in a very advanced mode: we challenge an opponent’s moves because we do not believe her or because we want to know why she asserts what she asserts; and we give reasons because we want to convince our opponents and because we want to explain to them why we assert what we assert. (We also have many more special linguistic tools than merely assertions, with which alone, we saw, a rudimentary form of GOGAR may be played.) However, this is an achievement of our developed form of the game, it is not characteristic of its more rudimentary stadia. Indeed, this would beg the question: explicit

The space of meaningfulness 157 reasoning, which today is part and parcel of playing the game, is more of a consequence of its development than its presupposition. Animals react to their environment in various ways. Indeed, the prosperous existence of an individual largely depends on the fact that it is able to react appropriately – to run away from predators, to gather food, etc. The environment of an individual of a social species, like us humans, is also constituted by its conspecifics, so it is necessary to react to them and to what they do, and also here it is vital to react appropriately. And our human environment is also increasingly constituted by the sounds our conspecifics emit, and we learn to react to them. We saw that the displays on which the rudimentary forms of GOGAR are based can be construed as this kind of reaction. When we hear the indication of an event, but we feel it is mistaken, then we react by challenging the indication. We can easily imagine that something like this can be helpful in that it refines group abilities to indicate the events – to root out mistakes and cheats. Similarly, when we perceive a challenge to our indication, we try to display some other indication, which may be more obvious for the challenger and from which the challenged original indication is inferable. Thus, though the GOGAR is all based on inferences, we must not imagine that from the outset it was inferences as we make them today, inferences by which we draw a meaningful conclusion from meaningful premises. The proto-inferences grew out of something more like conditional reflexes: we learned to react to various sounds with various other sounds; as this game grew more complex, the roles of the individual sounds grew ever more intricate, till they became what we now call meanings. Thus, meanings and GOGAR bootstrapped themselves into existence, mutually propping each other up. This bootstrapping can also be considered a case of niche construction that we mentioned in Section 9.4. We stressed that our human niche is predominantly a normative niche, and language plays a crucial part in it. Animals not only react to their environment but may also try to change it to make it more accommodating. While other animals do this to only a limited extent, we humans have come to excel at it. And we not only rebuild our material environment in an unprecedented way, we also create what could be called an immaterial environment, the sounds we emit being an important part of it. As Rouse (2015, p. 21) puts it: Language is a persisting public phenomenon that coevolves with human beings. Human beings normally develop in an environment in which spoken language is both pervasive and salient, while languages only exist in gradually changing forms that can be learned and thereby reproduced. Human abilities to acquire and take up the

158  The space of meaningfulness skills and discriminations that enable the ongoing reproduction of that phenomenon are integral to our overall practical perceptual responsiveness to our environment, which has thereby become a discursively articulated environment. The evolutionary emergence of this capacity and its ontogenetic reconstruction in each generation rely on the same close coupling with our discursively articulated environment that characterizes other organisms’ capacities for perceptual and practical responsiveness to their selective environments. Using kinds of sounds as the vehicles of playing GI and GOGAR brings about a process of familiarizing oneself with the sounds. Just like we familiarize ourselves with material tools we use for various purposes, we familiarize ourselves with the immaterial linguistic tools, and just like we come to see the material tools as embodied functions (we do not see a stick sharpened on one side, but rather a spear), we do not hear a particular sound, but rather its role aka meaning (we hear, e.g., an indication of danger). There is, however, a difference between the functions of tools like a spear and the roles of sounds within GI and its successors. For the latter – unlike the former – there is not only the de facto way of using them, but also the de jure way, determined by the incipient normative attitudes. And it is the de jure ways that turn out to be more salient – it is these ways that we call their roles and that we argue develop into what we call meanings. To be sure, for a tool like a spear, there may also be rules for how to use it. But these are “rules of engineering” (Section 5.3) that only mark correctness in the sense of effectiveness. The rules governing the use of linguistic items, in contrast to this, are a matter of merely normative attitudes – to violate them does not mean simply to be ineffective, it means not to play the corresponding game, or to play it wrongly. What is the function of a sound within GI? It is quite straightforward – it is given by a rule of the form (1). There are also rules concerning incompatibility and inference (of the forms (5) or (6)), but these appeared to emerge from the interplay of the basic rules and be fully derivative of them. However, we already saw that the situation is more complicated. Imagine that Tiger and Danger are tied to their respective situations. Those of these situations in which it is appropriate to voice Tiger, we can presume, are perceived as situations in which it is also appropriate to voice Danger. But we have already seen that this does not mean that the latter kind always necessarily involves the former kind. It does not follow that there is no circumstance in which it would be appropriate to voice Tiger, but not Danger. This need not be determined by the set of situations already encountered, not even by their straightforward projection into the future – there may be new situations utterly dissimilar to those already known.

The space of meaningfulness 159 Hence, the rule stating the inference may come to co-constitute the roles of Tiger and Danger in a nontrivial way; it excludes the possibility of a situation in which it would be warranted to utter Tiger, but not Danger; therefore, as long as this rule is in force, whatever is not dangerous is not a tiger. More generally, the roles of expressions will be determined not only by rules linking them to extralinguistic situations, but also by placing them within an expanding web of inferential relationships among expressions. We should say that talking about inferential relations at this rudimentary stage of language is a hyperbole. Inferential relations hold among sentences and have to do with truth-preservation. At the stage where language only consists of sounds like Tiger, Danger or Calm, we have no sentences proper and we cannot talk about their truth. But the hyperbolic talk about one of the sounds being incompatible with another one or being inferable from it is clear. The situation starts to become more complex when we move to GOGAR. The relationships among the sounds which get reinforced by its means are slowly coming to be called – deservedly – inferential without the hyperbole. Some of the sounds start to function as reasons for other ones, and some of them begin to work like challenges to other ones. Thus, the system of sounds starts to look like a (still rudimentary) language consisting of sentences only. Of course, to step from this rudimentary language to language proper crucially involves the incorporation of recursion and differentiating subsentential structures. This, however, is not something we will be dealing with here. What interests us more is giving language the essential structure that may be called logical and the appearance of logical vocabulary. The point is that it is this structure that forms the meanings of sentences into what can be called propositions and what can lead us to a new form of thinking and knowing, thinking- and knowing-that. What is a proposition? What distinguishes it from other kinds of meanings such that it enables a brand new kind of thinking and especially reasoning? It is not easy to say. A kind of answer might be that propositions exist within a “logical space”, or what Sellars (1956, p. 169) famously called the “space of reasons”. It is an entity which, speaking metaphorically, exists within this space like physical objects exist in space and time. Just like what it takes for an ordinary spatiotemporal thing to exist is to be in certain causal relations to other such things, what it takes for a proposition to exist is to be in certain, call them “logical”, relations to other propositions. Every proposition, for example, must have a negation – i.e. there must be another proposition that stands in a certain specific relation to it. Similarly, there must be a proposition that is a conjunction (or a disjunction, or an implication) of any two propositions. Thus, we can imagine that as ordinary things exist in their causal space, propositions exist

160  The space of meaningfulness in this peculiar logical space. This opens up a special chamber of the space of meaningfulness, the chamber of propositional meaningfulness. Wittgenstein famously claimed that language is like a labyrinthine city. Brandom, in contrast to this, argues that it is not a mere labyrinth, that it has a downtown and that it is constituted by GOGAR. As Brandom (2000, p. 14) claims, “inferential practices of producing and consuming reasons are downtown in the region of linguistic practice”. And we now come to see why he considers GOGAR as the “downtown” of language. It is certainly not the language game we spend the most time playing. But it is precisely this game that lets us forge propositions that paves our way to thinking-that. We can say that propositions are successors to what Kant called judgements and what Frege later called thoughts; they are characterized by the ability of being the relata of the relations of inference and incompatibility and being true or false. Shaping meanings of sentences into propositions also brings about the reform of the meanings of parts of sentences. Some of them then have been called concepts. (Propositional thinking then is tantamount to conceptual thinking.) Concepts, thus, are meanings of certain sub-sentential expressions, i.e. components of propositions. As both Kant and Frege stressed, it is the constitutive feature of concepts to be parts of propositions. Therefore, concepts, construed thus, exist too within the “space of reasons”. Thanks to the inferential interrelationships among sentences (and let us now subsume the incompatibility relationships under the inferential kind), sentences acquire what can be called inferential potentials. The inferential potential of a sentence is determined by the sentences from which it can be inferred and the sentences which can be inferred from it (together with other sentences in the role of collateral premises). The meaning of an expression, then, can be identified with its inferential role, i.e. with the contribution this expression brings to the inferential potentials of the sentences in which it occurs. In some cases, the sentences are also governed by rules that link them directly to extralinguistic reality, rules like Asserting “This is a dog” is correct when the asserter points at a dog. We have dubbed such rules quasiinferential because they articulate something that could be conceived of as inferences from the world to language (or the other way around), and we could call the potentials involving such rules quasiinferential. But we will put this aside so as to not complicate the exposition too much. Note that even sentences have inferential roles (that need not be identical with their inferential potentials) because even sentences can be part of other sentences. And both the inferential role and the international potential of a sentence must be distinguished from its inferential significance, which are its consequences given some collateral premises

The space of meaningfulness 161 (Peregrin, 2014a, §3.3). The same sentence can have different inferential significance for different speakers, given the different sets of beliefs of the speakers (which serve them as collateral premises). This constitutes what Brandom calls the perspectival character of concepts and which, according to him, provides for the fact that our concepts allow us to capture the world as objective and independent of us (we will discuss this in greater detail in Section 16.1).7 The inferential relations alone may already be enough for the establishment of a logical space. The point is that we can see a sentence B as a negation of a sentence A iff it is its “minimal incompatible”, i.e. if it is incompatible with A and it can be inferred from any other sentence that is incompatible with A. Similarly, a sentence C can be seen as a conjunction of A and B iff both A and B are inferable from it and any other sentence from which A and B are inferable is inferable from it. The introduction of the logical vocabulary then completes the establishment of the space. Words, such as “not” and “and” (“¬” and “∧” in the usual regimented form), are employed to produce “canonical” negations and conjunctions (and also to make it possible to produce a negation for every given sentence and a conjunction for every given pair of sentences). Thus, the “space of reasons” can exist before there exists any specifically logical vocabulary – this vocabulary merely makes this space complete and especially makes its contours explicit. This may seem to suggest that the role of logical vocabulary is merely auxiliary, but this would be an exaggeration. As Koreň (2018, pp. 214–215) argues, though the role of logical vocabulary indeed is expressive, this does not mean that its role within the constitution of the “space of reasons” is trivial: [L]ogic does not merely recapitulate preexistring inferential relations in a perspicuous idiom. Rather, by helping to establish norm-governed [GOGAR], it both shapes the very space of reasons and makes it increasingly articulate. This is consistent with it being the case that prelogical practices involve rudimentary intentionality and protolanguage games. Hence, it does not follow that logic forges completely new links between utterances and expressions where there have been none before, conferring content upon them from scratch. What elaboration of logical devices makes possible is individuation of full-fledged propositional contents relative to a space of reasons held in a common ground. I think this is an accurate description of the situation. Anyway, this construal of the role of logical vocabulary suggests a very unusual way of understanding logic: the expressivistic understanding, which will be discussed in the next chapter.

162  The space of meaningfulness 12.4 Truth In Section 8.2, we have seen that the criterion of being correct does not always derive directly from the common agreement on what is correct, but rather from the common agreement on the methods of how to show what is correct. This is the case of justification: a common agreement does not decide what is justified, but what is an acceptable justification of a claim. It does not decide who is the winner of an episode of GOGAR, but what are the rules the participants must observe. We have seen that the fundamental move of GOGAR is assertion. (And that a primitive version of GOGAR can be played solely in terms of [proto-]assertions.) When one makes an assertion, one commits oneself to justifying it when it gets challenged. If she possesses such a justification, then we can say that she is entitled to it. Such an entitlement is conditioned by the fact that the justification really exists, and in such a case, we can say that the assertion is correct. Now truth, as a crucial concept of GOGAR (and perhaps of other language games) amounts precisely to such a correct assertability. To extend the parallel with chess, compare asserting, on the one hand, with checking the opponent’s king, on the other. You can check with different pieces, in different ways, but only as a result of a correct move. Similarly, in GOGAR, you can assert various sentences in various circumstances, but only some such assertions are correct (namely those that may be justified). This is a concept of truth basically proposed already by Sellars (see, e.g., Sellars, 1968, p. 101). Saying that truth is correct assertability appears to endorse the so-called epistemic theory of truth. Such theories have been elaborated especially during recent decades, and they derive the concept of truth from that of justification. Their roots, however, go back at least to Peirce, according to whom truth is the ideal limit of justification (Misak, 2004); what is true is what is justified in the ultimate end of inquiry. As, however, the concept of “end of inquiry” is far from transparent (and it is not quite clear that it is coherent), it may be better to construe the link between justification and truth in different terms. Probably the most widely discussed philosopher deriving the concept of truth from that of justification (and thus endorsing the epistemic theory) is Michael Dummett (1959), who repudiated the correspondence theory of truth (endorsed by most of his contemporaries) as untenable. However, he insisted that we should not abandon the intuition that if a statement is true, then there is something in virtue of which it is true. But in contrast to the correspondence, he took the “something” to be nothing as a fact, but rather a justification, or a proof of the true statement in question. This leads him to adopting his intuitionistic standpoint claiming that “we are

The space of meaningfulness 163 entitled to say that a statement P must be either true or false, only when P is a statement of such a kind that we could in a finite time bring ourselves into a position in which we were justified either in asserting or in denying P; that is, when P is an effectively decidable statement” (p. 16). Dummett thus wants to tie truth with justification, but of course he cannot equate being true with having been justified – we fallible beings may fail to find a justification, even if it is easy. Hence, what is needed is something like being justifiable. This allows us to say This is not justified but perhaps it is true and indeed also This is justified, but perhaps it is not true (reflecting that we can mistake something for a justification). Rorty (1991) calls this the “cautionary” use of true. We know that however good a justification we can produce, there is always a possibility that it contains a mistake and that it will later turn out that it is not really a justification. (On the other hand, from the fact that we cannot find a justification we can never derive, with an absolute certainty, that none exists.) This is simply our human predicament. It is here that we see the birth of our complex concept of truth out of that of justification. Something may be true despite everybody thinking it is not true. This is because the criterion of being true does not derive from the common agreement on what is true, but rather from the common agreement on the methods of how to reach truth. And as I agree with Sellars that truth is correct assertability, this is just a special case of the fact that the criterion of being correct is not always common agreement on what is correct, but rather common agreement on the methods of how to find out what is correct. Anyway, I find it imperative to stress that if we want to call the Sellarsian approach to truth that I endorse an epistemic theory, then we must realize that the notion of justification it is based on is radically “non-solipsistic” in the sense of Sturgeon (1991). To clarify this, consider Wikforss (2001, pp. 205–206): Consider the case where I misperceive and utter “That’s a horse” of a cow. What semantic norm do I then violate? I see the animal, believe it to be a horse and, consequently, utter “That’s a horse”. Although I have made a false judgment, I have not broken any semantic norms. According to the notion of correctness entertained here, to assert That’s a horse when pointing at a cow is violating the rules of our language game, namely the rules determining which sentences are correctly assertable when. True, if we were to understand justification “solipsistically”, we might perhaps say that after the misperception, I was justified in asserting what I asserted and hence the assertion was warranted (for me) and hence correct (from my viewpoint). However, the concept of correct assertability

164  The space of meaningfulness I entertain disregards any subjective point of view; it is an entirely objective (or at least intersubjective) business. The assertion is correct if the assertor may possess a justification, which, in this particular case, amounts to showing that the item pointed at is a horse. It is not enough to believe that there is a justification (and it is not necessary to possess the justification) – just like to correctly check an opponent’s king, it is not enough to believe that one is making a legal move. But there is a difference between chess and GOGAR: in the case of the former, unlike that of the latter, every player who knows the rules is usually able to recognize whether a given move, in a given position, is correct (and sometimes it is not possible at all). As GOGAR may be quite labyrinthine, it is not always easy to find out which moves are correct, and hence it isn’t easy to determine which assertions are true. Hence, a great deal of problems around GOGAR concern the correctness of the moves. Thus, though we may know the rules of GOGAR very well, and though we may think that we are able to judge the correctness of a move (e.g., the justification of an assertion), there is always the possibility of an ignorance, mistake or an oversight – there is the possibility that the move is correct despite our firm conviction that it is not, or vice versa. There is always the possibility that what is in fact correct does not align with what we think is correct – that what is in fact justified and hence true is not exactly what we take to be justified. Note that truth, conceived in this way, is secondary to justification and hence secondary to inference. However, is not inference meant to track truth-preservation, and is it not so that inference must be secondary to truth? No, not if you accept the order presented here in which these concepts come into being. But note that this does not mean that we cannot say that inference tracks truth-preservation: if truth is correct assertability, or justifiability, then this only means that inference tracks assertability- or justifiability-preservation. And this, indeed, is (trivially) true: if B is inferable from A and A is assertable or justifiable, then B is such too. There is one more thing we should mention. Brandom endorses a version of the deflationary theory of truth and gives the impression that it is this theory of truth that goes with inferentialism. Is this another point in which we part ways with him? I think, not really. Given the “non-solipsistic” nature of justification (and correctness more generally) on which the Sellarsian conception is based, it is not incompatible with deflationism. Indeed, Sellars (1962b, p. 38) himself admits that “the word ‘true’ gets its sense from this type of inference” (viz. the T-sentences understood as bilateral inferences). And, importantly, he adds: “we must say that, instead of standing for a relation or relational property of statements (or, for that

The space of meaningfulness 165 matter, of thoughts), ‘true’ is a sign that something is to be done-for inferring is a doing” (cf. Shapiro, 2020). 12.5 Summary An all-important rule-constituted space that we have established is the so-called space of meaningfulness, the space constituted by the rules of our language games, in which we can produce meaningful utterances. To avoid “the problem of circularity” (that a theory of the emergence of language and meanings already presupposes meanings, and without such a presupposition it is difficult to make sense of it), we assume that the rule-governed practice evolved from mere habitual emittings of sounds in reaction to other sounds. No content, no meaning is present at the start. What then happens is that the habitual links between kinds of sounds are turned normative (especially are turned into inferences) by means of the emergence of normative attitudes. And with the growing complexity of such normative entangledness of the various sounds (potentially imbued with empirical content because some of the rules that govern them incorporate the world), the roles of the kinds of sounds become what we now call meanings. Thus, meanings and our language games bootstrapped themselves into existence, mutually propping each other up. This bootstrapping can also be considered a case of niche construction: our human niche is predominantly a normative niche, and language plays a crucial part in it. Languages tend to acquire a certain kind of structure, which forms the meanings of sentences into what can be called propositions and what can lead us to a new form of thinking and knowing, thinking- and knowing-that. This is effected especially by one of the “games” we play with language, GOGAR. And it is also in the context of this game that the concept of truth is conceived. Truth, according to Sellars, is “correct assertability”, and as GOGAR is incredibly complex, truth comes to look deeply enigmatic – to find out what is correctly assertable, and hence true, is often beyond our ken. Notes 1 A very detailed anatomization of the approach of Mercier and Sperber, as well as that of Tomasello and his school, with a focus on the various ways in which they can be seen as possibly circular, is given by Koreň (2021). 2 We have already seen that the inferential roles of empirical expressions cannot be based only on inferential rules in a narrow sense as they must also involve rules like “This is a dog” is only correctly displayed when you point at a dog. 3 There will most probably be other games interwoven with GI. Thus, atop of it there may be the game of warning, played with sounds like Tiger or Danger, with the additional rule: One should emit A always in circumstances C.

166  The space of meaningfulness 4 Quine talks about children entering language by way of ontogeny, rather than about the ur-speakers entering it by way of phylogeny. 5 But of course, they are still merely “proto-sentences”, not sentences in the standard sense of the word. 6 This construal of the function of logical words leads to their “expressivistic understanding” (Brandom, 2000; Peregrin, 2008b, 2014a, Chapter 9). 7 See Drobňák (2022) for a discussion of accounting for linguistic understanding in terms of inferentialism.

13 Logic

13.1  Material inferences and logical vocabulary The inferences that we portrayed as the most primitive, such as that from Tiger to Danger in our proto-language, are what we already called material (cf. Sellars, 1953a; cf. also Read, 1994). This means that they have as of yet nothing to do with logic. They interconnect “occasion sentences”, which grew out of what we called the game of indication. And we saw that it is no mystery how these relations come into being. To infer Danger from Tiger is probably reasonable, but there certainly may be animals that appear to be dangerous and are better avoided (“better safe than sorry”), which, however, later turn out to be more of a potential source of meat than danger. In such a case, to infer Danger from the indication of such a creature may be at first useful, but then ever less so. How can we discard an inferential rule that is in force? Of course, we can cease to follow it and cease to display the corresponding normative attitudes, but then we are likely to be seen as making mistakes, and to make it clear that this is not so may be a tedious process liable to a lot of confusion and misunderstanding. It would be much better if we could articulate the rule, convince our peers that it is to be rejected and reject it explicitly. Two things are needed for this to be possible. First, we need to have some means to make rules explicit, to articulate them as assertions. Second, we need to have the GOGAR up and running, to be able to give reasons for and against such explicit rules. We have already seen how (a rudimentary form of) GOGAR could have emerged as a superstructure of GI. However, so far we have considered what kind of linguistic means can provide for the explicitation of rules only cursorily. What we are going to argue is that such means are precisely those expressions of natural language which we tend to see as “logical”. The most basic tool we need is a connective akin to the English if-then: with its help we can build an explicitation of the inference from Tiger DOI: 10.4324/9781003388876-14

168 Logic to Danger: If tiger, then danger. Once we have this (proto)sentence, which can be asserted, we can argue, within GOGAR, for or against it. If we have a particle akin to the English not, we can assert If tiger, then not danger, which can be used to challenge it. Other “logical words” of natural language can be used to refine the explicitations: Thus, e.g., we can use and to explicitate inferences with multiple premises (If animal and raging, then danger), or we can use every to explicitate inferences from and to sentences already in the subject-predicate form (Every tiger is dangerous). Note that the new vocabulary is very different from the earlier one. While the function of words like Tiger or Danger was determined above all by the linkage between them and certain circumstances, the function of the new words is determined very differently. Take if-then. Its function is to produce sentences that are correctly displayed when certain inferences are correct, in particular if A, then B is correctly displayed iff B is correctly inferable from A, viz. when

A B

is a correct inference. Of course, as the correctness of the inference is likely to be derived from a certain regularity of the world (the correctness of the inference from Tiger to Danger is derived from the fact that tigers tend to be dangerous), the if-then sentence may be seen as expressing this fact about the world. But the fact is already expressed by the language, albeit implicitly: we reflect it by what we do (the material inferences), though not yet by what we say. And the if-then sentences allow precisely for this: for making the implicit explicit. Thus, while we can see Tiger or Danger as indicators of certain empirical circumstances, If tiger, then danger is to be seen as an indicator of the correctness of an inference. (To be sure, we can even take If tiger, then danger as indicating a circumstance, namely the circumstance that Danger has come to be inferable from Tiger, and consequently that tigers are dangerous. But this is a genuine circumstance if we view the language “from outside”. Viewed “from inside”, it is a necessity that cannot fail to be in place (as we have made it into such, subscribing to its rules). From the outside of English, the fact that This is an animal is correctly inferable from This is a dog is contingent and empirical, whereas for a speaker of English, it is constitutive of the game she plays and hence is inescapable – unless she quits the game.) More generally, if we allow for a list X of collateral premises, we have

If

X

A B

then

X if A, then B.

