4

Messages and meanings

JANET WOOLLACOTT

Ideology is the final connotation of the totality of connotations of the sign or the context of signs. (Umberto Eco, 1971, p. 83)

Interest in and discussion of the mass media has come from a variety of theoretical and disciplinary sources. Within these wide-ranging and sometimes contradictory approaches, the analysis of media messages has been seen as of varying importance. American concern with mass communications has tended to focus on a model of communication which stressed the relationships between the individuals involved. In this tradition the communication process was conceived of as a relationship between a sender of messages on the one hand and a receiver of messages on the other. The mass communication process merely converted the receiver from being one to being many individuals. Given this image of the workings of the mass media, the attention of researchers was directed at the psychological dispositions of the producers of mass media messages and at the effects of the message on the members of the audience. The analysis of the meaning of media messages came to be subsumed in these areas of study. Moreover, early Marxist studies of the media, whilst based on very different theoretical premises, tended to be more concerned with the overall ideological role of the mass media in capitalist societies and less concerned with the meaning of and the production of meaning within specific media messages. When such questions were addressed they were inflected with a form of cultural pessimism. Members of the Frankfurt School, for example, attempted to show that mass culture, and particularly, American mass culture with which they had acquired a forced familiarity, was a debased culture. Adorno and Horkheimer (1977) suggested that the culture of a society under monopoly capitalism was peculiarly repressive in that while bourgeois culture offered a better and more valuable world realizable by every individual from within, mass culture produced a more totalitarian state in which even the illusory advantage of inner freedom of the individual was lost.

In the culture industry the individual is an illusion not merely because of the standardization of the means of production. He is tolerated only so long as his complete identification with the generality is unquestioned. Pseudo-individuality is rife: from the standardized jazz improvisation to the exceptional film star whose hair curls over her eye to demonstrate her originality. What is individual is no more than the generality's power to stamp the the accidental detail so firmly that it is accepted as such. (Adorno and Horkheimer, 1977, p. 374)

Adorno's contention that the mass media provided the ideological counterpart to the economic development of capitalist societies, although very different from the propositions of researchers who concentrated on the ‘effects’ on audiences of violent or sexual aspects of the media, led to a similar intellectual lacuna in the concrete analysis of media messages.

The distinctive feature of production in the mass media as opposed to general economic production, is that it is concerned with the production and articulation of messages within specific signifying systems, the rules and meanings of which we tend to take for granted. The messages in the media are both composed and interpreted in accordance with certain rules or codes. When we see a news event on television or film, we do not see that event ‘raw’ but we see a message about that event. We can read the message and interpret it but we take for granted the rules and codes through which we read and interpret. The analysis of mass media messages and their discourses of meaning are clearly important for an understanding of mass communications. As Hall has suggested: ‘we must recognize that the symbolic form of the message has a privileged position in the communication exchange: and that moments of “encoding” and “decoding”, though only “relatively autonomous” in relation to the communication process as a whole, are determinate moments’ (Hall, 1973, p. 2).

The last ten years have seen an increasing concern with the formal semiological analysis of the mass media message, in news coverage advertising, feature films and television fiction and this has gone hand in hand with developments in theories of ideology. This article will seek to explore some of the developments and problems of this theoretical alliance.

The traditional method for the analysis of the meaning of mass communications messages was content analysis. Content analysts operated by establishing certain conceptual categories in relation to media content and then quantitatively assessing the presence or absence of these categories with varying degrees of sophistication. Content analysis as a method was used by researchers of quite different theoretical backgrounds and with varying degrees of success but the method inevitably stresses the manifest content of the message as the most important area for social scientific analysis. The manifest content of the message was taken to provide ‘the common universe between the emission, the reception and the researcher’ (Camargo, 1972, p. 124). While content analysis was much used in research operating on an individualistic model, it has also been successfully employed to support research on race and the area of news values (Glasgow University Media Group, 1976). Content analysis clearly has advantages for the systematic investigation of a wide range of material. Cantril's original study of the impact of The Invasion from Mars, for example, simply reproduced the script of the radio play, but if a researcher wanted to look at 200 plays this was no longer feasible (Cantril et al., 1940). At the same time, however, content analysis has usually proved to be quite limited in conveying the meaning of specific media messages.

More recent research has tended to conceptualize the problems of understanding media messages rather differently. Semiological or structuralist studies, deriving many of their theoretical premises from linguistics, reasserted a concern with media messages as structured wholes rather than with the quantified explicit content of fragmented parts of messages. Semiology as Burgelin points out, is not only rarely quantitative but also contains an implicit critique of the quantitative pre-occupation of content analysis.

But above all there is no reason to assume that the item which recurs most frequently is the most important or the most significant, for a text is, clearly, a structured whole, and the place occupied by the different elements is more important than the number of times they recur. Let us imagine a film in which the gangster hero is seen performing a long succession of actions which show his character in an extremely vicious light, but he is also seen performing one single action which reveals to a striking degree that he has finer feelings. So the gangster's actions are to be evaluated in terms of two sets of opposites: bad/good and frequent/exceptional. The polarity frequent/exceptional is perceptible at first sight and needs no quantification. Moreover we clearly cannot draw any valid inferences from a simple enumeration of his vicious acts (it makes no difference if there are ten or twenty of them) for the crux of the matter obviously is: what meaning is conferred on the vicious acts by the fact of their juxtaposition with the single good action? Only by taking into account the structural relationship of this one good action with the totality of the gangster's vicious behaviour in the film can we make any inference concerning the film as a whole. (Burgelin, 1972, p. 319)

This fairly familiar gangster plot, Burgelin contends, cannot be understood in terms of the quantification of its manifest content but necessitates an examination of the relationship of the different parts of the plot and the way in which they are articulated to form a specific and complex message with various levels of signification.

Some of the early sound gangster films such as Public Enemy (1931) aroused considerable official concern because of their violence. Similarly later television programmes on crime and other genres such as Westerns or spy films have aroused the kind of concern that led researchers (who in many ways knew better) to conceptualize the mass communications process in terms of a behaviourist model in which representations of violence were seen to effect in an unmediated way the opinions and actions of individual members of the audience. Content analysis was often used as a tool in this kind of research and its focus on a simple level of manifest content allowed a straightforward transition to be made between violence on the screen and delinquency, gangfights and muggings elsewhere. Semiological studies, on the other hand, focused on film as a discourse, on the film as a communication about violence rather than violence itself, and in that sense, reoriented research towards the system of rules which governed that discourse generally, and the gangster film in particular, rather than specific violent episodes. Within this kind of analysis the codes governing the genre of film noir gave different violent episodes different meanings. Indeed, the violent act could only be understood in relation to other elements in the film and in terms of the conventions of the genre. Such acts were no longer seen to have a single fixed meaning but to be capable of signifying different values and presenting different codes of behaviour depending upon how they are articulated as signs amongst other signifying elements within a discourse.

