Chapter 1: LCHC, a Pre History

In order to understand the mixture of academic and social issues that came together in the formation of LCHC, we begin by elaborating in a little more detail the socio-historical circumstances within which the lab emerged. We then summarize the basic research program of cross-cultural research that provided the initial set of academic issues motivating the formation of LCHC as a research laboratory.

The Social-Historical Context

During the 1960's, several social trends converged to focus both scholarly and broader social attention on the relationship between cultural experience and intellectual development. The first was a global focus on spreading and intensifying formal education as a mode of economic and social development a necessary accompaniment to the process of de-colonization. The second was the intense international conflict known as the Cold War in which the Soviet Union and its allies were pitted against the United States (U.S.) and its allies for global influence and, in public discourse, for global domination. In an era of ICBMs and Nuclear Weaponry winning this conflict was seen not only as a matter of national survival, of preserving "our way of life."

The International Context

The launch of Sputnik in 1957 challenged the U.S. in profound ways. The existing sense of technological superiority was shattered. In the aftermath of Sputnik, the U.S. initiated a wide variety of educational reforms focused science and technology (what are now referred to as the STEM disciplines) in an effort to “catch up” in a technological race that it thought it was leading.

The importance of elevating the educational capacities of “Third World” nations represented a point of convergence between the U.S. national desire to improve the education of its population as a mode of national survival and international efforts to promote economic and development of “under-developed” nations. The “Third World” comprised largely of former European colonies, was an important arena of Cold War competition as each side sought to extend its influence through programs of foreign aid that provided support for building formal educational infrastructures.

The international programs created to fulfill these educational goals shared a common understanding of their task, expressed as a basic premise in the planning documents of UNESCO, which was their role as a coordinating mechanism of otherwise contending nations. International academic and policy makers agreed that low levels of economic, political, and cultural development are associated with immature thinking that blocks the economic changes that the United Nations (UN) architects of the future were imagining for the world.

This premise linking "socio-cultural-economic development" with intellectual development came through clearly in the statements of UNESCO's founders when they spoke of the need for a "wide diffusion of culture" to the Third World. This "giving them culture" orientation comes through clearly when, in the same document, UNESCO writes that "...ignorance is not an isolated fact, but one aspect of general backwardness which has many features, like paucity of production, insignificant exports, poor transport and communications, deficient capital and income, [etc.]” (UNESCO, 1949, p. 4).

Much in the spirit of UNESCO's view, development economist Daniel Lerner argued that a key attribute of modern thinking is the ability to take another person's perspective and to empathize with their point of view (Lerner, 1958). Lerner was quite specific about the relationship between psychological modernity and modern economic activity. The ability to take another's point of view, he wrote, "is an indispensable skill for moving people out traditional settings... Our interest is to clarify the process whereby the high empathizer tends to become also the cash customer, the radio listener, the voter" (Lerner, 1958, p. 50).

If we put what these experts were declaring together, we see develop a belief that Third World peoples lacked critical forms of intellectual ability owing to deficiencies in their environment. As a consequence, so goes this logic, they could not enter into modern economic and political relations. The solution, then, was to spread formal, industrial-style education into the Third World, along with our “goods and services”.

The Domestic Context

In parallel with the international situation, social processes within the U.S. made education a focal point for discussions about race, ethnicity, and social inequality. The 1954 Brown vs. Board of Education decision started the process of desegregating American schools but it did not produce any visible change in the relative achievement of African American children. This situation, given the name of “the achievement gap”, remains a central issue in debates about American educational policies today.

In 1964, the U.S. initiated a “war on poverty”. One of the major weapons in that war was Project Head Start, a massive program of early childhood education and health care. Early childhood intervention, based upon a variety of reputable data on child development, it was believed, would show successive reduction of the achievement gap and children given a proper head start progressed through the grades. This hoped-for solution did not appear to work.

