REPLY TO ERNIE LEPORE AND MATTHEW STONE

There is a big difference between metaphor and allegory. An allegory presents us with a sequence of actions that would not be difficult to interpret literally. At the beginning of the Divine Comedy Dante encounters three animals in a wood: a leopard, a lion, and a she-wolf. It would not be a problem to understand the passage literally, yet legions of commentators, sensing an allegory, hastened to explain what those three animals stood for. In other words, we are looking at an allegorical interpretation of Holy Writ. Nevertheless, if some wish to interpret these allegories literally, they can do so perfectly well without any sense of shock, after all there are people who interpret every passage in the Bible literally. In addition, I would like to pause briefly to discuss an example given by Lepore and Stone, the Matt Groening quote: “Love is a snowmobile racing across the tundra.” At first sight it does not strike me as a metaphor but a simile, “love is like. . .” (the responsibility for which I leave to the author). It would be an allegory if the word “love” disappeared and all we had was a snowmobile racing across the tundra. The story could then be understood literally, and there would be no metaphor.

Let’s move on to metaphor. With metaphor things go differently. If someone says, to use the example provided by Lepore and Stone, “the girl is a birch,” any sensible person would have to agree that that someone is saying something that does not correspond to what we believe to be true—and therefore the speaker is either lying or mad or meant to say something else. If we want the act of communication to be successful we must first of all (i) decide whether the speaker meant something else, and then (ii) decide what she wished to say.

The two problems are different and it seems to me that much analytical philosophy has dwelt considerably on the first point but little on the second and I think that the objections made by Lepore and Stone suffer from this limitation.

There is no doubt, as the two authors say, that “the whole point of uttering something with the intention of getting information across to an audience is that what the speaker intends to communicate should be available to his audience”; but who says that a speaker always intends to communicate a clearly defined concept? What did Keats mean to say when he stated that “beauty is truth, truth is beauty”? We are not looking at a metaphor but certainly at a case in which the poet did not want simply to say (as a scholastic theologian would have said) that “verum et pulchrum convertuntur” or that the whole of being is at the same time true and beautiful. Frankly, it does not matter to me what Keats meant to say (I find his expression very interesting even though its meaning is rather obscure, but poets often intend to speak in an obscure fashion). I simply wish to acknowledge that the speaker often does not intend to communicate a clear, definite meaning and indeed takes pleasure in being vague and ambiguous—but despite that he wishes to communicate something to his audience, otherwise he would not have spoken.

Lepore and Stone connect this idea with the fact that I talk about open texts, and I would agree on that. Except that, strangely, they say that they do not understand why I consider the outcome of an approach to an open text to be an interpretation. I must say that I am astonished because this idea is in contrast with all that I have written on interpretation. But it is also at odds with good sense: perhaps they want to say that the many readings that can be made of many poetic texts do not represent acts of interpretation? In this sense the concept of interpretation is restricted to the understanding of literal expressions referring to everyday situations (such as “there’s a cat in the kitchen” or “pass me the salt”) and the possibility of their being paraphrased and translated.

I ask Lepore and Stone what different attitudes are implied by these two expressions: (i) “I do not like that pig” and (ii) “William was a bachelor.” In the first case, if the expression refers to a human being, it should be interpreted as a metaphor and one possible interpretation is that the one being talked about is a filthy person; but we might also think that he is a complete glutton, or that he has deplorable sexual habits. It depends on the context. In case (ii) we have to decide whether William was an unmarried man or a knight’s squire, and we have to interpret (ii) by deciding which is the contextually appropriate interpretation. And so we are looking at two processes of interpretation and I cannot see how there are substantial differences between them.

Lepore and Stone must surely know that I hold to the Peircean notion of interpretation whereby an interpretant, which translates a sign through another sign or another more complex expression, always says something more. To interpret the expression “Smith has killed Brown” does not just mean understanding that someone called Smith has used violent means to take the life of someone called Brown (apart from the fact that the mention of violent means already tells us something more), but also that Smith should be sentenced to a certain punishment, that from a moral standpoint he is a sinner, that from the standpoint of many revealed religions he will go to hell, and very many other things. If these things are not considered, interpretation is reduced merely to the discovery of synonyms, which represents a pretty rough and ready way of understanding the meaning of an expression.

If I ask Word to give me a synonym for to kill, the program suggests slay or eradicate, but I do not think Word is interpreting anything nor that it is explaining the meaning of to kill to us.

