PART ONE
QUALITY AND VARIATION OF COGNITIVE PROCESSES
Introduction
This first section attempts to stretch cognitive phenomena as widely as possible, to display them in a variety of cultural contexts. It is intended to help convince the sceptic that qualitative aspects of cognitive behaviours do indeed differ, that styles, processes and skills vary with language and ecology, values and functions.
Deeper than this mapping task, this display of variability, however, there is a more detailed goal; this is to attempt to relate (sometimes causally) this variability in cognitive behaviour to specific cultural variables. Thus, language and socialization in particular are frequently examined for possible relationships with logic, classification, cognitive style, belief, measurement, imagery and memory.
Chapters 1 and 2 introduce the concerns of anthropologists. Gladwin (Chapter 1) argues that different cognitive strategies and logical processes may be employed by people living in cultures with different problems to solve. Relying largely upon observations among Micronesian canoe navigators, he concludes that these differences are indeed real, and that psychologists must take them into consideration if they are not to remain āculture-boundā. This author has recently elaborated this point of view (Gladwin, 1970) in a more detailed analysis of Micronesian navigation. This emphasis on studying what people do, and how they go about doing it, is emphasized by Sturtevant (Chapter 2) in his advocacy of ethnoscience, the study of āthe system of knowledge and cognition typical of a given cultureā. Although only a tool or a means for the study of culture for the anthropologists, it may be considered an end or goal for the psychologist, in that it requires the comprehending of the cognitive categories of people living in other cultures. This knowledge of the categorization of experience by individuals permits the anthropologist to move on to study the culture, but since it is cognitive behaviour in itself, it comprises a major focus of study for the psychologist working cross-culturally.
A major tool in the ethnoscience approach is the study of language. How a language carves up the physical and social environment is thought to indicate the cognitive categories in use. Chapters 3 and 4 emphasize the importance of language in understanding cognitive behaviour, by outlining and systematizing the basic assertions of Whorf (Fishman, Chapter 3), and by carrying out experimental work (Maclay, Chapter 4) to evaluate them.
Another major set of cultural and cognitive variables is introduced with Chapter 5 (Witkin), in which a theory relating socialization to cognitive style is illustrated with reports of two cross-cultural studies of this perceptual-cognitive dimension. Specifically, the degree of differentiation (analytical approach to cognitive or perceptual problems) is shown to vary cross-culturally with child-rearing practices which are themselves related to the ecological setting of the cultural system. A qualification to this general finding is offered by Wober (Chapter 6) who argues that a high level of differentiation may be limited to a specific sense modality (e.g. vision in Western cultures or audition in some African cultures) within which cognitive growth may be largely confined. A further extension by Berry (Chapter 7), arguing from a position of ecological functionalism, demonstrates that across four ecocultural settings, high visual differentiation varies systematically with a cluster of ecological and cultural variables, including the importance of hunting, the severity of socialization, the quality of arts and crafts production, and the assistance from a complex vocabulary of āgeometrical-spatialā terms.
The final chapters deal with a variety of cognitive variables as they are exhibited in a number of African cultural settings. Beginning with beliefs and tying-in with the previous three chapters Jahoda (Chapter 8) explores the notion that there can emerge a change in cognitive structure in the direction of coexistence between African and Western ideas and beliefs. Cole, Gay and Glick (Chapter 9) examine in detail the cognitive behaviour involved in making quantitative judgements in Liberia, dealing sensitively with such subjects as geometric concepts, disjunction and conjunction, and volume, time, length and number estimates. This set of studies demonstrates the use of concepts discovered within the culture (ethnoscientific use of emic units) for designing field experiments in cognitive behaviour. More recent studies, employing this same strategy, are reported in Cole et al. (1971).
In cultural systems which function without the benefits of literacy, there have frequently been assumed to exist alternative ways of coding and remembering information. Doob (Chapter 10) explores the possibility that eidetic images (persistent and accurate image traces of the original stimulus) may be more used in nonliterate than in literate cultures. Although in the chapter presented here, Doob found a greater incidence of eidetic images among Ibo subjects than among Westerners, Doob (1966) has more recently discovered no consistently greater use of this memory aid among other samples from nonliterate populations. Another possibility, however, is that in cultures where writing is not available to record events, memory may be relied upon to a greater extent, and hence may be better under experimental conditions. Ross and Millsom (Chapter n) test this alternative in Ghana and New York, following up a long tradition of studies on the topic (Bartlett, 1932).
Throughout this first section then, the emphasis is on variability in cognitive behaviour, and on the cultural factors associated with this variability. Clearly these studies are concerned with processes and styles, and not with basic capacity or intellectual competence. Studies which comment on this latter issue will be sampled in the second section. However, while exploring the cross-cultural variability of cognitive performance, the reader may wish to consider the kinds of inferences about competence which might validly be made from these data.
1
T. Gladwin (1964)
Culture and logical process
W. H. Goodenough (ed.) Explorations in Cultural Anthropology: Essays in Honor of George Peter Murdoch (McGraw-Hill Book Co.), pp. 167ā77
In 1936, Gregory Bateson published Naven (1958), an admittedly partial ethnography of the Iatmul of New Guinea. The importance of this book lay in his attempt to formulate some theoretical constructs with respect to psychological process and social dynamics, constructs which would embrace his observations of the Iatmul, and would at the same time be of general validity. One of these, schizmogenesis, was developed extensively and ultimately led Bateson into cybernetics and the flourishing field of general systems research. The other really new concept in Naven, ados, has never enjoyed the equal development it deserves.
