History and Theory, Vol. 19, No. 3 (Oct., 1980), 245-276.
RETHINKING INTELLECTUAL HISTORY AND READING TEXTS
But if we see this circle as a vicious one and look out for ways of avoiding it, even if we just 'sense' it as an inevitable imperfection, then the act of understanding has been misunderstood from the ground up.
Heidegger, Being and Time
Over the last decade, intellectual historians have increasingly felt that their field is undergoing a crisis significant enough to reopen the question of the field's nature and objectives. Whatever its presumed causes (for example, the rapid rise of social history), one beneficial effect of a sense of crisis is the pressure it places upon practitioners of a field to be more articulate about what they are doing and why they are doing it. In response to this pressure, I shall attempt to define and to defend in relatively theoretical terms the approach to the field (more specifically, to modern European intellectual history) that I have come to find most fruitful. In setting forth this approach, I shall stylize arguments to bring into prominence a number of controversial issues. In so doing, I shall at times be forced not to practice what I preach, for I shall selectively treat the texts of other historians or theorists in order to highlight problematic positions as well as possible directions for inquiry.
In the course of its own history in this country, intellectual history has often patterned itself on other approaches to history, taking a framework of significant questions from somewhere else to orient and to organize its research. The desire to adapt to modes of inquiry immediately intelligible to some important set of historians, if not to all other historians, has characterized perspectives on the field that are frequently seen as competing or opposed options: the internal or intrinsic history of ideas (exemplified in the works of A.O. Lovejoy); the extrinsic or "contextual" view of intellectual history (exemplified in the works of Merle Curti); and the attempted synthesis of internal and external perspectives that has most often taken the form of a narrative of "men and ideas" (for example, in the works of Crane Brinton or H.S. Hughes). The problems generated by these options have become increasingly evident, and I shall return to some of them. They are exacerbated by the tendency of intellectual history either to become narrowly professional and even antiquarian by applying the internal method to increasingly
insignificant problems, or to become fixated more or less permanently on a popularizing and introductory level of understanding in narrating the adventures of "men and ideas." The more recent elaboration of a social history of ideas has seemed like an answer to these problems, for it goes beyond the older forms of contextualism in its rigor and methodological sophistication, and it promises to give intellectual history access to the remarkable achievements of modern social history. It is undoubtedly true that certain questions addressed in an impressionistic way by earlier intellectual historians can be cogently investigated only through the techniques of modern social history. But intellectual history should not be seen as a mere function of social history. It has other questions to explore in ways that may permit a better articulation of its relationship to social history. In the process, it may even suggest areas in which the formulations of social history stand in need of further refinement.
In the pages that follow, an obvious "territorial imperative" is at work in a manner that is modified by an active awareness both of the limits of intellectual history and of its relations to other perspectives. Thus my own argument is not motivated by the desire to establish a specious autonomy for intellectual history within historiography or within the disciplines in general. On the contrary, it is itself informed by an understanding of the subdiscipline of intellectual history that is in important respects transdisciplinary, and it defends what may be called the relative specificity of intellectual history. In more practical terms, it urges the intellectual historian to learn of developments in other disciplines addressing the problem of interpretation, notably literary criticism and philosophy. In fact, the argument I shall put forth constitutes a new twist to a rather traditional view of things but one that involves an at times disorienting critique and renewal of tradition through an insistence upon problems and interests that have been obscured in more traditional approaches. The concern I would like to reanimate centers around the importance of reading and interpreting complex texts the so-called "great" texts of the tradition and of formulating in a cogent way the problem of relating these texts to various pertinent contexts. This is a concern which does not have the place it deserves in historiography today including intellectual history, which would seem to be its "natural home." The approach I shall discuss is not, however, adapted solely to returning these texts to their rightful place. It is one that critically raises the question of why these texts are often objects of excessively reductive interpretation even when they are centers of analysis and concern. The primary form of reduction I shall discuss arises from the dominance of a documentary conception of historical understanding, because I believe that it, rather than other views (for example, the view stressing empathy), is most prevalent in the historical profession today. But the implications of my argument extend to all extreme derogations of the dialogue between past and present
dialogue that requires a subtle interplay between proximity and distance in the historian's relation to the "object" of study. (This "dialogical" relation between the historian or the historical text and the "object" of study raises the question of the role of stylization, irony, parody, self-parody, and polemic in the historian's own use of language that is, the question of the ways in which the historian's use of language is mediated by critical factors that cannot be reduced to factual predication or direct authorial assertion about historical "reality." Significant in this respect is the manner in which the historian's approach to the "object" of study is informed or "influenced" by the methods and views of other historians or "speakers.") In addition, the approach I shall defend is not motivated exclusively by the attempt to find order in chaos by familiarizing the unfamiliar; it is also sensitive to the ways in which the ordinary format for the acquisition of knowledge may be placed in question as the familiar is made unfamiliar especially when it is seen anew in significant texts.
What is meant by the term "text"? It may initially be seen as a situated use of language marked by a tense interaction between mutually implicated yet contestatory tendencies. On this view, the very opposition between what is inside and what is outside texts is rendered problematic, and nothing is seen as being purely inside or outside texts. Indeed the problem then becomes one of rethinking the concepts of "inside" and "outside" in relation to processes of interaction between language and the world. One of the more challenging aspects of the recent inquiry into textuality has been the investigation of the ways in which textual processes cannot be confined within the bindings of the book. The context or the "real world" is itself "textualized" in a variety of ways, and even if one believes that the point of criticism is to change the world, not merely to interpret it, the process and results of change themselves raise textual problems. Social and individual life has in part a textual structure and is involved in textual processes that are often more complicated than the historical imagination is willing to allow. In addition, the attempt to relate texts to other "symbolic," "representational," or "expressive" media (music, painting, dance, gesture) raises the problem of the interaction among modes of inscription the problem of "translating" from medium to medium in a process that entails both losses and gains in "meaning." To the extent that the historian or critic employs language to effect this "translation," he or she confronts the issue of textuality writ large. More generally, the notion of textuality serves to point to the fact that one is "always already" implicated in problems of language use as one attempts to gain critical perspective on these problems, and it raises the question of both the possibilities and the limits of meaning. For the historian, the very reconstruction of a "context" or a "reality" takes place on the basis of "textualized" remainders of the past. The position of the historian is not unique in that all definitions of reality are
implicated in textual processes. But the problem of historical understanding is distinctive. The more general problem is to see how the notion of textuality raises the question of relationship in a manner that may run the risk of aporia but, instead of eventuating in a vicious circle, creates the possibility of renewing modes of understanding. The more distinctive form this problem takes in historiography is that of the relationship between documentary reconstruction of, and dialogue with, the past.1
To the extent that the relationship between the documentary and the "dialogical" is a problem relevant to all historiography, the argument I shall develop is not restricted to intellectual history. In what follows, however, I shall generally forego extended discussion of the larger issues I may evoke in order to focus on the limited topic of the relation of a "textual" problematic to intellectual history. In even more limited fashion, I shall concentrate on the topic of written texts and, within this topic, on the problem of reading and interpreting the "great" texts of the tradition. These texts are not absolutely unique, and the processes they disclose are not altogether peculiar to them. But two reasons for focusing intellectual history on these texts are that the study of them tends not to be emphasized within contemporary historiography and that in them the use of language is explored in an especially forceful and critical way a way that engages us as interpreters in a particularly compelling conversation with the past.'
It is important to argue over the question of what works are to be considered great and to re-evaluate the "canon"' of works to which we devote special attention. I even see value in the argument that our understanding of a "canon" has been too ethnocentric in its confinement of the text to the book and in its exclusion of texts from other traditions and cultures. Indeed it is important to examine critically the very notion of a "canon" and certain of the functions it may serve. But I must confess that most often I agree with traditional authorities in identifying works to be included in any necessary
1. This article is best read in conjunction with the works of Hayden V. White, notably Metahistory (Baltimore, 1973) and Tropics of Discourse (Baltimore, 1978). White's work has been of immense importance in generating the current debate about approaches to historiography. My own discussion agrees with White's critique of traditional narrative and a narrowly documentary approach as inadequate to the tasks of intellectual history. But it is critical of the more "presentist" and "constructivist" tendencies that at times emerge in White's works, and it tries to provide a different way of understanding intellectual history as in part a dialogue with the past. For a discussion of White's Tropics of Discourse, see my review in Modern Language Notes 93, no. 5 (1978), 1037-1043. For a discussion of related problems as they arise in the interpretation of the works of Jean-Paul Sartre, the reader is referred to my Preface to Sartre (Ithaca, N.Y., 1978).
