F. s., and t. are parts of what is collectively called rhetorical or “figurative” lang. But the difficulty of distinguishing and relating these terms, or of giving a principled definition of any one of them, is well known, and writers from Quintilian (1st c. A.D. ) to Fontanier (19th c.) and Todorov (20th) have customarily begun by discussing the limitations of all previous positions. The present article considers first some of the attempts at defining t. in terms off., and f. in terms of t., then returns briefly to a consideration of s.
Discussions of f. and t. have traditionally used the philosophic and linguistic (semiotic) distinctions between word, or sign, and meaning to define t. as a delimited form of the more expansively conceived f. T. is defined by Quintilian (Institutio oratoria , ed. H. E. Butler, v. 3 [1920]) as “the artificial alteration of a word or phrase from its proper meaning to another,” and f. as “a change in meaning or lang, from the ordinary and simple form”; for Fontanier (Les Fs. du discours , 2 ?. [1827–30]), ts. are “all the fs. of discourse which consist of the divergent meaning of words, i.e. of a meaning more or less removed and different from their proper and literal meaning,” and fs. are “the characteristics, forms and turns … by which discourse, in the expression of ideas, thoughts and feelings, stands more or less apart from what would have been their simple and common expression.” Such classic formulations presuppose both a norm of “proper” meanings and “ordinary” usage from which ts. and fs. can then diverge (to which Todorov [1972] objects that, by this definition, “ordinary” lang, only becomes retrospectively—and inaccurately—viewed as lang, as it ought to be, “normal”), and also a qualitative distinction between changes in the meanings of words (ts.), and changes in the words and meanings of larger units of discourse (fs.). It would appear that there are only changes of meaning in the uses of individual words in such ts. as “rose” for a quality of affection or beauty in “My love is a rose” (metaphor), or “the White House” for the U. S. government (metonymy), or “hands” for seamen in “All hands on deck!” (synecdoche). But the words “rose,” “the White House,” or “hands” could not have any altered meaning as individual words (ts.) without an altered or at least specified context of words and meanings (fs.), following Austin’s argument that individual words alone can never have any meaning at all without context (1961). T. could thus appear to be a compressed instance of f., and f. to be an expanded t.
With regard to ts. and fs. as nonliteral, altered, or improper uses of lang.’s words and meanings, this is an aspect of how rhetorical lang, itself has always been viewed. The varying characterizations of this “rhetorical” difference in lang.’s employment bring with them different considerations of t. and f. Plato’s denigration of rhet. as manipulative and untruthful when measured against his model lang, of philosophy meant that ts. and fs. would primarily be characterized by falsehood (Phaedrus ). Aristotle maintains Plato’s view of rhet. as manipulative and less-than-philosophic but is appreciative of the pragmatic functions it serves; furthermore, he does not distinguish rhet. as “unlike” logic without at the same time noting the respects in which they are comparable or “like” (homos); , he also characterizes his syllogistic reasoning as one form or f. of speech among others (Rhetoric ). Shortly after Aristotle, rhetorical lang, receives its standard division into inventio (subjects, arguments, commonplaces), dispositio (arrangement of larger units of discourse such as exhortation, narration, peroration), and elocutio (choice and arrangement of specific words and phrases); fs. and ts. become simply parts of elocutio —which, after Ramus, we call “style” (q.v.)—constituting some of the verbal material of rhet.’s efforts at persuasion, incl. those of literary expression. Longinus similarly sees fs. in poetry as means for effecting emotional “elevation” of the audience (Peri hypsos ). In Cicero (De orat. )and the anonymous Rhetorica ad Herennium , fs. of speech (figurae dicendi )appear to mean “modes of eloquence” and to designate different levels of style. Dante refers to certain forms of allegory (q.v.), long classified as a f., as “truth hidden beneath a beautiful lie” (De vulgari eloquentia ), and Fontanier writes of “fs. of expression” (or words; see below)—which include personification, allegory, allusion, paradox, and irony (qq.v.)—as offering “an apparent illusory meaning in order to allow you better to find or grasp the real and true meaning.”
This brief survey of some notable conceptions of rhetorical lang, indicates that the alterations of “ordinary” usage in ts. and fs. may be variously analyzed as ones that change the true to the false, that shape or mold the unstylized into the highly embellished and specifically targeted, and ones that turn via the false or artificial to the true. The general insistence upon ts. and fs.’s divergence from a quasi-natural or basic norm is apparently preserved in the terms themselves, t. being from the Gr. tropein , “to turn,” “to swerve,” f. from the Lat. figura , “the made,” “the shaped,” “the formed.” But this view reaches a terminus ad quem in accounting for one of the briefest and most common of ts., catachresis (q.v), or the forced (Gr. “abusive”) attribution of a new meaning to a given word. Catachresis is not divergence from a literal meaning—in the sense of opposition to a word’s proper meaning—but rather its and its word’s extension to a place where there is no other sign (as in “leg of a chair” or “wing of a building”), which extension then itself becomes a necessary norm. Catachresis can thus be viewed either as the extension of literal lang, toward t., or as the entry of a t. (close to but more “basic” than metaphor) into “literal” lang. (Derrida).
