Human interaction, as opposed to aggregation, occurs in face-to-face groups. "Sociality theory" proposes that such groups have a nested, hierarchical structure, consisting of a few basic variations, or "core configurations." These function in the coordination of human behavior, and are repeatedly assembled, generation to generation, in human ontogeny, and in daily life. If face-to-face groups are "the mind's natural environment," then we should expect human mental systems to correlate with core configurations. Features of groups that recur across generations could provide a descriptive paradigm for testable and non-intuitive evolutionary hypotheses about social and cognitive processes. This target article sketches three major topics in sociality theory, roughly corresponding to the interests of biologists, psychologists, and social scientists. These are (1) a multiple levels-of-selection view of Darwinism, part group selectionism, part developmental systems theory; (2) structural and psychological features of repeatedly assembled, concretely situated face-to-face coordination; and (3) superordinate, "unsituated" coordination at the level of large-scale societies. Sociality theory predicts a tension, perhaps unresolvable, between the social construction of knowledge, which facilitates coordination within groups, and the negotiation of the habitat, which requires some correspondence with contingencies in specific situations. This tension is relevant to ongoing debates about scientific realism, constructivism, and relativism in the philosophy and sociology of knowledge.
1. Most behavioral and social sciences assume human sociality is a by-product of individualism. Briefly put, individuals are fundamentally self-interested; "social" refers to the exchange of costs and benefits in the pursuit of outcomes of purely personal value, and "society" is the aggregate of individuals in pursuit of their respective self-interests. To this view of "economic man," which long pre-dated Darwin, sociobiology added the idea that individual advantage could be measured in the currency of genes. In theories stressing the importance of group living, conspecifics are viewed largely as a class of objects, more unpredictable than others, but requiring substantial intelligence on the part of the actor to use these "social objects" to achieve genetic ends through alliances, manipulation, or exploitation (Byrne & Whiten, 1988). In contrast, by "social," I refer to a structural continuum of obligate interdependence, without which individual prospects for reproduction and survival to reproductive age are reduced.
2. Humans are obligately interdependent, not only for acquiring their daily bread, but also for the daily operation of their minds. When Adam Smith proposed the "invisible hand" of self-interest, he took for granted that the butcher, the brewer, the baker -- and their families who worked in the business -- were organized in workshops, which were in guilds, which were in villages, which were in districts, which had seasonal fairs and religious celebrations. In the 18th century, the skills for butchering, brewing, and baking were accumulated through generations, passed from adult to child, and repeated in daily, weekly and seasonal cycles of activity. Butchering, brewing and baking demanded finely tuned sensory and motor coordination; familiarity with variable materials, tools and methods; a marketplace, of course; and coordination among these physical, mental and social components. Today, telecommunications and transportation technologies expose the significance of this coordination. People still organize themselves in groups, but some of these no longer need to be face-to-face groups, constrained by space and time in a nested hierarchy of guild, village, district, etc.
3. Hull (1988) described a nested hierarchical organization in science similar to the one in village life; a "demic structure" composed of small research groups, "conceptual demes," and seasonal society meetings. This description accords very well with psychological research on the nested, hierarchical structure of social identity (Turner, 1987). I propose that the demic structure of science described by Hull is more general, and it is paralleled by a "demic structure" of mind. The parallelism suggests that the "mind's natural environment" can be more adequately specified in terms of functional organizational structure than it can be by invoking inclusive fitness theory and "life in the Pleistocene" (Tooby & Cosmides, 1992). Given a descriptive paradigm of the human evolutionary environment, it should be possible to derive testable, nonobvious hypotheses about human mental systems.
4. Before we can consider the "invisible hand" of demic structure, however, we need an evolutionary framework that relieves us of the individualistic assumptions built into the gene-centered view. The traditional single-level, gene-centered evolutionary analysis, based on genetic self-sacrifice, inclusive fitness, or number of offspring, does not lend itself to discussions of hierarchical organization for social structure or mental systems.
5. If evolution is conceived merely as changes in population gene frequencies, then the only factors of evolutionary interest will be those passed on in the zygotic DNA. Accordingly, environmental factors are, at best, secondary to evolutionary change, and development is a side issue (Oyama, 1992). For the study of most non-human organisms, this view has been useful, at least to the extent that a high correspondence between genotype and phenotype could be assumed. In the human case, that assumption has been much more troublesome. For genes to be of any interest in the study of mind and behavior, they must be conceived as much more than a step in protein production. They are "for" a useful attribute, as in "genes for" kin altruism or "genes for" rape, mathematical ability, mate preferences, or cognitive mechanisms specialized for producing the behavior. Although the environment is secondary, it may fill in gaps in "open programs," "shape" innate proclivities, or "shunt" behavior from one option to another. This is a vague gene-environment interactionism that privileges the gene and invites the useless imagery of the habitat as the environment for DNA. There are several problems with this version of evolution, not least of which is the difficulty of demonstrating genetic specificity for particular mechanisms, or reproductive variation corresponding to particular genotypic variation. If the appeal to genes is interpreted metaphorically, it presupposes aspects of behavior that are matters of empirical investigation: the universality of behavior, the difficulty of changing behavior, and the fixity of its form (Oyama, 1985).
