Prologue:
Change and Invariance

I

Occasionally we step back from the ordinary flow of engaged life to consider what we mean by this or that notion. For instance, what, we ask, is history really ? Surely history is distinctive of human affairs. Stop there! Does that signify that natural events such as the eruption of a volcano cannot rightly be said to have a history?

We are already in conceptually troubled waters. Is the fertilization of a human ovum or the biological growth of an embryo of Homo sapiens sapiens precluded from having (or being) a history? Is there an important question here, or is it more a matter of whether we take a narrative interest in the temporal span of particular events and processes no matter what they are? There's an inkling of the single most strategic option concerning the nature of history. Keep it in mind: it will haunt this account.

Certainly, human and natural affairs change, take place in time (as we so effortlessly say). But is mere change history? Or the mere record of change? And is the time of natural change the same as the time of human history? Or, if they are different, how can they be different, seeing that whatever has a history must undergo change, must persist through time, in the same sense in which time fits all natural events? We are tempted by the intuition that, whatever we now mean by history, things cannot have changed all that much since the ancient beginnings of philosophical reflection to justify or require a conception of history distinctly opposed

to whatever may have been meant by history in the past. But that would be a great mistake. It is undoubtedly true that the ancients speculated about history and change and time, but they did so in accord with conceptual convictions quite unlike our own.^[1] In fact, such differences may be a function of our changing history. Our theory of history may have to acknowledge that the theory of history—and history itself—have histories. The complications are beginning to swarm.

II

The pre-Socratic Greeks were distinctly fascinated by the problem of change. "All things come from water," says Thales at the beginning of our official record of Western philosophy. Parmenides, however, in the sternest of pronouncements, says, of What Is, you can only say that It Is; and of What Is Not, that It Is Not—meaning thereby, apparently, to deny the reality of change (since, as the jargon has it, change implicates "What Is Not," "Nonbeing")—that is, what is unsuited to discourse. Yet, the fragments that remain of his great poem confirm that Parmenides ventures a good deal about what (on the thesis) is not real at all—namely, change. And Heraclitus, possibly the most prophetic of the pre-Socratics in the way of anticipating the general line of thought that leads eventually to Aristotle's extraordinary vision of science, plainly compromises with those who insist on the reality of change (Thales, read metonymically) and with those who insist that reality is changeless (Parmenides, for a certainty).

Heraclitus holds that the changeable world (the one we perceive and act in) must be inherently subject to a changeless law of order (the Logos or the lesson of the logos ), which is unaffacted by change itself. On that view, we may suppose (thinking in a looser way than Parmenides), change is real enough, since it embodies or is governed by the changeless structures of What Is. We cannot be sure that that is Heraclitus's view, but it is certainly close to Aristotle's exemplary account of science in Metaphysics , Book Gamma. The whole of pre-Socratic thought may, in hindsight, be taken to be a sort of preparatory labor for Book Gamma (which is Aristotle's own view, as it turns out)—together, of course, with the more than casual contribution of the numerous alternative solutions invented by other gifted Greeks (not merely pre-Socratic), notably Plato, whom Heraclitus somewhat anticipates.

Where Plato is concerned, we can only map the plausible options of interpretation, because we cannot be entirely sure of Plato's intent in the

deep ironies of the Dialogues . On the sternest reading, the Parmenidean, we may say that Plato disjoins the orders of Being and Nonbeing, the changeless and the changing, and constructs rather pretty stories (but only stories) to suggest a way of linking the "one" with the "other" (or the "many"). Plato may not have meant "his" doctrine literally. We cannot say for sure. Our own world favors Aristotle in this regard because the notion of an invariant world of separable eternal Forms (whatever that may be interpreted to mean) is a formidably expensive conjecture we believe we can do without—in fact, we believe we had better do without.

