Preferred Citation: Margolis, Joseph. Interpretation Radical but Not Unruly: The New Puzzle of the Arts and History. Berkeley:  University of California Press,  c1995 1995. http://ark.cdlib.org/ark:/13030/ft209nb0kk/

Interpretation Radical but Not Unruly The New Puzzle of the Arts and History Joseph Margolis UNIVERSITY OF CALIFORNIA PRESS Berkeley · Los Angeles · Oxford © 1995 The Regents of the University of California

Preferred Citation: Margolis, Joseph. Interpretation Radical but Not Unruly: The New Puzzle of the Arts and History. Berkeley:  University of California Press,  c1995 1995. http://ark.cdlib.org/ark:/13030/ft209nb0kk/

for Coco, through good years
in an uneasy world
continually reinterpreted

PREFACE

This is the companion of a book, The Flux of History and the Flux of Science, which the University of California Press published during the fall of 1993. I am particularly pleased, therefore, to see this volume appear—and with California. The two books pursue a single theme from different directions. It is not, I should say, usual for American or Anglo-American philosophers, writing on mainstream topics, to write on either interpretation or history. Frankly, I regard that as a symptom of a certain conceptual impoverishment. History is almost completely neglected in nearly every (self-styled) "analytic" specialty; and interpretation seems marginal only because the analytic tradition is strongly disposed to believe that, essentially, we confront the world directly and have no deep need (except for accommodating the transient contingencies of ignorance and lack of evidence) to construe our cognitive interventions as interpretively freighted. That conviction is simply dead wrong.

The natural site for theorizing about interpretation is, of course, the world of human culture: the world of history and the arts. I am persuaded that its problems are intrinsic to the largest questions regarding knowledge, the sciences, human selves and societies, the very nature of reality. This is a view not generally favored among analytic philosophers. I am not aware of any full-fledged analytic account of interpre-

tation or history, or their conjunction, that ranges in a detailed way over the specialized literatures of the arts and history and the sciences. Or, for that matter, any account that enlarges the conceptual resources of analytic practices by enlisting (for instance, by interpreting in the English-language idiom) the best and most relevant work of continental European philosophies concerned with the theory of interpretation and history.

The fact is that, now, toward the end of the century, the mutual— often contemptuous—disregard of these two large groups of thinkers (and their allies in other disciplines) is beginning to thaw. Good will, however, is not enough, given the long chill of at least fifty years. What is needed are very strong, creditable, informed, compelling proposals for bridging the gap and for visibly enriching the technical resources of all by mingling or reconciling the strongest features of the alienated strategies of each. What I have written here I have written in this spirit: against actual professional practice but for the sake of its improvement. I believe the philosophies of the new century will come to see that they must occupy the middle ground between the extremes of the "two" schools.

The puzzles of interpretation were the very first topics I explored when I first entered the professional lists. It is extraordinary to find myself returning now to these first intuitions. They have apparently been active in a subterranean way through all my work and have surfaced explicitly at this late date to give fresh unity to the sprawl and scatter of earlier inquiries. In the history volume, I canvassed the salient theories of history and science belonging to the Western philosophical tradition; in this volume, I do the same for the theories of history and the interpretation of cultural phenomena (preeminently the arts).

I am especially keen on the fact that I have not (or so I believe) neglected to bring into my account, in a sustained and dialectical way, all the strongest currents of contemporary analytic philosophy. The familiar charge is that the "continentals" cannot understand its subtlety or power or cannot construct an argument of comparable rigor. My own view is that analytic philosophy has somewhat lost its way—with all its admirable skills. So I say that what I offer here, together with the history volume, is the work of a "reborn" analyst. I am not unaware of the perils of so saying. The fact is, I am persuaded that the intellectual future of the English-language world (which has profoundly penetrated the rest of the world) depends on the care with which the enlargement and correction of vision that I recommend (regardless of my

own success) demands a scruple that many will find descends to the professionally local. I admit the risk, but I insist on the necessity; I cannot see how the would-be improvement could ever gain ground without enlisting the formidable talents (and influence) of the "analytic" academy that, for the better part of a century, has turned its back on its principal themes.

A great many professional friends, more interested in what of value might emerge from a very long labor than in being merely kind, have encouraged me to believe that it would be well worth the effort to bring my reading of the issues to full form. I am more than happy to have followed their instruction. But of course I now see deeper questions that had been lying in wait until I could understand them rightly. No serious thinker would wish to be without such familiars. But that alone confirms that I can no longer identify all those who have helped me—patient companions on the way. I risk mentioning only two: Tom Rockmore and Michael Krausz, who, over the years, have never tired of obliging me to make my arguments clearer and stronger in every pertinent detail. There are others, of course, too numerous to mention. I hope they will forgive me for not thanking them explicitly.

I must, however, thank Raeshon Sykes in particular, who now reads my scribbled corrections better than I can and who prepared the manuscript through all its final stages; also, Edward Dimendberg of the Press, a most generous and most supportive editor. Still, no one but a book's author can fathom the inner odyssey by which it finally arrives at the public shore.

J.M.
Philadelphia

ACKNOWLEDGMENTS

Some of the essays in this volume have appeared in print, in quite different form. "Reinterpreting Interpretation" was the text' of my presidential address for the American Society for Aesthetics (1988) and appeared in the Journal of Aesthetics and Art Criticism 27, no. 3 (Summer 1989):237-251. "Interpretation at Risk" appeared in The Monist 73, no. 2 (April 1990):312-330. "Prospects for a Theory of Radical History" also appeared in The Monist 74, no. 2 (April 1991):268-292. And "Puzzles of Pictorial Representation" was first published in the third edition of my anthology, Philosophy Looks at the Arts (Philadelphia: Temple University Press, 1986), 338-357. These essays have, over the years, discovered their mutual affinities and separate weaknesses, and all have changed in important ways. The other essays, I believe, were all written for the present volume.

INTRODUCTION

I

As the twentieth century comes to a close, some will see in it the edge of a conceptual revolution. Others will gather to oppose the claim. To say we are on the "edge" is to say its lineaments are clear enough; but also, to be a proper "revolution," the themes perceived must have infected in a public way the theorizing cadre of our society well enough to produce a significant change of direction in its work and deeply enough to justify a sense of the fundamental redefinition of its problems. The second condition is not easily met, but it cannot ignore the role of interpretation.

The truth is, the revolution that is at least incipient has taken incipient form many times. It was voiced already at the beginning of Western philosophy by Protagoras. That Protagoras was cut down so quickly by Plato and Aristotle, ridiculed in fact (for no entirely satisfactory reason), does not alter the fact that the radical vision here intended has had an earlier inning (in a much different form) and that, there, it never achieved the reversal of conceptual fortune we may once again anticipate. It cannot quite fail in the way it once did, though it may never succeed in the way one dreams. It cannot utterly fail because so many of its executive themes have already been incorporated by respectable lines of professional theorizing. But it may not succeed in a grand way either, because its coopted lesser elements obscure by sheer number the supposed advantage of overthrowing the prevailing practice.

In effect, the sense of revolution has been dampened by its own piecemeal success; more curiously, the champions of each of the successive reversals of the supposed canon have characteristically pursued their own work (thus altered) as if it were a project defined by straightforwardly applying that "constant" canon correctly.

You may take this as an instructive fiction or a true history of twenty-five hundred years of philosophy. Either way, it offers the advantage of identifying the nerve of the complex debates of the whole of Western history. The master theme of philosophical dispute, I suggest, lies with this single question: whether reality has an invariant structure or whether it is a flux. If it is invariant, then any human aptitude for knowledge—for science, history, art, interpretation, morality, public policy, religion, education—must be governed and guided by constraints congruent with that supreme constraint; but if reality is inherently a flux—however stable it may be for inquiry, action, creativity, production, judgment, and government—then only the contingencies of human thought can determine what to make of reality and our capacity to understand ourselves and the world, and how to use "both" rationally for constructive purposes. There is, then, no principled demarcation between what is real and what is interpreted to be real. There you have the sense of Protagoras' original dictum: "Man is the measure." For man belongs to the flux he measures. Of course, I say this at the very start of our study. I have no right to suppose you will feel the least bit obliged to agree with me.

Over a span of twenty-five hundred years, it would be surprising if the meaning of this message had not taken a changing shape that gathered to itself the most compelling arguments it could find leading down to our own late day. I may perhaps risk, therefore, a drastic economy and declare straight out that, at the end of the twentieth century (also, foreseeably, at the start of the twenty-first), many strong philosophical currents have either embraced the following doctrines or now admit that they are, singly and conjointly, internally coherent and justified in claiming some plausibility:

1. Reality is cognitively intransparent: that is, all discourse about the world is mediated by our conceptual schemes, and there is no way to tell whether what we claim about the world directly "corresponds" with what is there, in the world, independent of the conditions of inquiry;

2. The structure of reality and the structure of human thought are inextricably symbiotized: that is, there is no principled means by which to decide correctly what the "mind" contributes to what we take to be the world's real structure or what the "brute" world contributes that makes its seeming intelligibility apt for our correctly representing whatever structure it has independent of inquiry;

3. Thinking has a history, is historicized: that is, all the supposed fixities, invariances, necessities, universalities of thinking and the world— for instance, of logic, rationality, laws of nature, principles of judgment and conduct—are contingent artifacts of the historical existence of different human societies; there are, thus, no necessities de re or de dicto, except in the sense of so appearing under the constraint of changeable history; and

4. The structure of thinking is preformed and self-modifying: that is, whatever appears compelling or salient with regard to inquiry or reality is tacitly formed by antecedent enculturing processes that we cannot entirely fathom, though, by participating in those same processes (as we must), we alter them, alter ourselves, and alter the conditions under which those who are yet to come will be encultured.

The extraordinary thing is that 1-4 are now very nearly commonplace (which is not to deny that they are also disputed). Taken in its most extreme form, the entire set accommodates Protagoras at a stroke and actually goes beyond him. For the most distinctive mark of modernity is associated with reflecting on the singular events of the French Revolution and coming to believe that human thought and existence have an inherently historical structure that affects our very competence to decide whether whatever we take to be true or real is timelessly or objectively true or real. In this sense, the radical notion that reality is a flux may already have won the contest. But it is true enough that very few believe it deeply: very few act and think as if the world were really intransparent, as if there were no de re or de dicto necessities, or as if the salient convictions of our world were merely horizonally confined to our own small history (or a little beyond). What remains is what usually remains: the need to work out as compellingly as possible the dialectical applications of 1-4.

In this spare sense, the theory of interpretation has, in the late twentieth century, undergone changes that could constitute a large revolution

of its own. In less than fifty years, all the supposed fixities of criticism in the arts and history have been strongly contested. The sense of objective rigor, so closely keyed to the themes that once dominated the theory of explanatory method in the physical sciences in the 1920s and 1930s, and, as a matter of course, were allowed to legitimate whatever might be seriously offered in the "weaker" disciplines (criticism and history), are now noticeably on the defensive. This is not to say that rigor is being bargained away. It is rather that it is being redefined.

In the philosophy of science, the change in conceptual orientation is quite startling. One can trace it in the early work of theorists like Gaston Bachelard and Ludwik Fleck and, closer to our own time, in Thomas Kuhn, Michel Foucault, and other historians of science—and, even more adventurously, in the recent work of Paul Feyerabend, Nancy Cartwright, Bas van Fraassen, Arthur Fine, Ian Hacking, and younger theorists. All this certainly brings us to the "edge." Still, "the revolution" remains marginalized within university and establishment circles—even as its piecemeal themes are enthusiastically embraced.

Philosophically, nearly every would-be canon has had its effective subversives: in phenomenology, Husserl has had to give way to Merleau-Ponty and Heidegger; in hermeneutics, Emilio Betti's masterful summary of the Romantic tradition has given way to Gadamer; Hegelianism has given way to pragmatism; Marxism, to early Frankfurt Critical theory; structuralism, to poststructuralism; positivism and the unity of science, to the work of the figures mentioned just above. Through the whole movement of Western thought, certain great subversives have always appeared. In the turn into the twentieth century, it is surely Nietzsche who is most honored in this regard. The greatest of these, across the centuries, have been the partisans of the flux. But our own late period has managed to produce a small army of freethinkers who have opposed, often irresponsibly, the presumptions of the would-be canon—and, within the space of interpretation and its theories, they have sought to reverse the vision of objective rigor that, as recently as the 1950s, was presumed to be in place in New Criticism, Romantic hermeneutics, and Francophone structuralism.

The wildest, most daring, possibly the most suicidal of these late thinkers have been the French philosophers: Derrida, Bataille, Barthes, Lyotard, Deleuze, and certain of the French feminists. The best of these is very probably Foucault, who was neither a mere wild thinker nor one to ignore the detailed connection between the newer views of the history of thinking and the strongest systematic conceptions of sci-

ence. In a true sense, therefore, the possibility of arriving at a comprehensive and perceptible revolution depends on the effectiveness with which the compelling subversion of the standard philosophies of science and logic may be joined to the compelling subversion of the standard theories of interpretation and history—in the name of the flux. It would not be far from the mark to suggest that, subliminally, many of the best minds of the day are busy testing that vision. It is even possible that never before has the complete subversion of the canon—of the ancient and modern canon, committed at least to the invariant structure of reality and the work of rational inquiries set to discern (said to be capable of discerning) such structures—been on the edge of being completely toppled. Doubtless, the canon will not be simply toppled. But the strengthening of the contest between these antagonists, the improved position of the partisans of the flux, is something of a surprise. I do not say the partisans will succeed—but they cannot be ignored; and philosophical honor requires that the challenge they represent be fairly met, without dismissing their charges as nonsense or illiteracy. It is in that sense that I offer the tally of claims—1-4—that I say would, if vindicated, constitute a genuine revolution in thought. That's all.

So seen, the radical possibilities of the theory of interpretation signify a larger conceptual estate than might have been imagined. The contest lends a sense of grandeur to disputes that would otherwise seem terribly local. At any rate, this change of focus captures the spirit of the argument that follows. It is persuaded that nothing the canon once held to be beyond dispute is still conceptually safe and that the doctrine of the flux can accommodate a sense of rigor, coherence, scope, plausible detail, resilience, provisionality, humor, objectivity, and the like adequate to the best work of the canon. Furthermore, the reverse cannot be shown (so I claim): the canon (the doctrine of the invariant structure of reality) cannot make room for the best work of the defenders of the flux. It cannot regard those as anything more than risked in misguided departures from the true way. The reasons are worth identifying.

Put more directly, the new puzzles of interpretation fasten on the single question of just how to demonstrate the coherence and viability and power of understanding artworks, texts, histories, theories, the real world, human existence itself, under the condition of the flux. That argument has never been attempted in a way that rightly answers the objections of the canon's best champions. It needs to be undertaken, since the point of the effort is to enlist the talents of a redirected

imagination: to produce, if possible, a sense of reasonable transition where, before, there was only bafflement and confrontation.

In the larger picture there are perhaps two decisive figures who have dominated the continuous revision of the canon, both in our understanding of science and logic and in our understanding of history and interpretation: Aristotle and Kant. The challenge intended is, in a fair sense, directed against both—or, better, against the conceptual commitment each has been taken to have made (but perhaps never actually meant to defend in the inflexible way their champions have imagined). Nevertheless, the target of the argument that follows is directed against the official doctrines of each as well as against the presumption that the objectives usually assigned their best projects may be asymptotically approached even if they were never actually completed.

In this sense, Aristotle is the advocate of the view that reality has an essential structure, a structure that is inherently changeless, that true science is directed to its discovery and governed by whatever constraints make that possible, including whatever makes the guidance of action in a changing world rational. Aristotle's canny caveat in the direction of the "possible" and the "probable" (for instance, in the Poetics ) therefore falls under the heading of the progressive, the self-corrective, the approximative, the verisimilitudinous. But if the first claim were defeated (that of strict invariance), then so would the second (the conformable objectives of an asymptotic science). Without the anchor of the invariant sciences, Aristotle's tolerance of the provisionality of the practical disciplines (notably, in the Eudemian Ethics ) would float out of "principled" control. The discipline of interpretation would be adversely affected as well. I speak here only of an intended strategy of argument: I claim to have established nothing substantive yet. Notice, however, that what holds for invariance and flux holds also for cognitive transparency and intransparency. For, invoking the distinction introduced in the tally offered just above, once we admit symbiosis and intransparency (1-4), it is conceptually impossible to treat cognitive conjectures as discernibly approximating more and more closely to the real structure of the independent world.

Kant, however, is the advocate of the view that reality is ultimately noumenal and that, under the conditions of phenomenal experience and reflection (known to be unable to capture the noumenal), we can nevertheless discover the strict necessities under which thought and intelligible world are uniquely constrained, constitutively and regulatively—hence, not symbiotically—in every sector of human interest.

(The single most decisive passage that illuminates both the strength and weakness of Kant is the preface to the second edition of the first Critique. ) Kant's admission that the noumenal world is really unknowable, and his admission that he himself was dissatisfied with all his efforts at formulating those synthetic a priori truths that the inherent structure of experience was said to manifest, have hardly dampened the (rational) hope that we could make progress transcendentally either in discerning the universal structures of the phenomenal world or in discerning the necessary sui generis structures of human understanding.^[1] But Kant nowhere satisfactorily explains just how, given our contingent cognitive resources, we can be sure that they harbor deeper cognitive powers that are not subject to the same vagaries that infect our contingent resources. That, ultimately, is what historicizing human thinking confirms.

In a profound sense, it is the convergence of these two still-vigorous (but failed) themes that has given the latest philosophical visions of the twentieth century their continued canonical force: for example, in Husserl, in the unity of science movement, in Habermas, in the unrelenting programs of physicalism and extensionalism so characteristic of American philosophy, in Saussurean structuralism. These are all at risk now, I claim. For the moment, I offer no supporting arguments. Whatever these visions take to be apodictic or invariant or universally binding, or progressively directed to grasping such structures, may be reclaimed in this or that more restricted context or may serve this or that more provisional interest. For example, the law of excluded middle is, as a logical principle, clearly not a necessary truth, as was once supposed (certainly by Aristotle and in the tradition that derives from Frege^[2] ); but no one would dream of disallowing, because the principle cannot be shown to be necessarily true, the regulation of at least a part of discourse and argument in accord with it.

Grant this much, and you will find yourself led to add a further dictum to our previous tally, namely:

5. The phenomena and entities of human culture are socially constituted or constructed, have no "natures," have or are only histories: that is, persons and selves, artworks, artifacts, texts, actions, institutions, societies, words and sentences, and the like cannot be characterized as falling under "natural kinds" (as either having assignably fixed essences or behaving nomologically in ways that may be explained by reference merely to what would explain the behavior of physical things);

they have only (predicatively), or are only (referentially), histories, narratized careers of a distinctive sort—ordered compatibly with the arrow of physical time but subject to forms of change peculiar to themselves.

Now, 5 is the radical thesis in our pack, the one still relatively unfamiliar and uncertain, the one still uncompromisingly disputed, denied, resisted, ridiculed, ignored. Were it to be admitted to full parity with 1-4, we should have gained a long march on the revolution sketched. There would then be no doubt that we were close to the "edge." The intriguing fact is that there are now strong theorists who embrace 5. Certainly, in theorizing about literary and historical interpretation, one cannot fail to find a clear advocacy of 5 in Roland Barthes, in Michel Foucault, in Harold Bloom, and also, somewhat less boldly but in a way more robustly, in the work of Hans-Georg Gadamer. There are others, of course. But in these four, it appears in a ramified, original, and decidedly influential form. This is not to accept everything any of them might say about 5. It is only to say that a very strong beginning has been made in our time to confirm the flux—even if, here and there (notably, in Gadamer and Thomas Kuhn), a certain symptomatic uneasiness about just where the doctrine leads also appears. Remember: I am merely reporting (here) the signs of the time. I am not even insisting on the adequacy of the philosophical work of these figures. (They are plainly controversial.) In any case, it is in the theory of interpretation as applied in the arts and history that the boldest gains favoring 5 have been made. It may serve to forestall misunderstanding (but not dispute) if I add a few further anticipatory distinctions here.

For one thing, I mean to speak of flux (or the flux) in a relatively uncomplicated way, as designating the reality of a world of change. In this I oppose two ancient doctrines: (a) the Parmenidean claim that change is simply unreal ("Is Not"); the other, (b), Aristotle's claim that, necessarily, change is real only if it is predicable of what, among real things (ousiai ), is changeless in them (predicating potencies with respect to their essences, or predicating attributes only in accord with a bivalent logic). I speak of the "flux," then, only in the sense (c) that a changing world may manifest stable and discernible structures, even structures that appear not to change over time, without its being the case that their apparent "invariance" is a necessary feature of the world (de re) or of our discourse (de dicto).

For a second, I mean to speak of the things of the world as arti-

facts or constructions in two different but compatible senses: (d) in the Kantian-like sense that the experienced world is "symbiotized" but only "intransparently," so that one can no longer claim (with Kant) that what, disjunctively, transcendentally, the understanding mind contributes to the intelligible structure of the phenomenal world is necessarily and discernibly invariant; and (e) in the relatively uncontroversial sense in which, within the terms of (d), some things are made or produced (artworks) or appear as what is done (deeds, actions) by way of specific human agency. This again catches up a classical distinction (Aristode's, regarding poiesis and praxis); but its more important instruction rests with the fact that what is "constructed" or "constituted" in either or both of these senses is not for that reason unreal, fictional, heuristic, imaginary, or "ideal" (in the philosopher's jargon). What is "brute" in the world remains acknowledged; but all post-Kantian thinking must come to terms with the benign paradox that what we take to be independently real (real apart from our inquiries) is itself an "artifact" of those symbiotized inquiries. Kant is already committed to a constructed world; against Kant, however, I read that same thesis in accord with symbiosis and the doctrine of the flux. That is what I mean by "Kantian-like" or "post-Kantian."

Finally, I mean to treat all discourse as contextual or contexted, in the sense (f) that discourse may be said to address an intelligible world "constituted" by human agents who are themselves artifacts of a culture that changes and lacks necessary invariance; and (g) that, relative to any such world, discourse could never convincingly pretend to master all possible conceptual schemes pertinent to truth-claims about such a world, and cannot, relative to any conceptual scheme, uniquely fix all the referents of our truth-claims. Whatever we say, therefore, is "contexted"—in the deep sense that reference is logically informal and inherently affects the discernibility of truth, and that predication reflects the discursive powers we acquire only contingently in becoming the culturally apt creatures we are. My point is that contexted discourse is ineliminable, insuperably Intentional, and intrinsically interpretive.

Now, then, the gist of the argument that follows is this: if you grant these last claims, (c)-(g), together with my first themes 1-4, and then add 5 (social constructivism), the most radical theories of interpretation favored at the end of our century but often thought incompatible with acceptable standards of philosophical rigor will prove entirely coherent and viable. Furthermore, given the conceptual tastes of our age,

I doubt that these most radical theories can be convincingly resisted for long.

II

Let us, however, set aside the red flag of "revolution." It is true that the items tallied—l-4—form a substantial part of the structured space of current philosophical argument, and it is true that they become increasingly radical and contested in the order in which they appear in the tally. But we may also view the puzzle of interpretation in terms of a number of smaller-gauge questions. In fact, the argument that follows rests essentially with a series of heterodox analyses of certain familiar notions: in particular, those of reference and predication, intentionality and intentionally complex properties, individuation and numerical identity, the nature of cultural entities, history and the historical past, and description and interpretation. I intend to explore all of these themes congruently with 1-4; but it will not be possible to give a full account of each. The running argument will support the view that all the linguistic distinctions mentioned—reference, reidentification, predication, interpretation—ultimately depend on a society's consensual memory of its ongoing practices: proving that would confirm that such matters were contextually bound to a society's historical experience in ways entirely opposed to standard views about the nature and description of physical entities. That would not be a negligible conceptual gain.

The treatment of physical and cultural phenomena would need to be reconciled, of course. So there would be a confrontation to work through. It would be seriously misrepresented, however, if it were put entirely in terms of that overworked contrast between so-called "analytic" and "continental" (or "continental European") philosophy. Everyone in the philosophical community has had to come to terms with those terribly muddy epithets. One has only to recall that Frege, Car-nap, Wittgenstein, Brentano, Meinong, and Husserl were at once "analytic" and "continental" philosophers and that what represents Anglo-American analytic philosophy best would be quite impossible without the contribution of these and similar thinkers.

No, the point of the contrast is to signal, obliquely, the strong programmatic disjunction between "analytic" philosophy thus augmented (think of Tarski, Hempel, Duhem, von Wright, and an army of others) and a loosely collected company of thinkers, chiefly contempo-

rary, whom—quite frankly—the "analysts" take to be dubious as philosophers. Chief among the latter are: Heidegger, Gadamer, the Frankfurt Critical group, Lukás, Merleau-Ponty, Sartre, Ricoeur, Derrida, Lacan, Deleuze, Lyotard, Foucault, Levinas, the French feminists, and similar-minded authors. Here and there, the harsh judgment has been attenuated. Merleau-Ponty and Foucault sometimes appear on the edge of respectability; Habermas and Apel may have made the grade. But the deeper complaint has more to do with discerning an alleged inability, among the authors mentioned, to "formulate a philosophical argument." It is a fair complaint and it will have to be fairly met.

Now, the notions collected a moment ago will form the target of the intended argument. I mean to defend a series of claims about reference, predication, intentionality, identity, and the like that are particularly apt for a systematic picture of interpretation in the arts and history. Wherever possible, the account will be reconciled with what is familiarly claimed for the physical sciences; where that is not possible, a rationale will be offered to justify the departure from "standard" views. Even here I shall have to pick and choose: more issues will arise than may be managed in a small space. Furthermore, I shall draw quite openly on the work of the doubtful figures mentioned a moment ago. But I should like to assure you that I intend to abide by two reasonable constraints: first, that whatever I endorse of such work can be shown to be philosophically rigorous; and, second, that the rigor of the "analysts" and the rigor of the "continentals" (not their favored doctrines, of course) are, among the better specimens of each, very much the same.

Several very large themes will need to be singled out particularly: intentionality may be the most strategic and the most quarrelsome of these. I shall have to economize here as well, to confine the argument to what is strictly needed and possibly a bit more. Elsewhere, it may be enough to lay out the sense in which particular terms are to be taken, so that readers may assure themselves that they have the lay of the land. For example, it may be enough to say, to distinguish between description and interpretation, that the difference between the two is not, as such, a difference between two kinds of speech act. Interpretation is chiefly concerned with certain predicates and the conditions under which they are rightly attributed to the referents of our discourse; whereas description is thought to yield statements or assertions that take bivalent truth-values or logically weaker values, or values ("accurate," "precise," and the like) that are more eccentrically

linked to the assignment of full-fledged truth-values. Obviously, then, interpretations may be descriptions and descriptions may or may not be interpretations. Also, we have as yet no reason to suppose that interpretation applies only to what is antecedently describable in non-interpretive ways. It is possible that statements that, for logical reasons, are interpretations may become so regularized that they are henceforth treated as descriptions. Still, on the argument being sketched, description must be interpreted at some level of analysis.

Something rather similar happens when value judgments (medical and legal judgments for instance) become so regularized that we treat the statements they subtend as factual statements. For the fact/value dichotomy is similarly a mixed classification: value judgments (moral, aesthetic, legal) are concerned with the use of certain predicates ("tubercular," "homicidal"); factual judgments, however, are bearers of truth-values regardless of whether their predicates are normative or not. We may leave the matter there for the time being. I believe that (in saying this much) I have prepared the ground for an important gain.

Reference and predication raise more difficult questions. What needs to be grasped is that neither reference nor predication becomes conceptually more difficult (than what would hold in physical contexts) merely by invoking the complexities of the cultural world. That may not be readily perceived or conceded. I shall claim that the individuation and reidentification of all the particulars of our world depend on our context-bound practices of reference and predication; furthermore, that, if we can rely on our consensual memory regarding these matters, then it cannot be logically necessary that the "natures" assigned to particulars be held invariant in order to succeed at reidentifying them. The constancy of their "natures" (I suggest) will be a function of consensual practice, and the constancy of their identity (in the face of a change of "nature") will be similarly ensured. This last suggestion will prove immensely important in developing the account of interpretation that follows. For the moment, I insist only that it is a coherent—and neglected—possibility. It is, also, plainly heterodox. Thus, if it were granted, it might well signify (I take it to signify) that there is a certain insuperable indeterminacy in determinate discourse itself that adversely affects every presumption of universality and invariance and that constrains consensus and communicative success in local, praxical, and contexted ways that are not at all the marks of contingent ignorance. (Here, I may say, I find the strongest intellectual support in

the originality of Charles Sanders Peirce and Wittgenstein and, in more historicized ways, in Marx and Foucault.)

Spelling matters out this way confirms the sense of the contest I say we cannot escape. Interpretation and history are bound to affect all the notions I collected a little while ago—if only we construe them in accord with items 1-5.

The most strenuous notion remains: that of intentionality. Fortunately, not a great deal needs to be said about the "intentional" or the "intensional." I shall read intensional in the standard way, as signifying the "nonextensional." The intensional is a feature of language initially, or at least preeminently, but not exclusively. It designates any form or structure of meaning, significance, sense, symbolic or semiotic or rhetorical or similar function or role assigned to a suitable vehicle (a sentence or semaphore signal or artwork or action or custom or text— or thoughts, if thoughts may be singled out). Such functions may not be merely structural or grammatical or syntactic in the sense that they may exhibit features that cannot be suitably analyzed in terms conformable with the constraints of the truth-functional connectives of first-order logic (for instance, as in the causal use of "because" or the use of the temporal prepositions "before" and "after"), or in the sense that they cannot be said simply to collect the semantic and pragmatic aspects of speech and related communicative vehicles disjoined from or added to an antecedently specified syntax. The intentional may be read, again more or less standardly, in the sense of "aboutness" developed by Brentano and improved by Husserl. I shall use it more in the Husserlian than the Brentanoesque sense, that is, as more concerned with the significative (or "intentional" or "noematic") structure of a propositional attitude (desiring, believing, or the like) than as concerned with so-called "intentional" (or "inexistent") objects of particular mental states (as in the fear of horses). The intentional is initially, or preeminently, a mental or psychological attribute or an attribute applied to thinking or cognitive activity, but not exclusively. The "in-tensional" and the "intentional" are quite distinct: each may occur without the other, though they are frequently combined, as in ordinary verbal expressions of belief. I assure you, however, that my use of these notions will be relatively standard.

I also introduce a term of art, the Intentional, to designate certain further features of interpretive discourse and interpretable phenomena generally ignored in the analysis of the more familiar notions. It is not an established term linked to the usual distinctions regarding the

"intentional." The "Intentional" (I shall say) incorporates both the "intentional" and the "intensional" insofar as they arise as features of culturally formed phenomena. The Intentional need not be confined to the mental or the linguistic (as in music, for instance, or conventional behavior). Broadly speaking, the Intentional = the cultural; it is characteristically articulated intensionally in phenomena or activities that implicate the intentional. It belongs primarily to the collective life of historical societies, and it appears as an ingredient in the properties of artworks, texts, institutions, traditions, actions, histories, theories, personal careers, linguistic utterances, customs, practices, and the like. So Intentional properties are to be regarded as real enough: they are not merely the marks of a façon de parler. I regard Intentionally qualified phenomena as inherently apt for interpretation; they are the possessors of intrinsically interpretable properties. Hence, then: "intensional," "intentional," and "Intentional" are to be used predicatively. Accordingly, where the properties they collect are ascribed veridically, questions arise about the "nature" of what exhibits them, since, as we suppose, there must be some appropriate fit between referents and their attributes. Furthermore, introducing the Intentional permits me to escape the decidedly solipsistic tenor of so much of "analytic" philosophy—regarding both the strong treatment of the intentional in Brentano and Husserl (and among their followers) and the dominant themes of Anglo-American philosophy featuring the psychological or the biological sources of psychological or "mental" competence (preeminently in cognitive claims). Interpretation, therefore, will be seen to play a most strategic role in reorienting inquiry in philosophy and philosophically sensitized sciences. Plainly, if the Intentional is as ubiquitous as I suggest, then interpretive questions will arise in all forms of discourse: it will not then be possible to segregate the physical and human sciences, for instance; all the sciences will be "human sciences." This gives a fair inkling of the potential scope of our topic, though I shall address that issue only obliquely. (Still, as I also say, to speak of what is real, within the terms of the flux and intransparency, is to speak interpretively.)

Realism regarding human cultures entails a realism regarding Intentional properties. There you have the principal, potentially quarrelsome, novelty on which the solution of the puzzle of interpretation rests—the one I intend. Everything depends on how the admission of the Intentional affects reference, predication, numerical identity, the "nature" of cultural things, the process of history, and interpretation

itself. The Intentional (I say) never occurs except as inseparably embedded ("incarnate") in physical or biological properties. I treat it as the interpretive feature of an indissolubly complex property—or simply as that property.

The entire argument that follows construes interpretation as addressed to the Intentional—in accord with 1-5 and under the assumption of the flux. My claim is that these conditions allow for the coherence and viability of a radical theory of interpretation and that that theory is more and more favored as we near the end of our century. The rest of the argument concerns how certain standard theories in philosophy and interpretive criticism are affected by the change considered and how broadly it may be applied.

I must also say that the argument that follows does not proceed in the manner usually favored by familiar philosophical examples. It is more oblique than that, because it means to recover the complexities of interpretation that the usual exemplars make no provision for. The indirection intended is more in the nature of a piece of tact than of a loss of discipline. It is simply impossible to find in the usual analyses of the philosophical topics mentioned—reference, reidentification, excluded middle, and the rest—sufficiently sensitive clues regarding the interpretation of paintings, historical events, individual careers, for instance, that would attract and persuade provisional practitioners of the interpretive arts in question. I am as interested in convincing the latter of the advantages of a change of interpretive model as I am of convincing philosophers that that change jeopardizes nothing that conceptual rigor would require.

As I see the matter, the single most strategic issue concerns the demonstration that discourse about the arts and history requires resources of reference and predication at variance with standard views that take physical objects as their paradigms. The actual discernible practices are the same, I am convinced, but they have been characterized in such a way as to make it extremely difficult to understand just how discourse about the arts and history could possibly function successfully if their pertinent referents did not exhibit "natures" very much like the natures of physical objects. What I wish to show is that these essential practices remain palpably coherent in accommodating the distinctive, even radical, possibilities of interpretive discourse. I believe the argument is widely thought to be impossible to defend. In any case, I have never seen the objections pursued in a sustained way. The question deserves an inning.

Let me add two final first qualifications. The first concerns the great scatter of views of what "interpretation" comes to. It is a fact that, in the philosophy of the physical sciences, the dominant tradition has long held that science aims at what is uniquely true of the independent natural world. Under conditions of relative ignorance and the absence of full evidence—perhaps ultimately irremediable—theories purporting to yield an accurate account of the world, or such an account suited to the rigorous explanation of the way the world behaves, are said to be "interpretive" in the manner of a faute de mieux strategy. This, pretty certainly, is Carl Hempel's view, and it has served analytic theorists of science well, even to some extent in the second half of the twentieth century. But many theorists of the physical sciences now believe, for a variety of reasons, that there is no principled sense in which that vision can be convincingly sustained: thinkers like Kuhn and Hacking and van Fraassen—whom I have already mentioned—emphasize one or another of the items of 1-5. They therefore treat interpretation as ineliminable in principle in the cognitive work of the physical sciences. Some hold (van Fraassen, for instance) that, since strict correspondence is not accessible, interpretation plays an ineliminable supplementary role in description, explanation, and prediction, once "empirically adequate" (but not uniquely true) theories are formulated; others (Kuhn and Hanson and Feyerabend, for instance) treat even empirical observation as theory-laden—hence, as inherently "interpretive." van Fraassen, I suggest, is inadvertently the most instructive of these, for van Fraassen admits that there is no theoretical disjunction between the observational and the nonobservational. But if there is none (as my original tally requires ) , then the function of interpretation in the sciences cannot be restricted (as van Fraassen believes) to a merely supplementary (however important) role, once theories "empirically adequate" (but not uniquely true) for preserving the data of observation are supplied.^[3]

My thought here is that the role of interpretation in the sciences is, under the constraints of my tally, exactly the same that obtains in the arts and history. That comes as a surprise. If, therefore, one denied that science seeks to capture, or at least approximate to, what is true of an altogether independent world (a world not constitutively affected by our conception and inquiry), one will have made room for an ineliminable interpretive function at every stage of cognitive inquiry. Furthermore, I should insist (against van Fraassen and an army of others less concessive than he is—Popper for instance) that, if we

cannot directly know what is true of the independent world, then we also cannot know what empirically "approximates" to (what is "adequate" to, though not yet true of) that independent world. Noting how (and why) that distinction collapses, I say, strengthens my claim about the presence of interpretation in all inquiry and science. (I shall later bring this finding to bear on the views of Davidson and Rorty, for example, as well as, more narrowly in the context of the arts and history, the views of Arthur Danto.)

The second qualification I have in mind concerns the relationship between theory and practice. For once flux, symbiosis, and intransparency are admitted, then, contrary to Aristotle, there cannot be a principled disjunction between theory and practice: theory is then a form of practice under the conditions of historical life. Not only does this ensure an essential role for interpretation in the practical disciplines (particularly with regard to the arts and conduct), it also marks the strong sense in which there cannot be a realist reading of physical nature if there is none for human culture. My own suggestion is that determinate reality—whatever we can make of it—is an artifact of the consensual practices of viable human communities, but not for that reason not "brute," not "external." I take this to be the minimal theme of the classic forms of pragmatism, which, however attenuated, has persisted to the present. It also draws attention to a neglected dimension of human thought (which I had barely remarked before), namely, that thinking is historically and collectively formed, and that, as a consequence, the entire world is open (in various senses) to interpretive work—to the extent, that is, that it is open to being cognitively grasped at all. I believe all the puzzles tethered to reference and predication make this clear. What it shows (when fully and rightly elaborated) is that our understanding of the world is, inherently, interpretive and that interpretive understanding falls within a historical tradition (which, of course, it alters by its own success). That affords the principal clue of what I mean by the "Intentional."^[4]

So let us make a start.

PART I

Chapter 1
Reinterpreting Interpretation

I

Give or take a little in the way of precision, there are at least three bits of advice that ought not be ignored in constructing a theory of interpretation of any size. First, it is impossible to disjoin the account of the nature or logic of interpretation from one's theory of the nature of what it is that may or must be submitted to interpretation. Second, there are only two sorts of pertinent theories of interpretation. One holds that interpretation is practiced on relatively stable, antecedently specifiable referents of some sort, and that the requisite account identifies the practice by which distributed claims about them are responsibly assigned truth-like values of some sort; the other holds that interpretation is a productive practice by which an entire "world" or what may be distributively referred to in that world is or are actually and aptly first constituted (not ex nihilo or by pure fancy but by Intentional technologies, by painting and sculpture, for instance) for certain further claims or use, possibly for interpretation in the first sense—but in ways that may be affected by referential puzzles due to the ongoing interpretation or reinterpretation of what has already been "constituted." These are not yet theories in their own right, I admit, but they are remarkably economical directives about what to explore. (They are no more than promises, of course.) The first sort of theory identifies the traditional genus of interpretation; the second is notably, even peculiarly, fashionable in our own time and is sometimes thought to disallow theories of the first sort.

The second option, separated from the first or some analogue of the first, cannot possibly be right, for the simple reason that there is no socially sustained discourse that does not provide for orderly reference and predication. Since that is so, the extreme contemporary worry that we may have to abandon altogether statement, assertion, judgment, claim, and the like on the grounds of the need to avoid any and all forms of cognitive privilege or transparency is a conceptually extravagant retreat that misses the point of the ineliminability of effective reference. One cannot do without reference, predication, description, interpretation, explanation, analysis, evaluation, although saying that disallows nothing in the way of arguable views about what may be described or interpreted or how description and interpretation actually proceed. The point would be entirely trivial except for the annoying fact that it is no longer unusual to hear it denied or implicitly rejected.

In any case, the first sort of theory is the classic one. It admits the complexity of what we interpret, but it does not extend the notion of interpretation to include the very constitution of that. It ranges over properties, but it does not extend to realism itself. In contemporary exchange, however, the cognitive intransparency of the world and the symbiosis of world and word oblige us to make room for theories of the second sort. (Bear in mind that I am using the terms "intransparency" and "symbiosis" in the sense marked in the Introduction. I do not deny that that sense may be challenged.) Characteristically, the work of such theories is inseparably linked to the viability of the work of theories of the first sort. You can appreciate, therefore, that the barest beginning of an account of interpretation plunges us at once into a conceptual swamp. For how can we interpret what has yet to be constituted and how can anything be constituted by way of interpretation? Still, there are no interesting theories of interpretation in our own time that do not—or will not—consider combining both senses of "interpret."

The third bit of advice reminds us that, whatever the slackness of linguistic usage, what are interpreted in either of the senses given are distinctly cultural (Intentional) phenomena of some sort, interpretable just in virtue of their having cultural features or because they are treated as having such features or because they have features sufficiently like cultural features to warrant being similarly treated.

These are all, of course, deliberately elusive but quite safe initial pronouncements that convey an air of imminent system and scope and a promise of detail that a streetwise audience is likely to be polite and

patient enough about while awaiting full delivery from the vendor. Also, it is not likely to be ignored that, regarding all three bits of advice taken together, the first sort of theory of interpretation identifies referents conceptually apt in some antecedent sense for interpretation, whereas the second sort of theory treats interpretation as a process of actually constituting things by interpretation, by a constructive activity by which certain phenomena or entities are first and merely posited. There need be no incompatibility or equivalence between these two sorts of theory. The kind of contribution the first might make can hardly be supposed to be entailed or precluded by the work of the second; and the point of the second would be entirely lost if it did not accommodate some form of the work of the first. Both are keyed to what would be distinctive of any plausibly realist account of the cultural world in a way that need yield nothing in favor of, or against, the reality of mere physical nature. The simple point is that to say that reality is "constituted" is to say at least that there is no principled distinction between the structure of the world and the structure of our conceptual schemes that must be in place prior to inquiry or that is reliably disclosed, within inquiry, to be such. It is entirely open to us (presumably that is the work of science) to propose (interpretively) what, within the terms of inquiry, we should treat as real or "independent" of inquiry. Hence, to admit that the world is in some sense artifactual is hardly to deny that its structures are real or to affirm that it is a mere fiction. (This is an accommodation of the supreme Kantian thesis. Of course it need not adopt what I construe to be Kant's undefended alternative to "symbiosis.")

I have moved rather quickly, here, into the neighborhood of the principal puzzles of interpretation that have taxed the most recent accounts in the arts and history, without much of a preamble. In that sense, what has so far been said is no more than a small clue, a warning of matters we must return to in good time. Still, there is an advantage in its being made at once clear that the question of the logic of interpretation is now very different from what it had been taken to be as recently as thirty or forty years ago. Also, this local change depends, we may anticipate, on larger conceptual changes that have affected the whole discourse of philosophy and science. In focusing on interpretation, then, I believe I am featuring the most strategically placed cognitive ingredient of the entire set of conceptual questions that confront us in a fresh way at the end of the century. What they are and how they are linked to the puzzle of interpretation have yet to be told. They

invite a certain patience and generosity of vision. For the idiom that favors the older account has not yet accommodated the new puzzle, and the sympathetic formulation of the new puzzle must bring in its wake a new vocabulary capable of escaping the limitations of the other. My "advice" is meant to mark what is salient in the transition.

At the very least, then, it comes to this: we must make provision for the assignment of truth-values in interpretive contexts, although to say that says nothing (as yet) about the nature of interpretable things or the conditions under which truth-values are rightly assigned. We must consider how interpretable things are constituted, seeing that they possess properties (what I am calling Intentional properties) that are not normally assigned physical objects (that are also both "brute" and "constructed"); and we must make provision for the sense in which both interpretable things and the interpretation of them obtain in the emerging "space" of real cultures. At bottom, therefore, the central questions are these: (1) in what sense are cultural phenomena real—as real as physical phenomena? and (2) what are the comparative advantages of a theory of interpretation committed to the flux rather than to invariance? Also, of course, if we admit that both cultural and physical phenomena are interpretable (which I support), then we are bound to account for the differences and similarities between the two practices.

Let me suggest that the first sort of theory is adequational, meant to assign (for the purpose of ensuring objectivity) a "nature" to the referents of our discourse, such that their features (their "natures") would be conceptually and evidentially congruent with our making and supporting interpretive claims about them. The adequational question is the question of the conceptual fit entailed in predicating this or that of things of this or that nature: whether, say, "smiling" may be falsely predicated of stones as of people, or whether it involves a sort of conceptual impropriety.1 Given the warning already collected, there is no reason to suppose that there is anything illicitly privileged in formulating an adequational theory. This is not to say that there is no metaphysical or epistemological bite to such a theory, only that it cannot reasonably be supposed that every metaphysics or epistemology necessarily violates the common injunction against privilege. Otherwise, since such discourse cannot be avoided (that is, discourse that raises the adequational question), and since it cannot proceed without a stable practice of reference and predication, it appears that it cannot fail to yield metaphysical and epistemological findings, privileged or not.

If one relativizes such findings to the saliencies (Erscbeinungen )

of our shared world, one may reject the tricks of cognitive privilege without giving up the benefits of an adequational theory. The point is modest enough, though enormously important. Indeed, it is nearly universally ignored. What it signifies is that the admission of reference and predication is the logical or formal admission of a need for the processes of description and interpretation: it is the admission that a world apt for interpretation must be stable enough to support such processes ("adequation"). It is a complete non sequitur, therefore, to suppose that admitting that much is tantamount to admitting some further metaphysical or epistemological privilege or fixity. Correspondingly, we may characterize the second sort of theory as constructive in the sense that we and the things of our cultural world may be taken to be constituted somehow, possibly serially reconstituted, as what they are, or thereby become, as a result of some initial (profoundly tacit or deliberately productive) interpretive act ("symbiosis"). Admittedly, the idea has an alien ring. But you may see in it the prospect of reconciling the requirements of discourse with the admission of the flux and, because of that, the ineluctability of metaphysical and epistemological questions ("adequation") otherwise fashionably dismissed (for instance, by postmodernists like Richard Rorty2 ) as both dispensable and unavoidably committed to some claim of cognitive privilege.

The reason for all this care is plain enough. Interpretation, in the strong sense in which it is practiced in the arts and history and human sciences, addresses the interpretable features of real things. But those features or properties are frequently treated as dubiously real. If they were spurious in this sense, then, since the adequational question would no longer seriously obtain, interpretation could never rise beyond a purely heuristic or rhetorical effort; hence, questions of objectivity and truth-value would have to be abandoned. However, if interpretive concerns were robustly sustained in the arts and history, then some sort of adequational account could not fail to be needed: some form of cultural realism would be required. Dialectically, so far, the matter remains neutral to the prospects of reduction and emergence.

Interpretation in the adequational sense must be referentially reliable, though it hardly requires, for that reason, that referents have fixed or unchanging natures; and interpretation in the second sense specifically admits an initial production, or a constitutive change in the nature, of things by virtue of some as yet unspecified activity, though it hardly requires, for that reason, that what is (thus) constituted be altogether lacking in natural properties—or be only doubtfully real.

Clearly, there is no reason to suppose there is a univocal sense of "interpret" that usefully serves both theories at once.

But if an account of interpretation may be fashioned for both sorts of theory—which seems both promising and generous—then it would be a considerable convenience to be able to identify, by the same term, the referents addressed in the first sense of "interpret" and whatever may be constituted in the second sense. Call such referents texts. Recapitulating what has already been said: texts and interpretation in the first sense must be adequated to one another; the ways in which texts are constituted yield referents apt for interpretation in the first sense; and texts are constituted as such by some suitable cultural activity, by interpretation in the second sense. Certainly, these very neat adjustments do not yet produce any noticeable incoherence or contradiction; and yet, they have somehow reconciled (on paper at least) the two senses of "interpret." I count that a welcome gain.

It helps to add that description and interpretation are not, in any obvious sense, alternative species of a genus of speech act or assertion. For "description" signifies at the very least the accuracy or validity of what is predicated of this or that referent, whereas "interpretation" signifies the imputation, to this or that referent, of attributes or predicables of a certain apt sort (Intentional attributes). Description concerns, we may say, the truth or falsity (or accuracy or aptness) of what is predicated, whereas interpretation (in the first sense) concerns only the use of a certain set of predicates. So interpretations may be descriptions, though they need not be; and both description and interpretation raise the adequational question. Putting matters thus avoids the inflexible view (favored in New Criticism and Romantic hermeneutics for instance) that interpretation presupposes description in a strong sense that disjoins the predicates of description and interpretation, or that the predicates of interpretive claims are logically identified with those that noninterpretive description admits.3 The advantage of these caveats will become clearer as we proceed.

This is all very general but still noticeably tighter than our first intuitions. I may perhaps add one further, quite preliminary distinction to save a little time down the road. The only other general constraints we need to impose on what we take texts to be—in order to accommodate interpretive discourse of a suitably comprehensive sort—are these: first, texts must be taken to be sufficiently unitary, in a logical sense, that is, individuatable and (thereupon) reidentifiable numerically, though this hardly settles the question of bow unified in a substantive way,

how fixed or unchanging, their internal "natures" must be; and second, however unified, variable, alterable, even enlarged or affected they may be or become as a result of ongoing interpretive activity, their "natures" must intrinsically include attributes of a suitably cultural sort (Intentional attributes) that render them apt for interpretation in the first sense ("adequation") and that account for their peculiar alterability and openness in the second ("flux"). They could not be unitary without some internal unity: so the warning about the distinction between a metaphysics and epistemology of privilege and a metaphysics and epistemology of discursive practice is well taken. But equally, they could not be texts adequated for interpretation if they did not intrinsically possess linguistic, language-like, semiotic, symbolic, representational, expressive, rhetorical, intentional, or similar properties: these are indeed just the sorts of property interpretive theories of the first sort and metaphysical theories of the second take for granted. My thought is that, metaphysically, texts are, in some important regard, intrinsically "indeterminate" (or, residually indeterminate) but determinable (in the productive sense) by interpretation. The hint comes from Peirce's extraordinarily perceptive account of vagueness and indeterminacy, though I am not prepared to subscribe to Peirce's own optimistic, evolutionary resolution of the puzzle of indeterminacy.4 In any case, the individuation of texts takes precedence over their reidentification and accommodates their "nature's" changing (without affecting their "unicity," their singleness as distinct "careers") under the practice of interpretation and reinterpretation.

I suggest we say, emphatically now, that texts possess Intentional properties intrinsically.^[5] Here again, I am trading in promises—this time, regarding a matter known to be difficult and contested. I shall return to it in a more effective way later, but for the moment I mean only to flag certain interlocking issues that must be grasped together. (I have already sketched, in the Introduction, a number of the distinctive features of the Intentional.) I should perhaps add that, by "unitary"— clearly a term of art—I mean to designate that feature of the changing career or persistence of a text (or anything else, for that matter: a tree, say) that remains one and the same through change. Bear in mind that the generic "nature" of any particular thing cannot, logically, determine its numerical identity (since, of course, many different individual things may share a common "nature"). The idea of the "unitary" accommodates that fact but also (on the argument intended) further facts about the persistence of things (texts) that lack a "nature" (in the

sense in which, say, camels are said to belong to a natural kind). I hasten to add, first, that, in calling things "texts" (accommodating symbiosis and the sense of "interpretation" that answers to the constituting function of locating determinate things within a symbiotized space), I do not mean to favor philosophical idealism or to construe "real" things as mere fictions; and, second, that, in distinguishing between the "unitary" and the "unified," I mean to draw attention to the fact that the logic of individuation is quite different from the logic of identity and reidentification.

A theory of interpretation, then, is a theory that: (1) accounts for originally constituting or reconstituting texts as such by constituting would-be referents possessing intrinsically Intentional natures; and (2) accounts for the interpretation of such texts in virtue of which pertinent claims about them may be assigned truth-like values and may be duly supported in an evidentiary way. It is a matter of considerable importance that texts may, on an opposed theory, be said to be produced by some sort of socially pertinent labor (poiesis) that is not originally interpretive itself—for instance, on the mimetic theory. In that case— classically, of course—only the adequational and assertive aspects of interpretation are needed. In our own time, because the world's transparency has been so radically denied, actually constituting what we can address by intelligent act or inquiry requires an original "mixing of cultural labor" with the physical world (the Kantian-like theme I have adopted) or the repeated reclaiming of such a mixed world by further mixing of the same sort (the fluxive doctrine Kant could not admit).

Those already engaged in the advanced argument will find these first remarks little more than a postponement of the essential issue. Still, managing things thus suggests very strongly that any alternative option is either defective, incomplete, inadequate, untenable, unresponsive, unconvincing, not pertinent—or worse. It is not an altogether innocent observation: I have in mind theories that mean either to preserve interpretation but deny the need for an adequational constraint in discourse about art and history (but not in the physical sciences, where interpretation is said not to apply) or to replace adequation with an entirely rhetorical strategy. (I shall return to these possibilities in due time—for instance, in Arthur Danto's company.) A great deal hangs on the presumption, of course, but we could never get started on an actual theory if we stopped to examine all the arguments that might lead in other directions. I anticipate an unavoidable barrage of puzzles that will need to be met.

II

So much for preliminaries. Now for a little scaffolding. Consider two very popular claims drawn from recent theoretically minded views of texts (or, more narrowly, artworks). In one, Rosalind Krauss, pressing into service what (reflecting on the views of Roland Barthes and Jacques Derrida) she takes to be the postmodernist intention to "blur the distinction between literature and criticism," speaks of "a kind of paraliterature," that is, a literature that is now neither criticism nor noncriticism but a sort of analogue of what criticism would have been for the modernist (preeminently, for Clement Greenberg) now that the distinction between artworks or texts and criticism has been blurred:

The paraliterary space [she says] is the space of debate, quotation, partisanship, betrayal, reconciliation; but it is not the space of unity, coherence, or resolution that we think of as constituting the work of literature. For both Barthes and Derrida have a deep enmity toward that notion of the literary work. What is left is drama without the Play, voices without the Author, criticism without the Argument. It is no wonder that this country's critical establishment—outside the university, that is—remains unaffected by this work, simply cannot use it. Because the paraliterary cannot be a model for the systematic unpacking of the meanings of a work of art that criticism's task is thought to be . . . there is not behind the literal surface, a set of meanings to which [the paraliterary] points or models to which it refers, a set of originary terms onto which it opens and from which it derives its own authenticity.^[6]

Clearly, in opposing the views of theorists like Greenberg, Krauss means to dismantle altogether (not merely to reverse) the high modernist the sis in all of its forms—for instance, as it appears in T. S. Eliot's famous remark that a work of art "is autotelic" and that "criticism by definition is about something other than itself" (a notion Eliot considerably changed in due course);^[7] or to oppose the thesis of New Criticism—for instance, as it appears in Monroe Beardsley's so-called Principles of Independence and Autonomy—"that literary works exist as individuals and can be distinguished from other things" and "that literary works are self-sufficient entities whose properties are decisive in checking interpretations and judgments."^[8]

Krauss's thesis is at least a first specimen of what I have called the constructive or "second" sense of interpretation, although to admit that is neither to support her particular thesis nor to suggest particular weaknesses in the second sort of theory as a result of weaknesses in her version of the theory. What is easy to miss, what Krauss misses, is that: (1) the rejection of the fixed disjunction between criticism or in-

terpretation and text or (2) the rejection of the fixed nature or fixedly bounded nature of texts independent of particular judgments or interpretation is not tantamount or equivalent to (3) the rejection of any functional (or logical) distinction between criticism and text—or, of course, not tantamount to simply rejecting all reference and predication. This may not be obvious, but it bears on the difference between description and interpretation and the logic of reference. It also bears, by example at least, on the distinction between postmodernism and poststructuralism.

Roughly put: the "paraliterary" need not—indeed cannot, logically—disallow, at any moment at which it is pertinently pursued, distinction (3). The very nature of assertion forbids it. Krauss risks—the evidence of her extended discussion indicates that she more than risks, she actually loses—the point of the paraliterary insertion itself. When she says that what is "left" is "drama without the Play, voices without the Author, criticism without the Argument," we may understand that she thereby opposes the pertinence or adequacy of modernist theories of the would-be referents of criticism—the objector distinctions (1) and (2); but, in dismissing them, she must hold (so must we all) to some version of (3), the logical distinction of paraliterary comments (criticism or interpretation, if you like), if, as she obviously also means, she means to speak (and does speak) of the work of Duchamp and Pollock and Stella and Serra and LeWitt and others. She must preserve reference, in short. She apparently does not grasp this and wrongly attributes the same opposition to Barthes and Derrida. I do not wish to deny that there are other uses of language besides that of making statements; but discourse concerned with truth and falsity cannot, I contend, avoid the resources of reference and predication. The matter is logically but not philosophically trivial. (I acknowledge that many philosophers—in our time, Quine, notably—have thought to retire reference in favor of supposedly adequate predicative resources. But, quite apart from the success or failure of such maneuvers, the intended work of ordinary reference is certainly admitted—for instance, by Quine himself. Krauss risks, therefore, impoverishing our discursive resources altogether, in a peculiarly profound way.)

Thus, reviewing a variety of postmodernist work, Krauss speaks of the "index," the "shifter," "traces, imprints, and clues," and similarly attenuated referential devices.^[9] She shows by her discourse that she cannot—she is hardly disposed to—abandon the devices of reference (captions and titles included); but her dialectical maneuvers against mod-

ernists are intended to leave the impression (there is reason to think she herself is convinced by the argument) that she has actually abandoned the logical referent we call "the play," in abandoning the high, complex, modernist entity "the Play"—as well as authors and criticism and the rest. Simply put: the logical distinction and pairing between interpretive discourse and interpreted referent is both entirely different in purpose from and perfectly compatible with the so-called postmodernist (or, more generally, the poststructuralist) insistence on denying an unbridgeable disjunction between criticism and text or artwork. The denial of the inherent fixity of texts and artworks is not equivalent to the denial of the heterotelic work of criticism; and modernist and (so-called) postmodernist criticism both rely on the same devices of reference and predication.

What Krauss fails to notice (or to acknowledge) is that the constraints of discourse, whether paraliterary or high critical, must retain an effective reference to what is unitary (is individuatable in a logical sense) sufficient for making such discourse pertinent and operative; also, that it makes no difference at all what we suppose is the internal unity of artworks, or what internal order remains when art departs from the high unity modernist usage presumes, so long as our theory (and practice) permits reference and predication to succeed. What post-modernists of Krauss's conviction confuse—which is not equivalent to an accusation against either Barthes or Derrida—is the difference between merely judging or interpreting artworks and trashing modernism, or the difference between formulating the difference between modernist and postmodernist art and judging or interpreting works of either sort, or the difference between favoring or opposing, for cause, particular theories about art of either sort.

The denial of a principled disjunction between artwork and criticism has to do with the possibility of changing the "nature" and properties of particular artworks as a result of interpreting them (or reconstituting them by interpretation). That, the so-called postmodernist theme (really more "poststructuralist"), is in complete accord with what I have termed the "second" view of interpretation. But interpretive work could never be pertinent or effective if we could not fix the referent to which it applies—even under a change of "nature." I admit that speaking of reidentification under such a change is unorthodox. That is, what is unorthodox is the idea that the "nature" of a text may be altered by being interpreted, not that a thing may change over time (without yet changing its nature). That may be what has misled Krauss.

Nevertheless, to abandon reference and reidentification is to court logical disaster.

Fussy though it may be, the quarrel is a strategic one. (I am, I confess, more interested in the ineliminability of reference than in doing full justice to Krauss's entire picture of the postmodern. What I wish to make clear is the neutrality of the referential matter to the quarrel between the modernist insistence on the fixity and autonomy of artworks and the "postmodernist" denial of such fixity.)

Krauss's intent is to reject the fixed demarcation between criticism or interpretation and text or artwork. Fine. The idea is that what, in the "paraliterary" manner, is said about a would-be artwork at time t may need to differ from what may be said about that "same" artwork at t' later than t, as a result of already baying defensibly interpreted (or commented on ) the work at t. That hardly precludes reference; it actually presupposes it. The first notion is indeed very close to Barthes's well-known contribution about "writerly" texts, but it is also entirely distinct from the matter of referential resources.

I agree with Krauss, then, to this extent: texts may be altered by being interpreted. That is a heterodox thesis, both because of the strong reading of intentionality (or, better, Intentionality) and because of the strong discounting of fixed "natures" in the cultural world. There are, nevertheless, two mistakes that Krauss commits. First of all, she wrongly supposes, in rejecting what a modernist critic offers in "a reading [of a particular painting] by [as she says] proper names," that she is also somehow committed to rejecting the need for proper names and other referential devices in critically discussing that painting or its details. For example, she shows, regarding Picasso's La Vie, a 1904 Blue Period portrait of Picasso's friend Casagemas, who committed suicide but whose portrait was modeled on an earlier self-portrait of Picasso himself, that a standard, somewhat psychoanalytic interpretation of the "meaning" of the work pretty well trades on what she herself wishes to avoid and rightly condemns as "the art history of the proper name." In context, she actually mentions and briefly discusses the principal philosophical theories of proper names and links them to what she terms disapprovingly "an aesthetics of extension."^[10] But she confuses the requirements of reference with the presumptions of (modernist) privilege. We do not automatically settle the nature, essence, or boundaries of artworks merely by ensuring that we identify and refer to them. Reference, we may suppose, is a grammatical distinction, al-

though many have taken it to have metaphysical import (or to convey metaphysical intent).^[11] But that is hardly necessary.

The truth is, the extensional function of proper names and referential devices readily obtains within complex intentional contexts (for instance, as in the fragmenting of BEAUJOLAIS in a Juan Gris collage— which counts against modernist simplification); and where, as with titles or captions, it serves to individuate an artwork, we need not suppose that the very nature, structure, Intentional detail or unity of the work is fixed or bounded by, or somehow determinately specified or specifiable in accord with, or unalterable with regard to, or unalterably linked to, that extensional function. The extensional function of proper names (naming La Vie, for instance) is not the same thing as fixing the extension of what the name names (whatever we may suppose that to be—the "painting," say); and the formal extension of a name (whatever that is) is not the same as, and does not determine, the Intentional complexities of what the name names (for example, what one or many nonconverging interpretations of La Vie may reasonably impute, synchronically or diachronically, to La Vie ).^[12] Of course, in saying this, we see that the force of these distinctions depends on what we concede to be interpretable—what I have labeled "Intentional." But for the moment, I am more interested in drawing the connection than in debating the scope of the Intentional. Nothing will be risked by that economy.

This, the complexity of interpretation, is precisely what is to be accommodated by distinguishing between the unicity and unity of an artwork, where what "unity" designates may be contested by modernist and postmodernist theorists of art, all the while some referential fixity regarding the bare logical "unicity" (or "career") of a work enables that contest to be actually and first joined. It is entirely possible that the purely referential function may be secured by paying attention to reliable markings that are not even part, in any pertinent sense, of the painting in question (a peculiar mark on the reverse side of a canvas, say). In a word, criticism and interpretation require referentially successful discourse; but providing for that says absolutely nothing about, and sets no significant constraints on (though it does require constraints on), the intrinsic "nature" of artworks and other cultural entities.

By "unicity," I mean to flag the fact that things having complex "careers" through time and change are individuatable, have "number," as such, and that they are reidentifiable in virtue of that, and have

"natures" (are "unified" in some measure) in accord with their admitted careers. It is an extraordinary fact that standard theories of interpretation have been focused on the "natures" of interpretable things, whereas I am recommending that they focus instead on the "careers" of interpretable things. (I take this to be central to Hume's deliberately primitive "official" objections to discourse about numerical identity under conditions of change.)

A theory of how to interpret the Picasso, eschewing a literal-minded "art history of the proper name" applied to the representational content of the painting goes no distance toward demonstrating that the use of referential devices for fixing the painting's identity, or even for fixing certain of its details, commits us to the doctrine that paintings have or must have fixed natures.

No, that is an utter non sequitur that draws us on to Krauss's second mistake, namely, her supposing that the play of paraliterary criticism in what she believes conforms with Barthes's practice is, in its own turn, incompatible with the mere referential fixity of the artwork itself or is capable of proceeding without genuine referential resources. The truth is that many have been wrongly persuaded (it is the error common to Krauss's two mistakes) that the extensional function of reference somehow fixes once and for all the substantive or Intentional complexities (the nature) of whatever (referents) are thus individuated—if, indeed, they actually are the sort of entity that possesses such (Intentional) properties. Take a moment more to identify Krauss's mistakes as clearly as possible: the one maintains that postmodernist criticism (but not modernist criticism) is actually able to forego referential practices altogether; the other maintains that postmodernist criticism actually entails the abandonment of such practices in abandoning modernist fixities. It is easy to suppose that the two are the same, but they are not.

The referential (grammatical) fixity of a text or artwork is a matter quite distinct from the substantive fixity of what may be thus fixed. The two are doubtless closely linked in the sense that nothing could be referentially fixed that did not exhibit a certain stability of nature; but how alterable (or by what means altered) the life of a person or the restored Last Supper or the oft-interpreted Hamlet or Duchamp's theoretically intriguing Fountain or the marvelously elastic Sarrasine may be is not a matter that can be decided, or that is actually determined, merely by marking such texts or artworks as the reidentifiable referents they are. Modernism does indeed appear to have been too naive

or too conservative about the conceptual link between the two notions, and postmodernism may have Liberated us in that respect; but, for its own part, postmodernism has failed (in Krauss at least) to acknowledge an ontic conservatism (a need for adequation) implied in referential success insofar as the possibility of such success constrains the very nature of texts apt for reference—which of course (sadly perhaps) the modernists were never even tempted to disown.

Unicity and unity are yoked concepts all right; but they need not, running in tandem, be taken for the same horse. By the same argument, to say that interpretation (in the sense of the first theory) presupposes description is not to say that description must, to be valid or true, be timelessly fixed or unchangeable or designate the fixed or unchangeable properties of whatever we go on to interpret. That would depend on the particular nature of what we mean to describe or interpret—for whatever we describe or interpret must have a "nature" of some sort, must submit to predication. Admitting description—or, better, describability—is, first, a purely logical concession to the minima of discourse; it is only secondarily, beyond that concession, disputatiously, a further—a hardly entailed—-concession to modernism or any other privileged metaphysics of art.^[13] If so, then the requirement of the first sense of "interpret" cannot be denied; the modernist thesis is at least not entailed by the concession. The formal fixities of discourse, of reference and predication, have nothing to do with deciding what the intrinsic nature of texts or particular texts may or must be—except for the fact (the hardly negligible fact) that whatever we say is the nature of a text must be compatible with so saying and with the interpretive discourse it is meant to support. Interpreted texts must have somewhat stable properties but they need not have altogether fixed natures,^[14] So we must go beyond Krauss.

III

Consider, now, a second claim, this time from Barthes's well-known essay, "From Work to Text," which is as close to a canonical formulation of what Krauss originally wished to borrow as one could possibly find:

In opposition to the notion of the work of art or literature [says Barthes] there now arises a need for a new object, one obtained by the displacement or overturning of previous categories. This object is the Text. . .. The Text must not be thought of as a defined object. It would be useless to attempt a

material separation of works and texts. . .. A very ancient work can contain "some text," while many products of contemporary literature are not texts at all. The difference is as follows: the work is concrete, occupying a portion of book-space (in a library, for example); the text, on the other hand, is a methodological field.^[15]

Two distinctions need to be made: first, there is no doubt that Barthes never means to abandon a reliance on referential facilities, all the while he clearly intends to subvert conventional views about reading a text (for instance, views somewhat like Monroe Beardsley's New Critical view of interpretive reading); second, there is no doubt that Barthes does mean to constitute, by a certain sort of reading and serial rereading, that "object" that thereby becomes (what he calls) the Text. The notion of the Text, for Barthes, therefore, is not the notion of an antecedent referent to which interpretation is directed but rather the notion of what is productively yielded by interpretively addressing "something else" that, in the ongoing (serial) process of reading and rereading, is uniquely affected by that very process.

