Formal Operational Thought and the Moroccan Adolescent

Douglas A. Davis1
Haverford College

In this paper I report some of the results of an investigation of problem-solving and cognitive development among adolescents in a Moroccan town. The cognitive data were collected as part of a comprehensive cross-cultural study of adolescence in a variety of national settings. The Moroccan portion of this study is the subject of a book by Susan Schaefer Davis and me. Since the larger study involved ethnographic participant-observation as well as structured interviewing and cognitive testing, I will discuss the results of the more narrowly cognitive testing in relation to the cultural, community, and personal context in which these young Moroccans performed the tasks given. My intention is to combine presentation of a particular contribution to the cross-cultural literature on cognitive development with a more general discussion of the problems which have beset the search for universal factors in the development of psychological abilities. The result will be neither a conventional presentation of empirical research on a particular aspect of development nor a traditional ethnographic treatment of a non-Western community, but a combination of both. I first briefly discuss the cross-cultural psychological literature on the development of formal-operational and post-conventional reasoning. After describing the Moroccan field setting in which the ethnographic observation and systematic testing were done, I present the results of tests of problem-solving and block design administered to a sample of Moroccan young people, showing the effects on their performance of sex, age, and years of formal schooling. Finally I discuss and elaborate on these finding by drawing on semi-structured interviews with some of the subjects in the larger study, using the latter discussion to raise general questions about the adequacy of existing cross-cultural models of development.

Culture, Cognition, and Adolescence

The major cognitive change alleged to occur during the adolescent years involves the emergence of what Piaget has called "Formal Operations" (Piaget, 1972; Inhelder & Piaget, 1958). Formal Operations are reflected in increased ability to reason hypothetically independently of concrete situations, and to describe one's own reasoning processes. In the samples studied by Inhelder and Piaget, Formal Operations are said to become established between the ages of roughly 11 and 15. Piaget notes, however, that the rate of progression through the developmental stages described appears to vary from culture to culture (Inhelder & Piaget, 1958, pp 337-347). Piaget also acknowledges that while Formal Operational reasoning is in principle independent of the content to which it is applied, empirical studies have indicated individual and subcultural differences in the contexts in which such reasoning is displayed. It was thus reasonable to assume that the wide range of socio-cultural settings in which the data for the Harvard Adolescence Project were collected would show divergences from the bulk of the US and European developmental literature in the levels and rate of achievement of various cognitive performances.

The Harvard Adolescence Project

The research on which our account of "Zawiya," the Moroccan town in which the data reported below were collected, is based was conducted as part of a cross-cultural and interdisciplinary study of adolescence in seven cultural settings under the auspices of the Department of Anthropology at Harvard University. The Harvard Adolescence Project was designed by Beatrice and John Whiting and Irven DeVore and co-supervised by a group of Harvard scholars from the fields of Cultural and Biological Anthropology, Cognitive and Personality Psychology, and Education. This project continued a long involvement of the Whitings and their students in cross-cultural studies, constituting what has been called the "Field Guide" approach (B. Whiting, 1963; Campbell & LeVine, 1970). Details of the Moroccan field setting are reported below. The general goals of the Adolescence Project were to sample cultures representing widely varied levels of complexity, ecological setting, and ethnic/religion background and to study the related physical, psychological, and social maturation of a group of young people between puberty and adulthood in each of these settings. While the studies undertaken were methodologically complex, and while each has evolved partially independently of the others, certain theoretical questions as well as a set of core methods were shared. These included the definition and testing of a sub-sample of "adolescents" from several of the settings on two cognitive measures: a test of logical problem-solving through the elimination of alternatives developed by Edith Neimark (Neimark & Lewis, 1967); and the standard Kohs block-design sub-test of the WISC. Each of these was selected because they had been shown to be strongly related to both maturation and intelligence in the Western literature, and because each seemed relatively free of cultural content. The settings studied include a Copper Eskimo settlement in the Canadian Arctic, an Aboriginal settlement in northern Australia, a Thai fishing village, two Ijo-speaking towns in southern Nigeria, a Kikuyu community in Kenya, a Romanian farming town, and the town we call "Zawiya" in central Morocco. The general goal of the cognitive-testing portion of the Harvard Adolescence Project was well articulated by Jerome Kagan, one of the consultant Harvard faculty, in a meeting with the researchers before their departure to the various field settings. Kagan noted that the general abilities subsumed under the label "Formal Operations" are probably necessary for adequate functioning as an adult member of every modern nation-state. Since, Kagan observed, clear evidence of widespread establishment of formal-operational levels of thinking has yet to be found for any non-Western setting, the literature to date must say more about the adequacy of our methods than of the reasoning of non-Western peoples. I will briefly review the cross-cultural literature before presenting our own findings.

