Being neither a literary scholar, critic, nor theorist of literature, and, further, being identified with that sort of event that is often said to be worth a thousand words, I have wondered how to approach a collection of words about metaphor as extensive, divergent, and thorough as those that have been assembled in this collection (Note 1).
I have considered whether I was chosen to comment on this series of papers because of what I was not (Note 2)--a researcher of words or of verbal behavior; or whether it was because of what I was--because the editor felt that the concept of metaphor was broader than the purely verbal, and that a scholar of visual communication might find some common elements within the largely verbal elements that most of the articles would have. I have taken the easy way out and will assume that I was asked to comment on metaphor for both reasons, for what I do not study as well as for what I do.
The striking thing to me about this collection of essays is that on the whole, and with a few exceptions, every author "assumes" that metaphor is a verbal event--a verbal "thing" of some sort. This
seems to be so deeply founded a notion as to make me think it might be made of the very stuff that culture is made of. What we absorb with mother's milk, and what we hardly ever question. My first effort, therefore, is to question that. Is the concept of metaphor one that can or must only be applied to verbal events? Has it, in fact, only been used in verbal behavior--in speech or in written materials? If the answer to these questions is not a clear yes (and I will try to show that it is a clear no) the second question that must be examined is why it seems to have been assumed by the writers represented in this issue that metaphor should largely be examined as only, or mainly, a verbal phenomenon or event. The answer to this last question is more complex than an appeal to context; that is, that the authors represented are writing for a literary journal. I think that in general all the authors are writing not only within the context of a literary journal but within the context of a literary and verbal culture as well.
I do not mean by this to imply the tired McLuhanesque refrain about "electronic media," the "film generation" of Buckminster Fuller, or variations on Rudolph Arnheim's "visual thinking." I also do not mean that the contributors to this issue of New Literary History are "behind the times," "retrogressive," or not "with it." I am suggesting that indeed they are with it; that many of the ideas expressed by people such as Arnheim and McLuhan about the centrality of the visual mode in our culture are simply not the case. I am suggesting that the contributors to this volume do represent the ethnographic present as regards thinking about a concept such as metaphor, and that the challenge is to explore some of the reasons why such discussions center around words, verbal discourse, and verbal texts.
First allow me to dispose of the duty to define metaphor. I shall, not only out of diffidence in the face of massive scholarship, but out of wisdom, refrain from anything but a cursory discussion of how I plan to use the word.
First, like almost all the contributors to this issue, I think of metaphor as a structure, rather than an artifact or a chunk of speech or writing. Metaphor is a structure composed of elements in any mode (as I shall try to show) related in certain ways. Most of the structures analyzed by such authors as Derrida, Ricoeur, Hamlin, Sparshott, and Todorov seem to fit, if not always agree, with my usage, and I have no general disagreement with them. In general, I will assume that although the issue is complex, by the time readers get to summary sections such as this one, definitional questions can be allowed to languish. In general, I like to think of metaphor as a structure that has inherent in it the Lévi-Straussian formalism A: B::C: D (read as "A is to B as C is to D"). The statement "the sun
is to the moon as man is to woman" leads to the commonly understood metaphor, "she is my moon," and to the somewhat shocking and richer metaphor, "he is my moon," the initial relationship of sun: moon :: man:woman being known as part of the things one knows in our culture.
Ernst Gombrich uses a similar notion in discussing visual metaphor. He starts with Homer's phrase about Achilles, who "leapt, a lion, on his foes," and carries this notion of verbal metaphor to its use in pictures, such as Jupiter carrying a thunderbolt, St. Catherine a wheel, and Charity, children. He continues: "In contrast to this use of images as labels, the lion . . . is not a code-sign. The image of the lion can be used in different contexts . . . to convey very different ideas. The only thing these ideas have in common is that they are derived from such traditional lore about the lion as is crystallized in bestiaries and fable. It is this lore which defines what may be called the area of metaphor" (1963a:12).
