Metaphoric Hypersignification, Metonymic Designification

Edward Jayne

by Edward Jayne
February 3, 1994

Published by The Centennial Review, Vol. 30, No.1 (Winter,1994), pp 9-32.

Perhaps the most neglected aspect of language is the interplay that takes place between memory and forgetting when we express ourselves in sentences. The meaning of words must be fixed in our long-term memories (Saussure's langue), but their fluent combination (parole) is mostly limited to the short-term memory and depends on our being able to foreground and then discard them to make room for others. Their steady syntagmatic accumulation as we use them is accordingly followed by a helter-skelter chaos of abandonments forced by their response competition with later words. In poetry the use of metaphor necessarily compounds this pattern, since it disperses and complicates the dynamics of signification. More thought must be processed in the same amount of time, increasing potential memory loss through response competition with later words many of which also impose metaphoric demands. Metaphor thus augments our attention and thereby sustains itself in our short-term memory, but without any particular success in penetrating the long-term memory. In fact, proportional to the range of implications expressed by metaphor, more is sacrificed to the limits of the short-term memory than for the ordinary transparent use of words.

In contrast, a literary text's "plotted" organization of experience based on Aristotle's unity of action tends to be simple and cumulative, organized in the long-term memory by both the accurate recollection of what has happened and the only marginally less accurate anticipation of an acceptable outcome. To this extent, a text's diachronic flow of words limited by the short-term memory plays a subordinate role to its more or less synchronic narrative unity that gathers itself in the long-term memory. Yet the medium for this narrative organization entirely consists of these words devoured by their use, and their transitory limitations afford a freedom that gives them a major role of their own in their relationship with narrative experience.

In this paper I shall try to demonstrate how this dialectic interplay between metaphor and formal organization, between memory and forgetting, and, more particularly, between the short and long-term memories bears more importance than generally recognized in defining literary experience. I want to show on one hand how metaphor and other figurative tropes overload signification with a surplus of connotations that draw attention to themselves for a few moments, but only to submit even more decisively to the process of forgetting. On the other hand, I want to show how the cumulative experience of a text accessible to the long-term memory brings closure that both "resignifies" and "designifies" its local metaphoric expressiveness. As a kind of terminal (or "end-stopped") metonymy, closure imposes a final signification that contracts and packages the text's earlier significations to guarantee at least their residual memory as an experience worthy of suspended disbelief. Metaphor effectively thickens signification, while closure subsumes its density to an appropriate sense of an ending. I argue that both are needed to reorganize verbal discourse as a recognizable literary achievement.

I also want to demonstrate that the referential aspect of literature exaggerated by metaphor tends to be highly indeterminate, as opposed to its structural aspect based on literary conventions linked with closure that tend to be characterized by a no less heightened sense of determinacy. The two literary devices--metaphor and narrative structure--are thus complementary, and any effort to eliminate either of them, as for example with postmodernism's imposition of metaphoric indeterminacy upon the plotted (hence metonymic) organization of a story, is likely to deprive it of an audience big enough to make a difference. Both functions are needed, letting metaphoric inclusiveness exaggerate the fullest resources of experience so closure can reciprocate by exaggerating a decisive outcome at odds with the uncertainties expressed by metaphor. Whereas metaphoric freedom evokes attitudes and impressions we cannot otherwise entirely acknowledge, the metonymic constraints obliged by closure subsume these attitudes and impressions to conventional literary "truths" whose message can be shared by a broad audience. Because of this interplay, I argue, metaphor and closure fall into complementary roles. Each distends experience after its own fashion, and the interaction between the two gives literature its unique appeal.

1. Words are useful only because they can be utttered, then swept aside to make room for other words. If a particular word could actually imbed itself well enough in the mind to be continuously active in one's short-term memory no matter what else is said ("lets/horse go/horse to/horse a/horse movie/horse"), its ultrasignification would necessarily impede communication; if all words bore a comparable effect, language would amount to little more than an unbearable jangle of concommitant verbiage. It is only because words may be retrieved from whatever zone of memory they inhabit (call it langue or Freud's preconcious), then used and set aside, that language succeeds as communication. Space must be created for the steady flow of new words, and this depends on forgetting earlier words, necessarily limiting them to a transitory role. Like a muscle, words may be flexed, but only to be relaxed so they may be retrieved and flexed again on other occasions. Repeat the word horse as rapidly as possible twenty or thirty times, and it temporarily loses its symbolic value as a word. But give it enough of a vacation to recover itself, and within, say, thirty seconds it regains its power of signification with remarkable elasticity. Every word is consequently both memorable and forgettable--memorable in the sense that it may be recalled time and again, but forgettable in the sense that each time it is recalled it may be set aside soon enough to accommodate the accretion of new words and word combinations.

What primarily impedes the retention of words in the context of a sentence is their response competition with other words as they are uttered in sequence.[i] As new words keep accumulating, earlier words lose their immediacy, scramble for retention, and soon enough--usually the time it takes to utter not more than a dozen or two successive words--are swept into oblivion. True, some words are transferred to the long-term memory, but most are forgotten along with the specific context in which they were spoken.[ii] The neat syntagmatic precision by which words succeed each other, like a string of pearls or the steady drip of a faucet, is thus offset by the rank confusion that ensues in the process of their being forgotten. The "swath" of forgetting that occurs as words are abandoned is predictably chaotic. Some words are forgotten almost immediately, others linger a while in the memory; some disappear individually, others in platoons, often with one or two reluctantly slipping from the memory only to precipitate the headlong flight of others closely associated with them. Those few hardy words that clamber into the long-term memory are more likely to be remembered as concepts than as significations, as demonstrated by our need to draw upon a different assortment of words to explain what has been said. Asked to repeat a passage from memory, we more often than not paraphrase it with other words that roughly express the same idea.

