Three—
The Declension of Progressivism

I

It is a fact that nearly all the latest versions of progressivism have been inspired by the work of Charles Sanders Peirce: some directly, like Putnam's, Habermas's, and Karl Popper's; others, surely through their mediation, like Kuhn's, Lakatos's, and Larry Laudan's. It is an even more important fact that Peirce's vision—his progressivism, not his general philosophy—is a conceptual disaster. No matter. It is its optimism that has carried its colors triumphantly through the entire Western world. That failed optimism, often overlooked in its generous acceptance and influence, has obscured an important part of the puzzle of history.

One might suppose that Peirce's doctrine was no more than a gloss on the old tale of loosing a band of monkeys to pound on typewriters an infinite amount of time, supposing that they would eventually produce the corpus of Shakespeare's plays. That would be a mistake. It was much more than that, and it was certainly not stupid. But it was and remains wrongheaded, impossible to redeem; and its failure affects the fortunes of the theory of history in a most profound way. For, in failing, it leads us to conclude that if there are no discernibly invariant structures in our symbiotized world at any contingent moment in a blindly historicized inquiry, then there cannot be any compelling (or plausible) reason for supposing that such structures are, in principle, discernible at the end of an infinite process of inquiry, no matter how thoroughly self-corrective and rational it may be on the way. For, on the argument,

self-correction is itself historically blind, and reason is no more than a (high) artifact of historical forces. Without an effective telos inhering in the continuous process of temporal change, Peirce's vision could not but fail. No mere society of human inquirers (no infinite "community" of cosmic inquirers—to favor Peirce's image) could rightly claim to have grasped, at any particular time, what would accord with that telic order that, on the thesis, dawns only at the infinite limit of history's completion—when, that is, the rational order of the whole may be judged to have been served piecemeal through the gathering movement of its temporal parts. This, of course, is just what Putnam's theory of the regulative function of truth could never overcome. It is also what Hegel's theory could never overcome if ever Hegel had meant—which is doubtful—to posit an actual telos at the end of history. (Hegel's theory of course is hardly an open book.) Peirce, we may conjecture, was somewhat victimized by the salient notions of evolutionary progress in his day and the irresistible charm of Emersonian transcendentalism.

This means that progressivism (fallibilism, in Peirce's idiom) is impossible to defend in any but an arbitrary way.^[1] On the "pragmaticist's" theory, some set of the self-corrective intervals of inquiry would have to count as a determinate (humanly determinable) gain over other (errant) intervals in the same process, if Peircean fallibilism were to be humanly vindicated as we moved, in ignorance, toward the limit of infinitely extended rational self-correction. Fallibilism could succeed only on the knowledge of the pragmatic equivalence (if the term be allowed) of the temporal ordering of the aggregated blind perspectives of every society's self-correction and whatever an omniscient reason at the end of time would require of the whole process.

The significance of this gnomic judgment is considerable, but it needs scope. It shows, for instance, that the first and greatest philosophical effort to defend progressivism—Hegel's immense project (if that was indeed its primary purpose)—could not but have failed. It cannot help Peirce's more recent argument, though it makes it nobler, to know that neither Hegel nor Peirce had held that there was a determinate telos governing the whole of history, discernibly effective in the distributed parts of the temporal process. The logic of Hegel's and Peirce's argument precludes the possibility. If they had presumed such a telos, their projects would have been palpable shams: there would then have been a changeless structure built into the very flux of history. In resisting that, they refused, however, the only grounds on which progressivism could thrive.

It was indeed Heidegger's arbitrary judgment that that was the point of Hegel's philosophical cheating.^[2] Heidegger fails to make good his claim or even to make it fully plausible: Hegel could not possibly have deceived himself so simply. But it serves us, now, to mention Heidegger's conception of history—to confront every historicism and progressivism with the unyielding logic of the flux. The result is startling. For, if it makes sense to speak of the inclusive history of the world, then it also makes sense to speak of interpreting the finite bits of time in accord with the proper parts of that encompassing history; and if that's conceded, then there cannot be a telic structure discernible at the end of history if the unfolding parts do not at least share already —as they contingently evolve through the blind interventions of human societies—a proper and humanly discernible place (a progressive place but not a finally telic one) in that ultimate structure . That is what Heidegger should have pressed but did not. That is the site of Hegel's mystery.

The importance of Hegel's account lies there. (Kierkegaard's complaint is essentially idle.)^[3] For what Hegel demonstrates is the insuperable contingency of the rational ordering of the whole of any human inquiry : the power of reason to recover the normative order of the real intervals of history must itself be a creature of the uninformed parts of such a history. But that is impossible without conceptual cheating. The glee with which we refute Hegel again and again obscures the deeper lesson of the sheer historicity of thinking. Hardly any theorist dares embrace the full implications of Hegel's "failure" (or Peirce's). Literal-minded progressivisms—those invented by Jürgen Habermas and Karl Popper, for instance—are no more than betrayals of Hegel's conceptual daring. But let that pass: Hegel failed, if he ever meant thus to succeed.

Gillian Rose, for one, has put the point effectively enough: "Hegel's philosophy has no social import if the absolute cannot be thought."^[4] Rose construes the lesson in terms of Hegel's anticipation of the failed presumption of the universalism (or universalized rationalism) of the neo-Kantians—Max Weber, Emile Durkheim, the phenomenologists, the Marxists, the Frankfurt Critical School, and, of course, Habermas read as a progressivist.^[5] Still more important is the discovery that we cannot step out of Hegel's philosophical shadow.