Keating (1980), discussing critically the literature concerning cognitive development in adolescence, notes deep and quite pervasive disagreement about both the extent to which Piagetian "formal operational" thought has been demonstrated and the extent to which it can be demonstrated given the complexity and the inadequate specification of the alleged components of reasoning. Commenting on this literature, Kagan (1972, p. 90) notes that "the Western mind is friendly toward the construction of discrete, abstract categories, each with its special set of defining characteristics." This "prejudice" Kagan contrasts sharply with the preference of the "classical Chinese" for viewing nature and experience as part of a "contained and continuous whole." The latter position is probably closer to that of non-Western peoples generally. As Kagan suggests, the Psychology of the West conceives of discrete developmental stages as natural, and it purports to discover evidence of stage-specific behaviors. The most influential twentieth century theories of personality development (Freud, 1905; Erikson, 1950), of cognitive development (Piaget, 1932, 1972), and of moral reasoning (Kohlberg, 1972) all treat psychological functioning as stage-specific. The consequences of such assumptions for the psychological study of persons have recently become the subject of important criticisms of each of these literatures (cf. Gilligan, 1982).

Some of the developmental literature has specifically addressed the concept of a "releasing mechanism" for the next stage of cognitive development. In a study using the Embedded Figures Test (Witkin et al., 1962) with a population of rural and urban Moroccans, Wagner (1978) attempted mathematically to partial out the effect of schooling on the EFT and found that the predictiveness of both age and the rural-urban ecological variable largely disappeared. Education may function, Wagner suggests [does/would he?], as a releaser variable for a variety of cognitive performances.

Cross-Cultural Piagetian Research

In a general review of this literature, Dasen (1972) noted that the bulk of the cross-cultural Piagetian work has been "descriptive" rather than "experimental." Most studies have found some support for the qualitative stages described by Piaget, but cultural factors typically appear to affect the rate of development.

Much of the cross-cultural research on the development of Formal Operations has shown cultural differences in the levels of reasoning displayed by members of various cultures, with schooled Western samples consistently showing superior performance. Sex differences have also been demonstrated by much of the cross-cultural research, with male subjects tending to score higher on a variety of measures of Formal Operational reasoning. Douglas & Wong (1977), for example, found that a sample of American 13 and 15-year-olds scored higher on three Piagetian formal-operational tasks than did a sample of similarly- aged Chinese youth, and that males performed better than females. Chinese females had the lowest scores on each of the tasks.

Cross-cultural Research on Moral Reasoning

While the first data to be reported here were concerned with an abstract measure of formal operational reasoning rather than with social cognition, the possible relevance of this cognitive domain to social cognition was a concern of the Harvard Adolescence Project which undertook the research. The most widely cited cross-cultural literature on social cognition concerns the dimension of moral judgment as articulated by Kohlberg and his associates (Kohlberg, 1964; Kohlberg, 1971; Kohlberg & Gilligan, 1971). Kohlberg's work has important implications for developmental psychology generally, since the internalization of cultural values is a topic with cognitive-learning, personality, and neuro-psychological implications [viz.?]. Kohlberg's treatment of moral reasoning draws heavily on Piaget's stage theory of the development of genetic epistemology (Kohlberg, 1962, pp. 394-400; Piaget, 1932), and similar difficulties have attended its cross-cultural application.

In a current review of the literature on the cross-cultural universality of Kohlberg's stages of social-moral development, Snarey (1985, p. 202) finds "striking support" for many of Kohlberg's assumptions, and also great difficulty discovering "post-conventional" moral cognitions in non-Western settings. As Snarey notes, one key assumption made by Kohlbergian research has been that the full range of moral stages (pre-conventional, conventional, and post-conventional) be present in all cultures.

However, while the full range of stages has been demonstrated repeatedly for Westernized urban societies both in Europe and elsewhere, no study reviewed by Snarey found evidence of Stage 4/5 or Stage 5 moral reasoning in a non-urban, non-Western setting. Salili, Maehr, and Gillmore (1976) compared evaluations of described moral and achievement situations by Iranian children ranging in age from seven to 18 years. They found roughly similar overall stages in the evaluation of moral situations. Iranian subjects, however, seemed to expect more moral behavior from persons of higher described ability than did similarly-aged Americans, seemed more ready to view described competence as an inherent good, and seemed more ready to reward effort regardless of outcome of a described activity. Salili et al. suggest that cross-cultural variations in such factors as the relative importance of ascribed social status may be important constraints on subjects' behavior in moral-judgmental tasks.

Using a new manual calling for matching responses to criteria judgments, Nisan and Kohlberg (1982) compared the stage levels and sequence of village and urban Turkish subjects between the ages of 10 and 28 with earlier work on Western populations. They argue that the results support the claim for structural universality in moral judgment. Village subjects showed a slower rate of development than city subjects, used lower levels of reasoning on all dilemmas presented, and tended to stabilize at Kohlberg's Stage 3 (Conventional). Village subjects also tended (regardless of stage) to justify their decisions in terms of social norms, while older city subjects were more likely to relate their positions to principles of justice. Nisan and Kohlberg suggest that the "conventionality" of the village subjects' responses may stem from the continued presence of a strong moral consensus in the village setting, and they argue that "the social ecology of the small village does not seem to call for the broader, generalized system perspective which is the hallmark of stage 4" (1982, p. 875).

It was thus with the ambitious and mixed goals of discovering formal-operational thinking in the narrower sense and of relating it to more contextual thinking about social and personal dilemmas that we set out for Zawiya in January, 1982.