2. The notion of historical understanding as a conversation with the past is developed in the works of Heidegger and in those of his more conservative disciple, Hans-Georg Gadamer. See especially Heidegger's "Onto-theo-logical Constitution of Metaphysics" in Identity and Difference, transl. Joan Stambaugh (New York, 1969) and Gadamer's Truth and Method (New York, 1975) and Philosophical Hermeneutics, transl. David E. Linge (Berkeley, 1976).
but not sufficient list of especially significant texts. What puzzles me at times, however, is the way in which these works are interpreted, for the interpretations may have little correspondence to the judgment that the work is great or at least of special significance. Here one may wonder whether something is elided in the passage from the judgment that identifies a great work to the discourse that interprets it. For the interpretation often treats these texts in terms that domesticate them by emphasizing their commonality with lesser works or with ordinary beliefs, desires, tensions, and values. This begs a number of crucial questions. Are great texts of special interest not in their confirmation or reflection of common concerns but to paraphrase Nietzsche in the exceptional way in which they address commonplace themes?3 Do they often or even typically engage in processes that both employ or refer to ordinary assumptions and contest them, at times radically? Is the judgment of greatness at times related to the sense that certain works both reinforce tradition and subvert it, perhaps indicating the need for newer traditions that are more open to disconcerting modes of questioning and better able to withstand the recurrent threat of collapse? Do certain works themselves both try to confirm or establish something a value, a pattern of coherence, a system, a genre and call it into question? Is there something sensed in judgments that may not be said in reductive interpretations that make certain works all too familiar? Are processes of contestation often or typically more powerful in certain kinds of texts for example, literary or poetic texts in comparison with philosophical or historical ones? How watertight are these higher-order forms of classification in relation to the actual use(s) of language in texts? What does a less reductive, normalizing, or harmonizing mode of interpretation require of the reader?
These are the types of question that are raised in what Heidegger calls "thinking the unthought" of tradition and Derrida, "deconstruction." (More precisely, what I take to be especially valuable in the approaches to textuality developed by Heidegger and Derrida is critical inquiry that tries to avoid a somnambulistic replication of the excesses of a historical tradition by rehabilitating what is submerged or repressed in it and entering the submerged or repressed elements into a more even-handed "contest" with tendencies that
3. See especially Nietzsche's Use and Abuse of History, transl. Adrian Collins (Indianapolis, 1957), 39. For Nietzsche's argument that the renunciation of interpretation and the restriction of scholarship to pure "truth" in its residual form as "truth-to-facts" constitute an expression of the ascetic ideal, see On the Genealogy of Morals, transl. Walter Kaufmann (New York, 1969), 151. For a more general discussion of Nietzsche's "genealogical" understanding of history, which combines the documentary and the critically reconstructive in a highly polemical perspective, see Michel Foucault, "Nietzsche, Genealogy, History" in Language, Counter-Memory, Practice, transl. Donald P. Bouchard and Sherry Simon (Ithaca, N.Y., 1977), 139-164. (The essays collected in this book represent an often critical supplement to the more well-known and somewhat doctrinaire "structuralist" positions which Foucault despite protestations to the contrary often develops in his principal works.)
are damaging in their dominant forms.) In the remainder of this paper, I shall try to put forth certain considerations that offer at least limited ways of addressing the questions I have raised.
I would distinguish between documentary and work-like aspects of the text.4 The documentary situates the text in terms of factual or literal dimensions involving reference to empirical reality and conveying information about it. The work-like supplements empirical reality by adding to, and subtracting from, it. It thereby involves dimensions of the text not reducible to the documentary, prominently including the roles of commitment, interpretation, and imagination. The work-like is productive and reproductive, for it deconstructs and reconstructs the given, in a sense repeating it but also bringing into the world something that did not exist before in that significant variation, alteration, or transformation. With deceptive simplicity, one might say that while the documentary marks a difference, the work-like makes a difference one that engages the reader in re-creative dialogue with the text and the problems it raises.
I shall return to this distinction and its implications in a somewhat different light. What I stress here is that the documentary and the work-like refer to aspects or components of the text that may be developed to different degrees and related to one another in a variety of ways. We usually refer to The Brothers Karamazov or The Phenomenology of Mind as works, and to a tax roll, will, or register of an inquisition as documents. But the work is situated in history in a way that gives it documentary dimensions, and the document has work-like aspects. In other words, both the "document" and the "work" are texts involving the interaction between documentary and work-like components in ways that should be interrogated in a critical historiography. Often the dimensions of the document that make it a text of a certain sort with its own historicity and its relations to socio-political processes (for example, relations of power) are filtered out when it is used purely and simply as a quarry for facts in the reconstruction of the past. (A register of an inquisition, for example, is itself a textual power structure with links to relations of power in the larger society. How it functions as a text is intimately and problematically related to its use for the reconstitution of life in the past.) Conversely, the more documentary aspects of a work are neglected
4. The notion of the "work-like" is of course indebted to Heideggei's discussion in "The Ougin of the Work of Art" in Poetry, Language, Thought, transl. Albert Hofstadter (New York, 1975), 15-87. The distinction between the documentary and the work-like may also be compared to J. L. Austin's distinction between the "constative" the descriptive statement that is measured against the criteria of truth and falsehood in "corresponding" to facts and the "performative" the doing of things with words that brings about a change in the situational context. On the approach I am suggesting, the constative and the performative are best seen not as generic types or sets of speech acts ("realms of discourse") but as more or less pronounced aspects of speech acts (or texts) that may be conceptually elaborated into analytic distinctions, ideal types, or heuristic fictions.
when it is read in a purely formalistic way or as an isolated source for the recovery of past meaning. Clearly, the larger questions at issue turn on the relations between documentary and work-like aspects of the text and between the correlative ways of reading it.
A dialogue with an "other" must be about a subject matter and convey information of some sort. But, as Weber and Collingwood have observed, a fact is a pertinent fact only with respect to a frame of reference that involves questions that we pose to the past, and it is the ability to pose the "right" questions that distinguishes productive scholarship. Heidegger has emphasized that these questions are themselves situated within a "context" or a "life-world" that cannot itself be entirely objectified or fully known. For Heidegger, moreover, it is only by inquiring into what a thinker did not think but might have thought that is, his or her significant "unthought" that one can gain some perspective on what he or she actually did think. Here anachronism is an obvious danger, but an imaginative kind of comparative history inquiring into the unexamined assumptions and unrealized possibilities of the past is nonetheless an important supplement to more empirical kinds of comparison in the dialogue between past and present. (Weber himself, it may be recalled, argued that the attribution of causal weight to an event or phenomenon itself depended upon its comparison with an imaginative rethinking of the historical process in which it figured. Only by hypothesizing what might have come to pass given the absence or significant variation of an event or phenomenon could one arrive at the understanding of transformational possibilities that enabled one to appreciate the fact that something occurred in the form it actually took.) Indeed, insofar as it is itself "work-like,'" a dialogue involves the interpreter's attempt to think further what is at issue in a text or a past "reality," and in the process the questioner is himself questioned by the "other." His own horizon is transformed as he confronts still living (but often submerged or silenced) possibilities solicited by an inquiry into the past. In this sense, the historicity of the historian is at issue both in the questions he poses and (pace Weber) in the "answers" he gives in a text that itself reticulates the documentary and the work-like. It may, finally, be argued that the dominance of either documentary or work-like tendencies generates tension, and this tension is neutralized only through processes of control and exclusion. These processes may operate both in the text being interpreted and in the text interpreting it. In relation to the texts of primary interest to intellectual history, they tend to operate more in our interpretations or uses of texts than in the texts themselves.5
A documentary approach to the reading of texts has predominated in general historiography and, in important respects, it has also characterized
5. This theme has been especially important in the works of Foucault and Derrida. For an application of it to the interpretation of Rousseau, see Jacques Derrida, De la grammatologie, transl. Gayatri Chakravorty Spivak (Paris, 1967), Part II.
intellectual history. If the dominance of this approach is open to question in other areas of historiography, it is perhaps even more questionable in intellectual history, given the texts it addresses." For certain texts themselves explore the interaction of various uses of language such as the documentary and the "work-like," and they do so in ways that raise the issue of the various possibilities in language use attendant upon this interaction. The Menippean satire is the most manifest example of a text openly exploring the interaction or dialogue among uses of language.7 But this issue may be raised in relation to any text in a manner that both opens it to an investigation of its functioning as discourse and opens the reader to the need for interpretation in his or her dialogue with it. Indeed there would seem to be something intrinsically wrongheaded in the idea of a purely or even predominantly documentary approach to a markedly "work-like" and "dialogical" text.
The predominance of a documentary approach in historiography is one crucial reason why complex texts are either excluded from the relevant historical record or read in an extremely reduced way. Within intellectual history, reduction takes the form of synoptic content analysis in the more narrative method and the form of an unproblematic identification of objects or entities of historical interest in the history of ideas.8 These entities are, of course, "ideas" ("unit-ideas" in the work of A.O. Lovejoy) or "structures of consciousness" or of "mind" (for example, in the work of Ernst Cassirer). Ideas or structures of consciousness are abstracted from texts and related to comprehensive, formalized modes of discourse or symbolic forms (philosophy, literature, science, myth, history, religion). How these structures actually function in complex texts is often not raised as an object of investigation
6. A critique of an exclusively or even predominantly documentary conception of historiography in general is developed by Hayden V. White in Tropics of Discourse. See especially the Introduction and Chapters 1-4. Curiously, White in this book does not explore the more specific applications of his arguments to intellectual history.
7. For an illuminating discussion of Fernand Braudel's Mediterranee et Ie monde mediterraneen a 1'epoque de Philippe II as a Menippean satire, see Hans Kellner, "Disorderly Conduct: Braudel's Mediterranean Satire," History and Theory 18 (1979), 197-222. Kellner, however, does not address two important issues: 1) the role of scientific discourse in Braudel's work and the problem of how it relates to other uses of language, and 2) the way in which the "Menippean satire" is not simply a category that allows one to identify the genre of a work but a multivalent use of language that may test the limits of genre classifications. These issues arise in a pointed way when one attempts to relate Northrop Frye's classical understanding of the menippea to the more carnivalized notion of Mikhail Bakhtin (notably in Problems of Dostoevski's Poetics, transl. R. W. Rotsel [Ann Arbor, 1973], 92-100). On Bakhtin, see Julia Kristeva, "Le Mot, Le Dialogue et Le Roman" in Semiotike (Paris, 1969), 143-173.
