The prefix 'post' suggests that 'structuralism' has now been supplanted by a new theory: indeed it has been confidently asserted that Derrida had 'brought the structuralist movement to an end' by his work on deconstruction in the late 1960's and early 1970's. From this perspective, the concepts 'structuralism' and 'post-structuralism' take on a relationship of binary opposition in which the latter term is privileged: the outmoded 'structuralism' has been replaced by the new, improved 'post-structuralism'. Apart from the fact that such binary oppositions are anathema to post-structuralists, it is in fact somewhat misleading to claim that a radical break took place and that the earlier phase was thereby invalidated.
Developments certainly occurred from within the original structuralist position and divergent tendencies gradually arose, but these were in part continuous or re-appraisals of lines of thought already inherent in earlier stages. As Derrida notes, `we are still inside structuralism in so far as structuralism constitutes an adventure of vision, a conversion in the way of putting questions to any object'. Furthermore, writers such as Barthes do not fit neatly into a single category (and would not wish to) and others (Lacan, Foucault) may be described as structuralist in one text and post-structuralist in another (Sturrock classes them as the former, Selden and Lodge the latter).
It is possible, however, to identify certain major differences between the two approaches: namely, where structuralism sought to establish a science or poetics of literature (or cultural signifying practices as a whole), post-structuralist thought, following Derrida's critique of the metaphysics of presence, has taken an anti-scientific stance and, pursuing the infinite play of signifiers, has resisted the imposition of any organising system.
In addition, a range of post-structuralist approaches are a synthesis of deconstruction and other theories derived from Marxism, feminism or psychoanalysis which produce a more historically and socially orientated critique of the text than was the case with the more ahistorical forms of structuralism. These latter developments contrast with the ostensibly apolitical brand of post-structuralism called `deconstruction', largely practised in the USA, which does not relate literary criticism to wider social concerns any more than did New Criticism.
A further tendency discernible in the later phase of structuralism and in post-structuralism is that the onus is increasingly placed on the reader or critic to produce meanings, rather than solely on the text itself.
Because post-structuralism chiefly evolved out of a critique of particular structuralist assumptions, it is first necessary to outline their shared foundation in Saussurean linguistics. This is where a central post-structuralist development occurs which departs from the initial structuralist position.
It was Saussure's linguistic theories, in particular his concepts of the bipartite linguistic sign, its arbitrary relationship to reality, and the diacritical nature of language which have specific relevance for all subsequent variations of structuralist and post-structuralist theory. According to Saussure, language is a system of signs, each of which consists of a signifier (sound image or written word) and a signified (the concept evoked by the signifier). Referents (actual entities) form no part of this relationship: the signified is not a thing but the mental concept of one, and the relationship between the sign and its referent is completely arbitrary, as is the connection between signifier and signified. The link between the sound image/word `cup' and the concept of a cup is a conventional (not a `natural') one. It is language which articulates the two continua of `jumbled ideas' and `vague sounds' to link signifier and signified, forming the units of meaning we term words. The signifier/sound image `cup' has meaning only in that it is phonetically distinguishable from `cap', `cut', cop', and so on; the signified `cup' depends on its semantic difference from related terms such as `beaker', `wineglass', `mug', `tankard' etc to produce its meaning. It is in this respect that language is said to be diacritical: it depends on a structured system of differences for its meaning.
This differential system organises all aspects of language in various relationships. The sequential or combinatory relationship between the three phonemes which comprise the sound image `cup' or that between the syntactical units of the sentence `the cup is overflowing' is termed `syntagmatic' by Saussure. Those relationships of absence which are brought into operation at the level of both signifier (the phonemes `cup' not `cut' etc) and signified (`cup' not `mug', `tankard' etc) are termed `associative' (later known as `paradigmatic'). Thus any sign can be regarded as the conjunction of a range of elements, linked to the wider system of language both by what is present and what is omitted. Because linguistic elements only acquire meaning according to their paradigmatic or syntagmatic relationships within the overall system and not as a result of a link between the sign and the referent or external reality, language is thus a closed, independent and self-sufficient structure of relations and can be studied as such.
This gives rise to Saussure's other major distinction: between `langue' (the complete system of language) and `parole' (the individual utterance which derives from it). `Langue' is the proper area of linguistic study, enabling one to identify the underlying principles by which language functions in practice.