What did the speaker mean to say when he said that the girl is a birch? That she is supple and graceful, or that, as Lepore and Stone suggest, “she bends under social forces”? But by doing this does it not happen that they are trying to interpret the metaphor?

Let’s go back to the analytic tradition. They cite Searle, who thought that when faced with a metaphorical expression we must (i) first of all identify its literal nature, then (ii) maintain that if we were to interpret the expression literally we would deem it inappropriate, then (iii) suppress the literal representation and look for “nonliteral forms of interpretations.” Very well, but it is point (iii) that begins to interest a semiotician who wants to explain what happens when we encounter a metaphorical expression. It is not enough to decide what the author’s intentions were, but one has to posit what the expression qua text may signify. The entire problem of the metaphor starts from that point onwards.

But from that point onwards it seems that Lepore and Stone are no longer interested in the problem. Yet it is precisely the moment of interpretation that becomes crucial. For example, many generative theories of the metaphor have been attempted, according to which the steps through which an author comes to construct a metaphor should be indicated. But the mechanisms of invention are largely unknown and sometimes a speaker produces new metaphors by chance, by quick association of ideas, or even by mistake. It is easier to take the opposite way, that is to say, to follow the procedures through which metaphors can be interpreted. From the steps of the interpretive procedure it is possible to infer the steps that the speaker would ideally have followed to produce the metaphor.

In the process of interpretation, we discover that it is not enough to take account of the similes that we can identify on the basis of our experience of the world; there are also those that have been established in the past by the entire encyclopedia of our knowledge. When Malherbe writes about a dead young girl who, a rose, has lived as long as roses do, (and that is to say) the “space of one brief morn,” on the one hand we must bear in mind the many cases in which a beautiful young girl has been defined as being like a flower, and then the current opinion whereby a rose blooms one morning and dies in the evening (which is not botanically true); and if we know that the rose does not live for one single day, we must then still interpret metaphorically the expression “the space of one brief morn,” understanding it as “a short time.”

It seems to me that, assuming a restricted notion of interpretation, we can take account only of semantic representations in dictionary format, and a dictionary can record only relationships between hyponyms and hyperonyms, or the relationships from genus to species that permit inferences by entailment. Unfortunately, on the basis of this model we can generate only synecdoches (pars pro toto or totum pro parte). Let us see, instead, what happens with the metaphors that Aristotle in the Poetics called of the third type.¹ Here we have a transfer from species to species (or from hyponym to hyponym) through the mediation of the genus (or of the hyperonymous term). But when one speaks of “the tooth of the mountain,” where peak and tooth are part of the genus “sharpened form,” it is clear that this example does not rest on a simple dictionary-like representation. In going from tooth to peak there is something more than a passage through a common genus. In dictionary terms tooth and peak have no common genus and the property “sharpened” is not at all a dictionary property of these words. The metaphor works because it singles out, among the peripheral properties of both tooth and peak, a common feature that is taken as generic only ad hoc—for the sake of that particular context. The same happens with Aristotle’s four-term metaphors. In the classical examples of the shield as the cup of Ares and of the cup as the shield of Dionysus, cup and shield would become interchangeable—in terms of a dictionary—only insofar as both are species of the genus “object,” which cannot explain the power of these metaphors. Instead we have to consider that—in terms of a certain description—both are concave objects. But the interesting thing about the two metaphors does not lie in the fact that shield and cup possess a common feature but in the fact that, in spite of this, we are struck by their differences. From the point of view of an imputed similarity we discover a contradiction between the properties of Ares, God of war, and those of Dionysus, God of peace and of joy (as well as a contradiction between the shield, an instrument of battle and of defense, and the cup, an instrument of pleasure). But in order to trigger all these inferences the metaphor must rely upon background knowledge, that is, not on a dictionary but on an encyclopedia. Let us consider a classical biblical example from King Solomon in Song of Songs: “Thy teeth are as a flock of sheep which go up from the washing.”² If we mean by sheep only “ovine mammal” we will never understand the beauty of the metaphor. In order to understand it we have to make some very complex inferences: (a) to decide that flock is a mass noun that must contain a feature like “plurality of equal individuals”; (b) to recall that for ancient aesthetics one of the criteria of beauty was the unity of variety (aequalitas numerosa); (c) to assign to sheep the property “white”; (d) to assign to teeth the property of being damp. Only at this point does the dampness of the teeth, white and sparkling with saliva, interact with the dampness of the sheep emerging from the water. In order to obtain this interpretive result, it was necessary to activate only some properties (among the more peripheral ones) while all the others (for example, the more than important fact that sheep are ovine and the young girl is human, rational, and two-footed) were narcotized. Naturally, a “satisfactory” interpretation remains merely virtual. There are readers who give a less dense reading of Solomon’s metaphor and others who could go further in establishing possible contextual interactions on the basis of a more complex encyclopedic competence. Since an encyclopedic competence is potentially unlimited—and since the idea of encyclopedia itself is a theoretical postulate—in order to interpret a metaphor an ideal reader is supposed to single out (and to organize in an ad hoc format) only the features that the context suggests as the most relevant for metaphorical interaction. Thus, (i) metaphor is above all the effect of a process of interpretation; (ii) this interpretation concerns above all the choice of a certain encyclopedic representation in which certain entries are focused on and others bypassed; (iii) this choice is determined by contextual hypotheses.