Semiological studies present their own ambiguities and difficulties not least because semiology, unlike content analysis, is not a method but constitutes a constellation of studies in art, literature, anthropology and the mass media which in some way developed or made use of linguistic theory. As a philosophy, as a theory, as a set of concepts and as a method of analysis, semiology has had many facets and has been subject to various interpretations, debates and polemics. Semiology emerged from the study of language problems and the structure of language. Barthes once defined structuralism as a method for the study of cultural artefacts which orginated in the methods of contemporary linguistics. The early structuralist studies attempted to uncover the internal relationships which gave different languages their form and function. Later semiological work took a broader view and attempted to lay down the basis for a science of signs which would include not only languages but also any other signifying system.

The contributions that linguistic analogies made to the study of other cultural forms did not rest solely on the blind application of the methods of one discipline to another, but developed in rather different ways in relation to different theoretical contexts. For our purpose, in working out the methods through which semiologists examine the mass media, it is worth noting some distinctive features of semiological studies in which there is a certain tension. Semiology is distinguished by its insistence on the importance of the sign. This involves the initial isolation of the signifier as an object of study from the signified. This is relatively easy to understand when the object of investigation is language but is perhaps less easy to comprehend when the object of research is a non-verbal sign system. One famous structuralist anthropological study is Lévi-Strauss's analysis of kinship. Here Lévi-Strauss (1969) considers marriage rules and systems of kinship in a number of ‘primitive’ societies as a ‘kind of language’, that is to say, ‘a set of operations designed to ensure a certain type of communication between individuals and groups’ (Jameson, 1972, p. 111). The message is made up of the women of the group who circulate between the clans, dynasties and families, whereas in language it would be the ‘word of the group’ which circulated between individuals. The priority of the language model suggested here is typical of semiological treatment whether it be of kinship systems, furniture and fashion, of films and television programmes or of toys and cars. Saussure, who laid down the outlines of semiology as a ‘science of signs’, contended that the advantage of the linguistic model was that it cut through the apparent naturalness of actions or objects, to show that their meaning is founded on shared assumptions or conventions. In this sense, the methods of linguistics compel the researcher to study the system of rules underlying speech rather than any external influences or determinants.

As the example of gangster films referred to earlier indicates, another characteristic feature of semiological analysis is that it appears to concentrate on the internal structuring of a text or message. Barthes points to this concern of semiologists with immanent analysis:

The relevance shown by semiological research concerns by definition, the meaning of the analysed objects: we consider the objects solely in relation to their meaning without bringing in, at least not prematurely, that is, not before the system be reconstituted as far as possible, the other determinants (psychological, sociological or physical) of these objects; we must certainly not deny these other determinants, which each depend upon another relevancy; but we must treat them also in semiological terms, that situate their place and function in the system of meaning…. The principle of relevancy obviously requires of the analyst a situation of ‘immanence’, we observe the given system from within. (Barthes, 1967, p. 95)

The internal relationships of any structure are therefore, what gives meaning to any element in the structure. Hence if a particular action is impolite, it is not because of its intrinsic qualities but because of certain relational features which differentiate it from polite actions. Structural analysis tends to stress binary oppositions of this type as a heuristic device, ‘a technique for stimulating perception, when faced with a mass of apparently homogenous data to which the mind and eye are numb: a way of forcing ourselves to perceive difference and identity in a wholly new language the sounds of which we cannot yet distinguish from each other. It is a decoding or deciphering device, or alternatively a technique of language learning’ (Jameson, 1972, p. 113).

The focus on the internal relationships of a text does raise certain problems. Many of the Russian formalist studies, for example, attempt to examine literature in terms of its internal structure. Propp's The morphology of the folk tale (1968) attempts to identify the narrative structure of the Russian folk tale. Propp was reacting strongly against the treatment of isolated elements in folk tales whereby tales were separated and classified according to whether their principal characters are animals, ogres, magical figures, humorous figures and so forth. Propp argues that the identity of character and landscape and the nature of obstacles is less important than their function. ‘Function is understood as an act of character defined from the point of view of its significance for the course of the action’ (Propp, 1968, p. 21). Propp establishes that the narrative of the folk tale follows a certain pattern. The story begins with either an injury to a victim or the lack of some important object and ends with retribution for the injury or the acquisition of the thing lacked. The hero is sent for on the occasion of the injury or the discovery of the lack and two key events follow:

1. He meets the donor (a toad, a hag, a bearded old man, etc.) who after ‘testing’ him, supplies him with a magical agent which enables him to pass victoriously through his ordeal.
2. He meets the villain in decisive combat or he finds himself with a series of tasks or labours which, with the help of his agent, he is ultimately able to solve properly.

The latter part of the tale may constitute a series of retarding devices before the ultimate transfiguration of marriage or coronation…. Propp identifies 31 narrative functions through which it is possible to classify folk tales.

The main problem of this focus on the internal relationships of a particular group of texts is that the specificity of any one text both in the context of its production and its reading, through which meaning is established, is lost. Russian folk tales become indistinguishable from the latest episode of The Sweeney, from Star Wars or from a Raymond Chandler novel. Indeed, Eco's analysis of the narrative structure of the James Bond novels which suggests that the novels are fixed as a sequence of moves inspired by a code of binary oppositions comes remarkably close to Propp's typical narrative. Eco suggests that the invariable scheme of the Bond novels is as follows:

A. M moves and gives a task to Bond.

B. The villain moves and appears to Bond (perhaps in alternating forms).

C. Bond moves and gives a first check to the villain or the villain gives first check to Bond.

D. Woman moves and shows herself to Bond.

E. Bond consumes woman: possesses her or begins her seduction.

F. The villain captures Bond (with or without woman, or at different moments).

G. The villain tortures Bond (with or without woman).

H. Bond conquers the villain (kills him or kills his representative or helps at their killing).

I. Bond convalescing enjoys woman, whom he then loses. (Eco, 1960, p. 52)

What takes Eco's analysis of Bond beyond some universal fairy tale is that Eco shows that this coded schema, which forms the basis for all the Bond novels (with the exception of The Spy Who Loved Me) is closely linked to a series of oppositions. So the opposition of Bond and the villain is accompanied by an opposition between the western world and the Soviet Union, between Britain and non anglo-saxon countries, between ideals and cupidity, between chance and planning, between excess and moderation, between perversion and innocence, between loyalty and disloyalty.