You will have noticed how our topic sidles, without warning, from history to reality. To theorize about history is to theorize about reality, even if only (with Parmenides) to deny that change is real. And yet, of course, it would be hopeless to attempt, in this small review of the question of history, to insert the entire stunning question of just what is real—and why. It will be enough to fasten on the matter of what to mean by treating histories as real—if only to oppose those who presume that they are not in the least real, unless in the pallid sense that stories (narratives, histories therefore) are real . . . stories . The issue is clear enough: we must ask ourselves whether we should admit that there are real "things" that have or are histories, which certain narratives (history in the reportorial sense) are about .

III

Whatever we make of the usual remark that Herodotus is the father of history, there can be little question that Thucydides is a splendid historian, a historian of a remarkably modern sort.^[2] There is every sense, in reading Thucydides, that he intends his narrative to represent, to correspond to, the actual history, the narratizable events (as we might now say—tendentiously), that he was engaged in relating. It is impressive, for instance, how many lengthy speeches and conversations Thucydides reports. There is the great speech he attributes to Pericles, the funeral oration for the first of those who fell in the war against the Peloponnesians,^[3] and the lengthy exchange between the Athenians and the spokesmen of the island of Melos just before the Athenians began the military campaign against the island.^[4]

Now, direct citation would, of course, have obviated narrative at the point at which it appears. But Thucydides sets every speech in the context of a larger narrative in which the cited material is clearly intended to be

congruent with a fuller account that is accurate regarding what is narrated: the cited material is meant to conform with (and perhaps confirm) the accuracy of the latter. Thucydides invariably gives the impression that the accuracy of what appears to be quoted is in accord with the speaker's intention at the time. So, for instance, when he gives the lengthy exchange between Nicias and Alcibiades regarding the advisability of the Sicilian campaign,^[5] Thucydides remarks at the close of a comment of Nicias's (who opposed the expedition): "These were the words of Nicias. He meant either to deter the Athenians by bringing home to them the vastness of the undertaking, or to provide as far as he could for the safety of the expedition if he were compelled to proceed. The result disappointed him."^[6]

In short, Thucydides construes direct discourse as a narrative device, guided by the spirit and intention of the speaker of his picture, possibly roughly accurate in some of its wording, but surely more seriously accurate in the way that belongs to his own purpose—which, we may fairly say, is narrative accuracy—and often instruction. The scruple of his entire work would make no sense unless Thucydides could not have been put off by the inevitably contingent inaccuracies of his reports of speeches just where he believes he is accurately reporting what took place. He must have supposed he was reconstructing the speeches themselves—accurately. There is no way to read his history without supposing the events he represents narratively are, accurately, narratively formed.

In this regard, there must be a measure of convergence between Thucydides' plan for his history of the Peloponnesian War and Aristotle's comparison, in the Poetics , between poetry and history. In fact, Aristotle offers a brief account in the Poetics that is expressly about the kind of device Thucydides uses. "From what we have said," he remarks,

it will be seen that the poet's function is to describe, not the thing that has happened, but a kind of thing that might happen, i.e., what is possible as being probable or necessary. The distinction between historian and poet is not in the one writing prose and the other verse—you might put the work of Herodotus into verse, and it would still be a species of history; it consists really in this, that the one describes the thing that has been, and the other a kind of t hing that might be. Hence poetry is somewhat more philosophic and of graver import than history, since its statements are of the nature rather of universals, whereas those of history are singulars. By a universal statement I mean one as to what such or such a kind of man will probably or necessarily say or do—which is the aim of poetry, though it affixes proper names to the

characters; by a singular statement, one as to what, say Alcibiades did or had done to him.^[7]

It would be hard to hit on a more apt defense of Thucydides' practice. For what Aristotle says of history he could have said—and in a way does say—of ethical judgment. It is also, by the way, not important for the Greeks that poetry might be about fictional characters, whereas history concerns itself with actual persons and events. The Greeks may never have supposed that Agamemnon and Oedipus were merely fictional. But Aristotle's intent is clear: even mere stories may capture what is changeless or essential in human nature (in a practical and normative sense). That explains the persistent theme (in Aristotle and his followers) that poetry functions to convey a high instruction.^[8] In fact, by this distinction alone, Aristotle sets the stage for a troublesome puzzle that has engaged a number of recent theories of history. For, although he does not himself develop such a theory, Aristotle does not rule out histories that strengthen their grip on the realist standing of what they claim by embellishing their chronicles with fictional or near-fictional approximations to what is "probably" essential to the life and nature of particular historical agents. This is what Thucydides' practice begins to make plausible. But in a world very much less sanguine about essentialism than was Aristotle, the idea of a realist history that openly invites the extended use of fictionalized narrative seems distinctly paradoxical.

