6—
Infrastructure and Superstructure:
Revisions in Marxist Sociology of Culture

Robert Wuthnow

Between the general theory of superstructures and their relationship to the base and the concrete study of each specific ideological phenomenon there seems to be a certain gap, a shifting and hazy area that the scholar picks his way through at his own risk, or often simply skips over, shutting his eyes to all difficulties and ambiguities. The result is that either the specificity of the phenomenon suffers . . . or an "immanent" analysis which takes account of specificity but has nothing to do with sociology is artificially fitted to the economic base. And it is precisely a developed sociological doctrine of the distinctive features of the material, forms, and purposes of each area of ideological creation which is lacking.
Mikhail Bakhtin (1928:3)

It is a commonplace that Marx turned Hegel's view of cultural determination on its head. "Life is not determined by consciousness, but consciousness by life," he and Engels asserted in a much-quoted passage (1846:15). Or as one student of Marx explains:

[T]here are only two ways to understand history. Either we start from consciousness; in which case we fail to account for real life. Or we start from real life; then we come up against this ideological consciousness that has no reality, and must

account for it. Historical materialism puts an end to the speculation which starts from consciousness, from representations, and hence from illusions (Lefebvre 1968:65).

In short, the material base of society determines the shape of its culture, not the other way around.

It is equally commonplace, however, to find assertions that hedge Marx's views, rendering them less vulgar, and pointing toward more complex, interactive relations between society's material base and its cultural superstructure. Raymond Williams (1977:79–80), for example, has argued that Marx's own usage of the terms base and superstructure was flexible and relational, and that it is mistaken to think of these concepts as enclosed categories or enclosed areas of activities. He also calls for a more active role for ideology in the shaping of the societal base, citing as evidence a letter to Bloch in September 1890 in which Engels states:

The economic situation is the basis, but the various elements of the superstructure—political forms of the class struggle and its results, to wit: constitutions established by the victorious class after a successful battle, etc., juridical forms, and even the reflexes of all these actual struggles in the brains of the participants, political, juristic, philosophical theories, religious views and their further development into systems of dogma—also exercise their influence upon the course of the historical struggles and in many cases preponderate in determining their form.

Another writer, Martin Seliger (1977:202), concludes from a lengthy survey of the ways in which culture has been discussed in the Marxist tradition that the evidence "calls in question the assumption that the contents and categories of thought can be unequivocally related to a specific class structure and the specific conditions of an epoch." In a similar vein, Göran Therborn (1980:viii) has expressed dissatisfaction with the manner in which Marxists themselves have dealt with the material determination of ideologies, either bypassing it "in embarrassed silence" or repeating it "in an endless series of Marxological exegeses" and has called for a wholly "new formulation of a theory of material determination and of class ideologies." And still another writer, approaching the subject as a Marxist literary critic, has noted that Marxist theory has often been interpreted in ways that have "failed to gain significant purchase on literature as a distinct, semi-autonomous social practice" (Sprinker 1987:238). The relation between culture and material conditions, it appears from these statements, is often more complex than the theoretical literature portrays. But if the complexity of this relation needs to be

recognized, it has been only in relatively recent work that efforts have been made to sort out this complexity.

The problems that appear most in need of further reflection in order to understand cultural criticism within this perspective fall into four broad categories. First, a number of scholars, both those concerned with culture and those oriented toward other aspects of Marxist theory, have suggested that the nature of the material base, and especially the concept of class, should be clarified. Second, recent discussions have challenged subjectivist concepts of culture and ideology and have suggested the value of respecifying these concepts as elements of practice, as varieties of social production, or as social formations. Third, greater attention has been called for in specifying analytic and empirical variations within culture itself. And fourth, a stronger emphasis on the dynamic or interactive relations between infrastructure and superstructure has been suggested.

The first of these issues—the question of class—can largely be dispensed with here because it has received ample treatment in the writings of Seliger, Therborn, and others. The value of moving from subjectivist to more objective concepts of ideology will be accepted as a given: rather than defend it here, we will examine writers who have emphasized an objective concept of ideology in general and of cultural criticism in particular. It is for this reason that we concentrate on writers whose contributions to the general discussion have come from literary criticism: in focusing on literature, they have inevitably stressed the concrete and explicitly produced character of cultural criticism. The main focus of the present discussion is on the remaining two issues: the problems of specifying the internal structure of cultural criticism more carefully and of relating these patterns to social structure more interactively.

These considerations provide a valuable perspective for better understanding cultural criticism. The contributions of Terry Eagleton, Fredric Jameson, and Mikhail Bakhtin provide useful starting points for considering these issues. After reviewing the main lines of development evident in these contributions, some concepts of a more synthetic nature will be presented.

Terry Eagleton

The most relevant aspects of Terry Eagleton's work for rethinking the Marxist sociology of culture are contained in his essays

"Categories for a Materialist Criticism" and "Towards a Science of the Text" (in Eagleton 1976). Eagleton's starting point is the work of Raymond Williams, which he argues is flawed by an overly abstract, subjectivist conception of culture that conflates it with its material base. As a remedy, Eagleton advances a more concrete conception of culture, one that relies on the more categorical way of thinking that Williams criticizes, but seeks to avoid oversimplifying the relation between base and superstructure by specifying several subdivisions of each. Culture—that which is to be explained—is conceived of not as some ephemeral orientation in society but as a specific material practice or object, namely, a text.

At the same time, Eagleton denies the possibility of relating a specific text directly to the social context in which it is produced. Intervening factors that must be considered include some of the more implicit beliefs and assumptions on which subjectivist approaches to culture have focused. Eagleton identifies six "major constituents" that must be related to one another: the general mode of production (GMP), the literary mode of production (LMP), general ideology (GI), authorial ideology (AuI), aesthetic ideology (AI), and text.

The first five of the constituents are the structures, Eagleton says, that produce the text. Their relations with one another and with the text do not, however, form a causal system (a recursive model of the kind sometimes imagined by statistical sociologists). Eagleton (1976:45) suggests instead that the task of criticism is to analyze "the complex historical articulations of these structures." The term articulation is, of course, reminiscent of Louis Althusser and Etienne Balibar (1970), although Eagleton makes no specific mention of them at this point.