Hegel's seemingly alien argument entails, without the least effort, the complete refutation of progressivism. If we turn back to Hempel's and Putnam's views, we cannot fail to see that Putnam's thesis is a pale instance (possibly the coyest) of a Peircean-like recovery of the unity of science program. If so, it must be internally inconsistent or arbitrary.

However, Hempel's inductivism is not in the least inconsistent; it is, nonetheless, incompatible with the flux and untenable under the conditions of symbiosis and intransparency. We of course are nearly all Kantians now, leaner by design than the Kant of the first Critique : that is, we mean to avoid (as that "first" Kant cannot) the progressivism (or "methodologism") of the neo-Kantian interpreters of Kant. Hegel shows the way. We abandon, with Hegel and against Kant, the very idea of a uniquely correct conceptual order within which the real structure and significance of the world can be humanly discerned. In that sense, Rose has exaggerated the lesson she draws from Hegel's ("failed") effort; for it may just be that Hegel means to instruct us about the infinite openness of every human effort to capture Absolute Spirit. That is, perhaps Hegel means to draw attention to the impossibility of historicized thought's grasping absolute history, though the human mind (Hegel's, for one) can conceive of the sweep of absolute history. (There is that equivocation in what Rose says, though her intention is clear enough. She does not give Hegel sufficient credit, however. Surely Hegel understood what Rose would explain.)

We begin, as Hegel observes, with our contingent perspectives and experience, and, in every attempt to encompasss an "absolute" history, we are bound to concede that there always looms a larger totality that we cannot fathom. Hegel's theme may be construed as signifying the historical relativity of all human cognitive claims—but not (alethic) relativism , that is, as signifying the contingency and diversity of truth-claims relative to our history, but not a "relational" definition of truth. In this, the Phenomenology serves to define the insuperable relativity of human claims: it does not define "absolute" truth as an achievement that is humanly accessible. Epistemologically, Hegel is a historicist—which is to say, a theorist who views ascriptions of knowledge as relativized to the resources of our cognitive horizon. Clearly, Hegel does not define (absolute) truth in any relativistic way; but his account is certainly compatible with a relativistic theory of truth (though not one that produces self-referential paradox). In a curious way, Hegel's insistence on Absolute Spirit and on defining "truth" from the "perspective" of Absolute Spirit confirms his own sympathy with a historicist account of knowledge .^[6]

Hempel's version of progress presupposes, as Putnam's does not, that there are invariant structures in the real world that science asymptotically grasps in its punctual work at particular stages of inquiry. Hempel "places" the structure of physical reality entirely outside history—out-

side the Kantian symbiosis altogether. Hempelian progress, therefore, is internally consistent, though bought at the heavy price of denying that history is an ingredient of reality; Putnam's progressivism, however, is internally inconsistent, wherever Putnam's increasing vagueness on the matter of methodology (and truth) functions as a sign of his Peircean-like acceptance of the real indeterminacies of the world.^[7] His progressivism depends upon it.

Hempel and Putnam have already made their appearance here as the champions of the two main forms of progressivism in science. Hempel is no historicist, but Putnam is. Hempel accepts the legible invariance of an independent physical nature (inductivism), whereas Putnam is sanguine about its recovery within the space of an asymptotic blindness. The space between them is made to yield (by Hans Reichenbach) still another form of progressivism. For Reichenbach falls back to the spare resources of a neo-Kantian inductivism—without benefit of Hempel's ontology.^[8] But we need to proceed more slowly in this matter.

II

The best-known exposé of Peirce's fallacy belongs to W. V. Quine. Peirce had characterized logic as "the doctrine of truth, its nature and the manner in which it is to be discovered." Logic, for Peirce, becomes the theory of inquiry—the theory, preeminently, of "abduction."^[9] Under this umbrella notion, Peirce supplies the famous pragmatist formula:

The real . . . is that (more exactly: the object of the opinion) which, sooner or later, information and reasoning would finally result in, and which is therefore independent of the vagaries of me and you. Thus, the very origin of the connection of reality shows that this conception essentially involves the notion of a Community , without definite limits, and capable of definite increase in knowledge.^[10]

This is a most carefully crafted progressivism. First of all, Peirce relies on the image of an entire self-correcting "community, without definite limits" to offset (to be "independent of") the particular inputs of "you and me." That already harbors a non sequitur, a desperate maneuver that Peirce cannot do without. Second, Peirce maintains that the infinite community of inquirers, working on and on without assigned limit, is quite capable of a "definite increase in knowledge," which, once again, is a completely gratuitous posit. "The final Interpretant [of what is real, Peirce adds,] does not consist in the way in which any mind does act but

in the way in which every mind would act." Here, "the whole communion of minds" may well include the intelligence of the creatures of other planets.^[11] Here, the play with actual and ideal minds and aggregated and collective consensus cannot be very reassuring.

Peirce intends no irony. I am tempted to say that there is an irony in Hegel, even in his extravagant philosophy of history, where the project of "the philosophy of world history" is taken most literally. For the seed is admitted in Hegel's reflection on two kinds of history:

The sole end of history is to comprehend clearly what is and what has been, the events and deeds of the past. It gains in veracity the more strictly it confines itself to what is given, and—although this is not so immediately evident, but in fact requires many kinds of investigations in which thought also plays a part—the more exclusively it seeks to discover what actually happened. This aim seems to contradict the function of philosophy; and it is this contradiction, and the accusation that philosophy imports its own idea into history and manipulates it accordingly, that I wish to discuss.^[12]

Hegel surely falls to prove, either as an empirical hypothesis or as a necessary philosophical truth, that "reason governs the World, and that world history is therefore a rational process."^[13] The theme, of course, is that of the "cunning of reason." Apart from my own temptation, I must say that J. N. Findlay offers the most succinct reminder of Hegel's excellent grasp of the "formal" inadequacy of the "logic" of teleology:

Hegel now points [in the Science of Logic ] to a radical defect in teleological mediation: it entails the same endless intercalation of Middle Terms that we saw in the case of the Syllogism of Existence. The End passes into execution only by way of the Means, but this implies that the Means must itself be "worked up to" by further Means, and so on ad infinitum : a Purpose, it might seem, cannot get under way at all, since every step in its fulfilment requires another step to be first made. (This is not intended to be a practical Teleology.) . . . Whatever we achieve, being only contingently related to the End, can embody it only in a one-sided, inadequate and temporary manner: it is itself merely a Means which will require further Means to be added to it, if the End is to be persistently pursued. . . . Strictly speaking, we have no finite Ends, but only "the End." This End, says Hegel, never finds a true Means, since it will always require a prior execution in order that this Means may bring it into being.^[14]

On this reading, progressivism can be no more than a peculiarly optimistic variant of Hegel's teleologism. Hegel, of course, is hardly indifferent to the conceptual puzzle of ephemeral constructions forming

determinate parts of a telic invariance completed and discerned only at the limit of history (or worse, discerned, however incompletely, in the flux of time itself).

There's no way in which Peirce could have validly recovered his "logical" optimism. But it is plain enough that he shaped his formula with an eye to eliminating every objectionable privilege "in the short run ." It's not the "short run" or the "long run" that betrays Peirce, however: it's the supposed conceptual connection between the two. At best, Peirce could have hoped that his optimism regarding the long run was formally coherent: it might have been, had he not muddled it with the extraneous idea that, in the short run, the long-run hope entails the inquiring community's capacity for "definite increase in knowledge ." That could never have worked. It's the same difficulty that worries Popper. It's the fallacy Quine identifies (in Peirce) in his elegant way:

Peirce was tempted to define truth outright in terms of scientific method as the ideal theory which is approached as a limit when the (supposed) canons of scientific method are used increasingly on continuing experience. But there is a lot wrong with Peirce's notion, besides its assumption of a final organon of scientific method and its appeal to an infinite process. There is a faulty use of numerical analogy in speaking of a limit of theories, since the notion of limit depends on that of "nearer than" which is defined for numbers and not for theories. And even if we by-pass such troubles . . . still there is trouble in the imputation of uniqueness ("the ideal result").^[15]

"Scientific method is the way to truth," Quine adds, "but it affords even in principle no unique definition of truth. Any so-called pragmatic definition of truth is doomed to failure equally."^[16] On the argument, every provisional gain in a developing inquiry remains forever risked in principle. There are no gains that, once made, are safely fixed come what may. Peirce could not have shown that the fallibilistic correction of whatever procedure appeared to lead in the short run to a significant gain would also hold in any longer run, or would politely fit in the ultimate long run into a single continuous series of incremental improvements convergently projected from any point in that same series. The very method of correction (in the short term) may require correction or invite a radical turn (in a later short run ). There is no "linear" improvement to be had. (Fallibilism can make no sense of it.)

In analyzing Peirce's notion of an "infinite progress," Quine inadvertently tumbles onto the deeper weakness of Peirce's "Hegelian" theme: for it must be the case, on Peirce's view, that some particular discoveries at some stages of human inquiry (but which ones and why?)

must form the henceforth reliable phases of the evolving process whose telos can be discerned only at the end of inquiry.

This is the same point Putnam had championed in the version of scientific realism he once shared with Richard Boyd.^[17] The doctrine of the (transhistorical) regulative function of truth is simply a commitment to the ultimate telos of (the history of) inquiry. But Putnam effectively disallows his own right to appeal—criterially —to such a telos. So Putnam equivocates (as does Peirce) about the relation between reason and the flux of history. Quine takes note of the fallacy but without the slightest interest in the theory of history: he simply reduces Peirce's pronouncement to a formal non sequitur.

Still, it is the same Quine who, in his own voice, had urged an indivisible holism (in science) deeper than any Pierre Duhem had ever dreamt of—one impenetrable to cognitive privilege of any sort. Quine even disallows a principled disjunction between analytic and synthetic truths.^[18] In that attenuated sense, he may be read as having compressed our historicized theme into a synchronic slice of scientific practice (for the sake of his formal treatment of the matter). This is hardly Quine's intention, of course: it's no more than a sly way of explaining the congruity between Quine's criticism of Peirce and our own recovery of Hegel's mastery of the historicist theme. But it would not be amiss to mention that the fatal weakness of Hegel's telic theme regarding the cunning of reason is the mate of the fatal weakness of the analytic theme of the cunning of extentionalist logic. There you have the compendious reductio of nearly the whole of Western thought. It's pleasant to record the charge, but we cannot stay to chat about it.

The weakness of Peirce's theory is precisely its commitment to determinate progress. History, then, is the record of that progress: hence, history has a telos, even if at any moment we cannot fathom its particular increments. Peirce's "thirdness," which counts toward the symbiosis of world and word, does not (as it does in Putnam) entail intransparency (or, radical intransparency) as well.^[19] Progress is palpable for Peirce; it is unclear how it could be for Putnam (once he parts company with Hempel and then with Boyd).^[20] Hegel construes history teleologically and progressively, but he does not hold that there is a telos in actual history or a discernible order of real progress in actual history. The telic and progressive are the normal structures of history all right, but they are what they are when retrospectively mapped by human historians onto the chronicle of appearances. That is the only conceptual option by which Hegel could escape Peirce's unfortunate mistake. However, there

is an explicit teleologism and progressivism in Marx, insofar as Marx contends that the real engine of history lies (progressively) in the dialectical contests of the class struggle. But, even in Marx, there is a deeper strand of thinking that, as in Hegel, is neither telic nor progressivist in any naive realist sense.^[21] (We shall return to the issue.)