The Intellectual Context

In light of the broad interest in using formal education as an engine of human development, a number of governmental and philanthropic organizations became actively interested in research that would help plan how to organize for mass education. During the late 1950's and continuing throughout the following decade, a number of social science research programs were mounted to provide empirical information about cultural variations in psychological development and how such variation should influence educational practices.

These programs of research sprang from different theoretical sources not only within psychology, but also within a broad range of developmentally relevant disciplines that now would be gathered under the rubric of developmental science.

By far the largest of the projects to come directly out of the discussions was motivated by broad interest in Piaget's theories. Early in his career, Piaget (1928) argued for the existence of important developmental differences in cognitive development associated with cultural variations. Citing the work of Levy-Bruhl (1910) and Durkheim (1912), Piaget distinguished between two kinds of societies each with its own mode of thought, which Levy-Bruhl characterized as "primitive" and "civilized" and which today might be distinguished as “traditional” and “modern”.

Piaget wrote an influential article calling for cross-cultural evaluation of his ideas (Piaget, 1966). In it he summarized then-extant research and concluded that education could be considered one of four basic factors associated with development. This represented something of a concession for Piaget because in some of his writings he had argued that formal education fails to promote development because schools are institutions where accommodation overpowers assimilation, resulting in inert knowledge. In 1966 he conceded that in so far as educational experience should provide children more overall experience relevant to discovering the nature of the world, true developmental differences could be created in either the rate or the final level of development.<br />

In general, the spate of Piagetian research begun in the 1960's found a good deal more variability among societies than Piaget had anticipated. In some cases, little or no difference in achieving Piagetian developmental milestones was observed. In other cases, however, researchers reported that children from traditional, non-industrialized societies lag one or more years in the age at which they achieve the stage of concrete operations as compared to children in Western industrialized societies. There were also a few studies in which 12 and 13 year old children and even adults failed to demonstrate understanding of the principles underlying conservation in the Piagetian tasks. This sort of finding led to the widely cited conclusion that perhaps some societies create experiences for their children (e.g. formal education) that speed up cognitive development and develop it to a higher level than others (See LCHC, 1983 for an extensive review of this literature and the other programs mentioned below)

For further reading see:
Cross-Cultural Article

Another prominent program of research began with the idea that cultures are "designs for living," which suggests that one should seek cultural influences on development in terms of cognitive styles --that is, pervasive tendencies to think about the world in ways that fit into the particular, unique, pattern of the culture.

Read more

In the 1960's, cognitive style research was part of a broad attempt to demonstrate that different styles can arise in quite different geographical environments depending upon key features of how the local culture has formed as a medium for dealing with its environment; socialization practices and features of the environment jointly entered into such theories. Schooling was not a central concern to cognitive style research because it was seen as part of a more encompassing socio-cultural environment.

By contrast, with either the Piagetian or Cognitive Style approaches, schooling was a central concern in the approach to culture and development that began to develop in Liberia in the mid-1960’s. LCHC can usefully be said to have begun its process of formation as a small part of an international educational research initiative sponsored by the Educational Development Corporation (EDC), an influential education research organization located in Cambridge, Massachusetts. At the time, EDC served as an organizational base for number of senior scientists in the country's leading institutions who had become concerned about the need to organize for a greatly expanded labor pool of scientifically and technically educated people to sustain American economic prosperity, including its international political and economic position in the world.

This group pioneered a number of the innovative, activity-centered curricula in the hard sciences and biology. One of its members, Jerome Bruner, was then leading the controversial curriculum effort, Man: A course of study, which adopted a view of culture as the historically constructed ways in which people deal collectively with their environment. The curriculum sought to illustrate this point and to serve as a provocation to think about how cultural variation, thus understood, can be made consequential for organizing successful education.

Surrounded as they were by influential Washington figures on the one hand and the massive effort to overcome the developmental disadvantages of poverty and ethnic discrimination on the other, these men (for they were all men) decided to determine if efforts at pedagogical reform in mathematics education could also be implemented in Africa, where mathematics education was running into difficulties. They decided to use their "New Mathematics" curriculum as the guiding logic of educational reform. In pursuit of this goal, they convened a conference in Entebbe, Uganda, to enlist interested African partners.

