WHY WE CAN'T UNDERSTAND THOUGHT FROM THE OUTSIDE 1 page
THE LAST WORD
New York Oxford Oxford University Press 1997
Oxford University Press
Oxford New York Athens Auckland Bangkok Bogota Bombay Buenos Aires Calcutta Cape Town Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madras Madrid Melbourne Mexico City Nairobi Paris Singapore Taipei Tokyo Toronto
Published by Oxford University Press, Inc. 198 Madison Avenue, New York, New York 10016
Oxford is a registered trademark of Oxford University Press
All rights reserved. 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 Oxford University Press.
Library of Congress Cataloging-in-Publication DataNagel Thomas.
5 7 9 8 6 Printed in the United States of America on acid-free paper
To Ronald Dworkin and Saul Kripke
The main intellectual influences on this book come from the two friends to whom I have dedicated it.
In the late 1970s I attended a seminar Saul Kripke gave at Princeton, in which he attacked various forms of relativism, skepticism, subjectivism, or revisionism about logic. He argued that classical logic could not be qualified in any of those ways, that it was simply correct, and that the only response to alternatives such as quantum logic, for example, was to argue against them from within classical logic. In any case, he pointed out, the skeptics all rely on it in their own thinking.
Since 1987 Ronald Dworkin and I have regularly taught together, and I have been exposed to his constant insistence that the only way to answer skepticism, relativism, and subjectivism about morality is to meet it with first-order moral arguments. He holds that the skeptical positions must themselves be understood as moral claims--that they are unintelligible as anything else. I would not go so far as that, but I have been led to the view that the answer to them must come from within morality and cannot be found on the metaethical level.
These two realist viewpoints, from two different regions of philosophy, have a great deal in common and have led me to the general conclusion that the last word in philosophical disputes about the objectivity of any form of thought must lie in some unqualified thoughts about how things are--thoughts that remain, however hard we may try to get outside of them or to regard them merely as contingent psychological dispositions.
I have presented portions of the material to various audiences. As is true of most of my recent work, it was discussed to my profit in several sessions of the Colloquium in Law, Philosophy, and Political Theory at New York University. In 1995 it provided material for the Carl Gustav Hempel Lectures at Princeton, the Alfred North Whitehead Lectures at Harvard, the Immanuel Kant Lectures at Stanford, and a Lionel Trilling Seminar at Columbia. I am grateful for critical attention from Paul Boghossian, Ronald Dworkin, Colin McGinn, and Derek Parfit. My research during the time of writing was supported by the Filomen D'Agostino and Max E. Greenberg Faculty Research Fund of New York University Law School.
New York T. N. April 1996
2 Why We Can't Understand Thought from the Outside
7 Evolutionary Naturalism and the Fear of Religion
This discussion will be concerned with an issue that runs through practically every area of inquiry and that has even invaded the general culture--the issue of where understanding and justification come to an end. Do they come to an end with objective principles whose validity is independent of our point of view, or do they come to an end within our point of view--individual or shared--so that ultimately, even the apparently most objective and universal principles derive their validity or authority from the perspective and practice of those who follow them? My aim is to clarify and explore this question and to try, for certain domains of thought, to defend what I shall call a rationalist answer against what I shall call a subjectivist one. The issue, in a nutshell, is whether the first person, singular or plural, is hiding at the bottom of everything we say or think.
Reason, if there is such a thing, can serve as a court of appeal not only against the received opinions and habits of our community but also against the peculiarities of our personal perspective. It is something each individual can find within himself, but at the same time it has universal authority. Reason provides, mysteriously, a way of distancing oneself from common opinion and received practices that is not a mere elevation of individuality--not a determination to express one's idiosyncratic self rather than go along with everyone else. Whoever appeals to reason purports to discover a
source of authority within himself that is not merely personal, or societal, but universal--and that should also persuade others who are willing to listen to it.
