Explaining the Normative E-Book

Stephen P. Turner

• Herausgeber: John Wiley & Sons
• Kategorie: Geisteswissenschaft
• Sprache: Englisch

Seitenzahl: 511

Veröffentlichungsjahr: 2013

Beschreibung

Normativity is what gives reasons their force, makes words meaningful, and makes rules and laws binding. It is present whenever we use such terms as ‘correct,' ‘ought,' ‘must,' and the language of obligation, responsibility, and logical compulsion. Yet normativists, the philosophers committed to this idea, admit that the idea of a non-causal normative realm and a body of normative objects is spooky. Explaining the Normative is the first systematic, historically grounded critique of normativism. It identifies the standard normativist pattern of argument, and shows how this pattern depends on circularities, assumptions about the unique correctness of preferred descriptions, problematic transcendental arguments, and regress arguments that end in mysteries.

The book considers in detail a paradigm case: legal normativity as constructed by Hans Kelsen. This case exemplifies the problems with normativist arguments. But it also shows how normativism was constructed as an alternative to ordinary social science explanation. The normativist argument is that social science explanations themselves are forced to rely on normative conceptsÑminimally, on normative rationality and on a normative view of ‘concepts' themselves.

Empathic understanding of the reasoning and meanings of others, however, can solve the regress problems about meaning and rationality that are central to the appeal of normativism. This account has no need for a parallel normative world, and has a surprising and revealing lineage in the history of philosophy, as well as a basis in neuroscience.

Leseprobe

EXPLAINING THE NORMATIVE

For Chairman Meow, and in memory of Socrates, Claiborne, and Bailey, who taught me about empathy.

EXPLAINING THE NORMATIVE

STEPHEN P. TURNER

polity

Copyright © Stephen Turner 2010

The right of Stephen Turner to be identified as Author of this Work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988.

First published in 2010 by Polity Press

Polity Press

65 Bridge Street

Cambridge CB2 1UR, UK

Polity Press

350 Main Street

Malden, MA 02148, USA

All rights reserved. Except for the quotation of short passages for the purpose of criticism and review, no part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher.

ISBN-13: 978-0-7456-5453-9

A catalogue record for this book is available from the British Library.

Typeset in 10.5 on 12 Sabon

by Servis Filmsetting Ltd, Stockport, Cheshire

Printed and bound in Great Britain by MPG Books Group Limited, Bodmin, Cornwall

The publisher has used its best endeavours to ensure that the URLs for external websites referred to in this book are correct and active at the time of going to press. However, the publisher has no responsibility for the websites and can make no guarantee that a site will remain live or that the content is or will remain appropriate.

Every effort has been made to trace all copyright holders, but if any have been inadvertently overlooked the publisher will be pleased to include any necessary credits in any subsequent reprint or edition.

For further information on Polity, visit our website: www.politybooks.com

CONTENTS

Introduction

Chapter 1

What Is the Problem of Normativity?

Chapter 2

The Conflict with Science and Social Science

Chapter 3

A Paradigm Case: The Normativity of the Law

Chapter 4

Lustral Rites and Systems of Concepts

Chapter 5

Communities, Collective Intentions, and Group Reactions

Chapter 6

Rationality or Intelligibility

Epilogue

References

Index

INTRODUCTION

My book The Social Theory of Practices was published in 1994, as was Robert Brandom’s Making It Explicit: the first was a small book with intent to demolish, the second a large and in every sense constructive book. The Social Theory of Practices was an argument against the appeal to what it called collective objects, with a modest constructive message about emulation as an alternative explanation. Making It Explicit was an account of linguistic normativity, in which the “it” being made explicit was a paradigm case of a collective object, but it bypassed tough problems about how the account fit into the world. I thought the arguments of The Social Theory of Practices applied to Brandom and, mutatis mutandis, to much of the flood of “normativity” thinking of the time. Joe Rouse wrote me a letter that claimed that they did not, because the structures Brandom and he were talking about were “normative.” Saying this amounted to giving these arguments a free pass from the questions raised in The Social Theory of Practices – questions about how collective objects could possibly work in the way they were supposed to, given the causal features of the world, our cognitive processes, ordinary features of social interaction, and the kind of feedback we get in the course of learning how to get around the world.

I found Rouse’s claim deeply puzzling, not least because so much of the argumentation in the texts of the normativists appealed to notions that were taken over directly – or slightly modified – from the social sciences and in particular from classical social theory. Wilfrid Sellars, for example, used Durkheimian notions. And one could not read philosophers like Philip Pettit without hearing Durkheim, for example in Pettit’s appeal to common knowledge. Saul Kripke’s account of rule-following in Wittgenstein wound up with puzzles about a problematic concept from social theory, namely community. And the borrowing from classical social science was explicit in such works as Margaret Gilbert’s On Social Facts (1989), with its discussion of Georg Simmel.

The puzzle extended to my own earlier writings. There was a strong relationship between the writings of normativism and the claims that I had made in my (Wittgensteinian) dissertation, published as Sociological Explanation as Translation (1980), in my (more Quinean) “Translating Ritual Beliefs” (1979), and in my critique of the Strong Programme (1981). Indeed, there were two columns of text (last column of 190 and the first of 191) in Brandom’s “Freedom and Constraint by Norms” (1979) that could have served as a summary of my own arguments in these texts – though in the next pages Brandom took them in the direction of Making It Explicit, where I would not have taken them, for reasons having to do with the social theory implicit in Brandom’s argument, particularly its own appeal to notions of community. The links between my earlier writings on translation and normativism and on the Strong Programme in the Sociology of Science (e.g., 1981) were close enough for me to be identified as a normativist myself (e.g., by Henderson 2002), though of a more Davidsonian kind.