We're meandering a little, but we've gained some ground. For one thing, if the Greeks (Aristotle and Thucydides in particular) thought of history as a diminished science concerned with the singular ephemera of human life in terms of what is possible or plausible for men of this or that recognizable kind—a cut, therefore, below poetry—history (for them) would still be addressed to what was invariant in human nature (and good, with respect to action, at least "for the most part"), though always only among the most local contingencies and only in accord with the precision the domain was able to support. For a second, for the Greeks, history could not possibly be mere change—for they discounted change as being of no intrinsic interest in itself: change is interesting when it draws our attention to certain significant invariances. It must attract a bona fide science; and only changes of a certain order could support a narrative history. Finally, only changes of the right kind, changes in human affairs that could attract a Thucydides, could yield a narrative interesting enough to invite questions about its accuracy. So the Greek conception would make no sense if there were no narratizable reality that historical narrative could accurately represent.

Hegel, in fact, has this to say about the Greek historians in opening his "popular" lectures on the philosophy of history:

As to the first mode [of history: "original history," the kind that provides the data for the other more advanced modes of history he mentions—"reflexive history" and "philosophical history"—which includes his own effort], the mention of a few names should give a definite picture of what I mean. Herodotus, Thucydides , and their like belong to this class—that is, to the class of historians who have themselves witnessed, expressed, and lived through the deeds, events, and situations they describe, who have themselves participated in these events and in the spirit which informed them. They have compiled a written record of these deeds and events, thereby transferring what were previously mere extraneous happenings into the realm of intellectual representation. What was originally mere existence thereby takes on an intellectual aspect and becomes a representation of the internal and external faculties of mind.^[9]

Hegel's remark that the "original" historians "share" the spirit of the age whose history they report signifies that the narratized structure of the events they consider belongs already to those events.

The Greek conception obliges us, then, to distinguish between mere change and the real thread of human history. It does so on the strenuous thesis that human affairs show an invariant structure of their own, one that answers to the strong requirements of science. We , on the other hand, necessarily plying our trade late in the twentieth century, are more than doubtful about essential human nature, the recovery of the Aristotelian notion of a science, the supposed laws of history (the Marxist laws, for instance),^[10] and even the separability of the laws of nature and historical contingency.^[11] Consequently, we cannot redeem a full-blown theory of history as easily as the Greeks, one that would admit, uncontroversially, real history and its accurate representation; or, if we claim such a theory, then, admitting those doubts, we are bound to construe it in a very different way. That's not to say: (1) that history and change are indistinguishable, or (2) that history is no more than a narrative, a pleasant fiction or heuristic device, concocted to represent a selected continuum of change in narrative form (judged from this or that point of view). But if history is more than either option pretends it to be, we need to explain just how the new possibility can be made out—why it matters and how it can be legitimated.

We need not subscribe to the Greek view, of course. But even the Greeks held different views of history and change: certainly Herodotus and Thucydides could not have agreed less in their practice; and Plato seems to have held—at any rate, he puts the thought in Socrates' mouth—that political history, transformation through time of the known political

constitutions, takes a generally cyclical (that is, invariant) form.^[12] Plato was keen enough about the kind of science Aristotle and Thucydides favored. They share the common genus of Greek history, though Plato, of course, offers a much more radical option than the other two. At its most extreme, where radical change is completely ungrounded, utterly lacking in invariance, history becomes (would be judged by the Greeks to be) entirely meaningless: history could then not sustain the minimal narrative structures apt for the formation of a good life. No science, Aristotle remarks, is concerned with the particular merely.^[13] Tragedy, then, constitutes a paradigm for intelligible history.