What happens in this process? In interpreting metaphor, as Aristotle wished, we know something more than we thought, so that the metaphor puts something before our eyes. Black said that metaphors often require us to reorganize our categories (“some metaphors enable us to see aspects of reality that the metaphor production helps to constitute”).³ In other words, a creative and original use of language obliges us to invent a new organization of our encyclopedia. In this sense a metaphorical interpretation—insofar as we build up tentative encyclopedic descriptions and make some predicated properties more pertinent than others—does not discover but rather posits or produces similarities. “It would be more illuminating . . . to say that metaphor creates the similarity than to say that it formulates some similarities antecedently existing.”⁴ In this vein I can mention my old reading of various metaphors from Semiotics and the Philosophy of Language and From the Tree to the Labyrinth.

Lepore and Stone agree that when Donne writes, “No man is an island; the belle tolls for thee,” he leaves the conclusions open for his readers. (By the way, this is a case in which the conclusions are not all that open; what Donne wants to say strikes me as being comprehensible in the same way for every reader.) But then Lepore and Stone conclude that “he’s not packaging them into meanings.” I really do not know what Lepore and Stone intend to say here by “meaning.” It seems to me that Donne wants to reveal one or more meanings to us and not none.

On reading Lepore and Stone’s text I think that they might agree with the various analyses I have suggested for some metaphors. The only point of disagreement is that for them the interpretation of a metaphor has nothing to do with their concept of the interpretation of literal expressions, and all that an interpreter may draw from a metaphor has nothing to do with meaning. We are faced with two notions of interpretation and meaning that cannot be compared. My objectors move within an analytic paradigm and I move within the theory of interpretation and meaning that emerges from Peirce’s semiotics. It is like getting a Muslim Arab and a German (Christian or not) to agree on the advantages of eating sausage with sauerkraut. Lepore and Stone assert that “the metaphor has its own mechanism that distinctively applies to metaphors and only to metaphors” and that metaphors produce not an interpretation but a kind of “excitation.” I do not understand why the excitation of the imagination cannot spring precisely from an interpretive effort.

Lepore and Stone insist several times on the fact that interpretation concerns only the speaker’s meaning whereas I have always insisted not on the speaker’s intention but on the intention of the text. A metaphor appears in a text even if the speaker (or the writer) invented it without realizing. Once we recognize it we must understand how many consequences could be drawn from it. And still following Peirce, I maintain that meaning concerns all the illative consequences that we can draw from a sign or a given expression. Lepore and Stone maintain that this has nothing to do with communication, but certainly my idea of interpretation goes beyond the idea of communication. I can interpret a cloud as a warning of rain even though the cloud had no intention of communicating anything to me. The interpretation of so-called natural signs has nothing to do with communication but it does have something to do with the mechanisms of interpretation. See my autobiography in this book for my objections to the objections of Gilbert Harman.

Perhaps we shall never manage to agree on this point even though, as I deduce from their text, Lepore and Stone have the same capacity as I do to make a metaphor work in order to draw excitation from it.

They insist that you cannot draw propositional content from the interpretation of metaphors and I wonder why, on analyzing some metaphors, I have drawn propositions that cast light on what they made me think about.

I have nothing else to say and I have tried to clarify my idea of metaphor. I am nonetheless grateful to Lepore and Stone for the care and scrupulousness with which they have analyzed my texts and explained their point of view.

U.E.

NOTES

  1. Aristotle, Poetics, 1457b1–1458a17.

  2. King Solomon, Song of Songs, 1.9.

  3. Max Black, Models and Metaphors (Ithaca, NY: Cornell University Press, 1962), 44–45.

  4. Ibid., 37.