We are so accustomed to what happens when words focus our attention, then disappear, that we find it difficult, if not impossible, to keep track of what is happening. Too much is going on, and meaning's incessant demands obscure how the words already expended make their exit from the short-term memory. Moreover, the contrast between the reception and abandonment of words seems utterly commonplace--too immediate, too personal, to be granted theoretical status. Nevertheless, the process can be studied with some degree of accuracy. As I have elsewhere tried to demonstrate, the dynamics of short-term retention and forgetting can be measured by giving groups of subjects--for example an entire classroom of students--the task of transcribing all the words they can remember from a recited passage up to an arbitrary cutoff point. Their transcriptions can then be graphed to calculate the frequency of those words retained by the group as a whole.[iii] What results is a concave horizontal figure with elevated primacy and recency effects (the latter more pronounced than the former) which respectively measure the improved memory of words at the beginning and end of the recitation. More important, an asymptote (or bathtub-like concavity) between these two elevations is filled with a variety of spikes representing the comparative salience of particular words and word combinations that are more or less resistant to memory loss. By experimenting in this fashion, I was able to measure the forgetting of particular words on what might be described as a micro-longitudinal basis by graphing a group of subjects' mnemonic response to a particular passage on three separate occasions with successively later cutoff points, each twenty words after than the last.[iv] The procedure was necessarily imperfect, but it let me map with rough accuracy the speed and sequence by which particular words were forgotten by the entire group of subjects.

I found most subjects able to remember between one and two dozen words well enough to be able to transcribe them to protocols. As to be expected, the so-called recency effect guaranteed a high level of accuracy over the duration of from five to eight words preceding the arbitrary cutoff points. But most words quickly disappeared from the short-term memory once they no longer benefited from the recency effect. Only a few bore any evidence of retention on protocol samples with cutoff points eight or ten words later. Concrete nouns fared well in resisting this memory loss, as did simple adjectives, active verbs, metaphors, rhymed words, striking images, and striking expressions. If more than a couple of normally salient poetic devices (for example rhymes and metaphors) converged in a passage, for example in a Shakespeare sonnet, those that were even marginally less salient seem to have been forgotten more quickly than they might have been in a different, less cluttered passage. As a rule, images and concrete expressions outlasted abstractions, hyperbole outlasted understatement, rhymed words outlasted unrhymed words, stressed words outlasted unstressed words, and the weird and bizarre outlasted the ordinary. But of all the words and poetic devices that stretched the short-term memory, very few, if any, could match names and simple ideas (e.g., King Lear is angry now) in crossing the threshold into the long-term memory. Most played their part, then disappeared, in contrast to the cumulative organization of experience that dominated the attention of readers as the story being told.

2. More than any other literary device, the metaphoric use of words seems to have stretched the short-term memory without gaining entry into the long-term memory. Its fresh signification imposed demands upon the reader's attention that repeatedly put it at the threshold of the long-term memory, but without entirely penetrating it. Why? It was my impression during my experiments that metaphor improves the temporary retention of words resulting from its use of an image, gestalt, or eidetic configuration to signify something else--either another word or an experience seemingly inaccessible to words. Additional effort is needed to grasp the intended meaning, since it transfers the function of signifier from the word itself to the image it represents.[v] However, this image bears connotations that can only be felt, then abandoned, as the reader's attention is drawn to new words and word combinations. Too much is happening in too brief a moment to be absorbed by the long-term memory, so the challenge of metaphor both guarantees its local salience and restricts it to the short-term memory.

As explained by I. A. Richards, metaphor's mental image (or "vehicle") conveys a fresh and different meaning (its tenor), and, as explained by Max Black, this meaning affords a new "focus" that is uniquely relevant to a "frame" comprising its relationship with nearby words and ideas.[vi] Since the image signified must be probed to grasp what is being said, more effort is needed to fathom its metaphoric connotations, and the attentive reader does in fact make this effort, as demonstrated by metaphor's improved retention in short-term memory tests. However, in most instances there does not seem to be enough time to fathom metaphor's fullest connotations, so it both entices and frustrates one's effort to grasp what has been said. More conscious activity occurs when metaphor is used, but under the pressure of response competition with nearby words and word combinations, less is discerned proportional to the range of additional connotations to be taken into account. Paradoxically, poets exert greater control of their words relevant to the subtleties they want to articulate, but less control relevant to the flood of connotations suggested both to themselves and their readers. As a result, more effort is needed to fathom the significance of metaphor, but the reader's interpretation necessarily remains tentative and without a final sense that metaphor's connotations have been fully grasped. Metaphor is therefore more salient than ordinary non-metaphoric words, but more vulnerable to forgetting proportional to the attention it draws to itself. It dominates consciousness for extra moments but seldom crosses the threshold into the long-term memory.[vii]

Of course an enormous literature has accumulated to explain metaphor, but to clarify the dynamics involved in its memory loss I want to propose here an expedient and relatively simple model combining I.A. Richards and Max Black's theory of metaphor with Roland Barthes's model of literature as a second-order semiological system.[viii] In what amounts to a double-order signification, the metaphoric use of words signifies an image (vehicle) which helps to depict words and ideas nearby. In itself this image is clear, but the new level of experience it signifies is often difficult--if not impossible--to decipher in its entirety. Its meaning should of course be obviously relevant to the context in which it is spoken, but there is room for the leakage of additional connotations that elude quick interpretation. A double el (or elbow) occurs that may be diagrammed as follows in linking Saussure's paradigm for the sign (stood on its head) with Richards's paradigm for metaphor to represent the tandem function of metaphor's image as both signified and signifier:

The two levels of signification indicated here, the signifier/signified interaction of the verbal sign and the tenor/vehicle interaction of metaphor, bear comparable vertical functions respectively depicting plain and figurative signification. Compounding the dynamics of signification, metaphor imposes an image as a second-order sign context to convey more elusive connotations relevant to the passage as a whole.[ix] The entire first-order sign context, including both signified and signifier, thus catapults signification into the second level well enough for the signified in the first instance to become the signifier (or vehicle) in the second as explained by Richards's tenor/vehicle distinction.