Neimark's Problem-solving Task

Neimark, who developed the problem-solving task used in our own research (Neimark, 1964), noted that subjects' calculated strategy score for performance on the eight-problem task correlated significantly with the rated quality of their verbal descriptions of the strategy used (Neimark, (1975b), as well as with mental age (Neimark, 1967). The task used involves matching a pattern concealed under a set of eight shutters arranged in a circle to one of a set of eight patterns of white and black dots (Neimark & Wagner, 1964; Neimark & Lewis, 1967). The optimal strategy for solving such a problem involves reducing uncertainty by one bit per move through selection of a shutter position at which half the still-eligible patterns have a black, and half a white, dot. [Insert Figure showing Neimark shutter board?] Each optimal move thus discovers one binary feature of the concealed pattern, and this has the effect of reducing uncertainty by half with each shutter-choice. The results to be reported below express test performance as an average "strategy score," expressed in the average content in bits of the subject's moves: 1.00 represents a perfect three-move solution to each of the eight problems in the test proper. A strategy score of 0.65 repreesents random performance, and mean performance among American subjects tested by Neimark increases from roughly 0.75 at age 12 to over 0.90 by age 17 (Neimark, 1967, p. 112).

Zawiya

All four of the field sites in which roughly comparable data to those to be presented here (Romania, Morocco, Canadian Eskimo, and Thailand) were collected are non-urban traditional communities, and three are also non-Western. In two of these settings (Romania, Morocco) a sample of adolescents were in addition to the Neimark testing asked a set of questions concerned with social cognition and moral values. These data (for the Moroccan case) will be briefly discussed below. More information about the domain of moral reasoning in these settings is provided in the monographs concerned with adolescence in each of the field sites making up the Harvard Adolescence Project. [2]

The community in which the ethnographic data presented below were collected has been described in detail by Susan S. Davis (1983). It is a large, traditional Muslim, semi-rural community of about 11000 citizens (1982 census) in north central Morocco only a few miles from a small city (which we call "Kabar") built during French colonialism. Susan Davis and I returned to Zawiya during ten months of 1982 as one of the field teams of the Harvard Adolescence Project. We had lived in the same neighborhood a decade before. The focus of our ethnographic interviewing, testing, and observation was roughly 50 families living in one neighborhood of Zawiya, and in particular their 150 children between the ages of nine and 21 at the time of the study. During eleven months of fieldwork in 1982 we lived in the neighborhood and took part in many of its public and semi-private activities such as weddings and circumcisions. We collected physical growth measurements and background family, educational, and travel data on over 100 adolescents. We also interviewed two dozen teenagers and young adults at length about their experiences and feelings toward family, friends, and self. These interviews typically included both the "social cognition" questions reported below and the presentation and solicitation of "moral dilemmas." Neimark's problem-solving task was administered during the summer vacation of 1982 to a sub-sample of 65 adolescents, and the Kohs block test was given to 46 of these in December of that year.

Cognitive Testing

Testing Situation. The Zawiya sample for Neimark's problem-solving task included 65 young people, 31 males and 34 females. Testing was carried out late in the 1982 Summer school vacation. Test administration was by me or by Susan Schaefer Davis, assisted by Hamid Elasri, a male school-teacher from the neighborhood.[3] All explanation and prompting was delivered in colloquial Arabic, and all testing was in a semi-private room in the researchers' house. Since the flavor of this testing may be important interpretation of the results, I quote briefly from my fieldnotes concerning the testing experience:

[26-Aug-82] Formal testing. After two days and about eight runs, the special vocabulary for the Neimark is getting clear, for example the spots are nuktat, ghta is "cover," and the basic idea of the "game" was expressed by Hafidas "You're going to try in order to open only two or three [shutters]. I had to explain to Hafid afterward that in fact three's the best possible with eight patterns. We have just tested two sons in the same family, and the older made no errors at all and could give a clear explanation of the rationale (i.e. that he should eliminate half the remaining possibilities each time), while his 13 year old brother was essentially random even by the end and after a great deal of prompting by my assistant Hafid. The older brother is a strong student in 5th year science, while the younger has flunked three years, so age and intelligence are probably confounded.

[28-Aug-82] With most of the sample males between 11 and 16 tested, it appears the age at which most have the ability on which success at the Neimark depends is much older than in the US. At least half the subjects never really get the idea that there is a correspondence between the picture covered by the shutters and the problem alternatives, such that one can eliminate possibilities steadily. We have had to remind several at each move to cover the dis-similar and not the similar dots, and roughly a third of subjects make mistakes until the end on which alternatives a shutter move entitles them to cover. We also are encountering a whole range of third-world distractions: older and younger sibling tagging along, previous Ss trying to kibbitz, wedding parties passing the house with loud ghiata music, our daughter's pre-school group reciting Arabic or playing in the courtyard, a goat wandering into the house.

Correlational results: Neimark Task

Despite these difficulties, which gave us in a few days a profound respect for the complications of cross-cultural testing, we were able to discover significant correlates of performance on this and the the block-design test. While we took pains to test roughly equal numbers of male and female adolescents at about the same ages, we did not attempt to match numbers of years of schooling for the two groups. Summary data for the sample are given in Table 1.

