or
LET'S ALL GO BACK TO BEING MODERN
© Ronald
de Sousa
Department of Philosophy
University of Toronto
Toronto M5S 1A1
E-mail: sousa@chass.utoronto.ca
One might think, upon holding one's finger to the philosophical winds, that Foundationalism is a dead issue. For most of us, it seems, it is no more a live issue than the existence of God, the validity of astrology, or consulting the entrails of sacrificial birds. However, multiculturalism being what it is, it's also true that for some people God, astrology, and perhaps even the consultation of birds' entrails are a live issue. And if something is a live issue for anyone, or at least anyone we respect, then that may automatically make it a live issue for the rest of us.
Indeed, philosophy is Lazarus land. No philosophical issue is ever dead for long. I think this follows, as I shall explain in a moment, from the very definition of philosophy, if such a thing can be constructed. So perhaps it is time for that dead issue to be resurrected. I may not go quite so far as to defend foundationalism, but I shall try to give you my own personal and perhaps excessively idiosyncratic view of what some discussions of this topic might be about.
The view I want to recommend was formulated essentially in response to the problem of addressing an apparently dead issue. In fact, to put it in a childishly simple way, the view I want to recommend is that the views that any philosopher formulates are always essentially a response to some question that presents itself as coming from the outside. Any view is a live option, so long as someone succeeds in raising a question about it in their own or someone else's mind. Just as belief in God is a projection of our infantile need for a perfect parent, so our search for foundations is an anxious, idealized quasi-Freudian projection of our need to be given a starting point before we can find anything to say. Call this `topic foundationalism,' if you like, or relativistic foundationalism. It's not that we need an indubitable basis for knowledge, as promised by most classical rationalist or empiricist philosophy, founding it either on reason or on phenomena of the sense. It's just that to get anywhere, we need to start somewhere. We need starting points, but those starting points needn't be here rather than there.
Apart from the dismaying obviousness of this idea, two things are likely to strike you about it. The first is that what I'm proposing is not foundationalism at all: relativistic foundationalism is an oxymoron if not a downright contradiction. You might say I'm trying to keep the bath water after throwing out the dead baby. The second is that what I am offering you is just a piece of psychologizing, and psychologizing of any sort, in matters of epistemology or metaphysics, is just a silly sophomoric mistake. From Philosophy 100 we've all been taught to avoid it, together with the naturalistic fallacy, evolutionary epistemology, and indeed all other avatars of naturalism. Well, I propose to flout both of these plausible sounding objections. I hope to persuade you that if you can bend a little on that second principle, and accompany me into a bit of psychologizing, then what I am proposing will start to look a bit more like foundationalism after all, or at least like the insight that underlies the quest for the holy grail of foundationalism.
The General Form of Foundationalism.
But before I distort the subject in my own way, let's look at the classic problem of foundations.
What, to begin with, are foundations supposed to be foundations to? There have been at least 4 different lines of thought:
1. Ontological or causal foundationalism: the search for first cause.
2. Epistemological foundationalism: the search for beliefs that can serve as the justification of other beliefs, and which themselves require no further justification.
3. Explanatory foundationalism: the search for propositions which bear the relation of explanans to other propositions, but the relation of explanandum to none.
4. Ethical foundationalism: the search for the ground of all value, for something that can be valued for its own sake without qualification, and for the sake of which everything else can be valued.
Cutting across all these and distinct from them is
5. Psychological foundationalism, which might be variously defined, but which could originally be characterized in terms of Aristotle's distinction between what is "more knowable to us" and what is "more knowable by nature" (An Post 71b33). This is most easily understood in terms of the two requirements on explanation. An explanation has to be satisfying subjectively and objectively. The subjective requirement reflects the fact that an explanation will only be judged fully successful if it eases a particular person's puzzlement about something they don't understand. If the explanans is itself found unintelligible, the attempted explanation has failed. But it might be true all the same that the explanans bears a genuinely explanatory relation to the explanandum: that is the objective requirement. These requirements are independent. Some people seem to be "satisfied" that they understand something on the basis of some astrological "fact": here the subjective requirement is satisfied, but the objective is not. On the other hand, if someone explains some phenomenon to me in terms of quantum mechanics, the explanation may well satisfy the objective requirement, but owing to my ignorance it will not satisfy the subjective one.
