Part One
Breaking Down Barriers through the Study of Culture in the Study of Mind
ONE
Challenging Developmental Doctrines through Cross-Cultural Research
ROBERT A. LEVINE
Harvard University
Graduate School of Education
If our nervous system as evolved through natural selection provides the universal hardware of human behavior, and our culturesâas historically transmittedâsupply its variable software, then psychology should offer a conception of the interactions through which children develop according to the divergent standards of human cultures. That there are many standards rather than one is a basic assumption in approaches to child development, supported by a large and growing body of evidence that anthropologists have accumulated. But psychologists have long been tempted to treat development as singular rather than plural.
G. Stanley Hall, a founder of child psychology, gave us the concept of adolescence (in 1904) but grounded it in the theory that ontogeny recapitulates phylogeny. The behaviorists claimed that the reinforcement principle could account for all behavioral outcomes. Jean Piagetâs ([1951] 1962); Flavell 1963) ambitious and comprehensive genetic epistemology assumed that the course of cognitive development in all humans could be generated from observations of children in Geneva during the early and middle twentieth century. Lev Vygotskyâs ([1934] 1962) theoretical formulations correcting Piaget opened the door to social influences but assumed them to be those that improve childrenâs development by pushing it toward goals prespecified by progressive social theory.
All these theorists thought it was possible to construct a model of development that could fit the human species as a whole largely on the basis of evidence about children in a particular place at a particular time. From the perspective of cultural psychology and psychological anthropology in the twenty-first century, this assumption is naive. A better assumption might be that of Karl Popperâs (1963) falsificationism, that scientific knowledge consists of those conjectures that have survived falsification by empirical evidence; science is constantly evolving as new evidence falsifies old conjectures. Child psychology, in trying to imagine the complex processes through which an infant becomes an adult in terms of behavior and mental development, is necessarily conjectural, but its conjectures should be subjected to a sturdy process of winnowing in which some generalizations survive to be recognized as (provisional) knowledge (Campbell 1974).
For most child psychologists, that winnowing process was provided by the laboratory experiment. They followed a pattern established during the behaviorist ascendancy in US psychology, roughly 1935â1960, when âthe laws of learningâ that applied to vertebrates in general were thought to be established through controlled experiments with albino rats that had been deprived of food for twenty-two hours. Psychologists from the 1960s onward conducted experiments with human children in which background factors were held constant so that the outcomes from factors manipulated by the experimenter could be detected. Thus, when Piagetâs observational assessments of childrenâs thought were subjected to experiments in the United States, the initial focus was on whether experimental conditions could accelerate childrenâs age-linked mastery of tasks, falsifying Piagetâs static portrait of age-linked stages. Then there was examination of whether the correlations among diverse abilities posited by his structural theory would hold up in empirical testing. The structural theory was declared dead at the Minnesota Symposium on Child Psychology in 1979 by John Flavell (published in 1982), who had done more than anyone to interest US psychologists in Piagetâs work some fifteen years earlier (Flavell 1963). This was a triumph of falsification through the laboratory experiment.
For developmentalists less imbued with the standards of the Journal of Experimental Psychology, the ânatural experimentsâ of Patricia M. Greenfield (1966) comparing schooled and unschooled children among the Wolof of Senegal and Douglass Price-Williams (1962) on the Tiv in Nigeria provided more convincing evidence to critique Piagetian theory. Greenfieldâs study of Wolof children strongly suggested that Piagetâs conservation task results in Geneva depended on experience in Western schools rather than simply reflecting an inherent developmental sequenceâan issue that could not be investigated in Geneva or anywhere else where schooling was universal. But for child psychologists inspired by a scientific ideal derived from the laboratory experiments of I. P. Pavlov and E. L. Thorndike rather than Darwinâs field research, her discovery was too exotic, too sociological, to be taken as seriously as it should have been.