8. I shall not discuss further synoptic content analysis which is both necessary and limited as a method of analyzing complex texts. But for one of the most successful and perspicuous narratives relying essentially on synoptic methods, see Martin Jay, The Dialectical Imagination: A History of the Frankfurt School (Boston, 1973). In an extremely ambitious and intelligent work. Mark Poster, with some misgiving, also
or given only marginal attention The ideas or structures (for example, the idea of nature or the great chain of being) may then be traced over time and used to distinguish between periods This approach is criticized as excessively detached a form of history of ideas "in the air" from a more socially oriented perspective9 But, on a very basic level, the social history of ideas often shares the assumptions of the approach it criticizes For it too may take ideas, structures of consciousness, 01 ' mentalities" as relatively unproblematic entities and not raise the question of how they function in texts 01 actual uses of language Instead it will look into the causes or origins of ideas and their impact or effect in history In brief, social history often adjusts a history of ideas to a causal framework and a conception of the social matrix without critically investigating what is being caused or having an impact10 It may also lead to the idea that the only things worth studying are those that indeed did have a social impact or effect in their own time, thereby depriving historiography of the need to recover significant aspects of the past that may have "lost out."
A different understanding of intellectual history as a history of texts may permit a more cogent formulation of problems broached by more established approaches and a better inter change with the type of social history that relates discourse and institutions On this understanding, what was taken as an assumption or elided in the perspectives I have mentioned becomes a prob-
practices intellectual history as a narrative relating synopses or paraphrases of the arguments in texts to contextual developments (Existential Marxism in Postwar France From Sartre to Althusser [Priceton 1976]) For an attempted analysis and critique of the role of the synoptic method m contextuahst interpretation, see my "Reading Exemplars Wittgenstein's Vienna and Wittgenstein's Tractatus,' Diacritics 9 (1979), 65-82
9. See the critique of Cassirer in Peter Gay, 'The Social History of Ideas Ernst Cassirer and After" in The Critical Spmt Esvas in Hanoi of Herbert Marcuse, ed. Kurt H Wolff and Bainngton Moore, Ji (Boston, 1967), 106-120 Gay nonetheless praises Cassirer for his emphasis upon structure and his ability to find order in seeming chaos Gay does not raise the question of the extent to which the order thus found is limited or even specious (The point of this remark is to suggest that the imposition of "order and perspicuity" in one of Gibbon's favorite phrases upon the historical record is misleading and that the objective of the historian should rather be to explore critically the ways in which the interaction between order and its contestatory "others" takes place )
10. For an extremely successful example of this approach, inspired by the methods of the Annales school, see Daniel Roche, Le siecic cles himieies en piovince academies et academineiis piovinciaiiY 1680-1789, 2 vols (Pans, 1978) Roche s approach to the texts of Rousseau, of which he does not provide an extended critical analysis, may be contrasted with that of Derrida in De la grammatologie. There are, however, signs at present that those affiliated with the Annales school are developing an expanded notion of "le travail de texte that reveals the limitations of narrowly documentary readings See, for example Michel de Certeau, L Absent de l'histoire (Pans, 1978) and Roger Chartier, Jacques Le Goff, and Jacques Revel, La Nouvelle Histoire (Pans, 1978)
lem for inquiry. One such problem at the very crossroads of the documentary and the "dialogical" is that of the precise nature of the relation between texts and their various pertinent contexts I would break this problem down into six intersecting and partially overlapping areas of investigation and, in discussing them, I shall stress certain points that are often neglected at present
Before proceeding to a seriatim discussion of these "contexts," it may be helpful to make my own objective clear My list is not exhaustive And my point is to emphasize the need to recognize that, in treating the i elation of texts to contexts, one does indeed have a problem and that what is often taken as a solution should be reformulated and investigated as a real problem An appeal to the context does not eo ipso answer all questions in reading and interpretation And an appeal to the context is deceptive one never has at least in the case of complex texts the context The assumption that one does relies on a hypostatization of "context," often in the service of misleading organic or other overly reductive analogies In relation to complex texts, one has a set of interacting contexts whose relations to one another are variable and problematic and whose relation to the text being investigated raises difficult issues in interpretation. In addition, the assertion that a specific context or subset of contexts is especially significant in a given case has to be argued and not simply assumed or surreptitiously built into an explanatory model or framework of analysis With these caveats in mind, the six "contexts" I would single out for attention are intentions, motivations, society, culture, the corpus, and structure (or analogous concepts)
1) The relation between the author's intentions and the text. I would not deny the importance of intentions and of the attempt to specify their i elation-ship to what occurs in texts or in discourse more generally But speech-act theory has lent support to the extreme belief that the utterance and, presumably by extension, the text derive their meaning from the author's intentions in making or writing them Quentin Skinner has argued forcefully that the object of intellectual history should be the study of what authors meant to say in different historical contexts and communicative situations u This view tends to assume a proprietary relation between the author and the text as well as a unitary meaning for an utterance At best it permits an overly
11 See especially Skinner's "Meaning and Understanding in the History of Ideas " History and Theory 8 (1969) 3 53 For a defense of authorial intention as providing the criterion of valid interpretation, see E D Hirsch, Validity in Interpretation (New Haven, 1967) and The Aims of Interpretation (Chicago, 1976) For a critique of Hirsch, see David C Hoy, The Critical Circle (Berkeley, 1978) Hoys book is a good introduction to the works of Gadamer who offers an extensive criticism of the attempt to center interpretation on the mens auctores. A more fundamental critique is provided by Jacques Derrida, notably in 'Signature Event Context," transl. Samuel Weber and Jeffrey Mehlman, Glyph 1 (1977), and 'Limited Inc abc,' transl. Samuel Weber, Glyph 2 (1977)
simple idea of divisions or opposing tendencies in a text and of the relationships between texts and analytic classifications of them. By presenting the text solely as an "embodied" or realized "intentionality," it prevents one from formulating as an explicit problem the question of the relationship between intentions, insofar as they can be plausibly reconstructed, and what the text may be argued to do or to disclose. This relationship may involve multiple forms of tension, including self-contestation. Not only may the intention not fill out the text in a coherent or unified way; the intention or intentions of the author may be uncertain or radically ambivalent. Indeed the author may in good part discover his or her intentions in the act of writing or speaking itself. And the "reading" of intentions poses problems analogous to those involved in the reading of texts.
It is significant that an intention is often formulated retrospectively when the utterance or text has been subjected to interpretation with which the author does not agree. The first time around one may feel no need to make one's intentions altogether explicit or one may feel that this is impossible, perhaps because one is writing or saying something whose multiple meanings would be excessively reduced in the articulation of explicit intentions. Along with the "projection" of a goal that in part directs the writing process, an intention is a kind of proleptic reading or interpretation of a text. A retrospectively formulated intention is more manifestly a reading or interpretation, for it is rarely a transcription of what the author meant to say at the "original" time of writing. Insofar as there is a proprietary relation between the author and the text, especially in cases where his or her responsibility is at issue (for example, in cases at law), one may want to give special weight to statements of intention, at least to the extent that they are plausible interpretations of what actually goes on in a text. But, even if one is content merely to extend the relevant analogy, one can argue that, to some significant extent, tradition "expropriates" the author. For the texts of the tradition have entered the "public domain." Here the intentions of the author have the status either of aspects of the text (for example, when they are included in a preface) or interpretations of it which the commentator should certainly take into account but whose relation to the functioning of the text is open to question.
The idea that authorial intentions constitute the ultimate criterion for arriving at a valid interpretation of a text is motivated, I think, by excessively narrow moral, legal, and scientific presuppositions. Morally, and even legally, one may believe that a person should bear full responsibility for utterances and have a quasi-contractual or fully contractual relation to an interlocutor. Scientifically, one may seek a criterion that makes the meaning of a text subject to procedures of confirmation that leave minimal room for disagreement over interpretation. At times responsibility may be great enough to meet moral or legal demands, although this eventuality would satisfy neither the
theoretical nor the practical conditions for full freedom or intentionality. In any case, to believe that authorial intentions fully control the meaning or functioning of texts (for example, their serious or ironic quality) is to assume a predominantly normative position that is out of touch with important dimensions of language use and reader response. The more scientific demand is closely related to the moral one. It might be acceptable if it were applicable. To insist upon its applicability is to sacrifice more dialogical approaches and to obscure the role of argument in matters of interpretation, including the interpretation of intentions themselves. It is, moreover, commonplace to observe that a sign of a "classic" is the fact that its interpretation doss not lead to a definitive conclusion and that its history is very much the history of conflicting or divergent interpretations and uses of it. It is less commonplace to apply this insight to the process of argument that engages its interpreters. Insofar as an approach supplements the documentary with the dialogical, informed argument in it is to be seen not merely as an unavoidable necessity but as a valuable and stimulating activity that is bound up with the ways in which interpretation may be related to forms of renewal, including the renewal of beliefs to which one is deeply committed. The point here would be to do everything in one's power not to avoid argument but to make argument as informed, vital, and undogmatically open to counterargument as possible.