Table 1

                Cognitive Test Results for Zawiya Adolescents
                  _______________________________________
                                            Mean
 
                         N            Neimark     Age    Schooling
 
                 Males   31             0.686     15.5       5.3 
                 Females  34            0.610     15.4       4.0
                  ______________________________________
 
 

The average age of males tested was 15.5 (range 12.5-20.5 years), and that of females 15.4 years (range 11.3-21 years). The 31 males who completed the Neimark test had an average of 5.3 years of schooling successfully completed (range 3-9 years), while the 34 females had an average of 4.0 years (range 0-10 years). Thus, while the mean Neimark test score for females (0.610) is significantly (p < .01) lower than that for males (0.686), this is almost certainly due at least in part to the significantly lower average schooling for the females. Girls in Zawiya are still only about half as likely as boys to continue beyond elementary school, although roughly equal proportions begin the five-year elementary school program.[4] It should also be noted that the Moroccan school system is very competitive, with a very high failure rate (almost one year for every two attended, in the Zawiya sample) and a difficult standardized examination for passage to secondary school.

As expected, the variables of age, completed years of schooling, and performance on Neimark's problem-solving task are corelated, and the pattern of correlations is quite different for males and females (see Table 2). The most striking difference is that age and years of schooling are not significantly correlated for females tested. Secondary schools have been built close to Zawiya in the past few years, and increasing proportions of girls are continuing beyond the first few years of elementary school. As a consequence, the group of older teenage girls tested in 1982 were not on the average more "educated" than their younger teenage neighbors and sisters. The significant correlation of years of schooling with Neimark performance indicated that the small group who were more schooled were indeed better able to complete the task than their school-dropout agemates. For males age and schooling are strongly correlated, and it is therefore not surprising that both age and schooling show significant positive relationship to Neimark performance.

Table 2 

          Intercorrelations of Neimark with Age and Schooling
          ____________________________________________________
 
                       Age   Schooling   Strategy
          Males (N = 31)
          ____________________________________________________
 
          Age           __     .74***    .37*
          Schooling             __       .49**
          Strategy Score                  __
          ____________________________________________________ 
 
          Females (N = 34)
          ____________________________________________________
 
          Age           __     .14       .45**
          Schooling             __       .21
          Strategy Score                  __
          ____________________________________________________
          *p < .05.  **p < .01  ***p < .001
 
 

When the strategy score is treated as a simultaneous function of sex, age, and schooling using multiple regression analysis, the effect of age is statistically significant (p = .01), that of years of schooling is marginally so (p = .10), and that of sex is still significant (p = .05). That is, older, more schooled, and male subjects tended to do better on the test when the other effects were statistically controlled. These results are consistent with those of other researchers in that performance on any measure of Piagetian formal operations is expected to increase sharply in the age range sampled (Hollos, Leis, & Turiel, 1986; Neimark & Lewis, 1968). Formal schooling appears also to be a key contributor to test performance. The poorer performance of girls tested is, as suggested above, probably due primarily to the fact that fewer of them (15%, compared to 39% of males) had reached the secondary school level at which formal reasoning is taught, although it is also possible they were more anxious at being questioned by males. Even among male subjects, however, the average level of performance is not much above chance. I kept detailed notes on the way each of those tested approached the task, and I was in several cases startled to note that a boy who was a skilled card or checkers player (both skills tapping similar problem-solving abilities) was utterly confused by the Neimark test.

Correlational results: Kohs Block Task

Table 3

          Intercorrelations of Kohs Blocks with Age and Schooling
          ______________________________________________________
                        Age    Schooling    Kohs 
          Males (N = 22)
          ______________________________________________________
          Age           __        .75***    .40*
          Schooling                __       .48**
          Kohs Score                         __
          ______________________________________________________ 
          Females (N = 24)
          ______________________________________________________
          Age          __         .24       .30
          Schooling                __       .34*
          Kohs Score                         __
          ______________________________________________________
          *p < .05.  **p < .01  ***p < .001
          _______________________________________________________

For the block design task the results show even more clearly the critical role of formal schooling, despite the smaller sample size. The average score for 24 females tested (18.6) was not significantly different from that of the 22 males (20.0). While age and years of successfully completed schooling were both moderately correlated with block design performance, when both variables were included in the regression analysis only schooling contributed significantly (p = .05). The deletion of two females with no schooling increased the strength of this finding (p < .01).

These results appear to confirm the strong role of structured schooling in establishing even modest levels of performance on standard measures of cognitive performance. To better understand the possible applications of cognitive capacities to real (or at least moderately realistic) situations we relied on semi-structured interviews with a subset of neighborhood youth. This attempt involved both asking for moral dilemmas and/or presenting a modified version of the "Heinz" dilemma used by Kohlberg, and the administration of a set of questions requiring logical thought or generalization from a concrete situation. The latter set of questions were modeled closely on a set used by Mitchell Ratner in his research in a Romanian town as part of the Harvard Adolescence Project. Our own set of questions were the following:

Social Cognition Questionnaire

A. HYPOTHETICALS

  1. What would happen if a sheep didn't have feet?
  2. What would happen if only boys were born in Zawiya?
  3. What would happen if the buses and trains no longer ran?
  4. Now can you make a question that begins, "What would happen if?"?

    B. EXHAUSTION OF POSSIBILITIES

  5. A family has a goat. One night the goat didn't come home. Can you explain why this might happen? Might there be another explanation? Another? (get 3)
  6. There once was a farmer who liked to plow, always with two different mules. That is, one day he would use the brown and the black, and another day he would use the brown and the white, and so on. If the man had 3 mules, how many days could he plow without using the exact same pair twice? What if he had 4 mules? And if he had 5 mules?