6. The restructuring I propose draws on three strands of thought: hierarchical views of evolutionary phenomena, which locate genes in their appropriate environment -- the cell; developmental systems theory, which emphasizes the contextual and contingent events in the flux living phenomena, and teleofunctionalism or "proper functions," which offers a historical definition of function and a taxonomy for the classification of function. The result is a notion of "repeated assembly," where nature-nurture and biology culture distinctions are eliminated. They are replaced with questions about the reliability and durability of replication rather than conflicts about "how much" can be attributed to genetic or environmental determinism.
7. A hierarchical, or expanded, evolutionary theory (Brandon, 1990; Buss, 1987; Campbell, 1974; Wilson and Sober, in press) recasts Darwinism in a form that can accommodate the human case without resorting to nature-nurture dualism. The nested, hierarchical organization of life allows for selection on multiple levels of organization -- macromolecules, genes, cell lineages, individuals, and on occasion, groups. Each of these entities is situated in its specific selective environment. Entities at one level can function as "environment" or context for entities at another level. Thus, the "gene-environment" interaction occurs between DNA and the local cellular machinery. Genetic variants are sorted in differential births and deaths, which are a function of organism-environment interactions; they are selected in the context of cellular machinery. The persistence, or continued replication, of lower level units is crucially dependent on the maintenance of the organized unit interfacing with the habitat. Hence, evolutionary analysis, distinct from other kinds of biological analysis, focuses on the interfaces between levels -- the coordination or lack thereof among levels of organization.
8. The hierarchical perspective re-orients us in two important respects. First, the focus is on functional relations between entity and environment, rather than just traits of the organism. Distinguishing between traits and functions is significant for adaptive explanation generally (Griffiths, 1992). In the human case specifically, the distinction allows that a functional relationship may be achieved through others or even artificially. For example, humans can have disabilities that in other species would be terminal, but humans may use "artificial traits" (e.g., wheelchairs, canes, eyeglasses) to achieve basic functional relationships such as locomotion or vision. The results of natural selection are not "traits" in the usual sense of static features possessed by an organism, but relational linkages between organism and environment. Second, these relations are concretely situated in specific contexts, available to empirical investigation. For example, there is no basis for assuming, as the trait view does, that an individual who may act socially dominant in a dyad will also be socially dominant in a group of five people working on a shared problem.
9. Social scientists are uncomfortable with the determinism suggested by genecentric evolutionary thought. For example, Crawford (1989), a committed evolutionist, reassured his colleagues in psychology that they need not believe that natural selection is currently occurring to use an evolutionary approach; they need only assume that evolution occurred in the past. Evolution, this view suggests, has stopped. Ideally, we would like to restate human evolutionary theory in a form that satisfies the general constraints of Darwinism (Endler, 1986; Brandon, 1990), facilitates "matches" between evolutionary theory and human psychological and social science data, and does not lend itself to mindlessly legitimating existing social and political inequities (Caporael & Brewer, 1991).
10. Contra Crawford (1989), organisms are the continuation of evolutionary processes rather than the result of them. This view allows relations between entity and contexts, over the lifespan and over generations, to function as both cause and effect over time (cf. Brandon, 1990; Gray, 1992; Levins & Lewontin, 1985; Oyama, 1985). We are no longer talking in the billiard ball universe of linear causes where we can imagine that genes cause or predispose traits, which are then modified by the environment. The analogy is closer to a shared language (relation) where (entity) speakers' utterances are both caused by and cause other (environment) speakers' utterances. Just as speakers are more interested in the "fit" between their utterances (rather than, say, the specific language they are using), so is our interest in the "fit" between entities and environments that feature in evolutionary processes (i.e., some relations persist, others don't).
11. "Repeated assemblies" are recurrent entity-environment relations composed of hierarchically organized, heterogeneous components having differing frequencies and scales of replication. This mouthful definition needs to be unpacked in pieces. To start, a mammalian zygote, for example, is the repeated assembly of two sources of DNA, centrosomes from the sperm, and other components, all of which, if in the right place, at the right time, automatically result in a zygote. There is no "genetic program" that directs or controls the assembly of the zygote. Instead, the analogy is to a chemical reaction, where, given a set of constituents, against a background of enabling conditions, an explosion, a precipitate or a vapor results -- or repeatedly assembles.
12. At the level of the organism, genes will always be among the multiple, heterogeneous resources of a repeated assembly (the organism), but they are inert without the epigenetic components at their own and other levels. Genetic action is dependent on the reliable recurrence of appropriate contexts, from cellular machinery to social events to constancies of atmosphere (Griffiths & Gray, 1994). Neither genes nor environment are privileged (although the microbiologist would still focus on genes and the educational psychologist might focus on classrooms). Some resources will themselves be repeated assemblies: ideas, customs, artifacts, learned skills, languages, and group configurations can all be repeatedly assembled.
13. Some repeated assemblies have, in Millikan's (1984) sense, "proper functions." That is, a relation repeatedly assembles ("is selected") because in the past (see below) the co-occurrence of organismic and environmental components in some particular relation (e.g., timing, frequency, contingency) contributed to the persistence of the assembly contingent on particulars of setting. Repeated assemblies are all but infinitely variable because they are concretely situated. No two digestive tracts are identical; sometimes there is a difference that makes a difference (e.g., lactose intolerance). No two breakfasts are ever exactly alike, but breakfast is repeatedly assembled; not by everyone, not every where in the world, but often enough to be extracted and named as a category for human minds. Change in any component, genetic or epigenetic, can alter an assembly, increasing or decreasing the likelihood of its future assembly, or having no particular effect at all. "In the past" may refer to time measured in terms of ontogeny (in "habits of a lifetime"), to cultural-historical time ("2,000 years of Western civilization"), or to geological- evolutionary time. Thus, assemblies may persist in evolutionary time, among Americans, in the Smith family, or in J. Smith.
