The Evolutionary Significance of Metaphor

The Evolutionary Signi= ficance of Metaphor

Lyle Steadm= an

Department = of Anthropology

Arizona State University

Human Behav= ior and Evolution Society Meeting

Northwestern University

Evanston, Illinois

June 30, 19= 96

&nbs= p; Today, I’m going to try and convince you skeptics that the use of metaphor in general, but especially in religion, is aimed at promoting cooperation. And it is this particular effect, I shall argue, that has led to its increased frequency through the generation= s. Obviously, we are the result of in= tensive cooperation among our ancestors. Metaphor may be one of the mechanisms used to promote this cooperati= on.

&nbs= p; “The feature of metaphor that has most troubled philosophers is that it is ‘wrong:’ it asserts of one thing that it is something else̶= 1; (Percy, cited in Geertz 1973:210). To understand a metaphor, one must first identify a statement as fal= se and then decode it as a simile. “That man is a rat” must be transformed into the simile, “he is like a rat.”= A metaphor is a disguised simile.<= /p>

&nbs= p; Geertz notes this when he writes, “when it works, a metaphor transforms a fa= lse identification . . . into an apt analogy” (1973:211). To be more explicit: it is not the metaphor that does the transforming, but the listener. The listener himself must transfor= m the false statement into a simile. Thus a metaphor depends on the listener’s willingness to go along with the speaker, to cooperate with him. When one uses a metaphor, because it is false, one risks rejection: “What do you mean he’s a rat; he’s a man.” Somewhere, Shelley wrote that poetry requires a willing suspension of disbelief. The acceptance of a metaphor communicates a willingness to suspend disbelief, a willingness to suspend skepticism.

&nbs= p; Analogies are used everywhere, presumably to facilitate communication. And there are many ways of making = an analogy, including simply assign one small sound to a word, such as rat-like, or elephant-like. All languag= es, apparently, provide for explicit analogies, and analogies may be true: a pe= rson may indeed be like an elephant. Why then do people, everywhere apparently, also use false claims to communicate analogies? What is gained by = the falsehood?

&nbs= p; Because the meaning of a metaphor depends on the listener’s willingness to ac= cept a false statement as true, accepting a metaphor constitutes a kind of collusion. “You’re right, that guy is a rat.” Collusion is defined as “a secret understanding involving deceit.” While a metaph= or is not aimed at deceiving or defrauding anyone, the listener nevertheless, by = accepting the falsehood, shows his willingness to cooperate with the speaker. We like people who accept our metaphors. I’m suggesti= ng that the effect of the use of metaphors to promote a kind of collusion, a f= orm of cooperation, may account for their persistent use in all human societies. To mention only one example: humans universally, apparently, use kinship terms metaphorically to promise, and thereby encourage, individuals to behave as if they were close kinsmen.

&nbs= p; But my purpose today is not to concentrate on the delightful consequences of ordinary metaphors. It is in religion that metaphor achieves apotheosis. The distinctive feature of religio= us behavior appears to be metaphor, but metaphor of a special kind: metaphor t= hat is explicitly denied as a metaphor. And such denial, often passionate, promotes a far more intensive cooperation, [and] altruism, than that achieved by an acknowledged metaphor= .

&nbs= p; Religious claims, like metaphors, are distinguished by the fact that they cannot be s= hown to be literally true. And rel= igious claims, like metaphors, are meaningful to their participants. Religious claims not only fit the definition of a metaphor, but those making the religious claim behave as if they were metaphorica= l in non-ritual contexts.

&nbs= p; Elsewhere, Craig Palmer and I (1995) have argued that religion everywhere is identified not by belief but by a particular behavior: the communicated acceptance by = one individual of another person’s supernatural claim. Obviously, not all supernatural cl= aims are religious. What constitut= es a religious claim is the communicated acceptance of the claim made by another person. But such a religious claim is not = simply a meaningless untrue statement. Nor is it a lie, for it does not attempt to deceive. The meaning of a religious superna= tural claim must lie in its interpretation as a simile. If this is true, a significant eff= ect of religion everywhere must be to encourage a collusion that is far more power= ful than that achieved by ordinary metaphor, or perhaps by any other means, for= the communicated acceptance of a religious claim includes the denial that it is metaphorical, that it is symbolic. To make this clearer, let me give = a few examples.