I think it's very likely that if I subjectively 'understand' an explanation about physics, the chances are all the greater that the explanation isn't really a good one. Conversely, the better an explanation is objectively the less likely I am likely to understand it. The best explanations for all sorts of physical phenomena may well be in terms that I cannot understand. In this sense, the two requirements are actually in tension. This poses a prima facie problem for any view that introduces psychological considerations into the question of foundationalism. I will later take up the challenge of explaining how to get around this objection.
For the moment, I want to point out that the four foundationalist lines of thought I have named are not equivalent, but share a crucial formal feature.
To see that the four are not equivalent, consider how foundationalism of each type might be compatible with the falsity of foundationalism in the other categories.
First, suppose that there is a first cause. It wouldn't follow from the mere fact of its existence that it could serve as the basis for an unexplained explainer of everything else, for we might still require an explanation of the fact that such a first cause had just those effects and not others. To account for that, we would need a set of explanatory principles distinct from the mere fact of the existence of the first cause. Nor would it follow, even more obviously, that we would have foundations for either knowledge or value.
Second, suppose something like phenomenalism were true. We could then justify everything we believed by tracing it to a set of beliefs about sense data plus contentless logical principles, none of which stood in any need of justification. But that would imply nothing about the existence of a first cause, nothing about the grounds for any value, and nothing about the explicability of anything (any non basic belief would have its analytic equivalent among the set of truth functions of basic beliefs, but any two beliefs, whether basic or not, might fail to be linked by any explanatory relation.)
The third case seems most obvious. If ethical foundationalism were true, nothing in particular would follow about ontological, epistemological, or explanatory foundationalism. [1]
As for the fourth, suppose we find that a unified field theory of some sort underlies every other scientific explanation. That would obviously be compatible with the nonexistence of a first cause (we might be unable to prove that the universe is not infinite in time, or simply uncaused). Nor would knowledge of such a theory be more obvious than anything else for ordinary mortals. That ultimate theory would be "first in the order of nature," but not "first in the order of knowledge." Nor, finally, would this have any bearing on whether our experience, or anything else, could ever provide us with an indubitable ground of value. Quite generally, the existence of a certain explanatory relation between two propositions says nothing necessarily about whether one could be known without the other, or even whether one could be true without the other.
Nevertheless, the different kinds of foundationalism all have a certain abstract form in common, a form which, captures the temptation of foundationalism, and yet makes it extraordinarily hard to see why anyone would ever have seriously believed it.
To get a common form, we need to abstract both from the type of inclusion into the privileged set appropriate to each category, and from the specific nature of the relation between the foundational statements and those grounded upon them. Call all types of inclusion into the privileged set acceptance, and call the different relations in question relations of grounding. Then the line of reasoning (first articulated by Aristotle) goes as follows:
Acceptance of any item requires grounding.
But if each statement must be grounded, there seemed to Aristotle (and many others) to be only three possibilities:
(a) there is an infinite chain of grounding
(b) the chain of grounding is circular
or (c) the chain of grounding stops in absolutely grounding or self-grounded items.[2] Foundationalists have been persuaded that only (c) is an acceptable alternative. But what is curious about this piece of reasoning is that although it is plausible, it is also a purely abstract existence proof, as it were: much of the history of philosophy after Descartes has consisted in the systematic debunking of one particular claim after another on behalf of different putative foundations. In the end, no actual candidate for the status of first principle has the kind of plausibility that one would want of something that is supposed to ground our acceptance of everything else. So if we find Aristotle's dilemma compelling, we automatically face another: should we believe that there must be a foundation, on the basis of the abstract argument, or should we believe that there isn't, on the basis of the absence of any plausible candidate?
But in fact there is worse. The actual form of argument for foundationalism seems to be inherently contradictory. It goes as follows:
6. SO, I'll accept something I can't prove.
6. SO, something exists without a cause, NAMELY GOD.
Zeno's Paradoxes and the Multivalence of Argument.