The internal oppositions within the text are obviously part of wider ideological discourses, notably the ideology of the Cold War. Eco makes this clear in the character of some of the oppositions, particularly that of Bond and the villain.

The villain is born in an ethnic area that stretches from central Europe to the Slav countries and to the Mediterranean basin: as a rule he is of mixed blood and his origins are complex and obscure; he is asexual or homosexual or at any rate, is not sexually normal; he has exceptional inventive and organizational qualities which help him to acquire immense wealth and by means of which he usually works to help Russia: to this end he conceives a plan of fantastic character and dimensions, worked out to the smallest detail, intended to create serious difficulties either for England or the Free World in general. In the figure of the villain in fact, there are gathered the negative values which we have distinguished in some pairs of opposites, the Soviet Union and countries which are not Anglo-Saxon (the racial convention blames particularly the Jews, the Germans, the Slavs and the Italians, always depicted as half-breeds), Cupidity elevated to the dignity of paranoia, Planning as technological methodology, satrapic luxury, physical and psychical excess, physical and moral Perversion, radical Disloyalty. (Eco, 1960, p. 44)

Moreover, Eco takes his concern beyond the structure of the text (the Bond novels) in other ways in examining the relationship between the ‘literary inheritance and the crude chronicle, between eighteenth-century tradition and science fiction, between adventurous excitement and hypnosis’ (Eco, 1960, p. 74). In seeking to establish relationships both with previous literary forms and more minimally and dubiously with audience response, Eco attempts to go beyond a predominantly inductive analysis such as that of Propp to place the narrative structure of the Bond novels within literary discourse and to suggest the necessity of placing the reading and understanding of the meaning of the novels in specific social practices.

Eco's analysis also indicates the tension in semiology between formal textual analysis and the realm of the signified and between different texts and between different signifying systems. It is in this area that semiology becomes vitally concerned with ideology. The principal conceptual tool of Saussurean linguistics was the sign and the concept of the sign distinguished between various elements in the process of speech, in the now classic formulation:

Signifier—Sign—Signified

The relationships involved here are not those between the word and the real world but between the signifier (an acoustic image, for example) and the signified (the concept). In this sense, semiology excludes consideration of the ‘real world’ but at the same time the notion of the sign inevitably suggests a reality beyond itself. There is then a certain ambivalence in semiological studies between the analysis of signifying systems such as the mass media as internally and logically structured and the search-for underlying structures. Different theorists have attempted to locate the underlying structures, in areas as different as ‘literariness’ or the universal qualities of the human mind. The theoretical alliance of semiology and Marxism in the study of the mass media has produced the argument that the underlying structure is that of ‘myth’ or ‘ideology’.

Roland Barthes’ Mythologies (1972) suggested both that semiotics could be applied to areas which had not previously been noted for their ‘meaning’ and that the results of such an analysis constituted an account of contemporary ideology, as in the following passage.

If we are to believe the weekly Elle which some time ago mustered twenty women novelists on one photograph, the woman of letters is a remarkable zoological species: she brings forth pell-mell, novels and children. We are introduced, for example, to Jaqueline Lenoir (two daughters, one novel); Marina Grey (one son, one novel); Nicole Dutreil (two sons, four novels), etc. What does this mean? This: to write is a glorious but bold activity; the writer is an ‘artist’, one recognizes that he is entitled to a little bohemianism. As he is in general entrusted—at least in the France of Elle—with giving society reasons for its clear conscience, he must, after all, be paid for his services: one tacitly grants him the right to some individuality. But make no mistake: let no women believe that they can take advantage of this pace without having first submitted to the eternal statute of womanhood. Women are on the earth to give children to men; let them write as much as they like, let them decorate their condition, but above all, let them not depart from it: let their biblical fate not be disturbed by the promotion which is conceded to them, and let them pay immediately, by the tribute of their motherhood for this bohemianism which has a natural link with a writer's life. (Barthes, 1972, p. 50)

Barthes was in no sense remarkable for his identification of ideological forms. After all Marxists had been describing paintings, novels and the mass media as ideological for many years, and rather more occasionally had analysed the meanings of specific ideological forms. What Barthes established was the use of semiology as a preamble to the study of myth or ideology and in so doing, he pointed to some of the specific problems of analysing the mass media as signifying systems.

In abstract it is not difficult to apply the central concepts of the structuralist conceptual apparatus, ‘sign’, ‘code’ (language) and ‘message’ (speech) to the mass media. Art historians for example, such as Panofsky, rapidly identified the fixed iconography of early movies.

There arose, identifiable by standardised appearance, behaviour and attributes, the wel-remembered types of the vamp and the straight girl (perhaps the most convincing modern equivalents of the medieval personifications of the vices and virtues), the family man and the villain, the latter marked by a black moustache and a walking-stick. Nocturnal scenes were printed on blue or green film. A checkered tablecloth meant once for all, a ‘poor but honest’ milieu, a happy marriage soon to be endangered by a shadow from the past was symbolised by the young wife's pouring the breakfast coffee for her husband; the first kiss was invariably announced by her kicking out with her left foot. (Panofsky, 1934, p. 25)

At the same time, however, there was a sense in which film and photography involved some crucial changes from preceding signifying systems, such as painting. Benjamin, in seeking to indicate the changes for works of art brought about by the process of mechanical reproduction, suggests that such changes can be illuminated by comparing the painter and the cameraman. The painter’, he states, ‘maintains in his work a natural distance from reality, the cameraman penetrates deeply into its web’ (Benjamin, 1977). Even naturalistic painting usually makes clearer the painter's presence, his techniques and codes, than does photography or film.

There is a tremendous difference between the pictures they obtain. That of the painter is a total one, that of the cameraman consists of multiple fragments which are assembled under a new law. Thus for contemporary man the representation of reality by film is incomparably more significant than that of the painter since it offers precisely because of the thoroughgoing permeation of reality with mechanical equipment, an aspect of reality which is free of all equipment. (Benjamin, 1977, p. 400)

The problem is that photography and film, unlike many other signifying forms, appears to record rather than to transform. Barthes suggests, for example, in his initial analyses of advertisements that the photographic component constituted the paradox of being ‘a message without a code’ (Barthes, 1971).

