ON THE CONTINUITY AND RANGE OF CERTAIN FIGURES OF SPEECH

Manuel Arce Arenales

Universidad de Costa Rica

Resumen: En este artículo argumento que no todas las metáforas pueden ser clasificables de la misma manera, independientemente de que uno asuma una posición como la de Grice o una aproximación

“directa” a la metáfora. En otras palabras, puede que algunas metáforas sean continuas respecto de fenómenos tales como la hipérbole o la aproximación, mientras que otras pueden ser de un tipo interpretativo distinto, específicamente uno que requiera de inferencias no lógicas dentro de un contexto dado. En cualquier caso, puesto que la metáfora es siempre no sólo dependiente del contexto sino dependiente de un contexto compartido, el conocimiento o la modelación de otras mentes es requisito necesario, y si éste es el caso no debería sorprender que los autistas, incluso si son capaces de manejar el habla literal ordinaria, no puedan comprender adecuadamente el lenguaje meta fórico.

Palabras clave: metáfora, tropo, inferencia no-lógica, dependencia del contexto, autismo.

Abstract: In this paper I argue that all metaphors may not be classifiable in the same manner, regardless of whether one takes a Gricean or a “direct” view of metaphor. In other words, it might be that some metaphors are continuous with such phenomena as hyperbole and approximation, whereas others may be of a distinct interpretive type, specifically one that requires a nonlogical inference within a given context. In any case, since metaphor is always not merely context dependent, but shared-context dependent, knowledge or modeling of another mind is requisite, and if this is the case not surprisingly autists, even if they can adequately handle ordinary literal speech, will not be able to adequately comprehend metaphorical language.

Key Words: metaphor, figures of speech, nonlogical inference, context dependency, autism.


29


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

In India, the Cow is a sacred animal and to this day is looked upon with great affection. A gentle, shy young girl is given the pet name Gaurie, little cow. She is addressed in these words:

Aav mari garib Gai (Gujarati)

Aau méri garib Gau (Hindi)

Literally “Come here my gentle cow.” The exact equivalent in English would be: “Come here my little lamb, my lambkin.” (Nanavutty, 1999, p. 26)

The metaphor “Sally is a cow” clearly has different interpretations in England, in Costa Rica, and in Hindu or Parsee India. It is equally true that the metaphor “Sally is a lamb” has very different interpretations in a country like Costa Rica as opposed to a country like Canada. According to the OED, ‘cow’ may be understood as Atimid, faint hearted person, a coward or the word may be Applied to a coarse or degraded woman. In Costa Rica, a person who ‘is a cow’ (es una vaca) is a particularly stupid or obtuse person, regardless of gender. On the other hand, someone who ‘is a lamb’ in Costa Rica (es un cordero) is a person (usually a man) who is submissive, a follower, someone without opinion of his own, who follows the direction(s) imposed by the group he happens to belong to at the time.

Metaphors are members of a class of linguistic constructions that rely for their correct interpretation not only on contextual data (talking of someone who ‘is a cow’, the person in questi on might have recoiled at the thought of crossing a swing bridge), but on shared [cultural] connotations (‘cow’ refers to a timid or faint hearted person). Notice that a metaphor is not simply a synonym: ‘cow’ is not a synonym of ‘faint hearted’ because when it is used to refer to a timid or faint hearted person it adds the nuance of a [gentle or domestic] animal that reacts instinctively. Out of a plethora of real or assumed characteristics of a named entity, in a given culture some are selected with the purpose of enhancing or enriching a particular meaning.

Cows of course are, of their own, neither coarse nor degraded nor particularly gentle or obtuse; it is a cultural bias that makes one associate any of these characteristics with this particular (female) mammal. In general, metaphors work on an assumed shared contextual body of meanings, associations, values, and sundry emotional links. Metaphors in fact range from stereotyped associat ed 30


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

meanings shared by all speakers of a given dialect (e.g. ‘estúpido’ [stupid] associated with ‘vaca ’ [cow]) to complex and subtle constructions built upon a dynamic interchange in which a shared context is created: such is the case in poetry, e.g.