These considerations bear upon the question of criteria for a "good" interpretation. The latter should of course resolve documentary matters that are amenable to ordinary procedures of verification, and it should in important ways seek mutual understanding on larger issues of interpretation. But it is also necessary to recognize that, in equally important ways, it does not settle once and for all the question of how to understand a work or a corpus. A "good" interpretation reactivates the process of inquiry by opening up new avenues of investigation, criticism, and self-reflection. This is not to say that one should "fetishize" the new or become a slave to current ideas about what is "interesting." But it is to say that basic differences in interpretation (or mode of discourse) rarely turn on simple matters of fact and that on certain levels these differences may have a value that is not entirely subordinated to the ideal of consensus in interpretation. For they may relate to processes of contestation that have a critical role at present and that one would want to retain in some form in any social context.
2) The relation between the author's life and the text. This approach is inspired by the belief that there may be relations between life and text that go beyond and even contradict the author's intentions. What is sought in a psychobiographical perspective is the motivation of the author which may be only partly known or even unconscious to him or her. A difficulty analogous to that found in the intentionalist view arises, however, when there is an assumption of full unity or identity between life and texts that allows both to be situated in parallel or homologous fashion in a cycle of development or a
pattern of breakdown.12 The temptation is then to see the text as a sign or symptom of the life process even when the resultant understanding of their relationship is left on the level of suggestion rather than elaborated into a full-blown causal or interpretative theory.
Here, as in the case of intentionalist interpretation, what is taken as a solution should be posed as a problem. There may of course be symptomatic aspects of texts. But life and text may also be both internally marked and related to one another by processes that place identity in question. A text or a life may question itself in more or less explicit ways, and each may question the other. Insofar as they are distinguishable, life and text may be characterized by patterns of development or by forms of repetition that are not simply coincident and that may even challenge one another. In any case, a problem common to a written text and a lived "text" may be worked or played out differently in each, and these differential relations pose important problems for interpretation. And we read significant written texts not only because they are compensatory but also because they are supplementary: they add something to the ordinary life that as a matter of (perhaps unfortunate) fact might not exist without them.
In addition, for a writer who takes what he or she is doing seriously (an attitude not necessarily divorced from a view of art or even of writing in general as a form of serious play or of jesting in earnest), writing is itself a crucial way of life. At times the writer may be more willing to defend the writings than other dimensions of the life. One may find this attitude in certain ways objectionable or "alienated," but one has to take it into account. It may (as with Kierkegaard) be proffered not to establish the innocence of the writings or to promote a vision of art for art's sake but to articulate a situation of which the writer is himself critical. In other words, the writer too may want a world in which writing is less distinctive because the text of life is itself written in a better way.
The general problem for an attempt to relate life and texts is to arrive at an understanding of the "text" of a life and of the use of language in texts, and of the relation between them, that is sufficiently nuanced to do justice to them. To believe that a relatively simple idea of identity or of breakdown does justice to a life may at times be plausible, although certain lives are rather complex. To believe that a relatively simple understanding of "real-life" problems provides the causal or interpretative key to the meaning of the texts or to the interaction between life and texts is altogether implausible. This belief almost invariably prefaces an interpretation of the texts and their relation to life that is excessively reductive. By contrast, the investigation of the relation between life and texts does more to complicate the problem of
12. This assumption and the difficulties attendant upon it affect even so careful and well-documented a study as Jerrold Seigel's Marx's Fate: The Shape of A Life (Princeton, 1978).
interpretation than it does to simplify it For it supplements the difficulty of interpreting demanding texts with the difficulty of relating them in a cogent way to existential processes The text to be interpreted then becomes larger and probably more intricate, for it includes the written texts, in cases where writing itself may be a highly existential process, and other dimensions of the life that are not simply external to these texts Simplification occurs only to the extent that it is plausible to lead texts, or aspects of texts, as secondary elaborations or rationalizations And here there is always the possibility that a psychobiography will tell us more about its author than about the author being studied 13
3) The relation of society to texts At this point in my discussion, the intersecting nature of the categories I am using becomes evident One cannot discuss the individual life without significant reference to society and vice versa But I shall attempt to focus on problems that have been taken to be more specifically social or sociological in nature (And I shall do so not from the perspective of a social history that inquires into the uses of texts for the empirical reconstitution of past society but from the complementary perspective of an intellectual history that inquires into the relationship between social processes and the interpretation of texts ) These problems have often been seen in terms of the "before ' and "after" of the text its genesis and impact
I have already indicated that the problem which is often elided or not emphasized in a social history of ideas is that of the relation of social to textual processes a relation that notions of "genesis" and "impact" may be inadequate to formulate Foucault, aware of this problem, has been suggestive in the elaboration of a notion of discursive practice which signals the interaction between institutions and forms of discourse But Foucalt has not been altogether satisfying in attempting to relate the discursive practice to the significant text or, even more generally, in articulating the relationship between more or less formalized modes of discourse and written or lived "texts " For he often treats written texts and other phenomena in a similar manner by falling back on the notion that they are instances or tokens of the discursive practice signs of the times In certain respects, this understanding may be accurate A text may include discursive practices or modes of discourse in a relatively straightforward way Marxist interpretation has often seen a similar relationship between ideology and text, and while Foucault's notion of dis-
13 This problem may arise in a rather subtle form The psychohistorian may make a dichotomous opposition between fully logical or rational arguments and illogical or irrational arguments in the text and assert that psychohistorical methods apply only to the interpretation of the latter For a well reasoned and careful exposition of this view, see Gerald Izenberg, Psychohistory and Intellectual History History and Theory 14 (1975), 139 155 The problem here is whether this extremely neat opposition applies to the text in question or whether it reflects the perspective of the analyst In any case, it obviates inquiry into the interaction between the "logical" and the "idelogical" in the functioning of the text itself
cursive practice is more comprehensive than the notion of ideology as false consciousness, it relies on an understanding of relationship comparable to the more orthodox Marxist one.
But in both Foucault and in certain Marxists a different possibility at times arises The text may then be seen not only to include discursive practices or ideologies in a relatively straightforward way but also to engage in processes that, whether consciously or not, render them problematic, at times with critical implications The question then becomes that of how precisely the discursive practice, deep structure, or ideology even the prejudice is situated in the text other than in terms of instantiation 01 simple reflection The locus classicus of this type of inquiry may still in certain respects be Lukacs' investigation of the relationship between conservative ideology and what the text discloses about social processes in the works of Balzac 14 But in Lukacs the understanding of language use and textual process was often not subtle or searching enough to account for the interaction between text and society The almost Platonic vehemence of his condemnation of modernist literature not only illustrates this point, it points to problems that are intimated in Lukacs' texts themselves but in a manner that remains "unthought" or not rendered explicit
This is where the work of Derrida may hold out possibilities for the type of inquiry into the interaction between text and social process that Derrida himself seems rarely to undertake in an overt way Deriida's elaborate critique of Foucault's reading of Descartes in Histoire de la Folie should not be seen as a simple rejection of Foucault's interpretation " Rather it directs attention to the question of precisely where and how the exclusion of madness takes place in Descartes' text and whether that text may be understood as a straightforward sign of the times Dernda's argument in "Cogito et histoire de la folie" must be seen in the broader context of his understanding of the long but tangled tradition constituting the history of metaphysics It must
14 See his Studies in European Realism (New York, 1964)
15 Michel Foucault, Folie et deiaison Histoiie de la folie a I'age classique  (Pans, 1972) with an Appendix in which Foucault responds to Dernda For Dernda's essay see LEcntme et la difference (Pans, 1967), 51-97 In "The Problem of Textuality Two Exemplary Positions," Ciitical Inquiry 4 (1978), 673-714, Edward Said argues that Denida's 'deconstructive" criticism remains within the text while Foucault's history of discursive practices takes one into "thick' historical reality where various "discourses of power' and related dominant institutions have ruled the production of texts In a valid attempt to stress the political importance of Foucault's concerns, Sdid ignores Derrida's extension of textuality" beyond the confines of the book, and he fails to see how Foucault's view at times tends to reduce the complex text to a token of a mode of discourse Nor does Said pose the problem of how the complex text may both "reflect' or inscribe dominant modes of discourse and also challenge them, at times with significant critical effects The view of modern history that emerges from this perspective veers toward a rather monochromatic story of repression in which the contestatory role of certain texts is not investigated The consequences of this limited view mark Said's own Orientalism (New York, 1978)
also be seen with reference both to the problem of relating a text to its times and to the way in which texts may place in radical question their own seemingly dominant desires and themes.
The manifest division between Derrida and Foucault takes place over the local interpretation of a passage in Descartes' first Meditation. Where Foucault locates the exclusion of madness that inaugurates or confirms its status in the Classical Age, Derrida sees a "pedagogical" and dialogical discursive process that does not exclude madness but includes it in a movement of increasing hyperbole. Derrida in effect gives a new turn to the very classical argument that one cannot take the passage in question out of context but must relate it to the overall movement of the text. The passage itself is too uncertain to allow for a definitive interpretation, and it may be impossible to decide whether Derrida or Foucault gives the better interpretation. There is something to be said for both accounts.
