V.

VEHICLE INTERPRETATION

THE PEIRCEAN INDEX IN
POETIC METAPHOR

The sort of metaphoric truth we have considered in chapters 2-4 is discovered by the reader through what Reinhart calls “focus interpretation” (1976). As I view this process, it consists of the discovery of the icon(s) and the literal object(s), often left implicit or unnamed by the poet, and a contemplation of the qualities or character which the object(s) and icon(s) share. As we have seen, the similarity between the icon and object may be a simple sense quality, an analogical proportion, or a universal congruence (first, second, or third Firstness). However, it is only because the Firstness of the similarity is embodied in the Secondness of the metaphor that the metaphorical possibility becomes forceful and actual to the consciousness. This Secondness therefore deserves careful attention in a complete interpretation or interpretive theory of poetic metaphor. Borrowing again from Reinhart, I will call this attention to metaphorical dualism “vehicle interpretation.”

Recall from chapter 1 the claim that the Two-ness of a genuine poetic metaphor must involve opposition between the two (or more) sign referents. It is the “two-sided consciousness of effort and resistance” which for Peirce characterized our sense of Actuality (CP 1.24). The dynamic linguistic actualization which poetic metaphor characteristically enacts upon vague possibility (see PW 81) is therefore a function of the metaphor’s semantic tension. The truth of metaphor is compelling, paradoxically, because it is embodied in a “lie”—an apparent impossibility which “flags our attention,” invites us to consider more carefully, and thus ultimately to sharpen our understanding of possibility. This sort of linguistic “red-flagging” function can be formalized and studied under the rubric of Peirce’s Index. Vehicle interpretation, therefore, I will undertake as a focus upon the Peircean Index.

Since a metaphorical index, as we have seen, does not exist for its own sake but exists first to “indicate” the icon from whose introduction in the metaphorical frame the index arises, does the index deserve to be studied in its own right? Given Peirce’s theory of meaning as triadic (Sign-Object-Interpretant), the answer is emphatically yes. That is, if the semantic tension is a genuine index which points us to an icon (its object), then the index is confirmed as Sign. But in Peirce’s scheme, all Signs which have Objects also have possible Interpretants (CP 4.536, 5.283; PW 32). Thus, if the metaphoric index reliably indicates its object, the index must also have its own interpretant. The “meaning” of the index has not been exhausted by the mere identification of its object (the focus). To assume that an index, metaphorical or otherwise, has no other meaning than that of signifying its object is to make the mistake of reducing meaning to a two-part or dualistic phenomenon; in short, it is to make the mistake of reducing meaning to mere reference, ignoring sense. Recalling figure 1.3, then, successful focus interpretation confirms the “Indexical Hypothesis”—namely, that the semantic tension is “meaningful.” Thus confirmed as a sign, the index must then be fully interpreted in its own right (vehicle interpretation). Of course, the interpretation of the index as sign in itself must still be carried out with reference to the icon it points to; but in vehicle interpretation, the icon is considered in its role as object of the index, rather than in its own role as sign of the metaphor’s literal topic.

This perhaps overly elaborate scheme is merely an effort to formalize in Peircean terms what experienced readers of poetry already know intuitively: Full literary appreciation and understanding of poetic metaphor consist not just of focussing on the icon or of guessing the literal object of an icon, nor just of contemplating the actual reality of the character they share, but of further considering how the semantic novelty of the trope (its indexical tension) functions aesthetically and semantically in the metaphor and in the poem as a whole. When we think of it, this is something we all should know about the interpretation of even non-literary indexical signs. Smoke, for example, is a sign of fire. Specifically, it is (mainly) an index of fire because it shows that there is a fire, and because it indicates roughly where the fire is. Now to the merely curious would-be spectator of the fire, that is probably the only “meaning” to be discovered from the smoke. But to the serious student of pyrochemistry, the smoke holds additional properties of its own. Its color, its density, its shape, the vigor with which it rises up, even its particle composition, can often reveal more about the fire itself than we can understand by merely staring at the flames. Similarly, a careful analysis of the semantic indexical tension arising from the vicinity of an icon in poetic metaphor can often yield fresh insights about the iconic character itself, refining our understanding and appreciation of metaphoric truth. Only when the results of vehicle interpretation (“figural displacement,” in my terms) are fully considered in relation to the results of focus interpretation (literal similarity) can we approach a Final Interpretant, and experience not just the reference of metaphor but the meaning of metaphoric truth. The truth of poetic metaphor is demonstrated to be a far-reaching truth by its ability to arise as truth out of the metaphor’s configuring “lie”; even its similarity condition only fully emerges for what it is when we carefully consider, as a second (or adjoining) step, the quality and condition of its paradox, its condition of dissimilarity

Further, it is possible that metaphoric dissimilarity is actually a first condition for the poet’s process of poetic invention (just as the semantic tension is an Immediate Interpretant which first challenges the reader to look for an icon). Shapiro and Shapiro write:

The preexisting condition acts as a springboard for the onset of the figural situation and simultaneously imposes a constraint on the combinability of the signata: they must differ in respect of at least one of the conditions. It is this constraint which prevents knife from being a trope for fork, or vice versa (ad Todorov 1974: 128), since both words do not differ in any of species, context, or rank. Note, however, that in aphasic (i.e., pathological) speech knife and fork can be commuted for each other (Jakobson 1971: 250). (Shapiro and Shapiro 1976: 7)

