Cognitive Metaphor Author(s): Charles O. Hartman Source: New Literary History, Vol. 13, No. 2, Narrative Analysis and Interpretation (Winter, 1982), pp. 327-339 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/468915 Accessed: 11/06/2010 06:53 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp.. JSTOR's Terms and Conditions of Use provides, in part, that unless http://www.jstor.org/page/info/about/policies/terms.jsp you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=jhup.. http://www.jstor.org/action/showPublisher?publisherCode=jhup Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact
[email protected].
The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to New Literary History.
http://www.jstor.org
Cognitive Metaphor Charles 0. Hartman ME
ETAPHORfascinates twentieth-century critics. It is the city to
which all our roads lead. Traditionally, we feel, it languished as one among the many figures of rhetoric. Now, perhaps unjustly distrusting the very word rhetoric, we refuse to relegate metaphor to the role of ornament. We ask it to transcend the old categories, to serve functions foreign to a "mere trope." Most important, we expect it to function cognitively. In many ways and many terminologies, modern critics propose that metaphor, somehow, is or creates or communicates knowledge. Yet this claim has troubled some philosophers and some critics. It encounters three major difficulties, each stemming from an implied definition. First, metaphor may be a condensed simile, "A is like B" having become "A is B." Aristotle, whose various mentions of metaphor anticipate or provide most subsequent definitions of the term, says in the Rhetoricthat "the simile ... is a metaphor, differing from it only in the way it is put"-the addition, in English, of "like" or "as."1 Both tropes state similarities. This is unobjectionable but not very exciting. Everything in the universe more or less resembles everything else. (Or, alternatively, nothing is like anything else.) The writer who notes similarity between objects says little about them, especially when he leaves their common characteristic-the comparisonunspecified. Howard Nemerov distinguishes metaphor from simile on just this ground, that only the latter "isolates for you the likeness in virtue of which the comparison is made."2 But in that case it is hard to see what metaphor adds to simile besides confusion, or why we should ever prefer it. Aristotle's answer-that "just because it is longer [simile] is less attractive"-seems trivial; at least, it evades the question of cognition. when the two things differ most, the basis of comparison is likely to be the least characteristic attribute of both-a lowest common denominator. Why, then, do we especially warmly favoring Metaphysical juxtapositions-those prize-so metaphors that "yoke by violence together" the most dissimilar things? We know ways around this objection-"surprise," the "harmony of opposites"-but they work by distinguishing poetic effect from ra0028-6087/82/130327-13$ 1.00/0 Copyright? 1982 by New LiteraryHistory, The University of Virginia
328
NEW LITERARY
HISTORY
tional knowledge. Such distinction, though surely valid, begs the question. Nothing in this view of metaphor encourages the critic who wants to ally it with cognition. We resort, next, to a more sophisticated definition: metaphor does not blandly remark a similarity but boldly declares identity. really is B. Yet the problem here is obvious. If this is what a metaphor does, then metaphor is a lie. It takes little philosophical training to see that a false statement will lead to a true one only by accident or revelation. If a rational or logical process begins if they happen somehow from false premises, its conclusions-even to be true-cannot reliably be called knowledge. (This suspicion underlies Plato's rejection of the poet.) At this point the critic may retreat into claims for a special kind of poetic knowledge. To do so, however, seems anachronistic. These claims are essentially mystical, relics of the Victorian attempt (made explicit by Arnold) to replace religion with art, to seek revelation in the books of men. The modern critic, whose defense of cognitive metaphor often suggests envy of the natural sciences, weakens his position by adopting this line of argument. Trying to save cognitive metaphor by insisting on new (and vague) meaning of cognitive, he shares the fate of those whose fervor outstrips their rhetoric. He may convert the gullible and gratify the sympathetic, but he does not convince the skeptic. We do not construct new definitions of knowledgefor psychoanalysis or our senses or experimental science; what would warrant our doing so for poetry? The objects of knowledge vary, and the ways of getting or giving it; but knowledge remains single in kind. Finally, from natural science itself (and the empiricist philosophy that serves it, and the psychologies that derive their principles from it) comes the objection that metaphor