For example, in the phrase, "a Harris Tweed cat" (borrowed from George Whalley's useful article on metaphor in the Princeton Encyclopedia of Poetry and Poetics), the words Harris Tweed convey an image of this tweed's color, texture and herringbone design that signifies a particular cat by featuring these as its most salient traits. A doubling thus occurs, since the image of the Harris Tweed fabric serves as both Saussure's signified and Richards's vehicle whose tenor consists of the felt experience of a particular cat as summed up by its coat's resemblance to the Harris tweed. Ironic detachment is probably suggested on the part of the speaker as clarified and reinforced by other words used in the same context: e.g., "Their Harris Tweed cat seemed intended to complete the living room's decor." The metaphoric connotations of such a remark may be conveyed primarily because the original signified (the image of a Harris tweed) becomes a signifier, and the transaction between these two verbal functions (as depicted by the intermediate equivalence in the paradigm) provides an eidetic core--in effect a figurative relay terminal--that both receives and dispenses signification. The habitual signified/signifier interaction typical of normal discourse expands to become a more challenging tenor/vehicle interaction that expresses a much fuller and more indeterminate range of connotations typical of poetic discourse.

In the paradigm I am proposing, a double elbow becomes operative (like moving from second to third gear on a stick shift) that compounds signification and obliges greater attention to fathom its implications, as indicated by the improved short-term retention of metaphors on protocol averages. With dead metaphor (for example, "to kill some time"), the intermediate signified/vehicle linkage is effortless, so the signifier (the word kill) automatically completes its circuit in representing its tenor (here, to "waste" or "use up"). One seldom thinks of homicide when using the word kill in this fashion, so the image immediately relays its meaning without intervening specifically as an image. But with live metaphor--and directly proportional to its life as metaphor--the intermediate linkage prevails, emphasizing the word's meaning as an image with implications that cannot be quite so readily contracted to an abstract meaning, or even to a simple analogy such as Aristotle proposed in Poetics.[x] With live metaphor the impact upon the short-term memory is different, but entirely predictable. On one hand, by energizing the connection between its first and second order referents, metaphor's image temporarily resists forgetting, since it dominates more than its share of the reader's attention. On the other hand, the experience signified does not occupy enough of the reader's time to be adequately interpreted relevant to its additional connotations implied by its image. As a result, metaphor is felt but forgotten, sacrificed to the ongoing necessity of processing new words. On an absolute scale, metaphor draws more attention to itself than the non-metaphoric language that surrounds it, and for this reason it tests the threshold of the long-term memory. However, relative to its fullest implications, it is also more vunerable to the dynamics of forgetting--there is more to be forgotten, and it is. Good metaphor splashes on consciousness with undeniable impact, but what it signifies is experienced without being clearly understood in its most inclusive ramifications. It is vividly brought to the fore, but only to be abandoned with even more of a sense of loss in the ongoing flow of new words.

The memory loss is probably negligible for dead metaphors, for example in the word combinations, "a horse-faced" individual, or "to horse around," the latter making a verb of the noun to imply a bumbling youthful playfulness. By itself, the word combination "to horse around" signifies pretty much the same as to play around or to kid around, so connotations are minimized that might augment the burden on the short-term memory. However, metaphoric overdetermination begins to be felt if the word combination is slightly expanded, for example in the warning, "Don't horse around with other women," possibly evoking the image of horses copulating or chasing each other in the field. This image is very likely reinforced if other words are added to extend the metaphor, for example in the sentence, "He is now harnessed, unable to horse around with other women." If still other words are added with comparable implications, a metaleptic complexity energizes the entire passage with an extra signification that might not readily be traced to the literal definitions of any of the particular words and that is more likely to draw upon the subjective experience of particular readers. This may be seen, for example, if a pun is added:

Jason bridles at his marriage--harnessed, unable to horse around with other women.

Here bridle could imply both the acceptance and rejection of marriage, and the words harnessed and horse around accordingly take on slightly different implications based on this choice. An additional spectrum of connotations presents itself, more suggestive than if each of these words, horse included, is interpreted alone. There is "hypersignification" in the sense that for the passage as a whole metaphoric overdetermination dominates the original sign situation. A wider, more expansive meaning suffuses the context of expressiveness typical of literary discourse, and this affords a kind of literary truth that transcends the raw factual content of nonliterary discourse. But, it turns out, this augmentation makes the word combination more vulnerable to forgetting if the reader's attention continues to be engaged in trying to fathom new hypersignifications in the context of a longer passage. For each metaphor there is momentary eidetic impact, but as with a steady succession of skyrockets on the Fourth of July, metaphors dissolve in their own glorious manifestation, replaced by others no less glorious and no less transitory. The threat of metaphor's elusiveness, that its penumbra of connotations remains opaque, thus "open" to vagrant and often dangerous expressiveness, is also, paradoxically, its advantage, that it escapes analysis (and potential censure) by being absorbed in the ongoing flow of words without being fully comprehended. It is therefore entirely to metaphor's benefit that it "images" ideas without elaborating them--letting them be felt, then forgotten, neither fully acknowledged nor adequately explained.

The same effect, diagrammed by the same double elbow, may also be produced by simile, by puns, by ambiguity, and by symbolism, in the latter instance through the cumulative signification of certain recognizable objects or relationships to suggest the ideas and feelings commonly associated with them. Whenever the signified usurps the role of signifier, a metaphoric interplay may be discerned that stretches immediacy at the expense of the long-term memory. Comparable doubling may be produced by ambiguity, for example, when two or more ideas are signified at once, or by irony when one idea is declared to signify another that is very likely its opposite, as to be recognized by the person(s) addressed. And finally, the same effect can be produced by localized metonymic constructions when parataxis brings to the fore associations independent of their syntactic context.[xi] For example, when one declares, "love blossoms--my horse gallops the plain," the cryptic linkage between these two apparently disconnected events may be resolved by assuming that the behavior of the horse implies what might be meant by the blossoming of love. For the purposes of this essay, I include all these literary devices under the category of metaphoric hypersignification, since they may be grasped as doubled references to focus the reader's attention on additional connotations within a relatively limited context smaller than the text as a whole.[xii] Often with symbols and image clusters, the intermediate image or idea that links the two levels of signification draws attention to itself well enough to be less vulnerable to forgetting than with metaphor. However, the double elbow still functions to the extent that it thickens expressiveness by bringing fugitive implications into play only to be overlooked because of response competition and the momentum toward a suitable outcome that dominates the long-term memory.