Fs. function not only as extensions of ts. into larger units of discourse, but also as intrusions of rhet. into thought. In Cl. rhet., the traditional distinction between fs. of words and of thought (figurae verborum and sententiarum ) attempts to address this problem. Quintilian claims that fs. of words always involve a changed aspect of lang, if not of thought (repetition, antithesis, asyndeton [qq.v.]), while fs. of thought always involve an artificial or contrived thought and may either leave lang, unchanged or be composed of several fs. of words. Fontanier similarly distinguishes between ts. (altered meanings of individual words), fs. of “expression” or words (in which the form, order, choice, or assortment of words is affected—e.g. allusion, litotes, inversion, apposition, ellipsis, apostrophe), and fs. of thought (among which he counts prosopopoeia (q.v.), concession, and description). But there could not be a f. of words such as litotes without its intruding upon the thought —that, for example, “not unlike” is weaker than or at least different from “like”; and there could not be an apostrophe to something absent or ordinarily incapable of being an object of audible address without this affecting the thought of the expression. (This difficulty is evident in Quintilian categorizing apostrophe as a f. of thought, but Fontanier as a f. of “expression” or words.) Irony may not alter a single word from an ordinary expression in order to be manifest in a context—“I really like that” may ironically mean that one does not like it at all—and thus Quintilian understands it as a f. of thought; but for the very reason that such a f. involves divergence from—indeed, opposition to—a customary meaning of words, Fontanier considers irony to be a f. of “expression.” Fs. of thought, then, may simply be extended instances of or views upon fs. of words: in each case, either words change, or meanings change, or both. Modern linguistics as well as ancient rhet. is hard put to find any instance of sheer fs. of words, in which only sounds or arrangements of words are involved (e.g. repetition, opposition, parallelism, alliteration, assonance [qq.v.]) that does not bring with it at least a minimal f.—alteration or artificing—of thought as well.
Ts. and fs. (with the possible exception of catachresis) may be said always to present two senses under the guise of a single verbal formulation: either they say something unexpected by way of a striking expression precisely because the expression or novel sense rebounds against an expected sense or usage which still attaches to the words in question; or they do not “say” it so much as they let it, contextually, be understood beneath (or even in spite of) the actual expression (as in allegory and irony). This effect of doubling a single word or expression’s appearance or meaning may be construed as alteration of a first meaning by the addition of a second (Genette summarizes Fontanier’s t. as “the change of meaning of a word”) or as “the substitution of one expression for another” (Genette on Fontanier’s f.). It allows lang.—which is necessarily thought at any given moment to have more or less customary and stable meanings and uses attached to it—to say and mean more or something other (to allegorize, from Gr. allos-agoreuein , “to speak otherwise”). Minimally, this suggests that ts. and fs. always involve at least the relating of other words, meanings, and usages to the ones at hand, or the comparing of various meanings for words (ts.) or of one arrangement or usage of words for another possible one (f.). In the 20th c., when figurative lang, has most often been studied as a part of poetics or literary stylistics (q.v.), this view has become part of a psychology of literary response, as in I. A. Richards’ position that a metaphor allows two or more ideas of different things to be carried by a single word or expression (the “vehicle”), its meaning (the “tenor”) resulting from their “interaction” (cf. Empson).
Opposed to these long-standing views of t. and f. as swerves from normal lang, usage, with their concomitant effects understood psychologically as an audience’s reaction to such divergence, there have also been treatments off. and t. as ineluctable discursive structures, fundamental to the organization of verbal experience and consciousness, and to the production of lang, and understanding themselves. Paul de Man considers the fs. of allegory and irony to be constitutive of interrelated treatments of temporality, allegory projecting a consciousness of meaning in relation to time along an axis of past and future—thus constituting narrative—irony doing so along an axis of a present moment divided between empirical and linguistic or “fictional” selves—thus constituting a divided self-consciousness (“The Rhet. of Temporality”). Roman Jakobson uses the structural linguistic (Saussurian) understanding that all lang, displays two axes of selection (similarity, the “paradigmatic” axis) and combination (contiguity, the “syntagmatic” axis) to argue that metaphor is essentially selective , metonymy essentially combinatory , and that all lang., incl. all lit., orients itself toward one or another of these principles of verbal organization.
With ts. (or fs.) such as allegory, irony, metaphor, and metonymy thus being held to generate literary texts and “original” verbal utterances, it is no longer a matter of t. being understood as a species off., and f. as a special (divergent) kind of lang., but rather one of t. and f. being used to name and analyze lang, in its fundamental structures. A lineage for this view can be found in such diverse writers as Vico, Hamann (“Über den göttlichen und menschlichen Ursprung der Sprache” [On the Divine and Human Origin of Lang.], 1772), Rousseau (”Essai sur l’origine des langues” [Essay on the Origin of Langs.], 1817), and Nietzsche, who variously see the “origin” of lang, in t. or f. such as catachresis or metaphor. When Nietzsche says that one can only speak figuratively about the relations between lang., fs., and their meaning (”Über Wahrheit und Lüge im aussermoralischen Sinn” [1873]), he recalls Vico’s effort to analyze historical “stages” of lang.-use with the names and structures of ts. and fs. (Scienza nuova [1725]; see also N. Frye, The Great Code [1982]) and anticipates de Man’s view that attempts at historical understandings of lang., its origins, and its meanings (Rousseau’s and Nietzsche’s among them) necessarily show themselves to be ts. and fs., e.g. allegory, metonymy, metalepsis (de Man 1979, 1983; see also H. Bloom, A Map of Misreading [1975], on lit. hist, as the tropological turns, displacements, or alterations of an earlier poet’s lang, by later ones). Erich Auerbach shows that, for a millennium, the late Cl. and medieval Christian doctrine of figural exegesis extended figura from a domain of eloquence narrowly construed into the interrelations of figurative lang, with identifications of historical meaning.