produces not knowledge but delusion. The usual term is "category mistake." Though the jargon and its philosophical basis are relatively modern, the accusation burdens many traditional definitions of metaphor, as when Aristotle says that metaphor is "giving the thing a name that belongs to something else."3 Though he approved of metaphor as "sign of genius"-and though, as we shall he provides elsewhere way to cut through these tangles-this definition resembles that with which began, in disappointing any hope that metaphor could produce real knowledge. If (as the definition assumes, and as both physics and psychology assure us) we perceive the world in terms of categories we have learned or inherited, and if rational action depends on our correctly applying these categories, then the habit of kidnapping word and cuffing it to an alien object endangers sanity itself. Despite a tradition of imagery, if I treat my guitar as a woman am insane. And if realize that my guitar is not woman, how can I learn anything from speaking as
COGNITIVE
METAPHOR
329
think do learn something, but this view of though it were? metaphor mystifies the process. Why, in short, in questing for knowledge (or attempting to communicate it), should we begin by pretending that the object is something it clearly is not? Perhaps, some critics suggest, comparison with things we know allows us to comprehend what we do not. Metaphors sometimes do this; but I know as well what guitar is, and what a woman is, before compare them as after. Nor do the confusions generated by this third definition of metaphor end here. Definitions stemming from Aristotle's imply that term from one context, serving as the vehicle, carries across to that of the tenor.4 Thus in "the ship plows the waves," a term from agriculture invades a nautical description. This might be confusing enough, though one assumes that the larger context of discourse-the story of an ocean voyage, for instance-would clarify relations between the categories involved, the centrality of one and irrelevance of the other. But many metaphors reverse the transference. In "A Valediction: Forbidding Mourning"-that primary text for modern treatments of metaphor-the stanzas depicting the compass (the vehicle of metaphor whose tenor comprises the pair of lovers) bulge with terms that normally inhabit context more closely neighboring on lovers than on compasses. The compass is said to "hearken" and come "home." These words describe a compass only indirectly, fact, metaphorically. Somehow, while the compass fancifully-in stands for the lovers, the lovers stand for the compass; the juxtaposed contexts grow ever more entangled, and the cognitive value of the metaphor becomes harder and harder to explain. Even the categories of metaphor itself, tenor and vehicle, start to confound themselves. Again, we recognize the phrase "a sea of faces" as a standard sea. But when, in "The crowd of faces resembles metaphor: Yachts," William Carlos Williams says, "It is a sea of faces," he inverts the metaphor: the sea has become a mass of faces. Once more, if the function of metaphor is to communicate knowledge, these poets' choice of method seems perverse. Compressed simile, forged equation, and category mistake are the three most common definitions of metaphor. By specifying certain characteristics of metaphor, each accounts for part of its effect in our experience. Yet each jeopardizes the claim that metaphor generates knowledge. By any of these definitions, "cognitive metaphor" bewant to suggest that a fourth definition obcomes an oxymoron. viates these difficulties. It is neither new nor mine; I owe its rediscovery to Hugh Kenner.5 In the Poetics, Aristotle lists the kinds of metaphor: "Metaphor consists in giving the thing a name that belongs
330
NEW LITERARY
HISTORY
to something else; the transference being either from genus to species, or from species to genus, or from species to species, or on grounds of analogy ... analogy is possible whenever there are four terms so related that the second (B) is to the first (A), as the fourth (D) to the third (C)."6We have already seen that the formula "giving the thing name that belongs to something else" undermines the concept of "cognitive metaphor." The last of Aristotle's list, however, repays attention. Briefly, it means that one properly generalizes metaphor, not as a C/D. This defiB), but as a proportion: A/B simple equation (A nition, which appears to luxuriate irresponsibly in mathematical symbols and to tempt the critic to trifle with diagrams, has been largely ignored. Many metaphors seem to offer simpler reductions as pairs of terms: "my love is a rose" reduces to "lover rose" and abstracts to "A B." Yet as Kenner points out, even so casual a metaphor as "the ship plows the waves" loses something vital when we reduce it to "ship plow." (That "my love is a rose" suffers similarly we will see later.) More accurately, the metaphor implies that "a ship does to (or is