There is no point in speculating about how the idea of exporting the new mathematics curriculum might have influenced the future of African mathematics if it had ever been seriously implemented. It never was. But even at the outset, one person was quite clear about his doubts concerning the utility of the new approach to the problems of African mathematics education as he had experienced it.

That man was John Gay, a one time a graduate student in mathematics at Princeton University. Gay left Princeton and opted for the ministry, ending up in Liberia where he taught mathematics at Episcopalian Cuttington College.

Supported by bits and pieces of money left over the from the project that brought many scholars together in Entebbe, Gay and Michael Cole were allowed to conduct some initial studies to ferret out the sources of poor performance based on apparent conceptual misunderstandings that Gay had identified. This work (Gay and Cole, 1967) attracted sufficient scholarly attention to enable them to obtain a rather substantial grant to follow up on their initial findings.

For this second round of work, they recruited the help of Joe Glick, a developmental psychologist with an interest in the issue of cultural contributions to development. Together with Donald Sharp, a former undergraduate assistant of Cole and Glick’s at Yale and a host of local Liberian college students, this team set out to see if they could put together a more adequate account of cultural differences in cognitive development, and how these differences were related to formal schooling.

Each project sparked a re-examination of how scholars warrant their conclusions about the role of culture in human development, the influence of particular of institutionalized formal education on cognitive development, and the nature of psychological research into these issues. We will treat them as "two steps" toward the formation of LCHC.

Refocusing the Discussion: New Mathematics and an Old Culture

At the time when Cole and Gay began their research, the idea that local tribal children suffered from specific forms of cultural deprivation was ubiquitous among the educationalists working in Liberia. As Cole summarized the situation:
The list of things that the tribal children could not do, or did badly, was very long indeed. They could not tell the difference between a triangle and a circle because they experienced severe perceptual problems. This made the tribal child's task almost hopeless when it came to dealing with something like a child's jigsaw puzzle, explaining why "Africans can't do puzzles." I heard a lot about the fact that "Africans don't know how to classify" and, of course, the well-known proclivity of African schoolchildren to learn by rote came in for a lot of discussion and help.
Although such conclusions drew upon a long history of 19th and 20th century scholarship not to mention UNESCO documents such as those cited above, the people who made them were basing their ideas entirely upon fantasy. They had the evidence of their own experience. Children who played around the living quarters at Cuttington College experienced what seemed like extraordinary difficulty doing jigsaw puzzles. Students attending the college appeared to experience extraordinary difficulty mastering the mathematics curriculum. Various forms of extreme rote learning were observed in elementary school mathematics classrooms. Yet the broad conclusions about cultural causality based upon such observations seemed overdrawn, especially since they all took place on the educator's home turf, culturally speaking. And the solutions offered "Flood the country with erector sets" were not particularly helpful.

Rather than jump to conclusions about Liberian child and adult intellectual disabilities, Gay and Cole approached this issue with the added common sense assumption that, although Kpelle children lacked particular kinds of experiences routinely encountered by children in the U.S., they were by no means lacking in experience. They explicitly began with the assumption that “we must know more about the indigenous mathematics so that we can build effective bridges to the new mathematics that we are trying to introduce” (Gay and Cole, 1967). This core assumption led them into an exploration of the way that numbers, geometrical forms, and logical operations are expressed in the Kpelle language.

A second, common sense assumption of this early work was that people would be skilled at tasks they had to engage in often as a part of their everyday activities. This statement may seem patently obvious or trivial, but its consequences are neither. It led directly to the conclusion that if they wanted to understand why Kpelle children have difficulty with the kinds of problem solving tasks American academics consider relevant to mathematics, they needed to consider the circumstances in which Kpelle people encounter something recognizable to us as mathematics. That is, they needed to study people's everyday activities that involved measuring, estimating, counting, calculating, and the like, as the precondition for diagnosing indigenous mathematical understanding in relation to schooling.