In one case, of course, we can decide to use the notion of nobility associated with lions and in another the notion of bestiality. In a similar but different way, "she is my moon" depends on our common social knowledge that the moon is gentle, soft, pale, and so forth, and that "she" is often, within our culture, associated with those attributes(Note 3). "He is my moon," however, demands that we structure the elements of that phrase within the same attributes we used in the previous similar structure. The change is not between different aspects of moon--that is, between different contexts--but between different pronouns attached to the metaphoric structure--that is, between different aspects of grammar.
It will become clear below that it is just this difference between grammar and context in similar structures that is at work in the visual mode, and that it is precisely these aspects of metaphor that are most richly manipulated by filmmakers and painters. I shall further argue that visual structures can clarify a general and abstract notion of metaphor.
Three major issues suffuse all the papers presented here: interpretation,
formalization (Note 4), and representation.
All of these issues seem to be considered as literary, or at least
as issues involved in forms of verbal use, either as discourse,
or as verbal or written articulation,
expression, or communication. Yet none of these issues is limited to the verbal mode, and all have obviously and explicitly been basic to the history, analysis, and criticism of the visual arts for at least the last three hundred years.
Most recently, the same three issues and concepts have become central to theories of semiotics and, in particular, to theories of the semiotics of film. Ricoeur, Sparshott, and Todorov particularly address themselves to, and lend themselves to a discussion of, questions of interpretation. What seems to emerge from all their discussions is the clarity with which they feel that metaphor must be interpreted. That is, that metaphor is not some event which is intrinsically known to the person who comes upon one. Almost all of the papers implicitly or explicitly understand metaphor as a formal device or structure demanding some kind of active interpretive behavior. (It might be important here to point out that David Edge conceives of metaphor in an entirely different way: one that not only doesn't demand interpretation in the way I shall be using the word but seems more akin to what Thomas Kuhn has called a paradigm. Not only does Edge use metaphor in a broader sense than anyone else; I believe that he so broadens the use of the term as to make it useless as a term describing a particular cognitive structure.)
It is clear that in this view, a "collection," "bundle," "bunch," or "gaggle" of signs, visual or verbal, in and of themselves and by themselves, have no (intrinsic) meaning, any more than the arbitrary sounds of language have meaning by themselves, without some sort of social agreement, as evidence, for example, in a lexicon (written or handed down verbally). In his famous Soviet film Strike, Sergei Eisenstein uses a sequence in which a factory foreman (who has informed the Cossacks about a coming strike) is shown in a close-up which Eisenstein then immediately follows with another close-up of a jackal running into the woods. Who, today, doubts that that film sequence is to be interpreted as something akin to the verbal notion of "the informer is a jackal"?
In Dusan Makavejev's film W.R.: Mysteries of the Organism completed in 1971, this Yugoslavian filmmaker uses complex structures of metaphor which seem to surpass the kind of structures we use, or perhaps can use, verbally. He combines commonly held knowledge about Wilhelm Reich (that he wrote about sex and politics) with commonly held notions about Stalin and current Chinese leaders (that they are dictators and powerful authoritarian figures), and produces metaphoric sequences that show people writhing in Reichian therapy, followed by a man writhing in electroshock, connected only by visual structure. As the metaphor continues to unfold, we see a
sequence of erotic art, and artists at work, which ends with a long series of shots showing a woman artist making a plaster mold of an erotic magazine editor's penis. This is followed by a shot showing the mold being opened and the emergence of a series of penises made out of red plastic. This is followed by a shot of hundreds of thousands of Chinese waving the little Red Book, followed by a sequence beginning with a close-up of Stalin taken from an old film; we (the audience) do not know if it is a documentary, and thus the "real Stalin," or a drama, and thus an "acted Stalin."