Clearly there were issues to be worked out, both as a matter of philosophy and as a matter of social theory, and for myself as well as in the larger community of readers of these literatures. During the decade after the publication of The Social Theory of Practices I wrote a number of articles and review essays on various aspects of the problem of normativity (mostly collected in Brains/Practices/Relativism: Social Theory after Cognitive Science, 2002, 74–107), watched Brandom present his views at various places, and presented my views on these issues in a rough form at an NEH Summer Institute with Brandom present. Among the elements of that discussion was the case of legal normativity. Brandom himself suggested to me at the time that this was a line of argument worth pursuing, and indeed it was. I was also paired with Joe Rouse at a number of philosophical events, reviewed his books (1989, 2005a), and commented on his notion that the concept of practices was a normative one (2007a). This book is the product of these exchanges, though the form of the text itself is an attempt to locate these issues in a much larger and different historical picture: not as a vindication of idealism and Kant, as it is usually presented, but as a story about the attempt to recapture ground that had been lost in the sometimes friendly, but often nasty, divorce between philosophy and social science in the late nineteenth century. Hans Kelsen was the poster child of this divorce, so he plays a central role in this text.

This material has been presented in a variety of places, as part of talks. I am grateful for the opportunities to discuss the arguments that attendees presented, the comments, the objections, and particularly the puzzled looks. I am in particular grateful for the response of students in the University of South Florida Philosophy department, who have put up with this little obsession for more years than I care to remember. I am also especially grateful for such old friends as Paul Roth, who have responded to these concerns in a philosophical language I understand, and George Mazur, who was kind enough to argue with me at length about Hans Kelsen. Eileen Kahl provided her usual competent support in putting the manuscript together. My family, decade upon decade, has put up with stacks of books and slips of paper in the wrong places. For their forbearance I am grateful. Thanks also to Gerhard Preyer for permission to use much of “What Do We Mean by ‘We’?” (Protosociology, 2003) in Chapter 5.

Pass-a-Grille January 31, 2009

1

WHAT IS THE PROBLEM OF NORMATIVITY?

Normativity pervades our lives. We do not merely have beliefs: we claim that we and others ought to hold certain beliefs. We do not merely have desires: we claim that we and others not only ought to act on some of them, but not others. We assume that what somebody believes or does may be judged reasonable or unreasonable, right or wrong, good or bad, that it is answerable to standards or norms . . . We find ourselves at sea because there is a huge disagreement about the source and the authority of norms on which we all constantly rely.

Onora O’Neill (1996, xi)

Normative facts (e.g. about who is committed to what) are just one kind of fact among others. Normative facts are facts statable only using normative vocabulary (compare: physical facts).

Robert Brandom (1997, 197n6)

No one bears away from this Sacrament more than is gathered with the vessel of faith.

John Calvin ([1536] 1960, IV.xvii.33)

Normativity is everywhere. The sign of this, as Onora O’Neill’s reasoning suggests, is that normative terms are ubiquitous, and we constantly and necessarily rely on them. Correct and incorrect, right and wrong, good and bad, rational and not rational, valid and invalid – the list is long. The normative is a special realm of fact that validates, justifies, makes possible, and regulates normative talk, as well as rules, meanings, the symbolic and reasoning. These facts are special in that they are empirically inaccessible and not part of the ordinary stream of explanation. Yet they are necessary in the sense that if they did not exist, ordinary normative talk, including such things as claims about what a word means or what the law is, would be unjustified, nonsensical, false, or illusory. To say that something has meaning requires that there be such a thing as a meaning. To say something is a real law is to say that there is something that validates the law as real.

But beyond this way of thinking about the normative as a realm of fact lies a vast muddle. What is the character of this normativity that is everywhere and signaled by the presence of these terms? Is it a nonnatural, noncausal property of things? A force that attaches to things, such as claims, that gives them some obligatory power? Are norms part of the furniture of the world, a part which is merely odd in some respects, or is it an aspect of things that are otherwise normal? Or is normativity something else entirely? Is it best understood as a kind of shadow system of rules, proprieties, scoring systems, presuppositions, and so forth that stands tacitly behind our normative usages and regulates and justifies them in a way that is a hidden analogue to the way that explicit rules, scoring systems, and the like regulate and justify? And if we are bound by these things, how are we bound? Do we bind ourselves under norms by our own commitments, or in some other way? A danger with these questions, pointed out by John Mackie, is that by answering them in the wrong way we could make normativity into something so queer that it could not be accommodated to the rest of our ideas about the natural, explainable world (Mackie 1977, 38–42, 48–49). But understanding normative language in terms of normativity forces us to ask such questions.

The usual way of getting at metaphysical questions of this kind is to ask about explanation. What does the stuff explain? What explains it? How does it explain? O’Neill alludes to the problem of explanation in her comments about the “huge disagreement” over sources of normativity. Are the sources natural, and if so does that mean that there is no special force answering to the name of “normativity”? Or are the sources normative? Does using normative terms commit us to some sort of complex metaphysics that goes beyond the natural and causal? O’Neill’s language points directly to this problem: “Sources” is an explanatory term, tracing something to its source is a way of explaining it. But what sort of explanation is this going to turn out to be? If the explanation is rooted in the causal world, how does it explain something that is part of the noncausal world? Is there is some kind of transformation from causal to normative? If there is some other kind of source of normativity – for example, moral intuitions – we are faced with questions about where the intuitions came from, and, inescapably, with questions about where and how their causes, if they are caused, manage to produce something “normative,” and, if they are not caused, how they came about.

These are problems about the explanation of the normative. But there are questions on the other end, about the norms or normative facts that normativity consists in, and what explanatory work these facts do. What do these facts, if they are facts, explain, if anything? And how do the kinds of explanations that involve norms relate to other explanations? Should anyone care about normativity who is not answering normative questions? Are normative considerations merely part of some sort of circular, closed system of thinking? Are they inescapable – part of the business of explanation itself? And does this inescapability imply that everything, or everything expressible, is part of a closed system – what Wilfrid Sellars calls the space of reasons – that is in the end normative? Or is their inescapability a feature of our subjective experience that is not itself factual – a feature of the task of explaining but not of the world that is explained?

