Back to the Table of Contents

Ideology, Discourse, and Cultural Studies: The Contribution of Michel Pêcheux¹

Martin Montgomery
Stuart Allan
University of Strathclyde

Abstract: This article offers an evaluative assessment of the potential contribution of Michel Pêcheux's research to a current movement within cultural studies to secure a conceptual framework for the critical discourse analysis of the linguistic mechanisms of ideology (examples of which are drawn from news accounts).

Résumé: Cet article propose une appréciation et une évaluation de la contribution potentielle qu'apportent les travaux de recherche de Michel Pêcheux à un courant actuel des études sur la culture qui vise à appuyer sur un cadre conceptuel toute analyse critique du discours et des mécanismes linguistiques d'une idéologie (des exemples sont tirés des compte rendus de nouvelles).

Introduction

The semiotic project offered the promise of breaking with approaches which reduced language to a ``neutral'' instrument through which ``reality'' is expressed. By foregrounding the arbitrary nature of the signifying fields in everyday life, it suggested fascinating new ways to think through the Williamsian maxim that what was at stake for cultural theory was the investigation of the complex of patterned relationships between elements of culture as a whole way of life (Williams, 1961). Moreover, semiotics allowed for the opening up of what had become a rather empty postulate, namely, that culture is inherently meaningful, so as to unpack the transparency or naturalness of real meanings rooted in practical social experience.

As has been well documented, however, ensuing attempts over the past two decades to re-centre the domain of cultural studies so as to make better use of this reflexive set of methodologies have culminated today in an acute crisis: how to define the limits of language? Is there a world outside of the text? Where does non-discursive reality end and language begin? (see, for example, Allor, 1988; Chevalier, 1990; Charland, 1990; Franklin, Lury, & Stacey, 1991; Grossberg, 1983; Hall, 1985; Johnson, 1987; Laclau & Mouffe, 1985; MacCabe, 1985; Morris, 1990; Nelson & Grossberg, 1988; Robinson & Straw, 1984; Spivak, 1987). Clearly, in some hands, the tools of semiotics have been wielded in such a way as to reify a view of language process antithetical to a critical understanding of the linguistic object. This is particularly so where language has ostensibly become divorced from its materiality, thereby effacing from the purview of study the very historical conditions of its existence.

In some cases, this impasse has stemmed from a reluctance to rework certain idealist precepts underpinning Saussure's (1966) own formulations, the ensuing theoretical commitments usually engendering a reductive division between ``form'' and ``content.'' That said, even those theorists who have endeavored to secure a symptomatic reading of Saussure frequently fall victim to the economistic trap of asserting that the declared limits of language are strictly determined vis-à-vis the prefigured needs of a society organized to reproduce a specific arrangement of power relations. Semiotics is then held responsible for allowing a rigidly functionalist view of the power/resistance dynamic to creep into the resultant conceptual frames.

Hence one possible explanation for the increasing attention being paid to linguistics and discourse analysis in cultural studies, at the expense of semiotics, is directly tied to this problem of accounting for contradiction, and thus struggle, in the play of ideology and power. Other possible explanations for the growth of interest in the concepts and categories of discourse analysis range from the difficulty of rendering Althusserian notions of ideology into operative concepts (Althusser, 1971, 1969), to the loss of linguistic specificity in the alternative categories borrowed from Foucault's work on the enunciative function of statements (Foucault, 1972, 1971), and, finally, to the growing prevalence of postmodernist arguments that the term ``discourse'' be purposefully stretched so as to encompass processes formerly identified as falling under the regions of representation, ideology and signification respectively (see Bauman, 1988; Murphy, 1988; Nicholson, 1990; Ross, 1988).

For those practitioners of cultural studies unwilling to allow the sharp edge of the term ``discourse'' to be blunted, yet forced to concede that no suitable alternative avenue is immediately available, a provisional step is often advocated: namely, that the social factors that condition or shape the meaning of any given utterance simply be recognized as dialectical. This maneuver is, of course, far from satisfactory; but will a turn to discourse analysis eventually reward cultural studies with the requisite means to explicate the materiality of discursive processes? The answer remains open to question: a number of serious difficulties associated with the ``borrowing'' of concepts will have to be surmounted first, yet in our view discourse analysis certainly holds promise as a means to reconstruct the problem of the consensual nature of meaning anew and in a much more elaborate fashion.

It is with this aim in mind that we begin the task of re-assessing the critical discourse analytic framework advanced by Michel Pêcheux (1988, 1983a, 1983b, 1982), for we believe that his approach is remarkably resonant with potential strategies for just such a cultural studies project. His work provides an incisive challenge to what has become, after certain postmodernist formulations, a politics of discursive surfaces while introducing, in turn, an elegant refutation of a traditional structuralist linguistics where the abstract system (the code) is accounted for at the expense of the actual implementation of the code in situations of use. This article will therefore proceed to examine the analytical agenda on offer to cultural studies by this theorist precisely as he speaks to the profound silences of Saussurean and Marxist linguistic orthodoxy alike. In addition, it is hoped that through this appraisal the general contours of a trajectory out of the conceptual impasse marked off by the language/reality paradox will also be highlighted.

Ideology in the Past

Today, Pêcheux's commitment to a research program derived from Louis Althusser's reconstruction of historical materialism has been thrown into sharp relief, especially when many former advocates will acknowledge only a vestigial loyalty to it, insisting as they do that the resultant political framework has proven intractable. Still, the influence of Althusserian Marxism on cultural studies has yet to be eclipsed, and to dismiss the import of this presence is to impoverish the discussion. A defining concern for cultural studies research continues to be its commitment to theorizing the ways in which social divisions and hierarchies are naturalized or ex-nominated (placed ``beyond discourse''), a project which has its roots in the Althusserian moment. Indeed, the struggle Althusser initiated over two decades ago to displace an essentialist logic mobilized to anchor a conception of ideology as a series of ideas separated from material practice (along with the parallel view to a rational, self-constituted unitary human subject) has, for the most part, been fought and won.