In fact "foundations" in the history of philosophy always have a sort of "emperor's new clothes" quality. No one who is not already corrupted by philosophy could really believe some of the rubbish that we expect, or used to expect, our students to think they didn't believe because they weren't smart enough. This is clearly the case with the cosmological argument: philosophers have spent enormous amounts of energy making it more subtle, which is another way of saying: making a simple argument complicated enough to hide the manifest contradiction. Other absurdities of this ilk include: it's good to do A iff it would be good for everyone always to do A; the only absolutely good thing is a pure will; everything that is distinguishable is distinct, etc.
In all these and many other cases of philosophical dogma, our first reaction was probably incredulity, and only subtle arguments persuaded us that our incredulity was merely a sign of insufficient understanding. So we, or if not we, then generations of other philosophers persuaded themselves that these things were true, in order to be able to claim that we understood them. But in all these cases the initial reaction is probably the right one. In support of it, one can advert to a wholly general feature of the use of reason. This general feature is the multivalence of argument. It can most brutally be put thus: it is that no argument ever compels [3]--no argument, in fact, ever proves anything in particular. The reason is that any argument merely gives us a set of alternatives: believe the conclusion together with the premises, or continue to reject the conclusion, but then also reject one or more premises. And in this situation, what is the most reasonable thing to do? The most reasonable thing to do is, surely, to believe the least incredible alternative.
Zeno's paradoxes provide an especially good illustration of this. Their lasting appeal comes from the fact that they appear to give us reason to question something -- the existence of motion -- than which little if anything is more obvious. A common but misleadingly unguarded description of how they do this is that they derive a contradiction from the existence of motion. But actually they don't do that: they purport to derive a contradiction from the existence of motion interpreted as committing us to certain other premises. Here, to illustrate what I mean, is one reconstruction of the argument about the racecourse, in the form of a reductio ad absurdum:[4]
1. [Suppose] there is motion from A to B in a finite stretch of time.
2. If anything moves from A to B, it must first pass through an intermediate point C.
3. But to get from C to B, it must still pass through an intermediate point D.
4. But 3 is applicable an infinite number of times.
5. Therefore, (LEMMA) to get from A to B, a moving object must traverse an infinite number of finite stretches.
6. To traverse a finite stretch must take at least a finite length of time.
7. THEREFORE to traverse an infinite number of finite stretches must take an infinite amount of time.
8. THEREFORE motion from A to B must take both a finite and an infinite amount of time.
9. 8 is a contradiction.
10. THEREFORE 1 is false.
In response to this argument, there are essentially three possible attitudes. The one Zeno reputedly intended us to take is to accept the reductio as it is, and reject the reality of motion. That is philosophical heroism; there have been other heroes in the history of philosophy, including Berkeley, who in the name of common sense rejected the common-sense belief in matter, and Bradley, who following the logic of an abscure argument about relational properties concluded that time is unreal. But my favourite is the possibly apocryphal story of Zeno the sceptic, who having decided that his senses were not able to provide him with any reliable information, walked most philosophically to his death by falling over a cliff, on the ground that there seeming to be a cliff in front of him did not make its actual existence more likely than not. But the Eleatics, Parmenides and Zeno, are certainly the first and among the greatest of philosophical heroes, rejecting on the basis of mere arguments what to everyone else are among the most basic and obvious truths of life.
Now it is a fine thing to be a hero, but as Falstaff observed in a different context, it is not necessarily always the most prudent policy. More precisely, it is sometimes positively irrational to follow "reason." (In my idiolect, it can't be irrational to follow reason. It can only be irrational to follow "reason", or "irrational" to follow reason. Other idiolects treat these issues differently: I shall defend my own usage later.) What then are the alternatives in the context of Zeno's argument, and why are they preferable?
Note that this argument, in ten steps, is unlikely to be the most elegant formulation possible of Zeno's thought. Nevertheless, it contains several indispensable premises. And the confidence trick here, is that some of the premises are clearly less obvious than the falsity of the conclusion. There are two ways of resisting the confidence trick: the know-nothing way, and the intellectually responsible way.