Logic  169 And as we do not want If A, then B to hold in case when B is not inferable from A, we have, conversely,

If

X if A, then B

then

X

A B.

And hence

X

A B

X if A, then B.

iff

This spells out the idea that if-then turns an inference into a statement. With the onset of logical vocabulary, the inferential landscape of the language becomes much richer and more exciting. All of a sudden, we can encounter nontrivial problems regarding what is inferable from what. Is, for example, If A, then B inferable from B? Given the inferential role we have assigned to if-then (and assuming some apparently evident properties of inference), it turns that it does. If we take X to consist of just B, we have

If

B

A B

then

B if A, then B

and as the antecedent holds, so does the consequent. This inference may seem strange and it may lead to the refinement of the inferential role of if-then. When we take X to consist just of If A, then B, we have

If

if   A, then   B then if   A, then   B

if A, then B B.

A

And again, as the antecedent holds, we have the inference known as modus ponens:

A

If   A, then   B B.

To be clear, though we discuss a possible (inferential) definition of ifthen, we are not aiming at establishing its meaning in terms of explicit definitions. We conjectured that a connective of this kind might have been necessitated, in the pre-history of language, by the need to help make its inferential structure explicit, and might have been brought into existence and underpinned by the speakers’ normative attitudes, which we are now trying to reconstruct in terms of explicit definitions. The fact that the English if-then (and similar connectives in other languages) is quite close to this appears to corroborate our conjecture.

170 Logic The new, logical inferences brought about by the introduction of logical vocabulary are very varied. Thus, if we have if-then, we can, for example, infer If tiger, then getaway from If tiger then danger and If danger then getaway, hence we have the rule If A, then B

If B, thenC If A, thenC.

If we have or and not governed by suitable inferential rules, we can infer Tiger from Tiger or lion and Not lion; hence, we have the rule known as disjunctive syllogism:

A or B

not B A.

Viewed thus, logical inferences are secondary to material ones, for the former are here as a by-product of the introduction of logical vocabulary, which, in turn, is here because we profit from having the latter in explicit forms. This contrasts with the more usual view that logical inferences underlie the material ones: indeed, that material inferences are only logical ones in disguise (see Peregrin & Svoboda, 2017, §2.2). Note that when we have the logical particles that let us make material inferences explicit and we have new inferences facilitated by the logical particles, we can make explicit even the very constitutive features of GOGAR. And GOGAR is what allows us to argue for or against – and eventually to uphold or to retract – a statement. But rejecting a statement explicitating a rule constitutive of GOGAR looks like scoring an own goal. We have, for example, taken for self-evident that an inference of the shape

B

A B

is correct; and hence, we accept its explicitation:

If A and B, then A.

Now somebody may argue against this, say in the following way: it is clear that If A, then A. So there is no point in adding the and B, and, indeed, if we add it, the sentence would not be reasonably seen as true, for it is not A and B that make A true, it is A alone. This may lead to a different kind of implication, no longer explicitating inference; or it may lead to the rejection of the prima facie correct inferences, thus changing what looks like a solid rule of GOGAR. Is this possible?

Logic  171 It turns out that while the rules constitutive of GOGAR may be tampered with only with extra care, it is not the case that they are untouchable. If we are able to amend a principle at the same time that we reject its explicitation, a modified version of GOGAR may be viable. We may, for example, reject the rule

B

A B.

And, indeed the principle

If

B C

then

B

A C

obtaining a version of GOGAR that is still feasible.1 I have discussed elsewhere some technical details of what minimum of logical vocabulary is needed to make all kinds of inferences explicit (Peregrin, 2008b), but this is not important for us here. (I reached the conclusion that the toolbox of logical constants fulfilling the requirement of explicitation of material inferences most straightforwardly corresponds to those of intuitionistic logic.) What is important here is that we may understand the rationale of the presence of logical vocabulary in natural language and, as a consequence, see the vocabulary as a kind of adaptation – an expedient providing for our control over the rules that govern our language games. 13.2  Formal logic This understanding of the logical vocabulary of natural languages paves the way for a thoroughly naturalistic understanding of logic, in particular an understanding of logic as a certain superstructure built on the basis of a proto-language. What usually stands in the way of such an understanding of logic is the conviction that logic cannot be so closely linked to natural language: that while the rules of natural languages are contingent, imprecise and haphazard, those of logic are necessary, rigorous and systematic. Logic thus has, so the story goes, more to do with our reason than with our language; it is not something that submits to empirical investigation. As, e.g., Glanzberg (2015, p. 114) puts it: “we do not find logical consequence relations in our natural languages”. Thus, the expressivistic account of logical vocabulary leads to a picture of logic that is not really compatible with the usual received wisdom concerning its nature. (Though we should say that various naturalistic construals of logic have recently been gaining importance.2) According to our picture, moreover, reason comes after language. And indeed what I

172 Logic am after is challenging the traditional construal. The road to logic as we understand the term today started, I insist, by the emergence of logical vocabularies in natural languages. True, the rules governing the items of such a vocabulary indeed were – and still are – contingent, imprecise and haphazard, so it truly was not the necessary, rigorous and systematic logic as we would think of it today. But the reason was not that the logical structure of natural languages was only a distorted shadow of a “true” structure that existed beyond the language and must have been excavated, but because the very necessary and rigorous logic came into existence only much later. We saw that logic is often considered something that has little to do with natural language: while the laws of logic are usually considered absolute and non-negotiable, what we find in natural language is seen as at most their feeble rehash. Thus, while the law of modus ponens,

(MP )

A→ B

A B,

is exceptionless, its closest counterpart in English,

( MP − Eng )

A

If A, then B B,

cannot be guaranteed to be such.3 However, according to our evolutionary story, (MP-Eng) is not an imperfect embodiment of a supernatural (MP), but rather (MP) grew out of our effort to account for (MP-Eng) in terms of an idealized theoretical model. From this viewpoint, logic, as we know it today, is our creation. It resulted from a two-step process: the first was the emergence of logical vocabularies of natural languages (as discussed in the previous section), while the second was our systematic effort to account for the structures emerging in the natural language thanks to this vocabulary. But do we not have to obey the laws of logic if we want to think properly, and hence are they not something independent of us that furthermore wield authority over us? Well, we must obey the laws of logic if we want to use certain extremely useful “cognitive gadgets” – such as implication and the hypothetical mode of thought going hand in hand with it. In the same way, we must obey the “laws of cycling” to ride a bicycle and not fall from it, but this does not mean that bicycles and cycling are not, in fact, our creations.4 The point is that after we became able to assume a reflective attitude toward our linguistic and inferential practices, we became able to formulate their explicit theories and indeed form their idealized models (Peregrin,

Logic  173 forthcoming). These models were sometimes perceived as upgrades on the natural languages that were their sources, and though it never happened that our reasoning largely moved from natural languages to these artificial ones, the artificial languages were perceived as a standard of rigor and clarity largely surpassing what we can encounter in natural languages. Logical constants of artificial languages and calculi were established by means of explicit rules embodying the principles of the kind hinted at in the previous section. There we pointed out that they were only idealized versions of the implicit rules that came to emerge in natural languages in a much more indeterminate and blurry way. Now the constructors of the artificial languages were free to bring the ideals to life (though there have turned out to be many competing proposals which their respective proponents held for ideal). One aspect of this movement was that constants of the artificial languages were used as tools of regimentation of the natural ones. In this way, logicians introduced, for example, a sign like “→” to upgrade the English “if-then”, the functioning of which, unlike that of “if-then”, is exactly stipulated in terms of a definition. In case of “→” it might be, for example, the notorious truth table stating that a complex sentence formed out of two sentences by its means is true if and only if the first sentence is false or the second one is true. If we want to bypass the concept of truth, we may resort to our definition from the previous section and say that the complex sentence is correctly assertible if the second subsentence is inferable from the first one. But a consequence of the movement was that the artificial languages started to appear as embodiments of the necessary, rigorous and systematic logic that cannot be found in natural languages. And then there was the idea that this is the “pure” logic that predates any contingent reasoning and any natural languages, for their (imperfect, but still) logical capabilities must be acquired by their participation in the true pure logic. And the recent formal languages of logic came to look like tools for the exact pinning down of this pure logic. 13.3  Logic out of GOGAR Logic is often considered giving directions for reasoning in the sense of a process taking place inside the mind. True, there is also a reasoning (or “reasoning”?) going on in the open – argumentation – but this is considered just a matter of making the “true”, private reasoning public and perhaps of its interindividual confrontation. It is important to see that the picture I offer not only challenges the notion of logic as something superhuman, but also its picture as – primarily – a matter of reasoning in this sense.

174 Logic Modern logic mostly converged on the aim of helping build our theories of the world, conceived as axiomatic systems. The goal was to pinpoint all the rules of derivation by which we can get from sentences codifying what we already know to those expressing new information entailed by it. Moreover, as the rules of derivation that logic was pursuing are also expressible as sentences (“logical truths”), logic itself was conceived as an axiomatic system; the goal was to reach a few axioms and rules of derivation from which the whole of logic would be derivable. (The pathway to this was prepared by the move of logicians from natural languages to the artificial ones discussed in the previous section.) It soon turned out that not all logicians are able to converge on a single, “genuine” logic and a lot of different logics started to flourish;5 however, in every (or almost every) case, the logics were presented as axiomatic systems.6 This supported the impression that logic is a matter of extending one’s knowledge, of getting from what one already knows to further knowledge. There were only rare attempts to capture logic as an interpersonal – rather than intrapersonal – enterprise. Probably the most significant of them was the so-called dialogical logic developed by the German logician Paul Lorenzen and his colleagues (Lorenzen, 1955; Lorenzen & Schwemmer, 1975; Lorenzen & Lorenz, 1978). Within this approach, logical formulas are associated with dialogical games, with exactly delimited rules, between a proponent and an opponent such that the (logical) truth of a formula is defined as the existence of a winning strategy in the corresponding game. Here logic is presented, from the start, as a dialogical, argumentative business (Rahman & Rückert, 2001). We have depicted logic as a product of GOGAR. Its mission results from the fact that it is useful to have an explicit articulation of linguistic rules, and hence it is useful to have means for the explicitation of inferences. Far from participating in some heavenly principles of logic, the job of the logical vocabulary of natural language is rather down-to-earth: to help us say what we were only able to do before. Thus, our conjecture is that logical vocabulary originated in the public space and so did any reasoning. Remember that we envisaged the birth of GOGAR, and hence of reasoning, without presupposing any private version of reasoning – not as an externalization of any private mental processes. Argumentation was rather depicted as resulting from emerging rules for reacting, by sounds, to outer circumstances and to other sounds. There is, therefore, no circle involved: private reasoning presupposes the pubic one (argumentation), but the latter does not presuppose the former. Hence, in contrast to the more usual view, I am convinced that public reasoning did not arise from the need to externalize the private one, but rather that private reasoning stemmed from the internalization of the

Logic  175 public one. Though uncommon, this view is far from unprecedented or without support. One of the theories which agrees with ours in respect to the primacy of public reasoning over the private one is that developed by Hugo Mercier and Dan Sperber. Mercier and Sperber (2011, p. 72) write: We view the evolution of reasoning as linked to that of human communication. Reasoning, we have argued, enables communicators to produce arguments to convince addressees who would not accept what they say on trust; it enables addressees to evaluate the soundness of these arguments and to accept valuable information that they would be suspicious of otherwise. Thus, thanks to reasoning, human communication is made more reliable and more potent. Mercier and Sperber (2017) further argue that the original point of reasoning was not helping us to make better decisions and more profitable dealing with the world around us. It was, they insist, social coordination and in particular defending one’s behavior if it gets out of social control. This explains why our reasoning is prone to make basic and systematic errors (as has been clearly and repeatedly documented – see, e.g., Kahneman, 2011). If the point of reasoning were extending our knowledge and making better decisions, then the deductive mode would already have been groomed to relative perfection, while the occurrence of elementary errors shows, according to Mercier and Sperber, that this kind of deduction is not what reasoning was selected for (which is not to say that it is not something for which it can also be used). A similar account of reasoning is given by Dutilh Novaes (2020): she tries to pinpoint the role of deduction, which she understands as a stepwise, truth-preserving process proceeding from a premises to a conclusion, in human affairs, and comes to the conclusion that this mode of reasoning is quite extraordinary, for it “does not seem to be a particularly suitable way to produce new information, given that it is non-ampliative, and it does not seem to be a reasonable guide for managing our beliefs and thoughts either” (p. 21). Like Mercier and Sperber, she concludes that the primary locus of reasoning is dialogue and argumentation. Dutilh Novaes, however, disagrees with the view of Mercier and Sperber. Concerning it, she writes: “Reason must be an adaptation, but if conceived as having the function of supporting the cognitive processes of the lone reasoner, it does not seem to perform this function very well. So, there must be a different function that reason is in fact responding to, given that it cannot be anything other than an adaptation” (p. 193). And further: “Prima facie, to argue for the adaptive nature of reason seems like a tall order in view of the numerous empirical findings suggesting that human reason is ‘biased and lazy’. […] But Mercier and Sperber go on to argue

176 Logic that these two features are in fact advantageous for the function of reason as socially conceived” (ibid.). I think that we must distinguish between two varieties of “reason as socially conceived” (or argumentation). Let me call the first of them “Socratic”: this is the argumentation which aims at an impartial seeking of truth, where reason acts “as a judge” (to use the metaphor of Haidt, 2001). The other variety of argumentation is “sophistic”: this aims at defending one’s pre-given views, cost what it may (it acts as a “defense lawyer”). And in my view, Mercier and Sperber do not claim that Socratic argumentation is an adaptation. As I read them, what they claim is that it is a kind of sophistic argumentation that is an adaptation, and that Socratic argumentation is its by-product. Defending one’s position (cost what it may) is an adaptation, and seeking flaws in another’s defense is a countermove to this adaptation. It is only when these two adaptations are played against each other that “Socratic” argumentation may arise. But fortunately, we need not delve into the quarrel of Dutilh Novaes with Mercier and Sperber, for what is important for us is what both these accounts share: namely the conviction that reasoning as a private effort of improving one’s informational status by drawing some yet unknown conclusion from already known premises is secondary to reasoning as argumentation. 13.4 Summary We can imagine that as it is desirable to be able to talk about rules, there appeared some means of making the rules explicit, of articulating them as assertions. Such means are precisely those expressions of natural language which we tend to see as “logical”. Once we have logical vocabulary, we can not only infer, say, C from A and B, but we can also say that it is so inferable: if A and B, then C. Aside from such logical vocabulary of natural languages, logicians, during the last century and a half, produced artificial regimentations of the natural logical vocabulary and subsequently constructed whole artificial languages consisting of such regimented vocabularies. These languages were sometimes perceived as upgrades on the natural languages that were their sources, and though it never happened that our reasoning largely moved from natural languages to these artificial ones, the artificial languages were perceived as a standard of rigor and clarity largely surpassing what we can encounter in natural languages. However, we propose to see the artificial languages as models in natural sciences – as idealized, purified and streamlined representations of a natural phenomenon, of natural language as the vehicle of “the game of giving and asking for reasons”. Hence, logic is not a mold into which we must squeeze our thinking and our language in order to make them

Logic  177 work optimally – “rationally”. Logic is one of the spaces we have built out of our rules. (It is not really a self-standing space, it is entrenched in the space of meaningfulness opened up by the rules of our language.) It follows that logic as a “public business” (interpersonal argumentation, as manifested by GOGAR) was prior to logic as a “private business” (intrapersonal reasoning). Notes 1 The result may be some version of the renowned relevant logic (Mares, 2004). 2 Most of them now go under the heading of “anti-exceptionalism” or “abductivism”; see Hjortland (2017), Hlobil (2021) or Martin (2021). See also Peregrin (2019) and Peregrin and Svoboda (2021). 3 Alleged counterexamples to (MP-Eng) have been presented, e.g., by McGee (1985). 4 Elsewhere (Peregrin & Svoboda, 2022), we argued that logic is a tool and that the laws of logic are nothing like natural laws, but rather something like the “rules of carpentry”. 5 See Cook (2010). See also Priest (2001) for an overview of some alternative systems. 6 Some of them being “Hilbertian” (their axioms are forms of sentences which are deemed to be “logical truths” and their rules of derivation derive sentences from sentences), others “Gentzenian” (their axioms being basic valid argument forms and their rules of derivation derive arguments from arguments). See Peregrin (2020b).

14 Cooperation and morals

14.1 Cooperation That we humans owe what we are to cooperation is something that has already popped up many times in this book. We are social animals, but this, of course, is not what makes us stand out from other social species. There are a lot of social animals, such as ants, wolves or our closest relatives, the great apes. What makes us stand out is the unprecedented extent and intensity of our cooperation. No other animal species has cooperated so as to produce any counterpart of our cumulative culture. No other species produced medicine to cure diseases, airplanes to fly in or books to read. What is even more striking is that the time needed to alienate ourselves to such an extent from the rest of nature was extremely short. It is only some eight odd million years that we have existed as a self-standing species. This may seem a long period of time; from the viewpoint of evolution, it is, however, negligible. After all, the great apes with whom we shared our last common ancestor before the parting of ways have not changed that much since that time. Most anthropologists nowadays conjecture that it was the need for closer cooperation during a certain period of our history that started our hasty development toward the unprecedented “ultra-social” species we are now. Thus Sterelny (2021, p. 15) notes: Perhaps sometime in the late Pliocene, near Pliocene/Pleistocene boundary, hominins gradually evolved a new lifeway centred on collaborative foragers targeting high value resources (Thompson et al., 2019). These resources are typically hard to find, heavily defended, or both. Animals defend aggressively; have vigorous and well-honed escape routines; hide, are camouflaged; rest in inaccessible places. Plants defend themselves mechanically (with thorns and shells) and chemically. Harvesting such resources depends on a blend of cooperation, technology and expertise. DOI: 10.4324/9781003388876-15

Cooperation and morals 179 The fact that it is intensified cooperation that marks the point of departure of the species Homo sapiens with all its specific features has profound consequences. If this is true (and I think it is), then, as we already indicated, it is not the case that we started to cooperate because evolution had made us smarter; it is rather the case that we became smarter because we had come to cooperate. Also, the nature of our smartness has to do with the requirements of cooperation. Our language and our reason have been molded by the needs of collaboration – hence, they both have the imprint of teamwork.1 What is the relationship between cooperation and rules? As a matter of fact, it turns out to be quite close. On the one hand, rules require some form of cooperation. We have stated that the form of an incipient rule may be just the resonation of normative attitudes, which, as such, does not yet seem to require any cooperation – it may come about just randomly. But for the resonance to remain at least minimally stable, some form of cooperation is needed, and this need is stronger once rules get any coordinated support. (And we also saw that some philosophers insist that even at the very inception, a mere resonance is not enough – that already there we need something as “collective intentions”.) On the other hand, cooperation would always have to be based on something like rules, or at least proto-rules. We saw that cooperation, in a nontrivial sense, means an investment, which is expected to give a return, with interest. But it works only if the investment does give a return – at least more often than not. Hence, returning the investment must be something that is taken to be an “ought to be” (in Sellars’ sense) – it must be taken as a rule. Therefore, cooperation and rules need each other, to bootstrap themselves into existence by propping each other up. We will see, in the next section, why there is a problem with the emergence of cooperation – free riding, i.e. accepting the investments of others and not returning them, pays off. Hence, there is a problem which impedes any rule-following. A rule, as with cooperation in general, typically involves some cost – it makes me do what I would not do otherwise. And this cost is acceptable because it brings about some profit – the enterprise of following the rules brings me some advantages. Thus, the problem of free riding is generally relevant for rules following as it is for cooperation. 14.2  Game-theoretical models Recent theories of cooperation suggest that its emergence has to do with overcoming the Prisoner’s Dilemma. Indeed, the dilemma is the most basic hurdle on the road to a genetic support for cooperation. It may be useful to review the problem of the emergence of cooperation as it is usually presented by means of simple game-theoretical models that are usually employed to pick up its core.

180  Cooperation and morals We (and many other species) do cooperate; hence, we are somehow genetically disposed to cooperate (maybe somewhat indirectly). Yet, at the dawn of evolution, there were no animals capable of cooperation, so the genetic support for cooperation must have emerged somewhere along the way between the dawn of humanity and our time. And evolution theory teaches us that it must have emerged so that the mutated individuals who displayed the tendency to cooperate were more fit than those who did not. And the gametheoretical models usually concentrate on the encounters between such mutated individuals and members of the non-cooperative establishment. The models therefore represent a confrontation between an organism disposed to cooperation and one not so disposed, examining who will come out of this confrontation as a winner. (We must keep in mind that what is modeled here is not an encounter between two intelligent creatures who choose their strategies, but rather between two simple organisms embodying the strategies.) And the result is disturbing, for it turns out that the cooperating organism gets the short end of the stick. Therefore, it was difficult to explain how cooperation could have expanded: it seemed that whenever there appeared a mutated individual disposed to cooperation, it must have been wiped out by the not-so-disposed establishment. Hence, the birth of cooperation is a kind of mystery; the general belief is that when we find out what makes it possible to break the barrier of the Prisoner’s Dilemma, we will understand a lot about the nature of cooperation. Imagine, quite generally, a confrontation between two organisms, one of which is disposed to cooperate (C), while the other is disposed to renounce its cooperation (R). Let us use the notation common in game theory (Binmore, 2007; Leyton-Brown & Shoham, 2008): we assume that the selection of strategies of the two players of the one-step game leads to a result which can be quantified so that each player is assigned “points”. (In this case, they are supposed to quantify the increase of fitness resulting from the confrontation, for example, when the modeled situation is that of splitting a quarry, it may be seen as “units of energy” contained in the part of the quarry acquired. In any case, the more the points, the better). Every cell of Table 14.1, in the intersection of a row and a column, gives the upshots of the first and the second player if the first chooses the strategy corresponding to the row, and the second that corresponding to the column: Table 14.1  Cooperation vs. rejection

C R

C

R

a,a c,b

b,c d,d

Cooperation and morals 181 Thus, upon the confrontation of two cooperating organisms, each of them gets a; upon that of two non-cooperating ones, each gets d. If a cooperating organism is confronted with a non-cooperating one, the former gets b, and the latter c. To the extent that the game is to model the confrontation of the two strategies, the numbers a, b, c and d cannot be just random, they must fulfill some restrictions. It would seem, first, that whatever kind of cooperation is going on, a cooperating organism should get more when it is confronted with an organism that reciprocates (i.e. is also cooperating) than when it is confronted with one that rejects cooperation, hence

a > b.