In anthropology at the time there was one place where an anthropological application of the conjecture-and-refutation model of science held sway: in John W. M. Whitingâs research group at Harvard. Whiting had tested hypotheses derived from Freudâs theory of psychosexual development in ethnographic reports from seventy-five cultures and found support for only some of them (Whiting and Child 1953). His students in seminar were urged to offer their own conjectures for explaining a given body of comparative data and suggestions for testing their conjectures against the cross-cultural evidence. The assumptions were that variations at the cross-cultural level could provide powerful, nonexperimental, evidence for and against developmental conjectures, and that generalizations based only on Western evidence were nothing more than speculations. For those of us trained in that research group, including several authors in this volume (Richard A. Shweder, Roy G. DâAndrade, Thomas Weisner, and me), the provisional character of monocultural findings was a fundamental methodological principle.
The Hausa Study
In 1968 I planned a long-term study of child development among the Hausa of northwestern Nigeria that had two parts: (a) exploring through ethnographic observations the social and cultural contexts of parents, infants, and children in a Hausa community; and (b) a program of replicating Western studies of developmental topics ranging from infant attachment to cognitive development with Hausa children in their natural social settings. If the studies, carried out by Nigerian university students in a child development training program, showed findings differing from those with Western children, the contextual data would help to explain why and might provide explanations making sense of the results in the Hausa context.
In the infant studies, Sarah LeVine observed mothers and babies in homes according to a schedule designed to assess conditions specified by Mary D. S. Ainsworthâs Baltimore research (Ainsworth, Blehar, Waters, and Wall 1978).
In the cognitive development research, Douglass Price-Williams and I designed assessments based on Piagetâs ([1928] 1968) early work in Geneva on the judgment and reasoning of the child (LeVine and Price-Williams 1974). We studied fifty-three Hausa children four to eleven years old; many had never been to school. (Our assistants gained their compliance in the assessment situation by letting them hear their voices on a tape recorder.) We adapted Piagetâs questions to fit the family environments of the Hausa, in which people live in walled compounds that often contain several families. The results showed that as children get older they become more able to say how others in the compound are related to each other instead of simply how each is related to the childâconsistent with Piagetâs model of development from an egocentric to a relativistic perspective. And this was correlated (controlling for age) with left-right reversibility, that is, a childâs ability to know that her right hand is on the âsameâ side as the left hand of the person facing herâsupporting Piagetâs structural theory of cognitive development.
At the same time, however, there were signs of environmental influence: Children tended to define the Hausa word for grandmother (kaka) as motherâs mother, reflecting the salience of the maternal grandmother, who typically lives in a different community, as the most likely primary caretaker when a child is ready for weaning. And the Hausa customs of name-avoidance and social distance from certain categories of kin living in the compound also were reflected in the childrenâs responses, suggesting enculturation of the growing child. Thus we could argue that a simple interview procedure focused on kinship could reveal the childâs developmental status in Piagetian terms, her acquisition of the local culture, and her personal experience in that cultural environment. Unfortunately, political conditions related to the Biafran civil war did not permit us to continue the research beyond the first year. But there are other bodies of data collected at the same time, and with my new colleague Richard A. Shweder, I published another article on the dream concepts of Hausa children (Shweder and LeVine 1975).
We were of course following in the footsteps of Margaret Mead, both in testing developmental conjectures with anthropological evidence and in some cases showing the culture-bound nature of Piagetâs findings. Mead (1928) had used her Samoan observations to challenge the universality of adolescent turmoil posited by G. Stanley Hall, and despite the much later attack by Derek Freeman (1983), her conclusions from that study about the culture-specific nature of adolescence still stand today. Her 1928 book Coming of Age in Samoa was revolutionary in demonstrating not only that development could vary across cultures but also that anthropological observations could be used to refute conjectures offered by Western psychologists. When she conducted fieldwork among the Manus of Melanesia a few years later, Mead examined how some of Piagetâs ideas hold up in that society. She found that the âanimismâ of Manus children increased with age, instead of decreasing as Piaget had found among children in Geneva, in accordance with their respective cultural environments, suggesting that what psychologists treat as âdevelopmentâ may also be âenculturation.â
Meadâs article on the Manus was published in 1932, more than thirty years before Piagetâs work became a recognized research topic among US psychologists, and for reasons that remain unclear, it appeared in the Journal of the Royal Anthropological Institute, where psychologists would be unlikely to read it. Perhaps this reflected Meadâs increasing alienation from the field of child psychology. But in 1973, when I was asked to contribute an article to a special issue of the (then) new journal Ethos dedicated to Margare...