3. Metaphor usually features an experiential immediacy resistant to narrative movement except as a bogged-down syntagmatic pleasure cruise among new figurations each of which reluctantly gives way to its successors. But this metaphoric density is only one aspect of literary experience, and it is necessarily offset by countervailing literary dynamics that steer consciousness toward an appropriate resolution. The sense of praxis, of something happening, of results to be achieved, plays a countervailing role that seems at least as important in fiction. Most obviously, this may be found in the story "plotted" toward an outcome, but it also occurs in the mounting complexity of characterization, in the recurrence of themes, motifs, and symbols, in the heightening of intentionality, and even in the transparent denotative signification of words and sentences that articulate the circumstances soon to culminate in acceptable closure. In all instances conventions and literary devices are engaged that both whet and fulfill the reader's collaborative participation in the story being told, for, indeed, every text is primarily a "story" with a narrative organization of experience.[xiii] Experience, feelings, and mounting expectations all converge in leading to a final effect that imbeds itself in the long-term memory as a "what happened" that makes a difference. Also important is the role of syntax --of sentences completing themselves--for, as has been argued by Paul de Man, there is dialectic interplay between the grammar and rhetoric (i.e., tropes, or figuration) of literary experience.[xiv] More inclusively, however, this interplay also pits rhetoric against itself, with metaphoric hypersignification offset by linear metonymic dynamics that bring a text to its appropriate resolution by imposing an "after" that signifies (or resignifies) its "before." Literary conventions, for example, are metonymic in the sense that they compel an appropriate resignification, as when villainy anticipate retribution, when romance and buffoonery anticipate the happy resolution typical of Menandrine comedy, or, at a simpler level, when a striking image or phrase prefigures a future turn of events. Most of these conventional expectations oblige readers to draw on their long-term memory to piece together the information conveyed by words as a text moves forward through time. Everything retained in the memory somehow augments the reader's collaborative participation, and by means of narrative outcome the fulfillment of this sharing effectively signifies all that has preceded.

The metaphoric thickening that impedes the reader's interest in a text's forward momentum is therefore counterbalanced by mounting interest in events, places, and characters whose transparent significations readily cross the threshold into the readers' long-term memory. These significations accelerate narrative momentum by both heightening the readers' sense of participation and drawing their attention forward to a suitable resolution, an outcome (or end-stopped metonymy) whose summing up declares a simpler and more acceptable version of experience. Any text with the reputation of being a "good read" necessarily yields its local significations to this metonymic push toward acceptable closure. Even in postmodernist fiction in which closure is supposedly banned, readers themselves try to impose a sense of an ending by weighing the story's culminating experience to see if it somehow provides a suitable resolution. This itself affords at least an abridged sense of closure acceptable to all but the most doctrinaire postmodernist true believers. As a virtually inevitable feature of fiction's second-order process of signification, closure both fulfills the cumulative meaning of earlier significations and reinterprets them by imposing a sense of an ending that is better and more desirable. Metaphor opens (i.e., expands or thickens) signification, whereas closure constricts and simplifies it, first promising and then delivering a conventional outcome that fully engages the long-term memory. This doubling may happen within the compass of a sentence or two, or, more inclusively, as the dominant organization of experience over the duration of an entire text, for example when Aristotle's unity of action brings a plotted reversal from one state of affairs to another which is appropriately its opposite--e.g., from indecisiveness to bold decision-making, from potential failure to success, and, more inclusively yet, from tension to pleasure.

For such a use of metonymy, the same double-elbow occurs as with both metaphor and the local use of metonymy already described. However, closure (an end-stopped metonymic operation) concentrates the reader's attention forward in time, i.e., on the horizontal axis of combination instead of the vertical axis of selection, and it dominates as the final impression of a text that cannot be superceded by other significations in their continuous advancement among themselves.[xv] Relatively free of response competition, it bears a final impact that effectively lingers in the memory as a summing up of all that precedes. Since Roman Jakobson's seminal paper, "Two Aspects of Language and Two Types of Aphasic Disturbances," most theories of metonymy have featured its potentially endless linear advancement among befores that incessantly anticipate their afters, or, with inverted but equivalent effect, afters that incessantly resignify their befores.[xvi] The dynamics of signification in any before/after metonymic sequence necessarily imposes at least temporary closure, since the after unavoidably gains its additional salience as a signification of before. And since the final meaning of a text imposed by closure is judged by whether it gathers in everything preceding it and asserts itself as a culminating signification, it may be described as being metonymic--not in the sense that its postponement may be indefinitely extended, as Jacques Lacan has proposed, but in the sense that it ends a text by revising and thereby resolving all its earlier significations. Lacan confessed his indifference to this function of narrative closure in his celebrated remark, "This signifying game between metonymy and metaphor . . . is played until the match is called, there where I am not, because I cannot situate myself there."[xvii] Here Lacan acknowledged his disdain for closure (and probably for literary experience as well), but he was not denying its importance for most readers as an end-stopped variety of metonymy, and in fact this use of metonymy has traditionally played an essential role in most literary experience.

Thus the importance of the dialectic interplay between metaphor and metonymy. For if metaphor is a more or less instantaneous doubling of signification based on a vertical use of the double elbow I have diagrammed above, end-stopped metonymy plays its complementary role as a postponed (or stretched out) doubling that resignifies one or more earlier significations. It need not be limited to a one-to-one relationship with a particular antecedent, but may extend the dynamics of signification to include many befores up to and inclusive of the text as a whole. As compared to metaphor, which focusses, then diffuses, signification by its reliance on images on the vertical axis of selection, metonymy gathers and resolves preceding significations on the horizontal axis of combination, and with its final act of resignification closure provides an eidetically coherent outcome that "re-images" all that has preceded. Transparency prevails in the appropriate sense of an ending that has long been anticipated, but its achievement, paradoxically, is through rhetorical doubling, earlier metaphor (intrinsically opaque) as revised by metonymy (lucid, hence retrospectively opaque). Here, like the negative sign, opacity cancels itself by its doubling: not (not x) = x. The opaque is "opaqued" for lucid results.