III

Karl Popper's falsificationism is even more transparent than Peirce's doctrine. There is evidence of Popper's interest in Peirce, but his references to Peirce tend to be confined to Peirce's indeterminism.^[22] The fact is that the falsificationist method makes no sense without something like Peirce's mythic speculation that the human mind is the cognitive organ of the whole of nature—which, of course, is needed to make Peirce's "abduction" plausible. (Peirce's doctrine is an ingenious version of the process of Geist's self-discovery.)^[23] Imre Lakatos hints at some dependence on Peirce, on Popper's part, in developing his "method of free, creative conjectures and empirical tests";^[24] he characterizes Popper's conception of scientific progress, particularly in The Logic of Discovery , as "increased awareness of ignorance rather than growth of knowledge. It is 'learning ' without ever knowing ."^[25] Lakatos's reading is a kindly letter with a devastating message.

Still, the question nags: how can we know that science is progressive if we are blindly bound to an infinite space of hypotheses, even if some limited conjectures appear (to us) to be falsified by our present resources? Popper answers by linking falsification to his curious doctrine of verisimilitude:

Intuitively speaking [Popper explains], a theory T₁ has less verisimilitude than a theory T₂ if and only if (a) their truth contents and falsity contents (or their measures) are comparable, and either (b) the truth content, but not the falsity content of T₁ is smaller then T₂ , or else (c) the truth content of T₁ is not greater than that of T ₂ , but its falsity content is greater. In brief, we see that T₂ is nearer to the truth, or more similar to the truth, than T₁ , if and only if more true statements follow from it, but not more false statements or at least equally many true statements but fewer false statements.^[26]

Apart from the utter unworkability of this proposal (the measures compared are obviously no more than transient artifacts of our historical horizon and verbal conventions), the point of the intended linkage rests with Popper's so-called "modified essentialism," which is itself an attack on essentialism proper (that is, the strong realist reading of the laws of

nature): "I do not think." Popper says that "we can never describe, by our universal laws, an ultimate essence of the world. I do not doubt that we may seek to probe deeper and deeper into the structure of our world or, as we might say, into properties of the world that are more and more essential, or of greater and greater depth."^[27] On the one hand, essences are taken to be noumenal; on the other, we somehow manage to approach these properties asymptotically through the work of science. It is as if Popper had economized on Peirce's extravagance; for Popper construes the laws of nature "as (conjectural ) descriptions of the structural properties of nature—of our world itself" but as neither "inherent in the singular things [of the world nor as] Platonic ideas outside the world."^[28] Popper advocates an inductivism manqué of his own falsificationist method. In fact, Popper has himself admitted, more recently, the unsatisfactory standing of verisimilitude, without (however) abandoning the notion itself.^[29] The result is the sheer drift of his own progressivism.

IV

Now, Popper has a theory of history as well, but it is equally baffling. In his historical note to the third edition of The Poverty of Historicism , Popper straightforwardly declares: "The fundamental thesis of this book is that the belief in historical destiny is sheer superstition, and that there can be no prediction of the course of human history by scientific or any other rational methods."^[30] But in the preface to the first edition, he had already concluded: "The argument does not, of course, refute the possibility of every kind of social prediction; on the contrary, it is perfectly compatible with the possibility of testing social theories—for example, economic theories—by way of predicting that certain developments will take place under certain conditions. It only refutes the possibility of predicting historical developments to the extent to which they may be influenced by the growth of our knowledge."^[31] Of course, wherever the formulaic view of verisimilitude fails, Popper can offer no measure of the "growth" of knowledge (or the "reduction" of ignorance): he would not then be able to show that "science" had any advantage over "history," he would not even be justified in dismissing the belief in "destiny." Remember: Popper has never shown that increases in verisimilitude accord with the "growth" of knowledge (via falsification).

One cannot be sure of Popper's meaning. If we take him to mean that some predictions may be "rational" though not "scientific," that to be scientific entails complying with falsificationism, then the note and the

preface are clearly incompatible. If, on the other hand, we favor Lakatos's reading (offered a moment ago), then we can never speak reliably of "the [positive] growth of our knowledge." The complaint against history would then be utterly fatuous: not even the prediction of physical events could then be shown to be "influenced by the growth of our knowledge."

In a recent review of his own long-held views, Popper claims that "the rationality of science and of its results—and thus of the 'belief' in them—is essentially bound up with its progress, with the ever-renewed discussion of the relative merits of new theories; it is bound up with the progressive overthrow of theories, rather than with the progressive corroboration (or increasing probability) resulting from the accumulation of supporting observations, as inductivists believe."^[32] But here, "progressive overthrow" surely means no more than "successive overthrow"; otherwise, it is as arbitrary as the inductive measure of progress itself. In that case, science cannot be shown to be "rational" at all. (Needless to say, both falsificationism and inductivism require a reliable measure of progress.) "No society," Popper insists, "can predict, scientifically, its own future states of knowledge."^[33] Of course he's right.

These quibbles are hardly negligible. They show how carelessly Popper dismisses the possibility of a "theoretical [or scientific] history."^[34] It was Popper, after all, who disallowed the realist's reading of the laws of nature and conceded (however equivocally) that historical prediction was possible. It's the same Popper who objects to every pretense at fathoming "the growth of our knowledge" and who yet affirms that "The course of human history is strongly influenced by the growth of human knowledge."^[35] We are lingering over a few lines of print, it's true; but if they help to recover the possiblity of a "scientific" history, they will have been worth our patience.