Much of the terminology has been recast, of course, in the light of insights drawn from the work of the neo-Gramscians, Foucaultians, deconstructionists, and most importantly in our view, feminist researchers. Significantly, many of the post-Althusserians and their critics still share Althusser's interest in elucidating the very obviousness that language ``transparently'' makes a word ``name a thing'' or ``have a meaning.'' In general, they are in agreement that this ``ideological effect'' is crucial to an understanding of how the parameters of ``legitimate,'' ``appropriate'' or ``authoritative'' discourse are policed. Pêcheux is thus one of several researchers who, following Althusser's lead, avoids positing the meaning of a word as existing ``in itself'' for a human subject to then ``decode.'' Equally inappropriate, it follows, is the presupposition that the word can somehow be analyzed irrespective of how discourses constitutive of this subjectivity (such as those of class, gender or ``race'') are themselves the terrain of ideological contestation. Nevertheless, it is precisely at this point that several of the most serious difficulties with this conceptual approach have arisen. For while Althusser appeared to be posing several of the most vital questions, the provisional nature of his answers was such that they have eluded rigorous application.

As many commentators have pointed out, one of the most perplexing aspects of Althusser's formulations on ideology is the absence of an elaborated stance regarding the possibilities for the realization of a resistant or counter-hegemonic politic at the level of the text-subject encounter. How to succeed in making ``common sense'' uncommon? More often than not, we would argue, Althusserian categories are being deployed in such a way as to render indistinct the imbrication of language with the conditions (institutional sites) of its use, thereby making it much more difficult to theorize the production of a particular range of (tendential) subject positions in relation to a socially contingent discursive formation. Alarmingly, the resultant claims are frequently made to rest on vague references to the ``manipulative practices'' of Ideological State Apparatuses (ISAs), of which the mass media are considered to be of key significance. For others, moreover, ``misrecognition'' comes to stand in for the terrain formerly marked off as ``false consciousness.'' This when Althusser was at considerable pains to show how it is that this ideological misrecognition is unavoidable: there simply is no ``true'' consciousness floating about in limbo.

Althusser is also to be credited, in our view, with pinpointing the need to problematize the procedures through which the ``dominant'' or ``preferred'' terms and definitions of ``ruling'' ideologies are ultimately translated as natural or inevitable. Still, this formulation of the ISA-interpellation linkage does not devote adequate attention to a strategic understanding of the linguistic mechanisms working to produce terms of address for a human subject as he or she negotiates possible identifications with a reality ``out there.'' This is a serious dilemma which other theorists, arguably the most conspicuous of which are the Foucaultians, have chosen to sidestep. Hence the import of Michel Pêcheux's theoretical intervention, as it constitutes a rare attempt to address the very site where the play of power and ideology in discourse, or--in the Lacanian idiom which he adopts, ``the politics of the signifier''--configure: the site of interpellation itself.

Here we will undertake to examine the implications of this radical move to locate the determining effects of heterogeneous socio-historical conditions in and through linguistic mechanisms, as opposed to an exclusive imposition at the level of the human subject proper. As will be shown, Pêcheux's integration of an Althusserian position on ideology with a fresh approach to discourse analysis creates the conceptual space necessary to begin the work of explicating the constitution and potentialities of the forces governing subjectivity, and furthermore, the conditions for an oppositional politics of meaning production. Considered in this light, his work promises to spark a decisive reappraisal of the basic assumptions underpinning recent efforts to secure new forms of critical discourse analysis for cultural studies.

Langue and Parole Revisited

To support this assertion Pêcheux must first clear the way to challenge certain Saussurean precepts with an eye to emphasizing the role of discursive process itself. Like other critical linguists before him (the contributions of the Bakhtin Circle also figure prominently), he strongly objects to the manner in which discourse has been reduced to the concrete acts of language use. In most instances, the situation of the utterance is dismissed as ``subjective'' or circumstantial, thereby ensuring that questions concerning the social determination of meaning are left largely unattended. Similarly, the langue/parole opposition has routinely been invoked so as to constitute the object of linguistic study in an unnecessarily constrained way, such that language as system becomes the centre of concern to the exclusion of issues of rhetoric, poetics, politics, and ideology. Linguistics in the Saussurean tradition, according to Pêcheux, ``is in the end condemned to regress behind the break which inaugurated it, by a kind of `return of the repressed' whose central element (forming its weakest link) is located in the region of semantics and articulated around the langue/parole couple'' (1982, p. 174).² He thus rejects Saussure's way of posing the opposition and reintroduces, in turn, the discarded areas of parole by way of a different pair of categories, namely, linguistic basis and discursive process. Accordingly, Pêcheux seeks to relate the operation of definable linguistic units outwards to larger formations, for as we shall see, the operation of particular compositional tendencies, such as relative clauses, are related to discursive and ideological formations and thus to the struggle for political hegemony.

To some extent, even when recasting the langue/parole distinction, Pêcheux seems to accept much that is implied in the traditional definition of langue. Every linguistic system, he stresses, ``as a set of phonological, morphological and syntactic structures, is endowed with a relative autonomy that makes it subject to internal laws which constitute, precisely, the object of linguistics'' (1982, p. 58). It then follows that every linguistic system is to be endowed with a relative autonomy from relations of class (narrowly defined) which, in turn, makes it subject to internal laws: the system of langue, after all, ``is indeed the same for the materialist and the idealist, for the revolutionary and the reactionary'' (1982, p. 58). Furthermore, it is on the basis of these internal laws that discursive processes develop (once again processes of substitution, synonymy and paraphrase) that, having been established through the unity of langue, are always inscribed (overdetermined) in relations of power and resistance. Perhaps not surprisingly, Pêcheux also looks to radically re-conceptualize the domain of parole by framing it in terms of this discursivity or discursive process. Discursivity is not to be conflated with parole: that is, ``it is not a `concrete,' individual way of inhabiting the `abstraction' of the langue'' (1982, p. 58). Still, discursivity may, admittedly, be predicated upon langue ``which is the indispensable prerequisite of any discursive process'' (1982, p. 58), since ``it is on the basis of these internal laws (of langue) that discursive processes develop'' (1982, p. 58).

At the same time, however, although language--the relatively autonomous linguistic basis (langue)--may be indifferent to political struggle, discursivity is decidedly not, because ``every discursive process is inscribed into an ideological class relationship'' (1982, p. 59). This is the crux of the matter for Pêcheux; and his claim here matches closely Bakhtin & Volosinov's (1973) well-known assertion that ``every sign is subject to the criteria of ideological evaluation,'' so that as a result ``sign becomes an arena of the class struggle'' (1973, pp. 10, 23). Indeed, Pêcheux's notion of ``relatively autonomous internal laws constituting the linguistic basis'' versus ``ideologically informed and differentiated discursive processes,'' which are predicated upon that basis but distinct from it, seems close in spirit to Bakhtin & Volosinov's distinction between meaning and theme, whereby meaning as the lower limit of linguistic significance--an abstract self-identical element--``is subsumed under theme and torn apart by theme's living contradictions so as to return in the shape of a new fixity and self-identity only for a while, just as it had before'' (1973, p. 106).