The know-nothing way actually has a naive and a sophisticated version. The naive simply consists in saying: I don't really follow that philosophical gobbledygook, but I know that its conclusion can't be right, because obviously there is motion, and so any argument that says it isn't has to be wrong. The more high-faluting way of saying the same thing is the claim that reason has its limits, and that beyond the limits of reason one must seek one's way by the light of faith. The most frequent use of this ploy is in defense of religious conclusions.
I am highly suspicious of the sophisticated route. Its employment should be rendered suspect right from the first, merely by the observation that no one ever appeals to the bankruptcy of reason when they have won the argument.
On the other hand, the naive rejection of obviously false conclusions of unrefuted arguments can actually be quite sensible, from a pragmatic point of view, taking a game theoretical perspective in which the costs and benefits of changing one's beliefs are taken into account. For the cost of looking for a refutation may not be outweighed by the risk of believing something false, even reckoned in terms of purely epistemic value.
The intellectually responsible response consists in accepting the challenge posed by the argument. This means, first, recognizing that intellectual disquiet must be resolved in one of three ways, and then setting out to find the way in which we can actually resolve it.
a) by accepting the unpalatable conclusion;
b) by finding a reason to reject one of the premises or
c) by the discovery of a fallacy in the form of the argument.
In the case of Zeno's argument, there is a premise that looks true, and is very close to something that really is true, but which in itself we have no reason to think either true or false independently of the argument at hand. That very argument, then, constitutes a sufficient reason to reject that premise. The premise in question is actually an unexpressed premise that lurks in the passage from 6 to 7. It says that
6' the sum of an infinite number of finite stretches is infinite.
The truth that it resembles is this:
6'') The sum of an infinite number of finite stretches of at least some definite dimension d is infinite.
Once we see the difference between 6' and 6'', we see that 6'' is so obvious that it would be hard to reject it, but 6' is dubious enough that the present argument is quite sufficient as a reason to reject it. Once we have done this, we have done the "intellectually respectable thing".
But there is actually one more possibility. Call it the "incommensurable" response. Our students are often tempted by it, and so, in a way, was Aristotle. It consists in constructing a new description of the situation, which simply does not address the earlier ways of seeing things, but which doesn't give rise to the problem any more. Thus Aristotle doesn't actually answer Zeno's argument. Instead, he says: "that's no problem, because time is also infinitely divisible in exactly the same sense, and so to every stretch of space there corresponds a stretch of space."[5] But actually that's no good, because it ignores the argument we have before us. Zeno's argument might easily be slightly reshaped to show that time can't exist, because to get to another moment of time, you need to traverse an infinite number of finite stretches of time.
So the incommensurable response is no good here. Still, in the history of science it sometimes seems to be just the right response. The most spectacular example is Newton's rejection of the Aristotelian problem of what keeps the arrow in flight. By saying: no explanation is needed for the arrow's remaining in flight, but only for the arrow's not remaining in flight, Newton effectively deals the the problem objective motion in an entirely coherent way, but we might say he fails to meet what I have called "intellectual responsibility". For it doesn't actually fulfil the Aristotelian demand to show "not only what is true, but also the explanation of why we originally got it wrong." [6]Yet surely, from the point of view of pragmatics or game theory, it is sometimes legitimate to do just this.
Here I want to point out something that is intended to soften you up for the part of this paper where I stick my neck out: in those cases where we fail to meet the strict requirement of intellectual responsibility by agreeing to incommensurability,[7] if we were to satisfy the Aristotelian demand, it would be precisely by showing how the belief originated in a certain way, that is, by going, roughly speaking, to the psychological level of explanation. We shall see later the significance of this fact. For the moment, I want to turn to another consequence of the kind of analysis I have just given of Zeno's argument: this consequence is that what I have called the "game-theoretical perspective" has introduced a subtle change of perspective. The kind of foundations we were formerly looking for were in some sense objective; they were independent of any particular person's worries, depending rather on being able to take on some sort of idealized sceptic. But in the present perspective, what a foundationalist needs to be able to do is to refute anyone who finds the denial of the putative foundations more plausible than their acceptance.