14. The proper function of a repeated assembly is not necessarily the normative outcome. Most sperm never fulfill their proper function; most acorns become humus. Some repeated assemblies have no function; they result from illusory contingencies, such as the superstitious circling of pigeons in a Skinner box. Other repeated assemblies may have had a function at one time, but become dissociated from it. Langer (1989) tells the story of a friend who would always cut a slice off the end of a roast before placing it in the oven. To make a long story short, the colleague said she did this because her mother did it; mother said she did it because her mother did it; grandmother shed light on the practice: the pan she used for roasting was too small to accommodate the whole piece of meat. Such dissociations may occur in evolutionary time, giving rise to vestiges, and in historical-cultural time, giving rise to customs that have no function or even incur costs (Boyd & Richerson, 1985).
15. Sources of heritability (i.e., acquired or innate traits) have no a priori significance in descriptions of repeated assemblies. Paradoxically, however, the notion of heritability is expanded. As humans, we inherit not only genes, but also attitudes, practices, place, nationality, expectations for behavior, and so forth. There may be change within lifetimes, but this simply indicates low reliability of repetition. Heritability itself is thus no indicator of the universality, fixity, or difficulty of changing organism-environment relations.
16. The difference between the gene-centered view and the expanded heritability of repeated assembly can be illustrated. Tooby and Cosmides (1992) argue, as do developmental system theorists (e.g., Levins & Lewontin, 1985; Oyama, 1985), that phenotypic features are fully codetermined by genes and environment in development throughout the lifespan; that the phenotype cannot be analyzed into separate genetically determined and environmentally determined components; that events at one stage of development are contingent on the products of prior developmental outcomes, and that both genes and developmentally relevant environments are outcomes of evolutionary processes. Yet, Tooby and Cosmides (1992) also assert that developmental programs in the genes create the relation between environmental conditions and developmental outcomes. Thus, a child raised in social isolation may not acquire language, but nevertheless will have a species-typical language acquisition device.
17. If we view language as a repeated assembly, however, we could allow that such a child might have genes for a language acquisition device, but lacking a critical component, a language environment, that device was never assembled. Both the genes and the language environment must be inherited. English speaking and Kikuyu speaking differ because the various elements have different cycles of repetition. One set, which includes genes, has a longer cycle of repetition relative to another set, which includes the language environment. Both sets are repeatedly assembled, but on different scales of time. The cycle of language environment (English or Kikuyu), in cultural-historical time, is nested within the cycle of other components, including genes, in evolutionary time. Although the frequency and scale of the cycles may be roughly distinguished, it makes little sense to separate language into an innate and an acquired component; both are parts of inherited resources. Nested within these two cycles of evolutionary time and historical-cultural time, is a third one, the sine qua non of variation and evolutionary process: the situated relation of "organism-in- setting," (e.g., a child raised in social isolation) where the conditions for an assembly to be repeated, be it later in the day or in succeeding generations, take place or fail to take place.
18. In the next section, we consider the relations humans repeatedly assemble in evolutionary time. These form the contexts for repeated assemblies of shorter cycle lengths in historical and lifespan time and the basis for superordinate coordination.
19. Human evolutionary theorists agree that group living was a critical feature in human evolution. They disagree about the causes and consequences. In my view, groups functioned and continue to function as an interface between individual and habitat. Although this is a substantial assumption, it does allow us to unburden ourselves from a host of lesser assumptions about the past and to construct a "minimalist scenario." We assume there would have been both minimum and maximum constraints on group size: too small a group would have a higher risk of perishing; too large a group strains the carrying capacity of the environment. The "envelope" for selection for sociality is thus a function of the physical parameters of the species morphology and ecology. The combination of minimum and maximum group size constraints has several important theoretical consequences.
20. First, to the extent that exploiting a habitat is more successful as a collective group process than as an individual process, not only would more successful groups persist, but so also would individuals better adapted to group-living. Because a group mediates individual contact with the environment, and the number of "niches" within groups is limited, fitness should have been correlated with the evolution of perceptual, affective and cognitive processes that support the development and maintenance of group membership (Caporael, Dawes, Orbell & van de Kragt, 1989; Brewer & Caporael, 1990). One consequence of these dynamics is a ratcheting, feed-forward evolutionary process where other constraints, including cognitive ones, constrain group size (cf. Dunbar, 1993; Wilson & Sober, 1994). We would expect humans to be obligately interdependent; that is, their prospects for survival and reproduction outside group contexts would be greatly diminished.
21. Second, "self-interest" at the individual level would be necessarily, albeit incompletely, constrained by requirements for group coordination and group fissioning. A corollary is that groups would reproduce by fissioning, not by dispersing individuals or dyads. Third, humans have not evolved to apprehend "true beliefs" in the sense implied by prescriptive rationality, realism, or objective knowledge, but rather to develop, maintain and negotiate face-to-face group membership. (Note that I am not saying humans cannot do these things, but rather that they have not evolved to do them.) Fourth, mental systems specialized for face-to-face interdependency in evolutionary time must be "reweavable" for the production of large-scale social coordination, which has appeared in historical-cultural time; that is, humans have no evolutionarily specialized adaptations for agricultural or urban living.
