WHY WE CAN'T UNDERSTAND THOUGHT FROM THE OUTSIDE 2 page
We constantly move from appearance to reality in this way. For example, suppose you think you see a friend who
died last year crossing the street in a foreign city. Logic tells you that he can't be there if he's dead and that several hypotheses would remove the inconsistency:
(a) that you've mistaken someone else for him;
(b) that you were misinformed about his death;
(c) that you've seen a ghost.
The choice among these hypotheses will depend on other evidence, further judgments of consistency, inconsistency, and probability--together with general beliefs about how nature works, which are themselves in turn the product of similar forms of reasoning. The aim is to locate your awkward experience in a world that makes sense not just from your own point of view.
That suggests a familiar way of filling in the domain of reason, but the abstractly described aim of the enterprise, to arrive at thoughts and beliefs that are objectively correct, leaves open various possibilities. The content of reason may be quite rich, including strong methods of empirical justification of belief and various kinds of practical reason and moral justification; or it may be very austere, limited to principles of logic and not much else. More or less of our thought may be about the objective framework, as opposed to being simply part of our perspective on the world. The actual content of rational justification depends on what emerges from the attempt to be self-critical and what we discover cannot be reconstrued as relative or subjective. We cannot expect these matters to be settled permanently, since it is always possible that someone will come up with a new hypothesis explaining the perspectival character of some hitherto unassailable form of reasoning--or, on the contrary, that someone will come up with a credible way to extend reasoning to a new domain, like aesthetics.
One of the most radically austere ways of making the division is Kant's: On the subjective side is the bulk of our
forms of reasoning, applicable only to the world as it appears to us; on the other side is the pure idea of the Ding an sich, about whose objective nature our reason can tell us nothing-and that includes ourselves, as we are in ourselves. 2. This is the model for all theories that the world, insofar as we can know anything about it, is our world. But even this view, which subjectivizes practically everything, preserves a nonsubjective frame in the idea that there is a way things are in themselves, and a way we are in ourselves, which together, even if we can have no knowledge of them, ultimately determine how things appear to us. I don't propose to discuss Kant's actual view; for example, I leave aside the mystery of why it should be possible to have a priori rational knowledge of the necessary properties of the phenomenal world, even if we suppose they are basically features of our own point of view, due to our nature and the conditions of the possibility of all human experience. Is it like knowing our own intentions? But the theory provides a limiting case of the division between perspectival and nonperspectival thought, with all contentful forms of reasoning falling on the perspectival side, and nothing but a pure idea of the way things are in themselves, of which we know nothing, on the nonperspectival side.
This seems to me to be too minimal an objective frame even to support the alleged phenomenal certainties of transcendental idealism. But I believe that if we separate the idea of reason from the idea that its results must carry absolute certainty, emphasizing instead its aspiration to universality, then it is possible to withhold any relativizing or subjectivizing qualification from much more of reason than Kant thinks. In spite of the abandonment of certainty and other obvious differences, the conception of the authority of human reason that I want to defend is very like that of Descartes.
I am talking about Kant's epistemology. Practical reason, he holds, tells us more.
I would explain the point of Descartes's cogito this way. 3. It reveals a limit to the kind of self-criticism that begins when one looks at oneself from the outside and considers the ways in which one's convictions might have been produced by causes which fail to justify or validate them. What is revealed in this process of progressively destructive criticism is the unavoidability of reliance on a faculty that generates and understands all the skeptical possibilities. Epistemological skepticism, like selective relativism, is not possible without implicit reliance on the capacity for rational thought: It proceeds by the rational identification of logical possibilities compatible with the evidence, between which reason does not permit us to choose. Thus the skeptic gradually reaches a conception of himself as located in a world whose relation to him he cannot penetrate. But skepticism that is the product of an argument cannot be total. In the cogito the reliance on reason is made explicit, revealing a limit to this type of doubt. The true philosophical point consists not in Descartes' conclusion that he exists (a result much more limited than he subsequently relies on), nor even in the discovery of something absolutely certain. Rather, the point is that Descartes reveals that there are some thoughts which we cannot get outside of. I think he was right-though I also think he might have upheld the principle more consistently than he did.