Let us examine this constraint more closely. As often happens in discussions about metaphor, no sooner does some scholar make a pronouncement about what can or cannot be a trope than some student thinks of evidence to the contrary. Linguists in the pragmatics (speech-act theory and generative semantics) schools of thought, which often hold that metaphor is a purely contextual or social phenomenon, are especially fond of creating possible contexts in which anything can be considered as a metaphor for anything else. Given Peirce’s notion that everything in the universe is a potential sign of something else (CP 5.448n), the creation of possible-world contexts is no trivial exercise. In the case of knife and fork, for instance, suppose we create a context in which the two words might become parts of a metaphor. Imagine, if you will, a bedraggled camper returning from an outing for which he was obviously ill-equipped. He tries to explain how he managed to eat his canned pork and beans without a mess kit: “My pocket knife was my fork and spoon.” Now, surely the expression is intelligently figurative, not aphasic. Ordinary speech is replete with such utterances which, in appropriate context, we accept as metaphorical (or as metonymic) rather than dismissing them as mistaken or pathological.

Yet this example actually confirms the Shapiros’ point, I think. What the invented context does for the example is to create a situation in which the slight differences between knives and forks and spoons become tellingly significant (we see the poor camper slurping bean sauce from a knife blade instead of a spoon). By contriving to increase the perception of dissonance, then, this context for the counterexample actually affirms the importance of dissimilarity as an allowing condition of metaphor.

Further, even the manufactured significance of dissimilarity in this case is not enough to make “My pocket knife was my fork and spoon” into a poetic metaphor. I will argue that the dual (or more) referents of a poetic metaphor typically differ not just in rank or in context but in species. Consider the following sentence: “Anchorage is Albuquerque.” This sentence is unpoetic, I think, even if it is generously interpreted as metaphor. Anchorage and Albuquerque, while different in geographical context, are perceived as members of the same species (city). Thus there is not the right kind of semantic tension to qualify as a genuine index pointing (reliably) at either city as a metaphorical icon of the other. In fact, unless and until we have a context suggesting the provisional assignment of figurative icon and object roles—say, the sentence is spoken by a tourist from Albuquerque who is noticeably disappointed to find a shortage of log cabins and igloos in Anchorage—then we might just as well consider the sentence a mistake or a sign of pathology.

However, when a dissimilarity in species is introduced, the results are clearly metaphorical and much closer to being poetic: “Anchorage is freeze-dried Albuquerque.” The species dissimilarity introduced by freeze-dried immediately qualifies this sentence as metaphor—without the help of special context, please note—because it is this that identifies freeze-dried Albuquerque as a figural icon for the literal object, Anchorage. (Of course it also identifies freeze-dried products as a figural icon for both cities.) To interpret the sentence, we do not need to know the context (it is a slightly modified version of something I read in John McPhee’s entertaining book about Alaska, Coming into the Country, 1977) for even if we do not know that the speaker is trying to be clever or metaphorical, the sentence contains a genuine metaphorical index (species dissimilarity) indicating a clear icon. I believe the sentence would be taken figuratively in almost any context. Even if it were spoken by a known aphasic, I would think the speaker’s prognosis was improving. The species dissimilarity is semantically interpretable even if it is not consciously intended; if it is a mistake, it is a very serendipitous one. (See Walker Percy 1975: 64-82 for a splendid account of “Metaphor as Mistake.”) While context is often a facilitating factor in interpreting a metaphor, it is not an essential condition to the identity of metaphor; what identifies something as a candidate for interpretation as metaphor is species opposition, for it is this that provokes the search for a figural icon, its object, and their similarity. If this search is successful, the utterance is confirmed as metaphor.

Once the similarity condition is discovered, however, the interpretation (if it is a fully appreciative and cooperative one) is by no means over. The careful interpreter is then led to consider the similarity in relation to the dissimilarity. In the case of Anchorage and Albuquerque above, the literal similarities might be that both cities have modern conveniences, shopping centers, flashing signs and billboards on the streets, and so forth. They have a similar ambiance and style. If that is all, the speaker might just as well have said, “Anchorage and Albuquerque are alike in several ways.” But that is not what the speaker said. He said that Anchorage was freeze-dried Albuquerque. This species dissimilarity is what makes an additional comment about the two cities’ similarity. First, it suggests an irony in the similarity in view of the radical difference in climate (freeze-). Second, it implies that the similarity between the two cities should further be interpreted as something like the “packaged artificiality” of their shared ambiance (-dried), a condition of “modern urban blight.” Further, the dissimilarity between freeze-dried products and both cities insinuates that such instances of urban blight are particularly “tasteless” or “hard to swallow”; for if freeze-dried comestibles are not palatable (being at least properly sized for human consumption), what of the rather large “menu” of urban blight one gets in a city the size of Anchorage or Albuquerque?

Thus, what species dissimilarity means in relation to literal similarity in metaphor is an important interpretive question. Without vehicle interpretation of the index, the ironies and subtleties of or about metaphoric truth are ignored.