22. Human face-to-face interaction, repeatedly assembled, generation-to-generation, over evolutionary time, is a continuation of the nested hierarchy of life's functional organization. I propose four "core configurations" of situated, face-to-face activity: dyad, family/workgroup or team, face-to face group or deme, and macroband or macrodeme. These labels are not intended to represent social roles, or even necessarily individuals: they represent kinds of interactions. A dyad is an interaction between two entities, one of which can be non- human. "Family/workgroup" does not specifically refer to families or workgroups, but to "small-group, common task orientation" interactions. Of course, such interactions are frequent in both families and workgroups, but, as we shall see, are not limited to such social groups.
23. Configurations are "core" because they are repeatedly assembled in hunter-gatherer groups (Birdsell, 1972; Jarvenpa and Brumbach, 1988) -- presumably in evolutionary time, but also in human ontogenetic sequences, and in day-to-day activity. Each configuration is associated with a group size, which should be considered more like a center of gravity than a fixed condition, and "modal tasks" for interacting with the environment. The general idea is that size/task configurations are stable and repeated generation-to-generation in evolutionary history as a functional consequence of the physical interaction between species' morphology and ecology. From a strictly evolutionary historical perspective, core configurations are affordances for the evolution of proper functions. That is, a size/task configuration is a specialized affordance for the repeated assembly and evolution of various proper functions. An affordance allows a proper function to occur, but there is no necessity that the proper function will assemble given the affordance.
24. Table 1 sketches a model of this structure. (I have used the terms "deme" and "macrodeme" to indicate a greater generality than the use of "band" or "macroband" by anthropologists.) The combination of size and modal task (not just size alone) constitute a specialized affordance for the evolution and repeated assembly of proper functions.
TABLE 1. CORE CONFIGURATIONS
Core Group Modal Proper Configuration Size* Task Function -------------------------------------------------------------- Dyad 2 Sex, infant interaction with Microcoordination adults & older children
Work/Family 5 Foraging, hunting, Distributed Group direct interaction cognition with habitat
Deme (Band) 30 Movement from Shared construction place to place, of reality general processing (includes folk and maintenance, psychology), work group social identity coordination
Macrodeme 300 Seasonal gathering, Stabilizing & (Macroband) exchange of standardizing individuals, language resources and information
* Except for dyads, these numbers should be considered as "centers of gravity," modal estimates in a range roughly plus or minus a third of the number.
25. There may be multiple proper functions, but for clarity and simplicity, one per configuration will do for this discussion. The dyad affords possibilities for micro-coordination (e.g., facial imitation in a mother-infant dyad, the automatic adjustment of gait that occurs when two people walk together); the workgroup affords possibilities for distributed cognition; the face-to-face group affords a shared, construction of reality ("common knowledge"), which also mediates interaction in macrodemes and other superordinate group configurations. Macrodemes afford the stabilization and standardization of language, signs and symbols. Higher-level configurations cannot be reduced to lower-level configurations, but once a proper function is available from higher-level configurations, it may be usefully employed in lower-level configurations. For example, there is no reason to propose that language evolved "for" dyadic interaction. However, once language has evolved, it can be used in dyadic interactions.
26. Human core configurations are also repeatedly assembled in an ontogenetic sequence. Infants develop microcoordination in dyads; as their coordination increases, they participate in workgroups (families), and through them, face-to-face groups (extended networks of kin, family friends, etc.). Rather than just independence, human development is increasing interdependence. There is a broadening of the range of social interaction accompanied by increasing requirements for reciprocity, skills, memory, social judgment and so on.
27. Psychological research on humans also suggests a nested hierarchical organization for social identity. Social identity plays an important role in ingroup-outgroup relations, the distribution of resources, self-categorization, and expectations for behavior. It is an automatic redefinition of "self" in terms of shared group membership (Brewer, 1991; Turner, 1987). To go back to Adam Smith's period, the brewer works in a small group for brewing; he is also a member of a guild, a village, and so on. Each of these may be distinct social identities, posing different conflicts and synergisms. Social identity is more complicated for many people in the modern world because they potentially belong to multiple "lineages" of groups: the modern brewer may also be a member of an environmental activist organization, which may be a chapter in a larger organization, which may be part of a federation of environmentalist groups. Is there a way of bringing together, from the one side, a vision of evolution as hierarchical, repeatedly assembled levels of organization, and from the other, a line of empirical research on the hierarchical organization of human social identity?
28. I propose that the topography for the evolution of human relations and practice is simultaneously social-organizational and cognitive. Dynamic psychological shifts in level of identity (which can result from various conditions including group size, shared fate or outcomes, and salient group boundaries; Turner, 1987) are hypothesized to maintain core configurations even when the physical constraints of time and space are reduced, as for groups of globally dispersed individuals linked by telecommunications technology.
29. Another way of saying this is that configuration units are group-level entities having proper functions. Their psychological correlates, that is, how they are cognitively maintained and repeatedly assembled, are redefinitions of "self" that minimize distinctions between self and others (Brewer, 1991). Thus, from one perspective, the "skin-bounded" organism consists of multiple selves; from another perspective, the skin bounded organism is "dissolved" into multiple individuals, that is, units of coordinated activity. Personal identity, the locus of conscious awareness of goals, plans and beliefs, is the product of multiple interacting and dynamic selves. These are two perspectives of the same phenomena (similar to the way that light can be described as waves or packets depending on one's perspective).
