Similarly, Kroeber (1909) found that just eight semantic elements (including those that distinguish sex, generation, and lineal from collateral kin) structure the kinship terminologies—diverse though they be—of many different societies. For example, the English terms “brother” and “sister” are distinguished by sex, “mother” and “grandmother” by generation, “son” and “nephew” by the distinction between lineal and collateral kin. Other elements have since been added; collectively they form a universal pool from which all kinship terminologies are drawn. The universal-pool kind of universal takes the form of “if A then some selection from B, C, D,…N,” in which A is any individual, society, culture, or language.
Hale (1975) discusses a kind of universal that has affinities both to those forming a universal pool and to near-universals. He gives counting as an example: peoples may have a very elementary system of numbers and yet have a full-blown ability to count (which allows them very quickly to adopt complex number systems when they become available and prove useful). Theoretically, he argues, numbers and counting could be absent among a given people (he gives no example), particularly if they had no need to count. And yet the ability to count is universal as an innate (and presumably specific) capacity of the human mind. Hale gives further examples from linguistics, where certain patterns are near-universals that probably result from innate tendencies that are only rarely blocked by unusual conditions.
Contrasting with “innate” universals are those that are “manifest” (Tooby and Cosmides 1989), and this is a distinction that must be borne in mind to make sense of some of the most influential statements on universals, such as those of Chomsky and Fox that were noted earlier. Consider the following: “Though the device of metathesis, like all linguistic universals, is in principle available to speakers in any language, it does not, of course, follow that every language must actually present examples” (Chomsky and Halle 1968:361). For Chomsky the interesting universals lie below the surface. Fox (1980:7) has taken an analogous position in anthropology. Thus the nuclear family, which is manifest on the “substantial, institutional level” rather than “at the level of process,” is not universal, while “the bonding processes on which it is based” are (Fox 1980:7). Most explicit anthropological discussions of universals are, as Fox also notes, heavily weighted toward those that are manifest.
Yengoyan (1978) suggests another contrast, by distinguishing the innate universals that Hale describes from “experiential” universals. An example of the latter is that all people have the experience of seeing that blood is red and, hence, that symbolic equations of red with blood are very widespread if not universal.
Another conception of universals is embodied in the “universal framework” or “universal model.” In this usage the universal or universals are not of central research interest but rather are tools for research on phenomena connected with them. The universal framework consists of abstract, analytical definitions, assumptions, and procedures that guide research on selected topics—in any and all societies. Malinowski’s (1944) framework for the analysis of cultures was probably the most famous in anthropology, but behind it was Wissler’s (1923) earlier and cruder model and perhaps one presented by Warden (1936). (Malinowski’s and Wissler’s models are described in chapter 3). Another example was Robert Redfield’s (1953) framework for the cross-cultural analysis of worldviews. Kearney (1984:39) summarizes most of the framework in a three-dimensional diagram whose parts are formed by such distinctions as that between self and other, we and they, human and nonhuman, nature and divinity—distinctions likely to give order to any people’s worldview. As the framework for his analysis of human kinship and marriage systems, Fox (1967:31) identifies four “principles” that underlie and give order to them all: (1) “The women have the children,” (2) The men impregnate the women,” (3) “The men usually exercise control,” and (4) “Primary kin do not mate with each other.”
The features of a universalistic framework are discussed by G. N. Appell (1973), who has developed one for the analysis of social structure in terms of the property relations that define and relate social statuses of all kinds. The validity of this kind of a framework rests upon the universality of its starting points: Appell’s framework rests upon the universality of social statuses and property. Universalistic frameworks are often proposed, used for a time, and discarded in favor of others.
Related to the conception of the universalistic framework is the distinction between “universals of classification” and “universals of content.” Such broad categories as religion or government are extreme versions of the former. To some anthropologists such categories seem meager in the face of the complexity to which they refer. Wissler’s (1923) “Universal Pattern,” discussed in the next chapter, consists mostly of universals of classification. A universal of content, on the other hand, is one in which the details of the phenomenon are themselves universal. Facial expressions of emotions and the coyness display (discussed later) are good examples of universals of content. Many universals stand between those that are purely classificatory and those whose content is specified in detail, precisely because it has been the aim of anthropological research to fill in at least part of the universal content of universals that initially were little more than delineated categories of research. The currently burgeoning studies subsumed under the heading of “gender” rest very largely on a universal of classification—the division of labor by sex—but have shown that there is at least a slender content to the universal. The low repute that universals of classification sometimes have, due to the variability of their contents, should be balanced against consideration of their fecundity or open-ended nature. The enormous range of what humans do with language, for example, in no way leads us to undervalue it.