Issues related to these questions dominate the present philosophical landscape, where they are often cast in David and Goliath terms. The normative is the small boy with the stone against the massive forces in modern philosophy of naturalism, materialism, physicalism, and causalism; forces that draw their power from the success of science. This imagery gives epochal significance to what might otherwise seem to be a parochial dispute among professional philosophers. Nor is the imagery misplaced. The long history of secularization, bound up with the history of modern philosophy, has largely consisted of desupernaturalizing explanation. Claims about “normativity” seem to imply that this project can never be completed, that the project of desupernaturalization will always be defeated by the small stone of normativity.

Robert Brandom captures this sense of the significance of the issue in a reference to Max Weber’s account of the process of rationalization in the West and the consequent disenchantment of the world. Weber located the precipitating cause of modern rationalization in the theological rationalization of the Protestant reformation. It is no accident that at the core of this rationalization was the rejection of the mystery of the real presence of Christ in the Host – the subject of the third epigraph to this chapter. But Weber traces this process back to the origins of the West, and to such rationalizing accidents of history as the alteration in oracular practice in ancient Judaism which forced the Priest to construct yes-no questions, thus sparing the West the less rational superstitious methods of discerning the truth of the sort that prevailed elsewhere. Brandom gives a different reading of the disenchantment of the world. For Brandom, disenchantment went too far, to the point where

the meanings and values that had previously been discerned in things are stripped off along with the supernatural and are understood as projections of human interests, concerns, and activities onto an essentially indifferent and insignificant matter. (Brandom 1994, 48)

Brandom proposes to re-enchant the world by reinstating the belief in normative powers, which is to say, powers in some sense outside of and distinct from the forces known to science.

Normativity against social science

The term “normativity” is a relatively novel addition to philosophy. But the question of the relation between the normative and nonnormative, with which this book is concerned, has a distinctive history of its own. Brandom points to this history by mentioning Samuel von Pufendorf, who defended a system of Natural Law against Thomas Hobbes’s disenchanting mechanistic account of sovereignty (Brandom 1994, 46–50; Pufendorf [1688] 1964). The philosophical discussion of law over the course of the subsequent centuries produced an extraordinary collection of variations on the conflict over the source and nature of the binding character of law. Much of this discussion involved regress problems. Pufendorf and his school ended this regress in norms. They argued that the normative fact of Natural Law prevailed before the State existed, that it supplies the guiding principles of all legislation, and that it binds the sovereign himself (Gierke [1880] 1939, 319). Hobbes thought the force of law followed from the nonnormative fact of sovereign power, which necessarily preceded law. The discussion was inseparable from the larger project of secularization. The theme of the critics of natural law was to rid discussion of the law of its superstitious, theological, and mystical elements, as physics had been cleansed.

The defenders of the older traditions pointed to the binding character of the law, and argued that for there to be normatively binding law, there had to be an ultimate source that was itself binding. Kant, who is usually given as a source of the idea of normativity, contributed to this discussion by arguing

that the absolute law of reason is binding on all rational beings by virtue of its rationality, that the state freely legislates the law of reason into positive law, and that it does so because the state’s own nature requires it to act rationally. (Gierke [1880] 1939, 321)

This reasoning is characteristic of normativism. Pufendorf, Kant, and the Naturrecht tradition that lies behind this play a double game. They explain the realities (of the state, in this case) in terms of a deeper reality hidden within (for example, in the form of an intrinsic nature). This hidden reality is systematically distinct from and different than the empirical reality – in this case, the reality of actual law and the actions done by actual states. The intrinsic features provide normative standards, which are systematically discrepant from what actually occurs. But this double game is also what gives the disenchanters their opening. They can deny that there is anything intrinsically there, or necessarily there. This is the core of the issue of normativity: normativity is a name for the non-natural, non-empirical stuff that is claimed to be necessarily, intrinsically there, and to in some sense account for the actual.

The social sciences play a special role in the twentieth-century part of this larger story of disenchantment, and place this particular philosophical literature in an unusual relation to “science.” The empirical material that is associated with the current problem of normativity, that is, the things that normative notions explain, was already addressed by the social sciences by the time the term “normativity” was introduced. There were social-science explanations of such phenomena as the state and law, and these explanations were designed to replace both folk and Naturrecht conceptions of the state. So present normativism, as I will call it, is a more or less self-conscious attempt to take back ground lost to social-science explanation.

One of the central points in the normativity literature is that normativism competes with social science explicitly. The empirical fact of normativity is called “sociological” to distinguish it from real normativity (Brown 2001, 160–61), and it is expressly argued that sociological normativity, which is often dismissed as “mere” sociological normativity, is not the subject of the normativists’ discussions, which in turn means that facts of sociology cannot refute claims about normativity. This is a way of conceding the explanatory point to the social scientists and taking it back at the same time. It sets up the problem of normativity in a certain way: the empirical sociological phenomenon of normativity is not denied, but it is not enough to explain what needs to be explained. The thing which must be explained varies, but the usual formulations involve obligation. The mere sociological fact that people believe a given practice to be obligatory does not make it so. It is the extra thing that does make it so that needs to be explained.

This line of reasoning preserves the double game. And it opens the door to a particular claim of dependence: that the normative concept of, for example, the law, is indispensable to empirical explanation of legal phenomena, that to speak nonnormatively about such a thing as the law is to “change the subject,” that “sociological” accounts of these things improperly presuppose, or, in present terms, “help themselves to,” normative language, and thus wrongly come to their naturalistic conclusions, and so forth. The same kinds of dependence or regress arguments appear, and are stronger, when it is a matter of the normative character of reason itself. How can one reason, even “sociologically,” about anything at all without reasoning according to reason in the normative sense of correct reason?

These are compelling arguments. But O’Neill’s mention of “huge disagreement about the source and authority of norms” should serve as a warning that all is not right with this reasoning.

These claims have regularly come to grief when the time comes to explain what normativity means, where it comes from, and why it is that what is normative in one social setting or intellectual context is not normative in another. The last issue, the problem of local normativity, is especially important. Different groups of people, different professions, users of different languages have different norms. The fact that what counts as correct, true, or valid differs from setting to setting compels us to recognize that this variation is, in some sense of this problematic term, “social.”