Similarly, it seems that two cooperating organisms should be better off than two non-cooperating ones; hence

a > d.

When a cooperating organism is confronted with a non-cooperating one, it is the latter one that should get more, for the former one invests something into the cooperation and it will be wasted. Hence, we have

c > b.

And finally a non-cooperating organism should not get less when it is confronted with a cooperating one than when it is confronted with another non-cooperating one; hence

c ≥ d.

What about the remaining relationships? Here clear restrictions no longer apply. Imagine, for example, that the cooperation concerns a project such that if it really comes into being, it brings to both the cooperators a benefit which would not be attainable for either of them alone. Then

a > c.

On the other hand, imagine that what is in question is the splitting of a kind of quarry. Then we can expect that during the confrontation of a cooperating individual and a non-cooperating one it is the latter that usurps the whole quarry, i.e.

c = 2a

182  Cooperation and morals and hence

c > a.

The situation is similar to the relationship between b and d. If we stay with the example of the splitting of a quarry, then if the confrontation of two non-cooperating organisms means that they will fight over the quarry, there can be two outcomes. One possibility is that though the fight will mean a loss, the loss will not exceed the gain of the part of the quarry they acquire through the fight (and we can assume that on average it will be one half), then it will be

d > b.

However, it can also happen that the fight will be so fierce that the loss effected by it will exceed the gain and it will hold that

b < d.

Thus, we have a couple of variants of games that can be considered when building models of the confrontation of strategies of cooperation and non-cooperation. Let us look at them one after another. First, we consider Table 14.2 summarizing the game with a>c and d>b: Table 14.2  Stag Hunt

C R

C

R

2,2 1,0

0,1 1,1

This is a game known as Stag Hunt (Skyrms, 2004). A legend to this might be the confrontation of two hunters, where C would be the intention to hunt a stag, which can be accomplished only together, while R would be the intention to hunt a hare, which can be done alone. This game is basically unproblematic because in it cooperation appears to be reasonable – mutual cooperation (hunting a stag together) constitutes the so-called Nash equilibrium (which is a state in which nobody can gain by unilaterally changing her strategy). Note, however, that this does not mean that cooperation would be a strategy optimal in the sense that it would bring the highest gain whatever the opponent does (mutual non-cooperation is also a Nash equilibrium). If I decide to cooperate

Cooperation and morals 183 (to hunt a stag), but the opponent decides not to (she is after a hare), my gain is smaller than if I were to be non-cooperative (I miss the opportunity to hunt my own hare). Let us now look at a variant of this game, with c>a, presented in Table 14.3: Table 14.3  Prisoner’s Dilemma.

C R

C

R

2,2 3,0

0,3 1,1

This is the well-known Prisoner’s Dilemma. Within the framework of modeling evolution, a similar game is usually called Hawks and Doves (McElreath & Boyd, 2008), though for the latter, in contrast to the Prisoner’s Dilemma, it is usually the case that c = 2a, as in Table 14.4: Table 14.4  Hawks and Doves.

C R

C

R

2,2 4,0

0,4 1,1

Here the legend might be, for example, two hunters meeting over a quarry, where C is the intention to share it, while R is the intention to fight for the whole of it. Hence, if two cooperating hunters meet, they peacefully split the quarry, while in the case of two non-cooperating hunters they also get a share of the quarry (on average, one half again), but now they suffer a loss imposed on them by the fight. If a noncooperative hunter (a “brawler”) meets a cooperative one (a “retreater”), he takes everything. The problem with this game is that its only Nash equilibrium is mutual non-cooperation, while we see that mutual cooperation would be much more profitable. The problem with the cooperation is that whenever one (unilaterally) changes the strategy from cooperation to noncooperation, she will profit – so there will always be the temptation to do this. (Such an abrupt switch from cooperation to non-cooperation would also mean a total loss for the opponent, and hence insofar as we see an opponent in this game as a partner, it may be seen as “letting her down”.)

184  Cooperation and morals The last possibility studied in the literature is the variant for which b > d, presented in Table 14.5: Table 14.5  Chicken.

C R

C

R

2,2 4,0

0,4 −1,−1

This is a game that is usually called Chicken. It is similar to the Prisoner’s Dilemma save for the fact that the fights between the non-cooperating hunters are so fierce that they do not pay at all. In this game, mutual cooperation is a Nash equilibrium. What is the moral that we should draw from the considerations of the possible models of confrontations of cooperation with non-cooperation? The bottleneck is the assumption that if I invest something, I will get it back (optimally enlarged). Cooperation can thrive if – and only if – this bottleneck is overcome. 14.3  Moral rules In current literature, the rules that have to do especially with cooperation are those that are usually called moral (Curry, 2016), for these are supposed to be the rules that hold the cooperating communities together. Recent theories of morals that are so naturalistic as to be compatible with the view of this book (Ridley, 1997; Joyce, 2006; Kitcher, 2011) mention the birth of morals in one breath with the establishment of human cooperation and so with overcoming the Prisoner’s Dilemma. Cooperation needs rules, and especially moral rules (while we have pointed out, rules need cooperation). So far we have dealt with various kinds of rules, but not yet especially with moral rules. This may seem surprising – are the rules of morals not the most important ones we have? Well, there is no doubt that they are important; they are (together with the rules of language) the firmest glue that holds our communities together. The reason why we do not pay any specific attention to them is twofold: first, I do not think they form a welldelimited category (I think they are just the most resilient social rules – see below); and second, moral rules have an aspect that goes far beyond the boundaries of this book. (It has to do with, in our terminology, the “inside” of moral rules – living within the rules, as we all do, is a peculiar and multifaceted experience, which gets much anatomized not only by philosophers, but also by artists.) But certainly we must say something about these rules.

Cooperation and morals 185 We saw that rules can be self-standing or integrative. The first of them are usually quite transparent and often make sense in the light of evolution (like You should always wash your hands before eating). We can possibly also find such rules – or their rudiments – in the communities of animals other than us humans. Integrative rules constitute a less transparent phenomenon; they are components of complex systems which may serve various aims. We have seen, many times, the example of chess: the game is constituted by a system of rules, which are skillfully wedged into each other, constituting chess pieces and opening up the space for them to engage in combat. We have also seen that complexes of integrative rules, more generally, are illuminatingly seen as opening up new spaces where we can do unprecedented things. The space of chess is specific in that it is not for everybody. Well, in principle, it could be for everybody; but in reality not everybody will play chess. Also, those who will play chess will not play it all the time – hence, the space of chess is not a place for permanent residence. It is visited by only some people, and they come and go. When you do not play chess, you can safely forget about its rules. Consider, in contrast to this, another of our frequent examples – language. It is for everybody, in the sense that the space of a particular language is for every member of the corresponding community. True, there may be people who are not able to learn the language – or to learn it perfectly – because of some disabilities, but such people are usually taught some substitute language (like sign language for people who are hearing impaired). Also language – or the “space of meaningfulness”, as we dubbed the space opened up by the rules of language – is suitable as a permanent dwelling.2 We do not leave it even when we cease to communicate. It is also a prism through which we see the world and with which it becomes so integrated with us that it would be difficult – if not impossible – to give it up.3 This has to do with the social import of chess (resp. language). Chess is what we call a game; it is not obligatory and not essential for the existence and life of a community. Language, on the other hand, is obligatory and essential – hence, it is much more important than chess. However, this may lead to the question whether it is not mistaken to put chess and language into one box (systems of rules or also institutions). Whether, in other words, the differences do not by far outweigh the similarities. I think not (and this is also one of the main ideas of this book). Without downplaying the differences between the two systems, I believe that the working of rules, which animates both of them, is the all-important common denominator, the capturing of

186  Cooperation and morals which gives us the key to understanding our ultra-sociality and hence us as a species. By not playing our language games, one excludes themselves from the life of our community. And now moral rules are rules that are as allencompassing as those of language and the violation of which works toward such an exclusion by others. In this sense, these rules form the normative backbone of the society; they are the glue that keeps it together. Some philosophers believe that moral rules are a sort of natural kind. We already saw that, for example, Brennan et al. (2013) maintain that moral rules are characteristically non-formal and are being taken to approximate some objectively valid principles of morality (see Section 5.1). Others deny that there would be a sharp boundary between moral and other social rules. We have already seen that this is the case, e.g., for von Wright (1963) or Kumar and Campbell (2022). Moreover, when we look at morality from a cross-cultural perspective (Berniūnas et al., 2022), we can see that the boundaries between what is taken to be moral (in contrast to what is arbitrarily conventional) are quite variable (it is even not clear whether the term “moral” has equivalents in all languages), which appears to support the view that moral rules are just the most general and most cherished of the rules holding the community in question together, without a natural border. Anyway, moral rules usually have to do with the very constitution of human communities: just like when you play chess you must respect the rules of chess (and if you do not respect them, you are excluding yourself from the community of chess players) to be a member of a community, you must respect at least the most basic rules generally respected by the community – and these are usually taken as the rules of morals. We saw that the rules of chess have the power of transforming pieces of wood into pawns, rooks or bishops. Similarly, the rules of morality, in tandem with those of language, form the backbone of a system of rules that has the power of transforming the organisms of the kind Homo sapiens into persons, centers of social acting, potential bearers of commitments and entitlements and individuals that are responsible for what they do (as we discussed this already in Section 9.3). Being a person is being a subject of the realm of moral rules – Kant’s “kingdom of ends”. 14.4  Normativity as a project We saw that the nature of the crucial confrontation of the strategies of cooperation and non-cooperation (which decides whether cooperation can survive as a viable strategy) is not so determinate as to lead to one definite model – there are variants. Some of the variants, like Stag Hunt, are more friendly toward cooperation. The trouble, however, is that the majority of

Cooperation and morals 187 theoreticians are convinced that it is instead the Prisoner’s Dilemma that renders the crucial encounters, and hence that we must consider.4 There are various models explaining how the dilemma could have been overcome. Given the current view on evolution (it is not individual organisms, but rather genes, that replicate), the so-called kin selection (Hamilton, 1963; Maynard Smith, 1964) is self-explanatory: seeming altruism at the level of organisms is due to sharing genes. This then could be perhaps generalized: if I am not able to recognize who shares my genes, then it may be a viable strategy to display some altruism to everybody – just in case. Such mechanisms as “altruistic punishment” (Fehr & Gächter, 2002; Boyd et al., 2003), viz. making life more difficult for non-cooperators, help of course (in that it makes it harder not to cooperate and easier to cooperate), but at the same time it only shifts the problem elsewhere, for there must be some “punishers” to implement it, and the dilemma of whether to be a punisher or a non-punisher suspiciously reminds us of the original dilemma of whether to be a cooperator or a non-cooperator, with all its troubles. If you assume that individuals can have strategies for not only single encounters, but also for iterated ones, you can introduce strategies like tit-for-tat, so that you start to cooperate and continue only if your opponent (partner?) reciprocates (Axelrod, 1984). Such strategies, it turns out, can be much more successful than simple cooperation. Moreover, if individuals can store the information as to who is reciprocating and who is not, one can continue to interact only with those that do (Spiekermann, 2009). This leads to an “ostracization” of the non-cooperators and the establishment of cooperating brotherhoods, which can then flourish. The weak point of this scenario is that it would seem that organisms capable of assuming such complex cooperative strategies must evolve out of simpler ones capable of cooperation. Therefore, the most promising solution may be connected with the socalled group selection account (Williams, 1970; Maynard Smith, 1976). This is something that, if construed too literally, leads us astray (and for some part of the previous century, it was therefore taken as just a mistake). The thing is that groups are not replicators in the sense of the theory of evolution and thus it is amiss to imagine that on their level there goes on a selection wholly parallel to the basic one on the level of individuals or genes. However, it is clear that groups may function better or worse (for example, from the viewpoint of warfare) and that being a member of a better functioning group makes me more fit than being a member of a worse-functioning one. True, what is the most rewarding is to be a “free rider” in a well-functioning group, so there is a danger that the group will always fall victim to its free riders. But it turns out that under suitable circumstances being a loyal member of a well-functioning group makes you more fit than a free rider risking that her group falls apart

188  Cooperation and morals (Bowles & Gintis, 2011). Hence, “group selection”, if understood as a suitable metaphor, can be explanatory. Insofar as this is correct, it may be essential for an individual to be part of a well-organized community. (Indeed, formulated in this way, it sounds more like a banality than a discovery.) And as we claim that communities are organized in terms of rules, we can expect that even this “wellorganization” will be a normative matter. And now morals, as the most robust infrastructure of the web of rules, appears, precisely from this viewpoint, to be quite essential. A close relationship between cooperation, morals and game theory is spelled out by McKenzie Alexander (2007, p. vii): Each of us has goals we would like to attain and ends we wish to achieve. However, your ability to attain your goals and achieve your ends is constrained by the fact that you are a social being. You live in a society where other people are trying to attain their goals and achieve their ends and, on some occasions, their goals and ends are incompatible with yours. The heuristics embedded within moral theories prescribe ways of acting so that the majority of people wind up sufficiently satisfied with their lot in life the majority of the time. This view embeds, very straightforwardly, morals into cooperation. What moral rules are about is simply finding mutually advantageous equilibria in our potential confrontations. Morals are just the statics of cooperation. I think this is much too simplistic (see, e.g., Beran, 2022); though, of course, I do not mean to deny that morals have a lot to do with getting along with each other and hence also with the critical confrontations that are represented by game-theoretical models of the Prisoner’s Dilemma kind. But morals are not merely a technology for resolving conflicts; it is, for each of us, an orientation system that anchors us in reality and provides for the mother board to all the normative spaces that constitute our indispensable niche. But here we have already arrived at an aspect of moral rules that goes beyond the boundary of this book and that I will not be discussing in any detail. Kitcher (2011) calls ethics a “project”, and this nicely chimes with some of the aspects of our explanation of rules. First, we have concentrated exclusively on rules that are our “projects” in the sense that they are here because of us, they are not products of a god and not even evolution as such. We have come to solve the problems of our coexistence – not only the serious, moral ones, but also lots of other, more marginal ones – in terms of rules. This, of course, is a project much more general than just ethics. However, we must keep in mind that when we talk about rules or ethics as a project, then we must not imagine a project that is intentionally

Cooperation and morals 189 devised and set into motion. The project was accomplished by being assembled piece by piece over generations, with nobody to oversee the whole mission. However, the individual pieces were not just put in place by evolution. They were at least partly set up by autonomous human agents. 14.5 Summary Like many other social species, we humans have come to cooperate. But we have come to cooperate in such an unprecedentedly complex manner that we have over time alienated ourselves from other social species. Our kind of cooperation both required and fostered rules, especially the rules that we call moral. Recent naturalistic theories of morals suggest that their birth had to do with the acceleration of cooperation and especially with overcoming the Prisoner’s Dilemma. This simple game-theoretical model made it difficult to explain how cooperation could have flourished, for it seemed that whenever there appeared a mutated individual disposed to cooperation, it must have been wiped out by the not so disposed establishment. However, it may be fitness-enhancing for an individual to be part of a well-organized community. And as we claim that communities are organized in terms of rules, we can expect that even this well-organization will be a normative matter. And now morals, as the most robust infrastructure of the web of rules (though they may not form a well-defined category), appear, precisely from this viewpoint, to be quite essential. Though morals have a lot to do with getting along with each other and hence also with the critical confrontations that are represented by game-theoretical models of the Prisoner’s Dilemma kind, moral rules are not merely a technology of resolving conflicts; they form, for each of us, the most basic matrix anchoring us within our, human world. Notes 1 Dor et al. (2014, p. 2) write: “Unlike perception in 3-D, [language] presupposes engagement with other minds. To be ‘language-ready’ (Arbib, 2006), the brain must be social to an unusual degree; and for the human brain to be that social, human society must have gone through an unusual evolutionary dynamic”. 2 This is reminiscent of the famous metaphor of Heidegger (1959): language as the “House of Being”. 3 Wittgenstein (1969, p. 110) writes: “So is the calculus [of language] something we adopt arbitrarily? No more so than the fear of fire, or the fear of a raging man coming at us.” 4 It is also possible to conclude that there is an irreducible plurality of the kinds of such encounters providing for the most basic elements of morality, which then can be combined in various ways (Curry et al., 2022).

15 Freedom

15.1  Flexibility vs. predictability We have discussed some ways in which we, individuals of the species Homo sapiens, may be seen to differ from the individuals of other species. One of the ways we have not touched on yet is our freedom: many philosophers would, beyond a doubt, consider precisely this mark of our existence as one of the most distinctive features of us humans. It is an obvious fact that our reactions to the stimuli of our environment are much less rigid than those of any other species. Glock (2012, p. 130) argues that besides language and our specific kind of sociality we are characterized precisely by this “special kind of plasticity: the capacity to adapt to highly diverse circumstances and environments through tools (technology) and rational deliberation (planning)”. In purely ethological – and very down-to-earth – terms, we can perhaps say that we excel in the flexibility of our behavior. We have already noted that the repertoire of the behavioral patterns we are able to display is obviously much richer than those of all other species. Dennett (2018b, p. 7) writes: We – and only we – must live in a world of our own creating that is orders of magnitude more complex and replete with opportunities (the degrees of freedom) than the lifeworld of any other living thing, and, with the help of evolution, both genetic and cultural, we have designed a system of higher-level cooperation that opens up modes of negotiation and mutually enforcible constraints, the civilization that makes life so worth living. Flexibility of behavior, of course, is something that makes a lot of sense from the viewpoint of evolution. If a creature is wired up so that it can react only in one way to a given stimulus, then if the reaction is inappropriate, the creature is doomed; whereas if it can try another reaction, it may fare better (Dennett, 1996, Chapter 4). (If the rigid reactions are fine-tuned DOI: 10.4324/9781003388876-16

Freedom  191 by evolution, then they are unlikely to be inappropriate – but this will not hold if there is a change in the environment with respect to which they have been fine-tuned.) Moreover, it fosters the “innovative behavior” that provides for the ability to explore the space of evolutionary trajectories, which we saw has been so important for us humans. Thus, flexibility of behavior might be seen as one of evolution’s tricks for increasing an organism’s fitness, a trick that has turned out to work very well for us humans. But we have already seen that flexibility has its own problems. Though the environment may change (and then the possibility of altering one’s way of coping with it may be welcome), it does not change all the time (and often it does not change significantly for a long time). In such situations, it would be disastrous if we were to try out all our available behavior patterns every time anew; this would, of course, frustrate our ability to deal with the world in a reasonable on-line manner. The primary solution to this problem is our ability of acquiring habits, of acquiescing in some “tried and true” ways of behavior. However, this would still leave us with the necessity for every individual to invent and establish their habits anew, perhaps by trial and error, which would still not be optimal as most changes in the environment do not proceed swiftly enough to substantiate this. The solution is obvious: we learn some “standard” ways of dealing with nature from our elders, our teachers and our peers. And this mechanism is immensely more elaborated for us humans than for any other species: young individuals of some other species do learn some things by imitating their elders, but this shrinks in comparison with our complex system of education and enculturation.1 But what is more important is that we are also social creatures, and our success depends also on how we cope with each other – and here again flexibility might be a hindrance. Cooperation and prosperous coexistence depends, to a large extent, on predictability, and the more flexibility of behavior you have, the less predictable you tend to be. Thus, it seems that if coping with the world leads us, by way of evolution, to an increasing variability of behavior, then coping with each other should lead us to some kind of – at least virtual – neutralization of the variability. And the idea I put forward here is that we have developed an elegant response to this prima facie schizophrenic situation – among all the courses of action which we can take we have singled out a limited number of those which should be taken. We have managed to convey our unpredictably flexible behavior into some “standardized” and publicly recognized channels.2 The result is the emergence of correctness and incorrectness as we have discussed it in previous chapters. There are cases in which “correct” means what we have dubbed “standard” above and which derives from natural facts as “rules of engineering”: for example, the correct ways of hunting a hare may be just those which are likely to bring us success. But there are

192 Freedom also cases of correctness which look quite arbitrary, such as various rituals. In the case of rituals, it may not be important what actual behavior is being proclaimed correct, it is only important that it is a behavior on which we all converge. This is the kind of “normative conformity” we talked about in Section 10.4. Of course, the boundary between correctnesses that are “forced by the environment” and those that are “arbitrary” (purely “conventional”) does not coincide with those that concern nature and those that concern the society. Surely not all the correctness in the social realm is just arbitrary – certainly “Thou shalt kill” is not as good as “Thou shalt not kill”. This gives rise to what becomes known as moral rules and what was discussed in the previous chapter. However, despite this, it is clearly the social realm that provides a lot of space where correctness can be stipulated purely “conventionally”. 15.2  Freedom and rules From what we have stated so far, it may seem that our freedom simply springs from the fact that our behavioral flexibility, the spectrum of possibilities of how to act which opens in front of us at nearly every moment, is vast and incomparable to what we find elsewhere in the animal kingdom. True, we saw we do need rules, but they seem to have little to do with freedom. Things, however, are not that simple. Every animal lives in a world delimited by certain physical boundaries. These interfere with the spectrum of behavioral possibilities that would otherwise be open for the animal thanks to the variability of the behavioral patterns it is endowed with. (An animal cannot go wherever it wants because of natural hindrances; it cannot eat whatever it might want; it cannot move faster than its physical constitution allows etc.) Humans have managed to expand their variability of behavior in an unprecedented way; however, they subsequently began introducing rules in order to narrow it back down. As we saw, this makes sense from the viewpoint of cooperation and predictability. Does not, however, the introduction of rules and the restrictions they institute again compromise our newly acquired freedom? Remember what we have said about the distinctively human form of life: our niche is normative and the reality into which we are born is institutional. Therefore, we find ourselves, from the beginning, in a situation different from that of any of our animal cousins. We do not merely learn how to behave effectively (escape from predators, catch prey), but rather how to act properly. Moreover, we learn this not only by trial and error and by the imitation of the older members of our species, we are taught

Freedom  193 how to act properly (or sometimes perhaps also how to act improperly), which is something our animal cousins lack almost completely. Of course, in the case of any animal, we can talk about freedom. When, say, a wolf who is captured and kept in captivity is let loose, we are likely to react Now she is free! and we may react in a similar way to a similar situation concerning a human. But we have seen that we humans reach maturity by acquiring the ability of operating in various normative spaces – so our abiding by certain rules becomes part and parcel of our becoming members of our communities and adopting our human form of life. So can we identify our freedom with a repudiation of all kinds of constraints, including the rules abiding by which made us fully fledged humans? Kant famously claimed that we humans, aside from belonging to the denizens of the “realm of the concept of nature”, also inhabit the “realm of the concept of freedom” (see Section 2.1). In particular, he also claimed that freedom is autonomy, i.e. the ability to freely choose the rules according to which to live. It follows that living by rules – by some rules – is a presupposition of achieving freedom, at least this kind of freedom. This may seem paradoxical, for rules appear to be what restrict us, what preclude us from doing certain things. It may seem that insofar as rules compromise our freedom, to be free we must apparently eliminate as many rules as possible. Rules and freedom thus seem to pull in opposite directions. However, Kant’s point was that this is a distorting view. Imagine a person following no rules whatsoever. Would it be a prototype of a free actor? No, because it would not be an actor in the first place. We have already seen (Section 9.3) that to act – in contrast to simply display a behavior – is to make a move in a rule-governed space, in which every move has reasons and consequences for which the actor is responsible. Hence, it presupposes a framework of rules. So rules are no enemies of freedom (cf. Brandom, 1979) – it is only a certain kind of “hostile” imposition of rules that compromises freedom. Thus, our Kantian kind of freedom to which rules help us is not greater in the sense of being simply broader, but rather in the sense of being “higher-level”. In fact, the kind of negative freedom which amounts to an absence of restrictions is not freedom in the Kantian sense at all. His freedom, autonomy, is something we can achieve only thanks to our ability to abide by rules, to be normative creatures. There are myriads of new things which we can do in the new world, and only in the new world. Think about the vast number of our actions that depend on the various kinds of normative, institutional frameworks we have established: studying at a university, getting married, playing football, buying goods etc. For better or worse, we humans partly evacuated the natural world in favor of the condominium of our normative spaces, aka institutions.