The active ingredient that vitalizes and gives aesthetic appeal to this resignification is usually a reduction of experience that amounts to designification, the imposition of a simpler, more tangible referent, the conventional sense of an ending that appeals to most readers.[xviii] The connotative density of the text in its forward momentum toward closure is overridden by a focussed theme or impression that calls attention to itself as the summing up of all that has preceded. There is a substitution of one plateau of experience for another which is its contrary--transparent instead of opaque, and resolved instead of loosely suggestive. Denial occurs in the sense that an earlier susceptibility to implied possibilities has been superceded, in effect reversed, but this denial (or designification) is best served by the affirmative effect of closure as a culminating vision of human potential acceptable to most readers. Results are positive, but relative to what has preceded something negative occurs--a cancelling has been obtained through the subjugation of alien connotations expressed or implied by metaphoric hypersignification. What metaphor articulates in the welter of local effects is first jostled from the memory because of response competition, and then its very trace is denied by metonymic designification. While metaphor tends to encourage expressiveness, the continuing flow of words obscures the fullest connotations of this expressiveness well enough for metonymic designification to package and sanitize the reader's final impression of the text. When closure finally happens, its signification might also be steeped in metaphor, but its end-stopped metonymic effect draws the reader's attention to itself with such additional impact that the full meaning of earlier words and metaphors become little more than a penumbra of fascinating connotations that cannot quite penetrate the threshold of memory.

Metonymic designification is the most dramatic when narrative resolution emphatically reverses or contradicts earlier significations at odds with its final effect. In such cases, the Aristotelian unity of action transcends its linear transition from beginning to middle to end, and instead becomes an organic manipulation of experience whose end consists of revising its origins through designification, and whose middle consists of the cumulative metaphoric expressiveness that makes such an outcome desirable. As the agent of metonymic reversal, plot is offset by metaphor's resistance to closure, and the two are pitted against each other to afford a more inclusive unity of action. A simple dialectic thus imposes literary form by steering the reader's experience into closure, both giving meaningful context to earlier hypersignifications and obliterating their memory through metonymic designification. For example, erotic risk as expressed by metaphor becomes little more than a memory of vague connotations preliminary to their thematic denial by the declaration of sentiments of chastity and Platonic rapport typical of both love poetry and the stories of chaste male and female bonding that persistently crop up in today's popular media.[xix]

4. The extent to which metaphor and metonymy depend on each other cannot be exaggerated. Metaphor assimilates fugitive experience, while end-stopped metonymy both consolidates and packages what has been assimilated. Metaphor stretches the signified at the expense of its signifier, thereby impeding (at least slowing down) one's collaborative participation in the forward momentum of the text. In contrast, closure's function as end-stopped metonymic designification stretches the signifier at the expense of its signified--the text as a whole preliminary to closure--by imposing a relatively simple outcome that both dominates the long-term memory and overrides preceding metaphoric distractions. Metaphor diffuses meaning in the motion (or detour) of the reader's attention from the signifier to its signified, while the anticipated metonymic signifier imposed by closure focusses and reduces the meaning finally obtained. Metaphor expands perception, followed by the use of closure to reconfigure expanded perception by an acceptable remembered outcome. A sort of doubled diffraction occurs, as when light passes through glass between two surfaces not necessarily parallel with each other. Metaphor bends outward and away from; closure bends inward again. As Northrop Frye might have argued, metaphor plays a centrifugal role by drawing upon tangential associations, whereas metonymy's role is centripetal by restoring one's focus on a text's motion toward closure as a determinate organization of experience. A story or poem may be predominantly metaphoric (front-loaded) or predominantly organized by the anticipation of closure (back-loaded), but it seems the most effective when it sets the two against each other on a more or less proportionate basis, with sufficient metaphor to expand consciousness but with sufficient metonymy to pack it back in again. Ideally, the feelings expressed by metaphoric hypersignification should demand, yet resist, closure, and closure (both as expected and as achieved) should help both to justify and undo their hypersignification by bringing a what-finally-happens that reduces its what-might-have-been-implied to mere forgotten suggestiveness.

Whereas the simple telling of a story features a transparent first-order verbal system borrowed from non-literary discourse, the interplay between metaphor and metonymy features a more elastic second-order organization of significations that guarantees the story's literary value. The Dionysian (polymorphous, or, as Camille Paglia would have it, chthonian) primary-process extravagance of metaphoric hypersignification is brought under control by Apollonian order, the convergence of theme, outcome, and literary "truth" that gives story its appeal with most readers. Metaphor is mostly limited to the short-term memory as the transient expression of latent connotations some of which are too unpleasant (too honest, too threatening) to be directly acknowledged, while closure temporarily overwhelms the long-term memory as a determinate organization of experience which guarantees satisfaction. Closure becomes all the more gratifying when it lets one cope with metaphor's spectrum of puzzling and potentially threatening experience; in turn, metaphor gains almost unlimited freedom of expression--with cathartic value of its own --because closure can be expected both to justify and override its rampant suggestiveness. By its doubled reference (image as sign), the metaphoric dimension reaches out to incorporate vagrant experience into the literary context--some at the very brink of consciousness--and by doing so it brings this breadth of experience into the normal use and forgetting of words. Then by imposing closure, the metonymic dimension subsumes this experience to conventional expectations that acceptably package it for the normal mind.

The same double-el (signified/vehicle) transaction is generic to both metaphor and closure (i.e., end-stopped metonymy), but it is put to use for different and basically complementary functions. Whereas the image (or eidetic core) of metaphor gathers, then disperses, signification, closure's metonymic impact as the text's final impression--likewise with an eidetic core--culminates the activity of signification by gathering everything that has preceded into a final indelible impression of what has happened as an event deserving of resignification. Like closure, metaphor is brought to focus--in its case by its image--but then it is diffused by its connotations and swallowed up by its competition with other words and metaphors nearby. Like metaphor, closure offers an accessibility to idiosyncratic associations, but it draws on these associations in order to revise them, rejecting (in effect denying) whatever lacks reinforcement in the long-term memory. The two thrive in their complementary roles. Like Laurel and Hardy, they exasperate but depend upon each other. Metaphor begs to be overridden, while closure gladly overrides it to justify its having been expressed. The interplay between them keeps the text at the edge of consciousness, both sane and exploratory, both extravagant in its associational versatility and conventional in the choices it finally makes.