Barthes then continues, however, to establish the codes through which advertisements and other mass media messages are constructed while simultaneously carrying the claim of having-been-there, the evidence of ‘this-iswhat-happened-and-how’. In his analysis of a Panzani advertisement, he points to the signs of marketing, the string-bag, stocked with Panzani tins, spaghetti and pepper and tomato, with the connotations of freshness of product and household use; to the colour tints of the poster (yellow, green and red) which signify Italianness reinforced by the Italian assonance of the Italian name, Panzani; to the assembly of different objects which suggest the idea of a whole culinary service and in which Panzani tins are equated with the natural products surrounding them; and finally to the aesthetic signified of still life. Barthes identifies three messages in the Panzani advertisement; a linguistic message, a coded iconic message and an uncoded iconic message. He suggests that one way of approaching the apparently uncoded message, that is, the literal image of the photograph, is to start with the linguistic message, then examine the literal image and finally examine the overall symbolic meaning of the message. This method of dealing with the uncoded iconic message (the literal image, the photograph) Barthes calls denotation while the analysis of the coded iconic message (the overall symbolic meaning of the advertisement) Barthes calls connotation. Clearly these modes of analysis are only analytically distinct in that there is no way to read a ‘literal image’ neutrally, which is not in some way dependent on coding and cultural conventions.

The distinguishing feature of Barthes’ formal readings of advertisements and other media messages is the identification of second-order meanings, meanings beyond those initially noted. In the case of the example from Elle, the connection between women, novels and children in the message signifies that women are only allowed to write if they have children—but it also goes beyond this in terms of the second-order meaning, whereby the complex of pictures and words and their meaning come to constitute a signifier for the idea that it is the natural place of women to produce children even if they also produce novels. Film and photography, Barthes suggests, operate upon us in a manner which suppresses and conceals their ideological function because they appear to record rather than to transform or signify. Hall uses this kind of analysis to establish the ideological character of news photographs:

New photos operate under a hidden sign marked ‘this really happened, see for yourself’. Of course the choice of this moment of an event as against that, of this person rather than that, of this angle rather than any other, indeed the selection of this photographed incident to represent a whole complex chain of events and meanings is a highly ideological procedure. But by appearing literally to produce the event as it really happened news photos repress their selective/interpretive/ideological function. They seek a warrant in that ever-pre-given neutral structure, which is beyond question, beyond interpretation: the ‘real’ world. At this level, news photos not only support the credibility of the newspaper as an accurate medium. They also guarantee and underwrite its objectivity (that is they neutralize its ideological function). (Hall, 1972, p. 84)

The analysis of news photographs is obviously very similar to that of Barthes but it perhaps registers more acutely because the conventions of news-reporting rely heavily on accepted canons of impartiality. Newsreporting presents itself as a selection of and impartial comment on ‘reality’ as it unfolds and uses photographs and films as evidence of reality ‘unfolding’. Yet a range of research studies on the position of women, on race, on the treatment of industrial relations and in particular on the role of the trade unions, would show quite clearly that such subjects have rarely been treated ‘impartially’ in news-reporting in the press or in broadcasting. A BBC survey conducted in 1962 showed that 58 per cent of the population used television as their main source of news and, even more significantly, that 68 per cent of the group interviewed believed that television news was a trustworthy medium. For this reason alone it could be seen to be important to establish that the claims of the news to ‘impartiality’ are dubious.

Semiotics, with its focus on the internal mechanisms through which meanings are produced in texts appeared to offer in relation to news coverage and many other areas, a way of engaging with the meaning of particular texts and of talking about more general ways through which signifying systems operate. Yet at the same time, if semiology was to be anything other than a set of formalist techniques, it had to be used and articulated within a general theory of ideology. I would suggest that semiology has been appropriated in a number of ways and has thereby been elided into a series of theoretical positions with which it is not altogether at one. I want to trace some of the problems of these theoretical elisions in relation to three positions, three arguments which take as a point of reference a semiological reading of a specific media message but which carry with them more general arguments about the nature of ideology.

Fiske and Hartley's recent introduction to reading television, for example, purports to be a first attempt to combine a theory of the cultural role of television with a ‘semiotic-based method of analysis whereby individual broadcast items can be critically “read’” (Fiske and Hartley, 1978). Fiske and Hartley appear at least initially to follow Barthes's ideas quite closely. Their own text is littered with concepts taken from Barthes although their argument about the role of television is very different, suggesting that while television may present messages with ‘preferred meanings’ and those preferred meanings ‘usually coincide with the perceptions of the dominant sections of society’, the form of television, its ‘constraints’ and ‘internal contradictions’, is one which allows ‘freedom of perception to all its viewers’. Essentially, Fiske and Hartley suggest that television functions to ‘de-familiarize’ the viewer precisely because the viewer is ‘spontaneously and continuously confronted’ with the necessity to negotiate a stance which will allow him to decode television programmes. Despite their expressed faith in the techniques of semiotics, these are largely eschewed. Lip-service is given to the terms but the authors proceed to analyse television programmes in a rather different way. Hence, the analysis of ‘News at Ten’ (7 January 1976) appears at first glance to follow closely Barthes's explanation of second-order signification in relation to the now famous example of the black soldier saluting the French flag on the cover of Paris-Match.

Thus the image in our film of a soldier clipping a magazine on to his rifle as he peers from his sandbagged bunker fortress in Belfast can activate the myth by which we currently ‘understand’ the army. This myth, as we shall show, is that the army consists of ordinary men, doing a professional and highly technological job. In order to trigger this myth the sign must be robbed of its specific signified, in this case, perhaps, of ‘Private J.Smith, 14.00 hours, January 4th 1976’. The sign loses this specificity and becomes now the second-order signifier; so the signified becomes oneof-our-lads-professional-well-equipped (not Private J.Smith) and the sign in this second order activates or triggers our mental ‘myth chain’ by which we apprehend the reality of the British soldier/army in Northern Ireland. (Fiske and Hartley, 1978, p. 42)

However, Fiske and Hartley move quickly from this to suggest both that myth meets our cultural needs and that those ‘needs require the myth to relate accurately to reality out there’.

Indeed, Fiske and Hartley go on to argue that their Belfast news film is part of a general process whereby television tests myths against reality and upon apprehending inaccuracy, initiates change.

Our news film from Belfast provides us with a particularly clear example of the way television can hint at the inadequacy of our present myths and thus contribute to their development. The sequence of army shots is followed immediately by a sequence showing the funerals of some of the victims of the violence. The last shot of the army sequence is of an armoured troop carrier moving right to left across the screen. There is then a cut to a coffin of a victim being carried right to left at much the same pace and in the same position on the screen. The visual similarity of the two signifiers brings their meanings into close association. The coffin contains the death that should have been presented by the soldiers in the troop carrier. Thus the myth of the army that underlies the whole army sequence has been negated by television's characteristic of quick-cutting from one vivid scene to another. (Fiske and Hartley, 1978, p. 44.)