It may be that, when officially affirmed by Popper, "growth of knowledge" never means more than what is meant by "progress," by increased "verisimilitude," "elimination of error . . . by nonviolent rational criticism."^[36] If so, then the elimination of a world-historical telos or destiny or special laws of history may actually advance the prospects of a scientific history—not, to be sure, in accord with Popper's view of science and history, and certainly not in accord with Hempel's. But there's reason to believe that those two gentlemen are wildly wrong about what the sciences can deliver. Hempel, you remember, had concluded that history is little more than a heuristic exercise: genuine science is heroically committed to eking out the invariant laws of nature. Popper was

always an opponent of that inductivist view; but his own doctrine of verisimilitude (that is, "modified essentialism") simply reinstates, along falsificationist lines, another realism for our nomological "conjectures" to be about . Neither maneuver seems entirely propitious, and neither quite vindicates its contempt for objective history.

V

Consider the following pronouncement of Popper's:

Sociology, to the historicist, is theoretical history. Its scientific forecasts must be based on laws, and since they are historical forecasts, forecasts of social change, they must be based on historical laws.

But at the same time the historicist holds that the method of generalization is inapplicable to social science, and that we must not assume uniformities of social life to be invariably valid through space and time, since they usually apply only to a certain cultural or historical period. Thus social laws—if there are any real social laws—must have a somewhat different structure from the ordinary generalizations based on uniformities. Real social laws would have to be "generally" valid. But this can only mean that they apply to the whole of human history, covering all of its periods rather than merely some of them. But there can be no social uniformities which hold good beyond single periods. Thus the only universally valid laws of society must be the laws which link up the successive periods . They must be laws of historical development which determine the transition from one period to another. This is what historicists mean by saying that the only real laws of sociology are historical laws.^[37]

We know Popper holds that what makes a conjectured hypothesis scientific is that it is genuinely falsifiable. Popper does not envisage that universal lawlike hypotheses in physics will be true ; also, he never shows (he cannot show) that generalizations restricted to historical "periods" cannot be treated as falsifiable. It's true enough that Popper opposes essentialism in all our nomological conjectures. That's the point of his war against "methodological essentialism" (which, he claims, "was founded by Aristotle, who taught that scientific research must penetrate to the essence of things in order to explain them").^[38] But even verisimilitude is really committed—committed even methodologically (against Popper's express denial)—to some such doctrine.

Furthermore, Popper never shows that a commitment to universal invariances is really needed to sustain (in at least a formal sense) the realist pretensions regarding physics or history; nor does he show that historicists could never formulate (if they cared) their own conjectures

in a vacuous universal form, all the while they simply tested their conjectures for the "periods" for which they were really intended. In fact, physics cannot do better in principle or in practice .^[39] Certainly, biology cannot.^[40]

Third, on Popper's say-so, historicists "accept" the "unity" conception, since they suppose that "success in sociology would . . . consist, basically, in the corroboration of predictions. It follows [Popper says] that certain methods—predictions with the help of laws, and the testing of laws by observation—must be common to physics and sociology."^[41] But if all this is true, then there is no principled disjunction between physics and sociology or history, though there are differences in effective prediction of course. Remember: on Popper's theory, the "degree of corroboration of a theory (which is something like a measure of the severity of the tests it has passed) cannot be interpreted simply as a measure of its verisimilitude. At best, it is only an indicator  . . . of verisimilitude, as it appears at the time t . For, the degree to which a theory has been severely tested I have introduced [he says] the term 'corroboration.' "^[42] So, a temporally indexed "corroboration" of a theory does not logically require the exceptionless invariance of its would-be laws, and it does not directly bear at all on confirming such laws (or confirming their verisimilitude). It certainly does not settle the question whether invariant laws are necessary to science, and it exposes how gratuitous the doctrine of verisimilitude is.

In any case, all this constitutes Popper's closest approximation to Peirce: it's the poor fruit of a strenuous labor. It confirms at a stroke that the method of corroboration at t may well need to be altered at t' in ways that cannot ever be rationally determined at t . So Popper is effectively reclaimed by Aristotle (whom he despises). There's a lot of flummery here.

VI

We are tracking the career of a strange concept—from Hempel's high discipline to certain increasingly arbitrary claims about invariant laws and real progress. Its declension depends on historicizing the accomplishments of science and on challenging the grounds for involving invariant laws. We are, in effect, defining a middle ground between the extravagances of Vico and Collingwood on the one hand and those of Hempel and Popper on the other. The first would save history by disjoining its invariances from the invariances of physical nature; the second

would diminish or dismiss history altogether by exposing its inability to bring its explanatory work into accord with the covering-law model. Progress in physical science becomes irrelevant to the first, and the second simply precludes history as a serious discipline.

The disjunction has proved intolerable since the appearance of Thomas Kuhn's The Structures of Scientific Revolutions , for Kuhn infects the analytic philosophy of science with the deep worry that human reason and inquiry must be as much subject to the flux of history as any particular picture of the world our science might confirm. In this, Kuhn himself was surprised, even dismayed, by the gathering appeal of the subversive possibilities of his theory. More locally, his intuitions about history—which are hardly systematically developed—became more and more entangled with all sorts of Hegelian, Marxist, Nietzschean, Weberian, Durkheimian, Foucauldian currents, possibly through a connection with the prescient inquiries of such figures as Ludwig Fleck and Gaston Bachelard.

