Some remarks entitled “Literature and Language,” R. A. Sayce, an English critic, voiced a complaint which I expect has recently diminished, and will continue to diminish, in force and frequency. Objecting to what he feels has become a dominant trend in poetic analysis, “This, then,” he insists, “is the first great problem of literary language–to discover the relation between technical studies of prosody or prose rhythm and the music of words.” I think that we may take Mr. Sayre’s implied request as demanding not philosophical clarification (e.g. of the general problem, “What is the connection between our knowledge about a process or thing and our normal uses of that process or thing?”), but instead, a justification of prosodic analysis in terms of its success in confirming and accounting for the almost magical effects upon a reader of the “musical” or non-semantic patterns of poetic structure.
One of the ways in which structural linguistics has been extremely helpful to literary criticism, I think, has been in helping to answer this kind of demand for a grounding of metrical studies in empirical knowledge. A structural grammar of a language does indeed segment the flow of speech and, temporarily at least, seems to hypostatize those segments, to treat them as things, and discuss the structure of more complicated things, utterances composed of them. But utterances are speech acts; it is only for the purposes of understanding their structure and function that they are treated as things at all. As far as linguistic theory is concerned, there is no chasm requiring bridging between the study of utterances and knowledge of their effects, meanings, significances, or specific ways of mediating the behavior of other people. The first leads to the second; and it may not be misleading even to say that the second can be defined in terms of the first.
