Geum.ru

Iv российский философский конгресс

Вид материалаДоклад

Содержание


First session
Подобный материал:
1   2   3   4   5   6   7   8   9   ...   25

Философия гуманитарных наук


First session


The Philosophy of Human Sciences


Джозеф Марголис (США). Концептуальные различия в моделировании физических и гуманитарных наук


Joseph Margolis (USA). Conceptual Differences in Modeling the Physical and the Human Sciences


The paper explores the relationship between the physical and the human sciences: opposes reducing human sciences to physical sciences; opposes a strong disjunction between the two sorts of science; and favors instead treating all would-be sciences as, in a deeper sense, human sciences. This affects the very concept of what it is to be a science; proves more promising and flexible that the other more standard views; and centers on the methodological aspects of our cognitive claims in both sorts of science. Emphasis is placed on the bearing of technological advances on our treatment of the methodological question, and the entire account is linked to certain saliencies in the history of modern philosophy.

The history of the comparison between the physical and the human sciences produced such meager results by the end of the 20th century that we now find ourselves obliged to review once again all our philo­sophical options by way of certain larger intuitions that cannot claim any easy advantage from comparing the distinct achieve­ments of the two sorts of science. This means that the nagging puzzle about their relationship — already a concession to the intuitions I have in mind — arises in the very context of admitting the indisput­able accomplishments of the one and the laggard but not entirely dismal record of the other. Even this concession obscures a deeper bias in the seemingly fair assumption favored by Carl Hempel, say, and other champions of the unity of science program, to the effect that the human sciences must be physical sciences that have failed to find the right reductive niche to which the targets of their own inquiries properly belong.

This line of reasoning has yielded so little of promise that it is now entirely reasonable to turn to what might be gained by reviving the intuition that we are dealing with two very different kinds of science — conceivably so different that it may indeed not be possible to devise a common methodology or model of explanatory reason or form of understanding or even criteria of meaning and truth that, admitting their essential differences, might be instructively fitted to each. This is the line of reasoning initiated by Wilhelm Dilthey, for instance, not altogether disjunctively (but disjunc­tively enough), since, on Dilthey’s view, both sorts of science are grounded in the same way in a common pre-scientific run of basic experience (Erlebnisse), although, in all other respects, the two sorts of disci­pline are rightly seen to be as different as they are said to be. The fact is, a proper brief in favor of the distinction between the two sorts of science cannot rely on contesting the productive superiority of the physical sciences, because, of course, a proper brief presup­poses the failure of any strong reductionism on grounds more funda­mental than those that might be drawn from any such comparison, and because, on the argument needed, differences in predictive and quantificational power, systematic simplification and the like, are themselves blackmail arguments of a demonstrably questionbegging sort.

There remains a third possibility almost never seriously broached, namely, that all would-be sciences are, in a deeper sense, human sciences: so that whatever difficulties rightly beset the human sciences (in the second sense) also profoundly affect the rigor and accomplishments of the physical sciences themselves; although such complications are often quite invisible, or not sufficiently grasped by those who analyze the physical sciences primarily. I’m afraid this last worry is a valid one; though it is difficult to put it in a manageable and convincing way. The third option has the virtue, nevertheless, of tempering whatever may be salvaged from the contrast offered in the name of my second model — at the same time we refuse to disallow, a priori, analogies between the work of the physical and the human sciences that draw on the pervasive complications of the latter. In this way, still speaking with Dilthey though against his limited disjunction, the human sciences may be thought to favor some form of Erklären and Verstehen as well as their interlocking use, as Paul Ricoeur, for one, has suggested.