For Althusser, the notion of articulation arises specifically in the context of discussing the relations among modes of production that are not entirely integrated with one another but that can be understood by the theorist as part of a single system. Their relations emerge not only in the historical evolution of capitalism itself but also in the theorist's efforts to reconstruct the internal mechanisms of capitalism. Rather than specific binary relations being most at issue, it is the larger system of interrelationships that Althusser claims determines the course of history. It is, of course, this notion of a broader system that distinguishes Althusser's approach from the more limited materialist interpretations of Marxism, and Eagleton follows Althusser in adopting this general perspective in dealing with culture.

More generally, the idea of articulation as a way of thinking about

relations among concepts also appears to underlie Eagleton's discussion. In this sense, articulation connotes dual, opposing tendencies, a fitting together of elements to form an interrelated system and yet a retention of distinct identities that prevent a system from becoming entirely integrated. As Maurice Bloch (1983:153) states, articulation refers to "a type of connection where what is joined does not consequently form a whole. The articulated elements remain fundamentally unchanged as if ready to detach themselves. The notion of articulation therefore stresses the idea of several elements whose different natures will lead to contradiction and therefore revolutionary change." Eagleton (1976:45) echoes this idea, although he places less emphasis on revolutionary change in suggesting that the primary relations of articulation that may occur among literary modes of production are "homology, conflict and contradiction." The idea of elements that remain unchanged, detachable, and contradictory is, as we shall see, evident in Bakhtin's discussion of the features of texts themselves, a theme that will prove helpful for understanding the specific nature of cultural criticism.

Eagleton passes quickly over the general mode of production, pausing only to assert that he considers it dominant, historically specific, and general only in the sense of referring to economic production broadly as opposed to literary production more narrowly. He thus assumes the Marxist position—that mode of production is the most salient feature of the social environment—as a starting point, rather than leaving this assumption open to historical investigation. In doing so, he of course leaves his discussion open to criticism on precisely these grounds. His analysis of the social environment in which cultural production takes place consists of asserting little more than that this environment is capitalist or bourgeois. Therefore, the value of his approach lies more in specifying the relations between cultural production and cultural products than in suggesting innovative conceptions of the relevant features of bourgeois society.

In his discussion of the concept of literary modes of production, Eagleton becomes more specific and of course more relevant to the immediate issue of cultural criticism. It emphasizes the fact, perhaps obvious yet often neglected in sociological theories of culture, that culture does not simply bear a kind of general affinity with its societal environment but is produced in specific contexts. It is to these contexts that closest attention must be paid to understand how social factors shape the form and content of culture. A literary mode of production, moreover, is not, in Eagleton's view, reducible to some mechanical or

technological determinant, such as the printing press and movable type. It is rather constituted by a set of social interactions among the occupants of concrete social roles: the relations among cultural critics, patrons, suppliers of materials, publishers, disseminators, opponents, and consumers. This is one of Eagleton's most useful observations, although it is by no means original to him. Whether one takes a Marxist approach to culture or some other approach, the point is the same: culture originates in specific social settings and is likely to bear the imprint of these settings.

In addition, Eagleton cites and distinguishes among some historical examples of the ways in which literary modes of production may differ.

The tribal bard professionally authorised to produce for his king or chieftain; the "amateur" medieval poet presenting to his patron a personally requested product for private remuneration; the peripatetic minstrel housed and fed by his peasant audience; the ecclesiastically or royally patronised producer, or the author who sells his product to an aristocratic patron for a dedication fee; the "independent" author who sells his commodity to a bookseller-publisher or to a capitalist publishing firm; the state-patronised producer (1976:47).

Clearly these correspond generally to the level of economic development (general mode of production), and they in turn limit the kinds of texts that are likely to be produced. Texts produced under ecclesiastical patronage are likely to be devoutly didactic, to take the most obvious cases, while novels produced for sale to a commercial audience of bourgeois consumers are more likely to cater to the private tastes of these consumers. Authors' conceptions of their audiences and the means of reaching these audiences will, Eagleton suggests, shape both the choice of genre and the actual content of the works produced.

Beyond the effects of the experienced social relations in which culture producers are engaged, texts are also shaped by the three varieties of ideology, which Eagleton identifies as general, authorial, and aesthetic. All three pertain to the specific social relations in which culture producers are embedded and differ primarily in terms of which social relations are at issue. General ideology is defined as "a relatively coherent set of 'discourses' of values, representations and beliefs which, realised in certain material apparatuses and related to the structures of material production, so reflect the experiential relations of individual subjects to their social conditions as to guarantee those misperceptions of the 'real' which contribute to the reproduction of the dominant social relations" (1976:54). Authorial ideology, Eagleton states, is "the effect of the author's specific

mode of biographical insertion into GI, a mode of insertion overdetermined by a series of distinct factors: social class, sex, nationality, religion, geographical region and so on." Aesthetic ideology, by comparison, is not defined formally but is described by typical examples of its content: "theories of literature, critical practices, literary traditions, genres , conventions, devices and discourses" (1976:60).

Each of these definitions, of course, raises a host of conceptual and theoretical questions that could become the focus of extensive discussion. Eagleton himself appears more interested in subsuming virtually everything that might shape the production of cultural texts under these rubrics than in delimiting the range of relevant considerations. Nevertheless, several important points can be extracted that may be useful in redirecting sociological thinking about cultural criticism and its relations to the social environment in which it is produced.

To begin with, Eagleton's conception of ideology differs radically from what might be considered in more standard sociological treatments of literary production to be the worldview, or even the internalized norms and values, of the literary producer. This point is worth underscoring because Eagleton does employ terms, such as beliefs, values, and misperceptions, and emphasizes what might be understood as the author's self-perception. All of this may seem reminiscent of a subjectivist conception of ideology. Nevertheless, Eagleton's emphasis is on a more objectivated form of ideology. In particular, he uses the phrase "discourses about values" rather than simply saying "values"; he asserts that these discourses are "realized in certain material apparatuses"; he discusses authorial ideology not as self-image but as literary expression, that is, as the way in which the author is represented within the text itself; and he again refers to concrete elements of discourse—conventions, genres, practices, and so on—in discussing aesthetic ideology.