It is impossible to pursue the theme without citing Barthes's famous distinction between the "readerly" and the "writerly" (the lisible and scriptible ) offered at the very opening of S/Z, which is close in spirit (and even language) to the paper just mentioned:

Why is the writerly our value? [asks Barthes.] Because [he answers] the goal of literary work (of literature as work) is to make the reader no longer a consumer, but a producer of the text. Our literature is characterized by the pitiless divorce which the literary institution maintains between the producer of the text and its user, between its owner and its customer, between its author and its reader. This reader is thereby plunged into a kind of idleness—he is intransitive; he is, in short, serious: instead of functioning himself, instead of gaining access to the magic of the signifier, to the pleasure of writing, he is left with no more than the poor freedom either to accept or reject the text: reading is nothing more than a referendum. Opposite the writerly text, then, is its countervalue, its negative, reactive value: what can be read, but not written: the readerly. We call any readerly text a classic text.^[16]

Of course, Balzac's Sarrasine is the classic readerly text that Barthes ingeniously shows us how to read as a writerly Text. In doing that, Barthes confirms: (1) that writerly reading does not eliminate readerly reading or its eligibility and contribution; (2) that reading of either sort presupposes a referent—the point of mentioning the "signifier," necessary for both readerly and writerly reading; (3) that the readerly text, which may have been derived, at some earlier time, from a writerly

Text, is now a fixed or bounded text the unity of which (in the modernist sense) has become a function of its particular interpretive history; alternatively, its (Sarrasine's ) being reread now (as a readerly text) commits us to recovering what constitutes it as a canonically fixed text; hence (4) that even a writerly Text is constituted (in our second sense) by interpreting something else—the "signifier," in Barthes's Saussurean usage; (5) that, for Barthes, the "Text," taken as the internal accusative of reading, is not an actual referent for further writerly reading (though the signifier is) but is collapsed into such a fixed referent only for readerly reading; and (6) that reading in the writerly way is not in the least incompatible with admitting readerly texts; in fact, it may be practiced on such texts.

A great deal of nonsense has been spread abroad maligning Barthes's intelligence, when what is wanted is a careful understanding of the remarkable thesis Barthes has bequeathed us. As it happens, it affords the best clue we are likely to find regarding the second sort of theory of interpretation. (I shall say nothing against Barthes's literary playfulness. Barthes was clearly impatient with or amused by the rigidities of structuralism. I see no evidence that his verbal games adversely affected his argument; and I see no evidence that Barthes ever played fast and loose with the constraints of coherence.) In any case, in his terribly freewheeling way, Barthes shows us how to entertain the idea that a text (in our sense, not quite in his, though congruently enough with his own notion) need not be presumed to have a fixed nature throughout a responsible reading (that is, in what we—once again, not Barthes—are calling interpretation) in spite of the fact that, however that nature may change, it remains a changing or changeable nature assignable (by reference) to this or that text (as we are prepared to say) or "signifier" (as Barthes would say). What shall we make of that?

Barthes does speak of interpretation but only either to dismiss it or to allude to what he calls "the Nietzschean sense of the word."^[17] What he means is that conventional interpretation (interpretation somewhat in my first sense) addresses readerly texts, texts construed as "products," referents with fixed natures; whereas writerly "texts" (texts treated in a "writerly' way) invite interpretation in the "Nietzschean' sense (in something close to my second sense), a sense applied to "production without product, structuration without structure": "To interpret a text [in this sense, says Barthes] is not to give it a (more or less justified, more or less free) meaning, but on the contrary to appreciate

what plural constitutes it. . .. This text is a galaxy of signifiers, not a structure of signifieds; it has no beginning; it is reversible; we gain access to it by several entrances, none of which can be authoritatively declared to be the main one; the codes it mobilizes extend as far as the eye can reach; they are interminable . . . their number is never closed, based as it is on the infinity of language."^[18]

Barthes is most exact here, despite the extravagance of his prose: "as nothing exists outside the text [he says: his remark follows Derrida's into print by about three years^[19] ], there is never a whole of the text [; that is,] for the plural text, there cannot be a [fixed] narrative structure, a grammar, or a logic."^[20] His meaning is plainly designed to preserve adequate resources for reference ("a galaxy of signifiers") but also to disallow a complete fixity of predicable nature ("a structure of signifieds"). He rejects the idea of exhausting the interpretive undertaking practiced in a cultural world that acquires new resources without limit, new "codes" of meaning drawn from its ongoing experience.

Interpretation in the "Nietzschean" sense subtends a responsive reading all right, but it is a reading that employs (as Barthes's own reading of S/Z shows) a selection from the "codes" of reading that gather and increase (somehow) in the life of our society—that do not and cannot lead to closure, to hierarchical preference, to mere correctness by way of reference to an antecedently closed textual nature. There is no longer an explication de texte, except by fiat: there is only a reading of the signifiers that thereby constitutes, reconstitutes, leaves indefinitely or "plurally" open to endless further reconstitution, the signifiers that acquire that interpretive history (that yield a text). This is the meaning of that otherwise impenetrable remark (playfully Rousseauesque): "narrative is both merchandise and the relation of the contract of which it is the object."^[21]

What Barthes means but will not say outright is that the interpretation of a "text" ("a galaxy of signifiers") is inseparable from a society's interpretation of its own open history. The meaning of whatever is singled out as a readable referent is, by that very act, judged apt for bearing whatever further signification its evolving "codes" dare ascribe to the evolving history of such ascriptions. That history is collected, referentially, as this "text" or that. The practice presupposes that texts may be individuated and reidentified as remembered histories (or careers), as referents that lack fixed natures. Barthes does not discuss the rigor of such a practice, it's true: he is more interested in its emancipatory power. But he never violates the limits of coherence, and we have (in S/Z ) the paradigm of the practice he recommends.

So seen, Sarrasine is not a standard story or a two-part story that contains a story within a story inviting explanation; it is a story "of a contract [says Barthes] of a force (the narrative) and the action of this force on the very contract controlling it": we are invited (in effect, by the writerly contract) to invent, by applying to certain signifiers the codes of reading of our world, whatever functional equivalences of structure may be imaginatively produced in an exchange of readings applied to the admitted structures of Sarrasine (the apparently "nested narratives").^[22]

It is helpful to notice that Barthes's way of reading—"writerly" reading—is entirely capable of its own characteristic discipline. It is indeed opposed to the exclusionary rights of "readerly" reading, though Barthes does not exclude such reading. The pertinent discipline depends on our sharing in some generous sense the consensual cultural resources of our own society. But that, as will gradually become clear, is also necessary for the preferred discipline of the "readerly' strategy. In short, the two ways of reading cannot be entirely disjoint. It is true, however, that Barthes is not interested in featuring any particular constraints—historical, traditional, intentional—on the interpretation of texts, in the manner favored, one way or another, by those who work within the hermeneutic tradition. There is that difference between the modernists and the poststructuralists.

It would be wrong to say that Barthes's sort of playful interpretation abandons reference or predication or a disciplined reading: it merely abandons the full fixity of texts favored in readerly readings in standard modernist accounts, and it shifts the focus of reference from finished text to enabling signifier. What Barthes alludes to is the neglect, among modernist theorists and literary commentators, of the changing context and history of reading—hence, of the contextually and historically changing "codes" of reading accessible to a living society. There is the clue to the infinite "plural" that is a Text, on Barthes's account. The very practice of reading—an entire society's practice, its Lebensform, we may say—ensures it. (I hasten to add that we must read "Lebensformen" or historical "traditions" in a much less conservative sense than Wittgenstein seems to have favored.) The "infinitude" of interpretations is simply the openendedness of a text's interpretability within a historical society's practice of reading. (This is also what Krauss attempts to assimilate.)

There are two pressure points in Barthes's theory of Texts pertinent to my second sense of interpretation. First, there is literally nothing to be interpreted (in the first sense of "interpret") until after a "text" is

"constituted" (Barthes's own term) by interpretive work (in the second sense of "interpret"); secondly, constituting a "Text" in Barthes's sense does not yield a "product," an ordinary text to be further interpreted (in the first sense of "interpret"). The point is regularly neglected by the would-be anarchists and irrationalists of interpretation. Remember: Barthes never disallows the disciplined option of readerly interpretation practiced on a textual product in accord with the first sense of "interpret."

Barthes's thesis holds only that the two sorts of reading arise together within the same societal practices (what he playfully identifies as its "codes" of reading) and may even be regarded as sequentially ordered phases of reading (or interpretation) within an increasingly normalized use of particular texts (or "galaxies of signifiers"). Barthes's emphasis is on the jouissance of (preferring) the writerly over the readerly, not the ineligibility of the latter.^[23] (The sexual joke is Barthes's, of course.) Liberty with texts or signifiers does not escape the normal constraints of discourse—only the presumptions of jejune literary theory. In a logical sense, reference has its rigor but is unavoidably informal; in a confirmatory sense, predication may be consensually apt or inapt but its possibilities evolve and are culturally openended.

Consider the following remark:

Reading a text cited by Stendhal (but not written by him) I find [says Barthes] Proust in the minute detail. The Bishop of Lescars refers to the niece of his vicar-general in a series of affected apostrophes (My little niece, my little friend, my lovely brunette, ah, delicious little morsel!) which remind me of the way the two post girls at the Grand Hotel at Balbec, Marie Geneste and Celeste Albaret, address the narrator (Oh, the little black-haired devil, oh, tricky little devil! Ah, youth! Ah, lovely skin!). Elsewhere, but in the same way, in Flaubert, it is the blossoming apple trees of Normandy which I read according to Proust . . . this does not mean that I am in any way a Proust "specialist": Proust is what comes to me, not what I summon up; not an "authority," simply a circular memory [that is, a memory that "circles" or stalks a text]. Which is what the inter-text is: the impossibility of living outside the infinite text—whether this text be Proust or the daily newspaper or the television screen: the book creates the meaning, the meaning creates life.^[24]

Reading in the writerly manner is a form of living, not a form of research; it involves know-how (savoir aller, not savoir, at least not in the sense of assuming the predicative fixity of objective texts). But it is disciplined. It involves a practice interesting to others (for instance, in S/Z ) only if the reader is really civilized, witty, inventive. We, then, re-

trace the play of S /Z in order to become similarly motivated and (perhaps as) expert. But reading in that way resists the (readerly) "bifurcation" of the reader/read text—in order to allow that same distinction to be made again in a freer (writerly) way. There is no "explaining" the Text (Barthes's "Text"), and there is no "knowledge" of the meaning of that Text: because, of course, there is (then) no definitive "text" and no one way of motivating readings "which would be definitive" of any meaning.^[25] Nevertheless, there are "galaxies of signifiers," socially habituated practices, disciplined options of reading, and above all the customary meanings of sentences and sedimented readerly texts. One sees at once Krauss's mistake—as well as the mistake of such postmodernists as Jean-François Lyotard.^[26] For, savoir faire and savoir lire do presuppose savoir—at least referentially. Writerly reading presupposes readerly reading—again, at least referentially.

Barthes effectively acknowledges the point: it is the only possible condition on which a complete chaos of reading (or of cultural life in general) can be avoided. It is in part at least what Wittgenstein means by "forms of life," what Bourdieu means by "habitus," what Marx means by "praxis" and "modes of production," what Hegel means by "Sitten," what Gadamer means by "wirkungsgeschichtliches Bewusstsein," what Husserl means by the plural of "Lebenswelt," what Foucault means by "epistemes." It is no more than the acknowledgment of the preformative historical practices by which culturally apt individuals first become apt. Their world is already culturally preformed for them: that is the reason they may be said to learn their native language and their native culture; that is the reason they can specify the "signifiers," the culturally (already) prepared materials, that, by interpreting (in the productive sense), they first constitute texts or artworks as such—what thereupon prove to be usable as referents apt for interpretation (in the adequational sense). In a word, they learn to share the consensual memory of an encultured society. If you press the point, you grasp in a single step the impossibility of modeling such interpretation computationally, for there is, and can be, no algorithmic accommodation of a genuinely emergent, historicized culture. If interpretation is open in anything like Barthes's sense, then, unless some very strong nativist account of concepts holds (Fodor's, for instance), which is most unlikely for the concepts favored in the interpretations of the arts, computationality must fail.^[27]

Barthes's emphasis, of course, is on the initial process of writerly reading. On the evidence (on his view), the process has been forgotten

or ignored or misconstrued. My own emphasis, for the moment at least, is focused rather on the option of continuing a critical discourse about whatever (a "text") is thereby so constituted—without in the least reneging on Barthes's fine lesson. On the argument, we preserve both themes merely by distinguishing with care the logical requirements of unicity or individuation from the prejudice of certain substantive (modernist) presumptions about unity or fixity of nature.

The constraints of reference and predication are not violated by Barthes, only displaced from produced or finished texts to interpretable (openendedly interpretable) signifiers—from "natures" to "careers," as I should say. Barthes himself does not tarry long enough to give us a theory of the social habituation of the practices of reading that support the distinctive discipline of readerly and writerly reading. He presupposes such a theory—or such theories—but he moves on to offer examples of what be recommends. For my part, we could easily pause to construct a theory of social practice—from Hegel or Marx, or Nietzsche or Foucault, or Weber or Lukács, or Husserl or Heidegger, or Adorno or Benjamin, or Lévi-Strauss or Althusser, or Wittgenstein or Bourdieu, or Gadamer or Kuhn.

The point remains quite constant, however: the waiving of texts in the sense suited to (a modernist view of) the first sort of interpretation does not eliminate referential discourse elsewhere (for readerly texts, say), does not preclude referential and predicative discipline within the writerly reading recommended (as in the identification of relevant signifiers, the identification of other readerly and writerly read texts, a certain civilized familiarity with the details or "codes" of one's culture); and it does not even preclude a rapprochement between readerly and writerly reading before and after the play of a particularly agile exercise of the latter sort (the charm of S/Z, say). Certainly, in yielding in an openended way at least minimally hospitable to speculations of Gadamer's and Foucault's sort, Barthes deliberately opposes any structuralist conception of interpretation (the point of the playful use of the term "codes"). He is, in effect, exploiting his own idiom of an earlier phase of work; and he is also accommodating (within an enlarged vision) all alternative practices of interpretation, so long as they do not entrench false presumptions of fixity or totalizing. (I shall have occasion to reconsider the matter when I examine Michael Riffaterre's late-structuralist model.)

In short, Barthes's preference of the writerly is not even a denial of the ontology of texts—or of the likely dawning of gradually normal-

ized texts for which such an ontology could be retrospectively constructed (if we wished); and it does not itself supply an adequate analysis of what a signifier is, or a practice of reading, or even a human being capable of reading in either the readerly or writerly way. It is one thing to grasp the fresh discovery Barthes bequeaths us; it is quite another to make a shambles of every effort to understand interpretation. After all, the "bifurcation" of the signifier and the would-be reader remains, after the provisional "bifurcation" of the readerly text and the reader (and author) is first disallowed—and then (of course) civilly permitted to be recovered again in Barthes's educated sense.

Barthes offers an instance (in S /Z )—after the fact of a readerly deposit of Sarrasine in the canon of conventional texts—of what it would be like before such a reading, to have read Sarrasine in the writerly way. The "galaxy of signifiers" lacks fixed meaning; but, as the competent readers we are, we possess the know-how for grasping what may be taken to be their meaning. Barthes suppresses this hermeneutic or habituative dimension of reading—but it is surely there. The semantic and semiotic potentialities of signifiers are already built into the minima of any socialized habit of reading and using language. Nevertheless, Saussure, whom Barthes had taken his original departure from (but now supersedes), had never successfully explained the "original" relationship between writing and speech or writing and thought that he insisted on; and without that "originary" source—or the effective replacement of it more perspicuously advanced by Wittgenstein and Gadamer, say—there remains a critical lacuna in Barthes's own account. Certainly, the familiarity of Wittgenstein's and Gadamer's views considerably domesticates the question of a "writerly" discipline. (Saussure's "failing," of course, is just what Derrida had so mercilessly exposed in Of Grarnmatology.^[28] )

But the deeper theme, missing also in Derrida, is this: that the deconstructive or poststructuralist or antimodernist rejection of the bifurcation of reader and text itself entails a competent practice or activity on the part of readers vis-à-vis something (signifiers, say) within a preformed or habituated cultural space in which (and by using the processes of which) what Barthes calls the "plural" or "infinite" Text is first constituted (and what we are now calling a "text" may be constituted for "infinite" or, better, indefinitely many interpretations). In a sense, "the deconstruction of hermeneutics" is therefore reversed and outflanked by being shown to require and presuppose a "hermeneutics of deconstruction."^[29] It is not, however, thereby disallowed

or repudiated. What the argument shows is that the rejection of a cultural world bifurcated between inquiring subjects and subjects inquired into—or between such subjects and what they do or produce (texts, in the idiom I have proposed)—is itself the work of subjects (competent selves) active in such a bifurcated world and affected by their own work in it.

We theorize in a critical moment about a preformative condition we cannot originally fathom (that Saussure thought he could fathom, that Husserl also thought he could fathom) within which the bifurcation of world and word (or text and reader) first arises. Barthes's splendid game of writerly reading tenderly texts serves a double purpose: for one thing, it affords a miniature exemplar of the impossibility of radically disjoining the double function of subjects as observers and observed (in much the same sense in which one cannot beat oneself at chess); and, for another, it subverts the fixities of privilege, of readerly reading, of the "metaphysics of presence," of all the bugaboos of failing to remember that the steady structures of our now-bifurcated world depend impenetrably on whatever we critically postulate as the preformed world within which our own salient "objective" world arises.

So seen, Barthes's invention is an attractive toy—no more than a toy, no more than a toy for Barthes himself: for we could easily (and would need to) interpose a conception of numbered, reidentifiable texts that could support interpretation in the first and second senses and that would, at the same time, subvert a "metaphysics of privilege" (the notion of fixed and bounded texts) just because—for reading purposes at least—texts do and must remain referentially accessible. Barthes's conceit of the infinite Text (that is not itself a referent) is, then, merely the deliberately posed extravagance of a disappearing limit for the more modestly interposed texts I am now recommending. Barthes nearly says as much:

The Text (if only because of its frequent "unreadability") decants the work from its consumption and gathers it up as play, task, production, and activity. This means that the Text requires an attempt to abolish (or at least to lessen) the distance between writing and reading, not by intensifying the reader's projection into the work, but by linking the two together in a single signifying process.^[30]

The point is, a theory of texts adequate for interpretation at the present time would favor Barthes's double lesson—but would do so in an ampler and more systematic way than Barthes actually does. We must:

(1) detach the full theory of the nature of texts (literary, visual, musical) from the mere referential and predicative constraints of discourse about them, so that all notions of fixity, essence, analogy with physical particulars are attenuated as far as possible or challenged as much as necessary; and (2) we must develop a positive theory of texts, of how texts (or culturally emergent phenomena and entities in general^[31] ) are actually constituted—first, from precultural physical materials and, second, from culturally prepared materials. Item (1) trades on the lesson drawn from Krauss and Barthes: namely, that unity and unicity are distinct though not altogether separable notions; item (2) requires an entirely fresh start and cannot fail to center on the peculiarities of Intentional properties and their incarnated relation to material properties.^[32]

All this may seem unnecessarily heavyhanded, because Barthes was so ponderously intent on being lightfingered. But there is no philosophical carelessness in what he says. On the contrary, what Barthes conveys by his practice is the logical viability of preserving reference and numerical identity and yielding on fixed natures. (He mistrusts theory, but he never violates the constraints of coherence.) I accept Barthes's innovation: I reconcile the two notions by treating texts as individuatable histories. (You will have noticed, by the way, two important terminological adjustments: first, that "interpretation" need not be confined to linguistic acts only—in the most literal-minded sense; second, that "texts" include any and all culturally constituted entities—intrinsically apt for interpretation. Thus, the ballet is interpretive, and persons are texts.)

IV

We may collect our findings a little more deliberately. An adequate theory of interpretation will explain: (1) how it is that we can referentially fix, identify, or individuate artworks or texts for interpretation without at the same time insisting that their nature, their collected properties, their essential boundaries must also be fixed, determined, changeless, or at least unaffected by merely interpreting them or commenting on them in the normal critical way; (2) how it is that artworks or texts are first constituted as such, so that they become the relatively stable referents of subsequent interpretive discourse; and (3) how it is that discursive interpretation can alter the "natures" of

individuated texts and artworks and, in doing that, reconstitute their natures or properties without disorganizing their numerical identity and (of course) without inviting total chaos.

What has been shown is the sheer coherence of the intended answer: the bare unicity of referents accommodates the absence of any fixed unity or fixed nature of the particulars thus identified. In the biological world, we capture the limits of tolerance for changing natures and fixed reference by adjusting our notions of natural kinds: spatiotemporal continuities, as Hume more or less admits, aid us in allowing fixity of identity to range over the shifting sequences of instantiated properties. In contemporary physics, among the quantum-mechanical puzzles of reconciling particle/wave anomalies, we exploit (with Heisenberg, for instance) punctuated identification for the sake of descriptive control and then permit identification to become as story-relative as can be tolerated at a theoretical level at which such identification would be altogether disallowed. In the cultural world, both with regard to persons and artworks, we borrow whatever similar conveniences we can; we maintain, for instance, wherever we may, "one person/one body," or "one sculpture/one block of marble," or "one poem/one inscription from a set of possible inscriptions." But texts and artworks do not form natural kinds and cannot be identified merely physically or as physical bodies. They differ from natural objects essentially in possessing Intentional properties. It is in virtue of that that, paradigmatically, texts are subject to interpretation in the two senses supplied and that those two senses are interrelated in the manner sketched. (These are the themes of adequation and symbiosis.)

I have now come to the most strenuous part of the thory needed. Since I cannot attempt a full account of the ontology of artworks or cultural phenomena in general,^[33] I may as well be candid about the upshot of what such an account would yield. It would, without endangering the rigors of numerical identity or the critical testing of particular claims or of coherent discourse in general, make possible our assigning numbered texts indefinitely many interpretations in principle, and (then) their entering into indefinitely many histories; also interpretations and histories assigned particular texts at t' later than t may be affected by interpretations and histories assigned at t. Barthes's notion would accommodate all that.

The fascinating thing is that it is possible to make such a notion coherent, manageable, even plausible and disciplined. It would require a number of substantial concessions regarding the logic of general dis-

course that would not be narrowly occupied with the theory of art or interpretation. They would include at least (1) abandoning as fixed principles the principles of excluded middle and tertiurn non datur; (2) admitting the adequacy of, and the impossibility of exceeding thelimitations of, story-relative reference; (3) admitting the viability of relativistic truth-values and the compatibility of distributively employing such values together with the (distributed) use of bivalent or bipolarvalues where wanted; and (4) challenging, if not repudiating, the adequacy of would-be strict extensional rules for regimenting all languages
descriptive of the real world. I freely admit that these are heterodox suggestions. My only point at the moment is that they are not incoherent and that the terrain of interpretation may actually favor them.

The project would also require a number of substantial concessions of an ontological sort reconciled with the adjusted logic of discourse. For instance, it would require: (5) denying the adequacy of all physicalisms (as opposed to materialisms—that is, theories of the fundamental "stuff" of the universe), whether reductive or not, in order to accommodate the reality of the artworld and human culture in general; (6) admitting cultural emergence, as distinct from physical emergence, as a process that yields indissolubly complex embodied or incarnated phenomena or properties; (7) admitting that what distinguishes artworks, texts, and other cultural entities from natural entities depends essentially on the complex incarnation of so-called Intentional properties; and (8) admitting, in addition, that Intentional properties are such that they can be constituted, altered, affected, generated by the processes of critical discourse or interpretation applied to given texts or cultural referents, without adversely affecting the numerical fixity of such referents and without altering merely physical properties. The beauty of holding to a theory in accord with 1-8 is that it is not a Cartesian dualism or in any way primarily centered on a disjunction of the mental and the physical. It is concerned rather with the difference and relationship between the natural and the cultural— within the space of which the other may be accommodated.

There are baffling and unanswered questions here. I don't deny it. I must leave them for the time being. I regard it as a considerable gain that the theory of interpretation that is dawning is not obviously incoherent. Many will be surprised. I believe the argument pretty well comes to this: first, that the distinction between culture and nature is different from, not reducible to, and in fact more inclusive than, the distinction between mind and body; and, second, that what is culturally

"emergent" with respect to physical nature is not tantamount to what is merely "supervenient" with respect to physical nature. I shall offer a further word about the second theorem in a moment.

If physicalism and extensionalism were philosophically correct, or at least adequate (in real-time terms) for all discourse about the cultural world, then everything so far said would be entirely pointless. That must be conceded straight off. But if those programs are neither correct nor demonstrated to be correct nor demonstrably correct nor even demonstrably adequate (in real-time terms), then we are left with a world for which theses 1-8 may be peculiarly apt, however quarrelsome and contested. I have been careful to put this option in the best light—to make it appear as a natural continuation of the familiar debate about interpreting texts. It is certainly true that attacks on physicalism and extensionalism are widely and honorably resisted. But it is also fair to say that there is no known demonstration that shows that opposition to those doctrines is incoherent, irresponsible, unfruitful, or calamitous. That is as honorable a stand as the other—probably a more resourceful one at the present time.^[34] In any case, the admission of 1-8 leads directly to the startling prospect that artworks or texts may be assigned indefinitely many interpretations and may enter into indefinitely many histories.^[35] This is just what Barthes was getting at when, notoriously, he affirmed that "the Text . . . practices the infinite deferral of the signified."^[36]

It is also, however, close to what a theorist like Gadamer means, speaking from the altogether different vantage of hermeneutic ontology, when he declares that

to understand a text always means to apply it to ourselves and to know that, even if it must always be understood in different ways, it is still the same text presenting itself to us in these different ways. . .. The linguistic explicitness that the process of understanding gains through interpretation does not create a second sense apart from that which is understood and interpreted. The interpretive concepts are not, as such, thematic in understanding. Rather, it is their nature to disappear behind what they bring, in interpretation, into speech. Paradoxically, an interpretation is right when it is capable of disappearing in this way. The possibility of understanding is dependent on the possibility of this kind of mediating interpretation. . .. Interpretation is contained potentially in the understanding process. It simply makes the understanding explicit. Thus interpretation is not a means through which understanding is achieved, but it has passed into the content of what is understood.^[37]

Notice that, unlike Barthes, Gadamer insists on the reidentification of one and the same text under plural, potentially infinite, interpretation

and reinterpretation. The openness of texts—in both an interpretive and historical sense (ultimately the same)—is ensured by the notion of reflexive "application": the Intentional import of a text essentially incorporates into its developing, endlessly reconstituted meaning what its recovery for our own historical experience and prejudice can make it out to be. Its meaning is heuristically schematized in the intersection between our present power of reading and what, from that evolving perspective, we posit as its collected past.

In this regard, my proposal about interpretable texts is closer to Gadamer's usage than to Barthes's. Quite unaccountably, however, Gad-amer is, at every step, much more reluctant than he ought to be to accommodate a frank relativism: he is arbitrary about the point; his thesis actually favors a relativism, although he denies it. In this regard, my substantive proposal about interpretation is closer to Barthes's vision than to Gadamer's. (But there is a measure of convergence that must be acknowledged.) Still, it is Gadamer rather than Barthes who answers the third of the three questions posed a moment ago, namely, how it is that discursive interpretation alters the nature of individuated texts without affecting their numerical identity and without producing conceptual chaos. Gadamer's answer depends on repudiating the Romantic recoverability of authorial intent, on reclaiming the historicity of human existence and cultural texts, on admitting the intransparency and preformative forces of the human world in which, in a Heideggerean sense, we are "thrown," and (most important) on featuring the natural or perspectival "prejudice" (Vorurteil) of all understanding and interpretation—in a word, on the function of "the fusion of horizons" (Horizontverschmelzung ). Thus Gadamer maintains: "It is part of real understanding . . . that we regain the concepts of an historical past [understand or interpret a text] in such a way that they also include our own comprehension of them. [This is what is meant by] 'the fusion of horizons'."^[38]

The meaning of a text, Gadamer says, is the "fusion" of its perceived past and its perceived present application to ourselves; but it is we who monitor both elements of the effort. Plainly, what is most important about Gadamer's proposed solution is that it is coherent, apt, and manageable. It shows us a way of answering our third question without loss of rigor. It opens the way, therefore, to a variety of alternative strategies. We shall have to weigh their merits, of course; but for the moment I am merely preparing the ground for a pertinent choice. I am not pretending that the full significance of Barthes's and Gadamer's options is easy to fathom. I claim only that no paradox of reference

or predication or individuation or numerical identity has surfaced as a result of invoking the second sense of "interpret." The gain seems to be a sturdy one. Barthes confirms by example (S/Z) and informal argument the sheer coherence of the general theory; Gadamer provides a generous but not infinitely plastic sense of the presence of a historical tradition within which interpretations may be tested and reasonably validated. Barthes does not actually address the conceptual puzzles I have broached, but he never violates any formal constraints on discourse. Gadamer does not explicitly formulate criteria for rigorous confirmation of particular readings, but turns instead to the authenticity of historical existence. I am trying to provide in a tactful way what is missing in both.

The upshot is: (1) that it is in virtue of the Intentional nature of texts that they require interpretation in order to be understood; (2) that since interpretable texts and textual interpreters exist historically, there cannot be a uniquely correct or uniquely convergent reading or interpretation of a given text; and (3) that since interpretation and understanding require the historicized recovery of the Intentional import of a given text, it is quite impossible to fix that recovery except in terms of the salient or convincing fusion of—or what, from the perspective of present interpreters, is posited as the shared or continuous or intersecting—horizons of the past and the present. This resolution of our third question, along Gadamer's lines, appears to reconcile the ontology of texts and the methodology of their interpretation.

We must remind ourselves that I have said nothing reliable about the "nature" of interpretable texts, beyond the veiled allusion to their "Intentional" properties. But it must also not slip our notice that the gains so far made have not depended in any way on the details of any particular theory. I emphasize the conceptual generosity of the distinctions so far advanced, the hospitality with which diverging and opposed views may be canvassed, rather than any prematurely urged argument in favor of this or that particular interpretive practice.

Gadamer, without the least defense (as already remarked), affirms that there is a "universal" or "classical" tradition that can always be historically recovered—indeed, that must be recovered (that was never, and could never have been, lost)—for a successful resolution of the hermeneutic task.^[39] Naturally, I reject his claim. But what could he have meant, having admitted the constructed nature of human selves?^[40]How could he, for instance, have disallowed the more daring conjectures Michel Foucault advances? For, puzzling over Velázquez's prob-

lem in representing pictorial representation within the Classical (that is, the baroque) canon (in Las Meninas ) , Foucault remarks: "Before the end of the eighteenth century, man did not exist. . .. He is a quite recent creature, which the demiurge of knowledge fabricated with its own hands less than two hundred years ago. [Before that time] there was no epistemological consciousness of man as such [he says]. The Classical episteme is articulated along lines that do not isolate, in any way, a specific domain proper to man."^[41]

The meaning of Foucault's remarkably apt perception (even if we should disagree with his interpretation of Las Meninas ) is that the historical past, which is both real and not the same as the physical past in which it is incarnate, can be retroactively affected (without violating physical time or physical causality) by future sensibilities that could not have been recognized as potentiated in a particular past present. Merely to mention the complication is to appreciate the task of a seriously contemporary theory of interpretation.^[42] (I shall return to this important issue.)

I must add here that, more or less in accord with the incipient theme of invariance just mentioned (the "classical" theme—regarding which my own reading is certain to be contested by many), there is evidence that Gadamer is in the process of retreating from the relativism implicit in his own theory (which he means to avoid) and of reinforcing a fairly conventional philosophical reading of poetry not terribly distant from the spirit of New Critical reading, both in terms of the "autonomy" of a poem and of a rather bland sense of historical change. (One might also say, I admit, that Gadamer has never been altogether consistent on the matter: perhaps, then, that he has not really "retreated."^[43] )

Notions of historicity, therefore, as variable as Krauss's, Barthes's, Gadamer's, and Foucault's strongly favor the need for, and the plausibility of, a theory rather like the one being sketched. The "canonical" view—what I have called interpretation restricted in the first sense (largely modernist)—does not permit these subtle questions even to surface; and interpretation in the sense here championed—that unites the first and second senses given—would secure the stability of texts and interpretation by way of the salient habits of life of a society rather than by way of a privileged discovery of independently fixed entities. The argument, therefore, leads directly to a reconciliation of the would-be objectivity of interpretive claims and a frank relativism about truth-claims (which is to say, not a relativism regarding the meaning of "true").^[44]

Without the sheer conservative contingencies of life itself, human history would be an utter chaos; the disorder of critical interpretation would be instantly matched by the loss of science and rational prudence. That admission is certainly much more than the recovery of a pragmatist aesthetics, but it is at least that. (I shall need to say more about relativism, of course.)

Before moving on to other topics, however, I must secure a theorem (broached a moment ago) that I regard as particularly strategic: namely, that the culturally "emergent" world of art and history is not equivalent to any mental world "supervenient" on physical nature. What I mean to oppose is the presumed adequacy of what has been called "nonreductive physicalism."^[45] The doctrine of "supervenience" (in contemporary philosophies of mind) is perhaps the best-known version of what I am identifying as nonreductive physicalism—which is normally not applied to the problems of history and interpretation but could be. Nevertheless, the notion of interpretation I have drawn from Barthes, and associated congenially (if somewhat loosely) with Gadamer and even Foucault, would be challenged along the lines of "supervenience" fashionably favored by philosophical physicalists who do not regard the mental and cultural life of humans as mere fictions or delusions of any sort. In the interest, therefore, of attracting a wider appreciation of the puzzles I have in mind, I must insist on the following: (1) that, as remarked, the Intentional is social or societal or collective or historical, in the sense in which language is its paradigm; also, that the psychological aptitudes of encultured persons are already characterized in cultural ways that cannot be explicated in terms of individual mental or neurophysiological states prior to their having internalized the Lebensformen of a particular society (in virtue of which they first "emerge" as selves); and (2) that, as a consequence, what is first identified in Intentional space is, however emergent with respect to the physical world, not supervenient on that world.

What I mean is this. There is no known procedure (rule, criterion, algorithm, law, or the like) by which, from a description of any physical events, we could infer in a reliable way any culturally significant events we spontaneously (normally) recognize; nor, for any culturally significant events, could we infer any reasonably detailed and pertinent physical events in which they would be embodied and by reference to which they could then be indexed. The implied difficulty was dubbed by Herbert Feigl, many years ago, the "many-many" problem. Feigl regarded this problem as the principal difficulty confronting any unity-

of-science or physicalist account of the mind; he did not consider extending the problem to cultural matters (interpretation and history, in particular), because he did not envision that the cultural world could not be reduced to the psychological (effectively, the solipsistic). But, for my present purpose, Feigl's admission is instructive enough.^[46] Now, Feigltreats the ("many-many") problem as empirically contingent. Those favoring supervenience do not favor contingency: they hold instead that there is a modal or necessary invariance (de re) between what "supervenes" and that from which it does so.

"Supervenience" was introduced, in its present usage, in Donald Davidson's well-known paper, "Mental Events." There, applying G. E. Moore's famous account of "good" as a nonnatural property "supervening" on natural properties, Davidson says the following:

Although the position I describe denies there are psychophysical laws, it is consistent with the view that mental characteristics are in some sense dependent, or supervenient, on physical characteristics. Such supervenience might be taken to mean that there cannot be two events alike in all physical respects but differing in some mental respect, or that an object cannot alter in some mental respect without altering in some physical respect. Dependence or supervenience of this kind does not entail reducibility through law or definition; if it did, we could reduce moral properties to descriptive, and this there is good reason to believe cannot be done; and we might be able to reduce truth in a formal system to syntactic properties, and this we know cannot in general be done.^[47]

Colin McGinn, who accepts Davidson's argument, emphasizes the modal nature of the claim: "if x and y have the same physical constitution, and x has mental property P, then y must also have P. " This McGinn calls the principle of "the supervenience of the mental on the physical."^[48]

Now, if even the "many-many" problem cannot be resolved in physicalism's favor, then the supervenience thesis is surely false; but it is also false if the modal interpretation of nonreductive physicalism cannot be confirmed: that would hold for both the mind/body problem and the culture/nature problem. It is for that reason that I say that, first of all, the admission of the coherence of Barthes's and Gadamer's theories of interpretation entails the falsity of the modal interpretation of nonreductive physicalism and, afortiori, the supervenience thesis. I draw the conclusion in a merely dialectical way in order to emphasize the pertinence of the interpretive question for a very large run of philosophical issues (of some importance currently) that would not

normally be perceived to be relevantly affected. But I should add at once that the interpretive question—in the arts, in history, in the human sciences, in language—is almost never brought to serious attention in terms that give priority to physical events by means of which the mental or the cultural is first identified as such. In short, the bare admission that the Intentional "natures" of texts are alterable under interpretation instantly renders untenable the supervenience thesis. This is a splendid and inexpensive gain, for one sees at once that any robust form of cultural realism (that is not reductive) ensures ontic and epistemic complexities that the "naturalizing" tendencies of current analytic philosophy cannot possibly accommodate.

It is our cultural fluency that enables us to recognize, individuate, identify, and reidentify spontaneously events and phenomena of the relevant sort—regarding which, thereupon, the supervenience thesis and its physicalist analogues are first pertinently broached. In that sense, the thesis is always raised after the fact of cultural fluency; whereas what I mean by "cultural emergence" is just that fact and whatever it entails—if nonreductive physicalism fails—regarding the structure of a science.^[49] For the moment, however, I mean to draw attention to the fact that, within the space of cultural life, our discursive fluency is capable of coherently supporting—both with regard to the rigors of reference and predication and with regard to larger philosophical issues touching on the mind/body and culture/nature problems—interpretive practices in accord with Barthes's conception of "writerly" reading and Gadamer's conception of the "fusion of horizons." There may be reasons to oppose Barthes's and Gadamer's theories, but I cannot see that they can be drawn from an analysis of our linguistic resources or from an analysis of what it is to be a science.

I have brought two themes together by introducing my heterodox proposal regarding interpretation: for one thing, I have now shown that the theory bears in a surprisingly direct way on certain presumptions favored in analytic forms of physicalism that might not have been suspected; and, for another, I have shown that the theory is patently coherent and promising. I want to emphasize, before I move on, that the nerve of the argument rests with challenging the idea that interpretable "things" (artworks, texts, histories, persons, utterances, actions) need, when "objectively" interpreted, to be assigned fixed meanings. Once the analogy with physical objects is abandoned (New Criticism), the fixity of authorial intent denied (Romantic hermeneutics), and the closed system of pertinent semiotic "codes" subverted

(Francophone structuralism), whatever objectivity interpretation may claim for itself depends entirely on the reasonableness with which the changing historical experience of interested societies may be tapped. The rest of the story requires attention to: (1) constraints on interpretive relevance; (2) the coherence of our theories of the "nature' and "career" of texts; (3) the viability of standard forms of reference and predication; and (4) the benign complexity of Intentional properties.

Chapter 2
Interpretation at Risk

I

Occasionally, someone wishes to play Roland to the Saracens of critical theory, claiming privileged knowledge of the villains of history. A few years ago, a redoubtable philosopher, Stanley Rosen, sensing a tide of alien thinkers threatening the territory of canonical interpretation, raised the following anthem to stiffen the spirit of the resisting troops:

If there is no human nature that remains constant within historical change, and so defines the perspectives of individual readers as perspectives on a common humanity, then reading is impossible. Whether one's primary orientation is ontological or philological, interpretation depends upon the initial accessibility of the sense of the text as independent of clarification and deepening by the subsequent application of theories, methods, and canons.1

Rosen's perception was an astute one, in that he unblinkingly fixed the point of all essential contests about the fate of competing theories of interpretation: that is, whether human nature—a fortiori, human texts— must be invariant and unchanging (through change) or whether, preserving the viability of coherent interpretation, that claim may be effectively denied. Rosen's exhortation gives way almost at once to a sense of distinct alarm: "the first [effect of the process of deteriorating theories of interpretation] is a progressive separation of our understanding of the obvious from an intensifying conviction that nature, and so human nature, is a historical myth that must be replaced by a scientific construction. . .. The second . . . is a product of the refined

sensibility of historical old age. Subtlety decays into ingenuity, and the speculative imagination, unrestrained by the standard of nature—or by the 'given', which is now unmasked as an epistemological error— slips into the dream world of fantasy."2

This is a marvelous piece of argument, partly because it is so completely aware of the real source of maximal danger to itself, partly because it mobilizes the defending troops by warning them of the wildest options of the fanatics of the enemy host. But it is a strategy that utterly fails—that fails in a most instructive way—unless mere pronouncement counts as success or unless what might be lost in the anticipated attack is so unthinkable that the victory of the defenders of what Rosen regards as the canon remains entirely assured.

I defend the Saracens here: because the thesis of "invariant nature" is undemonstrated and indemonstrable, because the charge of the "impossibility of reading" (or interpretation) is more than extravagant, and because (ulteriorly) the thesis I mean to defend depends in an essential way on the defeat of Rosen's sort of claim. I will not, however, rest the argument solely on the merits of Rosen's formula: as it turns out, the main lines of the argument are already to be found in Aristotle. (Frankly, I exploit Rosen as a proxy.) Also, Rosen's argument gives the impression that the denial of an invariant human nature leaves us only with the option that "human nature" must be some sort of fiction ("fantasy"); whereas the truth is that there are any number of intermediary conceptual strategies that could provide a basis for interpretive rigor that Rosen would not consider viable at all. Rosen is committed to two very doubtful (indeed, untenable) claims: one, that the denial of a fixed "human nature" produces conceptual chaos; the other, that there can be no basis for coherent interpretive rigor that is not grounded in the presumption of a fixed "human nature."

The reason for dwelling on Rosen's claim is not that it is so compelling. It is far from that. The truth is that, in the spirit of identifying what we "must" resist, Rosen may be unmatched in the clarity with which he publicly identifies what may no longer be easy to resist, what may well be on the point of overtaking us conceptually without any noticeable harm at all. Rosen's claim is that the loss of "the standard of [essential or invariant] nature" is tantamount to the loss of every possible criterion or ground of reasonable judgment:

When nature is replaced by history, and history is transformed into a methodological artifact, there is no criterion by which to select the correct artifact other than taste, or, as in fact happens, than by chance, disguised as the will to power.3

My counterclaim is that this is simply false and indemonstrable and that it is important to understand why. Of course, adopting this tone deliberately ignores the deep concern of thinkers like Rosen who genuinely worry about the fate of the world risked on some crazy academic's repudiation of the invariant norms of human life. But if the counterargument succeeds, then there may be more danger in the delusion that there are invariant norms to be read off human nature than in its being exposed. The complexity of the flux remains, of course; but, on the argument, it was always there. There is some comfort in that: because the unmasking of an important false claim cannot then be so easily accused of putting us at greater risk. Rosen's own claim certainly demands an answer. With due respect, then, the issue is more important than the man. In terms somewhat local to my argument, the defense of something like Barthes's conception of writerly reading entails a rejection of Rosen's (or better: Aristotle's) thesis and the recovery of a sense of the coherence of individuation applied to things that have no (fixed) "natures" and yet remain one and the same through their changing careers.

In any case, Rosen's argument raises a modal objection—as, in fact, does Aristotle's (the classic objection) and as does also the implicit argument against Barthes's proposal, as we have already seen. My own strategy, I may as well say, is to expose the modal claims as false or undemonstrated and to defend the contrary as more plausible for the issues raised. The new puzzle of interpretation and history is that concerning its coherence. Fair enough.

The reasons Rosen gives are certainly the reasons he believes the theories of interpretation recommended by Jacques Derrida and Hans-Georg Gadamer are "incoherent," however worthwhile their occasional observations may be. They deny, he says, any "theoretical substructure to reading or of writing: there is only the infrastructure of the reader and the writer."4 (This is also false or a palpable distortion.) More to the point, Rosen conflates the two sorts of "deterioration" he mentions. For, even if it were an extravagantly incoherent thesis to hold that there are no texts to be read (or interpreted) and that in their place we are to suppose there is only some sort of confrontation between reader and writer, it is neither true that Derrida or Gadamer eliminates intervening texts nor true that admitting texts restores an invariant human nature or nature's "standard." (This is partly what I have shown in the preceding chapter, in showing the coherence of Barthes's interpretive program.)

If, in philosophy in general, acknowledging a cognizable nature and cognizing agents need not entail the exceptionless invariances Rosen thinks necessary for science, then it is a foregone conclusion that the admission of readable or interpretable texts and the admission of human beings apt for reading or interpreting need not entail their having essentially fixed natures. There is, in fact, some reason to think that persons and texts are culturally emergent artifacts of some sort, subject to significant historical variability from age to age and within particular careers. Rosen does not countenance the possibility.

That texts are artifacts aptly produced for reading by humans (functioning as writers) for others apt for reading (functioning as readers) fixes the need for a conceptual congruity between the legibility of texts and those enabling powers. There is an obvious lacuna in Rosen's charge. The matter is complicated, of course, if we admit that, in an important sense, man himself is: (a) an interpretable text, (b) a reflexively interpreted text, (c) a text interpreted and affected or altered in the active interpretation of every other text, (d) a text the reflexive interpretation of which affects and alters the active interpretation of every other text, and (e) a text bound in a real symbiosis with the determinately many texts (including other persons) that man interprets. Items (a)-(e) are coherent enough, taken singly and conjunctively. Their coherence is the upshot of the argument of the preceding chapter. What I am undertaking to do now is, in effect, to assess the consequence of forcing a confrontation between its adoption and the strongest line of objection against its viability—in the context of examining the "nature" of persons, artworks, histories, and the like as interpretable texts. The "strongest line" is most certainly derived from Aristotle. Rosen simply shows us the way. But it is also true that Aristotle does not examine the objection in anything like its hermeneutic application. We are obliged to Rosen for a sense of that important adjustment. It is, after all, the one most pertinent to my present concern.

It does not seem possible to admit (a)-(e) consistently with Rosen's claim; at the least, it is a difficult trick. It is also difficult to resist the sheer coherence of (a)-(e). However, admitting (a)-(e) risks, without yet precluding, the reidentifiability of texts apt for interpretation and the objectivity and coherence of the particular interpretations we happen to offer. It is indeed distinctly contemporary to favor these items in increasingly radical—constructivist and historicist—ways, as Rosen acknowledges. But Rosen does not show us how to oppose the tendency effectively or what to offer in its place. My own conviction is that

there is no satisfactory argument against the denial of invariant nature and invariant texts—that is, Aristotle's modal claim—and that there is no reason to suppose that the radical reading of nature and texts being recommended (the denial of invariance) is intrinsically incoherent or deprived of the usual resources of effective reference and predication.

Rosen is content to say that if interpretation is construed in accord with conditions similar to (a)-(e), read at the radical end, then interpretation is impossible: reading is impossible, the informed transmission of civilization is impossible. He does not pause long enough to explain just why. But the charge is surely the standard one I must address if I am ever to vindicate the theory of interpretation I have begun to shape.

Among the more thoughtful recent theories of interpretation conformable with (a)-(e), Gadamer's metaphysics of human existence must surely count as one of the principal specimens. It would not be unreasonable to claim, for instance, that Gadamer's twin doctrines—Hori-zontverschmelzung (the fusion of horizons) and wirkungsgeschichtliches Bewusstsein (effective-historical consciousness)—are expressly meant to capture just such themes as (a)-(e).5 Gadamer's intention, in the second doctrine, is, I think it fair to say, to capture Hegel's notion that human agents do not (cannot) entirely understand what the historical significance of their own commitment comes to, at the same time he maintains that history has no telos. By the first, he means that our lives implicate a span of time that includes our own historical past, and that, at every present, we reinterpret the continuity of our changing past and present. Also, that, regarding neither doctrine, can we claim a privileged cognitive stance.

We need not endorse Gadamer's specific theory. In fact, it is notorious that Gadamer has no sustained account of the rigor with which interpretation can be pursued in accord with his own notions—say, with respect to reidentifying texts and confirming particular interpretations. He considers only the metaphysics of self-interpreting humans engaged in interpreting other relatively stable texts; he takes entirely for granted the discipline or practice of actual interpretation.6 He himself interprets poetry, but he offers no theory of how to confirm the validity of such work. This marks the fairness of Rosen's complaint, but it hardly confirms his charge—or the validity of his intended corrective. In fact, Gadamer is wrong to suppose that he can vindicate his hermeneutics without providing a sense of the rigor of interpretation: it will do no good, for instance, to take for granted an older philology whose sense

of objectivity was keyed to a metaphysics incompatible with his own— the Romantic tradition for instance; also, of course, the discipline of human self-interpretation clearly presupposes some account of the responsibility of all interpretive efforts. Gadamer simply fails us here.

In fact, a recent summary of Gadamer's hermeneutic theory unintentionally scuttles the very possibility of its accommodating interpretation as a cognitive discipline. Joel Weinsheimer, who is one of the translators of the revised edition of Gadamer's Truth and Method, says plainly, at the beginning of a study we have every reason to believe Gadamer would accept as accurate, that "for Gadamer understanding (philosophically considered) is an effect of history, not finally an action but a passion."7 This cannot be right (as a theory of interpretation): unless, per impossibile, judgment is a "passion" rather than an "action." The thesis ultimately entails the irrelevance of critical judgment in historical and interpretive contexts. Construed more generously, Gadamer neglects to bring his metaphysics of understanding into accord with the cognitive conditions of interpretation and self-understanding, that is, into accord with the assignment of truth-values. It cannot possibly make sense to treat such assignment as "passive," a mere "effect" of history, a matter entirely indifferent to the use of supportive evidence.

Furthermore, authors like Derrida, Michel Foucauit, Roland Barthes—or, among a rather different breed, Harold Bloom and Stanley Fish—are much more informal and allusive than Gadamer about both metaphysical and methodological considerations. And that is a pity. But to say that is hardly to vindicate Rosen's claim. It merely adds to the perceptible lacunae of some of the tantalizing theories of the day: it allows Rosen his due, but no more.

It needs to be said that radicalized accounts of (a)-(e) are not confined to what is merely current, to what Rosen finds infected with the same "postmodern" disease he pretends to find in Gadamer and Derrida. There is an older tendency that moves somewhat in the direction of the radical reading—in Charles Sanders Peirce's "pragmaticized" semiotics for instance, in the work of a near-contemporary who cannot be accused of being postmodern.8 Certainly, Peirce successfully combined the formal indeterminacy of interpretable things, their altered "nature" under successive interpretations, and the viability of referential and predicative discourse. It's true enough that Peirce does not pursue interpretation in the hermeneutic sense; but there is also an advantage in that. It's also true that Peirce believed that the pattern of

local indeterminacies and the effect of interpretation on the nature of things tended "in the long run" (and more or less irreversibly in the "short run") toward some uniquely determinate resolution. Peirce is, admittedly, an enthusiastic evolutionist. But the point remains that any fixity at the limit of infinite inquiry could not affect our judgments "in the short run."9 So Peirce's view is irretrievably flawed. Speaking strictly, there is, for Peirce, no discernible fixity or invariance in the experienced world. There is also of course the decisive, much less equivocal, ancient example of Protagoras.

Both are worth mentioning if for no other reason than that neither was influenced by Nietzsche or Heidegger. No one really doubts the coherence of Peirce's thesis; and Protagoras's may well be the most famous (and most famously contested) counterclaim to the entire sweep of Rosen's intended charge. What unites them—and, more informally, Gadamer, Foucault, Barthes, Bloom, and more and more contemporary thinkers—is a strong intuition (often more than that, as in Peirce and Protagoras) of the complexities surrounding the would-be logical necessity of the principles of noncontradiction and excluded middle when applied to an actual sector of inquiry: interpretation for instance.10 In Peirce, the critical issue concerns the moment of applicability of the principles of noncontradiction and excluded middle with respect to particular questions of predication; in Protagoras, the critical issue concerns, more generally, the inherent interpretability of a world that is a flux. When the two issues are joined—in rejecting the invariance of, and a telic progressivism with respect to, the real world—we are brought very close indeed to the thesis I am championing.

Excluded middle proves open to severe limitation, even elimination; and noncontradiction is inoperative except under the control of (interpretive) theories of the world through which it is applied. Peirce's insistence—that reality is inseparable from man's interpretive intervention (or, more strangely, the intervention of an Emersonian cosmic mind) and that, however much it is thereby made determinate, reality remains to some extent objectively indeterminate and vague—is a formula that is internally coherent, peculiarly apt for the cultural world, adequate to account for restricting the range of excluded middle and the pertinent application of noncontradiction, and suggestive about the sense in which the regularities of the physical world are themselves abstracted from a more inclusive interpretive milieu. (I am interested here only in the coherence and suggestiveness of Peirce's doctrine. I have no wish to champion it.)

It is the easy elision between formal necessity and "applied" necessity that is at fault. It may be hard to believe, but the contest Rosen takes up invokes the very same charge Aristotle turns against Protagoras. The entire dialectic of Western philosophy has found a fresh exemplar in the puzzle of interpretation.11 Peirce's contribution, here, bears more on the coherence of real individuals that have (in some respect) indeterminate natures (or partially indeterminate natures) than it does on the logic and practice of interpreting texts.

The point at stake is a profound one, easily misread. We are being reminded of one of the great subterranean contests of two thousand years of philosophy and rigorous discourse that has somehow come to a definitive confrontation in our time. Rosen is on the side of Plato and Aristotle, Descartes, Kant, and Husserl, not so much in the sense of favoring their particular doctrines as of sharing a loyalty to the generic principle of the constancy of nature and the sole intelligibility of what is constant or of what depends in a suitable way on what is fixed, determinate, and unchanging. So the theory of interpretation links up with the fortunes of that ancient quarrel: of course, Rosen saw the connection at once.

I cannot emphasize enough how neatly Rosen has grasped the dialectical crux of the entire question of interpretation in our time. Recall that, in the preceding chapter, I tried to demonstrate the coherence of Barthes's radical picture of interpretation—as far as the logic of reference and predication and of the reidentification of texts was concerned. Very simply, what Rosen manages to do is remind us that we cannot hold on to our would-be gains unless we can also show that a world (any sector of the intelligible world we care to explore) is indeed open to coherent inquiry even if it lacks an invariant structure. He exposes a presupposition of that argument, which, as it happens, belongs to the central quarrel of the whole (and most ancient part) of Western philosophy. I take up the challenge, therefore: I examine the presupposition of Rosen's countermove—at least its strongest version. Doing that, I mean to demonstrate how absolutely fundamental the quarrel about interpretation really is. If, in addition, I succeed in offsetting the charge—not merely Rosen's, but Aristotle's behind his (and, meto-nymically, that of every would-be recovery of the doctrine of invariance)—then I will have secured for the theory of interpretation (now taking form) its strongest possible footing. It may seem an extravagant labor, therefore, to turn to the ancients, but it is hardly a detour.

The irony is that both Plato and Aristotle had already grasped

the wider threat Rosen perceives—in Protagoras' dictum of "man the measure"—well before the historicizing of human nature and the artifactualizing of the whole of the intelligible world. They moved at once—Plato in the Theaetetus, Aristotle in the Metaphysics—to outlaw Protagoras' remarkable attack on the central theme of the canon (the necessary invariance of reality), which Rosen means to champion just as vigorously (in 1987) with regard to the theory of interpretation. In challenging Rosen, therefore, I am deliberately linking the themes of contemporary postmodernism, postructuralism, historicism, con-structivism, radical hermeneutics, and the like and the ancient anticipation, in Protagoras, of their generic claim.

All this is no more than a roundabout way of confirming that the contemporary quarrel regarding interpretation that has taken hold so palpably in the last half of our century does indeed implicate a grander campaign involving the canon of invariant reality and its invariant norms—all the way from Parmenides to the "Straussians," represented now by Stanley Rosen.12 Rosen is right to observe that the vindication of a theory of interpretation depends essentially on the theory of those sectors of reality that interpretation is intended to explicate. If the furniture of that world is truly invariant in structure, then interpretation cannot be as improvisational as I suppose. But if it is a world in flux, and real enough for all that, if there are no pertinent invariances to reclaim, then we must rethink what to make of the possible rigor of interpretation. Furthermore, if, along the lines of (a)-(e), human persons are also interpretable "texts," if they belong in an exemplary way to the flux, then the quarrel touches most profoundly on the fate of the whole of science and public policy. A theory of interpretation is nothing less than a theory of the human condition.

II

The issue is this. We are said to be confronted by a disjunctive conceptual choice: either the real world exhibits an invariant structure through every change, which alone renders it intelligible, or the world is a chaos, an unintelligible flux. Plato, in rather a risky way, had saved the intelligibility of the perceived world (which, on Plato's view, is not changeless in any way) by the device of invoking our dim recollection of the eternal Forms. (Or so it appears, on the usual reading of the Dialogues .) Somehow, in accord with the Forms, sensible phenomena are said to have been constructed. Otherwise, on Plato's thesis, the

world could not but be a chaos. There's the point of the labor of the Republic and the Timaeus. Plato's price is a radical separation between doxa and episteme — which is to say, Plato finds no invariance in the order of nature. Hence, in a remarkable way, if we dismissed Plato's doctrine of recollection in all its forms (for instance, in the recently fashionable innatist form of genetically hardwired concepts13 ), we would be left with something very much like Protagoras' doctrine. But even if we admit Plato's doctrine (or "doctrine," since it is hard to suppose there is no serious irony in Plato), it would be false to say, literally, that, for Plato, our world, the world we experience and act in, actually possesses or manifests an invariant structure. There is nothing to show that Plato ever held such a view: that's just the charm of his risky intelligence—when we compare him to the more sober and safety-minded Aristotle, who assigns invariance to our world. Put another way: on a literal reading of the Dialogues, the discerning of any congruity between the Forms and our perceivable world requires a skill that picks out a kind of order in the latter that is not an invariant order at all! This is too easily overlooked. (The same finding may be said to apply to Heraclitus.)

Aristotle is cannier, less daring. In Book Gamma of the Meta physics, for instance, he makes Protagoras out to be an idiot in logic, because, on his own reading, Protagoras violates the principle of non-contradiction and the law of excluded middle: he (Protagoras) does this (it seems) by reducing what is to what appears and by affirming that individual things (primary substances: ousiai, on Aristotle's reading) may possess contradictory or contrary properties.

We cannot recover Protagoras' own defense: all his works are lost. Also, I must remind you that I mean to pursue the matter only far enough to lay the groundwork for a rebuttal of the challenge regarding interpretation. But we can at least assure ourselves—I shall come to the argument in a moment—first, that Aristotle's counterattack against Protagoras is questionbegging and circular; and second, that what Aristotle offers as Protagoras' thesis can be made coherent, as against the doctrine of the invariance of nature. The result is that we may concede— if knowledge is indeed rightly so characterized only if it grasps invariant reality (by noetic intuition)—that man has no actual knowledge.

Plato had argued that even doxa must accord with the changeless Forms (to be what it is), even if humans are unaware of the connection. But, on that reading, Protagoras is decidedly sensible in denying

that there is any access to eternal Forms, in or beyond nature, by which to ensure "true belief." "Of all things," says Protagoras, "the measure is Man, of the things that are, that they are, and of the things that are nots that they are not."14

This means that Protagoras holds that man alone may decide what may be said to be the nature of the real world and what may be made to count as knowledge or true belief of it. There are no higher constraints. It's stunning, to say the least, to find that a local quarrel about what it means to read a text should vibrate so quickly through the entire history of philosophy. But there it is. All the parties to the skirmish are intuitively aware that they have been drawn back to the site of an ancient and unresolved war. The difference, now, is that the best warriors of the day count among their number more than a few who favor the protagorean forces that appeared to have been defeated in the ancient world. Furthermore, they construe the theorizing subject—man ("the measure")—in a profoundly historicized and social way.

The equally venerable canon (the one Rosen asks us to acknowledge) follows instead the immensely important dictum of Parmenides: the way of truth affirms that "it is the same to think and to be," only what "Is" is the possible subject of thought, "Nothingness is not possible [and not even possible to think]."15 In short, Rosen's huge charge obliges us to telescope the message of the entire history of Western thought within the terms of reference of the supreme contest of Western philosophy. We owe him a large debt, therefore, in having brought the matter to bear so directly on the larger fate of interpretive theory. Rosen enables us to see that the conceptual issues raised by her-meneutics and deconstruction are, however alien at first, straightforwardly occupied with the principal strife of the whole of the history of philosophy.

We must not, however, allow ourselves to be deflected too far from our original concern. As it turns out, we are close to the decisive finding. The fact that isolating it requires recovering the most ancient quarrels concedes the genuine importance of the sounded alarm and the sense in which (in spite of appearances) current disputes about the prospects of interpreting texts belong in the deepest way to the central tradition of Western thought. The upset encouraged by Derrida and Gadamer redeems, in effect, Protagoras' maneuver, which, therefore, must have been unsuccessfully outlawed in the ancient world. (It turns up so naggingly, you see.)

This is not to say that either Derrida or Gadamer actually approves

of Protagoras' strategy, but they would surely have known of its fate. What is more important: they are not committed to the ancient doctrine at all (the "canon"), although each in his own way is distinctly conservative. Derrida does not explicitly embrace the flux of his master, Nietzsche—or, for that matter, Heidegger; it would not be consonant with what he calls deconstruction, though he is committed to the irrecoverability of any changeless order. And Gadamer reacts rather fearfully to the radicalism he himself draws from Heidegger; he pretends that a historicized analogue of invariance can be recovered from the denial of real invariance, though that is surely impossible.16 In any case, if, quarrelsomely, one supposes that that is not the import of Gadamer's recovery of the "classic" values of humane Greece, then one cannot fail to be at a serious disadvantage, in championing Gad-amer, to find that he has no word to offer regarding the possible objectivity of the practice of interpretation and its "application" to oneself.

The issue is rather complicated. Rosen's double claim (a) that human nature is invariant and (b) that the very possibility of interpretive rigor presupposes (a), is a summary of the most conservative view of a literary "canon" (or a canon for any other related purpose). On the argument, the canon is that of the "classical" world—which, presumably, first discovered the invariance of the human. The tradition of literary theory has explored the possibility that both the "canonical" and the "classical" are subject to significant change over time, all the while they diverge from one another.17 Gadamer is plainly aware of all this and clearly rejects both (a) and (b). Nevertheless, he needs the "constancy" of the classical to justify the very idea of the aptness of our interpretive powers in discerning the meanings that rightly belong (or may be ascribed) to a literary or cultural tradition within the terms of the flux of history. The failure to find that constancy would lead him to relativism—or worse: he has caught himself in a dilemma of his own devising. (And so, I say, Gadamer favors a deus ex machina.)

Gadamer's hermeneutic formula seems quite straightforward: "Historical knowledge," he says, "can be gained only by seeing the past in its continuity with the present."18 Now he also regards his view of the matter as more or less in accord with Aristotle's account—if, that is, we acknowledge a caveat that applies to Aristotle as well as to his own doctrine:

If we relate Aristotle's description of the ethical phenomenon and especially of the virtue of moral knowledge to our own investigation, we find that Aristotle's analysis is in fact a kind of model of the problems of

hermeneutics. We, too, determined that application [self-understanding] is neither a subsequent nor a merely occasional part of the phenomenon of understanding, but co-determines it as a whole from the beginning. Here too application [self-interpretation] was not the relating of some pre-given universal to the particular situation. The interpreter dealing with a traditional text seeks to apply it to himself. But this does not mean that the text is given for him as something universal, that he understands it as such and only afterwards uses it for particular applications. Rather, the interpreter seeks no more than to understand this universal thing, the text; i.e., to understand what this piece of tradition says, what constitutes the meaning and importance of the text. In order to understand that, he must not seek to disregard himself and his particular hermeneutical situation. He must relate the text to this situation, if he wants to understand at all.19

As against Rosen, Gadamer is committed to: (c) the historicality of texts, and (d) the historicality of human nature. Hence, when he links his view with Aristotle's, when he speaks of the "traditional text" as a "universal thing," we cannot fail to ask ourselves how Gadamer means to recover the constancy of the universal.

The doctrine of the "classical" is the answer all right. But it is never explicitly or satisfactorily explained, and I believe it cannot be. The point is simply this: for Aristotle, although it is true that "practical reason" cannot proceed in the manner of theorizing nous, its own informality can claim to be guided by our theoretical knowledge of in-variant human nature; whereas, for Gadamer, that is not possible in principle. Hence, there is a profound puzzle that persists in Gadamer's account—that does not appear in Aristotle: namely, that if "the classic is alter et idem, "20 if "the classical . . . is certainly 'timeless,' but this timelessness is a mode of historical being,"21 then we are entitled to ask for the sense in which the "idem" can be recovered in the "alter," without invoking invariant universals.

I believe there is no account of this in all of Gadamer's writings— and that there cannot be. The "universal" cannot be captured either by a mere "feeling" or conviction of having recovered the past (which would, effectively, confirm Weinsheimer's reading of "understanding" as an "effect of history," a "passion") or by the sense of our having recaptured the past in terms of a hermeneutic continuity that we impose from our own vantage alone. Both would be self-serving.

What, ultimately, Derrida and Gadamer favor, what hardly depends on the felicity of their particular mode of attack, is the double thesis: (i) that reality is not (is not necessarily) invariant; and (ii) that coherent thinking need not be (is not necessarily) addressed to an invariant

world. Items (i) and (ii) are precisely the issues Protagoras shares with Parmenides, Heraclitus, Plato, Aristotle—and (now) Rosen: by way of championing (against the others) the contemptible option. The short summary of the "outlaw" thesis is this: there are no necessary truths de re or de dicto; any that appear to be such, so appear only under the conceptual bias of this or that particular society.22 You may resist the doctrine, but you cannot deny that it is coherent. Or, if you insist (with Rosen) that it is not, then you must recover Aristotle's argument from Metaphysics Gamma or offer a thesis of comparable power. I believe there is none to be had.

What I am tracking, then, is the impact of adopting the outlaw doctrine on the question of the rigor of interpretation. Make no mistake about it: the extraordinary issue that the new interpretive puzzle entails asks, in part, how, under the flux of history and the deep cultural contingency of every conceptual scheme, we can discern any intelligible order in any part of the apparent world—notably, in interpretive contexts, or in the sense in which every cognitive undertaking is interpretive. Viewed thus, the question before us is clearly an analogue of Plato's question, except that the Forms have been dismissed and we are left with the bold idea that every scheme of conceptual order is itself an artifact of the same inconstant condition that infects the phenomenal world and those who would understand it (ourselves).

Aristotle's argument against Protagoras is the supreme feint that has inspired the defensive line for twenty-five hundred years. It may be found in attenuated form in nearly every prominent philosopher's conceptual arsenal: for example, fashionably, if unexpectedly, in recent pronouncements by Hilary Putnam and Richard Rorty.23 But that has hardly discouraged philosophers from Aristotle's day to our own from insisting (why should one have to insist again and again?) that Protagoras was a fool or that an unqualified rejection of something like our specimen doctrines (i) and (ii) must be implicated somewhere in all rational argument—hence, in any viable theory of interpretation.