Eidos, in Batesonās terms, is āa standardization (and expression in cultural behaviour) of the cognitive aspects of the personality of individualsā (1958, p. 220). It is complementary to ethos, the āexpression of a culturally standardized system of organization of the instincts and emotions of... individualsā (1958, p. 118). Early in the history of research in culture and personality, Bateson thus made explicit the necessity for giving to the cognitive aspects of personality a weight and attention equal to that devoted to the emotional.
However, earlier exploratory thinking by Edward Sapir and others on the relationship between culture and personality had been rooted in the concepts of psychoanalysis, a system of theory anchored almost exclusively to the biological and emotional determinants of psychological process. Then, the year following the publication of Naven, Ralph Linton went to Columbia and soon became the dominant figure in a group of anthropologists collaborating with the distinguished analyst, Abram Kardiner. The impact on anthropology of their work was so great that since that time the main stream of research and theoretical development in culture and personality has virtually taken for granted the assumption that its primary data are to be found in the realm of emotion. It is ironic that of all the fields of inquiry into human behaviour, anthropology, with its primary emphasis on the regularities of behaviour as they are transmitted through culture from one generation to the next, is the one which most consistently ignores the cognitive learning involved in this cultural transmission.
True, the theory of learning formulated by Clark Hull has had considerable vogue in anthropology, but this theory is far more concerned with motivation and reward for learning than it is with the cognitive integrations accomplished in learning. Aside from Hullian theory, attention to the processes of learning and thinking and to the nature of intelligence has been minimal in anthropology. Bateson did carry eidos one step further, developing in collaboration with Margaret Mead the concept of deutero-learning (Bateson, 1942), referring to learning how to learn or to the content and logical process of learning. This was a useful concept, and although they did not develop it further, both Bateson and Mead are far more careful than most anthropologists to make explicit the nature of the learning process in the cultures they have described. More recently the emerging field of psycholinguistics and contributions to a few symposia have reflected a growing interest in cognitive categories and processes, but this remains a scattered effort.1 It is in no way comparable to the numerous but sometimes frustrating endeavours of anthropologists to derive a variety of personality types from common emotional experiences ordered within a Freudian framework.
I became acutely aware of this neglect by anthropologists some six years ago when, in collaboration with Seymour B. Sarason, I undertook a review of research in mental subnormality (Masland, Sarason and Gladwin, 1958). The first question of course was āWhat is mental normality?ā ā i.e. āHow do we define intelligence?ā This is obviously a difficult question, even disregarding cultural differences. Most psychologists in effect define intelligence operationally as being that which intelligence tests measure. People with high IQs are intelligent, and those with low IQs are variously dull, duller, or idiots. There has been much soul searching by psychologists on this score, but no more adequate definition has yet gained general acceptance.
I also found that numerous non-European peoples, many of whom do rather bright things, had been given intelligence tests by both psychologists and anthropologists with due attention to linguistic and other handicaps and had consistently come out with low IQs. This could mean either that we are of a master race, or else that we are in effect accepting an assumption that there is only one really good way to use the human brain, and that is our way ā whatever thai may be. Since the latter explanation seemed more reasonable, I endeavoured to find out what anthropologists had done to rectify the situation.
More particularly, I hoped to find in the work of anthropologists some substantive research which focused on cultural differences in modes of thinking and problem solving, research which could point to other criteria of intelligence than those valued by European-American professionals. Such studies would permit us to define ā and ultimately to measure ā different kinds of intellectual achievement, rather than, as now, to seek ever more ingenious (or ingenuous) ways of measuring our kind of achievement. From this in turn might come a more comprehensive and operationally useful definition of intelligence. However, my conclusion from this search was essentially the one set forth above; emphasis by anthropologists has been almost exclusively on the emotional determinants rather than on the cognitive aspects of personality.
I then turned to re-examine an earlier collaboration with Sarason, a personality and culture study of Truk (Gladwin and Sarason, 1953). We had tried a simplified intelligence test and had at least been enlightened enough to see that it was inappropriate to Trukese perceptual patterns rather than to conclude that their poor performance reflected poor brains. We, or particularly Sarason, had discussed at some length the very concrete, non-abstract nature of Trukese thinking. We earned an honourable mention for that bit of insight, but no medal. A medal was unwarranted for several reasons. First, in my fieldwork I had paid little attention to learning, and no attention to how the Trukese learn to think and learn (Batesonās deutero-learning), so we could offer only rather weak speculations as to how this style of thinking developed. Second, we had not developed our analysis of Trukese thinking in a frame of reference such as to command the theoretical attention of other psychologists or anthropologists. We did not, in other words, relate it to theories of cognition and intelligence in a meaningful way. Finally, in undertaking the analytic interpretation of Trukese personality development, I sat just as firmly at the feet of Freud as did any of my contemporaries, in spite of the realization that Trukese thinking and learning was significantly different from that employed by the patients or subjects of Freud or of contemporary psychologists. I did not yet appreciate that if behaviour is learned, as we culturally-oriented anthropologists keep insisting, the nature of the learning process must have something to do with the kinds of behaviour learned.
I cannot in this paper attempt to expiate all my past errors and omissions. I of course still have no data on how Trukese learn to think. Speculation on the relationship between the nature of learning and the kinds of behaviour learned is tempting, but could only be speculation. Instead I shall attempt to place in meaningful perspective the nature of Trukese intellectual achievement, differing as it does from our own in the kinds of thinking employed. Description of a variety of cognitive styles is a necessary first step toward the goal of studying how people learn to think.1
In our culture we value (and measure crudely with intelligence tests) relational or abstract thinking, in which bodies of knowledge are integrated and related to each other through unifying symbolic constructs. The Trukese seemingly do not, relying instead on the cumulative product of the adding together of a great number of discrete bits of data, summed together in accordance with predetermined parameters, to arrive at a desired conclusion. Both we and the Trukese operate within a gestalte a conception of the problem as a whole. However, we seek a unifyi...