Although ideology has an objective existence in the social world, it is analytically distinguishable from the social world, in Eagleton's view. Ideology is a discursive structure, a form of practice characterized by speech and the material artifacts of speech. It is discourse about the experienced world but not that world itself. The experienced world, perhaps ironically, is less observable for Eagleton than is the realm of discourse. Experience has a reality of its own, but it is observable only through the coordinates of the mode of production, on the one hand, and discourse, on the other hand. The mode of production, moreover, is a theoretical inference, an abstraction from the material conditions of the observed world. Thus the relation between base and superstructure

becomes for Eagleton, in effect, a relation between a theoretical construct concerning modes of production and the actual discursive structures observed in the texts produced within that context.

To state it this way, however, creates too sharp a distinction between infrastructure and superstructure. The social world is real, in Eagleton's outline, apart from discourse. But it is also incorporated into the realm of discourse. Or put differently, ideology does not simply reflect the social world; ideology draws the world of experience into itself, turning these experiences into categories that shape the production of specific texts. There is, in short, a social horizon built into the discursive structure of ideology. For Eagleton (1976:54), ideology is "strewn with the relics of imperialist, nationalist, regionalist and class combat." That is, political categories from the experienced world become discursive categories in ideology, both directly in the words that are used and indirectly in language and colloquial expressions. There are also "cultural" elements or, more precisely, antecedents drawn from the social environment about the ways in which knowledge is to be packaged and communicated, styles of discourse, genres, and literary standards. These are, to a certain degree, evident in the concrete texts of culture producers themselves.

The relation between ideology and infrastructure is, most importantly, mediated by the immediate social relations that Eagleton calls the literary mode of production . This is an apparatus that includes, he says,

the specific institutions of literary production and distribution (publishing houses, bookshops, libraries and salons), but it also encompasses a range of "secondary," supportive institutions whose function is more directly ideological, concerned with the definition and dissemination of literary "standards" and assumptions. Among these are literary academies, societies and book-clubs, associations of literary producers, distributors and consumers, censoring bodies, and literary journals and reviews (1976:56).

He also argues that the communications industry and educational apparatus play an important role.

Eagleton's point, however, is not to propose studies, such as those that have become common among both sociologists of culture and cultural historians, that focus primarily on the institutional aspects of printing, publishing, reading clubs, book fairs, literary academies, and the like. The point is rather to emphasize the form and content of the texts produced, to discover the categories that are emphasized and deemphasized, and to determine the ways in which these categories illuminate or obscure features of the world of social experience. For the last of

these especially, comparisons must be made between base and superstructure. Ideology, Eagleton assumes, will partially incorporate the experienced world into itself and will partially transform that world into something else. Ideology is not, therefore, simply a matter of false consciousness. As Eagleton (1976:69) explains, "Ideology is not just the bad dream of the infrastructure: in deformatively 'producing' the real, it nevertheless carries elements of reality within itself."

The elements of the real that are brought into the ideological realm are selectively processed through the institutional relations that comprise the literary mode of production. But again Eagleton stresses the interactive character of this process. He rejects a straightforward positivist view in which the real is seen as something external to ideology, a characteristic of the infrastructure that is merely reflected in ideology. The real is instead constituted by the discursive structure of ideology itself.

The real is by necessity empirically imperceptible, concealing itself in the phenomenal categories (commodity, wage-relation, exchange-value and so on) it offers spontaneously for inspection. Ideology, rather, so produces and constructs the real as to cast the shadow of its absence over the perception of its presence. It is not merely that certain aspects of the real are illuminated and others obscured; it is rather that the presence of the real is a presence constituted by its absences, and vice versa (1976:69).

Ultimately, then, the focus of analysis must be on the interplay of categories within the text itself. For it is here that the social world is reconstructed; here that reality is created—a reality that in some ways may correspond directly with the theorized character of the materialist infrastructure of history, but that also, by virtue of its textuality, escapes this direct form of determination, providing instead a defamiliarization of experience, "a sportive flight from history, a reversal and resistence of history, a momentarily liberated zone in which the exigencies of the real seem to evaporate, an enclave of freedom enclosed within the realm of necessity" (1976:72).

Fredric Jameson

Like Terry Eagleton, Fredric Jameson has emphasized the double determination of ideological texts—their determination through

the internal logic of the text's content and their determination through the broader contexts in which they are produced. For Jameson, the general mode of economic production and the class relations arising within this mode constitute the theoretical environment in which to situate the analysis of ideology. But he too is especially concerned with the ways in which the social environment becomes articulated within the structure of ideology itself.

Because Jameson's oeuvre is broad and has itself become the subject of a number of secondary works, only a sketch of those aspects that bear most directly on the sociology of culture need be considered here. Jameson's work, like Eagleton's, falls squarely within the Marxist tradition of criticism, and he is especially interested in the complex relations between infrastructure and ideology. As a literary critic, his work is perhaps limited to those features of ideology that gain expression in formalized texts. He is, however, sensitive to the importance of treating ideology as a kind of practice—"symbolic art" is a phrase he uses repeatedly—rather than as a vague set of beliefs or a taken for granted worldview. As symbolic acts, texts constitute a special kind of practice. They have an intentional or instrumental aspect; that is, they do things to the world, manipulate it, change it. At the same time, they leave the social world untouched. Their action remains within the realm of discourse. They change the world not directly but by drawing the world into its own texture. To an even greater extent than Eagleton, Jameson is thus interested in how the social comes to be incorporated into ideological practice (an interest he shares, as we shall see, with Bakhtin). Indeed, Jameson takes the work of Northrop Frye as a starting point because he recognizes the value of Frye's emphasis on community. Rather than work within Frye's framework, however, he attempts to subsume Frye into a broader Marxist perspective (Jameson 1981; Dowling 1984).