Into her lying down head

His enemies entered bed,

Under the encumbered eyelid,

Through the rippled drum of the hair-buried ear;

And Noah’s rekindled now unkind dove Flew man-bearing there. (Thomas, 1971, p. 125)

Here the word ‘dove’ has to assume cultural associations hallowed by a tradition specifically pointed at by the use of the proper name ‘Noah’. By sheer juxtaposition, the words ‘Noah’ and ‘dove ’ acquire specific meanings that distinguish them from other instances, say my friend Noah Westby or any ordinary pigeon. But, and this is one of the most interesting points, ‘dove’ cannot be construed here to mean A gentle, innocent person (Forbes Inc., 2000, p. 540), the “fixed” metaphorical rendering of the word. Instead, the reader is forced to create an altogether new (and, one suspects, unique) metaphorical sense for ‘dove’, aided by the adjectives ‘rekindled’, ‘unkind’ and ‘man-bearing’; the term of comparison is absent or, more precisely, it’s never explicitly designated. So, aside from the bounding adjectives, one has the boundaries of what ‘dove’ is not; for example, even though in a sense it is NOAH’s bird, it is not the Holy Spirit (another forced association in this cultural context). A metaphor such as this one is an n-dimensional web of associations with a center (usually a rather ordinary noun) connected with constructs of all kinds, including metaphorical constructs, and bounded by precise connections with what it is not. The connections, however, cannot possibly be arrived at by strictly logical means, indeed by any exclusively analytical method, and have to be built using not only the shared background of cultural knowledge, but patterns of nonlogical inference as well. The full sense of the metaphor, however, is neither illogical nor capricious, for the assoc iations with some of the things it is, could be or is not are precisely determined. How is the correct construction of sense and meaning achieved? Not only by isolating the relevant cultural indicators, but by correctly modeling the representation intended in another mind, namely the mind behind the poetic voice.

31


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852 In the case of stereotyped association, sometimes the metaphor eventually yields a common

word: thus burrada simply means “stupidity” or “foolishness”, and can no longer be understood as “something that donkeys do.” At the other end we have constructions such as the following:

Un puño tengo de corazón

bajo los pies distingo las hojas sueltas (Arce, 2000, p. 55)

A fist I have for heart

Under my feet

I make out the fallen leaves .

In this case the meaning “I feel as if I’m suffocating, my feelings are like a handful of dry leaves that may be trampled underfoot even by myself, and yet they feel compact and hard, dense within my chest, impenetrable, aggressive, heavy and hurtful” cannot possibly be assigned by anything other than a nonlogical inference; this nonlogical inference is part of what is conveyed by the poem without being part of what is actually said in the poem. Again, the inference is neither illogical nor capricious (in fact, it refers to a possibly common human experience), and it is reached among other things by virtue of correctly modeling the representation in the mind behind the poetic voice.

In the case of “fixed” metaphors, a so called ‘high functioning autist’ should have no problems, or at any rate should face problems of the same magnitude he has when dealing with “ordinary” literal speech. After all, Bernal es una vaca is practically equivalent to Bernal es un imbécil: what we would expect the autist to miss are associated representations like imagining Bernal peacefully (and needless to say stupidly) munching grass on some meadow, perhaps dolefully mooing every now and then. For clearly even the simplest metaphor is more than a mere substitute or shorthand for otherwise easily accessible meanings by ordinary means. In the case, however, of highly complex metaphors of the type illustrated by the poems quoted before, we should expect the autist, ‘high functioning’ or not, to face insurmountable problems. Furthermore, it appears to me that metaphors stretch on a continuum from “fixed” metaphors to “poetic” metaphors, so I would expect some metaphors in between to more or less weakly (or strongly) demand nonlogical context - dependent inferences for full understanding.