The larger strategies are important to keep in mind: they preserve a proper sense of our question’s history and conceptual resources. The pertinent practices of each general option are quite capable of assuming a great variety of promising forms that depart for local reasons from the spent options that have been favored up to now. In this sense, we remain entirely free to speculate along the lines of whatever comparably large intuitions we may now be able to bring to our question, without being bound to adhere to any prior canon or mere doctrinal loyalty. There simply is no inviolate essence of science: we find ourselves forced to invent all sorts of

candidate models of what a science should be like and then revise those accounts as often and as opportunistically as we must — which seems to be forever. Science and reason are not the kind of thing for which there could possibly be a discoverable essence: both are no more than idealized constructs infected by practical concerns and historically perspectived experience.

My own sense is that the first option offered about the essential nature of a science is dead in the water; that the second is too narrow and extreme (though more generous than the first) to fit the actual prac­tices and needs of the human sciences­ (arguably, to include economics and sociology and psychology as well as history and art criticism); and that the third is actually the most promis­ing of the lot, though it has been largely discounted or overlooked through the entire 20th century.

If you begin your speculations with no more than a sense of these three possibilities, the barest recovery of a few salient truths, and even fewer reasonable conjectures, you will find that you are already well ahead of the philosophical pack that once monopo­lized the discussion during the whole of the last century. For one thing, every attempt at a reasonably explicit methodology ranging over the physical sciences has simply failed to establish its concep­tual authority: for instance, about the very form and realist standing of explanatory models, about the translation of theoretical distinctions in observational terms, about the rela­tionship between empirical regularities and nomological universals, about the logical conditions constraining valid inductions, about the testing of hypo­theses, about the semantic extension of predicates of both observa­tional and theoretical kinds, about the determinacy of intra- and inter-theoretic reference made to unobservable entities, about the very meaning of theoretical statements, about discrepancies between actual laboratory practice and idealized conceptions of how a science works, about the meaning of scientific truth itself, about the logic of confirmation and disconfirmation, about the weight of doctrinal interests and consensual informalities in judging the promise and worth of ongoing inquiries, about the nature and admissibility of intentional complexities, about the methodological significance of palpable differences among the different sciences, about differences in the quantification of intentional and non-intentional and physical and cultural attributes, about explaining the conceptual relation­ships between theory and practice and between perception and thought, and, for that matter, about the bearing of our different conceptions of what a science is, or ought to be, regarding appraisals of the would-be rigor of any would-be scientific practice.

It’s true enough that one can overstate the importance of dis­tinctions of these sorts. But they cannot be neglected, and they signify the need for a certain latitude in reviewing the conceptual pretensions of the physical and human sciences. Consider, for instance, that the production of the first atomic bomb, viewed as a technological feat involving unobservable structures and forces interpreted in quantitatively explicit terms, risks the validity of the associated conjecture by relying on our witnessing the precise event of the predicted explosion. The very feat confirms the validity of the underlying methodology, rather than the other way around, and does so more in terms of practical adequacy than of theoretical truth. To be sure, the relation between these two notions must also be explained.

Thomas Kuhn had featured the troubling truth — pertinent to what he called the “gestalt switch” characteristic of revolutionary science — that notices that gains and losses of comparative concep­tual power cannot always be straightforwardly appraised in a period of paradigm shifts. If what has just been said holds true, then a similar difficulty probably infects normal science as well. Given the caveats I’ve just sampled regarding the methodological clarity favored in the heyday of positivism and the unity of science program, it may not be unreasonable to discount would-be philosophi­cal gains in grasping what a proper science is: its gains may be made more doubtful if we trust science to risk the validity of its own brief on some sort of formal rigor with which its best results can be demon­strably derived from an application of explicit methodological rules. For we realize that such derivations are rarely more than heuristic conveniences — not irrelevant for that reason but hardly secure enough to count on.