to) the waves what plow does to ground": ship/waves plow/ground. Though we have obviously abandoned elegance of expression, we have saved the essential information; and it is information that concerns us when we speak of metaphor as cognitive. Returning to "The Yachts," we can observe the same kind of metaphor emerging in the last three stanzas. The yachts in their race, contending, in the poet's vision, not against each other but against the sea through which they cut their way, become an emblem of those who own them. We might reduce this part of the poem, as metonymy, equation terms: yachts their owners, "the rich." But in those visionary stanzas Williams turns away from the yachts, attending instead to the "watery bodies" of the waves, the "sea of faces" which the yachts "pass over"-their indifference rescinding the celebrative imagery the first eight stanzas lavished on them. If the poem is a metaphor (it becomes impossible to read literally), it exceeds the complexities of the metonyfour-term analogy, however, allows us to include the sea on my. whose suffering the poet concentrates: yachts sea
rich poor
At least in describing what a metaphor contains, the four-term analogy surpasses the two-term equation. Before asking how this improves the case for metaphor as a cognitive process, we should clear a little more ground. Aristotle himself clearly intends his definition quite narrowly. For him analogy is only
COGNITIVE
331
METAPHOR
one kind of metaphor, presumably to allow for "simple" ones in two terms. But we have seen how the definitions available for two-term metaphor exclude it from the realm of the cognitive, and we might turn back to see if we can rescue it. Mathematics of a trivial sort will B," we can also help. If we can reduce "my love is a rose" to "A express that equation as a four-term proportion: A/C B/C. It stands essentially unaltered. And this corresponds more closely to what we understand in reading or hearing such a metaphor. By itself, "my love is rose" is merely odd. Instead, what affects us is the attitude it the speaker asserts himself only in "my." The embodies-though woman and the rose are not equated in vacuum; someone compares them, and he is the matrix in which they become comparable. "My love is a rose" means something like "I feel about my love as do about a rose" or "my love does to me what a rose does to me." (Compare "the ship does to the waves what a plow does to the ground.") The structure of "my love is a rose," then, is my love me
rose me
Though this diagram obviously does not replace "my love is rose," it rose."7 preserves more information than "lover If "my love is a rose" represents a reduction in the number of terms of the analogy-though only to three, not to two-many of the most interesting metaphors in literature require the proportion to expand beyond four terms. An example from Yeats: Memory One had a lovely face, And two or three had charm, But charm and face were in vain Because the mountain grass Cannotbut keep the form Where the mountainhare has lain.8 This remarkable poem deserves more attention than can give it here.9 But it unfolds from a central metaphor. The first two lines list various (noticeably vague) women; the third dismisses their virtues as futile (not only "vain" but "in vain," implying attempts-at what, we do not yet know); and the last three lines speak of hare, its nest in the grass, the grass itself, and the mountain on which it grows. Incongruity makes us-probably before we consider consciously-read the final explanation ("Because") as metaphor. We intuit the basic equation easily enough: the hare represents some one special woman who
332
NEW LITERARY
HISTORY
affected the speaker too strongly for other women's charms to attract him later, after she left. In short, "hare woman." But clearly this ignores most of the metaphor. The hare (within the context of the vehicle) associates with the form, the grass, and the mountain. By process of deduction which is, again, probably unconscious but which operates on the relations among the terms of the vehicle and transfers them to the tenor, we fill in the remaining terms: hare woman mountain speaker form memory (as in "a memory") memory (the faculty of memory) grass The complex process by which we arrive at these terms will concern us later; but it is clear that we must acknowledge eight terms here and not four. This example and that of "simple" metaphor suggest that four terms are only kind of model or norm and that in practice metaphor can reduce the number of different terms by one or expand it perhaps indefinitely. We ought also to notice in these examples the presence of implicit terms. Although in "my love is rose" the speaker who acts as the third term of the metaphor declares himself in the possessive pronoun, "the ship plows the waves" does not mention "ground" or "earth," the necessary fourth term. In "The Yachts," although the sea is said to be full of "watery bodies," what leads us to