To get outside of ourselves at all, in the way that permits some judgments to be reclassified as mere appearances, there must be others that we think straight. Eventually this process takes us to a level of reasoning where, while it is possible to think that some of the thoughts might be mistaken, their correction can only be particular, and not a general rejection of
Though the cogito is a philosophical Rorschach test, in which everyone sees his own obsessions.
this form of thought altogether as an illusion or a set of parochial responses. Insofar as it depends on taking the external view of oneself, the discrediting of universal claims of reason as merely subjective or relative has inescapable built-in limits, since that external view does not itself admit of a still more external view, and so on ad infinitum. There are some types of thoughts that we cannot avoid simply having--that it is strictly impossible to consider merely from the outside, because they enter inevitably and directly into any process of considering ourselves from the outside, allowing us to construct the conception of a world in which, as a matter of objective fact, we and our subjective impressions are contained.
And once the existence of a single thought that we cannot get outside of is recognized, it becomes clear that the number and variety of such thoughts may be considerable. It isn't only "I exist" that keeps bouncing back at us in response to every effort to doubt it: Something similar is true of other thoughts which, even if they do not always carry the same certainty, still resist being undermined by considerations of the contingency of our makeup, the possibility of deception, and so forth. Simple logical and mathematical thoughts, for example, form part of the framework within which anything would have to be located that one might come up with to undermine or qualify them--and thoughts of the same type inevitably have to play a role in the undermining arguments themselves. There is no standpoint we can occupy from which it is possible to regard all thoughts of these kinds as mere psychological manifestations, without actually thinking some of them. Though it is less obvious, I believe something similar is true of practical reasoning, including moral reasoning: If one tries to occupy a standpoint entirely outside of it, one will fail.
Thought always leads us back to the employment of unconditional reason if we try to challenge it globally, because one can't criticize something with nothing; and one can't criticize the more fundamental with the less fundamental. Logic
cannot be displaced by anthropology. Arithmetic cannot be displaced by sociology, or by biology. Neither can ethics, in my view. I believe that once the category of thoughts that we cannot get outside of is recognized, the range of examples turns out to be quite wide.
Having the cultural influences on our arithmetical or moral convictions pointed out to us may lead us to reexamine them, but the examination must proceed by first-order arithmetical or ethical reasoning: It cannot simply leave those domains behind, substituting cultural anthropology instead. That is, we must ask whether the proposed "external" explanations make it reasonable to withdraw our assent from any of these propositions or to qualify it in some way. The same thing is true whether the external standpoint is supposed to persuade us to withdraw a first-order judgment, or to recognize its subjective character (or the subjective character of the whole domain) without changing its content. These are questions within arithmetic or ethics, questions about the arithmetical or ethical relevance of the arguments.
To take some crude but familiar examples, the only response possible to the charge that a morality of individual rights is nothing but a load of bourgeois ideology, or an instrument of male domination, or that the requirement to love your neighbor is really an expression of fear, hatred, and resentment of your neighbor, is to consider again, in light of these suggestions, whether the reasons for respecting individual rights or caring about others can be sustained, or whether they disguise something that is not a reason at all. And this is a new moral question. One cannot just exit from the domain of moral reflection: It is simply there. All one can do is to proceed with it in light of whatever new historical or psychological evidence may be offered. It's the same everywhere. Challenges to the objectivity of science can be met only by further scientific reasoning, challenges to the objectivity of history by history, and so forth.
This doesn't mean that the results are unrevisable, only that revision must proceed by a continuation of the process itself. Any subjectivist proposal must survive as an addition to our body of beliefs, in competition with those it is trying to displace: It cannot claim automatic precedence. Since it is always an attempt to view ourselves partly from outside, it will inevitably have to provide us with a reason for abandoning or reinterpreting some of the unreduced thoughts that we continue to find plausible from inside.
It is customary to make a broad distinction between the Cartesian, foundationalist approach to the justification of knowledge and the much looser, more holist approach supposedly characteristic of actual science, which dispenses with self-evident, indubitable premises. But I think that this is a superficial distinction and that the ordinary methods of science are basically Cartesian. Where they depart from Descartes is in the relaxation of the requirement of certainty: Rational principles that play a foundational role at one stage may be superseded or revised as a result of rational criticism at a later stage. But the enterprise has a fundamentally rationalistic structure: It proceeds by the operation of methods that aspire to universal validity on empirical information, and it is an effort to construct a rational picture of the world, with ourselves in it, that makes sense of these data. However holistic the process, particular empirical observations can't overthrow general principles except in light of still other and superior general principles that give the observations the necessary leverage.