Ian Hacking has emphasized instead the compelling effect of certain technological successes, for instance, the discovery of Neptune in accord with quantified predictions of its physical pro­perties, the first atomic bomb, Mendeleev’s near-accurate prediction of hitherto unknown elements, Lavoisier’s experiments with the com­bus­tion of mercury in the absence of an adequate theory of the chemi­cal elements. The ability to formulate in a continually provisional way a productive model and one or another methodological sketch of a sector of the world around such contingent breakthroughs, which enable us to link in a promising way parts of the world that had seemed quite disparate before, may be more important than the admir­able (but risky) rigor of attempting to proceed the other way around — by ­fixed procedural rules. If so, then perhaps an ad hoc, improvi­sational, even opportunistic kind of reasoning may be the right paradigm for the art of doing science.

This last idea would be greatly strengthened if, for instance, the theory of science were to construe our theoretical claims as depending, strategically, on solving certain insistent practical problems. So that the longing for universal scope, so dear to the champions of covering law explanations might be convincingly (or indefinitely) postponed in the interest of practical relevance and practical success — perhaps not always, but certainly at those points at which it “really” counts — as in cases like those just mentioned. To think this way is to change profoundly the very conception of scientific rationality.

If you accept this line of reasoning, you must admit that science is inherently informal, however fairly and credibly it presents its conception of its own method in formalized terms regarding, say, its theory of meaning, testing, confirmation and discon­firmation, validity, truth, and universal scope — all the while it responds ingeniously and automatically to ad hoc, serially ideal­ized, quite opportunistic adjustments more than to any demonstrably effective top-down application of provisionally fixed rules or fixed procedures of any kind.

If you concede all this, you begin to see the sense in which the physical sciences are genuinely human sciences: they are, of course, human sciences for at least epistemological and methodological reasons. The seeming validity of the physicalist model of science — which requires in principle a reduction of the work of all the human sciences — manages to obscure the fact that there is no sui generis method that rightly applies to the physical sciences first, and then to any other human inquiry that aspires to scientific standing. This gets things backwards. On the argument being advanced, the reason is this: the rigor of the physical sciences is rightly and reasonably abstracted from various forms of discipline already native to the human sciences — under a convenient form of self-deception: namely, a belief that methodology does not itself enter, “constructively,” into the very constitution of what we count as our primary empirical data (theory-laden data, data penetrated by theory, by ideology, by belief, however much they remain empirically given) — thereby affecting their theoretical import, their relation to empirical confirmation and validity and truth, their realist standing, and similar measures. It begins to look as if it could not have been otherwise.

I have drawn this much of the argument from little more than a reflection on the import of 19th- and 20th-century philosophies of science. I’ve offered no larger intuitions about additional con­straints on the very nature of a science that might be drawn from other speculative sources. But there are other compelling consider­ations that would surely strengthen the hand of my third option. There are many intuitions that could be favor­ably mentioned and enlisted here. But I’m inclined to think that all of them, or nearly all, that would be useful to explore depend on two perfectly straight­forward theorems that are well-nigh impossible to deny. I want to show how to build the case effectively, without actually constructing the argument itself; and I want to show some of its most heterodox implications. It’s the rationale that counts, the rationale and what accepting it entails. But I must point out that these new considerations are meant to bridge two somewhat different lines of reasoning: one is dialectically apposite within the terms of the history of science and the history of the philosophy of science; the other derives from larger and more general reflections on the history of modern Western philosophy. The lesson to be drawn is that these two lines of reasoning converge very neatly in favor of the brief I am defending here.

Here, then, are the intuitions promised: they are, as you will see, terribly obvious. I take that to be a mark of strength. First of all, the human sciences are reflexive. Human beings are the only creatures capable of conducting an actual inquiry; and the human sciences examine aspects of the lives of just such creatures — “aspects” or “attributes” that are rightly deemed reflexive in the sense of linguistic utterance and communication, intelligent behavior, work and creativity informed by thought and feeling, and thought and feeling and reportable percep­tion themselves. On my view, it is self-deceptive to suppose that the physical sciences are not reflexive in the very same sense in which the human sciences are.