If that is the task before the foundationalist, they may reasonably despair. For the history of philosophy suggests that no substantive principle can be sustained in the face of a sceptical attack. (Descartes's sceptical doubt, notoriously, ought not to have been allayed by the considerations that apparently satisfy him.)
Fundamental Value Choices and the Uselessness of Argument
Another foundational argument is in the realm of value: it is what we might call the East-West dispute. Here the issue will be rather different, in that we are not tempted to reject the conclusion of an apparently good argument. Rather, we can think of no argument at all for a conclusion that is completely obvious to us, while its contradictory is equally obvious to something else.
A certain view of a certain Eastern philosophy, especially Buddhism, has it that what is wrong with life and must above all be corrected is desire. If we get rid of desire, we will get rid of suffering. Therefore we must get rid of desire. (This view is also to be found sometimes in Plato, who represents certain pleasures as necessarily linked to pain, and likens their satisfaction of certain desires on the mode of the futile filling of a leaky jar.).[8]
Against this is the view that at least some desires are worth the concomitant unhappiness.
What are we to say of this dispute? It is a little like the question: would you rather be immortal?[9] There really doesn't seem to be anything at all that can be brought into a reasonable argument on either side. Comparative methods such as the Plato-Mill idea of entrusting the judgment to a "competent judge" who has experience of both forms of life[10] are inapplicable when the choice is that fundamental, because to experience both desire and the absence of desire is a simple contradiction. The two sides are simply not juxtaposable. No one could possibly be a competent judge. Each person plumps for one or the other of these alternatives according to preconceptions that are at worst cultural and at best temperamental.[11]
So again the choice, which in a sense we all make, must be made "on faith", though again this is not the sense of faith which arises when it is opposed by reason, but rather it is the sense of faith that arises when reason can have no rational grip whatever.
Can we ask no question at all about it then?
Yes, we can. Once again, we can ask the (broadly speaking) psychological question: what causes one person to make this rather than that temperamental choice? Or, if the answer is a cultural one, what accounts for the force of this particular cultural force, what accounts for the staying power of this particular "meme."[12]
Going Procedural
In a moment I want to develop this idea of a psychological exploration of our deepest assumptions. But first, let me briefly return to note to Zeno's argument. One response which I listed but didn't discuss was to attack the form of the argument. It didn't seem promising, simply because the validity of a simple argument is usually too obvious to question. But that in itself suggests that we are quite confident of what good or bad arguments consist in. Could we then retreat to a weaker form of foundationalism, procedural foundationalism?
I mean by this any doctrine that assigns a foundational role to forms of inference, tautologies, or transcendental arguments: in short, to logic broadly conceived. This is the rationalist project. Spinoza's method of demonstrating his metaphysical system "more geometrico" exemplifies it. If all of science and metaphysics can be deduced from principles that are not empirical but a priori, then we won't have to have recourse to disputes about experience. In ethics, at least part of the Kantian project seems also to be of this nature, understood as an attempt to avoid substantive assumptions and base ethics entirely on a principle of rationality.
Unfortunately the rationalist project is plagued by the problem sketched above in connection with the duality of explanation. Either the particular foundations we are asked to admit -- the basic logical or quasi-logical principles -- are alleged to be subjectively obvious, or they are alleged to be objectively obvious. In the first case, there is no problem if they as obvious to us as they are to their proponents. But as soon as anyone challenges them, they entirely lose their charm, for their charm is precisely their obviousness. So in those cases one resorts to saying: if you deny that, the law of contradiction, for example, then you must be just crazy. Your denial just doesn't count, because the principle is in some sense objectively obvious and if you don't see it that's due to your failure of understanding, not to any failure in the principle itself. But this is little better than intellectual fascism: to adopt this objectivist view is to fail of the subjectivist criterion, and the subjective doubt that results simply ricochets back to cast doubt -- real, felt doubt -- on the legitimacy of the objective claim.
The moral seems to be that epistemology can't afford to ignore psychological realism. But this does not necessarily invalidate the procedural account altogether. Indeed, if there is any hope of rehabilitating some form of foundationalism, I believe it lies in this "procedural" account. To explain this, I want to take a somewhat roundabout way through some remarks about the nature of philosophy.