The scientific project, like Descartes's, brackets or sets aside naive impressions as mere appearances until they can be reintroduced into an overall conception on a firmer foundation, and this foundation requires an analysis of how such impressions arise from our interaction with the world. So science, as Descartes saw, requires that we step outside ourselves to the extent possible; but it also has to employ reason in doing
this, and in determining what to make of the data that result; and those thoughts we do not get outside of. At each stage of our inquiry there will be thoughts which, even as we acknowledge that they are themselves in some sense part of what is going on in the world, cannot be the object of an external psychological understanding that does not also employ them. There is nothing more fundamental to the construction of human knowledge than the reasoning that goes into the generation and elimination of scientific hypotheses suggested by the available evidence.
This, in outline, is Descartes's conception of knowledge-it is developed through interaction between the two poles of subjective appearances and nonsubjective reasoning to form a credible picture of an objectively existing world. Experience by itself does not produce scientific theories, and the reasoning that does produce them cannot be regarded by us as merely a more elaborate species of subjective impressions. To think that, we would have to take up the view from still farther outside ourselves, and the construction of any such view would have to rely on some thoughts that claimed objective validity in their turn. Such major conceptual revolutions are possible, 4. but they must be based on reasoning that actually engages us. The main difference between this and the Cartesian picture, as I have said, is that there is less reliance on indubitable judgments--though the closer you can get to certainty in scientific reasoning the better, and in its mathematical aspects that aim is achievable.
Descartes's general point remains correct: We discover objective reason by discovering that we run up against certain limits when we inquire whether our beliefs, values, and so forth are subjective, culturally relative, or otherwise essentially perspectival. Certain forms of thought inevitably occur
Einstein's special theory of relativity is an example. It revealed the space and time of Newtonian physics as subjective appearances.
straight in the consideration of such hypotheses--revealing themselves to be objective in content. And if we envision the possibility of coming to regard them after all as subjective, it must mean that we imagine making them the focus of other thoughts whose validity is truly universal. The idea of reason grows out of the attempt to distinguish subjective from objective. So far as I can see, that is what inevitably happens if one tries to take subjectivist proposals seriously--tries to determine whether they can be believed rather than just uttered.
This response to subjectivism may appear to be simply question-begging. After all, if someone responded to every challenge to tea-leaf reading as a method of deciding factual or practical questions by appealing to further consultation of the tea leaves, it would be thought absurd. Why is reasoning about challenges to reason different?
The answer is that the appeal to reason is implicitly authorized by the challenge itself, so this is really a way of showing that the challenge is unintelligible. The charge of begging the question implies that there is an alternative--namely, to examine the reasons for and against the claim being challenged while suspending judgment about it. For the case of reasoning itself, however, no such alternative is available, since any considerations against the objective validity of a type of reasoning are inevitably attempts to offer reasons against it, and these must be rationally assessed. The use of reason in the response is not a gratuitous importation by the defender: It is demanded by the character of the objections offered by the challenger. In contrast, a challenge to the authority of tea leaves does not itself lead us back to the tea leaves.
We are again on Cartesian territory here: Descartes is standardly criticized for the circularity of the argument by which he defends the authority of reason through a rational argument for the existence of a nondeceiving God. But leaving aside the weakness of his actual proofs of the existence of God, the procedure of replying to challenges to reasoning
with more rational argument seems to me blameless. Any challenge mounted against reasoning would have to involve reasoning of its own, and this can only be evaluated rationally--that is, by methods that aspire to general validity. 5.
This is the inevitable consequence of treating the proposal as something we are asked to think about; and what is the alternative? Those who challenge the rationalist position by arguing that what it appeals to at every stage are really contingent and perhaps local intuitions, practices, or conventions may attempt to apply this analysis all the way down the line, wherever a challenge to reason is met by further reasoning. But I do not see how they can terminate the process with a challenge that does not itself invite rational assessment.