conclude that it represents "the poor" is no linguistic evidence in the poem itself, but the structure of the metaphor. In "Memory" we had to supply the woman, insert the speaker who never mentions himself, and perceive the double significance of the title as naming both a faculty and what it contains or manipulates. In most cases the terms of the vehicle are explicit, but usually we must deduce at least one term of the tenor. The logical extension of this technique is what Hi Simons, speaking of Wallace Stevens, calls "radical metaphor"-metaphor that announces only its vehicle, the whole tenor being left to implication. Simons use of metaphor. As Steas the distinctively proposes vens's short poems show, it generates metaphors that yield multiple interpretations. One could defend any reading in which the relation among elements in the proposed tenor matches the relation among explicit elements in the vehicle.10 will not be thought to claim that by reducing these hope replace the originals. More happens in metaphors to proportions the metaphor of "Memory" than diagram even of eight terms identifies. The "hare" connotes swiftness and wildness, suggesting certain things about the absent woman it represents. For the speaker to figure
COGNITIVE
METAPHOR
333
himself as a mountain is to imply solidity and (since we see the mountain from its side, not from majestic vista) isolation. Such connotations are indispensable to the effects metaphors have on us. It would be silly to deny them. But cognition is a word that implies knowledge as distinct from emotion.11 Obviously, in poetry (as, ideally, anywhere else in our lives) the two cannot be separated. Yet when we ask of metaphor how and whether it produces knowledge, we plausibly look to the information it contains. Proportional analogy proves itself to be a credible way of representing the solid core of information in any metaphor. We need only remember, while searching for cognitive value, that metaphor has other values as well, and other ways of working. It seems possible now to answer the question with which began. suggested one fairly simple way in which metaphor is cognitive when I noted that each analogy involves implicit terms. These terms are, after all, bits of information. In "Memory," one of them allows us to replace the speaker at the center of a situation from which he has apparently removed himself to ensure a tone of resignation or indifference. The poem pretends to reflect in detachment on the workings of a mental faculty. Behind that facade, it nurses a personal wound. Told of the "mountain hare" that has (unnaturally, says the adjective) left its "mountain," we infer the speaker's suppressed pain, and perhaps anger, that the woman has deserted him. We need to discover him in the poem to understand this. But we accept his reticence as a necessary condition of his accuracy in feeling and comprehending the experience. His very silence affects us. Only the metaphor, in which he can remain an implicit term, combines this silence with the information we need to perceive its significance. In "The Yachts" the implicit terms of the tenor, which the structure of the metaphor bids us call social classes, associate the poem with a point of view closer to Marx's than Jay Gatsby's. And so on. The "information" in each case may involve or underwrite tone; but if the objectsof knowledge in poem are, like those in psychoanalysis, emotive, this makes it none the less "For some of the knowledge. Furthermore, as Aristotle points terms of the proportion there is at times no word in existence."12 We noticed this justification for metaphor earlier-that it makes known what was unknown. But we also recalled then that this breaking of new ground does not account for all or even most metaphors. Often the information these implicit terms convey seems trivial, as in "the ship plows the waves." Sometimes, as in "my love is rose," the metaphor eschews them entirely. And in any case they depend for their significance on the metaphor's structure, which leads us to posit them. It is hard to see
334
NEW LITERARY HISTORY
how they could detach themselves from that structure to operate as knowledge in any wider area. If they contribute to cognitive function of metaphor, it must be by the way in which we deduce them. Cognition must finally arise from metaphor as process of knowing rather than reside as a thing known; and we had best turn from the implicit terms to the metaphorical structure itself. And there, in fact, we find what we need. The very possibility of deducing from a metaphor terms which it does not explicitly present implies that the metaphor is a structure of relations. (The syllogism and the sorites are similar deductive structures.) This is why increasing the percentage of verbs-which carry the weight of action and sentence-almost relation in inevitably renders a passage more metaphorical. The mathematical form analogies take confirms the same point. metaphor equates, not