If this description sounds Cartesian or even Platonic, that is no accident: The topic may be ancient and well-worn, but it is fully alive today, partly because of the prevalence of various forms of what I (but not, usually, its proponents) would call skepticism about reason, either in general or in some of its instances. A vulgar version of this skepticism is epidemic in the weaker regions of our culture, but it receives some serious philosophical support as well. I am prompted to this inquiry partly by the ambient climate of irrationalism but also by not really knowing what more to say after irrationalism has been rejected as incoherent--for there is a real problem about how such a thing as reason is possible. How is it possible that creatures like ourselves, supplied with the contingent capacities of a biological species whose very existence appears to be radically accidental, should have access to universally valid methods of objective thought? It is because this question seems unanswerable that sophisticated forms of subjectivism keep appearing in the philosophical literature, but I think they are no more viable than "crude" subjectivism. 1.
To begin with the crude kind: The relativistic qualifier-"for me" or "for us"--has become almost a reflex, and with some vaguely philosophical support, it is often generalized into an interpretation of most deep disagreements of belief or method as due to different frames of reference, forms of thought or practice, or forms of life, between which there is no objective way of judging but only a contest for power. (The idea that everything is "constructed" belongs to the same family.) Since all justifications come to an end with what the
In general, I'll use the term "subjectivism" rather than "skepticism," to avoid confusion with the kind of epistemological skepticism that actually relies on the objectivity of reason, rather than challenging it.
people who accept them find acceptable and not in need of further justification, no conclusion, it is thought, can claim validity beyond the community whose acceptance validates it.
The idea of reason, by contrast, refers to nonlocal and nonrelative methods of justification--methods that distinguish universally legitimate from illegitimate inferences and that aim at reaching the truth in a nonrelative sense. Those methods may fail, but that is their aim, and rational justifications, even if they come to an end somewhere, cannot end with the qualifier "for me" if they are to make that claim.
The essential characteristic of reasoning is its generality. If I have reasons to conclude or to believe or to want or to do something, they cannot be reasons just for me--they would have to justify anyone else doing the same in my place. This leaves open what it is for someone else to be "in my place." But any claim that what is a reason for me is not a reason for someone else to draw the same conclusion must be backed up by further reasons, to show that this apparent deviation from generality can be accounted for in terms that are themselves general. The generality of reasons means that they apply not only in identical circumstances but also in relevantly similar circumstances--and that what counts as a relevant similarity or difference can be explained by reasons of the same generality. Ideally, the aim is to arrive at principles that are universal and exceptionless.
To reason is to think systematically in ways anyone looking over my shoulder ought to be able to recognize as correct. It is this generality that relativists and subjectivists deny. Even when they introduce a simulacrum of it in the form of a condition of consensus among a linguistic or scientific or political community, it is the wrong kind of generality--since at its outer bounds it is statistical, not rational.
The worst of it is that subjectivism is not just an inconsequential intellectual flourish or badge of theoretical chic. It is used to deflect argument, or to belittle the pretensions of the
arguments of others. Claims that something is without relativistic qualification true or false, right or wrong, good or bad, risk being derided as expressions of a parochial perspective or form of life--not as a preliminary to showing that they are mistaken whereas something else is right, but as a way of showing that nothing is right and that instead we are all expressing our personal or cultural points of view. The actual result has been a growth in the already extreme intellectual laziness of contemporary culture and the collapse of serious argument throughout the lower reaches of the humanities and social sciences, together with a refusal to take seriously, as anything other than first-person avowals, the objective arguments of others. I am not going to address myself directly to the manifestations of this attitude, but it is there as a source of irritation in the background--though I don't seriously hope that work on the question of how reason is possible will make relativism any less fashionable.
Many forms of relativism and subjectivism collapse into either self-contradiction or vacuity--self-contradiction because they end up claiming that nothing is the case, or vacuity because they boil down to the assertion that anything we say or believe is something we say or believe. I think that all general and most restricted forms of subjectivism that do not fail in either of these ways are pretty clearly false.
My own opinion is that there is such a thing, or category of thought, as reason, and that it applies in both theory and practice, in the formation not only of beliefs but of desires, intentions, and decisions as well. This is not to say that reason is a single thing in every case, only that certain decisive aspects of our thought about such very different matters can all be regarded as instances of it, by virtue of their generality and their position in the hierarchy of justification and criticism. I shall refer to this as the rationalist position. My aim will be to see whether it can be given a plausible form. How can one reconcile the unqualified character of the results at which
we aim by reasoning with the fact that it is just something we do?