The more interesting moment in Derrida's analysis comes when he discusses the point of hyperbolic doubt in Descartes, which seems to be open to the possibility of madness and to occur on a level that undercuts the opposition between madness and reason. But this point of extreme hyperbole is almost immediately followed in Descartes by a gesture that does seem definitively to exclude madness and to establish a firm foundation for reason. Thus for Derrida, Descartes also excludes madness but he does so in a manner that repeats in modified form both the traditional philosophical desire for a firm and fully unified foundation for reason and the hyperbole that, at least momentarily, seems to subvert or contest that desire. In fact, in Descartes the moment of hyperbolic doubt and radical contestation is more explicit and less masked than it is in many philosophers as Derrida himself interprets them. The larger question raised in Derrida's analysis is that of relating long but intricate traditions, such as the history of metaphysics, the specific period or time (including some delimited structural or epistemological definition of it), and the specific text. The attempt to delineate the mode of interaction among them requires an interpretation of the text in all its subtlety, and it indicates the importance for historical understanding of a notion of repetition with variation over time. In this respect, the relation among long tradition, specific time, and text cannot be determined through a notion of either simple continuity or discontinuity. Nor can the text be seen as a simple instantiation or illustration of either the long tradition or the specific time. Rather the problem becomes that of the way in which long tradition, specific time, and text repeat one another with variations, and the matter for elucidation becomes the degree of importance of these variations and how to construe it. The text is seen as the "place" where long tradition and specific time intersect, and it effects variations on both. But the text is not immobilized or presented as an autonomous node;
it is situated in a fully relational network.
This network is the context for one of the most difficult issues for interpretation: how the critical and the symptomatic interact in a text or a work of art. Only by exploring this issue in a sustained way can one avoid the one-sidedness of analyses that stress either the symptomatic and representative nature of art (as did even Lukacs or Lucien Goldmann for whom art was critical solely as an expression of larger forces) or the way in which "great" art is itself an exceptional, critical force for constructive change (as those affiliated with the Frankfurt school tended to argue). Stated in somewhat different terms, this is the issue of the extent to which art serves the escapist function of imaginary compensation for the defects of empirical reality and the extent to which it serves the contestatory function of questioning the empirical in a manner that has broader implications for the leading of life. One might suggest that texts and art works are ambivalent with respect to this issue, but that they differ in the ways in which they come to terms with this ambivalence. A criterion of "greatness'' or at least of significance might well be the ability of certain texts or works of art to generate a heightened sense of the problematic nature of this ambivalence and yet to point beyond it to another level of ambivalence where the very opposition between escapism and criticism seems to become tenuous indeed where oppositions in general founder and emerge in the passage between radical hyperbole and delimited structures. There is no ready formula for "decoding" the relations among the symptomatic, the critical, and what Derrida terms the "undecidable," but the attempt to interpret significant modern works forces one to confront the problem of what to make of these relations.
The question of "impact" is best seen in terms of complex series of readings and uses which texts undergo over time, including the process by which certain texts are canonized. Any text reaches us overlaid and even overburdened by interpretations to which we are consciously or unconsciously indebted. Canonization itself is a procedure not only of selection but of selective interpretation, often in the direction of domestication. We as interpreters are situated in a sedimented layering of readings that demand excavation. But the process of gaining perspective on our own interpretations does not exclude the attempt to arrive at an interpretation we are willing to defend. Indeed the activity of relating the existing series of interpretations, uses, and abuses of a text or a corpus to a reading that one tries to make as good as possible is essential to a critical historiography. Of course, this is not to say that the interpretation one offers is definitive and exhaustive. Not only is it open to revision through argument and reconsideration; it may actively address the issue of how the text itself resists the "closure" of definitive and exhaustive interpretation. And it should be agitated by the realization that we are inevitably blind to certain dimensions of our own perspective and what it fails to see. But we interpret nevertheless. And, unless we interpret, our reference to a text becomes purely nominal, and we trace the movement
of an "I-know-not-what" across time To proceed in this manner is to abandon all hope of attaining a critical understanding of what is involved in the "impact" of texts Indeed one might well argue that what is needed at present at the intersection of intellectual and social history is precisely an approach that relates an informed interpretation of complex texts to the problem of the way in which those texts have been adapted to and in certain ways have allowed important uses and abuses over time. The cases of Marx, Nietzsche, and Heidegger call for treatment of this kind l6
One important area for the study of impact that has not been sufficiently investigated is that of the readings of texts at trials The trial serves as one instance of social leading that brings out conventions of interpretation in an important social institution It is highly significant that in their basic assumptions about reading, the prosecution and the defense may share a great deal, and what they share may be quite distant from, or even placed on trial in, the work itself (One thing which a trial must repress is the way in which style, as Flaubert realized, may be a politically subversive or contestatory force that can be more unsettling and less easily processed than the revolu-
16 George Steiner's Martin Heidegger (New York, 1978) is one of the best short introductions to Heidegger's thought, and it raises the problem of the relation between that thought and Heidegger's life But the way Steiner addresses the latter issue is too summary and extreme to be altogether acceptable Steiner asserts an "organic" relation between the "vocabulary" of Being and Time (especially its later sections) and Heidegger's addresses of 1933 as well as "instrumental connections" between the "language" and "vision" of Heidegger's treatise and Nazi ideology (121-123) The mixing of metaphors to designate the nature of the relationships at issue itself indicates that they demand a more careful and extensive investigation than Steiner allows He provides some crucial pieces for this investigation in stressing the dangers of the more mesmeric sides of Heidegger's thought and the combination in it at times of near total criticism of the present and vague apocalyptic hope for the future These aspects of Heidegger's thought may well have helped to induce the belief that the Nazis were the bearers of the fundamental change in modern civilization that Heidegger desired And Steiner is, I think, justified in insisting upon the question of why Heidegger after 1945 remained publicly silent about his brief affiliation with the Nazis What is also significant is that Steiner, in spite of his strictures about Heidegger's relation to the Nazis, wants to argue that Heideggei's work remains a basic and valuable contribution to modern thought But the elucidation of the relation of that thought (notably Being and Time) to Nazi ideology and to Heidegger's own brief participation in politics requires an interpretation that inquires into the question of how the "same" themes or "ideas" function differently (or even in opposed ways) in different texts and contexts and how they may be both used and abused not only by others but by the -author himself in certain circumstances Steiner does not provide this extremely difficult kind of interpretation His own categorical response not only creates unexamined divisions in his own account (For Steiner himself argues that certain discussions in Being and Time constitute a radical critique of totalitarism ) It also leads him to the extremely dubious and unsupported assertion that the later sections of Being and Time (the ones presumably closer to Nazi ideology) are less persuasive and more opaque in comparison with the earlier portions of the text The specious but useful separation between "good" and "bad" parts of the text is, I think, too simple an answer to the admittedly intricate question that Steiner has the merit of raising
tionary message packaged in conventional forms.) In the modern West, it is writers of "literature" who have been placed on trial for their writings. This may be one sign of the more contestatory character of literature in comparison with other modes of discourse in modern society, at least on levels where a more general public may sense that something disconcerting is happening although the reasons given for trial often neutralize this sense by appealing to very conventional criteria of judgment, for example, "prurient interest." Writers of more "theoretical" or philosophical works in the West are tried more informally through critical responses to their works. This has happened to Nietzsche, Heidegger, Derrida, and even Wittgenstein philosophers who have perhaps gone farthest in challenging traditional understandings of philosophical discourse. In this sense, the history of critical response, including the book review, is an important chapter in the history of social impact, especially with reference to the constitution and development of disciplines. One can often learn more about the operative structure of a discipline from its book reviews and from their differential distribution in different sorts of journals than one can from its more formal institutional organization.
4) The relation of culture to texts. The circulation or noncirculation of texts among levels of culture is an intricate problem, and difficulties arise even at the stage of deciding upon terminology to name or identify "levels." The approach to intellectual history I have been defending is directed at what has been traditionally (and now often derogatorily) called "high" or "elite" culture. The dissemination of the "great" texts of at least the modern period to a larger audience is frequently at best a desideratum. At times it is actively opposed by important writers and intellectuals; one may wonder though to what extent this reaction is a defense against rejection by a larger public, for modern texts often make demands upon the reader that few readers even readers within the so-called educated class may be willing to accept. One crucial function of a more "recuperative" kind of intellectual history has been to disseminate these texts to the "generally educated" class in a "digestible" or "assimilable" form that may have little in common with the texts themselves and may even function as an excuse not to read them. Here I would note a general difference between a documentary and a "dialogical" approach to history. Insofar as an approach is documentary, it may validly function as a processing of "primary source material" that enables the nonexpert reader not to go to the sources or the archives themselves. But the very point of a "dialogical" approach is to stimulate the reader to respond critically to the interpretation it offers through his or her own reading or rereading of the primary texts.
To the extent that a text is not a mere document, it supplements existing reality, often by pointing out the weaknesses of prevailing definitions of it. In a more traditional context, texts may function to shore up norms and values that are threatened but still experienced as viable. For example,
Chretien de Troyes had the quests of his knights test and ultimately prove the validity of courtly values that were threatened in the larger society.17 In a more revolutionary context, texts may help to break down the existing system and to suggest avenues of change. But it is at times difficult to distinguish clearly between the traditional and the revolutionary context. And any text that pinpoints weaknesses in a system has an ambivalent function, for it may always be read against its own dominant tendency or authorial intention a "conservative" text being used for ''radical" purposes or vice versa. The fate of Marx in the hands of his liberal and conservative critics is illustrative in this respect.
