Content, Form, and Meaning in Literary Texts

C O N T E N T , F O R M , A N D M E A N I N G I N L I T E R A R Y T E X T S FRED BOWERS University of British Columbia I n a paper called “Style and Its Image” 1 Roland Barthes observes that the view of a text as a binary structure of form and content is probably incor­ rect; the structure, he suggests, is multiple: “within it there are only forms, or more exactly, the text in its entirety is only a multiplicity of forms with­ out content.”2 Instead of being like the apricot with a flesh of form and a kernel of content, a text is better seen as “an onion, a construction of layers (or levels, or systems) whose body contains, finally, no heart, no kernel, no secret, no irreducible principle, nothing except the infinity of its own envel­ opes — ■which envelop nothing other than the unity of its own surfaces.” 3 Because Barthes’s view is alien to that of many, if not most, literary crit­ ics, I should like to examine it by first looking at possible candidates for content in a literary text, by next considering instances of form, and by finally suggesting more rigorous definitions of the terms content and mean­ ing than their current use in literary criticism suggests that they have for those that use them. We ought, as a preliminary, to recall that whereas the notion of content has been dominant in literary criticism of the past two hundred years, it has never had much of a place in music or the visual arts except when some deliberate effort at programmatization has been made. The content of a symphony, painting, or sculpture is no different from its form. In the case of the verbal arts, however, we are strongly pulled to say that there must be more than form. In particular, we are attracted to the notion that words and sentences must have content in that they refer to things, states, and events in the real world. This attraction has gained strength in the twentieth century from the duality of signifié and signifiant which forms a dominant model of thought in language theory, and which still persists despite the difficulties of formalizing a model that postulates the autonomy of any one linguistic component. Experience within semantic and syntactic theories has led away from the apricot view of a semantics-less syntax to an onion model in which one component might have dominance in a particular part of a grammar but is never wholly disconnected from another. E n g l ish Stu d ies in C anada, vii, 2, Summer 198 Among the major components which a semantic theory defines, the prime candidate for a text’s context is the reference of words. Because of our everyday use of language to indicate objects in the real world, coupled with a reductive way of talking about meaning, it is common to assume that words refer to things and that, therefore, things are the content of words. Indeed, grammarians who ought to have known better established a di­ vision of the lexicon into content words and structure words on this very assumption. However, the assumption needs some scrutiny. First, a distinction between ‘indication’ and ‘reference’ must be made. In everyday language like Will you pass the salt? or It’s very foggy this morn­ ing, the speaker is using the words you, the salt, this morning to indicate things and states in the real world which are in the common view of the speaker and his hearer. This deictic use of words clearly relates available words with objects in the real world by convention; as a process of indicat­ ing, the deictic use of words is verifiable by ostensión to participants sharing the same space-time segment or having experiences and knowledge in com­ mon. Elsewhere4 I have argued that there is a spectrum of deixis from the private, which characteristically uses words such as this, here, now, ago, I, you, which are fully interpretable only when the hearer shares the spacetime of the speaker’s utterance...