Every major philosopher has had something to say about this. My own sympathies are with Descartes and Frege, and I will attempt to resist the limitation of the reach of human reason that is found in different ways in the treatments of Hume, Kant, and, on a common reading, Wittgenstein. More recently, versions of it are found in W. V. Quine, Nelson Goodman, Hilary Putnam, Bernard Williams, and Richard Rorty. These forms of subjectivism shrink from the apparently audacious pretensions of human thought and tend to collapse its content into its grounds, so that it doesn't reach as far beyond us as it appears to do. For the most part, I shall be arguing not against the positions of particular philosophers but against a general tendency to reduce the objective pretensions of reason, a tendency that manifests itself in many philosophical arguments--vulgar and sophisticated--and that is a constant temptation to those trying to make sense of the phenomenon. The position I oppose will be this form of subjectivism as I interpret its allure--a position which can sometimes seem the only possible account of the subject, given that we are who we are, but which I believe cannot be rendered intelligible.
We must distinguish between general philosophical challenges to the objectivity of reason and ordinary challenges to particular examples of reasoning that do not call reason into question. In order to have the authority it claims, reason must be a form or category of thought from which there is no appeal beyond itself--whose validity is unconditional because it is necessarily employed in every purported challenge to itself. This does not mean that there is no appeal against the results of any particular exercise of reason, since it is easy to
make mistakes in reasoning or to be completely at sea about what conclusions it permits us to draw. But the corrections or doubts must come from further applications of reason itself. We can therefore distinguish between criticisms of reasoning and challenges to reason.
If reasoning is what has been going on, then criticism of its results must reveal mistakes in reasoning, where these, too, are universally mistakes. Whenever we challenge a conclusion by pointing out a mistake in someone's arithmetic, or logic, or their failure to consider a possibility that is not ruled out by the evidence, or the disanalogy between two cases that have been lumped together, we remain within the territory of rational justification and criticism and do not cast doubt on whether our interlocutor is employing a generally valid method for reaching the objective truth. This internal type of criticism and evaluation imports nothing subjective.
There is an external form of criticism, on the other hand, which undermines the conclusion precisely by questioning the objectivity of its grounds. One important way of challenging from outside what is presented as a product of reason is to claim that it is not the result of reasoning at all, valid or invalid, but rather something else: the expression of a particular personal or cultural perspective of less than universal validity, perhaps artificially rationalized or objectified in an act of intellectual self-deception. Sometimes one can challenge a particular piece of ostensible reasoning in this way without implying any doubt that reason of that type is possible. The ordinary charge of "rationalization," like the exposure of errors in reasoning, does not question the claims of reason itself but rather presupposes them. It contrasts the sources of belief in this case with an alternative type of ground that would actually justify them, or demonstrate their truth.
But this type of diagnosis can also take a more general form and can be used to make a philosophical point. Depending on the case, the criticism may aim either to discredit the
putative rationally based claim altogether, or merely to show that it is something different, less universal but conceivably better founded than it would be under its rationalistic interpretation. If the aim is to show that reasoning is the wrong method for arriving at or backing up conclusions of the kind under discussion, then one would not describe the use of the correct, alternative method as a rationalization but would instead argue that calling it reason is a misinterpretation. This last strategy often plays a role in the attack on reason as part of the basis of ethics, when the aim is not to debunk ethics but to reveal its true grounds. 2.
On the other hand (to complete the spectrum of possibilities), such diagnoses can sometimes be offered neither as criticisms nor as alternatives but as reductive interpretations of what reason really is--namely, a contingent though basic feature of a particular culture or form of life. The usual set of moves among realism, skepticism, and reductionism occurs here as everywhere in philosophy: Reductionism (a subjective or relativist reinterpretation of reason) seems to offer a refuge from skepticism if realism (the strongly universalist position) seems too hard to sustain. 3. Being a realist about reason myself, I regard these reductive "rescues" as equivalent to skepticism; that is, they are forms of skepticism about the reality of what I myself take reason to be. Their proponents would describe them differently--as denials that my understanding of the nature of reason is correct.