Gaining attention to the challenges posed by these two assumptions appears, in retrospect, to be the cardinal contributions of this work. It was not sophisticated by any stretch of the imagination. But it caused people to stop and think.

Among the specific tasks that they created, one in particular sparked widespread interest in the academic community because it seemed to capture well the kind of research one would like to be able to carry out ––estimating amounts of rice. The Kpelle have traditionally been upland rice farmers who sell surplus rice as a way to supplement their very low incomes. Upland rice farming is a marginal agricultural enterprise and most villages experienced a “hungry time” in the two months between the time when their supply of food from the previous year was used up, but the new crop had not yet been harvested.

As might be expected from the centrality of this single crop to their survival, the Kpelle have a rich vocabulary for talking about rice. Of particular concern to Gay and Cole was the way in which Kpelle talked about amounts of rice, because they had been told rather often that Kpelle “can't measure” and wanted to see what would happen if they were asked to measure in a domain of deep importance to them.

Simple ethnographic investigation revealed that the local standard minimal measure of rice was a kopi, a U.S. tall tin can that holds one dry pint. Rice was also stored in boke (buckets), tins (tin cana), and bags. The kopi acted as a common unit of measure. It was said that there were 24 kopi to a bucket and 44 to a tin, and that two buckets were equivalent to one tin. People claimed that there were just short of 100 kopi to a boro (bag), which contains about 2 tins. The relationships between cups, buckets, tins, and bags are not exact by our standards, but they are close, and they reflect the use of a common metric, the cup.

Gay and Cole also learned that the transactions of buying and selling rice by the cup were slightly different in a crucial way. When a local trader bought rice, he used a kopi with the bottom pounded out increasing the volume in the container; when he sold rice, he used a kopi with a flat bottom. They surmised that this small margin of difference would make a big differences to people who refer to two months of the year as hungry time, so we decided to experiment on people’s ability to estimate amounts of rice in a bowl, using the local measuring tool, the kopi made of a tin holding a dry pint in U.S.

Using these materials, they conducted a “backwards” cross-cultural experiment modeled on Kpelle practices using Kpelle bowls and cups. The participants were working class Americans and Liberian farmers. Each person was presented with four equal size mixing bowls and asked to estimate the number of cups of rice in it. At the same time, they were shown the tin to be used as the unit of measurement. The bowls contained 1½ , 3, 4½ , and 6 kopi of rice. Kpelle adults were extremely accurate at this task, averaging only one or two percent error. American adults, on the other hand, overestimated the 1½-can amount by 30% and the 6-can amount by more than 100%.

These results supported their basic assumptions that people would develop cultural tools and associated cognitive skills in domains of life where such tools and skills were of central importance, as rice was central to Kpelle life. Whatever the cultural differences with respect to the domain of mathematics, a total lack of measurement concepts and skills did not appear to be one of them.

A conspicuous feature of the rice estimation task was that Gay and Cole were able to create an experiment which not only incorporated local knowledge concerning content (e.g., the Kpelle system of units for amounts of rice) but also local procedures under which such estimates would be relevant. At markets one often encountered women selling rice from large bowels containing various amounts. Consequently, this task provided Kpelle people with both relevant content and relevant procedures.

Not all of the methods Gay and Cole used achieved this kind of “groundedness”. Uncertain of the scope of the question they were being asked to address, they actually studied a spectrum of psycho-educational tasks, some imported directly from U.S. psycho-educational practices, others made up on the spot. This category included simple classification of objects, a simplified jigsaw puzzle, and a vocabulary test of the sort often found in IQ tests. When confronted with this sort of task, attending school appeared to improve people’s performance.

Overall, the data suggested that no generalized lack of mathematical, perceptual, or problem solving abilities stood in the way of mathematics education. When the materials and procedures used in assessment tasks were designed to match closely valued local practices, lack of ability could be replaced by apparent special ability. At the same time, schooling did appear to influence performance in tasks that were routinely used to measure cognitive development.