Jameson's view of ideology resembles Marx's earlier treatment of ideology as a response to the alienated life elements of capitalism more closely than it resembles Marxist theories, such as those of Althusser and Jacques Lacan, that have borrowed insights from Freud to examine the manner in which ideology functions as a form of false consciousness. For Jameson, ideology is not so much a distortion of a perceivable reality as it is an attempt to come to terms with and to transcend the unbearable relationships of social life. In short, ideology is vision, not simply a reflection of reality or a description of it, and as such it acts symbolically upon the world. Ideology is a form of storytelling (Jame-

son 1981:105). It is at least potentially liberating and redemptive. The manner in which ideology articulates with history and the experience of social life is therefore crucial: ideology identifies certain features of experience and yet extricates itself from experience in such a way that experience can be transformed.

Though opting for a Marxist perspective, Jameson immediately declares his dissatisfaction with standard interpretations of the relations between infrastructure and superstructure. The latter does not merely reflect social structures that are somehow more basic, Jameson argues. The two are not levels of empirical reality than can be related causally or historically. Jameson, like Eagleton, is an Althusserian in that he takes Marxism as a theoretical framework that creates a certain view of reality rather than as a scientific description of something external to itself. He is also an Althusserian in stressing that the categories of Marxist thought should be related, not as elements of a causal system, but as components of a totality. Base does not, as it were, determine superstructure. The two, rather, articulate with one another, and the manner of their articulation is consequential for understanding the totality of society and for predicting the direction of its development. Most generally, Marxism is a source of "insight." It provides a conceptual scaffold for reconstructing relevant relationships. The concept of mode of production, for example, does not refer to actual historical development but serves as a model for understanding historical development. And Jameson adds to it a conceptual framework of his own that is aimed at revealing more of the internal structure of texts themselves.

A text can be rendered intelligible—or, we might say, meaningful—by apprehending it within three concentric frameworks, or "horizons." These consist of (1) political history, (2) the relevant social context, and (3) "history now conceived in its vastest sense of the sequence of modes of production and the succession and destiny of the various social formations, from prehistoric life to whatever far future history has in store for us" (1981:75). Each horizon may be understood in two ways: as a framework that the analyst imposes on the text and as a framework built into the text itself or definable as a set of relations among the text's elements. It is the second of these that most concerns us here. Jameson provides not only a general description of each of the three horizons but also a discussion of their relations with one another and with the external social environment.

The first horizon may be likened to the historical setting of the text that would be of interest in ordinary literary criticism. It is constituted

by actors and events with which the text is concerned. These of course may be the same as those composing the author's contemporary experience, or they may be quite different, as in narratives about the past or about some imagined future. In either case, some parallels are likely to exist between the social environment of the author and the social horizon depicted in the narrative. Neither will exactly reflect the other, but certain resemblances, analogies, and dual meanings are likely to be present. Jameson argues especially for careful examination of narrative content as a starting point, but his discussion implicitly acknowledges the importance of knowledge about the context of the narrative's production as clues to this analysis. The conflicts and contradictions that can be observed within the narrative are particularly important according to Jameson.

At the second level, Jameson turns at once to a more general notion of social order within the text and a more specific notion of the salient features of this social order. Social order is essentially a matter of struggle between classes and among various class fractions. At a crude level, the conflicts evident in the text are read through the lens of class antagonism—an exercise that clearly runs into dangers of misinterpretation in dealing with many kinds of texts. The more useful point Jameson makes is that the relations among actors in the text, whether conflictual or harmonious, can be understood as a kind of dialogue. That is, social order at this level can be conceived of as discourse. And this discourse is likely to reveal both the substantive antagonisms or polarities around which discussion takes place and the shared codes that make discussion possible in the first place. The emphasis here on discourse is clearly the key. We can look at the institutional settings in which narrative action takes place and even compare these narrational settings with their historic counterparts. But over and above that, it is necessary to focus on the categories of speech that actors in these settings use and the ways in which these discourses are presented in the text. In addition, discourse always implies counterdiscourse. What is said must be read against what is not said, especially if we are interested in comprehending the hegemonic character of a dominant culture. Just as in Eagleton's discussion, then, Jameson emphasizes the negations and polarities that construct discourse.

Jameson's third framework or horizon is least clearly formulated and perhaps of less immediate programmatic relevance to the sociology of culture. It invokes the broad evolution of history conceived of in Marxist theory itself and asks the interpreter to seek traces of historical evolu-

tion within the various systems of discourse evident in the text. In one sense the specification of this horizon does nothing more than challenge the analyst to see the class antagonisms that have already been posited at the second level as part of a larger evolutionary process. In a different sense the analyst may be encouraged to recognize that historical evolution is no orderly or linear process but is a movement of fits and spurts, of continuous transition involving negotiation among class fractions and competing modes of discourse. From this perspective it appears that Jameson's third horizon, while partially evident in textual content, is primarily a horizon that the analyst must develop for himself or herself from an independent study of history.

Jameson's second horizon, then, offers the greatest degree of innovativeness to the study of ideology. Working within the larger historical-evolutionary framework suggested by the third horizon, and taking into account the standard social elements incorporated into the text itself (the first horizon), the analyst then engages in an act of discourse analysis that identifies basic categories of discourse, conflicts and polarities, and an internal dialogue among these polarities. This task is, of course, not merely an inductive operation, but an interpretation, a search for "an abstraction nowhere completely present in any body of texts or utterances and something that must always be reconstructed from partial evidence" (Dowling 1984:132). It is, nevertheless, an abstraction that organizes—and is organized by—the actions, events, and utterances that comprise the text itself.

The purpose of this reconstructive task is partially, for Jameson, to illuminate the ways in which basic class conflicts gain expression as intratextual dialogue. There is, however, a deeper purpose as well: to understand how it is possible for certain things to be said and others not to be said. In other words, the utterances of which ideology is comprised are shaped, not by the social infrastructure, but by the possibilities and limitations built into the framework of discourse itself. Characters in the text, therefore, do not emerge simply as actors appropriated directly from the experienced world; they are transformed into characters by being placed within a particular framework of discourse, and their speech and actions are oriented primarily to this framework.