Whether they are frankly skeptical or accommodatingly reductive, these sorts of diagnoses challenge the strongly rationalistic--Platonic or Cartesian--ideal. They may be di-
See, for example, Philippa Foot, "Morality as a System of Hypothetical Imperatives" ( 1972), reprinted in her collection Virtues and Vices ( Blackwell, 1978), and Bernard Williams, Ethics and the Limits of Philosophy ( Harvard University Press, 1985).
I discuss this triangle in The View from Nowhere ( Oxford University Press, 1986), pp. 68-9.
rected either at a particular category of claim, such as legal or ethical or scientific reason, or they may be more general. A further distinction, of greater theoretical interest, has to do with the type of diagnosis such criticisms offer of what is really going on under the heading of reason. Reason purports to offer a method of transcending both the merely social and the merely personal. And a critic of the rationalist conception, believing such double transcendence to be impossible, may say either that what is being appealed to is really an aspect of the shared practices of one's social, intellectual, or moral community (perhaps a particularly deep aspect) or that it is a deep but nevertheless individual feature of one's personal responses. In either case the claim to unconditional universal authority would be unfounded.
As I have said, such criticisms can be offered in a rationalistic framework. Then one will merely be pointing out that this particular allegedly rational justification of a conclusion does not in fact work, while assuming that such things are certainly possible. The same applies when the target is gradually broadened. Even someone who is doubtful about the claim to rationality in an entire domain of thought can continue to accord validity to the claim more generally and can even rely on it in the course of his criticism. But I want also to discuss the problem posed by the broadest type of attack: by the position that no faculty of such universal application and validity could be found within us to test and support our judgments.
I shall argue that while it is certainly possible in many cases to discredit appeals to the objectivity of reason by showing that their true sources lie elsewhere--in wishes, prejudices, contingent and local habits, unexamined assumptions, social or linguistic conventions, involuntary human responses, and so on--interpretations of this "perspectival" or "parochial" kind will inevitably run out sooner or later. Whether one challenges the rational credentials of a particular judg-
ment or of a whole realm of discourse, one has to rely at some level on judgments and methods of argument which one believes are not themselves subject to the same challenge: which exemplify, even when they err, something more fundamental, and which can be corrected only by further procedures of the same kind.
Yet it is obscure how that is possible: Both the existence and the nonexistence of reason present problems of intelligibility. To be rational we have to take responsibility for our thoughts while denying that they are just expressions of our point of view. The difficulty is to form a conception of ourselves that makes sense of this claim.
WHY WE CAN'T UNDERSTAND THOUGHT FROM THE OUTSIDE
Because of the way in which doubts about reason are raised, the issue is connected with the limits we encounter in trying to understand ourselves from the outside.
Typically, in challenging an appeal to reason not as an error of reasoning but as a rationalization, we point out that conviction is due to some source other than argument that there is no reason to accept: something that produces conviction in this case without justifying. In other words, we are in the mode of psychological explanation; once we recognize the cause, we see that alternative responses would be equally eligible, or perhaps superior.
In offering such a diagnosis to someone else, we provide an explanation of his beliefs, attitudes, conduct, or whatever in terms that he may or may not be able to accept. If he accepts the explanation, then he may find it necessary to give up the conviction, or he may retain it but withdraw his former interpretation of it--deciding, for example, that it is a conventional judgment that can legitimately express a particular and nonuniversal point of view. What we have offered him is an external view of himself, or at least of some of his judgments and attitudes.
We can also apply this sort of criticism to particular regions of judgment of our own which we have assumed to be based on reasoning whose validity is unqualified or universal. Sometimes we may conclude that, after all, that is not so. It is
not unusual in this way to come to believe that some of our moral or political convictions are more personally or socially subjective in origin than we had thought. Whether or not it leads to revision of those convictions, it is an important form of self-awareness.