Nevertheless, the importance of figural tension as a sign in itself has frequently been ignored, underestimated, and often even denied. Michael Reddy, for example, presents instances of “metaphor” which he says involve no formal semantic tension or oddity at all; a typical case he presents is this sentence (1969: 240-251):

1. The rock is becoming brittle with age.

Reddy correctly points out that this sentence would be taken literally in the context of a geological expedition, but that it would be taken figuratively if spoken by one of a group of college students walking out of the office of some staunch old Professor Emeritus. Therefore, Reddy would have us believe, metaphor does not entail any logical semantic difference; its identity as metaphor is purely a contextual consideration; and formal semantic accounts of metaphor must be replaced with a pragmatic contextual theory.

What Reddy does not seem to have considered, as Robert Matthews points out (1971: 413-425), is that his sentence 1 above is not itself a “metaphor.” Matthews’s rebuttal in no way denies the importance of context; he admits that it is the context of the students with the Professor Emeritus which would cause any sensitive listener to construct a mental metaphor such as the following:

2. The old Professor Emeritus is a brittle rock.

It seems clear that this construct, a reasonable entailment from Reddy’s sentence 1 in the appropriate context, does indeed involve the sort of “strangeness” or “oddity” which Reddy wants to banish from any account of metaphor. In fact, the novelty (species dissimilarity) of sentence 2 is probably the only thing that makes Reddy’s sentence 1 worthy of inclusion in an account of metaphor.

In fairness to Reddy, what he really wants to banish from metaphor study is the notion of “deviance,” in the Chomskian sense of “violation of selectional restriction.” Though I believe Chomsky’s grammar has been unfairly maligned in this regard, I now consider the choice of the terms “violation” and “deviance” by those of us in the Chomskian school of syntax to be unfortunate. Metaphor is natural. Ordinary discourse, not just poetic discourse, is transfused with it. It is not to be relegated to the ungrammatical fringe of language, where it is about to fall off the edge into the pathological void; it is rather at the center of language. Peirce himself expressed this view very aptly:

If a logician had to construct a language de novo—which he actually has almost to do—he would naturally say, I shall need prepositions to express the temporal relations . . . , I shall need prepositions to express the spatial relations . . . , and I shall need prepositions to express motions into and out of these situations. For the rest, I can manage with metaphors. (CP 2.290n)

Truly, language itself is inherently (though broadly) metaphorical, a fact confirmed by the continual growth of the lexicon within language via metaphorical extension and narrowing (see chapter 7).

The only aspect of language that metaphor (in a narrow sense) may be said to “violate,” in a creative way, is the set of boundaries which all of us, including Reddy, recognize as marking it off from strictly literal discourse. Without some notion of “boundary” between the literal and the figural, however unconscious it may be, even ordinary communication would be virtually impossible. We would never know when someone was speaking literally as opposed to figuratively. Almost worse, without such a boundary the delight of metaphor would be gone, because fresh metaphor as such would be gone, or at least be reduced to simple counterfactual. Consider these two sentences:

3. The stars are glowing asteroids.

4. The stars are trembling diamonds.

Without a boundary of some kind between literal and figural discourse, these two sentences would have the same evaluative status. That is, if only one set of habits applies to the interpretation of all sentences—presumably a set of merely contextual habits, if Reddy and others are to have their way—then both sentences are simply wrong. In the absence of any immediate linguistic co-text, the only “context” in which we could interpret (and assign a truth value to) the sentences would be our prior knowledge of the world. Measured against that knowledge alone, both sentences are merely counterfactual. Worse yet, if we choose to dismiss truth value from interpretation altogether, whereby communication becomes a sort of relaxed, pluralistic sharing of thoughts, with no contractual obligation to speak true, then sentences 3 and 4 must be considered equally acceptable and interesting. We would have to admit that the creator of “glowing asteroids” had found an interesting way of thinking about stars; and John Keats, who thought of the stars as “trembling diamonds” (chapter 3), would deserve no better.

All of us, I think, know better, however. If a child wrote that the stars are “glowing asteroids,” a good science teacher would offer gentle correction, approving the child’s effort to generalize about heavenly bodies, but carefully explaining the difference between stars and asteroids. Now what kind of a teacher—of science or otherwise—would correct a child who wrote that stars are “trembling diamonds”?

Perhaps there is some context in which “The stars are glowing asteroids” would be acceptable and even insightful. However, I will venture that “The stars are trembling diamonds” would be interesting and insightful in any context (even if “inappropriate” in a context which demands strict adherence to literal language). For instance, even if the sentence were spoken by a known schizophrenic in a science class, I would think he was making progress. It is, if nothing else, the sign of a healthy imagination. It is of course literally false, but somehow we know to (temporarily) ignore the literal “lie” in this case and assign instead a figurative interpretation. How do we know? How, unless we recognize that the sentence—whether or not the speaker intended it to do so—crosses a recognized boundary between literal and figural truth?