Unsurprisingly, answers to questions about the source and authority of norms traffic in sociological concepts. Community plays a central role in Saul Kripke’s discussion of rule-following in Wittgenstein (1982, 56, 79–81, 89–109 passim) and in subsequent discussions of the problem. Collective intentionality is central to Sellars, who reinterprets Kant’s notion of rational beings in terms of a rational collectivity of rational beings whose collective intentions are the source of the binding character of reason (DeVries 2005, 266–67; Sellars 1968, VII § 144: 225, 226; 1967, 411).1 Do these usages mean that the end of the regress is in sociological facts? Kripke’s seems to – it ends, or appears to end, in actual communities. Sellars’s community of rational beings, however, is entirely virtual. The normativist can deny that there is an issue here and say that the relevant sociological concepts, such as community, are themselves dual concepts, with both a normative and a sociological sense. This allows them to say that the sense of community needed to account for the normativity of rule-following is the normative sense rather than the sociological. Or they can argue that such concepts as “practice” are normative rather than causal (Rouse 2002, 19–22).

Each of these arguments has problems. But the dual-character argument opens a door that normativists do not want to open; a door to a dualism in which the normative has no explanatory or metaphysical significance at all. The issue appears in a particularly sharp form in connection with the program known as SSK, the sociology of scientific knowledge. SSK operates with what it takes to be a naturalistic notion of reason, in which reasons figure as causes of the beliefs of scientists. But SSK wishes to remain neutral with respect to questions about validity, and objects, in the words of David Bloor, to “the intrusion of a non-naturalistic notion of reason into the causal story” ([1976] 1991, 177, cited in Friedman 1998, 245). This insistence, Michael Friedman suggests,

rests on a misunderstanding. All that is necessary to stop such an “intrusion” of reason is mere abstinence from normative or prescriptive considerations. We can simply describe the wealth of beliefs, arguments, deliberations, and negotiations that are usually at work in scientific practice, as Bloor says above, “without regard to whether the beliefs are true or the inferences rational.” In this way, we can seek to explain why scientific beliefs are in fact accepted without considering whether they are, at the same time, rationally or justifiably accepted. And in such a descriptive, purely naturalistic enterprise, there is precisely enough room for sociological explanations of why certain scientific beliefs are accepted as the empirical material permits. Whether or not philosophers succeed in fashioning a normative or prescriptive lens through which to view these very same beliefs, arguments, deliberations and so on, is entirely irrelevant to the prospects for empirical sociology. In this sense, there is simply no possibility of conflict or competition between “nonnaturalistic,” philosophical investigations of reason, on the one hand, and descriptive, empirical sociology of scientific knowledge on the other. (1998, 245)

This is a very rich statement, and it provides an alternative picture, with a distinctive but nevertheless impeccably Kantian pedigree, for an account of the relation of the causal or sociological and the normative that does not invoke normativity as something in the same explanatory domain as the causal, that is to say as a fact which is relevant to explanation.

Friedman prefaces this statement by asking “why SSK represents itself as in conflict or competition with traditional philosophy. Why do we not simply acknowledge the fundamental divergence in aims and methods and leave it at that?” (1998, 244). The question has the effect of putting the normativist, as distinct from the normative philosopher of science, out of business, for it treats the normativity that is all around us as something we impose on a causal world that is unaffected by our “fashioning a normative prescriptive lens” through which to view it. If Friedman’s picture is right, there is no problem of finding the place of normativity in the causal world because the idea that there is such a place is simply false.

The presence of normative language and normative belief, from Friedman’s point of view, tells us nothing about the presence of some sort of shadow realm of normativity. We are the sources of normativity. The normativity we find in the world, “the meaning and values that had been previously discerned in things,” were put there by us, and we had no effect on the things themselves. Normative reasoning is and should be trapped in this circle of imposing, and then falling for the illusion that we are discerning that which we have imposed. The normativist rejects this picture, and even on the basis of the little that has been said here about the idea of normativity, one can see why. Friedman, in his first sentence, insists that we may abstain from normative or prescriptive considerations. The normativist, in contrast, says that such considerations are ineliminable and are an integral and inextricable part of the phenomena to be explained, and part also of the explanations themselves. The normativist denies that we can simply describe the wealth of beliefs, negotiations, and the like that go into scientific practice without reference to normative considerations, such as whether an inference is rational. Even identifying the beliefs requires us to attribute rationality to the scientists we are studying. Explaining their inferences without reference to their rationality, the normativist would say, is not possible. Applying decision theory to account for their errors, for example, requires an appeal to the normative model of decision theory itself. Science, the normativist would say, is a normative concept, not a sociological one, and distinguishing science from voodoo requires the normative sense of science, something that the causalist as Friedman describes him cannot appeal to. The SSK model of explanation uses the notion of belief as a cause. But belief is itself, normativists routinely claim, a normative concept, as is the concept of concept. And the SSK model relies on the idea that explaining a belief can be done on the basis of beliefs alone. But the explanation of scientific beliefs as science necessarily involves the normative relation of truth between the objective world and the beliefs in question, since science aspires to truth, not merely to the satisfaction of one or another subjective belief about truth. The normativity that is everywhere is a matter of some kind of fact, or a condition of facticity itself: the idea that normative considerations can be understood as a matter of “fashioning a normative lens” is anathema.

The door Friedman’s argument opens is to a way of making sense of normative language as normative without invoking a shadow world of normativities, or a dualistic metaphysics of the normative standing in some problematic relation to the natural. The normativist wants this door closed, because it looks like an attractive option when we begin to inquire into the difficulties of normativism. It is these difficulties that are the focus of this book.

The standard argument form

My concern in this book will be with the common form of argument underlying various assertions about the necessity or indispensability of the normative, and therefore of “normativity.” O’Neill’s comment about disagreement, it turns out, is partly true and partly false. What is true is that there are a variety of normativity arguments, each of which accounts for a different kind of normativity, which are not easily reconciled with one another. There are also different ideas about the sources of particular kinds of normativity, such as legal or semantic normativity. But there are, nevertheless, extremely strong family resemblances between these arguments. And the family exhibits a large number of genetic defects. In the rest of this chapter I will briefly describe a standard-type or ideal normativity argument and then describe the defects to which this family of arguments is prone. In the following chapter I will do something similar for the naturalistic, social-science explanations of the normative. In Chapter 3, I will discuss a specific, paradigmatic case – as it happens, the case of one of the progenitors of this family – in detail.