objects, but relations or sets of relations. If we reduce it to B, we are hard put to make sense of the doubled hyphen between the two letters. But in A/B C/D, the sign of equation means what it says. (It is hard to resist the frivolity of remarking that this holds even for the abstract equation when we take it "literally": precedes B in the alphabet-a definable relation between the two-just as precedes D.) Metaphor does not state a similarity of objects, as the first objection to calling it cognitive would have us think, but an identity of relations.13 Nor does the metaphor lie, as the second objection complains. What a ship does to the waves is, literally and exactly, what a plow does to the ground. An action is one kind of relation that connects two things, an agent and an object. The metaphor detaches the action from the things and applies it to or it in new things. If its precision is limited, so is the accuracy of all means of knowing. Its limits are those of language. We need not charge metaphor with falsity. The statement in which it begins is as true as the knowledge we hope to see it communicate. It might be possible-though beyond both my scope and that of this essay-to show that all accretions of knowledge take this form. This is how we learn "what things are," how we form categories. confuses that a This returns us to the third category mistake is dangerous because it betrays its viccategories. tim into taking one thing to be another. But metaphor, once we see it as analogy, does not threaten this at all. Rather than confuse one category with another, it demonstrates the identical relations obtaining within both. It does not invite us to act irrationally-as I would if believed my love to be a rose-unless we misread it. In fact, to define metaphor as category mistake is to confuse two categories, poetry and insanity. In "Creative Artists and Daydreaming," Freud commits this error. Hugh Ormsby-Lennon has eloquently sketched its history in an
COGNITIVE
335
METAPHOR
essay entitled "Metaphor and Madness."14It results from mistaking metaphor, "poetry is a divine (or other) madness," for simple equation. The real statement to be verified is more like this: that the relations among the objects of poetry are identical to those among the contents of consciousness that typify a madman. One could explicate the details of these relations, like those in the similar metaphors "a poem is a dream" and "a poem is joke." But the language distinguishes poems from dreams and jokes; nor need we confuse the poet-as poet-with the madman. Though Aristotle, conceiving of metaphor as analogy, has helped us refute the objections that first encumbered our search for cognitive metaphor, examples of the striking similarity between understanding metaphor and reasoning cognitively in general may clarify the question. Following the modern tradition, take as my first text a stanza from Donne's "Valediction": Our two soules, therefore, which are one, Though must goe, endure not yet breach, but an expansion, Like gold to ayery thinnessebeate. One could reduce this metaphor to B: "our souls are gold" or, more fully, "our souls are like gold." The simile suggests that the lovers share with gold some common property, and the obvious candidate is "high value." But Donne's treatment of the metaphor intrigues us in just this regard, that he bypasses the obvious meaning, taking up instead more arcane property of gold, its ductility. Reducing the metaphor to simile, we lose this. Taking the stanza as whole, we discover several sets of mutually opposing terms. Even before the last line, Donne establishes a kind of analogy: breach two soules one expansion (He also links the second and fourth terms by rhyme.) When we carry this information into the last line, we find three terms of the metaphor already given, and the fourth easy to deduce: breach
_(lead) gold expansion (In metallurgy as well as in commerce, gold is the opposite of "dull" when "beate.")15 we lead, which have moved from an analogy that straddles the line of metaphor to one that confines itself within the vehicle. To move back into
336
NEW LITERARY
HISTORY
metaphor, we must incorporate an antithesis that pervades the whole poem: "our soules" stand against others, those of "dull sublunary lovers." From this (as it cooperates with our first analogy) we conclude that two soules others' souls one ours This analogy resides within the tenor. But in such a chain of analogies, substitution becomes possible. Things equaling the same thing equal each other, and from these three proportional equations we can derive (among others) this fourth: others' souls ours
gold