The Materiality of Ideological Struggle

If it is accepted that the relations of contradiction-unevenness-subordination between different ISAs are the stake in the ideological struggle (given the struggle to impose inside the complex set new relationships of unevenness-subordination), for Pêcheux the effectivity of hegemonic ideology is to be characterized as the ``winning out'' of the reproduction of social divisions over their transformation (1982, p. 100). Ideological struggle between two antagonistic classes or groups is, once again, not symmetrical: this process of ``winning out'' is achieved largely through obstruction or suppression. Neither seeks ``the same thing'' as the other, thus the reproduction/transformation relation is characterized by movement and not an objective form of inertia (1982, p. 101). To begin the task of resolving the principal difficulty at issue here: namely, the question of how to secure for analysis the ideological conditions of the reproduction/transformation of social hierarchies in ideas, Pêcheux moves to reconfigure Althusser's (1971) controversial notion that ideology interpellates individuals as subjects.

First then, the very ``self-evidentness'' of the human subject as he or she is ``hailed'' into hierarchical relations of subjugation needs to be treated with much greater precision. Crediting Althusser's notion of how the human subject ``works by itself'' as the principal point of departure, Pêcheux suggests that it is through the examination of the means by which unconscious repression and ideological subjection are materially linked that critical efforts may best transcend the limitations of earlier formulations which held that ideologies are ideas (not material forces) that have their sources in human subjects. It is therefore his intention to problematize

the processes of ``imposition-concealment'' that constitute that subject by ``putting him [or her] into place'' (by signifying to him [or her] what he [or she] is) and by concealing from him [or her] at the same time that ``putting into place'' (that subjection) thanks to the illusion of autonomy constitutive of the subject, such that the subject ``works by himself'' [or ``by herself'']. (1982, p. 91)

Consequently, having affirmed that the articulation of ideology with the unconscious is concealed within a ``web of subjective evident truths,'' Pêcheux locates the linkage between the constitution of meaning and that of the constitution of the human subject within the figure of interpellation itself. It is there that the discrepancy in the formulation individual/subject (the ``paradox'' by which the human subject is called into existence) may best be designated. Pêcheux notes that Althusser's formulation carefully avoids presupposing the existence of the human subject on whom the operation of interpellation is performed. As opposed to stating that the ``subject is interpellated by ideology,'' Althusser suggests that the non-subject is interpellated/constituted as a subject by ideology. Herein lies the aforementioned paradox, however, as Pêcheux argues that interpellation so conceived possesses a ``Munchausen effect'': that is, given that every individual hailed by ideological discourse is ``always-already a subject,'' a circular or ``retroactive'' effect results. To illustrate this point, Pêcheux calls forth the image of the immortal Baron Munchausen who, according to legend, lifted himself out of a bog and into the air by pulling ``with all the strength of one arm on a lock of his own hair.''

How to escape this difficulty, then? As will be shown, for Pêcheux the initial step is to establish that ``the effect of the preconstructed as the discursive modality of the discrepancy by which the individual is interpellated as subject [...]while still being `always-already a subject' '' nevertheless operates by contradiction (1982, p. 107). This manoeuvre then allows the ``transparency of language'' to be grappled with concretely: that is, this apparent contradiction between the formulation, after Lacan (1977), of human subjects as processes of representation (``caught'' in a network of signifiers) and the recruitment of subjects (via interpellation) from individuals who accept as evident the meaning of what they hear, say, read and write as ``speaking subjects,'' will be opened up for investigation (1982, pp. 108- 109).

The Transparency of Language

What Pêcheux is offering here is a formal account of discursive processes both within discourses and between one discourse and another, rather than a substantive account of particular ideologies and discursive formations in a concrete, situated fashion. According to this definition, a discursive formation seems best understood as a set of regulative principles that underlie actual discourses but remain separate from them. This formulation suggests then that words, expressions, and propositions obtain their meaning from the determinate discursive formation in which they are produced (the linguistic elements selected, how they are combined), thus meaning becomes an effect upon an active human subject, and not a stable property. Once again, an expression or proposition does not have a meaning ``of its own'' perpetually attached to it. Pêcheux stresses the attendant point that given the emergence of this ``matrix of meaning,'' individuals are then interpellated ``as speaking subjects (as subjects of their discourse) by the discursive formations which represent `in language' the ideological formations that correspond to them'' (1982, pp. 111-112). Such a conception leads Pêcheux to declare that the human subject is ``forgetful'': he or she misrecognizes or occludes the ``cause'' or determination of his or her discourse, thinking instead the he or she is its author ``in reality.'' By highlighting how the naturalness or obviousness of words or expressions will, in turn, ``change their meaning'' as they ``slide'' or ``slip'' from one determinate discursive formation to another, Pêcheux substantively refines the notion of intertextuality (the passage from one sign system to another) posited by Bakhtin. Here, the interweaving of elements between contending discursive formations is to be specified as the outcome of the dictates of hegemonic struggles traversing the social field.

The second thesis Pêcheux advances dictates that ``every discursive formation, by the transparency of the meaning constituted in it, conceals its dependence on the `complex whole in dominance' of discursive formations, itself imbricated with the complex of ideological formations'' (1982, p. 113). At first sight, articulating discourse to ideology in this way seems to have the merit of opening up the study of ideology to a more concrete, socio-historical analysis. If discourse instantiates or enunciates ideology then the latter is made tangible and present to study in particularly practicable ways. In the elaboration which follows, however, Pêcheux refuses to define the precise nature of this correspondence beyond the notion of an imbrication. He is unable to be specific about the nature of the relationship between ideological formations and discursive formations. At moments he seems to suggest that ideological formations provide principles of coherence that underpin the intelligibility of their corresponding discursive formations; but the relationship of one to another remains persistently vague. Similarly, his conception of discursive formations remains ill-defined. It is not clear whether they are best understood in terms of institutional provenance (e.g., ``the discourse of the defence establishment'') or topical scope (e.g., ``the discourse of nuclear deterrence''). Nor, crucially, is it clear at what level of abstraction from actual utterances they should be recognized as operating on. And since no hint is given as to how the boundaries of any discursive formation may be identified, it is difficult to see how the substantive constituents of a given discursive formation can be specified in practice. Thus, a major putative gain in the delineation of ideologies in concrete situations is thrown away.⁴

