An intuitive reading of Prose (we will come back to it in greater detail) shows the unusual accumulation of phonic groups in each line, their repetition in a given stanza and their reprise in other stanzas having a synonymic or antonymic signification. So that this phonic insistence imposes itself as the preponderant organizational principle in the text, which goes so far as to eradicate the subjacent metric processes (ceasuras, enjambment, quatrains, octosyllables, etc.): in the so-called free verse and after, this phonic network remains as the only constraint, unconsciously exercised, rather free on the whole and inherent to language, which restores a unity to the process. Statistical, or even probabilistic, studies of the frequency of phonemes in a text relative to the frequency of the same phonemes in everyday language, uphold the notion that poetry uses the typical sounds of a language more and that it emphasizes their presence while suppressing raresounds (See Jiri Levy, Mathematical Aspects of Theory of Verse, in L. Dolezal and R. W. Bailey, in Style. New York: American Elzevier, 1969; pp. 95-112), while other studies dispute this finding (See Bailey, Statistics and the sound of poetry. Poetics I, 1971; pp. 16-37); but the principle of the accumulation of phonemes, as well as the structuring role which the distance (the gap) between two occurrences of the same phoneme plays in the poem, no longer seem to be disputed (See H. SpangHanssen, The Study of Gaps between Repetitions, in For Roman Jakobson. Hague: Mouton, 1956; pp. 492-502). One can observe, through this research and in Mallarm6's text itself, that the repetition and the redistribution of the phonic and semantic potentialities proper to the language produce new structures of signification. But we can go further. If repetition accentuates the semiotic role of the phonematic code of a language, repetition also tends to cross over the limits of what one understands by a phonematic or morphophonemic code. crossing occurs in two directions. On the one hand, the increased frequency of a given phoneme, or the accumulation of phonemes of a single group, or the drifting between phonemes of neighboring groups, produce an effect which is foreign to the common usage of the natural language; they tend to move, not toward a universal phonetism, embracing all languages,' but rather toward a pre-phonematic, shall we say phonetic, state, which can be observed in children who have not yet acquired the sounds of one language but are capable of producing all possible (non-linguistic) sounds. On the other hand, each of these phonemes carries semes with it, so that the morpheme or the lexeme to which these phonemes belong is dislocated, and the phoneme, acquiring semantic value in the process, tends to constitute a semantic constellation made up of all the lexemes containing this phoneme. first of these directions divests the phoneme of its phonematic character, as an element of a system of distinctive traits, and reconnects it to the phonetic, that is to the articulating body: initially the articulatory apparatus and then, through the drives (pulsions), to the body as a whole. second direction, on the contrary, utilizes the distinctive, phonematic character of the sounds of the language to establish agreements, one could say applications (in the logical sense of the term) between semes belonging to different morphemes or lexemes. mixed func-
Read full abstract