32


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852 Let us now examine examples of the following type:

• Well, Harry is [almost] human today.

• Not to worry: Harry is simply being an ape .

Consider now two different contexts for these examples: (I) Harry is a member of the species homo sapiens sapiens; (II) Harry is a member of the species pan troglodytes. If we are operating within context (I), the first sentence forces us to focus on things that Harry is not: he is not considerate (say), nor ordinarily thoughtful or sensitive, characteristics that we freely and somewhat high - handedly bestow upon ourselves generally. The understanding is, of course, that Harry is by nature overbearing or uncouth (we could alternatively be pointing out that Harry today is surprisingly tidy, quiet or graceful). Notice that, depending on who is uttering the statement and why, it could be ironic or not (it could even be tender or affectionate in certain circumstances). On the other hand, if we consider the second sentence within the same given context, it directs our attention to (perhaps unexpected) aspects of Harry’s personality: it could be a mere statement of fact (hominids certainly are, after all and strictly speaking, apes), or it could be forcing us to focus on certain charac teristics that Harry has and that we do not tend to associate with ourselves: coarseness, lack of consideration for other people’s views or feelings, or else boisterousness or vulgar deportment. In the first case we could be merely explaining or even justifying Harry’s behavior; in the second case we could be dismissing Harry’s behavior as repulsive or else we could be accepting its repulsiveness while at the same time appealing for a compassionate attitude in view of his unintentional animalism. In any o f these cases, the crucial element that is needed to correctly understand the intended meaning is precisely the intention of the speaker: we are dealing here not merely with ad hoc concepts HUMAN * or APE*, but rather with ad hoc shared contexts .

The concept of ‘ad hoc concept’ is vital for relevance theory (henceforth RT), since “it is posited that a single explanatory process—the construction of ad hoc concepts—suffices to explain metaphor as well as a range of other phenomena of both (pre-theoretically) literal and figurative kinds. No special resources are required for the metaphorical case” (Wearing, 2006). In turn, “According to the relevance theoretic view, metaphorical content is not to be found merely in the implications of an utterance. Instead, it constitutes (at least in part) what is explicitly communicated when a speaker utters a sentence. For example, if a speaker says ‘Harry is a bulldozer’, she does not assert the proposition that Harry is a BULLDOZER (let us suppose that BULLDOZER picks out the concept encoded by the word ‘bulldozer’), but rather the proposition that Harry is a BULLDOZER*, where BULLDOZER * 33


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

is a modification of the encoded concept. The proposition expressed is therefore not the ‘literal’ content that we might associate with the sentence, but rather a proposition involving the ‘ad hoc concept’ BULLDOZER*, a concept capturing the metaphorical sense of the expression” (Wearing, 2006). The construction of ad hoc concepts, conceived within the central “relevance-theore tic commitments” of under determination of explicit and indirect communication and the determination of explicit and implicit content in terms of considerations of relevance (how to maximize cognitive effect Which roughly means, in Wearing’s words, “information gained” while minimizing processing effort), clearly envisions a linear, individual-centered process, but shared contexts are non- linear, collaborative products, for which the determination of an interlocutor’s intentions becomes a necessary input.

If we now turn our attention to context (II), the first sentence forces us to focus on things that Harry (unexpectedly) is: [unusually] considerate, thoughtful or sensitive, or else tidy, quiet or graceful. In this case it is very difficult to posit a possible ironic intention: instead, the most likely scenarios include admiration or pleasant surprise. The second sentence directs our attention to characteristics that Harry does not have but that we perhaps wishfully hope he had: in this case Harry’s nature excuses his behavior. As before, to correctly interpret the utterances we need to perceive the intention of the speaker, and therefore we need to have access to a shared context.