Metaphor and closure are comparable in the sense that they both compound signification, but they serve different and essentially complementary functions. Everywhere, in every context of literary experience, the two must therefore be understood to offset each other. Metaphor disperses signification to be resolved by context, while closure guarantees context to help resolve the significations already dispersed by metaphor. Metaphor affords referential multiplicity in the thick of competitive significations, while closure terminates signification with simplistic finality by resignifying (and therefore designifying) this multiplicity. Metaphor is primarily visual (or eidetic), while closure provides a more sweeping convergence of theme, affect, and the conative fulfillment of expectations. Metaphor verges on being confessional, while closure voices conventional public certitudes. Metaphor renders the experience signified more vivid, yet more forgettable, while closure reinforces this memory loss by all the more effectively designifying metaphor's vagrant implications at odds with the final impression authors and readers find acceptable. The most marginal and potentially embarrassing expressiveness may accordingly be acknowledged in local contexts only to be denied by the narrative organization of experience that brings an acceptable outcome. Metaphor tends to be subjunctive in the sense that it features both elusiveness and vivid suggestiveness, while closure, in the declarative mode, features an affirmative summing up whose simplification supercedes and therefore rejects hypothetical possibilities. Metaphor scatters and multiplies signification, while closure reduces it to unity; metaphor encourages extravagant flights of primary process expressiveness, while closure, as reinforced by literary convention, draws the reader's attention forward to a socially acceptable outcome; and, finally, metaphor disorients experience, while closure reorganizes it, the one by exposing a mood or predisposition to be resolved by imposing the other.

Metaphor tends to be indeterminate, since it encourages intuitive consciousness whose truth is difficult to fathom with any degree of exactitude, while its metonymic denial through closure tends to be determinate, since it imposes a clear-cut vision of the truth, gratifying but too simplistic to be entirely true. Metaphor features the inchoate and apparently inexpressible resources of language, to which closure affixes a well-defined and appropriately literary finality. Through metaphoric indeterminacy, a greater diversity of interpretation becomes possible among both readers and individual readings, as explained by the notions of aporia, "freeplay," and "interpretive acts." But through metonymic determinacy as reinforced by literary convention, this diversity must inevitably be squeezed and brought to focus by a collaborative participation in the dynamics of closure that unites readers as an audience in agreement with their authors.

Wolfgang Iser and others have suggested that indeterminacy is the product of textual omissions (or "gaps") that permit the reader to draw upon personal associations independent of the text itself. Here I am proposing the antithetical thesis that the freedom of the reader to draw upon these associations primarily derives not from any real or imagined hiatus, but from the density of connotations typical of metaphoric hypersignification. Thickness (Freud's overdetermination)--not any kind of a void--declares subjectivity, and metaphor is the primary vehicle of this thickness in literary expression. As earlier indicated, this referential overdeterminacy produces a paradoxical effect, that authors gain more freedom in communicating their experience, but that readers likewise gain more latitude in measuring this experience against their own. For both, metaphor expands the access to individual consciousness, hence the personal idiosyncracy of any particular reader's response to the text. However, this rampant subjectivity is eventually superceded by literary conventions which impose a relatively uniform audience response--very similar to that of a partisan crowd at a sports event--by steering this expanded consciousness forward to what might be perceived as a satisfactory outcome. The indeterminate candor of metaphor is thus supplanted by closure's pseudo-determinate conventional "truths" that can be shared by authors with an expanded audience.

Of course metaphor and closure feature alternative versions of the truth which differ from the warranted truths of non-literary discourse that take into account logic, descriptive accuracy, probability, the role of exceptions, etc.[xx] Closure's truths are usually variants of the grand and empty certitudes that all turns out well, that relationships can be made simple, etc. In contrast, the opaque and indeterminate freedom of metaphor features more subversive truths through its acknowledgement of feelings otherwise inaccessible to language. However, its confessional integrity is first jostled from the memory in favor of the text's majority of transparent, hence false, significations, and then denied by closure's conventional outcome as an after that contradicts its before.[xxi] What justifies closure's misrepresentation of raw, non-literary truths (for example, that retribution seldom occurs, that virtue too often expresses timidity, that most of our lives are in a rut and we're glad of it, etc.) is its role in granting a voice to metaphor's indeterminate truths that would otherwise be altogether lost to the reader. The two complement each other in their literary reorganization of experience, one by its candid enlargement of expressiveness, the other by its prudence in framing this enlargement based on acceptable aesthetic constraints. They thrive by feeding on each other, each invigorated by its dialectic interaction with the other. Only because of Shakespeare's conventional endings, for example, does his metaphoric honesty gain its hearing; but, then again, only because of his metaphoric honesty do his conventional endings deserve our suspension of disbelief.

Metaphor, metonymy, and closure have usually been treated as independent rhetorical strategies, separate devices that perform separate functions in the separate genres of poetry and fiction. But as specific modes of signification, they tend to fall into complementary roles in virtually all the literary genres--metaphor understood in its most inclusive sense as a preliminary function of stretching and giving breadth to verbal expressiveness, and closure as an end-stopped version of metonymy that accents conventional "truths" at the expense of too much breadth of experience and the inclusion of too many fugitive implications. In primarily metaphoric texts the dynamics of ambiguity emphasized by William Empson predominate; in narrative texts dominated by literary convention the archetypal dynamics emphasized by Northrop Frye predominate. But both play a part, and neither is entirely effective without a sufficient trace of the other to catalyze the experience of literature. Six categories may accordingly be proposed for combining these two primary features of literary discourse in what amounts to a hierarchy of literary effect inclusive of all works of literature:

1.The plain style that lacks both metaphoric hypersignification and metonymic designification except on an obviously random basis. This discourse possesses the potential of unending significations as illustrated by Pepys's Diaries and the Journals of Andre Gide. With enormous skill, Frank O'Hara's poetry tests the limit of this category and yet contains a sufficient trace of both metaphor and closure to remain valid in its poetic expressiveness.