Here we ought to notice two things. First, we have ended with a version of the initial equation (our soules gold), but have enriched it. The value of gold is complicated by the various contexts through which the metaphor has passed it. The juxtaposition of commerce and metallurgy suggests third context, that of alchemy. Alchemists sought to derive gold, the noblest metal, from base lead. In that system, again, gold has high value as compared with lead, but on less materialistic grounds than we could originally have seen. That a refined or spiritual value can replace a more obvious one is, of course, theme that unifies the "Valediction." This is how the line's metaphor fits the poem in which it appears. Second, the way we arrive at this conclusion or comprehend the several implications of the line-a way which our chain of analogies has imitated formally-resembles the chain of syllogisms, a sorites. We cognitive process of following should, however, probably distinguish Donne's perpetual play among the forms of logic from the real knowledge that underlies it. The cognitive process of the metaphor is not precisely the argumentative one it enacts and continues, but the one we perform in understanding it. "The Yachts," at which we have already glanced, yields more direct view of metaphor functioning cognitively. What strikes us most in the way the poem develops during its first eight stanzas is the which are nevertheless standard density of metaphors-metaphors and seem to be offered casually, ornamentally. The pervasive trope is personification. Williams populates his lines with figures half-realized by the language. He calls the "lesser and greater craft" on which spectators crowd "sycophant." "The sea is moody," the ocean beyond the "well-guarded arena" "ungoverned." The waves in the protected bay where the race takes place "lap" the yachts' "glossy sides, as if
COGNITIVE
337
METAPHOR
feeling for some slightest flaw." When, in the last three stanzas, "the horror of the race dawns staggering the mind" of the speaker, we see these half-hidden metaphors burst out into full light: Arms with hands grasping seek to clutch at the prows. Bodies thrown recklesslyin the way are cut aside. It is a sea of faces about them in agony, in despair ...6 As we follow the poem, we are watching metaphor blossom into grotesque vision, its cognitive potential realized as knowledge of the "horror" a social system can breed. I want to offer a final example to suggest direction these speculahe calls tions might take. It comes again from Williams, a short simply "Poem": The rose fades and is renewed again by its seed, naturally but where save in the poem shall it go to suffer no diminution of its splendor17 Though more could be said about the poem, want to concentrate on one small question: According to what principle does Williams make the apparently flat and innocuous phrase "shall it go" a line by itself? So much attention seems odd even when we see that syntactically it rules the poem's second clause. But the reason for this emphasis becomes clear as soon as we realize that the line necessarily embodies metaphor. Sessile roses can "go" nowhere. If they could move, they could not "go" to the poem, which is not place in the same way as a garden or vase. What "fades and is renewed again by its seed" is flower; but the "rose" that inhabits the poem is word. In retrospect, of pun (not the last in the poem) first discloses paradox. But the pun reveals the heart of poetry, and of language itself. The word rose, says the poem, serves as metaphor for the flower by naming it. This recalls statement that critics and others often make: that all language is metaphorical. While this has become nearly an article of faith, what we have seen of analogy suggests logical foundation. To we need only examine (as with retrieve the other terms of the any vehicle and tenor) the contexts to which Williams's two "roses" belong. One is an element in the domain of words, the other dwells in
338
NEW LITERARY
HISTORY
the realm of things. Language becomes the vehicle of vast metaphor. In English it comprises not two or four but about half million terms. On the other side we find the tenor of the metaphor in the system of things that constitutes the world. Once again, the operative word is "system" or "structure." Ever since de Saussure, linguists have assumed that word does not mean directly or inherently but by relating to other words within structure of differences. Saying that "language is metaphor," we imply that the relation certain word has to the system of all words is precisely the same relation certain thing (the one the word "names") has to the system of all things. This raises a further question, which I hinted at before only to decline it, as must do again: How does the bric-a-brac of things we call "the world" become a world, cosmos, system? possible answer is that our minds, our senses, and our language create that system by process one ought to call metaphorical. This answer would take its first authority from the Critiqueof Pure Reason, and it is the one suggested by the work of gestalt psychologists. But its testing must remain their work, however valuable the critic's evidence; it is not our central concern. Cognitive metaphor is analogical, and so far as we can express metaphor as