That is surely Plato's lesson as well. From our present vantage, however, it involves a drastic choice, the entrenchment of the one and only range of possibilities the classical figures were disposed to explore—with the single exception, always, of Protagoras, whose works are completely lost. For, in recovering history in the modern world, we can either forge a link between change and changelessness, or claim that there is a stable order in change itself that is not invariant , or else confess the world to be an utter chaos. The first option is the one Plato and Aristotle and Thucydides share (in different ways); the surprise is that that vision dominates the modern world as well. The third belongs to Parmenides, but eliminates history (together with familiar nature) altogether. The second has no clear place in canonical Greek thinking, though something like it plainly tempted the Sophists.^[14] The fact is that the entire history of Western thought has, with regard to narrative history, science, rationality, understanding in general, traversed twenty-five hundred years of laborious theory moving in a direction away from the first and third options and distinctly toward the second . The former have never been abandoned, however, and the second has never been convincingly worked out.

There you have the central puzzle of the theory of history. All the intellectual currents of late twentieth-century thought deny or contest or admit to be contestable every form of invariance. They also affirm the coherence of a science addressed to, and infected by, radical change. But what are the possibilities of history according to the second option? And can the first still be recovered consistently with contemporary views of change?

IV

The Greeks considered history, then. But even Thucydides, who has a much more informal view of principles and causes than Aristotle, claims that a grasp of history—an understanding of how it is that "Athens is the school of Hellas"^[15] —falls under a generous conception of science,

and requires a grasp of what is invariant or probable relative to what is invariant. Otherwise neither the example of Athens nor of Oedipus Rex would be thought paideutic. That is the plain sense of the opening sentence of Aristotle's Nicomachean Ethics , the link between practical reason and science: "Every art and every inquiry, and similarly, every action and pursuit, is thought to aim at some good; and for this reason the good has rightly been declared to be that at which all things aim."^[16]

But we need to understand the connection between change and history—that is, in our own world, where we have largely abandoned the ancient forms of essentialism and where the most insistent recent philosophical efforts have sought to retire every strong conception of invariance. If human nature had an essential structure (as Aristotle preeminently affirms), then history could claim a measure of scientific rigor by bringing the ephemera of life into accord with the narratizable structures native to the species. It could not be a science (in whatever measure we might press) unless the generic human career inherently possessed, whatever its vagaries, a narrative form that governed the valid explanation and direction of all individual lives.

Deny that thesis and three intriguing alternatives spring into view: (1) narrative history is no more than a concocted story answering to certain extrinsic interests and imposed for that reason on the recorded changes of particular human lives; (2) history is no more than the orderly chronicle of selected actual changes; and (3) narratizable structures are validly attributed to real lives and social events in the absence of invariance. Notice that the first and second options are compatible with admitting real invariances. They are also particularly favored in modern theories of history. The third option offers a radical departure from the ancient and modern canons. It is in fact favored in certain salient views in the late years of our century. But there is no fully articulated philosophical account of what a reasonable and coherent history of this sort would be like. Views of the first two sorts may be convergently associated with the work of Carl Hempel, Karl Popper, Thomas Kuhn, and Arthur Danto; the most radical versions of the third are associated with Friedrich Nietzsche and Michel Foucault, and a more moderate version may be reasonably associated with the work of Hans-Georg Gadamer.

There is reason to believe the ancient canon is no longer compelling. There is also reason to doubt that the modern option regarding invariance is sufficiently robust: for it characteristically denies the possibility of an objective history of cultural processes or reduces historical objectivity to no more than what would serve a science addressed to physical nature

"apart" from man. (We shall pursue the matter in good time.) But there is also some doubt as to whether theories of history of the third sort can be coherently formulated. Our own age seems determined to find the answer. If it proved coherent now—when it is particularly attractive to theorists of all sorts—we should have reversed the entire history of conceptual thought running from Thales to the very threshold of the next century. And we should have altered thereby the trend of the history of the theory of history. Philosophically, that would be a piece of enviable good fortune.