The structure of normativity arguments can be summarized in terms of a series of steps with a major disjuncture in the middle: the point where the normative fact appears. The background to normative facts is ordinary, involving the kinds of facts that are part of the ordinary stream of explanation. There is nothing binding, compelling, or constraining about these facts. So these new normative facts constitute a rupture in the world of ordinary fact. The normative, however, arises out of ordinary facts: meanings, obligations, rationality, and so forth come into existence through actions, learning, and the like but have the special added properties of norms: of binding, constraining, and the rest. Once the norms are established, they have consequences for behavior. They do not directly cause behavior, but they regulate it normatively, by specifying what is the right way to say something, what obligations one has, what one owes to others as a result of one’s meaningful actions, and what is justified for others to do in response to your actions.

The beginning point of the structure is that which can be naturally or causally explained. One can, as a kind of shorthand, refer to these as dispositional explanations, because they typically involve dispositions of some kind – expectations, for example, that are produced by some finite process such as learning. It is important to note that normativists ordinarily do not deny the existence of these causes, or say that they explain nothing. Instead they claim that the phenomena to be accounted for, which the normativist takes as distinctively normative, are more than these kinds of considerations can explain. Expectations may have, for example, a role to play in the creation of obligations, but obligation is more than expectation. The “more” is taken to be the distinctively normative content, which cannot be accounted for by the causal or dispositional explanations at hand, such as the causal explanations of learning that account for the dispositions that produce people’s expectations.2 For convenience, one can keep a few of these special objects in mind: meanings, binding laws, obligations, rules, and so forth. None of them can be fully accounted for by the available causal accounts, or so it is argued.

Typically there is an issue of description with respect to these facts, an issue over what I have called the double structure of genuine and sociological or empirical. The issue may take a variety of forms. In the case of law, as we will see, the question is whether what can be explained causally or nonnormatively is “really law.” Sometimes the issue is formulated in terms of qualitative differences or differences between the things that need to be explained and the explainers. Semantic normativity, for example, or the rules governing meaning, is infinite in character (in the sense that the meaning of my statement today does not change in the future and I am thus in some sense committed to this meaning in perpetuity), while the explanatory facts that make up such things as the dispositions to use words in given ways, which are the result of learning, are finite, as the learning process itself is. Often this special character is simply assumed or asserted. But the problem of description becomes apparent when a social-science explanation already exists, and is criticized by arguing that the explananda does not match the object to be accounted for. In short, the normativist claims in each case that there is a novel thing which social science can’t explain, and can’t explain because it is of a novel kind.

The novel thing that can’t be accounted for in the normal (social science) way must be accounted for in another way. Typically this is done by transcendental arguments, in which conditions of the possibility of this thing are identified. These conditions can then be asserted to have been met by virtue of (the uncontroversial fact of) the existence of the thing itself, such as a “meaning.” These conditions for possibility are necessarily different in kind from the conditions that ordinarily operate in causal contexts, or else there would be no need to appeal to the notion of normativity. As noted earlier, the existence of these things is typically supported by tu quoque arguments to the effect that anti-normativists “help themselves” to concepts, considerations, and so on that are the property of the normativist, or that they speak about them in a way that “changes the subject.” A typical example is the notion of concept. The anti-normativist will refer to concepts or necessarily employ concepts in their own thought, and thus concede the normativists’ point: that normativity is real and necessary. Similarly for meaning: the normativist can invoke tu quoque arguments: that, for example, to insist that a rejection at the notion of meaning would be incoherent because it would imply that the rejection itself was without meaning. Or the anti-normativist will make the reflexive error of referring to a “good theory” and thus concede the reality of goodness.

Non-philosopher anti-normativists – for example, cognitive scientists who are concerned to account theoretically for the kind of thinking called conceptual – are unimpressed by these arguments for the simple reason that they consider the concept of “concept” to be up for grabs and a matter for future science to determine through the normal process of testing theoretical hypotheses against data. And this was the response of the Strong Programme as well: we are just scientists. In science, nothing hangs on the essential definition of terms; what counts as real depends on the findings and the theories that explain them. And this is also the attitude of philosophical naturalisms, such as Quine’s. This is the point at which the conflict occurs. The normativist relies on an argument of conceptual necessity: if one says x, one must accept its conceptual precondition, y. The anti-normativist, typically, argues that there is nothing out of the ordinary stream of explanation, so there are no special facts of the kind normativism insists on. Whatever “meaning” and “concepts” are, accordingly, they are not going to have the properties normativists ascribe to them, such as the power ascribed to semantic rules of binding into the infinite future. The normativist reply to this is that they must have these properties to support inferences, to justify, and so forth: to abandon these special properties is to give up on intelligibility itself.

This is the core issue. Why? Because the claim of the anti-normativist is that the normative moment in the middle of the standard story, the source of all the issues, is eliminable. Can the social scientist provide explanations that get from the inputs, like learnings, dispositions, beliefs, and so forth, to the outputs, that is to say, the empirical facts of communication and behavior which need explanation, without appealing to the normative concepts of meaning, obligation, reason, and the like? If this is the case, normativity needs to be understood not as fact but as an explanatorily inert aspect of facts. Joseph Raz speaks in a way congenial to this when he says that there is a normative aspect to reasons, and that the goodness of a reason or its normative significance is an aspect (Raz 1999, 113). This way of thinking of normative significance leaves a great deal of scope for reflection and for fashioning a normative lens to better understand this aspect. It leaves no scope for grounding anything on normativity understood as fact. As facts, they are facts solely in the sense that they appear as such in a particular lens. They do no explanatory work and are not part of the mechanism.