Most modern writers of note have seen their period as revolutionary or at least as "transitional." Indeed they have often been "alienated" from what they perceive as the dominant society and culture. Even significant conservatives, such as Burke and Maistre, do not simply defend a status quo, but often inveigh against it in defense of values they believe to be under full siege and rapidly disappearing. They may advocate a context in which the attachment to values, norms, and communal groups is prereflective or quasi-instinctive, but they are forced to become highly reflective intellectuals in spite of themselves. More often than not, the modern conservative is a divided self who may even harbor rather radical tendencies. This tension is quite evident in Dostoevsky and Balzac. And, for conservatives and radicals alike, the very notion of a popular culture to which they might relate arose largely as an ideal, critical fiction, or goal to be opposed to "modern" forces that jeopardized those at times vestigial forms of popular culture they judged to be desirable.
One might argue that the global society or culture is too large and undifferentiated a unit for the investigation of the most relevant community of discourse for intellectuals. The more delimited school, movement, network of associations, or reference group would seem to provide the more immediate complex of shared assumptions or pertinent considerations that operate in a tacit or explicit way to shape the intellectual's sense of significant questions and modes of inquiry. Hence intellectual history should be a history of intellectuals, of the communities of discourse in which they function, and of the varying relations ranging in often complicated ways from insulation to openness they manifest toward the larger culture. Thinkers as diverse as T.S. Kuhn, Quentin Skinner, R.G. Collingwood, and Michel Foucault may be drawn upon in the attempt to elaborate and to apply this view. This view has much to commend it, but I would like to raise at least two problems it sometimes generates.
First, it may be used to restrict historical inquiry to the historicist and documentary attempt to recreate the dialogue of others but to prohibit the
17. The example of Chretien de Troyes is discussed by Wolfgang Iser in The Act of Reading (Baltimore,1978), 77-78.
extension of that dialogue to include the interpretations of the historian, perhaps on the grounds that epochs are primarily if not exclusively dissociated in their forms of understanding. (The latter grounds give rise to damaging aporias that are too well known to be rehearsed here.) Even more often, the assumption is made that historical understanding itself is (or should be) purely "objective" and that the very notion of an informed dialogue with the past is absurd or at least nonhistorical. This position not only identifies the historical with the historicist and the documentary. It may also construe the notion of dialogue in a simplistic way (for example, in terms of drawing immediate lessons from the past or of projecting our own particular or subjective concerns onto it). One may, however, argue that the reconstruction of the dialogues of the dead should be self-consciously combined with the interpretative attempt to enter into an exchange with them that is itself dialogical only insofar as it actively recognizes the difficulties of communication across time and the importance of understanding as clearly and as fully as possible what the other is trying to say. To the extent that the past is investigated in terms of its most particularized aspects, the dialogue with it becomes minimal. But the subsequent question is whether historical research should be directed primarily to aspects of this sort which restrict the historian's use of language to predominantly informational and analytic functions. When it is applied to the works of major figures, an approach that attempts to be exclusively documentary is often deadly in its consequences. And when we historians who are trained to believe in the primacy of the documentary ideal do venture to put forth interpretations or critical judgments, the latter may well be of little interest, for they are not the products of a richer and more varied discursive background. Here, of course, we face the traditional problem of how the educator is himself to be educated.
Second, the focus on communities of discourse must be cogently related to the problem of textual interpretation. It is not enough to establish influence or the existence of a shared "paradigm" through the enumeration of common presuppositions, questions, themes, or arguments. One must elucidate in a more precise and detailed way how the borrowed or the common actually functions in the texts in question. To document common assumptions or lines of influence may suffice to debunk the myth of absolute originality. But this procedure easily gives rise to its own forms of self-deception or even mindless chronicling when debunking goes to the extreme of not recognizing why there is, for example, a major difference between a Fliess and a Freud. Influence studies are of minor interest unless they address the issue of how common ideas function differentially in different texts and corpuses, and even the attempt to dethrone a reigning "great" must face up to the problem of interpreting his works in their complexity. All too often the focus on the community of discourse leads the historian to limit research to minor figures or to highly restricted and unsituated aspects of the thought of a major figure
(for example, Nietzsche's elitism, Marx's utopianism, or Freud's biologism). In addition, the delimited "communities" in which major modern intellectuals participate may themselves be made up of more of the dead or absent than of the living or present. In other words, the most significant reference group may most prominently include dead or distant (even future) "others" who become relevant largely through their works which the "creative" intellectual helps bring to life in his or her own works through emulation, selective appropriation, parody, polemic, anticipation, and so forth. The more literal person-to-person group may have a lesser significance for the actual production of "ideas," and in any case its role is always supplemented by the relation to others through their texts or other artifacts. The dialogue with the text may even be experienced as more immediate and engrossing than most conversations. Indeed one of the re-creative implications of reading might well be to attempt to create social and cultural conditions in which the literal conversation and the general text of life are more like the processes stimulated by an encounter with a great text.
This last consideration provides a limited avenue of re-entry into the question of the relation between "great" texts and general or popular culture. The processes which Mikhail Bakhtin discusses in terms of "carnivalization" help to identify at least one kind of popular culture activated or reactivated in the texts of many significant writers and often desired by them as the larger context to which their writings might relate. "Carnivalization" as employed by Bakhtin is epitomized in carnival as a social institution but not restricted to it. In its larger sense, "carnivalization" is an engaging process of interaction through which seeming "opposites" body and spirit, positive and negative, high and low, seriousness and laughter are related to one another in a contestatory interchange that is both literally and figuratively "re-creative." It is set within a more encompassing rhythm of social life, and one might argue that its functions depend upon that larger setting. Carnival itself is an ambivalent "safety-valve" that may include forms of protest reaching rebellious proportions in certain oppressive situations.
For Bakhtin, a lively interchange among carnival as a social institution, popular culture, and high culture did exist in Rabelais' Renaissance. While aspects of elite culture were closed to the common people (for example, works written in Latin), the elite participated in popular culture, and aspects of that culture affected high culture. Thus Bakhtin can interpret the works of Rabelais as drawing upon, and feeding into, a rich and vital popular culture. The modern period witnessed the decline of carnival, the separation of the elite from popular culture, and the detachment of processes of carnivalization in literature from significant public institutions. Indeed great literature is for Bakhtin the primary repository of the modern carnivalesque in its more restricted state. Social forms of carnivalization have tended to withdraw to the private sphere, for example, in the domestic celebration of
holidays And the modern literary carnivalesque has itself often gravitated toward more reduced extremes, such as largely negative irony and hysterically shrill laughter Bakhtin s own analysis of the works of Dostoevsky, however, itself indicates the more re-creative possibilities of carnivahzation and "grotesque realism" in modern literature ls
The decline of carnival was directly related to religious reform and more indirectly related to the multiple processes gathered together under the label of "modernization " The withdrawal of the elites from popular culture was a long-term process extending from 1500 to 1800, and what had been everyone's second culture was rediscovered as an exotic residue of the past In the nineteenth century, the turn toward folklore and other forms of popular culture was often one aspect of vanous responses to the perceived "excesses" of the Enlightenment In post-Enlightenment writing, there is, moreover, a critical relation to the dominant society in which "carnivalization' is underplayed or repressed Here, for example, we have one basis for Nietzsche's critique of positivism as a eunuch-like flight from carnivalesque contestation the Nietzschean version of the "betrayal of the intellectuals " We also have one way of viewing Flaubert's "post-romantic" notion of "puie" art as an ironic and stylistically insuirectionary variant of the tiansformed caini-valesque More generally, the notion of carnivalization provides one way of interpreting the contestatory styles with political overtones that have been so characteristic of modern writing Indeed processes of carnivalization might even be i elated to social and political action in part inspired by a desire for a context of "lived experience" more open to revitalizing, contestatory forms One need not look to the distant past for examples of this phenomenon Not only may aspects of the work of recent French figures (for example, Foucault, Deleuze, Sellers, Kristeva, and Derrida) be seen in terms of processes of carnivalization, but the events of 1968 have been interpreted in this way, often with the term "carnival" being used in a pejoiative sense by opponents and at times with the vision of "carnivalization" becoming a pretext for romanticization in defenders of the evenements 19 The larger problem these
18 See Mikhail Bakhtin, Rabelais and His World, transl. Helene Iswolsky (Cambridge, Mass 1968) and Problems of Dostoevsky's Poetics transl. R W Rotsel (Ann Arbor, 1973) Foi an application of certain of Bakhtin's views, see Natalie Z Davis, Society and Culture in Earl Modem France (Stanford 1975).
19 For a treatment of early modern culture (or cultures) with a discussion of the role of carnival its decline over time, and the withdrawal of the elites from popular culture, see Peter Burke, Popular Culture in Earl Modern Europe (New York, 1978) For a study of French culture emphasizing the problem of carnivalization, including its role in 1968, see Maurice Crubellier, Histoire culturelle de la France (Pans, 1974) A more detailed discussion of the history of the carnivalesque and its relation to various writers would require qualifications and discriminations that I have not provided For an analysis of Flaubert and carnivalization, see Arthus Mitzman, ' Roads, Vulgarity, Rebellion and Pure Art The Inner Space in Flaubert and French Culture, Journal of Modern History 51 (1979), 504 524 It must, however, be noted that the danger
considerations bring out is that of the way in which innovatory "elite" responses in the modern period may appeal to older "popular" culture in the radical critique of what is perceived as the dominant socio-cultural context.