Universals are importantly distinguished as either emic or etic. In anthropology, “etic,” by analogy with phonetic analysis in linguistics, refers to analyses in terms of cross-culturally valid, scientific frameworks (universalistic frameworks).4 “Emic,” by analogy with phonemic analysis, refers to the way the natives conceptualize things. In English, to illustrate the original linguistic conceptions, the p of “pin” and “spin” are phonemically the same. English speakers, unless they are also phoneticians, do not meaningfully distinguish them. But to a phonetician, these are two different sounds: the p in “pin” is aspirated (accompanied by a puff of air), the other is not. (Hold a thread or thin strip of paper closely in front of your mouth while pronouncing these two words, and the difference between the p’s will be readily apparent.)
To say that a universal is an emic universal, then, is to say that it is a part of the conceptual system of all peoples. It is a part of their culture that is meaningful to them and that is probably more or less integrated into their worldview. Applying this standard to religion, Hockett (1973) denies that it is universal: the Menomini Indians (and no doubt many other peoples) do not conceptualize religion as a distinct sphere of their culture. It is a commonplace in anthropology that institutionally distinct spheres in one culture may be merged in others. In general, and to simplify, sociocultural evolutionary advance has been marked by the progressive emergence of, say, distinctive governmental, religious, and economic spheres from an originally diffuse kinship sphere. In cases where this institutional proliferation and specialization have not occurred, we usually say that economic functions, for example, are discharged by kinship institutions; we do not usually say that economic functions do not exist if the category “economy” is not a part of the native system of categories (even though we do want to note this, and it may well be a significant factor in understanding native life).
By the same token, it would be an unusual linguist who would say that if a people does not have a conception of grammar they or their language has no grammar. Hockett is a linguist, and in most cases I treat universals the way he would treat grammars: if it is there it is there, whether the natives are aware of it or not. For example, chapter 4 is entitled “Incest Avoid
ance” rather than “The Incest Taboo” because the former is an etic concept that may well be universal, whereas incest taboos have numerous emic variants and pretty certainly are not universal. In short, I generally define universals etically, but at times they are also conscious, verbalized, emic parts of culture. Emic universals may be of special interest, but there is no reason to confine the discussion of universals to them alone. (Failure to realize that one party is speaking of emics and the other of etics is a fertile source of disagreement among anthropologists.)
A distinction not currently in the anthropological literature, but that I believe will be useful, divides universals of “essence” from those of “accident.” “Intrinsic” and “extrinsic” universals might also be suitable terms. Chomsky and Halle (1968:43) use the terms “essence” and “accident,” and they appear to overlap with the distinctions noted earlier between formal and substantive universals. Universals of essence are those that could not be eliminated except by unnatural interventions (e.g., by genetic engineering, or in concentration camps). Universals of accident are those that we can more easily (and realistically) imagine not being universal. The purest examples of the former are those strictly biological features of the species that do not generally receive much anthropological attention (Malinowski and Fox providing notable exceptions). Spiro illustrated the latter by arguing that if the Oedipus complex is universal, it is so in part by accident: we just never find the (easily imaginable) conditions that would eliminate it. Fire and cooking also illustrate the latter type.
Universals of essence at the level of the individual collectively constitute human nature, or at any rate would be important ingredients in its definition. Universals of essence at the level of society and culture—if such universals exist—constitute the nature of society and the nature of culture and would probably express the logic of sociocultural integration and development. Possible examples at the cultural level would be provided by those implicational universals that in a generalized form assert that wholes do not precede parts (an example, previously mentioned, is that the locomotive does not precede the wheel). Perhaps it will prove more useful to see the distinction between essential and accidental universals as parts of a continuum of more or less fixed universality. I take it that much of the current debate concerning male and female differences turns round the issue of whether certain of the universal differences are essential or accidental.
Another term that needs to be introduced is “new” universals: those traits or complexes that were not present in all the societies known to ethnology but that have become universal in all extant societies. Tobacco is a very likely candidate (Aginsky and Aginsky 1948). The dog and metal tools are among the equally likely candidates, and such items as plastic containers, phosphorus matches, and machine-manufactured clothing do not (alas for the romance of anthropology) seem far behind. With each decade the new universals must swell in numbers, as the once rich diversity of cultures is steadily eroded.
It might also be useful to designate “former universals” for all those experiences that were once the common lot of humanity but have since been eliminated in some populations. A few related examples are high infant mortality, relentless childbearing and nursing for most women, and relatively rare experiences of menstruation (Harrell 1981; Ward 1963:37), but one could also note transportation by foot only, the virtual absence of impersonal relationships—and very much more. Since many former universals were part of the environment in which human nature evolved, keeping them in mind will be particularly useful in any attempt to understand that human nature.
Finally, it should be noted that universals are sometimes stated in the negative. Thus it is said, for example, that no society is a matriarchy (Bamberger 1974) and that no society’s music is composed exclusively of notes of equal length (Nettl 1983:40). This list could also go on for quite some length. Chomsky often states universals in the negative and with the implication that at some deep level there not only are no exceptions but could not be exceptions (see, e.g., Chomsky and Hampshire 1968).