194 Freedom Hence, Kant’s conception of freedom goes very well with our characterization of our human form of life. Throughout the previous chapters, I have elaborated on the idea that we, modern humans, have largely relocated from the space of nature into the various spaces we have built up from our rules. We have concluded that it is only within such normative spaces that we can act as persons who are responsible for their deeds – and it is precisely this that our form of life amounts to. It follows that living by rules is something that must underlie any other achievement of us humans as rational animals. Thus, it is, of course, presupposed by achieving freedom – freedom, that is, in the human sense explicated by Kant. The point is that while, of course, even mere behavior can be called “free” (like when a captured animal is let loose), freedom in the fully fledged sense applicable to us humans amounts to free action. 15.3  Navigating our normative world freely From the viewpoint of evolution, behavioral flexibility is our principal weapon – we have developed it just like fish have developed their gills and birds their wings. But we saw that the vast broadening of the spectrum of behavior available to us calls for some regulation, for it is too vast. It has come to be regulated by rules, which did elevate our behavior on a new level, on the level of acting. I think that it is instructive to reconsider a problem which Kant left to his followers and which became, according to Pinkard (2007), one of the deepest problems of German idealism. If we are autonomous beings and if autonomy is choosing one’s own laws, then the crucial question is how do I choose the rules to follow? Where do the rules, by which we, as actors and as rational beings, abide, come from? The point is that there seem to be two possibilities. Either we chose the rules reasonably, which, however, means, by Kantian standards, by some rules, and hence we can choose rules only if we already have some other rules, and we start an infinite regress; or we chose them just arbitrarily, which does not seem to lay an acceptable foundation for reason. As Pinkard (2007, p. 26) puts it: In the Groundwork Kant had declared that “the will is therefore merely subject to the law, but is so subject that it must be considered as also giving the law to itself and precisely on this account as first of all subject to the law (of which it can regard itself as the author)”. However, as Kant himself clearly saw, taken literally, this would be paradoxical; a lawless will cannot obligate us, and a choice of practical principle made on the basis of no law would be simply arbitrary,

Freedom  195 so if the will were to give itself a law, it would need another law to choose that law; but if only a self-legislated law was obligatory, that prior law could not itself be obligatory. Now a reader of this book should already foresee how we solve this conundrum.3 That we cannot have rules for the establishment of all rules is obvious – the imminent infinite regress is unacceptable. Hence, do we admit that we establish the rules just arbitrarily? Not really. There are “reasons” for the specific choice of the ground-level rules we follow, but they are not our reasons. They are, as it were, the reasons of the evolution that has brought us into our current shape and armored us with our reason. It is only because these very rules founded something as uniquely useful as our reason and reasoning that they were supported by evolution and that they have come to govern us. Of course, the “reasons” of evolution are not reasons in the normal sense of the word. Clearly, evolution itself does not reason; it is only in hindsight that we can say that we have evolved as if reasons were in play. This is connected with what we have stated when discussing the naturalization of rules (Section 6.4): the most foundational rules are not correct/ incorrect, they are just useful. There is one more problem we should return to. Evolution, we said, equipped us with a broad spectrum of possible behaviors, while rules tend to narrow it back down. Do not rules wholly neutralize the achievement we gained in this way, do they not rob us of the most valuable achievements of our becoming free, such as the possibilities of being spontaneous, creative or playful? Consider, once again, playing chess. When I accept the rules of the game, I voluntarily restrict the possibility of my behavior: I refrain from moving the bishop otherwise than diagonally etc. But it is via this refraining (of me and my opponent) that the pieces are turned into bishops, knights etc., and I can enter the virtual space of chess games, the space which makes it possible to play the game. However, this space becomes a very specific kind of arena, which lets me do very specific things I am not able to do outside of it. It is only within the arena that I am able to check an opponent’s king. I also acquire a new kind of freedom: the freedom to attack my opponent in such or another way, to choose ways to defend myself against her attacks etc. Of course, part and parcel of the game is that I can choose my moves freely. I am restricted by the rules, but the restriction still leaves a vast space in which I can operate. (It is so vast that though we know, and knew as long ago as Zermelo, 1913, that there exists an optimal strategy, we are – fortunately – not able to find it.) And if I were forced to choose the moves according to some prescription, the whole enterprise can lose its point.

196 Freedom Hence, even in a rule-delimited space, there may be the freedom of movement – and the freedom may be so important that the crux of the whole space may stand and fall with it. Now humans, we saw, have established multifarious normative spaces in which we can perform similarly unprecedented kinds of actions, save for the fact that they are not mere moves in a game, but often socially much more “serious” actions. The most important space is the space of our language, the “space of meaningfulness”, which we discussed in previous chapters. But by no means is it only the rules of language that constitute a normative space which we inhabit. We have a plethora of other, smaller or bigger, either less or more important, either less or more widespread, normative spaces. We live great parts of our lives in this system of interconnected normative spaces; this is our specific human kind of niche – the normative niche. 15.4 Summary Our reactions to the stimuli of our environment are much less rigid than those of any other species – we excel in the flexibility of our behavior. The repertoires of the behavioral patterns we are able to display are obviously much richer than those of all other species. Flexibility of behavior, of course, is something that makes a lot of sense from the viewpoint of evolution: if a creature is wired up so that it is fine-tuned to its environment, then if the environment changes, the creature may be able to adapt. But though the environment may change, usually it does not change very much and it does not change too quickly, so it is useful to have some “standard” repertoire of reactions. The primary solution to this problem is our ability of acquiring habits which get upgraded to rules. We humans reach maturity by acquiring the ability to operate in various normative spaces – so our abiding by certain rules becomes part and parcel of our becoming members of our communities and adopting our human form of life, and hence our freedom does not consist in the repudiation of all kinds of constraints but rather in the ability to freely choose the rules according to which to live. It follows that living by rules – by some rules – is a presupposition of achieving freedom, at least this kind of freedom. Rules are no enemies of freedom – it is only a certain kind of “hostile” imposition of rules that compromises freedom. Notes 1 See Laland (2017), who puts stress on education as a human specialty that has granted us our unique ability to learn from others and thus to establish our cumulative culture.

Freedom  197 2 We saw (Section 13.3) that Mercier and Sperber (2017) argue that this is also the situation where our argumentative practices took origin – originally, they argue, these practices were tools for the neutralization of potentially disturbing non-standard behavior. 3 Pinkard discusses, in detail, Sellars’ attitude to the problem, so the reader may compare it with ours.

16 The world

16.1  Objective truth and the objective world How do people start to live in an objective world, knowing that it is independent from the view of each of them? We have seen that due to the birth of the primordial, Protagorean normativity out of being “in tune” with each other, there emerges a sense of an intersubjectivity; this, however, is still not objectivity. Hence, we need to see how we have reached the notion of objectivity, according to which the world is not only independent of the views of any one individual, but also of the intersubjective view of all the individuals. Thus, at the beginning of the time span that we are interested in, we find our ancestors emitting various kinds of sounds in an uncoordinated way. At the end of that time span, we find us – people living in the shared objective world – talking about it, and arguing about what is true. What happened in between these two points, and how did we work our way to the objective reality in which we all live? I think that inferentialism can tell a plausible story about this. Of course, it is not intended as the infamous “just so story” that tells us what happened on the basis of no evidence; it’s just a parable documenting that such a development was possible. Let me first summarize how this problem is handled by Brandom; then I will argue why I hold his account to be insufficient. Brandom’s basic units of normative entanglement are commitments and entitlements. We are always committed and entitled to various things, and participating in rule-governed activities (and especially linguistic activities) crucially involves acquiring (and losing) various kinds of commitments and entitlements; it also involves keeping track of the commitments and entitlements of other participants (“deontic scorekeeping”). One may live up to the commitments they have, or one may fail. If one does not live up to them, it may be because they are not able to, or not willing to, or else they do not accept the commitment. (One may become committed without wanting, or even without knowing about it.) Because of the last possibility, one’s DOI: 10.4324/9781003388876-17

The world 199 scorekeeping tends to record not only what others are committed to, but also what they take themselves to be committed to. This distinction forces the scorekeeper to have “two sets of books” (Brandom, 1994, p. 488): one for commitments and entitlements which she ascribes to others, and one for those she recognizes they acknowledge. Special kinds of commitments and entitlements are those gained and lost in the course of our ongoing linguistic activities, the central one of which is, according to Brandom, “the game of giving and asking for reasons”, GOGAR. Its principal moves are assertions, which count as expressing propositions and as making claims. Here again one can fail to do what she should do, in particular one can assert something that is not correctly assertable – viz. something the assertion of which violates the rules of GOGAR. As an assertion – typically – purports to present how things are, the act of asserting something that is not correctly assertable is to present how these things are not, either deliberately, in which case it is to mislead, or inadvertently, in which case it is oneself that is misled or mistaken about how these things are. Of course, to judge somebody as taking something to be not as it really is presupposes knowing it as it really is, and every human judger can only approximate this by way of her own convictions. Hence, the scorekeeper may be in error, too; from the viewpoint of somebody else, it might be that what she holds is or should be the case is not what really is or should be the case. So what she sees as the confrontation of appearance with reality is merely the confrontation of another’s views with her personal view. A “reality” in the normal sense of the word is something that can differ from everybody’s views. Thus, everybody is merely able to compare somebody else’s beliefs with her own beliefs – no objective reality is in play yet. How do we reach it? Brandom is convinced that this raw material is enough for the fabrication of the notion of an objective reality. In GOGAR, the incipient rules state the correctness of its moves vis-à-vis other moves and vis-à-vis extralinguistic circumstances. As the moves mutate into claims, the rules can be seen as regulating the correctness of the claims; some claims come to follow from other claims in the sense that the commitment to the former brings about that to the latter. In the general case, the commitments that belong to the first book are, according to Brandom, those claims “that, according to the scorekeeper, actually follow from the claims made (given how things really are) regardless of whether the one making the claims realizes that they follow or not” (p. 488). Those that go into the second book are such that are recognized by “the one making the claims” (ibid.). Brandom (1994, p. 597) hopes to be able to get this notion of objective reality from the distinction between attributions and acknowledgements

200  The world of concepts, via the fact that they are “perspectival” and that the distinction itself is “structural”: On this account, objectivity is a structural aspect of the socialperspectival form of conceptual contents. The permanent possibility of a distinction between how things are and how they are taken to be by some interlocutor is built into the social-inferential articulation of concepts. The distinction is in the first instance available to each scorekeeper regarding the commitments of others, of those to whom the scorekeeper attributes commitments, for it just reflects the difference in perspective provided by the different inferential significance of claims in the context of auxiliary premises provided by commitments acknowledged (and so undertaken) by the scorekeeper and commitments that are attributed to others by that scorekeeper. Although grounded in essentially social, other-regarding scorekeeping, however, the possibility of a distinction between how things actually are and how they are merely taken to be by some interlocutor remains a structural feature, even, as will be seen below, in the case of attributions to oneself. What Brandom has in mind here are not only outright conflicts between the claims of the scorekeeper and those of the “scorekeepee”, there are also much more delicate discrepancies that are in play. As our language, and hence the system of claims that are at our disposal, is intricately inferentially articulated, the significance of asserting the same sentence by the scorekeeper may differ from that of the scorekeepee. The significance consists in what follows from the sentence, and what follows from it is influenced by all kinds of collateral premises – viz. other claims that the person in question accepts. The significance of Tiger coming from the mouth of somebody who is additionally convinced that it is an attacking fierce beast is very different than when it comes from the mouth of somebody who thinks this is just their pet returning from an outing. Brandom thinks that such different perspectives are what create the kind of threedimensionality that provides for the objectivity of the real world. The idea seems to be that the illumination from many visual angles provides for the determinacy, persistency and plasticity required of reality. I do not believe that this account suffices for the establishment of objectivity. True, the root of the appearance/reality distinction may be seen as borne out by distinguishing what should be done from what is done – if we have a plausible enough story about how the notion of what should be done comes into being. However, Brandom, who is not really interested in the developmental aspects of normativity, does not provide us with such a story. Moreover, I do not think that explaining the appearance/reality

The world 201 distinction is the whole story to be told. I think that what we need in addition to this is an explanation of how the relative version of this distinction (where my “reality” is somebody else’s appearance and vice versa) becomes so stabilized that it yields an absolute notion of reality, which is the same for everybody. 16.2  Non-Protagorean correctness revisited I think there is something else that is crucial here, crucial for the very notion of objective reality; but to articulate this, we must return to the very beginnings of the common language. Let us return then, once more, to Section 7.2, where we pondered the possibilities of the emergence of rudimentary rules. We argued that the displays of people may have come to be useful for one’s peers, and that this may have given rise to the displays being adjusted by means of normative attitudes. We also conjectured that the choice whether we try to modify the displays by means of the attitudes or rather change our expectations accordingly might have been influenced by the corresponding choices of our peers – which is to say that the people wielding the normative attitudes tend to align the attitudes across the community. This, we may now further conjecture, provides for the alignment of what should be done across the community. We saw the yawning gap between what someone does and what should be done, and now we see that what should be done is no longer idiosyncratic for every member of the community: it tends to merge and produce something as the objective should. Hence, with respect to what should be done, “reality” is no longer merely a parochial notion, the “reality” of one person being mere appearance for another one. There emerges the “real” should as something relatively stable and independent of the parochial views of individual members of the society. Thus, it is already when forming the Protagorean attitudes that the protagonists recognize something as an intersubjective stance. A stance is not of any particular person, though it is likely to coincide, or roughly coincide, with the stances of the majority. But an intersubjective stance is not yet an objective one – it does not provide for the possibility of the majority being mistaken. Therefore, it is satisfactory for the Protagorean rules but not for the host of non-Protagorean ones that grow out of them. We are still not able to explain how something can be correct or true despite the fact that the majority – or, for that matter, everybody – does not think so. From what we already saw, it follows that the key to overcoming this conundrum is in how we set up our rules. When the problem is what the word dog means, it is enough to check the attitudes of relevant speakers (hence here the relevant objectivity does not surpass intersubjectivity),

202  The world while when the problem is whether a concrete animal is a dog, it may be necessary to research it, which may yield a result surprising for everybody (see Section 8.3). How we research it depends on the context, sometimes it is enough to look at it; in some other cases, we need to do a closer check to make ourselves sure that it is not a wolf, or perhaps a cat; in any case, it is these procedures that come to be associated with the word dog and make the rule To say “This is a dog” is correct when pointing at a dog more precise. The foundation of our normative edifice which innervates our social life is constituted by the Protagorean attitudes, with respect to which agreement with the majority is the only thing that makes them correct. However, as we saw already in Section 8.2, these attitudes may give rise to other attitudes that already need not be Protagorean. In their case, agreement with the majority is not the ultimate measure of correctness. There may be procedures the results of which need not directly elicit a common agreement, which must then be reached in terms of GOGAR. (Thus, e.g., the Protagorean agreement on what it takes to be a dog – i.e. what dog means – provides for some animals being dogs even when the majority of people do not see this.) It is part and parcel of the current variety of English that it is correct to say This is a dog when pointing at a dog. Why is it correct? Oversimplifying a little, we can say that it is because this is held to be correct by the speakers of English. There is no deeper reason, and none is needed. Consider, however, a specific case of this rule. I see a dog, point at it, and say This is a dog. Is this correct? Most probably it is – unless I mistake some other animal for a dog. The danger of this kind of error increases with the complexity of the concepts involved. This is a case of Covid may necessitate some laboratory tests or other research, which introduces a lot of space for mistakes. In general, while the assignment of meaning to a linguistic item is a straightforwardly Protagorean matter, the meaning may be such that it introduces a non-Protagorean correctness. In the prototypical case of an application of a concept (dog, Covid), there is a – bigger or smaller – space for misapplication. This means that though people will tend to assume normative attitudes to such applications, these may be divergent and may be subject to change, especially on the basis of GOGAR. Indeed, GOGAR (especially some of its developed variants, like, e.g., polemics in professional journals) is what will lead us to a prevailing verdict regarding the correctness of the application. Anyway, what is crucial for the emergence of objectivity not reducible to mere intersubjectivity is the step from the Protagorean correctness to the non-Protagorean one, viz. correctness as mere resonating with others to the more sophisticated versions, according to which what counts is not

The world 203 the resonance with others with respect to the result, but rather with respect to the ways in which it is reached. In this way, some animal may be a dog even if nobody believes it is a dog, because this may be the result of an examination carried out by the method everybody takes for appropriate. Hence, it is this step which provides for the passage from intersubjectivity to the genuine objectivity. We believe that things may turn out to be different from how we all believe them to be, because we believe there are ways to find how they really are, which may go counter to anyone’s beliefs. Thus, we believe there is an objective truth regarding how the world is, independent of how we all believe it is. 16.3 Objects That we live in a world in which something holds, or is true, objectively does not yet mean that we live in a world of shared objects. But precisely this appears to be characteristic of us humans: we live not only in a shared and objectively existing world, but we live, more specifically, in a world of shared objects. How does this come about? Quine (1969, pp. 8–9) famously claimed that what opens the door to the world of objects is our language: It is only when the child has got on to the full and proper use of individuative terms like “apple” that he can properly be said to have taken to using terms as terms, and speaking of objects. […] How can we ever tell, then, whether the child has really got the trick of individuation? Only by engaging him in sophisticated discourse of “that apple,” “not that apple,” “an apple,” “same apple,” “another apple,” “these apples.” We therefore learn to live in the world of objects by learning our language with its “individuative apparatus”. The question, of course, is why we have such an individuative apparatus in our language. How did we get from Tiger and Danger to, say, This tiger is dangerous, where this tiger points out a certain individual? Again, there is no definite answer to this question; there are, nevertheless, some plausible speculative ones (ScottPhillips, 2015; Planer & Sterelny, 2021). The idea is that an organism, of course, perceives its environment; however, to perceive it as consisting of objects having properties and standing in relations one needs to be a speaker of a language with the requisite tools. What comes to mind is the “cookie-cutter metaphor” of Putnam (1991, Chapter 7), according to which reality is a shapeless dough until one comes with the “cookie-cutter” of language and gives it the shape of a world.