These results seemed intriguing and worthy of more systematic attention, particularly since many were based upon untested procedures that had been cobbled together in the field. Gay and Cole clearly had conducted far too little ethnographic work to provide more than a hint of what might be learned beyond the domain of mathematics. The population differences relating performance to age and education were unsystematic.

Fortunately, the attention attracted by this initial foray into cross-cultural research was sufficient to obtain the funding needed to pursue the agenda opened up by the “New Mathematics” project. Equally important, Gay and Cole obtained the collaboration of Joe Glick, who was trained in classical developmental psychology, adding a central area of expertise that was lacking.

Struggling with Context: Ethnographic Psychology & Experimental Anthropology

This second round of research was funded by NSF in 1965. That support was sufficient to sustain 4 years of fieldwork. After two further years of gestation, an account of the new findings was published as a monograph.

Some Matters of Social and Academic Context

In the years that passed between the initiation of the “New Mathematics” project and the completion of what we will call the Cultural Context Project, the U.S. underwent a fundamental shift in its foreign and domestic policies that brought the civil rights movement into convergence with opposition to the war in Vietnam. Martin Luther King's dreamers of the early 1960’s had lost their great inspirer, and his death just one of many manifestations of the violence that wracked the U.S.

The authors of The Cultural Context of Learning and Thinking (Cole, Gay, & Glick, 1971), a monograph that resulted from The Cultural Context Project, were, like a great many academics, caught up in the conflicts of the time, which re-framed their work in a significant way. For reasons we return to in the next chapter, this cross-cultural work entered into debates at the national level about the value of the massive head start program and debates at the academic level fostered by Arthur Jensen’s publications attributing difference in school success to genetically determined factors associated with race.

This framing of the work is important to consider because it explains why research on cultural differences between people living in rural Liberia and suburban California seemed relevant for the American scene. When imported from Liberia to New York, results showing great variety in the way rural Liberian farmers solved intellectual problems disrupted the idea that any sort of general deficit afflicted these people. It also struck at the idea that the apparatus of experimental psychology, as then practiced, could legitimately claim to reach conclusions about learning ability. This line of thinking came to be associated with what is referred to as a “difference” interpretation of cognitive development and cultural milieu. Within the American context this difference interpretation was contested by those who assumed that ethnic minority/poor children grew up in environments that were deficient, also known as the “cultural deficit” position.

liberia guys  1.jpg
Cultural context research team (Don Sharp, ?, Tom Ciberowski, Paul Mulbah

housebuilding (1).jpg
House building

Both points of view exist to the current day, even as increased knowledge about biological-cultural interconnections has made the debates more complex. We will return in later chapters to consider the current situation, but we pause here long enough to give a flavor of the additional kinds of evidence that several years of research using a broad variety of methods added to conversation.

Putting Together a More Adequate Methodology

A general message that we took away from the “New Mathematics” project is clearly stated in the preface to the new monograph, The Cultural Context of Learning and Thinking: “People will be good at doing that things that are important to them and that they have occasion to do often” (Cole, Gay, & Glick, 1971, p. xi). This meant that ideally, we would spend a few years doing a comprehensive “cognitive ethnography” of Kpelle society and use it as a matrix against which to conduct systematic studies of culture, schooling, and cognitive development. In reality, we could not wait for an ethnography of such a sort to be completed. We needed to deliver results within three years ––ethnography, schooling, testing, the whole ball of wax.

The approach that emerged to meet this impossible set of demands began, so to speak, from both sides at once. On the one hand, we did have ethnographic information and did use it to assess the kinds of thinking that people engaged in. But at the same time, we also had experimental procedures that seemed to be of a simple sort, easily communicated, that could at least get us started on trying to figure out how age and education were implicated in the levels and modes of performance that people engaged in.

This age/education nexus was of keen interest in Euro-American psychology at the time because there appeared to be broad agreement that a marked change occurs in cognitive processes in the age period from 5-7 years (for more, see the 5-7 shift). For example, this is the age period that marks the advent of concrete operational thinking in Piaget’s work. It was also the object of attention of American learning theorists, who linked these changes to marked changes in opportunities for learning that children began to experience during this age period, whether in school or outside of it (White, 1965).