Notwithstanding this difficulty, however, Pêcheux demonstrates in a convincing manner the advantages of theorizing meaning as a function, not of particular words or wordings, but rather of the discursive formation in which such expressions occur. This is not to suggest that meaning is purely accidental and contingent. Although variable, certain generalized and stable mechanisms or processes may be seen as underlying this productivity, all of which fall under the domain of the notion of discursive process outlined above (briefly, a system of relationships of substitution, paraphrases, synonymies, and so forth, which operate between linguistic elements or signifiers in a given discursive formation). These processes work to secure the play of meaning within the ``complex whole of the discursive formations'' into a universal (or naturalized) hierarchy. More precisely, any instance of enunciated discourse has its intelligibility ensured, at least in part, by the operation of rules of inclusion and exclusion or interdiscourse. That is, the means by which a specific arrangement of discursive formations, itself the localized expression or product of ideological hegemony, is mediated with a text is via these rules. The declared ``meaning'' of a given text is in this way removed from the author as speaking subject exclusively (usually aligned with a view to his or her intentionality) and is instead confirmed as the effect of relationships of language use within and between this ``complex whole of discursive formations.''

At issue now is the need to push forward from this contention that the human subject's misrecognition of autonomy and centredness (the illusion of a unified self-identity) is anchored within the play of his or her ``own'' discourse, and thus those of the dominating discursive formations dispersed unevenly across the social field. Pêcheux proceeds by looking to further unravel the ``thread of the discourse of the subject.'' Here he discerns two principal elements of this ``thread'' or interdiscourse, namely, the ``preconstructed'' and ``articulation,'' both of which are embedded in the materiality of the human subject and his or her Other. Briefly, the preconstructed in this formulation corresponds to the ``always-already there'' of the ideological interpellation as it supplies/imposes already available positions: that is, the ``raw material'' of ``reality'' and its ``meaning'' in the form of universality (the ``world of things,'' the ``evident facts'' or ``what everyone knows to be real'') or obviousness (1982, pp. 115, 121). Thus the preconstructed stands in contrast with that which is said to be ``constructed'' by the utterance. Alternatively, the element of articulation (support effect) sustains the human subject in a relation to that meaning (``as I have said before'' or ``as I shall say afterwards'') as if it were something exclusively internal to him or her. In this sense, their domain of operation is the syntagmatic chain of discourse, a plane of relationships that Pêcheux may now refer to as intradiscourse or the operation of discourse with respect to itself.

Intradiscourse, then, is always the ``discourse of a subject,'' while interdiscourse is embedded in previous statements: the word, symbol or concept, to be recognizable as such, must be linked with a pre-existent ``given'' before being absorbed into a particular text. However, with respect to the meaning effects produced along the horizontal plane of discourse (intradiscourse), Pêcheux notes how these intradiscursive relationships may be impinged upon, and affected, by discourses from elsewhere. That is, from somewhere within the complex whole in dominance of discursive formations; in other words, from interdiscourse. He is thereby signalling that it is interdiscourse which determines the domination of the subject form by making it take up positions as a speaking subject, thereby ensuring the human subject an experience of unity that would otherwise be denied (1982, pp. 115, 121). At particular crucial points in the plane of intradiscourse, elements from interdiscourse may erupt as ``already there'': these preconstructed elements are lateral reminders of material established in another discursive formation. In this way, one ``line'' or ``plane'' of discourse may intersect with another, providing tacit support from elsewhere in interdiscourse to an intradiscursive enunciation. When the intelligibility of an intradiscourse leans for support on its intersection with pre-established discursive material from interdiscourse, Pêcheux chooses to call this phenomenon transverse discourse.

Two mechanisms stand out as being heavily implicated in the operation of transverse discourse, viz. determinative and explicative relative clauses. As grammatical constructions (corresponding to the distinction in grammatical description between defining and non- defining relative clauses) these could be seen as part of what Pêcheux has described as the linguistic basis. They provide him, however, with instructive examples of the way in which, on the basis of these internal laws (of langue), discursive processes develop. For both kinds of construction provide points where intradiscourse is susceptible to the workings of transverse discourse: in other words, each construction in its own way allows material to infiltrate by means of transverse discourse into the enunciated intradiscourse. This can be seen more clearly by concrete illustration.

Explicative and Determinative Relative Clauses as Discursive Process

A purely grammatical approach to the sentence would claim that the relative clause (in italics) here merely explicates or adds information in a contingent fashion about some element of the main clause (see, for example, Sinclair, 1972). Thus, the relative clause here adds information about the referent ``Napoleon'' of the main clause. Pêcheux, however, argues that the information (or the proposition) of the subordinate clause is not in this case of a purely accessory or contingent nature. On the contrary, articulating the two propositions together through the use of an explicative has, in this case, the effect of implying a causal relationship between the two, such that: Napoleon led his guards against the enemy position because he recognized the danger to his right flank. Indeed, the subordinate clause (who recognized the danger to his right flank) does, in this reading, express more through its connection with the main clause than it would in isolation. For the causal relationship to be activated, however, requires the recognition of some general background assumption, such as: If (being a general, or being Napoleon) one recognizes a danger threatening, one must oneself lead the attack to ward it off.

Explicative clauses, therefore, act according to Pêcheux as ``lateral reminders,'' prompting a kind of ``return of the known in thought.'' In the illustration above it might be argued that a discourse of motives and intentions (from somewhere in interdiscourse) intersects with a discourse of strict historical narration.

Attention now turns to consider the workings of a determinative (defining) relative clause, an illustration of which is provided with the following sentence: He who discovered the elliptical orbit of the planets died in misery.

The relative clause in this example (Pêcheux borrows it from Frege's article ``On Sense and Reference'') does not so much add information about the referent of a constituent in the main clause but actually determines, restricts or defines who that referent is. Determinative, or defining, relative clauses, in fact, typically constitute part of the constituent itself, rather than being in a weaker appositional relationship to it. Pêcheux claims that this form of embedding allows for the insertion into intradiscourse of elements preconstructed elsewhere (precisely where ``elsewhere'' is, Pêcheux never says). More particularly, he claims that in the example given above the discourse of scientific history erupts in a preconstructed fashion into the discourse of personal biography. In this case, regarding the German astronomer Johannes Kepler, Pêcheux points to the material conditions in which he died in 1630; ``a reality which has little to do with the discovery of the laws of planetary motion, except of course in a religious or moral perspective for which misery is the counterpart of genius, and a punishment for knowledge seen as transgression'' (1982, p. 63). Certainly, it is reasonable to claim that constructions of this type present a logically necessary entailment (``someone discovered the elliptical orbit of the planets'') as part of a syntactic nominalization rather than as an independently asserted, and therefore more easily contested, proposition. Accordingly, the preconstructed surfaces as ``transparent,'' an ``always-already-there,'' in which some segment of reality is invoked as if already given in a preconstituted ``world-of-things-as-they-are.''