The question of how we know ways to make interpretations of such complex visual metaphors in films seems to revolve around questions of the complexity of learning and of describing interpretive strategies for verbal structures. But the notion of complexity often is a red herring, akin to concepts that linguists used to have that some languages were more complex than others, and that of course the more complex (read "civilized") state was the end development toward which the simpler (read "primitive peoples") aspired. In comparing filmic events to verbal events, we might sometimes be tempted by the same simplistic concept of complexity. But a moment's thought should show us that it is not complexity, but different forms, rules, and procedures which are at work. For example, "she is my moon" offers no difficult interpretive syntactic problems for native speakers of English. "He is my moon" offers no syntactic problems either. Native speakers of English are certain of the grammatical rules to apply to both metaphors. One may argue about or reflect upon the "traditional lore," the special cultural meanings, or the particular social and aesthetic principles displayed in those two verbal metaphors, but not their syntax.
In a visual metaphor, whether one as seemingly simple as the Eisensteinian double close-up of foreman and jackal or the one just briefly and incompletely described as used by Makavejev, the syntax is much more at issue than the cultural and traditional lore which forms the boundaries within which the structure of metaphor operates.
Interpretation of visual metaphor raises questions about the interpretation of "language" or syntactic forms for nonverbal matters for which we have very little social agreement, thereby raising the question of the semiotics of metaphor. That is, is metaphor a metastructure dealing with a code which is transformed upon a variety of modes visual, musical, verbal, and even mathematical? Must metaphor therefore be studied across modes in order to extract the particular cognitive functions which make it so powerful a structure of human thought?
This question of the semiotics of interpretation which I have
dealt with more lengthily elsewhere can be briefly summarized for the purposes of this paper by mentioning several central issues of interpretation (see chapters 5 and 7). First, interpretation of any sign system takes two distinctive and opposed forms, one interactive, the other communicational. The interactive (see Figure 5-2) is related to strategies of interpretation leading to attribution. The other strategy is communicational and is related to structures of implication and inference. Metaphor is a communicational code depending upon the recognition of structure and the assumption of intention on the part of the "articulator," "artist," "producer," or "creator" of the form we are to treat as metaphor.
Space obviously does not allow a full exposition of this theory of interpretation, but the mention of its significant terms might be sufficient to allow the reader to follow these remarks. Attribution as a strategy of interpretation in any mode puts onto the interpreted event meanings that come from outside the event only and does not use rules and grammatical "instructions" from within the form. Communicational inference uses both knowledge from outside and knowledge from syntactic forms. Communicational strategies make use of social knowledge, as opposed to purely personal or psychological belief. We may personally love men with black hats, but in dealing with Western film metaphor, we "know" that the man with the black hat or the black horse is the bad guy. We may think lions are cute, but we "know" about the symbolic use of the lion as ferocious, noble, or as king of beasts. In a film sequence of "hunter stalking his prey silently" and "lion stalking his prey silently," we would have to be careful (since filmic syntax is still relatively unclear) to arrange our shots in such a way that the metaphoric rather than the narrative aspect of the sequence was apparent. That is, we would have to arrange the shots in such a way as to make it clear that we weren't meaning (because one shot of a lion followed a shot of a hunter) that "the lion is hunting the hunter," or in a reverse arrangement that "the hunter is hunting the lion." Rather we would have to structure our film so that it was clear that we meant the metaphoric structure, "the hunter is a lion." In further sequences, we could clarify whether we are invoking concepts of nobility, ferocity, and so on.
The semiotics of interpretation demands a delineation of culturally accepted rules. In the interpretation of dreams, as well as of movies and poems, cultures must and do provide us with syntactic, semantic, and communicational rules. Whether these are based on Freudian principles, the oracles, or (as with the Walbiri of Australia) a semilinguistic form of dream interpretation employing drawings in the sand is not the important factor. The important point is that every culture provides its "native speakers" in any mode with a code for
interpretation. These formal metasemiotic structures which deal with all forms of cognitive organization within the context of mode, genre, and culture seem to offer the most exciting new roads to the understanding of metaphor.
I would like to jump now to a possible set of rules exemplified
by Sparshott's conception of metaphor as talking about something
as it is not, since he also is one of the few authors who mention
the possibility of metaphor in other modes such as music and painting.