an analogy, it is cognitive. Still, cognition does not account for all the functions and effects of metaphors. It may not even be the primary one. One of the least satisfactory defenses of metaphor holds that it reveals the true nature of particular thing, a nature ordinary language cannot reach. But if all language is metaphor, there is no "ordinary language" in this sense. Metaphor tells us the truth of things no better and no worse than other uses of language. The way it does seem to exceed the ordinary capabilities of language lies in service that partakes at least as much of emotion as of cognition, and above all unites the two. Some call this service "defamiliarization"; the original force of "realization" aims in the same direction. Metaphor not only lets us know, it helps us to know what we know: to understand. UNIVERSITY OF WASHINGTON
NOTES Aristotle Rhetoric (tr. W. Rhys Roberts) 3. 10. 1410b16. Is this sentence of Aristotle's metaphor? Are all propositional statements metaphors? Northrop Frye's distinction between "identification as" and "identification with," though it does not answer these questions, can inform thought about them. See Anatomy of Criticism (1957; rpt. New York, 1968), p. 123; and, more fully, "The Realistic Oriole: Study of Wallace Ste-
COGNITIVE
METAPHOR
339
vens," in Wallace Stevens: Collection of Critical Essays, ed. Marie Boroff, Twentieth Century Views (Englewood Cliffs, N.J., 1963), pp. 170 ff. Howard Nemerov, "On Metaphor," Virginia QuarterlyReview, 45, No. (Autumn 1969), 623. Also reprinted in Nemerov, Reflexionson Poetryand Poetics (New Brunswick, N.J., 1972). 3 Aristotle Poetics (tr. Ingram Bywater) 21. 1457b6. This has become the most common definition of metaphor, and covers the subject as far as many dictionaries, glossaries, and encyclopedias are concerned. use the terms tenor and vehicle for lack of better ones. They have a use in distinguishing the two contexts that are undeniably brought together, somehow, in metaphor. That the view of metaphor they imply needs refinement and qualification is part of this essay's contention. 5 See The Poetryof Ezra Pound (Norfolk, Conn., 1951), esp. pp. 87-88, but also chs. through 12. Owen Barfield, in Poetic Diction (Middletown, Conn., 1973), seems to start from the same point, as do other critics. The reader interested in the history of thought about metaphor cannot do better than to consult Warren A. Shibles, Metaphor: An reference list of those Annotated Bibliographyand History (Whitewater, Wisc., 1971). who have, one way or another, taken metaphor to be analogical can be found on p. 372. Aristotle Poetics (tr. Bywater) 21. 1457b7- 17. am indebted to Michael Mesic for much of this paragraph's substance. 8 W. B. Yeats, "Memory," in CollectedPoems of W. B. Yeats (New York, 1956), p. 147. 9 But I should say two things. First, "form" has a technical meaning-the nest a hare makes in grass-which can be missed. Second, if one misreads "Cannot but keep" as equivalent to "Cannot keep"-as Jon Stallworthy does in Between theLines: Yeats'Poetry in the Making (London, 1963, p. 18)-the poem adamantly refuses to make sense. 10 Some valid interpretations exceed others in plausibility for reasons external to the of an author's habitual themes, for instance. review metaphor itself-considerations of critical comment (by Edward Kessler, Joseph N. Riddel, and Lucy Beckett) on Stevens's "Valley Candle" demonstrates both the multiplicity and the convergence of plausible interpretations. Note that critics other than Simons use the phrase "radical metaphor" in very different way; see Shibles, Metaphor, p. 402. 11 See, for instance, "Cognition," EncyclopediaBritannica, 11th ed. 12 Aristotle Poetics (tr. S. H. Butcher) 21. 7. In this instance Butcher is more helpful than Bywater, whom have used elsewhere. His translation may be found in Hazard Adams, Critical TheorySince Plato (New York, 1971). 13 Just so, we may now be able to rescue simile as well as "simple" metaphor by including the former in the latter. Many similes deal, like metaphors, with relations, disguising their equations with a modest "like." The example from Donne's "Valediction" given below supports this point. So does Aristotle; see Rhetoric (tr. Roberts) 3. 11. 1412b34. 14 Hugh Ormsby-Lennon, "Metaphor and Madness," Etc.: Review of GeneralSemantics, 33, No. 3 (Sept. 1976), 307-18. 15 According to the Encyclopedia Britannica, 11th ed. (articles under the metals' names), gold shares with lead a high specific gravity and "is nearly as soft as lead." But "it is also extremely ductile; single grain may be drawn into wire 500 ft. in length, and an ounce of gold covering a silver wire is capable of being extended more than 1300 m." As for lead, "Its breaking strain is very small: a wire 1/10th in. thick is ruptured by charge of about 30 lb." 16 William Carlos Williams, "The Yachts," in SelectedPoems (New York, 1969), p. 72. 17 Williams, "Poem," in Picturesfrom Brueghel (New York, 1962), p. 39.