This is the point, incidentally, at which Jameson's discussion intersects most directly with Eagleton's. The possibilities for intratextual dialogue between different discourses are, to a degree, contingent on the genre of the text, and the genre employed will invariably reflect the relations between culture producers and consumers that constitute the

prevailing mode of literary production. Jameson points out, for example, that the modern novel is particularly suitable for mixing generic conventions and setting up dialogue that represents alternative points of view. It incorporates older genres as raw materials, combines them with realistic accounts of everyday life, and yet imposes its own ideological form as an organizing device. The general point is that genre becomes relevant to the analysis of ideology, both as a substructure on which surface conflicts are overlaid and as a mode of articulating the relations between texts and historical events (see also Eagleton 1981a).

Mikhail Bakhtin

In Mikhail Bakhtin we confront a theorist not only from a different era but also from a school of criticism to which some Marxists themselves have taken exception (see Carroll 1983; De Man 1983; and Young 1985, for overviews). The difference between these approaches is, however, considerably less severe in the case of Althusserian Marxists such as Eagleton and Jameson than it has been in the past. Indeed, it is Bakhtin to whom Jameson (1981:84) refers in suggesting the importance of examining the internal dialogue between opposing discourses in texts, and Eagleton (1981b, 1982) has also favorably appraised Bakhtin's work. For our present purposes, Bakhtin helps us understand the structure of cultural criticism by providing additional insights about the ways in which social horizons become incorporated into literary texts.

Although his early work stands half a century closer to Marx's than to Eagleton's or Jameson's, Bakhtin stresses the objective, material aspects of ideology and draws on Marx's scattered remarks about the production of culture in much the same way as Eagleton and Jameson, rather than subjectivizing ideology in relation to its infrastructural roots. In fact, Bakhtin argues explicitly for a perspective that emphasizes the materiality of cultural objects. He states, for instance, that "all the products of ideological creation—works of art, scientific works, religious symbols and rites, etc.—are material things, part of the practical reality that surrounds man" (Bakhtin and Medvedev 1928:7). They may have, as he says, "significance, meaning, inner value." But these meanings are "embodied in material things and actions." He also indicates specifically

that his remarks should not be construed to pertain only to the more explicitly produced, tangible artifacts of culture:

Nor do philosophical views, beliefs, or even shifting ideological moods exist within man, in his head or in his "soul." They become ideological reality only by being realized in words, actions, clothing, manners, and organizations of people and things—in a word: in some definite semiotic material. Through this material they become a practical part of the reality surrounding man (7).

It is fallacious, he asserts, to accept the prevalent social psychological approaches that regard ideology as the creation and comprehension of individuals. Terms such as meaning and consciousness should be rejected as elements of bourgeois ideology itself and replaced with a sociological approach that emphasizes the relations between social interaction and ideology as practice. Ideology, he concludes, "is not within us, but between us" (8).

If superstructure cannot be distinguished from infrastructure by its subjectivity, it is nevertheless, for Bakhtin, distinguishable. Ideology should not be approached like any other variety of social production. The danger is to treat it so much as a material object that it stands entirely apart from social interaction. This is the tendency, Bakhtin claims, in utilitarian positivist theories—even in utilitarian positivist interpretations of Marx. Cultural products should not be regarded as objects that have only uses—"external purposes"—that can be understood entirely from the outside and primarily in terms of technical organization. Cultural products are distinguished by their expressive dimension, by the fact that they become important in the relations among people. It is for this reason that terms such as ideological intercourse, dialogue , and especially discourse dominate Bakhtin's work.

What Bakhtin seeks is a richer, more variegated understanding of the "ideological environment," as he calls it. The ideological environment is a layer of reality that should be understood as the critical mediating context between economic modes of production on the one hand and human thought and action on the other.

Social man is surrounded by ideological phenomena, by objects-signs of various types and categories: by words in the multifarious forms of their realization (sounds, writing, and the others), by scientific statements, religious symbols and beliefs, works of art, and so on. All of these things in their totality comprise the ideological environment, which forms a solid ring around man (14).

Our experience, in his view, is shaped, not directly by our existence, but through the medium of this ideological environment.

Bakhtin's perspective is thus similar in one sense to theories of reality construction and symbolic interactionism that emphasize the symbolically mediated realities in which we live. The ideological environment can be understood as the same world of symbols that has been stressed in these approaches. In another sense, however, Bakhtin's emphasis is quite different. His concern is not principally with explaining individual consciousness or even with showing how individual worldviews are constructed from symbolic material. It is with the contours of the ideological environment itself. This environment has a unity of its own, a structure, and it is in constant flux, constant interaction, always influenced by and influencing the economic and political relations with which it is associated.

Bakhtin is particularly helpful in demonstrating how the characters and events that compose literary texts are related to the broader ideological environment and to the economic base. He points out that these characters and events cannot be understood by projecting them directly into life, that is, by imagining they reflect actual characters and events in the real world. Nor can they be understood strictly in terms of some broad philosophical outlook of a certain period. A literary figure is instead an "ideological refraction of a given social type" (Bakhtin and Medvedev 1928:21). We must know something of the actual historical circumstances in which a literary figure is situated, and Marxist theory provides a starting point for ascertaining which of these circumstances to emphasize (for example, feudal relations or bourgeois class fractions). There is also a kind of semantic horizon, identifiable at several levels (the levels Jameson discusses), that can be seen in literary texts associated with these historical circumstances. This ideologeme , as Bakhtin calls it, sets the ideological context in which a literary figure can be expressed. It consists of genre, of certain conceptions of action and agency, of artistic standards, and more substantively, of dominant problems, constraints, and resources that shape the literary figure's actions. Character becomes characteristic within this framework and action becomes representative. Bakhtin insists that characters and events be understood as structural elements of the artistic work itself. The analyst's task, therefore, must always be twofold: to examine the relations between the broad economic modes of production that make up the societal infrastructure and the structure of the ideological system that Bakhtin variously refers to as the ideological horizon, ideologeme, semantic

horizon , or artistic totality (all of which have been specified and drawn systematically into the models proposed by Eagleton and Jameson) and to examine the relations between these broader structures and the specific characters and actions that become defined within these structures. Both are appropriate foci of sociological investigation.