In semiotic terms, Sparshott's "first impulse" that
a portrait "is literally pigment and metaphorically person"
is simply a confusion in terminology. As "literally pigment"
it is not a "portrait," since "portrait" is
a social term that at various times has various meanings, none
of them "literally pigment." In thinking of photography,
however, C. S. Peirce fortunately has given us some terms to handle
the "portrait" problem. Some photographic faces can
be thought of both as a "portrait" and as an "index."
It is when we think of a portrait as an "icon," as not
having a point-to-point similarity--being an index --with its
referent, but as having something called a resemblance or a similarity--being
an icon--that the closeness which Sparshott recognizes between
a metaphor and a portrait becomes central to the point we also
are trying to elucidate. When we look at the portrait of Churchill
that Sparshott presents to our imaginations, we must, in order
to see it as a portrait, assume certain intentions on the part
of the portrait maker (Note 5). We must assume
not only an intended portrait but an intended structure,
an intended relationship between the elements on the canvas and
the elements connected to the sitter. This connection could be
perceptual (how he looked--cigar and all) as well as cognitive
(how he was thought to be--by the painter or photographer, as
well as by the public).
A portrait is not like a cloud that looks like a camel or a flower that I see as my love, but like "a flower that is my love." That is, a portrait is like a sentence; it is a structure of a certain kind, and could
be a metaphor if its form had the structure of metaphor. When, then, can a portrait be a metaphor, or what, then, would it mean to say that a particular portrait was a metaphor? Here we must go back to the "common lore" which Gombrich talks about. "Red," he continues in a later passage, "the color of flames and of blood, offers itself as a metaphor for anything strident or violent." He later mentions such things as gold and glitter as becoming metaphors of value in certain periods, and other art historians have pointed out that because of their cost, the actual pigments (ultramarine, for example) became metaphors, or stood for importance, wealth, and value in painting.
If metaphor is "talking" about something as it is not, a portrait is somewhere smack in the middle, and this points up why I believe that an examination of the issues of metaphor in visual codes will clarify the issues raised for verbal codes.
A portrait clearly is an event that is not the "real thing," but it is clearly as if it is the real thing. Rather than being like a cloud that looks like a camel, a portrait is a "camel" that looks enough like a camel so that people can say that it is that camel. A picture of Churchill is enough like Churchill to be recognized as Churchill. This, however, is not really what is interesting in either metaphors or portraits.
It is clear, but has not been discussed enough, that some intended portraits or some paintings of faces do not "come off." No one recognizes Churchill. It is also clear that some portraits of Churchill are more like Churchill, or better of him, than others. It is also clear that some metaphors do not come off and that some are better than others, more exciting, more revealing, and more illuminating of the condition or situation they mean to communicate.
It seems to me that understanding the nature, or the "meaning" (as Ricoeur uses that term) of a picture as a portrait and as a metaphor requires of us not only an understanding of what it means to make something look like something else (as in iconic structures) but also to understand those iconic structures that look like, but also deliberately are in many ways not like, the thing with which they have an iconic relationship. I speak here of course about that form of portrait we call a caricature. The argument I shall be making is that metaphor might most fruitfully be understood in comparison with theories and concepts of caricature, rather than with theories of representation. Theories of metaphor as representation, such as Sparshott's (Note 6), are in my opinion possible only because they use verbal labels and
philosophical traditions that historically deal solely with the analysis of words. That is, arguments as to whether a verbal sign (spoken, written, and so forth) is true or false, or speaks of things as they are, or as they are not, apply only to those modes for which truth tables, concepts of true-false, and the formal ability to assert the negative in propositional forms may be applied. Pictures do not have these abilities to make true-false statements or to assert the negative: they cannot say ain't, either (Note 7). Music is neither true nor false, and metaphor can best be understood outside the largely verbal traditions of true false and negation.