With respect to disentangling the factors involved in the 5-7 shift, what made the situation in Liberia so attractive was that schooling was still restricted geographically primarily in relation to towns built along roads constructed by foreign concessions for the extraction of natural resources. Moreover, in most places, schooling had only arrived recently so many people of all ages had not experienced formal education. This set of historical circumstances made it possible, at least in theory, to separate the influence of the child’s chronological age from the kinds of changes in experience they underwent in connection with schooling as well as indigenous activities.

Examples of “Ethnographic Psychology”: Experimental Anthropology

From the large set of studies conducted over a four year period, we have select three examples of strategies we used to combine ethnographic investigation of Kpelle language and culture with the analytic apparatus of experimental psychology, which purported to provide evidence about mechanisms of learning and problem solving.

A) Learning about Leaves: In a poignant description of her early attempts to learn Igbo during her fieldwork in Nigeria, Elenore Smith Bowen (Laura Bohannan), recounts her great difficulty in learning the names of the different leaves that people used in their everyday life (Return to Laughter). Beryl Belman, then a graduate student and our project’s “cognitive-ethnographer”, had the same problem when he sought to study local medicines. Leaves are central to local medical practice, and Beryl could not distinguish them, practice though he might.

We used this observation to create a learning experiment in which the task was to classify leaves. The leaves were of two kinds: those taken from local vines and those taken from local trees. In the most straightforward case, people were told that they would be shown a set of leaves, one at a time, and they were to divide them into vines and tree leaves. Under these conditions, all of the uneducated Kpelle adults tested made the correct classification either right at the start, or after one pass through the set of leaves. No problem. After 10 cycles through the leaves, each identified as leaf or vine each time, American Peace Corp volunteers could not learn to classify twelve leaves.

As an indicator of how important the categorization of the leaves by local kind would be to those who learned the categories, we included a condition where the two kinds of leaves were assigned to categories at random. In this case, the Kpelle subjects (psychologese for the people who are asked to engage the task) as well as the Americans experienced great difficulty, although the Kpelle gave some evidence of learning while the Americans did not.

These results are more or less what the ethnographic examples would lead us to expect. However, a third experimental condition greatly complicated matters. In this condition, the two categories of leaves were again to be sorted into categories of vine versus tree leaves. However, in this case, the experimenter did not designate the leaves by referring to them as vine or tree leaves. Instead, people were told that half the leaves belonged to a man name Togba and the other half to a man named Sumo. In this case, the Kpelle participants failed to learn to sort the leaves.

This left us with a puzzle. Why, if we know that local people can sort those leaves when they are asked to do so in a straightforward manner, do they fail to do so when the categories are signaled not by their habitual labels, but by a fictitious “person” said to posses them?

B) Categorical Organization and Remembering: One of the most extensive series of studies carried out as a part of the Cultural Context Project concerned the way in which people learn and organize sets of objects or lists of words. The specific procedure is referred to as “free recall” because the items to be remembered, although presented in sequential order, can be remembered in any order the person chooses. Interest then is focused on the order in which the words are remembered and how many are remembered. In particular, if the set of items is made up of clearly recognizable and familiar categories, will people remember them in the order presented or “cluster” them by categories? After a good deal of preliminary work establishing the items and categories to be used, the studies were carried out in ways that contrasted age and schooling to the extent possible.

In the most straightforward replication of procedures used in the U.S., people were told to remember a list of common nouns consisting of familiar objects: clothing, food, utensils, and tools. They were read the list one item at a time and asked for recall in any order. This procedure was repeated five times with the order to the items scrambled each time.

These initial “list reading” studies indicated that traditional Kpelle farmers recalled relatively few items and failed to learn many new items with repeated presentations of the list. The sequence in which words were recalled conformed to no recognizable pattern of organization. This generalization held even though the experiment was modified to include monetary incentives, different kinds of words, concrete objects, and a variety of other stratagems aimed at producing good recall.