These linguistic constructions are from one perspective, by virtue of the grammatical constraints which govern their operation, located in the linguistic basis. At the same time, however, when implemented in any intradiscursive enunciation they open up spaces for the operation of transverse discourse, so that the latter sustains the former, supplying supportive connections by a process of implication between the propositions of intradiscourse. Inasmuch as explicatives and determinatives are susceptible to the lines of force running along transverse discourse from interdiscourse, they are from this altered perspective never less than discursive processes.

It would seem that, for Pêcheux, the relative clause is important because it can be used to illustrate how a detailed description of the grammar of such constructions cannot in itself account for how discursive relations of implication can be set up between main and subordinate clauses. Interpretation of these relations involves a discursive process that leads beyond coded elements of the sentence into surrounding discursive formations and hence into the sphere of ideology. For Pêcheux, the discursive relations that an interpreter comes to recognize as connecting a relative clause with its main clause draw upon what we know already from elsewhere in a taken-for-granted fashion: so that for an interpreter to recognize an uncoded relationship amounts to a confirmation or ratification of the already-known. However, some important lacunae and uncertainties remain in Pêcheux's account. It is not clear, for instance, if he is claiming that every instance of a relative clause will carry an ideological implication; or, indeed, whether all ideological implications will necessarily be carried only by relative clauses.

Crucially at stake here is the degree to which a discursive effect stands in a one-to-one relationship with a grammatical process. A fundamental premise of discourse studies as they have developed over the last 15 years is that no simple correlation is to be expected between grammatical (sub-sentential) processes and discursive processes: it is not possible, in other words, to read off discursive effects from sentential structures in any direct, one-to-one fashion. Thus, the same discursive effect may be achieved by a variety of grammatical realizations. Although Pêcheux rightly argues for a separation of the two domains of grammar and discourse (in his terms ``linguistic basis'' and ``discursive process'') he fails to follow through the full logic of his argument, which is inhibited by his tendency to work from grammar to discourse, in line with his principle that discursive processes develop ``on the basis of these internal laws'' (1982, p. 58). Pêcheux does, admittedly, raise the important issue of how ideological considerations may underpin discursive processes. But a more radical step, reflecting more fully the consequences of his position, would entail working from discursive processes as such, conceived in a more autonomous fashion, and noting the range of grammatical processes that potentially may be associated with them.

In any case, relative clauses of the type identified by Pêcheux are relatively rare in everyday discourse. Even a quick glance at the fairly formal prose of a ``quality'' newspaper finds that they may occur with a frequency of less than one in 10 sentences. The focus on relative clauses seems, in consequence, to be extremely limiting. In order to develop Pêcheux's substantive insights about ideology in language, then, it is important to widen the range of constructions considered susceptible to what might be called ``ideological effectivity.'' In this respect, a key starting point is provided by the range of constructions that may stand in the place of full relative clauses and serve a similar function to them.

Relative Clause Type Constructions in a Sample News Text

A further illustration of a reduced explicative (``egalitarianism'') may be found in the following example, where again particularly contentious material may be identified both in this explicative and the full relative clause which succeeds it. ``In all these institutions a single damaging philosophy has prevailed since the 1960s: egalitarianism, which holds that all children are equal and so none should be allowed to `fall behind.' '' Some assertions of a disputatious nature which are implicated in this statement can be identified as follows: #(1) A single damaging philosophy has prevailed since the 1960s in all these institutions (i.e., colleges, education departments, the schools inspectorate, educational journals, and the DES itself ). #(2) This philosophy is egalitarianism. #(3) Egalitarianism holds that all children are equal and that none should be allowed to fall behind. Whilst the first assertion is potentially questionable, it becomes so only in light of the second and third assertions and these significantly are presented in a non-prominent position and in reduced form. Although Pêcheux generally describes the effect of explicatives as effecting ``the return of the known in thought,'' in many cases they might more aptly be understood as transforming the questionable and the contentious into the obvious and the self-evident.

Full determinative relative clauses, like the foregoing explicatives, may be similarly replaced by a range of alternative constructions. Probably the most frequent is a clause introduced by a non-finite verb instead of a relative pronoun, as italicized in the following example:

The class, festooned with posters and pictures, is divided into two chattering groups playing with paper shapes and plastic triangles. [...] But they are not youngsters. They are 21 and in the fourth year of a teacher training course specializing in the progressive ``child-centred'' approach to learning.

Thus the range of constructions which can act like full determinative relative clauses is equally as wide as those that can act like explicatives. To some extent their respective roles might be seen as overlapping. Consider, for instance, the following sentence: ``Those loudest in defending the system have been those with the responsibility for improving it--the `educationalists,' whose empire includes colleges, and education departments, the school inspectorate, educational journals, and the DES itself.'' The phrase, ``those loudest in defending the system,'' may be seen as derived from ``those who are the loudest in defending the system.'' The role of such reduced determinatives bears close comparison with Pêcheux's general description of determinatives as giving existential solidity to some segment of reality which is invoked as if already given in a preconstituted world of ``things-as-they-are.'' In this case, if we include the succeeding explicatives as well as the opening determinatives of the sentence, we can trace a complicated series of propositions, such that: #(1) Educationalists exist (as a definable group). #(2) They have responsibility. #(3) Their responsibility is to improve the system (of education). #(4) They defend this system. #(5) They are its loudest defenders. #(6) They have an empire. #(7) The empire includes colleges, and education departments, the school inspectorate, educational journals, and the DES itself.

In the context of this particular opinion column these prove to be quite crucial assertions, all the more so for being proposed in non-salient positions. How otherwise can the columnist, an avowed ``Right-wing Professor of Aesthetics,'' writing in the Sunday Express, a decidedly pro-Conservative Party newspaper, attack the existing educational system without tarnishing the reputation of the Conservative Government which has had effective control of it for well over a decade? The answer lies in attributing responsibility to a completely different group, prominent members of which are identified subsequently by name as a ``Socialist dogmatist,'' a ``Communist Professor,'' a ``Professor at London University's pernicious Institute of Education,'' and so forth.