A caricature, like a picture, is neither true nor false, but, like a metaphor, is a structure that reveals a set of meanings intended to communicate a certain set of relationships within some understood or understandable context and bounds. A caricature is a structure that means neither that something is as it is or as it is not. A caricature is, rather, a structure that relates several elements on one level (in short hand--that of "reality") with elements on another (the symbolic). It puts things together both as they are and as they are not, and the point of caricature, like that of metaphor, is that neither is only a "portrait." A metaphor is more than "as it is," but it cannot be totally "as it is not." A tracing of a photograph of Nixon will sometimes make a recognizable portrait of him, but most often not. A caricature of Nixon will always include some arrangement of his famous jowls, his hairline, his nose, his ears, and one or two other distinguishing features. But not all drawings of people with jowly cheeks, long ski noses, and receding hairlines will be recognized as Nixon. If labeled as such by lettering, audiences will often treat this as a sign of incompetence on the part of the artist, assuming that he knew it didn't look like Nixon and therefore needed to use a written label. The famous Escher drawings of buildings and stairs in strange perspectives are labeled by viewers as "impossible," rather than as untrue or false. The word portrait, just like the word metaphor or caricature, is a label we place upon a structure. In the verbal mode, this structure may allow us to make meaning "from a clash between literal meanings" (cf. Ricoeur). On a semiotic level of analysis which relates all modes, metaphor may be a system that helps us to provide meaning to structure through the juxtaposition in space or in time of layers and levels of elements in a variety of codes, modes, and contexts. Metaphor may be that special code that allows us to put together that which seemingly doesn't go together, and which also, and at the same time, provides us with the rules or lore which
tell us that everything that "isn't" does not necessarily go together.
I am suggesting here that the rules governing the structure of metaphor in any mode are designed with as much to prevent putting the wrong things together as to help us put the right things together. In the hunter and the lion sequences, the conventions and rules of film metaphor are there to prevent us from interpretations such as "he is hunting the lion" and to facilitate interpretations such as "the man is to nobility, grace and hunting ability as the lion is to those very qualities."
Let me illustrate this with quite a different example taken from the work of David Perkins (1974). A caricature of Samuel Beckett by Levine shows Beckett with a huge beaked nose, a long neck, and a spindly, almost legless, body. The nose is drawn more like the beak of a bird and around the neck we notice what might be a collar of feathers. Somehow this drawing suggests a kind of bird of prey, a vulture out of some strange bestiary. And yet the picture also looks like Samuel Beckett. Compared to a photograph of Beckett, the resemblance is extraordinary. But the photograph doesn't intend to mean anything about Beckett the poet. The beaklike, vulturelike drawing of a swooping bird of prey, however, fits one cognitive concept we may have of Beckett's plays and novels. Somehow, not only does this caricature look like Beckett while also looking like a kind of bird which obviously doesn't look human or like Beckett, but it also reveals its meaning to us in the "clash" that Ricoeur mentions between the physchological attributions we make about Beckett the poet and about buzzards as birds of prey. The same caricature also relates something about the quality of Beckett's work that is succintly put together in this clash of meanings between what Beckett looks like and what a bird of prey is like. Here L»vi-Strauss's structure that I suggested in my opening pages helps to explin as film montage not only verbal metaphor, but caricature as well. It is as if in one drawing we have put together the following structure: "This drawing of Samuel Beckett is to the way the 'real' Samuel Beckett looks as the way we fell about buzzards and birds of prey is to the way we feel about certain aspects of the work of Samuel Beckett." Nothing here is like it is, everythign clashes with our recognition of oppositions and relationships, and yet nothing is the way it is not, either. It all seems to fit.
The problem of metaphor is therefore somehow the problem of fit--of iconicity. How much similarity do we need in order to make something similar, and how much do we know that something is similar enough for us to say that it is a "picture," or a "description, " or a
"representation," or an "instance" of X? Metaphor, like painting but unlike written language and music, has no notation system. As Nelson Goodman, in Languages of Art (1968), has so clearly pointed out, we cannot have a performance of a painting, but we can have a performance of a Bach sonata, of a play, or of a poem. We cannot have a performance (or, again unlike painting, a forgery) of a metaphor, either. And yet the problem of how we know that some metaphors in any mode "work" is akin to how we know that certain performances of a Bach sonata have reached a point where we say, "It just is not that sonata anymore," or further, "I don't know what you are playing." It is the problem of how we know when a variation on a theme is a variation and not a new theme. Often, after hearing variations one through twenty consecutively, we can tell that variation number twenty comes from (is a transformation of) variation number one and is structurally related to it. But listening to or looking at one and twenty without knowing the intervening steps, we are often unable to discern the structure that, through transformations, made the latter a variation.