Why, you may ask, do I not accord “glowing asteroids” the same benefit of doubt? That is an excellent question. My answer is that it shows no figural indexicality in itself. It crosses no boundary I know of which marks off literal discourse from figural. (Stars and asteroids, unhappily, are members of the same species in ordinary usage.) In fact, from a purely linguistic point of view, the proposition “The stars are glowing asteroids” is analytically correct. Morphemically, asteroid is aster- (star) + -oid (like). So the purely linguistic sense of the proposition is analytic: “The stars are glowing star-like [bodies].” Thus even the linguistic form, or mode of discourse, of this sentence is entirely literal. It stays well within the boundaries that circumscribe all sentences that I, as a speaker of English, have the habit of evaluating as literally true or false. For this reason, I apply to it no test of truth other than a literal test. In so doing, I find its sense circular (stars are star-like) and its reference faulty: The symbol asteroid, by convention, actually refers to heavenly bodies which are like stars, in certain respects, but which differ from stars in certain technical respects; these technical differences lack the quality or sufficient breadth to cause any engaging metaphorical tension, and so I conclude that it is more reasonable to consider the sentence a technical mistake. This judgment accords, by the way, with what Peirce said in the “Ethics of Terminology” about the metaphorical use of a symbol to apply to “different conceptions” from its original. Such use, he said, can be “rather helpful,” but only under two conditions: first, the two meanings (original and new) must be “strictly analogous in their principal suggestions”; and second, the two meanings must be “remote from one another, both in themselves and in their occasions of occurrence” (CP 2.222, emphasis added). In other words, I think Peirce, too, would take the terminology “glowing asteroid” for star as mistaken usage rather than as metaphor, precisely on the second provision above.

Conversely, “The stars are trembling diamonds” presents an immediate figural indexicality because of the dramatic and instantaneous clash of the fully lexicalized semantic markers on the sense of its words. We need not consult the context or our technical knowledge of astronomy in order to identify the source of this tension. It is there at once, in the very nature of species dissimilarity as reflected in the language. We feel it, we are delighted by it, we play with it; it points beyond the literal, crosses the boundary to the figural, and we follow it with imaginative pleasure.

These boundaries between the literal and the figural, then, I believe to be real, but not in the sense that they offer any restriction or obstruction to crossing. The tension we feel when we cross them provides the energy we need to go ahead in the search for a higher truth. The lines described by these boundaries are usually clear (in the case of poetic metaphor, at least), but they are like the lines of latitude and longitude on a map: They are there only to mark the progress of the metaphorical journey, not to impede it.

On a Peircean map of metaphor, that is exactly the function of the boundaries. I believe that there is a whole network of these boundaries and that we can map their major outlines in such a way as to plot the direction, the magnitude, and the implied final destination of the journey the poet makes in the vehicle of metaphor. This is what I mean by “figural displacement” as a function of the metaphorical index, the target of vehicle interpretation. In chapter 6, I will posit a preliminary version of such a map, based loosely on Peirce’s Categories and his theory of Being, and I will suggest how figural displacement occurs, as well as what further meaning it initiates.

For the present, let us consider the hermeneutical necessity of such a map (or maps). We must have some sort of device to measure the “semantic distance” (causing indexical tension) between a figural icon and its literal object in order to “fix the range” of both the iconic and indexical Interpretants. Specifically, if the semantic distance between the figural icon and its object is small, the range at which an Interpretant may be reliably fixed is also small (by analogy to the principles of triangulation). Conversely, if the distance between the figural icon and object is large, then the Interpretant may be reliably fixed at a greater range. Although this is only an analogy, my days as a land surveyor convince me of its accuracy: To fix the range of a distant mark, one must “turn the angles” between two distant points of reference; if the points of reference are too close together, one cannot get a reliable “fix” on any but a “shallow” mark.

I am suggesting that when the figural tension of a metaphor (a function of the semantic distance between its icon and object) is slight, one had best not look too far out for an Interpretant. On the other hand, if the semantic distance between icon and object is large, then the reader’s prerogatives are greater to go further with the interpretation. The Final Interpretant of a radical trope (one in which there is a great distance between icon and object) must stand at a radically distant range, and the triadic area circumscribed by this icon-object-interpretant relation must be very large, accommodating a great number of (valid) Dynamic Interpretants. But the Final Interpretant of a conservative trope (one in which there is only a small distance between icon and object) is much closer to the Immediate; hence, there is less semantic space in such a triadic relation for a variety of Dynamic Interpretants.

This is nothing new. Sapir has said (1977: 31):

The more remote the terms are from each other . . . the greater are the possibilities for making a variety of nonarbitrary connections. This fact was fully understood by the surrealists . . . who hold as a major tenet the value of arbitrary juxtapositions as a way of achieving “a spark” leading to a fuller, more than real, insight into reality.

Given my position on the nature of genuine metaphorical similarity (chapters 2-4), I would offer only one slight qualification to Sapir’s statement: The juxtapositions in a poetic trope are not “arbitrary.” Nevertheless it is clear that readers of an oxymoron, say, in which the distance between icon and object is huge, have equally huge prerogatives of interpretation. As Marvin Ching has shown in his fine studies of oxymora (1975a; 1975b; 1980), widely divergent reader interpretations of oxymora all exhibit clear co-textual, contextual, and textual validity. On the other hand, “mild” metaphors—as in conversational examples, in which the icon-to-object distance is perceived as negligible—do not encourage or justify great diversity of interpretation. The general scope of interpretation, therefore, is at least partly a function of the semantic distance between icon and object.

