Denotational/indexical sound symbolism
Deux sortes d'iconicité (iconism) selon Silverstein

12 février 2015

John J. Ohala, Leanne Hinton and Johanna Nichols, Eds.
Sound Symbolism, New York, CUP, 1994

/p.39/ Chapter 4
Relative motivation in denotational and indexical
sound symbolism of Wasco-Wishram Chinookan

Michael Silverstein

4.1. Introduction

To many, sound symbolism would appear to be at the margins of how — as Jakobson and Levi-Strauss view inherently social facts — language is an "intervention of culture in nature" (Jakobson and Halle 1956: 17). To such a view, in fact, just as there were logically equivalent "Ding Dong," "Pooh Pooh," or "Bow Wow" theories of yore (see Whitney 1867: 426-427), cast in the idiom of language origin, there is a startling persistence or resurgence of essentially pre-structural views on the matter of sound symbolism. There appear to be many proponents of a notion of primordial sound iconism (in the technical, Peircean sense) for whom — notwithstanding the fact that language and culture are specific, organized semiotic systems — such iconic relations atomically motivate certain lexical forms in respect of what they denote, on grounds independent of any such sociohistorical facts of linguistic and/or cultural semiosis.

This kind of logically pre-linguistic and pre-cultural motivation of lexical denotation is, of course, what Saussure was talking about under the rubric of the "symbolic" (i.e. iconic) aspect of absolute motivation. The entire first two parts of the Cours (Saussure 1916: 97-192), by contrast, are devoted to demonstrating two truths. The first is that anything that is seriously a (denotational) sign in human language, from word-stem up through syntactic phrase, is so much more "arbitrary" in its semiotic properties than anything else, that we might as well axiomatize linguistics with this stipulation. The second is that the useful and productive opposition of "arbitrary" and "motivated" in language is a system-internal matter of degrees of relative motivation, from relatively arbitrary (or lexical) denotational signs to relatively motivated (or grammatically formed) ones, depending on a sign's regularity of value in, or rule-governed determination by, the whole system of language, its grammar.

In many respects we are more sophisticated today, eighty-five years later, in our ability to articulate the Saussurean lessons, at least in certain areas of the problem. /40/ For example, in the very analysis of the signal-forms, we have applied the Saussurean lessons in the realm of phonology, and can give a fairly precise descriptive account, in terms of primitives and conditions on their concatenation and projection, of what emerges at the surface as inventories and combinatorics of phonological segments that seem to comprise phenomenal denotational units — morphemes, words, phrases, sentences.

So we can examine putative sound-symbolic denotational units in any language with respect to the degree of phonologicization of pre-linguistic, pre-structural "sound substance," and with respect to the degree of violation of expectable phonological combinatorics under the Saussurean assumptions. Further, we can articulate with some precision the relationship of grammatical constructionality to semantic compositionality, given a rich enough understanding of morphosyntax of denotational language, and determine thereby what is, as a grammarian might say, "merely" lexical (hence, from a systemic point of view, totally arbitrary, and indeed the fit subject matter of non-grammatical study of sound symbolism). We can even understand something of the poetics of linguistic expression as a functional plane distinct from denotation as such, to determine the contribution of (broadly speaking) "metrically" organized form as one of the determinants of at least the native speaker's feeling of sound symbolism attached to certain expressions.

But even with such advances in being able formally to describe denotational iconism in language, we remain at the same impasse of understanding that Saussure himself faced. For we are operating along the single dimension of signs as being denotationally iconic or denotationally arbitrary and, as it were, equating specific-system determination with arbitrariness. Hence, on a higher plane of abstraction, many writers look for cross-systemic (i.e. cross-linguistic) universal generalizations; they immediately see in these absolute or statistical tendencies of form-denotatum correlation transparent evidence for psychological or more broadly biological motivation for linguistic form. (Haiman 1985; Givon 1984: 29-45; 1989 are two exemplars among many.) This is merely ding-dongism operating with due respect to the power of the structuralist perspective on the analysis of denotational structure. It is not really making use of the distinction between absolute and relative arbitrariness/motivation, even on the denotational plane.

For the point about sound symbolism in particular is that phonological or even phonetic shape is far from being merely the non-denotationally-correlatable level of "interpretation" (Chomsky and Halle 1968: 7), "representation" (Hockett 1961: 33, 41–42; Lamb 1964), or "articulation" (Martinet 1964: 22-27) of the real linguistic units, the morphemes, words, phrases, sentences — as in the Saussurean and subsequent grammatical views of the matter. In sound symbolism, such form is endowed with its own plane of meaning, one that, in the classic instance of denotational iconism, supersedes the power of the Saussurean grammatio-semantic system at its own functional game, symbolically mediated reference-and- /41/ predication. In other words, we must view denotational iconism as one of the "breakthrough" modes of semiosis, in which a system of sound structure (with its own, merely distributional functions of making segmental form), normally subordinated to virtual zero autonomous power with respect to reference-and-predication in the doubly articulated structure of language, undergoes a functional rankshifting (to use a tagmemics term apt in this context) into the plane of referential-and-predicational function. That is, sound as sign becomes independently endowed with apparently denotationally relevant value qua signifier.

If such is the case, we must ask, what is/are the usually latent function(s) of such a system of sound signifiers? How are such signifiers organized as units in some structure? How do they manifest the same kind of relative motivation, mutatis mutandis, in their own functional sphere(s) as gramrnatico-semantic units like segmental phonological units do in theirs? Finally, what other kinds of "breakthrough" modes implicate this structure of sound as well as denotational iconism?