The disagreement involves both what is to be explained and how it is to be explained. The anti-normativist accepts that there may be a need for theoretical terms, as in science generally, thinks that the thing to be explained is the observables (such as human behavior), and thinks that normal explanations, such as causal explanations, will suffice, and have to suffice. If they do not, we have a genuine mystery, which doesn’t help the normativist either. The normativist thinks that this misses the point. There are things that need to be accounted for that are not only not mysterious, but absolutely central to our life and thought, and thus have to be accounted for. Moreover, the account should occupy a central place in our metaphysics. As Sellars puts it, the problem is to fuse the scientific image with the normative one. If the conditions for the possibility of these essential things are recherché, and involve exotic objects, that is a price that must be paid: no one said philosophy would be easy or correspond with common sense. Paul Boghossian puts this issue clearly when he lists a series of peculiarities that arise because “meaning properties appear to be neither eliminable nor reducible,” and then suggests that “perhaps it is time we learned to live with that fact” (1989, 548). For the anti-normativist, this is mystery-mongering.3

The “does it matter” problem

The key to the appeal of the concept of normativity is that normative facts are irreducible, meaning impossible to explain in nonnormative terms, and at the same time ineliminable. But in what ways are they ineliminable? The normativist needs to say that they are ineliminable from explanation: if “a normative must is to have a distinctive place in the world” (Railton 2000, 4; emphasis in the original), the “must” needs to have some explanatory role. Does it? Scientists may be unable to talk about their theories without using normative terms like “good” and “elegant” or even “must.” But this normative language does not, in any simple or direct way, require us to believe in normativities. It would be a different matter if we could not explain their ways of talking about their theories without appealing to the notion of normativity, and this is what the normativist needs to claim: that a special fact of normativity is presupposed or required in some sense by this talk or to make this talk intelligible. But there is a generic problem with such claims. The problem is that ordinarily the explanation of actions involves beliefs. The validity of the beliefs, normative or otherwise, is not explanatory in itself. Beliefs in nonexistent things, such as ghosts, also explain people’s behavior, their other beliefs, and so forth.

The issue is not a new one. Weber argued that religious rationalizations of the problem of theodicy had profound causal effects on European history, and, indeed, that this problem set in motion a chain of effects that led to the same disenchantment of the world that Robert Brandom mentions. Weber claimed that the reflections of John Calvin on the fundamental questions of life and the universe, framed, as one would expect, in terms of the theology of the time, were driven by a resolute attempt to make Christian theology rational and consistent, especially with regard to the central (and as Weber argued, for salvation religions, universal) problem of theodicy. This argument, however, seems to involve a normative concept of rationality in which a human drive for consistency can have powerful effects. Weber, it could be argued, has fallen inadvertently into normativism, helping himself to the normative notion of consistency. And this normative notion, in this context, seems to do some important explanatory work. Alasdair MacIntyre made precisely this argument, though not in these terms. He argued that the relationship between Calvinist doctrine and the effects Weber described was “logical” (1962, 55), and credited Weber for having discerned it, but argued that Weber made a mistake by thinking it was a causal explanation. If the explanation was not causal, logic would be doing explanatory work.

But did Weber make this mistake? A final answer to this question will need to be delayed until the last chapter. It will suffice here to note that rationality is another example of a concept that has a dual nature. It is a normative concept. There is also a philosophical tradition that treats rationality as a disposition (Hempel 1965, 469–87; cf. Davidson [1976] 1980, 273). Weber’s notion of rationality seems to be dispositional as well as normative. But the dispositional aspect is all that is needed to do the explanatory work. The normative consideration of whether the conclusions drawn by Calvin’s followers were genuinely ‘logical’ seems to make no difference at all, and to add nothing, to the explanation, as an explanation. As Friedman’s argument suggests, explaining can make do by treating beliefs as causes, and can leave judgments about rationality or theological correctness – in this case the same thing – to the theologians.

The explanatory peculiarities of normativity

The argument form I have outlined here is a composite of various normativity arguments. Not all the arguments in the family of normativity arguments concern themselves with the whole set of claims. Many are concerned only with the transcendental arguments that can be employed to account for normative language, and ignore even the distinction between the sociological and the genuine sense of “norm.” In some ethical writers, for example, the normative character of this and that is taken as self-evident (Nagel 1986, 159–60; Korsgaard 1996, 41), and the existence of alternative accounts of norms is not even acknowledged. Ignoring them amounts to taking a free pass on the central problem of normativism: the strangeness of the explanations given by the normativist. But some of the literature on normativity frankly addresses, or at least acknowledges, that the notion of normativity is unusual and problematic. The response to this unusualness has been to say that one must learn to live with it (Boghossian 1989, 548), to embrace it (Brandom 1979, 192), to make it a foundation of a radical normativist antidualist ontology so that everything is ultimately normative (Rouse 2002), to minimize the oddity by contrasting it to even more spooky entities, like those of Platonism (McDowell 1996), or to attempt to separate it from the ordinary stream of explanation (Korsgaard 1996). There are multiple aspects to the problem of this strangeness, which I will simply list and briefly elaborate here.

Queerness

The claim that there are such things as normativities and normative powers is characteristically bound up with claims about special and puzzling objects: self-authorizing principles; rules, understood as tacit rules behind the explicit rules we might invent or propose; mathematical objects, which are not realized in the physical world; objective values; dictates without dictators, the dictates of reason itself; performative utterances which have the property of being able to create a norm, through the act of utterance of a special ritual kind; the nonspatial space of reason; the constraining ideal structure of thought experiments; intentions without intenders, except for supposed collectives, which, despite being wholly analogical, have the power to create and warrant norms; commitments that are unlike ordinary commitments in that they are made without anyone knowing that they were made, in which the person undertaking the commitments could not have understood them as commitments when they made them, or perhaps even afterward. David Lewis, in the paper from which Brandom takes the notions of commitments and score-keeping, proposes a kinematics of presupposition, in which if anything needing a presupposition is said in time t, the necessary presupposition, if it did not previously exist, comes into existence at time t (Lewis 1979, 340). Lewis formulates a “rule of accommodation for permissibility” as follows:

If at time t something is said about permissibility by the master to the slave that requires for its truth the permissibility of certain courses of action, and if just before t the boundary is such as to make the master’s statement false, then – ceteris paribus and within certain limits – the boundary shifts at t so as to make the master’s statement true. (Lewis 1979, 341)

The range is from these objects whose lifetimes extend only through the utterance of a sentence to immutable Fregean concepts and the Platonic forms themselves. These odd objects, it may be noted, have analogues in social theory and social science, often with equally mysterious features, such as the collective consciousness in Durkheim, or the notion of the central value system of a society in Talcott Parsons, or the “essentially unavoidable and irremediable haecceity of immortal ordinary society” of Harold Garfinkel’s ethnomethodology (1991, 10, 19n10).