Independent of these problems of a metalanguage of history being addressed to lang., nonhistorical analytic or descriptive lang, about lang, also appears to involve t. and f. Quintilian, defining t. generically as “transfer of expression,” uses the technical term for metaphor (Gr. metapherein , “to carry over,” “to transfer”) to describe the set of which metaphor would be a member; and Fontanier concedes that to speak of “fs. of discourse” at all is itself metaphoric , using a metaphor (or perhaps a catachresis) from the traits, form, and contours of the body for those of lang. Derrida has analyzed this persistence of f. and t. in philosophy’s discussion of the very terms. Jakobson, observing that “similarity in lang, connects the symbols of a metalanguage with the symbols of the lang, referred to” and that “similarity [the metaphoric pole] connects a metaphorical term with the term for which it is substituted,” concludes that this dichotomy “appears to be of primal significance and consequence for all verbal behavior and for human behavior in general.”
Whether or not all metalanguage is metaphorical—which would mean that all rhetorical (and grammatical and logical) treatises and also all linguistics are figurative—it is undeniable that all commentaries upon fs. and ts. are metalinguistic, i.e. substitutive, and thus, in the view running from Quintilian to Jakobson, tropaic. From the medieval trivium to Fontanier and much modern linguistics, understandings of figurative lang, have traditionally subordinated rhet. to grammar and grammar to logic, but this has not diminished the recurrent power of analyses of fs. and ts. to uncover a disruptive questioning of the ostensible “logic” of explanations off. and t., and of the tropological forces at the basis of both lang, and metalanguage (de Man 1986). The argument that fs. and ts. are constitutive of and coextensive with lang, has as one of its consequences that lit. becomes not merely an occasion for “foregrounding” lang, or bringing it to consciousness—a view found from Hegel to some Rus. Formalists and Czech and Fr. structuralists (e.g. Riffaterre 1971, 1978)—but also a privileged vehicle or avenue for access to the problems lang, poses for understanding, with de Man “calling ‘literary,’ in the full sense of the term, any text that implicitly or explicitly signifies its own rhetorical mode and prefigures its own misunderstanding as the correlative of its rhetorical nature; that is, of its ‘rhetoricity’” (1986).
S., in its relation to f. and t., enters here as a summary instance of the breadth and complexity of figurative lang, and its understanding. S. in antiquity often means fs. of words and of thought (schémata lexeós and dianoia ), and Cicero, the Ad Herennium , Quintilian, and later Ren. rhetoricians frequently appear to use the term for fs. of speech or levels of style. But in Cicero’s Lat. trs. of Aristotelian and other Gr. uses of s. for outward appearances, semblances, or perceptible forms and shapes, it can also broadly represent thought and lang, in their relations to reality, its true perception and understanding. Thus Aristotle writes of his schémata syllogismou , and of the schéma tés ideas (Metaphys. ), and the atomists Democritus and Lucretius consider ss. to emanate from bodies and to be like images (Auerbach). Kant uses s. to name an order of thought and lang, that cannot be known perceptually or empirically, but in relating s. to such fs. as hypotyposis and symbol, he reintroduces its imagistic and figurative (”shaped” or “formed”) aspects (Critique of Pure Reason, Critique of Judgment ). When one asks colloquially after “the s. of things,” this trad, of forms and meanings is summarized as one invokes a nonobvious, non-visual pattern or sense which attaches to things or events, but which nonetheless can be brought to lang, and thought in the form of a verbal image or narrative.
S. is scarcely used any longer to speak of figurative lang, in rhet. or lit. But it is schematization, in the wider sense of the tropaic rendering of appearances and meanings, that is recalled when de Man analyzes the t. of prosopopeia—from Gr. “to make or give a person or face,” and once categorized as a f. of thought or a s. —as both materially a textual f. and, in its catachretic projections, a means by which “face” (Fr. figure )and voice are inscribed for lang., cognition, and consciousness (”Hypogram and Inscription,” “Anthropomorphism and T. in the Lyric”). Prosopopeia and f., s., and t. are not unique or proper to certain forms of the lyric such as the elegy, the ode, or ekphrastic poetry, nor to lit. in general. Rather, poetry and poetics, to the extent that they can be said to be constructed exclusively of lang., are sites among others where f., s., and t. have been deposited as materials of lang, and modes of thought.
User Comments Add a comment…