Kuhn confesses that he too, once upon a time, subscribed to the covering-law model of history. (He does not actually name Hempel.)^[43] He observes that historians cannot be denied access to the covering laws of their own field of interest—if there are any such laws. He claims only that the usual specimen "laws" are trivial and that the historian's explanatory task lies elsewhere:

When [social scientists, imitating the physical scientist] draw examples from historical writings, [he says,] the laws they educe are at once obvious and dubious: for example, "Hungry men tend to riot." Probably, if the words "tend to" are heavily underscored, the law is valid. But does it follow that an account of starvation in eighteenth-century France is less essential to a narrative dealing with the first decade of the century, when there were no riots, than to one dealing with the last, when riots did occur?

Surely, the plausibility of a historical narrative does not depend upon the power of a few scattered and doubtful laws like this one. If it did, then history would explain virtually nothing at all. . . . I am not claiming, [he goes on,] that the historian has access to no laws and generalizations, nor that he should make no use of them when they are at hand. But I do claim that, however much laws may add substance to an historical narrative, they are not essential to its explanatory force.^[44]

There is a distinct gain to be collected here, although Kuhn himself has "no alternative philosophy of history or of historical explanation to offer."^[45] For one thing, Kuhn affirms that the historian's task is not the same as the social scientist's: it is, he says, a narrative task. Second, he favors a realist sense of narrative. Third, he insists that the historian is

subject to constraints of rigor and may be said to explain the events he collects. Finally, he offers a sketch of the useful sorts of explanatory work in history that cannot be provided by the physical or human sciences, that play a consequential part in the ongoing work of the sciences themselves and the philosophy of science. That is something of a small windfall.

We are taking advantage of Kuhn's resistance—as an insider—to the automatic application of the covering-law model (and any of its analogues) to history. Kuhn admits he has no ramified theory of history. (Some would say he has no theory of science.) He believes there are laws of nature, but he has no argument to show that there must be invariant laws. He is candid enough to favor the possibility of objective (narrative) history,^[46] but he does not work out the relationship between science and history. He adds very little more. He mentions inviolable "rules" that "limit but do not determine the outcome of . . . the historian's task": the historian must not falsify the facts to be explained; he works (somewhat as a child does, putting the pieces of a puzzle together) "to fit [those pieces] to form a familiar, if previously unseen, product." Kuhn says the historian is neither a prophet nor a seer. "If history is explanatory," he says, "that is not because its narratives are covered by general laws. Rather it is because the reader who says, 'Now I know what happened,' is simultaneously saying, 'Now it makes sense; now I understand; what was for me previously a mere list of facts has fallen into a recognizable pattern.' I urge [Kuhn adds] that the experience he reports be taken seriously."^[47] He offers little else.

Kuhn calls the historian's work "global." It incorporates discrete events into an inclusive pattern that: (1) resembles the narratizable patterns of ordinary life, and (2) strikes us as explaining, by making familiar , the scattered historical facts that otherwise form "a mere list." Thus far, there is no real difference between Kuhn's and Collingwood's accounts, except that Collingwood is quite clear about his theory of human reason, and Kuhn is not. What we have is rather inchoate as a theory, but the pieces are important.

There is a troubling element in Kuhn's account—the jigsaw puzzle analogy. Apparently Kuhn believes: (1) that historical "facts" are relatively independent of the historian's theorizing intervention in piecing them together in his narrative ; and (2) that the historical facts are facts in no more complicated a sense than are the facts of the physical sciences . In the Preface to The Essential Tension , he also risks an admission to the effect that he has only begun to realize he is practicing "the hermeneutic

method."^[48] But he hardly seems to have grasped that this itself calls into question the jigsaw model in regard to items (1) and (2). (Of course, his view of the physical "facts"—remember Priestley and Lavoisier!—requires a much more radical conception of the role of history than the jigsaw model would ever allow. "Facts" are theoretical constructions. Surely Kuhn sees that.)

So from these small potatoes, a thicker brew begins to form. Even where Kuhn admits that causal laws "enter science as net additions to [empirical] knowledge and are never thereafter entirely displaced," he admits that "it is no longer clear just what it would be for a law to be purely empirical." He also says (but does not stay to explain): "Nevertheless, as an admitted idealization, this standard account of empirical laws fits the historian's experience quite well."^[49] We may, therefore, suppose that "idealization" collects physical events within the scope of a covering law analogously to the way in which the historian's narrative collects historical events. But that already invites a sense of the hermeneutic method richer than any Kuhn supplies.

VII

Kuhn's deep prejudice lies in this: he is committed a priori to some ultimate invariant structure of nature; but his own historicized picture of the actual work of science makes it increasingly difficult to ensure any realist (or "idealist") reading of such invariance.

The obverse side of his conviction appears in his discussion of Karl Popper. First, he acknowledges that neither he nor Popper is an inductivist; then he criticizes Popper for speaking so easily of "testing." "revolutionary" theories by his falsificationist method (or speaking of anyone's simply "learning from our mistakes" in replacing "Ptolemaic astronomy, the phlogiston theory . . . or Newtonian dynamics"):^[50]

Quite possibly Sir Karl's sense of "mistake" can be salvaged, but a successful salvage operation must deprive it of certain still current implications. Like the term "testing," "mistake" has been borrowed from normal science, where its use is reasonably clear, and applied to revolutionary episodes, where its application is at best problematic. That transfer creates, or at least reinforces, the prevalent impression that whole theories can be judged by the same sort of criteria employed when judging the individual research applications of a theory. The discovery of applicable criteria then becomes a primary desideratum for many people. That Sir Karl should be among them is strange, for

the search runs counter to the most original and fruitful thrust in his philosophy of science. . . . I fear that he is pursuing a will-o'-the-wisp born from the same conjunction of normal and extraordinary science which made tests seem so fundamental a feature of the sciences.^[51]

"Testing" and "mistake" prove to be no more than vague bets made within the transient process of historical science—however "salvageable" may be the fruits of a fallibilism continuously committed to the "normal " science of some past interval. In this sense Kuhn implicitly undercuts both Peirce's and Popper's claims. But in doing that, he inadvertently undercuts his own optimism. The difficulty is this: Kuhn cannot draw a principled line between normal and revolutionary science. Progress is impossible to measure. (The reasons are no more than a gloss on Hegel's text.)