However, the pursuit of self-awareness breaks down if we try to extend this kind of external psychological criticism of ourselves to the limit--which must happen if we entertain the possibility that nothing in human thought really qualifies as reason in the strong sense I wish to defend. For we are then supposed to consider the completely general possibility that there are contingent and local explanations of the sources of all our convictions, explanations that do not provide justifications as strong as reason would provide, if there were such a thing. And the question is, what kind of thought is this? It purports to be a view of ourselves from outside, as creatures subject to various psychological influences and prey to certain habits, but what are we supposed to be relying on in ourselves to form that view?
Suppose, to take an extreme example, we are asked to believe that our logical and mathematical and empirical reasoning manifest historically contingent and culturally local habits of thought and have no wider validity than that. This appears on the one hand to be a thought about how things really are, and on the other hand to deny that we are capable of such thoughts. Any claim as radical and universal as that would have to be supported by a powerful argument, but the claim itself seems to leave us without the capacity for such arguments.
Or is the judgment supposed to apply to itself? I believe that would leave us without the possibility of thinking anything at all. Claims to the effect that a type of judgment expresses a local point of view are inherently objective in intent: They suggest a picture of the true sources of those judgments which places them in an unconditional context. The
judgment of relativity or conditionality cannot be applied to the judgment of relativity itself. To put it schematically, the claim "Everything is subjective" must be nonsense, for it would itself have to be either subjective or objective. But it can't be objective, since in that case it would be false if true. And it can't be subjective, because then it would not rule out any objective claim, including the claim that it is objectively false. There may be some subjectivists, perhaps styling themselves as pragmatists, who present subjectivism as applying even to itself. But then it does not call for a reply, since it is just a report of what the subjectivist finds it agreeable to say. If he also invites us to join him, we need not offer any reason for declining, since he has offered us no reason to accept.
Objections of this kind are as old as the hills, but they seem to require constant repetition. Hilary Putnam once remarked perceptively on "the appeal which all incoherent ideas seem to have." In spite of his perennial flirtation with subjectivism, Putnam himself has restated very forcefully the case for the incoherence of relativism. 1. It is usually a good strategy to ask whether a general claim about truth or meaning applies to itself. Many theories, like logical positivism, can be eliminated immediately by this test. The familiar point that relativism is self-refuting remains valid in spite of its familiarity: We cannot criticize some of our own claims of reason without employing reason at some other point to formulate and support those criticisms. This may result in shrinkage of the domain of rationally defensible judgments, but not in its disappearance. The process of subjecting our putatively rational convictions to external diagnosis and criticism inevitably leaves some form of the first-order practice of reasoning in place to govern the
"Why Reason Can't Be Naturalized," in Hilary Putnam, Realism and Reason: Philosophical Papers, vol. 3 ( Cambridge University Press, 1983). The remark about incoherent ideas is on p. 194 of the same book. See also Peter van Inwagen , Metaphysics ( Westview Press, 1993), pp. 65-8.
process. The concept of subjectivity always demands an objective framework, within which the subject is located and his special perspective or set of responses described. We cannot leave the standpoint of justification completely, and it drives us to seek objective grounds.
It is not just that in criticizing each part of our system of beliefs we must rely on the rest, whatever it may be. The thoughts that enter into such criticism must aspire to a universality that is lacking in the thoughts criticized. Some of these standards have to be discovered; others are simply those basic and inescapable forms of reasoning that enter into all possible criticism--even when some examples of them are among the objects of criticism: The serious attempt to identify what is subjective and particular, or relative and communal, in one's outlook leads inevitably to the objective and universal. That is so whether the object of our scrutiny is ethics, or science, or even logic.
Is this in the final analysis just a fact about how we think? Or can we affirm that the authority of reason is something independent, something of which the hierarchy of our thoughts is an appropriate reflection? I am convinced that the first alternative is unintelligible and that the second must be correct.
The claim has two aspects. First, the outermost framework of all thoughts must be a conception of what is objectively the case--what is the case without subjective or relative qualification. Second, the task of bringing our thoughts within such a framework involves a reliance on some types of thought to regulate and constrain others, which identify general reasons and thereby advance objectivity. This introduces a hierarchy in which reason provides regulative methods and principles, and perception and intuition provide reason with the initial material to work on.