Is there a way to measure such icon-to-object “distances” systematically? Probably there is no single “map” of literal/figural boundaries to serve this purpose; most likely, sensitive readers construct detailed maps in ad hoc fashion to fit the possible world of the literary text they are reading. For instance, things which seem normal in Alice’s Wonderland would seem outlandish in The Old Man and the Sea. I believe it is possible, however, using Peirce’s Categories and Metaphysics, to construct the general outlines of a “master map” which subsumes the possibility of many finer contextual maps, and which will work well for plotting the overall index of major figural displacements in poetic metaphor.

I will begin by positing three conditions on which the degree of semantic distance between icon and object depends:

1. the nature of the boundaries crossed;

2. the number of boundaries crossed;

3. the direction of crossing(s).

A fourth condition might be added, namely, the number of times a given boundary has been crossed previously in the language or in the special context of a literary work, since such repeated crossings tend to “erase” the boundary (permanently in the language or temporarily in the special context of a given work). When a boundary begins to be erased, however, crossing the boundary produces only a moribund metaphor; if and when the boundary is ever completely erased, crossing the boundary produces only a dead metaphor—that is, it is really no “crossing” at all. Since I wish to consider relatively fresh poetic metaphor, I will not consider the factor of boundary erasure.

Since the number of extant boundaries (condition 2 above) and the direction of the crossing (condition 3) both depend upon the arrangements of the boundaries on the “map,” I will defer those considerations until chapter 6, where I will offer a hypothetical model of such a map. In the remainder of this chapter, I will consider only the first condition above: the nature of the boundaries crossed.

Peirce’s Categories considered ontologically would seem to suggest three kinds of boundaries that might be crossed in figural displacement. These are represented in figure 5.1

Figure 5.1

Crossing a conceptual boundary (kind 1) produces the highest degree of tension because it predicates what we perceive as a literal impossibility, a negation (at the indexical level) of what Peirce called “Formal Logical Possibility” (CP 3.527). (The negation corresponds to my “Paradoxical Possibility,” kind 1, figure 3.3.) Initially, we may get some sense of what it means to cross such a conceptual boundary by considering anomalous expressions like the now famous “colorless green ideas sleep furiously.” It is conceptually impossible, under our present limits of mind, for something to be literally green and colorless at the same time, and so forth. However, even when such an anomaly is devoid of any interpretation out of context, it can be “rescued” by a context which gives it a diagrammatic frame. (I understand that someone has written a poem to give Chomsky’s famous anomaly an analogical context—I am sorry that I have not seen the poem; my apologies to the author.) To take another example, “triangularity barks” crosses a clear conceptual boundary, and might for that reason seem to be utter nonsense, an indexical “tin finger” pointing nowhere in an art mobile, and I think it very nearly is. Still, it can be rescued, if we are of a mind to do so, by placing it in a context which suggests a diagrammatic frame (this, I confess, is one of my own best efforts):

Circularity whines

Rectangularity growls

Triangularity barks.

Note that the context helps us to interpret (I hope) the high indexical tension by duplicating it three times in parallel but incremental fashion, thus making it seem purposeful; it points to a redeeming “similarity” between the relative “sharpness” of sounds (whine, growl, or bark) and the relative “acuteness” of angles in the various geometric shapes (circle, rectangle, triangle).

I seriously doubt that this sort of “rescue” operation is the ultimate purpose of genuinely poetic metaphor, even if some paradox or another is the first thing which occurs in the actual chronology of events in the poet’s mind. The paradox may be first as a psychological means, but it is not First as a semeiological goal. Poets typically cross conceptual boundaries, thereby creating paradoxes in the contextual frame, during the intuitive discovery of natural icons; they do not manufacture icons out of the contextual frame in order to rationalize natural paradoxes. The paradox is the vehicle; the icon is the focus. Nevertheless, if we are not prepared to do both focus and vehicle interpretation, then we may underestimate the magnitude of the “conceptual leap” the poet has made in this sort of metaphor, thereby underestimating the interpretive possibilities of the iconic ground itself.

Carefully considered, Tom Sexton’s metaphor (discussed in chapter 3) twice crosses a conceptual boundary; yet the iconic possibility it captures in so doing gives the metaphor an aesthetic and semantic ease which belies its conceptual effort:

Floating words and syllabic nets are no more “possible” in Peirce’s “Formal Logical” way than are barking triangles. The metaphor takes a big “conceptual leap” which is easily lost sight of because it “lands” on something subtly true. (Contrast the absurdist Squircle!—square + circle—which leaps but apparently never lands on solid iconic ground.) Still, the semantic tension of the Sexton metaphor is very high when considered in itself. And more importantly, when the tension is considered in relation to the iconic content, an additional purpose or function perhaps emerges. Specifically, the indexical tension brings into sharp relief that sense of “delicacy” in the syllabic nets of frost; the paradoxical impossibility of the index also reinforces the subtle sense of ineffability which arises from the whole figure (as an iconic possibility): its feeling of evanescence, of words that escape the frozen breath, of thoughts which cannot be captured in words, just as words cannot be captured in “syllabic nets.” Here, then, the crossing of conceptual boundaries not only allows a discovery to be made beyond the boundary, but the conceptual tension of the crossing itself seems to enhance the discovery.

Next, consider the more moderate tension created by the crossing of an existential boundary (kind 2 in figure 5.1). By this I mean a boundary imposed by the accidents of existence or actuality and not by any logic of possibility. Peirce wrote: “Of those [combinations] which occur in the ideal world some do and some do not occur in the real world; but all that occur in the real world occur also in the ideal world. . . . [For] the sensible world is but a fragment of the ideal world” (CP 3.527). For those inclined to smirk at Peirce’s rather blatant Platonism here, consider the following paradigm in figure 5.2.

