Structuralist semioticians tend to focus on the internal structure of the text rather than on the processes involved in its construction or interpretation. Where those working within this tradition do theorize beyond the text, they tend to argue that communication (particularly mass communication) is a primary process of reality construction and maintenance whereby positions of inequality, dominance and subservience are produced and reproduced in society and at the same time made to appear 'natural'. The 'New Critics', W K Wimsatt and M C Beardsley, whilst not structuralists, advanced the formalist argument that meaning lay within the text and defined as 'the affective fallacy' the notion that the meaning of a poem depended on the 'subjective' responses of the reader, which they saw as 'a confusion between the poem and its results (what it is and what it does)' (Wimsatt & Beardsley 1954, 21). Such accounts tend towards 'textual determinism', assuming that texts are invariably read much as was intended by their makers, leaving little scope either for contradictions within and between texts or for variations amongst their interpreters. Monolithic theories of this kind ignore what Saussure had referred to as 'the role of signs as part of social life' (Saussure 1983, 15; Saussure 1974, 16).
Contemporary semioticians refer to the creation and interpretation of texts as 'encoding' and 'decoding' respectively. This unfortunately tends to make these processes sound too programmatic: the use of these terms is of course intended to emphasize the importance of the semiotic codes involved, and thus to highlight social factors. For semioticians, there is no such thing as an uncoded message, so that - for those who argue that all experience is coded - even 'encoding' might be more accurately described as 'recoding' (Hawkes 1977, 104, 106, 107). In the context of semiotics, 'decoding' involves not simply basic recognition and comprehension of what a text 'says' but also the interpretation and evaluation of its meaning with reference to relevant codes. Where a distinction is made between comprehension and interpretation this tends to be primarily with reference to purely verbal text, but even in this context such a distinction is untenable; what is 'meant' is invariably more than what is 'said' (Smith 1988, Olson 1994). Everyday references to communication are based on a 'transmission' model in which a sender transmits a message to a receiver - a formula which reduces meaning to 'content' which is delivered like a parcel (Reddy 1979). This is the basis of Shannon and Weaver's well-known model of communication, which makes no allowance for the importance of social contexts and codes (Shannon and Weaver 1949).
Whilst Saussure's model of oral communication is (for its time) innovatingly labelled as a 'speech circuit' and includes directional arrows indicating the involvement of both participants (thus at least implying 'feedback'), it too was nevertheless a linear transmission model (albeit a 'two-track' one). It was based on the notion that comprehension on the part of the listener is a kind of mirror of the speaker's initial process of expressing a thought (Saussure 1983, 11-13; Saussure 1974, 11-13; Harris 1987, 22-25, 204-218). In this model there is only the briefest of allusions to the speaker's use of 'the code provided by the language', together with the implicit assumption that a fixed code is shared (Saussure 1983, 14; Saussure 1974, 14; Harris 1987, 216, 230).
In 1960 another structural linguist - Roman Jakobson (drawing on work by Bühler dating from the 1930s) - proposed a model of interpersonal verbal communication which moved beyond the basic transmission model of communication and highlighted the importance of the codes and social contexts involved (Jakobson 1960). He noted elsewhere that 'the efficiency of a speech event demands the use of a common code by its participants' (Jakobson & Halle 1956, 72). He outlines what he regards as the six 'constitutive factors... in any act of verbal communication' thus:
Jakobson proposed that 'each of these six factors determines a different function of language' (ibid.):
||expressing feelings or attitudes
||It's bloody pissing down again!
||Wait here till it stops raining!
||establishing or maintaining social relationships
||Nasty weather again, isn't it?
||referring to the nature of the interaction (e.g. genre)
||This is the weather forecast.
||foregrounding textual features
||It droppeth as the gentle rain from heaven.
This model avoids the reduction of language to 'communication'. Referential content is not always foregrounded. Jakobson argued that in any given situation one of these factors is 'dominant', and that this dominant function influences the general character of the 'message'. For instance, the poetic function (which is intended to refer to any creative use of language rather than simply to poetry) highlights 'the palpability of signs', undermining any sense of a 'natural' or 'transparent' connection between a signifier and a referent. Jakobson's model demonstrates that messages and meanings cannot be isolated from such constitutive contextual factors. In its acknowledgement of social functions this is a model which is consonant with the structuralist theory that the subject (here in the form of the 'addresser' and the 'addressee') is constructed through discourse.
Whilst these earlier models had been concerned with interpersonal communication, in an essay on 'Encoding/Decoding' (Hall 1980, originally published as 'Encoding and Decoding in Television Discourse' in 1973), the British sociologist Stuart Hall proposed a model of mass communication which highlighted the importance of active interpretation within relevant codes. Justin Wren-Lewis insists that Hall's model, with its emphasis on coding and decoding as signifying practices, is 'above all, a semiological conception' (Wren-Lewis 1983, 179). Hall rejected textual determinism, noting that 'decodings do not follow inevitably from encodings' (Hall 1980, 136). In contrast to the earlier models, Hall thus gave a significant role to the 'decoder' as well as to the 'encoder'.
Hall referred to various phases in the Encoding/Decoding model of communication as moments, a term which many other commentators have subsequently employed (frequently without explanation). John Corner offers his own definitions:
Hall himself referred to several 'linked but distinctive moments - production, circulation, distribution/consumption, reproduction' (Hall 1980, 128) as part of the 'circuit of communication' (a term which clearly signals the legacy of Saussure). Corner adds that the moment of encoding and that of decoding 'are socially contingent practices which may be in a greater or lesser degree of alignment in relation to each other but which are certainly not to be thought of... as 'sending' and 'receiving' linked by the conveyance of a 'message' which is the exclusive vehicle of meaning' (Corner 1983, pp. 267-8).