30. This unfamiliar way of thinking about social cognitive phenomena can be intuitively illustrated: five people in an elevator are an aggregation, not a group; that is, they are not a core configuration. The same five people stuck in the elevator between floors are a group (core configuration), the "glue" being shared fate. When group identity is salient, personal identity is not. Although we can access the experience of self at different levels of organization, it is, in effect, partially encapsulated.
31. Some reviewers complain that by introducing a dual perspective, I have not completely dashed dualism. That was not my objective: some minimal dualism is essential because without it, we have no way of conceptualizing (or perceiving) contrasts. I have relocated the panoply of dualisms that we tacitly assume exist in nature (nature-nurture, individual-society, etc.) and put them in the researcher's project, where they properly belong; have been all the time; and -- when placed into nature -- have caused no end of confusion and dispute among realists and relativists.
32. Once a proper function, associated with a core configuration, evolves, it can be extended to other domains of interaction; that is, it can become part of another behavioral assembly. For example, dyadic micro-coordination (involving perception of movement dynamics, emotional expressiveness, etc., whose proper function probably includes sex and infant-elder interactions) can be extended to interaction between animal trainer and animal, and human and artifact (where it becomes the basis for embodied knowledge, tinkering or highly skilled performance). Language is one of the most salient examples of extension, having both upward and downward causal effects, although these do not seem to be equivalent. Language has considerable impact on band-size humans, and largely allows modern humans to be members of multiple macrodemes presumably an innovation in recent history. Nevertheless, the proper function of macrodemes persists. Various fields of work, from dog grooming to biochemistry, have specialized languages, and one purpose of professional meetings is to standardize terms and references that have no ordinary usage. Though language is still important for workgroups, these smaller groups rely more than higher level groups on gesture, visual aids, and "private language" for telegraphing information. The rhythmic features of language, more than content, are probably significant in microcoordination.
33. With an appropriate bridge, proper functions specialized for one level of organization may be evoked at others. The rhythm of a marching band, for example can generate microcoordination among hundreds of marchers. We would not expect the evolution of a capacity, "group-marching ability" to evolve for marching in large groups, as there is no reason to suppose that either large groups or marching bands (or an equivalent) were repeatedly assembled until recently in human history. Technology, crucial for achieving agriculture and settlement life, also bridges domains of application and the psychological correlates of coordination, particularly where technology eliminates the constraints of time and space that have traditionally characterized human social organization (Caporael, 1987). We "reach out and touch someone" thousands of miles away by using a telephone. Radio and television can be used to extend group identity to millions of people by invoking national category boundaries rather than ethnic, religious, or professional ones.
34. Specialized affordances may be combined (e.g., a heart surgery team combines affordances for microcoordination and distributed cognition). There may be other recognizable configurations for human activity in addition to core ones, but all human action, and all selective action at the organism environment level, is situated in group contexts plus the solitary or individuated state (e.g., being alone in a crowd). For analytic purposes, affordances can be decomposed. Piloting an airplane involves highly evolved microcoordinative abilities, and, counterintuitively, would not be evolutionarily novel. Getting into an airplane -- sharing a small contained space with hundreds of strangers -- is evolutionarily novel.
35. David Hull (1988) provides an excellent description of scientific organization that can be used to illustrate how repeated assembly, core configurations, and their psychological correlates could be used. He proposed that scientists competed for citations, priority, and recognition, but they had to cooperate to achieve their ends. He also saw that the social organization of science, its demic structure, was very important. My interest here is to use his excellent observations to illustrate an analogy between the demic structure of science and that of traditional societies.
36. Hull (1988) identified a scientific demic structure consisting of the individual scientist, research groups, "conceptual demes," and seasonal society meetings. These correspond to the typical organization among nomadic hunter gatherer groups: a nested hierarchy composed of work groups, microbands, and seasonal macrobands (Jarvenpa & Brumbach, 1988). To these configurations, we must add the dyad. In both instances -- in science and among hunter-gatherers -- group-size numbers at these four levels are fairly constant; dyads, about 3-5 individuals, 30-50 individuals, and 100-500 individuals (Birdsell, 1972; Hull, 1988). In both cases, isolates have reduced viability, and must be part of a subgroup with sufficient "critical mass" to persist. As Hull noted, isolated scientists are not very productive. Also, scientists do not usually begin a new research group as social isolates; they must bring some of the old group members with them or attract new recruits.
37. Learning to fashion specialized tools, in stone or other materials, or to use specialized laboratory or food preparation equipment and skills is a hands-on, situated activity (like golf or tennis) requiring feedback from materials, equipment, and an experienced user to develop finely-tuned sensory-motor microcoordination. The most intense conceptual interaction with the material environment occurs in workgroups, the activities of which are influenced by the microband (conceptual deme) and feed back into it. Small research groups or intensely social small hunter-gather groups working on the environment seem to do two things. They distribute cognition; that is, they share tasks such as perception, classification, inference, memory, and contextually-cued responses in the conduct of practice and interpretation of data, or in interaction with uncertain habitat features. In the process of doing so, they draw on and contribute to a socially constructed and shared "world-view," which permits groups (hunter-gatherers or scientists) to interface with nature (cf. Jacobs and Campbell, 1961; Amann & Knorr-Cetina, 1990).