The first and most obvious point about the demonstration of universals is that it is never done by exhaustive enumeration, showing that a phenomenon exists and existed in each known individual, society, culture, or language. There are too many known peoples to make this feasible, and there are too many shortcomings in the descriptions of “known” peoples. Thus all statements of universality are hypotheses or arguments based on various limited kinds of evidence (this is not to say that confidence in the validity of the hypotheses or arguments does not vary).
One way of constructing an argument for universality results from wide reading in the anthropological literature, which gives weight to statements that one has never seen reliable reference to an exception to some proposed universal. This is often coupled with the stated or tacit view that no convincing reason is known as to why the trait shouldn’t be universal. The anthropologist best illustrating this kind of argument was George Peter Murdock, founder of the Human Relations Area Files (HRAF). These files contain coded information on a wide range of societies and were specifically designed to sample and quantify the range of ethnographic traits and complexes (the sample size was originally 400, which was thought to be slightly more than 10 percent of all societies known to ethnography and history [Murdock 1975:xii]). Murdock’s broad command of ethnographic literature lent authority to his list of human universals (1945).
Murdock’s feat can now be duplicated by use of the HRAF to test the hypothesis that something is universal (for an example, see Otterbein 1987). The result is not to show that it occurred in all societies but that in none of the societies sampled is it found not to exist, which allows the reasonable inference that if not universal it is at least nearly universal. The larger the sample of societies from the HRAF, the stronger the inference.
Most anthropologists, it should be noted, do not have that overall familiarity with ethnography that Murdock possessed, and they don’t use the HRAF either. Yet on some specific topic they may read as widely as Murdock did, so that on that topic their assessments of universality carry considerable weight.
Another quite different kind of argument rests on a few, sometimes only two, ethnographic observations—but the observations are carefully conducted. In this argument the societies must be very different in most respects, so that when some trait or complex is shown to be present in each, one suspects that it may well be common to all societies. When Paul Ekman and his associates (1969), for example, wanted to provide a particularly convincing test of the idea that the emotional meanings or determinants of certain facial expressions are universal, they visited tribal peoples of New Guinea who had only recently made their first contact with Westerners and showed them photos of Western persons displaying various emotions through facial expression. As noted in the previous chapter, Ekman’s team found that most of the natives could accurately identify several emotional expressions. Ekman labels these emotions as “basic” and posits that they are universal. Most anthropologists, by virtue of their experience with other peoples, can confirm this universality, and Ekman’s claim has not been challenged in spite of the extremely limited sample he used (though it should be recalled that he had previously tested his findings among several more accessible peoples).
The recent film “First Contact,” compiled from footage produced in the 1930s when numerous New Guinean natives first met the white man (Australian prospectors), allows almost anyone to re-create the sort of experiment Ekman’s team conducted. The New Guineans and Australians had strikingly contrasting cultures—as the film amply documents—and yet the attentive viewer can see numerous ways in which the two peoples were fundamentally the same: facial expressions, gestures, and aesthetic appreciation being only a few of the obvious examples (see also Connolly and Anderson 1987).
Eibl-Eibesfeldt’s (1979) photos of a coyness display by an African native woman is a persuasive demonstration of the universality of the display, even though it provides e
vidence from only a single people, which of course we implicitly compare with the evidence provided by ourselves and other peoples we may know. Given the complexity of the display and the conditions that elicit it, the odds are simply too heavily against it developing independently and by chance in the same form among disparate peoples. (It is as if we discovered that in unrelated languages people expressed the idea of “I love you” with phrases that actually sounded like the English “I love you.”)
Another mode of arguing for universality consists of dispensing with those cases that have been set forth as evidence against universality. Several examples were given in the last chapter.
A potent kind of argument for universality might come about as a byproduct of a determined but unsuccessful attempt to show that a trait or complex is not universal. Feminist interests, for example, led to diligent searches for examples of female-dominated societies. When feminist anthropologists now say that there is no substantial evidence that such societies ever existed (Bamberger 1974, Ortner 1974), their conclusion carries a certain weight.
An important ingredient in some arguments for universality consists of providing a convincing explanation for the universal. When we can understand the conditions that produce a universal, and get a sense for the ubiquitousness of those conditions, we then more readily accept the universality of the trait or complex.
Even if all methods of demonstrating the universality of a trait or complex have been employed, it bears repeating that universality has still not been proven. However, the likelihood that the phenomenon is at least a near-universal, and for that reason a significant part of human nature or the human condition, may then be very great. Most of the phenomena I discuss (particularly in chapter 6) have not been demonstrated to be universal by all available means, and renewed or continued attempts at such demonstration are now in order.
Human Universals Page 8