5) The relation of a text to the corpus of a writer. The notion of context provided by other texts is itself apparently textual in nature (although one must recall that "corpus" here may also mean "body"). And it raises the problem of the relationship between a text and the texts of other writers as well as other texts of the same writer. For what is at issue in this problem is precisely the unity or identity of a corpus. Often the corpus is seen in one of three ways: continuity among texts ("linear development"); discontinuity among texts (change or even "epistemological break" between stages or periods); and dialectical synthesis (the later stage raises the earlier one to a higher level of insight). The corpus is thus unified in one way or another (developmental unity, two discrete unities, higher unity). Seen in these ways, the corpus is like a single text writ large, for the single text may be interpreted through the use of these categories. The question, however, is whether these categories are too simple for the interpretation of the functioning of both the complex text and the corpus of complex texts. The relation among aspects or elements of a text, and a fortiori among texts in a corpus, may involve uneven development and differing forms of repetition or displacement that place in question simple models of intelligibility. Indeed the "corpus" of a writer may be at least partially dismembered, at times in ways that are intended or explicitly explored by the writer himself. Carnivalization as described by Bakhtin involves dismemberment or creative undoing that may be related to processes of renewal. One strategy of dismemberment is the use of montage and quotation through which the text is laced or even strewn with parts of other texts both written texts and elements of social discourse. In Flaubert, for example, the text is punctuated with parodic citations from other novels and from the cliches of daily life. In Mann and Joyce, the montage technique assumes panoramic proportions in its ability to piece or graft together various uses of discourse. In Sellers or Derrida, dismemberment
in Flaubert is that the leveling tendency he saw operating in modern culture affected his own approach in an uncritical way, leading him at times to a homogenized and nearly nihilistic condemnation of both modern society and humanity in general. In his most lapidary and famous formulations of the ideal of "pure art," Flaubert "sublimated" the carnivalesque into its opposite: an ascetic negation of reality and an attempt to transcend it into an inviolate sphere of beauty or absolute style. The more compelling and subtle dynamic in his stories is that whereby a vision of pure art is not simply exemplified but contested and even "carnivalized," notably through the treatment of analogous forms of the quest for the absolute and the empathetic-ironic modulations of narrative voice in the so-called "free indirect style." More generally, art was for Flaubert the most important commitment in the world and the work of a clown. It is important to recognize that he affirmed both these conceptions of art with great intensity and that there is in his works a variable tension between a pathos of belief and critical forms such as irony, parody, and self-parody.
(involving textual distribution of the "self") may at times seem to attain Dionysian heights or depths. The larger question raised by these strategies is that of the interaction between the quest for unity, which may continue to function in direct or parodic ways, and challenges to that quest operative or experimented with in the texts themselves. The texts, however, do not become hermetically closed upon themselves: they both differ from and defer to other texts both written and lived.
6) The relation between modes of discourse and texts. In the recent past, considerable attention has been paid to the role of more or less formalized modes of discourse, structures of interpretation, and conventions or rules. Many theorists have argued that writing and reading are informed by structures or conventions that should be a primary if not exclusive center of critical interest.20 In his own work, Hayden White has attempted to arrive at a level of deep structure that undercuts the opposition between literature and history to reveal how modes of emplotment inform all coherent narratives and how tropes construct the linguistic field. He has also indicated that figurative uses of language connect the levels of description and explicit interpretation or explanation in prose narratives. This last point serves to raise again a question that has not been sufficiently explored in various structuralist accounts of discourse: the question of how various modes of discourse, rules, or conventions actually function in texts or extended uses of language. In this regard, the reading of "minor" texts is certainly important for the attempt to establish what the dominant rules or conventions of a genre were at a given time. But the relation of a "great" text to genres both the ones it is placing in question and the ones it is helping to establish is always problematic, and even the "minor" text may hold out some surprises here. Often, however, it is assumed that this relation is one of coverage on the part of structures and instantiation on the part of texts. This view (which may actually be held by theorists who attack "positivism" in other respects) leads to the belief that there are unproblematic realms of discourse that are illustrated by texts that fall within them.
This view is misleading as it relates to the status of analytic distinctions or structural oppositions and to the question of how these distinctions or oppositions function in texts.21 Analytic distinctions such as those drawn between history and literature, fact and fiction, concept and metaphor, the serious and the ironic, and so forth, do not define realms of discourse that unproblematically characterize or govern extended uses of language. Instead what should
20. For one especially forceful elaboration of this view, see Jonathan Culler, Structuralist Poetics (Ithaca, N.Y., 1975). For an equally forceful exploration of some of its limitations, see Jonathan Culler, Flaubert: The Uses of Uncertainty (Ithaca, N.Y., 1974).
21. I try to develop this assertion with reference to the thought of Jurgen Habermas in "Habermas and the Grounding of Critical Theory," History and Theory 16 (1977), 237-264.
be taken as a problem for inquiry is the nature of the relationships among various analytically denned distinctions in the actual functioning of language, including the use of language by theorists attempting to define and to defend analytic distinctions or oppositions in their conceptual purity. To say this is neither to advocate the obliteration of all distinctions nor to offer a purely homogeneous understanding of a mysterious entity called the "text." It is rather to direct attention to problems that are obscured when one relies uncritically on the concept of "realms of discourse." For example, it is common to distinguish history from literature on the grounds that history deals in the realm of fact while literature moves in the realm of fiction. It is true that the historian may not invent his facts or references while the "literary" writer may, and in this respect the "literary" writer has a greater margin of freedom in exploring relationships. But, on other levels, historians do make use of heuristic fictions and models to orient their research into facts, and the question I have tried to raise is whether historians are restricted to the reporting and analysis of facts in their exchange with the past. Conversely, literature borrows from a factual repertoire in multiple ways, and the transplantation of the documentary has a carry-over effect that invalidates attempts to see literature in terms of a pure suspension of reference to "reality" or a transcendence of the empirical into the purely imaginary. Even when literature attempts to "bracket" empirical reality or to suspend more ordinary documentary functions it engages in a self-referential work or praxis through which the text documents its own mode of production. The very prevalence of literature about literature or art about art raises the question of how to interpret self-referential activity with respect to a larger historical context. Thus there certainly are distinctions to be drawn, but the problem is the manner in which these distinctions function in texts and in our reading or interpretation of them.
In the last-mentioned respect, there are different possibilities from the dominance of a given analytic distinction or type to more open interplay and contestation among various uses of language. But dominance implies some form of subordination or exclusion, and how this relationship is established must be investigated. Any critique of pure identities, pure oppositions, and attendant hierarchies must pay close attention to the way these categories function, for they have indeed been of decisive importance in thought and in life. One has certainly witnessed a quest for pure fact, pure fiction, pure philosophy, pure poetry, pure prose, and so forth. For those committed to one or another variant of it, the quest is taken at face value and defended. It may also be institutionalized in disciplines that organize themselves around conventions and rules that restrict language to certain uses and prohibit or sanction the attempt to raise questions that problematize these restricted uses. One of the largest of these questions is whether the quest for purity and the direct projection of analytic categories onto "reality" are related to a "meta-
physics of the proper" whereby one's own identity, propriety, or authenticity is established through the identification of a totally different "other" or outsider who may even become a pariah or scapegoat. In any case, the problem is how seeming purity (or unmarked identity and unity) is established and whether the quest for it in language use is contested by other aspects of the text or more general linguistic context in which this quest takes place. A text itself may of course seek purity by engaging in procedures of exclusion or domination that tend to neutralize or reduce its more disconcerting or contestatory movements. These procedures provide points of entry for interpretations or entire disciplines that "found" themselves upon the purity and autonomy of "realms of discourse" putatively emanating from certain master texts. But complex texts may well involve other movements that test the desire for unity in a variety of ways that demand detailed investigation. Indeed certain texts that seem to rely exclusively on one function or abstract dimension of language, for example, analytic dissociation or simulated denotative usage and attendant metaphoric deprivation in Beckett, may involve parody and stimulate in the reader a contestatory awareness of other possible uses of language. In fact the question is whether any text that seems successful in the sustained reliance on one function or analytic aspect of language for example, the accumulation of facts or of theoretical reflections is engaged in intentional or unintentional parody or self-parody, or at least may always be read in that way.
These points indicate that analytic distinctions are useful for purposes of clarification and orientation as pursued on an ideal level in their pure or "laboratory'' form, but that they never function "as such" in actual discourse or in texts. When they do seem to be used purely as such, other processes are at work or at play. The critique of these processes is prone to its own excesses (irrationalism, political quietism, anomic disorientation, the quest for full liberation either of or from libidinal demands). At its best, however, this critique may raise the problem of a more viable interaction between forms of language and forms of life. The exploration of this problem in the great texts of the tradition constitutes an especially engaging adventure that at times involves a strangely disconcerting way of making us think seemingly "alien" thoughts that are in fact within us and that may well return in unmitigatedly destructive ways when they are simply repressed or excluded. This problem may not be seen or appreciated when texts are read in an excessively reductive way or delegated exclusively to discrete disciplines. No discipline has the imperial right to dominion over a Freud, Marx, Nietzsche, or Joyce. (The practical advantage of intellectual history in this respect is that it can, without excuse or subterfuge, explore the problem of reading various texts together and thus raise questions about their functioning as language that might not otherwise be apparent.) Indeed, as I have already intimated, a discipline may constitute itself in part through reductive readings of its
important texts readings that are contested by the "founding" texts themselves in significant ways.^ These readings render the texts less multifaceted and perhaps less critical but more operational for organized research. Here the more decisive role of certain disciples and practitioners lies not in the fine-tuning of a paradigm enunciated in "founding" texts but in the active reduction of those texts to their paradigmatic level.