There are a number of problems with this reading. Although the basic claim that Fiske and Hartley make (that television news is critical of institutions) is not implausible, the reading that supports it is. Buscome calls it ‘nothing more than a piece of free association’, pointing out that in order for the two shots of the army and coffin to have the meaning claimed, it would be necessary to show that it is a general rule of television news-editing that two subjects moving in the same direction across the screen will be read as ‘linked’ by more than space and time, or alternatively that the interpretation offered of the inter-cutting is in some way marked in the text, if for example, the second track said something like ‘Where the army goes, death is not far behind’ (Buscombe, 1979, p. 88). Basically Buscombe argues that the readings given by Fiske and Hartley are not ‘semiotic’ because they are not dependent on the idea of a set of structured relationships but are dependent on notions of similarity of content.

Moreover, Fiske and Hartley frequently seem to treat ideology as a functional, if mediated, reflection of reality.

The myths…cannot themselves be discrete and unorganized, for that would negate their prime function (which is to organize meaning): they are themselves organized into a coherence that we might call a mythology or an ideology. This, the third order of signification reflects the broad principles by which a culture organizes and interprets the reality with which it has to cope. (Fiske and Hartley, 1978, p. 46)

Television overall, they argue, is better than the literary traditions of the past both at using this area of mythology or ideology and disrupting it. Television, they suggest, performs a ‘bardic function’ operating as a mediator of language, producing messages not ‘according to the internal demands of the text’, ‘nor of the individual communicator’ but ‘according to the needs of the culture’. Similarly they explain the centralized institutionalization of television as a response to the culture's ‘need for a common centre’ and the ‘oral’ quality of television as a compensatory discourse for cementing the ‘non-literate’ working class into a culture which places ‘an enormous investment in the abstract elaborated codes of literacy’ (Fiske and Hartley, 1978, p. 86). Semiology is taken over and into this set of arguments with rather curious consequences. The systems of signification embodied in television are handed over to a formulation quite alien to what one would have thought were the first principles in a semiological ABC. The meanings of television programmes are seen to be structured not in accordance with any internal logic but in response to reality ‘out there’. Television as an historically specific social institution which constitutes the material base for specific discourses is reduced to the ‘needs of the culture’; and finally and most strangely for the work of two such enthusiastic espousers of the semiotic cause, the authors suggest that the overall form of television, with its contradictory and ‘de-familiarizing’ effects, operates to give the audience the ‘freedom to decode as they collectively choose’ (Fiske and Hartley, 1978, p. 193). Although, of course, it is perfectly possible to decode oppositionally in the sense of reading a television text while disagreeing with and reversing its ideological message, it is certainly not the case that the audience is free to decode as it wishes. Oppositional readings are dependent upon an accurate decoding in the first place. (Buscombe, 1979)

Fiske and Hartley clearly set out to avoid a crude and reductionist analysis of the ideology of television forms, but their own position involves certain ambiguities. The confusion between freedom to decode and freedom to read oppositionally is echoed throughout their work. Hence while accepting that television performs an ideological function at a general level, they are anxious to avoid either a conspiracy theory on the part of media professionals, on the one hand, or a view of media audiences which sees them as mindless dupes. They are compelled therefore to assume that the ideological function is a general one in which the material practices of the television industry have no part and that the ideology of television is also avoidable by collective aberrant decoding on the part of the audience. Their own readings, if somewhat erratically, stress the ideological meanings of television programmes but they attempt to use these readings in a pluralist theoretical framework which stresses the universal character of the ideologies involved (ideology answers ‘cultural needs’) and glosses over specific ideological forms. This sometimes seems to lead them into the worst forms of the reductionism they sought to avoid. Their reading of ‘The Sweeney’, for example, focuses on the relationship between Carter and Regan, comparing them with their West Coast counterparts, ‘Starsky and Hutch’. At the same time, their view of cultural needs is drawn in, and on the basis of the way in which Regan and Carter work together with Regan dramatically privileged and in a higher position in the police hierarchy, they suggest that The Sweeney’ tells us that ‘in a period when real life offers us wage restraint, inflation and a fall in living standards, there is no need for class hostility’ (Fiske and Hartley, 1978, p. 188). Although Regan and Carter are in different positions in the police hierarchy, it seems curious to assume that the Regan and Carter relationship represents class relationships when both men share within the series the same class position and both articulate populist resentments against a system which inextricably entangles class and crime, against them. It is one thing to suggest that The Sweeney’, like ‘Starsky and Hutch’ and ‘Ironside’ operates to personalize status relationships: quite another, to suggest that ‘The Sweeney’ ‘presents a society where class divisions are overcome because both “classes”—Regan and Carter—share the same outlook on life, methods and language’. This kind of dubious leap tells us little about the specific ideological message of The Sweeney’ and assumes a relationship of reflection between television and society. The Sweeney's mythology of defensive determination, we are told, is peculiarly appropriate for a society in a period of recession.

Since ideologies operate in a manner generally concomitant with the needs of the culture and since audiences are free to decode as they will in the Fiske and Hartley formulation, there is little need to examine specific developments and changes within ideological and televisual discourses or the relationship between mass media texts and systems of production or the inter-connections between the media, the state and the class system. The production of ‘readings’ becomes an end in itself, an exercise in establishing different interpretations in a manner not dissimilar to certain traditional forms of literary criticism, although without the search for excellence which normally preoccupies those forms.

Marxist negotiations with and appropriations of semiology as a linguistic paradigm have taken different directions in the sense that semiology has been articulated with an existing and a developing theory of ideology. For Marx, ideology constituted a specific part of his theory about the nature and internal dynamics of capitalist society. Marx never wrote systematically about ideology and culture but nevertheless a theory of ideology is contained within his work and scattered throughout his work are a series of programmatic outlines. Marx's concept of ideology rested on a substructure/superstructure model which is clearly set out in the much quoted passage from the Preface to a Contribution to the critique of political economy.