I don’t deny that the objects the physical sciences posit for examination are hardly confined to the study of human beings, and I don’t deny that the physical examination of human beings normally does not (and need not) trade on those features that rightly count as reflexive in a sense appropriate to the work of the human sciences. No, I concede all that. ­­Still, the physical sciences do make cogni­tive claims, regardless of what they choose to examine. Accord­ingly, if what I have already said about the way methodology, con­cepts, and theory enter into and affect the very constitution of whatever we claim to investigate — and if we now add that thought and perception are themselves culturally “penetrated” by methodology and theory, very probably also in historically evolving ways — then it cannot be denied that the work of the physical sciences is inelucta­bly and profoundly reflex­ive. Of course, all this is unhesitatingly denied as a result of rendering the very role of thought and percep­tion neutral — or, conceivably, invisible — within the space of scientific inquiry. The role of thought will then appear to be entirely external to scientific findings, or else the content of thought will be detached from thought itself and converted into neutral data.

But, of course, if anything like Kuhn’s idea that scientists working with theories as different as Aristotle’s and Galileo’s (with respect to the pendulum) must “live in different worlds” — or, less controversially, if perception and thought are “enlanguaged” and “encultured” by the various forms of conceptual penetration drawn from our theories and methodologies — then the study of any objective data ineliminably implicates the reflexive features of such data. There’s no escape there.

In just this logically elementary but scientifically momentous respect, every science must be a human science. If, then, you add to the “penetration” question all the puzzles I’ve col­lected from the history of the failure of positivism and the classic forms of the unity of science program — for example, the fact that there is no formal algorithm by which to fix reference, predica­tion, context, meaning, the relation between theory and practice, the relation between perception and conception, and the determination of empirical truth and the like, you see how unlikely it is that anyone would think the physical and human sciences could possibly be methodo­logically or epistemologically disjoint.

The physical sciences are demarcated by way of a plausible abstraction drawn from within the larger purview of the human sciences, or from whatever are the cognitive sources on which the human sciences depend. But they are abstractions — deliber­ately restricted human sciences that, by and large, suppose they can avoid the distinctly reflexive problems of the latter: because they feature physical systems other than the human; because, even in the human world, they eliminate all non-extentional attributes; and because, in doing that, they neglect to consider the very formation and preforma­tion (Vorurteil) of their perceptual, theoretical, and methodo­logical practices. Those powers, of course, are subject in the deepest way to whatever is reflexive in the work of the human sciences them­selves. Seen this way, the admirable precision, the fine-grained quantification, the prediction and technological control of the physical sciences remain a tribute to the intrinsic discipline of the reflexive sciences — that is, the human sciences! For the avoidance of non-extensional attributes among the objects of study in the physical sciences is apparently enough to ensure their distinctive virtues. Obviously, they need not deny their cognitive and cultural encumbrances: they literally cannot possibly do so. There’s a reversal of the usual verdict regarding the difference between the two sorts of science. It begins to look as if the human sciences have their own undeniable rigor, since it is they, (or their cogni­tive resources) that make the success of the physical sciences possible, at the same time they constitute a discipline that deliber­ately features the non-extensional attributes most characteristic of human life and culture and history and language from which our reflexive theories of our cognitive powers are themselves drawn — that is, all the so-called intentional, and therefore intensional, attributes of the reflexive human world. There’s no defect there, only the conditions for the success of the physical sciences.

The human sciences cannot avoid the reflexive features of human life. Neither can the physical sciences. It’s the condition of human life itself that requires the admission of non-extensional attributes affecting the physical and human sciences alike. It’s just that the physical sciences present themselves in a way that such attributes do not appear among the objective features of whatever they explicitly concede they examine. Nevertheless, according to the argument being advanced here, neither the non-extensional cognitive conditions of perception and thought nor the conceptual dependence of the extensional vocabulary of the physical sciences on the very sources of enlanguaged and encultured life need ever threaten their distinctive rigor. Why then should we think it a source of weakness among the human sciences that they elect to explore whatever pre­cision may be had in the analysis, explanation, and understanding of the phenomena that provide the sine qua non of the admirable work of the physical sciences, though they themselves resist reduction? They choose to examine what the physical sciences exclude!