204  The world There are a number of objections to this picture. One kind is based on the conviction that to be able to discern objects, at least on a basic level, one does not need language. Psychologists undertook sophisticated experiments indicating that pre-verbal infants are already able to keep track of individual objects even when they are out of sight (Soja et al., 1991). A very different kind of objection is brought out by Donald Davidson, in many respects Quine’s close fellow-traveler. He put forth a sharp criticism of the “cookie-cutter” view (Davidson, 1974) as he was convinced that any dualism of a scheme (like a language or a conceptual apparatus) which shapes an amorphous content (reality) is untenable. He argued that, despite appearances, we cannot make sense of the notion of a conceptual scheme incommensurable with our own one or a language untranslatable into ours. The reason is that similarity with our conceptual scheme or with our language constitutes the backbone of the criterion of something being a conceptual scheme or a language – hence the mere fact that something is a conceptual scheme or a language entails that it cannot be incommensurable with our scheme or our language. Davidson (2001) thus developed his own account of our world becoming the world of objects – an account that is not so tightly tied with language. The key concept here is that of triangulation (Myers & Verheggen, 2016): an object is what emerges as the focus in the intersection of the lines of attention of two different individuals. Davidson himself describes the situation in the following way (p. 83): The possibility of thought as well as of communication depends, in my view, on the fact that two or more creatures are responding, more or less simultaneously, to input from a shared world, and from each other. We are apt to say that someone responds in ‘the same way’ to, say, wolves. But of course, ‘same’ here means ‘similar’. Our grounds for claiming that a person finds one wolf similar to another is the fact that the person responds in similar ways to wolves. This prompts the next question: what makes the reactions similar? The only answer is, someone else finds both wolves and the reactions of the first person similar. This of course only puts the basic question off once more. Nevertheless, it is this triangular nexus of causal relations involving the reactions of two (or more) creatures to each other and to shared stimuli in the world that supplies the conditions necessary for the concept of truth to have application. Without a second person there is, as Wittgenstein powerfully suggests, no basis for a judgement that a reaction is wrong or, therefore, right. Hence, once one realizes that there are perspectives different from her own, she begins to triangulate: she notices that the perspective of somebody else

The world 205 intersects with hers at a point which may become the focus of their common attention. It solves two problems. One is that connected with the term triangulation1: you can identify a point in space or a spatially limited object. Another one is the alignment of categorization: if you find wolves similar to each other, and if you find somebody else’s reactions to wolves similar to each other, you may take wolf as an intersubjective category. Although the reactions to the objects (wolves, in his example) that Davidson is interested in are mostly verbal, they are crucial for learning language and hence for the pre-linguistic stage. Thus, according to him, individuation of objects is not facilitated exclusively by the linguistic structures one masters as urged by Quine. The term “triangulation” is taken over also by Tomasello (2019), who, however, does not seem to see it as directly connected to language at all (p. 45): Great apes do not distinguish subjective and objective, in my view, because this is not an insight that individuals can come to on their own. An individual cannot come to it either by inventing a clever theory or by simulating another’s experience, and they cannot come to it by comparing their past to their current experience. To understand the distinction between subjective and objective, an individual must triangulate (to use the term of Davidson, 2001) on a shared situation with another individual at the same moment: we both see X, but you see it this way, and I see it that way. That is, the participants must come to understand that the two of us are sharing attention to one and the same thing, but at the same time we each have our own perspective on it. This is the basic cognitive organization of the so-called dual-level structure of shared intentionality. Tomasello thus comes to the conclusion that what is the key precondition for taking the world to consist of objects is not language, but rather our ability to share attention, which is unique to our species: “skills for putting one’s head together with a partner to form a joint goal with joint attention, creating the possibility of thinking about things in terms of perspectival cognitive representations and socially recursive inferences” (Tomasello, 2019, p. 17). A question that is often posed in this context is whether the structure of our language reflects the structure of our world, or whether rather we imprint the structure of our language onto our world; whether our sentences have the subject-predicate structure because we found our world to consist of objects and their properties, or whether we tend to parse the world into objects and properties because we have happened to acquire such languages. It is the former option that comes more naturally, but we saw that Quine went for the latter one, and he was certainly not the only one.

206  The world (Thus, Goodman, 1960, p. 56, for example, urges us to see that “there is no such thing as the structure of the world for anything to conform or fail to conform to”.2) I think that the question whether language reflects the world or rather shapes it is one of the most stubborn philosophical discussions, apparently leading to no resolution. In the twentieth century, it acquired the shape of a dispute between “realists” and “antirealists” (Alston, 2002). I think that these debates often result from the way that the structure of language is bound to correspond to the (independently scrutable) structure of the world. Then it is almost inevitable to conclude that one of the structures must mimic the other one, anything else would be, as Putnam (1975) puts it, a “miracle”.3 Well, it is clear that language, with its structure, developed within our world and in this sense, it is bound to “fit” into it. However, it is not at all clear why this fitting should acquire the form of a correspondence, representation or mirroring.4 We are already familiar with the Wittgensteinian idea that elements of language should be seen as tools (and their meanings as the ways they are usable). True, they must have the peculiar properties that make them usable, but this need not be anything like correspondence with elements of the world. (For example, the model of a lock and key seems to be much more apt than that of an object and its mirror image.) Moreover, we saw that language, more than being directly usable in the natural world, comes to be usable in the virtual normative worlds that codevelop with it. With this in mind, we must conclude that there are no two structures matching each other (be it by a miracle or not), there is only one that bootstrapped itself into existence oscillating between language and the world. True, the structure is most explicitly discernible in the language, but the structure of the language is not just arbitrary, the language has it because it has developed to fit with the world; hence, there is a clear sense that it is due to the world (though not as a copy of a pre-existing structure of the world). 16.4  Symbols in the world of shared objects Thus, our natural world is the world of shared objects, a world that has become our common, intersubjective arena after a lot of alignment of normative attitudes and after a lot of triangulation. Atop of this world, as we saw, we build numerous normative spaces, spaces in which we can lead our specific form of life, which is quite different from those of other animals. In fact, even ordinary objects become denizens of the all-encompassing normative space produced and continually reproduced by our natural

The world 207 science and depicted by Sellars as his scientific image. The original immediacy of our Lebenswelt, which we shared with other animals, has been replaced by this normative space, often only loosely interconnected with our immediate bodily experience. Husserl (1936) saw this as a crisis, but I tend to consider it as more of a natural development of the environment by the normative creatures of our kind. What is, however, more interesting are the normative spaces I termed virtual, especially that which I dubbed the space of meaningfulness. It is this which brings our being the normative species we are to full fruition. In terms of this space, our interaction becomes what may be called “symbolic” because it is facilitated by “symbols”, tokens that have no real impact in the natural world but acquire substantial impact in our space of meaningfulness. But the nature of this “symbolic” interaction is usually misunderstood. The usual idea is that symbols originate by our interconnection of sounds and inscriptions with things, by our “letting the sounds stand for the things”. Here we offered a viable alternative to this view. Symbols do not arise out of our letting sounds or inscriptions stand for something, but out of our assembling of normative spaces, where sounds or inscriptions may become vehicles of new kinds of “symbolic” actions. The spaces are constructed in such a way that subtle differences that do not make a big difference in the natural world make a big difference in them. They become such a useful tool of not only our interaction, but also of our thought, because they are embedded within worlds that provide for this. Krebs and Dawkins (1984) described the emergence of such a situation by talking about a “conspiratorial whisper”. According to them, human symbolic communication originated in three phases. First was “mind reading”: individuals started to predict the future conduct of others on the basis of hints in their behavior. This is something that natural selection, no doubt, put a premium on. But once this was on board, there started to appear a countermeasure, “manipulation”: animals whose minds were “read” started to manipulate the “readers” by deliberately displaying the hints that made the “readers” expect the corresponding behavior. Now in the third stage, the development may take two different routes: when the manipulation is to the detriment of the manipulated, it will require an ever more energetic backup to be successful, while if it is, on the contrary, to their profit, the amount of energy will be reduced to the point of being close to zero. In the latter case, what is born is the “conspiratorial whisper”. What is common to this story and the one told in this book is that the nature of “symbols” is not derived from our ability to let items stand for other items, but rather from the ability to provide for new kinds of actions (like “symbolic interaction”). What I stress here, however, is that the

208  The world “symbolic dimension” of our lives becomes fully potent only when rules are on the scene, only when the emerging symbols not only come to be used in various ways but acquire roles vis-à-vis the rules, i.e., become the equipment that allows us to play the games implemented by the normative spaces. In this way, we have changed the course of our evolution once we developed rules and the mechanism of rule-following. We mastered the art of interlocking rules into constructions delimiting vast spaces which made us completely change our life form. We no longer merely adjust our lives to the causal forces of our environment (that offer opportunities, like food, and threaten with dangers, such as predators); our (ultra-)social organization redistributed all of this in such a way that an individual of our species must adjust her life much more to the normative forces of the society. Thus, we became a truly normative species. 16.5 Summary How do people start to live in an objective world, knowing that it is, as it is, independent of the view and will of each of them? We have seen that due to the birth of primordial normativity out of the alignment of normative attitudes with each other, there emerges the sense of an intersubjectivity; this, however, still is not objectivity. True objectivity and the notion of an objective world enters the scene when our normative attitudes shift, as it were from result to process: once we accept that something may be correct not because a majority hold it for correct, but because it is rendered as correct by some method that is rendered as correct. In this way, there emerges the notion of objective correctness (and especially objective truth) such that something may be objectively correct (true) despite nobody believing it is correct. Aside from living in a shared objective world (a world which does not obediently respect our wishes) we live, more specifically, within the world of shared objects, viz. a world which is adequately describable in subject-predicate sentences. That such sentences turned out to be useful for describing our world seems to be beyond doubt; it makes, however, no sense to ask whether these sentences are useful because the world consists of objects and properties (and relations), or whether we tend to parse the world into objects and properties because we have found the subject-predicate sentence useful. Hence, our natural world is the world of shared objects, a word that has become our common, intersubjective arena after a lot of alignment of normative attitudes and after a lot of triangulation. Atop of this world, as we saw, we build numerous normative spaces, spaces in which we can lead our specific form of life, which is quite different from those of other animals.

The world 209 Notes 1 If you are being localized by your mobile phone, you need to find an antenna tower that “sees” your phone. But it can only determine a direction, a line stretching away from it somewhere on which you are. It is only once you have another tower that “sees” you that they can start to triangulate: you are at the intersection of the two lines. 2 The extreme version of this view leads to so-called linguistic relativism (Goodman, in contrast to Quine and Davidson, being one of its most outspoken exponents). According to it, speakers of different languages may, as language shapes thought and hence the world, live in worlds that are not accessible to each other. 3 The so-called no miracles argument developed on this basis claims that as our best scientific theories are extraordinarily successful, they must be true and hence must correspond to reality (Barnes, 2002; Frost-Arnold, 2010). 4 In this spirit, Rorty (1979) famously proclaimed that to see our theories as mirroring the world is essentially misguided.

17 Conclusion We have become a normative species

17.1  We have come to follow rules Wilfrid Sellars (1956) finished his most well-known piece, The Myth of the Given, with the following litany: I have used a myth to kill a myth – the Myth of the Given. But is my myth really a myth? Or does the reader not recognize Jones as Man himself in the middle of his journey from the grunts and groans of the cave to the subtle and polydimensional discourse of the drawing room, the laboratory, and the study, the language of Henry and William James, of Einstein and of the philosophers who, in their efforts to break out of discourse to an arché beyond discourse, have provided the most curious dimension of all. I think that the story told in this book, and to be summarized in this final chapter, should be seen as a similar kind of myth. We have no direct evidence about the emergence of rules or language or reason, so all we can do is speculate, and all we can present are only little more than myths. The rationale behind such mythmaking is the hope that it is through such myths as the Sellarsian one that the reader can glimpse “Man himself in the middle of his journey from the grunts and groans of the cave to the subtle and polydimensional discourse of the drawing room, the laboratory […]”. The point is that certain things are very hard to comprehend when they are not embedded in a story. Moreover, it was not my intention to put forward any stories not based on evidence as something the reader was to unquestionably believe. I put forward my stories as examples of how we humans could have gotten “from the grunts and groans of the cave” to our “subtle and polydimensional discourse”. I think that here the aim is not choosing the most credible of a spectrum of stories, but to document that there is such a coherent story at all. Thus, where does our myth begin? DOI: 10.4324/9781003388876-18

Conclusion  211 Every animal assumes certain practical attitudes to elements of its environment, treating them as food, as dangerous predators, as hindrances to be overcome, as instruments to be used etc. Of course, insofar as its conspecifics are also part of the environment, this applies to them too: they can be treated, e.g., as “friends” or as “foes”. For some reason, however, we humans started to assume a specific kind of attitude to our peers and especially to what they are doing. Here is a summary story (needless to say that it is speculative and hypothetical) for how this could have happened. Aside from assuming the usual kinds of attitudes to what our conspecifics do, i.e. counteracting what is harmful to us and supporting what is profitable, we came to concentrate on the harmfulness and profitability of acts as independent of our own perspective; we concentrated on harmfulness and profitability, as it were, per se. If somebody hurts me, I defend myself and try to counteract what she is doing, but this new kind of attitude makes me try to counteract whenever anybody hurts anybody else. This new kind of attitude also comes to be aligned across our communities, thus creating something that can be called the attitudes of the communities, not only of its individual members. Thus, hurting anybody comes to be subject to counteracting not only for me, but also for other members of my community, and hence for the whole community. Tomasello (2014, p. 113) comments on the establishment of such an “agent-neutral” (as he calls it) perspective and the rules based on it in the following way: Modern humans faced different kinds of coordination problems, namely, those involving unknown others, with whom one had little or no personal common ground. The solution on the behavioral level was the creation of group-wide, agent-neutral conventions, norms, and institutions, to which everyone expected everyone, in cultural common ground, to conform. The acts that come to be counteracted in this new vein may well be termed incorrect, while those that are supported correct; and the attitudes of counteracting and supporting may be called normative. Normative attitudes thus came to constitute rudimentary implicit rules. In the simplest case, such a rule is just a cluster of aligned normative attitudes. If my community comes to counteract hurting others, in a coordinated way, we can say that the community follows the rule that prohibits hurting. Why and how did normative attitudes come into being? The inevitable part of the answer to this question is because it was useful, i.e. it helped (in such or another way) enhance fitness (or was a by-product of something that did etc.). There seem to be two different ways in which the attitudes may come to be helpful for our fitness. First, our species, as well as other

212 Conclusion social species, profited from social coherence and hence displayed a strong tendency to conformity. Thus, stepping out of line may well have come to be the subject of discouragement, of negative normative attitudes. This might have given rise to the most rudimentary version of normative attitudes. Once we reach, in some or another way, some uniform ways of behavior, this uniformity may start to be felt as correct and rewarded, while disrupting it may come to be felt as incorrect and punished. This is the correctness we termed “Protagorean”, for here to be correct is just to do what is positively sanctioned. Second, a social species faces the pitfalls of the environment collectively, letting the individuals play their particular roles and letting the environment exert its feedback on them also only indirectly. It is not so, at least not typically, that an individual hunts alone, having to collect the result of her effort alone (to suffer or die of hunger or to survive with a full belly). Typically, hunting is a collective enterprise, with the partakers assuming different roles and then facing the result also collectively. They share the quarry in a way that may have more to do with their social hierarchy than with the contributions to the hunt. Hence, another kind of incentive for normative attitudes is handling the environment collectively. We saw an example: emitting the same kind of sound in case of danger may usefully act as a warning. More generally, if doing something in some kind of circumstance may come to forward the benefits of the society, normative attitudes are likely to appear and to fix this act as something correct. The way we have just sketched may seem to presuppose group selection in a naively straightforward way: if something is profitable for a group, then it is likely to be sustained. Of course, things are not that simple. But on the other hand, recent models of the evolution of social species (Bowles & Gintis, 2011) have made it plain that profit for groups count (even given that groups are not replicators in the sense in which genes are), for being in a more prosperous group is something that is beneficial for an individual and can even outweigh her natural tendency to take a free ride and thus undermine the group. I do not pretend that I am able to describe, even just hypothetically, all the details of the mechanism that brought about the normative attitudes, and hence rules, into being. But we can see they did come into being – we see rules constituting the “social infrastructure” of our societies. Moreover, if we believe that rules, as I have argued throughout the book up to now, constitute the basis also of our language and our reason, then we must see that acquiring the ability to create, maintain and follow them must have constituted a crucial switch in our evolutionary trajectory. It is notoriously difficult to say something exact and

Conclusion  213 indubitable about our pre-history; the most we can expect is the indication that something like this could have happened. And this is what I have tried to do. Let me summarize what had to be realized in order for us to become rule-followers in the way that I suggest we did. First, we must have started to assume normative attitudes – evaluative and regulative attitudes directed at the behavior of others (and consequently ourselves) concentrating only on the kind of behavior, not on its protagonists. Second, we must have had the tendency to align these attitudes across our societies. Third, we must have developed the tendency to turn our normative attitudes into what we called “supernormative attitudes” (see Section 10.3) – the tendency, that is, to perceive the normative attitudes not merely as something to submit to, but rather also as something that “ought to be”, i.e. that one should help enforce. If this is satisfied, then rules and their correlates can be handed down from generation to generation, without the help of genetics, and the virtuous circle of “cultural inheritance” can get going. 17.2  We have come to build institutions So far we have concentrated on rules and rule-following, but an objection may be raised that non-human animals also have rules. This is a notoriously complex question, partly terminological, but let us not fight a trench war about it now. Let us move from individual rules to systems of rules, where our uniqueness, I believe, is utterly beyond doubt. Aside from individual rules, we have come to build systems of interconnected rules, the constituent rules of which are not self-standing, but lean on each other. Some such systems are known as institutions; especially those that play some important role within our societies and that are explicitly stated and backed up by an administrative apparatus. How can such systems have come into being? Again, we can only speculate. It might be that some originally self-standing rules come to interconnect and lose their self-standingness. This was our example with the rudimentary rules of language and with the game of indication. Here rules regulating sounds of display which were originally useful one by one (as warnings or for other purposes) came to interconnect and constitute the rudimentary system of language. Of course, when language was already on the scene, there emerged the possibility of producing systems of rules by fiat, viz. by constructing and stipulating them. Hence, the system of the rules of language may be the mother of all our institutions. What is remarkable about such systems is that they created, as it were, inside spaces, spaces delimited by the rules and with different

214 Conclusion kinds of dynamics than the outside world. It is not easy to find somebody who pointed out this aspect of rules explicitly, but here is Hart (1961, pp. 55–56): A social rule has an “internal” aspect, in addition to the external aspect which it shares with a social habit and which consists in the regular uniform behaviour which an observer could record. This internal aspect of rules may be simply illustrated from the rules of any game. Chess players do not merely have similar habits of moving the Queen in the same way which an external observer, who knew nothing about their attitude to the moves which they make, could record. In addition, they have a reflective critical attitude to this pattern of behaviour: they regard it as a standard for all who play the game. Each not only moves the Queen in a certain way himself but “has views” about the propriety of all moving the Queen in that way. These views are manifested in the criticism of others and demands for conformity made upon others when deviation is actual or threatened and in the acknowledgement of the legitimacy of such criticism and demands when received from others. For the expression of such criticisms, demands, and acknowledgements a wide range of “normative” language is used. “I (You) ought not to have moved the Queen like that”, “I (You) must do that”, “That is right”, “That is wrong”. The main merit of systems of rules is that they allow for new kinds of actions which can be carried out only within them. For example, in the space of meaningfulness that constitutes our language, it is possible to make reports, ask questions, express doubts, cry for help etc. Some of these actions may make things happen without the energetic investment that would be necessary for this outside of these spaces (Knight, 2008). Within the space of meaningfulness, we can make other people do complicated kinds of things, doing nothing more than displaying a suitable sound. Other institutions then allow us to marry, to study at university, to practice a favorite sport or to take out an insurance policy. The institutions turned out to be quite suitable for the development of the kind of ultra-socially characteristic of us humans. Based mostly on the space of meaningfulness, we have developed a lot of other spaces and we have moved our lives to a great extent to these new virtual habitats. We thereby completely changed our form of life because we have come to peruse the possibilities offered by our new virtual environments and have largely forfeited the possibilities offered to us by Mother Nature. In this way, we have exacted, from evolution, a certain space where we can act partly independently of it, ahead of it, and sometimes even counter

Conclusion  215 to it. Creating a system of rules, we need not make every single rule fitnessenhancing (for otherwise evolution would wipe it out before we created other ones); in the “sheltered workshops” of our normative spaces, we can produce things, such as new systems of rules, without the direct interference of evolution, and let the results be sanctioned by evolution only in a longer run. But is this rebellion against evolution not a mere fantasy? We know that everything in the biological world is inevitably governed by evolution, so how could we possibly come to avert it? Of course, what happened was no reversion of evolution, but merely a realignment of the impacts of its forces.1 We have managed to create relatively sheltered spaces (where evolution does not impact directly) by absorbing most of the impacts at their boundaries. In a less metaphoric idiom, we can say that as we face evolution collectively (“ultra-collectively”, in comparison with other social animals), we have specialists for the tasks that it makes us face and we also have specialists who do not face it directly, but do other useful things that can then be employed by those who do. 17.3  We have come to build the space of meaningfulness Systems of rules, we saw, can open up their “inner spaces”, with their own affordances and dynamics. And such systems constitute not only what we tend to see as our institutions (and which we usually have no problem seeing as our creations), but also, I argued, such enhancements of our species as our language and our reason. This has to do with the establishment of the most important of our systems of rules, opening up the space of meaningfulness, the space that offered the possibility of meaningful communication. Its emergence vastly enhances the possibilities we have, including the possibilities of building more systems of rules. There are various ways in which the rudiments of language could have come into being. We saw a hypothetic example: rules for displaying sounds that were originally useful individually came to interlock in a way that made them dependent on each other. I do not mean to suggest that this is the way a basis of language truly came into being; there are alternative stories that may be more plausible. But the one I presented is transparent and hence useful to show the possibility of such a development. (Let me sketch one alternative, which seems to me to be more plausible than the one given above, but which I am, at present, not able to articulate with sufficient clarity. I think that our pre-linguistic ancestors may have used the sounds they emit as a kind of weapon, as something to fight with in some ritualized combats. (The ritualization, of course, was a matter of moving combat from an open into a rule-governed space.) At first, it might have been that the louder and scarier the sound is, the more effective a

216 Conclusion weapon it is, but then various kinds of sounds may acquire various roles (due to the rules of the combat and perhaps also of some other activities outside the combat in which they also figure) and their effectiveness within the combat may start to derive from these roles – from their incipient content.) In any case, the language game that was crucially important and that made our language into what it is the game of giving and asking for reasons, GOGAR. It established what Sellars aptly called “the space of reasons” (Sellars, 1956, p. 169) as a powerful laboratory in which we can, in Popper’s (1972, p. 244), memorable words, “permit our hypotheses to die in our stead”. Therefore, the kinds of sounds (and later also inscriptions) that were to become expressions of natural language were identified especially by the roles they assumed within GOGAR – their inferential roles. True, empirical expressions, like tiger, dog or bicycle, must be linked to the world – to tigers, dogs or bicycles, respectively. However, it would be a misunderstanding to think that this must take place outside of GOGAR, that GOGAR is self-contained and severed from the world outside of it. GOGAR is not self-contained in this sense – it is an enterprise that incorporates the world; many of its rules involve the world essentially. (It is correct to say “This is a dog” when pointing at a dog.) The crucial role assumed within our societies by GOGAR and other language games also could not stop short of affecting our cognition. The cognition (or its upper layer) thus became discursive – we have come to articulate our thoughts in terms of the concepts we carried out from public discourse and we learn to reason by internalizing imaginary discursive and argumentative partners. Language and GOGAR also completely changed our enterprise of establishing rules by building spaces delimited by them. There emerged the possibility of explicit rules. We could make the existing implicit rules explicit, and we could bring new rules into being by prescribing them. (Of course, this presupposes also an already existing system of authorities, so that the one making the prescription has the authority to back it up.) In a literate society, institutions can also be fortified by various kinds of measures – from authorities who decide who has broken the relevant rules and who not, to those who systematically punish those who did. This completely upgrades the possibilities of rule-governedness available up to this point, and our rule-governed behavioral patterns, practices and societies, further alienating us from the rest of the biological world. We have become the unique linguistic, literate, historical and

Conclusion  217 rational species, and for the description of our antics we also need, over and above biology, “higher level” theories like sociology, history, psychology, linguistics or logic. 17.4  We cannot live outside of rules In Section 9.1, we quoted Sellars (1949, p. 315): “To talk about rules is to move outside the talked-about rules into another framework of living rules. (The snake which sheds one skin lives within another.)” Why, moving outside of a system of rules, must we end up within another system of rules? Why can we only “shed our skin” after we have another one under it? The short answer to this question is that skinless we are not able to face the environment, that rules and institutions have become our “second nature” to such an extent that we are not able to exist without their coordination system. But let us give a slightly extended version of the answer. A human being, from its early childhood, is raised to observe and abide by rules; she is literally made to grow into our rule-governed, institutional world. This is the niche she learns to inhabit, to cope with and to feel at home in. We have speculated that at first the normative boundaries of the world (what should not be done) blend with the causal ones (what cannot be done), but we saw that relatively early a child becomes able to identify normative boundaries as something that not only restrains her, but that also counts on her support. It is relatively easy in the practical sense: in the sense that the child comes to not only abide by the rule herself but also assumes the corresponding normative attitudes to make others abide by it. This is what we dubbed the supernormativity of our human rules. However, aside from this practical recognition of rules, there is a reflective recognition, which is a much deeper achievement. This amounts to the understanding that rules are of our making and that they can be changed – some as a matter of principle, and some for real. This is grasping the fact that we humans are beings that crave for Kantian autonomy – for having a say in which rules we follow. And if we are lucky, viz. if we do not live under the yoke of a despot, we are autonomous. This reflective recognition of rules is made possible by our language, which makes it possible to make our rules explicit. Before we have the means of such an explicitation, we can only abide by – or, as the case may be, resist abiding by – the rules, but once we are able to make them explicit, we can make rules subject to GOGAR, weigh their pros and cons and eventually modify them or possibly wholly abandon them. In this way, we become self-reflective beings. (Needless to say, this is the process given so much attention by Brandom, 1994.)

