Escaping language: Roman Jakobson and Abhinavagupta.

• Author: Gerow, Edwin

PART I--ROMAN JACOBSON

In their now classic work. Fundamentals of Language, (1) Roman Jakobson and Morris Halle seek, in a brief second section (pp. 69-96), to extrapolate to a more general level of linguistic truth certain verities that follow from their important grammatical distinction between syntax and paradigm--syntax, the study of words in proximity, never coinciding and therefore at least two, ordered by a principle of sequence; and paradigm, the study of words never in proximity, but necessarily exclusive, occupying the same "slot" in the sentence by a principle of selection.

Jakobson and Halle close their discussion with an examination of two wider "linguistic" issues, speech pathology and "poetic" language, both of which show the imprint of the distinction referred to above. One might hazard the guess that pathology occupies something of the level "beneath" language, while "poetics" is clearly "above" it, using its "resources" both consciously and willfully. However that may be, speech pathologists have long recognized, according to our authors, that aphasia has two basic aspects, which are realized in different subjects and constitute two "disorders"--one a "similarity" disorder, the other a "contiguity" disorder. In the former, words of the same "class" are easily confused or misapplied--but "word-order" is generally respected; in the latter, words, which may be suitable to the context, are "put together" helter-skelter, producing a syntactic jumble--sentences become, in the extreme case, monosyllabic. Clearly, the patient of the first sort is having trouble with paradigms, while the latter patient is deficient in "syntax." The two principles of "order" underlie the two forms of "disorder."

The true import of the contrast is brought out in the few pages devoted to "poetic" language--where the terms "similarity" and "contiguity" tend to replace "paradigm" and "syntagm." Jakobson and Halle here advance the interesting thesis that the two basic "figures" of speech, metaphor and metonymy, realize at the stylistic level the distinction between the two types of order--metaphor being a usage that derives its sense from a notion of similarity, that is, replaceability (or even "identity"); and metonymy, one that derives its sense from connections and relations to things and ideas other than the thing itself. The former tends to appositional utterance (or to word replacement based on such predication), "my love's like a red red rose," or simply, Abie's "wild Irish rose." The latter tends to verbal predication, for a syntactic disjunction is essential to understanding the displacement at the heart of "metonymy"--"Washington was all agog over it" or "the grandstands are cheering." In both these cases, the disjunction depends on the perplexing combination of an inanimate subject with an animate verbal predicate. No "equation" of course is intended, but rather a difference, one that forces a novel reinterpretation of one of the terms, so as to save the "sentence": "Washington" means (by a synecdoche of "part" and "whole") "all those in Washington who have some interest in the government" and "grandstands" means (by a relation of "container-contained") "men occupying the grandstands."

Jakobson and Halle further relate these two basic "figures" to the two principal modes of figurative discourse, "poetry" and "prose," and even to "styles." such as the "romantic" and the "realistic." that they see as exploiting one or the other "figure" in demonstrable excess, or to the quasi-exclusion of the other.

My purpose is not to object to this basic structure, but to call attention to two qualifying factors, drawn from the Indian traditions of poetics, that, first, establish that Jakobson and Halle have been largely anticipated by several centuries in the formulation of these parameters; and secondly, that these same Indian traditions (with their peculiar emphasis on the significance of language itself) force interesting reexamination of some of the more lapidary contrasts ("poles") implied by Jakobson and Halle's outline.

Primo, the boundary between their "metonymy" and "metaphor" is not only recognized, but is even more aptly conceived in the Indian texts; secundo, the preponderance attributed by Jakobson and Halle to the "metaphorical" pole in Western literary studies is not at all evident in the Indian world, which pays considerable attention to "figures" not based on simile.

