This is an advance summary of a forthcoming article in the Oxford Research Encyclopedia of Linguistics. Please check back later for the full article.
The linguistic study of literature addresses the ways in which language is differently organized in verbal art (literature). Form is added to language, altered, attenuated, and differently grouped. These different kinds of organization are normatively subject to limits, some derived from limits on general or language-specific linguistic form. However, linguistic form can, in principle, be altered in any way at all, in avant-garde texts, for instance, or to produce artificial languages for literature; this possibility raises the general question of whether some organizations of literary language are cognitively transparent and others are cognitively opaque.
Of the various added forms, the most extensively studied has been metrical form, which requires the words of the text to be grouped into lines. Metrical form combines a non-linguistic counting system with a rhythmic system, which adapts the rhythmic systems of ordinary phonology. The metrical line may have a special status, as a cognitively privileged level of grouping, possibly because it is fitted to working memory. Rhyme and alliteration are two common kinds of added form; most linguistic interest has been in what counts as “similarity of sound” between two words, whether at a surface or underlying level. Rhyming and alliterating words are distributed relative to the grouping into lines and other constituents. The other major kind of added form is parallelism, where two sections of text are structurally similar, usually in syntax and vocabulary.
All literary texts have a discourse structure, which includes division into various types of group or constituent, including the division of a narrative into episodes, exploiting verbal cues of episodic boundaries. Narratives also require the tracking of referents such as people and objects across the discourse, which draws on the study of pronominals. Literary texts may also have a distinctive vocabulary, borrowing or inventing words to an unusual degree, and engaging in various kinds of word play.
Literary texts have “style” and “markedness,” ways in which the language varies in noticeable ways but without coding a different linguistic semantics. These stylistic variations are sometimes treated as having determinate interpretations, but there are also approaches to stylistic variations in literature, which treat them as having a non-determinate relation to meaning. Literature cannot have a different semantics or pragmatics from ordinary language, but meaning can be “difficult” in literature in ways not characteristic of much ordinary language (but in common with ritual speech and other ways of speaking).
A major mode of linguistic investigation involves corpora, over which statistical analyses are undertaken. This has a relation to the question of whether our literary-linguistic knowledge has a probabilistic basis, a question that ties the study of language to questions of expectation in aesthetics (e.g., music) more generally. Literature exists in various modalities—writing, oral literature, and signed literature—and linguistic approaches to literature have been sensitive to this, as well as to the special questions about how texts are set to music in songs.