&nbs= p; Some Australian Aborigines claim to be an actual kangaroo (or a witchety grub, or an owl, etc.) because their father was one, and his father etc. They say = they won’t eat kangaroo meat because it would be eating their kinsmen. Anthropologists are virtually unan= imous in agreeing that these Aborigines truly believe they are kangaroos. For example, the anthropologist Ar= am Yengoyan, who has spent some time studying an Australian tribe, told me, looking me straight in the eyes, that he knew Aborigines [who] believed they were kangaroos. [Yengoyan sta= ted that] “[o]ne old man was so sincere that he cried when he told me.” And it is probably= true that an Aborigine would pass a polygraph [or] a lie detector test. He’s not lying.

&nbs= p; Nevertheless, their non-ritual behavior belies their claim. No human kangaroo has ever been re= ported trying to mate with an actual kangaroo, none has ever been reported as resisting the killing and eating of actual kangaroos, as they certainly wou= ld if their kangaroo clan brother were killed and eaten. In fact, the human kangaroos perfo= rm a ritual which they claim encourages the reproduction of kangaroos so that ot= hers can hunt and eat them. Furthe= rmore, because clans are exogamous, one’s parents must be of different clans= . Thus, if your father is a kangaroo= , your mother must be a crow, owl, or witchety grub, etc. Everyone knows kangaroos are born = only from kangaroos. Thus, the beh= avior of human kangaroos belies their claim.

&nbs= p; Why do they claim to be real kangaroos? To identify a set of co-descendants—kinsmen—as kangaroos emphasizes that they are so much alike and so different from others they ar= e like a separate species, or catego= ry of nature. Such talk serves to encourage cooperation between the human kangaroos. A football team calls itself rams = or bears for the same reason, to emphasize their distinctiveness from other te= ams and their closeness to one another, as if they were a separate species. All of this is metaphorical. The crucial difference between spo= rting teams and Australian clans is the denial made by the clansmen that the clai= m is metaphorical, achieving thereby, much greater collusion.

&nbs= p; Roman Catholics claim that at communion they are consuming the actual blood and b= ody of Jesus. Yet Catholics know = what they are consuming comes from the vineyard and the bakery. And they cannot distinguish by the= ir senses the wine and bread that has been allegedly transubstantiated into Je= sus (by some words of a priest) from the wine and bread that has not. When they take communion, they anticipate that the “blood” will taste like wine and the “= ;body” like bread, and would probably vomit if it actually tasted like blood and flesh. And they deny being cannibals. Furthermore, they = would agree that the amount of “blood” and “body” consumed throughout the Catholic world weekly is far greater than the presumed dimensions of Jesus. Therefor= e, on the basis of this knowledge and behavior, we must conclude that the claim is metaphorical—that its meaning depends on implicitly converting the cl= aim into a simile: that it is like incorporating Jesus into one’s body.= Interestingly, Protestants often distinguish themselves from Catholi= cs by citing this very claim, saying that the wine and bread are obviously onl= y symbolic, that they are only symbolically incorporating Jesus. Thus, while both Protestants and Catholics use the claim metaphorica= lly, Catholics explicitly deny its metaphorical status and thereby, with that denial, achieve greater collusion. Of course, they both use other supernatural metaphors, which they de= ny as metaphor, to achieve intensive cooperation.

&nbs= p; The most widespread religious behavior, found apparently in all tribal societie= s, is ancestor worship (Steadman, et al 1996). What distinguishes ancestor worshi= p as religion is the communicated acceptance of the supernatural claim that a de= ad ancestor can influence, and be influenced by, his or her living descendants. This claim appea= rs to be metaphorical, for on the basis of the senses it is untrue, and is acted = upon as a simile. The communicated acceptance of such a claim implies a promise to act as if the dead ancestor were still alive and sentient, and still c= ared about us, even though one’s senses and experiences shout the opposite. And that is exactly how the followers act ritually.

&nbs= p; A ritual can be defined, that is identified, as stereotyped cooperation; its function, to promote un-stereotyped cooperation. The worship of a common ancestor i= s a ritual of co-descendants, kinsmen, aimed at encouraging their future cooperation. Such worship also encourages respect for their common ancestor, and his behavior. We propose that the identifiable e= ffect of ancestor worship that has led to its universal persistence is that it encourages cooperation between living kinsmen and respect for traditions, t= he two fundaments of human societies everywhere. Note that ancestor worship implies= the soul, for it is the dead ancestor’s soul that is the object of communication. And because ev= eryone at birth is a potential ancestor, everyone is said to have a soul. Thus, talk of souls, like ancestor= worship, is universal in tribal societies.