With reference to the still somewhat ill-defined discipline of history, the "great" texts should be part of the pertinent record for all historians. They are certainly part and parcel of a general historical culture. Something excessively reductive has already taken place when they are assigned to the sub-discipline of intellectual history, which may then function as a park or reserve for them. But at least within that subdiscipline, they should be read with an eye for the broader and occasionally uncanny processes they engage and that engage us. One such process is precisely the interaction between the desire for unity, identity, or purity, and the forces that contest it. The investigation of this process does not imply a simple rejection of conceptions of unity or order in a mindlessly antinomian celebration of chaos and dismemberment. What it calls for is a rethinking of the concept of unity and its analogues in more workable and critical terms. It also requires a sensitivity to the way in which these concepts are related to their "adversaries" in the texts we study and in our own attempts at theoretical self-understanding. One practical implication of these considerations is the possibility of reconstructing norms and conventions in forms that may be more durable precisely because they enable us to contend better with criticism and contestation. In this respect, a function of dialogue with the past is to further the attempt to ascertain what deserves to be preserved, rehabilitated, or critically transformed in tradition.
I would like to conclude by returning to the distinction I drew between intellectual history as a reconstruction of the past and intellectual history as a dialogue or conversation with the past a distinction that should not be taken as a purely dichotomous opposition The reconstruction of the past is an important endeavor, and reliable documentation is a crucial component of any approach that claims to be historical. But the dominance of a documentary conception distorts our understanding of both historiography and the historical process. Indeed I have tried to suggest that a purely documentary conception of historiography is itself an ideal type or heuristic fiction. For description is never pure, in that a fact is relevant for an account only when it is selected with reference to a topic or a question posed to the past. The simplest fact a dated event relies on what is for some historians a belief and for others a convenient fiction: the decisive significance of the birth of Christ in establishing a chronology in terms of a "before" and "after."
22. This view motivates my study of Durkheim in Emile Durkheim: Sociologist and Philosopher (Ithaca,N.Y., 1972).
Yet a purely documentary conception may function as an unexamined assumption or it may give rise to a paradoxically self-conscious and sophisticated defense of a "naive" idea of the historian's craft. In any case, insofar as it achieves a position of dominance, a documentary conception is excessively restrictive, especially in the results it yields in the analysis of significant texts. And it obscures the problem of the interaction between description and other uses of language in an account. The idea of a purely descriptive, objective rendering of the past can allow for uses of language that escape it only in terms of the exiguous category of unavoidable bias or particularistic subjectivity. This category may apply to certain aspects of historiography. But the simple opposition between self-effacing objectivity and subjective bias fails to accommodate the range of language uses in any significant history.
The purely documentary view of historiography often coincides with a historicist definition of the historical that identifies the object of study as changing "particulars" in contrast with extratemporal or synchronic types or universals. This venerable view ignores the historical process of repetition with variation or change that functions to mitigate the analytic opposition between the particular or unique and the typical or universal. Yet it is this historical (and linguistic) process that is operative in the past and that raises the problem of the historicity of the historian in his or her attempt to come to terms with it. A documentary historiography that tries to exclude interpretation or to see it only in the guise of bias, subjectivity, or anachronism also has a bizarre consequence. It presents historical truth in an essentially non-historical way. For, by attempting to restrict historiography proper to the description and analysis of verifiable facts (ideally in the form of a definitive and exhaustive account), it strives for an unchanging representation of changing "particulars" that would itself transcend the historical process. As the work of Ranke amply documents, the narrowly historicist and the ahistorical are extremes that meet in the ideal of a purely documentary historiography. And the desire to transcend history reappears in a form that may be invisible precisely because it has become so familiar. Indeed a belief that historiography is a purely documentary or descriptive reconstitution of the past may be prone to blind fictionalizing because it does not explicitly and critically raise the problem of the role of fictions (for example, in the form of models, analytic types, and heuristic fictions) in the attempt to represent reality. The result is often a tacit reliance upon the most conventional narrative structures to combine documented fact, vie romancee, and unsubstantiated judgments about the past or flimsy analogies between past and present.
With specific reference to intellectual history, I would argue for a more "performative" notion of reading and interpretation in which an attempt is made to "take on" the great texts and to attain a level of understanding and perhaps of language use that contends with them. This notion, which valorizes the virtuoso performance in reading, may easily be abused when it becomes a
license for reducing the text to little more than a trampoline for one's own creative leaps or political demands. In this respect, certain recent French thinkers (notably Roland Barthes) are at times dubious guides. Certainly, the act of interpretation has political dimensions. It is not an autonomous hermeneutic undertaking that moves on the level of pure meaning to establish a "fusion of horizons" assuring authoritative continuity with the past. In some relevant sense, interpretation is a form of political intervention that engages the historian in a critical process that relates past, present, and future through complex modes of interaction involving both continuities and discontinuities. But it is equally misleading to pose the problem of understanding in terms of either of two extremes: the purely documentary representation of the past and the "presentist" quest for liberation from the "burden" of history through unrestrained fictionalizing and mythologizing. In relation to both of these extremes (which constitute parts of the same larger complex), it is necessary to emphasize the status of interpretation as an activity that cannot be reduced to mere subjectivity. A significant text involves, among other things, creative art, and its interpretation is, among other things, a performing art. But art is never entirely free, and the art of the historian is limited in specific ways. He must attend to the facts, especially when they test and contest his own convictions and desires (including the desire for a fully unified frame of reference). And even when he attempts to think further what is thought in a text, he cannot reduce the text to a pretext for his own inventions or immediate interests. The belief in pure interpretation is itself a bid for absolute transcendence that denies both the finite nature of understanding and the need to confront critically what Freud discussed in terms of "transference."
The genuine alternative to a purely documentary and contemplative conception of the past "for its own sake" is not its simple opposite: the futile attempt to escape the past or to identify it through projection with the present. Rather texts should be seen to address us in more subtle and challenging ways, and they should be carried into the present with implications for the future in a "dialogical" fashion. Historiography would be an exercise in narcissistic infatuation if it amounted to a willful projection of present concerns upon the past. The notion of "creative misreading" is itself misleading when it legitimates one-sided, subjectivist aggression that ignores the ways in which texts may actually challenge the interpreter and lead him to change his mind. Even if one accepts the metaphor that presents interpretation as the "voice" of the historical reader in the "dialogue" with the past, it must be actively recognized that the past has its own "voices" that must be respected, especially when they resist or qualify the interpretations we would like to place on them. A text is a network of resistances, and a dialogue is a two-way affair; a good reader is also an attentive and patient listener. Questions are necessary to focus interest in an investigation, but a fact may be
pertinent to a frame of reference by contesting or even contradicting it. An interest in what does not fit a model and an openness to what one does not expect to hear from the past may even help to transform the very questions one poses to the past. Both the purely documentary and the "presentist" extremes are "monological" insofar as they deny these possibilities. Indeed the seeming "anomaly" should be seen as having a special value in historiography, for it constrains one to doubt overly reductive interpretations and excessively "economical" shortcuts from understanding to action.
The conception of the field that I have tried to defend complicates the task of the intellectual historian. But it also keeps intellectual history in touch with questions that are raised in "great" texts and that are forever old and new in a manner that cannot be reduced to some philosophia perennis or to a subjectivistic relativism. And it defines intellectual history more in terms of a process of inquiry than in terms of rules of method or of a body of information about the past. This is the most fruitful kind of "definition" possible for an approach that both addresses historical problems and understands itself as historical. The demand for documentation serves to keep responsive interpretations from becoming irresponsible. But to use this demand to attempt to escape our own dialogical relation to the past is to attempt to escape our own historicity. The problem is to understand more clearly what is involved in a relationship that is dialogical and historical without being either "historicist" or "presentist." The historian who reads texts either as mere documents or as formal entities (if not as Rorschach tests) does not read them historically precisely because he or she does not read them as texts. And, whatever else they may be, texts are events in the history of language. To understand these multivalent events as complex uses of language, one must learn to pose anew the question of "what really happens" in them and in the reader who actually reads them. One of the most important contexts for reading texts is clearly our own a context that is misconstrued when it is seen in narrowly "presentist" terms. I have only alluded to ways in which this context involves the reader in an interaction among past, present, and future having a bearing on both understanding and action. But it is precisely here that intellectual history opens out to other modes of interpretation and practice. This "opening" relates to the way in which the power of dialogue and of reflection is itself effective only when it comes with the "working through" of existential problems that are perforce also social and political problems.23
23. I would simply mention one limited way in which I believe intellectual history should address this issue. The intellectual historian should, I think, recognize his or her audience as a tensely divided one made up of both experts and a generally educated public. The intellectual historian is required to come as close as he or she possibly can
to an "expert" knowledge of the problems being investigated But a goal of intellectual history should be the expansion of the "class" of the generally educated and the generation of a better interchange between them and the "experts " This means helping to put the generally educated in a position to raise more informed and critical questions It also means attempting to prevent expertise from becoming self enclosed within its own dialect or jargon In these senses, intellectual history faces complex problems of "translation," and its own concerns bring it into contact with larger social and cultural questions One such question is how to resist the establishment of common culture on a relatively uncritical level and to further the creation of a more demanding common culture that, within limits, is genuinely open to contestation.