How then does Bakhtin suggest going about these dual analytic tasks? Only the more general guidelines can be mentioned here, but his own substantive discussions are replete with specific insights. As a starting point, we must recognize that the social situations in which ideological products are produced do not remain entirely external to these products themselves. The situation, Bakhtin asserts, "enters into the utterance as a necessary constitutive element of its semantic structure" (in Todorov 1984:41). This does not mean, of course, that the social situation will be reflected accurately in textual utterances; indeed, the text inevitably transforms the materials it draws from the external social world. Nevertheless, the sociologist must be attuned to the existence of a social world within the text as well as to the social world outside the text. Bakhtin suggests several important features of this intratextual world: the spatial and temporal horizons that constitute the situation in which action and dialogue occur, the social relations signaled by words such as I and we , the objects that are known and appropriate for discussion, and the relations between actors and what is happening, especially those that suggest an evaluative stance.

Further, the social relations incorporated into ideological products can be understood as a complex, multilayered series of discourses. At a minimum, primary discourse in the text may take the form of verbal dialogue, written correspondence, or a series of gestures exchanged between two or more of its characters. There may also be secondary levels of discourse: discourse about discourse, such as the direct quotation of other actors' (external or literary) discourse, commentaries on previous discourses, or even discussions about styles and modes of discourse. Then a number of implied discourses may have to be inferred from parallel forms of construction and other literary devices: a discourse between the author and an intended reader, discourse between the author and his or her own representation or voice within the text, discourse between either of these and other textual voices, and discourse between textual voices and the reader. All are forms of social interaction to which sociologists should be especially attentive. They also constitute much of the structure of texts, that is, the essential categories, frames, voices, contrasts, and parallels that impose order on the text (Bakhtin 1981:284). Further-

more, Bakhtin stresses the importance of considering the degree to which these structures envelop themselves to form a tight, monolithic sense of form and content or the degree to which heterology is evident in the variety of social forms, voices, and opinions expressed.

The analysis of discursive patterns is for Bakhtin, therefore, the most fruitful arena in which to explore how social conditions are incorporated into literary products and how these products define representative actions and agents. The dialogue contained in texts is likely to be framed concretely within various authoritative languages: "the language of the lawyer, the doctor, the businessman, the politician, the public education teacher and so forth" (1981:289). These may be elements of the class relations of the broader society that are expressed textually. They may also set up certain oppositions within the text that are artistically resolved, heightened, or objectified in the actions of representative characters. Again, it is essential to recognize that these intratextual relations are objects worthy of investigation in themselves. They do not precede the text but are created by it. Markers must distinguish alternative voices or positions, symbols of authority must be attached concretely to particular characters or opinions, and even categories of thought must be objectively signaled to the reader by parallel constructions, binary oppositions, polemical negations and counternegations, and so on.

Bakhtin, then, does not go so far as to champion a kind of superstructuralism that neglects the importance of infrastructural relations. He remains broadly within the Marxist framework. But he stresses the existence of a social infrastructure within the superstructure itself in addition to the external one. And he points to the importance of a closer reading of texts themselves. Ideology cannot be understood in terms of either its general thematic content or the patterns of beliefs and values that form an individual's consciousness. Ideology has its own internal structure, a structure that is at least as complex as the structures composing the economic base of society and that is as fully relevant and amenable to sociological investigation.

Causal Determination and the Problem of Meaning

Before turning to the task of extracting some common substantive themes from Eagleton, Jameson, and Bakhtin, some note

must be made of the essential theoretical logic that emerges in their work. In Eagleton and Jameson particularly, standard assumptions about causality and the social determination of culture come explicitly into question. Rather than being superstructural elements shaped in some temporal or mechanistically causal way by elements of the infrastructure, the elements at both levels stand in an articulated relation to one another as parts of a larger totality. This manner of speaking may appear to undermine traditional Marxist theory, converting it to a kind of bland Hegelianism, at the same time that it tries to preserve the language of conventional Marxism. The basic challenge, however, should be a critique of empirical positivism rather than an attack on Marxist theory. Even Jameson, who has been accused most strongly of turning Marx upside down, holds firmly to the assumption that the historical movement among capitalist modes of production, as posited in Marxist theory, constitutes the most fundamental starting point for any analysis of cultural production. The materialist infrastructure, however, is no longer conceived of as something merely evident in the empirical world and discoverable through the positive application of scientific methods. It is rather an analytic feature of reality that becomes a reality only in interaction with the application of Marxist theory itself. Only with the benefit of this theoretical framework does the totality of the base and superstructure components become evident. This perspective more nearly represents a post-positivist or hermeneutic understanding of the relations between evidence and interpretation. It in no way rules out the study of empirical relations to determine how they relate to the posited totality and, within this totality, to one another.

The point of examining relations between base and superstructure, therefore, ceases to be one of explanation, in the sense of finding the true causes of ideological expressions, or of reduction, in the sense of showing that these expressions are mere epiphenomena in the larger scheme of objective social structural relations. It is essentially a task of interpretation, a task accomplished by determining whether the various elements of base and superstructure relate to one another to form a single, dominant totality or whether there are internal contradictions and conflicts, or even openings, that provide room for creative, redemptive—dare we say, revolutionary—alternatives. For this purpose, the notion of articulation appears to provide a convenient summary term. It connotes mutual accommodation or adaptation, a fitting together, as it were, of the various puzzle pieces to form a coherent whole. Base and superstructure interact dynamically in this process on many levels. Ideology may not ultimately

transform the prevailing mode of production, but its internal structure shapes the degree of articulation that can develop with elements of the infrastructure, and it acts to control the resources and actions that its adherents can take. At the same time, pressures against articulation must be recognized. Base and superstructure are prevented from forming a wholly integrated system by the various contradictions, the alternative logics, that impinge upon them at any moment in historical development. Indeed, the very notion of totality must, since Althusser, be taken less as an integrated whole than as a conception that emerges from the complex, sedimentary or layered, interactions of what is present and what is absent.