Here is what Aristotle offers as his knockdown principle: "the same attribute cannot at the same time belong and not belong to the same subject in the same respect."24 On its face, it is impossible to resist the claim. Nevertheless, it is not the abstract principle it appears to be. Aristotle is claiming that one and the same "subject" (or particular "substance": ousia, thing) cannot possess contradictory properties; and that that signifies that real "substances" cannot lack certain properties, that certain attributions that do not appear contradictory would

(if they were allowed) entail contradictory attributions nevertheless. (Protagoras implicitly disallows that complex claim in advance of Aristotle.) Aristotle is candid enough about the matter: one cannot, he says, treat all attributes as "accidents"; "there must be something which denotes substance. And if this is so, contradictories cannot be predicated at the same time."25

Aristotle's point is that discourse restricted to accidents will entail contradiction somewhere; but he never shows that that is so. It is this thesis that he conflates with the formal principle of noncontradiction. Together, they constitute the principle of the invariant nature of substance and the principle that thinking is invariantly addressed to substance—which, of course, (i) and (ii) jointly oppose. The same rationale supports Aristotle's adherence to the principle of excluded middle: for, since (on Aristotle's view) substances cannot have merely "accidental properties," all opposed properties are disjunctively predicable of substances. (In a deep sense, this is what Peirce subverts.) Restricting ourselves to accidents or appearances is said to entail either contradictory ascriptions or intolerable indeterminacies regarding contraries.

Very neat—but hardly compelling. Nothing so far said actually shows that substances (or referents of any sort) must have unchanging natures in order "to be" or "to be thought." Nothing shows that Protagoras was a fool to affirm that intelligible discourse can function coherently—while denying Aristotle's "standard" of nature. Afortiori, nothing shows that Barthes's and Gadamer's and Foucault's and Bloom's interpretive practices—all of which entail the dictum that interpretable texts are not (and need not be) changeless in their essential structure— are incoherent.

Aristotle's argument is this: (1) if all attributes are "accidental," then it will be impossible to avoid contradiction (even appearances will have to accord with the unchanging attributes of substances; so not all attributes can be mere appearances); alternatively, (2) if "substances" can be ascribed merely accidental features, then substances must possess changeless properties (whatever may possess a property must have a nature, a "definition" or "essence," but to have a nature is to possess unchanging properties).

Items (1) and (2), however, are non sequiturs. Item (1) misses the point that "accidents" are always and only relationally or contextually or phenomenally ascribed, so that it is never in principle self-contradictory to admit that x appears F (under circumstances C) and not-F (under circumstances C'). And (2) begs the question—the ques-

tion Rosen begs, supposing Aristotle to have supplied the argument— namely, that the referents of interpretive discourse (texts, or persons as texts) must possess invariant properties or natures.

Aristotle's argument appears compelling only because the would-be formal rules of noncontradiction and excluded middle are already antecedently construed as supporting (1) and (2). They set the canonical reading of the logical principles; but they themselves are not logically necessary truths. Consequently, it is entirely possible to disjoin the formal principles from these metaphysical interpretations. Once we do that, we save an inning for Protagoras (and ourselves): we demonstrate that the interpretive practice we are endorsing is not incoherent in the least. The immediate consequence of this long aside is this: cultural referents may, coherently, be said to have (or be) histories—may, coherently, be said to lack natures. There is no compelling argument in Aristotle to the effect that discourse cannot be restricted to appearances or that the "natures" of things cannot be abstracted from (while remaining no more invariant than) the phenomenal features on which they depend. Aristotle holds that existing particulars possess the attribute of "being," which is: (a) invariant, and (b) such that a particular's "nature" may be construed as an articulation of its "being." Hence, since its "being" is invariant, its "nature" is invariant. The modern view is that "being" is not a determinate predicable of any kind (or at least not a first-order predicate) and, consequently, the thesis that particular things have fixed natures cannot be strengthened in any pertinent way by reference to its "being."

III

I shall now drop the busy convenience of stalking the doctrines Rosen promulgates. There is no incoherence the ancient canon could have vouchsafed, and there seems to be no modern alternative. The solution simply requires that the constancy of what we refer to (making predications of substances or texts) be sufficient for effective reference and predication, without ever entailing that our referents must have invariant natures. That's all. It appears that I have taken the long route. All in all, it was not a bad idea: it did manage to confirm that the viability of the interpretation of texts is nothing but a special case of the viability of discourse in general. Both with regard to a thing's individuation (its possessing a "nature" or quiddity) and with regard to a thing's haecceity (its being this particular thing and not that), it now

appears that no particular in the real world, no natural or cultural referent, logically requires an invariant nature. Relative stability of structure and property and slow-enough change of structure and property (even if continual) are quite enough to secure conceptual conditions adequate for viable reidentification. That's all that interpretation requires. That's also all the doctrine of the flux requires.

This may seem a tame conclusion, and in a way it is. One cannot find many champions of the specifically Greek canon. But the point of the argument is suppler. All the apparent concessions to the doctrine of the flux that set aside the strict invariant structure of things (essences) can often be found—in our own time—to balk in ways that directly bear on the would-be coherence of the radical view. Doubts about the coherence of interpretive theories like Barthes's and Bloom's may be fairly regarded as vestigial forms of the objections put by Aristotle to Protagoras. So it is not entirely responsive to point out merely— however correctly—that the admission of real indeterminacy, the limitation or abandonment of a bivalent logic, insistence on empirical contingency, the rejection of essentialism, the acknowledged symbiosis of world and word are characteristic of our own age. All that may be conceded: we must still address all the puzzles of the logic of interpretation—under a hospitable reading of the inconstancy of nature.

We may guess at how this comes about, if we remember that there are powerful versions of the doctrine of invariance that, even as they yield to dialectical pressures similar to those that could have been mounted against Aristotle, carry in their wake commitments regarding the constancy of the conditions for reference and predication that the new puzzle of interpretation means to subvert. There is the point of this longish aside. We must remember that the commitment to the strict nomologicality of nature, as in the unity of science program, and the claims of Saussurean structuralism, as in the views of Claude Lévi-Strauss (specifically designed to offset the unmanageable flux of history), afford ample reason to suppose that there is a continuous (indeed, a perfectly pertinent) thread of argument that links Parmenides, say, to the disclosure of the putative incoherence of Barthes's account of writerly reading.26

Still, there are all sorts of considerations that may be mentioned in favor of the flux and the embattled overview of interpretation being offered. For instance, in biological evolution (in Aristotle's backyard, so to say), it has been shown that the traits by which one natural kind or species is distinguished from another must be capable of changing

over time, in a way that is consistent with the individuation of particular members of those kinds and the biological viability of the evolving kinds to which they belong.27 (So much for quiddity.) Again, no one can formulate (unless trivially—or problematically) the putatively fixed set of unique general attributes by which any and every particular can be differentiated from everything else in the universe. (So much for haecceity. Leibniz certainly knew this well enough.28 ) Hence, we have no reason to believe that the mere change or constancy of all attributes, or the fixity or nonfixity of a thing's nature, either ensures or disallows the numerical identity or reidentifiability of anything under the conditions of actual discourse.29 One excellent confirmation of this appears in E. D. Hirsch's theory of genres—in effect, a late Romantic hermeneutic companion to Rosen's essentialism (also cast in Aristotelian terms). Hirsch believes that the fixity of reference in interpretive contexts (the fixity of authorial intent) depends on the discernibly changeless genres within the terms of which interpretation may claim to be rigorous. Nevertheless, Hirsch's own candor forbids his treating genres as true entelechies or as ever explicitly cognizable. In effect, Hirsch claims the advantage of the Aristotelian tradition, but he concedes its failure at every critical moment.30

These are very deep problems. To do them full justice requires an argument too ramified for our present purpose. But I venture to say (perhaps I must) that (on Leibniz's perceptive reading of the problem of denotation and reference) only a God capable of knowing the whole of the world in one fell swoop could possibly know—and therefore use, in actual discourse—any and only general predicates to individuate, as unique, any particular thing. The purely formal solution that everything "must" be uniquely distinguishable in principle by the sole use of general predicates cuts no ice at all regarding the deliberate use of natural languages.31 Reference is preeminently a cognitive matter, one of knowing which particular things are which. To "solve" it formally—say, syntactically, without regard to cognitive assurance or reliability—is, frankly, both irrelevant and questionbegging. At any rate, without pursuing the matter in cognitive terms, there can be no basis for supposing that its resolution either entails or would support any substantive invariance affecting the logic of interpretation. (For example, to think, with Quine, that the proper name "Socrates" could, in principle, be replaced, predicatively, by some uniquely instantiated predicate, "socratizes," is to insist on invariance, even if not on essentialism.)

There can be no doubt that if these and similar considerations hold reasonably well for things in general, they must hold even more persuasively for discourse about persons, artworks, texts; for it is (now) certainly plausible, certainly not demonstrably incoherent, to treat cultural phenomena as if they possessed changing histories rather than es-sentialized natures. That is, it is (now) plausible to treat them as culturally constructed or emergent, and as subject to intrinsic change via interpretation. The objection that that is impossible cannot but fail now, if Aristotle's argument fails (that is, the modal argument). The numerical identity of persons and texts depends very largely on contingent contexts of discourse and culturally stable continuities of change, and, above all, on a society's narratized memory of its actual referential and predicative practices.32 Numerical identity requires only a sufficient measure of stability for the sake of logical or grammatical reference and reidentification: it does not require invariance. That is a stunning departure from the standard view.

Let me put this finding a little more formally: discursivity is not and does not entail essentialism with regard to natural kinds or haecceities. The operative grammatical stabilities sufficient for reference and predication are not invariant and do not entail any ontic invariances or in-variant conceptual resources. The basic reason still rests with the non sequiturs of the ancient canon: namely, that, on the argument that Parmenides, Plato, and Aristotle share, "to be" and "to be thinkable" are the same, inseparable, committed to an invariant order of reality. There you have the stalemate of twenty-five hundred years of philosophy.33 Surprisingly little has changed on that score. The implied quarrel between Rosen and Barthes, for instance, is no more than a variant of the quarrel between Aristotle and Protagoras. Extraordinary!

Here, I remind you that we have drawn on two distinct strategies: one, just now pursued, confirms that the denial of invariance and the use of reference, predication, criteria of individuation and reidentification under the condition of the flux are not incoherent or impossible to defend; the other, broached earlier, confirms that it is not incoherent to construe the human self as a historically formed construction of some sort: hence, it is not incoherent to construe human cognitive powers as themselves contingent artifacts of a changing and preformative history. Invariance may be denied in two ways then: first, against the fixity of referents (a fortiori, interpretable referents); and second, against those claims that ignore the formative contingencies affecting our cognitive powers (a fortiori, our interpretive powers). What ensures the success

of discourse regarding both matters are the collective resources of a natural-language community's habits of life (our Lebensforrnen, we may say)—without prejudice to their supposed structure. On the present argument, they are as much needed by the partisans of invariance as by the partisans of the flux.

IV

I have allowed myself to be deflected from directly pursuing the theory of interpretation. But I have outflanked the deflection, by relocating the function of interpretation in a larger space of discourse. I was obliged to follow the demur simply because there are almost no contemporary theorists of interpretation who are both distinctly influential and not committed to the doctrine of the flux or a radically contingent and changeable cultural world, and because the conceptual reclamation needed to buttress the theory is so strenuous that very few have been willing to suffer the exertion. The upshot is that entrenched philosophical disbelief about the informality of reference and predication tends to appear too easily as an effective reductio of the radical option.

One sees the point of the required effort in examining the views of Derrida and Gadamer, regardless of how they themselves present the facts. Rosen was right to force the question, though it was a foregone conclusion that he could not redeem the canon merely by attacking Derrida's and Gadamer's way of worrying it. He can certainly trace their doctrines back to the denial of invariance, but that is not enough for his purpose.

Consider Derrida's claim: "il n'y a pas de bors-texte. "34Everything discursible, Derrida maintains, is "texts," intelligible only insofar as it is located within the conceptual terms of our operative categories: texts, interpretations, interpreters, the world. Those categories are, first of all, contingent and radically variable; secondly, not "originarily" grounded, not linked to any known absolute beginning of all categorization; thirdly, not apodictic, not known without qualification; fourthly, not "totalizable," not known to enter into any idealized set of complete and inclusive categories forevermore adequate for the whole of (all) possible experience. That is the heart of Derrida's notorious marker, différance ("neither a word nor a concept").35

A related but much less purple version of what Derrida intends (without the full-blown deconstructive program) is afforded in a

characteristically mild version of pragmatism by Hilary Putnam, who explicitly subscribes to the following thinned-down Kantian-like themes: (1), favoring Kant, the indissoluble constructive link between world and word; (2), against Kant, the intransparency (at every point) of the world thus cognized; and (3), beyond Kant, the impossibility of closure with respect to all possible categories of conceptual analysis (symbiosis).36 Most recent pragmatisms show the same tendencies.

You will have noticed, of course, that 1-3 are very nearly the ubiquitous themes of all current philosophical movements—effectively, the post-Kantian equivalent of what, against the Greek canon, we collected earlier as: (i) the thesis that reality is not changeless, and (ii) the thesis that thinking may be coherent even if addressed to an inherently changeable world. Construe 1-3 as historicized, horizonally restricted, tacitly preformed, intransparently penetrated by reflexive thought: you will have placed Rosen's double theme beyond recovery. For there is no known strategy by which to establish that reality is ultimately changeless, or what, qua changeless, its invariant structures are; or to demonstrate that, otherwise, all discourse (including interpretation) is impossible. The decisive point, remember, is that the supposed necessity of the canon's doctrine of the structure of reality and thought is not itself assured. That's all: stalemate here is tantamount to victory. That is, very plainly put, I am claiming only that there are no necessary invariances—anywhere in the world. Perceived regularities and apparent invariances are easily accommodated (wherever they appear) under the flux of historical experience; but the piecemeal—modal— necessity of such structures can be coherently abandoned. (In this, I agree with W. V. Quine's famous attack on the "dogma" regarding would-be analytic and synthetic distinctions.37 )

In any case, Gadamer, in a spirit very different from Derrida's, also concedes the contingent, changing, historicized conditions under which human beings emerge in the social space of their particular culture. But if they are formed in distinctly different contingent milieux, then their most deliberately formed artifacts (texts, artworks, and the like) must similarly be infected with the same changeableness. So Gadamer says:

True historical thinking must take account of its own historicality. Only then will it not chase the phantom of an historical object which is the object of progressive research, but learn to see in the object the counterpart of itself and hence understand both. The true historical object is not an object at all, but the unity of the one and the other, a relationship in which exist both the reality of history and the reality of historical understanding. A

proper hermeneutics would have to demonstrate the effectivity of history within understanding itself. I shall refer to this as "effective-history." Understanding is, essentially, an effective-historical relation.38

If, now, you add to this Gadamer's explicit insistence: (a ) that "we are always subject to the effects of effective-history. It determines in advance both what seems to us worth enquiring about and what will appear as an object of investigation,"39 (b ) that "language is not just one of man's possessions in the world, but on it depends the fact man has a world at all,"40 (g ) that "language speaks us, rather than that we speak it,"41 and (d ) that "there is nothing like an 'I and thou' at all— there is neither the I nor the thou as isolated, substantial realities. They are only marked off within a common understanding [Verständi-gung ] [that] always precedes these situations [a shared world, a shared language, a shared tradition],"42 then: first of all, the plausibility of replacing an (essentialist) nature by a (radically contingent) history (or careers) becomes entirely understandable; and, second, the plausibility of dismissing Gadamer's own somewhat "essentialist" misgivings about having adopted the first also becomes plain.

Gadamer says both:

The historical movement of human life consists in the fact that it is never utterly bound to any one standpoint, and hence can never have a truly closed horizon. The horizon is, rather, something into which we move and that moves with us. Horizons change for a person who is moving,

and

when our historical consciousness places itself within historical horizons, this does not entail passing into alien worlds unconnected in any way with our own, but together they constitute the one great horizon that moves from within and, beyond the frontiers of the present, embraces the historical depths of our self-consciousness. It is, in fact, a single horizon that embraces everything contained in historical consciousness.43

These remarks are ultimately incompatible: the second is meant by Gadamer to be a sort of very bland advance toward recovering an effective (or near-) invariance of norms in history against the inherently inconstant reality acknowledged in the first. That is the point of Gadamer's conception of the "classical." Its abiding mistake is this: on Gadamer's argument (the openendedness of history), the "one great horizon" he speaks of cannot be assigned number (cannot be individuated)—in that sense in which, for whatever reason, we retrospectively individuate the contingent "horizons" of this thinker and that.

The recovery of the "classical," therefore, cannot but be relativized to "one" horizon or another; but, that is to say, its universality (normative invariance) is a phantom of Gadamer's own horizonal hopes. It cannot be specified and it cannot rightly be claimed. The "universality" of the universe of discourse has nothing as such to do with the universality of the categories and norms of understanding. It is a kind of spurious Hegelianism.

The charge may be doubted. Nevertheless, there is a puzzle about "the one great horizon." It may signify that all discourse falls within "one" horizon—in a merely formal sense that lacks number altogether (permits no plural) and is indifferent to all content. On that view, even the discernment of an "alien world" or the posit of an incommensurable conceptual scheme falls within the "one" horizon. (The point is Husserl's.)44 Or, it may signify more robustly that whatever falls, synchronically, within "one" horizon is, as intelligible, necessarily such in virtue of being in accord with universally applicable categories— which entails a confusion between the would-be inclusive scope of our understanding (in the first sense) and the mere conceptual consensus of all those who interpret what (in historically overlapping ways) falls within the "one great horizon." There is no other pertinent sense that Gadamer could supply, and there is none that would serve him. Now, it is true that Gadamer never explicitly subscribes to the doubtful claim. In fact, he regularly opposes it. It is more nearly the view of Jürgen Habermas.45

But Gadamer is very fond of deliberately intruding the humane notion of a "common nature" of mankind where his own hermeneutics would put such a presumption in doubt. For instance, he favors Hölderlin's remark that "the thoughts of our common spirit are completed quietly in the soul of the poet." He says that "in a composition belonging to the lyric genre, it would not be the I of a poet [that is, the first-person discourse of the lyric poem] if it were not to become the I of everyone."46 I admit that this is open to diverse readings; but the sense of it, I think, is that, insofar as the poet moves "all of us" in whatever way the poet does, through changing history, his power attests to some stable normative constancy in human nature itself. This seems to be related to Gadamer's more recent drift back to the "autonomous" object of philosophical interpretation (the "continental" analogue of New Critical interpretation).

You will find the same theme more explicitly favored in Gadamer's account of the "idea of the good" in Plato and Aristotle. For there,

Gadamer not only explicates the classical notion but assimilates its advantage to his own hermeneutic vision. Gadamer speaks for both Aristotle and himself (without adopting Aristotle's invariant world) when he says

insofar as the world of human practice is located within the entirety of what exists, the whole sphere of human praxis (action) and poiesis (doing) has its place within the realm of nature. Not only art imitates nature. Human practice does so too insofar as it aims at nothing other than the highest fulfillment of human existence itself.47

In anticipating this finding, Gadamer explains that "the good" or the "one" (in Plato and Aristotle) "is that which on any given occasion provides what is multiple with the unity of whatever consists in itself. As the unity of what is unitary, the idea of the good would seem to be presupposed by anything ordered, enduring, and consistent."48

All that can be said in Gadamer's favor (now) is that he fears what Rosen fears. Rosen is more consistent than Gadamer, because he refuses at the very start to replace man's invariant nature by a constructed history, whereas Gadamer believes (quite unaccountably) that he can hold on to a historicized human existence and, at the same time, to an effectively changeless order of meanings and values (or better: a historically contingent but fortunate set of conceptually universal norms of understanding and humane values that history somehow never erases or deforms).49

Alternatively put: Gadamer believes that the very fact of communication signifies the universality of human nature at every point of effective understanding. But he does not explicate what we are to understand by successful "understanding." If understanding is historicized and only consensually confirmed—pragmatic or honorific in that sense—then universal criteria of communicative success are irrelevant. Even what is judged the failure of one culture to understand another could, on Gadamerian grounds, be said to confirm the universality of human nature (the "one" horizon, the "classical"). The difference between Gadamer and Habermas here is merely that Gadamer takes it that the universality of the human is never lost, whereas Habermas believes that determinate universal norms of communicative success are progressively worked out by a rational, dialogic critique of any posited norms. The first (Gadamer's view) is utterly vacuous; the second (Habermas's) is surely more than doubtful. (The issue will arise again.) For our present purpose, it is enough to say that the "universality" of

communication is indifferent to the difference between a genuinely neutral language that all may acquire and the piecemeal practical success of local communication, generalized for all such occasions. No one can formulate a neutral language; and no one can show that the informal features of communicative success entail a universally shared semantics.

Gadamer does indeed lay out the general lines of a radical account of interpretation, but he is much too fearful of its implications to risk pursuing it in either metaphysical or methodological terms. We need not follow him in this. It is, first of all, impossible to reconcile a universalism of invariance and a radical historicism: Rosen grasps the point perfectly, whereas Gadamer either does not or will not risk discussing it.50 Second, it is, as I have argued, conceptually gratuitous to claim, as Rosen does, that any departure from a universalism of invariance is impossible to defend. It follows that a theory and practice of interpretation thus "deprived" cannot be shown to be incoherent.

The truth is, Gadamer abandons airing the radical possibilities at the very moment he gives them form; also, although Derrida never descends to any such explicit contradiction, his own deconstructions are too thin and too parasitic, and his "constructive" work too indifferent, to help us meet Rosen's challenge. For Derrida simply means to leave things pretty much as they are (whatever that comes to), all the while he demonstrates again and again that his own textual interpretations cannot be derived from différance. In a word, Derrida has no theory in terms of which to test (in any sense at all) any of the bold interpretations he actually offers in the body of his work. The pragmatists, Putnam and Rorty in particular, have nothing to say about the rigor of interpretation; and would-be deconstructive critics, Christopher Norris for instance,51 have little to say in the way of a "constructive deconstructive" mode of interpretation.

But we are still whistling in the dark. We need a better sense of just how the intended replacement works. Let there be no misunderstanding. No genuinely distinctive theory of interpretation in our age denies most of what has been affirmed in this chapter: that is, (i)-(ii) the rejection of invariance, (a)-(e) selves as texts, radical historicity, constructivism, the disjunction of discursivity and essentialism, the incompatibility of universalism and historicism. The countercharge claims that to embrace these doctrines renders all theory impossible, incoherent, self-contradictory. I have shown that there is no satisfactory demonstration that the charge holds, and I have also shown that there is a

certain tendency among the would-be champions of the radical cause to wobble in advancing it. What I have not yet managed to do is identify the formal conditions that would make the radical option reasonably viable, or specify the substantive conditions that could yield a genuinely disciplined interpretive practice. Fair enough.

The required clues are in the neighborhood, however. They have been obscured in the scatter of our first skirmish. So let us have them explicitly before us: (A) the conditions of discursivity are not, or do not entail, the constraints of essentialism or of an invariant nature; and (B) this (A) holds for physical nature just as much as it does in the context of human culture. The pretty thing about this argument is that (A)-(B) do not entail any determinate theory of texts or interpretation and do not entail (or forbid) the replacement of "natures" by "histories." In the context of modern physical science—where, although theories of nature may be said to have histories, the pertinent phenomena and posited theoretical entities either lack histories or possess histories only extrinsically—the most advanced work hardly entails essentialism or strict invariance.52 It may tolerate essentialist accounts here and there; but no one believes any longer that it is positively incoherent to construe science along nonessentialist lines. Some theorists, notably Karl Popper, actually hold that a plausible theory of physical science must be opposed to essentialism and determinate universal in-variances.53 Science prospers, Popper believes, only by resisting such doctrines, both methodologically and metaphysically. But, of course, if this much holds for physical nature and physical science—where truth-claims are clearly at a premium—then it cannot fail to hold as well for interpretive discourse.

V

I come, finally, to the constructive moves required. It is an essential part of the canon—certainly a part of Aristotle's influential version— that things are individuated with respect to their generic nature and remain this particular rather than that in virtue of whatever belongs to their haecceity. If, now, with respect to persons, artworks, and other cultural artifacts, we agree to substitute individual (individuating) histories for generic invariances (against Aristotle) or invariant haecceities (say, against Saul Kripke), we relieve ourselves at once of what is usually regarded as an insuperable contradiction. The dictum, that a thing cannot change its "nature" without becoming another thing,

proves to be utterly vacuous; for, now, persons and texts have no fixed natures (either generically or qua "thisness") that they can violate. Gadamer, as we have already seen, allows for self-change through hermeneutic practice; he is opposed to fixed essences, though he is coy enough to avoid saying whether hermeneutically affected self-transformations can constitute a change of "nature." But Michel Foucault is clearly disposed to say that man can change his "nature," in the very process of the radical freedom Foucault was drawn to in his later work.54

The point is that, since he replaces "nature" by "history," Foucault cannot be accused of the usual contradiction the canon would insist on; and since, as I have shown, the canon is not unavoidable, there can be no a priori objection to such a replacement. Correspondingly, if a text or artwork has no fixed nature, then its ongoing interpretation—the historicized process of interpretation itself—may change its "nature" without producing incoherence or self-contradiction and without jeopardizing or losing its identity. (I am preparing the ground for admitting the reality of history and interpretable referents.)

Roughly, on the substantive side, all that is required is a reasonably continuous process of interpretation over time that preserves the public identity of "what" is affected and the sense in which its "nature" may be altered from step to step. Numerical identity among texts and artworks is, in context, a matter of the narrative history of description and interpretation (what I am calling "unicity," numerical identity through change). Identity is Intentional; a Leibnizian account is impossible; reliance on remembered context is unavoidable; and strict invariance is lacking. Why then should the adjustment be disallowed? The purely formal objections have proved indecisive; the ontological constraints (Aristotle's) have turned out to be questionbegging. The truth is that, once we extend the argument so that it applies to the puzzles of predication as well as to those of reference (that is, to the reality of "universals"), it becomes clear that a society's consensual recollection of its own history of predication leads directly to the finding that the invariant "natures" of physical objects (presupposed by a realist reading of the laws of nature) are themselves artifacts of an ongoing science. But then, that too subverts the supposed de re and de dicto necessities I am opposing; and, by resisting the disjunction of science and its history—the disjunction of (worlded) discourse and (discursible) world, the disjunction of theory and perception—it suggests a sense in which the real world is itself interpretively complex. (You will find that Peirce has anticipated the point.)

There is a very strong intimation of this argument to be found in the famous case of the ship of Theseus, which lacks the inherent Intentional complexities of the interpretive problem (though it does rely on intentional considerations). Again, Saul Kripke's solution of the problem of haecceitas in terms of causal or "christening" continuities is already a narratized, contexted, and external ground for fixing identity, and is recognized by Kripke himself as subject to being overridden by deeper intentional considerations.55 Thus, if an error arises in the individuation of the christening circumstances, it might be satisfactorily corrected by reference to other contextual factors.56 Kripke never offers criterial advice about the determinate haecceity of anything in particular; and the assurance that particulars may be "rigidly designated" (in every world in which they exist) is itself managed narratively, that is, without reliance on the use of known rigid designators and in accord with a society's consensual memory of how, contextually, it has ensured reference and reidentification.

In any case, if reference cannot be retired by way of uniquely instantiated general predicates—not because the idea is impossible but because no human speaker could ensure its adequacy in any or every instance—then no appeal to quiddities or haecceities will serve that purpose. The first option I associate with Aristotle's never having properly brought his account of essences to bear on the logic of individuation; the second, with Duns Scotus's failed attempt to do so by a restriction of essences. It needs to be remarked that difficulties of the reverse sort also obtain. That is, if I identify a particular thing— whether by proper name or indexical expression ("Peter" or "that book")—it does not follow either that the properties of what is thus identified are criterially determined by the concepts by which I successfully identify it, or that I must have identified it by reference to some natural kind of which it is an instance or that natural kind of which it is normally said to be individuated as the particular instance it is, or even that I must have intended to invoke in a criterial way any of a range of natural kinds or kinds of which it is an instance.57

Perhaps a reference to the interpretive literature will help us see the bearing of these subtleties on the reidentification of texts. Harold Bloom's way of reading offers obvious exemplars. His little book, The Breaking of the Vessels, the inaugural volume of the Wellek Library Lectures, may be too gymnastic for easy citation. But the following, which conveys Bloom's use of the notion of "transumption," may fix conveniently enough the theme of poetry as a kind of open history, a

history linking poetic acts or utterances under the force of influence— that is, reading under "anxiety," reading as a form of "misreading" or "misprision," "reading forward" (in Kierkegaard's sense of "repetition"), the role of "negation" and "defense" in the poet's functioning as poet. "When I was young," says Bloom, "I did not understand that one was reading [Walt] Whitman, whether one knew it or not, when one read fully in [Wallace] Stevens or in [T. S.] Eliot, or in [Hart] Crane."58

With that clue in hand, consider this:

Influence, as I conceive it, [Bloom says,] means that there are no texts, but only relationships between texts. These relationships depend upon a critical act, a misreading or misprision, that one poet performs upon another, and that does not differ in kind from the necessary critical acts performed by every strong reader upon every text he encounters. The influence-relation governs reading as it governs writing, and reading is therefore a miswriting just as writing is a misreading. As literary history lengthens, all poetry necessarily becomes verse-criticism, just as all criticism becomes prose-poetry.59

We need not bother our heads (here at least) about Bloom's exotic thesis that "Lurianic Kabbalism [Isaac Luria, a sixteenth-century Kabbalist] [is] the ultimate model for Western revisionism [the search for an original source of truth through a precursor text] from the Renaissance to the present."60 The important thing is this: "To live, the poet must misinterpret the father [in a poetically pertinent Freudian-like sense], by the crucial act of misprision, . . . the re-writing [revision, transumption] of the father [that is, the precursor text]." In this sense, "Poetic influence, in the sense I give to it [Bloom says], has almost nothing to do with the verbal resemblance’s between one poet and another."61 "Strong poets," he adds, are those, within the space of poetic influence, who "make [poetic] history by misreading one another, so as to clear imaginative space for themselves."62

Bloom's model is not meant to be the exclusively apt model of all reading or interpreting. It is deliberately idiosyncratic. It is aware of other possibilities, which it shuns. For instance, Bloom is very clear about his repudiation of T. S. Eliot's "malign influence" in offering up the model of reading recommended in "Tradition and the Individual Talent." But that, Bloom declares, "was hardly [the voice of the strong poet,] Eliot [himself]."63 (His own rejection of Eliot, of course, is the work of a "strong" critic.)

Possibly one of the most striking of Bloom's illustrations appears in his analysis of Nathanael West's Miss Lonelyhearts, which he reads

as "another displaced version of the Miltonic/Romantic crisis-poem": "Miss Lonelyhearts is Jesus Christ in Depression America, and the sublime Shrike is a very American Satan. The precursor work is certainly Milton's Paradise Regained, a poem for which Milton accurately implied The Book of Job was ancestor. Miss Lonelyhearts is one of Job's descendents as are also Wordsworth's The Borderers and Blake's Milton. Again, whether or not West ever read it, Paradise Regained is uncannily close to Miss Lonelyhearts, which I would insist is the superior work, and perhaps the strongest of Job's varied progeny. West's strength as compared to Paradise Regained is that Shrike is a far more resourceful and subtle tempter than Milton's late Satan, who hardly deserves the name of the great villain-hero of Paradise Lost. "64

Remember, I am not concerned here to decide the best form of criticism and interpretation. I wish only to count as viable and plausible (against Rosen's charge and others of a seemingly more temperate sort) a radical conception of interpretation that eschews essentialism and denies the primacy of invariant properties, that replaces nature by history and permits the transformation of the "nature" of a poem under interpretation, that tests structural similarities between poems, and assesses texts comparatively—and does not sacrifice numerical identity. Bloom—and Richard Rorty, who follows Bloom—fails, it is true, to discuss the nature of the form of rigor that accords with his own model of critical reading.65 But that is what I am attempting to supply.

In fact, once we have Bloom's model in hand, we can invent a whole series of intervening models, more restricted than Bloom's but committed to certain minimal conditions on interpretation opposed to Rosen's constraint. That is all we need. On Bloom's thesis, poems or texts are no longer "objects," "things," Aristotelian substances; yet they remain individuatable and reidentifiable as real referents over a span of interpretations of Bloom's sort. For example, the Greek tragedies— Euripides' plays, in what is very nearly a modern sense (say, reading Iphigenia or Medea or Phaedra ) —lend themselves admirably to being construed as "misprisions" of the Iliad and other epic materials, even if under more restrictions than are countenanced in Bloom's own game. That is, Aristotle's treatment of the tragedies, in Poetics, can easily be relaxed along Bloom's lines, without jeopardizing the syntactic requirements of discursive reference and predication—or reality.

The poem or text, though it is not merely the string of words (or worse, the physical marks designating the words) on which a given

text depends, can, in the context of a relevantly functioning culture, effectively fix the numerical identity of such a text for interpretive purposes of the Bloomian sort.66 And, if it can do that, it can do the same for any of a vast array of interpretive practices in which some suitable subset of the following conditions obtain: (a') the text (or poem), though Intentional in nature, has no fixed or invariant or essential nature; (b') every text possesses a relatively stable verbal medium (the gathering past of a living language) by the use of which it may (normally) be individuated and identified, and by reading or interpreting which the "nature" of what is thereby constituted may be discerned; (c') the (Intentional) "nature" of any particular text is itself open to change as a result of an ongoing process of interpreting the precursor (interpreted) phases of that same text; (d') the interpretation of a text is an openended function of the power of its verbal medium and history of interpretation to absorb further evolving saliencies of the enveloping culture in which it is read; (e') there are no invariant constraints on admissible interpretations of particular texts that can be legitimated by reference to the invariant generic nature or haecceity of those same texts. Conditions (a')-(e') effectively entail replacing "natures" by "histories" or "careers"; but they also countenance various sorts of useful stabilities, regularities, empirical or "indicative" invariances (as we may term them) of verbal order, habits of reading, traditions, and the like ("forms of life" if you care)—by means of which discourse (intelligible reference and predication in particular) may be counted on to work reasonably well.67

To admit all this is hardly to endorse particular critical practices— Bloom's or Eliot's or Aristotle's or Barthes's or Derrida's or Gadamer's or Stanley Fish's or E. D. Hirsch's. But that is another matter altogether. The issue here is that we cannot disallow any of these or similar alternatives simply on the extravagant supposition (Rosen's) that an accommodation of (a')-(e'), by one or another interpretive practice, instantly consigns that practice to incoherence, contradiction, collapse, or utter chaos. The point at stake—it is worth repeating—is simply that all the discursive functions of interpretive criticism are, effectively, functions of a viable societal practice reflexively aware of (capable of recovering) its own past habits with regard to reference, explanation, evaluation, and the like. There are no rules without consensual memory; there are no rules that can replace such memory; and such memory functions without rules itself.

Let me remind you, then, of the quarrel of the opening chapter. I

am in effect saying that Barthes's theory of writerly reading is coherent, does not sacrifice at any point the pertinent rigor of reference and predication, is a paradigm, therefore, of indefinitely many alternative models of interpretation favored at the end of our century. Furthermore, Krauss's recommendation to abandon reference is neither required by Barthes's sort of theory nor defensible in its own right. To press the point: I suggest that Barthes is writing as a poststructuralist and Krauss, only as a postmodernist. They converge, in the important sense that both abandon the notion that a text must: (i) manifest a fixed nature, (ii) exhibit determinate boundary properties, and (iii) possess properties that (in any case) are not affected by interpretation alone. They differ in that, adhering to (i)-(iii), poststructuralists, but not postmodernists, do not sign away the resources of discourse (reference, predication) by means of which truth-claims (a fortiori, interpretive claims) may be processed at all. What we begin to glimpse, therefore, is that, although there must be some sense of the discipline of interpretation, some fairly regularized practice, there need not be any uniquely correct such practice and there may be many diverging and differently motivated practices compatible with exploiting and sustaining a common history but not necessarily with one another. Ultimately, the argument rests with the fact that both reference and predication are inherently, even radically, informal. There is no fixed rule by which reference can be secured, and predication cannot be decided by appeal to changeless universals.

The Saracens of criticism now live among us. They do have an eye for empire, no doubt. But they are not barbarians.

Chapter 3
Prospects for a Theory of Radical History

I

In an affectionate aside on Giambattista Vico's conception of history, which in our own time might have led Vico himself to a fresh reading of technology, Hannah Arendt observes:

In the modern age history emerged as something it never had been before. It was no longer composed of the deeds and sufferings of men, and it no longer told the story of events affecting the lives of men; it became a man-made process, the only all-comprehending process which owed its existence exclusively to the human race. Today this quality which distinguished history from nature is also a thing of the past. We know today that though we cannot "make" nature in the sense of creation, we are quite capable of starting new natural processes, and that in a sense therefore we "make nature," to the extent, that is, that we "make history." . . . We can do in the natural-physical realm what Vico thought we could do only in the realm of history. We have begun to act into nature as we used to act into history . . . technology has emerged as the meeting ground of the natural and historical sciences.1

There are difficulties with Arendt's account of science that need not detain us, except to note that she mistakenly supposes that technological intervention in nature, of whatever profound sort, somehow entails a breach of "inexorable [causal] laws."2

Not only is this a conceptual blunder, it obscures a deeper sense in which we do "make" or constitute nature. In fact, Arendt obscures the sense in which we also make history and the world of human culture.

For, we make history not because we "act into history," but because history is a human construction. We also make or "constitute" nature—not merely because we literally transform nature by technological intervention but because the only nature we address intelligibly is a "nature" already (somehow) formed (or preformed) in accord with whatever way human understanding functions; and because, by such "constructions" (that, pace Descartes, Kant, and Husserl, cannot be assuredly sorted between the would-be invariances of the mind and the would-be invariances or contingencies of the "brute" world), we equilibrate the apparent structures of the one and the apparent structures of the other. (I say that language and the world are "symbiotized," that we are all "Kantians" to that extent—Arendt included— or, better, "post-Kantians," since we cannot fail to repudiate Kant's uncompelling disjunction.)

Arendt herself affirms, at the beginning of The Life of the Mind: "Being and Appearing coincide,"3 by which she means to unify her reading of the entire history of Western thought and, in particular, to join the executive themes of Kant, Hegel, and Heidegger. Her intention is not to bifurcate appearance and reality but (following Cato's mot— "never is a man more active than when he does nothing, never is he less alone than when he is by himself"4 ) to reserve a place for a higher contemplation of the meaning of history. Human deeds are confined to the domain of appearance; but since "Being and Appearing coincide," we must (she insists) be alert to the hidden meaning of what we do.

Here, she favors the vision common to Kant and Hegel: "it is not through acting but through contemplating that the 'something else' [that is, the significance of what men do, apart from what they intend and hope and fear, which cannot be 'consciously willed by the actor'], namely, the meaning of the whole, is revealed" (whether, as with Kant, to a plural "spectator" or, with Hegel, to the singular "spectator" who "becomes the organ of the Absolute Spirit").5 The vulgar activism of history, she remarks, the presumption of immediate progress, the assurance that the meaning of what we do rests entirely in the intentional structure of our immediate plans and deeds—which she associates particularly with Marx—utterly betrays the deeper contemplative possibilities reflections on the French Revolution have made possible.

She now pursues her reading of the laws of nature, technological intervention, the distinction and unity of nature and history, in order to ensure the more profound lesson. But there is no need (then) for the extravagance remarked. One small adjustment brings the entire

argument into accord with what has already been said: the would-be higher "spectatorial" meaning of history (that entails "a withdrawal from the world as it appears and a bending back toward the self"6 ) is itself no more than a contingent "appearance" within the same apparent world. There you find the clue to what is more challenging in Marx and Foucault. It's not the vulgarity of political activism (though there is that); it's rather the threatening vulgarity of the continued presumption of a hidden world embedded in appearances, revealed to the modern successors of the ancient oracles—the last vestige of invariance in the flux itself.

The theme of history is deliberately skewed by Arendt in this sense; although, quite frankly, the correction I have in mind is also meant to alter Vico's characteristic insistence on the human production of the cultural life of nations as well as Kant's insistence on the species-wide necessities of human understanding. Its intended benefit, therefore, will have to wait a bit to be made entirely clear. For the moment, we may note that the constructive nature of history is different from the constructed nature of physical nature. The failure to mark the distinction risks confounding, utterly, the nature of human history, human understanding, human interpretation.

If, thinking along these lines, we construe history not only as a narrative but as the actual temporal process that is human life and, by extension, the distinctive career of cultural events and artifacts, then it is still true that history is a human construction—which is to say, man makes and remakes himself and his cultural world. If "nature" is, in some measure, an artifact of that same (symbiotized) process, then nature too is "made." But it would remain true nevertheless that, apart from its artifactual standing, nature (as distinct from culture) could not be the setting of history in the same complex sense in which the human world is inherently historical. That is why Arendt's odd confusion between history and nature needs to be resisted: it makes a shambles of both.

Of course, one is hardly bound to say, saying that, that nature, human nature, human culture, history, artifacts of any kind are manufactured merely, constructed through and through—composed entirely of ideas or fantasies, merely idealist. At every point of human intervention, we acknowledge the resistance of a brute world—secondness, as Peirce perspicuously remarks7 —including our own existence and the things of our cultural milieu, the articulated structure of which is (once again) specified only artifactually; and if we cannot force a prin-

cipled distinction between "the" world and what we verbally impute as the world's structure, then the world is a "construction"—but "independent" for all that. Our intervention permits no such extreme disjunction as Arendt suggests, even between physical nature and human history. That is more than Vico bargained for. But it is no more than what remains of the Kantian thesis deprived of Kant's optimism regarding the mind's distinct contribution and the presence of noumena. (Needless to say, I am extending the range of the quarrel of the previous chapter.) A provocative way of making the point is this: currently, by historicizing the symbiosis of world and word—against Kant himself—there is no longer any strong sense in which "realism" and "idealism" can be applied disjunctively to the achievements of science any more than to the achievements of the interpretation of the arts.8

That we insist nevertheless on the distinction, that we are tempted by "inexorable laws" in science and deny to nature anthropomorphic traits (human intentionality in particular), testifies to our sense that, though we "make" the world, we cannot make it any way we please. We open our eyes and see a world we cannot ignore: still, what we see is due to what we are; and what we are we are as a result of our continuous self-formation and transformation within a larger history and the larger processes of nature. So the "resistance" of the encountered world is not at all incompatible with its being "constituted." On the contrary, a temperate realism now requires such a compatibility. Put in the simplest terms: the apparent invariances of physical phenomena are just such invariances noted under the changing, formative conditions of human history. We posit them, from within that history, as realities independent of that history.9 (In Peirce's extraordinarily concise account, the brute resistance of the world—its "secondness"—is both uninterpreted as such and not explicable as the mere effect of any interpretive intervention. That is at least the underlying intuition of what exists, as opposed to what is real [but is not real in the way of existing], the predicable natures of particular things for instance. What is real in the latter regard is and must be interpreted—implicates "third-ness," in Peirce's terms. I take the distinction to mean, among other things, that no constructed scheme of individuated things can fail to be brought into accord with secondness; but that that alone cannot displace the determinate work of thirdness. I accept Peirce's distinction in this spirit.)

One is tempted to say that conjectures of this sort cannot but be incoherent—if plumbed sufficiently deeply. But it is not true. One way of

securing the point is this: the very laws of nature are artifacts of a historicized science—both in the sense that their supposed invariance is an idealization no sustained inquiry can ever discern, and in the sense that the phenomenologically lawlike regularities we do perceive, that "strict" laws idealize, are themselves identified only in the context of the tacit theories by which our perception is horizonally directed in the first place to what "resists" us and other things (that is, to what exists). If this holds for physical nature, it cannot fail to be even more compelling when applied to cultural phenomena. That the apparent significance of finite segments of human history harbor this or that "universal meaning" can hardly be more than an expression of our own prejudices regarding the norms and essential needs of the entire race.

When, therefore, within the context of a hermeneutic history, Hans-Georg Gadamer declares that

the classical is fundamentally something quite different from a descriptive concept used by an objectivizing historical consciousness. It is a historical reality to which historical consciousness belongs and is subordinate. What we call "classical" is something retrieved from the vicissitudes of changing time and its changing taste . . . it is a consciousness of something enduring, of significance that cannot be lost and is independent of all the circumstances of time, in which we call something "classical"—a kind of timeless present that is contemporaneous with every other age. So the first thing about the concept of the classical (and this is wholly true of both the ancient and the modern use of the word) is the normative sense . . .10

he must be whistling in the wind, he must be betraying his own conception. In any event, that the invariance of the laws of nature and the universal significance of human history are artifacts of our own careers is easy enough to entertain, without being obliged to suppose they are, for that reason, merely arbitrary posits. They are posits of some original and sustained symbiosis of cognized world and worlded cognition in which we cannot determinately segregate (on independent "realist" grounds) the contributions of mind and world. (In short: symbiosis is a Kantian-like—even a Kant-inspired—thesis pointedly directed against Kant himself.)

Gadamer is entirely justified in searching for "classical" invariances every bit as much as the physicist is entitled to search for the exceptionless laws of nature. The point is that whatever either finds, he finds within a texted, historically perspectived space; hence, whatever he finds is subject to its contingencies. The question remains as to what those contingencies entail: the physicist admits that inquiry has a his-

tory, but he claims that the nature he fathoms is not historied in the same sense in which his inquiry is; Gadamer insists on the historicity of the human world and the peculiar intimacy of our reflexive inquiry into it, and yet he seems to claim, there, some historicized analogue of natural invariance. He offers no supporting argument. The truth is that he seems to believe, in spite of the historically constructed nature of human persons, that the interpretation of the texts of the historical past already ensures the universal community of mankind (which the "classical" captures in a normative regard). But the interpretation of the sense of what the past deposits does not entail that there is a prior sense that interpretation grasps or grasps correctly. The thesis is impossible to defend, and the universality of the conditions of human understanding is a most unlikely conviction—under historicity. Gadamer holds that horizonal understanding entails a "fusion" of horizons— which is a "universal" condition of mankind; also, that that "fusion" confirms our belonging to "a single horizon that embraces everything contained in historical consciousness."11 The first is true enough, but vacuous in any interpretive sense that ranges over particular texts; the second is simply a very poor (entirely arbitrary) inference.

A good deal hangs on conceding the point. Hannah Arendt is perhaps too much absorbed by the great evil of the political events her remarks were meant to illuminate. She does not grasp the full implications of her own modernity. She admits that history is made; but not having correctly fathomed what that signifies, she does not quite grasp what history is about, what a "living" history is, what it is the history of, what the sense is in which histories are made. Ultimately, she misses the radical "nature" of history itself, that is, that history bas no nature: that what is historical "in its nature" has no nature, no quiddity, no assignable essence, no de re necessity. This is the radical theme I mean to confirm, the important lesson I draw from her confounding of history and nature.

There is a predictable bafflement such remarks set off. At some point it must be met in two quite different ways: in one, to convince us that the idiom is not incoherent or self-contradictory or unusable; in the other, to convince us that it is a distinctly apt idiom, one to be preferred in our own time as a consequence of the changed conceptual options we have come to consider at the close of the century. We have made some gains in both regards.

Arendt says "the modern concept of process pervading history and nature alike separates the modern age from the past more profoundly

than any other single idea."12 But this is almost certainly mistaken. Her sense of moral unease—of terror—comes from the extreme unpredictability of nature and history now that "we have begun to act into nature, [now that] we have manifestly begun to carry our own unpredictability into that realm which we used to think of as ruled by inexorable laws."13 (She has the Nazi example in mind, of course.) This means that she believes that nature, like history, now but not always, is in danger of not being bound by the "laws of nature."

The argument is unacceptable but it is strategically important—and not because it is merely foolish to think that there are no real laws of nature, no inexorable regularities that obtain everywhere without exception (no so-called nomic universals). For, though in the heyday of the unity of science movement it would have been unthinkable to deny the operation of the laws of nature, it is now quite fashionable (and coherent enough) to affirm straight out that nomic universals are fictions or idealizations of some sort, introduced in accord with the supposed requirements of explanation in science.14 This, of course, is the way any constructivist would redeem the apparent stabilities or invariances of the formal and empirical sciences—but it is not Arendt's way. Empirical regularities need not be denied, but they need not be taken to drive us to any sort of essentialism or progressivism or the like. That is, the denial of nomic necessity is not tantamount to the confusion of nature and history. That's what we must be clear about. Causality may obtain in history as well as in nature (ultimately, the two cannot be disjoined but they are also not the same); furthermore, to admit that much is not yet to admit nomologicality. For, for one thing, causality does not strictly entail nomologicality; and, for another, the nomological need not be causal.

The unpredictability of history would remain unchanged even in a completely deterministic world: that couldn't possibly be what "separates the modern age from the past more profoundly than any other single idea." No, what separates the "modern age"—the present age— is the dawning notion that what has an intrinsic history has only a history, no "nature" at all; and that, as a consequence, what we posit as the "nature" of (historied) things is altered and affected (in as intrinsic a way as can be imagined, once we abandon fixed natures) by the historical process itself, by interpretation and purposive acts reflexively and historically applied to what is historied by "nature." This op tion—radical history, as I shall call it—has barely been recognized, certainly not supported, in the larger part of contemporary philoso-

phy. It is, for instance, almost completely absent from the Anglo-American literature. I take it to be coherent, viable, even preferable over alternative theories of history, well-nigh ineluctable under standard conditions still to be mentioned. I confess I am recommending it. (Bear in mind, however, that the force of the claim is hostage to the fate of the concept of history. I am not pretending that it needs no defense.)

I am also suggesting, of course, that Gadamer violates the notion (of "radical history") in his own way. Arendt thinks that massive technological intervention may deform the laws of nature, if there are laws; Gadamer thinks that, although human nature is a historical artifact, all its changing forms must belong to one legible humanity, since humans understand one another through history. Arendt fears a history that does not respect the invariances of nature (and of history); and Gadamer is confident that the variations of history confirm that human understanding is universally apt through all its changes. The true poet's voice (Paul Celan's for instance) is, for Gadamer, literally, a universal voice accessible to every historical age adjusted to its own horizon.

Let us move a little more slowly: a great deal is at stake. Arendt wishes to resist the idea that even history is merely made. The idea has its obvious Vichian source. Arendt believes that the thesis she opposes lies at the heart of totalitarianism: "the conviction that everything is possible," (as she says), the conviction that the laws of nature can (now) be overturned (as she fears).15 Perhaps it does lie there; or perhaps the idea that "history is made" is not the same as that "everything is possible"—in history or in physical nature. She appears to blame the conceptual slide on Marx. For, she claims, "Marx was only the first—and still the greatest, among historians—to mistake a pattern for a meaning"; "only patterns can be 'made'," she says, "whereas meanings cannot be, but, like truth, will only disclose or reveal themselves [in the larger scheme of things]."16 She means that the meaning of (true) history dawns on us unbidden and compellingly— hence, "unmade"—as Vico and Hegel thought, arising out of the historical "process as a whole," always beyond "the 'narrow aims' of acting men," consequently never (as in Marx, on her reading of Marx) explicitly pursued and realized, effectively, pragmatically, deliberately, in either a utilitarian or totalitarian way, in (merely in) the "intended aims of political action."17 (She blames Marx for Stalin's excesses.) Consequently, "acting into history" and "acting into nature" represent,

for Arendt, the substitution of a history of "patterns" (of manageable projects, of any projects at all) for a history of grander, holistic, "universal" meanings ("world history"). But that is to confuse the Intentional, artifactual, real nature of history with what exists in a brute way in nature. What Arendt finds vulgar in the Marxist is, metaphysically, no different from what she finds noble in the Vichian and Hegelian: historical meaning is a horizonal construction in each, but we are entitled to our different interpretive tastes.

The meaning of history, she fears, is now pluralized, distributed, rendered "particular" (placed within effective reach), "led to . . . complete meaninglessness" by thinking that historical meaning accrues as the result of achieving any intended objective at all. The deeds of history, like the events of nature, are deprived of meaning whenever they are disconnected from the "universal process" in which they are rightly whatever they are (the Hegelian or Vichian process: paralleling the work of the laws of nature); "they have ceased to make sense [she claims] without a universal process in which they are supposedly embedded."18 She means that technology and practical politics may achieve their intended goals, but they cannot similarly control what evolves as their palpable meanings historically. But that hardly signifies that history imperiously obliges us to acknowledge "its" meaning (its universal meaning) for what it is. We have no need for any Hegelian (or Vichian) Providence; we need only admit that we cannot fathom the future, that our own horizon makes it impossible to anticipate the historical imagination of others (even of ourselves at a later time), and that the horizonal nature of our thinking makes it impossible to grasp the complete meaning of even "local" historical events. Mere differences in scale are philosophically pointless here. Once you admit historicity, you cannot reclaim universal laws or universal meanings or universal norms modally: the option is there, but it is no longer conceptually necessary. (And that discovery is not itself a necessary truth.)

What I am urging is simply that there is no principled distinction between "universal" and "local" history: all histories acquire meanings through future interpretations; final closure is as impossible in the global as in the local; and, in a profound sense, pace Vico and Hegel, they are ultimately indistinguishable. What I am recommending is a "radical" reading of local history—which obviates the need for universal history (or at least the disjunction). Local history is universal history, since the total context of any and all events cannot be fathomed. That is just what Hegel meant (in his best moments) in speaking of the

closure of world history: man is incapable of fathoming the meaning of the whole of history. (If he meant more—I cannot deny that he may have meant more—then I cannot see how he could have legitimated the claim.) By contrast, the idea of an inclusive history of mankind entails neither Arendt's Kantian theme nor Gadamer's hermeneutic: two very doubtful forms of universality.

I take this, then, to fix the defensible meaning of Hegel's "philosophical" notion of world history: that is, "the idea that reason governs the world, and that world history is therefore a rational process."19 One can puzzle forever about the intricacies of Hegel's thought. But in any plausible sense in which passages like the following are interpreted,

what the spirit is now [the World-Spirit], it has always been [being always actualized]; the only difference is that it now possesses a richer consciousness and a more fully elaborate concept of its own nature [than at an earlier phase of its career]. The spirit has all the stages of the past still adhering to it, and the life of the spirit in history consists of a cycle of different stages, of which some belong to the present and others have appeared in forms of the past. Since we are concerned with the Idea of the spirit and look upon everything in world history as a manifestation of it, we are invariably occupied with the present whenever we review the past, no matter how considerable that past may be. For philosophy is concerned with what is present and real,20

it is impossible to deny that finite histories pertinent to Hegel's project acquire ampler meanings in the evolving present (that is, future-presents) than they possess in the past (past-presents). Thus, as Hegel explains:

World history . . . represents the development of the spirit's consciousness of its own freedom and of the consequent realization of this freedom. This development is by nature a gradual progression, a series of successive determinations of freedom which proceed from the concept of the material in question, i.e., the nature of freedom in its development towards self-consciousness. . .. All that need be noted here is that each step in the process, since it is different from all the others, has its own peculiar determinate principle. In history, such principles constitute the determinate characteristics of the spirit of a nation. Each historical principle, in its concrete form, expresses every aspect of the nation's consciousness and will, and indeed of its entire reality.21

What I suggest is that the "world history" of a people and the history of "local" events are equally restricted to the past ("no matter how considerable that past may be") and that the "present" meaning

of a history cannot be fathomed from its own "past," since every "present" has "its own peculiar determinate principle" (even if, on Hegel's view, Spirit is always "actual"). If we do not expressly invoke Hegel's "Idea" of the historical process, but cleave to his intuition, then: first, there is no principled cognitive difference between "local" history and "world" history; second, the meaning of history is incomplete in any of its past phases; third, the meaning of the past phases of any history is reinterpreted in its evolving present; and, fourth, human historians interpret their "material" without assuming that there must be a telos or absolute meaning or contextless space within which alone their re-interpretations of the historical process may legitimately claim objective validity.

Hegel was indeed tempted by his own scheme: he appears to presume at times that he alone may have grasped the progress of the World Spirit. But in his best moments he uses the vision essentially to confirm the relic structure and aspiration of finite human histories, under the condition that no one can fathom what is historically "actual" in the whole of history. I believe that Hegel anticipates here the theme of "radical history" but is badly misread by his enthusiasts. He develops his "radical" theory by way of an analogue of negative theology. If he intends his thesis more robustly—if, for instance, he means either (a) that there is a discernible "absolute" history or (b) that finite histories make no sense except in the context of correctly discerned "absolute history"—then, quite frankly, I cannot see the force of his argument. There is no reliable sense in which the unique meaning of the world-historical process is ever simply "disclosed." It is its scope, not its privilege, that counts. It must be parasitic on more fundamental models of history. It cannot be more reliable than they.

Give up the invariant laws of nature, concede the historical formation of human consciousness: you must abandon as well the universality or necessary invariance of human norms of reason, of validity, of truth, of coherence, of objectivity, of value—and of history's meaning. These options, as we shall see, are part of the radical possibilities of which nature and culture and cognizing selves are made. In short, "universal history" does not require genuinely universal (that is, strict, invariantly apt, normative) historical meanings and is not incompatible with the constructed nature of history, the human world, nature, human persons themselves. This is what makes Gadamer's claim about the "classical" so puzzling. It cannot be normatively universal, binding on all mankind, just like that. It is the will-o'-the-wisp of every strong

historicism. To have proclaimed its accessibility so firmly is to have promised a full disclosure of its rationale; but Gadamer never offers the necessary key. At best, "universal" is equivocal: as between "in-variantly binding" on all times and places and "intended inclusively" for an entire people or all humanity (whatever its content). The first is the hermeneutic analogue of the unity model of science: the discovery of the would-be essential norms of human nature; the second is the sense dramatically favored by Marx for instance, speaking of the Gattungswesen, and by Hegel, speaking of "world-historical" societies. In Marx, it embraces the whole of humankind because of the evolved power of technology, but it makes no realist claims about moral norms.22 In Plato, Aristotle, Kant, and Hegel, the two senses are mingled, though not always in the same way.23 This helps to fix our theme.

II

My claim is that, yielding in the direction of Arendt's reading of Vico and Hegel, we risk missing what is most radical and most profound in recent conceptions of history. The compelling clue rests with two utterly different notions of history (that implicate a third).

On one notion, the historical significance of temporal events is entirely due to the extrinsic interest we take in those events; the events themselves have histories only in the sense that their occurrence extends over a span of physical time; their historical meanings, however, are relationally attributed to them only as a result of our interest in them. On that view, the events and the persisting particulars they involve have no historical significance intrinsically; historical meanings lack a realist standing altogether. To say that things have histories is a façon de parler equivalent to saying that they occur in or persist through a span of real physical time. There is then no reason to deny that mere physical events have (real) histories as well as human events: nothing depends on the nature of what enters into a history. The attribution is nothing more than a compendious way of collecting the temporal spread of an event. The time of history and the time of physical nature are one and the same. Actual histories and imputed historical significance are two entirely different matters.

On the opposed alternative, some events (only) and some particulars, but not all that may be assigned a temporal spread, have histories: to have a history is, in addition to occupying a temporal spread, to have a "nature" intrinsically apt for possessing historical properties

or meanings; to possess such properties is to be open to interpretation and to "have" (for that reason) no "nature" (that is, no essential or "natural-kind" nature). In particular, cultural phenomena (events and "texts"—the writing of a novel and the novel written, the waging of a war and the war waged) have histories because, on the argument, they are suitably complex. They have Intentionally complex "natures" apt for interpretation, apt for having histories ascribed to them. On the second view, there is no difference between baying a history and possessing historical meaning. Hence, historical time is fundamentally different from physical time—to which it must conform and which it will most certainly embody—because historical time has an inherently Intentional structure that physical time lacks.

Both views are coherent. The question is: which is the more resourceful and more plausible? (I must add that I am frankly trading on the gains of the first two chapters. I take it that reference, predication, individuation, reidentification are viable under the condition of denying that reality must be invariant and particulars must have fixed natures.)

Now, whatever else is true, the two notions sketched could not be more opposed. The first treats historical significance as extrinsic, relational, rhetorical, not metaphysically (or objectively) grounded at all; the second refuses to disjoin history and historical significance: it treats history and historical meaning as real, as intrinsic to real phenomena. (They are inseparable.) The entire question of the rigor and objectivity of interpretation cannot fail to be affected by favoring one or the other conception.

One of the clearest examples of the first appears in Adolf Grün-baum's well-known diatribe against psychoanalysis. The same theme appears in his attack on the views of Hans-Georg Gadamer, Paul Ricoeur, and Jürgen Habermas. Grünbaum never addresses the conceptual possibility that the time of human history and the time of physical nature may be of logically different sorts: he simply takes it for granted that Intentionality does not affect historical time or cultural phenomena in a way that would support a principled difference between the physical and the human sciences. He affirms the following with characteristic vigor:

The physical theory of classical electrodynamics will now enable me to show that Habermas and Gadamer have drawn a pseudocontrast between the nomothetic and human sciences. For that major physical theory fea-

tures laws that embody a far more fundamental dependence on the history and/or context of the object of knowledge than was ever contemplated in even the most exhaustive of psychoanalytic explanatory narratives or in any recapitulation of human history.24

The question looms, of course, as to whether the first conception of history can accommodate the puzzles that the second is intended to address. The odd thing is, Grünbaum never examines the possibility. The obvious answer is that it cannot.

Call the first view the modernist conception of history. (Allow the term an inning. It has an affinity with Krauss's use of the same term in speaking of "modernist" theories of art.) Essentially, it opposes a strong role for a fully Intentional treatment of context: on its own reading, reference to context signifies no more than a lack of information about that part of the world in which certain events occur, not the impossibility of satisfactorily rendering reference or individuation in extensional terms (favorably, say, to the unity of science model).

The "modernist" camp includes such prominent views of history as Grünbaum's and C. G. Hempel's25 which embody different versions of the unity conception. It must be said that "modernists" regarding history need not specifically subscribe to the unity canon. Thus, Arthur Danto is a coherent modernist, but he does not subscribe to the unity thesis. It is true, however, that he inclines very far in its favor. He is, I should say, committed to "naturalizing" history—in that sense of "naturalizing" that, influenced by Quine's proposal in "Naturalizing Epistemology," has pretty well swept most of contemporary analytic philosophy: Danto, therefore, intuitively resists the second conception of history, because it is incompatible with the naturalizing strategy.26 In this sense, he need not actually subscribe to the unity of science program in order to be able to reconcile his account of history with the dominant currents of analytic philosophy. But he neglects all the interesting metaphysical, epistemological, and methodological issues in so doing. In particular, he has no basis for a realism of historical meaning. (I return to Danto in chapter 7.)

Modernists—Habermas and Ricoeur as well as Danto, but certainly not Gadamer—tend to identify physical and historical time as one and the same, and tend, as a result, to theorize about assigning meaning to history extrinsically. That seems to be the common denominator. Modernist views may vary considerably regarding the other distinctions mentioned (for instance, the vagaries of context). If so, then it

may be better to think of modernist theories as a range of very different views grounded in a common perception of time and a common perception of the conditions for ascribing historical meaning.

Let me call the second view the poststructuralist conception, though the term, like the first, is certainly tendentious and possibly not entirely helpful. Gadamer is not a poststructuralist, for instance. There are affinities between his view and that of the poststructuralists; but, ultimately, he is opposed to poststructuralism. Vico and Gadamer are more nearly dualists about nature and history (as, also, is Arendt, in her stronger moments), though none of these is a Cartesian dualist; whereas the poststructuralists treat physical time and physical nature as no more than convenient abstractions (fictions or heuristic instruments) drawn from within the all-inclusive context of historical time. They are hardly disposed to dualism, and they are not favorably drawn to metaphysics. Perhaps Foucault may serve to fix the sense of the latter distinction. (In a much less radical, but still quite robust, way, the same theme appears in Dewey's pragmatism.27 )

This helps to define the third view of history I hinted at a moment ago—which, again unorthodoxly, may now be called the hermeneutic conception (shared, however thinly, by Gadamer and Vico)—which emphasizes the disjunction of physical nature and human history and (therefore) the "made," the interpretable or narratizable, form of the latter. Heidegger, we may say, appears more "poststructuralist" than "hermeneutic" in Being and Time, insofar as he prioritizes Zeitlicbkeit (the inherent distinction of Dasein ) over all particular human categories; but he is also "hermeneutic" insofar as he segregates the "on-tic" and the "ontological" and distinguishes the existential complexities of the human condition from the Vorhandenheit of mere things. These remarks may indicate the sense in which our categories are terms of art; for Heidegger is obviously too early to be a poststructuralist in the French sense, and Vico, too early to be a hermeneut in the German.

The reason for preferring these plainly tendentious labels is simply that the general descriptions so far given (of the different sorts of history) do not quite capture the point of what I am calling "radical history." Also, they do not reflect any well-established catalogue of types of history. Let me say quite candidly that the distinctions offered are a little forced and are meant to be gradually replaced. I shall redeem them shortly, but it is not unhelpful to have a first rough partition

among the figures we must bring together. At the very least, it motivates the trimmer distinctions being promised.

I am not, as you may surmise, putting much store in mere classificatory schemes, and I apologize for tossing so many disparate thinkers into the same pot. The schema makes good sense, but its full lesson will need a little more time. I am bent on reclaiming a theory of history ("radical history") that could serve as the counterpart of the theory of "radical" interpretation I originally sketched from an examination of Barthes's account of "writerly reading." The trouble is, the distinctions marked—modernist, hermeneutic, poststructuralist histories—are not really standardized. Still, I think they are reasonably drawn. I shall take the liberty of defining their features several times with increasing precision. But with each improvement, I mean to define a conception of "radical history" that is bolder and more resilient than the others.

III

My original purpose, you remember, was to isolate the radical conception of history I found obscured in Arendt, which the hermeneuts have also missed. My argument was meant to show that the modernist conception is simply inadequate for human history, and that the hermeneutic conception fails to concede the resources of the poststructuralist option continuous with its own. Once this is made clear, one sees that "universal history" need not be deprived of a modernist or a poststructuralist reading: the conceptual eligibility of universal history has nothing to do with the different conceptions of history I am gradually defining; it remains intact for each as long as each remains coherent, but it cannot fail to be infected by their particular differences.

Let me then, for the sake of a certain tidiness, collect the principal alternatives again. The modernist conception distinguishes sharply between actual histories and historical meanings (the meanings of antecedently identified histories). Anything that persists through physical time ("real time"), anything that has a "temporal spread" (events), has a history merely as such; the distinctive natures of any and all such referents do not bear on their merely having (or being able to have) histories, though the natures they have will of course affect the particular meanings their histories are ascribed.

Presumably, on Grünbaum's view, electrodynamic events have

histories. But those histories do not have historical meanings intrinsically; or if they have "meanings," their meanings are whatever we extrinsically assign their physical histories. (Their natures are not Intentionally qualified, you see.) By contrast, the poststructuralists and hermeneuts oppose the disjunction between an event's or a particular's having a history and its having historical meaning. (They agree with Hegel in this.) The hermeneuts hold that only certain "kinds" of things (persons, artworks, actions, institutions, language, texts, narratives, cultural phenomena in general) are inherently "historied," intrinsically open to interpretation, Intentionally qualified. The poststructuralists are bolder, because, though they agree about the point just made, they go on to treat the very distinction between the physical and the cultural as no more than a heuristic abstraction within a common, otherwise undifferentiated discourse. The poststructuralists, therefore, would never endorse Arendt's or Gadamer's or Vico's confidence in universal meanings.