The question points to the second theorem promised: namely, that the reflexive features of the human or encultured world are not, or cannot be shown to be, reducible to the extensional properties the physical sciences admit in the world they choose to examine. Hence, to insist on disallowing the reflexive dimension of the human world is tantamount to impoverishing the competence of science itself; or, faute de mieux, to claim (without demonstration) that the non-extensional is reducible to the extensional wherever we claim to be investigating the real world. To take the more extreme option — to deny the realty of non-extensional attributes — is, effectively, to deny our own existence.

This counts as a provisional reductio. I have now conditionally demonstrated, for one thing, that the physical sciences are them­selves restricted forms of the human sciences; further, that they cannot easily disallow (or disallow at all) the examination of the non-extensional dimension of the actual world, since the admission of the reflexive human world is itself a necessary condition for the work of the physical sciences; and, finally, that it is a completely unreasonable, utterly para­doxical complaint that insists that the human sciences should abandon altogether the study of the non-extensional dimension of the human world, that is, in their study of the human world! That would make no sense at all. For, on the strength of my second theorem (my intuition), the reflexive features of the human world are irreducibly non-extensional, though they are essential to the retrieval of the other.

Sometimes it is argued — fatuously, as Daniel Dennett does — that “Science” itself requires that we abandon all first-person discourse in favor of third-person discourse: which is to say, abandon all intentional distinctions in favor of extensional dis­course. But that’s a blunder: because, of course, third-person discourse is every bit as reflexive as first-person discourse. In effect: to say, “The cat is on the mat” is to implicate a reflexive avowal or assertion of some kind (let us say, a perceptual asser­tion), which clearly (then) depends on some actual non-extensional features of asserting what is asserted, features instantly sup­pressed in acknowledging the putative non-intentional content of the assertion uttered. It’s the act of asserting (or thinking) that implicates the intentional complication. Nothing of first-order evidentiary importance can be secured for the physical sciences from such a diminished example.

But my argument requires at least one further step to bring closure to our original query. How, we ask ourselves, might we strengthen our second theorem decisively? Could it be done with a very small labor? I believe it can. I said, you remember, that both of my intuitions were “obvious.” And so they are. But even the obvious can be disputed: the second theorem is indeed strongly con­tested in contemporary philosophies of science. Other­wise, there would be no army of reductionists to take on. Happily, what’s needed amounts to no more than an academic reminder drawn from a source quite different from the dialectical considerations I’ve been favor­ing here. Ultimately, these two strategies converge in the most natural way.

Consider only this. The whole of early modern Western philoso­phy prior to Kant rests on an insuperable paradox (if it doesn’t explicitly affirm it) that can be overcome only by adopting the master stroke by which Immanuel Kant first resolved the Cartesian paradox. That is, only by rejecting every principled disjunction between cognizer and cognized within epistemologically defined contexts in which we seek to gain the advantage of some form of direct realism. Kant argues masterfully that, in cognitive contexts, we cannot, as far as experience is concerned, separate the “contribu­tion” of cognizing subjects and cognized objects with regard to whatever counts as realist knowledge. It’s true that Kant goes on to claim, more than improbably, that we can reconstruct the necessary a priori conditions on which empirical knowledge is conceptually possible; and, of course, it’s also true that Hegel shows us a way to preserve Kant’s original, seemingly unsurpassable insight, while abandoning every pretension regarding its strictest transcendental truths. But the point of importance remains unchanged for present purposes, if we may borrow from both Kant and Hegel: namely, that the solution to the Cartesian puzzle (whether read in terms of direct or representational realism) entails the indivisible unity of what we claim is real and what we claim to know to be such. The consequence is that it becomes impossible to investigate the physical world without implicating the reflexive dimension of the human world. This single corrollary sets a strenuous barrier against the advance of reductionism that cannot be waved away in the manner favored, say, by Paul Churchland and Wilfrid Sellars. Our previously conditional reductio now seems more robust.