Dependence on transcendental arguments

The claim that these things exist usually rests on a particular type of argument which is transcendental, or presuppositional. The claim is that such and such an object, space, intention, or the like, is a condition for the possibility of something that is itself indispensable, obviously true, constitutive of our humanity, reason, or the like. The problem with these arguments is their use: they seem to be stand-ins for causal explanations, but fail to live up to any of the standards of explanation that causal explanations are normally held to.

Construction by analogy

Many normative objects, especially those that figure at the end of regression arguments, are analogical. The objects include rules that are not explicit rules, but rather tacit rules hidden behind rules; commitments that are not commitments in an explicit intentional sense, but something given preintentionally or tacitly; laws, such as the Grundnorm, which is not itself enacted as law and so forth. Analogical usages are common enough in philosophy – “true” and “norm” are terms from carpentry, for example. But these devices used in these arguments have an odd common structure: they are hidden or tacit forms of explicit things, and are needed to do work, such as justification, that the explicit things cannot do, namely stop a regress.

Externality

The thingness of normativity is the source of its normative force – it is in some sense outside us and inside us at the same time. When we recognize it fully we are at the same time recognizing that we must obey. There is a parallel with the ontological argument for the existence of God here. To merely treat the normative as conventional, pragmatic, or sociological is to fail to fully understand it. To fully understand it is to believe, to accept, or to acknowledge our pre-acceptance of it and dependence on it.

Acknowledgment

The common thread in these claims about normativity is in our relation to it. It is something to be acknowledged and accepted as a constraint or obligation – something that is, in the Heideggerian phrase, always already there, but which we only express or recognize through reflection, or admit the power of when we reach the age of reason. McDowell formulates this clearly:

The idea is that the dictates of reason are there anyway, whether or not one’s eyes are opened to them; that is what happens in a proper upbringing. We need not try to understand the thought that the dictates of reason are objects of an enlightened awareness, except from within the way of thinking such an upbringing initiates one into: a way of thinking that constitutes a standpoint from which those dictates are already in view. (1996, 91–92)

Duality and Circularity

These objects trespass across the line between is and ought. They are facts that warrant values or oughts or that are values or oughts, or from which normative language can be derived. But are they normative themselves? The question typically produces a dilemma, which can be seen in the case of Kripke’s appeal to community standards of correctness with respect to rule-following. Are these “community” facts empirical or normative? The community here can’t be the empirical community, for the whole of the actual membership of the actual community might be in error. Nor can it be the empirical community after the error is pointed out, because the members of the community might also be wrong about this. To appeal to the empirical community leads either to a regress or to the abandonment of an idea of normativity beyond the empirical fact of community assent. But the community can’t be a normative fact either – such as the community of those who are correctly admitted to membership in the community by virtue of having correctly grasped the rule. This results in circularity – which is also a problem for the hypothetical community of rational beings.

The claim of surplus force

A major underlying issue with these odd objects – such as notions of meaning or conceptual content – is this: there is a gap, surplus meaning, or what I have called a disjuncture, between the naturalistic explanations – dispositions, paradigmatically – and the thing to be explained. The problem is with what dispositions or anything else could possibly explain: if they explain law as systematic rules with sanctions, they don’t explain why the rules are obligatory. That seems to require something more: normative force. By making this distinction, it seems we create a novel ontological realm beyond explanation, which has mysterious properties. But what necessitates the existence of this realm? A real force? Or our desire to appeal to obligations and the like as justifications?

The question of whether linguistic normativity serves as the fons et origo of all normativity

There is more than a hint of this idea in much of the writing on normativity (Brandom 1979, 190). But it runs into a stubborn issue within linguistic normativity, which undermines any extension to other domains. What is the logical relation between counts as normativity and obligations to speak correctly in accordance with these conventions? (Brandom 1997, 193; Kusch 2006, 51–55). What is the normative force of the conventions? Are they merely conventions, coordination devices, in David Lewis’s phrase (1969, 42), which we adopt pragmatically, like driving on the left or right, which derive their normative force from other, in this case moral or legal considerations, such as the fact that by driving on the incorrect side one is likely to kill or be killed or incur sanctions? Or do they derive directly – so that by acknowledging that there is a correct and incorrect way we immediately acquire or admit an obligation to act accordingly? This is McDowell’s picture:

To learn the meaning of a word is to acquire an understanding that obliges us subsequently – if we have occasion to deploy the concept in question – to judge and speak in determinate ways, on pain of failure to obey the dictates of the meaning we have grasped. (McDowell [1984] 2002, 45)

The problem of separating normative from pragmatically successful

For some philosophers, the issue of normativity comes down to the issue of semantic normativity, which can be construed as the source of all normativity. But there is a problem here, which is revealing about the larger question of how various notions of normativity (and various types of normative language) relate to one another. The core reasoning is this: semantic normativity is signaled by the distinction between correct and incorrect application of terms, and the problem of correctness is intrinsic to language, which makes language and anything linguistic by definition normative. The core kind of normativity here is constitutive: it is about what counts as something – what counts as a horse or a duck. But what is the difference between using an utterance successfully, for example when ordering dinner in a foreign language, and using it correctly in the normative sense? And why should one think there is a difference between the two?