On the poststructuralist account, cultural phenomena possess historical properties "natively." As a trivial consequence of their changing history, they may be ascribed historically changing meanings— afortiori, historically changing "natures." Some hermeneuts and most poststructuralists accept this much—by way of post-Heideggeran hermeneutics for instance and Foucauldian genealogies. But the poststructuralists also oppose the dualism of history and nature and all essentialist thinking: they collect the "narratives" of science for instance, within which, so they believe, the "physical sciences" themselves are no more than provisionally abstracted.

The usual hermeneuts never go that far. Furthermore, poststructuralists deny that anything has a fixed nature: discourse is itself hostage to history. Hermeneuts remain dualists regarding history and nature—not, however, regarding mind and matter; whereas poststructuralists outflank all dualisms by treating conceptual schemes as radically provisional. (In this, they resemble the pragmatists.)

What is radical about the poststructuralist conception (Foucault's, for instance) is that, in addition to sharing with late-hermeneutic views (Gadamer's, in particular) the thesis (1) that whatever is intrinsically historical may be directly altered by the historical process, by interpretation and by intentional action, it is also committed to the thesis (2) that what is thus affected has no fixed nature at all, has only a "historied nature," a "nature" that can impose no invariant or universal or

nomic constraints of any kind on the historical meanings such things may acquire.

The hermeneuts and their allies ("modernists") dispute this in different ways (Gadamer and Vico and Hegel and Habermas—and Arendt); and of course the question arises as to whether the poststructuralist conception is genuinely coherent. But, at the very least, the modernist and poststructuralist conceptions could not be more opposed; nor, frankly, could the hermeneutic and the poststructuralist. Gadamer (as we have already seen) arrests the possibilities of historical novelty by imposing on all histories the mysterious normative constraint of the "classical." There's the point of honor between the hermeneut and the poststructuralist. It's also the point of honor between Gadamer and Habermas, though for quite different (local) reasons.28

These distinctions, you must remember, are still provisional and approximate. There are fairly explicit theories of the "modernist" sort (which are often not thus identified), but there are no comparably explicit poststructuralist theories. Foucault shows us a bit of what to expect,29 but the poststructuralist is impatient with systematic theories.

Nevertheless, what is most distinctive about these notions is this. The modernist (i) disjoins having a history and possessing historical meaning; (ii) reduces historical time to physical time; and (iii) treats the conditions for ascribing meanings to histories relationally, primarily or only in a way that depends on the extrinsic interests of human historians. Histories, then, have realist standing, but historical meanings do not. By contrast, the poststructuralist (i') opposes the modernist's disjunction; (ii') identifies a range of phenomena (cultural phenomena) that intrinsically possess histories; (iii') distinguishes (therefore) between physical and historical time (the principal hermeneutic issue); and (iv') acknowledges that whatever possesses a history changes "nature" as a result of its ongoing history (a thesis shared with late-hermeneutic theorists: uneasily, as it turns out, with Gadamer).

Furthermore, the modernist regards his distinctions as straightforwardly true about the world—as essentially objective (even "objectivist"), certainly not the work of shifting history; whereas the poststructuralist treats his distinctions as no more than heuristically useful, even if provisionally compelling, within this or that episteme—cer tainly not fixed for all times and places. Any "realist" concession on the part of the poststructuralist is, therefore, distinctly relativistic. The difference centers on the role and fixity of "context." The extraordi-

nary thing is that these distinctions are nowhere adequately sorted. The analytic modernists (Hempel, Grünbaum, Danto) explicitly deny the symbiosis of world and word; or, if they admit it, then, doing so or-thodoxly (as full-fledged Kantians: Apel and Habermas for instance), they oppose the indissoluble union of the two, symbiosis and intrans-parency. (This, of course, is Kant's maneuver.) The poststructuralists contextualize or historicize the union of both notions (Barthes, Foucault). You can see, therefore, the difficulty of classifying Gadamer.

IV

Try another tally. (We are making good progress.) On the modernist view (1) the natures of things have no histories; natures do not change, though particulars of this or that nature do; (2) all particulars that persist through time, all events that occupy a temporal spread, have histories (in an objective sense) merely as such; (3) historical narratives or historical meanings are ascribed—only or primarily relation-ally—to things that have natures and histories, in virtue of a certain extrinsic interest we take in them; (4) historical time is identical with physical time; (5) if real, Intentionally complex (or interpretable) properties enter into objective histories in the same way non-Intentional properties do—which raises, of course, questions of consistency with (3); (6) historical narratives are Intentionally ordered representations (or interpretations) of actual histories, by which a unified meaning is assigned such histories congruently with their real temporal spread and the extrinsic interests of a human narrator; (7) the objectivity of historical narratives and assigned significance depends on the use of selected criteria in virtue of which pertinent interpretations correspond to certain elements of actual histories (agents' intentions, typically, or the Geist of this age or that)—but this again raises questions of consistency with (3). The unconditional admission of (5), the admission of Intentional properties as fully real, is plainly the decisive card that must be played or withheld.

To forestall misunderstanding about the possible conflict between (3) and (5), theorists like Hempel regard inquiries into historical meanings as stop-gap measures; theorists like Grünbaum treat intentional complexities as signifying no difference that affects the methodology of science; theorists like Danto treat meanings as entirely extrinsic to the histories of things; theorists like Ricoeur and Hayden White think of the meanings of historical events equivocally, as at once both

"real" and metaphorically ascribed.30 (I shall return, much later, to Ricoeur and White.) Yet, all these thinkers are modernists. So we have unearthed an essential aporia for modernism: a conflict between (3) and (5).

It would be fair to say that the most ramified recent account of history in accord with the modernist model appears in Arthur Danto's Narration and Knowledge. The model itself is coherent and spare.31 Its principal strength rests with the fact that a history (the narrative) is taken to be objective if it accurately represents the "history" (the persistence) of (and change in) anything at all through physical time, when it is also construed as being in accord with the (external) interest a historian actually takes in it. Anything that merely persists has a history; historical narratives are interpretive stories about the bare continuum of a thing's persistence; and in every other respect no difference arises between the general concerns of science and history. Danto and Grünbaum agree about this. Grünbaum, however, treats psychoanalytic (Intentional) phenomena as if they behaved causally (and nomologically) in the same way physical events do (because the causal is nomological ); Danto treats the historical interpretation of the arts and other cultural phenomena as extrinsic to objective histories (hence, as no more than rhetorical). A "small" question remains unanswered regarding the objectivity of historical meanings. Danto's remarks about the meaning of this or that history suggest that he regards such meanings as objective, but his theory ultimately disallows the possibility.

Two difficulties confront the modernist model. For one thing, there is no compelling argument to show that the principal referents of history (persons, texts, artworks, actions, cultural phenomena in general) have natures in the strong sense items (1) and (2) of the last tally affirm; on the contrary, they cannot be easily thus construed, unless some form of physicalism obtains—which seems doubtful on independent grounds. They belong to "cross-category" kinds, which tend not to support direct nomic regularities;32 also, they possess Intentionally complex properties, which tend to thwart extensional ("naturalizing") treatment. According to the other difficulty, the referents of history risk violating (7), in that their "natures" change, contra (1), as a result of intervening action and interpretation. If these difficulties cannot be overcome, the modernist conception must utterly fail.

Recall, now, the logical features of the "Intentional." It signifies a complex attribute: (a) real insofar as it is incarnate in physical or biological attributes; (b) culturally emergent with respect to such attributes

and ascribable (intrinsically) only to what thus emerges; (c) indissolubly one with respect to ascriptions of "aboutness" (the "intentional") and the semantic or semiotic or language-like "content" of such "about-ness" (the "intensional"); (d) not, however, restricted to the mental or psychological but ascribable as well to the nonsentient public artifacts, texts, artworks, products, institutions of culturally significant or social labor; (e) primarily descriptive of the collective features of encultured life, instantiated in acts and artifacts; (f) designated by monadic or one-place predicates; and (g) intrinsically interpretable. Its admission proves an insuperable barrier to physicalism, extentionalism, supervenience, methodological solipsism, and related doctrines; and it would make no sense to admit poststructuralist or hermeneutic histories without admitting the Intentional as fully real. So the contest is joined.33 Thus: if human persons are real, then Intentionality is real; and if Intentionality is real, then the cultural world is real; and if the cultural world is real, then a realism of history must extend to a realism of historical meanings. And, if that is true, then the prospects of a radical theory of interpretation are essentially the same as those of a radical theory of history. The modernists, I say, fail to admit at least (a) and (g) of the tally just given; they therefore lack a realist account of historical meaning.

Counterarguments against modernism are of two sorts. For one thing, no sanguine view of physical nature ensures the modernist's reductive intent about historical time and objective history; on the contrary, doubts about the realism of invariant laws, difficulties with reductionism, the import of historicizing science threaten to subordinate the physical sciences to the human sciences—or at least to deny their independence or methodological priority. For a second, the Intentional phenomena of history resist the forms of rigor drawn from the description and explanation of mere physical nature. (I should be willing to assign these lessons to Kuhn's account of science, but Kuhn himself is uneasy about acknowledging them.)

By contrast, the poststructuralist model precludes (1)-(7). On that view: (1') particulars and events that genuinely have histories lack fixed natures, have only histories or "historied natures," or are themselves only individuated histories; (2') such particulars have histories (or careers) in the sense that their numerical identity is manifested as a narratized persistence through physical time; they are culturally emergent and continually reconstituted by interpretive consensus; they have no independent existence, but are not constituted merely by our inter-

est in antecedent physical particulars; (3') historical meanings are not merely relationally ascribed to physical things in virtue of extrinsic interests; (4') historical time is fundamentally different from physical time, with which it must accord and in which it must be incarnate; (5') things that have histories inherently possess Intentional properties, in virtue of which they are constituted as histories and incorporate particular parts of the physical world; (6') historical narratives are interpretations of things that possess "historied natures" effectively altered by interpretation; (7') there is no principled distinction between historical narratives or the interpretations of things that have histories (or historied natures) and those histories or "natures"; such distinctions are grammatically drawn within a continuous history. The last item signifies what is so problematic about the realism of the world of human culture.

The second model is consistent, though hardly spare. It is distinctly heterodox in fact and requires considerable explanation. There is no tidy specimen of its kind. The tally just given is as brief a statement of the model as one can find. It is also more or less in accord with the interpretive practice of Roland Barthes, Michel Foucault, and Harold Bloom, without being bound to their particular extravagances.

V

One strategic distinction remains to be made clear. It concerns the notion of a thing's "nature" and the possibility that some things have no natures but only histories or "historied natures." The idea is a complicated one, but it may be put in a simple way. First of all, all discourse concerned with truth or the assignment of truth-like values must provide linguistic facilities for effective reference and predication: we must be able to identify what we are speaking about and what, regarding that, we mean to affirm or deny. There is no plausible reason to suppose that that is not as true of history and art criticism as of physics and morality. (This, you remember, was the point of the complaint against Rosalind Krauss's "postmodernism.") Second, it is a staple of the twenty-five hundred years of Western philosophy that referents must be logically fixed (not necessarily in essentialist terms) if predication is to function successfully. Even Quine's clever (but epistemically futile) account of displacing referring expressions with uniquely instantiated predicates clearly invokes the notion of logically constant designations.34 (I shall come to Quine's account in a moment.) Third,

there is no theory of reference confined to the formal features of language that accounts for the operative success of referential discourse. The fixity of reference appears to elude all principled efforts to specify its operative procedure: which is not to say that reference fails, only that it is not theoretically tidy. There is no algorithm for ensuring or testing successful reference: there cannot be. Context is simply not manageable algorithmically, and context is ineliminable in successful reference.

On Aristotle's account, as has already been shown, real particulars possess invariant natures, in virtue of which they are real. Aristotle believes the denial of this thesis to be incoherent; but, although he attacks Protagoras for having denied it, he nowhere shows that his charge is true.35 In any case, the individuation and reidentifiability of particulars (said to possess a common nature) cannot be settled by attending merely to the nature (quiddity) they share; and, of course, Aristotle has no companion principle to offer for securing, in cognitive terms, the individuation of different particulars said to share this or that common nature. There is no account, in Aristotle, of how hyle functions criterially for the purpose of distributed reference; there is only the claim that individuation depends on the distribution of hyle .36 Aristotle has no way of fixing reference.

A cognate problem arises for W. V. Quine's Leibnizian (really: more than Leibnizian) solution regarding reference. Quine believes that we may dismiss reference in principle, since whatever can be referred to can be denoted as instantiating a unique (general) property compounded from however many such properties. (Quine never explains why this is so.) He simply recommends "reparsing . . . names [and singular terms] as general terms"—for instance, by treating "'Socrates' as [a] general term [that] would be true of one and only one thing" or by similar means. The name "Pegasus," as in "Pegasus flies," may then be reparsed as a general term that, if true of anything, is true of one and only one thing.37 But Leibniz had already seen that no mere human could ever apply the notion (in effect, Quine's) in a cognitively satisfactory way: no human could possibly know the unique property that every thing instantiates (possibly, the property answering to whatever is true of each particular thing), and so no human could know the unique (general) property that any thing instantiates. (Quine's program promises principled grounds for precluding the ontic importance of context.)

Reference, therefore, proves to be ineliminable under real-world cir-

cumstances; it cannot but be hostage to history (and, redundantly, to variable context or horizon).38 It cannot be supplanted by predication and the use of quantifiers alone. Only God could possibly grasp the nature of the universe in which all particulars are found and may be explicitly differentiated; and, even with regard to God, Leibniz will not risk a principled solution in terms of general attributes (that is, one restricted by considerations of formal consistency alone). The ineliminability of context is simply the counterpart of the failure (in cognitive terms) of Quine's and similar programs for retiring reference by predicative means. That failure also signifies that we must treat reference as an Intentionally complex practice, which is precisely why the eliminability of reference is so much prized.

Now, it may also be reasonably argued, though the effort would be a busy one that would take us too far afield, that the failure of reference to yield in the manner indicated (similarly, the insolubility of the problem of "universals" along formal lines) is already tantamount to the need to admit historical practices, conceptual schemes, and the like, as Intentional tertia —reconciled with what I have characterized as in-transparency and symbiosis.39

Causal (or causal/christening) theories of reference are clearly subject to the same sort of laxity that confronts Aristotle's and Quine's options; and, in fact, Saul Kripke, one of its principal champions, is entirely explicit about that weakness.40 Nor can contextual theories of reference, Wittgensteinian or Austinian in spirit (for instance Gareth Evans's), overcome the limitation indicated; such theories are already content to accept the inherent informality (contextedness) of referential practice.41

The reason for pursuing the matter is entirely straightforward: (a) there is no formal procedure for fixing reference uniquely, regardless of one's theory of real particulars; (b) reference succeeds in a practical way without requiring any such rigor; (c) there is no antecedent reason why reference should fail, if our inability to specify in general terms the haecceity of every particular does not entail the failure of reference; and (d) all truth-seeking inquiry depends on reliable referential practices.

Doctrines (a)-(d) form a conceptually stunning challenge to the entire philosophical tradition running from Aristotle to the present; for it has been standardly maintained that individuation and reference presuppose a certain suitable fixity of nature (metaphysically essentialized: as by Aristotle; logically conventionalized: as by Quine) relative

to which those quasi-logical processes succeed—notably, without invoking contextedness and Intentionality. In short, if reference could be retired in favor of predication, or if reference could be syntactically fixed for any space of application by means of a formal algorithm, then questions of context would not substantively affect the meaning and truth of whatever we utter. But if such programs fail, then, by a reasonable inference, the analysis of language—in all its uses— is , ineluctably, historicized. What I am claiming, then, is that since such programs do fail, the plausibility of the radical notion of interpretation I am urging cannot but be enhanced.

Denotation and reference, the conservative argument goes, presuppose individuation; individuation presupposes natures; natures are natures, for the purpose in question, only if, logically (if not also onto-logically), they remain fixed and predicatively constant for all successive episodes of repeated denotation, reference, and identification. That multiple claim now turns out to be false or indemonstrable; but, in any case, it is not necessarily or demonstrably true (de re or de dicto). Also and independently, the haecceity of any and every particular cannot, as I have argued, be reliably specified by the sole use of general predicates. Furthermore, the viability of radical history depends on the coherence and persuasiveness of (a)-(d). There you have the essential conceptual connection, the one so easily and so regularly ignored. For we must surely ask ourselves: What are the conditions of successful reference? What could replace Quine's epistemically inadequate program?

If we agree that history is inherently an interpretive discipline (concerned to assign meaning or significance to things that have histories), then the admission of (a)-(d) has the force of disallowing any logical advantage favoring modernist histories over hermeneutic, poststructuralist, or other more radical accounts. What holds here holds for interpretive practice in general.

There could hardly be a more decisive gain for the revolutionary possibilities I have been collecting in the name of the end of our own century and the beginning of the next. The radical notion, the master notion at stake, is simply this: all the conceptual puzzles of science, morality, history, art, language, technology, interpretation are now to be recovered under the banner of radical history. By radical history, I mean at least: first, that we have no reason to think the real world is invariantly structured in any fixedly essential or necessary way; second, that, if that is so, knowledge and inquiry never discern an in-

variant order or an order not already dependent on the contingent symbiosis of knower and known; third, that knower and known are themselves historically preformed in a tacit and blindly perspectived way; fourth, that reference and predication are effective in practice despite our lacking any principled criteria for fixing the unique identity of this and that and despite our lacking changeless universals in virtue of which (alone) particular ascriptions and interpretations are rendered intelligible and true; fifth, that the referents of discourse need not possess, or have assigned to them, fixed natures, but may change (in) the "natures" or histories or "historied natures" they are assigned, without disabling reference or reidentification; and sixth, that, as a consequence, even the physical sciences, even logic and mathematics, are, as far as their lawlike and formal invariances are concerned, similarly bound by the conditions mentioned.

No discernible or recommended uniformities—formal, substantive, methodological, cognitive—need be disturbed by this way of locating invariance. It may be a little brusque to say, but these six themes, clearly favored at the end of our century, could well produce a historical turn as fundamental as that ushered in with the turn of the Renaissance and the turn of the French Revolution. The potential is certainly there.

VI

The line of thinking being favored is close to (but also a move distinctly beyond) the intent of one of Vico's master themes—the principle of "verum = factum "—which Vico narrowly applies to human history but which I now wish to apply (contrary to Vico's thesis), beyond the production of the cultural world, to include the "production" of knowledge and the knowable world itself:

[T]he first indubitable principle [says Vico] above posited [§331] is that this world of nations has certainly been made by men, and its guise must therefore be found within the modification of our own human mind. And history cannot be more certain that when he who creates the things also describes them. Thus our Science proceeds exactly as does geometry, which, while it constructs out of its elements or contemplates the world of quantity, itself creates it [as fictions]; but with a reality in proportion to that of the orders having to do with human affairs, in which there are neither points, lines, surfaces, nor figures. And this very fact is an argument, O reader, that these proofs are of a kind divine, and should give thee a divine

pleasure; since in God knowledge and creation are one and the same thing.42

This is certainly the clue to Vico's anti-Cartesian stance. For, as Vico says a bit earlier,

the world of civil society has certainly been made by men, and it is beyond all question that its principles are themselves to be found within the modifications of our own human mind. Whoever reflects on this cannot but marvel that the philosophers should have bent all their energies to the study of the world of nature, which, since God made it, He alone knows; and that they should have neglected the study of the world of nations or civil world, which, since men had made it, men could hope to know.43

Vico's thesis is being deliberately deformed here, to bring it into strong accord with contemporary notions of radical history, which can only partially accommodate Vico's correct (but also inadequate) insistence that the natural world cannot possibly be a (mere) human construction. Of course, this caveat is not at all the pivot of Vico's intention. But Vico is a pre-Kantian, above all; and he is also implicitly opposed to radical history. In holding that mankind is the creator of history, Vico intended to emphasize primarily that the intelligibility of history belongs to its mortal creator: Vico did not thereby mean to deny the invariances of history any more than the invariances of nature; consequently, he did not mean to deny that human history has a significance beyond the ordinary ken of human projects. The conviction is of course shared by Vico and Hannah Arendt (and even Gadamer)—on quite different grounds.

If we allow the point, then, curiously, it becomes quite easy to make room for a "tempered" version of what Vico calls the "providential" history of the world—which, in the relevant respect, is analogous to what may be saved of Hegel's and Marx's and similarly historicized teleologies (though Hegel and Marx are not dualists in the way Vico and Gadamer are—that is, they are not disposed to disjoin nature and culture). In fact, by his own example, Vico offers a formula (the matter may be contested) by which to capture the essential point of Hannah Arendt's complaint against Marx and in favor of Hegel:

It is true [says Vico] that men have themselves made this world of nations . . ., but this world without doubt has issued from a mind often diverse, at times quite contrary, and always superior to the particular ends that men had proposed to themselves; which narrow ends, made means to serve wider ends, it has always employed to preserve the human race upon this earth.44

Providence or Reason, or the putative telos of human existence, is now, by way of radical history, simply a conjecture about the narrative direction of man's symbiotized world, viewed on a scale (whether by Arendt or Vico or Hegel or Marx—or St. Augustine or Foucault or Habermas, for that matter) as "superior to the particular ends that men had proposed to themselves." That is, universal history raises questions of manageable scope, not of new sources of knowledge. I suggest, in fact, that, by "universal history," we mean no more than the narrative unity of the history of an entire society, or its interpreted meaning cast in collective terms, or both, discerned through the local deeds of single and aggregated individual agents. In the hands of Vico, Kant, Hegel (on Arendt's reading), Heidegger, and Arendt herself—but not Marx, not Gadamer, and not Foucault—universal history is privileged in some ulterior "contemplative" sense.

The point is this: "universal history" can be reconciled at a price with each of our distinctive types of history—modernist, hermeneutic, poststructuralist, and radical. To account for the formal features of universal history is not (as Vico and Hegel had supposed) to make substantive claims about truth and knowledge, or about certainty and necessity. I have, therefore, used Vico unceremoniously to separate questions regarding the narrative and interpretive structure of history and questions regarding their cognitive prospects. "Universal history" proves viable but relatively unimportant—conceptually. The argument does not adversely affect the work of any such history.

On the contrary, the feature Arendt had singled out (against Marx) as the mark of "universal history"—the dawning of the meaning of any human project beyond its immediate completion—is now seen to be the mark of every local history (except on the modernist's externalist theory); it has nothing to do with the scope of local and universal history. Or, if it does, it signifies only the constructed nature of history and the inherent indeterminacy of any "absolute" (universally inclusive) context of all contexts. So seen, Hegel's theory of history confirms that every local history has "universal" pretensions and that those pretensions cannot be humanly satisfied.

The decisive issue lies elsewhere. The issue is this: the condition of successful reference and predication cannot be made to favor modernist conceptions of history over other theories; and the advocacy of hermeneutic and poststructuralist conceptions are viable (as far as reference and predication are concerned) on precisely the same grounds that modernist theories can be shown to be obliged to admit. Hence,

these latter accounts need not worry about any threatened incoherence in their practice of individuation and reidentification. Those practices ultimately depend on a society's consensually remembered history of reference and the like; their adequacy accommodates the denial of the fixity, the invariant structure, the determinate boundaries that "modernist" theories have regularly insisted on.

Beyond that, the resolution of the puzzle of reference and predication draws on precisely the same resources that ensure the viability of hermeneutic and poststructualist theories of history—which the modernist opposes. The modernist fails, therefore, to grasp that the reality of history and the objectivity of historical meaning cannot be severed (as he claims): their union is already implicated in the discursive resources he shares with his opponents. Since those resources cannot in principle ensure an algorithmic solution to the problem of reference, it follows that there is no principled cognitive distinction between local and universal history.

Vico was more sanguine about the disjunction between nature and history than we need be. In any case, I have demonstrated the formal viability of a "poststructuralist" history and a "poststructuralist" mode of interpretation. The arguments that favor radical histories over "modernist" and "hermeneutic" histories depend on their dialectical advantage in answering questions about the nature of knowledge and reality. I shall not pursue these directly. They concern the salient philosophical themes of the end of our century: the cognitive intransparency of the world; the inseparability of science and its history; the social construction of culture and nature and human selves; the collective, historically contingent preformation of human consciousness and understanding; the indemonstrability of necessities de re and de dicto; the horizonal nature of legitimation; the impossibility of totalizing our conceptual and legitimative resources; and the artifactual contingency of all relevantly universalized claims. Given mounting concessions along these lines, there may be no way of resisting the poststructuralist alternative. It favors, I suggest, the conceptual projects of the next century.

VII

Nevertheless, the foregoing argument is a little too glib. For one thing, the poststructuralists—Foucault, Derrida, Lyotard, Deleuze, Irigaray—are largely preoccupied with versions of a single primitive theme: the suppression or repression of the "other" (l'autre ). By this, I mean:

philosophical poststructuralist is very nearly spent in a kind of critique (call it deconstructive or genealogical, possibly even phenomenological or pragmatist) that is entirely of a negative or disruptive or subversive sort, one hardly ever concerned to recover the disciplined rigor of a science or history or interpretive practice, or a philosophy that has suitably adjusted its sense of discipline to fit poststructuralism's principal question—that is, the question, whether, or how, its cognitional practices would need to change under the condition of radical history. I must still supply that sense, to bring the argument to a proper close.

The privative theme is simply this: no ordered scheme of conceptual categories can accommodate every viable set of distinctions answering to the salient interests of all disadvantaged populations (workers, blacks, women) already adversely affected by socially entrenched practices, or answering to "all possible" conceptual schemes that might otherwise evolve if those in place were not in place.45

There is reason to think poststructuralism exhausts itself in endlessly exposing the "terrorism" and "hegemony" of "totalized" such schemes. (These expressions, "terrorism" and the like, are urged by the poststructruralists themselves.) There is, therefore, a sense in which the poststructuralist obliges us to concede that, ultimately, there is no principled disjunction between philosophy and politics, between savoir and pouvoir (in Foucault's sense). All conceptual schemes are interpretively (constructively) restricted. I freely acknowledge the poin—it is regularly ignored or denied in analytic and modernist philosophies. Many will see it as an unsatisfactory finding. Still, although it signifies the permanent contribution of poststructuralism, it does not actually touch on the coherence of cognitive undertakings under its own auspices. The question continues to nag.

Apart from current disputes about the nature of knowledge and reality and apart from the poststructuralist puzzle itself, the conventional (analytic) account of reference and predication has proved epistemically hopeless in its own terms. It hardly requires the challenge of radical history. But the radical theory of history I am gradually shaping would itself benefit from recovering whatever is needed for a genuine theory of reference, that is, one that addresses the operative concerns of actual speech. Also, successful reference now begins to appear to be a thoroughly historicized achievement. There's the radical theme in a nutshell. Reference makes no sense except in terms of the mastery by apt speakers of the actual historied practice (of reference) that belongs to a particular society: the absence of any adequate algorithmic

strategy of reference, the ineliminability of context, and the thoroughly historicized sense of referential allusion converge on the same lesson. Furthermore, what holds, paradimatically, for reference holds, metonymically, for the whole of discourse and the lingual skills that depend on discourse.

The problem, you remember, is this: reference to reidentifiable particulars, for predicative purposes, requires that those particulars be individuatable. This, it has been conventionally claimed, entails the fixity of their natures: ontologically, according to Aristotle; merely logically, according to Quine.46 The notion of radical history upsets such fixities; or better: it fits very well with their demonstrated indefensibility. In fact, it does so in a way that, on the radical argument, we can find no way of bettering. I have already demonstrated that successful reference cannot be shown to entail the fixities claimed. At best, we must have been pretending we were relying on what was never accessible in the first place. I have been led to wonder, therefore, whether there is any other viable way to confirm that "natures" must be changeless de re or de dicto. I find that there are none—or that none are known.

Remarkably, the rest of the story proves to be quite straightforward: successful reference is achieved in all instances by narrative devices, by socially shared histories of interpretation, by internalizing the consensual practices of our Lebensformen. The decisive point is this: the solution to the problem of reference (and predication, individuation, reidentification, and the like) is not syntactic or linguistic in the narrow sense; it is lebensformlich, it is grounded in our habits of life, it is informal but effective, it is also tacit and inexplicit.

Imagine that some particular text—Shakespeare's Hamlet, say47 — identified (let it be noted) not, first, physically but only by a story-relative48 array of words possessing a historically assignable sense, is then continually reinterpreted: Hamlet's meaning, say, is found to change and evolve in accord with, and under the influence of, its ongoing (consensual history of) interpretation. But these are the terms of "radical history." I have already argued that some sort of narratized (or contexted) numerical identity is presupposed in every successful referential act. Success here is entirely compatible, therefore, with our lacking denotatively unique predicates (equivalent to fixed natures or haecceities: socratizing, say). I am now claiming that if narratized reference is successful in general, then it must also be possible (by similar means) to reidentify texts that undergo substantive changes in the

assignment of their histories and interpreted meanings. We hardly jeopardize such success by merely conceding the changed "nature" of the "historied" things we thus interpret (Hamlet, say, or the French Revolution).

The argument may be put in a single line: only a referential practice in accord with radical history could make sense of such interpretation; but it is entirely up to the task. Only devices of narratized history could sustain the referential requirement. Nothing else will do. The "nature" of interpretable (Intentional) things is multiply altered by their being interpreted. Only the historied memory of the course of such changes could possibly ensure the success of reidentifying particulars thus interpreted. But then, only a society's narrative memory of its Lebensformen, its referential and predicative practices, could possibly secure the reliable reidentification of (even) the particulars favored in the physical sciences. Hence, a narrative memory is required everywhere. Nothing is risked by the admission, except the pretense of an all-inclusive extensionalized system of the world. I have pressed, therefore, beyond poststructuralist histories to begin to sketch a more affirmative form of radical history. (Let me now officially dub this positive view the radical conception. It is hardly that radical, conceptually, if, in effect, we have always used it or something like it. But it will seem radical to those who object to the philosophical theory that supports it. Let that be.)

Two distinctions (beyond poststructuralism) mark its advantage: one I have just developed; the other remains to be mentioned. The first, (v')—to catch up the earlier (poststructuralist) tally of the present chapter—maintains that serial interpretations proceed by the devices of narrative history, in order to secure, and to be able to secure, reference under condition (iv'). The second, (vi'), adds that, if (v') holds, then histories may support indefinitely many interpretations, and historied particulars may possess indefinitely many histories.49 They may all be sorted (as many as we can pursue) by individuating the narratives by which we fix the (interpreted) "natures" of the referents whose numerical identity we fix conformably. No doubt there is conceptual inconvenience here; but there is no incoherence.

The point is that (vi') is by no means an idle or exotic recommendation: interpretation as actually practiced is entirely congruent with the radical conception. We hare been making reference this way. We hare been abandoning the fixities of modernist histories. The prospect of indefinitely many histories and indefinitely many interpretations is

already upon us. For example, a fairminded theorist like Morris Weitz, having scrupulously canvassed the large literature of Hamlet interpretation—partly in accord with his Wittgensteinian bent, but without venturing an opinion about its bearing on the different theories of history—draws the following conclusion: "A philosophical examination of our survey of Hamlet criticism reveals the undeniable multiple linguistic character of the criticism as well as its pervasive assumption that this linguistic multiplicity is reducible to a logical univocity. This assumption, I believe, is false. Its falsity can already be detected in the ostensible multiple character of the language of Hamlet criticism, which itself reflects the logically multiple character of its discourse."50

What Weitz means is that, taking Hamlet criticism as a paradigm of literary criticism, possibly a paradigm of criticism in all the arts, we cannot deny an ineliminable "multiplicity—of procedure, doctrine, and disagreement" in all criticism.51 Weitz offers at least twenty-four sets of "main issues," regarding which his own study confirms the fact that "criticism is many things, not just one"—it is never "logically univocal."52 In fact, Weitz continues,

interpretation is not the autonomous procedure that some critics claim it is but is the same as explanation; . . . the meaning of a work of art or of what is central in a work of art are not descriptive but explanatory concepts; . . . evaluation is often argument, and . . . good evaluative argument is neither deductive nor inductive argument but simply the employment of unchallengeable criteria in support of praise or condemnation; . . . dramatic (or artistic) greatness is not a property; praising and condemning are not describing; and, finally, . . . criticism need not state, imply, or presuppose a true poetics of drama or tragedy or an aesthetics of art in order to render intelligible or to justify its utterances about Hamlet.53

Weitz arrives at his finding by the most minimal of tests, by adhering as conservatively as possible to the constraints of the "modernist" orientation. (For example, by favoring an "emotivist" view of values, he finds that he need not give up excluded middle.) Hence, any radicalization of his finding would encourage a greater tolerance for radical history and radical interpretation. (Weitz does not raise any of the referential and predicative complications. Emotivism is not obliged to raise such issues. But that is an abdication of sorts.)

By contrast, Jean Baudrillard, who is already committed to conceptions of reference and interpretation of the poststructuralist bent (doubtless, somewhat incoherently, since there are no stabilities of reference and predication in his theory of the simulacrum) claims:

It is by putting an arbitrary stop to this revolving [manipulative] causality [as in Watergate] that a principle of political reality can be saved. It is by the simulation of a conventional, restricted perspective field, where the premises and consequences of any act or event are calculable, that a political credibility can be maintained (including, of course, "objective" analysis, struggle, etc.). But if the entire cycle of any act or event is envisaged in a system where linear continuity and dialectical polarity no longer exist, in a field unhinged by simulation, then all determination evaporates, every act terminates at the end of the cycle having benefited everyone and been scattered in all directions.54

Baudrillard's point, which deliberately oversteps the usual constraints of coherence in order to make itself heard, is that "simulation threatens the difference between 'true' and 'false', between 'real' and 'imaginary'." It marks "the death sentence of every reference." It signifies the "generation by models of a real without origin or reality: a hyperreal [or simulacrum ]." It introduces, arbitrarily, "the fiction of the real."55 (Notice, please, the extraordinary glibness of this last pronouncement.)

Baudrillard brings all this to bear on the play of interpretations (afortiori, on the play of histories):

Is any given bombing in Italy the work of leftist extremists, or of extreme right-wing provocation, or staged by centrists to bring every terrorist extreme into disrepute and to shore up its own failing power, or again, is it a police-inspired scenario in order to appeal to public security? All this is equally true, and the search for proof, indeed the objectivity of the fact does not check this vertigo of interpretation. We are in a logic of simulation which has nothing to do with a logic of facts and an order of reasons. Simulation is characterized by a precession of the model, of all models around the merest fact—the models come first, and their orbital (like the bomb) circulation constitutes the genuine magnetic field of events. Facts no longer have any trajectory of their own, they arise at the intersection of the models; a single fact may even be engendered by all the models at once.56

Here, Baudrillard nearly throws conceptual caution to the winds, though (to be sure) not the conditions of coherence and plausibility that make his own text so engaging. So he suggests the possible explosion of radical history and interpretation, by favoring an untenable description of what is genuinely possible —pour épater Foucault perhaps. He is the polar opposite of Weitz, therefore, who senses that the same explosion must exceed the conservative resources of the modernist (the Wittgensteinian) idiom of reference and interpretation he (Weitz) favors. The better option lies between these extremes.

My own concern in all this has been to demonstrate the coherence and viability of the radical conception as well as the reasons for thinking that its practice cannot continue to be refused. But in pursuing its possibility, we have discovered that the conditions under which it must be admitted are the very same conditions under which reference and predication everywhere depend. There's the surprise and the argument's strength.

The best examples of the radical effort, doubtless—those that preserve the conditions of reference—are to be found in the interpretive exercises of Roland Barthes and Harold Bloom.57 But, in a larger sense, the ongoing work of all interpretation and history may be construed in accord with it. As already remarked, successful reference everywhere depends on the same conditions that make radical history viable. As soon as one sees the force of this, one sees how easy it would be to generate, in reverse, from the vantage of a captured future, the various poststructuralist, hermeneutic, and modernist conceptions of history and interpretation. For each of these imposes an increasing conceptual restriction on items (iv')-(vi') of the tally just completed. Hence, the inquiry has actually yielded an orderly declension of all the principal theories of history.

I am aware, I may say, that the drift of the argument will be alarming to some. For it appears to have proceeded from a strong adherence to the rigors of an analysis of reference, predication, individuation, and numerical identity, and then to have veered off (unaccountably, some may say) to support a certain "radical" conception of history and interpretation at variance with the usual philosophical findings regarding the former. I admit that this is partly so. But I insist that I have not departed in any way from the best practice of analytic philosophy. I bare departed from its usually associated convictions: for instance, in admitting that thinking is inherently historicized and that the "natures" of cultural entities may be altered by interpretive intervention. My own claim is that, in adhering to these doctrines and pursuing the problems of reference and the rest, philosophical rigor ineluctably leads to radical history and radical interpretation. Even if you oppose those prior doctrines, you must see that this last finding is both coherent and follows from them. That is a very large gain.

I must add a final word to clinch that gain. Everything so far said, you remember, bears on the individuation and reidentification of one and the same text under changes effected by interpretation. It won't do to claim that interpretation in accord with the model I have elicited

from Barthes and Bloom, and now from Weitz, posits a numerically distinct text for each member of every pair of admittedly valid interpretations that are plainly incompatible on a bivalent logic. For, of course, you must antecedently individuate that text to which a valid interpretation is assigned. It follows at once, therefore, that to admit the foregoing argument is, effectively, to embrace a form of relativism.58 On the view I propose, relativism requires that what would ordinarily be incompatible interpretations may be validly ascribed to one and the same text. There are those who, through one retreat or another, either substitute (implicitly) different referents for such (incompatible) interpretations or deny that, where they are thought to be "justified" or "appropriate," their being such is not a question of truth or the like. The first group of theorists often call themselves "pluralists" (rather than "relativists"); the second often call themselves "noncognitivists" (rather than "cognitivists").59 1 do not share their conceptual fears. (I shall return somewhat later to the logic of relativism.)

PART II

Chapter 4
Puzzles of Pictorial Representation

If cultures are the actual milieux in which humans live and flourish and make and do the things they do, then there is no convincing sense in which art and nature can be disjoined as unreal and real, or as embodying orders of reality inferior or superior one to the other. The banal mystery of human existence insists that persons belong to both "worlds" at once—the natural and the cultural—but there is no generally accepted account of how this should be understood. Jean-Paul Sartre is perhaps the best-known contemporary theorist of the arts who has confused that issue (the issue of whether cultural phenomena are genuinely "real") with the more profound one of distinguishing, within the one real world, the mark of the cultural and the mark of the natural.1

Sartre had supposed the difference rested with the difference between the perceptible and the imperceptible, whereas one might better claim that the visible may have to be construed in several ways, but that in none of them can it be convincingly supposed that what we report as perceptually accessible is ever free of the complexities of our cultural habits of perceptual discrimination. The reality of our cognitive powers is inseparable from the effective reality of the enculturing society in which they are first formed and exercised. The immediate consequence is plain: if artworks and cultural phenomena are "unreal," as Sartre signifies by his perverse elevation of the arts, then human selves, the culturally apt cognizing and active subjects we take ourselves to be, are also "unreal." And then, the point of uttering the

first pronouncement is completely lost—or threatened. There must be some conceptual adequation between truth and knowledge and reality. The issue is not altogether unrelated to Arendt's rhetoric regarding the concealment of Being in the revelations of Appearance, but Sartre's is more paradoxical. Arendt pretends to reclaim the "sempiternal" lessons of history in its own transience; Sartre produces what appears to be a self-defeating doctrine.

To move too quickly to this conclusion, however, is to obscure the odyssey by which it is found to be so. We need something of the record of the journey. For, otherwise, it would be impossible to offset the entrenched dogma that the real world is unaffected by "mere" perception, that it has whatever invariant structure it has apart from human inquiry, that claims of truth are always simply true or false, that thinking and perception have no history (or have no history in common). The point of the caveat is to concede that the epithet "real" is meant univocally as between nature and culture but that, nevertheless, we cannot discern the distinction of what belongs to those two "worlds" by nature's means but only by culture's.

The lesson draws on that most attenuated sense in which we are (still) "Kantians," that is, the sense in which we are committed not only to the conjuncture but also to the symbiosis of our world and our conceptual powers and the intransparency of the world thereby disclosed.2 (Perhaps some terminological distinctions will help: the "pre-Kantian" utterly disjoins cognized world and cognizing agent and relates them epistemically by correspondence; the "Kantian" conjoins brute world and cognizing mind in the constitution of the phenomenal world but insists he can discern the separate contribution of the mind; the "post-Kantian" construes that conjuncture as an indissoluble symbiosis under intransparency; and the "poststructuralist" historicizes symbiosis. On this reading, Kant illicitly betrays the role of noumena in his own "critical" doctrine.) Putting the point this way permits us to feature two very useful gains: first, it overcomes the lurking suspicion of some sort of "Cartesian dualism" of mind and body; and second, it confirms the ubiquity of the Intentional throughout the one intelligible world. For the first invites an analysis of the sense in which "cultural" attributes are indissolubly incarnate in "natural" or "physical" attributes (or, the associated sense in which "cultural" entities are embodied in "natural" or "physical" entities).3 The second entrenches the essential differences between the uses of the "Intentional," the "intentional," and the "intensional." Thus, if the Intentional attributes of

artworks are real, then it is impossible to confine intentionality to the mental or psychological Or subjective. Introducing the Intentional enlarges our resources considerably: it helps to identify the sui generis conceptual space of the immense world of human culture; and it permits an altered use of the standard categories of the "intentional" and the "intensional." In short, we need to match the metaphysics of the Intentional world and whatever we suppose interpretive judgments objectively confirm.

But the argument needed is a strenuous one—and somewhat unfamiliar. We shall need some patience. The master theme is this: that the intelligible structure of the natural world is inseparable from the perceptual and linguistic competences we acquire, as cognizing selves, by internalizing the enabling Lebensformen of our culture; and that, as a consequence, even the non-Intentional predicates applied to the physical world are tethered to our symbiotized powers. If so, then the fate of the complexities of the artworld—for instance, the Intentional distinctions of representationality in painting—are inexorably bound to the complexities of our own existence as culturally evolved agents. We cannot dismiss the first if we cannot eliminate or neutralize those features of our own cultural life that the usual treatment of the physical world does not call into play. I take this for granted. It is certainly contested. But the point of insisting on the connection is to enlist a measure of hospitability for the conceptual puzzles of the artworld. They repay patience, but they are likely to inspire the opposite. I claim: (1) that the artworld cannot be satisfactorily analyzed except in terms of its Intentional complexities; (2) that those complexities are the inseparable mates of what is required in understanding ourselves as selves; and (3) that those same complexities make possible the discernment of the nonintentional properties of the physical world. Doctrines (1)-(3) are somewhat heterdox: they clearly go contrary to the physicalisms of our time. I believe they cannot be avoided, but I am concerned here primarily to introduce doctrine (1) in the most sympathetic way. It is (1) that I mean to pursue, though always with an eye to its larger import.

Still, admitting (1) already subverts the pretense of Sartre's pronouncement: for, if there is no intelligibly structured world apart from the intelligibilizing powers of human inquiry, then to "elevate" art and human culture as "unreal" relative to the physical world is to defend an incoherent option. The temporal priority of the physical is itself a (thoroughly reasonable) posit within a symbiotized world. Sartre

offers no more than a florid "continental" analogue of the strongly "pre-Kantian" temperament of the physicalisms of analytic philosophy, except that his intention (but not that of the analysts) is the reverse of reduction. (I suggest that Sartre's maneuver is also the essential clue to Danto's account of history and interpretation.)

Remember: the divided powers of cognizing subjects and cognized world are inferred to have been artifactually posited (and serially reformed) by historically placed, similarly preformed, conjecturing subjects. Such subjects can claim (at no point in their own endeavor) any privileged source of knowledge about a noumenal world or about their own original powers. This is the rough postulate of "symbiosis" that is at once "Kantian-like" and (subversively) "post-Kantian," in disallowing epistemic or ontic fixities regarding the functions of "subject" and "object." Relative to this division, the real world is "external," "independent" of the particular experience of you and me; but that experience and that independent world are conceded to be what they are, within the undifferentiable "space" in which they are what they are. What is said here can only be expressed privatively: it has no other discursive role to play but to disallow privilege and fixity. In this sense, Sartre's conceptual perversity and the physicalist's presumption overstep our cognizing competence in the same way. Our picture of the physical world indissolubly implicates our picture of the cultural world (and vice versa): we cannot have one without the other.

I

Among animals, there is no reporting of sensory experience. Among humans, there is no reportable experience that is not culture-laden—interpreted. There is, therefore, no reportable comparison between sub- or de-cultured and culturized perception.4 That is part of what it means to be a "Kantian." Hence, those glib disputes about the natural and conventional status of linear perspective and so-called naturalistic painting—made prominent by the disagreement between E. H. Gombrich and Nelson Goodman—are somewhat mismanaged on both sides. The independence of the natural world, its "objective structure," is a dependent (but reasonable) posit posited within a "world" whose own assignable structure is already Intentionally symbiotized. This is our Kantian heritage, provided we eschew altogether the noumena and the fixed disjunctive contribution of mind and brute world Kant could never completely abandon.

I believe we escape the untenable Kantian themes by way of two assumptions that Kant could not accept: one, that human experience is inherently "historied" (horizoned), tacitly restricted by the same particular cultural resources by which it is empowered to affirm and support truth-claims; the other, that cognitively apt persons or selves are socially constituted as such in historically diverse ways. This means that there can be no a priori assurance that the conditions of perception, understanding, and reason are invariant in any species-wide sense. The new "Kantian" orientation includes a commitment to (i) symbiosis, (ii) intransparency, (iii) historicity, and (iv) social construction; that commitment entails that all cognitional powers are inherently interpretive.5

To put things this way is to prepare the ground for a favorable reception of Intentional complexities. I will not pretend that the conceptual connection between nature and culture is not a strenuous one. But to address the matter in the attenuated "Kantian" way I have is to ensure the initial pertinence of all forms of intentionality. There are well-known philosophical efforts to disallow altogether the bare admission that there are any real phenomena that are irreducibly intentional.6 I am not concerned to quarrel about that sort of reductionism here. My purpose is much more modest. The complexities of interpretation I have already discussed—and those I mean to introduce shortly—would make no sense without also admitting: (a) that the cultural world is real, and (b) that the cultural = the Intentional. What we must look at more carefully are the distinctive complexities of the Intentional world of art. I shall take (a) and (b) for granted; but I shall also suppose that a demonstration of just how deeply entrenched the Intentional idiom is in the artworld must inevitably bear on the prospects of reductionism.

Once the cultural world is placed, provisionally, on an equal ontological footing with the physical and the natural, the objective standing of interpretive judgments makes sense. The entire theory of art—in particular, the prospects of the theory of representational art and naturalistic painting—are affected. I have, then, by the expedient of retreating to the least disputable commonplaces of our age, postponed an inevitable question.7 But I have so far only promised an economy.

Even so, an advantage suggests itself. If, for instance, we make the easy concession of construing Brunelleschi's great dome of the Santa Maria del Fiore as an undeniable artwork, we thereby grasp at once that art is, actually exists, whether or not it represents anything; that it cannot represent anything if it is not real; that it need not represent

anything despite being real in the way of an artwork; and that it may yet have curious properties (as an artwork, as a cultural entity) that, whether they include representational properties or not, are not likely to be characterizable in physical terms alone. One may indeed wonder whether it is not true (it is true) that the same generic puzzle arises in different ways in pairing sound and language, stones and cathedrals, nature and culture, movement and action, time and history, the biology of perception and perspectival drawing, Homo sapiens and persons. (I infer from this that what I have to say about intra- and intercultural forms of art holds with equal force for intra- and inter-cultural interpretations of human life and history. I see no principled difference between anthropological and art-historical inquiries, though I am obliged to confine myself to an economy of sorts.)

If the pairings mentioned hold, then it must dawn on us that the local puzzles of pictorial representation offer an easy entry into the master puzzle of all philosophy and science. Opposing Sartre's extravagance, then, opposing physicalism, reformulating the dispute between Gombrich and Goodman, viewing these options as versions of the same fundamental philosophical decision, we may suppose that an oblique issue like that of natural perspective offers more than meets the eye. We need to be aware of what the predicative resources of the Intentional idiom are, before seriously entertaining philosophical economies that would preclude them. That, at any rate, is the intended benefit of my proposal.

A Sartrean might be misled by the following thought: if language is "about" the world, then it represents the world and is either not real in the sense in which the world is real (language being "about" it, but not belonging to it), or it exhibits some order of reality other than that of the real world it makes reference to. Both conclusions are extravagantly untrue; both arguments are non sequiturs; and both disorders are entirely in accord with Sartre's characterization of art. Sartre's motive seems to have been to elevate art by treating it as "unreal," that is, as superior to the (merely) perceivable world. Others have sought to eliminate language altogether—together with the whole of culture—lest they seriously complicate the high work of genuine science.8 I mean to steer a middle course.

If we say (as I have urged) that the cultural is the Intentional and the Intentional, the cultural,9 we should perhaps add that, among the arts, the representational is a commonplace instance of the Intentional. But what that means is hardly clear. The "Intentional" signifies a semi-

otic or linguistic or intensional property ranging smoothly over the collective practices of human societies—over the specifically purposive and conscious features of personal life. It cannot be easily dismissed and it cannot be captured by the (global) representational idiom of the pre-Kantian empiricists or Kant himself.

What is essential to a metaphysics of culture rests with recognizing that the Intentional is: (a) real, (b) not in any obvious way reducible to nonintentional or nonintensional elements of any sort, (c) symbiotically implicated in whatever we take to be the discernible features of the world, (d) inclusive of whatever is intrinsically significative as a result of cultural work, (e) apt for that reason for interpretation, (f) public and objective in a consensual or collective sense, and (g) inherently historicized. These distinctions form an intuitively reasonable and appealing set, but there can be no doubt that their admission has seemed to many theorists to be subversive of the best instincts of a rigorous science. (At any rate, there is no convincing a priori model of what a science is.)

Brunelleschi's dome is certainly Intentional in at least this sense: its properties invite interpretation. But there is no reason to suppose that it is specifically representational in addition to being, say, expressive of, or endowed with or originally used in accord with, the gathering spirit of Brunelleschi's perception of the Florentine world; and even if it were representational—in the same general sense, say, in which one speaks of the drive in the Gothic vault to achieve or represent the divinely infinite—no one such (Intentional) function could reasonably supplant every nonrepresentational but still semiotized function.

All this (I say) is put at serious risk in saying—insouciantly enough— that art and language are "about" the world. We assume we know what that means, but, quite honestly, it is not always clear. "About-ness" may be a near-synonym for my "Intentionality." But if it is, then "aboutness" is not, in the logical sense, transitive; is not, in the grammatical sense, necessarily relational; and is not, in the global metaphysical sense aptly fitted to the world of culture, necessarily or invariably representational.10 It is also obviously not restricted to the mental; and it cannot be discounted as unreal. Vermeer's Lady Reading a Letter at an Open Window—- undoubtedly "about" the seventeenth-century Dutch world already so altered from Rembrandt's, and understandably offered as a specimen of representational painting—does not in any obvious sense represent that world, the actual Dutch world (reference to which does indeed facilitate our recognizing it as a cultural

artifact of the Dutch world and our recognizing in it what is represented there ). It's true that there is represented in it a young woman, a casement, a letter, a table, a carpet, a drape, a chair. These, represented, need not be taken to denote anything belonging to the actual world. But what is represented in the painting is represented only on the sufferance of the larger "aboumess" or Intentionality of the painting itself. One of Kandinsky's familiar Improvisations represents nothing, neither denotatively nor nondenotatively (as we say), though it "discloses" an artifactually visible (entirely Intentional) "world." It is significant in this sense, it is "about" that world, it is Intentionally so structured. In that respect, there is no difference between the Vermeer and the Kandinsky; but what we mean by "that respect" remains a palpable puzzle.

II

There is indeed a familiar theory that would ensure the representational nature of all painting and (if necessary) all art: namely, the theory that art (as Nelson Goodman seems to hold) is inherently self-referential, that art must exemplify (as by expression or self-reference) the properties it possesses. Cloth samples do something like this, it is true; but surely, it is only we who make things serve as samples. There is nothing that is a sample tout court. Only by an extravagance—for instance, a stratagem for saving a certain semiotic theory of art—could one say that every painting or every artwork refers to the (or certain) salient properties it possesses (whatever we suppose the complexity of such properties to be);11 or to say, more cautiously, that insofar as an artwork is said to possess "expressive" properties (which, perhaps, both the Vermeer and the Kandinsky do, in being "about" the visual "worlds" they respectively disclose), they reflexively represent or refer to what they express.12 It is easy to imagine a more economical formula than that of self-referentiality: the mere possession of real Intentional properties, for instance! Obviously, the notion of self-referentiality externally imposed on artworks is intended to relieve the artworld of any intrinsic Intentionality: the ulterior analysis of semiosis is then construed functionally, entirely in terms of human behavior and production. Whether this is sufficient to preclude any robust ontology of the Intentional tends to be neglected by those who espouse the strategy.13 It ignores the ontic similarity between selves and artworks. In a sense already supplied, it is meant to be a "naturalizing" strategy

for the semiotic features of artworks. (If it were metaphysically more ambitious, it might well be tempted to duplicate Sartre's solution.)

If, of course, what is represented in a painting need not exist in the real world, then (on the standard view: the view that actual reference is made only to what is real), in order to hold that a painting refers to what it thus represents, one must also hold that it does refer there—magically or by way of metaphor, or fictively, or by some other extravagance; or that, in making reference in the ordinary way, the properties it then "possesses" (that serve such reference) it possesses magically, metaphorically, or fictively. (Goodman favors the second option.) Similarly, if what a painting expresses it cannot literally express (since, on the standard view: the view that, being merely physical and inanimate, a painted canvas cannot literally possess expressive properties), then paintings are expressive only by way of an extrinsic metaphor answering to our interests. Perhaps, then, paintings are expressive only because we refer to them thus (as "exemplifying" such properties metaphorically). The vacuity of the idiom begins to betray itself; but its motive is clear enough: it means to deny that there are any real things that straightforwardly possess Intentional properties. (This is essentially Danto's view: that is, of artworks—but not, I think, of persons. Here, Danto and Richard Wollheim diverge, though difficulties arise for each.)

We face a disjunctive choice: we may (try to) say that (mere) physical objects or physical media, have, or can support, no expressive or representational attributes as such, though, extrinsically, we ascribe such properties to them ("metaphorically"); or we may say that artworks are not mere physical objects but culturally complex phenomena that intrinsically do possess such properties. The ontology of art, therefore, is inextricably linked to what we take our own "nature" to be. The decisive datum at the bottom of the story is this: we cannot attribute Intentional properties to ourselves —extrinsically. But if that is so, then, on the argument favored, there is no conceptual gain in denying such properties to artworks, since the cultural world is ultimately consensual and real in reflexive terms.14 It is this very space in which we emerge as ourselves.15

There are "objects" represented in the Vermeer, then; there are none in the Kandinsky. But both are "expressive" in the generous sense conceded, the (grammatical) sense of being intransitively "about" the artifactual visible "world," the one made accessible only through the two-dimensional painted surface of each of those canvases. The "aboutness"

of each uniquely controls our access to their perceivable "worlds": perceiving them pertinently is (pertinently) perceiving their visible "worlds." But if that is so, then, whatever other sense we assign the term "representation," no painting is primarily a representation, in that it first discloses (must first disclose)—to those rightly trained in the pertinent (appreciative) practices—a visible "world" not otherwise accessible: a "world" in which a drape or lady may then be seen to be determinately represented, though (once again) without presuming that the painting somehow refers to whatever is thus represented in it.16 The Intentional extends to that facilitating process as well as to the discernible structure of what it discloses. Both are (culturally ) real.

The represented world disclosed in the Vermeer and the nonrepresentational world of the Kandinsky are accessible for viewing only through the culturally prepared medium of the brushwork. The worlds thus "Intended" cannot be detached from their enabling medium any more than pains or subjective qualia can be detached from a human subject's mental states (or, those states from the same subject's biological states). "Representations" and qualia are not like rubber balls in physical space: they cannot be located as real independently of the inseparable housing of their respective "medium." There you have the fatal (and common) fallacy of "naturalizing" mind and culture, by way of neurophysiology or by restriction to a purely physical medium.17

Three very difficult themes form a crossroads here. First of all, the relationship between the cultural and the natural cannot be satisfactorily reduced to that between the mental and the physical; on the contrary, in the human sphere, most of what is "mental" or "psychological" cannot be adequately fathomed if not in terms of what is culturally pertinent (the linguistic, for instance). The decisive factor lies with the collective nature of the cultural—in the sense intended by Wittgenstein's Lebensformen or Gadamer's "traditions" or Foucault's epistemes.18 The collective predicables that constitute the Intentional, that qualify the mental or subjective attributes of individual human agents and that are themselves incarnate in physical attributes, cannot be reduced to either the mental or the physical. I claim that there is no successful reductive resolution of this matter. Treat that as a philosophical bet.

Second, the objectivity of interpretation depends on whether we are able to provide a reasonable sense in which cultural phenomena form a real part of the world, stably and distinctively what they are in a way

sufficient to support a disciplined practice of truth-claims. I claim that the cultural world is real in the same sense in which human selves or persons are real, real in sharing a "form of life" in terms of which they recognize and understand one another. Hence, conformably, their deeds, their work, and what they produce are real. If that be conceded, it follows: (1) that the metaphysics of human existence and the existence of a world of art and culture and history are conceptually inseparable; and (2) that the objectivity of interpretive claims about history and art cannot be satisfactorily modeled on the supposed observationality of a mere physical world.

That is the crux of the problematic relationship between the so-called Naturwissenschaften and Geisteswissenschaften. That is not to say, however, that the Vermeer and the Kandinsky cannot be "observed" or "perceived" but only that the perception of them is inseparable from an (interpretive) "understanding" of their Intentional structure; also, that what can be sensorily discerned in them cannot be disjoined from, or given priority over, what can be thus understood. This too is a philosophical bet. It signifies that the best way to construe the objectivity of interpretive judgments is in consensual terms—though not, for that reason, by construing consensus criterially. Our cultural aptitude, rightly exercised, first "discloses" the world of the Vermeer in which the represented lady and the casement window can be "observed." Observation is an artifact.

Third, since human existence and the reflexive understanding of the human world are historical and historicized (that is, since existence and understanding change over time, change Intentionally, and—as a result—ineluctably alter our cognizing powers), we must consider whether and why we should subscribe to the usual bivalent truth-values physicalisms have regularly favored. Here, the salient puzzles concern the fortunes of historicism and relativism.19 This too is a philosophical bet: I claim that some accommodation of relativism is inescapable. The question of fitting a logic and a methodology to this or that domain of inquiry is, I assume, an empirical question. That is: once you admit the complexity of interpretive practices and the realism of the cultural, you cannot then convincingly avoid admitting some moderate relativism at least. (That, at any rate, is an important part of the argument intended.)

For the time being, I merely draw your attention to the strategic importance of the three distinctions mentioned—the collective, the

consensual, and the historicized—which, I say, form the deeper concerns of this entire inquiry and bring into focus the important quarrels of the philosophical tradition regarding the reality of the human world.

There is a natural misunderstanding that threatens these distinctions. We must be careful. I have just said that mind and culture cannot be "restricted to a purely physical medium." On an unsympathetic reading, this will be construed as reinstating a Cartesian dualism of body and mind. But I oppose dualism. I am not claiming that there are two different "stuffs" in the world, but only "one" (whatever its analysis may finally confirm): that is, I subscribe to "materialism." But I am also insisting that cultural phenomena have properties that are not and cannot be suitably expressed as purely physical properties: that is, I reject "physicalism." To grasp the difference between materialism and physicalism is to grasp the conceptual difference between the mind/ body problem and the culture/nature problem. (By "materialism," I mean no more than that I take the real world to be "composed" of "matter"—whatever matter is found to be, by way of the natural sciences: suitable for quarks and quanta and living processes and the like—in whatever sense the world is composed of some fundamental "stuff." But I do not mean by that that whatever is so composed cannot also exhibit intrinsic properties that are not physical or material or not merely physical or material. On the contrary, my mention of "incarnate" properties is meant to accord with materialism but not with physicalism.)

I cannot do full justice to these grand issues here. But I can offer a small clue sufficient for my purpose: the properties of the culturally prepared medium of each of the particular arts (words and sentences, dance steps, musical tones, brushstrokes) are "incarnate" in the properties of the physical medium in which each matched art is similarly "embodied" (sounds or marks, bodily movements, colored paints and canvas). I do not need to invoke psychological predicates here, whatever may be the culturally enabling conditions of art.

Thus, the represented "world" that a (representationally functioning) artwork is Intentionally "about" (whether painting or literature or music) cannot, logically, be identified apart from that artwork; similarly, the felt qualia of psychological states (pleasure and pain, for instance) cannot be separated from the neurophysiological states in which they are incarnate. Hence, for broadly analogous reasons, neither can be rightly characterized in terms restricted to the physical medium

(that, ex hypothesi, exhibits no Intentionality) in which both artworks and psychological states are indissolubly embedded.

This is also the sense in which Intentional properties are monadic. Once admitted, their (Intentional) "content" may be articulated in as complex a way as you please; but they themselves are directly attributed, much as simple qualities are, to the things they are the intrinsic properties of. Thus: "being red" is attributed to a painted (physical) surface, whereas "representing a lady at a window" is rightly attributed to the Vermeer. The question of "adequation" between property and object is the same in the two cases.

Here, then, is the sticking point of speaking of aboutness or Intentionality (or, more narrowly, of intentionality in Brentano's and Husserl's sense), the point of the systematic difference between the natural and the cultural. It is not, as Sartre supposes, that the first is perceivable and the second, not; it is rather that the perception of the second is more complicated than the perception of the first. In the second, certain nonsensory structures of Intentionality control both the enabled perception of a "world" uniquely disclosed by the executive devices of a particular painting (or musical composition, for that matter) and also (dependently) the "objects" discernible in that then-disclosed world. It needs to be said that (on the attenuated Kantian thesis I have acknowledged) the perception of the physical world is also controlled by the devices of aboutness (by conceptual categories and linguistic practices), but its proper objects are free of all Intentionality.20

A represented chair in a Van Gogh interior is seen to possess a certain existential pathos; but a "natural" stone, catalogued as a stone in accord with one physics or another, completely lacks aboutness, but, as a pendulum, it does possess inertial properties. This is the immensely important point of Thomas Kuhn's otherwise baffling remark that, since "Lavoisier . . . saw oxygen, where Priestley has seen dephlogisticated air and where others had seen nothing at all, . . . Lavoisier saw nature differently"; and that, as a consequence (not that every difference of perceptual belief has this effect), "Lavoisier worked in a different world" from Priestley's.21 We are so taken with the contrast that we begin to suppose that a "natural" chair (like a stone) is also nothing but an ordered pack of physical goods lacking aboutness, to which, inessentially, extrinsically, we somehow ascribe a certain "function" answering to an "alien" human interest. (The discussion of the preceding chapter begins to apply here.) The strange thing is that

we are then also tempted to speculate that the entire world of art is nothing but the physical world to which, again extrinsically, we assign a certain Intentionality.22 The threat is already present in Sartre's formula; it is also favored by Goodman and Danto.

The Intentionality of the physical world (in effect, its symbiotized conceptual conditions, not the range of properties it admits) is executively controlled by the same aptitudes that create the world of art. This is the essential point of Kuhn's remark, although Kuhn himself was clearly troubled by it. Only the elimination of that strange being, the human person, the self—the agent of both art and science, the denizen of both nature and culture—could possibly vindicate the finding that art and history (the site of interpretable things) are nothing but a façon de parler regarding the physical world. For, admitting the symbiosis of world and word, there is no science if there is no language or art.23

Of course, in spite of all this, paintings often are representations. The general rationale is prophetically captured in Roger Bacon's Opus Majus. Bacon promotes a strict knowledge of science and geometry as a way of understanding how God's grace orders the created world: "It is impossible," Bacon says, "for the spiritual sense to be known without a knowledge of the literal sense." Hence, as he goes on: "Since. . . artificial works, like the ark of Noah, and the temple of Solomon and of Ezechiel and of Esdra and other things of this kind almost without number are placed in Scripture, it is not possible for the literal sense to be known, unless a man have these works depicted in his sense, but more so when they are pictured in their physical forms; and thus have the sacred writers and sages of old employed pictures and various figures, that the literal truth might be evident to the eye, and as a consequence the scriptural truth also. For in Aaron's vestments were described the world and the great deeds of the fathers."24

Whether particular paintings are actually emblematic in this way, in representing pictorially what may be propositionally extracted as a lesson—an issue that has been notably disputed in the context of Dutch painting25 —is a contingent question at best, hardly one that would decide whether, in failing the emblematic criterion, a would-be painting was not an artwork at all or did not exhibit the required aboutness. Bacon's concern centers on the privileged reality that is said to be "represented' in nature itself and in suitably compliant art. That is, by construing nature as "representing" God's providential order, Bacon captures, by way of analogy, the theme of symbiosis: in both the me-

dieval and post-Kantian doctrines, the natural world manifests (what I am calling) an Intentional structure. My point is simply that the "worlds" artworks disclose are tethered to a certain generic "about-ness" that, distributively, they intrinsically possess. Bacon's thesis holds (extravagantly, I should say) that nature manifests a similar "about-ness." I deny that— with Kant—but agree (against Kant) that the intelligible structure of the natural world is itself inseparable from the "aboutness" of the cultural world. In short, the "worlds" artworks disclose, like the "worlds" Kuhn assigns to Lavoisier and Priesdey, are plural because of our intervening interpretive tertia (our conceptual schemes). The sense in which there is "one" world (not "one" because of number) is the vacuous (but not unimportant) sense (Husserl's sense) in which whatever we speak about belongs to the same ("one") world "we are speaking about." Finally, the sense in which nature functions representationally, for Bacon, presupposes the possibility of multiple Creations and God's having disclosed the interpretive rule by which our world's representationality may be deciphered.

Several important theorems suggest themselves here. For one thing, to deny a realist reading of the world of human culture is tantamount to denying a realist reading of ourselves (ourselves as selves). For a second, to adopt a Kantian-like view of the natural world (symbiotized, intransparent, historicized) is to admit that a realism of phenomenal nature is the conceptual analogue of a realism regarding the cultural world, regardless of their discernible differences. And, for a third, the symbiosis of the natural and cultural worlds implicates, holistically, the collective Lebensformen of communicating societies; hence, it cannot be characterized in merely mental (Kantian) or subjective (Husserlian) terms. In short, unless a suitable reductionism were con-firmed—which I believe unlikely—the realism of nature and of culture must be seen to be indissolubly linked.

The important thing about the emblematic interpretation of (Dutch) painting, mentioned a moment ago, is that it offers a coherent possibility even if it is a mistaken one.26 It helps us understand that representation may take many forms and that "aboutness" is not exclusively representational. The emblematic thesis holds, as Svetlana Alpers summarizes the matter (in order to reject it), that "Dutch Art . . . only appears realistic . . . is [rather] a realized abstraction."27 (The most familiar emblematic device is that of illustrated proverbs, but the claim is surely more far-reaching.) Nevertheless, the representation of meaningful scenes is not the emblematic representation of

meanings. Right or wrong, the thesis points to the importance of distinguishing between the primary aboutness of paintings (their generic Intentionality) and the further possibility of a specifically representational function. The emblematic possibilities of Breughel's Proverbs may not bear at all on the representational possibilities of Breughel's Seasons; and the representational possibilities of van Eyck's portrait of Arnolfini and his bride may not capture the ulterior (Intentional) function of the painting (functioning thus) as a documentary utterance. These are oblique ways of acknowledging that interpreting artworks is hardly logically uniform. Only if we grasp the Intentional structure of the different ways in which artworks function can we venture anything like a plausible theory of interpretation.