    C. COUNTERFACTUALS

  7. People began to use cars (tumubilat) in 1900. Sultan Moulay Ismail lived in the 17th century (1600's). Is it possible that in a Meknes museum where they keep old things so that we [now] may see them there is a car that Moulay Ismail used?
  8. One person builds his house of brick (yajur), and another builds his house of stone. Is it possible for a house of brick to be heavier than a house of stone? (Can press with 20 rooms brick vs. 3 stone) [If no: What if the brick house were much larger?]
  9. Three farmers met at the market. Each had a sheep to sell. Farmer A. looked at the sheep of B., and he thought to himself, "My sheep is larger than B's, I'll get a higher price." Later Farmer C. looked at the sheep of A. and thought, "My sheep is larger than A's, I'll get a higher price." Then Farmer B. looked at the sheep of Farmer C., and he thought to himself, "My sheep is larger, I'll get a higher price." What do you think? Is it possible for all the farmers to be right in what they said?
  10. a) Let's say there is a village called "Duwwar Zwin" and in Duwwar Zwin there are only black cows. On the way to Duwwar Zwin you met a woman from "Duwwar Zwin" and she said she had a cow at home. Would you know what color her cow was? What color?
    b) On another day, you met on the road a woman from a neighboring village, called "Duwwar Xayb," and she told you she has a cow at home. Would you know what color her cow was? What color? Why?
    c) If you happened to meet a black cow on the road, would you know for sure what village it was from?
    d) And if you happened to meet a brown (one not given above) cow on the road, would you know for sure what village it was from? What village?
    e) If the answer is "no": would you know for sure what village the cow could not be from?

    D. SOCIAL COUNTERFACTUALS

  11. Is it possible for a schoolteacher to teach something that isn't true?
  12. Is it possible for a father to tell a lie to his child?
  13. Is it possible for a policeman to steal?
  14. Is it possible for a wife to be very dear to her husband and [yet] he beats her?
  15. [A question used by Ratner concerning whether a person could be both a Christian and a Communist was skipped.]
  16. Is it possible for there to be mother whose children are not dear to her?
  17. Is it possible there is a religious scholar who studies the Quran and is married, and [yet] he goes out with [other] women?
  18. Is it possible for a judge (qadi) to put an innocent man in jail?

Nine males and 13 females between the ages of 11 and 21 were asked these questions late in the fieldwork year as part of a general round of interviewing concerning their attitudes and experiences. As a rough way of capturing the expected increase in sophistication during this age range I simply noted the number of questions with a clear right answer (e.g., "Yes," to "Is it possible for a house of brick to be heavier than a house of stone?") each subject responded to correctly. For this group of subjects the number of correct responses to the social thought questions is significantly correlated with years of successfully-completed schooling (r = .45, p < .025). This effect appears to have produced solely by the male respondents, however: when the data are dichotomized by sex, both age and school effects are significant for males while neither is for females. The correlation matrices for these social thought questions are presented in Table 4.

Table 4

       Intercorrelations of Social Thought with Age and Schooling
       ______________________________________________________
 
                 Age   Schooling   N-Correct  N-Yes
       Males (N = 9)
       ______________________________________________________
       Age        --       .47       .61*      .94*
       Schooling            --       .85**     .42
       Number Correct                 --       .64*
       Number Affirmed                          --
       ______________________________________________________
 
       Females (N = 13)
       ______________________________________________________
       Age        __        .26      .19      -.35
       Schooling             __      .25       .11
       Number Correct                 __       .02
       Number Affirmed                          --
       ______________________________________________________
       *p < .025.  **p < .001

It seems plausible that this effect too is a consequence of the lower persistence of females into secondary school, and indeed schooled subjects mentioned classroom parallels to some of the questions. The manner of responding to these questions showed the effects of schooling in a striking way for several of the male subjects. For example, three boys in secondary school puzzled over the mule-combination question, then asked for paper and pen and drew Cartesian products, explaining that they had learned this in math!

For the last set of questions, concerned with whether various role models may behave inconsistently with their roles, I have simply noted the number of times each subject responded "yes." This number increases with age and schooling, but the correlations are significant only for male subjects, among whom older youth gave more affirmative answers.

Individual subjects' produced a number of responses which are of special interest in the context of social reasoning in Zawiya (and perhaps in other traditional communities). Younger children tended to answer each "social counterfactual" question in the negative: a teacher couldn't teach falsehood, a policeman can't steal. Older adolescents not only realized that such generalization could easily be false in a particular instance, but they also frequently couched their answers in terms of a realistic (if rather cynical) awareness of the tenuousness of moral behavior. Thus 19-year-old Abdelaziz pointed out when asked whether a teacher could teach something that wasn't true that the teacher might teach something he himself believed (giving fortune-telling as an example) but which was not really so. He illustrated this principle with a quotation from the Koran condemning fortune-telling. Indeed, older adolescents often responded to the questions about whether authority figures could violate their trust with variations on, "Are you kidding?" When 18-year-old Kabiri was asked whether a judge could jail an innocent man he said, "Yes. Think how many innocent men are in jail." Several younger boys and girls, however, gave a simple "no" to all seven of these social counterfactual questions. Only one person, a girl of 12.5, thought it was possible for a woman's children not to be "dear" to her, and she said this might be because her husband hit her because of their behavior.