38. The product of that interface feeds into higher levels in the group system. Bands and conceptual demes are "staging communities" (Jarvenpa, 1993), which serve as "general processing and maintenance centers" for information and resources retrieved from the smaller dispersed groups; they are also loci of shared group identity. For both scientists and hunter-gatherers, seasonal macroband meetings (or yearly conventions) are important for the exchange of myths, gossip, and information about more distant areas and groups. Macrobands are also arenas for competitive games as well as the affirmation of common worldviews, the maintenance of languages (hunter-gatherers) and idiolects (scientists), and the exchange of people (mates, new PhDs, or disgruntled members). For science, unlike hunter-gatherer groups, the production of articles in journals may find their way into the popular press and eventually be incorporated as "common knowledge" (e.g., almost everyone believes in atoms, germs, and innate sex differences), which remains unsituated (e.g., atoms) or is realized in interaction (e.g., sex differences).
39. The correspondence between these two demic structures, scientists and hunter-gatherers, suggests some counterintuitive hypotheses: divergence in scientific ideas may not lead to fissioning; fissioning or rather too large a group size may lead to divergence in scientific ideas (or, alternatively, bureaucratic suppression of productivity in too large a group). Likewise, the common practice in academic departments to hire faculty for coverage of a discipline may come at the cost of having critical mass for effective research groups.
40. The system of meanings and practice implied by terms like "culture" or "society," must be realized in specific organism environment interaction, which becomes increasingly elaborated during development. Superordinate coordination partitions the "blooming, buzzing confusion" of situated activity into conventionally shared, orderly descriptions. These in turn influence situated activity. Such interactions result from, and contribute to, specialized domains of repeatedly assembled belief, knowledge, and practice, such as those distinguishing between sexes, generational cohorts, technical specialties and cultures.
41. For example, a young child may be puttering around in the kitchen "mixing stuff," as children typically do, but a boy's mixing is more likely to be described by adults as evidence for an incipient interest in "chemistry" and a girl's as incipient interest in "cooking." For both children the situated activity is the same. Yet the adults' description of activity, based on common cultural assumptions, is different. One child is directed to ponder a future career, the other a future meal.
42. Coordination occurs to the extent that knowledge and practice domains overlap or are complementary. I suggest that values serve as a medium. Humans live in a value-saturated environment; values are known from interactions with people, natural objects, and artifacts, all of which interact with social life and customs in complex ways (Caporael, Panichkul & Harris, 1993; Merchant, 1980).
43. The word "values" tends to trouble psychologists. I am not suggesting there exists a set of unalterable "shoulds," moral provisions, or ethics that can be revealed by biology or the neurosciences. By values, I am referring to the evaluative (positive/negative) dimension of aggregate social preferences, usually tacit and invisible, shared by members of a community, diagnostic of shared social identity, and, to some interesting extent, channelling behavior in some directions and away from others.
44. Values could not play a coordinative role unless they were both accessible and automatic. Accessibility allows tacit values to be exposed and used to negotiate new prescriptive values. Automaticity is critical; values could not serve coordinating functions if we always had to contemplate them before acting. We should, therefore, expect a tradeoff or compromise between values (for coordination) and experience (of situated activity), which in turn should result in a discrepancy between actual experience and the description of it.
45. Obviously, humans can describe situated activity, but the circumstances for doing so are rather rare (e.g., a radio sportscaster delivering a "real time" description of a baseball game, a field researcher gathering "real time" observations of animal or human behavior). Instead, people summarize experience or concatenate it, in widely shared, conventional ways. These descriptions are not just a shorthand, expandable to a full description of behavior. People are often unaware of their actual experience, and instead they describe it in terms of widely-shared expectations (Nisbett & Ross, 1980). Descriptions of other people's behavior is often no more precise. Even a very short time delay can produce descriptions of behavior that correspond more highly to measures of conventional linguistic meaning than to actual observed behaviors (Shweder & D'Andrade, 1980). Expectancy effects can result in self-fulfilling prophecies about others' behavior (Eccles, Jacobs & Harold, 1990; Rosenthal & Jacobson, 1968), as well as structure access to physical spaces, knowledge and objects (Caporael, Panichkul & Harris, 1993). Research on eyewitness testimony, self-description, social judgment, causal attributions, as well as other social psychological phenomena, suggest that discrepancies between actual behavior and descriptions of behavior may be quite common. Because cognition is considered to be fundamentally "asocial," these phenomena are typically explained in terms of side-effects of a tradeoff between accuracy and cognitive efficiency rather than between accuracy and sociality.
46. The discrepancy between situated activities and their conventional descriptions allows perception and talk to function in the production of coordinated activity despite the novelty of every living moment. At least two interacting psychological systems appear to be involved in superordinate coordination. One is based on folk psychology as "value-making" talk and the other, following Hodges & Baron (1992), is based on perception (of objects, people, things) as "value-realizing" (Hodges & Baron, 1992). Both systems would operate in face-to-face groups, but easily extend for higher levels of coordination.
47. Psychologists and philosophers have extensively debated the role of folk psychology as an adequate or heuristically useful description of human cognition (Greenwood, 1991). However, the prescriptive content of folk psychology (e.g., the evaluative dimensions of trait terms) suggests another interpretation. Folk psychological talk about beliefs, intentions, and desires -- including judgments concerning which are "natural" and which are not -- develops a context for action that limits and entrains which actions are conceivable, possible, desirable, essential, tolerated and forbidden. Folk psychology can serve for making heuristic predictions of behavior because, to some extent, it produces the behavior it predicts. The coordination involved is not simply directing behavior (as in the sense implied by norms), but in its social cognitive construction.