The larger theme that looms behind the link between change and history and change and invariance is the notion of the flux. The pre-Socratics are obsessed with it and ask how it can be conceptually tamed. Parmenides denies that it is real at all. Plato makes it "resemble" as best it can the invariance of the Forms that never change and have no causal role. His story (in Timaeus , for instance) requires a deus ex machina , the Demiurge. The flux has no form of its own. This seems to be Heraclitus's view as well. Aristotle, catching up the causal concerns of the pre-Socratics, indissolubly embeds the invariant structures that make the flux intelligible, orderly, amenable to discourse, efficacious, and real, in the materia of the flux itself. Parmenides' flux (if we may call it that) "Is Not." Plato's lacks inherent structure altogether; so it would be "Not" in Parmenides' sense, were it not for the fact that it can (somehow) be stably formed but never imprinted with invariance. Aristotle's flux (again, if we may speak thus) is nothing but what, in scientific retrospect, we specify as the "distributed" portions of a hypothetical, formless materia (hyle ) capable of sustaining the ordered changes of particular things that are themselves (the changes, that is) governed by the invariant structures of those same individual things. Hence, at one stroke, Aristotle retires (in part) both Plato's Forms and Plato's flux. We, now, at the end of our millennium, are tinkering with a more daring option that was already glimpsed in the ancient world—and outside the Western world, of course—namely, that the world is a flux, but a flux in which the discernible structures of things and the power to discern such structures are themselves inconstant, inseparable from one another, and apt (by reflexive testimony at least and at most) for sustaining a science loyal to the flux—where Plato's and Aristotle's science is not.

The thread of the argument that follows is straightforward enough, but it leads through a baffling labyrinth. The labyrinth is the flux. We can only acknowledge it as the house of the argument that follows. But the thread is a question: how (we may ask), if thinking is inherently

historicized, can we recover the objective history of historied things? There's a puzzle to conjure with.

We cannot turn at once to offer a plausible account of histories in accord with the third option. A proper ground must be laid. The matter rests with two considerations: to succeed, we must, first, dismantle (if we can) the entrenched conviction that reality is invariant in its essential properties and structure—the conviction that we already possess adequate evidence to the effect both that reality is invariant and that denying that produces intolerable paradox; and we must, second, demonstrate (if we can) that the discursive resources a narrative history would need—serving reference, predication, individuation, reidentification, and the like—could be coherently applied even as we strengthen the denial of invariance. The required argument is less concerned with the actual structure of a history of the third sort than with its conceptual credentials; ensuring those credentials would—will—channel our labors in the direction of a great many questions practicing historians are likely to find irrelevant to their own close work.

This is indeed our task. It will take some ingenuity and tact to gain its end. The presumption of invariance has been so massively entrenched in the entire span of Western thought that the very idea of opposing it, as well as the patient preparation of a viable alternative, may strike many as bordering on the bizarre. But if we concede that the resources of argument are not entirely settled beyond standard deductive arguments, then the plausibility of the labor may benefit from a sustained reflection on the foundations of Western philosophy. The very meaning of argumentative validity is at stake. For if the perceived rigor of philosophical argument rests on our perception (and interpretation) of how the world is structured, the "logic" that serves invariance and the "logic" that serves the flux may prove quite different—wherever arguments are "dialectical," "ampliative," "pragmatic," nonmonotonic," "legitimative," "inductive," "abductive," or the like. The structure of valid reasoning and the structure of the intelligible world could not then be altogether independent of one another; and, where arguments are conceded to be inherently informal, our sense of rigor could not then be altogether independent of the discursive practices of our society. In that sense, argument and rhetoric are consensually linked and may be artifacts of history. Here, at the beginning of our tale, it is enough to broach the possibility. For, if thinking is itself a historically constituted competence, then the ineluctable question of the bearing of the nature of history on the logic of science and philosophy will be acknowledged.