Western Moral Methodology, Zawiya Reality

These, then, are some illustrative results of the attempt to evoke adolescent cognitive judgments in a non-Western traditional setting. As important as any inferences about the validity of Western theories of development to which these findings might be applied, however, is the lesson concerning the methodological requirements of such research. Such work will, I believe, necessarily require tolerance for variations in testing procedures: samples which are small and non-random, and a mixture of numerical analysis and single-case explication which offends the clear distinctions most of us learned in graduate school. The fundamental tension I am describing is between rigorous and replicable methods involving grouped data, and a clinical/ethnographic style of selecting and presenting case histories. Accordingly I have moved in this account, as in the book manuscript from which these results are partly extracted (Davis & Davis, forthcoming), from an account of a regression analysis of cognitive test data for a sample of 65 Zawiya subjects (themselves a sub-sample of the 150 neighborhood sample) to a more "clinical" attempt to understand the reasoning of several members of that first sample.

Moral Reasoning in Zawiya

We also asked our clinical informants about moral dilemmas, both ones posed by us and ones growing from their own recent experience. The results are impossible to summarize adequately here, but several of aspects of social self-perception were clearly illustrated by the responses to these moral questions.

We posed the problem Kohlbergians have often used (cf. Kohlberg, 1973; Gilligan, 1982), the dilemma of a man whose wife is dying of a disease for which the only medicine is controlled by a druggist who demands more money for it than the man can raise. We did not, as has been usual in Kohlberg's research, ask directly whether the man would be justified in stealing the medicine, but rather asked our adolescent informants to suggest what he might do. It was striking to us that Zawiya adolescents typically did not mention the possibility of the man's stealing the drug until all other options had been exhausted. Most often it was suggested that the man could surely reason with the druggist and get him to take payment over time, or that relatives would lend the necessary money.

We also asked each adolescent to suggest a moral dilemma of their own, and the responses gave us some insight into personal styles and preoccupations. Several secondary school male students responded by recalling instances in which a friend wanted to copy their work in school. These examples were discussed, not in terms of universalizing conceptions of the general good ("What would happen if everyone cheated?"), but rather in terms of a realistic assessment of the relative interpersonal cost of disappointing a friend or of being punished by a teacher. Abdelaziz, when asked to make up a dilemma said:

You might know somebody who's fleeing the police and comes to your house. If you tell the police they'll send him to prison for 12 years, and people will say you did a bad thing; but if you don't tell the police might punish you, perhaps with three years in prison.

In this "prisoner's dilemma" imagined moral question, as in his real examples concerning school, Abdelaziz casts the issue in terms of whether to follow a societal rule and both hurt a friend and suffer censure from the community, or to break the "law" and be severely punished. Such responses do not earn Abdelaziz membership in the highest Kohlbergian stages, but they reflect vividly the actual content of Zawiya ethics.

Formal thinking, informal thinking

Here, for example, is `Abdelkhalq , a boy of 13 at the time we talked. `Abdelkhalq has six years attendance at school behind him, three of which he passed. I had first asked him for a personal example from the previous year in which he had not known what was the right thing to do, and he could think of no example. I then posed the modified "Heinz" dilemma, and got the following answer (transcribed from notes taken during the interview):

He should give him (the druggist) half, and then work to get the rest of the money. [And if he doesn't have the money?] Borrow. [And if he's tried everything and can get only half the money, should he steal the medicine or leave his wife without?] He should steal. [Why?] So his wife will live. [But is the druggist also in the right?]

This last question seems hard for `Abdelkhalq to understand. He struggles with it and finally says of the druggist, "Yes, he's right too." My next question is concerning "moral behavior" in school:

[Do kids sometimes copy in class?] Yes, lots. [What does the teacher do?] Well, our French teacher is very tough , and she hits the kids with a hose. [But do kids say it's OK to copy?] If it's their friend. Otherwise they get angry. Older kids even pay to copy.

The immediately following question is one I have puzzled about for years in relation to Morocco:

[If you see a boy who's nice and quiet at home but hits others with stones in the street, why do you think that is? `Abdelkhalq repeats my question, and I ask "Any idea why?"] No. [But say, why is he nice at home?] So his father will say, "He's nice." Then if he hits a kid in the street his father will say, "Mine's a good boy, he doesn't hit."

What are we to make of this as moral self-reflection? There are of course two very different readings for the (hypothetical?) boy's motives in behaving with careful correctness in front of his father. It seems obvious that we could take this to be evidence of a cynical or of a naively "concrete" and "conventional" way of behaving in each situation in a manner that minimizes personal loss and wins approval of parents or peers. On the other hand `Abdelkhalq's motives, while not ignoring the crass advantages of such behavior in the not-unlikely event his father has to defend him in public (or at least order his older brothers to do so), may be more complex. He may indeed begin to believe that he is what he shows himself to be in front of Dad, that this is the social role he will grow into and make his own. We can hardly judge on the basis of this small behavbior sample, but I have known `Abdelkhalq's older brothers for years and I think they have undergone such a maturational process.