In context (I) we would expect an autistic person to miss the full meaning of the first sentence if it is not intended literally: this failure can be (at least partially) explained either by taking a Gricean or a RTapproach. Let us remember that, according to the Gricean implicature approach, metaphorical content is implicated by the saying of something which, if taken literally, is conversationally inappropriate. Therefore, metaphor becomes a pragmatic phenomenon of the same type as irony and indirect speech, and one can expect autistic speakers able to grasp literal meanings to have difficulties in understanding metaphors, precisely because they lack the necessary faculties to adequately interpret pragmatic aspects of communication. On the other hand, an autistic speaker’s misunderstanding of the first sentence in context (I) could fit tidily with Wearing’s claim that “what distinguishes the interpretation of metaphor from cases of literal speech is the relative importance of being able to manipulate information about what is not the case in the process of metaphorical interpretation”(Wearing, 2006); in this case, that Harry is not [fully] human. The case is similar to what we would expect to happen for the second sentence within context (II): if not meant literally, we can expect an autistic person to miss the idea that Harry is more than “just” an ape.

34


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

Unfortunately, both explanations miss what is one of the deciding aspects of metaphor: the determination of the intent behind the speaker’s utterance. In the case of the second sentence within context (II), a literal interpretation is possible if the speaker is trying to reassure a visitor somewhat alarmed by Harry’s display of ape-like behavior; its non-literal interpretation is possible if the speaker is trying to justify Harry’s antics in the presence of someone who has been made to expect something more from Harry than plain old ape-like conduct. In this case both the speaker and his interlocutor take Harry to be more than a mere monkey, so a strictly literal interpretation is not possible. However, a third party watching either exchange cannot possibly determine the correct interpretation without taking into consideration the speaker’s intent, in other words, without having some sense of the state of another mind, i.e., without the capacity to model the minds of others. On the other hand, a literal interpretation of the sentence ‘Well, Harry is almost human today’ in context (I) is only possible if the speaker actually believes Harry to be subhuman, something that is impossible unless we are capable of perceiving this belief in another’s mind, since the fact is that Harry is of course fully human, uncouth as he may be. A non-literal interpretation is possible if one perceives the speaker’s intention of drawing attention to Harry’s special behavior [today], or else to point out (by contrast) Harry’s usual unbecoming conduct.

The second sentence in context (I) presents a slightly different situation: if our autistic speaker is familiar with basic primatology, his understanding will be the simple statement of fact that Harry belongs to the super-family hominoidea. If the sentence is not meant literally, he will fail to understand that Harry is behaving in a way that makes him be less than what he can be expected to be. The first sentence in context (II), by contrast, is not subject to a literal interpretation: the focus is not so much the concept HUMAN or an ad hoc concept HUMAN*, but rather Harry himself, who now appears to be more than one could reasonably expect him to be according to his nature. In a nutshell, then, an RT approach seems capable of dealing with some instances of metaphor, but not with all of them or not with their whole range, something that is consistent with the claim that metaphors are not all classifiable in the same way. The same holds true for a Gricean interpretation.

Having arrived at this point, it is necessary to remember that “the relevance theoretic explanation treats metaphorical interpretation as importantly continuous with such phenomena as hyperbole and approximation, rather than as a distinct interpretive type” (Wearing, 2006). This might be true of some cases, but metaphors such as the ‘dove’ metaphor in Dylan Thomas’ poem seem to be of a different type than hyperboles or approximations: among other things, let us recall that in this 35


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

metaphoric construction the term of comparison is absent, in the sense that it must be created or constructed by the interlocutor on the basis of connections in turn based upon shared cultural contexts, and bounded by associations and specific indicators (such as the adjectives ‘rekindled’, ‘unkind’ and ‘man-bearing’). Hyperbole and approximation, on the other hand, have to always include a term of reference: that which is being exaggerated or that which is being approximated. In fact, in hyperbole and approximation one is compelled to start from the perspective of the characteristic one wishes t o exaggerate or from the perspective of the meaning one wants to draw near.