Defining Philosophy.
It's worth thinking about little about the paradox that philosophy, the most disputations and uncertain of disciplines, has taken it upon itself to discover the basic certainties that ought to guide us in every domain. Philosophy has perhaps not worried enough about this curious fact of intellectual life, presumably because it has armed itself with the distinction between the subjective and the objective. But I believe the paradox actually stems more deeply from the very nature of philosophy.
Here, then, is the characterization of philosophy which, throwing caution to the winds, I am offering for your consideration:
1. Philosophy differs from science in two respects, resembles it in one:
i. Unlike science, philosophy is concerned not with specific knowledge but with overall vision.
ii.Unlike science, philosophy does not have any definitive methods dictated by the nature of any intelligible question. In any field of science, understanding the question generally suffices to be able to judge whether any particular proposal will qualify as an answer, or whether any proposed method is likely to yield an answer. In philosophy this is not so. That is what, apart from the subjective-objective ambiguity already noted, accounts for the fact that philosophical questions are never quite put to rest: since there are no universally accepted methods, any given proposal about how to solve a problem can be challenged.
iii But philosophy, like Science, is committed to rationality: to arriving at its conclusions via argument.
Philosophy's relation to religion is the mirror image of its relation to science. Like religion, it is concerned with overall vision rather than with specific knowledge, and it has no definitive methods dictated by the nature of any intelligible question. Unlike religion, however, philosophy systematically repudiates tradition, faith, and revelation. For philosophy, as for science, nothing is sacred. In fact, it might not be too bizarre to claim that nothing is sacred to philosophy, except the principle that nothing is sacred.
This is not to say that all those who call themselves philosophers would be happy with this characterization. On the contrary, whether we should have any concept of the sacred, or whether it merely remains a remnant of a primitive mentality, is an issue that profoundly divides philosophers. But note that it divides them according to philosophical temperament, and is therefore subject to inquiry not at the level of argument but at the deepest psycho-philosophical level.
Assuming this characterization of philosophy does make sense, however, we can understand the "post-modernist" rejection of reason itself -- the position that Larry Laudan characterizes as "relativist".[13] It is a position that effectively uses philosophy's self characterization against itself.
But now if rationality is the one sacred thing in philosophy, we can construct an antinomy.
Thesis:
Philosophy's claim to the rejection of the sacred is phony. For its reliance on reason is itself a case of holding something sacred.[14]
Antithesis:
The second defining feature of philosophy -- the absence of any absolute and definitive method for the resolution of any of its questions -- allows us to reject reason, as far as that goes, without betraying the ideals of philosophy.
The principle that nothing is sacred is being followed even as it is being betrayed, since the rejection of reason amounts to rejecting the one thing that philosophy might be suspected of holding sacred. Besides, it doesn't have to reject reason in favour of any other putatively sacred ideal. The moral -- the transcendental synthesis, if you will -- seems to be that philosophy's claim to procedural rationality can survive its own critique.
These considerations are rather trite. But their point is just to emphasizse that it is not so easy as it might at first appear to reject foundationalism altogether.
In fact, they entail a rather startling conclusion, which is that even the post-modernist rejection of reason upholds the foundational thesis of philosophy. So insofar as what motivates that rejection is a desire to be the baddest boys and girls in philosophy, it won't work.
The Big Greek Idea
What I have called "procedural foundationalism" contains a further idea, which has been much questioned of late. This is the idea of universality as a defining ideal of reason and philosophy. A certain ideal of universality lies, I believe, at the heart of the "modern" revolution effected by the Greeks. I would like, within the limitations of my "relativistic foundationalism", to say a word in favour of this idea.
Before the presocratics, a cosmogony was just a kind of soap opera. Why did this or that happen? Well, Cronos was told he would be overthrown by one of his children, so he ate them all at birth, but Rea hid Zeus, and gave him a stone to eat instead, etc. So, you see, that's why it all happened.... The modern conception of explanation started with the Presocratics: it replaces soap opera as explanation with a demand for universal principles of explanation that can be generally and uniformly applied. For a modern mind, an explanation has to make it necessarily or at least plausible that something happened, because under the circumstances detailed in the explanans it would happen this way at any time in any place. If that's not the case, then we are missing some parts of the explanation, just those parts that make this place special. But what makes it special has to be more than the mere fact that it is this place and time.