`Abdelkhalq had helped six months previously by locating and bringing to our house for testing the younger boys and girls included in the cognitive sample. I asked him what he had done with the $50.00 I had paid him then. He told me he still had it , that he wanted new trousers and shoes but they were expensive in the Fall. Do we credit him with foresight and self-restraint in this instance, or assume his parents or oldest brother have impounded the money? I don't know the literal answer, but it seemed to me he was expressing a personal belief. More importantly, his answers makes sense in the context of Zawiya, where boys need street toughness and at least the appearance of good behavior at home to make their way in an uncertain world where too few resources[5] are claimed by too many young people.

Conclusion

We began interviewing and testing in Zawiya with good reason to believe that the youth there would not perform as Western subjects have, both because of limitations of the tests developed for US and European use and because of lower levels of schooling in Zawiya. We do not believe the low average scores of persons tested in Zawiya or other non-Western settings can be used to infer differences in native ability. From long association with the adults of Zawiya, I can assure readers that they reason as subtly about their social world as do Americans about theirs. They do not, however, typically express this social intelligence in easily-recognizable expressions of the formal didactic or syllogistic thinking Piagetians have described; and this "failure" seems to me primarily due to the low level of Western-style formal schooling in Zawiya. These results are thus consistent with Scribner's (1976, pp. 5-6) observation concerning the critical effect of schoolong on the emergence (or at least the dsplay) of Piagetian formal operations:

In all cultures, populations designated as "traditinal" or "nonliterate" have just somewhat better than a chance solution rate across all types of problem material.... Within each culture there is a large discrepancy in performance between schooled and non-schooled. With schooling, there is little between-culture variation in performance for the cultures studied.

The teen years in Zawiya are clearly a time of changed self-perception, and indeed of the formation of a newly sophisticated sense of oneself as a social actor. The successful older adolescent will be aware how s/he comes across to others, will have a clear conception of how one's own behavior is judged by the context in which it occurs and by the company one keeps, and will know how to enlist the support of neighbors and to avoid conflict. While these skills may not translate into skilled performance on standardized tests, they are indicative of a well-honed intelligence which allows the Moroccan adult to see a wide range of possible motives and consequences of one's own and others' actions, and to take moral considerations into account in a practical way (cf. Rosen, 1984). Despite their often impoverished circumstances, the youth of Zawiya aspire to a much more affluent life, perhaps one taking them far beyond the scope of their semi-rural community and family ties. While they view many of the traditional practices of their parents with ambivalence, they remain on the whole committed young Muslims and proud Moroccans. Indeed they fit very well a picture of adolescence as a time of tension between rampant optimism and sometimes self-contradictory thinking about how to achieve goals. What they fail to show in formal reasoning in structured tests, or in the class-room, they can often demonstrate with great subtlety in their relations with peers and family. Living with them helps us appreciate that there is much still to learn about the social imbeddedness of reasoning. As Inhelder and Piaget have observed:

[T]here is more to thinking than logic. Our problem now is to see whether logical transformations fit the general modifications of thinking which are generally agreed--sometimes explicitly but often implicitly--to typify adolescence (Inhelder & Piaget, 1958, p. 335).

Like other cross-cultural studies of the development of formal thinkings, these Moroccan results suggest only a partial and attenuated movement through the stages delineated by Piaget, Kohlberg, and other Western theorists. It does seem fair to say that when given material of more obvious relevance to their own goals and problems Zawiya youth often respond with a subtlety one would not predict from their performance on imported structured tests. I believe some of the lessons to be learned from such work are obvious: let us understand the goals and dilemmas actually occuring in non-Western settings and build some of our measures aound them (cf. Greenfield, 1976); and let us, as Valsiner has suggested (1984), learn to see the person-setting interaction as the necessary focus of our investigations.

References

Adelson, J. (Ed.). (1980). Handbook of adolescent psychology. New York: Wiley-Interscience.

Campbell, D.T., & LeVine, R.A. (1970). Field-manual anthropology. In R. Naroll & R. Cohen (Eds.) A handbook of method in cultural anthropology. New York: Columbia University Press, 366-387.

Cole, M., Gay, J., Glick, J.A., & Sharp, D.W. (1971). The cultural context of learning and thinking: An exploration in experimental anthropology. New York: Basic Books.

Dasen, P.R. (1972). Cross-cultural Piagetian research: A summary. Journal of Cross-Cultural Psychology, 3, 23-40.

Dasen, P.R. (1975). Concrete operational development in Canadian Eskimos. International Journal of Psychology, 10, 165-180.

Davis, S.S. (1982). Patience and power: Women's lives in a Moroccan village. Cambridge, MA: Schenkman.

Davis, S.S., & Davis, D.A. (1989). Adolescence in a Moroccan town: Making social sense. New Brunswick: Rutgers University Press.

Douglas, J.D., & Wong, A.C. (1977). Formal operations: Age and sex differences in Chinese and American children. Child Development, 48, 689-692.

Hallpike, C.R. (1979). The foundations of logical thought. New York: Oxford University Press.

Hollos, M, Leis, P.E., & Turiel, E. (1986). Social reasoning in Ijo children and adolescents in Nigerian communities. Journal of Cross-Cultural Psychology, 17, 352-374.

Inhelder, B., & Piaget, J. (1958). The growth of logical thinking from childhood to adolescence. New York: Basic Books.