Samuel Edgerton, addressing the direct consequence of views like Bacon's on the emerging worldliness of the Renaissance some centuries later, remarks that, as far as he can determine, in the Brancacci Chapel of the Santa Maria della Carmine, two frescoes, Masaccio's Tribute Money and Masolino's Raising of Tabitha and Healing of the Cripple, "are the first in the entire history of art to illustrate the scientific-artistic principle of [what Edgerton terms] horizon line isocephaly" (the convention of representing the relative distances of persons in groups by the horizontal ordering of their heads).28 Edgerton does not say that, in employing (as they do) Brunelleschi's rules for the use of perspective, these artists were representing those rules or their use, or perspective itself, or, emblematically, the grace of God through perceptual geometry. (Neither isocephaly nor the vanishing point principle are actually explicit in Brunelleschi's handbook.) And even if we supposed that, given the occasion, there was some application of Bacon's instruction, such an application must already have been fading into a new perspectival habit regarding aboutness that need not entail the telos of nature.

III

Aboutness is a monadic (but complex) rather than simple feature of predicable structure: what is identified, in understanding pertinent utterances (in speech and painting, for instance), is relationally identified within and only within the scope of that one-place structure. Such a structure is Intentionally (or culturally) "emergent," not "supervenient" (in the current sense)29 —but real enough—and not committed, as such, to relations (that is, not actually polyadic). The pertinent dif-

ferences lie with this: (a) what is "supervenient" (what is psychological or cultural) is (must be) open to extensionally equivalent truth-claims regarding selected physical phenomena, but this need not be true of what is (culturally) "emergent"; (b) what is "supervenient" is first indexed by reference to a physical particular to which it is extensionally bound, but this is neither required nor characteristically true of what is "emergent" (Intentional); and (c) what is "supervenient" may be explained in accord with some (nonreductive) nomological account (perhaps in accord with a version of the unity of science program), whereas what is culturally "emergent" is sui generis and cannot be explained in such terms. Hence, supervenient phenomena are not discursively manageable except in terms of a reasonably adequate physicalism; whereas emergent phenomena are manageable in terms of the consensual lebensformlich practices of an encultured society—which are themselves not explicable in purely physicalist terms. The contrast is very strongly drawn.

I venture to add that the supervenience thesis is demonstrably false if the Intentionality of art and culture is admitted; for, on that admission, the Intentional features of relevant referents are: (1) not fixed or unique, (2) variable under interpretation, (3) plural in an openended way, (4) capable of being "incongruent" in the relativistic sense (see chapter 7), and (5) subject to change as a result of historical and interpretive processes. These traits are plainly incompatible with supervenience—but not with emergence.

The seeming detachment of meanings and representations from those culturally freighted utterances (locutionary acts, executed brushstrokes) that control access to them is of course misleading. There are no independent meanings or representations that human speakers and artists somehow collect or select or contingently work with: that is the way of platonism.30 Meanings and representations are Intentionally inseparable from the "utterances" in which they are "incarnate"—hence, they are always and only interpretively imputed.

There are two constraints that I support here: first, "meanings" are inseparable from ("incarnate" in) the physical features of the "utterances" (speech acts, paintings) of which they are the monadically attributed content; and second, such utterances, which are Intentionally qualified in being open to such ascriptions, are themselves (correspondingly) indissolubly "embodied" in some physical or biological phenomena with respect to which they are culturally emergent. That is why meanings and artworks may be "adequated" to one another.

The only escape from platonism—if we mean to escape—is to treat aboutness monadically: that way, the difference between the real and fictional referents of any Intentionally accessible "world" becomes an independent matter; the monadic dependence of any such emergent "world" (its "discernibility") is no longer at issue. Only the proper interpretation of particular Intentional structures remains an open question. Otherwise, the "world" represented (for instance, in Rembrandt's using Hendrickje as a model for "rendering" Bathsheba) would require that Bathsheba exist (and "be as the painting shows her to be") even though "Hendrickje need not exist."31 No, representation partakes of aboutness essentially, monadically—as an intrinsically ascribable property of Rembrandt's painting, not of the physical canvas in which it is emergently embodied. The further question of whether whatever is identified within the scope of that monadic structure occurs or exists independently of its being represented remains an entirely fair question. That, of course, is a distinct advantage.

Monadic representation, as I am using the term, is, however, an equivocal notion: it signifies either the generic aboutness of a painting or other artwork or (as for Goodman) the representation of what is imaginary. Both monadic representation (in the second sense) and dyadic representation (as in portraiture) signify logically complex findings relating what is represented (as in paintings) to what may be found in the actual world.32 But that does not affect the reality of the representing structure of such paintings. I claim that it is not enough to construe the representational structure of paintings in terms of the representational intentions or activities of painters.

A felicitous distinction for managing these extraordinary difficulties is afforded by the old notion of natura naturans and natura naturata: in the space of human culture, in the space of shared Lebensformen, pertinent phenomena must first be bodied forth (rendered accessible: understood and thus perceived, Intentionally "disclosed") in order to function in whatever way they do. Aboutness or Intentionality serves in this dual way. The first way is presupposed by the functioning of the second, but they need not be sequential or sequentially discerned. A painting is "about" a world internal to itself—in the monadic sense in which what is emergently complex within its intransitively uttered structure ("naturata") is identified thus only on the condition of its having first been suitably uttered ("naturans").

It is suitably "uttered," I claim, in merely conforming to the histori-cized lebensformlich practices of a viable society. Nothing more strenuous is required. (Paintings need not "refer" to what they are thus

"about.") Sappho's uttering her poems raises the question of what they, the poems (thereby uttered), represent. The strokes produced by Rembrandt, studying or conjuring up Hendrickje in his imagination in order to represent Bathsheba, do not represent themselves or Hendrickje in the sense required; and the assumed parity between the image and the associated proverb in the Steen interior (Easy Come, Easy Go) merely shows the double function of the image, in the sense of about-ness or representational uttering (naturans ) and of aboutness or what is represented (naturata )

No one thinking along these lines could doubt that the validity of interpretive claims is determinately affected by what we may say about the ontology of art. One now sees the sense in which the objectivity of interpretive judgments is a function of how the Intentional structures of the collective life of a society provide evidential support for the attribution of particular such structures to particular artworks or texts. There is no plausible sense in which Intentional attributes can or must be extensionally correlated with physical attributes (as in the physicalist's view of "supervenience").

The representationality of a painting, then, is a functionally specialized role of a particular painting's essential aboutness. Its "essential" aboutness is simply its being a certain significative referent—an artwork or text uttered and apt for interpretation— because, being such, it intrinsically (monadically) possesses semiotic or similar properties.

Also, of course, the representations that may be found in paintings (the Vermeer, for instance) presuppose their essential aboutness (naturans ), for it is in virtue of that that they function as the details they do (naturata ). Only thus are natural marks "transfigured" (as Danto puts it) into significative marks, made to bear (intrinsically ) incarnate meanings or to embody culturally significant forms. Still, even to speak of "transfiguring" physical marks into significant (or significative) marks is misleading—is already too concessive to a failed physicalism.33 (It favors supervenience over emergence and it ultimately denies the sheer reality of art.)

Danto's formula is quite explicit in warning us that artworks are not metaphysically distinct, that "they" are no more than the metaphoric illusions produced by viewing "commonplace" objects (objects that are, metaphysically, accessible to an idiom that does not countenance distinctive cultural—culturally emergent—entities). Speaking of Andy Warhol's "Brillo Box," for instance, Danto remarks, in closing The Transfiguration of the Commonplace: "In the end this transfiguration of a commonplace object transforms nothing in the artworld. It

only brings to consciousness the structure of art which, to be sure, required a certain historical development before that metaphor was possible. The moment it was possible, something like the Brillo Box was inevitable and pointless. It was inevitable because the gesture had to be made, whether with this object or some other. It was pointless because once it could be made, there was no reason to make it."34

The "metaphor" is art itself. The daring of Warhol's "philosophical" act was to "transfigure" a "commonplace" artifact into an artwork. "As a work of art," adds Danto, "the Brillo Box does more than insist that it is a brillo box under surprising metaphoric attributes. It does what works of art have always done—externalizing a way of viewing the world, expressing the interior of a cultural period, offering itself as a mirror to catch the conscience of our kings."35 Danto offers no way of accounting for the identity and reidentification of artworks as referents, or indeed for "their" natures or stable attributes. His mastery of the artworld permits him to range with assurance through many an interpretive and historical detail; but he collects his interpretations, ultimately, by means of an idiom closer to "supervenience" than to "emergence": there are, for Danto, no such entities as artworks; they are conversationally fixed in whatever way they are (in a spirit not unlike Sartre's) by a cultural rhetoric practiced on the "commonplace" world. But the rigor of history and interpretation demands a more reliable sense of an "objective" world.

More than that, the constraints of predication cannot be different for the "commonplace" world from what they are for the artworld. Reality is not a two-story world (contrary to Sartre's sly compliment). But if it is not, then, Danto cannot sustain his thesis; for, per impossibile, he requires an ulterior order of reality relative to which, extrinsically, he can invoke the "metaphor" of ascribing literal physical attributes to artworks. If you add to this the puzzle of how a society of encultured selves characterize themselves as real, you see at once the aporia of Danto's account. It is no more than a version of the aporia of physicalism (whether eliminative or reductive or nonreductive), even though Danto does not actually subscribe to physicalism (and even though "commonplace" objects include diagrams, ties, painted canvasses). Also, of course, the theory that interpretation imputes attributes to (commonplace) "objects" extrinsically is the mate of the modernist theory of history. For modernists, as we have seen, history is also extrinsic to physical time.

The general Intentionality of paintings confirms the logical perti-

nence of interpretation. The specific devices of representation problematize the sense in which its actual practice is judged objective. Certainly, not grasping the documentary function of Jan van Eyck's Giovanni Arnolfini and His Wife impoverishes what can be perceived in the "world" it discloses;36 or, for that matter, not perceiving the difference between Flemish perspective during van Eyck's period and the perspectival rules of Brunelleschi's Florence; or, even failing to consider Michel Foucault's speculation about Velizquez's Las Meninas, wondering (say) whether "perhaps there exists, in this painting by Veláz-quez, the representation as it were, of Classical [baroque] representation, and the definition of the space it opens up to us," so that "representation, freed finally from the relation that was impeding it [the relation of resemblance to actual figures, according to the canon], can offer itself as representation in its pure form."37 But, after all, the same would be true of perceiving a Balinese cockfight.

IV

Representationality is an extraordinarily complex affair. It has its intentional side and its perceptual side. Both are Intentionally constrained and, accordingly, interpretively dependent on the historically shifting background against which particular paintings (or other apt artifacts) are examined. Representational paintings actually help to clarify our predicative practices in perceptual contexts and, in doing that, ineluctably fix the strategic role of interpretation. The puzzles about perceptual similarity are easily misread, however, wherever overly familiar categories like those of colors, natural kinds (dogs and cats), and highly conventional artifacts (furniture) are favored.38 But in turning to perceived similarities in paintings, we are instantly confronted with the role of representational conventions. Let us see where this leads.

It is quite usual to speak of representation as either relational or nonrelational, dyadic or monadic, as depending on whether what is represented exists or not. If, however, representation may be of propositional referents, then, of course, both the monadic/dyadic distinction and the notorious question of resemblance cannot fail to be merely local issues. Monadic (pictorial) representation, "representation-as," cannot visually resemble what does not exist, is fictional, or merely imagined (though it may aptly represent what it actually represents, for instance in accord with a descriptive text);39 but the representation of a maxim does not even raise the question of resemblance (as op-

posed to aptness) and is not suitably captured, conceptually, by the notion of the pictorial representation of what is itself sensorily con-ceived—whether monadically or dyadically (Silenus or Arnolfini).

Pictorial or visual resemblance, however, obtains independently of representation, although visual resemblance may be enhanced here and there by culturally entrenching the function of particular representations—even where it is not intended. One may, for instance, have spotted Matisse women in the New York subway or Modigliani women at fashionable restaurants. And Picasso's famous remark to Gertrude Stein on the occasion of having painted her in her prime is less a prophecy of how she would age than an insight into the influence of the practice of painting and of the perception of paintings on our perceptual and interpretive habits. What, then, of the famous question of the natural resemblance between painting and object, or of the allegedly conventional nature of similarity? The validity of interpretation is plainly at stake.

The question has two foci: the first, largely neglected; the second, largely ill-presented. They are, nevertheless, closely linked and, together, they enable us to collect in a coherent and compendious way nearly all the proliferating issues regarding representation. The first is the famous problem of universals,40 which bears in an important way on the dual question of naturalistic painting and natural perspective. The deepest clue rests with the question of resemblance or perceived similarity.41 There are two arguments to press: one to the effect that resemblance is a cognitively pertinent concern (usually perceptual, as in painting); the other to the effect that resemblances, like any other cognizable features of the world, are not cognizably transparent. The one bears on the fortunes of nominalism; the other draws on the overwhelming contemporary consensus that realism and idealism are indissolubly linked. Both issues affect and are affected by the historicity of thought and perception.

The point about nominalism is elementary: it is impossible to account for the spontaneous extension of perceptual and linguistic practice beyond exemplars—as in the use of natural languages and related cultural habits—by any strictly nominalistic theory. The reason is simply that nominalism has no way of accounting for the extension of predicative distinctions first introduced by way of paradigms. The extension of predicables cannot be satisfactorily effected by merely extending the scope of their predicates (the "names" of the predica-

bles), for that would be a wild gesture, one without any cognitive credentials at all (nominalism proper); nor can their extension be effected by applying any supposedly general concept subtended and first introduced by such a name or predicate (conceptualism), for that would reintroduce a kind of realism. This suggests that interpretation is ineluctably improvisational. But, on the argument, so is perceptual judgment.

The upshot is surprising: the aptness of all predicates applied to the natural world are as much dependent on our cultural (our lebensform-lich ) habits as are the predicates applied to artworks. In short, there is no gain in objective rigor that results from favoring "supervenience" over "emergence": our predicative practices require—everywhere—the consensual objectivity of our forms of life. Danto's treatment of art and criticism in terms of a rhetoric imposed on the "commonplace" suggests that, in the commonplace world, discursive (predicative) practices firmer than any rhetoric could ever ensure may be counted on. But that is nowhere shown to hold, and it is (I believe) mistaken. If so, then the rigor of the physical sciences cannot fail to be dependent on the practices that treat the artworld as consensually real.

These considerations confirm the essential weakness of Nelson Goodman's extreme insistence—for example, against E. H. Gombrich—even when Goodman agrees with Gombrich, that "dyadic likeness between particulars will not serve to define those classes of particulars that have a common quality throughout."42 It is true that both Goodman and Gombrich reject "the myths of the innocent eye and of the absolute given" as, as Goodman puts it, "unholy accomplices."43 But, adhering to his well-known nominalism, Goodman will not and cannot concede an effective biological or cultural basis for spontaneously recognizing resemblances. As Goodman puts it: "Realism is a matter not of any constant or absolute relationship between a picture and its object but of a relationship between the system of representation employed in the picture and the standard system . . .. Realistic representation, in brief, depends not upon imitation or illusion or information but upon inculcation. "44

The trouble is, Goodman does not admit the full (and fatal) force of his own concession: namely, that even the use of a "system" of representation (or "inculcation," in the sense he concedes: for instance, under "laboratory" conditions) defeats the strict nominalist. For what could possibly account for the successful use of any such system? It is

the entrenched interpretive practices of a society's form of life that alone ensures the "proper" extension of predicates: for, first of all, the application of a predicate in new cases obtains symbiotically, within some Lebensform —so there is no way to fall back to strict nominalism or realism (regarding "universals"); and, second, we validate such extensions only in consensual ways—though validation is not mere consensus. In this sense, interpretation extends to the entire range of predication. Alternatively put: there is no problem of real "universals"; there is only the genuine problem of real generality. (That is, what is at stake is not a question of the existence of abstract particulars: "universals"; it is a question of the reality of predicables: "real generals.") We cannot determine the "largeness" of large things (as in Plato's joke) or the "whiteness" of white things; but we do decide the question of which things, in context, are large and which are white. We do so in accord with our aggregated memory of our collective practice.

Gombrich understands Goodman's thesis to be that "there is no such thing as resemblance to nature."45 He is right in a way, but he is not precise enough. For Goodman goes on to say: "That a picture looks like nature often means only that it looks the way nature is usually painted" and "Resemblance and deceptiveness, far from being constant and independent sources and criteria of representational practice are in some degree products of it."46 By this qualification, correct though it is, Goodman betrays his nominalism: because he cannot insist on the point Gombrich assigns him, and because he cannot concede that his own formulation entails some habituated basis for extending the use of predicative distinctions. There is no other way to see the matter.

Gombrich's thesis, which concedes (or insists on) the need to learn the representational conventions of a society, opposes "extreme [mere] conventionalism"—and endorses the basically biological thesis that "recognizing an image is certainly a complex process and draws on many human faculties, both inborn and acquired. But without a natural starting point we could never have acquired that skill."47

The trouble with Gombrich's claim is the reverse of Goodman's; for Gombrich apparently means that there is a formulable sense in which representation may be made to resemble more and more closely some "natural image"—which is the point of his study of Constable's oeuvre48 —whereas, now, even on neurophysiological grounds, the biological basis for discerning resemblance is thought to be routinized by shifting molar interests, affect, contexts of past practice, and other historicized factors.49 There simply are no contextless paradigms of

predicative resemblance. Resemblance is a tacit invention of tolerable (societal, horizonal) categorizations, not a biological discovery. Resemblance is biologically grounded all right, but not in any criterial sense; it is abstracted, reflexively, from a transient but stable practice selected for survival and viability. In the human world, resemblance is inherently interpretive. Goodman cannot account for the effectiveness of interpretive conventions of resemblance; and Gombrich cannot explain how there could be a discernible (natural) basis for doing so. (In the misapplied analogy: there is no strict similarity by which to account for the frog's perceptual invariances and for perceptual generality among humans: the first is hardwired, or else learning is quite minimal there; in the second, generality is inherently interpretive, however biologically incarnate.50 )

Both Goodman and Gombrich ignore the symbiosis of predicative practices: against Goodman, symbiosis ensures a "realism" grounded in our Lebensformen; against Gombrich, symbiosis precludes any independent criteria of "realist" resemblance. The point of the lesson is that predicative practices are essentially no different in the physical sciences from what they are in the interpretation of history and the arts. On the contrary, predicative effectiveness (what Peirce calls "real generality".51 ) is an artifact of the cultural world. Hence, to ensure a realism regarding the physical world, one cannot refuse a realism regarding the cultural world— that is, always assuming the "Kantian-like" themes I have been invoking from the start—symbiosis, intrans-parency, historicity, and the social construction of selves. All predicates are Intentionally specified, whether they are Intentional or not.

V

The second question about similarity asks whether resemblance is or is not conventional. Both Goodman and Gombrich agree that it makes no sense to speak of perceptual similarities in the real world independent of human perception. This is the point of Goodman's denying "any constant or absolute relationship between a picture and its object" and any "constant and independent sources and criteria of representational practice." It is also the point of Gombrich's rejection of the "innocent eye."

But Goodman rejects as well the unique "success" of the evolution of linear perspective in two-dimensional representations; and Gombrich insists on it.52 For Goodman, there is no empirical basis for

claiming that particular modes of representation are rightly favored over others—species-wide: no "natural" resemblances are ever discernible. For Gombrich, there is a great deal of favorable evidence.

Goodman challenges the logic of the claim—both Gombrich's and J. J. Gibson's.53 Gombrich concedes: "Nelson Goodman is certainly right when he protests that the behavior of light does not tell us how we see things."54 But Gombrich moves on, more promisingly, to theorize about the way in which ecologically shifting perception (in Gibson's sense) inclines us to draw our pictures into the phenomenal world in which we live: although "the world does not look like a picture [he says] a picture can look like the world."55 Both miss what is essential to the relationship between perception and interpretation: namely, that spontaneously perceived resemblance is possible and, in effect, unavoidable; it is the consequence of a culturally entrenched (but biologically viable) form of life within a symbiotized world. Put another way: the biological constraint is primarily holistic; any set of operative predicates that accord with the conditions for the survival of the race and, more narrowly, the survival of particular societies is suitably "realist" in its mode of functioning. Beyond that, within our symbiotized world, we may conjecture about our perceptual proclivities in favor of certain notably stable resemblances. Furthermore, in the arts and history, perceived similarities tend to rely less and less on subcultural constraints. This is another way of confirming the sui generis forms of cultural emergence.

Once we grasp the argument, both the history of linear perspective and the theoretical quarrels about pictorial representation fall into place. As Gombrich firmly and correctly remarks, natural perspective and the rendering of physiognomic likeness are guided by the phenomenal appearance of the world—not by the geometry of light. But the phenomenal appearance of the world is itself elastic, variably shaped by the different perceptual and nonperceptual practices of different historical cultures.

There's the remarkable lesson, the one constantly in danger of being lost in the refined air of Gombrich's and Goodman's quarrel. The perception of nature is as much Intentionally "prepared" as the perception of art; it is only that, in the space of a painting, further complex "objects"—perceptual representations—are discernible "there" that are not "discernible in nature." Even there, the success of spontaneous recognition cannot be merely an artifact of "representational systems [or] pictorial purpose alone."56 There is no need to deny whatever biologi-

cally favored perceptual dispositions obtain. But, relative to the praxis of our predicative categories, "natural" resemblance is always initially inchoate and ultimately interpretive. Gombrich seems unwilling to admit the point. (Still, the biology of perception does make its contribution—which Goodman will not admit.)

Speaking of art and nature suggests a fixity that can never be found. Naturalism in painting is a matter of salience only. Pictorial realism tolerates considerable fuzziness regarding the "rules" of representation, which Gombrich somewhat ignores. But that tolerance, it must be said, is consensual, historicized: it neither converges toward objective resemblance nor establishes similarity by fiat. That is why the classic forms of realism and nominalism (regarding "universals") fail: there is no ulterior ground for extending predicates beyond consensual practice. Yet admitting that does not disallow "correction." Thus, in the matter of horizon line isocephaly (lining up the heads of human beings in the horizonal distance, as in the Masaccio and Masolino frescoes), familiarity with the isocephalic rule, together with fixing the vanishing point on the isocephalic line itself, enables us to tolerate (through our effective memory of a general practice) small but quite definite variations in the legibility of the compared sizes of particular figures: we are often aided more by opportunistic occlusion and the perception of intended spatial relations than by the requirements of strict scale. The "natural" is itself a culturally tolerated perceptual precipitate.

One could make an amusing puzzle of Velázquez's Las Meninas along these lines—by intruding a sense of trickery regarding isocephaly. Thus, the use of the dwarf and the midget at the right of the painting, more or less in the same isocephalic line as the Infanta, raises odd questions about the relative size of the figures represented—a matter that is complicated by the kneeling of one maid on the Infanta's right and the apparent breach of isocephaly by the tall maid on her left, as well as by the "witty" uncertainty of the size of the dog. A similar isocephalic puzzle appears in the line that links Velázquez himself, the mirror image of Philip and Maria Ana, the court officer in the doorway, and the other two adults at the right of the painting; for, now, the illusion is clearly an illusion, since the distance of the mirrored image baffles the isocephalic order if not first correctly interpreted.

In fact, a number of Velázquez's paintings seem to query the significance of creatures of anomalous size in realistic representations. They also query the bearing of the imputed size of representations of

representations (mirror images, other paintings) lined up with "actual" creatures (the painter and the Infanta) represented within one and the same visual image. Velázquez's own training, through Pacheco, makes the imputation entirely reasonable.57 These games favor our construing Veláizquez's painting as a representation of representing by painting (Foucault's point) rather than (or merely) as a representation of a represented scene, a would-be painting (as has been recently suggested by John Searle).58 It also confirms, therefore, the usefulness of the distinction between aboutness and representation, and it warns us of the deep informality, the contextedness, of interpretation itself. (I favor an analogy here between Foucault's treatment of Las Meninas and Barthes's treatment of Sarrasine— in terms, that is, of an adequate theory of interpretation.)

But we must not lose sight of the point that the biological grounding of natural perspective does not entail the unique validity of any particular perspectival canon. If, as Gombrich originally supposed (following Gibson), linear perspective and visual resemblance depend primarily or essentially on the geometry of light, then a uniquely correct canon might have been possible. But if, as Gombrich concedes to Goodman—and, in conceding, supersedes Goodman's conventional-ism—these phenomena depend on the regularities of phenomenal perception, then there is room enough for incommensurable modes of perspectival rendering that may be judged to be equally "correct" in naturalistic terms. Thus: van Eyck's perspective precedes Brunelles-chi's; but if, as Panofsky observes, the Arnolfini has a number of distinct vanishing points, it hardly follows that the Arnolfini is not perspectivally correct. One might even argue that ecological perception, even in paintings, may sometimes be more favorable to alternative vanishing points than to a fixed point.59 This goes against Gombrich's use of Gibson's thesis.

The puzzle about perspectival realism begins, of course, with Brunelleschi's famous mirror experiment.60 The Brunelleschi experiment depends on the (historically original) surprise that follows when—standing in an appropriately fixed place inside the middle door of the Santa Maria del Fiore, one peeps through a small hole located in the back of Brunelleschi's own painting (now lost) of the Santo Giovanni (the Florentine Baptistery) at the represented spot corresponding to one's actual place (one's eye)—one first sees what purports to be a per-spectivally accurate picture of the Baptistery reflected in a mirror and then, when the mirror is removed, one sees the Baptistery itself, which

seems not to have been altered or to be different from the painting in any essential particular. Brunelleschi's experiment, even admitting the limitations of fixed-point perception, certainly confirms the "naturalness" of the perspective he introduced; and certainly, the fixed point of the peephole should not be confused with the representationally fixed vanishing point of Masaccio's and Masolino's frescoes—influenced by Brunelleschi's instruction though they are.

Their work clearly shows the import of Brunelleschi's discovery for a moving or ecological perspective represented two-dimensionally at a pictorially fixed point. It goes without saying that pictorial representation rarely tries to capture the difference between focal and peripheral vision at a fixed point.61 Nevertheless, regarding Brunelleschi-like possibilities involving a moderately restricted ecological perspective in an ambient world, one can find a devilishly clever confirmation in René Magritte's Les Promenades d'Euclide. Magritte's device (which is an improvement on Brunelleschi's) helps to clarify the sense of natural similarity, without yielding to the extremes of conventionalism or transparency; and, perhaps by allusion, it helps to confirm the sense in which Velázquez's painting might be a witty commentary of some sort on Brunelleschi's original experiment rather than a merely realistic representation. How could we rule out such an option, particularly if we incline (however mildly) in the direction of the radical possibilities endorsed by Gadamer's hermeneutics and Foucault's genealogies? Fou-cault's interpretation of the Velázquez begins to look like the mate of Barthes's interpretation of Sarrasine, which, of course, is a welcome economy. For it means that we have now brought the theory of sensory perception sufficiently into line with the theory of predication; the puzzles of historicizing interpretation can now be seen to infect painting and literature (and history) in the same fundamental way, and the sciences and arts can be seen to function, predicatively, in exactly the same way.

In this same sense, it is the insistence on the causal sources of the photograph that threatens its potential inclusion in the artworld. For that insistence privileges a certain regulative model of "natural" perception and thereby excludes the Intentional and artifactual mode of functioning of the photograph's pictorial powers. To speak of art in these terms would then yield nothing but a fictive (or heuristic) internal accusative answering to a certain purely extrinsic interest in some (properly) "real"—some physical—natural, "commonplace" things.62

Chapter 5
Textuality and Intertextuality

I

In one of his characteristically straightforward demonstrations of the semiotics of literary texts, Michael Riffaterre inadvertently betrays the deep vulnerability of his grand account of interpretation. As theorists of what a rigorous interpretation is said to require, or allow, we are the beneficiaries. For Riffaterre offers, in a notably sustained practice of great skill, one of the most carefully reasoned attempts to isolate the bona fides of critical reading distinguished from every otherwise casual or misapplied interest in literature. This is not to oppose particular readings of particular poems or the validity of Riffaterre's interpretive practice rightly applied to certain kinds of poems. It is rather to challenge his sense of the very range and adequacy of the interpretive undertaking. He believes he can identify essential constraints on the logic of literary interpretation, in virtue of which the intrinsic norms of a correct reading of poetry may be defined. The simple candor and explicitness of his thesis demand a respectful assessment. The question arises, of course, as to whether he is right in what he specifically claims. But, more than that, we are pressed to ask ourselves whether it is likely that any attempt to disjoin the "worlds" of poetry and ordinary life could possibly be compelling—or even viable.

I am bound to say, matching Riffaterre's candor, that I take him to be unqualifiedly wrong in his theory, though often stunningly instructive in his actual practice. He is, I should say, one of the ablest champions of an interpretive practice diametrically opposed to the theory and

practice that inform Roland Barthes's S/Z, which I endorsed at the start of this account. In addition, both Riffaterre and Barthes have formed the rationales for their own interpretive discipline by way of an initial acceptance of the general lines of Francophone structuralism. Reviewing Riffaterre's theory, therefore, promises a certain appropriate closure; although readers accustomed to "analysis" in the Anglo-American sense may find the elaboration of the argument somewhat alien and even bizarre. (I find myself obliged to pit Barthes and Derrida against Riffaterre, and I actually favor their notions.)

I remind you of the deliberately "mixed" nature of the theorizing specimens I have been gathering, without strain, from French and German and English-language sources. My own taste in philosophical argument is "analytic" but my sense of the range of conceptual resources is definitely sympathetic to the "continental." I need your patience, therefore, in reconceiving what should be consulted in forming a genuinely pertinent overview of the logic of interpretation (and history). (I claim, you remember, that the essential puzzles of interpretation and history are almost entirely neglected in English-language philosophy and, for instance, in New Criticism and theories of history favorable to analytic philosophy. I am, I confess, fashioning an argument to enlarge our sense of the standard texts to be consulted at the same time I am committed to pressing my own particular claims.)

Riffaterre's theory and practice support one another. In fact, the self-imposed limitations of his practice would make no sense apart from his theory. But the practice of a good many other sorts of interpretive criticism cannot be reconciled with his theory, and the theory cannot rightly justify their exclusion. Furthermore, the strongest features of his own work do not actually require his strenuous claims; although, if they were abandoned, there would be no way of justifying the restrictions he chooses to place on his own practice. These are strains not unusual in Francophone structuralism, as one may confirm by a quick scan of the systematic studies of kinship, myth, and the preparation of food offered by Claude Lévi-Strauss.

The matter is of considerable importance, since Riffaterre's structuralism purports to specify the proper field of Intentionally pertinent referents for any objective form of poetic interpretation. On its assumption, the practice of both Barthes and Bloom, for instance, appear as unruly caricatures of the structuralist constraint; and the seemingly careful work of the New Critics and Romantic hermeneuts appear to risk all rigor by relying on interpretive evidence drawn from sources

beyond the pale of legitimate allusion. The result of pursuing this line of argument would be no more than a hothouse dispute, were it not for the fact that Riffaterre has formed one of the strongest, most persuasive conceptions we have of what we should mean by textuality and intertextuality in the context of literary interpretation. His concern is a question that cannot be neglected. (It may force us to meander a little, for instance in reflecting on the vagaries of fact and fiction; but, even in this, our speculations will prepare the ground for the issues of the next chapter. And in allowing a certain generosity of reference, we may persuade ourselves that we have indeed balanced our assessment of the principal lines of distinctly French and German and English-language speculation.)

Unfortunately, so I shall argue, Riffaterre's conception is profoundly problematic—on internal grounds. It is also, however, a specimen of one of two master theories we are bound to choose between. (I risk something by this frontal assault, of course, but conviction and consistency at this point in the argument, and the need for economy and closure, demand it.) Also, though I shall be speaking of poetry particularly, I mean to bear in mind as well the application of the notion of textuality to painting. I broached the question, in the preceding chapter, in terms of the puzzle of representationality, and I shall try to lead us—gently—in the direction of an inclusive finding.

The truth is, Riffaterre's theoretical model is not, as he sometimes supposes, common ground for structuralist and poststructuralist interpretive strategies (or, more loosely, modernist and poststructuralist options). His is rather an extreme modernist alternative that (I claim) cannot possibly be vindicated. There is no way to hide the fact that the view I favor is bound to collide with Riffaterre's proposal—and for important reasons. The charge is a brusque one, and you may well wonder whether the conceptual gains to be had justify such a beginning. The verdict rests with the sweep and economy of its intended harvest.

The opening paragraph of one of Riffaterre's most important papers contains all the pertinent clues:

If we try [he says] to arrive at the simplest and most universally valid definition of the representation of reality in literature, we may dispense with grammatical features such as verisimilitude or with genres such as realism, since these are not universal categories. Their applicability depends on historical circumstances or authorial intent. The most economic and general definition, however, must at least include the following two features. First,

any representation presupposes the existence of its object outside of the text and preexistence to it. . ..Second, the reader's response to the mimesis consists in a rationalization tending to verify and complete the mimesis and to expand on it in sensory terms (through visualizations, for instance). The metalanguage of criticism accordingly prolongs and continues the text's mimetic discourse, and critics evaluate representation in terms of its precisions and suggestive power. Both processes—presupposition and rationalization alike—assume that referentiality is the basic semantic mechanism of the literary mimesis.1

The meaning of this passage is not entirely obvious, though it does indeed capture the nerve of Riffaterre's thesis. Riffaterre goes on to draw our attention to a certain paradox in literature, namely, that "the represented object eschews referentiality yet refuses to vanish altogether, becoming instead the verbal vehicle of an interpretive activity that ends up by making the object subservient to the subject."2 This catches up an extremely fashionable but quite mistaken view that words and texts are not real, do not sustain reference in the same sense physical objects do; and that, therefore, the "reduced" sort of reference in literary criticism to words and texts and the intertextual analysis of texts are utterly unlike normal discourse about the natural world. Riffaterre then springs his own suggestions on us:

Critics fail to explain this paradox [he claims] because they stick to referentiality as the only law governing representation and assume that the reference on which the mimesis is based is from words to things, from the verbal to the nonverbal domain. I propose that reference in such cases is from words to words, or rather from texts to texts, and that intertextuality is the agent both of the mimesis and of the hermeneutic constructions on that mimesis.3

This is an important pronouncement. But the reason it is, quite frankly, is that it is subtly misguided and ultimately incoherent—and that its central mistake is just the one that has captured so many earnest efforts to penetrate the mystery of reference, the distinction between representation and reality, and the distinction between fiction and reality, all of which dominate literary and art theory at the present time.4

In opposing Riffaterre, I am claiming that this is just the wrong place to begin: both the distinction being pressed (reference to words or texts as distinct from reference to mere "things") and the kind of distinction being featured (the bare reality of referents) have almost nothing to do with explaining textuality and intertextuality or with

clarifying how interpretation functions intertextually. What I am flagging for closer scrutiny is Riffaterre's claim: (1) to have distinguished between the reality of the natural world and the "reality" of the cultural world (the difference between "things" and "words"); (2) to have exposed the inappropriateness of construing the reference to nature and the "reference" to poems as being of the same kind (the question of mimesis); and (3) to have clarified the necessary restriction of interpretive criticism to "hermeneutic constructions" that depend on other texts (the rationale of the intended structuralism, that rests on avoiding a supposed aporia regarding the mixing of kinds of reference).

You may notice in all this, therefore, a curious similarity to Danto's philosophical strategy regarding history and interpretation. (I shall return to Danto in chapter 7. We shall then be able to collect a strong but unsuspected analogy between structuralism and analytic philosophy.) What I am recommending, in advance of my analysis of Riffaterre's theory, is a robust insistence on one world embracing both physical nature and human culture, the joint dependence of discourse addressed to each on the symbiotized resources of our Lebensforrnen, and the inseparability of objective truth-claims from the historicized conditions of our own cognitive competence. I find these considerations completely ignored, or disallowed, by Riffaterre. Riffaterre's extravagance confirms, therefore, the indissoluble conceptual linkage between legitimating a particular interpretive practice and whatever theory of reality we are prepared to defend. In fact, the initial plausibility of Riffaterre's practice depends on the disjunction between the two sorts of reference he identifies: without that, we might well wonder just what was the justification for precluding reference to the ambient world of human history and culture not already regimented within some putative literary canon.

There is a genuine puzzle about the nature of reference, of course; but it is entirely an artifact of certain presumptions about how a compliant theory could be made to serve certain ambitious conceptual programs: for instance, how it might serve the analytic philosophies of the twentieth century that construe the semantics of truth in an entirely extensional way, or that mean to retire reference altogether in favor of so-called uniquely designating general predicates, or that link the success of referential acts to the actual existence of their purported referents. Such undertakings, we now see, are open to serious challenge for reasons entirely independent of Riffaterre's project—for example, because reference has proved ineliminable and because the Intentional features of fiction are peculiarly stubborn.

What is generally ignored in the current literature is that the same mistaken presumption (about reference) is invoked in the same way in modernist, postmodernist, and poststructuralist theories. They need not differ in their treatment of reference; and their common error (if it is an error) draws the very effort to theorize about interpretation away from its most promising conceptual pivot. (I have touched on the matter lightly in speaking of Quine and Krauss and Baudrillard.) The confusion I wish to pinpoint—which obscures the deeper and more important distinction of textuality—is that of denying that reference in literature can (and must) function in a perfectly standard way that extends uniformly to history and science. For my part, reference, whether in literature or science, is ineluctably mediated by our conceptual schemes and historical world. Reference functions uniformly in the world: Riffaterre denies this. Furthermore, as I argue, reference is inescapably context-bound and therefore Intentional: the leading arguments of analytic philosophy deny this. The two questions are linked and their resolution is of decided importance.

Consider a specimen view. Linda Hutcheon correctly summarizes the drift of the strongest philosophical accounts, observing: " Reference is not correspondence, after all. Can any linguistic reference be unmediated and direct?"5 (Plainly not, is the implication.) Nevertheless, by way of a complete non sequitur, Hutcheon moves from this relatively safe pronouncement to another that is more than doubtful (but appears to resemble what she has just affirmed): "What a novel like Rudy Wiebe's The Temptations of Big Bear suggests, by its very form as well as its content, is that what language refers to—any lan-guage—is a textualized and contextualized referent: the Big Bear we come to know is not really the Big Bear of actuality (for how can we know that today?) but the Big Bear of history texts, newspaper accounts, letters, official and unofficial reports, but also of imagination and legend. . ..The novel is both a referential inscription and an imaginative invention of a world."6 This won't do at all. For the minimal conclusion the analysis of cognitively informed reference can yield is simply that the actual world is "textualized and contextualized," not that it is (therefore) unreal or unknowably noumenal, or that literature (or any other use of language) must refer to a completely invented "world" utterly different from the real world.

Somehow, Hutcheon (Riffaterre as well) believes that "a textualized and contextualized referent" signifies a referent that is not a "real' referent (or one that is not real in the same sense in which physical objects are real: in particular, a referent that is not like anything that was

real in "actuality," that was actual in the past but is now no more). But if (as Hutcheon herself suggests) no "linguistic reference [is] unmediated and direct," then, on the argument she favors, we never succeed in referring to the real world at all. Physical nature is in no better position than literature.

There you have a sign of the serious confusion between referentiality and textuality. We do know the real world, but what we designate as the real world (that we know) we designate as such within the textualized (the symbiotized) world in which we live and make inquiry. The conclusion to be drawn is the same (so far at least) as the one Thomas Kuhn arrives at in the context of the physical sciences—that is, in the context of the history of their practice, from which, of course, the practice is inseparable.7 (I shall, in chapter 7, consider Donald Davidson's strong challenge to this view.) We claim to know the real world, but we know it only contextually. The Kantian-like noumenon that Hutcheon implicates is a vestige of past philosophy that no one can now afford, that is no longer instructive, that serves only to distort our understanding. Kuhn, I may add, is somewhat tempted to read his metaphor of the "different worlds" Priestley and Lavoisier inhabit, as literally distinct real worlds.8 He is initially tempted, therefore, to con-flare fiction and reality—which he ultimately resists. Two caveats are all he really needs, however: one, that the real world (what we posit as real, on the best evidence) is symbiotically constrained—therefore constructed; the other, that the interpretation of what is real is context-bound and historicized—therefore open to plural, possibly nonconverging characterizations.

Hutcheon implicitly adopts Riffaterre's distinction regarding the language of literature: "Literature [she says] does not mean by any process of external reference."9 This alone adversely affects her account of the difference between the historical novel and "historiographic metafiction."10 But she goes on as well to claim (untenably) that "real," "external" reference can be retired altogether. Postmodernist critics like Rosalind Krauss and poststructuralist critics like Jean Baudrillard are just as much inclined (as we have seen) to resist referentiality and the straightforward distinction between fiction and reality as are the fashionable "late" modernists;11 and philosophers like Richard Rorty, Michel Foucault, and Jean-François Lyotard are either unhelpful on the question or quite willing to court incoherence as well.12 There is no way to show—I claim—that reference to physical objects and reference to texts employ the notion of reference in two different senses; or

that if we admit we can refer to fiction, we must then deny the familiar difference between fiction and reality.

I take Riffaterre to be mistaken, therefore, in supposing that reference is the essential clue to the theory of interpretation. It is not reference (or referentiality) but textuality (or, as current philosophical jargon has it, cognitive "intransparency," or the "constructed" nature of the intelligible world) that is decisive. The reason is quite straightforward: insoluble paradox results from encumbering the grammatical resources of reference and predication with the metaphysical differences between fiction and reality; for then, the textual complexities of interpretable literature are too easily confounded with the artificial difficulties produced by the first maneuver.

Riffaterre supposes that reference and predication are in some essential way ontologically freighted.13 Reference is committed (he believes) to the independent existence or reality of its referents, and those can be of certain ontological sorts only. True reference precludes literary texts. Riffaterre concludes that insistence on "referentiality" in literature is due to a misguided adherence to the "fallacy of reference" which, in effect, replaces texts by things that lack textual structure. He favors instead a diminished form of reference in poetry, since what is being "referred" to, in interpreting poems, is what continues to be open to mimetic "constructions" drawn from a pool of similarly constituted "referents" (texts). Perhaps he believes that the imputation of these hermeneutic accretions is incompatible with the antecedently determinate nature of the "things" that we make true reference to. If so, then, as I have already argued, the caution is pointless: as we have seen, successful reference does not require the fixity of the nature of its would-be referents. The deeper lesson of this quarrel, however, lies with the ease with which, on Riffaterre's argument, it then seems plausible to confine legitimate interpretation to some constructed space of literary "texts"—that are themselves said to be segregated in a principled way (by way of the supposed paradox of referentiality) from the unmanageable contingencies of an evolving historical experience. But that would mean that interpretation in history and the arts was, necessarily, of fundamentally different kinds.

Actually, Riffaterre's thesis is more extreme than Umberto Eco's, which defines the would-be "fallacy." Eco holds only that "the referential fallacy consists in assuming that the 'meaning' of a sign-vehicle has something to do with its corresponding [real] object."14 What Eco means is: first, that significant discourse may obtain despite the

absence of "real" referents (as in fiction); second, that what a "sign-vehicle" means cannot always depend on the real referents that are its "corresponding object"; and, third, that "referents" within the space of a "theory of codes" (suited, say, to interpretation) need not be treated (or need not be able to be treated) "extensionally" (as either "actual" or "possible" in the real order of physical nature) but only as "convey [ing ] a cultural content "15 —by which expression Eco intends a somewhat dubious Fregean adjustment. Riffaterre counters that "the mimesis proper [in literary representation] refers not to reference but to elementary representations of these. The mimetic text is not composed of words referring to things but of words referring to systems of signs that are ready-made textual units."16

The reason Riffaterre cannot possibly be right is this: on his own theory, there are "ready-made textual units"; they are real enough, as real as any physical referents. Hence, reference to them necessarily violates the very constraint Riffaterre imposes. (The same difficulty infects Eco's theory, just where Eco speaks so beguilingly of individuated "cultural content.") The fact that "texts" have features unlike those of physical things does not as such affect the further fact that we do make reference to this poem and that. The world does not admit of grades of reality or compartmentalized sectors of reality. Texts belong to the world in the same way trees do, but they are not the same kind of "thing."

What is missed in both accounts is this: (1) reference, being only grammatical, conveys no ontological import at all (contrary to the strong tradition of analytic philosophy both Eco and Riffaterre pay homage to); (2) there is no need to resist mere referentiality, in resisting anything like the "axiom of existence" said to be entailed by reference;17 and (3) the intertextual nature of literary interpretation, which Riffaterre espouses, must be disjoined (to be instructive) from any encumbrances drawn from the mere logic of reference—a fortiori, disjoined from whatever can be made out to be the disastrous consequences of subscribing to the putative "fallacy of reference." I shall return to this complex matter in a moment. But I hurry to take advantage of a muddled quarrel, to draw the simple conclusion wanted: nothing follows regarding textuality or intertextuality from the analysis of mere reference; nothing follows from it that would privilege particular theories of interpretation—neither

Riffaterre's nor any other. The entire maneuver is an extravagance.

II

There is another preliminary distinction that may be drawn from Riffaterre's account that points us in a more promising direction. Part of what Riffaterre says is certainly right—but for the wrong reasons. "The literary representation of reality, then, for all its objectifying stance, is [he says] essentially an interpretive discourse."18 What is wrong is due to the intrusion of the connective "then." For what Riffaterre means to say is that literary representation is interpretive because (and only because) it is not referential (or, alternatively, because it is "referential" in that severely reduced sense in which, in the passage cited, the "represented object eschews referentiality but refuses to vanish altogether"). It is not, he thinks, objective or "objectifying" in the strict sense. It does not implicate the "objective" (the real) world; interpretive practice about its "represented object" is, however, capable of being "objective" (rigorously correct: "objectivist") in its own correctly segregated way. (The equivocation is deliberate.)

Once again it is the conflating of referentiality and cognitive transparency that skews Riffaterre's notion of interpretation. On the corrected view (emphasizing intransparency)—which, by the way, is common ground for postmodernism and poststructuralism (against modernism)—verbal representation, whether in science, history, or literature, is inherently interpretive. Riffaterre believes that classical mimesis is not interpretive, because it and its corresponding mode of criticism are committed to true "referentiality"; whereas literary mimesis is interpretive—hence, contrary to the classical view, it is not referential at all. Strictly speaking, referential and interpretive acts exclude one another. (This helps to explain the sense in which Riffaterre's project is structuralist and hermeneutic, not Aristotelian.)

As soon as you grasp the argument, you cannot fail to appreciate the difference between two utterly dissimilar notions of intertextuality: between, say, Riffaterre's notion and Roland Barthes's or Jacques Derrida's. Also, pursuing that difference, you cannot fail to see that it is the latter, not Riffaterre's, that captures the larger conceptual tendencies of late twentieth-century theories: such diverse conservative and alien voices, for instance, as Hans-Georg Gadamer's and Barthes's and Derrida's (and, of course, Michel Foucault's, Jean-Francois Lyotard's, Luce Irigaray's, Harold Bloom's, Paul de Man's, Geoffrey Hartman's).

Riffaterre wrongly assumes a measure of convergence holding

between his theory and interpretive practice and those of certain of these others; whereas the truth is that each subscribes to a different conception of textuality, and each favors an entirely different account of interpretation. In particular, Riffaterre is an "objectivist" about interpretation and none of the others are. Nevertheless, his objectivism entails disjoining the interpretation of texts and the interpretation of the (texted) world—which the interpretive practices of the others mentioned do not. Of course, in saying all this, I am setting the stage for what is important to say about textuality and intertextuality. I am clearing away some conceptual debris that threatens to make an insoluble mystery of interpretation—that cannot even be seen to be debris without these preliminary reflections.

III

Riffaterre regards literary criticism as interpretive only when these two features are present together: one, when "reference" is made to "ready-made textual units"; the other, when the mimetic function of literature is restricted to the representation of such "ready-made textual units." The "world" of literary reference is interpretive because it is "textual." But as soon as that much is conceded, Riffaterre would have us believe that interpretive criticism is also "intertextual," that is, concerned one way or another to interpret the specific textual signs of a poem in terms (only ) of the entire set of (what he designates as) a culture's (already stored) system of similar textual signs (that is, other poems). The result is literary representation in the proper sense. No provision is made, therefore, for "textual" interpretations along Marxist or Frankfurt Critical or Freudian or feminist or similar lines, that is, for criticism that goes beyond the merely "verbal" in order to locate the culturally significant import of literature in the evolving habits of actual life, in the traditions and practices of a historical culture. Extraordinary!

There you have the ulterior purpose of Riffaterre's account of "referentiality": it restricts the proper range of "textuality." There is all the difference in the world, therefore, between Riffaterre's (structuralist) and Gadamer's (hermeneutic) conceptions of interpretation. For, for Gadamer, literary texts are indissolubly part of the same real tradition (or cultural world) to which human beings belong: to interpret the one is to interpret (or reinterpret) the other. They cannot be disjoined.

Riffaterre offers no explanation for the curious limitation he favors, except for the arbitrary claim that, in literature, "words" refer only to "words." Why this should or must be so, he never says. Our entire linguistic practice is against it. Even the Intentionality of texts functions as such only within the space of the consensual forms of life of a particular society. On that reading, there cannot be a valid disjunction between literary texts and the ambient cultural world. Structuralism cannot countenance the admission, but it offers no convincing rationale in its place.

Now, there is no question Riffaterre's interpretive practice is sometimes extraordinarily skillful. It also fails at times. For example, at a Modern Language Association symposium in New York in 1987, Riffaterre attempted to demonstrate the force of his own method by applying it to W. H. Auden's "Musée des Beaux Arts."19 The choice proved a complete disaster. No one could have been more clever in choosing a decisive counterinstance for his own method. The choice was his own. But if he had not proposed it, it would eventually have surfaced, with the same effect and the same lesson. (This is what I hinted at at the beginning of this chapter.)

The trouble with the Auden case is simply that Riffaterre's sort of intertextuality is largely irrelevant to the intertextuality of the poem: we cannot possibly gain any idea of how to interpret it by mastering Riffaterre's strategies. Riffaterre plainly neglects the antecedent question of just what sort of intertextuality the poem may be said to exhibit. The wealth and precision of the semiotic details of the ancient treatment (in existing poems) of the Icarus myth, for instance, cannot possibly be central to the chatty gauge of the language of the "Musée" poem. The issue is not simply the interpretation of the "Musée" poem; it's a matter of the rationale for precluding from interpretation the resources of the ambient culture that is not already compressed into the functionally closed system of poetic "texts" Riffaterre is willing to allow. There's the significance of his notion of textual "reference."

In a formal sense, the issue is very much like the one that looms so disastrously over every automatic attempt to extend Erwin Panofsky's iconographical and iconological procedures to the analysis of contemporary paintings, that is, to paintings that are no longer produced in the strongly traditionalized world in which Panofsky's methods appear at their best. (Riffaterre's and Panofsky's methods are entirely different, of course. But both raise questions about disjoining, for the sake of rigorously correct interpretation, "an artworld" and the "real

world.") The ambiguously casual reference, in the Auden poem, to the Breughel Fall of Icarus, however apt it may be—seemingly chanced upon in drifting through the museum, settling in a knowing way on this or that painting as we stroll (that is, the speaker and his companion, in the poem)—offers no plausible reason for abandoning the salient conversational tone (of the poem) for the sake of the Virgilian "homologues" Riffaterre insists upon. There is, I admit, an obvious structural similarity between the marginalizing of Icarus' fall in the painting and the reference in the poem; but that is another matter altogether.

A weakness similar to Riffaterre's, relating to Panofsky's altogether different interpretive practice, was demonstrated by Meyer Schapiro, many years ago (in the late 1940s or early 1950s), in a famous debate (with Panofsky) regarding Panofsky's methods, at the Metropolitan Museum of Art: Schapiro was able to show very easily that a Mondrian abstraction—a composition in diamond form that lacked iconographic meaning, that was not representational in the required way—nevertheless did still manifest semiotic features, features that could not be analyzed in Panofsky's way.20 (Again, Schapiro's counterinstance could have been invented, to the same effect, if the debate had never taken place.)

Auden's poem is certainly textual, certainly intertextually rich, certainly representational, certainly ironic, certainly self-referential, and certainly who knows what else; but it is not intertextual in Riffaterre's sense. We cannot force these matters. There is no reason why we should feel the least bit obliged to proceed (as interpreters) in the way Riffaterre recommends. For, first of all, Riffaterre decries favoring merely contingent, historically tethered interpretive moves—genre studies, for example. He favors instead interpretive strategies that (as he supposes) depend on the "universally valid" aspects of the representation of reality—protected, it should be noted, from the uncensored intrusion of immediate historical reality. Second, as we have seen, critical interpretation (for Riffaterre) is governed by the textuality of the "world" we experience in literature—a world restricted to "ready-made textual units." Third and most important, the interpretation of the textual world is said to be necessarily intertextual— in the special sense Riffaterre intends, which (we may assume) is supposed to catch up the universally valid features of any understanding of literary representation. On the argument being mounted, Riffaterre has defeated himself out of his own mouth (the Auden poem). In any event, Riffaterre's pre-

sumptions cannot be more than a structuralist's prejudice. The very relevance of his own critical practice, where it is indeed successful, depends not (as he supposes) on any universally inclusive, totalized "world" of "ready-made textual units" to which interpretation may be confined but rather on the contextually pertinent Intentional exploration of some sector of the encompassing cultural world in which his own favored explications gain credibility. There is no universal, formally isolable "world" of "textual units."21 The very relevance of his semiotic practice is a function of the interpretive habits and histori-cized horizon of his own literate society. That is what he denies—which makes both his practice and his refusal to countenance certain alternative hermeneutic practices appear arbitrary.

Here is what Riffaterre says (it will pay to have the entire passage before us):

An intertext is a corpus of texts, textual fragments, or textlike segments of the sociolect [the repository of social myths] that shares a lexicon and, to a lesser extent, a syntax with the text we are reading (directly or indirectly) in the form of synonyms or, even conversely, in the form of antonyms. In addition, each member of this corpus [of texts] is a structural homologue of the text: the depiction of a stormy night may serve as an intertext for a tableau of a peaceful day; crossing the trackless sands of the desert may be the intertext of furrowing the briny deep.

In contrast, intertextuality is not just a perception of homologues or the cultivated reader's apprehension of sameness or difference. Intertextuality is not a felicitous surplus, the privilege of a good memory, or a classical education. The term indeed refers to an operation of the reader's mind, but it is an obligatory one, necessary to any textual decoding. Intertextuality necessarily complements our experience of textuality. It is the perception that our reading of the text cannot be complete or satisfactory without going through the intertext, that the text does not signify unless as a function of a complementary or contradictory intertextual homologue. In a given poem, under certain verbal conditions, a peaceful day will make sense—literary sense—only as the contrary of a stormy night, in opposition to countless depictions where day and peace are represented without eliciting such a dual perception, and without our feeling a need for it.22

The telltale features of the theory are these: first, there is a definite structuralized way of entering the intertextual world interpretively—by way of a binary use of synonymous or antonymous homologues for a given text, drawn from the determinable world of a given culture's "ready-made textual units"; second, that way of proceeding is inviolably required of any responsible literary interpretation. But it doesn't work in the Auden poem! The intertextual features of the "Musée des

Beaux Arts" confirm the ease with which Riffaterre's thesis is outflanked, all the while we honor his practice (wherever it succeeds). All we need do is: (1) construe interpretive strategies as contextually contingent, historicized, incapable of the universalized legitimacy Riffaterre insists on; and then (2) admit that it is not only possible but very likely that the invention of new sorts of poems cannot but subtend (in a generous sense) the invention of correspondingly new forms of interpretive practice.

Surely, the denial of (1) and (2) would be counterintuitive at this point in late twentieth-century thought. Their affirmation would also be disastrous for any canonical structuralism or near-structuralism; for, since structuralism is committed to the doctrine of internal relations, that is, treats signifiers only relationally, by virtue only of their place in a total system of relata, any enlargement of the would-be system would disorder all the relations already collected within it. The theme is a favorite of Derrida's of course.

In fact, the notion of textuality (or intransparency) developed by theorists like Barthes and Derrida and Foucault undercuts the very possibility of invariant, necessary, totalized, transhistorical methodologies of any kind. So there must be at least two entirely different notions of textuality and intertextuality at stake here.23 The difference is this: Riffaterre's structuralism views the textual world as at once constructed, real qua "mimetic," and totalized: on that account, modernism need not be incompatible with a form of social constructionism. Barthes's and Derrida's constructionism, by contrast, is inherently incompletable, not fixed in Riffaterre's sense of a synchronic realism, and not perceptually legible in the sense of Riffaterre's deliberately self-isolating objectivism. Their vision is distinctly poststructuralist.

Still, the Auden example challenges Riffaterre's structuralist hermeneutics without calling on any absent alternative. For the "Musée" is a poem that palpably resists Riffaterre's strategy. The intertextuality of Auden's poem, which is real enough, is simply inhospitable to its exclusive use. But, you will ask: what other option is there?

The answer stares you in the face. We might admit the openended textualized thought and behavior of a historically changing, improvisationally apt culture. Why not? We could then draw on all the resources admitted in Wittgensteinian, Marxist, Freudian, Nietzschean, Heideggerean, Foucauldian, Gadamerian, Frankfurt-Critical, Lacanian, feminist, and similar accounts. We should then not find ourselves confined to "ready-made textual items" (homologues) or restricted to

textual choices within the space of an already stored bank of (inter-textual) alternatives. Harold Bloom's practice, for one, can be neither accommodated nor convincingly disallowed by Riffaterre's model. Certainly it could never be reconciled with it.24 Riffaterre's limitation appears to be arbitrarily imposed.25

Even if we admitted narrowly "textual items" of the sort Riffaterre invokes, we could, on wider intertextual grounds, always defend interpretations not confined to such "ready-made" items (that is, to texts already completed and deposited in the historical past); we might even produce new texts (as critics or incipient poets) that embodied the ongoing inventive work of the present, bent on interpreting the evolving historical significance of what it supposed was the meaning of its own past texts.26 Gadamer's Horizontverschmelzung, for instance, productively eludes Riffaterre's limitation: the interpretation of a text, you remember, is, for Gadamer, always a form of self-interpretation.27 In the same vein, the interpretive practice of a critic like Walter Benjamin could not possibly be constrained by Riffaterre's rubric; but that is no reason for disallowing it. Also, of course, to admit these alternatives is to admit the possibility of altering the interpretable, the historical, past—a possibility that goes beyond merely adding to the interpretation of a fixed past. (I return to this issue in chapter 7.) The diachronic nature of interpretation cannot ensure that it makes sense to suppose that it itself obtains only within a synchronically totalized semiotic world.

In a word, there is no demonstrably compelling reason to construe interpretive criticism as merely "heterotelic" (in the sense T. S. Eliot once introduced and then abandoned): there is no principled distinction between poet and critic, though saying so generates a great many conceptual puzzles. Also, there is no compelling reason for thinking of criticism as addressed to words, texts, or otherwise antecedently isolable "textual items" as opposed to the speech acts or even more complex actions or practices enacted or instantiated by the culturally apt agents of this or that society—through which "ready-made textual items" count only as a sort of static approximation, somewhat in the same way in which a line (in Bergson's sense) impoverishes (though not without relevance) the process of the living movement it represents.

Riffaterre has the gravest doubts about the ability of literary texts to make direct reference to the real world; or, correspondingly, the ability of intertextual interpretation to make reference, in making sense of a text, to the intertextuality of the real world in which poets and

critics (finally) live (together). But we need not share his doubts. On an argument already supplied, the entire intelligible world may be said to be "texted," in being symbiotized.

In any case, Riffaterre certainly does not deny that ordinary discourse is used to make reference to the "commonplace" world: he imposes an ontological disjunction on literary interpretation only "after the fact." But if, as I have been arguing, language is not an isolated or autonomous aptitude among humans, if the linguistic is, everywhere, an abstraction from ampler forms of intelligent action and intervention in the real world, then it is unreasonable either to disjoin literary reference from the forms of reference in the commonplace world or to disallow the historically evolving resources of meaning drawn from that world to enrich the interpretive possibilities of literature itself. That is what the "Musée" poem signifies.28 That is what I have extracted from Barthes's and Gadamer's practice. The linguistic, I should say, is already lingual; consequently, both reference and interpretation may be lingual and not merely linguistic in the narrow sense. (This is often ignored.)

IV

It would be useful to have before us the two notions of textuality I have been hinting at. The most notorious formulations of the missing alternative are, of course, those crafted by Derrida and Barthes. "Il n'y a pas de hors-texte" —Derrida's formula—does not mean that everything is part of a supreme fiction, or that fiction and reality are indistinguishable.29 Derrida does not say: "Il n'y a [rien en] dehors [de] texte." Some readers profess to see his remark as simply ungrammatical. He plainly says: "Il n'y a pas de bors.texte. " His meaning is not that there is a "texte" that nothing escapes—a supreme fiction perhaps, in the manner more obviously favored by Paul de Man; or, that there are nothing but texts. It is rather that there is nothing real that is not "texted," constructed, symbiotized, contexted—interminably. As soon as the second reading is construed in Wittgenstein's or Gadamer's or some related sense, one cannot fail to see Riffaterre's arbitrariness. (Needless to say, the extension to the lebensformlich is not Derrida's game; for, for Derrida, the hermeneutic and the phenomenological are as open to deconstruction as is the structuralist.30 The "hors-texte" line is as close as deconstruction could possibly come to Kantian-like transcendental conjectures.)

It would not be necessary to press this warning, except for the fact

that a literary theorist and critic of Paul de Man's stature, who is quite explicit about his admiration for Derrida's deconstructions (who is himself said to be a deconstructive critic), has managed to work extremely hard (one may say) at confusing the difference between fiction and reality, as if doing that were somehow to supply a corollary of the formula just given. De Man's reason for drawing the corollary is not merely to muddle the difference between fiction and reality (which he deliberately does). On the logical side, that alone would be a disaster for any pretense at preserving the notions of validity and truth. But he also rather slyly presents his referential confusion as if it were tantamount to, or subtended by, the radical intransparency (or "textuality") of the intelligible world according to the deconstructionists.

The corrective is easy enough to state but difficult to appreciate: de-construction has nothing to say, distributively, about either reference or truth-claims: truth-claims obtain in whatever way they do—blindly, without any possibility of "external" correspondence, confined within the unstructured space within which assertions are tested and supported. (Derrida most certainly acknowledges ordinary texts and ordinary reference, though many readers doubt that that is so.) Deconstruction has no objective or objectivist pretensions at all; it affords no more than an undifferentiated global sense of the impossibility of ever fixing reference in a way that is free of our conceptual schemes, or of ever fixing our conceptual schemes with respect to all possible such schemes so that we can be certain of precisely when discourse provides an invariant or closed system and when it does not. Unfortunately, many have been happy to adopt de Man's sort of gloss: the sheer "rigor" of deconstruction has suffered accordingly.

In point of fact, de Man's view is, in a way, motivated in much the same spirit as Riffaterre's, though it is of course as different as it could be from Riffaterre's Cartesian practice. Both explicate textuality by conflating it with referentiality. In Riffaterre's case, a clean bifurcation of referents leads to an intertextuality of homologous "ready-made textual items"; in de Man's case, a deliberate confusion of real and fictional referents permits a strong preference for unresolved paradox even where a seemingly scrupulous attention to actual literary texts adds a certain disingenuous rigor to interpretive intertextuality. (One might in fact argue that de Man is more Heideggerean than Derridean, more hospitable to Gelassenheit regarding Sein than to the denial of any "hors-texte.")

It is extraordinary that Derrida's formula, which is clearly a pronouncement about what we should mean by textuality, is offered in a

chapter titled ". . . That Dangerous Supplement . . ." and developed in terms of the question of interpretive closure and of the "referent" or "signified" of any piece of discourse. It could easily suggest that there really is a strongly convergent perception shared by Derrida and Riffaterre. But that would be a mistake, because Derrida means to complicate the "relationship" between signifiers and signifieds (by way of the disruptive function of infinitely many supplements at every interpretive point, that cannot be assigned to any aggregate of "ready-made" past "textual items"), and to disallow as well any "referent" external to the texted worlds subtended by the pronouncement just given (the so-called "transcendental signified"). Both of these maneuvers are plainly inimical to Riffaterre's intended practice.

I have, I may say, permitted the discussion to shift in a natural way from Riffaterre's original thesis. I see a danger there—not so much in finally shaping the larger argument but rather in drawing readers on in the short run, through unfamiliar territory, and in holding their attention. I began with the supposed conceptual link between referentiality and textuality, moved on very briefly to the distinction between fact (or actual history) and fiction, and have now returned to two competing views of textuality. But I know how fatiguing Derrida can be, and how unsympathetic many (particularly analytically minded) readers would find an extended discussion of Derrida and de Man to be. The best I can say is that the discussion is needed, will be no longer than needed, and may even serve to expose a groundless prejudice or two regarding Derrida's philosophical acumen. Derrida is often right, I find, on the essential issues; but I confess he is often tiresome. Still, there is no sustained discussion of textuality and intertextuality in the analytic literature.

To put the matter trimly: in Riffaterre, textuality permits—nay, insists on—an initial, fully articulated disjunction between signifiers and signifieds (which are, of course, "textual items"); in Derrida, who is explicitly opposed to any structuralist confidence, textuality disallows any antecedent disjunction between signifiers and signifieds. (They are no more than contingent artifacts developed, for this or that particular convenience, within the global space of textuality itself.)

Riffaterre is a true structuralist. He accepts the initial, steady, reliable, objective binarism of signifier and signified as being never seriously at risk in the flux of history and cultural improvisation. He dismisses authorial intent and genres, it is true: he is, after all, an opponent of Romantic hermeneutics. But he denies any and all conces-

sions to historical transience. So he believes it is possible (against Gadamer, say, who also opposes the Romantic interpreter) to reclaim a certain hermeneutic objectivity. All that is required apparently is to graft interpretive practice onto the referential fixities his own structuralism permits.

Derrida rejects such "logocentrisms." Textuality is not (as it is for Riffaterre) the local predicative color of whatever is suited to interpretation. It is also not, in any prior sense, the legibly ordered space of interpretation. For Riffaterre, intertextuality signifies the determinable scope and conditions of success of objective interpretation regarding textual things. For Derrida, intertextuality signifies, at best, the pragmatically objectifying plural practices of interpretation—improvised within an encompassing textuality that has no determinable structure or boundaries of its own. For Riffaterre, interpretive objectivity applies to what is antecedently fixed in the world: texts in particular. For Derrida, the intelligible world is itself an artifact, somehow, of an originating "posit" that we can neither fathom nor test; or, in whatever way we do test our interpretations of would-be determinate referents, we and they are already complicitously (and blindly) regularized for such testing. For Riffaterre, what lies beyond the intertextual world of texts is the nontextualized, referentially accessible real world; for Derrida, what lies beyond the textualized world is nothing at all (the mythical hors-texte), what (unguardedly) are supposed to be noumena (as in Hutcheon's account, as we have seen).

For Riffaterre, objective interpretation is the benefit of a rigorous intertextual methodology applied to a narrowly textual world (disjunctively contrasted with the natural world). For Derrida, objectifying interpretation is simply the provisionally successful stability assigned (without "totalized" legitimation) to any fragment of our functioning within any larger fragment of what we take the world to be. There is no actual, determinable, prior structure in reality that we can claim our interpretations match or correspond to; but every binary matching (as in Riffaterre or Lévi-Strauss) can be tested by creatures like ourselves who are already wedded to the practice, being ourselves (somehow) in-discernibly preformed for that, preformed within a "world" similarly preformed to receive our efforts hospitably.