Theorists who seem as far apart as Freud and Chomsky sound precariously similar when they discuss how we know that one thing is related to another by some form of similarity. The knowledge a child has that the statements "John hit the ball" and "the ball was hit by John" have similar, if not identical, meanings comes in part, according to Chomsky, from the fact that both sentences were generated from some basic form, and that children know these principles of generation. He further proposes a theory of mind which suggests (and often states) that the human mind is "wired," or has developed in such a way that certain features of "Language" called "deep structure" are inborn. Individual "languages" are transformations from this basic deep-structure "Language," and the reason all children learn to speak their surface-structure specific languages is that transformations from what is inborn are relatively simple.
Freud (to my consternation) often sounds like Noam Chomsky. He seems to suggest throughout his work that there are certain "things"--love, death, sex, sex organs, and so forth--that everyone is unconsciously concerned about and that everyone deals with in dreams and other unconscious behaviors. These finite subjects or things (like Chomsky's linguistic rules) have infinite symbolic representations (like Chomsky's sentences). That is, for Freud there is a finite group of unconscious materials we work with, which have an infinite number of symbolic representations. Thus, for example, "penis" is a member of the finite group, but "chimney," "pencil,"
"snake," "cigarette," and on and on, are members of the infinite set of things with which we symbolize penis. And so with "vagina" as "viola" and "violin," and so on.
Freud, however, becomes much more relevant to a theory of metaphor as caricature. His answer to the question, "How do we know that snakes 'mean' penis?" is disarmingly simple. He posits an innate, unconscious ability on our part, never quite clearly described, to tell which things are recognized as similar to other things. In modern Chomskian terms, Freud posits a theory of innate iconic recognition. Our brains must be "wired"; we "know" somehow that certain things are similar to other things. We don't need to learn it; in fact we don't need to consciously know it.
If we accept these theories of innateness, the problem of metaphor is solved. We are born with the ability to recognize certain similarities in verbal language, in images of dreams, and presumably therefore in pictures, and probably in music (who would want to postulate so unparsimonious a brain as to leave music out of any theory of innateness?). The problem, however, is that some peoples "see" (think of) bananas as well as many other so-called phallic symbols as fertility symbols and as indisputably female, and that rules of deep structure, like computer translation of Shakespeare, are a dream that seems to be slight and fleeting.
We are still left with the question of how we know the similarities that metaphor brings together. Here I find the early part of Ricoeur's paper the most congenial, and the one open to expansion across modes. He mentions Richards, Black, and Beardsley as those with whom he agrees as to the meaning of metaphor. It might be interesting to conclude by describing some of the theoretical writings on film that I mentioned earlier, and to use their explanations of what is called "montage" in film to explain both the filmic examples described previously and to amplify the concept of metaphor as defined by Ricoeur. Ricoeur uses several key concepts: opposition, clash, and shift of context. All of these were described most succinctly in the early writings of Sergei Eisenstein and others (Eisenstein 1963; Jarvie 1970; Worth and Adair 1972). I shall not quote here, but rather paraphrase to save space. Film meaning, Eisenstein suggested in 1925, occurs by a collision of filmic ideas. This collision occurs in the form of "shots" (which are, he suggested, the basic units or elements of film form) in opposition. Each shot must correspond to a concept or feeling, so pictured as to present to the audience an idea. The following shot in the sequence presents to the audience another idea, but one in opposition, one designed expressly to create a conflict or collision of ideas. From this collision of ideas comes a synthesis--a
new idea, depending upon the previous shots, but not being merely additive.