One final instance of the preconstructed may be identified as nominalization, a phenomenon where processes capable of being encoded by a verb are encoded by a noun group.⁵ For instance, to assert that ``standards in state schools have declined'' is a questionable truth claim. However, to write: ``The decline of standards in state schools has been apparent to university teachers for years, but our warnings have been dismissed as the rantings of an `elitist' minority'' not only converts a full clause into a noun phrase, but simultaneously has the effect of transforming the questionable truth claim into a preconstructed presupposition. Thus it may be seen that a variety of grammatical structures realize ideological and discursive effects of the kind that Pêcheux sought to draw attention to, even if he narrowly located then in relative clauses.

Modalities of Subjective Operation

In Pêcheux's account, this problem of definition leads to acute difficulties when exploring the role of transverse discourse. The sustaining effect which he attributes to this kind of discourse is predicated upon the intersection of one type of discursive formation with another. This effect, however, is difficult to demonstrate without clear criteria for distinguishing one kind of discursive formation from another. And in practice, it seems quite possible for a sustaining effect to be achieved without insisting upon the background presence of a rival discursive formation. For example, in one of the Sunday Express items referred to above, it would seem that a recurrent use of explicative relative clauses (or appositional constructions of an associated type) involves the juxtaposition of one temporal state of affairs with another. This may also be seen in the following examples drawn from a Sunday Times account: #(A) ``Wright, whom they [his former colleagues] now despise, was one of their most trusted officers.'' #(B) ``Wilson, once so garrulous about MI5, will no longer discuss it...'' #(C) ``But Barbara Castle, then a cabinet minister, said last week that he [i.e., Wilson] had been incensed by the smear campaign.''

This kind of juxtaposition does, following Pêcheux's argument, create, as it were, an intradiscursive implicational space which can be filled in various ways. In (A), for instance, the contrast between present contempt and past trust casts some doubt on the reliability of the judgments exercised by Wright's former colleagues. In (B), the contrast between Wilson's former garrulity on the subject of MI5 and his present taciturnity suggests some unstated intervening event which causes the change in attitude. In neither case, however, do the possible implications seem to derive necessarily from a rival discursive formation. On the contrary, it is quite possible to argue that they issue from within the very discursive formation upon which the text is predicated. Thus it seems that transverse discourse, especially when discursive formations themselves are designated in arbitrary terms according to shifting criteria, is not necessarily dependent upon a rival discursive formation to fill the spaces of intradiscourse (cf. Chevalier, 1990).

Here it is crucial to recall that implicit to Pêcheux's notion of intradiscourse is an active subjectivity. Much effort has been expended to illustrate how different forms of selection and combination render the contradictory text ``coherent.'' The intradiscourse-interdiscourse dynamic has also been thrown into sharp relief, as it has been shown to be capable of providing critical discourse analysis with the rudimentary elements to elucidate the interpellation of the individual as the ``forgetful'' subject of its discourse (and in this way refine substantively Althusser's view to the imaginary relation). Interpellation, it has been noted, is achieved by the identification of the human subject with the discursive formation that dominates him or her: that is, the relation in which her or she is constituted as a subject and where his or her ``complete freedom'' as a speaking subject is located (1982, pp. 114, 125). It is this relation of identification which, in turn, always acts to reinforce the imaginary unity of the human subject--or does it?

Here an intriguing problem may be brought to the fore. It would appear that this relation provides ``already available'' subject positions for the good subject who, realizing his or her subjection in the form of the ``freely consented to,'' ``spontaneously'' assumes the position offered by the universal Subject ``in all liberty'' (1982, pp. 114, 156). This subject accepts the image of self projected by the dominant discourse. Pêcheux provides a historical example: ``France is in danger / We are all Frenchmen / We are at war! [or] `A French Soldier never retreats!' `To the last man!' '' (1982, p. 165). But what of the subject who does not accept the preferred image of self as it is inflected?

The discourse of what then becomes a bad subject ``turns against'' the dominant identification, primarily by taking up a position that consists of initiating a separation, challenge or revolt against ``what the `universal Subject' `gives him [or her] to think': a struggle against ideological evidentness on the terrain of that evidentness, an evidentness with a negative sign, reversed in its own terrain'' (1982, p. 157). That is, Pêcheux argues, the ``trouble-making'' subject does not recognize those meanings lived by the good subjects as being ``obvious'' or ``natural,'' but rather as achieved contradictorily, and therefore the identity on offer is refused. The philosophical and political forms of a counter-discourse will then produce in the ``bad'' subject a counter-identification with the discursive formation imposed on him or her by interdiscourse, yet one where the evidentness of meaning remains complicit with it, in this case to be rejected (1982, p. 157). For example: `` `It is always the same people who get killed' / `Down with war!' / `Long live peace!' '' (1982, pp. 165-166).

Pêcheux thus elects to introduce into his scheme a third modality of subjective operation, that of disidentification, which accounts for the taking up of an antagonistic or non-subjective position, an effect working to transform or displace (but never escape entirely) the dominant practices of ideological subjection (1982, p. 159). By operating in ``reverse'' or ``on and against itself,'' ideology as the process of interpellation threatens to ``rearrange'' or ``overthrow'' both the complex of ideological formations and the overlapping discursive formations (1982, p. 159). To re-organize the relationships concealed in the opposition war/peace indicative of the above examples, it would be recognized that ``a struggle for peace which [is] not also at the same time a struggle for socialism is a nonsense, because pacifism is an illusion so long as socialism has not been established'' (1982, p. 168).

How, then, to continue with this work of theorizing the reactive, contradictory or antagonistic nature of this process as distinct from ``dominant ideology''? If, as the Althusserians would suggest, to consider the question of ideology from the standpoint of reproduction also implies, by definition, considering ideology from the standpoint of resistance to reproduction, then for Pêcheux (1983a) analysis must be re-situated so as to account for ``the multitude of heterogeneous resistances and revolts which smoulder beneath dominant ideology, threatening it constantly'' (1983a, p. 26). He thus calls for critical research to undertake a re-evaluation of dominated ideologies, suggesting that they be treated as a series of ideological effects to be identified as they emerge from domination and work in opposition to it through its inherent gaps and failures (1983a, p. 27). Similarly, his observation that analyses must look toward what is happening ``below'': namely, in the space that constitutes the ordinary of the masses, is an astute one (1988, p. 645). ``It is becoming increasingly obvious,'' Pêcheux writes, ``that we must learn to listen to this often silent speech enclosed within the urgency of survival. [...] We must hear the articulations embedded in the `ordinary way' of meaning'' (1988, p. 645).

