Dark Matter of the Mind
Page 13
Rokeach spends much time discussing another set of values, what he labels “immanent values”—values that hold universally by virtue of their very nature: “Thou shalt not kill,” “Thou shalt not commit adultery,” “Treat the weaker fairly,” and so forth. It is far from clear to me, however, whether such values actually exist. Therefore there is little more to be said about them here. They fall mainly in the domain of axiology and are orthogonal to our current objectives of understanding how dark matter arises and comes to guide our actions and thoughts.
This is enough to get our discussion of values, culture, and dark matter under way. But for values and value ranking to be of utility in the anthropological enterprise, we need to say how they might be studied—the methodological requirements for a study of values. The first methodological question that arises is how one goes about identifying the values of individuals and groups. One obvious way is to follow the method employed in the famous Rimrock study, headed by Clyde Kluckhohn and his wife, Florence.
In 1948 the Laboratory of Social Relations at Harvard University planned and undertook groundbreaking research entitled the “Comparative Study of Values in Five Cultures,” or more familiarly known among those close to the research as the “values study.” The site of the study was one of the most austerely beautiful regions of the United States—Rimrock, New Mexico—where Clyde Kluckhohn, the nominal leader of this project, had conducted research among the Navajos since 1936. The values study emerged from his (and others’) “value orientation theory.” This Harvard research group launched a pioneering series of studies of human values, completing the values study in spite of Kluckhohn’s sudden death in 1960 at age fifty-five.
In the Harvard study, the research team examined five separate cultures in the Rimrock area, cultures of particular interest because they shared the same environment but with significantly different ways of life. It was hypothesized that since the material environment was identifical, differences in the ways of life would derive exclusively from cultural differences, in particular to differing value systems. Linguists call differences in the same environment contrast—as /p/ and /b/ contrast in the same environment of pat and bat (spelling notwithstanding, looking only at the phonemes). The five cultures were Navajo; Zuni; Mexican-American; Texan and Oklahoman farmers and ranchers; and Mormon. Each of these represents a separate community and culture.
Kluckhohn prepared a statement on the project that included the following: “There is general agreement among thoughtful people today that the problem of ‘values’ is of crucial importance, both practically and from the point of view of scientific theory” (Vogt and Albert 1966, 1).
Following the publication of Vogt and Albert 1966, however, the study of values in anthropology eventually petered out. According to D’Andrade’s (1995, 13ff) assessment, “The results [of the Harvard Rimrock study] were generally agreed to be disappointing. The major problem seems to have involved the identification of values. If a universal framework was used, like Florence Kluckhohn’s universal framework for the analysis of values, specific cultural values were left undescribed and unanalyzed. But no procedures had been developed to determine specific values.” D’Andrade’s criticisms are cogent and appropriate. And I suggest an answer directly.
The problem, as D’Andrade describes it, is reminiscent of the criticism of Chomsky’s research program in linguistics. For example, there is no good treatment in that theory for either (i) the nonoverlap of supposedly universal principles in either abstract or concrete structures of specific languages (e.g., no two “passives” work exactly alike) or (ii) there are no overt universals at any interesting level (see N. Evans and Levinson 2009).5 Universalist theories of socially affected cognition and behavior or linguistics often run aground on the reefs of the details. Boas ([1911] 1991) was one of the first to point out in modern times that we ignore diversity at our scientific peril.
Gallagher (2001, 1–2) summarizes the Rimrock study in the following way:
They hypothesized that “. . . there are a limited number of common human problems for which all societies at all times must find some solution . . . How a group is predisposed to understand, give meaning to, and solve these common problems is an outward manifestation of its innermost values, its window on the world: its value orientation.” The five common human problems, posed as questions, that provided the most useful “value orientations” in creating a cultural typology were:
What is the temporal focus of life? (Time orientation)
What is the modality of human activity? (Activity orientation)
What is the modality of a person’s relationship to others in the group? (Relations orientation)
What is the relationship of people to nature? (Person-nature orientation)
What is the character of innate human nature? (Human nature orientation)
. . . From their research they deduced that societies would respond in one of three ways to each of the five questions or orientations (figure 1 [figure 2.3 in this volume]).
Figure 2.3
This passage at once illustrates the innovative and very important nature of this research, as well as its fatal flaw: universalism. Assuming that all people share the same values or that similarly labeled values are the same values is unwarranted. The paragraph that follows immediately below is interesting in that it shows that the Harvard group had grasped the idea of value rankings, though not the concept of violable values.