Mass media codes offer their readers social identities which some may adopt as their own. But readers do not necessarily accept such codes. Where those involved in communicating do not share common codes and social positions, decodings are likely to be different from the encoder's intended meaning. Umberto Eco uses the term 'aberrant decoding' to refer to a text which has been decoded by means of a different code from that used to encode it (Eco 1965). Eco describes as 'closed' those texts which show a strong tendency to encourage a particular interpretation - in contrast to more 'open' texts (Eco 1981). He argues that mass media texts tend to be 'closed texts', and because they are broadcast to heterogeneous audiences diverse decodings of such texts are unavoidable.
Stuart Hall stressed the role of social positioning in the interpretation of mass media texts by different social groups. In a model deriving from Frank Parkin's 'meaning systems', Hall suggested three hypothetical interpretative codes or positions for the reader of a text (Parkin 1972; Hall 1973; Hall 1980, 136-8; Morley 1980, 20-21, 134-7; Morley 1981b, 51; Morley 1983, 109-10):
This framework is based on the assumption that the latent meaning of the text is encoded in the dominant code. This is a stance which tends to reify the medium and to downplay conflicting tendencies within texts. Also, some critics have raised the question of how a 'preferred reading' can be established. Shaun Moores asks 'Where is it and how do we know if we've found it? Can we be sure we didn't put it there ourselves while we were looking? And can it be found by examining any sort of text?' (Moores 1993, 28). Some theorists feel that the concept may be applied more easily to news and current affairs than to other mass media genres. David Morley wondered whether it might be the 'reading which the analyst is predicting that most members of the audience will produce' (Morley 1981a, 6). John Corner argues that it is not easy to find actual examples of media texts in which one reading is preferred within a plurality of possible readings (Corner 1983, 279). As Justin Wren-Lewis comments, 'the fact that many decoders will come up with the same reading does not make that meaning an essential part of the text' (Wren-Lewis 1983, 184). And Kathy Myers notes, in the spirit of a post-structuralist social semiotics, that 'it can be misleading to search for the determinations of a preferred reading solely within the form and structure' of the text (Myers 1983, 216). Furthermore, in the context of advertising, she adds that:
Just as a reductive reading of Hall's model could lead to the reification of a medium or genre, it could also encourage the essentialising of readers (e.g. as 'the resistant reader') whereas reading positions are 'multiform, fissured, schizophrenic, unevenly developed, culturally, discursively and politically discontinuous, forming part of a shifting realm of ramifying differences and contradictions' (Stam 2000, 233).
Despite the various criticisms, Hall's model has been very influential, particularly amongst British theorists. David Morley employed it in his studies of how different social groups interpreted a television programme (Morley 1980). Morley insisted that he did not take a social determinist position in which individual 'decodings' of a text are reduced to a direct consequence of social class position. 'It is always a question of how social position, as it is articulated through particular discourses, produces specific kinds of readings or decodings. These readings can then be seen to be patterned by the way in which the structure of access to different discourses is determined by social position' (Morley 1983, 113; cf. Morley 1992, 89-90). Morley's point about differential access to discourses can be related to to the various kinds of 'capital' outlined by Pierre Bourdieu - notably 'cultural capital' (to which Bourdieu relates the construction of 'taste') and 'symbolic capital' (communicative repertoire). An 'interpretative repertoire' (Jonathan Potter, cited in Grayson 1998, 40) is part of the symbolic capital of members of the relevant 'interpretative community' and constitutes the textual and interpretative codes available to them (which offer them the potential to understand and sometimes also to produce texts which employ them). Morley added that any individual or group might operate different decoding strategies in relation to different topics and different contexts. A person might make 'oppositional' readings of the same material in one context and 'dominant' readings in other contexts (Morley 1981a, 9; Morley 1981b, 66, 67; Morley 1992, 135). He noted that in interpreting viewers' readings of mass media texts attention should be paid not only to the issue of agreement (acceptance/rejection) but to comprehension, relevance and enjoyment (Morley 1981a, 10; Morley 1992, 126-7, 136).
The most basic task of interpretation involves the identification of what a sign represents (denotation) and may require some degree of familiarity with the medium and the representational codes involved. This is particularly obvious in the case of language, but may also apply in the case of visual media such as photographs and films. Some would not grant this low-level process the label of 'interpretation' at all, limiting this term to such processes as the extraction of a 'moral' from a narrative text. However, David Mick and Laura Politi take the stance that comprehension and interpretation are inseparable, making an analogy with denotation and connotation (Mick & Politi 1989, 85).
Justin Wren-Lewis comments that 'given the wealth of material using semiological tools for the analysis of film and television, it is remarkable that so little work has been done on the practice of decoding' (Wren-Lewis 1983, 195). Whilst social semiotics stakes a claim to the study of situated semiotic practices, research in this area is dominated by ethnographic and phenomenological methodologies and is seldom closely allied to semiotic perspectives (though there is no necessary incompatibility). A notable exception is the research of David Mick in the field of advertising (Mick & Politi 1989, McQuarrie & Mick 1992, Mick & Buhl 1992).