Such a structure shows itself constantly in the actual procedures of our thought--in its phenomenology, so to speak. The question is, what does it mean? What should our attitude toward it be? How can we reconcile it with the recognition that we are biological specimens, fallible creatures subject to a great many influences that we may not understand, and formed by causes over many of which we have no control? If we look at how people actually think, we find that the claim to objective content is pervasive. It is found even in aesthetic judgment, which (though it is not a form of reason because it does not follow general principles 6. ) cannot as a whole be displaced by sociological, psychological, historical, or economic explanation of its sources. But I shall concentrate on reasoning, logical, empirical, and practical. It cannot, I believe, be regarded as merely a psychological or social phenomenon-because that would mean trying to get outside it in a way we cannot do. The question is, how can one regard it otherwise?
See Bernard Williams, Descartes: The Project of Pure Inquiry (Penguin, 1978), pp. 206-7. If anything, I think Descartes gives the challenges too much credit; see the further discussion in chapter 3.
See Mary Mothersill, Beauty Restored ( Oxford University Press, 1984).
What kind of self-understanding would make our capacity to think comprehensible?
I believe there is no informative general answer to this question, because the authority of the most fundamental kinds of thought reveals itself only from inside each of them and cannot be underwritten by a theory of the thinker. The primacy of self-understanding is precisely what has to be resisted.
There is a kind of inequality in disputes about the unqualified authority of reason: Its attackers, in one version, content themselves with saying the same simple thing over and over, while its defenders have to resist with something different and more complicated for each type of application of reason. The resistance must be piecemeal even if there is also a general argument that not everything can be subjective; for that argument does not tell us where in the cognitive network universal validity or rational authority is to be found. (Indeed, the boundary may shift as a result of reasoning.) We have to discover the answer by seeing which kinds of substantive judgments overpower a perspectival interpretation of themselves. There is no alternative to considering the alternatives and judging their relative merits.
The subjectivist's all-purpose comment, applicable to anything we say or do, including any procedure of justification and criticism, is that it is ultimately the manifestation of contingent dispositions for which there is no further justification. Justification proceeds only within the practices which those dispositions support--practices that reflect the common forms of life of our culture or our species, but nothing more universal than that. This argument, whatever it may be worth, can be made about anything. It is always possible to say, after the final justification has been given, "But that is only something that satisfies you, something you say with the conviction
that it requires no further justification: and all that you say is merely a manifestation of the contingencies of your personal, social, and biological makeup. The end of the line is not the content of your reasoning but rather the fact that for you, justifications come to an end here; and that is a natural fact."
The reply to this cannot be equally general, since sometimes the claim to a more universal objectivity can be shown to be spurious, through an alternative explanation of the process. The defender of reason must therefore mount his defense in each domain of thought separately, by trying to show, from within a form of reasoning, that its methods are inescapable and that first-order engagement with them resists displacement by an explanation of the practice in other terms that do not employ those methods. The general challenge at the metalevel must be reinterpreted as a set of proposals about the subjectivity of particular forms of ostensible reasoning, so that it can be met by multiple particular responses at the ground level. Those responses must show, for the case of mathematics, or ethics, or natural science, that the methods internal to that form of inquiry have an authority that is essentially inexhaustible, so that their results cannot be bracketed or relativized in the way proposed. It must be shown that we cannot have the subjective without the objective in this case.
This means that a sufficiently facile and persistent critic of the claims of universal reason has an easier time than their defender. The former can just say the same thing again and again; but the only way to defend the objectivity of ethics, for example, is by ethical argument at the ground level--by showing that it is impossible to get entirely behind it or outside of it. To turn the tables on the subjectivist one must take his proposal seriously, not as an empty formula that can be applied to anything but as a specific claim about the area of thought whose unrelativized authority is being challenged. Only in that way can the clash between the inner substance of the
thoughts and the relativizing external view of them be brought into the open.
Here are two flagrant examples of the interpretation of reason as consensus, both of them from philosophers. I admit they are easy targets, but the view expressed is very common. Sabina Lovibond refers to
our lack of access to any distinction between those of our beliefs which are actually true, and those which are merely held true by us. No such distinction can survive our conscious recognition that some human authority has to decide the claim of any proposition to be regarded as true--and, accordingly, that the objective validity of an assertion or an argument is always at the same time something of which human beings (those human beings who call it 'objectively valid') are subjectively persuaded. 7.