In my opinion, Eisenstein was describing a theory of how metaphor works. His films and his writings detail this theory, going on to describe how shifts of context by the filmmaker could make the same image change meaning. By that he meant, in a simple filmic sense, that if you put a shot of a weeping woman (in a film in which the dramatic action concerned a son jailed for revolutionary activity) in between several shots of ice crashing and breaking up in a swiftly flowing river, you would get an entirely different meaning than from a sequence in which the same shot of the woman was placed between shots of birds gracefully flying through the air and children playing happy games in a sunny, wooded countryside.
But although a filmic theory of montage, editing, sequence, or interpretation is, in my opinion, essentially a semiotic theory of metaphor in general, its understanding depends not upon some innate notion of similarity between our receptors (ears or eyes) as postulated by Chomsky, Arnheim, or Freud, but rather on a variety of theories related to how we learn to organize the world around us so that it makes a little more sense to us now than it did when we arrived in it with that Jamesian sense of a "booming, buzzing confusion."
Theories of metaphor depend not only on our knowledge of film structure and linguistic structure, as described by such linguistic theorists as Whorf, Sapir, Hymes, and others, but upon social theorists ranging from George Herbert Mead to Erving Goffman. Knowing about metaphor means knowing how to organize the universe within our minds, knowing systems of myth, of grammar, of behavior, value, and art as they are defined by our group now, and have been in the past. I am afraid we learn about metaphor not only through "art" but also through symbolic behavior in general, through learning how to speak, by watching "bad" movies and reading "bad" books and "cheap" comic books. We learn about metaphor as we learn to stand close or far away from others and to do the million and one symbolic acts that our particular culture, in a particular context, teaches us as the traditional lore.
It is this traditional lore of symbolic and communicational behavior that helps us to know that in only some contexts are bananas to be thought of as penises, and violins as vaginas; and that in some contexts only, is an eye a gazelle. But most important, we know about intention, and the rules for structuring metaphor--in films, in speech, and in poems. That cultural knowledge makes it possible for us to use this marvelous structure of human thought.
It is the fact that we learn the agreed-upon rules for the intentional creation of meaning within specific contexts that makes metaphor possible. We learn the structure of metaphor, and we learn the schemata, the conventions, and grammars by which each creator in each mode of symbolic behavior intends meaning through transformations of a basic structure within which elements and contexts are shifted, made to clash and collide, and thus made to illuminate a complex and difficult world. A metaphor, like a great caricature, hits us like a blinding flash and forever reorganizes the world.
1. This paper originally appeared in New Literary History 6 (1974):195-209, as an invited discussion of a special issue on metaphor--Ed. Return.
2. A notion suggested to me by my reading of Professor Sparshott's paper and his intensive discussion of the concept of metaphor as a way of talking about something as it is not. Return.
3. The fact that this conception of "she" is at issue at present confirms the power of the cultural stereotype employed in these structures. Return.
4. Space does not permit me to develop any discussion about the problems of abstraction or formalization across visual and verbal modes of metaphor. Return.
5. The very notion of "portrait" is much more complicated than I can hope to deal with here. Historically, "portraits" are labeled not only by a known resemblance, but, as in the case of late Roman mummy portraits, by an assumed intention to present a likeness. In other historical contexts, a portrait can be "identified" by certain stylistic features: a face in a painting containing pimples, uneven ear sizes, or defects in physiognomy, which are assumed by art historians to have "meant" that the face was a portrait of "someone" rather than an idealized version "made up" by the painter. My point is that a "portrait" as a metaphor is a structure used as a label within a specific context. Neither portraits nor metaphors can be considered anything other than cognitions of a certain kind. (Much of my understanding of the complexity of the seemingly simple concept of "portrait" I owe to lengthy discussions of this subject with Debora Worth Hymes.) Return.
6. I single out Professor Sparshott's paper not because I mean to quarrel with it, but rather because it is both clear and presents almost all issues I wish to discuss. While I disagree to some extent with some of his formulations, I admire their originality and have found them profoundly stimulating to my own thinking. Return.
7. This point is developed in chapter 7, below--Ed. Return.