In a word: for Riffaterre, intertextuality is the course of objective interpretation regarding antecedently discerned textual "referents"; for Derrida, intertextuality is the course of interpretive discourse (effectively: all discourse) formed, applied, tested, and judged to be

objectively valid (if we care) within the apparently fathomable space of a world that depends, unfathomably, on the ultimate intransparency of an (alleged) originary link between word and brute world. (Even this way of putting matters is a distortion, a way of effing the ineffable. But if we must speak, there is no other course.) Certainly, in the Derridean world, the historical past cannot be any more assured than the assured fixity with which a determinate "signified" can be matched to a putative "signifier." There's a lesson there that we shall have to reclaim.

Derrida actually says (quite succinctly) all that I am attributing to him here, though it is more fashionable to suppose he is really an intellectual anarchist, a spoiler, a purveyor of the most hopeless paradoxes. "If it seems to us in principle," he says quite mildly, "impossible to separate, through interpretation or commentary, the signified from the signifier, and thus to destroy writing by the writing that is yet reading, we nevertheless believe that this impossibility is historically articulated: It does not limit attempts at deciphering in the same way, to the same degree, and according to the same rules."31

Nothing could be plainer: all referential truth-claiming discourse is constrained in a way that cannot "itself" support reference or truth-claims (by way of an adequate theory of reference or truth); but that "constraint" can never be rightly construed as setting any determinate limit or barrier or referential, truth-claiming, or interpretive practices. Those practices go on as they must: the constraint is "entirely negative."32 The "reading" Derrida proposes cannot (as it does, with Riffaterre)

consist [for example] of reproducing by the effaced and respectful doubling of commentary [that is, by way of correspondence], the conscious, voluntary, intentional relationship that the writer institutes in his exchange with the history to which he belongs thanks to the element of language. This movement of doubling commentary should no doubt have its place in a critical reading. To recognize and respect all its classical exigencies is not easy and requires all the instruments of traditional criticism. Without this recognition and this respect, critical production would risk developing in any direction at all and authorize itself to say almost anything. But this indispensable guardrail has always only protected, it has never opened,a reading.33

I cannot see how this passage can be read (particularly the cautionary remark) without implicating Barthes's reflection on the conceptual linkage between "readerly" and "writerly" reading. On Derrida's view,

"il n'y a pas de hors-texte" signifies, by a sort of radical negation, "the absence of the referent or the transcendental signified" as a criterial guide for any would-be objectivist ("logocentric") reading:34 "Our reading must be intrinsic and remain within the text. This is why, in spite of certain appearances, the locating of the word supplement is here not at all psychoanalytical, if by that we understand an interpretation that takes us outside the writing toward a psychobiographical signified, or even toward a general psychological structure that could rightly be separated from the signifier."35

Textuality, for Derrida, is not objectivist in the least; it cannot provide any criteria of objectivity at all though it does not disallow the quest for its viable conditions. On the contrary, every effort to arrest the interpretable world for the sake of a regularized practice is doomed: first, by the irrecoverability of the "originary" surd on which all discourse ultimately depends if it is to be characterized in terms of correspondence; and then, by the uncertainty with which one and the same referent may be fixed by the use of conceptual schemes that are forever changing, forever imperceptibly evolving, forever supplemented by new distinctions, forever impenetrably affected by the course of discourse itself, forever incomplete and incapable of capturing any all-inclusive conceptual scheme within which any particular fragmentary practice (our own, say) can find its reliable niche. But Derrida's notion of textuality and intertextuality hardly disallows stable practices of interpretation—or even the legitimacy of claims to have entrenched an objective practice. In a sense, it actually makes room for Riffaterre's practice (as for Barthes's), provided only (which is a lot, and which of course Riffaterre would resist) the textual referents Riffaterre introduces are, now, only pragmatic artifacts posited from within the unfathomable textuality Derrida posits, and provided only the appeal to textual homologues is no more than a contingent practice—certainly not one grounded before and independently of raising particular inter-textual questions.

V

The foregoing may be summarized fairly neatly. On the one view of textuality, we discern an objective disjunction between interpretable and interpreting texts: they are and must be discernible prior to interpretation—which is intertextual only in the sense that, one way or another, we bring "textual items" from other parts of an encompassing

culture to bear (in an appropriate way) on the determinate text we mean to interpret. Riffaterre's well-defined practice is an example of an indefinitely large family of objectivist practices that include such utterly opposed specimens as E. D. Hirsch's quite conservative Romantic her-meneutics.36 On the other view of textuality, the intelligible world is itself textual, texted, contexted, suffused through and through with the structuring concepts of human understanding. That is not, however, a merely Kantian or Husserlian concession, because the duality of cognized world and cognizing agent (requisite for discourse itself) is never more than a "late," largely tacit artifactual distinction hostage to the deeper (impenetrable) symbiosis Derrida posits (the theme marked by différance ) that we may believe, at the point at which we admit the duality, transiently infects whatever we take to be the universal conditions of objective discourse and the supposed structure of that duality. (Derrida's treatment of textuality, therefore, is a deconstruction of Kant as well. Nevertheless, it is also very close to Kant in the post-Kantian way.)

Riffaterre makes his point by way of an objectivist account of the referential division between signifier and signified. Derrida makes his point by way of the mythic intransparency of all we imagine "there is," which he formulates in terms of the symbiotic inseparability of signifier and signified. By this he draws attention to the hopelessness of relying on any strategy of binary reference not already interpretively implicated in the "originary" intransparency of the world. (I am casting the opposition between the two in as many ways as possible, because of the notorious difficulty of Derrida's prose. But the intended lesson is surprisingly straightforward.)

It is hard to see how much is at stake in these two options. The full consequences have yet to dawn on us. But the essential clue is this: on the first view, determinately many "textual items" are already antecedently in place for interpretive (intertextual) work relating one to the other; on the second, there are no such prior "textual items," but (intertextual) interpretation, proceeding opportunistically, posits the individuated texts to be interpreted and the potential options of their interpretation by way of individuating other "items" within a pertinent cultural milieu. The first requires the fixity of interpretable individual texts; the second "constructs" interpretable texts de novo as it constructs their interpretations. The first claims an objectivism grounded in the reality of a cultural world largely independent of our individual interventions, however it may have been "constructed." The second

treats objectivity as an artifact posited from within an ultimately blind textuality. The continuity of reference and predication and orderly truth-claims is not disallowed on either view. The first is committed to an ideally valid interpretation of given texts relativized to the culture in which such work is pursued; the second is committed to an indefinite plurality of viable openended constructions at any point in the reflexive history of interpretation, by which in turn the enveloping culture is forever being reconstructed.

The first presupposes the fixed priority of description over interpretation; the second relativizes that distinction—temporally. The first is committed to the regulative ideal of a fixed system adequate for interpreting all pertinent texts; actual interpretive schemes, enlarged by a continuing practice, are then construed as progressive, fallible approximations to the totalized system intended. The second disallows the presumption of any such ideal ab initio; conceptual schemes are the artifacts of a continuing interpretive practice obliged to rely on fragmentary, local, transient, incompletely rationalized experience, subject to abrupt, diverse, potentially conflicting and overlapping lines of categorization. The first eliminates context and history; the second radicalizes the indeterminacy and flux of history. The first is structuralist and the second, antistructuralist—poststructuralist. They cannot be reconciled. But the interpretive practice legitimated in the first can find a diminished support in the second. There is, as remarked, a strong analogy here with Barthes's contrast between the resources of "readerly" and "writerly" reading. (The second subverts the modal claims of the first.)

The first notion of textuality does not admit that the "textual items" of interpretation are textual in accord with the second notion. The second notion disallows any determinate structures, including the individuation of textual referents, that are not themselves textual in its own sense. But it does not, for that reason, disallow the discursive division of interpretable and interpreting texts. They may be conceded, provided one makes no claim to have captured, through them, the fixed "transcendental signified" (the hors-texte )— that is, any referent or referents external to all textuality (in the second sense) and known to be such.

To make this last concession is to problematize at a stroke all the cognitively pertinent elements of discourse. For instance, it is to construe the reidentification of textual referents, through the course of interpretation, as forever risked and recovered through the evolving

reconstitution of such referents. It also historicizes the individuation of texts and the discipline of interpretation. It is not specifically deconstructive. It cannot be, being addressed to the work of interpretation. But it is hospitable to deconstruction. Deconstruction is not itself a method of inquiry or criticism; it is a parasitic "adverbial" attitude that attaches to one or another method. It cannot be a method, since it "posits" (only to "erase" what it posits, since there is literally nothing to posit: différance, "radical alterity") in order to mimic (parasitically) what other would-be methods of inquiry and criticism would themselves posit—by a palpable impossibility. Derrida may be fatiguing in this exercise, but he is not mistaken.37

It is just here that certain conceptual excrescences—de Man's, for instance—incubate, because it is easy to suppose that if something like Derrida's notion of textuality obtains, then history, literature, science are artificially distinguished parts of a single encompassing fiction ("il n'y a rien en dehors de texte"); or, that the very positing of the real world is itself a work of fiction. A fair sense can be given to such verbal tricks, but they plainly encourage contradiction and incoherence, if not the deliberate satisfaction of scotching every effort at understanding our world.

There is a fashionable line of argument that pretends to allow, in principle, the distinction between fiction and reality but that, thereafter, cleverly falls back to an insuperable ignorance as to how to defend the view that we actually know, in particular cases, whether we are dealing with the real or the fictive. Frankly, the important point about this strategy is that it is irrelevant to the issue at stake and misguided about what is essential to the distinction. But it is certainly fashionable.

As it happens, a convenient specimen is offered by Linda Hutcheon, while developing her own notion of "historiographic metafiction" (a literature that deliberately parodies and makes problematic literary and historical devices of reference). Happily, her proposal helps us to see what is so abusive about de Man's strategy (and what is so doubtful about Riffaterre's):

To what [she asks] does the very language of historiographic metafiction refer? To a worm of history or one of fiction? It is commonly accepted that there is a radical disjunction between the basic assumptions underlying these two notions of reference. History's referents are presumed to be real; fiction's are not. But, . . . what postmodern novels teach us [for instance, John Fowles's A Maggot, E. L. Doctorow's The Book of Daniel,J. M.

Coetzee's Foe] is that, in both cases, they actually refer at the first level to other texts: we know the past (which really did exist) only through its textualized remains. Historiographic metafiction problematizes the activity of reference by refusing either to bracket the referent (as surfiction might) or to revel in it (as nonfictional novels might). This is not an emptying of the meaning of language, as Gerald Graft seems to think. The text still communicates—in fact, it does so very didactically. There is not so much "a loss of belief in a significant external reality" as there is a loss of faith in our ability to (unproblematically) know that reality, and therefore to be able to represent it in language. Fiction and historiography are not different in this regard.38

One cannot fail to observe straight off that Hutcheon's protest about how we might claim to know "a significant external reality" harkens back to another pronouncement of hers (that was cited earlier) plainly committed to a noumenal view of "reality." Also (countering Riffaterre again): if, as Hutcheon says, postmodern novels refer to other texts and we can genuinely know that they do, then she has lost the point of the argument by her own candor; for, presumably, both the novel in question and the texts to which it "refers" are real texts and known to be such. But if that is so, then there cannot be a deeper difficulty regarding reference to the "nontextual" (that is, the nonliterary) world (on her own and Riffaterre's view of reference).

There is of course an equivocation here. Terms like "text," "textual," "texted" may be used to signify what is distinguished from the natural world, which is a "nontext" in that it is not (say) a literary text: literary texts and other interpretable artworks are located in the world at large. Nevertheless, although the world is not a text (in that sense), it may be said to be "textual" or "texted" in the further sense that determinate claims about what is found in the world are subject to some Kantian-like symbiosis in which the functions of cognizing subjects and cognized world cannot be sorted in a principled way.

I suggest that de Man treats the world as a supreme text—hence a fiction—within which the distinction between fiction and history founders ("rich en dehors de texte"), and that Derrida treats all discourse as interpretive, because there is no way to capture the noumenal cognitively ("pas de hors-texte"). If both conceptions are forms of deconstruction, they are still very different. There is nothing in Derrida's best analyses (for instance, in Of Grammatology ) that either supports the first reading (de Man's) or is compatible with it; and, oddly, it is the first that has attracted literary critics and literary historians (like Hutcheon). The second is, by this time, a bland option. It has been

given a certain arresting form by Derrida because, in Derrida's hands, deconstruction is somewhat analogous to negative theology. For what it's worth, I suggest that the dictum "il n'y a pas de hors-texte" is a philosophical thesis akin to Kant's and Husserl's, but that deconstruction is not a philosophical strategy; it is (rather) a deliberately parasitic form of insinuation by the use of denials hospitable to the other.

In any case, if we accept something like the second sense of textuality (even if not as floridly characterized as Derrida would have it), then it turns out that we cannot in principle separate the discursive use of reference and predication from our knowledge or justified belief about the world that such a use embodies. The simple truth is that, on something like a Derridean view, there is no sense in which we can meaningfully ask whether what we claim to be true really does correspond to an entirely independent external world: reference, predication, interpretation, truth, validity, knowledge, confirmation and disconfirmation, correction of beliefs and improvement of conceptual categories, reality and fiction, mimesis and representation, are all and only internal artifacts within an encompassing textualized world. That verdict holds as firmly of science as of literature. (The "hors-texte" line is meant to preclude noumena, cognitive privilege, apodicticity, totalizing, modal necessity, and the like.) Nevertheless, in unsettling the fixity of all these distinctions, deconstruction does not disturb the pursuit of objectivity or the distinction between the real and the fictional. There's the deeper lesson. Derrida's extravagance does not disturb—is not meant to disturb—"objective" practices! It offers a picture of how to regard objectivity and its fruits holistically. We may resist if we wish, but objectivism, Riffaterre's or another's, cannot be recovered as a result of effectively resisting Derrida.

(It may interest some, if I intrude here the gratuitous observation that Derrida's maneuver bears as directly and as adversely on the strategies of analytic philosophy as it does on structuralism. Thus, for instance, if deconstruction is admitted, then the so-called disquotational theory of truth, which Quine endorses and which has tempted David-son, though Davidson ultimately rejects it, would be rendered entirely untenable. The reason is this: viewed episternically, the disquotational theory presupposes that what is true about the world is, however variously interpreted by our conceptual schemes, totally unaffected by such schemes. In this sense, Quine's "analytical hypotheses" are, ontically, as benign as the structuralist's binary classifications.39 Both policies suppose that the objective world is ultimately unaffected by the condi-

tions of human inquiry and determinately structured apart from those conditions. I call that thesis externalism: it is incompatible with what I have been calling symbiosis and holds instead that the relationship between cognizing subjects and cognized objects is entirely "externalist," disjoint. Symbiosis, of course, admits that referential and predicative discourse obtains and requires "external" relations between subjects and objects [in a functional sense]; but the demarcation of their sites and local properties is itself an artifactual posit assigned within its own space. You may take it, therefore, that, in this long aside, I am trying to recover the important notions of textuality and intertextuality for analytic philosophy and the theories of history and interpretation "analytic" thought might countenance. The fact remains that it is an uphill struggle, not so much because of the impenetrability of thinkers like Derrida and Gadamer and Foucault but more because of the patent lack of hospitality, among "analytic" thinkers, for conceptual resources of these sorts. There's the pity, as I see things.)

VI

In the narrowest sense, "fiction" and "reality" are formal terms that bear on what we take the import of truth-claims to be: fictional utterances concern only what we imagine to be real, which we believe or know to be not real. That there are "stories," "novels," "narratives," even "fictions" (or stories shelved in the "fiction" section of a bookstore) does not affect in the least the sheer logic of the contrast intended. It does of course raise questions about the compatibility of melding fictional and real elements in a novel, and it does raise questions about what considerations we may rightly regard as decisive for sorting real and imaginary things. But, first of all, there would be no point to the logical distinction if we did not have a viable practice for sorting fiction and reality; and second, on the second view of textuality (Derrida's), whatever distinction we do allow is necessarily an internal artifact of the textualized world we live in. ("Il n'y a pas de hors-texte.")

Once we have these rather bland distinctions in hand, we may move to the decisive consideration: discourse about both fiction and reality rightly employs referential and predicative resources and, in doing that, makes particular truth-claims or truth-like claims about either. Interpretation, whether of actual historical events or of fictive events, is clearly assertive in this sense. But the textuality posited on the second

view cannot support such assertive acts about " itself ": the inclusive textuality of the world is not a determinate object or text of any kind; it is not an "item" in a context; "it" utterly lacks number; "it" is, rather, the supposed context of all contexts within which reference and predication, description and interpretation, obtain. Derrida's "textuality" marks only the holism of the world, as if "it" were bounded. But "it" disqualifies all the exclusionary presumptions of methodological rigor of the first model (Riffaterre's), without yet disqualifying its local practice and without disqualifying the possible rigor of such practice (referential, predicative, evidential, interpretive) that does not pretend to reach "externalist" sources of cognitive assurance (regarding truth, certitude, universality, all possible worlds, critique, or the like) and yet proceeds (distributively) beyond the holism of the second model. I say only that the second model accommodates the distinction between history and fiction; also, it is not clear at all how Riffaterre's conception of referentiality could handle fictions like War and Peace (which are plainly not "historiographic metafictions").

Hutcheon will have none of this: so she falls back to the self-defeating maneuver about the noumenal world. And de Man will have none of this: so he falls back to deliberately confusing the fictional and the real in order to justify, in the limit, any strong interpretation of history and literature that he happens to favor.

In fact, in what may be de Man's most famous pronouncement, our entire diagnosis begins to grasp the force of its own argument. It is hard to imagine how anyone (but de Man) could have contrived to frame a statement that would confirm the confusion I am attributing to him. (Again, if de Man had not uttered it, it could have been invented to convey his meaning.) On the one hand, there is the genuinely pretty notion de Man offers in support of interpretive subtlety, that catches up the title of his book, Blindness and Insight: "The rhetoric of crisis states its own truth in the mode of error. [Here, de Man opposes Husserl—correctly, if on the strength of something like the second sort of textuality.] It is itself radically blind to the light it emits." On the other hand, de Man explicates what this ultimately means, by a formula that confounds both the first and second notions of textuality:

[T]he statement about language, that sign and meaning can never coincide, is what is precisely taken for granted in the kind of language we call literary. Literature, unlike everyday language, begins on the far side of this knowledge; it is the only form of language free from the fallacy of unmediated expression. All of us know this, although we know it in the mislead-

ing way of a wishful assertion of the opposite. Yet the truth emerges in the foreknowledge we possess of the true nature of literature when we refer to it as fiction.All literatures, including the literature of Greece, have always designated themselves as existing in the mode of fiction. . .. The self-reflecting mirror-effect by means of which a work of fiction asserts, by its very existence, its separation from empirical reality, its divergence, as a sign, from a meaning that depends for its existence on the constitutive activity of this sign, characterizes the work of literature in its essence. It is always against the explicit assertion of the writer that readers degrade the fiction by confusing it with a reality from which it bas forever taken leave. . .. One entirely misunderstands this assertion of the priority of fiction over reality [de Man had just cited a suitably paradoxical line from La Nouvelle Héloïse], of imagination over perception, if one considers it as the compensatory expression of a shortcoming, of a deficient sense of reality.40

There are two blunders here: in one, de Man conflates textuality with fictionality: that is, the global intransparency that something like Derrida's "hors-texte" dictum insists on and the complete erasure of the distinction between fiction and reality in favor of a supreme fiction (which, de Man says, "has forever taken leave" of empirical reality);41 in the other, he loses thereby all possibility of the stable resources of reference on which interpretation of any kind depends.42

The corrective I have offered is quite straightforward: textuality on the second reading (Derrida's) is never treated as distributed in the logical sense. It is always and only global, holistic, "mythic"; whereas discourse about fiction and reality, in any sense that catches up literature, history, and science, is invariably distributed with regard to reference and predication and truth-claims. Taken literally, de Man's suggestion is a conceptual disaster—if it is not, as some have hinted since the disclosure of his early journalistic career, something far more sinister. In any case, once these distinctions are made clear, it hardly matters whether one favors the language of a supreme "fiction" or not: for, if one does, one must still attend to the equivocation that results; and if one does not, then one escapes that particular equivocation. In either case, no injurious paradox results.

I concede that de Man has put his finger on an essential puzzle. But it is not the logical puzzle of distinguishing in principle between fiction and history or fiction and reality, and it is not the epistemic puzzle of determining whether what is said about the meaning of a literary text or about the historical meaning of a human life can be shown to be true or false. It is the puzzle of distinguishing between the reality of the cultural and natural "worlds" to which man jointly belongs. Hu-

man life is constructed and reconstructed by the evolving meanings of our Intentional life. There is a puzzle at the critical and interpretive level, I don't deny; and de Man's subtle interpretations of particular literary texts have surely contributed to our grasp of its complexity. But it cannot be perspicuously captured by confusing the metaphysics of fiction and reality. That maneuver threatens to disorder our understanding—utterly. (Of course, resistance to a realism regarding the cultural world plays into de Man's hands.)

Certainly, against de Man, we need to recall that, on Aristotle's view of reality, poetry is more "philosophic" than history (that is, it captures what is essential to what is real rather than what is merely ephemeral). So Aristotle does not regard tragedy as fictional in the predicative sense. What is fictional in the referential sense does not come in for pertinent discussion. De Man fails to see that be needs to make everything fictional in the predicative sense, but that would produce incoherence. In analytic philosophy, by contrast, one usually hears it said that fictions are real in some sense. The motive, as I have hinted, is to improve the chances of an extensional treatment of reference. But all such accounts confuse and conflate the fictional as an intentional "object" (that is, what we merely imagine to exist but believe does not) with the actual textual construction (an author's delineating a "character," for instance) by the proper use of which we refer to what is imaginary.43

A cognate adjustment is required regarding the meaning of "interpretation." For where interpretation is construed in accord with the first notion of textuality (I mean something like Riffaterre's notion, or the New Critical view, or the Romantic hermeneutic view— not de Man's view), description always takes priority over interpretation; that is, the interpretation of a "textual item" is (in principle) always made of something individuated and referentially identified before that interpretation is submitted to assessment and evidentiary confirmation. That is the thesis of objectivism: any deviation from it, the objectivist (Riffaterre, for instance) regards as incoherent. But where interpretation is construed in accord with the second notion of textuality (Derrida's or Barthes's), where there is, to begin with, no antecedently fixed disjunction between signifier and signified (between interpreting and interpreted texts), what may (now ) be called interpretation both posits a provisionally viable such disjunction and comments, relative to that, on the meaning or semiotic import of the "first" in terms of the "second." (Here, I believe, I have come full circle back to the account of

interpretation I introduced in my opening discussion of Barthes's theory. I am quite prepared to admit that, though they are drawn to something like the second notion, theorists like Gadamer and Bloom are not at all explicit on the issue being raised. It is easy to read what they offer in terms congenial to the second notion, but I am unwilling to assert flat out that that is their own view of the matter.)

I am still concerned to demonstrate the coherence of the second view of interpretation, as well as its advantage over the first. But, in opposing objectivisms, I have no wish to undermine the objectivity due reference and predication—or the distinction between description and interpretation, or between fiction and reality, or between what is interpretable and what is not. Make no mistake about that. The uncanny thing about Riffaterre's theory is that his structuralist option threatens all of these large distinctions.

The point at issue rests with the admission of both real texts and "mere" physical things within a textualized world. All objectivists about interpretation—Monroe Beardsley (speaking for the New Critics, in the shadow of the unity of science model), E. D. Hirsch (speaking for Romantic hermeneutics, in accord with a generally Diltheyan conception of science), Michael Riffaterre (speaking for structuralist literary criticism in the spirit of that softly platonized order of universal relata so dear to structuralist science44 )—never address the deep puzzle of just where to locate, as accessible, the meanings or other semiotized elements they claim for their own. There is no transhistorical space in which such elements remain legible and constant to commentators active in one historical milieu or another.

Hence, if we choose to speak of interpretation in both respects simultaneously—the constructive and the referential—we must tame the equivocation. The trope that best fits the practice (as I have earlier suggested) is Spinoza's: the distinction between natura naturans and natura naturata. On the second notion of textuality, "interpretation" first constitutes, within a certain cultural space, the duality of interpreted and interpreting texts; on the first notion, reference and predication are then reclaimed within the space of the second. The advantage lies in this: nothing the first notion could possibly supply would ever be lost by the terms of the second (except its objectivist presumption, its insistence on some prior de re necessity); and, by the terms of the second, a greater interpretive freedom would be gained than by anything in accord with the first alone. (The first, viewed as adequate to all the complexities of interpretation, without admitting its own

artifactuality within the space of the second, is what I called a moment ago the "externalist" account of epistemic questions. The account is strongly favored in analytic philosophy—hence, notably, in Beardsley's theory of literary criticism.)

At least two specific benefits may be drawn from the second notion: for one, on the second (but not on the first), interpreted texts may be changed in their "nature" through the very work of interpretation; for another, in producing valid interpretations, we need not draw only on the resources of a "ready-made," completed, closed world of "textual items." We may now improvise from every new future-present: we may alter the meaning of what any past-present has posited; we may, in the present, fuse meanings drawn from different "historical pasts."

No doubt these heterodox possibilities raise questions of reidentifying interpreted texts over time, and of confirming and disconfirming interpretations actually proferred. But, for the moment, I am concerned more with the coherence and plausibility of the proposal than with the details of whatever novel sort of rigor it may require. I suggest that we may now begin to see as well the intimate conceptual connection between the theory of interpretation and the theory of the limits of practical human life; the prospects of human freedom are bound to be affected by the plausibility of adhering to, or opposing, something like Riffaterre's constraints on critical reading. The reason is plain: self-understanding and the understanding of "texts" are, both theoretically and practically, equilibrative, mutually confirming and mutually enabling. (I return to this important question in chapter 7.)

VII

I have now assembled the principal considerations meant to justify and soften the initial rudeness of my charge against Riffaterre's model (and similar models) of textuality and intertextuality. Radicalize intransparency in every cognitively pertinent way (in particular, regarding subjects and objects): the first model of textuality becomes completely untenable and we are forced to explore new forms of interpretive rigor within the terms of reference of the second. It was in the process of shifting from the one to the other that I located the slack opportunism of de Man. In itself, de Man's maneuver is little more than a pale echo of Nietzsche's robust condemnation of the pretensions of history: to put it tamely, history (for Nietzsche) is entitled to

claim neither a scientifc memory of the past nor the life-giving illusions of literary and artistic fictions.45 The hierarchizing of fiction over history is one of the steadily strengthened themes of the entire movement of metaliterary refection from the late nineteenth century to the present moment. (I noted it in passing, you may remember, in Sartre and Danto: it appears in an unguarded way as the prioritizing of rhetoric over argument.) It is astonishing how many important novelists and novelists turned theorists and historians and theorists of history have been drawn to something like de Man's pronouncement.46

The prospects of an objectivist history, or at least a history that captures the "truth" of the human condition without yielding to fiction, lie in three directions—all decidedly unconvincing by this time: (1) along the lines of the Aristotelian notion that man has an invariant essence, so that poetry (or a poeticized history, possibly in Thucydides' sense) could be recovered without dissolving ("ideographically") into mere fiction;47 (2) along the lines of the presumed nomological in-variances of genuine science, whether physical or human, that, in its success, comes to treat history as at best provisional and heuristic and at worst hopelessly muddled and misguided;48 (3) along the lines of a post-Hegelian grasp of the true Geist of each historical age, in terms of which (so-called classical historicism) the inevitable imprecision and play of particular details are finally judged benignly negligible.49 (Structuralism, Francophone—not Slavic, of course—completely "supersedes" history.50 )

The second notion of textuality is incompatible with each of these options; but they themselves have been severely challenged on entirely independent grounds. For example, the radical historicizing of the physical sciences (which, in the Anglo-American tradition, may be approximately dated from the appearance of Kuhn's The Structure of Scientific Revolutions but which is itself indebted to the French sociology of science and to such originals as Ludwik Fleck)51 has recently begun to yield compelling objections to both the supposed realism of, and the supposed necessity for conceding a realist reading of, the in-variant laws of nature.52 Should, then, the laws of nature be treated as mere fictions—in the sense in which they are artifactually generated in constructing a theoretical interpretation of the entire intelligibilized domain of physical nature? The extravagance is plain enough. The only pertinent advice is: say what you want but be careful. (For convenience only, then, the first notion of textuality may be labeled

"externalist"; and the second, "symbiotized." I believe these terms catch up all the distinctions we require. But I hasten to add that they are not standard categories in any sense.)

The point of distinguishing between fiction and reality is at least to resist any disabling logical paradox. Textuality of the second sort ensures the interpretability of the entire order of nature (an option usually opposed to what is said to be entailed by the first sort of textuality). Notice also the effect of adopting de Man's formula: de Man's erasure of the distinction between the fictional and the real, the advocacy of a superior Fiction that "has forever taken leave" of reality, would completely disorganize the pretense of the interpretive rigor de Man himself claims to preserve.53

It is true enough, as recent theorists of history have begun to appreciate, that "no one lives or acts—even in the course of a day—in a single structural reality."54 Structures are heterogeneous, discontinuous, historically transient, affected by interpretation. But that hardly explains the rigor of interpretation or the distinction between fiction and historical reality.

It would not be suitable to call this adjustment "relativism" either, or to claim it improves our model of history because it acknowledges the formative force of "ideology." For, of course, it ignores the self-referential paradoxes such a supposed relativism would draw in its wake. For example, the historian, Fred Weinstein, claims to favor relativism in history and literature in the following way:

By relativism I mean the notion that historical events change as generational needs and expectations change—the novelty of each future demanding a novel past, as George Herbert Mead put it. For, if all we have is, as Gore Vidal has recently stated, "a mass of more or less agreed-upon facts about the illustrious dead and each generation tends to rearrange these facts according to what the times require," then the authors who express these requirements must be making judgments about people's intentions different from the ones people themselves made at the time.55

The crux of the problem—in our time, it is really Michel Foucault's problem56 —concerns how it is that "truth" is an artifact located within one episteme or another and yet is also known and fixed as such by some practicing historian (Weinstein or Foucault), who, by hypothesis, must (per impossibile ) be working both inside and outside the episterne he is identifying among others. There is a coherent, ramified resolution of the puzzle, but I cannot supply it in its full form here.57 The penny answer is this: at the point of making truth-claims,

there is (and can be) no operative distinction between what we take to be the conditions of truth and what merely reflects our distorting "ideology"; only at some remove, temporally (even if in our own voice) or synchronically (in another's), can the distinction be maintained. Still, reflecting on the course of history, we theorize, holistically, that we too must be subject to the transient effects of the horizonal conditions of knowledge; and, of course, every would-be correction of our distorting conceptual scheme will be subject to the same seeming disorder. In any case, Weinstein and Foucault merely acknowledge the puzzle. They do not resolve it. It must be able to be resolved if the second notion of textuality is not to be rejected; and, of course, if that notion is adopted, then the impressive successes of the physical sciences must be able to be favorably interpreted in its own terms. There you have the dialectical clue to the recovery required.

We must remind ourselves, however, that, although the resolution suggested permits us to recover communicative success and the objectivity of science and the disjunction between fiction and reality, the anticipated solution need not (in the manner, say, of the alternative progressivisms of Habermas, Husserl, Peirce, and Popper) require in-variances of reason or of the world. Any would-be invariances may be coherently posited entirely within the terms of the second notion of textuality. No bona fide science need be slighted.

With regard to Foucault's aporia, all we need concede is that the genealogical stance with respect to which truth is viewed as an artifact of this or that "regime" of knowledge must be holistic, cannot be permitted to function criterially with respect to particular truth-claims. Alternatively: to judge what is true regarding the formative forces that orient the truth-claims of another age or society is not to make a judgment in the same genealogical sense in which truth in general is taken to be an artifact of given episternes. We suspect that we are similarly encultured, but we cannot judge the matter determinately. To affirm the claim in particular cases is effectively to deny that that condition pertinently bears on the claim advanced. But it is not to privilege that claim either.

In an important sense, we "privilege" our own cognitive orientation, but it is only a Pickwickian sense; we literally have no alternative. We can never say (in that sense) what the proper demarcation is between ourselves and "others." This is the pons in terms of which, within analytic philosophy, both Quine's unintended parody of the field linguist's work and Rorty's explicitly ethnocentric prejudice may be

seen to be conceptual failures. Neither Quine's nor Rorty's philosophical program makes the least sense without a principled demarcation between the collective "us" and the alien "them" of which our shared practices enable us to decide what is "objective" (Quine) or what is simply "loyal" to our local beliefs (Rorty). Neither addresses or solves the problem. That failure—the perceived failure in first-order discourse—vindicates (against Quine and Rorty) the ineliminable need for (second-order) legitimative discourse. Failure here is tantamount to solipsism (Quine) or arbitrariness (Rorty). Hence, both Quine's "naturalizing" stance and Rorty's "postmodernism" cannot but be defective. Put another way, our Pickwickian privilege is, once historicized, critically empowered to challenge our own earlier epistemic commitments and the epistemic commitments of other societies besides our own. Quine and Rorty ignore the fact: our community harbors disagreements as strenuous as any we may find between ourselves and "alien" societies.58

There is, therefore, a holist and a distributive sense in which the genealogical thesis is applied: on the first, we envisage a historically unpredictable (holist) context in which truth-claims are pursued; in the second, we pursue particular claims in a way that subverts the reliability of the claims of another episteme. Failure to distinguish between the holist and the distributed use of genealogy (or the sociology of knowledge) produces self-referential paradoxes. In a fair sense, this is the essential key to the coherence of that most contemporary form of relativism: historicism. For historicism takes seriously the thesis that knowledge is a construction of changing epistemes, although, in the practice of any particular age or society, that conviction can play no reflexive criterial role. (This is what I had in mind in challenging, earlier in this chapter, the so-called disquotational theory of truth. For the disquotational theory does not satisfactorily admit: (a) the distinction between the different functions of the theory of truth and the ascription of truth-values; or (b) the sense that ascriptions of truth-values is a function of our changing epistemes. In short, the admission of (a)-(b) precludes the disquotational theory itself or trivializes it. The critique of Riffaterre affords a soft analogy.)

Historicism is relativist only in a holist sense. If it were more, it would be self-defeating. Relativism proper, however, is concerned with distributed truth-claims. Hence, its minima cannot be formulated in historicist terms.59 But, by the same token, no relativism restricted to

its logical minima can be interesting where orderly explanatory practices are wanted.

In a curious way, the foregoing argument obscures the most elementary intuitions on which the second notion of textuality and intertextual interpretation rests. Barthes may have captured one such intuition—in advocating "semioclasm":

[I]t is the sign itself which must be shaken; the problem is not to reveal the (latent) meaning of an utterance, of a trait, of a narrative, but to fissure the very representation of meaning, it is not to change or purify the symbols but to challenge the symbolic itself. . .. In an initial moment, the aim [of the science of the signifier] was the destruction of the (ideological) signified; in a second, it is that of the destruction of the sign: "mythoclasm" is succeeded by a "semioclasm" which is much more far-reaching and pitched at a different level . . . whose operational concepts [are] no longer . . . sign, signifier, signified and connotation but citation, reference, stereotype.60

This supplies the essential clue to Barthes's overly playful opposition between "work" and "text": "The intertextual in which every text is held, it itself being the text-between of another text, is not to be confused with some origin of the text: to try to find the 'sources', the 'influences' of a work, is to fall in with the myth of filiation; the citations which go to make up a text are anonymous, untraceable, and yet already read; they are quotations without inverted commas."61 (The passage effectively identifies Riffaterre once again as the metonymic champion of the practice to be superseded.)

But what clue can I supply that would irresistibly show the power and plausibility of the radical "change of object" being recommended? Consider the following instances drawn from the world of painting— which, after all, should respond to our proposal in the same way as the world of literature and history.

One often hears it said that Matisse failed, in the well-known l'Atelier rouge (Museum of Modern Art, New York), to give a sense of a perspectivally modeled three-dimensional world by means of a strong monochrome red applied to the entire space of the studio's interior. Sometimes the point is made that the blue of the equally well-known Blue Window (also in the Museum of Modern Art, in fact hung close enough to invite comparison) is better suited for—hence, is more successful in—permitting an even bolder but quite legible departure from a conventional three-dimensional space: that is, in creating a guise of volume by means of the modeling of the color.

But the wit of The Red Studio could be made out to be, precisely, the "demonstration" that the mere placement of what are read as familiar objects (Matisse's own canvasses, a table with some familiar things on it, other such furnishings) helps us recover a "normal" three-dimensional world in the absence of (in spite of the absence of) any such modeling. If one doubts this, then, on independent grounds, it cannot be doubted in the instance of Matisse's great Café maure (Hermitage), about which Sergei Shchukin was so ecstatic, since it produces the same sort of interpreted spatial perception through an extraordinarily dreamy and mistlike white that completely erases every sense of volumes built up out of "color," without at all erasing the overriding sense of a legible three-dimensional world. What Matisse achieves here is a demonstration of just how (along the lines of at least one mode of interpretively imaginative perception) what we see, what we discriminate perceptually, is inseparable from our gathering knowledge, our narrative memory (antecedently primed but apt for instant adjustment under the effects of a powerful new experience), of the way the world is and the way it is standardly represented (in particular, in accord with Matisse's earlier experiments with first baffling and then recovering a conventional picture space).

When you see (in The Red Studio ) the clever way in which the line drawing of the table and its placement in the left front part of the painting are related to the equally clever diagonal "placement" of a Matisse canvas (already familiar as rectangular) in the corner of a room that is not represented as the corner of a rectangular room by the modeling of the monochrome red or by any other fully articulated conventional perspectival clues, you cannot help but admire the skill and tact and economy of the entire trick. Similarly, the canonical foreshortening of the legs of a suitably placed seated man in the café (Moorish Café ), where the fretwork at the top of the painting could be either a pointless ornament or a perspectivally significant series of small arches (therefore placed on the floor) at the rear of the café, again lacking in any clue that could be drawn from the ubiquitous white or from any familiar Renaissance space, you see the same spare intelligence at work. The space is built up by our memory of a practice (or of the pieces of many practices), not by an established or fixed visual convention or by a science of perspective. (So much for E. H. Gombrich and J. J. Gibson.62 )

Now, if you bear in mind that, in accord with the second notion of textuality, there is no fixed way in which the world "is" and no fixed

way in which it is represented either, then it is not difficult to entertain the possibility of interpretive rigor within the terms of reference of the second notion. That there may be many such conceptual and imaginative possibilities (even somewhat at odds with one another), functioning both synchronically and diachronically, makes little difference. On the contrary, the fact that, on the second notion, we can no longer claim to fix the way the world is or fathom its representation in any unique way testifies to the pivotal importance of the clue I have just introduced.63 Now, at the end of the century, we are in fact distinctly apt at improvising, with every new turn of art and history, some small new interpretive strategy keyed to whatever seems saliently novel in our changing experience, without the least presumption of permanence or universality and without the least expectation of avoiding conflicting, diverse, indefinitely many similarly generated inventions. What seems important, here, are viability, communicability, plausibility, scope, and the like, not invariance or fixity or cognitive privilege or the necessity of the prior closure of a critical "world." All privilege is surely gone by now. For one thing, Riffaterre can be easily seen to have obscured the historical conditions under which his own practice succeeds (wherever it does succeed)—or can be seen to hold to a practice that may (or must: on empirical grounds) be supplemented or superseded. And, for another, the interpretation of particular paintings of Matisse's (viewed metonymically for the whole of critical practice) entails an interpretation of the historical import of the ordered sequence of the actual phases of Matisse's practice as a painter. Large new inventions in painting, like Matisse's, introduce new ways of seeing paintings: hence, new ways of theorizing about the play of interpretation. The same is true in all the arts, in history, and in the explanatory theories of science.

Thus, regarding Matisse, the witty use of an unfinished painting of a woman (Matisse's own) in Piano Lesson (1916) cannot but be interpretively clarified by, say, the placement of the paintings in The Red Studio (1911), which helps us to see how Matisse erases the marks of conventional three-dimensionality without erasing three-dimensionality itself. My point is not merely that the one painting (there are many that are closer to Piano Lesson ) illuminates the structure of the other, but that that interpretation has a historicized structure within the space of the artist's inclusive oeuvre. We see Piano Lesson as we do because we relate it to Matisse's experiments with spatial paradox, convention, two-dimensionality, ambiguity, and equivocation. One cannot

mistake this, for example, in comparing the very clever differences between Nasturtiums with "Dance" (I ) and Nasturtiums with "Dance" (II ) (both from 1912) and the original Dance (I ) (1909). Both of the "Nasturtium" paintings deliberately "cut" the larger painting Dance represented in the later paintings. But, in Nasturtiums (I ) , the stand on which the nasturtium vase is placed is itself equivocally in the "Dance" painting represented; whereas in Nasturtiums (II ) , the stand is represented three-dimensionally in such an equivocal way that, though we know that Dance is represented, the figures could be construed as (active) dancing figures represented in the same space as the stand, which (then) merely happen to resemble the represented figures of the representational Dance.64

The essential point is this: we are no longer merely correcting a mistaken interpretation of the past when we appeal to the disclosures of an emerging history; they need not merely afford "more" or "more complete" information about a fixed past. We may be actively changing the historical past in terms of our evolving historical present. We need not be objectively fixing once and for all the finished past; we may be introducing a new kind of objectivity of plural reconstitutions of the interpretable past in accord with the ever-new schemes of significative coherence made possible for the first time by every new future-present. We are not changing the physical past; we are allowing only for a change in the Intentional structures of the historical past. (This is not quite the same issue as the one posed by Kuhn's admitting that Lavoisier lived in a "different world" than Priestley, though Kuhn's thesis also affects the would-be fixity of physical nature.) But it does constitute a fresh strategy against Danto's presumption regarding the end of art.65

The same clue manifests itself in Cézanne's utterly novel mode of representation. In fact, Cézanne telegraphs his extraordinary attenuation of representational clues in his landscape watercolors, particularly those that are just barely begun. For, what we see are more or less uniformly squarish patches of diverse greens placed without regard to representational specificity in different parts of the paper, as if Cézanne were loath to wash his brush until he had placed a given green wherever in the developing painting it might be thought to be wanted. The effect is quite extraordinary. We find ourselves mesmerized by the evolving effect of color patches that are emphatically not representational, that play on our cultural habituation and our capacity to extend and alter that habituation with regard to interpreting repre-

sentations, that we share with Cézanne, that in some way he knows we share.

The meaning of Cézanne's brushwork changes with its continuous unfolding within a painting and within his oeuvre. It affects the meaning of the historical past of his previous work. The interpreted sense of this or that patch is not merely corrected by construing it asymptotically in terms of the presumed relic ordering that will emerge when it is "finished." It could be read that way—which is simply the analogue of what Barthes calls "readerly" reading. But in another spirit (analogous to "writerly" reading) the interpretive work does not come to an end when Cézanne stops reworking his paintings. The meaning of whatever is discerned in the continuously reidentified piece is open to whatever our own evolving present may perspicuously provide; and then what we say of the "finished" Cézanne may continue to alter the meaning of the historical past of the original composition and of other compositions already "completed." The logic of our option is clear enough. Also, something of the vertigo it may produce.

We are caught in a certain perceptual inflexibility that our culture has entrenched—a sort of second nature somewhat more artificial and more plastic than that of the familiar textbook cases of visual illusion, though they too seem to be a little more culturally (rather than neurophysiologically) entrenched than was once assumed. Also, in the great final oils of the Mont Sainte-Victoire, the repeated image of the mountain and its approach, carefully built up in Cézanne's lifetime of representing the world, adjusted by our watching Cézanne at his work, is triggered in a completely reliable and apt way by an accumulation of paintings that have, along another extraordinarily imaginative path, attenuated every familiar structural clue that has itself been gradually abstracted as far as possible from more explicit clues favored here and there in Cézanne's continuous career. We see because we know what we have gradually learned to see by submitting to the narrative career of Western painting in which Cézanne plays such a daring role— colored by our own differentiated habits of perception and conception in the changing and fragmentary world at large. This, it may be said, is either utterly at variance with Merleau-Ponty's famous reading of Cézanne—or at least with the usual Husserlian reading of Merleau-Ponty's essay.66

What we learn with each new perception is how to reinterpret the seemingly completed past of the entire corpus of Cézanne's art. The perceptual question is itself historicized in our punctuated perception:

that is what is usually missed. That is, the perception of the Intentional structure of the late Cézanne paintings (incarnate in the oils) is itself Intentionally structured by our evolving sense of the history of Cézanne's having painted his earlier work leading to the other. Hence, that evolving history alters, by reinterpreting the Intentional structure of its own emerging phases, the historical past of those paintings and the perception of those paintings. That cannot be captured in any classically phenomenological account (for instance, by Merleau-Ponty), for classical phenomenology is ahistorical.

All that has been said is true (if it is true) with but one proviso: there is no single or completed historical narrative to master; at every point in the exercise, new lines of interpretive perception are generated, as others fall away, within the consensual tolerance of our diversely ordered experience. That tolerance, of course, is itself transient, "plural," fragmentary, and, above all, emergent in the open-ended idiosyncratic experience of one's culture. (Here, you may sense, again, the defect of Gombrich and Clement Greenberg.)

The master theme is this: if the second notion of textuality is at all apt for all our forms of life, then what congeals as our science, our history, our interpretive uniformities, our sense of salience, reliable predication, reasonable conduct, and the like—in short, the apparent fixities that fall within the space of the first notion of textuality (and within the natural world on which they rest)—are themselves already articulated within the terms of the second notion. If they are reasonably reliable and realistic, within the horizonal limits in which they function, then there is no reason at all to fear any deeper arbitrariness or incipient chaos merely by having frankly placed the fruits of the first notion within the context of the second (where they may also be gathered). If anyone doubts this, let him demonstrate how the supposed cognitive privileges of the first notion can ever be recovered.

PART III

Chapter 6
History and Fiction

I

One of the great puzzles posed by what we call human history concerns whether there is or is not a principled—also, a discernible— difference between fiction and real history. Surprisingly, the question is not posed in Aristotle's Poetics, though it is set by it in part, and though, understandably, many have supposed it to be addressed there. Aristotle does distinguish between history and poetry, but he takes poetry to have a stronger grip on reality than history—tragedy, for instance—because the merely transient particularities of human events (the historical) may not capture what is essential (or more probable relative to what is essential) regarding human nature. Capturing the latter, Aristotle believes, is capturing what is "real" and intelligible: the work of genuine science. The pertinent passage is a famous one:

[T]he 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. . .. [T]he one describes the thing that has been, and the other a kind of thing that might be. Hence poetry is something 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.1

In a fair sense, Aristotle never really concerns himself with the difference between the fictional and the real: for one thing, he does not,

as we do, construe the distinction in terms of reference; for another, he distinguishes, as we are less and less inclined to do, between history and science in terms of essences (quiddities). Tragic poetry that addresses what we treat as fictional may nevertheless capture (from Aristotle's viewpoint) what is essential (or probable) regarding human nature. And so, curiously, it may be closer to science than history, though it concerns only what we imagine rather than what has actually occurred. This, for instance, puts Thucydides's well-known license in a favorable light. Reference in suitable fictions becomes a device for instantiating universal invariances or probabilities relative to them. So the tragedies have something of the scientific about them.

Accordingly, an interpretive criticism of literature that relied on the instructive model developed in the Poetics came to be thought to hold the promise of a genuine science of interpretation. The definition of tragedy, for instance, afforded the supreme example of a trope or genre in accord with which meaningful human careers could be structured for interpretation in such a way as to conform with the essential lines along which life is meaningful. Northrop Frye, of course, is the best-known contemporary champion of such an Aristotelian interpretive science. His contribution lies primarily in expanding the primary categories of such a science in a plausible and unforced way suited to the systematic ordering of a great sector of the literary canon.2 The scheme has, by a natural extension, been subsequently applied to the analysis of history as well, most notably by Hayden White.3

The assumptions underlying both efforts are these: (1) interpretive criticism is not autonomous but guided in an objective way by the essential structure of human life; and (2) the serious practice of interpretive criticism, in literature as in history, cannot be disjoined from the conditions of human self-understanding. Hence, at a stroke, Aristotelian criticism is seen to be committed to principles diametrically opposed to anything like the structuralist poetics of Michael Riffaterre. Still, Aristotle's essentialism generates its own difficulties.

Aristotle has no developed conception of history. He has a notion of narrative, which the Poetics provides, and he has a notion of time, or at least of the enigma of physical or natural time, which the Physics provides.4 But he never brings the two together to yield anything recognizably like a theory of history in the modem sense; for the modern sort of history requires attention to the reality of the past and to actual historical referents, and both of these demand a systematic distinction between the real and the fictional. Furthermore, Aristotle's sense of the

puzzle of time invites no distinction between the physical past and the historical past; and no modern theory of history can afford to neglect that question, regardless of how it is ultimately resolved. "Time," says Aristotle, is "'number of movement in respect of the before and after,' and is continuous since it is an attribute of what is continuous."5 Fine. But that goes no distance at all toward clarifying the nature of history. In a deeper sense, it goes no distance toward clarifying the nature and unity of time; for, of course, Aristotle either never explains the meaning of "the before or after" or "explains" it by reference to what is changeless in the nature of things.6

Aristotle's notion of time is intriguing but impoverished. For one thing, although it addresses the ordered succession of temporal moments and distinguishes them from spatial orderings by construing them as real structural features of physical movement, it does not construe such orderings directionally, that is, in anything like the sense in which such successions have a certain necessity or strong weighting of some kind in the ordering of unified time itself. It construes "before" and "after" only relationally and only within the space of a timeless plenum in which motion obtains, so that it lacks the full notions of past and future, of what existed but no longer exists and of what will exist but does not yet exist. It lacks the notion of singular past referents. For past and future belong to a strong conception of the unified and singular directionality of time (even of the necessity or "near-necessity" of time's direction), where time is taken primarily to measure movement among particulars, without (necessarily) supposing or invoking an encompassing timeless and changeless order of reality. Time does not "take" time to unfold, and even the persistence of things requires a sense of time that cannot be explained in terms of potency and actuality. For those notions (the Aristotelian categories) are explained by Aristotle in terms of the invariant essences of distinct kinds of things, which invite a different sense of time.

Second, even though Aristotle construes time as encompassing "everything"—since, as he quaintly puts it, even the "now" and the "before" and "after" are in time, and even the essences of things are "in" and are measured by time7 —reality (for Aristotle) has structures (essences) that are impervious to whatever processes time measures. It is these structures, apparently, that are the decisive ones for any science.8

The modern theorist of science tends to agree with Aristotle, at least here, as in subscribing to the view that nature is everywhere subject to invariant universal laws. It is just that nothing that exists is now

thought to be eternal. Changelessness is in the laws, but nature changes within the constraints of an inclusive temporal order. The thesis is generally thought to be coherent: it is, in fact, the canonical view that has dominated the Western world through most of its modern history. But it may not prove ultimately defensible—that is, the modal claim, the claim that, necessarily, nature has a changeless structure. Here and there, it is true, a daring theorist like Charles Sanders Peirce treats nomic in-variances as the conjectured asymptotic limit of the evolving orderliness ("habits") of all natural processes,9 which, of course, Peirce intends in a peculiarly sanguine spirit—but without demonstration. This is, precisely, the weakness of his theory: he treats the lawlike regularities of natural things (in ordinary time) as "habits," as intrinsically subject to change, as not at all necessary; but he also imposes an in-variant and telic inflexibility on the direction of these same habits, leading to exceptionless laws in the limit of time. What we owe Peirce, therefore, is a sense of the coherence of the contingency of the would-be laws of nature, even if it is not possible to ensure a realist reading of strictly invariant laws. On the contrary, to abandon Peirce's teleologism is, in effect, to entertain a genuinely radical conception of science. More recently, theorists of science have become noticeably more willing to entertain the notion that there are no nomological invariances in the real world, however much we suppose we must assume or invent them for the sake of scientific explanation.10 Hence, it is implicitly conceded that the modal thesis is no longer compelling. Some theorists have even supposed that the directionality of time or, more substantively, of entropy, is a structural (not an anthropomorphized) feature of the physical world itself.11

The important thing about this line of speculation is that it is problematic. Hence, most of the disputed hypotheses about the directionality of time are probably reconcilable, one way or another, with some part of our intuitions regarding human history. On the most strenuous reading, when they are construed as invariant and exceptionless, the laws of physical nature are said to be formulable without invoking any referential devices or referentially individuating indices.12 (It may be doubted that there are any such laws.) But the directionality of time seems not even expressible except as indexed to our world (or, euphemistically, to other worlds like "ours"). And yet, both views accommodate a relatively intuitive notion of human history. For both a world of changeless structures relative to which particular things change in a temporally invariant order of change (preeminently, as in Aristotle's

biology—a biology of the repeatable actualizing of potentialities— however oddly extended to the world of inanimate physical movement and human art13 ) and a world in which strict invariances are denied can accommodate reference to real pasts.

Aristotle's biologized intuition is meant to capture a "circular" order, what is repeatable in the same way from generation to generation: which is to say, whatever directionality Aristotle grasps—biological process and human agency certainly feature some sort of directionality—he never generalizes about the ubiquitously unified and uniform directionality of time over all its intervals of tensed, relationally defined "befores" and "afters." So it is true enough that Aristotle has some notion of directed time (of generation and corruption, of potentiality and actuality), but he has no notion of past and future beyond "before" and "after" or beyond the episodic, repeatable order of potentiality and actuality.

One could formulate an Aristotelian conception of history, but Aristotle does not. That is what is so telling about his pronouncement that "actuality is prior to potency":

From our discussion of the various senses of 'prior', it is clear that actuality is prior to potency. And I mean by potency not only that definite kind which is said to be a principle of change in another thing, or in the thing itself regarded as other, but in general every principle of movement or of rest. For nature also is in the same genus as potency; for it is a principle of movement—not, however, in something else but in the thing itself qua itself. To all such potency, then, actuality is prior both in formula and in substantiality; and in time it is prior in one sense, and in another not.14

Aristotle has the notion of the inclusiveness and unity of time, but only as the unity of the formal measurability of piecemeal movement or nature (physis), as far as the merely tensed distinctions "before" and "after" are concerned. He also has some notion of the directionality of time, but only distributively, only regarding what is cyclically generable and corruptible or what may be conceptually extended (by analogy) from such processes. Failing to combine the two, Aristotle has no concept of history or of a real past in the modern sense.

Alternatively put, if history is thus confined, then there will be no difference between tensed order ("before" and "after" and "now") and the seemingly unified temporally directed order of "past" and "future"; for, as many physicists are inclined to believe, the laws of nature are indifferent to the formal ordering of past and future. Natural processes are reversible in principle, they think, as far as any merely tensed

order is concerned. In that case, physical and historical time would be one and the same; there would be no fundamental difference between past and earlier or between past and changed or between past and destroyed; and history would simply impose a narratized direction on tensed physical time answering to the perceived directionality of mortal life, human memory and expectation, and deliberately planned activity. Broadly speaking, this would be the physicalist's view of history, whether reductively pursued or not.15 It would be a view congruent enough with at least part of our intuition about history.

The physicalist's is certainly a coherent conception, but it may not yield an entirely compelling view of actual history. In any case, it is possible to square a theory of history with the merely tensed ordering of physical time: either by simply denying that the events of human culture (the locus of history proper) are analyzable entirely in terms confined to physical processes, although their distinction does not (and need not) violate the laws of physical processes, and although physical and historical time may or may not prove to be the same; or by affirming that physical nature, essentially or nearly essentially, already exhibits the directionality of time and, in doing that, captures history.

The problematic nature of the question—whether the physical world is temporally directed—makes both of these options initially eligible in an intuitive sense. Hence, in supposing that the time of human history is "necessarily" unidirectional—in supposing that there are real pasts (what once existed or obtained but now no longer do), in supposing that past and future are not (merely) provisionally tensed orderings of natural movements, in supposing that past and future cannot be reversed (even though particular states, whose onset and decline may be marked in a tensed way, may sometimes be restored)—we are successfully isolating the regulative notion of human history.

The most familiar intuition holds that the directionality of time is a necessary or "nearly" necessary form of order. (The strong entropists say there is some probability favoring the reversal of entropy.) But if that turned out to be false, we should still not be obliged to abandon the notion of history altogether. We could always retreat (if we cared) to the physicalist's view; or else treat history as a heuristic account of special interest to human beings, even if delusive. In any case, Aristotle has no conception of even that much. Since the matter has not been satisfactorily resolved (but does not disorder our intuitions about human life), the notion of the directionality of time may indeed be taken to limn the regulative notion of history, without our being clear as to

whether it is also a constitutive principle of nature or of human consciousness or of both.

Two important constraints may be drawn from these conjectures— in spite of not having favored any one theory of history over any other: first, we must distinguish between past and future (in a suitably strong sense of the encompassing directionality of time) and merely tensed, relationally ordered discrete temporal continua; and second, we must concede that the directional sense of time is at least strongly, possibly even necessarily, grounded in the characteristic mode of experience of human beings (possibly grounded in the directionality of physical movement itself—but not, or not yet, demonstrably known to be such). We cannot (yet) show that the directionality of time is, everywhere, a necessary or inviolable constraint; but we also cannot show that its assumption is incoherent, or even dispensable, at the same time we acknowledge the normal range of human thought and experience: that is, regarding the mortality of life, regarding memory and expectation and the structure of Intentional human activity.

If we accept this much, we need not insist on the ineluctability of time's direction, or the ineluctability of the consciousness of time as a directed process—in any sense prior to, or detachable from, the incarnate human experience of the natural world. Changes in that world, or changes in our perception of that world, may conceivably alter our sense of the alleged necessity. This seems to be the heart of the objection Paul Ricoeur mounts against Edmund Husserl's strategy regarding the necessary structure of internal time-consciousness.16 Our theory of history may need (and may be able) to presuppose no more than the salience (rather than the necessary invariance) of the directionality of time. The truth is we cannot yet be sure which way to turn: the fact that that is so is of enormous importance, conceptually. Our best intuition urges that whatever changes we entertain regarding time's directionality (as distinct from the apparent irreversibility of particular natural processes) should be tested so as not to render incoherent our system of temporal reference.17 We cannot demonstrate the apodictic necessity of the directionality of time—a fortiori, the necessity of admitting a uniquely historical structure in any reasoned use of human concepts. But, as we understand ourselves at present, the directedness of time does seem invariant and unavoidable. (Some, Kurt Gödel for instance, have not thought the reversibility of physical time incoherent.18 It is its inseparability from physical processes that makes the reversibility of time problematic.)

The issue is extraordinarily subtle and extraordinarily strategic regarding questions already raised. It may pay to spell out its significance a little. For what our uncertainty reveals is that our sense of what is necessary or conceptually invariant is inseparable from our actual experience: whether, say, Husserl's sense of the necessary structure of "internal time-consciousness" is genuinely invariant, universally binding, is, on the argument, something Husserl could not possibly know—and could not possibly know he was approximating. Maurice Merleau-Ponty raises similar doubts. He worries, for instance, that Husserl could not have accommodated well enough the fact that so-called primitive peoples lack the sense of historicity (characteristic of the European understanding) Husserl presupposes in developing his phenomenology. He reports Husserl's acknowledging something of the sort in a letter to Lucien Lévy-Bruhl (11 March 1935). Along related lines, he ponders whether and how Husserl could determine the extent to which, say, Freud's conception of sexuality captures the invariances of sexual intentionality.19

This is also very close to the nerve of Derrida's deconstruction of phenomenology and structuralism and to the "hors-texte" theme— hence, both to Derrida's philosophical challenge to modal universalities and to the deconstructive play of différance. Derrida applies his most sustained philosophical analysis to the study of Husserl's "Origin of Geometry." There, he poses a genuinely fatal question—which he directs against Merleau-Ponty as much as against Husserl—and which he could have directed as compellingly against Gadamer (and others)— namely: how, if reason (whether phenomenological, structuralist, hermeneutic, or "Enlightenment") is horizonally constrained by the contingencies of natural experience, could we possibly claim to discern, or approximate to, the necessary limits, the eidetic invariances, the normative or pragmatic necessities, of conceptual imagination? There can be no entirely satisfactory answer. Yet, in the "Origin of Geometry," written within a year of the letter to Lévy-Bruhl, Husserl seems (on Derrida's account) to have sustained his confidence that, although phenomenological reflection cannot be regarded as "autonomous," "pure," detached from the vagaries of history, it remains possible nevertheless to make progress regarding the universal limits of conceptual variation.20 How is that possible? Derrida asks. Derrida's own literary efforts are directed in large part to exposing the utter indefensibility of any such claim.

I should say that what Derrida has discovered here is a version of

the master theme that links (in the least ambiguous way) Hegel, Marx, and Foucault: namely, regarding what may be called the praxical nature of theory, the fact that our search for cognitive origins, apodictic certainty, exceptionless universality, modal necessity, and genuine totalizing cannot be more than horizonal, artifactual, contingently constrained (but not for that reason useless or unrewarding). I am bound to say that my own argument is in complete accord with this important finding. (It was, for instance, the point of querying Gadamer's appeal to the "classical.") We are reminded here of the historicity of thinking, although that is certainly not Derrida's thesis.

II

Let us return to history and fiction. Insofar as the past is "no more," reference to the past is, in a formal sense, hardly different from what is absent from, or indiscernible in some respect within, any immediate field of referents. This holds for what once was but is no longer (the past), what is to come but is not yet (the future), what we cannot perceive but may still play a causal role with respect to what we do perceive (theoretical entities and processes), what may, for all we know, instantiate this or that general attribute (possible particulars), what could under other circumstances be discerned but not now (the contingently imperceptible), and what we merely imagine to exist but does not (fictional entities). The differences among all these sorts of possible referents need not adversely strain our referential practices. If we concede that we make reference to instances of any and all of these kinds, then such reference is uniformly empowered by our capability of referring to what, here and now, we can discern (what is actual and present). There is reason, therefore, to say that reference is inherently Intentional and that the nature and existence of what we make reference to is a matter indifferent to and different from that of the grammar of reference itself.21

Even if we were inclined to resist the general thesis, we could not easily deny that reference to the real past, reference to the contingently absent, reference to the inherently imperceptible cannot fail to be irreducibly Intentional. But if that is so, then there cannot be a compelling objection to our making reference to what is only (coherently) imagined to exist but does not. The concession is contested, of course, but it has never been shown to be intrinsically incoherent.

The difference between history and fiction remains unaffected by

mere referentiality (that is, the mere "grammar" of reference), for history (we say) is intrinsically concerned with the real past, and fiction is concerned with what we merely imagine to exist but does not; and (on the argument) both are open to reference in the same way and in the same sense—indifferently to their real differences. So it would be an enormous extravagance to claim that the past was inherently a piece of fiction: simply because it is now imperceptible (exists no longer) though (we claim) it once was perceptible, all the while we "pretend" to make reference to it now (though it is gone). If the extravagance just proffered were adopted, then, by a very small adjustment, we should have to hold that all theorizing about imperceptible physical processes was also indistinguishable from fiction; perhaps, then, even the world we suppose persists while we sleep would be a fiction.22

The mistake is plain enough. So are the stakes. We need the resources of theory and imagination in speaking of reality every bit as much as we do in speaking of fictions. In that sense, interpretation is as relevant and as unavoidable in science as it is in history and art. In fact, ever since the Kantian revolution, we find ourselves unable to pry the "brute" or "independent" world apart from what we take to be the discernible world, the world captured in our concepts. Since the middle of the eighteenth century, our sense of that difficulty has deepened—even against Kant—because we cannot deny a certain contingent, historically artifactual, horizonal, synchronically diverse cast to our various theories of what is fundamental to conceptual imagination and the intelligible world. We are no longer as sanguine, conceptually, as Kant (or Husserl) about the invariant structures of experiences and the human mind that is their source. What, determinately, we take the real world to be is, with regard to its cognitive source, an artifact of our changing history. Similarly, every would-be corrective of any particular theory of the world is also an artifact of that same history. But saying that does not entail that the real world is a fiction or that, however real it is, certain of its imputed structures (for instance, those suggested by explanations in science) are or incorporate fictions.23

The distinction between history and fiction is, as I have said, a distinction between different kinds of referents. That difference is a mutually exclusive and correlative one: what is fictional cannot be real, and what is real cannot be fictional. We may have difficulty in distinguishing between would-be instances (the status of King Arthur and the patriarch Abraham, for instance), and we may construct stories in which both fictional and real events and persons are jointly involved

(Sherlock Holmes and Gladstone, perhaps). But that has nothing to do with whether history, since it involves reference to what (ex hypothesi) no longer exists, or to what was but is no longer a phase of what still exists, must involve the ineliminably fictional. What is theoretically conjectured with respect to the real is not at all identical with what is merely imaginary with respect to the same. Perhaps an overly zealous emphasis on the difference—possibly a supposed discrepancy—between the noumenal and the phenomenal is responsible for thinking that all serious discourse concerns the Fictional. But that would be an empty extravagance, for we should still require, within its precincts, some disjunction between the "real" and the "fictional" as we ordinarily use these notions.

The most egregious confusion between the real and the fictional may be found in Paul de Man's pronouncements—as I have already remarked. It is extraordinary how many theorists who address the question of the nature of history, or the difference between history and fiction, subscribe to de Man's formula—and how many may have been directly influenced by it. I have already examined de Man's mistake— namely, the conflating of an operative distinction between reality and fiction on the basis of literature's allegedly being "the only form of language free from the fallacy of unmediated expression."24 Symbiosis and intransparency, I have argued, are common ground for history and fiction: their differences need not be lost. But perhaps another specimen of the recent de Manian eroding of the distinction between history and fiction may help us here—this time from the side of history.

As it happens, a recent discussant, Fred Weinstein, offering an overview of the theory of history and historical interpretation, begins at precisely the same point as does Linda Hutcheon's account of metafiction (already noted): namely, at the breakdown in "the distinction between the real and the imagined, 'between formal fiction and the actual, palpable sense of life as it is lived,' and especially between literature and history as art forms."25 Weinstein is content to report the paradox and problem rather than resolve them: he retreats behind a barrage of accumulating citations that "make the case" rather than press any of their false leads. In fact, after much ingenious sleuthing, Weinstein ends by announcing only that "there is no definite answer, of course" to the question of the distinction between history and fiction.26 But since he had already given the initiative to the "postmodernists," at the beginning of his report (the same ones Hutcheon favors), we are obviously supposed to believe that the postmodernists

have indeed made a compelling case and have obliged us to abandon the "unacceptable . . . approaches and proposed solutions [of objective history] the plausibility of which had hitherto been more or less taken for granted." Weinstein's final recommendation (if that is what it is) is just this: "It will be interesting to see [he muses] how theory and practice are affected by [the reexamination of the nature of history that is required]."27 Well, that's hardly a book's worth.