And she credits Wittgenstein:
Thus Wittgenstein's conception of language incorporates a non-foundational epistemology which displays the notions of objectivity (sound judgement) and rationality (valid reasoning) as grounded in consensus--theoretical in the first instance, but ultimately practical. 8.
Richard Rorty puts the same point this way:
We cannot find a skyhook which lifts us out of mere coherence--mere agreement--to something like "correspondence with reality as it is in itself." . . . Pragmatists would like to replace the desire for objectivity--the desire to be in touch with a reality which is more than some community with which we identify ourselves--with the desire for solidarity with that community. 9.
Realism and Imagination in Ethics ( University of Minnesota Press, 1983), p. 37.
Ibid., p. 40 .
"Science as Solidarity," in Rorty Objectivity, Relativism, and Truth ( Cambridge University Press, 1991), pp. 38-9.
Such views have a self-evident air if they are not examined too closely, which may account for their greater popularity outside philosophy than in it. But if one takes them seriously, they turn out to be inconsistent with the very consensus on which they propose to "ground" objectivity. What human beings who form scientific or mathematical beliefs agree on is that these things are true, full stop, and would be true whether we agreed on them or not--and furthermore that what makes that true is not just that we agree to say it! The only way to deal with such a general subjectivist slogan is to convert it into a specific, substantive claim about arithmetic, or physics, or whatever, and see how it holds up. I believe it will usually turn out to be inconsistent with the content of statements within the discourse under review, and considerably less credible than they are, in a direct contest.
The standard response of a subjectivist to such arguments is that he is not saying anything that conflicts with the content of ordinary mathematical, or scientific, or ethical judgments and arguments. Rather, he is simply explaining how they really work. Here is another passage from Rorty (I am not making this up):
What people like Kuhn, Derrida and I believe is that it is pointless to ask whether there really are mountains or whether it is merely convenient for us to talk about mountains.
We also think it is pointless to ask, for example, whether neutrinos are real entities or merely useful heuristic fictions. This is the sort of thing we mean by saying that it is pointless to ask whether reality is independent of our ways of talking about it. Given that it pays to talk about mountains, as it certainly does, one of the obvious truths about mountains is that they were here before we talked about them. If you do not believe that, you probably do not know how to play the usual language-games which employ the word "mountain." But the utility of those language-games has nothing to do with the question of whether Reality as It
Is In Itself, apart from the way it is handy for human beings to describe it, has mountains in it.10
But he can't escape so easily. The claim that there is nothing more to objectivity than solidarity with your speech community, even if it is extended to the things your speech community says would be true whether they said so or not, directly contradicts the categorical statements it purports to be about--that there are infinitely many prime numbers, that racial discrimination is unjust, that water is a compound, that Napoleon was less than six feet tall.The contradiction comes from adding a qualification that is incompatible with the unqualified nature of the original. The strongest objection to these ideas is the most obvious. The subjectivist may insist that he is not denying any of the following commonplaces, but he cannot really give a sensible account of them:
(1) There are many truths about the world that we will never know and have no way of finding out.
(2) Some of our beliefs are false and will never be discovered to be so.
(3) If a belief is true, it would be true even if no one believed it.
Simply to say that such statements are part of the "languagegame" with which we seek solidarity does not render them intelligible. It is as if someone said "There is nothing more to wrongness than being contrary to the laws of my community," and then added, "Of course, the laws of my community specify that not everything that is wrong is illegal."
These forms of subjectivism are radical positive claims, and not, as their proponents represent them, merely the rejection of metaphysical excesses. To take such a claim seri-
"Does Academic Freedom Have Philosophical Presuppositions?" Academe, November-December 1994, pp. 56-7.
ously, one has to try to interpret it as a genuine alternative-something we are being asked to believe about our relation to the world--and then it will inevitably engage the mechanisms of rational assessment. Such a proposal cannot be exempted from the requirements of intelligibility and credibility: It is a statement after all, and supposed to be true. How else can we decide whether to accept it but by thinking about it? In most cases we will then conclude that reason and objectivity are not grounded in consensus, but on the contrary, that where consensus is available, it arises from the convergence among different individuals, all reasoning to get at the truth. There is a consensus on the nondenumerability of the real numbers because the demonstration of it is conclusive, and not vice versa.