Now of course that requirement of universality comprises dangers. In his recent book Cosmopolis, Stephen Toulmin remarks that "one aim of 17th-century philosophers was to frame all their questions in terms that rendered them independent of context". [15] He complains that this "modern" ambition led them to disclaim "any serious interest in four different kinds of practical knowledge: the oral, the particular, the local, and the timely," (p. 30) and urges a "post-modern" turn away from the universality of knowledge. But while it is true, of course, that we must avoid the oversimplifications that sometimes pass for "rationality," I don't see how we can do away with the ideal of universality in science. In this, my own instinct is resolutely modern. I think I can understand the suggestion that knowledge, or at least acquaintance, is of particulars; but understanding is essentially general.
The claim that reason is essentially universalizing should not be confused with the claim that we can always arrive at universally valid conclusions. That would be a mistake very similar to confusing determinism in natural law with predictability in practice. Even if the world were totally deterministic, chaotic effects, which in essence involve enormous disproportion between causes and effects, would prevent us from making the calculations required for accurate prediction. Similarly, the universality of reason must not deprive us of our humility in the face of the diversity of concrete situations. To say that reason is by nature universal is quite compatible with the admission that we can probably never hope to have a complete and correct explanation or justification of any particular thing at all.
Back to Modernism: A Modest Foundationalism.
Human reason is the faculty of self-transcendance: it is philosophy, in the very precise sense that it recognizes no sacred rule or principle, other than the principle that nothing is sacred. Contrary to the view of anthropologists who see civilization as bred in taboos and self-justifying rituals, I see the philosophical ideal of civilization as moved by what we might well still like to call the Natural Light of Reason. That is the methodological key to civilization. We climb upwards by that light, or at least that is how it seems: perhaps from some outside point of view our motion would resemble not a spiral but one of those Escher pictures where circular stairs seem perpetually to be climbing. And though it seems to me virtually impossible to slip down from a more inclusive to a lower level of awareness, I know that people can, alas, decline from awareness into self-deception, delusion, even fanaticism. The conditions for such "climbings down" are all too material. Torture will surely do it to all but a few, and so may repression, extreme stress, and isolation: witness Salman Rushdie's recent and all too futile "conversion" to Islam. But conceivably there might also be such a thing as intellectual corruption: the seduction of facile faulty inference to forms of scepticism that seem to lift the weight of intellectual responsibility. Such, in my arrogant opinion, is the influence of much philosophy that calls itself "deconstructionist" or "post-modern", which seems to infer from difficulty of making valid inferences to its impossibility, from the narrowness of some thinkers's conception of reason to the bankruptcy of the notion of reason itself, and, in a kind of rhetorical version of the doctrine that might makes right, from the success of their own rhetoric in English departments filled with miseducated philosophy enviers, to its legitimacy as a tool of thought.
When we use reason not as a tool of rhetoric, not as a mere means for the persuasion of others, but as a genuine method for the sorting out of what to believe, we also encounter that other "modern" idea, the idea of freedom. The idea of freedom confronts us in a paradoxical way. What I have called the "self-transcendance" of reason has revealed to us the fact that what seems most obvious to us at any particular moment might not actually be true. The paradox consists in this, that in trying to judge what is true we cannot avoid trusting the obviousness of something: we have to treat something, at least for the moment, as quasi-foundational.
We may decide to trust only what seems evident to our own natural light of reason; we may decide to trust Science, even if what science tells us is quite literally unintelligible to us (surely most of us are reduced to this humiliating stance whenever topics like Bell's inequality or the non-locality of quantum effects are mentioned). Or we may even decide that Big Brother Knows Best, in one form or another, whether Big Brother be represented by the Pope, or the Ayatollah, or Willard van Orman Quine, or by Tradition, that annoying mix of time tested rules of thumb and pseudo-reasons for doing things for which no good reason exists. But in every case, when confronting an argument, we have to make a decision about the relative obviousness of the alternatives open to us at that moment. We are, as Sartre memorably put it, forced to be free.