48. Like folk psychology, perception itself can also tacitly limit the conceivable, possible, desirable and essential. When presented with a stimulus, humans are capable of judging whether or not they like it even if they cannot identify it -- evidence of the primacy of value in perception (Zajonc, 1980). Moreover, such evaluations correlate with prior exposure to a stimulus: familiarity breeds liking. Hodges and Baron (1992) go beyond the notion of values as preference. They argue that values are criterial and embodied in public, objective affordances. This view allows that perception can be coordinative on a large scale. For example, a simple artifact such as a toddler's spouted cup, is weighted on the bottom to discourage spilling and encourage drinking (the dual affordances of a cup filled with liquid). It also connotes a value on independence and self-sufficiency consistent with a culture with early weaning and training for independence (Hodges & Baron, 1992). The device, which is only one of many possible designs for transporting liquids to the mouth, contributes to the direction of parents' child-rearing activity, influences the child's development, and reverberates through the culture, constraining other activities. Most Americans would disapprove of a woman breast-feeding a two-year old toddler, even in private, although it is common in nonindustrial cultures and was throughout human evolutionary history. In perception, "the social" actively interacts with "the material" through the translation of value in situated interaction (cf. Pacey, 1983).
49. The descriptive paradigm presented in this article is preliminary; it is a target to shoot down, a thesis to test. Flaws that are invisible to me shout for recognition from critics, and much more empirical and conceptual work remains. My claim is that, in a fundamental sense, the natural environment of humans has not dramatically changed. It is maintained by social/cognitive mechanisms and a structure of activity through which humans "parse" the stimulus information of complex environments into a small-group "grammar" throughout the lifespan (Caporael, 1987). The persistence of nested demic structure, in industrial as well as hunter-gatherer societies, the existence of novel, cross-cutting "bureaucratic" organization (typically indifferent to issues of group size and task structure), and now, the existence of groups unsituated in space and time, linked through email, fax, phone, and "groupware" provide unprecedented opportunities for both pure and applied investigations by evolutionarily-oriented researchers.
50. Although I have stressed sociality, almost as a synonym for coordination, I am not claiming that human cognition is always a group-level process: individual cognition is an important source of variability. Sociality theory most definitely is not a claim that humans are basically altruistic, prosocial or selfless. I have not denied that humans can be self-interested, competitive, altruistic, individualistic or intentional, as the case may be. However, these terms may be less useful as analytic categories for science than they are for the critical function of negotiating, coordinating, and constructing the human social environment (Caporael, 1994). If there is a distinction, as I have argued, between the "real-time" description of situated activity and the conveyance of expectation (value) through unsituated descriptions, there may be a genuine failure to discriminate between scientific activity as a description of human nature from scientific activity as the construction of human nature. If this is the case, then there may be no transcendental "human nature," but only historical sequences of descriptive paradigms, of which self-interest and sociality are two, comparable in their generativity more than in their truth value.
51. Whether repeated assembly is of any value to biologists depends to a large extent on how useful and appropriable it is for the study of other species and the development of evolutionary theory. To the extent that genotype and phenotype can be collapsed into a single unit of analysis for some research purposes, repeated assembly will likely be an unnecessarily complicated description of evolution. Still, this rephrasing seems to pose a challenge -- to recent notions, at least. One is the interactor/replicator distinction (Hull, 1988), which divides entities (or their attributes) into those things that interact and those that replicate. Repeated assembly, coordinated over levels of organization, implies that repeated interaction is replication, or at least the source of replication. Similarly, there has been a trend to divide evolution into two domains of inquiry: the process of evolution and the results of evolution (Reeve & Sherman, 1993). If organisms (and other entities) are the continuation of evolutionary processes, than process and product are indistinguishable; the durability of repeated assemblies, or repeated assemblies nested in other repeated assemblies, replaces the process-product distinction.
1. The term "deme" comes from the Greek word, demos, referring to district, and it is used in biology to refer to a breeding population. Interestingly, in ancient Athens, there was a law requiring a citizen to use his deme name rather than his family name. The explicit reason for this requirement was to foster ties with fellow citizens over kin.
I am indebted to Reuben Baron, Marilynn B. Brewer, P. Thomas Carroll, Glen Culbertson, Elihu Gerson, Maryanne Garry, Paul Griffiths, Cecilia M. Heyes, Susan Oyama, Henry Plotkin, Michael G. Shaftoe, and David Sloan Wilson. This work was made possible by by the National Science Foundation, Grant No. SBR-9321461. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.
Amann, K. & Knorr Cetina, K.(1990) The fixation of visual evidence. In Lynch, M. (Ed.), Representation in scientific practice, 85-121, MIT Press.
Birdsell, J.B. (1972) An introduction to the new physical anthropology. Rand McNally.
Boyd, R. & Richerson, P.J. (1985) Culture and the evolutionary process. University of Chicago Press.
Brandon, R.N. (1990) Adaptation and environment. Princeton University Press.
Brewer, M. (1991) The social self: on being the same and different at the same time. Personality and Social Psychology Bulletin 17: 475-482.