Even in the case of utterances such as ‘Harry is a gorilla’, though, which could be either hyperbole or metaphor if Harry is a member of the homo sapiens species, an understanding of intent and shared context is necessary in order to correctly interpret the utterance. If the shared context is one of gorillas as gentle giants, one might want to emphasize Harry’s physical strength (hyperbole) or else one might want to point out Harry’s salient characteristic as one of strength and tenderness intertwined (metaphor). However, if the shared context is one of gorillas as immensely strong but insensitive brutes who always resort to the use of physical force in order to impose their will u pon others, the speaker of such an utterance probably wants to emphasize Harry’s meanness (hyperbole) or else to select aggressive brutality as Harry’s salient characteristic (metaphor). The correct interpretation is of course also dependent on correctly assessing the speaker’s intention as one of endearment or one of animosity.

Let us now turn to a subject that is crucial in the context of this discussion: the relation between simile and metaphor. Apparently, there is evidence that ‘high functioning’ autistic people do not have problems with similes the way they do with metaphors, or at least that they encounter when dealing with them the same degree of difficulty they experience with “ordinary” literal speech (Happé (1993). Wearing, however, would prefer to treat similes as continuous with phenomena such as metaphor, in keeping with standard relevance theoretic assumptions, although she accepts Happé’s data in support of considering irony as a cognitively more complex process than metaphor. But it could very well be that similes, like metaphors, are not all classifiable in the same way, and that simple similes of the type John is like a bull have more in common with “fixed” metaphors than they do with examples of the following type:

The night came like a great lady, slowly dragging a long black coat studded with diamonds. 36


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852 How do we come about understanding that the twilight was extended in time, that the sky

was very dark but very clear, and that there were many stars that shone brightly? Even more, how do we understand that this particular night was of impressive majesty, and that there was an animated quality about it that helped one to feel how our ancestors were able to conceive it as a deity? It seems to me that the cognitive difficulties posed by a simile such as this one are not significantly less than those posed by a [complex] metaphor. In fact, one would expect an autistic speaker to fail in interpreting that the night is seen as a lady, that the sky is perceived as a coat, and that the stars are understood as diamonds. All of these perceptions are the result of non-logical inferences, made within the scope of a shared cultural context. However, the sense of awe that one feels is the result of sharing that very feeling with the mind behind the poetic voice. In fact, our sense of being humbled by an overpowering majesty is the reflection of the speaker’s feelings (the feelings expressed by the poetic voice), to which we would have no access were we to lack the capacity of modeling our mind and the mind of others. Assessing a state of mind, which in some of the examples of metaphor previously examined is a necessary condition for the correct understanding of the intended meaning, is in this case perhaps even more necessary, since an essential part of the meaning is the state of mind itself. A metaphor can, of course, embody this same quality, as we may verify by reexamining a pr evious example:

Un puño tengo de corazón

bajo los pies distingo las hojas sueltas

A fist I have for heart

Under my feet

I make out the fallen leaves .

Rather than fathoming the state of mind [the state of heart] behind the poetic voice in order to capture the purport of the poem, we must fathom the state of mind expressed because it is the meaning. Another interesting issue is the question of how exactly does the simile of our example

differ from the corresponding metaphor:

37


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

The night was a great lady that slowly dragged a long black coat studded with diamonds.

This direct anthropomorphization of the night in fact diminishes the perception of grandeur and mystery that we attach to it within the simile: night, after all, is more than any human could be, and the comparison with a great lady is done to bring attention to its living or animated quality, and to point out avenues or glimpses of superhuman deportment. Which brings us to what is perhaps the most obvious difference: the simile allows us to emphasize movement in a way that the metaphor cannot, and the quality of moving, of being able to move [perhaps purposefully], is essential to our perception of night’s awe-inspiring characteristics. Where the metaphor favors closeness in the viewer’s perspective, the simile favors distance. Furthermore it is possible to say that, generally speaking, in a metaphor the two component terms are fused, as it were, whereas in a simile they must remain clearly distinct.