2.Discourse that persistently hypersignifies, but without adequate closure. The poetry of Ammons, Ashbery, and Merrill is often successful in testing the limit for this particular category.

3.Discourse that brings closure without preliminary hypersignification, as, for example, in pulp realism and the narrative verse of Longfellow and Masters.

4.Discourse with both closure and hypersignification, but with no clear relationship between the two. This category lacks Aristotle's unity of action, and it may be expected in run-of-the-mill creative writing assignments.

5.Discourse with both closure and hypersignification but with metaphor limited to the supportive foreshadowing of closure. To a certain extent all good fiction taps this process of anticipation, but any text that exclusively depends on it--for example with tantalizing sexual images that culminate in the act of seduction--tends to fall short of the fullest literary achievement.[xxii]

6.And finally discourse that sets metaphoric hypersignification against its metonymic designification. Metaphor draws upon otherwise inaccessible experience, but only to be revised and superceded by means of an acceptable resolution.[xxiii] Shakespeare, Milton, and the romantic poets best illustrate this synthesis, though it may be found in the best work of most major authors.

When both metaphor and closure are entirely missing, signification is non-literary; and when one or the other is missing, signification is only partially literary despite the claims of purity, mimetic honesty, or greater audience appeal--all of which have been made at one time or another. When the two are independent (as in category four) or totally in harmony with each other (as in category five), literary experience does take place, and often with genuine aesthetic effect, but also with a simplified design perhaps the most appropriate for popular audiences. Only if there is adequate dialectic interaction between the two can literature be appreciated at its most challenging level. One or the other might be favored, for example in the poetry of Asberry and Merrill (primarily metaphoric through a localized use of metonymy) as opposed to that of Longfellow and Masters (primarily metonymic with an end-stopped emphasis). But whatever the author's choice, both metaphor and metonymy should be felt well enough to guarantee the text's literary value as a second-order intrareferential achievement that transcends the ordinary use of language.

Footnotes

[i] Terms such as response competition, decay, primacy effect, and recency effect are commonplace in memory theory. My use of them is based on their explanation in Human Memory: Theory and Data, by Bennet B. Murdock (New York: John Wiley & Sons, 1974).

[ii] Without response competition, simple "decay" in the short-term memory of words is slow enough to improve the prospects for long-term memory retention. For example, words uttered immediately preceding a period of imposed silence --say, a thirty second interlude devoid of words--can be remembered more easily, at least until the flow of language resumes. However, most words are both preceded and followed by other words, so they are more likely to be forgotten.

[iii] "Psychostylistics: The Possibilities of a Behavioral Science," Style 18, n. 1 (Winter, 1984): 83-97.

[iv] The recitations were conducted approximately three weeks apart from each other without my disclosing to the group that the same passage was being used. I also recited other passages throughout this period to minimize the recollection of the passage I wanted to measure in this fashion, and of course the progressive advancement of its text in its readings augmented the portions already used with entirely new words. Of course all of these variables need to be taken into account in arriving at any final determination for the process of mnemonic decay I was measuring, but the pattern of attrition I was able to measure seemed predictable both for the group as a whole and for the "average" single member of it.

[v] "Psychostylistics," pp. 92-93, and my recent book, Negative Poetics (Iowa City: University of Iowa Press, 1992), pp. 111-13.

[vi] See I. A. Richards, The Philosophy of Rhetoric (New York: Oxford University Press, 1965), pp. 89-138; Max Black, Models and Metaphors (Ithaca: Cornell University Press, 1965); and Paul Ricoeur, The Rule of Metaphor (Toronto: University of Toronto Press, 1977), pp. 76-90.

[vii] The exception to this rule might be provided by those with near-perfect photographic memories. Harold Bloom, for example, once told the author he had been able as a college student to recite from memory Hart Crane's Collected Poems backwards word for word starting at any part of the text, and without having ever tried to memorize Crane's poetry. However, individuals blessed with this capacity are a very tiny minority of the reading public and may be discounted in trying to establish the actual dynamics that occur when the rest of us focus our attention on literature.

[viii] Roland Barthes advances his model of literary discourse as a second-order semiological system in Mythologies (New York: Hill and Wang, 1972), pp. 114-15. I am proposing here a tandem horizontal linkage between two essentially vertical models of representation: as a first-order system, Saussure's model for signification combining signified and signifier, and, as a second-order system, I. A. Richards's model for metaphor combining tenor and vehicle. This relatively simple model may readily be expanded to accommodate more elaborate explanations of metaphor, for example George Lakoff and Mark Turner's theory in their recent book, More than Cool Reason: A Field Guide to Poetic Metaphor (Chicago: the University of Chicago Press, 1989). Based on the double elbow proposed above, all metaphor is reducible to the "image" or "basic" metaphor defined by Lakoff and Turner, and the tenor/vehicle interaction proposed by Richards is necessarily implicit in the "schemas" Lakoff and Turner respectively map between "target" and "source" domains.

[ix] Here Saussure's sign may be profitably expanded as explained by the concept of a "sign situation" as explained in C. K. Ogden and I. A. Richards's The Meaning of Meaning (London: Harcourt, Brace, & Co. Inc., 1956), pp. 4-12.

[x] Aristotle's explanation of metaphor in The Poetics, XXI, includes its function as an analogy that combines four terms, for example by describing a cup of wine as the shield of Dionysus, since the cup of Dionysus (for drinking wine) is equivalent to the shield of Ares (for defense in personal combat). The analysis of metaphor based on a simple equation, in this instance between a cup and a shield, has encouraged numerous theories that box metaphor as a basic catachretical substitution of names, one for another. However, the impact of live metaphor derives from its creative use of a signifier that stimulates complex experience otherwise inaccessible to signification. Here metaphor, as it were, escapes the box.