In the social production of their life, men enter into definite relations which are indispensible and independent of their will, relations of production which correspond to a definite stage of development of their material productive forces. The sum total of these relations of production constitutes the economic structure of society, the real foundation on which rises a legal and political superstructure and to which correspond definite forms of social consciousness. The mode of production of material life conditions the social, political and intellectual life process in general. It is not the consciousness of men that determines their being but, on the contrary, their social being that determines their consciousness. At a certain stage of their development, the material productive forces of society come in conflict with the existing relations of production, or—what is but a legal expression for the same thing, with the property relations within which they have been at work hitherto. From forms of development of the productive forces these relations turn into their fetters. Then begins an epoch of social revolution. With the change of the economic foundation the entire immense superstructure is more or less rapidly transformed. In considering such transformations a distinction should always be made between the material transformation of the economic conditions of production, which can be determined with the precision of natural science, and the legal, political, religious, aesthetic or philosophic—in short ideological forms in which men become conscious of this conflict and fight it out. Just as our opinion of an individual is not based on what he thinks of himself, so can we not judge of such a period of transformation by its own consciousness; on the contrary, this consciousness must be explained rather from the contradictions of material life, from the existing conflict between the social productive forces and the relations of production. (Marx and Engels, 1962, pp. 262–3)

This passage has often been read as an economically determinist view of ideology in which both the ‘ideological forms’ and the ‘consciousness of men’ are moulded by the economic substructure. This has been the justification for the focus in Marxism on the problem of transforming the capitalist infrastructure. Yet, Marx was well aware that the superstructural forms—the organization of the state, religion, etc.—could exert considerable influence on the course of events and his empirical work often points to the relative autonomy of these areas of society in specific historical circumstances. A great deal would seem to hinge on the use of the word ‘determines’. The term ‘determination’ can be used to suggest rather different forms of relationship. Williams has argued that it is possible for the term to indicate either a process ‘of setting limits and exerting pressures’ or a quite different process in which ‘subsequent content is essentially prefigured, predicted and controlled by a pre-existing external force’, and that Marx uses the term in the former sense (Williams, 1973, p. 4). Marx certainly uses the language of determinism but it is worth noting that he was writing in opposition to idealist and theological accounts of the world, in which the language of determinism was the expected form. It is noticeable that it is in statements that reverse received propositions that Marx uses the word ‘determines’ most forcibly as in: ‘It is not the consciousness of men that determines their being but rather their social being which determines their consciousness’ (Marx and Engels, 1962, p. 363).

Debates within Marxism have consistently revolved around the problems associated with economic determinism, although Williams accurately identifies and develops the elements within Marxism which militate against a crude determinism. Marx certainly emphasizes the necessity both for specific historical analysis and for viewing capitalist society as a totality in which the tendencies of the determining base are mediated at other levels. Some formulations, however, have, in stressing the class basis of ideology, lent themselves to various forms of reductionism:

The ideas of the ruling class are in every epoch the ruling ideas: i.e. the class which is the ruling material force in society is at the same time its ruling intellectual force. The class which has the means of material production at its disposal has control at the same time over the means of mental production so that thereby, generally speaking, the ideas of those who lack the means of mental production are subject to it. The ruling ideas are nothing more than the ideal expression of the dominant material relationships grasped as ideas: hence of the relationships which make the one class the ruling one, therefore, the ideas of its dominance. (Marx and Engels, 1970, p. 64.)

The idea of ruling-class ideology is a well-known one and has coloured a great deal of thinking about the mass media. Traditional Marxist accounts of the media reveal two important characteristics in the conceptualization of the media. Firstly, ideology is conceived as ‘false consciousness’. The work of members of the Frankfurt School, for example, gives to the mass media and the culture industry a role of ideological dominance which destroys both bourgeois individualism and the revolutionary potential of the working class. Secondly, the base/superstructure model applied to the mass media generated a continuing concern with the ownership and control of the mass media which gives the signifying capacity of the media a second place, an essentially reflective place, within its theorizing.

Structuralist or semiological investigations of the media allowed a temporary, or in some cases permanent, suspension of involvement in these problems of determination associated with the base/superstructure metaphor. The representation of the media within Marxism as, on the one hand, a purveyor of ruling-class ideology and on the other hand, the inculcator of false consciousness was, in any case, threatened by the Althusserian reformulations of a theory of ideology. Although the Althusserian view retained the notion of determination in the last instance by the economic, it also stressed the autonomy and materiality of the ideological and effected a decisive break with ideological reductionism of an economist and reflective nature. The Althusserian conceptualization of ideology as the themes, concepts and representations through which men and women ‘live’ in an imaginary relation, their relation to their real conditions of existence also involved a shift away from problems of determination in favour of articulation between the parts in a structure in the focus on the terrains, apparatuses and practices of ideology. In this theoretical context, there have been a number of efforts to combine and synthesize Marxist studies with a semiological paradigm.

The project of the periodical Screen, in attempting to generate the theoretical basis for film and television studies in the education system, has involved just such efforts. In particular there has been continuous attention within Screen to the dominant codes of narrative cinema, one focus of which has been the contention that such codes are ‘realist’ and that this form of realism has to be critically engaged with in order to understand the ideological character of the cinema and in order to effect changes within it. Of course, the ‘window on reality’ effect of photography, film and television has become almost a commonplace of media research. By the early 1970s, there was a general recognition of the inadequacies of a conceptualization of the media which stressed its neutral and reflective role. The arguments in Screen surrounding MacCabe's identification of a ‘classic realist text’ had a rather different focus. MacCabe argued that the ‘realism’ of the cinema is tied to the characteristics of a particular type of literary production—that of the nineteenth-century realist novel. MacCabe defined the ‘classic realist text’ as one in which there is ‘a hierarchy among the discourses which compose the text and this hierarchy is defined in terms of an empirical notion of truth’ (MacCabe, 1974, p. 10). The essential features of the ‘classic realist text’ MacCabe proposed, were, firstly, its inability to deal with the real as contradictory and secondly its positioning of the subject in a relationship of ‘dominant specularity’. The dominant discourse in a classic realist text effects a closure of the subordinated discourses and the reader is placed in a position ‘from which everything becomes obvious’. This is achieved through the effacement of the text's signifying practice, through the concealment of its construction. MacCabe used the notion of the ‘classic realist text’ in an illustrative example of the analysis of Pakula's film, Klute. The dominant discourse in Klute, according to MacCabe, is the unfolding of the narrative as reality revealing itself. Against this can be measured other subordinate discourses, notably the subjective account provided by the heroine, Bree, talking to her psychiatrist, in a series of fragmentary scenes throughout the film. This subjective discourse in which Bree talks about her desire for independence is seen to be illusionary in relation to the dominant discourse.