In conclusion: the full understanding of both simile and metaphor demands a complex shared context, many times culturally determined, rich in [shared or equivalent] emotional associations, and therefore necessitating access to another’s actual, projected, or imagined inner state(s) or scenario(s); it also often demands access to another’s intention. It is therefore not at all surprising that autistic persons are totally or partially incapable of understanding metaphorical meaning, and either Gricean or RT explanations of metaphor (and possibly of simile as well) must be expanded to accommodate the full range of complexity it presents.

This article is based on an oral presentation given at the WORKSHOP ON LANGUAGE, CONTEXT AND COGNITION that took place in May of 2006 in Punta del Este, Uruguay. I take the opportunity to thank Rob Stainton and Carlos Caorsi for being wonderful hosts, and Catherine Wearing for providing me with stimulating ideas

Referencias

Arce Arenales, Manuel (2002). Visitas al desván. San José, Costa Rica: Editores Alambique. Arce Arenales, Manuel (2004). De leguas y minutos. San José, Costa Rica: Editores Alambi que.

Arce Arenales, Manuel (2005). Las huellas del zapatero. San José, Costa Rica: Editores Alambique. 38


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

Arce Arenales, Manuel (2005). Murciélagos de fuego. San José, Costa Rica: Editores Alambique. Bezuidenhout, A. (2001). Metaphor and what is said: A defense of a direct expression view of metaphor. Midwest Studies in Philosophy, 25:156- 186.

Davidson, Donald (1978). What Metaphors Mean, reprinted in Inquiries Into Truth and Interpretation (1984). Oxford: Oxford University Press.

Forbes Inc. (2000). The American Heritage Dictionary of the English Language. Boston: Houghton Mifflin Company

Grice, H. (1989). Logic and conversation. In Studies in the way of words, pages 3-57. Harvard University Press, Cambridge, MA.

Happé, F. (1993). Communicative competence and theory of mind in autism: atest for relevance theory. Cognition, 48:101- 119.

Happé, F. (1995). Understanding minds and metaphors: Insights from the study of figurative language in autism. Metaphor and Symbolic Activity, 10(4):275- 295.

Lakoff, George and Mark Johnson (1980). Metaphors We Live By. Chicago: Chicago University Press.

Lakoff, George (1990). Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: Chicago University Press.

Nanavutty, Piloo (translator and commentator) (1999). The Gathas of Zarathushtra: Hymns in Praise of Wisdom. Ahmadabad: Mapin Publishing Pvt. Ltd.

Oxford University Press (1985). The Compact Edition of the Oxford English Dictionary. London: Oxford University Press

Ricoeur, Paul (1977). The Rule of Metaphor. Trans. Robert Czerny. Toronto: University of Toronto Press.

Sperber D. and Wilson D. (1995). Relevance: Communication and Cognition. Oxford: Blackwell, 2nd. Edition.

Sperber D. and Wilson D. (2002). Pragmatics, modularity and mind-reading. Mind & Language, 17(1-2):3- 23.

Stern, J. (2000). Metaphor in Context. Cambridge, MA: MIT Press. Thomas, Dylan (1971). Collected Poems. New York: New Direction Books. Wearing, C. (2006). Metaphor and what is said. Mind and Language .

39


REV. HUMANITAS, 2008, 5(5): pp. 29-40, ISSN 1659- 1852

Wearing, Catherine. (2006). Autism, Metaphor, and Relevance Theory [DRAFT], presented at the WORKSHOP ON LANGUAGE, CONTEXT, AND COGNITION, May 2006, Punta del Este, Uruguay.

Wilson, Deirdre and Dan Sperber (2005). Handbook of Pragmatics. En Horn, Laurence. And Gregory Ward (eds.). Relevance Theory (pp. 607-632). Oxford: Blackwell .

40