A Postmodern Other?

Still, to acknowledge the benefits of a series of disruptive strategies is not to endorse the resultant conceptual or political project. What some theorists offer in their willingness to abandon the ``the dialectic between conditions and consciousness'' is frequently purchased, increasingly in the name of postmodernism, at the price of valourizing the power-resistance dynamic as a principle and not as a force contingent on concrete historical circumstances. Here we would want to affirm the value of Pêcheux's view to the dialectic of divergent practices, meanings and identities within the codes or regimes of signification naturalized as common sense, as the reality of ruling ideology. To confront the hierarchies of ``otherness,'' of the ``marginal,'' the ``alien,'' or the ``illegitimate,'' Pêcheux's insight that the resistance points of a counter-hegemonic politic must be made ``graspable'' precisely as they are implicated in what he aptly calls the ``ordinary way'' of meaning is crucial.

Therein lies the necessary basis for a politically reflexive form of analysis, as it places the re-articulation via linguistic mechanisms of social identification and otherness squarely on the agenda. Despite the limitations of Pêcheux's notion of discursive formation that we have documented above, its capacity for articulating the relationship between, on the one hand, the normative ``rules'' of ideology and, on the other hand, aspects of its subjective appropriation or negotiation, offers significant heuristic advantages. As he himself characterizes his project: ``All my work finds its definition here, in this linking of the question of the constitution of meaning to that of the constitution of the subject, a linking which is not marginal (for example the special case of the ideological `rituals' of reading and writing), but located inside the `central thesis' itself, in the figure of interpellation'' (1982, p. 105). How, then, to develop further the connection between the constitution of meaning and the constitution of the human subject? At first glance, the most straightforward approach would be to extend an adapted Whorfian or Saussurean account, whereby experience (subjectivity) remains inchoate and in flux until the entry into the symbolic order, language, which makes available categories, terms and processes, not only for the rendering of experience, but also for its active shaping and organization.

The problem, however, with developing the linkage in these terms is that the symbolic order--the language--is conceptualized within Whorfian and Saussurean approaches as a unified totality, implying in consequence that ideology (here as world view) would be the same for all members of the language community. Pêcheux's distinction between linguistic basis and discursive process is precisely designed to offer a different way of articulating this interdependence. Subjectivity is not evenly constituted ``on the linguistic basis'' in terms of a unified symbolic order. Instead, as we have seen, Pêcheux proposes that particular meanings are constituted in particular discursive processes; and since ``every discursive process is inscribed into an ideological relationship,'' subjectivity is constituted in uneven and contradictory ways, depending upon the discursive formation that the processes are locked into. Once again, interpellation is thus predicated upon the discursive operation of the ``preconstructed'' and ``the sustaining effect.''

In more concrete terms, what seems to be at stake is as follows. Any enunciation within intradiscourse opens up implicational spaces either for the operation of the preconstructed or for the sustaining effect. These spaces (inferential gaps) require completion by the subject to secure the intelligibility of whatever has been enunciated. The subject supplies the sense of the enunciation by recourse to transverse discourse. Accordingly, supplying the inferential links or filling the implicational space within interdiscourse is at the same time to be recruited to the terms of that very transverse discourse which provides the grounds of its intelligibility. But, insofar as the subject in this way makes the enunciation intelligible, he or she has, by this very act, been interpellated. Crucially, of course, this may occur outside the level of conscious awareness. It is as if ideology, along the axis of transverse discourse underpins the obviousness of the enunciation; and--in the act of recognition that subscribes to this obviousness--the subject is interpellated.

Pêcheux, by focusing particularly on the role of language in this process, supplies more detail about how ideology is subjectively appropriated in the act of interpellation. His treatment provides a way of developing the notion of interpellation in such a fashion that it extends beyond the more abstract characterizations of the positioning of the human subject within ideology typical of much work in cultural studies. One notable limitation of the Althusserian account, especially when it is rendered in its most schematic form, is that the role and practice of specific ideologies becomes obscured under an ahistorical rubric governing the operation of Ideology-in-general: in consequence, his emphasis falls upon the constitution of the Subject as such (this being justified by reference to it being ``the elementary ideological effect'') at the expense of exploring particular interpellations into distinct ideologies. In this respect, Pêcheux's emphasis on discursive process as a mode of interpellation allows precisely for a greater degree of specificity. As long as it is possible to delineate the precise parameters of concrete discursive formations, it should be possible to go beyond merely formal accounts of the positioning of the human subject within ideology, and to address instead the interpellation of him or her by particular ideologies in enunciated discourses.

To close, then, we are calling for the positive elaboration of Pêcheux's provocative discussion of the materiality of language. We envision this work proceeding along three crucial trajectories: first, to argue that the notion of discursive formation is necessary to a materialist theory of language and ideology is, at the same time, to acknowledge that it still requires a more rigorous definition. If a number of problems derive from the sheer sketchiness of Pêcheux's demonstration, the most pressing difficulty noted above is the reductive sense of determination at work, one which operates unidirectionally ``downwards'' from the utterance to social organization. Second, the notion of discursive process needs to be developed to include a wider range of practices than those associated with relative clause constructions. Finally, further attempts must be undertaken to mobilize the heuristic, analytic and strategic possibilities of this rich, if inchoate, category of disidentification.