If the fundamental point of studying ideology ceases to be one of causal explanation, then it can no longer be taken simply as a matter of interpreting the meaning of cultural symbols. Jameson, for instance, near the beginning of his discussion of literary texts, argues that a whole new approach to ideology must be found that goes beyond the quest for meaning. He cites a passage from Gilles Deleuze and Félix Guattari (1977), who assert a "general collapse of the question 'What does it mean?'" and argue that studies of linguistic structure have advanced only "to the extent that linguists and logicians have first eliminated meaning." Only in the broadest sense, that in which the interpreter provides an interpretation of base and superstructure by relating them to the theoretical framework provided by a Marxist (or some other) conception of history, can it be said that meaning has been ascribed to the cultural elements under scrutiny. Eagleton, Jameson, and Bakhtin emphasize the internal relations among the elements of texts and broader ideologies and the relations between these internal ones and those that constitute the social contexts in which the cultural products arise. The quest is for patterns of intelligibility rather than meaning.

These theorists do not go so far as to become what E. D. Hirsch (1976) has termed, referring to the deconstructionists, cognitive atheists , that is, interpreters who deny entirely the possibility of knowing what texts mean. Their view of meaning, however, focuses on relations rather than on some substratum of substantive truth. Jameson (1972:215), for example, specifically denies the simple semiotic method of discovering meaning by looking for the deeper "signified" that underlies the "signifier." Instead, he posits an "infinite regress from signifier to signified, from linguistic object to metalanguage." The shift is from meanings to sentences, from the content of substance to the content of relations. Rather than attempt to identify the meaning of a text, he writes about

"meaning-effect" and "meaning-process." He rejects the notion of prediscursive meaning and focuses instead on meaning as the possibility of transcoding from one level of structure to another (Mohanty 1982:35). Meaning from this perspective is thus deprived of any ontological status; it is approached through "a basic reformulation of substance into process and form, into structured movement and production" (Mohanty 1982:36).

Common Directions and Problematic Foci

For a sociology of culture, particularly one claiming ties to the Marxist tradition, the relation between infrastructure and superstructure must remain central. There are, however, several emphases in the approaches, of which Eagleton, Jameson, and Bakhtin have been taken as representatives, in recent Marxist literary analysis that merit greater consideration from a sociological perspective. A useful starting point is to conceive of culture as practice, process, or product rather than as something more general, vague, or even subjective that bears only the broadest analytic correspondence to the social environment. This orientation is evident in Eagleton, Jameson, and Bakhtin and has been prominent in recent thinking about culture among sociologists as well. It is particularly appropriate to conceive of culture in this manner when the focus of inquiry is a particular ideology or ideological movement, the advent of a new idea or mode of discourse, or the social role of genre, discourse, or communities of artistic, literary, or religious expression. Less clear, of course, is whether this perspective can be applied successfully to studies of broad cultural orientations, such as individualism and rationality, or to patterns of generalized value orientations, such as universalism and particularism. At the very least, two considerations may be important when these broader conceptions of culture are at issue. First, the most generalized cultural patterns still become most real when they are manifested in concrete social situations. They are produced and disseminated, gain expression in concrete social processes, and influence behavior insofar as they govern discourse and other social resources in institutional settings. Second, traces of more generalized cultural patterns are likely to be observed both in the intellectual antecedents that shape specific cultural processes and in the theoretical perspectives that shape the interpretation of these processes.

An emphasis on specific cultural practices merely directs attention to the ways in which these patterns actually are manifested as traces.

An explicit feature of this approach is the importance it attaches to specifying the institutional contexts in which culture is produced and has its effects. Eagleton's distinction between general modes of production and literary modes of production is most helpful in this regard. However much the process of cultural production may be shaped by its general location within a social environment that may be termed bourgeois, modern, industrial , or whatever, its character will also be affected by the social relations that arise among producers, patrons, audiences, and wielders of power. These relations connect the social infrastructure, as it were, with the production of superstructural elements. The two are distinguishable, not in the sense of one being somehow real and the other being somehow distorted or epiphenomenal, but as manifestations of the Marxist framework itself. Infrastructure consists of historic behavioral patterns that the observer reconstructs from evidence at least partly, if not primarily, separate from the cultural products that arise from these patterns. At the same time, superstructure consists of the behavioral patterns in discourse, texts, and other symbolic-expressive media, which are also reconstructed by the observer according to certain preconceived theoretical categories of analysis. The analyst, then, also reconstructs a certain kind of articulation between infrastructure and superstructure by identifying homologies and noting patterns of covariation. This reconstruction, however, constitutes only the crudest level of analysis in determining the modes of articulation between base and superstructure. Neither category is hermetically sealed from the other. Superstructure becomes a feature of infrastructural relations, as it is said, when life imitates art, or when genres of discourse become, as it were, genres of social interaction. Similarly, and of greater interest in the present literature, features of the infrastructure become incorporated into the superstructure, comprising elements of its content and form and gaining textual transformation in the process.

As a guide to sociological investigations of these dynamic relations between base and superstructure, we might simplify the more nuanced discussions of Eagleton, Jameson, and Bakhtin by suggesting the term social horizon to encompass those features of the social environment that become incorporated into the cultural products under scrutiny. The social horizon of a text, therefore, encompasses the textual representation, in Eagleton's terms, of the mode of production, class relations, the author, and the author's relations to significant actors and events consti-

tuting the literary mode of production. In Jameson's terms, it encompasses the specific elements of political history and the more general symbolizations of class relations and discourse about class relations. In Bakhtin, the social horizon is expressed mainly in reflective levels of intratextual discourse and discourse about discourse. Different theoretical interpretations clearly will shape the particular ways in which the social horizon is reconstructed. At a minimal level of interpretation, primary emphasis should be given to the ways in which social events, actors, and relations are modeled in the text as settings, plots, characters, poignant examples, and so on.