Behind the maneuver, we may still glimpse the causes of Weinstein's paralysis. Weinstein has a robust sense of the reality of the historical past, but it is worn through by the rhetoric of the more persuasive postmoderns. For instance, he cites admiringly the subversive hint offered by E. L. Doctorow on that author's own novel, Ragtime: "What's real and what isn't—I used to know, but I've forgotten. The book [Ragtime ] gives the reader all sorts of facts—made-up facts, distorted facts—but I happen to think that my representation of historical characters is true to the soul of them."28 Apart from its echo of de Man's formula, the remark (taken on its face) obviously (misleadingly) suggests Aristotle's assessment of history and poetry—and of course Thucydides's actual practice. But the Greeks, you recall, bad a firm sense of invariant human nature that poetry could represent, that Weinstein (and Doctorow [on Weinstein's reading] and de Man) has entirely given up. Certainly, though he offers it in what is nearly a third-person report, Weinstein's commitment makes the recovery of Aristotle quite impossible:

The claim that historical interpretations are fictions, inventions, that they are always ideologically contaminated, intended to serve present needs, closure rather than truth, has become routine. On this view, the constructs that historians employ to interpret the significance of events are themselves no more than subjectively interested ways of talking about the social world, which is too heterogeneous, fragmented, and discontinuous to be reduced to such constructs.29

Weinstein may not appreciate that his reading of Doctorow's insistence is more Nietzschean than Aristotelian. But there is no question that, in recent years, there has actually begun to develop an odd sort of convergence between a much-domesticated Nietzsche (caught up with the values of bourgeois freedom and character) and a largely de-essentialized Aristotle (still able to legitimate past canons of social life). Perhaps the best-known instance of this may be found (largely, it must be said, by what is denied and what is left unsaid) in Alasdair Maclntyre's After Virtue; but it may also be glimpsed as well, more

circumspectly, in Alexander Nehamas's Nietzsche, Life as Literature and, of course, in Martha Nussbaum's Love's Knowledge.30 (In any case, theorizing under the grip of an ideology is not the same as inventing fiction.)

Weinstein also cites (again approvingly) Hayden White's favorable gloss on a recent statement by Nicola Chiaromonte: "It is only through fiction and the dimension of the imaginary that we can learn something real about individual experience. . .. I felt that this could only be done on the basis of that particular kind of historical truth which is fiction and, more especially, great nineteenth-century fiction, whose avowed purpose was to provide the true, rather than the official, history of the individual and society."31 White's own gloss runs as follows: "These literary artists give us truer, because they are more honest, representations of the human experience of historical events than do historians themselves."32

White's reading falls somewhere between an existentialist and a hermeneutic sense of truth (not unrelated to Heidegger's and Gadamer's sort of theme, say) and a full-fledged Nietzschean sense of life-giving illusions. But it is definitely not Aristotelian (though it derives from it), for the simple reason that, on White's view, "it is not a matter of choosing between objectivity and distortion [in history], but rather between different strategies for constituting 'reality' in thought so as to deal with it in different ways, each of which has its own ethical implications."33 The operative word is "constituting 'reality'": it catches up both the intransparency of reality and the doctrine of the "heterogeneity" of the ``constructed" real world. Both themes preclude Aristotle, though first impressions are against such a reading. (White's use of ``truer" is problematic.) For his part, Weinstein holds that "no one lives or acts—even in the course of a day—in a single structural reality"; on the contrary, people are regularly engaged in "construct[ing heterogeneous, discontinuous, conflicting] versions of the world in terms of their own needs and interests that are more solid in imagination than they ever are in reality."34 So the distinction between history and fiction is threatened again.

Several important findings fall out here. First of all, if fiction is "true to the soul of [historical characters]," and if reality is "constructed"—variably, under changing history and changing ideological needs—then it is impossible to construe fiction's "truth" in Aristotelian terms. (That is de Man's formula.) Second, the perceived need to save some sort of fictional truth (in the postmodernist spirit) and to

count it as superior to history depends entirely on construing history and science as incapable of grasping what actually is true about reality (or about the reality of the human world); but that presupposes that reality is ultimately inaccessible and that the imaginative "construction" of the world is no more than the construction of a habitable imaginary world. (That is also de Man's formula.) Finally, if these two findings are adopted, then the "truth" of fiction can only be construed in terms of life-giving illusions or some sort of existential conviction of authentic life. (But, of course, that is de Man's final formula.)

There is a certain flabby quality in all this that we can dismiss at a stroke. One perfectly elementary distinction completely undercuts these flounderings. It is this: what, symbiotically, we "construct" as the structure of the intelligible world we may get wrong and still correct, but only within the terms of that same ultimately impenetrable symbiosis we acknowledged at the start. It is only the conceptually illicit, "wild" insistence on an inaccessible noumenal world (to which, alone, ordinary truth "rightly" applies) that discredits the possible realism of the "constructed" features of our recovery of the historical past— together with the constructed features of whatever we take the real world to be.

Fiction is not, as theorists like Weinstein (and Hutcheon) seem to believe, the construction of an imagined world that takes the place of a noumenal reality we can never reach; it is, rather, the "construction" of a wholly imagined world explicitly taken as unreal in contrast to the real world we inhabit, however intransparent (and "constructed") the real world may be. The notion of fiction is logically inextricable from the notion of reality—by way of exclusion. By itself, that has nothing to do with the intriguing possibilities of historical fiction and "historiographical metafiction." Follow the argument to its end: there is no reason to suppose the historical past is a fiction, and there is no way to make it one if human lives are histories.

Another way of forcing the distinction is this: the "construction" of the real world is meant as a symbiosis in which (along "somewhat" Kantian or Husserlian lines) the ultimate constituting elements of thought and "brute" world cannot be designated at all-or cannot be designated except horizonally within that same impenetrable symbiosis. The "construction" of a fiction, by contrast, is a construction out of whole cloth—a construction of what is merely imagined to exist but does not. Hence, however incompletely determinable the properties

and boundaries of the real world are when tested, they are determinately testable against the saliencies of experience ("secondness," for instance, in Peirce's sense); whereas fictional worlds, being unreal, are determinate only as imagined and imaginary. Ontically, fictions are constructed in the (real) world.

Furthermore, under the constraints of intransparency and symbiosis, it is true that the (historically) constructed nature of reality (and human thought) cannot ensure any uniquely correct account of the real world. Realism yields in the direction of alternative—possibly nonconverging, even incompatible—"constructions" of the world. Fiction, in the logical sense, concerns only the reality of our would-be referents— not the constructed nature of the world and not the plurality of valid theories of the world. Since what we judge to be determinately real is a theoretical posit made within the holist constraints mentioned, the concession does not entail the conflating of reality and fiction at all. It is by this time a theoretical commonplace. For example, in Quine's Word and Object, though we are encouraged to be realists about science, we are said to be unable to resolve in principle the irreducibly plural ontologies that we fit to our experience of the world: "there is no fact of the matter" as to what the real world is—uniquely.35 In that sense, contrary to de Man's claim, mere intransparency does not entail the fictionalizing of reality.36 The plurality of the "constructed" forms of fiction and the plural interpretations of reality hardly erase the distinction between fiction and reality. It is, rather, the resistence of what "exists" (Secondness, as Peirce has it37 ), the role of prediction, technical intervention,38 and the like, that serve convergently to disjoin the real and the fictional.

Please note: in pressing the disjunction (between the real and the fictional), I do not mean to ignore the complexity of the Intentional world. The point is, of course, that both human history and fiction belong, predicatively, to the Intentional. That accounts for our cognitive difficulties. The predicables we employ belong to the real world of our historical culture; but whether what we predicate of what is historical and of what is fictional is true depends on a robust distinction between real and fictional referents and between what, given the constraints of historical and fictional discourse, may rightly be predicated of each. So the fact that historical and fictional referents "share" the same Intentional predicates does not as such entail any confusion between fictional and real referents or between the "natures" ascribed to each.

That Gladstone and Sherlock Holmes "are" men does not at all entail any aporia regarding reality and fiction. To think otherwise is, I'm afraid, a dreadful confusion.

III

I keep returning to these minima of history: first, that historical time is directional, whether physical time is or is not, that it is cast in terms of a unified past and future, not a sequence of tensed relations; second, that the difference between history and fiction is essentially a difference between two kinds of referents, real and imaginary; and third, that the historical past is real, despite the fact that, like reality in general, it is "constructed." Hence, by parity of reasoning, there is no basis for supposing that the proliferation of plural histories is itself a sign of the confusion of history and fiction.

But what, we may ask, is the historical past? What is the difference between physical and historical time? The short answer is: Intentionality. The directionality of historical time, the difference between real and fictional referents, the constructed nature of fiction and the constructed nature of reality are all Intentional distinctions. So is intransparency, that is, the condition under which historical claims are said to be true or false as opposed to the condition (also Intentional) under which fictions are merely imagined to be true.

The standard answer to the questions just posed is of course that a history, whether real or fictional, whether true or merely imagined to be true, is a narrative structure of some kind: with regard to both the referents of a history and their representations. The issue of narrativity has attracted its own share of dubious theorizing, as much perhaps as the questions already put to rest. But to answer the new question effectively promises to bring our previous answers to bear on the full puzzle of interpretation.

History still affords the most legible clue, chiefly because a realist account of history seems particularly plausible—perhaps even necessary. We cannot, therefore, evade these well-known queries: (1) how, in accord with Foucault's important aporia, can truth be an artifact of this or that historical episteme and yet at the same time legitimately claim to characterize that episteme?39 also, (2), how can logically irreconcilable interpretations of the ("same") historical past still be jointly validated consistently with construing that past as genuinely real? For the moment, these complex questions are not my concern. In any case,

they cannot be answered without laying the kind of foundation I am now attempting to provide. Also, the answer lies scattered in what has already been said.

Consider the theory of narrative history. The curious thing is that the mixing of fiction and history (as in postmodern "historiographic metafiction") does not affect the conceptual fortunes of narrative structure at all. That may come as a surprise. There is no principled difference between the narrative structure of history and the narrative structure of fiction, even if there are "events" in a fiction that would be impossible in a real history. The formal structure of narrative is neutral as between fiction and reality, just as the grammar of reference is. The difference between fiction and history, I claim, is a difference between possible grammatical referents and what we may say of them. (Fiction is both referential and predicative.) But the general structure of a story must be linked with the kind of narrative in which human beings reflexively remember (the past), anticipate (the future), and recount their own lives (in a temporally directed way centered in the present). (This much already appears in St. Augustine.) Once such narratizing is conceded, of course, there need be no limit to its variant forms. Also, mastering the narrative structure of human life is nothing more than whatever is entailed in mastering social life itself (some Lebensform ). Mastering the narratives of life is like mastering natural speech: one need not be able to characterize the skill.

Only if there were essential human types or an essential human nature, could there be a universal structure of historical narrative (or universal literary genres). But if, distinguishing between human persons or selves and members of Homo sapiens, human "nature" (the "self") were itself a historicized artifact, then the structures of narrativity would be whatever Intentional sequences were compatible with the directionality of time, with memory and expectation keyed to that directionality, and with the perceived biological cycle that belongs to Homo sapiens.40

These are surely the minimal constraints on narrative structures. They are not particularly strenuous constraints. For one thing, the contingent interests of culturally formed selves dominate narrative, not the bare biological limits of Homo sapiens; for another, even in the real world (think of Annaliste histories41 ), narratives readily address referents other than individual human persons (societies, traditions, entire regions); for a third, the structures of historical and fictional narrative embody in an endless variety of ways any of an endless variety

of fragments drawn from the life of historically diverse societies; and, for a fourth, there are no known constraints on the narrative possibilities of fiction, apart from the contingencies of conceptual tolerance, for even contradictions may be fictionally entertained (as in time travel), and some of these possibilities (not mere contradictions) may prove historically viable. If we concede that selves are culturally constituted and their "natures" historically alterable, then the structure of history can never yield denumerably closed alternative narrative structures. Furthermore, as in Annaliste narratives, the referents of histories need not be the agents of history: the "Mediterranean world," for instance, cannot be an intelligent agent of history, though it may be treated as a historical referent; nor is that to say that there can be no effective forces in history other than those of agents. On the contrary, Marxist and Hegelian histories are rightly impressed with the importance of nonagental forces.

Considerations of these sorts confirm the openended nature of narratized interpretation—a fortiori, of interpretation in general. They therefore expose the doubtful tidiness of so-called essentializing and totalizing theories. These sorts of theory are very different from one another, to be sure; but, in the context of history, they share a fatal assumption: namely, that the real world exhibits no narrative or semiotic structure at all, although the human world (which is real) cannot be rightly understood except through the terms of one or another such structure. Hence, depending on which exclusionary theory is favored, the argument insists that narrative or semiotic structure must be imposed on mere chronicle or biological process. Francophone structuralism is the result of adhering to the latter strategy, presciently anticipated in Saussure's strong disjunction between parole and langue ;42 and a kind of fictionalized hermeneutics of history results from the assignment of narrative meaning to a world that does not intrinsically possess such meaning, as in the accounts of Hayden White and Paul Ricoeur—and, more instructively, in Arthur Danto's quite different theory. Saussure's worry was that parole was too heterogeneous and complex to support a genuinely universal science. Nevertheless, in imposing the binarism of the sign (signifiant/signifié ), Saussure obviously meant that that structure could be found in parole. If we regard parole metonymically then, the paradox of all these strategies leaps to the eye: the human world is real and meaningful, they claim; but its meaning cannot be found in it. Meaning must be imposed from above:

mythically, Iévi-Strauss affirms;43 alternatively, by way of borrowing from the entrenched resources of literary genres, by which, fictionally, our society has already narratized our lives.

You may rightly see in this an unexpected—but not at all accidental—convergence among the theories (of art or history or intelligible cultural life) offered by such disparate thinkers as Iévi-Strauss, Sartre, Ricoeur, and Danto. Each, for his own reason, "privileges" the physical or natural world and disjoins the cultural or historical or Intentional world (or the principal part of it) from the other. Iévi-Strauss does so in order to ensure the prospects of a structuralist science; Sar-tre does so in order to ensure the ontic superiority of art; Ricoeur does so in order to make his obeisance to the strong claims of analytic philosophy and its allies; Danto does so in order to remain safely within the "naturalizing" strategies of analytic philosophy. None of these thinkers bothers to legitimate his own line of argument in any sustained way. All of their views are, I say, untenable and for the same reason: they fail to explain how it is that a realism regarding the human condition can fail to entail a realism regarding history and Intentionality. (Or, they trivialize "history" as continuous physical change and disconnect historical meanings from nature.)

Hayden White's maneuver is more instructive than the structuralists', both because White explicitly acknowledges the puzzle of historical narrative and because he means to resolve it in a way that adheres (in a realist's sense) to what (metonymically) parole itself already designates. But the bifurcation remains. Thus White says:

Precisely insofar as the historical narrative endows sets of real events with the kinds of meaning found otherwise only in myth and literature, we are justified in regarding it as a product of allegoresis. Therefore, rather than regard every historical narrative as mythic or ideological in nature we should regard it as allegorical, that is, as saying one thing and meaning another.44

Here, White attempts to reduce the bifurcation to the slimmest intrusion of narrative meaning; but it is still too alien: narrative structure cannot be found, it is said, in "real events"; it must be imputed Intentionally, from above, as the "content" of what we say and do.45 Why? The only reason for appealing to this deus ex machina is the altogether extravagant supposition that the human sciences must accord with the model of the physical sciences and that "meanings" are not "real" in that world. The argument collapses as soon as it is made clear

that the physical sciences already presuppose the reality of human discourse and human action (metonymically: parole ).

White, therefore, is the author of his own reductio:

[A] narrative history can legitimately be regarded as something other than a scientific account of the events of which it speaks—as the Annalistes have rightly argued. But it is not sufficient reason to deny to narrative history substantial truth value. . .. Narrative historiography may very well . . . "dramatize" historical events and "novelize" historical processes, but this only indicates that the truths in which narrative history deals are of an order different from those of its social scientific counterpart.46

This cannot be right: first, because it is already committed to a realism regarding societal life; second, because it fails to recognize the Intentional structure of the "historical events" and "processes" it concedes. We see in White's labor, therefore, another version of the conviction that the historical and the fictional cannot be disjoined. This time (opposing de Man, say), it is sanguine about the prospects of "social science." But it remains an extravagance nevertheless.

The picture has suddenly become very complicated. White and the structuralists—in their very different ways—mean to save the rigor of both the human sciences and the ascription of "meanings" to human events (narrative, semiotic, hermeneutic, allegorical). The structuralists invent a new science of signification by bifurcating the natural and the human sciences; White admits the bifurcation and assigns "meanings" outside the ken of any scientific practice. Both strategies mean to discipline their attributions by reference to certain "codes" of meaning each (disjunctively) supposes pertinently structure our cultural world. Here again White's account is exemplary. Thus, in an effort to understand the notion of "narrative truth" implicated in Marx's interpretation, in "The Eighteenth Brumaire of Louis Bonaparte,"—the "farcical" reenactment of the "tragedy" of 1789—White introduces "the threefold distinction between the chronicle of events, the explanation of them given in direct discourse as commentary, and the narrativization of the events provided by allegoresis ":

It is not fact that legitimates the representation of the events as a farce [he says], and it is not logic that permits the projection of the fact as a farce. . .. [The] transition is effected by a displacement of the facts onto the ground of literary fictions or, what amounts to the same thing, the projection onto the facts of the plot structure of one or another of the genres of literary figuration. To put it yet another way, the transition is effected by a

process of transcodation, in which events originally transcribed in the code of chronicle are transcribed in the literary code of farce.47

One cannot fail to see in this a plausible analogue of the strongly conventional (consensually normalized) "codes" of "readerly" readings Roland Barthes acknowledged—and then displaced. You will not find the analogue of "writerly" reading in either White or Lévi-Strauss. In the context of recent theorizing about the interpretation of history, the alternative is of course already impressively championed by Michel Foucault. But Foucault presents his own option in very much the same extravagant language that Barthes does: he is writing, you must remember, under Nietzsche's influence. So he is deliberately paradoxical:

[I]f the genealogist refuses to extend his faith in metaphysics, if he listens to history, he finds that there is "something altogether different" behind things: not a timeless and essential secret, but the secret that they have no essence or that their essence was fabricated in a piecemeal fashion from alien forms. . .. What is found at the historical beginning of things is not the inviolable identity of their origin, it is the dissention of other things. It is disparity.48

More explicitly (but in the same purple style), Foucault concludes: "From the vantage point of an absolute distance, free from the restraints of positive knowledge, the origin makes possible a field of knowledge whose function is to recover it, but always in a false recognition due to the excesses of its own speech."49

I find three themes here that can be straightforwardly recovered and defended: first, the interpretation of history need not presuppose the invariance of the real world if it means to invoke truth-values; second, science (of any kind) is a normalized and normalizing practice sustained within the span of some history, that a later interval of history may displace or alter; and, third, the "genealogist's" exposé of the first two themes is itself a judgment, from a later historical vantage, regarding the epistemic fixities of some earlier episterne. Actually, Foucault himself introduces a distinction between "traditional' history and "effective" history that, in the most literal sense, is already the complete analogue of Barthes's "tenderly" and "writerly" reading:

"Effective" history [he says] differs from traditional history in being without constants. Nothing in man—not even his body—is sufficiently stable to serve as the basis for self-recognition or for understanding other men. . .. Knowledge, even under the banner of history, does not depend on "rediscovery," and it emphatically excludes the "rediscovery of ourselves."

History becomes "effective" to the degree that it introduces discontinuity into our very being. . .. The final trait of effective history is its affirmation of knowledge as perspective.50

Return, now, to the paradox of White's proposal. White insists on a closed set of narrative tropes in order to prepare (or "prefigure," as he says) human life for history. But a realist reading of history entails a "correspondence" between the real structures of history and the narrative structures we impute to the human world. On White's view, science cannot vouchsafe that correspondence; on Foucault's view, the issue is no more than trivial, since the human world is itself an artifact of "effective" history. White cleaves to the invariances of traditional science and traditional history. Foucault supersedes these. On the strength of an earlier argument, there can be no fundamental difference—either epistemically or in realist terms—between the interpretation of art and the interpretation of history and human life.

I should say that Foucault agrees with Gadamer here—and both of them, with Hegel: history is "effective" (therefore, "discontinuous," for Foucault) because historical agents generate by their own commitments meaningful events the full historical significance of which they cannot fathom in the process. Thus seen, the radical theory of history and interpretation I am trying to fashion is (eccentrically) Hegelian. But what I wish to emphasize particularly is that the florid authors of the "continental" world can all be translated in a favorable way in terms of the familiar puzzles of analytic philosophy—which, I'm bound to say, the analysts ignore. An important part of this entire undertaking is to show the strong and beneficial relevance of continental European philosophy for the seemingly insoluble puzzles at the heart of current analytic philosophy.

IV

I need to bring these sprawling themes together and pin down the decisive arguments. For one thing, I have sketched a connection between (a) the confusion between history and fiction on referential and predicative grounds and (b) the complex conviction that human life is meaningful only in narrative terms despite the fact that narrative content cannot be found in the "real events" of the human world. And, for a second, I have noted the essential disagreement among (c) those who believe that the real world is invariantly ordered and that truth answers (in the "externalist" sense) to what we can discern of such

structures and (d) those who believe that the human world is a historicized artifact of one episteme or another and that the assignment of truth-values behaves conformably (within the same "symbiotized" space).

The two themes are independent of one another, but a good many theorists seem unable to keep them apart. The reason is this: different discussants cry "fiction" when: (1) human life is interpreted narratively; (2) reference proves to be context-bound; (3) truth-claims are characterized as horizonal or perspectived; (4) our conceptual schemes are historicized; (5) there is no uniquely correct characterization of what is real; (6) the intelligible world is symbiotized, intransparent, constructed, and continually reconstructed; and (7) the world is a flux and lacks any necessarily invariant structures. I claim that none of these numbered reasons is sufficient to justify conflating history and fiction, since none—taken singly or together—implicates the difference between actual and imaginary referents. But, in a deeper sense, the confusion is a symptom of a genuine philosophical worry; for what the charge of fiction masks is a concern about the coherence of any inquiry committed to anything like the items of this last tally.51

I cannot hope to persuade you, here, that there is no incoherence threatening. But I invite you to consider the matter in the following terms. I have cast the entire argument in such a way as to suggest that we may replace every plank of conceptual fixity by an admission of indeterminacy and flux, without shipwrecking reason. My guiding thought has been this: that the entire history of Western philosophy has ineluctably yielded ground on every salient doctrine of invariance, necessity, determinate structure, and certainty regarding truth, and has gradually replaced each of these (second-order ) fixities by alternatives committed to the master themes of our age. The latter are hardly uncontroversial, but they are straightforward enough: (i) symbiosis; (ii) intransparency; (iii) historicity; and (iv) social construction. I have presupposed these themes in all that has gone before; I have also tried to strengthen their plausibility. But my argument has been focused entirely on the need to reinterpret the practice of interpretation itself—moving from the conceptual vision of invariance to the vision of flux—within the terms of (i)-(iv). The argument may be cast—heuristically-in terms of what it means to abandon Aristotle's Metaphysics Gamma and to favor Foucault's Nietzschean papers on history.

My analysis has not been centered on Aristotle or Foucault, of course. But they are surely the most visible markers of the extraordinary

change in conceptual vision that proceeds from the ancient world to our own late day. The theory of history and interpretation cannot fail to have changed conformably. The present chapter is in fact concerned to bring the notion of historical narrative into line with that of the Intentional structure of art and literature, so that a uniform theory may be provided for both. In effect, I've proceeded in accord with an argumentive declension developed in three stages. First, I have shown that reference and predication are inherently informal, context-bound, successful only in the sense of being consensually entrenched in the Lebensformen of viable societies. (The point is that it hardly matters what one's theory of reality is: reference, predication, individuation, and reidentification cannot exceed the resources of human inquiry.) When, therefore, at the second stage of the declension, these cognitive functions are also admitted to be historicized or horizoned, the additional complications are seen to accord already with the generic concessions of the first stage. The fact is that the historicizing of thought can only be admitted holistically, that is, on the hypothesis—which, in a way, Hegel, Nietzsche, Wittgenstein, and Foucault share—that we cannot (synchronically) specify the determinate limitations of our own understanding, although we may reasonably suspect that we also are not exempt from the historied conditions of thinking.

Furthermore, when we identify the cognitive limits of an earlier episteme ("genealogically"—in Foucault's idiom), we do so from the vantage of our own horizon, the cognate limits of which we cannot (as I say) similarly penetrate. Hence, there is no incoherence that threatens merely as a result of historicizing thought: the effort to ensure coherence proceeds synchronically. That, of course, is the solution to Fou-cault's aporia: the sense in which the "truth" is an artifact of history is not the same sense in which, genealogically, we state what we take to be true about an earlier episteme in which truth-claims are said to be processed this way or that.

Finally, at the third stage of the declension, any symbiotized and in-transparent world, subject to the logical informalities of the first two stages and deprived of any necessary invariance de re or de dicto, will have to accommodate the possibilities of conceptual incommensurability, the lack of criteria for fixing any uniquely correct analysis of the real world, the eligibility of relativism, real indeterminacies, and the Intentional alterability of interpretable things. I have been testing the coherence and viability of these last themes within the terms of refer-

ence of the larger vision summarized in the tally marked as (i)-(iv). I am not supposing that the encompassing vision is securely in place. I am urging instead that the adjustments required be shown to be fully congruent with that vision. Furthermore, a large part of the argument—what, in fact, is claimed in the first stage of the declension—is already conceded in the strongest currents of analytic philosophy that have nothing to do with the further puzzles of lntentionality, history, or interpretation that I have been pursuing. That, frankly, is what I have relied on.

I trust the converging import of the somewhat different issues explored in the name of the relationship between fiction and history may now be grasped. I need to return to them to round out the details of this discussion. The following claims, I judge, are both reasonable and reasonably defended: (1) history and fiction are, logically, mutually exclusive, in the sense that their respective referents are, disjunctively, real and imaginary; (2) the fact that the real world and fictional "worlds" are, in different senses, constructed or artifactual does not entail the denial of (1); (3) there is no principled difference between the predicates suited to history and to fiction; and (4) the real referents of the cultural world uniquely possess Intentional attributes—in particular, narrative structures. I should say that the upshot of (1)-(4) is simply that to admit the special problems of historical interpretation is not to admit that interpretation necessarily fictionalizes history.

We can do no better than follow Paul Ricoeur's lead here—with reservations. It collects all the puzzles and all the dead ends. For, first of all, Ricoeur believes that "White's subtle but often obscure analyses . . . constitute a decisive contribution" to our understanding of the narrative structure of real history. "White's tropological analysis," he continues, "is the sought-for explication of the category of the Analogous. It tells us but one thing: things must have happened as they are told in a narrative such as this one [that is, one that accords with White's scheme of narrative alternatives]. Thanks to this tropological filter, the being-as of the past event is brought to language."52 (What Ricoeur means by "the Analogous" will become clear eventually.) But, on a second reflection, Ricoeur fears that "White's recourse to tropology may run the risk of wiping out the boundary between fiction and history. . .. If [for instance] we cannot reestablish the primacy of [the] referential intention [to recover the real past], we may not say, with White himself, that the competition between configurations is at the

same time 'a contest between contending poetic figurations of what the past might consist of.'"53 Ricoeur believes, nevertheless, that White's contribution is recoverable.

Ricoeur's point, if I read it correctly, is this: it makes little difference whether White holds to the reality of the historical past (which he apparently does) if he (also) treats as mere metaphorical possibilities the narratized predications made of what we are referring to in (or as) the historical past—possibilities merely "of what the past might consist of," where we cannot ever say what it does consist of. No account of the narrative structure of the past may be construed in a realist way unless such a structure is supposed to be, or is supposed to encompass, the actual historical past. Ricoeur is certainly right there. White risks fictionalizing the narrative structure of the past, even if he does not intend to fictionalize the past. (Of course, the two are inseparable.)

But then, unaccountably, Ricoeur adopts an equally (if not an even more) bizarre view of narrative structures. For, as he says, White has got things exactly right in seizing the decisive importance of the "poeric of narrative": "Rhetoric [he says] governs the description of the historical field just as logic governs argument that has an explanatory value: 'for it is by figuration that the historian virtually constitutes the subject of the discourse.'"54 How, we may ask, does the historian do that? Ricoeur's answer is rather complicated.

For one thing, Ricoeur wishes to hold fast to Ranke's formula, wie es eigentlich gewesen ("the facts as they really happened," as he has it55 ) rather than "what the past might consist of." Still, in spite of the seemingly steady grip on the real, Ricoeur goes on to endorse White's "tropology":

"Before the historian can bring to bear upon the data of the historical field the conceptual apparatus he will use to represent and explain it, he must first prefigure the field—that is to say, constitute it as an object of mental perceptions." [White, whom Ricoeur is citing here, goes on, in the same passage, to say—which Ricoeur does not cite—"This poetic act is indistinguishable from the linguistic act in which the field is made ready for interpretation as a domain of a particular kind. That is to say, before a given domain can be interpreted, it must first be construed as a ground inhabited by discernible figures. The figures, in turn, must be conceived to be classifiable as distinctive orders, classes, genera, and species of phenomena. Moreover, they must be conceived to bear certain kinds of relationships to one another, the transformations of which will constitute the 'problem' to be solved by the 'explanations' provided on the levels of emplotment and argument in the narrative."] "In order to figure out 'what really happened' in the past, therefore [this is still White speaking—as cited by Ricoeur], the

historian must first prefigure as a possible object of knowledge the whole set of events reported in the [pertinent historical] documents." The function of this poetic operation [this is now Ricoeur speaking] is to outline possible itineraries within the "historical field" and thus to give an initial stage to possible objects of knowledge. The intention here is certainly directed toward what really happened in the past, but the paradox is that we can only designate what happened prior to any narrative by first prefiguring it.56

Here, Ricoeur invents his own unnecessary alternatives to the equally unnecessary maneuvers White affords. He resists White's construing the narrative emplotment of the historical past as no more than fictional or imaginative possibilities; but he also means to use White's tropology to recover the narrative structure of the historical past. He explicitly says: "This recourse to tropology is imposed by the unique structure of historical discourse, as contrasted with mere fiction."57 But how is it possible to recover that realist sense of historical narrative if the narrative structures imputed to the historical past depend on the free or arbitrary imposition of one or another of the classical tropes? White borrows the scheme from Northrop Frye—it was originally applied by Frye to fiction—because, on White's (and Ricoeur's) argument, the historical past cannot be literally said to possess the structures that the rhetorical tropes (metaphor, metonymy, synecdoche, and irony) impute to it. Why must that be so?

Ricoeur's complex answer is this: first, of the four tropes, only metaphor has "an explicitly representative vocation" (in respect of which the others are mere "variants" of it);58 and second, metaphor's mode of functioning "is not merely rhetorical but also ontological [which gives it 'a referential import']"—that is, it functions by way of a strong analogy between "being-as" and "seeing-as."

How's that? you ask. Well, says Ricoeur, you don't deny, do you, that "the past is what is missing, a 'pertinent absence'"? If you grant that, then you must also grant that it is only by the historian's use of metaphor that we can ever recover the past through the idiom (at once rhetorical and ontological) of "being-as"—that is what is needed (and is thereby supplied): "We have to combat the prejudice that the historian's language can be made entirely transparent, to the point of allowing the things themselves to speak; as if it sufficed to eliminate the ornaments of prose to be done with the figures of poetry."59 (Transposed to the phenomenon of "seeing-as," this would mean that every report of what was perceptually real was irreducibly metaphoric, or that the perceptible world was itself nothing but the way the noumenal world

appears to us "metaphorically." Both options are plainly extravagant, on the assumption—Ricoeur's own—that we are speaking of the perception of the real world. Please observe that my account, in chapter 4, of the forms of Intentionality found in the artworld is pursued in an altogether different way. Here, in Ricoeur and White, either the real world has no Intentionality at all or the human world lacks the Intentionality of historical meanings, which must then be imposed on it; there, the whole of reality is already symbiotized, the special forms of Intentionality of art found within it must simply be learned.)

There is no tactful way of softening the blow. What Ricoeur is claiming is impossible to defend—and entirely unnecessary. First of all, he identifies the historical past as "a pertinent absence" (das Gegeniiber ), something "missing" (palpably opposed to what is present, by way of a "trace"). Yes of course; but, so far at least, the historical past is not "missing" in any more profound sense than is a peach that once matured and has now rotted away (or, for that matter, the moment of life each of us endured ten milliseconds ago). The sense in which the past is gone has not been shown to entail anything more than the bearing of the passage of physical time and natural causes on things of a certain nature. The narrative tropes are obviously not required in admitting the (real) demise of the (real) peach. Ricoeur does not contend that physical science is inherently metaphorical in giving an account of past events in nature. Why not? If metaphor is not required in science, then it cannot be required in history. Or so it seems.

Second, Ricoeur (somehow) conflates the loss of the historical past and/or of what is past in nature with the cognitive inaccessibility of "things themselves" (noumena, things-in-themselves). He speaks of combatting the "prejudice" of supposing that "the historian's language can be made entirely transparent" in the context of Ranke's formula. But, of course, if the world is not "transparent" (Ricoeur's own term), then the past is no more missing than is the present. ("Il n'y a pas de hors-texte!") The entire Kantian paradox has been unnecessarily resurrected. Surely the present is not "missing." Or, is it also a noumenon we cannot know "in itself"?

Third—this is the most complicated part of Ricoeur's argument— the noumenal world, "things themselves," can, in spite of everything, be reclaimed! The trick is done by recovering them metaphorically, in the mode of "being as": "prefigured" by the narrative tropes that (cleverly, you must admit) have a connection with the world that "is not merely rhetorical but also ontological"—that is, the tropes that serve

to explain (as Ricoeur explains) the "really" of Ranke's formula by way of the "as." This is the strange sense in which Ricoeur insists: "Thanks to this tropological filter, the being-as of the past event is [first ] brought to language." It is "brought to language" in a sense, but surely only in that deliberately mythic sense that Derrida for one celebrates in his occasional deconstructive frenzy. It is not "brought to language" metaphorically (Ricoeur's intention is ultimately Heideggerean60 ) in the sense that, distributively, we can show (contrary to what Ricoeur claims) that the structure of our narrative predicates consensually corresponds to the (real) structure of the historical past.

Ricoeur apparently supposes that metaphor (the narrative tropes) rhetorically "transfigure" the physical time of real events for the sake of historical meaning: metaphor, he says, provides the necessary fictions by which real histories are first made discernible.61 He does not explain why history requires this unusual linguistic solace. There can be only one explanation: the actuality of human history belongs (it seems) to the reality of physical time; the essential meaning of history, however, belongs not to physical time but to the literary tropes in accord with which we are accustomed to reflect on the course of life. The two are disjoint, but the "metaphor" of narrative history brings them together—"metaphysically." For some inexplicable reason, the "metaphysics" of human existence requires that history combine the reality of physical time and the fiction that results from applying the tropes of literature to the actual structure of such time.62 In short, Ricoeur embraces the modernist conception of history: he disjoins the realism of the human condition (historicity) and a realism of the meaning of history (historical interpretation).

No, the only way to proceed is to retreat from any commerce with the noumenal world. Quite simply then, historical narratives may validly represent the historical past because the past possesses a narrative structure. (I am persuaded, I may say, that Ricoeur's thesis is extremely close to the theme of Danto's account of interpretation in art and history. That is, Danto's insistence on "transfiguration" and Ricoeur's on the "metaphoric" meaning of historical narrative seem to me to play very much the same philosophical role in their respective accounts. The language and philosophies of each are, of course, worlds apart.)

The theory of history, like the theory of interpretation, demands a unified account of narrativity and reference in Intentional and realist terms. Neither White nor Ricoeur succeeds, because, puzzled by the reality of the historical past, they cannot bring themselves to regard

history's narrative structure as as real as whatever they treat as physically real. There's the fatal mistake. Each fails to provide for the conceptual adequation between what we claim to be true historically and what, in human life, could legitimate such claims.

Weinstein had gone wrong because he treated narrative structures as confined to "no more than subjectively interesting ways of talking about the social world."63 That's what Weinstein meant by "ideology"64 and by prioritizing the material order over the ideological. For Weinstein, narrative is a rhetorical flourish imposed on an order of nature in which physical time is sufficiently rich for whatever we can recover of objectivity.

Hutcheon goes wrong because she ultimately loses the distinction between historical and fictional reference. Between Weinstein and Hutcheon, then, we glimpse what the "postmodern" signifies in current literary and historical circles: (a) the rejection of any principled distinction between history and fiction; and (b) the vindication of (a) in terms of the "subjective" blurring of a robust sense of realism.

V

The argument now appears remarkably straightforward: there is no coherent conceptual strategy by which the historical past may be treated as real, unless it is also open to description; hence, if, conceding the directionality of time and the constraints of memory and expectation and the like, it must be treated as possessing a narrative structure, then it must possess such a structure. That's all! Ricoeur's detour is ineffectual and unnecessary.

The reason we cannot ignore Ricoeur's maneuver is due not only to its considerable influence but also to the uncanny way in which it concentrates in one formulation all the errors of those who either (a) erase the distinction between history and fiction or (b) indissolubly unite history and fiction in the narrative interpretation of human life. The first group (Weinstein and Hutcheon and de Man) insist that there is no point in speaking of history in realist terms; the second (White and Ricoeur) insist that there is indeed a point, but, as they explain, "real" history must, in the pertinent sense, already have been fictionally "prefigured." Here, then, is Ricoeur's essential formula:

Thanks to this power, which I spoke of as redescription [that is, the power of poetic language, by the use of which 'being itself has to be metaphorized in terms of the kinds of being-as"—in accord with the tropes], we [are able

to achieve] the refiguration of time by narrative—which is the heir of this metaphorical redescription. . .. In the hunt for what has been [the past], analogy does not operate alone but in connection with identity [the same] and otherness [the Other]. The past is indeed what, in the first place, has to be reenacted in the mode of identity, but it is no less true, for all that, that it is also what is absent from all our constructions. The Analogous, precisely, is what retains in itself the force of reenactment and of taking a distance, to the extent that being-as is both to be and not to be.65

The entire argument rests on an avoidable confusion: that between what is present and what was present and that between what is no longer present and what was never real at all. The past is what it is, and is not what it is not; it is past, hence it is not what is present qua present; but it is also not what never existed at all, and so it is the past as opposed to nothing. (This sounds for all the world like Aristotle's exposé of Parmenides' confusion between what Is Not tout court and what is not this or that but something else: namely, change.66 ) But the "is" of reference and the "is" of predication are entirely neutral to distinctions of tense, to the directionality of time, and to the distinction between fiction and reality. Otherwise, we should never capture the specious present.

Nevertheless, in bringing this part of the argument to a close, I cannot forebear remarking that historical time and the historical past do have an internal structure quite different from that of mere physical time and the merely tensed pasts we recover from the nonhuman world. I am not yet venturing a new argument here, merely giving notice of additional options.

Historical time, I suggest, is "incarnate" in physical time: it is the indissolubly complex dimension and medium of things, and of what happens to things, that manifest both physical and Intentional features at the emergent (cultural) level at which histories are acknowledged.67 For the moment, I say only that the distinction is coherent—and adequate to the complexities of interpretation and history.

The "meanings" of the events and artifacts of human life belong to the world of human culture if they belong anywhere. Merely to admit the use of natural language, then, is to admit the question of how to locate the Intentional import of the things of the human world. Whatever answer we may venture is likely to be constrained by the following considerations at least: (a) that the cultural world is consensually real; (b) that it is minimally distinguished as the "space" of Intentional properties; (c) that such properties are complex, indissolubly incarnate

in the physical world; (d) that their detection and ascription depend on the reflexive aptitudes of the members of given societies; (e) that the validity of their ascription is consensually grounded and historically alterable; (f) that, therefore, the historical past may be interpretively altered, and "incongruent" interpretations (that is, those that, on a bivalent logic but not now, would yield incompatible claims) are admissible as coherent and viable; and (g) that the objectivity of interpretive claims depends on the collective memory of a society regarding its practice of reference, predication, individuation, reidentification, and the like.

I put it to you that resistance to items (a)-(g) signifies that, if the Intentional world is real, then "meanings" must be fixed in historical time (afortiori, in the past) in a way that accords with the familiar intuition that physical events that have once occurred cannot be "undone" (cannot be such as to "become," in the physical past, what never occurred at all). This is not to say that states of affairs can never be altered. (Perhaps technology will find a way of unscrambling scrambled eggs.) But no one, to my knowledge, who has denied that the historical past (as opposed to the physical past) can be altered has ever bothered to supply a metaphysics of "meanings" in accord with which that denial makes sense. (For my own part, I say the disjunction of history and historical meaning makes no sense.)

These distinctions obviously justify a new beginning.

Chapter 7
Interpretation and Self-Understanding

I

We are drawn to two very persistent intuitions about ourselves: one, that we have no adequate conception of our "nature" in being the "selves" we theorize we are; the other, that we cannot, in general, not really know what we think and believe and feel and seem to see as we reflect on our ongoing experience. The first is a shocker but true enough. Our sciences have simply never progressed beyond the obviously primitive modeling of the self that marks the whole of Western theory from Plato's psyche to Freud's functional triad. There are excuses enough, but the fact speaks for itself. (I don't deny that there are many theories that hold that human selves are not real at all; but I confess I have never seen any sustained argument that showed that that was even a coherent possibility, let alone a confirmed truth. Assertions along these lines are remarkably unguarded, arbitrary, and nearly impossible to debate.) The second intuition is a deep discovery almost invariably misrepresented as cognitive indubitability—the Cartesian heritage, let us say—that we have never entirely shaken off.

The two are linked in the work of interpretation. For we cannot have an adequate theory of the self if we have no adequate theory of knowledge and self-understanding: the nature of the self, to speak incautiously, is just to be the competent agency of effective understanding and knowledge. There can be no science of the self if there is no science of science, no deeper and surer foundation for knowledge than

what the usual sciences collect. For the same reason, there can be no science of any part of the world if there is no science of the self; for to know what is true of the world is to know what it is that we can and do know. The thesis is naively put but compelling. All knowledge is or entails self-knowledge—in a double sense: first, because knowledge is inherently reflexive; second, because, on the thesis that philosophy can only deepen but no longer abandon its Kantian-like themes, self-knowledge and world are indissolubly symbiotized.

Of course, there is no such science! There are only—ineluctably— second-order legitimative speculations: a knowledge of what knowledge is, a science of true science, is, on the most eligible reading, a distinctly second-order concern, however shorn it may or must be of transcendental pretentions regarding a privileged form of self-knowledge. We form our theory of the self in order to reflect what we understand our capacity for knowledge and understanding to be. Furthermore, if we admit that a realist account of history and the art-world makes sense only in terms of the self-interpreting life of an aggregate of selves sharing a viable Lebensform, then there can be no adequate account of objective history and objective criticism without a convincing theory of the human sell Also, on the argument being advanced, neither can there be an adequate account of the sciences. Any serious weakening of our grip on the nature of the self, I say, signifies a weakening of our grip on the competence of science. The bridge between the two is the work of interpretation.

These rather unrefined remarks are strong enough to ensure at least three prima facie constraints that collect (not altogether uncontroversially) the entire history of Western thought: first, that theorizing about the self is a second-order reflection on our cognitional powers; second, that we have no grounds for assuming that the difference between first- and second-order competences is a difference in cognitive resource or reliability; and third, that there is no principled distinction between self-knowledge and knowledge of the world. I do not believe that any of these theorems can be convincingly denied in our own time, or that any of them is otiose, or that there are any other theorems about the self of comparable power that are less controversial. Nevertheless, there is one further theorem at least that needs to be mentioned that captures the most salient theme of all Western thought reflective of the French Revolution and gathering force as we approach the end of our century: namely, that thinking—a fortiori, knowledge— has or is a history. Cognitional competence is to some extent impene-

trably preformed. The most distinctive theories of the self in our own time tend to focus on the import of the fourth theorem applied to the other three. To accept all four is to place the theory of history and interpretation at the very center of speculative thought.

The strange thing is that current Anglo-American philosophy is largely opposed to these concessions. To be perfectly frank, I think such resistance is a serious mistake—an altogether unnecessary impoverishment of our conceptual resources. Furthermore, of course, it effectively pits the account of interpretation I have been developing against the principal strengths of analytic philosophy. There's no use in pretending otherwise. You will understand, therefore, that in bringing this account to a proper close, I find myself obliged to confront the implied objections that spring from the very practice I should have liked to enlist. I say this at the start of this chapter, so that you may have some reason for patience with a protracted inquiry that may otherwise appear to deflect us from the essential issue. I cannot hide the fact that if those objections were sound, my own undertaking would be completely pointless. By the same token, to stalemate or defeat them is to strengthen the argument in my own behalf and, also, to cast some doubt on the most salient tendencies of American and British philosophy.

I have cast this summary in as neutral a way as possible. Still, at the risk of appearing to veer off in a merely partisan way, let me add that, without endorsing his entire hermeneutic theory, I take the gist of these remarks to be quite close to what Gadamer has in mind when he says "all reading involves application"; "a person reading a text," he says, "is himself part of the meaning he apprehends. He belongs to the text that he is reading."1

I have no intention of explicating Gadamer's sentences here: they cannot be rightly understood except from the vantage of his own complex doctrine. They do, however, capture (opportunistically) the essential thrust of the foregoing tally. I claim that every adequate theory of the self will provide an analogue of this theme of Gadamer's: that even the spectator's perception of physical nature implicates some form of self-interpretation and that self-understanding implicates an understanding of the world and other selves; that is, that there is no principled difference between understanding the way the world is, understanding bow others understand themselves and the world, and self-understanding, and that that admission generates no vicious or untoward paradox.

Resisting this complication, I say, is illicitly advocating at some point

in our theory of the self some form of cognitive privilege. Furthermore, the thesis that thinking has a historical structure, that it is formed among human cohorts by the contingencies of societal life (whatever biological regularities may constrain divergent conceptual horizons) entails the alterability of the very structure of thinking. The conceptual resources of thought change as a result of their being exercised. I share this benign paradox with Gadamer. I believe it is implicit in Wittgenstein's notion of a Lebensform and Quine's notion of "analytical hypotheses,"2 though neither addresses it or history directly. It is of course essential in Hegel, Marx, Nietzsche, and Foucault, and it is ignored without cause in recent analytic philosophy.

This will not be conceded by many analysts. For instance, it is a standard feature of current Anglo-American theories of knowledge that they explicitly reject all forms of transcendental thinking—Kantian or Husserlian, say3 —and construe their doing so as directly yielding (or even requiring) a "naturalized" theory of knowledge.4 In saying this, I mean to flag the fact that, in dismissing "transcendental" thinking, analytic philosophy is strongly drawn to dismiss all "legitimative" or "second-order" thinking as well. (The difference is widely ignored in "naturalized" epistemologies.) For the moment, let me simply say that there is a lacuna in the argument. If that proved not to be so, if the paradox proved to be pointless, then (I should concede) history and interpretation would not be entitled to the central role I am according them. The generalized "Gadamerian" thesis would simply be false, and some sort of objectivism and externalism would doubtless be adequate for our cognitive needs. You cannot then fail to grasp the dialectical thread of the argument, regardless of how you might account for the self and knowledge.

With his usual ingenuity and wit, Charles Sanders Peirce had fashioned the following compressed model of mind and self that catches up in a most suggestive way—perhaps the most effective any American philosopher has ever fashioned—the connection I favor between the two intuitions with which I began:

[T]he mind is a sign developing according to the laws of inference. . .. [M]en and words reciprocally educate each other; each increase of man's information involves and is involved by, a corresponding increase in a word's information. . .. [T]here is no element whatever of man's consciousness which has not something corresponding to it in the word; and the reason is obvious. It is that the word or sign which man uses is the man himself. For, . . . the fact that every thought is a sign, taken in conjunction

with the fact that life is a train of thought, proves that man is a sign. . .. Thus my language is the sum total of myself; for the man is the thought. . .. Furthermore, the individual man, since his separate existence is manifested only by ignorance and error [Peirce means here that personal awareness or "self-consciousness" dawns first by way of the effect of incongruities felt regarding the beliefs and behavior of others], so far as he is anything apart from his fellows, and from what he and they are to be, is only a negation.5

Peirce combines here a sense of the social emergence of the self, its distinctive interpretive power, the primacy of its cognitional role over the merely mental and private, its somewhat collective nature, its inherent indeterminacy (or determinability) within its conditional determinateness, its self-transformation by way of interpretation and the influence of other semiotized (or Intentional) forces (other selves) in an ambient culture, and the inadequacy of disjoining selves and their world exclusively as observers and observeds (though selves do indeed function as observers). Here and elsewhere, Peirce offers what may be the most powerfully drawn metaphysics of the self: as an "interpretant" in an endless continuum of inter-interpreting signs. What his account lacks, however, is a developed sense of the fully historicized nature of the interpretive aspects of the human condition and a detailed study of the dialectical complexities of collective life. (Peirce is a partisan of evolution, not of radical history.) Otherwise, there is a clear convergence between Peirce and Gadamer on the paradox I say we must take up.6 (I say that analytic philosophy impoverishes itself by neglecting the matter.)

"The" self is not just another specimen for an expanding zoo of observable things, it is the site of the aptitude for conceiving any such zoo. So it is not "nothing," but it is also not anything that could be straightforwardly inspected—or introspected. However, if we cannot begin our reflections with indubitable truths, we must begin with what we find most difficult to doubt. What that is is only eccentrically caught by our private thoughts. For what we gradually realize is that we cannot call into question—all at once, massively, hyperbolically (in the Cartesian manner)—the sheltering practices and beliefs of the society that forms us in forming that aptitude. (We can, of course, call all of our beliefs into question piecemeal.)

None of these admissions commits us to the Cartesian extravagance: either to the cogito or to mind/body dualism; and none obliges us to fall back to a merely "naturalized" account of knowledge and self.

Quite simply, the self's "nature" is essentially its cognizing competence; hence, the self cannot fail to be characterized in terms of its second-order powers. Nevertheless, this simple admission need not reinstate transcendental privilege: the postmodernists (Rorty, especially) are quite wrong to insist that legitimation must be transcendental if it is anything, or that philosophy cannot be philosophy unless it is legitimative in this sense. As a consequence, it cannot be supposed that the "naturalizing" of mind and knowledge is the only viable alternative to transcendentalism or cognitive privilege. That strategy (the "naturalizing") would require that our second-order powers can be analyzed "psychologically" or "causally" or in some other apt first-order way. But, of course, the inseparability of first- and second-order powers may well signify that first-order cognition cannot be entirely naturalized; put more cautiously: it cannot be naturalized if our second-order powers cannot be naturalized.7 "Naturalizing" signifies: (a) retiring second-order inquiries in favor of first-order inquiries and (b) construing explanation as causal or congenial to the causal. (It is an irony, therefore, that Rorty's dismissal of traditional philosophy is no more than the shadow side of the naturalizing preferences of current analytic philosophy.)

Curiously, W. V. Quine does not examine this issue in his famous paper, "Epistemology Naturalized,"8 which set off the contemporary search for such a theory—although he touches on it. Actually, what Quine says is this: "If the [naturalizing] epistemologist's goal is validation of the grounds of empirical science, he defeats his purpose by using psychology or other empirical science in the validation."9 Quine seems to understand the question only in terms of "deducing science from observation,"10 which would be unpromising on any grounds. So he acknowledges a distinctive question of "validation" which cannot (on his own vision) be "naturalized"; but he instantly impoverishes the point of admitting the question so that it cannot be made to accommodate the search for legitimation. Elsewhere, he seems more sanguine but gives no sustained account.11 The "naturalists" move too quickly here. They dismiss legitimation on the mistaken grounds that legitimation would "validate" knowledge by deriving it from foundationally privileged sources. But we now see that to be a caricature of the facts.

On the argument intended, the following triad—cast in a Kantian-like idiom—appears unavoidable: (1) first-order truths without second-order legitimation are "blind"; (2) second-order legitimation without first-order truth is "empty"; and (3) the distinction between first-order

and legitimative discourse is itself a second-order distinction. No serious inquiry can fail to raise and answer questions about the grounds on which it is reasonable (against reasonable challenge) to view its own work as yielding knowledge of the way the world is. Only if first-order discourse were accorded some sort of cognitive privilege—which Gadamer's paradox precludes—could second-order questions be effectively reduced to any first-order questions (hence, naturalized). But there appears to be no compelling naturalized reading of such notions as truth, knowledge, existence, and reality, and no compelling reason for disallowing their analysis. The opponent of a naturalistic epistemology: (i) opposes psychologizing knowledge, (ii) admits that first-order nor-mative distinctions may be naturalized, (iii) is not committed to cognitive privilege, (iv) insists that second-order legitimation is unavoidable, and (v) demands that it be shown how the normative work of legitimation can be naturalized.12

Modernity (in the contemporary sense) complicates these reflections by historicizing thought and human existence. So the "self" must be construed as a history: as the historically formed site of understanding the historically changing processes of understanding. On the foregoing argument, the self cannot be constant, fixed, invariantly structured. The "natural" science that shadows it "observationally" cannot fail to be insufficient; whatever it may utter as its firmest finding will be inextricably bound to the tacit collective habits of the living society from which it emerges and within the implicit terms of which its legitimative reflections will be abstracted. That is why legitimation cannot be naturalized; a fortiori, that is why first-order knowledge cannot be naturalized.

One sees this very clearly, for instance, in Richard Rorty's notorious—philosophically deflationary—use of the distinction between ("ethnocentric") solidarity and ("externalist") objectivity: favoring the first, Rorty pretends to be able to demonstrate the philosophical fatuousness of theorizing about truth and reality; he retires the whole of second-order philosophical inquiry at a stroke, as entirely unnecessary, ineffectual, even paradoxical and self-defeating. Nevertheless, at the still center of his own labor, he declares that "the pragmatist holds the ethnocentric third view [between relativism and realism]," which is to say:

The view that there is nothing to be said about either truth or rationality apart from descriptions of the familiar procedures of justification which a given society —ours—uses in one or another area of inquiry.13

You see therefore the enormous power Rorty invests in his notion of solidarity, his notion of loyalty to a ("our") community's consensual practices. It comes as a shock, therefore, to discover just how arbitrary and how much of an artifice it is, on Rorty's part, to identify "our community." Rorty's "community" is definitely not anything like a Wittgensteinian Lebensform or a Foucauldian episteme. "There are," he says, opening the essay, "Solidarity or Objectivity?,"

two principal ways in which reflective human beings try, by placing their lives in a larger content, to give sense to their lives. The first is by telling the story of their contribution to a community. This community may be the actual historical one in which they live, or another actual one, distant in time or place, or a quite imaginary one, consisting perhaps of a dozen heroes and heroines selected from history or fiction or both. The second way is to describe themselves as standing in immediate relation to a prehuman reality.14

Notice that Rorty thinks of "our" community as no more than a collection of individuals that we, privately, admire or wish to adhere to, that may be entirely fictionally constructed, idiosyncratically projected. He never entertains the question of just how we are able to construct this or that harmonious vision in the first place or how we are ever to know that we are actually adhering to its would-be norms. Elsewhere, he invokes Wittgenstein's philosophical practice, but he never explains the epistemic linkage between his adherence to Wittgenstein and his projection of the conditions of "solidarity"; and (now) he needs to. He cannot supply an answer, I suggest, because, if he did, he would have to implicate (once again) the distinction between first-order and legitimative discourse. It hardly matters, for what we are compelled to understand is that Rorty obscures (by his own invention) the deeper sense in which be cannot possibly extricate himself from the labors of "traditional" philosophy. That is to say: first-order discourse is "blind" if it is not inextricably caught up with the reflections of second-order theorizing (which is itself "empty" when separated from first-order discourse). In short, Rorty compresses the whole of philosophy into the conditions for defining and regulating our "loyalty" to the norms of a community (we have chosen), which he then simply discards—in dismissing philosophy. What I say, therefore, is that Rorty has, in this, produced an intractable self-defeating paradox for his own enormously influential (philosophical) interventions; and that he has inadvertently demonstrated that we cannot escape a similar fate if we go his way. Rorty offers no recognizable grounds for social or epistemic "solidarity."

Interpretation in the largest sense, then, is the power to understand the power to understand. Its principal work is to reconcile, reciprocally, its cognitional and effective aptitudes and what, provisionally, within our shifting histories, we "cannot" doubt are its strong achievements.

Once this vision dawns, we see the presumption and promise of speaking of "objective science"; for knowledge is an artifact—a product—of contingent history just as much as the self is. This is not to relativize the bare meaning of "true," but only to relativize the conditions for the ascription of truth-values.15 "True" cannot be coherently relativized, in that its meaning cannot be given as "true-in-L₁ " (that is, disjunctively, for some language L₁ , or L₂ . . . or L_k ) unless subordinated (for limited reasons) to a use of "true" not thus defined; otherwise, it would generate insuperable (self-referential) paradox.

Most discussions of relativism are rightly concerned with relativizing the (epistemic) conditions of knowledge rather than the (alethic) meaning of "true"; that is, they are unwilling to be restricted to alethic questions only, they go on to epistemic matters. But relativizing the conditions of knowledge (cultural relativity) is not, as such, tantamount to relativism (in any interesting sense). Relativism requires the (alethic) replacement of standard bivalent truth-values by some set of many-valued values, such that: (a) judgments that, on a bivalent logic but not now, would yield contradictories, may be jointly confirmed (call them "incongruent" judgments); and (b) any pertinent epistemic practice would relativize the conditions of knowledge (as by historicism or some admission of diverse Lebensformen ) in accord with (a).16 An alethic relativism entails (as yet) no epistemic thesis, and cultural relativity is not yet a bona fide relativism.

On my own view, the best of the relativistic logics is one that: (1) replaces a bivalent logic with a many-valued logic; (2) acknowledges the compatibility of a bivalent and a many-valued logic, if the set of judgments each ranges over may be suitably segregated from one another; (3) treats truth and falsity asymmetrically, so that, although truth is retired as a truth-predicate within the space of relativistic judgment, judgments may be shown to be false where incompatible with admitted evidential data; (4) introduces truth- or truth-like values like "reasonable," "plausible," "apt," and the like, possibly graded, as truth-values, used therefore not to signify the abdication of truth-claims altogether; (5) concedes, as a consequence, the possible validity of "in-congruent" judgments (as just explained); (6) admits that considerations of consistency, coherence, noncontradiction, and the like obtain

in many-valued logics as they do in bivalent ones (hence, obtain in a way that do not themselves implicate a bivalent logic); (7) construes the rationale for any such policy as legitimated in accord with a theory of the ontic structure of the particular domain to which it applies; and (8) invokes relativism piecemeal rather than globally—for instance, admits the viability of a nonrelativistic defense of relativism.

Thinking, as I say, has, or (better) is, a history. This too is not to make science impossible. On the contrary, it confirms that even science has its interpretive origin and that that "origin" is an ever-changing projection from an ever-changing science that we cannot doubt en bloc.

These are heady speculations, no doubt; but they are becoming more and more insistent. The theory of interpretation searches for a suitable conceptual home. That cannot involve anything slimmer than a guess at the entire human condition. Certain puzzling constraints have now surfaced in a way we can no longer ignore. They come to rest in our having to decide the adequacy of what has been called "naturalistic" (or "naturalized") epistemology—which, invoking Gadamer's paradox, I construe to be the equivalent of a "naturalistic" theory of the self. That, finally, is the clue to the "new puzzle" of interpretation.

I say only that the self cannot be merely a describable denizen of the world, because the linguistic aptitudes of functional societies of human selves play a constituting role in the construction and continual reconstruction of the intelligible world. That is the paradox Gadamer caught sight of—which, for his own purpose, he casts in terms of the paradox of self-understanding and self-transformation. I conclude from this that: (i) naturalistic epistemologies and naturalistic theories of the self are fatally flawed; (ii) all perception, knowledge, understanding are inherently interpretive; and (iii) the paradox of understanding is entirely benign—for it appears and is resolved in terms of the historicized nature of human existence. I repeat, therefore, that understanding what there is in the world, understanding other selves, and self-understanding are inseparable aspects of a single undertaking; that that undertaking is interpretive, historicized, and legitimative; and that any departure from these constraints is a concession to cognitive privilege. That at least is my charge. In a word, symbiosis requires either transcendental powers or a historicized critique of truth-claims. Tertium non datur— if we are not to be skeptics. Kant favors the first; frankly, I favor the second, and the second is a radical replacement for the first.

All this may now explain the contemporary import of David

Hume's notorious candor about having tried to capture the self by introspection, and failing; and, Ludwig Wittgenstein's barely disguised contempt for the vestiges of the Cartesian and Lockeian need for cognitive certainty. Wittgenstein's impatience has been transmuted into an entirely new (naturalized) assurance that any massive run of ordinary beliefs must ("for the most part") simply and surely be true; and Hume's candor has been transformed into an excuse for eliminating all reference to the "self."

For example, Donald Davidson offers the following pertinent pronouncement:

My argument has two parts. First I urge that a correct understanding of the speech, beliefs, desires, intentions and other propositional attitudes of a person leads us to the conclusion that most of a person's beliefs must be true, and so there is a legitimate presumption that any one of them, if it coheres with most of the rest, is true. Then I go on to claim that anyone with thoughts, and so in particular anyone who wonders whether he has any reason to suppose he is generally right about the nature of his environment, must know what a belief is, and how in general beliefs are to be detected and interpreted. These being perfectly general facts we cannot fail to use when we communicate with others, or when we try to communicate with others, or even when we merely think we are communicating with others, there is a pretty strong sense in which we can be said to know that there is a presumption in favor of the overall truthfulness of anyone's beliefs, including our own. So it is bootless for someone to ask for some further reassurance; that can only add to his stock of beliefs. All that is needed is that he recognize that belief is in its nature veridical.17

Given the unguarded generosity of Davidson's remark, it may seem churlish to complain that it suffers from the fallacy of division. It is true "in a sense" that, relative to our survival, "most of a person's beliefs must be true." Taken in a strictly "holist" sense (Davidson's preference)—that is, taken in a sense that yields no particular truths, a sense that cannot function criterially—it is a breach of elementary philosophical logic to conclude " and so there is a legitimative presumption that any one of them . . . is true." It is also doubtful that the optimism can be taken in any other way but in terms of the survival of large populations, certainly not of single individuals. Also, the argument on which the claim rests cannot but be a second-order legitimative argument of some sort, cannot but be unclear and doubtful as it stands, cannot but fail to justify the abandonment of further legitimative concerns.

You may see in Davidson's remark a step in favor of naturalizing

the self and certifying the mass of ordinary beliefs, managed in such a way that, down the argumentative pike, it might eventually become unnecessary to develop a theory of truth or a theory of the self. Well, perhaps. But the argument is missing.18

Davidson, who is distinctly influential on these matters, is also plainspoken about the elimination of any and all interpretive tertia between our ordinary beliefs about the world and the world's independent features. In this, he departs from Quine's appeal to intervening "analytical hypotheses" (which relativize truth-value ascriptions but not truth) and Quine's questionable disjunction (another tertium ) between evidential appeal to direct sensory experience and appeal to socially entrenched beliefs. "Since we can't swear intermediaries to truthfulness," Davidson observes, "we should allow no intermediaries [tertia ] between our beliefs and their objects in the world. Of course [he adds] there are causal intermediaries. What we must guard against are epistemic intermediaries."19 (Quine is of two minds on the matter; and Davidson, opposing Rorty, is unwilling to abandon the theory of truth.20 )

The trouble is this: once we admit that the causal intermediaries (we acknowledge) must also be the correct ones epistemically, we see that the "naturalizing" maneuver (the move against tertia ) cannot stand on its own. It itself presupposes the pertinence of some critical and normative resources, some tertium or other, without yet insinuating any "ultimate source of evidence."21 Our beliefs are true in the realist idiom Davidson prefers, if they are caused by "their [corresponding] objects in the world" in some nonmisinforming way. But what is that way? A failure to answer and an insistence that the mass of "belief is in its nature veridical" are equally committed to an illicit privilege. Davidson has not escaped the need for an interpretive tertium quid. Nothing substantive follows from the conjecture about "most" of anyone's beliefs, whatever constraints of coherence may be favored. The stalemate that results is much more than the upshot of a local quibble. Have patience.

II

It is true that Hume says that "what we call a mind is nothing but a heap or collection of different perceptions, united together by certain relations, and suppos'd, tho' false, to be endow'd with a perfect simplicity and identity"; it is true that he says that human selves "are

nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement."22 But Hume also says (in his deliberately thin idiom) that "all sensations are felt by the mind, such as they really are," "the mind is a kind of theatre, where several perceptions successively make their appearance; pass, re-pass, glide away, and mingle in an infinite variety of postures and situations. There is properly no simplicity in it at one time, nor identity in it at one time, nor identity in different; whatever natural propensity we may have to imagine that simplicity and identity."23 It seems fair to say that what Hume really objects to is the "simplicity" and "identity" of the mind (on his own severe reading of those notions), not the ineliminability of mind or self, or the mind's cognitional or active function in undergoing experience, or even the qualia of mental life. Certainly, there is no eliminationist argument in Hume's account.

Kant's fascination with Hume undoubtedly centers on what we should treat as the work of the "theatre" of the mind. Hume is obviously a pre-Kantian whom Kant thought had anticipated something of his own discovery. But even if that were not so, contemporary philosophers cannot, without defense, favor pre-Kantian philosophy in a pre-Kantian manner. They are bound to come to terms with (1) the symbiotized world and (2) the historicity of thinking. Read thus, Hume's candid reliance on empiricist considerations and his scrupulous rejection of de re necessities legitimate the avoidance of neither (1) nor (2). The upshot, I claim, is that the would-be naturalizing of epistemology and selves cannot be strengthened by anything like Hume's program (which we may conjecture, lies—inspirationally at least—behind Quine's finding). The paradigms for such accommodation (however primitive) are already supplied in the early work of Russell and Moore (which similarly owed much to Hume though they are not specifically Humean).24

The fact is that, in the first two decades of the twentieth century, Russell and Moore plainly defined the main lines of the new "analytic" philosophy that would preclude or marginalize the need for interpretive tertia. The marks of this "externalism" (directed, remember, largely against idealism and British versions of Hegelianism) include a tendency to favor: (a) direct perception, foundational beliefs, and cognitive privilege; (b) the epistemic neutrality or irrelevance of context; (c) the correspondence theory of truth or congenial surrogates (disquotation, for instance); (d) the eliminability of definite descriptions, names,

and indexicals of all sorts and their replaceability by predicative devices;25 (e) a constructivist view of commonsense, macroscopic, or "folk-theoretic" entities in terms of sense data, sensibilia, universals, syncategorematic functions, and the like, that do not themselves require interpretation; (f) a neglect of intentional and Intentional complications; (g) a disregard for sustained analyses of the nature of cognizing selves; and (h) a rejection of transcendental necessity.

Such has been Hume's immense influence, however, that many Anglo-American philosophers, struck by the famous passage in the Treatise in which Hume confesses that "when I enter most intimately into what I call myself, I always stumble onto some particular perception or other," wrongly take the words to signify that Hume denies that there is a self or mind.26 Many go on to affirm in our own time that the "self" is a delusive shadow cast by some sort of "folk" idiom that a thoroughgoing physicalism will in time erase. They fail to see that an answer to the empiricist question cannot possibly be an answer to the second-order question of the import of relying solely on empiricist devices. (My argument is meant to be metonymic here.)

I am suggesting, in short, that the "naturalizing" turn is a retreat, in post-Kantian times, to a pre-Kantian confidence via Hume or similar spirits, made plausible by the refusal (preeminently, Hume's) to admit de re necessities. In fact, Moore's famous paper, "Proof of an External World," sketched for the first time the principal strategy by which Kant's critical philosophy was, in the new Anglo-American analytic tradition, to be reinterpreted in pre-Kantian terms—which effectively signified the marginalizing of the self in epistemic contexts and, coordinately, the entrenchment of externalism.27

This entire line of argument, however, remains in limbo, once it is grasped that legitimative questions do not (though they do indeed in Kant and Husserl) require transcendental answers of an apodictic sort.

