[note that this paper should not be read in its printed version as the paragraph order in that version was mangled]
In: Council on Anthropology and Education Quarterly, Vol. V, No. 2, May 1974
L'Image d'imagination n'est pas soumise à une vérification par la réalité. 
The following is an attempt to understand, from a cultural point of view, the structural and transformational interrelationships between processes of classification, segmentation, ranking, and segregation in a suburban high school. Some of the events I shall use as data would be classified by some sociologists as belonging to the domain of stratification, while others would be considered aspects of American mythology. I shall refuse to differentiate a priori between these so-called domains and shall start my analysis from the "cultural" ground up; from the informants' perception of their social world, the quasi-mythical "texts" through which they communicate with each other about each other. The eventual purpose of this research  is to provide a unified cultural theory of quality and inequality in the United States that will allow us to go beyond the 'traditional dichotomy with which native American sociologists studying their own society believe they are confronted—the dichotomy between the myth of democracy and the reality of social classes.
For the last two or three decades it has been widely accepted among social scientists that America is not a totally equalitarian society, but that it is stratified into a certain number of classes. At the same time, it became accepted also that this class system could be seen within the confines of schools, that adolescents arranged themselves following the way their parents were arranged in the organization of the community. But this "knowledge," by now almost common sensical, has not always been with us. Indeed, the first systematic studies of stratification in the United States at the town or school level, those of Warner and Hollingshead, were couched as "discoveries" of a previously unknown reality. Warner titled one of the initial chapters of the "Yankee City" series as "How the several classes were discovered'; (1951, ch. 5, 80ff.). Hollingshead wrote:
Chapters 1 and 2 of the book tell the story of the way the study began; . . . how we learned that Elmtown had a class system; how each family was located in this class system. (1949:1)
Few scholars seem to have been bothered by the fact that such a thing as a "class system" could remain "unknown" for so long. It had been discovered using the most modern methods, particularly the ethnographic anthropological method, and, of course, native perceptions were irrelevant. But are they really? Doesn't the anthropological method rely first and foremost on the informant's perception, their ability to name the groups which form their social structure and to specify the relationships which exist between them? And even when it can be shown that the natives have a double system, one contradicting the, other, as Americans could be said to have, the responsibility of the anthropologist remains to trace the relationship between the two systems. 
The dilemma of American studies is well known. On the one hand, the observer is confronted by informants who regularly rehearse statements to the effect that they believe that all men are equal, that none is inherently better than the others. They say they believe that their institutions reflect this belief and that America is the most complete embodiment of a certain religious and political ideal. On the other hand, the informants seem then to turn around in the same breath and savagely rank each other, deny each other opportunities, and segregate their private associations through any means, straightforward or insidious, which they can devise. These processes of stratification have been studied in great detail at the whole society level as well as at the school level, which will interest us here.
Since it had become accepted that schools were stratified, fewer studies were made of the actual process of stratification than of the process by which teachers produce objective differences in achievement through differential treatment and expectation. Furthermore, these are not random but are significantly correlated with the place of the parent within the status structure.
All this is well known and satisfactory as far as it goes. It is inadequate, however, because it rests on an uncriticized theory of stratification in the United States and, by extension, in its schools. Stratification is seen purely as the result of unequal access to goods and services, compounded by segregative tendencies on the part of those more or less at the same level of the scale against people at other levels. Furthermore, it is denied that stratification as it is empirically seen in a specific situation has anything to do with true ability. If there are objective differences, they are the result of the stratification process rather than what it is about. If teachers pretend that their evaluation of a child is based; on his performance in the classroom, it is retorted tot. them that they are blinded by their own prejudices, that it is they who create this very difference in performances If the teachers were not prejudiced, or if the system was not skewed, performance ranking would be spread randomly throughout all the non-natural-categories such as Black and white, rich and poor, which American society and culture have created because- in their "nature" all human beings are equivalent. Nobody seems particularly bothered by the fact that this way of seeing stratification is shot through by-individualistic—indeed, democratic—ideals- too sacred among sociologists to be criticized.
I am not myself going to criticize the notion of the perfect equivalence of human beings. It is very sacred for me, too. What I want to say is that we cannot really understand the empirical system of stratification seen in American institutions if we-do-not go beyond a simple derivation of the ideological premises which we carry in ourselves as natives. We must understand that stratification systems exist within a cultural context, that there is more to them than the inequality to which the formal definition would restrict them. We must study stratification systems as total social facts, including not only statistical evidence of differential-- treatment, but also the total environment, social and ideological, in which specific examples of segregative and stratifying behaviors are found.
The social characteristics of the town of Sheffield  in the suburbs of New York City are the result of: well-documented segregative patterns in housing. It is lily-white, Waspish, clean, sophisticated, expensive, and conservative: the proper domain of those who are often called the "upper middle-class," managers, professionals, or owners of small businesses, below the level of the true leaders of the capitalistic world.
Sheffield High School, whose catchment area is exactly co-terminous with this social area, is attended by the sons and daughters of these people. It is rather small-680 students—well supported, mild. It is free from serious tensions, economic or social. This is not to say that students, teachers, and administrators do not complain. Students are lobbying for an "open campus"; teachers do not always feel as supported by the administration as they would like to be; some would like a more open policy with regard to "free days"; others complain about financial difficulties which prevent the renewing of textbooks; some gripe about the fact that the high school building is not air-conditioned. The administration is bothered by students smoking in the bathrooms, the rowdiness of the freshman class, the teachers' seeming unwillingness to innovate or just simply to "be reasonable," the superintendent's rigidity on certain administrative questions, and the failure of the Kiwanis-sponsored "career night."