Finally, Saussure's methodology had further implications for structuralism in that he advocated the synchronic investigation of language. During the nineteenth century, the opposite procedure had prevailed: linguistic study had been diachronic, in that the history of language was traced back through time to discover phonetic variations or etymologies. Saussure regarded this type of research as speculative: it was his contention that, while the diachronic approach should not be relinquished completely, only by adopting a synchronic mode of analysis whereby language was studied as a system of relationships functioning at a given period of time (not as it evolved) would linguistics be placed on a scientific basis.
Structuralism was founded on a similar methodological and scientific basis. It set out, following Saussure, to identify the signifying patterns, codes and conventions underlying all human cultural practices. Benveniste, for example, says structuralism constitutes its object as `a system whose parts are all united in a relationship of solidarity and dependence' and asserting the `predominance of the system over the elements', defines the structure of the system `through the relationships among the elements'. Not only language and literature, but myth, fashion or kinship systems could be examined from a synchronic, ahistorical perspective to explain their functioning, rather than by diachronically tracing their historical development.
The early aims of structuralist literary criticism were to found an analytical discourse or `metalanguage' which would operate scientifically, identifying the systems of codes and organisational principles of all literary texts to create a `second order' level of understanding. For Barthes, in his Elements of Semiology (1967), it was the discourse of semiology which could perform this metalinguistic function, so as to analyse the connotative systems of `first order' `natural' language or any other `cultural artifacts'. In the case of literature, the connotative potentiality of the first order language can be exploited by the critic who actively engages with the text to articulate one or more of its plural interpretations: in Barthes' words, the literary "work is `eternal', not because it imposes one meaning on different men, but because it suggests different meanings to one man".
However, for the majority of its early theorists, structuralism was an essentially formalist method which focused on literature's signifying structures rather than on its content. Just as Saussure emphasised that signs depend on their differential relationships with other elements in the system in order to produce meaning and not on actual entities, it therefore follows that a structuralist analysis of literature will not be concerned with the liberal-humanist view that the text expresses a `truth' about the `real world'. Investigation will centre on the literary system (equivalent to langue) as a whole, of which the individual text, (parole) is a constituent part, governed by the system's organisational principles. The author of a text and authorial intention correspondingly decline in priority: all the author's role consists of is selecting elements from the pre-existing `already written' system and producing new texts which combine these elements in different ways'.
A common factor to the approaches adopted by many structuralist theorists is their use of the fundamental signifying function of binary oppositions. Indeed these oppositional orderings are perceived by Levi-Strauss to form the basis of the `socio-logic of the human mind, which structures nature/reality in its own image.' Saussure's paired categories have already been mentioned (signifier/signified; langue/parole, etc); language itself has been shown to be structured at its most basic level, the phoneme, in this way (eg, voiced/unvoiced; nasalised/non-nasalised etc). In accordance with the belief that fundamental linguistic structures necessarily control the basic structures of literature, theorists have constructed their narrative grammars or organisational categories in terms of these binary oppositions.
Levi-Strauss' use of the `mytheme' corresponds to the phoneme of language as a basic unit of narrative organisation operating in paired opposites and Greimas divides narrative roles into three sets of binary oppositions.19 On a larger scale, Barthes' work is pervaded by distinctions such as `lisible'/`scriptible', `plaisir'/`jouissance', `écrivain'/`écrivant' etc.
Jakobson, too, has developed Saussure's original theory of syntagmatic and paradigmatic axes of combination and selection in terms of his metonymy/metaphor distinction,20 one which has proved influential for the characterisation of different modes of literature (Romanticism and Symbolism reveal "the primacy of the metaphoric process ... the predominance of metonymy underlies and actually predetermines the so-called `realistic' trend")21
It is precisely the use of binary oppositions, linguistic analogies and the total concept of the existence of a metalanguage of scientific objectivity that Derrida and his post-structuralist followers have called into question.
Barthes had anticipated a critique of the authority of his own semiological discourse as early as the 1967 Elements of Semiology: he admitted that it had the potential of becoming the `language-object' of a new metalanguage which would analyse semiology in its turn.22 Hypothetically, the process could be continued to infinity.
To return to Derrida, however; his theories are difficult to summarise or define because he consciously resist the desire to construct a unified, systematic theory of his own with which to master other discourses, but certain key concepts may be outlined in order to discuss his major areas of concern and their implications for subsequent post-structuralist developments.