Popper too is caught, but for different reasons:

Since I have used the word "progress" several times [he says], I had better make quite sure . . . that I am not mistaken for a believer in a historical law of progress. . . . The history of science, like the history of all human ideas, is a history of irresponsible dreams, of obstinacy, and of error. But science is one of the very few human activities—perhaps the only one—in which errors are systematically criticized and fairly often, in time, corrected. This is why we can say that, in science, we often learn from our mistakes, and why we can speak clearly and sensibly about making progress there.^[52]

The difference between Kuhn and Popper is this. Kuhn believes in progress because he believes in an invariant order of reality that we somehow fathom; but he never imagines he could construct a methodology valid for different histories; and he never ventures an assured claim about any large invariances. He does, it's true, adopt certain generic marks of what it would be reasonable to construe as scientific reason.^[53] Popper, however, advances a timeless methodology, though the actual process of science depends on the sheer contingency of our candidate hypotheses. He insists, therefore, that any successful string of falsifications accumulates in such a way that we may claim to approach, progressively, the noumenally inaccessible invariances of the real world. Kuhn's progressivism is a mere hope that never functions criterially; Popper's is criterally determinate—and therefore untenable.

For Kuhn, the physical laws are inseparable from their encompassing histories, though they may be "salvaged" moving from one theory to another.^[54] Popper has no sense at all of the historicized nature of the

hypotheses he means to test by falsification. Hence, like Peirce, his argument depends on a sort of diminished teleologism that he never perceives.

Kuhn has a more powerful insight into history:

When [the historian] looks at a given period in the past [Kuhn says] he can find gaps in knowledge later to be filled by empirical laws. The ancients knew that air was compressible but were ignorant of the regularity that quantitatively relates its volume and pressure; if asked, they would presumably have conceded the lack. But the historian seldom or never finds similar gaps to be filled by later theory. In its day, Aristotelian physics covered the accessible and imaginable world as completely as Newtonian physics later would. To introduce the latter, the former had to be literally displaced. After that occurred, furthermore, efforts to recapture Aristotelian theory presented difficulties of a very different nature from those required to recapture an empirical law. Theories, as the historian knows them, cannot be decomposed into constituent elements for purposes of direct comparison either with nature or with each other. That is not to say that they cannot be analytically decomposed at all, but rather that the lawlike parts produced by analysis cannot, unlike empirical laws, function individually in such comparisons.^[55]

There may be difficulties with Kuhn's conception of science; but if there are, the pertinent corrections would strengthen his picture of history's function. For what he means here is that scientific "theories are in certain essential respects holistic" (relatively undecomposable),^[56] "global" (meant to unify a society's entire experience), and already narratized in structure (apt for indicating the train of rational thought by which the scientists of a given age made just the conceptual linkages they did among observable data, would-be laws, and explanatory theories). As he says, "[t]he overwhelming majority of historical work is concerned with process, with development in time. . . . About these processes we know very little, and we shall not learn more until we learn properly to reconstruct selected theories of the past. As of today, the people taught to do that job are historians, not philosophers."^[57]

Theories, then, have two salient interests: one is concerned to link explanatory laws to "empirical" laws (themselves connected, though not inductively, to the scattered data of some scientific work); the other is concerned to preserve, by way of a synchronic summary, the historical movement of the rational, self-critical process by which the first comes to be favored (in its time). The historical pattern of past science is plainly hidden, partially erased, difficult to fathom. So history has always had its distinctive role to fill. Where it succeeds, it explains, as Kuhn says, by making the pattern of reasoning "familiar." Hence, Kuhn's excursion

into the theory of history inclines him to dismiss the invariant (real) structures of nature (as in Hempel) and the invariant (real) structures of human reason (as in Collingwood and Reichenbach). But he relents, at least far enough to recover his party credentials.

What Kuhn shows by indirection is the coherence and plausibility of joining a realist reading of narrative history, on the concession that the world may be a flux, with the conviction that human reason cannot claim an invariance greater than the other's . There you have a summary of the entire thrust of Western philosophy—focused, in terms of methodology, at the end of the millennium.

VIII

Finally, we need to offer a better sense of the tension in Kuhn's conception of the history of science, since there is every reason to believe related difficulties will surface in other accounts. The clue is a nice one: it appears almost at once in the Introduction to The Structure of Scientific Revolutions . One could, Kuhn says there, construe the "history of science [as] the discipline that chronicles both [the] successive increments [to an ideally adequate science] and the obstacles that have inhibited their accumulation."^[58] To go that far would be to agree with Hempel and Popper. But then, as he also observes, that very effort is hopelessly stalemated by the fact that, if '[o]ut-of-date theories are not in principle unscientific because they have been discarded "—that is, if, say, Aristotelian dynamics "can be produced by the same sorts of methods and held for the same sorts of reasons that now lead to scientific knowledge"—then "science has included bodies of belief quite incompatible with the ones we hold today."^[59]

Here, Kuhn anticipates the account of history he gives in The Essential Tension : it leads, he says, to a "historiographic revolution in the study of science," in which, as in the work of Alexander Koyré (whom he features), historians "insist upon studying [not the relation of Galileo's views to those of modern science], but [the opinions of Galileo's group, i.e., his teachers, contemporaries, and immediate successors in the sciences, and] the opinions of that group and other similar ones from the viewpoint—usually very different from that of modern science—that gives those opinions the maximum internal coherence and the closest possible fit to nature."^[60] So a strong narrative conception was already in place in Kuhn's first account. On the summary we have been pursuing, it must have signified that Popper completely falsified

the historicized nature of the very concepts on which the progress he reports depended. He cannot have been aware of it; and yet its recovery requires in principle the same "historiographic" attention due Aristotle and Galileo.