The terms metonymy and metaphor have rarely been used carefully. Sometimes one or the other term is taken as the genus of all figuration, the remaining term being just a variety thereof. Aristotle seems so to privilege "metaphor" (if his Greek may be taken at face value--after all, his usage is the source of the term itself). A "metaphor" is a term that takes on (inf., metapherein) another meaning by virtue of its context. This could be based either on similarity or on contiguity. In my own writings (2) I tend to use "metonymy" as the generic term, reserving "metaphor" for substitutions based on similitude. A "metonym" is a term used "in place of" another, for whatever reason, which, in principle, covers similes. Sometimes, as with Jakobson and Halle, an attempt is made to dichotomize the figurative universe by applying the terms contrastively, and when this is done, "metonymy" is usually reserved to figuration based on relations of contiguity, "metaphor" to figuration based on relations of similitude. This contrastive understanding is very likely influenced by the usage of C. S. Peirce, whose well-known symbolic triad; icon, index, and symbol. (3) depends on a radical disambiguation of "likeness" (in Peirce's terms, "firstness," which precedes or transcends distinction) from "contact," etc. (in Peirce's terms, "secondness," which depends on distinction, on one "influencing" the other). The weathervane, as an indicator of the weather, is an "index," whereas the rose, as an indicator of "my luv," is an "icon."

Here, it suffices to point out that the early Mimamsakas, in their theory of word and sentence signification, insisted on the difference between a transfer of meaning based on "secondary" (gauni) qualities and that based on positive "indications" (laksanika). Their discussion makes it abundantly clear that the issue is the principle of relation involved--in the former case, it is similitude (merely), while in the latter case, it is, in effect, any other relation. The reasons for so differentiating similitude among the "relations" are quite parallel to Peirce's: for a "similitude" to obtain, no requirement of mutual coexistence need be involved--which amounts to basing the "similitude" on the "eye of the beholder," and releasing it from all objective measurement. Similitudes are notoriously imponderable--whereas "cause and effect" or "container-contained" are neither located in the observer nor are they immeasurable, unspecifiable. This distinction has been defended in various forms by many later Indian semiologists, though it has also been contested--oddly enough, chiefly by poeticians, to whom similitude is the relation par excellence. (4) The "poetic" version of any relation--take "cause and effect," for example--involves a "distortion" of some sort--as when the cry of the peacock is represented as provoking the rains. There is, at this level, little reason either to privilege or to disparage "similitude."

In this context, I wish to direct your attention to the surprising remark of Jakobson and Halle that "the study of poetical tropes" has been "directed chiefly toward metaphor" (p. 96). They add: "nothing comparable to the rich literature on metaphor [a work of Sutterheim is noted] can be cited for the theory of metonymy" (p. 95). While this might be true of the Western traditions Jakobson and Halle are concerned with, nothing could be less true of the Indian poetic and semiological traditions. As mentioned above, Indian semiologists were very concerned, indeed preoccupied, with the notion termed laksana--that deviation from the literal and the direct that is both ubiquitous in language and appears to compromise its objective authority. Whenever an assertion--Vedic assertion most centrally--appears to depend on the faculty of interpretation brought to it by the hearer, as is the case in any metonymic usage, its claim to authority is diminished. What are we to make of the Vedic "assertions" that "trees came to the sacrifice" or that the "rocks wept Soma"? Can such language be faultless and immemorial and thus a sure guide to our obligatory actions? That such usage is parallel in every way to mundane figurative usage ("the grandstands are cheering") does not add to its stature either. This problem led to an extensive discussion of "figurative" usage and to its classification, in later treatises, such as Vrttivarttika, into minutely differentiated subtypes, such as whether the base meaning (to which the figurative refers) of the word or phrase survives in any way in the resultant understanding, or whether the referent of the metonymy is subsumed in the metonym or retains at least verbally its separate status, etc. Nine or more such "types" are discussed--by theorists such as Appaya Diksita, who are also poeticians of note, and quite aware that such studies are essential precursors to the study of poetic language itself--where the "figures of speech" and the notion of "indirection" or "suggestion" (dhvani) occupy center stage. But the original purport of the discussion is never lost sight of--every attempt is made to link "metonymic" usage to its literal base and there rationalize it, so that the subjective "interpretive" factor is ruled out. (5) Such an "objectification" of language and its semiosis is unparalleled and deserves to be noted in any account of theories of language. Not to do so serves to confer an illegitimate aura of novelty on the work of those who thus ignore the rich pre-modern and Oriental...