218 Conclusion Another achievement of our self-reflection is that we can switch between our view of rules from within and from without. We are anchored in our rules, and given this, they are absolute determinants of our form of life. Certain things are right, others are wrong – not just right or wrong for us, they are right or wrong, full-stop. However, insofar as we are able to rise above our rules (and here Sellars points out that this can be done only by leaning on some other rules), we can see that their validity for us does not have any metaphysical underpinning, that from inside we see them as absolutely valid just because they are ours. Richard Rorty (1989, p. 189) makes this point particularly vividly: I have been urging in this book that we try not to want something which stands beyond history and institutions. The fundamental premise of the book is that a belief can still regulate action, can still be thought worth dying for, among people who are quite aware that this belief is caused by nothing deeper than contingent historical circumstance. Once we give up metaphysics and religious faith, we are left with the picture of our norms, including the moral ones, sketched in this book. The rules are contingent creations of ours – true, we have not quite developed them randomly but in response to the environment in which we lived and in reaction to the fact that we needed to come to terms with each other, but they are not imprints of any necessary structures shaping our form of life from without. We cannot live wholly outside of rules, as most of the rules that we accept are like the air that we need to breathe. And the most basic rules that we accept are akin to our need for our sources of air. True, we may hold our breath for a while and immerse ourselves in another medium, but this is just a temporary exercise from which we must quickly remove ourselves and surface back into our rules. 17.5 Summary By some strange quirks of evolution, we humans have come to follow rules and thus make rules into indispensable parts of our world. Moreover, we have come to interweave these rules into institutions, which we have adopted as our virtual dwellings, essentially changing our form of life. We have developed a space of meaningfulness in which we can use a wide variety of sounds as precise tools of communication with others. In this way, we have become normative animals.

Conclusion  219 Note 1 This problem seems to be structurally similar to that in which the second law of thermodynamics holds and yet we are able to build systems that are organized and hence appear to decrease entropy. Of course, this does not happen because we would be able to violate the laws of physics, but rather because we can locally decrease entropy if we increase it somewhere else – we can exact, on the laws of physics, a space where we can work locally ignoring them, because we are able to compensate for this on its boundaries and behind them.

References

Alston, W. P., ed. 2002. Realism and Antirealism. New York: Cornell University Press. Andrews, K. 2020. “Naïve Normativity: The Social Foundation of Moral Cognition.” Journal of the American Philosophical Association 6 (1): 36–56. Arbib, M. A. 2006. “The Mirror System Hypothesis on the Linkage of Action and Language.” In Action to Language via the Mirror Neuron System, edited by M. A. Arbib, 3–47. Cambridge: Cambridge University Press. Axelrod, R. 1984. The Evolution of Cooperation. New York: Basic Books. Baker, G. P., and P. M. S. Hacker. 1984. Scepticism, Rules and Language. Oxford: Blackwell. Barnes, E. C. 2002. “The Miraculous Choice Argument for Realism.” Philosophical Studies 111 (2): 97–120. Beisecker, D. 2013. “‘From the Grunts and Groans of the Cave….’ Presidential Address.” Southwest Philosophy Review 29 (1): 1–11. Beran, O. 2022. “The Moral Luck of Rules.” Philosophical Investigations 45 (1): 21–39. Berniūnas, R., V. Silius, and V. Dranseika. 2022. “Moralization East and West: Moralizing Different Transgressions among Chinese, Americans, and Lithuanians.” Asian Journal of Social Psychology 25 (2): 185–97. Berwick, R. C., and N. Chomsky. 2016. Why Only Us: Language and Evolution. Cambridge, MA: MIT Press. Bicchieri, C. 2005. The Grammar of Society: The Nature and Dynamics of Social Norms. Cambridge: Cambridge University Press. Binmore, K. 2007. Game Theory: A Very Short Introduction. Oxford: Oxford University Press. Blackburn, S. 1984. Spreading the Word. Oxford: Clarendon Press. Blackmore, S. 1999. The Meme Machine. Oxford: Oxford University Press. Boghossian, P. 2014. “What Is Inference?” Philosophical Studies 169 (1): 1–18. Boghossian, P. A. 1989. “The Rule-Following Considerations.” Mind 98 (392): 507–49. Bowles, S. 2009. Microeconomics. Princeton: Princeton University Press. Bowles, S., and H. Gintis. 2011. A Cooperative Species: Human Reciprocity and Its Evolution. Princeton: Princeton University Press.

References 221 Boyd, R., H. Gintis, S. Bowles, and P. J. Richerson. 2003. “The Evolution of Altruistic Punishment.” PNAS 100 (6): 3531–5. Boyd, R. P., J. Richerson, and J. Henrich. 2011. “The Cultural Niche.” PNAS 108 (Supplement 2): 10918–25. Boyle, M. 2016. “Additive Theories of Rationality: A Critique.” European Journal of Philosophy 24 (3): 527–55. Brandom, R. 1979. “Freedom and Constraint by Norms.” American Philosophical Quarterly 16 (3): 187–96. Brandom, R. 1994. Making It Explicit: Reasoning, Representing, and Discursive Commitment. Cambridge, MA: Harvard University Press. Brandom, R. 2000. Articulating Reasons. Cambridge, MA: Harvard University Press. Brandom, R. 2008. Between Saying and Doing: Towards an Analytic Pragmatism. Oxford: Oxford University Press. Brandom, R. 2009. Reason in Philosophy: Animating Ideas. Cambridge, MA: Harvard University Press. Brandom, R. 2019. A Spirit of Trust: A Reading of Hegel’s Phenomenology. Cambridge, MA: Harvard University Press. Bratman, M. E. 1992. “Shared Cooperative Activity.” The Philosophical Review 101 (2): 327–41. Bratman, M. E. 2013. Shared Agency: A Planning Theory of Acting Together. Oxford: Oxford University Press. Brennan, G., L. Eriksson, R. E. Goodin, and N. Southwood. 2013. Explaining Norms. Oxford: Oxford University Press. Campbell, J. D., and P. J. Fairey. 1989. “Informational and Normative Routes to Conformity: The Effect of Faction Size as a Function of Norm Extremity and Attention to the Stimulus.” Journal of Personality and Social Psychology 57 (3): 457–68. Carruthers, P. 2011. The Opacity of Mind: An Integrative Theory of SelfKnowledge. Oxford: Oxford University Press. Carruthers, P. 2013. “The Distinctively Human Mind: The Many Pillars of Cumulative Culture.” In The Evolution of Mind, Brain, and Culture, edited by G. Hatfield and H. Pittman, 325–46. Philadelphia: University of Pennsylvania Press. Casler, K., T. Terziyan, and K. Greene. 2009. “Toddlers View Artifact Function Normatively.” Cognitive Development 24 (3): 240–47. Castro, L., L. Castro-Nogueira, M. A. Castro-Nogueira, and M. A. Toro. 2010. “Cultural Transmission and Social Control of Human Behavior.” Biology & Philosophy 25 (3): 347–60. Ceci, S. J., and W. M. Williams, eds. 1999. The Nature-Nurture Debate: The Essential Readings. Malden: Blackwell. Chemero, A., and M. Turvey. 2007. “Complexity, Hypersets, and the Ecological Perspective on Perception-Action.” Biological Theory 2 (1): 23–36. Chomsky, N. 1955. “Logical Syntax and Semantics: Their Linguistic Relevance.” Language 31 (1): 36–45. Chomsky, N. 1986. Knowledge of Language. Westport: Praeger. Chomsky, N. 2000. New Horizons in the Study of Language and Mind. Cambridge: Cambridge University Press.

222 References Chong, I., and R. W. Proctor. 2020. “On the Evolution of a Radical Concept: Affordances According to Gibson and Their Subsequent Use and Development.” Perspectives on Psychological Science 15 (1): 117–32. Christiansen, M. H., and N. Chater. 2016. Creating Language: Integrating Evolution, Acquisition, and Processing. Cambridge, MA: MIT Press. Christias, D. 2015. “A Sellarsian Approach to the Normativism-Antinormativism Controversy.” Philosophy of the Social Sciences 45 (2): 143–75. Claidière, N., and A. Whiten. 2012. “Integrating the Study of Conformity and Culture in Humans and Nonhuman Animals.” Psychological Bulletin 138 (1): 126–45. Clément, F., S. Bernard, and L Kaufmann. 2011. “Social Cognition Is Not Reducible to Theory of Mind: When Children Use Deontic Rules to Predict the Behaviour of Others.” British Journal of Developmental Psychology 29 (4): 910–28. Cook, R. T. 2010. “Let a Thousand Flowers Bloom: A Tour of Logical Pluralism.” Philosophy Compass 5 (6): 492–504. Cosmides, L., and J. Tooby. 2005. “Neurocognitive Adaptations Designed for Social Exchange.” In The Handbook of Evolutionary Psychology, edited by D. M. Buss, 584–627. Hoboken: Wiley. Cresswell, M. J. 1973. Logic and Languages. London: Methuen. Curry, O. S. 2016. “Morality as Cooperation: A Problem-Centred Approach.” In The Evolution of Morality, edited by T. K. Shackelford and R. D. Hansen, 27–51. Cham: Springer. Curry, O. S., M. Alfano, M. J. Brandt, and C. Pelican. 2022. “Moral Molecules: Morality as a Combinatorial System.” Review of Philosophy and Psychology 13: 1039–58. Davidson, D. 1970. “Mental Events.” In Experience and Theory, edited by. L. Foster and J. W. Swanson. London: Duckworth; reprinted in Davidson: Essays on Actions and Events, 207–43, Oxford: Clarendon Press, 1980. Davidson, D. 1974. ‘On the Very Idea of a Conceptual Scheme’. Proceedings and Addresses of the American Philosophical Association 47; reprinted in Davidson: Inquiries into Truth and Interpretation, pp. 183–98. Oxford: Clarendon, 1984. Davidson, D. 1994. “The Social Aspect of Language.” In The Philosophy of Michael Dummett, edited by B. McGuinness and G. Oliveri, 1–16. Dordrecht: Springer. Davidson, D. 2001. Subjective, Intersubjective, Objective. New York: Oxford University Press. Dawkins, R. 1976. The Selfish Gene. New York: Oxford University Press. Dennett, D. 1996. Kinds of Minds. New York: Basic Books. Dennett, D. 2013. “The Evolution of Reasons.” In Contemporary Philosophical Naturalism and Its Implications, edited by B. Bashour and H. D. Muller, 47–62. New York: Routledge. Dennett, D. 2018a. “The Phantasy of First-Person Science.” In The Map and the Territory. The Frontiers Collection, edited by S. Wuppuluri and F. Doria. Cham: Springer. Dennett, D. 2018b. “Facing Up to the Hard Question of Consciousness.” Philosophical Transactions of the Royal Society B 373 (1755): 20170342.

References 223 deVries, W. A. 2005. Wilfrid Sellars. Chesham: Acumen. Dewey, J. 1925. Experience and Nature. New York: Dover. Dor, D., and E. Jablonka. 2014. ‘Why We Need to Move from GeneCulture Co-Evolution to Culturally Driven Co-Evolution’. In The Social Origins of Language, edited by D. Dor, C. Knight, and J. Lewis, 15–30. Oxford: Oxford University Press. Dor, D., C. Knight, and J. Lewis. 2014. ‘Introduction: A Social Perspective on How Language Began’. In The Social Origins of Language, edited by D. Dor, C. Knight, and J. Lewis, 1–12. Oxford: Oxford University Press. Drobňák, M. 2022. “Normative Inferentialism on Linguistic Understanding.” Mind & Language 37 (4): 564–85. Dummett, M. 1959. ‘Truth’. Proceedings of the Aristotelian Society (New Series) 59; reprinted in Dummett: Truth and Other Enigmas, 1–24. London: Duckworth, 1978. Dutilh Novaes, C. 2020. The Dialogical Roots of Deduction: Historical, Cognitive, and Philosophical Perspectives on Reasoning. Cambridge: Cambridge University Press. Evans, J. S. B. 2003. “In Two Minds: Dual-Process Accounts of Reasoning.” Trends in Cognitive Sciences 7 (10): 454–59. Evans, J. S. B., and D. E. Over. 2013. Rationality and Reasoning. Hove: Psychology Press. Fehr, E., and S. Gächter. 2002. “Altruistic Punishment in Humans.” Nature 415 (6868): 137. Fernald, A., and V. A. Marchman. 2006. “Language Learning in Infancy.” In Handbook of Psycholinguistics, edited by M. J. Traxler and M. A. Gernsbacher, 1027–71. Amsterdam: Elsevier. Fitzpatrick, S. 2020. “Chimpanzee Normativity: Evidence and Objections.” Biology & Philosophy 35 (4): 1–28. Fodor, J. A. 1975. The Language of Thought. Cambridge, MA: Harvard University Press. Fodor, J. A. 2008. LOT 2: The Language of Thought Revisited. Oxford: Oxford University Press. Frost-Arnold, G. 2010. “The No-Miracles Argument for Realism: Inference to an Unacceptable Explanation.” Philosophy of Science 77 (1): 35–58. Gauker, C. 2007. “The Circle of Deference Proves the Normativity of Semantics.” Rivista Di Estetica 34: 181–98. Gauker, C. 2011. Words and Images: An Essay on the Origin of Ideas. Oxford: Oxford University Press. Gergely, G., and G. Csibra. 2006. “Sylvia’s Recipe: The Role of Imitation and Pedagogy in the Transmission of Cultural Knowledge.” In Roots of Human Sociality: Culture, Cognition, and Human Interaction, edited by S. C. Levinson, 229–55. Oxford: Berg. Gibson, J. J. 1979. The Ecological Approach to Visual Perception. Boston: Houghton Mifflin. Ginsborg, H. 2011. “Primitive Normativity and Skepticism about Rules.” The Journal of Philosophy 108 (5): 227–54.

224 References Gintis, H. 2011. “Gene-Culture Coevolution and the Nature of Human Sociality.” Philosophical Transactions of the Royal Society B: Biological Sciences 366 (1566): 878–88. Giromini, J. (forthcoming): ‘The Logical Structure of Normative Attitudes’. Philosophia. Glanzberg, M. 2015. “Logical Consequence and Natural Language.” In Foundations of Logical Consequence, edited by C. R. Caret and O. T. Hjortland, 71–120. Oxford: Oxford University Press. Glock, H.-J. 2012. “The Anthropological Difference: What Can Philosophers Do to Identify the Differences between Human and Non-Human Animals?” Royal Institute of Philosophy Supplements 70: 105–31. Glüer, K., and Å. Wikforss. 2009. “Against Content Normativity.” Mind 118 (469): 31–70. Goldfarb, W. 1985. “Kripke on Wittgenstein on Rules.” The Journal of Philosophy 82 (9): 471–88. Goldhaber, D. 2012. The Nature-Nurture Debates: Bridging the Gap. Cambridge: Cambridge University Press. Goodman, N. 1955. Fact, Fiction, and Forecast. Cambridge, MA: Harvard University Press. Goodman, N. 1960. “The Way the World Is.” The Review of Metaphysics 14 (1): 48–56. Greene, J. 2008. “The Secret Joke of Kant’s Soul.” Moral Psychology, Vol. 3: The Neuroscience of Morality: Emotion, Disease, and Development, edited by W. Sinnott-Armstrong. Cambridge, MA: MIT Press. Hahn, H., O. Neurath, and R. Carnap. 1929. Wissenschaftliche Weltauffassung: Der Wiener Kreis. Wien: Wolf; English translation The Scientific Conception of the World: The Vienna Circle in S. Sarkar and R. Nozick, ed., The Emergence of Logical Empiricism: From 1900 to the Vienna Circle. London: Routledge, 1996. Haidt, J. 2001. “The Emotional Dog and Its Rational Tail: A Social Intuitionist Approach to Moral Judgment.” Psychological Review 108 (4): 814–34. Haidt, J. 2012. The Righteous Mind: Why Good People Are Divided by Politics and Religion. New York: Pantheon. Hamilton, W. D. 1963. “The Evolution of Altruistic Behavior.” The American Naturalist 97 (896): 354–6. Hanfling, O. 2013. Philosophy and Ordinary Language: The Bent and Genius of Our Tongue. London: Routledge. Hart, H. L. A. 1961. The Concept of Law. Oxford: Oxford University Press. Hattiangadi, A. 2003. “Making It Implicit: Brandom on Rule Following.” Philosophy and Phenomenological Research 66 (2): 419–31. Haugeland, J. 1982. “Heidegger on Being a Person.” Nous 16 (1): 15–26. Haugeland, J. 2000. “Truth and Rule-Following.” In Haugeland: Having Thought, 305–62. Cambridge, MA: Harvard University Press. Heft, H. 1989. “Affordances and the Body: An Intentional Analysis of Gibson’s Ecological Approach to Visual Perception.” Journal for the Theory of Social Behaviour 19 (1): 1–30. Heidegger, M. 1927. Sein und Zeit. Halle: Niemeyer.; English translation Being and Time. New York: Harper & Row, 1962.

References 225 Heidegger, R. 1959. Unterwegs zur Sprache. Pfullingen: Günther Neske.; English translation On the Way to Language. New York: Harper & Row, 1982. Henrich, J. 2015. The Secret of Our Success: How Culture Is Driving Human Evolution, Domesticating Our Species, and Making Us Smarter. Princeton: Princeton University Press. Henrich, J., and F. J. Gil-White. 2001. “The Evolution of Prestige: Freely Conferred Deference as a Mechanism for Enhancing the Benefits of Cultural Transmission.” Evolution and Human Behavior 22 (3): 165–96. Hildebrand, D. 2018. “Dewey, Rorty, and Brandom: The Challenges of Linguistic Neopragmatism.” In The Oxford Handbook of Dewey, edited by S. Fesmire, 99–130. Oxford: Oxford University Press. Hjortland, O. T. 2017. “Anti-Exceptionalism about Logic.” Philosophical Studies 174 (3): 631–58. Hlobil, U. 2021. “Limits of Abductivism about Logic.” Philosophy and Phenomenological Research 103 (2): 320–40. Hofstadter, D. 1995. “To Seek Whence Cometh a Sequence.” In Fluid Concepts and Creative Analogies, 13–86. New York: Basic Books. Hudson, W. D., ed. 1969. The Is/Ought Question. A Collection of Papers on the Central Problem in Moral Philosophy. London: Macmillan. Hurford, J. R. 2007. The Origin of Meaning. Oxford: Oxford University Press. Hurley, S. 2008. “The Shared Circuits Model (SCM): How Control, Mirroring, and Simulation Can Enable Imitation, Deliberation, and Mindreading.” Behavioral and Brain Sciences 31: 1–58. Husserl, E. 1936. die Krisis der europäischen Wissenschaften und die transzendentale Phänomenologie: Eine Einleitung in die phänomenologische Philosophie. Belgrad: Philosophie.; English translation The Crisis of European Sciences and Transcendental Phenomenology: An Introduction to Phenomenological Philosophy. Evanston: Northwestern University Press, 1989. Jablonka, E., and M. J. Lamb. 2014. Evolution in Four Dimensions. (revised edition). Cambridge, MA: MIT Press. Joyce, R. 2006. The Evolution of Morality. Cambridge, MA: MIT Press. Kahneman, D. 2011. Thinking, Fast and Slow. New York: Farrar, Straus and Giroux. Kalish, C. 1998. “Reasons and Causes: Children’s Understanding of Conformity to Social Rules and Physical Laws.” Child Development 69 (3): 706–20. Kant, I. 1785. Grundlegung zur Metaphysik der Sitten. Riga: Hartknoch.; English translation Groundwork for the Metaphysics of Morals (Practical Philosophy The Cambridge Edition of the Works of Immanuel Kant), edited by M. J. Gregor, trans. P. Guyer and A. W. Wood. Cambridge: Cambridge University Press, 2012. Kant, I. 1790. Kritik der Urteilskraft. Berlin: Lagarde und Friedrichz; English translation Critique of the Power of Judgment (The Cambridge Edition of the Works of Immanuel Kant). Kelly, D., and T. Davis. 2018. “Social Norms and Human Normative Psychology.” Social Philosophy and Policy 35 (1): 54–76. Kendal, J. R. 2011. “Cultural Niche Construction and Human Learning Environments: Investigating Sociocultural Perspectives.” Biological Theory 6 (3): 241–50.