In 1961, Kluckhohn and Strodtbeck published their theory and findings in their book, Variations in Value Orientations, in which they proposed that the rank-order of preference—from most to least—gave the society its cultural character. The different patterns of rankings allowed one culture to be distinguished from other cultures. It was this rank-order of preferences, they argued, that was the foundation for the more-visible cultural values, beliefs, norms, and actions—and even heroes, rituals, songs, etc.—of the society. They also proposed that, although a society may have a general preference that is dominant, there is a great deal of diversity within cultures and all cultures will express all possible dimensions at some time or through some individuals. Carter (1990) added to these propositions with his finding that cultures could share the same rank order of dimensions, but differ substantially if there was relative difference of preference for each of the dimensions. (Gallagher 2001, 3)
Numerous researchers (see D’Andrade 2008 for a detailed criticism and review) have criticized the Kluckhohns’ study for conflating values into an artificial set of “orientations.” Rokeach’s body of work addresses, in my opinion, many of those shortcomings, however, and could serve as a useful basis for a typological comparison of values. On the other hand, that said, there is no reason to suppose that all values—nor even most, apart from biological values—will be universal. (And in fact, perhaps not even all biological values are universal given the enormous variation in their rankings and interpretation cross-culturally.) The Rimrock study failed, if this is even the correct judgment, because of a failure to sufficiently articulate its theory of value ranking and, most important, for its unwarranted assumption that values are universal. Yet the study has rightly been seen as pioneering. It is an extremely important milestone. In fact, it may have been due to this study that even philosophers began to recognize the significance of values and their rankings.
What seems common to both Harris’s inadequately developed notion of cultural contradictions and the Rimrock study, in retrospect, is that both treated values like lists of inviolable constraints (this is also true of newer studies such as Graeber 2001). In other words, if two values conflicted (say, honesty and accomplishment) a quandary arose—how to understand why one value was respected while another was not. People understood, of course, that not all values received equal priority. Davidson even refers to “value ranking,” in the following:
Now suppose that our judge owns the house coveted by B and C and that she has decided to sell it to one of them. We can imagine that ther
e are . . . distinct steps in her reasoning, insofar as it involves the desires of B and C. First, she determines what she can of their preferences, their value rankings, perhaps on an interval scale. Second, she compares these preferences, her judgment or judgments being of the kind just mentioned. (2004, 60; emphasis mine)
Conceptually, at least, values are ranked fairly easily, as can be seen in the following example. Assume that we are comparing the values of the inhabitants of two cities—say, Paris and Houston. Let us further assume that Parisians and Houstonians value “good food,” however they define good and food locally. And let us suppose that both of them value being in good shape. Now, for the sake of discussion of this simplified value system, we propose the following rankings:
Parisians: GOOD SHAPE >> GOOD FOOD
Houstonians: GOOD FOOD >> GOOD SHAPE
It seems fair to say that the different rankings of just these two values could produce different body shapes (add to that a finer analysis of what each group considers to be “good food”—fried chicken and mashed potatoes vs. coq au vin, etc.—and the differences grow), in spite of the fact that it is true to say that the two cities have the same values. In this case, it is not the values but their relative ranking that makes the difference. Thus it is essential not only to have some idea of what a group’s values are, but—as the Rimrock study presciently observed—also the prioritization of the values (and the rankings themselves are also dynamic, changing according to situation, subgroup, etc.). And since I do not accept the idea of a “universal value set,” my claim is that we must first discover values and their rankings in each group separately.
How do we come up with reasonable hypotheses of a people’s values and their rankings? Questionnaires and interviews are not enough, though they can provide some useful data. We crucially require participant-observer records accompanying the former. We need to study a variety of texts, conversations, and interactions. How are values portrayed, described, lived out, discussed, reacted to? Actions and words often are in conflict. People can imagine what their values are, while their actions may reveal different values, different rankings, or finer nuancing.
Consider interviewing two populations. Both claim that to be vegetarian. The first group never eats meat. The second group eats meat in the home of friends who do not know they are vegetarians. Thus although they share the value of “avoid meat, eat vegetables,” the first group seems more consistent. Is the second group hypocritical? In a theory of inviolable values, yes. But in the real world of violable values, no. Here is how we might characterize the difference:
Group 1: AVOID MEAT >> PLEASE FRIENDS IN THEIR HOME
Group 2: PLEASE FRIENDS IN THEIR HOME >> AVOID MEAT
Notice that the question of cultural relativity does arise in this context, but not in a naive way. Societies must have values, along with structuring of roles and knowledge. And these values and structures can differ in ranking, identity, complexity, number, and so on. But there will always be some degree of overlap, due to the ecology of adaptation—we all have to get by, make a living, and follow Shannon’s (1949) basic model of communication (see also D. Everett 2012a; Sterelny 2014). We never reach what Davidson (1973) referred to as incommensurable “conceptual schemes.” This overlap between humans need not be ascribed to innate knowledge, but to external organizational pressures, as well as innate emotional structures, our physiology, and so on.
Nevertheless, just as there are many aspects of theoretical linguistics that are vital still for understanding languages—especially of the Platonic formal variety (Katz 1972), construction grammars (Goldberg 1995, 2006; Croft 2001), and role and reference grammar (Van Valin and LaPolla 1997)—the problems with Kluckhohn’s program do not reduce the utility of values in understanding culture.