&nbs= p; I’m arguing that people do recognize implicitly, although they deny explicitly, that religious statements are metaphors, and they therefore behave accordingly. For example, to = accept the claim that a person has not died when he has died, is to communicate one’s willingness to act (in certain, usually ritual, respects) = as if he were still alive. However, because the statement is understood as a metaphor, those who communicate th= eir acceptance of the claim that they will never die do not take more deadly risks than individuals who deny the cla= im.

&nbs= p; I have attempted to account for the use of metaphor by arguing that all metap= hors promote cooperation, and that religious metaphor, because it is explicitly denied as metaphor, promotes much more intensive cooperation than ordinary metaphor. But I have also tri= ed to show you that religious behavior, and its persistence, can be studied objectively, thereby exposing the hypothesis to disproof. Up to now, the study of religion h= as been an attempt to explain why people ha= ve the religious beliefs they claim to have.&= nbsp; It is usually asserted that people have these beliefs because they m= ake them feel better, reduce fear or anxiety, or give them confidence, despite the fact that many religious c= laims would seem to increase anxiety and fear, such as talk of hell, witches, and sorcery. And these alleged beliefs seem to be almost infinite, limited only by our imagination. But beliefs of others cannot be identified by our senses; hence claims asserting their existence cannot be checked. This is probably why= the study of religion has not significantly increased our understanding of human behavior, and why many scholars of religion have called the study stagnant (Geertz 1996:1), or contradictory (Evans-Pritchard 1965:18), or muddled (La= wson 1989:305), or itself religious (Murdock 1971). Studying religious belief is like studying angels on the head of a pin.

&nbs= p; Humans do not identifiably respond to one another’s beliefs, but rather to t= heir behavior, including their talk of supernaturals. Regardless of beliefs, the communi= cated acceptance of a supernatural claim is not only crucial in religious rituals= but is mandatory for all followers. Because the beliefs of others cannot be identified, they cannot be acquired from others. Two peo= ple exhibiting identical religious behavior may or may not have the same beliefs. And no one can tell.= We must conclude that beliefs are irrelevant in the study of religion.

&nbs= p; We can objectively study the behavior, and hence the evolutionary function, of ancestor worship, magic, divination, shamans, priests, prophets, modern religions, the killing of people said to be witches, etc., if we rigorously limit ourselves to those hypotheses checkable, and hence disprovable, by our senses. The pretense that we = are studying other people’s beliefs call to mind George Peter MurdockR= 17;s (1971) indictment of anthropology’s theories as more like religion th= an science. We can do better.

BIBLIOGRAPH= Y

Evans-Pritchard, E.E.

1965 Theories of Primitive Culture. Oxford: The Clarendon Press.

Geertz, C.

1973 The Interpretation of Culture. New York: Basic Books.

1966 “Religion as a Cultural System,” in Banton, M., ed. Anthropological Approaches to the Study of Religion.&n= bsp; London: Tavistock Publications.

Lawson, E.T.

1989 “Explaining Religion: A Symposium: Introduction.” Religion, 19 (4): 303-4.

Murdock, G.P.

1971 “Anthropology’s Mythology.” Proceedings of the Royal Anthropologic= al Institute of Great B= ritain and Ireland. The Huxley Memorial Lecture, p= p. 17-23.

Steadman, L. and C. Palmer

= 1995 = “Religion as an Identifiable Traditional Beha= vior Subject to Natural Selection.” Journal of Social and Evolut= ionary Systems, 18 (2): 149-64.

Steadman, L., C. Palmer, and C. Tilley

= 1996 = “The Universality of Ancestor Worship.” Ethnology, 35 (1): 63-76.

[Addendum]

(Use somewhere?)

If the argument I have made above is accepted, that religious claims are metaphors denied, then every religious claim should be translatable into its simile. Thus, the acceptance of the claim distinguishing magic, that a certain technique = can supernaturally cure or kill, communicates that the technique is like curing= or killing a person. The communi= cated acceptance of the claim distinguishing divination, is understood and acted = on as if the divination technique provides the best or true answer. The same can be shown for all reli= gious claims.

(Add?)

Does an actor need to believe he is actually the chara= cter he portrays in order to give a convincing performance? The same might be asked of a person denying a crime. Clearly, hum= ans are capable of communicating great sincerity regardless of their personal beliefs.