Of course, this hardly confirms that certain much-debated features of the human world — those that are most salient and charac­teristic — are indeed irreducibly non-extensional. It does, however, show compel­lingly that there is no way to advance the reductive cause (against the enthusiasts of the human sciences) except by directly confronting an essential epistemological question that the entire history of early modern philosophy has shown to be unavoidable. The reasoning involved, which bears on the fortunes of the philosophy of science, affects every conjecture regarding realist knowledge, demon­strates that we cannot defend any pertinent form of the realist claim itself without first establishing the cognitive credentials of the sources on which the human sciences depend and what finally should count as the right relationship between the physical and the human sciences.

I freely admit that there is no hands down argument by which to refute the bare possibility that reductionism is true. I doubt that that could ever be shown. But there is also no known or remotely promising strategy in its favor. Every serious reductive sketch has proved completely inadequate to its task — nearly always for very simple reasons — running from Freud to Carnap to Hempel to Sellars to Feyerabend to Skinner to Quine to Churchland to Dennett or to any more recent champion you may care to name. But if so, then there’s no need to encumber the analysis of the human sciences any longer under reductionism’s impoverishing threat or promise. We must proceed faute de mieux. The conceptual distinctions of the human world drawn from its cultural and historical side have been delayed in analytic quarters for at least a century. Truth to tell: there is almost no sustained discussion of the human sciences among the strong specimens offered in recent philosophies of science that explore the ontology, epistemology, or methodology of pertinent inquiries openly opposed to the reductive thesis. In fact, the dearth of arguments in favor of the two principal alternatives to the reductive model is itself an obvious consequence of the hegemony of the first. It’s time to move on.

I’ve hardly done more here than sketch the barest clue to the distinctive work of the human sciences. I’ve barely touched on its special problems, although they are easily enough identified by probing any of the puzzles I’ve already mentioned — which, as I say, the physical and human sciences share in somewhat different ways, but which the physical sciences tend to ignore or fail to notice. In any case, they are all articulations of the generic, the paradigmatic puzzle of the uniquely reflexive dimension of human life in its cultural, historicized, self-conscious, intentional, interpretable, significative, cognitive, effective, and creative manifestations. That’s an immense world in itself, of course, which the physical sciences cannot address, except dependently, if my second theorem is true or cannot be shown to be false. So I’ve attempted nothing more than a preamble to the human sciences in the way of an instruc­tion regarding where one might go wrong.

I can now put the entire argument in a nutshell. The best work of early modern Western philosophy, which spans the interval that joins Kant and Hegel, yielded the single most important of the sine qua nons of any ­future analysis of the actual world and of our know­ledge of it — affecting our account of any responsible and coherent science. That is the condition that, in one accepted use of the term, is now called “constructivism” or “constructive realism.” It signifies that what is admitted to be “given” in experience in any way that bears on presumptive knowledge without tendentious presup­po­sitions of any kind — prior to all theoretical speculation but surely already infected by whatever enculturing forces “penetrate” our per­ception and thought (perspectived, therefore, even prejudiced, but not for that reason philosophically privileged) — cannot be cogni­tively decomposed in order to deter­mine the separate subjective and objective contributions to the work of any of the sciences.

Any realist reading of the world we claim to know cannot, then, be more than “constructed” or “constructivist”: which is to say, cannot possibly justify any principled disjunction among the epis­temological resources of the physical and human sciences, cannot recover the noumenal world of the pre-Kantian realists. This is not to confuse the distinction between realism and idealism; rather, it draws attention to the complicating fact that the very idea of classical realism was itself irretrievably corrupted by the Cartesian paradox that Kant and Hegel resolved.