Brewer, M.B. & Caporael, L.R. (1990) Selfish genes versus selfish people: Sociobiology as origin myth. Motivation and Emotion 14: 237-242.
Buss, L.W. (1987) The evolution of individuality. Princeton University Press.
Byrne, R.W. & Whiten, A. (1988) Machiavellian intelligence. Clarendon Press.
Campbell, D.T. (1974) "Downward causation" in hierarchically organized biological systems. In Ayala, F.D.,& Dobzhansky, T. (Eds.), Studies in the philosophy of biology, 179-186, Macmillan.
Caporael, L.R. (1987) Homo sapiens, Homo faber, Homo socians: Technology and the social animal. In Callebaut, W. & Pinxten, R. (Eds.), Evolutionary epistemology: A multiparadigm program, 233 244, Reidel.
Caporael, L.R. (1994) Of myth and science: Origin stories and evolutionary scenarios, Social Science Information 33: 9-23.
Caporael, L.R. & Brewer, M.B. (1991) The quest for human nature, Journal of Social Issues 47: 1-9.
Caporael, L.R., Dawes, R.M., Orbell, J.M. & van de Kragt, A.J.C. (1989) Selfishness examined: cooperation in the absence of egoistic incentives. Behavioral and Brain Sciences 12: 683-739.
Caporael, L.R., Panichkul, E.G. & Harris, D.R. (1993) Tinkering with gender. In Rothschild, J. (Ed.), Research in Philosophy and Technology. Vol. 13. Technology and feminism, 73-99, JAI Press.
Crawford, C.B. (1989) The theory of evolution: Of what value to psychology? Journal of Comparative Psychology 103: 4-22.
Dunbar, R.I.M. (1993) Coevolution of neocortical size, group size and language in humans. Behavioral and Brain Sciences 16: 681-735.
Eccles, J.S., Jacobs, J.E. & Harold, R.D. (1990) Gender-role stereotypes, expectancy effects, and parents socialization of gender differences. Journal of Social Issues 46: 183-201.
Endler, J. (1986) Natural selection in the wild. Princeton: Princeton University Press.
Gray, R.D. (1992) Death of the gene: Developmental systems strikes back. In Griffiths, P.E. (Ed.), Trees of Life, 165-209, Kluwer Academic.
Griffiths, P.E. (1992) Adaptive explanation and the concept of a vestige. In Griffiths, P.E. (Ed.), Trees of Life, 111-131, Kluwer Academic.
Griffiths, P.E. & Gray, R.D. (1994) Developmental systems and evolutionary explanation. Journal of Philosophy, vol. XCI, 277-304.
Greenwood, J.D., Ed. (1991) The future of folk psychology. New York, Cambridge University Press.
Haraway, D. (1989) Primate visions. Routledge.
Hodges, B.H. & Baron, R.M. (1992) Values as constraints on affordances: Perceiving and acting properly. Journal for the Theory of Social Behaviour 22: 263-294.
Hull, D.L. (1988) Science as a process. University of Chicago Press.
Jacobs, R.C. & Campbell, D.T. (1961) The perpetuation of an arbitrary tradition through several generations of a laboratory microculture. Journal of Abnormal and Social Psychology 62:649 658.
Jarvenpa, R. (1993) Hunter-gatherer sociospatial organization and group size, Behavioral and Brain Sciences 16: 712.
Jarvenpa, R. & Brumbach, H. (1988) Socio-spatial organization and decision-making processes: Observations from the Chipewyan. American Anthropologist 90: 598-618.
Kitcher, P. (1985) Vaulting ambition: sociobiology and the quest for human nature. MIT Press.
Langer, E.J. (1989) Mindfulness. Addison-Wesley.
Levins, R. & Lewontin, R. (1985) The dialectical biologist. Harvard University Press.
Merchant, C. (1980) The death of nature. Harper & Row.
Millikan, R.G. (1984) Language, thought, and other biological categories. MIT Press.
Nisbett, R. & Ross, L. (1980) Human inference: strategies and shortcomings of social judgment. Prentice-Hall.
Oyama, S. (1985) The ontogeny of information. Cambridge University Press.
Oyama, S. (1992) Ontogeny and phylogeny: A case of metarecapitulation? In Griffiths, P.E (Ed.), Trees of life, 211-239, Kluwer Academic.
Pacey, A. (1983) The culture of technology. MIT Press.
Reeve, H.K. & Sherman, P.W. (1993) Adaptations and the goals of evolutionary research. Quarterly Review of Biology 68:1-31.
Rosenthal, R. & Jacobson, L. (1968) Pygmalion in the classroom. Holt, Rinehart and Winston.
Shweder, R.A. & D'Andrade, R.G. (1980) The systematic distortion hypothesis. In (ed), Fallible judgment in behavioral research. New Directions for Methodology of Social and Behavioral Science, No. 4, 37-58, Jossey-Bass.
Tooby, J. & Cosmides, L. (1992) The psychological foundations of culture. In Barkow, J. H., Cosmides, L. & Tooby, J. (Eds.), The adapted mind, 19-136, Oxford.
Turner, J.C. (1987) Rediscovering the social group: a self categorization theory. Basil Blackwell.
Wilson, D.S. & Sober, E. (1994) Reintroducing group selection to the human behavioral sciences. Behavioral and Brain Sciences, 17: 585-654.
Zajonc, R.B. (1980) Feeling and thinking: preferences need no inferences. American Psychologist 35: 151-175.