226 References Kenward, B., M. Karlsson, and J. Persson. 2011. “Over-Imitation Is Better Explained by Norm Learning than by Distorted Causal Learning.” Proceedings of the Royal Society B: Biological Sciences 278 (1709): 1239–46. Kern, A., and H. Moll. 2017. “On the Transformative Character of Collective Intentionality and the Uniqueness of the Human.” Philosophical Psychology 30 (3): 319–37. Kitcher, P. 2011. The Ethical Project. Cambridge, MA: Harvard University Press. Knight, C. 2008. “Language Co-Evolved with the Rule of Law.” Mind and Society 7 (1): 109–28. Koons, J. R. 2021. “Sellars on Rational Agency as Presupposing Collective Attitudes.” In Groups, Norms and Practices, edited by L. Koreň et al., 189–213. Cham: Springer. Koreň, L. 2018. “Propositional Contents and the Logical Space.” In From Rules to Meanings: New Essays on Inferentialism, edited by O. Beran, V. Kolman, and L. Koreň, 196–218. New York: Routledge. Koreň, L. 2021. Practices of Reason: Fusing the Inferentialist and Scientific Image. New York: Routledge. Koreň, L., H. B. Schmid, P. Stovall, and L. Townsend, eds. 2021. Groups, Norms and Practices. Cham: Springer. Krebs, J. R., and R. Dawkins. 1984. “Animal Signals: Mindreading and Manipulation.” In Behavioural Ecology: An Evolutionary Approach, edited by J. R. Krebs and N. B. Davies, 380–402. London: Blackwell. Krikorian, Y., ed. 1944. Naturalism and the Human Spirit. New York: Columbia University Press. Kripke, S. A. 1982. Wittgenstein on Rules and Private Language: An Elementary Exposition. Cambridge, MA: Harvard University Press. Kumar, V., and R. Campbell. 2022. A Better Ape: The Evolution of the Moral Mind. Oxford: Oxford University Press. Kusch, M. 2002. Knowledge by Agreement: The Programme of Communitarian Epistemology. Oxford: Clarendon Press. Kusch, M. 2006. A Sceptical Guide to Meaning and Rules: Defending Kripke’s Wittgenstein. Montreal: McGill-Queen’s University Press. Laland, K. 2017. Darwin’s Unfinished Symphony. Princeton: Princeton University Press. Laland, K. N., and M. J. O’Brien. 2011. “Cultural Niche Construction: An Introduction.” Biological Theory 6 (3): 191–202. Lance, M. N. 1998. “Some Reflections on the Sport of Language.” Philosophical Perspectives 12: 219–40. Lau, H., and D. Rosenthal. 2011. “Empirical Support for Higher-Order Theories of Conscious Awareness.” Trends in Cognitive Sciences 15: 365–73. Lewens, T. 2017. “Human Nature, Human Culture: The Case of Cultural Evolution.” Interface Focus 7 (5): 20170018. Lewis, D. 1969. Convention: A Philosophical Study. London: Blackwell. Lewis, D. 1979. “Scorekeeping in a Language Game.” Journal of Philosophical Logic 8 (1): 339–59.

References 227 Leyton-Brown, K., and Y. Shoham. 2008. “Essentials of Game Theory: A Concise Multidisciplinary Introduction.” Synthesis Lectures on Artificial Intelligence and Machine Learning 2 (1): 1–88. Lorenzen, P. 1955. Einführung in die operative Logik und Mathematik. Berlin: Springer. Lorenzen, P., and K. Lorenz. 1978. Dialogische Logik. Darmstadt: Wissenschaftliche Buchgesellschaft. Lorenzen, P., and O. Schwemmer. 1975. Konstruktive Logik, Ethik und Wissenschaftstheorie. Mannheim: Bibliographisches Institut. Lycan, W. G. 2001. “A Simple Argument for a Higher-Order Representation Theory of Consciousness.” Analysis 61 (1): 3–4. Mandelkern, M. 2020. “How to Do Things with Modals.” Mind & Language 35 (1): 115–38. Mares, E. D. 2004. Relevant Logic: A Philosophical Interpretation. Cambridge: Cambridge University Press. Martin, B. 2021. “Identifying Logical Evidence.” Synthese 198 (10): 9069–95. Maryanski, A., and J. H. Turner. 1992. The Social Cage: Human Nature and the Evolution of Society. Stanford: Stanford University Press. Maynard Smith, J. 1964. “Group Selection and Kin Selection.” Nature 201 (4924): 1145–7. Maynard Smith, J. 1976. “Group Selection.” Quarterly Review of Biology 51 (2): 277–83. McDowell, J. 1984. “Wittgenstein on Following a Rule.” Synthese 58 (3): 325–63. McElreath, R., and R. Boyd. 2008. Mathematical Models of Social Evolution: A Guide for the Perplexed. Chicago: University of Chicago Press. McGee, V. 1985. “A Counterexample to Modus Ponens.” Journal of Philosophy 82 (9): 462–71. McKenzie Alexander, J. 2007. The Structural Evolution of Morality. Cambridge: Cambridge University Press. Mead, G. H. 1934. Mind, Self, & Society from the Standpoint of a Social Behaviorist. Chicago: University of Chicago Press. Mercier, H., and D. Sperber. 2011. “Why Do Humans Reason? Arguments for an Argumentative Theory.” Behavioral and Brain Sciences 34 (2): 57–111. Mercier, H., and D. Sperber. 2017. The Enigma of Reason. Cambridge, MA: Harvard University Press. Millikan, R. G. 2013. “The Tangle of Natural Purposes That Is Us”. In Contemporary Philosophical Naturalism and Its Implications, edited by B. Bashour and H. D. Muller, 73–84. New York: Routledge. Millikan, R. G. 1984. Language, Thought, and Other Biological Categories. Cambridge, MA: MIT Press. Millikan, R. G. 1989. “In Defense of Proper Functions.” Philosophy of Science 56 (2): 288–302. Millikan, R. G. 1995. “Pushmi-Pullyu Representations.” Philosophical Perspectives 9: 185–200. Millikan, R. G. 2005. “The Father, the Son, and the Daughter: Sellars, Brandom, and Millikan.” Pragmatics & Cognition 13 (1): 59–71.

228 References Misak, C. J. 2004. Truth and the End of Inquiry: A Peircean Account of Truth. Oxford: Clarendon. Myers, R. H., and C. Verheggen. 2016. Donald Davidson’s Triangulation Argument: A Philosophical Inquiry. London: Routledge. Nichols, S. 2002. “On the Genealogy of Norms: A Case for the Role of Emotion in Cultural Evolution.” Philosophy of Science 69 (2): 234–55. O’Shea, J. 2007. Wilfrid Sellars. Cambridge: Polity. O’Shea, J. 2012. “Prospects for a Stereoscopic Vision of Our Thinking Nature: On Sellars, Brandom, and Millikan.” Humana Mente 5 (21): 149–72. Odling-Smee, F. J., K. N. Laland, and M. W. Feldman. 2003. Niche Construction. Princeton: Princeton University Press. Okrent, M. 2017. Nature and Normativity: Biology, Teleology, and Meaning. New York: Routledge. Packer, M. J., and M Cole. 2019. “Evolution and Ontogenesis: The Deontic Niche of Human Development.” Human Development 62 (4): 175–211. Partee, B. 2018. “Changing Notions of Linguistic Competence in the History of Formal Semantics.” In The Science of Meaning: Essays on the Metatheory of Natural Language Semantics, edited by D. Ball and B. Rabern, 172–96. Oxford: Oxford University Press. Peirce, C. S. 1998. The Essential Peirce: Selected Philosophical Writings Volume 2 (1893–1913). Bloomington: Indiana University Press. Peregrin, J. (forthcoming): “Logic as a Science of Patterns?” The Journal of Philosophy. Peregrin, J. 1995. Doing Worlds with Words. Dordrecht: Kluwer. Peregrin, J. 2001. Meaning and Structure. Aldershot: Ashgate. Peregrin, J. 2008a. “Inferentialist Approach to Semantics.” Philosophy Compass 3 (6): 1208–23. Peregrin, J. 2008b. “What Is the Logic of Inference?” Studia Logica 88 (2): 263–94. Peregrin, J. 2010a. “Language, the World and the Nature of Philosophy.” In The Philosophy of Richard Rorty, edited by R. E. Auxier and L. E. Hahn, 225–45. Open Court: La Salle (Ill.). Peregrin, J. 2010b. “Logic and Natural Selection.” Logica Universalis 4 (2): 207–23. Peregrin, J. 2012. “Semantics without Meaning?” In Prospects of Meaning, edited by R. Schantz, 479–502. Berlin: de Gruyter. Peregrin, J. 2014a. Inferentialism: Why Rules Matter. Basingstoke: Palgrave. Peregrin, J. 2014b. “Rules as the Impetus of Cultural Evolution.” Topoi 33 (2): 51–545. Peregrin, J. 2016a. “Should One Be a Left or Right Sellarsian? (And Is There Really Such a Choice?).” Metaphilosophy 47 (2): 251–63. Peregrin, J. 2016b. “Social Normativism.” In Normativity and Naturalism in the Philosophy of the Social Sciences, edited by M. Risjord, 68–85. London: Routledge. Peregrin, J. 2017. “Is Inferentialism Circular?” Analysis 78 (3): 450–4. Peregrin, J. 2018. “Davidson and Sellars on ‘Two Images.’” Philosophia 46 (1): 183–92.

References 229 Peregrin, J. 2019. “Logic as a (Natural) Science.” In The Logica Yearbook 2018, edited by I. Sedlár and M. Blicha, 177–96. London: College Publications. Peregrin, J. 2020a. “Normative Mindshaping and the Normative Niche.” In Groups, Norms and Practices, edited by L. Koreň et al., pp. 85–98. Cham: Springer. Peregrin, J. 2020b. Philosophy of Logical Systems. New York: Routledge. Peregrin, J., and V. Svoboda. 2017. Reflective Equilibrium and the Principles of Logical Analysis: Understanding the Laws of Logic. New York: Routledge. Peregrin, J., and V. Svoboda. 2021. “Moderate Anti-Exceptionalism and Earthborn Logic.” Synthese 199 (3–4): 8781–806. Peregrin, J., and V. Svoboda. 2022. “Logica Dominans vs. Logica Serviens.” Logic and Logical Philosophy 31 (2): 183–207. Phillips, J., and J. Knobe. 2018. “The Psychological Representation of Modality.” Mind & Language 33 (1): 65–94. Pinkard, T. 1996. Hegel’s Phenomenology: The Sociality of Reason. Cambridge: Cambridge University Press. Pinkard, T. 2007. “Sellars the Post-Kantian?” Poznan Studies in the Philosophy of the Sciences and the Humanities 92 (1): 21–52. Pinker, S. 1995. The Language Instinct. New York: Harper Collins. Pinker, S. 2010. “The Cognitive Niche: Coevolution of Intelligence, Sociality, and Language.” PNAS 107 (Supplement 2): 8993–9. Planer, R., and K. Sterelny. 2021. From Signal to Symbol: the Evolution of Language. Cambridge, MA: MIT Press. Popper, K. R. 1972. Objective Knowledge. Oxford: Oxford University Press. Poundstone, W. 1992. Prisoner’s Dilemma: John von Neumann, Game Theory, and the Puzzle of the Bomb. New York: Doubleday. Premack, D. 1983. “The Codes of Man and Beasts.” Behavioral and Brain Sciences 6 (1): 125–36. Price, H. 2011. Naturalism without Mirrors. New York: Oxford University Press. Priest, G. 2001. An Introduction to Non-Classical Logics. Cambridge: Cambridge University Press. Punčochář, V., and C. Gauker. 2020. “Indicative Conditionals in Objective Contexts.” Theoria 86 (5): 651–87. Putnam, H. 1975. “What Is Mathematical Truth?” Historia Mathematica 2 (4): 529–33; reprinted in Putnam: Mathematics, Matter and Method: Philosophical Papers, 60–78. Cambridge: Cambridge University Press, 1979. Putnam, H. 1991. Representation and Reality. Cambridge, MA: MIT Press. Quine, W. V. O. 1960. Word and Object. Cambridge, MA: MIT Press. Quine, W. V. O. 1969. Ontological Relativity and Other Essays. New York: Columbia University Press. Quine, W. V. O. 1995. From Stimulus to Science. Cambridge, MA: Harvard University Press. Rahman, S., and H. Rückert. 2001. New Perspectives in Dialogical Logic. Dordrecht: Kluwer. Rakoczy, H., and M. F. Schmidt. 2013. “The Early Ontogeny of Social Norms.” Child Development Perspectives 7 (1): 17–21.

230 References Rakoczy, H., F. Warneken, and M. Tomasello. 2008. “The Sources of Normativity: Young Children’s Awareness of the Normative Structure of Games.” Developmental Psychology 44 (3): 875. Ramstead, M. J., S. P. Veissière, and L. J. Kirmayer. 2016. “Cultural Affordances: Scaffolding Local Worlds through Shared Intentionality and Regimes of Attention.” Frontiers in Psychology 7: 1090. Rawls, J. 1955. “Two Concepts of Rules.” Philosophical Review 64 (1): 3–32. Read, S. 1994. “Formal and Material Consequence.” Journal of Philosophical Logic 23 (3): 247–65. Richerson, P. J., and R. Boyd. 1998. “The Evolution of Human Ultra-Sociality.” In Indoctrinability, Ideology and Warfare: Evolutionary Perspectives, edited by I. Eibl-Eibesfeldt and F. K. Salter, 71–95. New York: Berghahn. Ridley, M. 1997. The Origins of Virtue. London: Penguin. Risser, J. 2009. “The Demand of Freedom in Kant’s Critique of Judgment.” Comparative and Continental Philosophy 1 (1): 89–104. Rorty, R., ed. 1967. The Linguistic Turn. Chicago: The University of Chicago Press. Rorty, R. 1979. Philosophy and the Mirror of Nature. Princeton: Princeton University Press. Rorty, R. 1989. Contingency, Irony, and Solidarity. Cambridge: Cambridge University Press. Rorty, R. 1991. Objectivity, Relativism and Truth (Philosophical Papers vol. I). Cambridge: Cambridge University Press. Rosenberg, A. 2013. “Disenchanted Naturalism.” In Contemporary Philosophical Naturalism and Its Implications, edited by B. Bashour and H. D. Muller, 17–36. New York: Routledge. Roughley, N., and K. Bayertz, eds. 2019. The Normative Animal?: On the Anthropological Significance of Social, Moral, and Linguistic Norms. Oxford: Oxford University Press. Rouse, J. 2007. “Social Practices and Normativity.” Philosophy of the Social Sciences 37 (1): 46–56. Rouse, J. 2015. Articulating the World. Chicago: University of Chicago Press. Schloss, J. P. 2014. “Darwinian Explanations of Morality: Accounting for the Normal but Not the Normative.” In Understanding Moral Sentiments: Darwinian Perspectives, edited by H. Putnam, S. Neiman and J. P. Schloss, 81–121. New Brunswick: Transaction. Schmidt, M. F., L. P. Butler, J. Heinz, and M. Tomasello. 2016. “Young Children See a Single Action and Infer a Social Norm: Promiscuous Normativity in 3-Year-Olds.” Psychological Science 27 (10): 1360–70. Schmidt, M. F., and H. Rakoczy. 2019. ‘On the Uniqueness of Human Normative Attitudes’. In The Normative Animal?, edited by N. Roughley and K. Bayertz, pp. 121–35, Oxford: Oxford University Press. Schmidt, M. F., H. Rakoczy, and M. Tomasello. 2011. “Young Children Attribute Normativity to Novel Actions without Pedagogy or Normative Language.” Developmental Science 14 (3): 530–9. Schroeder, M. 2008. “What Is the Frege-Geach Problem?” Philosophy Compass 3 (4): 703–20.

References 231 Searle, J. R. 2010. Making the Social World: The Structure of Human Civilization. New York: Oxford University Press. Scott-Phillips, T. 2014. Speaking our minds: Why human communication is different, and how language evolved to make it special. Basingstoke: Palgrave. Sellars, W. 1949. “Language, Rules and Behavior.” In John Dewey: Philosopher of Science and Freedom, edited by S. Hook, 289–315. New York: Dial Press. Sellars, W. 1953a. “Inference and Meaning.” Mind 62 (247): 313–38. Sellars, W. 1953b. “A Semantical Solution of the Mind-Body Problem.” Methodos 5: 45–82. Sellars, W. 1954. “Some Reflections on Language Games.” Philosophy of Science 21 (3): 204–28. Sellars, W. 1956. “The Myth of the Given: Three Lectures on Empiricism and the Philosophy of Mind.” In The Foundations of Science and the Concepts of Psychology and Psychoanalysis, edited by H. Feigl and M. Scriven, 127–96. Minneapolis: University of Minnesota Press; reprinted in Sellars: Science, Perception and Reality, 127–96. Atascadero: Ridgeview, 1963. Sellars, W. 1962a. “Philosophy and the Scientific Image of Man.” In Frontiers of Science and Philosophy, edited by R. Colodny, 35–78. Pittsburgh: University of Pittsburgh Press. Sellars, W. 1962b. “Truth and Correspondence.” The Journal of Philosophy 59 (2): 29–56. Sellars, W. 1963a. “Abstract Entities.” The Review of Metaphysics 16 (4): 627–71. Sellars, W. 1963b. Science, Perception and Reality. Atascadero: Ridgeview. Sellars, W. 1968. Science and Metaphysics: Variations on Kantian Themes. London: Routledge. Sellars, W. 1969. “Language as Thought and as Communication.” Philosophy and Phenomenological Research 29 (4): 506–27. Sellars, W. 1974. “Meaning as Functional Classification.” Synthese 27 (3–4): 417–37. Shapiro, L. 2020. “Sellars, Truth Pluralism and Truth Relativism.” In Wilfrid Sellars and Twentieth-Century Philosophy, edited by S. Brandt and A. Breunig, 174–206. New York: Routledge. Shapiro, S. 1997. Philosophy of Mathematics: Structure and Ontology. Oxford: Oxford University Press. Shapiro, S. J. 2006. “What Is the Internal Point of View.” Fordham Law Review 75: 1157–70. Sinha, C. 2014. ‘Niche Construction and Semiosis: Biocultural and Social Dynamics’. The Social Origins of Language, 31–46. Oxford: Oxford University Press. Skyrms, B. 2004. The Stag Hunt and the Evolution of Social Structure. Cambridge: Cambridge University Press. Sloman, S., and P. Fernbach. 2018. The Knowledge Illusion: Why We Never Think Alone. London: Penguin. Soja, N. N., S. Carey, and E. S. Spelke. 1991. “Ontological Categories Guide Young Children’s Inductions of Word Meaning: Object Terms and Substance Terms.” Cognition 38 (2): 179–211.

232 References Spiekermann, K. P. 2009. “Sort Out Your Neighbourhood.” Synthese 168 (2): 273–94. Sripada, C., and S. Stich. 2007. “A Framework for the Psychology of Norms.” In The Innate Mind, vol. 2: Culture and Cognition, edited by P. Carruthers, S. Laurence and S. Stich, 280–301. Oxford: Oxford University Press. Sterelny, K. 2021. The Pleistocene Social Contract: Culture and Cooperation in Human Evolution. Oxford: Oxford University Press. Stovall, P. 2021. “Rationality, Autonomy, and Obedience to Linguistic Norms.” Synthese 198 (9): 8955–80. Stovall, P. 2022. The Single-Minded Animal. New York: Routledge. Sturgeon, S. 1991. “Truth in Epistemology.” Philosophy and Phenomenological Research 51 (1): 99–108. Svoboda, V. 2018. “The Nature and Diversity of Rules.” In From Rules to Meanings: New Essays on Inferentialism, edited by O. Beran, V. Kolman, and L. Koreň, 261–77. New York: Routledge. Thompson, J. C., S. Carvalho, C. W. Marean, and Z. Alemseged. 2019. “Origins of the Human Predatory Pattern: The Transition to Large-Animal Exploitation by Early Hominins.” Current Anthropology 60 (1): 1–23. Tomasello. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press. Tomasello, M. 2014. A Natural History of Human Thinking. Cambridge, MA: Harvard University Press. Tomasello, M. 2019. Becoming Human: A Theory of Ontogeny. Cambridge, MA: Belknap Press. Tomasello, M., M. Carpenter, J. Call, T. Behne, and H Moll. 2005. “Understanding and Sharing Intentions: The Origins of Cultural Cognition.” Behavioral and Brain Sciences 28 (5): 675–90. Tuomela, R. 2013. Social Ontology: Collective Intentionality and Group Agents. Oxford: Oxford University Press. Turner, J. 1997. The Institutional Order. New York: Longman. Von Wright, G. H. 1963. Norm and Action. New York: Humanities Press. Whiten, A. 2021. “The Burgeoning Reach of Animal Culture.” Science 372 (6537): abe6514. Wikforss, Å. M. 2001. “Semantic Normativity.” Philosophical Studies 102 (2): 203–26. Williams, G. 1970. Group Selection. New York: Routledge. Wittgenstein, L. 1922. Tractatus Logico-Philosophicus. London: Routledge; English translation Tractatus Logico-Philosophicus, London: Routledge, 1961. Wittgenstein, L. 1953. Philosophische Untersuchungen. Oxford: Blackwell; English translation Philosophical Investigations, Oxford: Blackwell, 1953. Wittgenstein, L. 1969. Philosohpische Grammatik. Frankfurt: Suhrkamp; English translation Philosophical Grammar, Blackwell, Oxford, 1974. Wittgenstein, L. 2016. Lectures, Cambridge 1930–1933, From the Notes of G. E. Moore. Cambridge: Cambridge University Press. Wrangham, R. 2019. The Goodness Paradox: The Strange Relationship between Virtue and Violence in Human Evolution. New York: Pantheon.

References 233 Wray, A. 1998. “Protolanguage as a Holistic System for Social Interaction.” Language and Communication 18 (1): 47–67. Zachník, V. 2020. “When and Why Conventions Cannot Be Social Institutions.” Philosophia 48 (3): 1235–54. Zahavi,A., and A. Zahavi. 1997. Handicap Principle: A Missing Piece of Darwin’s Puzzle. Oxford: Oxford University Press. Zermelo, E. 1913. “Über eine Anwendung der Mengenlehre auf die Theorie des Schachspiels.” In Proceedings of the Fifth International Congress of Mathematicians, edited by E. W. Hobson and A. E. H. Love, 501–4. Cambridge: Cambridge University Press.