By contrast, Kpelle who attended high school remembered well, learned rapidly, and manifested a high degree of conceptual organization in the way they recalled the lists. One implication of this work was that developmental changes in free recall and clustering observed in the U.S. might well be measuring something about the influence of years of formal schooling, not age. But interesting as that possibility might be, it does not explain why people who had not attended school appeared to be so inept at remembering what were, corroborating studies showed, familiar items and categories to the people involved.

A final study in this series further complicated matters. Drawing upon the Kpelle practice of story telling and riddling, we created a story that mimicked a local genre but contained the same to be remembered categories of items as in the prior research. When integrated as part of the story, the order in which items were recalled conformed to the categories into which they had been grouped in the story.

Again we have a case where, when deliberately designed psychological task is presented, Kpelle rice farming people appear to experience extraordinary (to us) difficulties –except when the categorical/conceptual relations within the materials they are asked to learn and think about are made clear either through explicit linguistic designation or mediated by a familiar cultural practice (in this case, using a local story telling practice).

C) Making Simple Inferences: Our last example illustrates that following the procedure of iteratively modifying pre-scripted cognitive tasks to tease out the factors associated with low or high performance fundamentally changes interpretations of cognitive development based on research in the U.S.

The task, in this case, was to learn to make a simple inference by piecing together three easily learned associations, and inferring what combination of two of those associations will produce a reward (candy for children, money for adults). We selected this task for two related reasons. First, it had been used in research with American children and appeared to show the kind of “5-7 shift” in modes of responding associated with stage theories of cognitive development. Consequently, it seemed pertinent to our goal of separating age and schooling in the development of cognitive processes. ßSentence structure. Second, it was part of a suite of tasks devised to test the idea that the changes in performance are symptomatic of a general intellectual change in which language contributes to thought processes in a new way (White, 1965; Vygotsky, 1962). Since features of language use in classrooms were then, and remain, prominent in psychological debates about schooling effects, the experimental procedure seemed a promising way to address these concerns.

We began this set of studies using the device that Tracy and Howard Kendler used in their own research in the U.S. (Kendler & Kendler, 1962). (Figure 1). The apparatus was made of tin and when used in the U.S., it operated electronically. Somewhat fearsome, we thought. We learned to operate the apparatus by hand, but the idea was to stay as close to a real replication with the original apparatus as possible.
Figure 1: Infernal Box
Figure 1: Infernal Box

The problem is presented as follows: First, subjects are taught that pushing the button on the left-hand panel will yield a marble, which drops into a tray below. Next, they are taught that pushing the button on the right-hand panel will yield a ball bearing. Then, with the two side panels closed, they are taught that putting (say) the ball bearing in a hole in the center will yield a piece of candy. When all of this has been learned, all three panels are opened at once, and the subjects are instructed to obtain the candy, which they can keep and eat. The question is: will they choose to obtain the marble or ball-bearing to get the candy, a task they have not been explicitly taught?

Described in this manner, the task facing the subjects was absolutely trivial. But, implemented in real life, whether in Santa Barbara or rural Liberia proved challenging for some of the groups of people studied. In the U.S., the Kendlers had found that even at third grade (roughly 8 years of age) only half the children in their studies would immediately start pushing the correct button and then put the correct object into the hole to get the candy. Older groups of children instantly solved the problem.

When this same task was presented to groups of traditional Kpelle (both children and young adults), performance was very unimpressive. For example, only 15% of the young adults who had never attended school spontaneously solved the problem. School attendance was associated with higher performance, but even 10-14 year school children spontaneously solved the problem less than half the time. Overall, people seemed somewhat confused and even fearful of the apparatus.

As could be expected, we began by creating a logically equivalent task, but using locally familiar materials ––two distinctive match boxes, each with a distinctively colored key in it. One of the keys, when dropped in a wooden box, produced a suitable reward. These changes produced the expected results. When the local materials were used, performance shot up, even for the younger children who had not been to school.