Figure 5.2

It seems clear to me that the items in column 1 could occur in the actual world, though to my knowledge they have not occurred there. (These correspond to my “Negative Possibility,” kind 2, figure 3.3.) Because, so far as I know from experience, they have not occurred in actuality, they cross only experiential or “existential” boundaries and thereby obtain a moderate degree of semantic oddity or tension (creating the sense of a new experience, but not requiring a modification of conceptual possibility). Unlike the items in column 2, which cross conceptual boundaries, the items in column 1 are ideally possible, although they are probably only “negative possibilities,” which consist in my own “ignorance.” Peirce said, “If we do not know that there are not inhabitants of Mars, it is subjectively possible that there are such beings” (CP 4.573). Thus crossing an existential boundary does create a moderate surprise in metaphor, requiring moderate imaginative effort, but the tension and the effort are not so great as when we cross a conceptual boundary. “We know in advance of experience,” Peirce wrote (emphasis added), “that certain things are not true, because we see they are impossible.” Conversely, we can “by ideal experiments” imagine that “certain combinations [can] occur,” also in advance of experience (CP 3.527). However, we must cross an existential or experiential boundary to so imagine them. Leaping a “gap” in Actuality requires less imaginative effort, producing less semantic tension, than does leaping a gap in conceptual Possibility.

For this very reason, the more moderate tension of crossing an existential boundary helps to initiate the nonpoet reader into the more radical dynamics of the conceptual leap. In Sexton’s metaphor, for instance, I think “nets of frost” crosses only an existential boundary. I have never actually seen what I could identify as a net made out of frost; I have seen nets made out of string and rope and wire, but never of frost. Under a microscope, of course, frost seems to be composed in a network, but Sexton’s metaphor is the reverse of that—it is not frost made of nets, but nets made of frost. He asks me to imagine my frozen breath forming a fine lattice in the cold air, something I have never experienced, or at least never noticed before—but something I can instantly “see” as an image. This possibility makes me more at home with the whole metaphor. What is more, the fact that this possible image is not actual, not experientially so, creates a moderate imaginative tension which makes the radical tension in the rest of the metaphor more purposeful: “Nets of frost,” a metaphorical image, a new experience, helps build up the level of imaginative energy needed to interpret the abstract metaphorical diagram of “syllabic nets,” the bigger leap, a new conception. Thus we see that there are different levels of tension and that they work together—rather than against each other, for shock effect alone—in a really fine poetic metaphor.

In other metaphors, rarely “poetic,” yet a third kind of tension obtains: the tension of crossing a merely conventional boundary, a boundary created by habitual association of a given predicate with a certain range of objects in ordinary usage. The predicate bark, for instance, is habitually associated with dogs, less often with seals (in zoology), and even less often with birds (only in ornithology). Thus you will probably feel only a very low level of figural tension when I tell you, quite honestly, that I have actually heard a man bark. This person, in an actual experience of mine, apparently thought he was a dog, and he was not merely saying “Bow-wow!” which of course would not qualify as an actual bark at all. Instead, he was making a sound for which there is no more accurate (albeit mildly metaphorical) description than “barking,” and it was much closer to the usual and immediate reference of barking than is the sound made by a seal or by a squirrel or, certainly, by any bird that I have heard. The figural tension you feel, probably in the form of amusement, when I describe this man’s behavior as “barking” obtains from your habitual associations with bark—that is, from conventional usage. No boundary of actual experience (contrast a barking butterfly) or of possible conception (contrast a barking triangularity) is crossed by the predication of a barking human. The boundary crossed is merely a matter of linguistic convention. This seems to be true of most tropes which predicate animal behavior of human objects. These metaphors produce only a very minor tension, arising from the fact that many “animal” predicates are conventionally reserved for non-human animals. But both in our actual experience and in our logical conceptions, human beings really are animals; thus to say that a human barked, or did anything else normally associated with non-human animals, is not necessarily to cross a boundary of the empirically actual, and it is never to cross a boundary of the conceivably possible. It is usually only to cross a conventional linguistic or social boundary which reflects our habitual usages and cultural attitudes.

I believe it is the rather low tension, and thus the imaginative ease, with which such merely conventional boundaries are crossed that has led some students of metaphor to the false conclusion that all figurative “crossings” are simply of the unconventional kind. Henle, for example, defines metaphor as a mere transgression of convention (1958). Bickerton (1969) and Reddy (1969) fall into the same trap. This is an easy mistake to make if we consider only conversational metaphor, which is most often a mere unconventional stretching of usage. It is also a natural mistake for another reason: Conventional linguistic norms are not always purely arbitrary. That is, Peircean Habit, because of its Thirdness, includes—along with arbitrary rules—conventions which are themselves linguistic or social reflections of the laws of experience or of conceptual possibility. Thus all crossings of conceptual or of existential boundaries can be expected to cross conventional boundaries as well—that is, redundantly. We must not, however, assume the converse; not every crossing of a conventional boundary necessarily entails crossing an existential or conceptual boundary.