In contrast with social horizons, the term discursive fields might be used to designate an underlying polarity, framework, or set of categories around which discourse is organized. Discursive fields are objectified in the internal dialogue that Bakhtin identifies and in the opposition of voices to which Jameson refers. If a Lévi-Straussian argument is followed, it may be argued that discursive fields are evident in all texts insofar as thought and speech are inherently dependent on binary oppositions. Discursive fields, however, constitute a more elaborate feature of texts than mere binary oppositions. They cannot be reduced to mutually exclusive conceptual categories or to dialectical moments that are ultimately synthesized and resolved. Discursive fields are instead defined by alterity, by poles that remain apart, and, indeed, by an opposition of words, genres, contexts, and voices that are, as Bakhtin suggests, set against each other dialogically. Especially in ideological discourse of the kind that Marxist theory has generally found interesting, discursive fields appear to represent one of the more fundamental structures that articulates (and disarticulates) the relation between base and superstructure.

A final conceptual component that is worth special attention in this context consists of what might be referred to as figural action . Within the symbolic polarity defined by a discursive field, ideological discourse typically identifies representative, prescriptive behaviors that reflect and resolve the dilemmas inherent in the symbolism of the discursive field itself. These behaviors, however, typically receive objectification over and above mere abstract idealization through an association with characters and characteristic actions that serve as exemplars or models. These figurative constructions, it appears, play an important dual role: on the one hand, they draw ideology into a specific relation of relevance with the social setting in which it is produced; on the other hand, they elevate ideology—emancipate it—by suggesting more complex, textually referential modes of action.

In conclusion, the perspectives on ideology evident in the writings of Eagleton, Jameson, Bakhtin, and other Marxist literary critics differ significantly from the dominant sociological conceptions of the relations between infrastructure and superstructure. At its worst, sociological theory has posited a broad homology or isomorphism between the stages of capitalism and the varieties of ideology alleged to be associated with these stages. At its best, sociology has gone beyond such general assertions mainly by specifying more detailed conceptions of social structure and suggesting relations between these contextual factors and various attitudinal themes or ideological dispositions. The sociological poetics of Bakhtin and the post-Althusserian formulations of Eagleton and Jameson allow for the development of more nuanced theories of the relations between infrastructure and superstructure. The general epistemological perspective of this literature focuses on theoretically constructed categories and the reconstruction of relations—relations of articulation and disarticulation—among these categories, rather than on raw empiricism, induction, or positivist methods. Ideology is conceived of as a form of social practice rather than as a subjective feature of consciousness, and questions of meaning are referred to the reconstructive or interpretive process itself. A more nuanced view of social structure is encouraged, particularly one that privileges the institutional contexts in which tangible ideological practices and objects are produced. The relation between base and ideology becomes more interactive as well, but not simply in terms of mechanical or historical causal influences between the two. Closer attention is paid to the fact that social contexts become incorporated into ideological texts rather than remaining external to these texts. The internal structure of texts, therefore, imposes its own mode of organization on this material and is in turn organized by it. A sociological theory of culture must, for this reason, demonstrate greater awareness of textual construction itself: of genre, methods of objectification, voice, dialogue, interpolation and interpellation, textual authority, redundancy, embedding, parallels, and contrasts. Discursive fields, it has been suggested, may be a particularly important textual pattern to observe. It has also been suggested that many of the specific behavioral prescriptions—models of ethical behavior, representative social actors, characters to emulate—emerge in ideological constructions, less as the direct prototypes of the social world and more as the products of the particular discursive fields that frame their options and provide the problems with which they struggle. In this sense, the focus of ideological analysis shifts from a metaphor of reflection—

superstructure as a reflection of infrastructure—to a metaphor of space. The thematic orientations that constitute the moral discourse of ideology, this perspective would suggest, bear an indirect or mediated relation to their social environment. The critical mediating link is a discursive field that creates the space, as it were, in which figural action can be identified.

References

Althusser, L., and E. Balibar. 1970. Reading Capital. London: Verso.

Bakhtin, M. M. 1981. The Dialogic Imagination: Four Essays. Austin: University of Texas Press.

Bakhtin, M. M., and P. M. Medvedev. 1928 [1985]. The Formal Method in Literary Scholarship: A Critical Introduction to Sociological Poetics. Cambridge: Harvard University Press.

Bloch, M. 1983. Marxism and Anthropology. Oxford: Oxford University Press.

Carroll, D. 1983. "The Alterity of Discourse: Form, History, and the Question of the Political in M. M. Bakhtin." Diacritics 13: 65-83.

Deleuze, G., and F. Guattari. 1977. The Anti-Oedipus. New York: Viking.

De Man, P. 1983. "Dialogue and Dialogism." Poetics Today 4: 99-107.

Dowling, W. C. 1984. Jameson, Althusser, Marx: An Introduction to the Political Unconscious. Ithaca: Cornell University Press.

Eagleton, T. 1976. Criticism and Ideology. London: Verso.

———. 1981a. "The Idealism of American Criticism." New Left Review 127: 60-65.

———. 1981b. Walter Benjamin, or Towards a Revolutionary Criticism. London: New Left Books.

———. 1982. "Wittgenstein's Friends." New Left Review 135: 64-90.

Hirsch, E. D. 1976. The Aims of Interpretation. Chicago: University of Chicago Press.

Jameson, F. 1972. The Prison-House of Language. Princeton: Princeton University Press.

———. 1981. The Political Unconscious: Narrative as a Socially Symbolic Act. Ithaca: Cornell University Press.

Lefebvre, H. 1968. The Sociology of Marx. New York: Vintage.

Marx, K., and F. Engels. 1846 [1947]. The German Ideology. New York: International Publishers.

Mohanty, S. P. 1982. "History at the Edge of Discourse: Marxism, Culture, Interpretation." Diacritics 12: 35-44.

Seliger, M. 1977. The Marxist Conception of Ideology: A Critical Essay. Cambridge: Cambridge University Press.

Sprinker, M. 1987. Imaginary Relations: Aesthetics and Ideology in the Theory of Historical Materialism. London: Verso.

Therborn, G. 1980. The Ideology of Power and the Power of Ideology. London: Verso.

Todorov, T. 1984. Mikbail Bakhtin: The Dialogic Principle. Minneapolis: University of Minnesota Press.

Williams, R. 1977. Marxism and Literature. Oxford: Oxford University Press.

Young, R. 1985. "Back to Bakhtin." Cultural Critique 2: 71-92.