Seeking to pinpoint the part of the process that seemed to be disrupted by the unfamiliar apparatus, we conducted another series of studies, this time including American children of different ages. We focused on identifying where in the inference process the unfamiliar apparatus disrupted the thought process. The details of the study involved Rube Goldberg arrangements where keys from matchboxes were dropped in the fearsome metal box and other maneuvers that fit the research conventions of the time. The long and the short of the outcome was that if the first part of the task was carried out with familiar materials such as match boxes and keys, even if the keys had then to be dropped into the fearsome metal box to get a reward, the problem was solved with relative ease. But if the problem began with the metal box, performance was depressed to the level observed with that initial, “standard” apparatus alone.

A bonus in this work is that by bringing our new research question about familiarity of the apparatus and assessment of different forms of learning and reasoning into the American developmental arena, a genuinely new finding was obtained. When presented with the match box/keys/drop the key in the box version of the task, American third graders now solved the problem (Cole, Gay, & Glick, 1971). The cultural context of learning and thinking an exploration in experimental anthropology.

The implications of this finding took several years to percolate into the American psychological literature. There the methodology in our cross-cultural developmental research coincided with the cascade of research showing that young children's performance on a whole range of tasks used as benchmarks of cognitive development tasks produced the effects.

An Initial Summing up of the Evidence, Circa 1971

When it came time to summarize our results, we were hard pressed to formulate an overall explanation of the results. We had succeed in several cases to combine ethnographic evidence about common local cultural practices with experimental- psychological analysis. In many such cases, it proved possible to demonstrate that local people were capable of engaging in various cognitive activities just fine so long as the tasks were, in various senses, familiar. (Later, Donaldson would coin the term tasks that “make human sense” to capture the idea here). In other cases, it appeared that the kinds of tasks drawn from schooling activities posed difficult challenges to people who had not encountered school, even when all the measures we could think of could not change the basic pattern of results. With so many missing parts to the puzzle and being so far from fitting neatly into a single, coherent picture, we confined ourselves to concluding:
Cultural differences in cognition reside more in the situations to which particular cognitive processes are applied than in the existence of a process in one cultural group and its absence in another. (Cole, Gay, Glick, and Sharp, 1971, p. 233)
Note how neatly hedged this summary is. We did not deny the possibility of specific cognitive differences between cultural groups arising from differences in specifiable experiences. After all, we were continuing to argue that the effects of culture on development should be sought through analysis of the kinds of activities that people engage with and are expected to be competent at accomplishing. Schooling is an everyday activity in the lives of children where it is present and they attend even for a few years. But schooling, on the basis of our research, did not make a general cognitive difference. Rather, it made a difference with respect to those intellectual processes that schooling practices engage.

All of this may seem sufficiently banal to make it a curiosity, but what we claimed about schooling we thought to be a general principle so long as the research was careful to include both the activities of schooling and those of making local medicines. This proviso meant that using the scientific means at our disposal we could not legitimately make generalizations about “culture and memory ability” or “culture and the development of categorization”. We could, however, make warrantable inferences about the specific kinds of activities we studied. The origins of a “cognitive ability” by this view are tightly connected to particular activities or contexts of use.

This work, taken as a whole, cast doubts not only about generalizations concerning culture and cognitive development but also about the conclusions concerning processes of cognitive development that dominated the professional literature in the industrialized world. Conclusions about age-related performance differences in the U.S. based upon experimental tasks such as those illustrated above no less than comparative studies in Liberia appeared to be badly compromised by hidden biases that crept into their design.

These doubts, transposed into the arena of American academia in 1971, provided the initial social impetus to create LCHC.

comment.JPGVisit the Chapter 1 Discussion Page

Chapter 1 References and Additional Resources
More Pictures for Chapter 1

If you have additional changes or suggestions for the wiki content please send to

Next →
Continue reading with Chapter 2.
Or, move ahead to Section 2.