For instance, suppose an eccentric old bachelor calls the veterinarian to announce that his pet parakeet is “decidedly uncomfortable!” The mild figural tension we feel, in the form of amusement, obtains from the crossing of a mere conventional boundary of linguistic usage. There is absolutely no reason, when we think of it, not to believe that birds really are comfortable or uncomfortable at any given time. It is just that—given our egocentric habits—we do not normally concern ourselves to that extent with the feelings of the poor things. The old bachelor’s parakeet probably is uncomfortable (in actual experience). Surely no one would deny that a bird could be uncomfortable (in conceptual possibility). We simply do not normally put it that way. The usage is unconventional.

Conversely, we must not equate the figural displacement of all metaphor with its (often) redundantly unconventional use of language. Let us examine this principle of redundancy more closely. It is easy to see, in the poetic examples of metaphor we have been considering, that every crossing of a conceptual boundary necessarily entails the crossing of existential and conventional boundaries as well. When I say that my “words float in syllabic nets of frost,” of course I am using unconventional language. But the mere oddity of the usage hardly accounts for the strong indexical tension of the trope. I do not have the habit of talking about words in this way for a simple reason: I have never experienced a word literally floating in a syllabic net of frost. Further, I have not experienced such a thing because there cannot (literally) be such a thing. Words (not the acoustic tokens of words) are abstract symbols; abstractions do not, because they cannot, literally have mass or material shape (more at chapter 6). In this case, then, a conceptual paradox expresses itself (redundantly) as both an existential anomaly and an oddity of convention.

Thus metaphoric tension, which in verbal metaphor always includes unconventional uses of language, cannot be fully explained in terms of such, unless we are to abandon the hope of measuring those variations in semantic distance between the figural signs and literal objects of different tropes, on which variations their interpretants partly depend. Further, defining metaphoric indexical tension solely in terms of a departure from linguistic convention obstructs inquiry into the metaphorical dimensions of other than purely linguistic signs and objects in a literary work. The characters, the actions, the very structure of the work itself can all be indexical and iconic (as we have seen in chapter 4 with a sonnet by Shakespeare). An adequate theory of metaphor must therefore be semeiotic, not just linguistic.

Peirce’s semeiotic—based upon his Categories of Firstness, Secondness, and Thirdness, subsuming (respectively) Possibility, Actuality, and Habit—allows us to distinguish the varying degrees of conceptual, existential, and conventional (including linguistic) tension in the metaphorical index. So far, we have considered two reasons why it is important to make these distinctions: first, because the metaphorical index points to an icon which it contains, the different kinds of tension shape the icon—or our perception of the icon—in various ways, thereby coloring or changing our evolving view of the icon and its object, as well as of the similarity condition they share. Second, the various kinds and degrees of indexical tension interact among themselves, frequently producing a gradual “build-up” of semantic indexation, thereby facilitating the nonpoet reader’s interpretive progress towards higher levels of linguistic, existential, and conceptual possibility (in the icon) which lie just beyond those same levels of impossibility in the index.

With regard to this second phenomenon, it should perhaps be noted that twentieth-century poetic metaphor (as well as metaphor from the metaphysical school of poets in the seventeenth century) often employs the different kinds and degrees of indexical tension to different effect: Rather than a gradual “build-up” of tension, we get sudden, jarring shifts which seem almost calculated to distort, or even to overthrow, the hierarchy of the conceptual, the actual, and the conventional. In such cases (passages from Eliot and Donne come to mind), the strange or surrealistic effect can often be studied to advantage by isolating the unexpected fluctuations in level, as well as the non-sequential shifts in kind, of semantic tension. I will offer some suggestions about this in chapter 6.

Poets in the nineteenth century, however, seem “kinder” to the uninitiated reader, gradually increasing or decreasing the level of semantic tension in such a way that all the different kinds of indexation work together to produce a single unified Index, orienting the reader by pointing in one direction only. I conclude this chapter with a discussion of one such example, from John Keats again (Endymion, I, 453-456):

The passage begins in low tension, with two crossings of conventional boundaries: the Magic sleep and the comfortable bird. As already suggested, “comfort” is conceptually and actually predicable of birds; it is simply an unconventional predicate for non-human animals. The tension is therefore so mild as to be hardly noticeable, effecting only a slight figural displacement of the human ego (the habitual scope of comfort) into the image of the bird. The result is thus a subtle initiation of Keats’s “Negative Capability”—his theory of poetic projection of the human spirit into that of other bodies or objects. “Magic sleep!” is also mildly figural. To those who know Keats, this is no ordinary sleep; it is probably that trance-like state which he associated with poetic visionary experiences. However, even the uninitiated reader of Keats might sense something unusual about this sleep: Magic is normally associated with potions, charms, incantations, spells, and the like, not with sleep. “Magic sleep” is therefore an unconventional association (except in the special world of fairy tales); it does not, however, necessarily cross any boundary of actual experience or of conceptual possibility. Given the existence of magic to begin with, at least as an actual psychic or anthropological phenomenon, there is no difficulty in believing that sleep could be—and in the case of a religious trance, actually is—the result of magic as opposed to ordinary fatigue. The language in the opening line, then, is merely unusual, creating only a low level of figural tension.