[xi] Here my discussion of metonymy somewhat differs from Roman Jakobson's distinction between metaphor and metonymy in his seminal paper, "Two Aspects of Language and Two Types of Aphasic Disturbances," in Fundamentals of Language (The Hague: Mouton & Co., 1956). Jakobson associates metonymy with syntactic exaggeration as explained by the similarity disorder, counterbalanced by the "word heap" typical of metaphoric exaggeration as explained by the contiguity disorder. The usefulness of Jakobson's model cannot be denied, but the grammatical interaction he finds among words in their one-dimensional sequence on the horizontal axis of contiguity seems to be declared, not implied, by the semantic function of the words with which they are grammatically conjoined, as opposed to the purely metonymic interaction among words whose relationship isn't otherwise declared by either syntax or meaning. For example, in the sentence, "she bought candy because she was sick," the meaning of the subordinate conjunction because declares a cause-and-effect relationship between the two clauses. But if because is eliminated, for example in the paratactic construction, "She bought candy--she was sick," the absence of a declared linkage encourages the reader to explore associations between the two clauses more along lines of the tenor/vehicle interaction typical of literary figuration. Judged on this basis, syntax may be granted to be metonymic as opposed to the metaphoric emphasis of signification, but less metonymic than word combinations in which sequence alone dominates interpretation. Thus in essence paratactic freedom must be treated as being more purely metonymic than hypotactic grammatical complexity.

[xii] Jonathan Culler gives essentially the same priority to metaphor in The Pursuit of Signs: Semiotics, Literature, Deconstruction (Ithaca: Cornell University Pres, 1981), pp. 188-91--cited by Jill Matus in "Proxy and Proximity: Metonymic Signing," University of Toronto Quarterly 58, n. 2 (Winter, 1988/9), pp. 305-26.

[xiii] See Negative Poetics, in passim, pp. 67-85, for my treatment of Gertrude Stein's poem, "a rose is a rose is a rose is a rose" as a narrative organization of experience.

[xiv] Paul de Man, "Semiology and Rhetoric," Allegories of Reading (New Haven: Yale University Press, 1979), 3-19. De Man's model of literary experience is basically the same as the one I am proposing here once his category of rhetoric on the axis of selection is identified with figuration--or, more specifically, with metaphor--and once his category of grammar on the axis of combination is expanded to include all literary conventions that draw the reader's attention forward to an acceptable resolution, as I try to do by somewhat revising the concept of metonymy. Probably the most basic difference between De Man's model and mine is that his is almost entirely spatialized and does not take into account the interplay between the short and long-term memories as one reads a text from beginning to end. As a result, he depends on a broad and relatively cumbersome concept of deconstruction to explain the interaction between the rhetorical (i.e. metaphoric) and grammatical (i.e. metonymic) dimensions of literary experience rather than the more specific conscious dynamics that actually transpire when reading literature.

[xv] See Roman Jakobson, "Closing Statement: Linguistics and Poetics," in Thomas Sebeok, ed., Style and Language (Cambridge, Mass.: the MIT Press, 1960), p. 358.

[xvi] in Fundamentals of Language, by Roman Jakobson and Morris Halle (The Hague: Mouton & Co., Printers, 1956): 55-82. See also Leon Surette, "Metaphor and Metonymy: Jakobson Reconsidered," University of Toronto Quarterly 56, n. 4 (summer, 1987): 557-74; Gilbert D. Chaitin, "Lacan's Letter," MLN 103 (December, 1988): 995-1011; Jill Matus, "Proxy and Proximity: Metonymic Signing," University of Toronto Quarterly 58, n. 2 (winter, 1988-89): 305-26; and Murray Krieger, A Reopening of Closure: Organicism against Itself (New York: Columbia University Press, 1989).

[xvii] Ecrits: A Selection (New York: W. W. Norton & Company, Inc., 1977), p. 166.

[xviii] My discussion in this context is based on the concepts of an affirmative fallacy and its principle of negationsnegierung (or double negative) explained in Chapter Seven of Negative Poetics.

[xix] My first treatment of this possibility was the article, "Up Against the Mending Wall," College English, 34, n. 7 (April, 1973), pp. 934-305--later included as Chapter 4 in Negative Poetics.

[xx] Here I use the word warranted in reference to John Dewey's concept of "warranted assertibility" as the verification of scientific truths based on their relative probability. See Dewey's Logic: The Theory of Inquiry (New York: Henry Holt and Company, 1938), pp. 7-9, 143. See also my discussion of this principle as applied to non-scientific discourse in Negative Poetics, pp. 277n.2, 280n.20.

[xxi] Paradoxically, a literary text's transparent significations (those words used to tell its story) can only deceive, since they represent events as if they are actually happening, which they are not, and since they are primarily used to organize outcomes different from our normal expectations in life.

[xxii] An exception to prove the rule for this particular category would be the play Hamlet as I try to interpret it in Negative Poetics. On one hand, the repeated misogynistic figuration encountered throughout the play primarily seems to foreshadow the symbolism of the respective deaths of the characters in the final scene, Hamlet dying by the sword of Laertes and in the arms of Horatio, Gertrude dying by the chalice, and the King, a thing of nothing, dying by both the chalice and the sword. However, it is also possible to shift the story's essential denouement to its very end, when it turns out that the gender ambivalence that made life impossible for Hamlet has now been resolved the only way possible, by his acceptance of death despite his earlier apprehension as expressed in his important "to be or not to be" soliloquy. The more I teach and read the play, the more I prefer this second and more basic interpretation.

[xxiii] Coleridge's "Kubla Khan," for example, begins with a private symbolism (metaphoric) whose androgynous recapitulation of Milton's Garden of Eden (a mound of pleasure, a cave that spouts, another that leads to a sea of ice, etc.) is both resignified and designified by his second stanza's conventional gender distinction between the poet and a maiden who has bewitched his imagination with such a vision. In effect, the first stanza front-loads bizarre androgynous imagery, and in response the second stanza back-loads the more acceptable theme and imagery of a male poet inspired by his female muse. Both imageries are of course metaphoric, but the second also imposes closure in the sense that it both justifies and erases the more disturbing aspects of the first. Animus has accordingly been separated from anima, and the poet from the implications of the hypnopompic fantasy that first inspired him. Coleridge's ambiguous imagery affords pleasure, but so, too, does its denial through conventional affirmation. Something has been both said and unsaid relevant to a temptation too threatening to be directly confronted.