If self-examination is at the heart of philosophy, then philosophy can be conceived as as the search for the deep structure of our thought. 'Deep', that is, in the sense of buried, important, and, yes, philosophically fundamental. That is why, for want of a better word, I have called this view relativistic foundationalism: it addresses not so much the foundations of knowledge, value, or science, as the foundational strategies of our inquiries.
I began by pointing out the Janus quality of our demands for explanation: we want to understand things "in relation to us", subjectively, but we also want those explanations to be valid "in relation to nature itself", objectively. The view I have sketched is meant to articulate the way that the two quests can actually turn out to be one and the same. By looking further and further for the roots of my own convictions, I am, in a way, delving into my own subjectivity. But I am doing so in ways that take me far away from the original content of my own awareness, from the conscious nature of my justificatory thoughts. And so, by exploring the mechanism of my own beliefs, I can say I am becoming more objective. By doing psychology in this extended sense, I am escaping from the limitations of my own psychological make-up, just as scientists escape the limitations of their senses and intelligence by using instruments devised by human intelligence with the aid of human senses. This is an old idea, but it is a modern idea: the idea of reason as self-transcending.
(Here is a practical corollary for students of the history of philosophy: look for bad arguments in great philosophers. A bad argument is essentially one that has a false lemma: the lemma, and its falsity, is hidden from the philosopher propounding it. In a great philosopher, the worse an argument is, the more it will tell us about what that philosopher is really fundamentally thinking.)
The type of foundationalism I am advocating is "modern", as opposed to post-modern, in its espousal of an universalizing reason. But it is not, to be sure, a position that the Greeks, or for that matter the classical empiricists and rationalists, would have found comfortable. For it recognizes no substantive basic propositions. In effect, I've gone along with the insight that rationality is essentially procedural: we look for a direction of thought, not for a terminus of thought. On the other hand, the idea that philosophy is a kind of deep psychology is perhaps implicit in Hume; for as I read his "solution" to the problem of induction, it consists in saying: don't look for a justification of induction: look only for the actual laws that govern the mind's activity in inductive reasoning. That search can take us into different directions: into considerations about the cognitive mechanisms of the brain, or about their evolutionary origins. But all are, in the broad sense I intend, psychological rather than strictly normative or epistemological.
Is this Dialectics?
At the end of this talk, I find myself rather surprised to be espousing something which may, in fact, bear a name with I have never tended to take very seriously: perhaps what I have oxymoronically called relativistic foundationalism is really dialectics.
Let me conclude with a bit of autobiography. I recall two incidents in which someone tried to tell me about dialectics, but I was not ready to hear it. The first was when I was a schoolboy. I had a Marxist teacher called Monsieur Weber, and I asked him once what was the sense in trying to improve the lot of our beloved proletariat, if the best possible outcome was that they would all turn into something indistinguishable from our hated bourgeoisie. At this he puffed on his pipe and mused: "Now we are for them in their struggle. When they have won, then we will be against them. The proper role of an intellectual is to be always against."
I went away puzzled. Surely, either a given state of affairs was good, and then we should work for it now and then bask in it when it comes, or else it is not good, in which case we should oppose it if it comes and not work for it now.
Much later, my then colleague Emil Fackenheim once walked into the Philosophy Faculty lounge, puffing on his cigar, and uttered a piece of vaticination which I have always fondly remembered as a hilarious philosophical joke: "The task of philosophy, he pronounced, "is to find out the limits of knowledge, and then transcend them." Well, I naively thought, if we can transcend them then they're not limits, and if we have truly found the limits of knowledge then they can't be transcended.
Today I feel, without altogether being relieved of my puzzlement, that I have finally come to advocate, if not to understand, what both Fackenheim and my high school teacher were perhaps trying to teach me: the meaning of dialectical thinking.
Thus, even if I have taught you nothing today, I can comfort myself, at least until I apply my methods of self-criticism to this potentially self-deceiving consolation, with the hope that it might turn out, in twenty or thirty years, that I have said something interesting after all.