The founders of the science of signs were Saussure and the American logician and mathematician Charles S. Peirce, who called it sémiologie and, respectively, semiotic.  However there are important differences between their premises and between their conclusions: Peirce’s theory relies in the first place on the methods of logic, rather than on linguistics; for him the sign is triadic, it is not based on a twofold division, as in Saussure’s view. Thus Peirce’s sign is regarded in its connection with the object (the referent) and with the interpretant (the idea begotten by the sign). Saussure was less concerned with the referent and instead emphasized the arbitrary nature of the relationships between the signifier and the signified, as well as the role of the differences between signifiers in the production of meaning.
It is difficult to establish basic differences between structuralism and semiotics, because their objects and methods are usually similar: as Jonathan Culler has pointed out, in the study of cultural products with the help of linguistics there are two fundamental insights, which yet are not divergent:
|first, that social and cultural phenomena are...objects or events with meaning, and hence signs; and second, that they...are defined by a network of relations, both internal and external. Stress may fall on one or the other of these propositions but in fact the two are inseparable. /SP 1/|
However, structuralist analysis may deal with models which are not signs proper, as for instance the kinship systems, but which can be regarded as sign structures. This explains why after the 1960s the term “structuralism” began to be gradually superseded by “semiotics” or “semiology”, a tendency which evinced the broadening of the field scope: various other branches appeared now, such as zoosemiotics, kinesics, and so forth.
A semiotic analysis of a literary text deals, instead of themes and general meaning, with the way in which meaning is produced by the structures of interdependent signs, by codes and conventions.
It is well known that from a semiotic point of view signs appear within a threefold process (semiosis), which includes syntax (the study of the relationships among signs), semantics (dealing with the relationships between signs and the objects signified), and pragmatics (dealing with the relationships between signs and their interpreters). If in the past the phonemes, morphemes or sentences were regarded as basic signs, later on Peter Hartmann viewed the text as the “initial linguistic sign”- a reference point for the other linguistic elements.
A detailed analysis of the text as sign as well as of literariness was provided by the German professor of English, Heinrich Plett, in his 1975 study Text Science and Text Analysis. Plett points out that, as a linguistic sign, the text has an arbitrary nature and is based on social conventions. In accordance to the semiotic triangle, it has: a signifier (the sounds / letters), a signified (the meaning related to it), and a referent (the reality it points to). As referents one can find abstract concepts, current, historical or imaginary events. The text is actually a “super-sign”, which includes psychological and sociological prerequisites underlying the specific communication situations. The code itself is such a prerequisite, which differs from age to age and from person to person.
The textual coherence may be ensured most often by syntactic factors (such as anaphoric and/or cataphoric elements, connectives, the sentence order, and so on). When these are missing, the pragmatic coherence may take their place, as in Ezra Pound’s short poem In a Station of the Metro:
The apparition of these faces in the crowd;
Petals on a wet black bough.
It is the reader’s role to constitute these lines as a text, by supplying the missing connections and establishing a direction toward meaning.
As for semantic coherence, it is dependent on the compatibility of the referential features within the text. If the latter are missing, there may appear a “coreferential ambiguity”: this may be a flaw of the text itself (caused for instance by pronouns used ambiguously), or, conversely, an intentional feature (as in difficult poems, for example).
The literariness of the text is much harder to define, even if the tools and methods of semiotics are employed. Plett admitted that semiotic aesthetics was in a burgeoning stage, and as long as it was not constituted as a definitive system, the only solution to attempt at a definition of literariness was to establish “an aesthetic threshold”, with the help of the DEVIATION concept. Deviation is akin to the formalists’ idea of defamiliarization, but actually it is as early as the 16th century that a poetician, George Puttenham, asserted that figurative speech, specific to literature, was estranged from the ordinary manner of talking and writing. There are four proposals for defining aesthetic deviation that Plett examines in detail: two are qualitative, and two, quantitative: non-grammaticality, equivalence (i.e. repetition, resemblance, analogy, redundancy), and, on the other hand, occurrence (i.e. statistically infrequent appearance) and recurrence (statistically frequent appearance). The first two definitions, based on the principle of grammar, are regarded as more significant by the German author: if they are considered together, they represent a solid starting point for the analysis of the literary function of language. Considering its perspective, Plett’s model of analysis can be defined as a rhetorical-stylistic approach to literariness.