The final scene [it is suggested] is particularly telling in this respect. While Klute and Bree pack their bags to leave, the soundtrack records Bree at her last meeting with her psychiatrist. Her own estimation of the situation is that it most probably won't work but the reality of the image ensures us that this is the way it will really be. (MacCabe, 1974, p. 10)

This analysis allowed MacCabe to dispute contemporary critical accounts of the film which stressed the realistic and liberated character of the heroine, played by Jane Fonda. Rather MacCabe contended that the hero, Klute, the detective, played by Donald Sutherland, is privileged within the narrative as a character whose discourse is also a discourse of knowledge. As a man and a detective, he both solves the problems of his friend's disappearance and comes to know the truth about Bree, thereby guaranteeing that the essential woman can only be defined and known by a man. Moreover, this possession of knowledge is also shared by the reader of the film as the narrative unfolds: ‘if a progression towards knowledge is what marks Bree, it is possession of knowledge which marks the narrative, the reader of the film and John Klute himself’ (MacCabe, 1974, p. 11).

The linguistic paradigm, the form of ‘immanent’ analysis familiar to us from earlier examples, is clearly present in this type of reading. What distinguishes MacCabe's argument, however, is the setting up of the category of ‘classic realism’ as the dominant mode of film and television production and endowing that category with certain essential ideological characteristics. MacCabe does not suggest that classic realist texts cannot be progressive but he does argue that such texts can only be progressive in so far as they espouse an ideological or political position which is at odds with the status quo. Realist texts remain unprogressive in their form in the sense that realist texts always interpellate or pull in spectators as unified noncontradictory subjects in a position of dominant specularity. In Klute, for example, there is a process of identification involved in the progression of the narrative and the sequence and form of shots which positions the viewer in relation to the narrative in a position of knowledge, which makes it appear as if he or she knows reality. But this position of knowledge is created by the film rather than produced by the viewer. The classic realist text is, in MacCabe's formulation highly ‘closed’. It is for this reason that MacCabe favours, as progressive texts, certain avant garde films in which there is no dominant discourse but on which the reader has to work and produce a meaning for the film.

There are a number of problems with the notion of the ‘classic realist text’ not least of which is the extent to which films and television programmes conform to and effectively realize a classic realist project. The rather general nature of the category also raises difficulties. The idea of the classic realist text has a tendency to conceal as much as it illuminates inasmuch as it becomes difficult to distinguish between a nineteenth-century novel and a Hollywood movie or between different groups of Hollywood films. At the same time, the thesis of the ‘classic realist text’ and the subsequent debate around the term, did have the considerable merit of bringing to the forefront of discussion the formal and ideological characteristics of film and television. It is worth remembering that the implicit modes of pluralist mass communications research against which Screen and MacCabe wrote conceived of the media as transparent and neutral communicators and that early semiological inquiries focused on individual texts and general categories of ideology. MacCabe's argument in its suggestion that texts embodied even in terms of their formal characteristics a political signification moved beyond a view of the media as passive transmitters and beyond the reading of single texts. It also undoubtedly led to an élitist concern with avant garde texts and with texts which reflected upon themselves.

The Screen arguments around realism also involved an explicit rejection of traditional Marxist views of the media as reflective. MacCabe makes it clear that his own work ‘does not understand cinema to have an ideological function determined by its representational relationship to other ideological, political and economic struggles’ (MacCabe, 1978, p. 32). The theory of ideology which lies behind this takes as its central conceptions the notions of discourse and the subject. The idea of discourse focuses attention on the internal characteristics and processes of signifying systems. Relationships between discourses are conceived of in terms of articulation rather than determination. This use of the linguistic paradigm would replace the operations of the base/superstructure metaphor and in an extreme form suggest that there is a necessary non-correspondence of all practices. A central concern of this theory of ideology has been the development of theorizing the neglected area of the subject and subjectivity, using Lacanian psychoanalytic concepts to indicate how subjects are constituted in language and other discourses as a non-unified and contradictory set of positions.

The debate around realism and the analysis of realist texts moved the conceptualization of ideology closer to a linguistic or structuralist model of society. There are advantages here in terms of the internal coherence of the conceptual apparatus employed and in the space provided for the concrete analysis of particular ideological and discursive formations. However, there have also been attempts to register the autonomy of discursive practices and signifying systems within a Marxist framework. Policing the crisis, for example, represents a formidably ambitious attempt to reconcile a reworking of Gramsci's theory of hegemony with an analysis of the signifying practices of the media in an account of ‘a crisis in hegemony’ in post-war Britain (Hall et al., 1978). The authors attempt to map out the shifting ideological configurations of the period, characterizing them as culminating in a crisis in hegemony. The study is not confined to an analysis of the ideological superstructures but involves tracing ‘the “passage” of a crisis from its material base in productive life through to the complex spheres of the superstructures’ (Gramsci, quoted in Hall, 1978). Beginning from the orchestration by the media of mugging as a ‘moral panic’, the writers attempt to establish that the mugging panic represents a movement from a ‘consensual’ to a more ‘coercive’ management of the class struggle which in itself stems from the declining international competitiveness of the British economy following the post-war boom. The analysis suggests that there was a form of hegemonic equilibrium in the immediate post-war period, the erosion of which led to attempts to secure ‘consent’ by more coercive although ‘legitimate’ means. The immediate post-war period saw the construction of a consensus as the condition for the stabilization of capitalism in the circumstances of the Cold War and this provided a period of extensive hegemony in the 1950s. Economic decline triggered the disintegration of this ‘miracle of spontaneous consent’ and there was an attempt to put forward a ‘Labourist’ variant of consent to replace it. The exhaustion of this form of consent, however, combined with the rise of social and political conflict, the deepening of the economic crisis and the resumption of a more explicit class struggle culminated in the ‘exceptional’ form of class domination in the 1970s through the state (Hall et al, 1978, p. 218). (There is a further discussion of Policing the Crisis in chapter II, pp. 30–55.)

The media appear to play a particularly important part in this analysis. They are described as ‘a key terrain where “consent” is won or lost’, ‘a field of ideological struggle’ (Hall et al., 1978, p. 220). The media are also the focal point for the authors’ conception of the autonomy of the superstructure for, while rejecting the idea of a ‘set of monolithic interpretations systematically generated by the ruling class for the explicit purpose of fooling the public’, Policing the Crisis does contend that the media serve to reproduce—although through their own ‘constructions and inflections’—‘the interpretations of the crisis subscribed to by the rulingclass alliance’ (Hall et al., 1978, p. 220). The crisis has its basis in changes in the economy. Although avoiding a heavily determinist stance and relying more on a culturalist view of determination, Policing the Crisis retains a hierarchy of determinations, while at the same time seeking to establish the specificity and relative autonomy of the media signification system.