There are no grammatical errors in a preliterate society – there is simply being understood and not being understood. In one respect – the empirical, pragmatic one – semantic normativity is governed by the same principle: the correctness of semantic use is no more or less than being understood. So how does genuine semantic normativity – correctness beyond mere usage – arise? Or is there such a thing? In the case of syntax, we produce the normativity of grammar by writing grammars and then using them as a guide or norm of correct speech, regardless of what people actually understand. It is an add-on, in which some model of speech is taken as normative, or in which someone’s theory of what is right is imposed on actual, divergent, but intelligible speech to stigmatize something as bad, erroneous, and the like. Is genuine semantic normativity more than this? Presumably it is. So we have two basic models of meaning: one is of an ideal object, collectively held, that has the power of distinguishing correct and incorrect usage infinitely into the future; the other is a notion of meaning as works in interaction, as speech behavior, a functional summary of the actual behavioral usages that produce a specific set of similar results when used with certain people in diverse situations, and which is open-ended and changes through adaptation and extension.

The transition problem

Each account of normativity that makes a distinction between the normative and the nonnormative – that is, accounts other than those that collapse everything into the normative – involves some sort of transition between a nonnormative or pre-normative state and the normative one. This is even a problem for the metaphysical normativist, who believes that everything is normative, and argues that the evidence for this includes the presence of various kinds of normativity. When this normativist tries to account for the production of one kind of normativity out of another, a form of the transition problem arises. There is a standard successful model for creating norms by relying on norms: as in the case of the norm (such as a constitution) that gives authority to a norm-giver to create norms (such as a legislature). But the chain of normative justification needs to end somewhere. Does it end in something nonnormative, such as Wittgenstein’s “what we do,” or something normative which is basic – Kelsen’s Grundnorm, for example? If it is the former, how does something nonnormative produce something normative? Does it have to end in a norm-creating act of will, such as the will of the people, or in norm-creating commitments, on the model of obligations freely entered into? How do these transitions work, exactly?

These accounts model normative commitment and the like on law. But even in the case of law there are problems. H. L. A. Hart recognized that Kelsen’s answer did not work – that even Grundnorms needed to be grounded in something more basic (Hart 1961, 117; cf. Postema 1987, 89)4. J. L. Austin’s notion of performative utterance5 might seem to be a solution to the problem, at least the problem of transition from non-law to law. A law was something said to be law by the appropriate person in the appropriate setting. But this is a purely formal solution. The explanation is missing. One needs to know something about why the person or setting is appropriate. Since “appropriate” and similar terms seem to be normative concepts, the problem of normativity is not solved by appeal to them. It is just pushed a step back, and the answers there are problematic. The idea that by going through the ritual of voting and signing pieces of paper, legislatures and executives or parliaments transform utterances by magically infusing them with the collective will, thus making them obligatory, was ridiculed in some of the funniest writing in the history of philosophy by the Scandinavian legal realist Axel Hägerström (1953, 74–116), as Hart knew well. Hägerström’s point was that the whole appeal to will in these cases was fanciful. The people who were supposedly doing the willing into existence of laws, such as the sovereign, were ordinarily unaware of the content of the law, much less engaged in an act of infusing them with normative force. He concluded that ideas about will make more sense as political fictions than as true accounts of the nature of law.

The problem of transition arises in a variety of forms other than this paradigmatic one. In McDowell, there is the point at which the child comes to recognize the bindingness of norms of rationality. In Kripke, there is the point at which a person is taken into the community as a competent rule-follower. As with Kelsen, the issue haunts those who seek to evade it, such as Brandom, who places himself in the paradoxical position of basing his conception of the normativity of score-keeping (thus of language proper) on the normative fact of commitment, but denies that prelinguistic conduct is properly intentional, which seems to mean that it cannot include commitments.

The problem of belief

The epigraph from Calvin that prefaces this chapter refers to the central episode in the disenchantment of the world, the denial that “Jesus Christ is corporeally, really, and in fact entirely and personally in the flesh contained and concealed in the species of bread and wine, as grand and as perfect as if he were living in the present” (Holt 1995, 18). Calvin, as the epigraph indicates, believed that what you got out of the Eucharist was what you put into it in the form of faith. The normativist does not merely urge us to believe and make it so, in the fashion of Augustine, but rather insists that the normativity in question is real apart from our beliefs, and a matter of discovery or acceptance, or that something different in kind, a normative fact, is created by our actions.

The problem of description

According to an old joke, a Texas Baptist was asked whether he believed in infant baptism. “Believe, hell,” he replied, “I seen it done!” The joke works, if I may belabor the obvious, because there is a difference between the normative and theological sense of “I believe in infant baptism” and the factual sense. In the normative sense, the Texan would have been in error: a true Baptist believes that infant “baptism” so-called is not baptism at all, so in fact he could not have seen such a thing. In another and equally familiar sense, of course, he is correct. He had seen what a non-Baptist would call, and he could also, speaking in non-Baptist terms, call, “infant baptism.” The normativist is constantly forced to take the theologians’ side of this argument. The arguments for the existence of normative things, and for the attribution of powers to them, begin with establishing a privileged description of the things to be explained. This is a traditional problem of the philosophy of social science, perhaps one of its central problems. Peter Winch’s The Idea of a Social Science (1958) is based on the insistence that to talk about human behavior in terms other than the (normative) conceptual terms of ordinary action is to change the subject.

The short “scientistic” reply to such arguments is that scientists are under no obligation to employ any particular vocabulary, and it should be no surprise that explanations may be available in one vocabulary and not another. Changing vocabularies is part of the explanatory enterprise. Social scientists, however, may be in a different position, as Weber (1949, 111) argued, and need to use value-related concepts to define their explanatory problems. But they can use terms such as “norm” or “law” in a purposely nonnormative way to denote the empirical equivalent of the term, to bracket the normative aspect, and thus free themselves of the normative implications of the concept. This possibility is not limited to scientists. The Texan could have saved himself from the joke by explaining that what he had seen done was what people mistakenly believed was baptism. And this raises an important point: we use false descriptions that we disagree with all the time. We use them along with an error theory or error explanation which accounts for the difference between the descriptions used by others and those we use.

This response bears on the tu quoque