Even without Hume, the issue rests with the fate of all those strategies that provide for the "reduction" and "elimination" of the self. Paul Churchland, for example, straightforwardly declares that "since folk psychology is an empirical theory, it is at least an abstract possibility that its principles are radically false and that its ontology is an illusion." Churchland supposes that reductionists are simply too sanguine about their own theories: "Folk psychology is a radically inadequate account of our internal activities, [he says,] too confused and too defective to win survival through intertheoretic reduction . . . it

will simply be displaced by a better theory of those activities."28 But Churchland ignores two elementary difficulties: one, that the theory of the self cannot be regarded as a mere "empirical [first-order] theory," straightforwardly capable of being found to be "radically false" (though it is certainly open to change); the other, that his own theory about the self's eliminability implicates a legitimative function (of the self) in a way that the first does not touch on at all and that he sees no need to justify.

No one has ever been able to explain how to treat "folk psychology [as] an empirical theory," falsifiable in the first-order terms Church-land intends, while, at the same time, supposing that viewing matters thus does not commit us to a vicious regress (in terms of the self's constitutive or legitimative function). For example, no one has ever shown that the devices of linguistic reference and predication and the use of language to individuate and identify and reidentify whatever there is in the world do not already implicate the "folk psychology" Churchland finds so tiresome. If so, then the eliminationist's claim may be impossible to defend. The charge is consequential.

Furthermore, in acknowledging the collective sources of the linguistic aptitudes that form and constitute the self, the issue cannot be confined to the reality of individual (solipsistic) selves: it extends to the whole of human culture. The discursive and legitimative powers of the self (the very powers Churchland invokes) confirm its emergence in accord with the collective practices it internalizes. Once that is admitted, first-order discourse cannot be disjoined from second-order powers. Under the conditions collected a moment ago as (1) and (2)—that is, symbiosis and historicity—there is no way to preclude interpretive tertia. So you see the direct relevance of the theory of the self for interpretation. The seeming detour is simply a Peircean geodesic. Peirce's suggestion has the advantage of preserving the effectiveness of our reference to the interpreting and self-interpreting sites we call selves at the same time it treats the self's "nature" as maximally open to self-transformation through collective resources.

These spare reflections (I can afford no more) give considerable precision to the notion of the self. Provisionally, the self may be characterized in terms at least of: (1) existence, (2) cultural emergence, (3) membership in a society sharing a natural language, (4) self-understanding and other cognitive powers, (5) the historical preformation of its conceptual competence, (6) the collective source of its "nature," and (7) the

causal effectiveness of its initiatives among similarly formed selves. These items constitute a distinctly plausible set of constraints regarding the self's mode of functioning. At present, philosophy offers not the slightest inkling of how any of these may be compellingly denied, or, if admitted, admitted without also acknowledging the Intentional, second-order complexities adduced. Hence, items (1)-(7) cannot be analyzed in mental or biological terms alone. Selves, I say, are cultural artifacts: self-interpreting "texts." The puzzle of interpretation cannot be confined to the vagaries of the mental or the processes of the brain.29 But if it cannot, then "naturalizing" fails.

Churchland's maneuver, therefore, draws attention to an unlikely conceptual connection. Once the "folk" conception is seen to include the self's legitimative competence, one sees that that conception depends on the fate of the theory of truth, and vice versa. If, say, the predicative use of "true" cannot be naturalized along the lines of Davidson's "holism,"30 it follows at once that the self cannot be a "natural-kind" entity.

If all this be conceded, then selves have no "natures," have (or are) only histories. I take Davidson's opposition to tertia, therefore, to signify that there are no serious philosophical questions about truth (or selves)—which, in his own view, is not to deny the philosophical importance of the concept of truth. But there remains a difficulty in Davidson's account that betrays his naturalizing maneuver. He offers us an account of knowledge (not a "theory") "that [he says] would make interpretation possible," would answer the question "what we do know that enables us to interpret the words of others."31 (In speaking of "interpretation" here, Davidson means to resolve the problem of other minds, without admitting interpretive tertia at all.) His view is this:

What makes interpretation possible . . . is the fact that we can dismiss a priori the chance of massive error [in belief]. A theory of interpretation cannot be correct that makes a man assent to very many false sentences: it must generally be the case that a sentence is true when a speaker holds it to be. So far as it goes, it is in favor of a method of interpretation that it counts a sentence true just when speakers hold it to be true. But of course, the speaker may be wrong; and so may the interpreter. So in the end what must be counted in favor of a method of interpretation is that it puts the interpreter in general agreement with the speaker: according to the method, the speaker holds a sentence true under specified conditions, and these conditions obtain, in the opinion of the interpreter, just when the speaker holds the sentence to be true.32

I should, in all fairness, say that, just before making this remark, Davidson had warned that the conditions of intelligibility by which we might identify a false belief among the ancients (for instance, the belief that the earth is flat) "must depend on a background of largely unmentioned and unquestioned true beliefs." Hence, he claimed, no single mistaken belief "necessarily destroys our ability to identify further [presumptively true] beliefs."33

Certainly this seems reasonable. Still, in the central account of "radical interpretation," Davidson offers an altogether different (less doubtful) reading, namely:

[The] method [of devising a theory of truth for an unknown native tongue] is intended to solve the problem of the interdependence of belief and meaning by holding belief constant as far as possible while solving for meaning. This is accomplished by assigning truth conditions to alien sentences that make native speakers fight when plausibly possible, according, of course, to our own view of what is right. What justifies the procedure is the fact that disagreement and agreement alike are intelligible only against a background of massive agreement. Applied to language, this principle reads: the more sentences we conspire to accept or reject (whether or not through a medium of interpretation) the better we understand the rest, whether or not we agree about them.34

This second account is not the equivalent of the first; it undermines it. For, in the first account, Davidson claims that "massive error" in belief must be impossible; whereas, in the second, all that is needed is the operative policy that we assume that there is "massive agreement" (among the "natives")—relative to what we believe to be true. The second account actually admits an interpretive tertium, whereas the first does not. The first rejects all tertia, retreating to a form of "externalist" certitude.

Notice that, on Quine's view—which Davidson has in mind—the problem of intersocietal agreement and interpretation is the same as that of intrasocietal agreement and interpretation.35 Hence, Davidson cannot fall back to the impossibility of "massive error" (among the natives). Put another way: translational and "interpretive" endeavors (of Davidson's sort) cannot solve the question of what is true: they presuppose a favorable answer. The second reading of radical interpretation does not "naturalize" epistemology.

The force of this distinction may not be immediately evident. Davidson means to use his argument in defending his application of Tarski's semantic conception of truth to natural-languages ("Convention T").36

But he does not reckon with the missing step—the step by which to go from the second reading to the first. Thus, the following formulation, which makes sense in terms of the first reading, fails to do so in terms of the second:

Knowing that [a speaker] holds the sentence [uttered] to be true, and knowing the meaning, we can infer his belief; given enough information about his beliefs, we could perhaps infer the meaning. But radical interpretation should rest on evidence that does not assume knowledge of meanings or detailed knowledge of beliefs.37

The reason the strategy fails—and with it the important assumption that Tarskian "T sentences" may be used to "interpret" the sentences of a natural language—is this: "massive agreement" in beliefs entails that the beliefs of the speakers conform with what we believe to be true. Hence, contrary to what Davidson requires, "radical translation" cannot be free of "knowledge of meanings or detailed knowledge of beliefs." Q.E.D. (Incidentally, the expression "radical interpretation" is meant to preclude "interpretation" in the ordinary sense.)

This leads to some important conclusions: first, tertia have not been eliminated in Davidson's program of radical translation; second, there is no clear prospect that they can ever be; third, unless they can be avoided, radical translation cannot be counted on to naturalize epistemology; fourth, on that concession, "radical interpretation" falls back to ordinary interpretation; fifth, a theory of truth can claim no priority or precedence with respect to interpretation; and sixth, a theory of truth (or interpretation) is tantamount to a theory of the self's legitimative competence. Davidson intended to treat "a theory of truth [his generalized treatment of Tarski's account] modified to apply to a natural language as a theory of interpretation."38 But that now proves impossible: not merely because Davidson fails to make his own case, but (more to the point) because no plausible strategy suggests itself by which to ensure a "massive" body of truths in terms of which to weaken the importance of both the theory of truth and the theory of interpretation.39

Richard Rorty favors a more radical solution in the light of having grasped the vulnerability of Davidson's project. Espousing something close to Quine's holism (so as to disallow legitimative questions), Rorty holds that the theory of truth no longer has any use: "'True' has no explanatory uses," he says; explanatory uses are all concerned with causal connections, and truth has no such connections; furthermore,

other would-be kinds of explanation ("normative" ones, in particular40 ) cannot save truth, since they are not causal either.41 But this is unsatisfactory. The argument is not compelling.42

It is one thing to say that truth has no explanatory use in the sense that it affords no causal explanation and quite another to say that truth need never be itself explained, or that explanation is always and only causal. Truth, after all, is not a natural-kind term. It is at the very least (as Tarski clearly acknowledges) a metalinguistic attribute; as such, it cannot fail to require an explanation of its proper function. Furthermore, if truth were explicated along the lines Davidson favors (which Rorty construes in accord with his own opposition to epistemological questions), then the causal question would itself inexorably generate a deeper legitimative question of its own. And then, "true" would have to have an explanatory use, in the sense that causal explanations would invite and require a legitimation of their explanatory function. Call that an "explanatory use" or not, as you please, but it cannot fail to count very heavily against treating truth as "basic," as not needing to be explained. It would count, for instance, against Davidson's well-known proposal (in "Radical Interpretation"). Distinguishing his own position from Tarski's, Davidson there offers the following conceptual strategy:

In Tarski's work, T-sentences are taken to be true because the right branch of the biconditional is assumed to be a translation of the sentence truth conditions for which are being given. [A T-sentence is a sentence of the form "s is true (in the object language) if and only if p," where that sentence is "obtained by replacing ' s ' by a canonical description of s, and ' p ' by a translation of s." The "canonical" description is a metalinguistic description, the metalanguage is assumed m include the object language, and its logical properties are assumed to accord with the extensional constraints imposed by the would-be canonical description.] But we cannot assume in advance that correct translation can be recognized without preempting the point of radical interpretation [that is, judging speakers' intentions and the like in declaring their beliefs]; in empirical applications, we must abandon the assumption. What I propose is m reverse the direction of explanation: assuming translation, Tarski was able to define truth; the present idea is to take truth as basic and to extract an account of translation or interpretation. The advantages, from the point of view of radical interpretation, are obvious. Truth is a single property which attaches, or fails to attach, to utterances, while each utterance has its own interpretation; and truth is more apt to connect with fairly simple attitudes of speakers.43

This is meant to dovetail, of course, with the thesis (his own) regarding the truth of beliefs "for the most part." But if the latter fails, then

there cannot be any way of vindicating the proposal now before us. Remember: these are maneuvers designed to "naturalize" knowledge; hence, to minimize the complexity of the nature and function of the self. (I assure you, we are close to the end of this harangue.)

The essential point is this: on the argument, there can be no principled disjunction between the theory of truth, the theory of meaning, the theory of knowledge, the theory of thought and belief, the theory of linguistic competence, and the theory of the self. No logical priority can be convincingly assigned any of these concerns over any of the others. I conclude, therefore, that the problem of interpretation infects the whole of man's intelligible world in a uniform way.

Rorty grasps the plain fact that "true" and "false" have no use as yet in the context of holism; savoring that, he hurries to conclude that they have no philosophical use at all! What Rorty must have realized is that Davidson's adherence to Tarski's semantic conception of truth— Davidson's extension of Tarski's conception to the whole of natural language, well beyond anything Tarski would or could ever have countenanced—was already a relic of an earlier era in which holism had not yet won the day; for the Tarskian conception is not philosophically neutral (as Davidson seems to suggest): it is rather a vigorous defense of an exceptionless extensionalism (within a restricted portion, to be sure, of a suitably formal language).44 Applied even more boldly to the whole world in Davidson's way, it would mean that every science and every inquiry must be congenial, everywhere, to that same extensional-ism! Here, Davidson simply endorses Quine's view in Tarski's name, but he offers no supporting argument;45 and Rorty erases the telltale inconsistency between that thesis and Davidson's would-be holism. Hence, Rorty "naturalizes epistemology" simply by retiring the philosophical role of truth altogether. But he overlooks Davidson's commitment to Tarski and his unwillingness to abandon truth.46 That line of thinking cannot, as we have seen, escape legitimative queries—and tertia.

Rorty argues, against legitimative theories of truth, that "the lines of evidential force . . . do not parallel the lines of referential direction"47 : the "justificatory story" and the "causal story" (of science and knowledge) do not mesh. This is why (Rorty suggests) "philosophers" suppose they need some tertium quid (Quine's faulty device, for instance: "analytical hypotheses") in order to bridge the gap between language and world. Rorty supposes that Davidson will not traffic in such devices, because he offers the theory of "radical interpretation."48 But Rorty fails to reckon with Davidson's use of Tarski's theory (which

is another tertium quid ). He also wrongly insinuates that all interpretive tertia must illicitly breach the original holism. That is surely false. Remember: a holism is simply a generic form of cognitive intransparency. Hence, some tertium cannot be avoided—although, contrary to what Rorty and Davidson apparently believe, tertia can acknowledge the weight of the encompassing holism without transcendental flights. Legitimation entails neither the fixity of the world nor any cognitive privilege.

These are vexed matters. Let me isolate the most important findings for our present purpose. First of all, all discourse about the beliefs of the members of another society is (contrary to Davidson's "Radical Interpretation") subject to interpretive tertia, simply because they are read entirely in conformity with the terms of intelligibility of our own beliefs. Second, determining the truth and falsity of our own beliefs is (contrary to Davidson's "A Coherence Theory of Truth and Knowledge") subject to interpretive tertia, simply because no argument from holistic assumptions to distributed truth-claims can be anything but fallacious, because Tarski's theory of truth is itself interpretive, and because legitimative questions about truth cannot be retired or ignored. Furthermore, for any complex community—for any community in which there are profound disputes about the truth of particular claims ("our" community, in Rorty's phrase)—the epistemic "normalizing" of a "community's" boundaries is itself an interpretive artifact. The upshot is: (1) knowledge is inherently interpretive; (2) interpreting beliefs so that they count as knowledge is inherently legitimative; hence (3), neither the self nor knowledge can be satisfactorily "naturalized."

III

I have been trying, you remember, to remove the objections, and to answer the counterproposals, of those current analytic philosophies that would disallow the epistemic role of history, context, intransparency, Intentionality, interpretation, horizon, prejudice, radical contingency, incommensurability, the nonnaturalized structure of self and culture. I claim that holism entails the "Gadamerian" paradox. You have to think of the factors of human life that confirm that finding irresistibly. Let me offer a specimen—it collects in the blink of an eye the entire force of my objection to "naturalizing." In a survey of the distinctive features of the human sciences, Alasdair MacIntyre neatly observes: "The notion of the prediction of radical conceptual innova-

tion is itself conceptually incoherent."49 Of course. That single fact cannot be managed, at any time, by a naturalized view of explanation; and, there is no plausible way to deny or ignore it.

If, now, you review the trajectory of the argument up to this point, you will see that it has proceeded by way of an orderly declension: from invariance to flux, from privilege to intransparency, from the naturalizing to the legitimative, from the extensional to the Intentional, from correspondence to symbiosis, from language as representation to language as praxis, from bivalence to a many-valued logic, from the independently observed to the consensually constituted, from the objectivist to the historicized. I must now bring the choice between these options to bear on the narrower concerns of the theory of interpretation. The argument has been entirely too global.

I do believe I have stalemated the strategies for "naturalizing" epistemology and for eliminating or "naturalizing" the self. But I must still identify a suitable "analyst" prepared to "naturalize" interpretation in the context of the arts and history.

I trust the general lines of my own proposal are clear enough. Let me gather them quickly, however, so that I can bring this argument to a proper close. I subscribe to the following doctrines—and I take it that the argument in their favor shows that a naturalized account of interpretation must fail and must produce a paradox in linking the theory of interpretation with the theory of the self: first, the cultural world is, intrinsically, Intentionally complex; second, the cultural world has "emerged" (in some way) from the physical and biological world; third, its reality is entailed by the reality of "selves," themselves culturally emergent and empowered; fourth, reference to, and predications made of, cultural phenomena indissolubly implicate the (non-Intentional) phenomena of physical and biological reality; fifth, cultural attributes and entities are consensually discerned by selves apt for doing so, without our ever being restricted to imputing (only extrinsically) Intentional import to physical phenomena; sixth, all conceptual distinctions, even those regarding the physical world (from which the cultural is thought to have emerged), are artifacts posited by cognizing selves functioning at the cultural level at which they first emerge as the competent agents they are; seventh, cognizing selves and cognized world are symbiotically linked and intransparent; and, eighth, selves and their cognizing powers are historically preformed in lebensformlich ways and alterable under the conditions of history.50

As a result, there is no insuperable paradox generated by the paired

intuitions: (a) that human life and human culture have evolved from a lifeless physical world, and (b) that the Intentional complexity of the human world cannot be explained in terms solely of the attributes and causal powers ascribed to whatever may have been its original physical source. The resolution of the seeming paradox lies with the obvious fact that (on the argument) that continuum is itself a posit made at the level at which inquiry first obtains and only in terms of the conceptual options accessible at that level.

I see no reason why it should not be true that science entails the tacit, endogenous limitations of reflection at the culturally emergent level. It is entirely possible that, effectively (not necessarily), we cannot fathom how we have evolved as the creatures we are. After all, we have no real understanding of how language has arisen. Hence, on the argument intended, it is not at all necessary—it is a mere vestige of the "naturalizing" strategy—to suppose that cultural phenomena must ultimately be explained in physical and biological terms. On the contrary, it is enough that they should be explained in culturally adequated terms, all the while we agree that the cultural is embodied or incarnate in the physical.

The contest I have in mind therefore is just the one I introduced earlier regarding the different strategies of "emergence" and "supervenience." Hence, the conceptual objection I mean to bring against the supervenience option (or, indeed, against any other "externalist" option that begins with the priority of the physical, even if it neglects or abandons strict supervenience—say, along Davidson's lines) is simply this: (1) supervenience cannot justify avoiding a realism of history and art if it cannot (or does not) account for such an avoidance, while at the same time admitting the real existence of selves; (2) there is no principled difference between a realism of emergent selves and a realism of their environing world; and (3) the interpretation of history and art and the reflexive interpretations of selves and their world are inseparable parts of the same undertaking. I have already argued in favor of this thesis; and I have of course associated it with Gadamer's paradox.

As it happens, there is only one reasonably explicit theory of the "naturalizing" or "externalist" sort to be found in the analytic literature regarding the interpretation of the arts and history, namely, Arthur Danto's.51 It employs a sort of methodological dualism deliberately weaker than Dilthey's division between the Naturwissenschaften and the Geisteswissenschaften; for Dilthey's account rests squarely on the supposed real differences between physical nature and human

culture. There is no developed metaphysics in Danto's account of history and art—which you may well doubt—but that is indeed the conditio sine qua non on which Danto's intended reconciliation of nature and culture depends. Dilthey, too, as Gadamer has neatly observed, constructed his account of the human studies by way of assuming, first, the independent rigor of the physical sciences.52 Dilthey fails, therefore, to incorporate (before the fact) what is genuinely radical in Gadamer's insistence on the universality of the hermeneutic problem.53 (The same criticism applies to the views of Jürgen Habermas.54 ) We are glimpsing, here, the European counterpart of the American penchant for giving pride of place to those salient features of the physical and formal sciences already remarked in Hempel, Tarski, Quine, Davidson, and Rorty. (In the last chapter, I broached the issue in the context of reviewing Ricoeur's and White's theories of history.)

If, now, we turn to Danto's theory, it would not be amiss to say (as a first pass) that his account of history is broadly congruent with Hempel's externalism, though not with the details of Hempel's physicalism: for the actual work of history and interpretation is (quite separately) conceived by Danto, with great skill, to accord with a deliberately restricted rhetoric (not a metaphysics) favoring a version of Romantic hermeneutics.

Danto's theory of art history is explicitly Hegelian, but deliberately cast in rhetorical terms. His theory of the interpretation of art is strenuously and explicitly antimetaphysical. Danto believes the interpretation of artworks is bound by norms of objective truth and correctness; yet his sense of the inadequacy of a purely physicalist language formed for interpreting the cultural world is never specified in realist terms. This is not easily perceived. There is, I claim, an aporia here.

Danto's analysis of action clearly draws him in the direction of a strong physicalism; but his disciplined discussions of art and history never rely on any explicit physicalism, and what he offers in interpretive terms is treated only rhetorically—that is, as a culturally entrenched idiom that lacks any realist presumption of its own. This is the proper meaning of Danto's insistence that interpretation "transfigures" but does not "transform" "commonplace" objects or "mere real things" as artworks. I have already mentioned the distinction. Here, it is helpful to see that the point of Danto's insisting on it is to yield in the direction of "supervenience" (supervenience manqué ). The result is a kind of free-floating discourse: knowledgeable, to be sure, always

perceptive, but never explicitly grounded in terms of truth or objectivity. ("Rhetorical," by the way, is Danto's own choice of epithet.)

You will find that Danto's Analytical Theory of Action offers his most sustained attempt to bring the description and interpretation of cultural phenomena (actions, primarily) into accord with physicalist themes.55 More recently, however, speaking directly about the interpretation of art, Danto quotes himself from the Action volume: he mentions for instance Giotto's tableaux in the Arena chapel in Padua (representing episodes in Christ's career) in a way that unmistakably signals the extension of the physicalist model to art and history. He concludes there, after raising the helpful but false lead of Wittgenstein's famous question about what is "left over when, from the fact that you raise your arm, you subtract the fact that your arm goes up," that:

The difference between a basic action and a mere bodily movement [roughly: Danto's reading of Wittgenstein's contrast] is paralleled in many ways by the difference between an artwork and a mere real thing, and the subtractionist query may be matched with another one here, which asks what is left over when we subtract the red square from "Red Square."56

The reference to Red Square is a reference to what is very nearly Danto's signature in the philosophy of art. He offers, just before the comment cited, the conceptually amusing prospect of six red canvases indistinguishable from one another "perceptually" but still six different referents: some, genuine artworks; some, not artworks at all ("mere real things"). One, he says, might be a "painting of the Israelites crossing the Red Sea," by Sf ren Kierkegaard; another, also by a Dane, might be titled Kierkegaard's Mood; another, "a clever bit of Moscow landscape," is titled Red Square; another, with the same title, is "a minimalist exemplar of geometrical art"; and so on until we come to "a surface painted, though not grounded, in red lead: a mere artifact I [Danto] exhibit as something whose philosophical interest consists solely in the fact that it is not a work of art and [whose] only art-historical interest is the fact that we are considering it at all: it is just a thing, with paint upon it."57

The fact remains that physicalism cannot pertinently solve the problem of individuating these paintings (as distinct from the canvases), and the interpretive comments about the different paintings cannot rightly be characterized as objective apart from a (second-order) theory about

what belongs to the Intentional "nature" of the paintings rather than to the canvases. So the philosophy (the physicalism) and the art criticism (the somewhat Romantic hermeneutics) are not satisfactorily matched. But if they are not, then Danto cannot possibly (as it is clear he supposes he can) have any principled reason for advancing his own view over an alternative ontology of history or interpretation. He simply does not address the issue; and yet, his own referential and predicative remarks presuppose an objective policy.

Danto's solution to the indiscernibility puzzle—he mentions approvingly Borges's little story of Pierre Menard—argues, first, that artworks are not "mere real things" even if they are perceptually indistinguishable from them and even if, as in the case of Duchamp's Fountain, a particular "work . . . might have been elevated from a mere real thing to an artwork."58 (The expression "elevated" is clearly Sartrean.) "Real things" or "mere real things" are, for Danto, ordinary "material objects" or ordinary "artifacts" (snow shovels and diagrams).59 So the sense of these remarks is that "being an artwork" is a kind of (rhetorical) status real things have or acquire; that they ("real things") do not change, as far as their being real or possessing numerical identity is concerned, when they acquire that status; but also that to have such status is not merely to be a "mere real thing." (They are metaphorically "transfigured.") In fact, Danto clearly invokes what Davidson and Rorty would regard as an objectionable tertium (the "theory" of the artworld).

Let me make Danto's paradox a little more evident. Danto does successfully individuate his "indiscernibly" different paintings, not just the canvases. But the paintings are nowhere to be found until the canvases are "transfigured": in fact, once the canvases are "transfigured," the paintings become discernibly different! There can be no insuperable "perceptual" confusion among the red paintings though there will certainly be the threat of "perceptual" confusion among the "mere" red canvases. I submit that Danto must be equivocating in his use of the term "perception": restricting "perception" in the minimal sense, sensorily, we must admit that confusion may arise among the canvases; but then, in that sense, there are literally no paintings to be confused with one another. Allow a sense of perception robust enough to admit the perception of paintings: the indiscernibility problem trivially dissolves. So Danto's famous puzzle is a complete nonstarter.

Furthermore, Danto argues that artworks implicate artists' inten-

tions and the history of their actual creation. In short, artworks are discernible as such only in the space of an "artworld":

To see something as art requires something the eye cannot decry—an atmosphere of artistic theory, a knowledge of the history of art: an artworld.60

The formulation (the italicized terms) clearly shows Danto's disinclination to move in the direction of reductive physicalism. By his maneuver (his rhetorical turn), what "is" an artwork is and remains (numerically identical with) the "mere real thing" one learns to "decry" as art. Danto goes on to say: "To mistake an artwork for a real object is no great feat when the artwork is the real object one mistakes it for."61 Hence, he concludes, the mimetic theory of art is false.

"The artworld," Danto explains, "stands to the real world in something like the relationship in which the City of God stands to the Earthly City. [He means the ordinary earthly world.] Certain objects, like certain individuals, enjoy a double citizenship, but there remains, the RT notwithstanding, a fundamental contrast between artworks and mere objects."62 (By the "RT," Danto means the theory, advanced by Roger Fry, among others, against the "imitation theory" which encouraged a certain sort of illusionism. RT maintains that artists create real things, not imitations of real things.) Danto clearly enters the metaphysical debate about art to this extent: he opposes the metaphysics of the "RT"; but he avoids offering any positive theory of his own. The result is that his position remains theoretically untethered, however much we may admire this or that interpretive aperçu. Nevertheless, since he favors the objectivity of Romantic criticism (in favoring the authority of the original artist's intention), he cannot rightly ignore the metaphysics of art (for he is unwilling to ignore altogether the metaphysics of selves). There is no consistent resolution of this aporia.

What is dialectically important is that Danto opposes "RT," refuses to collapse artworks into "mere real things." He offers no "metaphysics" of art. Nevertheless, he insists on the objectivity and conditions of correctness of interpretation. There is a serious lacuna there.

For Danto, artworks are not independent, real referents. They are identified as such and referred to only "by the 'is' of artistic identification" when applied to "mere real things." That, I say, is a form of the "supervenience" theory (manqué). We are warned not to confuse the "is" of artistic identification with the "is" of numerical identity, but we are never told what the difference is. The importance of the caveat

is this: Danto admits the objectivity of history and interpretation but resists any form of cultural realism. Hence, though we may admire Danto's interpretive skills, we are deprived of any theoretical or evidentiary (or legitimative) grounding for judging whether this or that interpretation is true.

Recently, Danto has explained his position somewhat more fully, in opposing Susan Sontag's theory of interpretation. What he offers in this regard is the clearest formulation he has yet supplied:

My theory of interpretation [he says] is instead [that is, as opposed to Sontag's "explanatory" account] constitutive, for an object is an artwork at all only in a somewhat logical way. Interpretation in my sense is transfigurative. It transforms objects into works of art and depends upon the "is" of artistic identification. . .. Mine is a theory which is not in the spirit of science [that is, is not explanatory, as by the use of psychoanalytic or semiotic codes: Sontag's preference] but of philosophy. If interpretations are what constitute works, there are no works without them and works are misconstituted when interpretation is wrong. And knowing the artist's interpretation is in effect identifying what he or she has made. The interpretation is not something outside the work: work and interpretation arise together in aesthetic consciousness. An interpretation is inseparable from work, it is inseparable from the artist if it is the artist's work. . .. The possible interpretations [of artworks] are constrained by the artist's location in the world, by when and where he lived, by what experiences he could have had. . .. here is a truth to interpretation and a stability to works of art which are not relative at all.63

There is a very strong statement favoring interpretive objectivity. But it is still an obiter dictum (and it clearly invokes tertia ). It fits certain familiar convictions, to be sure (those, roughly, of the Romantic hermeneut), and it does indicate that there is an objective order of some sort involving artists' intentions and the history of art's creation. But it does not supply an explicit argument, and it does not explain the logical relationship between the "is" of "artistic identification" and the "is" of numerical identity. How, we may ask, can objectivity be secured if the world of human culture is not real? How are we to construe the stability and public discernibility of the peculiar "properties" of artworks?

Danto offers no answer, and I believe there can be no answer consistent with his thesis that Intentionally complex artworks and historical acts (open as such to interpretation) are numerically identical with "mere real things" and "basic actions" or mere physical movements (that cannot as such support objective interpretations). In a word: in

order to make the "transfigurative" reading of artworks operative in interpretive terms, Danto would need to deny that selves and human history are also "transfigured"; but if he acknowledged that they are not transfigured, there would then be no basis and no need for the transfigurative theory of art. Alternatively, if he held history to be trans-figurative as well (which, in effect, he does), then he could not escape a vicious regress. I take this to be a reductio: he offers no basis at all for his own referential and predicative acts regarding artworks and history.

IV

Read metonymically, the defeat of Danto's theory of history leads to the heart of an extraordinary issue. I suggest that we fix our sense of the contest by something like the following illustration. Consider the recent events in (the former) Yugoslavia that have pitted the Croats and Muslims and Serbs against one another. One theorist, Danto, might say that the full meaning of the original division between the Eastern and Western Roman empires, or the later expansion of the Ottoman Empire into southern Europe, or the nazification of Croatia, can only have been learned by reviewing the further external relationship between those original events and the hostilities of the summer and fall of 1991 and their aftermath in the context of the collapse of the Soviet empire and similar events. Others, ourselves, might say instead that the meaning of the original break between the Eastern and Western empires is itself affected and altered by our present interpretation of the emerging hostilities between the Bosnians and the Serbs—possibly also our interpretation of the recent Gulf War and the developing import of the relationship between the Christian West and the Muslim East. (I remind you of my discussion of Matisse and Cézanne.)

The important point is that both conceptions of how historical interpretation works are coherent on their face; they are also radically opposed in practice and in what they take to be the historical past. Without a theory of the nature of real history, it would be impossible for Danto to legitimate his own picture of the historian's work or to disqualify the other. Furthermore, on a favorable view of the second, even the practice of the first might be allowed: by viewing the first as a restriction, for certain purposes, of the full practice of the second. This would provide a straightforward analogue of what, at the very beginning of this study, I offered as a way of reconciling Barthes's

distinction between "readerly" and "writerly" reading. It also accords with the declension of the paired options I mentioned a little while ago: that is, supposing that the fixity of the historical past may be treated as a special constraint placed on its being (otherwise) alterable by being interpreted; whereas the reverse is not possible.

There you have the full puzzle of interpretation. For, should we say that the original meanings of historical events are fixed once and for all at the time of the original occurrence of those events, although their complete meanings are never rightly exhausted by any such first occurrence; or should we say that the meanings of any and all historical events are first constituted from the vantage of some present interpretation, and, rightly, therefore, reconstituted (that is, the meaning of the "original past" and the past) through an unending series of future presents? Should historical meanings be treated as ampliative, fixed originally and then added to only by external, relational accretions through time; or should they be treated always as monadic but never fixed at any time, consensually constructed at a given time and open to reconstruction at later times? These options yield very different accounts of historical realism. They also encourage very different views about the objectivity and realism of the interpretation of artworks. The two issues go hand in hand. (At the risk of an intrusion, I suggest that Peirce's theory of the self-transformative potencies of the self is completely separable from his parochial adherence to nineteenth-century evolutionism. Simplifying matters in this way, we may take Peirce to have [implicitly] shown Gadamer how to reconcile "the fusion of horizons"—which has no telic pretensions at all—with a disciplined treatment of interpretive truth-claims. The absence of such a connection is the principal weakness of Gadamer's account, just as Peirce's relic optimism is the inadmissible weakness of his.)

We now see why Danto calls his theory of history "philosophical": he means it to accord with his version of the Hegelian view; but he means it "rhetorically" rather than "metaphysically."

There are two principal foci in Danto's theory of history. The first has already been broached:

[T]here is [he says] a class of descriptions of any event [the Thirty Years War, for instance] under which the event cannot be witnessed, and these descriptions are necessarily and systematically excluded from the I.C. [the "Ideal Chronicler," the one who could give "a full description of an event E": "a set of sequences which, taken together, state absolutely everything that happened in E "]. The whole truth concerning an event can only be known after, and sometimes only long after the event itself has taken place,

and this part of the story historians alone can tell. It is something even the best sort of witness cannot know. What we deliberately neglected to equip the Ideal Chronicler with was knowledge of the future.64

This signifies that part of the pertinent description of historical events (Danto does not speak, here, of historical interpretation) is rightly fixed at the time of the original event. It acquires additional properties—or meanings—in virtue of new relationships that the original (intrinsically fixed) event enters into with later events.

The alternative picture of historical interpretation (the one I favor) holds, instead, that historical events have no "natures," are not natural-kind phenomena at all, cannot be straightforwardly described (by analogy with the description of physical nature), and must be admitted to possess Intentional properties that cannot be fixed in the way we suppose their physical past may be fixed. Put another way: Danto has no philosophical reason at all for insisting on fixing the meaning of historical events: if he had, he would have a metaphysics of history; and if he had that, he would have a metaphysics of art and its interpretable properties.

No anarchy or irresponsibility need follow from the second picture: for, interpretation will still accord with the consensual practice and tolerance of historically viable societies. (The parallel with Barthes's distinction is clear enough.) The curious thing is that Danto puzzles about the past in ways that suggest some hesitation on his own part. He says, for instance, that "knowledge of the past estranges us from times other than our own!"65 But does that mean that he is willing to admit (in a somewhat Foucauldian way) that our "knowledge" of the past is a construction of our present Lebensform: that we understand the past only through the terms of our present conceptual orientation? If he does, then, of course, he has accepted something like Gadamer's notion of Horizontverschmelzung; and if he opposes it, then it seems he is obliged to offer us a suitable metaphysics of history. The concession, you see, suggests nothing less than the inseparability of the interpretation of history and self-interpretation: historicize the life of the interpreting subject (the point of Gadamer's objection against the Romantics), and you gain the reductio of Danto's (Romantic) option.

The second focus of Danto's account is more involved, but also more decisive. It is briefly specified in the following remark:

[T]he past for us must be regarded as the past tout court by us. To believe that Aeneas courted Dido, is not just to know what Virgil says about Dido and Aeneas; it is to hold what he said as true, correspondent to an erlking of dalliance in history-as-reality66

Danto is right, of course, but he has missed the full import of his own comment. He reads it—rightly—as a reductio of the "relativist's" position: that is, of the relativist who supposes that "true" simply means "true-for-us" (or, "true-in- L_k "). That, I say, produces insoluble paradox: thus, it becomes impossible to identify, as true , the fact that this group operates with the rule "true-for-us1 " and that group, with the rule "true-for-us₂ "; or, if it is indeed possible to make the distinction, then it is just false that everyone is restricted to some such rule as those just mentioned.67

Still, this does not settle the matter of the historicized context of truth-claims. It is true that "the past for us must be regarded as the past tout court by us"; but that is still an artifact of our changing history, as a Foucauldian genealogy would insist. We occupy no fixed point at which it is possible to ensure that what is "past tout court" will be the same for everyone, epistemically and interpretively, no matter where they are historically "placed." Rejecting the formula "true-for-us" as an analysis of "true" sans phrase does not in any way ensure the fixity of objective interpretation or the "nature" of interpretable texts. The connection is a non sequitur.

It is not a question of interpretive disagreement. It is a question rather of the horizonal change in the interpretive saliences of different times and of a historical meaning's being an artifact of just such change. It is not yet a question of relativism, though it is a question on which the fate of relativism depends. Of course, if this analysis of the second theme be allowed, then Danto's stand at the first becomes either untenable or impossible to confirm exclusively. Hence, there is no historical past that (necessarily) can be fixed once and for all. The historical past, as distinct from the merely physical past, can always be altered under changing interpretations and intervening histories. (That is because it is Intentional.) Dating events in physical time is not (and does not entail) fixing the historical meaning of those events. (Remember: I am opposed to a disjunctive dualism between the mental and the physical, between the cultural and the natural, between the Intentional and the non-Intentional, and between history and its meaning.)

If the argument be admitted, then the most startling possibilities will be seen to be coherent: whatever is interpretable is open to indefinitely many interpretations and indefinitely many histories; and the imputed "natures" of interpretable texts, or the historical pasts of human events, are intrinsically alterable under further interpretation and history.68 Either, then, we must now understand "objectivity" (in interpretation and history) differently from the "naturalist's" sense of

the term, or we ought to abandon the term because of its entrenched meaning; in either case, we need not abandon the idea of the consensual rigor with which interpretations and histories may be constructed.69

Now, then, I have only to add that what is true of interpretable and historical texts is true of the self; for, first of all, the self is a history or interpretable text; and second, the interpreting text that is the self is a self-interpreting text subject to changes in the meaning of its own history and imputable Intentional properties—as, on the argument, is every text. This agrees in part—but not entirely—with Gadamer (and Gadamer's criticism of Dilthey). For Gadamer says:

Self-reflection and autobiography—Dilthey's starting-points—are not primary and are not an adequate basis for the hermeneutical problem, because through them history is made private once more. In fact, history does not belong to us, but we belong to it.70

We "belong" to history because we are histories; but it is not also true that there are (independent) histories and we belong to them! There is no assignable priority to history and tradition or to "self-reflection and autobiography": each implicates and affects the other, and neither is fixed. Interpretation and history are constructed in the flux of the world by creatures that are similarly constituted and reconstituted. They (that is, we) are inseparable from one another. Hence, we find the meaning of our own lives dialogically. But our lives have no interpretive telos and they support no necessarily unique interpretation. Furthermore, since the world is "constructed" and "reconstructed" (by cognizing selves) in accord with their own emergent history, selves are entities that manifest powers that: (1) cannot be confined to the first-order attributes of whatever "world" they (at any given time construct; (2) allow for the relativization of truth-claims to the "genealogized" constructions of their successive epistemes; (3) disallow any principled disjunction between the interpretation of whatever falls within their "worlds" and their self-interpretation; and (4) preclude any uniquely correct interpretation of the "world" or of themselves, and are hospitable to "incongruent" interpretations.

To concede this much is, incidentally, to defeat all "naturalizing" strategies, whether (as with Davidson) with regard to science or (with Danto) with regard to art and history. This is the pared-down sense, two hundred years after the fact, in which the Hegelian theme (the historicity of thinking) has been liberated from its own impossible idiom—made legible for the new century.

NOTES

Introduction

Chapter 1: Reinterpreting Interpretation

Chapter 2: Interpretation at Risk

Chapter 3: Prospects For a Theory of Radical History

Chapter 4: Puzzles of Pictorial Representation

Chapter 5: Textuality and Intertextuality

Chapter 6: History and Fiction

Chapter 7: Interpretation and Self-Understanding

Index

A

"Aboumess," 108 , 133 -134, 135 -136, 138 , 141 , 142 , 144 -145

Action, Danto on, 257 , 260 -261

Adequational, 27 , 41 , 46 , 303 ;

defined, 24 -25;

interpretation as, 25

Allegoresis (White), 220

Alpers, Svetlana, 141

Analysts (analytic philosophy), 10 -11, 235

Annaliste , 217 , 218

Arendt, Hannah, 91 , 97 , 99 , 102 , 103 , 114 -115 passim, 128 ;

"The Concept of History," 88 -89, 93 -97 passim;

The Life of the Mind,89 -90

Aristotle (Aristotelian), 5 , 6 , 7 , 8 , 9 , 17 , 57 -60 passim, 63 -64, 67 -69 passim, 72 -74 passim, 78 -79 passim, 81 -83 passim, 85 , 110 , 111 , 118 , 186 , 189 , 201 -206 passim, 212 , 213 , 223 , 231 , 296 , 297 ;

Metaphysics,65 , 69 -71, 205 ;

Physics, 202 -203, 205 ;

Poetics, 201 -202

Art:

Danto on, 145 -146;

as Intentional, 147 ;

logical and ontological concessions regarding, 46 -50 passim;

mimetic theory of, 28 , 165 ;

as real, 131 -132, 134 ;

representational, 131 , 133 -134, 135 , 136 , 138 , 140 -142, 143 -145, 147 -148;

Sartre on, 127 -128, 129 -130;

semiotic theory of, 134 ;

uttered, 144 -145

Artifacts (constructions):

cultural phenomena as, 7 ;

defined, 8 -9;

laws of nature as, 92 ;

observations as, 137 ;

world composed of, 8 -9, 90 -91, 265

Artworid (Danto), 259 -260

Auden, W. H.:

"Musée des Beaux Arts," 167 -168, 169 -170, 172

Augustine, St., 217

Autotelic (Eliot), 29

Axiom of Existence (Searle), 164

B

Bacon, Roger, 140 -141, 142

Bakhtin, Mikhail, 277 -278

Balzac, Honoré de: Sarrasine , 36 -37, 39 , 43 , 154 , 155

Barthes, Roland, 8 , 29 , 30 , 31 , 34 , 45 , 46 , 49 -54 passim, 58 , 87 , 103 , 109 , 122 , 123 , 154 , 155 , 157 , 170 , 172 , 176 , 177 , 179 , 186 , 187 , 221 , 261 -262, 263 , 270 , 272 -273, 292 ;

Change the Object Itself," 193 ;

"From Work to Text," 31 -36, 43 -44, 48 ;

S/Z, 36 -37, 40 -43 passim;

"The Photographic Message," 290 -291;

The Pleasure of the Text,40 -41

Baudrillard, Jean, 162 ;

Simulations,120 -121

Benjamin, Walter, 171

Bennington, Geoffrey, 293

Bergson, Henri, 171

Black, Max, 300

Bloom, Harold, 8 , 85 , 109 , 122 , 123 , 152 , 171 , 187 , 269 -270;

The Breaking of the Vessels,83 -85;

A Map of Misreading , 84

Brentano, Franz, 13 , 139

Brunelleschi, Filippo, 142 , 154 -155;

Santa Maria del Fiore, 131 -132, 133

Brute world, 90 -91, 210

C

Career(s), 33 -34, 42

Causality, 94

Cavell, Stanley, 278

Céanne, Paul, 196 -198

Chiaramonte, Nicola, 213

Churchland, Paul M., 246 -248 passim

Classical (Gadamer), 50 , 51 , 67 , 68 , 78 -79, 92 -93, 98 -99, 209 , 227

Computationality, 41

Constable, John, 150

Contexted (contextual), 105 , 106 ;

de-

fined, 9 ;

Derrida on, 76 ;

discourse as, 9 ;

Gadamer on, 76 -77;

Kuhn on, 162

Continentals (continental European philosophy), 10 -11

Criticism:

Beardsley on, 29 ;

distinct from literature (artwork), 29 -32, 34 -35;

Eliot on, 29 ;

Krauss on, 29 -32, 34 ;

modernist and postmodernist theories of, 31 -35 passim;

Riffaterre on, 166

Cultural attributes, incarnate, 128 , 138 , 143

Cultural entities, embodied 128 , 138 , 143

D

Danto, Arthur C., 101 , 106 , 130 , 135 , 140 , 145 , 149 , 160 , 189 , 218 -219 passim, 229 , 255 -265 passim, 284 , 286 , 296 , 305 ;

Analytical Theory of Action , 257 ;

"The Artworld," 259 ;

Narration and Knowledge,107 , 162 -164;

The Philosophical Disenfranchisement of Art , 260 ;

The Transfiguration of the Commonplace , 145 -146, 257 -258, 287

Davidson, Donald, 182 , 253 , 255 , 271 -272;

"A Coherence Theory of Truth and Knowledge," 243 -244, 253 ;

"Mental Events," 53 , 287 ;

"Radical Interpretation," 249 -250, 251 -252, 233 , 304 ;

"Thought and Talk," 248 -249

Deconstruction, 180 , 208 , 293 , 306

de Man, Paul (de Manian), 172 -174, 180 -181 passim, 186 , 188 -190 passim, 211 -215 passim, 220 , 230 ;

Blindness and Insight,184 -186

Dennett, Daniel C., 271 , 284

Derrida, Jacques (Derridean), 29 , 30 , 58 , 66 -67, 68 , 76 , 80 , 170 , 173 -176, 178 , 180 -186 passim, 208 -209, 229 , 293 , 297 -298;

Of Grammatology , 43 , 75 , 172 -173, 176 -177

Description/interpretation, 11 -12, 26 , 179 , 186

Dewey, John, 102

Dilthey, Wilhelm (Diltheyan), 255 -256, 265

Discursivity and essentialism, 74 , 81 , 86

Doctorow, E. L., 212

Duchamp, Marcel: Fountain , 258

Duns Scotus, 83

E

Ecu, Umberto, 163 -164

Edgerton, Samuel, 142

Eliot, T. S., 29 , 171

Emergence:

cultural and physical, 47 ;

distinct from supervenience, 48 , 52 , 54 , 142 -143, 145 -146, 255 , 287

Emotivism, 120

Epistemes (Foucault), 41

Essentialism, 74 , 81 , 100 , 277

Excluded middle, law of, 7 , 47 , 62 ;

and charge against Protagoras, 65

Extensionalism, 48 , 279

Externalism, 182 -183, 188 , 245 -246, 255 , 256

Extrinsic meaning, 103 -104

F

Fact/value dichotomy, 12

Feigl, Herbert, 52 -53

Fiction/reality, 180 , 183 , 189 , 191 , 214 -216;

Aristotle on, 186 , 201 -202;

de Man on, 180 -181, 184 , 186 , 190 ;

Hutcheon on, 180 -181, 184

First-order/second-order, 238 -239;

and naturalizing, 239

Fish, Stanley, 293

Fleck, Ludwik, 189

Flux, 2 , 5 -6, 8 , 27

Folk psychology (Churchland), 246 -248 passim

Foucault, Michel (Foucauldian), 4 , 8 , 82 , 90 , 102 , 104 , 105 , 109 , 117 , 147 , 154 , 155 , 163 , 170 , 183 , 190 , 191 , 209 , 216 , 221 -222, 223 -224 passim, 249 , 263 , 284 -288, 290 ;

"Nietzsche, Genealogy, History," 221 -222;

The Order of Things , 50 -51, 52 , 286

Frege, Gottlob, 7

Freud, Sigmund, 233

Fry, Roger, 259

Frye, Northrop, 202 , 227

Fusion of horizons (Gadamer), 49 , 60

G

Gadamer, Hans-Georg (Gadamerian), 8 , 51 -54 passim, 58 , 60 -61, 66 -67, 75 , 79 -80, 95 , 98 -99, 102 , 104 -106 passim, 114 , 155 , 166 , 171 , 172 , 175 , 183 , 209 , 222 , 236 , 237 , 242 , 253 , 255 , 256 , 262 , 272 , 273 , 277 ;

"Are the Poets Falling Silent?" 78 ;

The Idea of the Good , 78 -79;

Truth and Method , 48 -50, 67 -68, 76 -78, 92 -93, 253 , 265

Gasché, Rodolphe, 293

Geertz, Clifford, 285

Genealogy (Foucault), 191 , 192 , 224 , 265

Genres (Hirsch), 73

Gibson, J. J., 152 , 154 , 194

Gödel, Kurt, 207

Gombrich, E. H., 130 , 132 , 149 -153, 154 , 194 , 198 , 290

Goodman, Nelson, 130 , 132 , 134 , 135 , 140 , 144 , 149 -153, 154 , 300 , 301 -302

Greenberg, Clement, 29 , 198

Grünbaum, Adolf, 103 -104, 106 , 107 ;

The Foundations of Psychoanalysis , 100 -101

H

Habermas, Jürgen, 78 , 79 , 105 , 256 , 277 , 305

Habitus (Bourdieu), 41

Haecceity, 71 -72, 81 -82, 83

Hanslick, Eduard, 286

Hegel, G. W. F. (Hegelian), 78 , 89 , 95 -97 passim, 99 , 104 , 114 -115 passim, 209 , 218 , 222 , 245 , 256 , 262 , 265 ;

Lectures on the Philosophy of World History , 97 -98

Heidegger, Martin (Heideggerean), 102 , 173 , 301

Hempel, C. G., 16 , 101 , 106 , 256

Hesse, Mary, 300

Heteroglossia (Bakhtin) 277 -278

Hirsch, Jr., E. D., 73 , 178 , 187 , 270

Historical past' alternative views of, 62 ;

as Intentional, 216 ;

as incarnate, 231 -232;

narrative structure of, 216 -217

Historicism, 192

Historicity, 93 , 137 , 208

Historicized, defined, 3

Histories and natures, 7 , 81 -82, 109

History:

Arendt on, 88 -89, 93 -94, 95 -96;

Aristotle on, 201 -203, 205 ;

Danto on, 107 , 255 -256, 260 , 262 -264;

distinct from symbiotized nature, 84 , 90 ;

distinguished from fiction, 209 -211, 216 , 217 , 222 -223, 225 , 229 ;

Foucault on, 221 -222;

Grünbaum on, 100 -101, 103 -104;

has no nature, 93 -94;

Hegel on, 97 - 98;

hermeneutic theory of, 102 , 104 -105;

and historical meaning, 99 -100;

a human construction, 89 , 90 ;

and in-variance, 98 ;

local and world (universal), 96 -99, 103 , 115 , 116 ;

modernist theory of, 101 -102, 103 -104, 105 -108, 116 ;

narrative, 109 , 119 , 217 , 219 -222;

poststructuralist theory of, 102 , 104 , 108 -109, 116 ;

physicalist conception of, 206 ;

radical, 94 , 95 , 98 , 103 , 112 -113, 117 -119, 122 ;

Ricoeur on, 225 -229;

Vico on, 113 -115, 116

Hölderlin, Friedrich, 78

Horizon (Gadamer), 77 -78

Human beings (selves) as texts, 59 , 60 , 64 , 80 ;

Gadamer on, 60 -61, 68 ;

historicized, 137 -138

Hume, David (Humean), 243 , 244 -246 passim

Husserl, Edmund (Husserlian), 13 , 44 , 78 , 139 , 141 , 178 , 182 , 184 , 207 -208, 210 , 214 , 236

Hutcheon, Linda, 175 , 184 , 211 , 214 , 230 , 292 ;

A Poetics of Postmodernism , 161 -162, 180 -181

Hyle (Aristotle), 110

I

Indeterminacy, 27 , 61 -62, 72

Indiscernibility puzzle (Danto), 258 -259, 284 , 286 , 287

Individuation, 27 , 71

Innocent eye, myth of, 149 , 151

Intensional(ity), defined, 13

Intention, author's, 54 , 270 , 272

Intentional(ity), 47 ;

Davidson on, 271 -272;

defined, 13 ;

Dennett on, 271

Intentional(ity), as term of art, 33 , 128 -129;

collective nature of, 136 ;

as cultural 14 ;

defined, 13 -14, 107 -108, 133 , 268 , 284 ;

as incarnate, 15 ;

interpretation addressed to, 15 ;

as monadic, 139 ;

realism of, 14 , 108

Internalism, 279

Interpretation:

addressed to the Intentional, 15 ;

as adequational, 246 ;

in arts and history, 25 , 194 -195, 196 , 263 ;

Barthes on, 38 -42;

as constitutive, 25 ;

constraints on theory of, 21 -22;

Davidson on, 248 -253 passim, 271 -272;

and description, 11 -12, 26 ;

and human condition, 64 ;

and invariance, 56 -58;

narratized, 217 -218, 219 , 223 , 226 -229 passim;

new puzzle of, 5 , 23 -24, 58 , 242 , 261 -262;

objectivity of, 136 -137;

Peirce on, 61 -62;

and physicalism, 54 ;

and referential reliability, 25 ;

Riffaterre on, 165 ;

Rosen on, 56 -60;

in science, 16 ;

and texts, 86 , 186 -187;

theory of, 3 -4, 9 , 21 -23 passim, 28 , 45 -46;

and truth-value assignments, 21 , 26 ;

two sorts of, 21 -22, 24 -25, 187 -188;

van Fraassen on, 16

Interpretive tertia :

Davidson on, 248 -250 passim, 279 -280;

Rorty on, 250 -251

Intransparent, defined, 2

K

Kandinsky, Wassily: Improvisation , 134 , 135 -137 passim

Kant, Immanuel (Kantian), 6 -7, 9 , 23 , 28 , 76 , 89 -92 passim, 106 , 128 , 130 -

131 , 133 , 141 , 151 , 162 , 178 , 182 , 210 , 214 , 228 , 238 , 242 , 245 , 246

Kantian-like, 9 , 28 , 76 , 130 , 172 , 234

Krauss, Rosalind, 29 , 39 , 41 , 45 , 51 , 97 , 101 , 109 , 162 ;

The Originality of the Avant-Garde , 27 -32, 34 -35

Kripke, Saul A., 81 , 83 , 111 , 276

Kuhn, Thomas S., 8 , 108 , 139 -141 passim, 162 , 189 , 196

L

Laws of nature, 92 , 94 , 189 , 204 , 277

Lebensforrnen^[*] (Wittgenstein), 39 , 41 , 52 , 118 , 129 , 141 , 150 , 160 , 224 , 234 , 240 ;

equivalences in other authors, 41

Lehenswelt (Husserl), 41

Legitimation, 238 -239

Leibniz, G. W. (Leibnizian), 73 , 82 , 110 -111

Lévi-Strauss, Claude, 157 , 175 , 219

Lévy-Bruhl, Lucien, 208

Lingual (as term of art), 172 ;

defined, 268

Lyotard, Jean-François, 41 , 162

M

McGinn, Colin, 53

MacIntyre, Alasdair, 212 , 253 -254, 272 -273, 283 -284

Magritte, René, 288 ;

Les Promenades d'Euclide , 155

Many-many problem (Feigl), 52 -53

Marx, Karl (Marxist), 89 , 90 , 95 -96, 99 , 114 -115 passim, 218 , 220

Masaccio:

Tribute Money , 142 , 153 , 155

Masolino:

Raising of Tahitha and Healing of the Cripple , 142 , 153 , 155

Materialism, 47 , 138

Matisse, Henri:

l'Atelier rouge , 193 -194, 195 ;

Blue Window,193 ;

Café maure , 194 ;

Nasturtiums with "Dance" (I),196 ;

Nasturtiums with "Dance" (II) , 196 ;

Piano Lesson,195

Meanings, 143 , 231 -232

Medium, physical and cultural, 138

Merleau-Ponty, Maurice, 197 , 198 , 208

Metaphor, Ricoeur on, 226 -229, 230 -231, 300 -301

Metaphysics of culture, 133 , 141

Misreading (Bloom), 83 -85

Modernism/postmodemism, 29 , 39

Moore, G. E., 53 , 245 , 246

Moral realism, 283 -284

[*] I have taken the liberty of construing "lebensformlich" as an English adjective; hence, I have dropped the inflectional forms.

N

Narrative, 216 -219 passim;

Foucault on, 221 -222;

Ricoeur on, 225 -229, 230 -231;

White on, 219 -221, 222

Natura naturans/naturata (Spinoza), 144 , 145 , 187 -188

Natural kinds, 7 , 46 , 100

Naturalizing (Quine), 54 , 101 , 134 , 192 , 236 , 237 -238, 239 , 242 , 244 , 254 , 255 , 265 ;

Danto on, 255 -261 passim;

Davidson on, 243 -244;

Goodman on, 301 -302;

Quine on, 228 ;

Rorty on, 239 -240

Nature(s), 15 , 24 , 27 , 35 ;

affected by interpretation, 31 -32;

Aristotle on, 69 -71;

Gadamer on, 68 ;

and histories (careers), 7 , 42 , 77 , 81 -82, 85 -86, 109 ;

invariance of, 12 , 46 , 60 ;

invariance of human, 56 -57;

Rosen on, 56 -58

Nehamas, Alexander, 213

New Criticism (Beardsley), 29 , 54

Nietzsche, Friedrich (Nietzschean), 4 , 37 , 188 , 212 , 223 , 224

Noematic, 13

Noncontradiction, principle of, 62 ;

Aristotle on, 69 -71;

and charge against Protagoras, 61 , 69 -70

Noncognitivists, 128

Nonextensional, 13

Nonreductive physicalism, 52

Norris, Christopher, 80

Noumenal (Kant), 6 -7, 128 , 130 , 228 , 229

Numerical identity, 74 , 82

Nussbaum, Martha, 213

O

Objectivism, 182 , 186 , 187 ;

in history, 189

P

Panofsky, Erwin, 154 , 167 -168

Paraliterary (Krauss), 29 -30, 32 ;

as postmodernist, 29

Parmenides (Parmenidean), 8 , 64 , 66 , 74 , 231

Past, physical and historical, 196 -198 passim, 222 , 228 , 231 , 263 -264

Peirce, Charles Sanders, 27 , 61 -63, 90 , 91 , 151 , 203 , 215 , 248 ;

Collected Papers,236 -237

Perception, 139 ;

of art and physical nature, 152 ;

Danto, 258 -259;

Intentionality of, 152 ;

Kuhn on, 139

Perspective (pictorial), 154 , 155

Photograph, 155

Physicalism, 47 , 48 , 52 , 54 , 107 , 129 , 138 , 257 , 286

Picasso, Pablo: La Vie , 32 , 33 , 34

Plato, 1 , 63 , 64 -65, 74 , 78 -79 passim, 150 , 233

Platonism, 144

Pluralists, 123

Popper, Karl R., 81 , 227

Post-Hegelian, 189

Post-Kantian, 9 , 76 , 89 , 128 , 130 , 141 , 246

Postmodernism/modernism, 29 , 31 , 32 , 39 , 109

Poststructuralism, 116 -117, 128

Pragmatism, 144

Praxis (Marx), 41

Predicates, 26 , 148 -149, 151

Predication, 11 ;

applied to cultural phenomena, 15

Preformed, defined, 3

Prejudice (Gadamer), 49

Pre-Kantian, 114 , 118 , 130 , 133 , 245 , 246

Proper names:

extensional function of, 33 ;

Krauss on, 32 , 34 ;

Quine on, 73 , 110

Protagoras, 1 -2, 3 , 62 , 63 -71 passim

Putnam, Hilary, 69 , 76 , 80 , 280 , 299

Q

Quiddity, 71 -72, 83

Quine, W. V., 30 , 31 , 73 , 76 , 101 , 109 , 110 -111, 112 , 118 , 182 , 191 -192, 215 , 236 , 244 , 245 , 249 , 250 , 299 ;

"Epistemology Naturalized," 238

R

Ranke, Leopold von, 226 , 228

Readerly/writerly (Barthes), 36 -37, 39 -41, 42 -43, 44 , 179 , 197 , 261 -262

Realism, physical and cultural, 151

Realism/idealism, 91 , 151 , 215

Reality:

and fiction, 201 -202, 211 -212, 213 -215, 224 -225;

Aristotle and Kant on, 6 -7;

as artifact, 17 , 23 , 214 -215;

as flux, 3 ;

as intransparent, 2 ;

as symbiotized, 3

Reductionism, 131 , 246 -247

Reference, 11 , 73 -75 passim, 111 -112, 224 ;

Barthes on, 36 ;

cannot be retired, 83 , 111 ;

causal theory of, 111 , 276 ;

a cognitive matter, 73 ;

fallacy of (Riffaterre), 163 -164;

to cultural phenomena, 15 ;

as grammatical distinction, 32 -33, 34 , 164 , 210 ;

histori-cized, 117 -119;

to history and fiction, 209 -210;

Hutcheon on, 161 -162;

as Intentional, 209 ;

Krauss on, 29 -32 passim;

and Leibniz, 73 , 110 -111;

to literature, 161 ;

and Quine, 73 , 110 -111;

Riffaterre on, 159 -160, 163 ;

success of, 75 , 111 ;

univocal sense of, 161 -163

Reidentification, 27

Relativism, 51 , 105 , 123 , 137 , 190 , 192 -193, 241 , 264 ;

logic of, defined, 241 -242

Relativistic logic, 241 , 242

Rembrandt:

Bathsbeba , 194 -195

Representational art, 133 -134, 135 -136, 138 -139, 140 , 141 , 144 -145, 147 -148

Resemblance (similarity), 148 , 150 -151, 152 , 153 ;

Foucault on, 288 ;

Gombrich on, 150 -151, 152 ;

Goodman on, 149 , 152 ;

Russell on, 288

Ricoeur, Paul, 106 , 218 , 219 ;

The Rule of Metaphor,300 -301;

Time and Narrative,225 -229, 230 -231

Riffaterre, Michael, 156 -158, 161 -164 passim, 167 -168, 171 -172, 173 -178 passim, 181 -182 passim, 184 , 186 -188 passim, 192 -193 passim, 195 ;

"Describing Poetic Structures," 291 ;

"Intertextual Representation," 158 -160, 165 -166, 169 -170

Romantic hermeneutics, 54

Rorty, Richard, 25 , 69 , 80 , 85 , 162 , 191 -192, 238 , 240 , 250 -253 passim, 272 ;

"Pragmatism, Davidson and Truth," 250 , 252 -253;

"Solidarity or Objectivity?" 239 -240

Rosen, Stanley, 61 -64 passim, 71 , 73 , 74 -80 passim, 85 -86 passim;

Hermeneutics as Politics , 56 -60

Russell, Bertrand, 245 , 288

S

Samples (Goodman), 134

Sartre, Jean-Paul (Sartrean), 127 -128, 129 -130, 132 , 135 , 139 , 146 , 189 , 219

Saussure, Ferdinand de, 43 , 44 , 218

Schapiro, Meyer, 169

Searle, John R., 154 , 290 , 303

Secondness (Peirce), 90

Self-knowledge, knowledge as, 234

Selves (persons), 233 , 235 , 242 , 265 ;

Gadamer on, 265 ;

Hume on, 243 , 244 -246;

nature of, 237 -238, 247 -248;

Peirce on, 236 -237;

and second-order powers, 238 , 265

Semioclasm (Barthes), 193

Shakespeare, William: Hamlet, 118 , 120

Signifiers (Barthes), 38 , 43

Simulacrum (Baudrillard), 120 -121

Sitten (Hegel), 41

Social constructivism, 9

Socially constituted (constructed), defined, 7

Solidarity (Rorty), 239 -240

Sontag, Susan, 260

Spinoza, Baruch, 187

Steen, Jan:

Easy Come, Easy Go,145

Straussians, 64

Structuralism (Francophone), 54 -55, 157 , 170 , 218 -219;

and New Criticism, 157 ;

and Romantic hermeneutics, 157 ;

Riffaterre's version of, 170 -171, 174 -175

Supervenience, 48 , 52 , 54 , 143 , 145 , 255 ;

Davidson on, 53 , 287

Symbiotized (symbiosis), 23 , 46 , 128 , 151 ;

as post-Kantian, 128 ;

defined, 3

T

Tarski, Alfred (Tarskian), 245 -253 passim, 278 -280

Tertium non datur , 47

Text (Barthes), 36 , 37 -41, 44

Texts:

adequated to interpretation, 26 ;

altered by interpretation, 31 , 32 , 45 -46, 87 , 172 , 196 , 197 -198;

as artifacts, 28 , 43 -45, 59 ;

Barthes on, 37 -38, 39 -40;

Bloom on, 83 -86, 269 -270;

defined, 26 ;

de Man on, 181 ;

Derrida on, 293 ;

Gadamer on, 49 -50;

as Intentional, 27 , 50 , 54 , 86 ;

interpretive history of, 38 ;

as interpretively open-ended, 39 , 86 ;

as intrinsically indeterminate, 27 ;

mimetic (Riffaterre), 164 ;

normalizing of, 42 , 43 ;

not natural kinds, 46 ;

as referents, 26 , 122 ;

unity and unicity of, 46

Textual/intertextual, 166 , 169 -170, 178 - 179;

Barthes on, 193 ;

Davidson on, 279 -280;

Riffaterre and de Man on, 173 ;

Riffaterre and Derrida on, 172 , 174 , 175 , 177 -178;

Riffaterre and Gadamer on, 166 ;

Riffaterre and Panofsky on, 167 -168;

structuralist and poststructuralist views of, 179

Textuality:

and referentiality, 159 -140, 162 , 163 , 164 ;

Barthes on, 193 ;

de Man on, 172 -173, 180 -181, 184 -186;

Derrida on, 72 , 172 -176 passim, 177 , 180 , 182 , 184 ;

Riffaterre on, 158 -159, 163 , 174 -175;

two notions of, 172 , 188 -190, 198

Theory/practice, 17

Thucydides, 202 , 212

Time:

Aristotle on, 202 -203,205;

directionality of, 203 -205, 206 -207, 216 ;

incarnate, 231 ;

physical and historical, 206 , 207 , 231

Transfigured (Danto), 145 -146

Truth:

as artifact, 190 -191, 216 ;

disquotational theory of, 182 , 192 ;

Fou-cault's aporia regarding, 216

Truth-values (claims), 11 -12, 191 , 192

T-sentences (Davidson), 251 , 279 -280

U

Understanding, 235 ;

and interpretation, 241 ;

historicized, 242

Unicity, 27 , 33 -34, 35 , 46

Unified (unity), 26 -27, 28 , 32 , 35 , 46

Unitary, 27 -28, 32 ;

defined, 26

Universals, problem of, 111 , 148 , 150 , 152 ;

and nominalism, 148 -149;

and "real generals" (Peirce), 150 , 151

V

van Eyck, Jan, 154 ;

Giovanni Arnolfini and His Wife , 149

van Fraaseen, Bas C., 16 , 270

Velàsquez:

Las Meninas, 51 , 147 , 153 -154, 155 , 290

Vermeer, Jan:

Lady Reading a Letter at an Open Window , 133 -134, 135 -137 passim, 139 , 145

Vico, Giambattista (Vichian), 88 , 90 -91 passim, 95 -96 passim, 99 , 102 , 116 ;

The New Science of Giambattista Vico , 113 -116

W

Warhol, Andy:

"Brillo Box," 145 -146

Weinsheimer, Joel, 61

Weinstein, Fred, 191 , 211 -214 passim, 230 ;

History and Theory After the Fall , 190 , 211 -212

Weitz, Morris, 120 , 121 , 123 ;

"Hamlet" and the Philosophy of Literary Criticism , 120

West, Nathaniel:

Miss Lonelyhearts , 84 -85

White, Hayden, 106 , 202 , 213 , 218 -222 passim, 225 -228 passim;

The Content of the Form , 219 -221

Wiebe, Rudy:

The Temptations of Big Bear,161

Wirkungsgeschichtliches Bewusstsein (Gadamer), 41

Wittgenstein, Ludwig (Wittgensteinian), 120 , 172 , 236 , 243 , 257 , 273

Wollheim, Richard, 135 , 286

World:

as constructed, 90 -91;

as independent, 91 ;

as textualized, 161

Preferred Citation: Margolis, Joseph. Interpretation Radical but Not Unruly: The New Puzzle of the Arts and History. Berkeley:  University of California Press,  c1995 1995. http://ark.cdlib.org/ark:/13030/ft209nb0kk/