He does not use the concept of deviation to refer to all the elements making up the poetic medium (the sequence of incidents, conventional practices, and so forth), like previous theorists, but he restricts it to language as such. This is the first level of reductionism which he admits to have been forced to use (by excluding the psychological, sociological components, and others). The second one consists in leaving aside the semantic level of semiosis, because of its referential complexity, as well as the pragmatic level, impossible to use in a single study, owing to the multitude of extralinguistic factors involved.
So Plett’s analytical system is based on two fundamental elements: the linguistic deviation and the linguistic unit. As a breach of conventional norms, deviation may involve four kinds of aesthetic alterations: the addition, the subtraction, the permutation, and the substitution of signs. The fifth type, the equivalence of signs, represents a reinforcement of the rule. As concerns the linguistic units, Plett mentions the phoneme, the morpheme, the sentence, and then he adds the grapheme and the sememe; accordingly there are five kinds of deviations: phonological, morphological, syntactic, graphemic and semantic.
The complex structure of Plett’s rhetorical system is therefore made up of a web of aesthetic alterations based on the combinations between the linguistic units and the linguistic deviations.
As a sample we will present here the chart of morphological figures drawn out by Plett. They are ranked in this order:
1. Addition metamorphs, with several subtypes, of which examples are: Sampson-syrup-gold-maned (D. Thomas); the unchilding unfathering deep (G. M. Hopkins); fishnetzveil (J. Joyce).
2. Subtraction metamorphs: 2.1 clippings: 2.1.1 common ones: ad, pram; 2.1.2 unique ones: the achieve (Hopkins); 2.2 portmanteau words.
3. Permutation metamorphs: upjump (Joyce); how dearly ever parted (Shakespeare).
4. Substitution metamorphs: almonthst; prapsposterous (Joyce)
5. Equivalence metamorphs: repetition, polyptoton (And singing still dost soar, and soaring ever singest - Shelley), paronymy (But me no buts!), plays upon homographs or homophones, historical rhymes, and so on.
In a similar way Plett classifies the other groups of figures, rounding up his system. He is aware that his perspective is twice reductionist, as we have seen, and therefore yields to the temptation of trying to present the model of a linguistic analysis of two deviations (the unrelated pronoun and the metaphor) from an unreduced semiotic perspective (referring not only to the syntactic components, but also to the semantic and the pragmatic ones). In a scene from T. S. Eliot’s play, The Family Reunion, where the protagonist, Harry, appears on the stage for the first time, the pronoun they is used without a clear referent: this deviation from the semantic textual norm is a case of hyposemy; later on, when Harry refers to the same entities using metaphorical terms, the case turns into hypersemy (or plurisignation). So Plett’s model analysis has now also taken into account the “pseudo-referential” character of this literary text, namely the “fictionalization” of the object (within the semantic perspective), as well as the reader’s creative role in “producing” the text (in other words the pragmatic perspective).
semiotic methods were used successfully during the structuralist decades in
the study of theatre, poetry and the narrative, by such authors as Patrice Pavis,
Yury Lotman, Michael Riffaterre, A. J. Greimas, and others. In the 1970s, however,
strong challenges to the structuralist thought made their way into the critical
world, and semiotic studies, as a branch of literary criticism and theory grew
less important, although general semiotics, as a transdisciplinary science,
 From the Greek word semeion (sign).
 P. Hartmann, “The Text as a Linguistic Object”, in W. D. Stempel (ed.), Contributions to the Linguistics of the Text (Munchen, 1981), 9.