Index

a priori 53, 54, 58n2 affordance 145–6 agreement 3, 5, 13–14, 86, 107, 109, 111–12, 118, 135–6, 162–3, 202 Alemseged Z., 232 Alfano, M. 222 Alston, W. P. 206, 220 altruism 187 Andrews, K. 114, 220 Arbib, M. A. 189n1, 220 argumentation 84, 106–7, 110–2, 141, 173–7; see also game, of giving and asking for reasons Aristotle 72 arithmetic 55 assertion 14, 34, 156, 162–4, 199 atoms in the void 54 attitude: normative 2, 3, 56, 58, 62, 70, 73–8, 81–8, 90–4, 96–9, 101, 103–12, 115, 121, 123–5, 132–6, 138–41, 143, 147, 149–54, 158, 165, 167, 169, 179, 201–2, 206, 208, 211–13, 217; supernormative 133–4, 136, 138, 213 autonomy 51, 58, 193–4, 217 Axelrod, R. 187, 220 Baker, G. P. 22n1, 220 Barnes, E. C. 209n3, 220 Bashour, B. 220, 222, 227, 230 Bayertz, K. 62, 230, 231 bedrock 82, 84, 87, 105, 111 behavior: pattern-conforming 30; pattern-governed 29–31, 36; rule-enacting 30, 37n11;

rule-following 2, 7, 14, 16–17, 19, 22, 29, 30–1, 33, 37n11, 46, 91, 136, 138–9, 213; rulegoverned 1, 4, 21–2, 28–31, 52, 74, 76, 85, 138, 165, 193, 198, 215–17; rule-obeying 30, 37n11 behaviorism 24, 36n2, 49n5 Behne, T. 232 Beisecker, D. 95, 149, 220 Beran, O. 188, 220, 226, 232 Bernard, S. 222 Berniūnas, R. 186, 220 Berwick, R. C. 131, 220 Bicchieri, C. 48, 50, 220 Binmore, K. 180, 220 Blackmore, S. 130, 220 Boghossian, P. 22n1, 150, 220 bootstrapping 31, 44, 77, 115, 157, 165 Bowles, S. 118, 188, 212, 220, 221 Boyd, R. 38, 43, 122, 183, 187, 221, 227, 230 Boyle, M. 45, 221 brain 27, 42, 44, 48, 189n1 Brandom, R. vii–viii, 1, 3, 15, 26, 28, 34, 41, 51–3, 58, 73–4, 76–81, 84, 86, 88, 90, 103, 110, 121–2, 139, 147–8, 155, 160–1, 164, 166n6, 193, 198, 199–200, 217, 221, 224–5, 227–8 Brandt, M. J. 222, 231 Bratman, M. E. 75, 150, 221 Brennan, G. 61, 74, 186, 221 Butler, L. P. 231

Index  235 Call, J. 133, 232 Campbell, J. D. 135–6, 221 Campbell, R. 46, 118, 186, 226 Cantor, G. 24 Carey, S. 231 Carnap, R. 224 Carpenter, M. 232 Carruthers, P. 47, 143, 148n3, 221, 232 Carvalho, S. 232 Casler, K. 91, 221 Castro, L. 99, 221 Castro-Nogueira, L. 221 Castro-Nogueira, M. A. 221 Ceci, S. J. 130, 221 Chater, N. 37n9, 131, 222 Chemero, A. 146, 221 chess 16, 21, 35, 52, 60–1, 71, 77, 82–4, 99, 115–20, 145–6, 151, 153, 162, 164, 185–6, 195 Chicken (game) 184 Chomsky, N. 27, 131, 220–1 Chong, I. 146, 222 Christiansen, M. H. 37n9, 131, 222 Christias, D. 79, 222 Claidière, N. 137n2, 222 Clément, F. 92, 222 cognition 6, 13, 40, 43–5, 49n4, 143, 148–9, 216, 220 Cole, M. 119, 126n3, 228 commitment 55, 87, 121–2, 198, 199 conditional 14, 24, 59n6, 61, 157 conformity 98, 135–6, 192, 212, 214 conjunction 159, 161 constant, logical 171–2; see also conjunction; disjunction; implication; negation convention 62, 98, 135 Cook, R. T. 177n5, 222 cookie-cutter 203–4 cooperation 4, 6, 54, 68, 100, 122–3, 178–84, 186–90, 192 coordination 47–8, 56, 94, 97–9, 101, 123, 175, 211, 217 correctness 3, 5, 14, 64, 68–72, 76–7, 81–4, 87, 89–100, 103–13, 158, 163–4, 168, 191–92, 199, 201–2, 208, 212; Protagorean 103, 105–8, 110, 112–13, 198, 201–2, 212 Cosmides, L. 102n3, 222

Cresswell, M. J. vii, 222 criterion 3, 107, 108, 112, 162–3, 204 Csibra, G. 100, 223 culture 38, 43, 47, 53, 68, 125, 129, 130–1, 136, 146, 178, 196n1, 223–4 Curry, O. S. 184, 189n4, 222 Davidson, D. 8, 52, 59n4, 204–5, 209n2, 222, 228 Davis, T. 74, 225 Dawkins, R. 130, 207, 222, 226 deduction, natural 83 Dennett, D. 54, 63, 135, 151, 190, 222 deVries, W. A. 34, 36n1, 55, 223 Dewey, J. 52–3, 58n1, 223, 225, 231 disjunction 83–4, 159 disposition 85, 140 Dor, D. 72n1, 131, 189n1, 223, 231 Dranseika, V. 220 Drobňák, M. 166n7, 223 Dummett, M. 162–3, 222–3 Dutilh Novaes, C. 175–6, 223 education 32, 100, 191, 196n1 Einstein, A. 24, 210 emotion 46 enculturation 142, 191 entitlement 120, 121, 162 environment 17, 23–4, 40–1, 50, 58, 66, 81, 92–3, 95, 97, 101, 118, 123, 125, 129–30, 132, 134–5, 145–6, 157, 190–2, 196, 203, 207–8, 211–12, 217–18 equilibrium 98, 182–4 Eriksson, L. 221 essence 19, 21, 140–1, 147 ethics 6, 45, 188; see also morality Evans, J. S. B. 44, 223 evolution 3–4, 13, 29, 31, 44, 48, 53, 66–9, 79–80, 82, 96, 114, 122–3, 125n2, 127, 129–32, 134, 136, 147, 151, 154, 175, 178–80, 183, 185, 187–91, 194–96, 208, 212, 214–15, 218, 223 expectation 86, 93, 96 faculty, language 6, 27, 37 Fairey, P. J. 135–6, 221

236 Index Fehr, E. 187, 223 Feldman, M. W. 228 Fernald, A. 37n9, 223 Fernbach, P. 49n2, 231 Fitzpatrick, S. 114, 223 Fodor, J. A. 40, 44–5, 223 football 3, 47, 52, 100, 115, 118–20, 193 freedom 4, 24, 26, 65, 116, 135, 190, 192–6 Frege, G. 160, 231 Frege-Geach problem 88, 89, 231 Frost-Arnold, G. 209n3, 223 F-rule 62, 63, 70 future 9, 24, 26, 36, 56, 58, 79, 86, 94, 96, 119, 122–5, 130, 145, 154, 158, 207 Gächter, S. 187, 223 game: language 1, 4, 7, 8, 19–22, 34–5, 41, 52, 81, 116, 139, 142, 145–6, 151–3, 160, 162–3, 165, 171, 186, 216; of giving and asking for reasons 34, 42, 106, 108, 110, 152, 160, 165, 176, 199, 216; of indication 152, 167, 213 game theory 122, 180, 188 Gauker, C. 45, 53, 59n6, 79, 223, 229 gene 131 Gergely, G. 100, 223 Gibson, J. J. 145–6, 222–4 Gil-White, F. J. 137n1, 225 Ginsborg, H. 74, 140–1, 223 Gintis, H. 131, 188, 212, 220–1, 224 Giromini, J. 90n2, 224 Glanzberg, M. 171, 224 Glock, H.-J. 190, 224 Glüer, K. 71, 224 Goldfarb, W. 22n1, 224 Goldhaber, D. 130, 224 Goodin, R. E. 221 Goodman, N. 9, 10, 206, 209n2, 224 grammar, universal 27, 60 Greene, J. 46, 224 Greene, K. 221 Grice, H. P. 52, 149 grue 9, 10, 87 habit 28, 214 Hacker, P. M. S. 22, 220

Hahn, H. 58n2, 224, 228 Haidt, J. 46, 176, 224 Hamilton, W. D. 187, 224 Hanfling, O. 52, 224 harmony, democratic 13, 17, 87, 103, 105, 111 Hart, H. L. A. 48, 73–4, 90n1, 214, 224 Hattiangadi, A. 76, 80, 86, 224 Haugeland, J. 22n1, 74, 224 Hawks and Doves (game) 183 Heft, H. 146, 224 Hegel, G. W. F. 144, 148n4, 221, 229 Heidegger, M. 146, 189n2, 224, 225 Heinz, J. 231 Henrich, J. 94, 137n1, 221, 225 Hildebrand, D. 52, 225 Hjortland, O. T. 177n2, 224, 225 Hlobil, U. viii, 177n2, 225 Hofstadter, D. 11, 12, 225 holism 115 Hudson, W. D. 88, 225 Hume’s thesis 88 Hurford, J. R. 42, 225 Hurley, S. 148n5, 225 Husserl, E. 207, 225 image: manifest 24–6, 36n4, 51, 55–7, 121; scientific 24–6, 36n4, 51, 55–7, 121, 207; see also vision, stereoscopic imitation 48, 100, 136, 192, 226 implication 84, 159, 170, 172 inference 34–5, 43, 48, 88, 152, 154, 158–60, 164, 167–70; material 168, 170–1 inferentialism vii–viii, 1, 3, 73, 90, 150, 164, 166n7, 198 infinity 10, 77, 85, 86 information 10, 42–3, 129, 135, 146, 152, 174–5, 187 institution 99, 101, 118 intention 29, 96, 118, 182–3, 210 intentionality: collective 75, 142; shared 47, 150, 205 interaction viii, 19, 30, 43–4, 53, 58, 66, 73, 86, 94, 108, 132, 150, 207 interpretation 8, 15, 17, 22, 38 intersubjectivity 104–5, 198, 201–3, 208

Index  237 Jablonka, E. 72n1, 129, 131, 223, 225 Joyce, R. 184, 225 justification 9, 34, 162–4 Kahneman, D. 175, 225 Kalish, C. 93, 225 Kaluziński, B. viii Kant, I. 24, 28, 36n3, 160, 186, 193–4, 224–5, 230 Karlsson, M. 226 Kaufmann, L. 222 Kelly, D. 74, 225 Kendal, J. R. 125n2, 225 Kenward, B. 101n2, 226 Kern, A. 100, 226 Kirmayer, L. J. 230 Kitcher, P. 184, 188, 226 Knight, C. 147n2, 189n1, 214, 223, 226 Knobe, J. 101n1, 229 knowledge 34, 42–3, 49n2, 53, 58n2, 82, 117, 146, 148n3, 150, 174, 175, 221 Koons, J. R. 75, 226 Koreň, L. viii, 53, 75, 161, 165n1, 220, 226, 229 Krebs, J. R. 207, 226 Krikorian, Y. 53, 226 Kripke, S. A. 6, 8–10, 13–14, 17, 19, 140, 224, 226 Kumar, V. 46, 118, 186, 226 Kusch, M. 17–19, 22n1, 72n2, 226 Laland, K. 32, 68, 94, 100, 125n2, 153, 196, 226, 228 Lamb, M. J. 129, 225 Lance, M. N. 52, 226 Lau, H. 143, 226 Lewens, T. 132, 226 Lewis, D. 98, 122, 189, 226 Lewis, J. 189n1, 223, 226 Leyton-Brown, K. 180, 227 life 5, 23, 43, 64, 66, 70, 81, 94, 99–100, 109, 117–19, 121, 136, 143–4, 148, 154, 173, 185–7, 190, 192–4, 196, 202, 208; form of 2–4, 23, 36, 46, 48–9, 58, 69, 81, 117, 119, 145, 192–4, 196, 206, 208, 214, 218 logic viii, 4, 6, 34, 58n2, 83–4, 161, 167, 171–4, 176–7, 217;

dialogical 174; intuitionistic 171; relevant 177n1; see also deduction, natural; space, logical Lorenz, K. 174, 227 Lorenzen, P. 174, 227 L-rule 62–3, 70, 73 Lycan, W. G. 143, 227 Mandelkern, M. 59n6, 227 Marean, C. W. 232 Mares, E. D. 177n1, 227 Marchman, V. A. 37n9, 223 Martin, B. 177n2, 227 Maryanski, A. 141, 227 Maynard Smith, J. 187, 227 McDowell, J. 22n1, 227 McElreath, R. 122, 183, 227 McGee, V. 177n3, 227 McKenzie Alexander, J. 188, 227 Mead, G. H. 43, 227 meme 130 Mercier, H. 141, 150, 165n1, 175–6, 197n2, 227 metaphysics 58n2, 218 metarule 77 Millikan, R. G. 36n5, 67–8, 72n2, 118, 227, 228 mind 6, 14, 22–4, 26, 28, 40, 46–7, 53, 58, 65, 74, 91, 101, 111, 118, 134, 141, 143, 148–9, 173, 180, 188, 200, 203, 206–7 mind-reading 47 Misak, C. J. 162, 228 modal 59n6 modus ponens 84, 169, 172 Moll, H. 100, 226, 232 morality 46, 62, 143, 186, 189n3; see also etrhics Muller, H. D. 220, 222, 227, 230 Myers, R. H. 204, 228 myth vii, 210 naturalism viii, 53–5, 58 naturalization viii, 77, 84, 195 nature (world) 2, 4, 24, 28, 38, 50, 53, 63–4, 135, 145, 177, 191–4 nature (essence) 1–4, 6–7, 14, 16, 24–5, 27, 31, 38, 40–1, 46, 54, 57, 62, 73, 92, 96, 129–31,

238 Index 134–6, 145, 147n1, 151, 164, 171, 175, 178–80, 186, 207, 217 negation 83–4, 159, 161 Neurath, O. 54, 224 niche 43–4, 117, 119, 123, 126n3, 130–1, 147, 157, 165, 188, 192, 196, 217; cognitive 43–4; normative 123, 126n3, 157, 165, 196 niche construction 123, 130, 131, 157, 165 Nichols, S. 46, 228 norm: formal 61; linguistic 62, 125; technical 61, 97 normative (idiom) 56–7, 78–81, 88–9, 124–5 normativity 1–3, 13, 22–4, 58, 63–65, 67, 69, 71–2, 74–5, 84, 88, 95–6, 103–107, 111, 123, 125, 198, 200, 208, 220; primitive 140; promiscuous 93; see also attitude, normative; psychology, norm objectivity 104–5, 113n2, 198, 200–3, 208 O’Brien 125n2, 226 Odling-Smee, F. J. 123, 228 Okrent, M. 66, 68, 72n1, 114, 153, 228 ontogenesis 31, 101n ontology 58n2 O’Shea, J. 36n1, 59n4, 228 ought-to-be 32–33 ought-to-do 32–33 Over, D. E. 44, 94, 223, 226 Packer, M. J. 119, 126n3, 228 Partee, B. 37n8, 228 Peirce, C. S. 144, 162, 228 Pelican, C. 222 person 10, 26, 29, 33, 36n4, 53, 71, 75–7, 98, 100–1, 103, 109, 118, 121, 140, 143–4, 186, 193, 200–1, 204 Persson, J. 226 Phillips, J. 101n1, 203, 229 phylogenesis 31 physics 43, 47–8, 54–5, 219n1 Pinkard, T. 148n4, 194, 197n3, 229 Pinker, S. 42–3, 131, 229

Planer, R. 47, 153, 203, 229 Popper, K. R. 216, 229 Poundstone, W. 122, 229 practice 4, 15–18, 34, 60–1, 63, 121, 140–1, 144–5, 147, 151, 160, 165, 214 Premack, D. 45, 229 prescription 65, 195, 216 Price, H. 113n2, 229 Priest, G. 177n5, 229 primate 146 Proctor, R. W. 146, 222 proposition 14, 19, 159 proto-language 152–4, 161, 167, 171 psychology 47, 55, 74, 118, 143, 145–6, 217; belief-desire 41–42; norm 94, 142 Punčochář, V. 59n6, 229 punishment 74; altruistic 187 Putnam, H. 203, 206, 229, 231 quaddition 8–10, 84–6 quasiinference 152, 154 Quine, W. V. O. vii–viii, 52–4, 58n1, 153, 166n4, 203–5, 209n2, 229 Rahman, S. 174, 229 rail 86 Rakoczy, H. 52, 73, 91, 93, 229, 230–1 Ramstead, M. J. 146, 230 Rawls, J. 60–61, 140, 230 Read, S. 167, 230 reason 4, 6, 15, 22, 34–5, 38–9, 41–5, 47–9, 51–2, 55, 58n2, 65, 67, 72, 76, 80, 95, 97–101, 114, 117, 125, 129, 135, 150–1, 153, 155–6, 171, 175–6, 179, 184, 194–5, 202, 204, 210–12, 215–16; see also space, of reasons reasoning 35, 41–5, 49n4, 84, 102n3, 140–1, 150, 157, 159, 172–6, 195; dual-process theory of 44 reducibility, causal vs. logical 79 regularity 12, 29, 30, 76, 84–6, 92, 98, 104–5, 168 regulation 35, 61, 143, 194 regulism 15, 28, 29, 73 representation 20, 28, 31, 101n1, 206 responsibility 121–2, 125, 156

Index  239 reward 74 Ridley, M. 184, 230 Richerson, P. J. 38, 221, 230 Risjord, M. viii, 228 Risser, J. 36n3, 230 role 1, 7, 14–15, 19–20, 22–3, 36, 40–1, 43, 45–6, 53, 56–7, 61–2, 65, 71, 74, 89–93, 100–1, 107, 111–3, 118–19, 121, 135–6, 150, 158, 160–1, 169, 175, 213, 216 Rorty, R. 52, 55, 58n3, 163, 209n4, 218, 225, 228, 230 Rosenberg, A. 54, 230 Rosenthal, D. 143, 226 Roughley, N. 62, 230,y 231 Rouse, J. 53, 140, 147, 157, 230 Rückert, H. 174, 229 rule: constitutive vs. regulative 60–1; implicit 2, 49, 60, 73–5, 78, 96, 115, 172, 211, 216; instrumental vs. noninstrumental 46, 61, 63–4, 66, 68, 97; integrative 114–15, 118, 121, 125, 185; moral 4, 46, 79, 184, 186, 188–9, 192; restrictive vs. prescriptive 71; social 52, 62–3, 72, 77, 97, 103, 118, 184, 186, 214; see also behavior, rule-enacting; behavior, rule-following; behavior, rule-governed; behavior, rule-obeying rule following 1, 2, 16–19, 29, 31, 46–8, 57, 65, 132, 179, 208 sanction 64, 74, 76, 86, 87, 90 Schloss, J. P. 79, 231 Schmid, H. B. viii, 226 Schmidt, M. F. 52, 73, 92–3, 101n2, 229, 231 Schroeder, M. 89, 231 Schwemmer, O. 174, 227 science viii, 24, 38, 50–5, 57, 59n5, 82, 91, 207 Scientia mensura 25, 51 scorekeeping 122, 198, 200 Searle, J. R. 60, 231 selection: group 187, 188, 212; kin 187; natural 29, 66, 68, 127, 129–30, 134, 151, 207

self-consciousness 51, 139, 143 self-governance 142, 143 self-reflection 144, 218 Sellars, W. 1–2, 23–6, 29, 31–8, 41, 43, 48–9, 51–3, 55, 59n5, 63, 73–5, 79, 87, 90n3, 116–17, 120–1, 139, 151, 153, 159, 162–5, 167, 179, 197n3, 207, 210, 216–18, 223, 226–9, 231 semantics vii, 1, 4, 9, 40, 153 Shapiro, L. 165, 231 Shapiro, S. 120, 231 Shapiro, S. J. 90n1, 231 Shoham, Y. 180, 227 Silius, V. 220 Sinha, C. 39, 231 Skyrms, B. 182, 231 Sloman, S. 49n2, 231 Soja, N. N. 204, 231 Southwood, N. 221 space: logical 159, 161; of meaningfulness 4, 41, 90n3, 116, 142, 145–6, 149, 160, 165, 177, 185, 196, 207, 214–15, 218; of reasons 151, 159–61, 216 Spelke, E. S. 231 Sperber, D. 141, 150, 165n1, 175–6, 197n2, 227 Spiekermann, K. P. 187, 232 Sripada, C. 142, 232 Stag Hunt (game) 182, 186, 231 Sterelny, K. 47, 153, 178, 203, 229, 232 Stich, S. 142, 232 Stovall, P. viii, 30, 37n7, 53, 226, 232 Sturgeon, S. 163, 232 Svoboda, V. viii, 62, 70, 170, 177n2, 229, 232 syllogism, disjunctive 35, 83, 170 symbol 15, 23, 39, 144 task, false-belief 92 teaching 31, 139 technology 38, 178, 188–90 telephone 98 Terziyan, T. 221 Thompson, J. C. 178, 23 tit-for-tat 187 Tomasello, M. 42, 52–3, 100, 141–2, 149, 165n1, 205, 211, 230–2

240 Index Tooby, J. 102n3, 222 Toro, M. A. 221 Townsend, L. viii, 226 traffic 14–15, 26, 28, 60–1, 76, 97, 134 truth 5, 66, 81, 87, 89, 105, 108–9, 113n2, 125, 150, 159, 162–5, 173–6, 198, 203–4, 208 Tuomela, R. 75, 232 turn, linguistic 58n3 Turner, J. H. 118, 141, 227, 232 Turvey, M. 146, 221 ultrasociality 38 Veissière, S. P. 230 Verheggen, C. 204, 228 Vienna Circle 58n2, 224 vision, stereoscopic 55–6 vocabulary, logical 156, 159, 161, 167, 169–71, 174, 176 von Wright, G. H. 61, 232

Warneken, F. 230 warning 97, 99, 105–6, 165n3, 212 Weber, M. 25 we-intention 75, 96, 125 Whiten, A. 134, 137n2, 222, 232 Wikforss, Å. 71, 163, 224, 232 Williams, G. 187, 232 Williams, W. M. 130, 221 Wittgenstein, L. vii, 2, 6, 7–10, 14–23, 26, 31, 34–6, 38, 41, 49, 50, 52–3, 56, 58n1, 63, 65, 75–6, 85, 101, 144, 160, 189n3, 204, 224, 226–7, 232 world: possible vii, 24, 89; real 26, 106, 118–200 Wrangham, R. 144, 232 Wray, A. 153, 233 Zachník, V. 125n1, 233 Zermelo, E. 195, 233