Let’s consider again the pioneering work of Milton Rokeach. Rokeach defines the notions of value and value-system as follows:
A value is an enduring belief that a specific mode of conduct or end-state of existence is personally or socially preferable to an opposite or converse mode of conduct or end-state of existence . . . “A happy life is preferable to a sad life,” for example would fit this whereas “A happy life is preferable to fried eggs,” would not.” And also, “A value system is an enduring organization of beliefs concerning preferable modes of conduct or end-states of existence, along with a continuum of relative importance. (1973, 5ff)
Using this conceptualization of values, I suggest a methodology for the study of values.
A METHODOLOGY FOR CULTURE STUDY
There is a minimum threshold of research to incorporate an understanding of values into a particular ethnography, which includes several steps:
1. Through observation of behavior (we must begin with what people actually do and what they avoid), identify potential etic values. How are resources (money, time, fuel, food, etc.) used? How are activities and ends related? And so on.
2. Identify professed values: What do people say that they value, in Rokeach’s sense of values?
3. Posit emic, tacit values using the standard distributionalist methodology of linguistics (Pike 1967, etc.). For example, if stray dogs are treated with violence while domesticated dogs are treated lovingly, then we see different values of “dog” and distinct manifestations of “lovingly” (presumably it is different toward family members and friends.)
4. Test emic values: Conduct experiments to see how people behave relative to the predictions of the proposed value system.
5. Propose potential value systems via experiments, interviews, and observations. (How are values ranked relative to one another in systems such as food, religion, philosophy, relationships, employment, language, etc.? How do values shape or affect different roles of an individual in society?)
6. Test value systems: For example, how do values interlock and vary within and across communities such as food, government, religion, and so on? Are potentially useful predictions made? Do they hold up? This could have been done, for example, in the Rimrock study; and if it had been, they might have seen contrasts and similarities across systems.
7. Test value rankings.
8. Look for individual variation in value inventories, value systems, and value rankings.
9. Write up findings across value systems.
Here the write-up would follow the basic principles of the definition of culture suggested earlier. Emic values become visible and differentiated from etic values when we ask questions such as, “What values do native speakers pay attention to?” “How are different values distributed among the population?” “How do different etic values get grouped together by native speakers as the same emic values according to circumstances?”
Once we have undertaken such initial studies, we can attempt to craft an ethnographic description of the target culture. This will be a very similar effort to the description of a new language. Though each language or society may be relatively independent of other languages or societies, there may be causally significant interaction affecting pairs of languages via contact between societies at a macrolevel or by individual bilinguals/biculturals. Both kinds of structures are relatively stable, outlasting the lives of the individual members of the society or the speakers of a language. For example, a language may have dialects distributed in space or distributed in terms of social levels, time, style, idiosyncracies (idiolect). There are subgroupings and interlocking groupings (where members belong simultaneously to different groups, yet share values and rankings to some extent), roughly identifiable along the lines of linguistic isoglosses.6
It is with the work of the linguist Kenneth Pike (1967) that notions of culture begin to capture the dynamics and structure missing from other definitions, because Pike’s research is informed by his research in language and linguistics. Using his understanding of language, Pike argued that a society is “a structured group of individuals sharing in . . . behavior [whatever social behavior one chooses to examine].”
This notion of structure is fur
ther clarified by Pike and Pike (1976) as a “grammar of society” and, in my terms, a “grammar of culture.” In this sense, a culture is partially a C(ulture)-grammar. Like any grammar, a C-grammar can be revealed and tested only by a solid methodology and rigorous testing of hypotheses. A C-grammar is like a linguistic grammar—a link between the forms instantiating it and the meanings derived from the resultant forms and their distributions.
D’Andrade (1995) correctly observes that earlier treatments of cultures as grammars were too simplistic to be of much use. But the claim I am making here is that culture is only partially constructed grammatically, or alternatively, that a culture contains many grammars produced and interpreted by the culture’s values. Whether the entire culture can be said to be a grammar is not the point of the discussion here. But it nonetheless seems clear that every member of a culture has a variety of ever-changing roles that are sanctioned by, produced by, valued by, and understood by the culture in question.
Society and culture are of course more than merely grammars—but they are connected and constructed in grammar-like ways and especially in their local contexts, groupings, and actions. A Bostonian investment banker and an Amazonian hunter find their place (or if preferred, “make their place”) by occupying a series of grammatical nodes in society. These nodes are rarely invented. One cannot be a professional musician without an entire technology, social role, and payment structure produced by society over time. And the structures and roles and fillers of the sociocultural grammatical system into which we are born themselves emerge from the ranked values and supporting beliefs of the culture. In this sense, if we take culture as beliefs, knowledge, and values, and society as the roles and structural relationships between them, with members of society as “slot fillers,” then we begin to develop an idea of culture as meaning projecting society as structure/grammar, from which individual “meaning” and purpose emerge in conjunction with the small vocabulary of individual psychologies.