The key to the media's involvement in the construction of consent lies in the authors’ analysis of news as performing a crucial transformative but secondary role in defining social events. The primary definers are those to whom the media turn, their accredited sources in government and other institutions. Although Policing the Crisis emphasizes the transformative nature of media newsreporting in the selection and inflection of items and topics, the conception of the media role is one of ‘structured subordination’ to the primary definers. Further the ‘creative’ media role serves to reinforce a consensual viewpoint by using public idioms and by claiming to voice public opinion. Thus in the crisis described the media have endorsed and enforced primary definitions of industrial militancy, troublesome youth cultures, mugging, student protest movements as part of a ‘law and order’ problem pulling discrete and local events into an amplification spiral and registering them all within a discourse of ‘law and order’.

The thesis put forward in Policing the Crisis raises certain problems in relation to the siting of signifying systems within a Marxist theory of ideology. It is clear that the autonomy of media significations within the argument is very limited. Basically, the media serve, in the specific historical conditions analysed, to reproduce and reinforce ‘primary definitions’. They are assumed thereby to signify a crisis which already exists for the primary definers, a crisis already in operation in the realm of politics and economics. Moreover, given this view of the operations of the media, it is difficult to see how the media operate as ‘a field of ideological struggle’. Since the news is read as ‘the media’ and the news is characterized by its reproduction of primary, ‘dominant’ definitions in a consensual form, struggle, along with those primary definitions, would seem to lie outside this area of media signification. The area of ‘struggle’ or opposition would seem to lie in Policing the Crisis, insofar as it lies anywhere, in the areas of class experience and the cultural forms through which men and women live that experience; but those cultural forms are largely neglected in favour of the focus on ‘news’. Some of the difficulties present in Policing the Crisis undoubtedly stem from the attempted synthesis of this form of Marxist culturalist theory, inflected through Gramsci, with an Althusserian conception of the media as an ideological state apparatus largely concerned with the reproduction of dominant ideologies and with an attempt to recognize the autonomy and specificity of the media. With this kind of multiple ‘grafting’ going on, it is not entirely surprising that some shoots do not flourish. In this case, attention to the internally ordered characteristics of the media suffers, since the media is conceived of as representing ‘reality’ in a manner inflected in the interests of dominant groups. In effect a sophisticated version of the notion of ‘false consciousness’ is proposed; ‘by consenting to the view of the crisis which has won credibility in the echelons of power, popular consciousness is also won to support too the measures of control and containment which this version of social reality entails’ (Hall et al, 1978, p. 221).

Semiology or structuralism and in particular the semiological analysis of media texts have been woven into various formulations of a theory of ideology with a range of subsequent problems in the internal coherence of such theories. It is through the endless thinking through of this kind of incoherence, that intellectual work proceeds. The problems raised in the texts discussed here indicate the general difficulty of reconciling semiotics with any theory of ideology which conceives of the media as essentially reflecting the ‘real’. Yet the treatment of systems of signification as autonomous, not bound in a relationship of reflection or representation to an external reality, does not exclude relationships of articulation between different forms of signification nor does it necessarily exclude the analysis of the determinations of signifying systems. Indeed, an effective theory of signification would necessarily involve examining the overall pattern of signifying systems and the configuration of ideological practices. In this sense, Policing the Crisis, with all its problems, suggests the theoretical ambitions which a materialist theory of signification and ideology should have.

REFERENCES

Adorno, T.W. and Horkheimer, M. (1977) ‘The culture industry: enlightenment as mass deception’, in Curran, J. et al. Mass Communication and Society, London, Edward Arnold.

Barthes, R. (1967) Elements of semiology, London, Jonathan Cape.

Barthes, R. (1971) ‘The rhetoric of the image’, Working Papers in Cultural Studies, Spring.

Barthes, R. (1972) Mythologies, London, Jonathan Cape.

Benjamin, W. (1977) ‘The work of art in an age of mechanical reproduction’, in Curran, J. et al. Mass Communication and Society, London, Edward Arnold.

Buono, E.del and Eco, U. (eds) (1966) The Bond Affair, London, Macdonald.

Burgelin, O. (1972) ‘Structuralist analysis and mass communication’, in McQuail, D. (ed) (1972) The Sociology of mass communications, Harmondsworth, Penguin.

Buscombe, E. (1978/9) ‘Unscrambling semiotics’, Screen Education, (29).

Camargo, M.de (1972) ‘Ideological analysis of the message: a bibliography’, Working Papers in Cultural Studies, 3, Autumn.

Cantril, H. (1966) The Invasion from Mars, New York, Harper Torchbooks.

Curran, J., Gurevitch, M. and Woollacott, J. (eds) (1977) Mass Communication and Society, London, Edward Arnold.

Eco, U. (1966) ‘Narrative structure in Fleming’, in Buono, E.del and Eco U. (eds) The Bond Affair, London, Macdonald.

Eco, U. (1971) A Estrutura Ausente, S.Paolo, Perspectiva.

Fiske, J. and Hartley, J. (1978) Reading Television, London, Methuen.

Hall, S. (1972) ‘The determination of news photographs’, Working Papers in Cultural Studies, 3.

Hall, S. (1973) ‘Encoding and decoding in the television discourse’, CCS occasional paper.

Hall, S., Critcher, C, Jefferson, T., Clarke, J. and Roberts, B. (1978) Policing the crisis: Mugging, the State and Law and Order, London, Macmillan.

Jameson, F. (1972) The Prison-House of Language, New Jersey, Princeton University Press.

Lévi-Strauss, C. (1970) The Elementary Structures of Kinship, London, Social Science Paperbacks in association with Eyre & Spottiswoode.

MacCabe, C. (1974) ‘Realism and the cinema’, Screen, 15 (2).

MacCabe, C. (1978) ‘Discourse, cinema and politics’, Screen, 19 (4).

McQuail, D. (1972) The Sociology of Mass Communications, Harmondsworth, Penguin.

Marx, K. and Engels, F. (1962) Selected Works, London, Lawrence & Wishart.

Marx, K. and Engels, F. (1970) The German Ideology, London, Lawrence and Wishart.

Panofsky, E. (1934) ‘Style and medium in the moving pictures’, reproduced in Talbot, D. (1967) Film, An Anthology, Berkeley, Cal., University of California Press.

Propp, V. (1968) The Morphology of the Folktale, Austin, Texas, and London, University of Texas Press.

Saussure, F. de (1966) Course in General Linguistics, New York, McGraw-Hill.

Williams, R. (1973) ‘Base and superstructure’, New Left Review, 82.