Nevertheless, at the end of the original monograph, Kuhn raises and answers the same question of progress. He tries to save the notion and, simultaneously, to oppose Popper's doctrine of verisimilitude. Progress in science must meet at least "two all-important conditions," he says: "First, the new candidate must seem to resolve some outstanding and generally recognized problem that can be met in no other way . Second, the new paradigm must promise to preserve a relatively large part of the concrete problem-solving ability that has accrued to science through its predecessors."^[61] Here, Kuhn certainly eclipses Popper's doctrine; for, even for progress, we need only "preserve a relatively large part" of science's past achievements. The deliberate vagueness of the phrasing scuttles any hope of precision in measuring verisimilitude:

A sort of progress will inevitably characterize the scientific enterprise so long as such an enterprise survives. [This is apparently because, within the scope of an enterprise for solving the problems or puzzles that its paradigms define, such solutions must inevitably be progressive; and because "(scientific) revolutions close with a total victory for one of two opposing camps."] In the sciences there need not be progress of another sort. We may, to be more precise, have to relinquish the notion, explicit or implicit, that changes of paradigm carry scientists and those who learn from them closer and closer to the truth.^[62]

Some difficulties remain. For one thing, Kuhn still insists on a very strong sense of the victory of one paradigm over another; but though that shores up a realist reading of progress all right, it cannot dispel the impression of the transience of the "total victory" it claims to discern. In this regard, Kuhn is a lesser ally of Peirce's and is caught up in the same "Hegelian" web. For another, Kuhn signals that we may have to avoid speaking of "truth"—in the sense of "verisimilitude"—when assessing the outcome of contests between divergent, incompatible, sometimes incommensurable paradigms. But he does not seem to realize that, in principle, he cannot disjoin the two notions: progress in the "short run" implicates and is implicated by verisimilitude in the "long run." He is doomed to be Popper's apprentice here. In fact, there is no evidence that the contest of paradigms ever yields a truly "total victory"; or that, among the advanced sciences, "recognized problems . . . can be met in no other way" than by applying this or that exclusive paradigm.

Furthermore, in the important Postscript to the second edition of The Structure of Scientific Revolutions , Kuhn considerably weakens his declared sense of the logical play of "incommensurable viewpoints":

There is no neutral algorithm [he says] for theory-choice, no systematic decision procedure which, properly applied, must lead each individual in [a given scientific] group to the same decision. In this sense it is the community of specialists rather than its individual members that makes the effective decision. . . . Two men who perceive the same situation differently but nevertheless employ the same vocabulary in its discussion must be using words differently [within the same community, perhaps from two different communities]. They speak, that is, from what I have called incommensurable viewpoints. How can they even hope to talk together much less to be persuasive.^[63]

This surely means that the fate of the notions of "progress" and "verisimilitude" go hand in hand: Kuhn's instinct for abandoning the second should have instructed him to abandon the first as well. (It is, we may say, Hegel's revenge.) Kuhn himself concedes the point obliquely: "the superiority of one theory to another is something that cannot be proved in the debate . . . each party must try, by persuasion, to convert the other."^[64]

The center of gravity shifts now, away from the question of progress to that of communication and linguistic understanding—that is, shifts to the incommensurability question. Our puzzlement concerns the fact that

because the words about which difficulties cluster have been learned in part from direct application to exemplars [roughly: from "the concrete problem-solving solutions that students encounter from the start of their scientific education"], the participants in a communication breakdown cannot say, "I use the word 'element' (or 'mixture,' or 'planet,' or 'unconstrained motion') in ways determined by the following criteria." They cannot, that is, resort to a neutral language which both use in the same way and which is adequate to the statement of both their theories or even both those theories' empirical consequences. Part of the difference is prior to the application of the languages in which it is nevertheless reflected.^[65]

Let us be open enough to admit that objective judgments of progress are no longer possible here. Kuhn falls back to an unacceptable objectivism—in fact, he sounds remarkably Quinean. He claims for instance that those who experience "communication breakdowns must . . . have some recourse," and he offers the following reassurance: "The stimuli that impinge upon them are the same. So is their neural apparatus, however differently programmed."^[66] But the recovery can be no more

than an illusion, for the "neural apparatus" thought to be cognitively pertinent must be subject to the same quarrels, the same incommensurabilities, the same divergence of educative "exemplars" Kuhn has just been telling us about. The incommensurabilist issue is much deeper than he can admit—or fathom.

We need another moment to collect our gains. We have been tracking all the progressivists to their lair. They cannot make their doctrine plausible: certainly, they cannot make it compelling. Behind all the verbiage, what we glimpse is the not-so-secret insistence on the invariances of nature. Kuhn remains "a convinced believer in scientific progress," but he also confesses in effect that he must worship progress from afar. At any rate, there's no necessity in it (in the Hegelian sense): there's "no theory-independent way to reconstruct phrases like 'really there'; the notion of a match between the ontology of a theory and its 'real' counterpart in nature now seems to me [Kuhn admits] illusive in principle."^[67] Invariance, then, is no more than a provisional posit within a fluxive inquiry; and science collects our best conjectures without ever exiting from history. Progress, therefore, is not a principled distinction. The theory of history must take note of that.