If we read the first line to mean that “sleep [is a] bird,” then the indexical tension would be very high (a crossing of a conceptual boundary). As it turns out, that is ultimately a connection which Keats wants us to make, but he does not ask us to make that connection in the first line; rather, he merely places the object (sleep) alongside the eventual icon (bird) in loose parataxis, without predicating the one as the other. It is only in the second line, with “broodest o’er the troubled sea,” that the indexical tension necessarily increases to the next level—the crossing of an existential boundary. The image of a bird brooding over an entire sea is more, I think, than a mere unconventional stretching of conventional usage. Even in its current figurally extended sense, brood means “to hover envelopingly.” The primary sense of brood is that of a bird sitting on eggs or enclosing the young under its wings protectively. The primary sense is no arbitrary assignment of this predication to birds: Actual experience finds birds doing just that, protecting the eggs or settling the young chicks beneath the wings. Keats, however, has this bird brooding over an entire sea, hovering over it as if to settle its waves (“troubled sea”). Not only in ordinary word usage but also in actual existence, birds do not do such a thing. It is conceivably possible, of course, that there could be a bird with a wingspan that huge or a mission that halcyon-like (thus no conceptual boundary is crossed, here), but such is contrary to our extra-linguistic experience with and knowledge of birds. (Recall that the notion of a “comfortable bird,” conversely, is contrary only to linguistic convention.) Thus the semantic tension of the metaphorical complex has increased to a second level in the second line, significantly expanding our perception of the focal icon (bird) and orienting us toward even higher imaginative flights to come.

With “sea of the mind/ Till it is hushed and smooth,” the tension increases to a third level, as a clear conceptual boundary is crossed: mind ⇒ // ⇒ sea. Students of Peirce may here recall his notion of the “Quasi-mind” as consisting of the whole universe, or his objective idealism, the theory that all matter is but effete mind (CP 6.25). For Peirce, then, the predication “sea of mind” might not have crossed any conceptual boundary at all. In view of Keats’s theory of Negative Capability, the “conceptual leap” by which mind is reconstituted as sea might have seemed quite a small or natural step for him as well. For me, though, and I suspect for most readers, this part of the metaphorical index requires a modification of conceptual possibility, not just the imagining of a new experience. Indeed, that is precisely the value of this powerful index: Beyond the seeming impossibility of mind becoming a sea, or of a sea being a mind, there is perhaps a very real possibility which Peirce and Keats understood and would have us understand. In fact, I believe the icon within this index possesses metaiconic status: “mental growth ≅ spatial expansion” or the like, here in the token of a poet’s mind acquiring the expansiveness of a sea (more at chapter 6).

In any case, a bird brooding—hovering envelopingly—over a sea is at least a spatio-conceptual possibility, however unconventional the language or contrary to actual experience the image may be; but to stretch mind to the limits of sea and to reduce mind to the substance of sea simultaneously is no mere unconventional usage, no mere suggestion of a new experience; it is a philosophical leap, a radical reconstitution of conception itself. Moreover, with this conceptual boundary crossing, Keats at least triples the effect by crystallizing the entire complex diagram in a single instant; “sea of the mind” becomes the typological key to unlocking a series of interlocking analogies:

sleep / mind :: bird / sea

mind / troubles :: sea / [waves]

sleep / settles troubles :: bird / [smoothes waves]

This all happens at once because, as noted earlier, the sleep-bird connection is merely paratactic in the first line of the complex; only when we discover in the second line that the bird is floating over a sea-mind are we asked to imagine, by analogy, sleep “floating” over the human mind; this completed connection is necessary before we can construct the analogy in the third line, that of sleep settling one’s (mental) troubles as some huge halcyon’s wings might smooth or suppress the waves of the sea as it hovers over them. All three analogies involve the crossing of conceptual boundaries, so the indexical tension mounts exponentially, driving the metaphor to a transcending metaiconic possibility.

Not only does this mounting tension thereby significantly reconfigure our perception of the iconic possibilities, but it continues to serve its purely indexical function of pointing still further in the same direction of increasing imaginative and conceptual complexity: “O unconfined/ Restraint! imprisoned liberty!” Here there can be little doubt about the crossing of conceptual boundaries, for the cross-predications of restraint as lack of confinement, and of liberty as imprisonment, are in precisely opposite directions. Except for the way in which Keats’s indexical tension builds up to these radical oxymora, they might seem hardly more than “squircles” or “colorless green ideas.” Within the gathering trend of Keats’s indexical patterning, however, the radical oxymora represent the perfect aesthetic and semantic culmination of the metaphorical complex. Specifically, the building indexical tensions of the whole figure find a simultaneous culmination and release, both in what the oxymora are (natural paradoxes) and in what they say (that sleep is at once a release and a restraint of neuro-muscular tension, energy, and motion). Hence, the distinctly kinesthetic index of the oxymora constitutes, in itself, something of an iconic overlay for the entire metaphor.

We see, then, that carefully reading the indexical tension helps us to understand how the whole passage, in its very indexical pattern of organization, is a superb icon of sleep: The feeling of sleep is first mildly stretched to include the comfort of a brooding bird; then the mind in sleep is both expanded and sublimated in the magnitude and vast unconsciousness of the sea; finally, as if its longed-for “magic” were at last found out, the liberating restraint of sleep is captured and at once released in pure conceptual abstraction. To be captured in such sublimity, of course, is to be imprisoned by freedom.