As the rigid disciplinary boundaries separating discourse analysis from the realm of cultural theory are slowly dismantled, cultural studies researchers will be well poised to acquire a new set of linguistic tools for the explicit purpose of theorizing this play of ideology and power in discourse. At a time when the conceptual terrain formerly marked off by ``structures of feeling'' and ``interpellation'' are consistently being rewritten into the superficial terms of aesthetic gestures, the struggle to develop and refine a new critical vocabulary for speaking the world is intensifying. We have therefore endeavored to show how Pêcheux's research framework is of continuing relevance to cultural studies, possessing as it does much potential for analyzing concrete instances of ideological and discursive effectivity. 1 We would like to express our thanks to Jacques Chevalier, Line Grenier, and William Straw, as well as to the editor and anonymous reviewers of this journal, for their thoughtful critiques of an earlier version of this manuscript. Funding for this research was provided, in part, by the Social Sciences and Humanities Research Council of Canada. This is a familiar criticism, one also to be found in Firth (1957), Hymes (1964), and Halliday (1978). Interestingly, it also recalls Bakhtin & Volosinov's (1973) strictures on abstract objectivism: ``The actual reality of language-speech is not the abstract system of linguistic forms [...] but the social event of verbal interaction implemented in an utterance or utterances'' (1973, p. 94). Here care must be taken so as to ensure that a key dimension is not sacrificed in this shift, namely, the ability to trace the concrete trajectory of the sign's multi-accentuality for the explicit purpose of identifying both possible and actual points of discontinuity and rupture across the field of hegemonic articulations (see Laclau & Mouffe, 1985). Significantly, one important attempt to apply aspects of Pêcheux's work cuts this Gordian knot merely by collapsing the plane of ideology into that of discourse by speaking of the ``ideological-discursive formations'' (see Fairclough, 1989). A further example of how the question of nominalization in news text may be addressed is found in Hodge & Kress, 1988. In their account, nominalization may be seen as an extreme case of grammatical reduction inasmuch as nominals, such as ``negotiations'' or ``management,'' are derived from underlying predicate structures built around a verb. Thus, an example such as: ``Negotiations took place at ACAS'' is treated by Hodge & Kress as if derived from an underlying structure such as the following: X negotiated with Y at ACAS where the verb (``negotiate'') functions as a two place predicate in an underlying clause in which the associated participant roles (X and Y) require specification for the clause to be structurally complete (such is also the case with: ``The miners negotiated with the Coal Board at ACAS''). Accordingly, they argue, when a structure such as this is nominalized (e.g., ``negotiations''), one important consequence is that the participant roles (e.g., ``miners'' and ``the Coal Board'') may easily be deleted, where the grammatical process of reduction leads to a reification and objectification of the underlying process.

Inasmuch as the process is truly one of objectification, where an action represented by a verb becomes rendered ``thing-like'' by nominalization, then it bears some comparison with Pêcheux's account of the ``preconstructed,'' in which entities (e.g., ``He who saved the world by dying on the Cross'') assume existential solidity, even where their existence is denied elsewhere in the same enunciation (e.g., ``He who saved the world by dying on the Cross does not exist''). A recent account of ``discourses'' attempts to resolve this kind of ambiguity in the following way: ``Any account of a discourse or a discursive practice must include its topic area, its social origin, and its ideological work: we should not, therefore, think about a discourse of economics, or of gender, but of a capitalist (or socialist) discourse of economics, or the patriarchal (or feminist) discourse of gender. Such discourses frequently become institutionalized, particularly by the media industries, in so far as they are structured by a socially produced set of conventions that are tacitly accepted by both industry and consumers. In this sense we can talk about the discourse of the news, or of advertising: these discourses still exhibit our three defining characteristics--a topic area, a social location, and the promotion of a particular social group'' (Fiske, 1987, pp. 14-15). This proposed solution, however, conflates several different and contradictory levels of abstraction and merely insists by assertion on the existence of those very discourses it should be at pains to define. Indeed, it is Allor's (1988) contention that postmodernism's extended disavowal of signification and mediation is itself rapidly forming the site of a new paradox: namely, that in collapsing the depth model characteristic of semiotics (where the play of signifiers links with a hierarchy of levels or instances), postmodernism has brought about an increased focus on discursive and textual analysis (1988, p. 300). That is, this dismissal of a problematic centred on conventions of human practice (meaning, representation) elides the contradictory effectivity of social practices, thereby engendering a mode of critique that will ultimately be forced to find refuge in functionalism. This type of analysis, he argues, can not stand without recourse to ``epochal logics'' to explain the arrangements of ``surfaces'' under examination.

References

Allor, Martin. (1987). Projective readings: Cultural studies from here. Canadian Journal of Political and Social Theory, 11(1-2), 134-138.

Allor, Martin. (1988). Theoretical practice in the post. Communication, 10, 295-303.

Althusser, Louis. (1969). For Marx. New York: Vintage.

Althusser, Louis. (1971). Ideology and ideological state apparatuses: Notes towards an investigation. In Lenin and Philosophy (pp. 127-186). New York and London: Monthly Review.

Bakhtin, Mikhail. (1981). The dialogic imagination. Austin: University of Texas Press.

Bakhtin, Mikhail, & Volosinov, V. N. (1973). Marxism and the philosophy of language. Cambridge, MA: Harvard University Press.

Bauman, Zygmunt. (1988). Is there a postmodern sociology? Theory, Culture and Society, 5, 217-237.

Charland, Maurice. (1990). Rehabilitating rhetoric: Confronting blindspots in discourse and social theory. Communication, 11(4), 253-264.

Chevalier, Jacques M. (1990). Semiotics, romanticism and the scriptures. Berlin and New York: Mouton de Gruyter.

Cousins, Mark. (1985). Jokes and their relation to the mode of production. Economy and Society, 14(1), 94-112.

Eagleton, Terry. (1991). Ideology. London: Verso.

Fairclough, Norman. (1989). Language and power. London: Longman.

Firth, J. R. (1957). Papers in linguistics 1934-1951. London: Oxford University Press.

Fiske, John. (1987). Television culture. London: Methuen.

Foucault, Michel. (1971). Orders of discourse. Social Science Information, 10(2), 7-30.

Foucault, Michel. (1972). The archaeology of knowledge. New York: Pantheon.

Fowler, Roger, Hodge, Bob, Kress, Gunther, & Trew, Tony. (1979). Language and control. London: Routledge & Kegan Paul.

Franklin, Sarah, Lury, Celia, & Stacey, Jackie (Eds.). (1991). Off-centre: Feminism and cultural studies. London: Harper Collins.

Frow, John. (1986). Marxism and literary history. Oxford: Basil Blackwell.

Goodrich, Peter. (1982). Materialism and linguistics. Radical Philosophy 32.

Gramsci, Antonio. (1971). Selections from the prison notebooks. New York: International.

Grossberg, Lawrence. (1983). Cultural studies: Revisited and revised. In Mary S. Mander (Ed.), Communications in transition (pp. 39-70). New York: Praeger.

Hall, Stuart. (1982). The rediscovery of ``ideology'': Return of the repressed in media studies. In Michael Gurevitch, Tony Bennett, James Curran, & Janet Woollacott (Eds.), Culture, society and the media (pp. 56-90). London: Methuen.

Hall, Stuart. (1985). Signification, representation, ideology: Althusser and the post-structuralist debates. Critical Studies in Mass Communication, 2(2), 91-114.

Halliday, M. A. K. (1978). Language as social semiotic. London: Edward Arnold.

Hartley, John, & Montgomery, Martin. (1985). Representations and relations: Ideology and power in press and TV news. In T. A. van Dijk (Ed.), Discourse and communication (pp. 233-269). New York: Walter de Gruyter.