Kagan, J. (1972). A conception of early adolescence. In J. Kagan & R. Coles (Eds.). Twelve to sixteen: Early adolescence. New York: Norton.

Kagan, J., & Coles, R. (1972). (Eds.). Twelve to sixteen: Early adolescence. New York: Norton.

Keating, D.P. (1980). Thinking processes in adolescence. In J. Adelson (Ed.). Handbook of adolescent psychology. New York: Wiley-Interscience.

Kohlberg, L. (1964). Development of moral character and moral ideology. In M.L. Hoffman and L.W. Hoffman (Eds.). Review of child development research (Vol. 1). New York: Russell Sage Foundation.

Kohlberg, L. (1971). From is to ought: how to commit the naturalistic fallacy and get away with it in the study of moral development. In T. Mischel (Ed.). Cognitive development and epistomology. New York: Academic Press.

Kohlberg, L. (1973). Continuities and discontinuities in childhood and adult moral development revisited. In L. Kohlberg, Collected papers on moral development and moral education. Moral Education Research Foundation, Harvard University.

Kohlberg, L., & Gilligan, C. (1971). The adolescent as a philosopher: The discovery of the self in a post-conventional world. Daedelus, 100, 1051-1086.

Lancy, D.F. (1983). Cross-cultural studies in cognition and mathematics. New York: Academic Press.

Neimark, E.D. (1964). Information gathering in diagnostic problem-solving as a function of number of alternative solutions. Psychonomic Science, 1, 329-330.

Neimark, E.D. (1967). The development of logical problem-solving strategies. Child Development, 38, 107-117.

Neimark, E.D. (1975).Longitudinal development of formal operations thought. Genetic Psychology Monographs, 91, 171-225.

Neimark, E.D. (1975). Intellectual development during adolescence. In F.D. Horowitz (Ed.). Review of Child Development Research (Vol. 4). Chicago: University of Chicago Press.

Neimark, E.D., & Lewis, N. (1968). Development of logical problem-solving: A one-year retest. Child Development, 39, 527-536.

Neimark, E.D., & Wagner, H. (1964). Information gathering in diagnostic problem solving as a function of number of alternative solutions. Psychonomic Science, 1, 329-330.

Neimark, E.D., & Lewis, N. (1967). The development of logical problem-solving strategies. Child Development, 38, 107-113.

Nisan, M., & Kohlberg, L. (1982). Universality and variation in moral judgment: A longitudinal and cross-sectional study in Turkey. Child Development, 53, 865-876.

Piaget, J. (1948 [1932]). The moral judgment of the child. Glencoe, IL: The Free Press.

Piaget, J. (1972). Intellectual evolution from adolescence to adulthood. Human Development, 15, 1-12.

Rosen, L. (1984). Bargaining for reality: The construction of social relations in a Muslim Community. Chicago: The University of Chicago Proess.

Salili, F., Maehr, M., & Gillmore, G. (1976). Achievement and morality: A cross- cultural analysis of causal attribution and evaluation. Journal of Personality and Social Psychology, 33, 327-337.

Scribner, S. (1976). Modes of thinking and ways of speaking: Culture and logic reconsidered. unpublished manuscript.

Snarey, J.R. (1985). Cross-cultural universality of social-moral development: A critical review of Kohlbergian research. Psychological Bulletin, 97, 202-232.

Valsiner, J. (1984). 'Intelligence' as person-environment interaction in structured action contexts. Paper presented at the symposium Changing Conceptions of Intelligence and Intellectual Functioning, XXIII International Congress of Psychology, Acapulco, Mexico.

United Nations (1983). World statistics in brief: United Nations statistical pocketbook. New York: United Nations.

Wagner, D.A. (1978). The effects of formal schooling on cognitive style. Journal of Social Psychology, 106, 145-151.

Witkin, H.A., Dyk, R.B., Faterson, H.F., Goodenough, D.R., & Karp, S.A. (1962). Psychological differentiation. New York: Wiley.

________________________________________________________


[1]Douglas A. Davis is Professor of Psychology, Haverford College, Haverford, PA.
Copyright (c) Douglas A. Davis, 1987-1998. All rights reserved.

The comparative research on which this paper is based was made possible by grants from the National Institute of Mental Health (#MH 14088) and the William T. Grant Foundation. I am grateful to Susan S. Davis and Jaan Valsiner for comment on an earlier draft.

A revised and extended version of this paper appeared as:

Davis, D.A. (1987). Formal operational thought and the Moroccan adolescent. In J. Valsiner (Ed.) Cultural context and child development: Towards a culture-inclusive developmental psychology. Hofgrefe, 1988.

[2]Four volumes have been published as separate volumes in the series "Adolescents in a Changing World," Rutgers University Press.

[3]Test materials and a computer program for calculation of strategy scores were prepared by Mitchell Ratner and Ann-Mari Gemmill, who conducted the Romanian study in the Harvard Adolescence Project.

[4]Roughly one-third of the secondary-school age Zawiya children in the study neighborhood were still enrolled in 1982. Schooling data reported here are for the number of years passed by each individual.

[5]Morocco's per capita Gross Domestic Product was $716 in 1982, the year these data were collected (United Nations, 1983). Half the Moroccan population is under 18 years of age.