All these things are grist for the social organizational processes which structure relationships within the high school. They are the material which is used to create a situation of scarcity which allows differential rewarding systems to be put into effect. They are used to maintain and symbolize a certain social structure, while at the
—.same time the fact that these materials do not concern basic survival items makes even unequal relationships rather relaxed. The students know that they are-. privileged by comparison to most youths from nearby towns. The teachers know that they are among the best paid in their profession, and that they have the easiest students to manage that could be found. The administration is deeply satisfied with its staff on one hand,. and with the community on the other.
Yet, however mild the stratification processes- that exist in the high school may appear to be by comparison to what happens in other places, they are still at work. In- other words however homogeneous the high school might appear to be from a socioeconomic, demographic point of view, absolute equality or sameness is not what is evident when one first sees the school, or even, after one has come to •know it intimately. This is tme of the students. This is also true of the teachers, who differentiate among themselves and who are differentiated even more readily by students and administration alike. The Sheffield High School student body or teacher body is not an undifferentiated, unified group; it-is one that is highly organized. What are the principles of this organization?
As with any other classification system, this one uses events outside itself—natural, psychological, or sociological events—in order to make a basically cultural point.  What are these events and how are they organized? Let us first- list haphazardly some of the things used by the-teachers to characterize students.
All these events  can be organized into three broader categories according to the occasions and contexts in which they are used. First, there are things which are considered obvious, inescapable, and irrelevant to the mission of the school. Health, wealth, ethnic background, sex, like the color of eyes or hair, or size and weight, are the biological endowment of the person. As such, they are considered inevitably to have some. influence on certain behavior, such as to whom you feel "instinctively" attracted, whether you can be a member or the basketball team, or whether you use the boy's or the girl's bathroom. But to let oneself go wholly to those instincts, or to judge somebody according to these characteristics, is not appropriate for an individual and is illegal for the school as an institution. That one is doing so cannot be openly confesses, except maybe in a joking or angry manner.
The second group of criteria for classification involves performance, mostly scholastic and athletic, though it may also involve performance in the arts, music, and drama. This criteria is based on a performance—in opposition to a state of being—or on an act on the part of the student. This performance is "objectively" measured through tests and expressed in grades. The notion of objectivity is important here—it implies that the performance of the student is considered as an object rather than as a subject. In other words, what is judged is not the student's performance in relation to capacity—the attempt to reach a certain result, it is the result itself, an object which the student has made. The further implication that is made is that it is not the properly human part of the student which is primarily responsible for his grades, though this may enter into account in another way, as we shall see presently. What the tests are supposed to measure is the relationship between what the teacher has taught the student, the content, and the student's ability to restate this teaching. Not everybody is expected to perform equally. The tests are indeed designed to measure an aptitude to study. This aptitude is thought to be eventually related to the intelligence of the student which is a given of the biological endowment. The school, from this point of view, is the institution which develops this endowment to its potential and establishes objective ratings of students for future reference. It is important to note that this classifying process is a matter of absolute ranking from 1 to n; it is not a matter of associating certain students with certain categories. Colleges or employers may decide that they will only accept students with a score above 700, or 600, or whatever, and in the process may create a group or class, but the nature of this group or class is not given as constituent of the social world. Only the individual rank is.
The third system of stratification involves things such as dress, spirit, willingness to participate, morale, attitude, and the like. These are generally considered by social scientists to be secondary phenomena, used by the informants to express something else than what they appear to be saying, to hide prejudices or to express it... It would thus seem that one should go to the deeper causes rather than to the rhetoric about them. To follow such a route would violate our methodology. For the informants such things as "school spirit," "fair play," "lady-like" appearance, "being dressed like a bum," were real events recognizable in the world. Informants would often disagree among themselves as to whether a particular student was a bum or not, but they knew what each other way saying when he was talking about bums. On the other hand, these events were treated differentially from the other two sets which we explored earlier. There are no tests which can determine how much of a bum or a lady a student is. Similarly, being a bum or a lady is not part of the biological or sociological endowment of the student. This makes the classification absolute and subjective, rather than relative and objective. A child was not born with the behavior exhibited in the high school. And it is not because the parents are bums or nice people that the child itself is one of these things, though the parent's example in the home may influence the child in one direction or another. The child could conceivably change this behavior, willingly, with the help of a psychologist, or under the influence of peers. Intelligence cannot be improved, but will to learn and morale can. Far one reason or another a student may not want to change a behavior to fit a model approved by an interlocutor; the student may want to remain a freak in the middle of jocks or a jock in the middle of freaks. To do so is within students' rights. It also makes them responsible for their acts and means they cannot justifiably complain if certain people do not like the chosen behavior and refuse interaction. It may be improper for a person to segregate out another because of race or wealth, but it is proper to do so because the person is not dressed to one's taste or one just does not "feel good" with the other.
One can see this process as one of restriction in the field of social action. Biological endowment, or performance derived from this endowment, remains non-specific as a determinant for social action. The field is wide open to random association. It becomes culturally restricted and, indeed, organized in a specifically human way through those action in which one and one's interlocutors are "free," those which involve lifestyle decisions. One aspect of this process is passive insofar as one is not specifically penalized for having made a certain choice. At most one is refused participation in certain activities in which one might have wanted to be admitted but to the total atmosphere of which one is not willing, or not able, to yield.
But one can also be actively punished for certain lifestyle decisions. One may decide to smoke in the bathrooms, pull the fire alarm, or skip school. For any such acts, or for any others which include the breaking of a formal rule or regulation, the school, is entitled to actively punish the student. The school's function is to teach. [91 It can create and enforce regulations which it believes necessary to accomplish this task. It can punish students for performing an act which it believes disruptive. It should not penalize a student for what he "is," only for certain things which he does—the breaking of rules.
At a more general level, we can see that two sets of :e recognized by the people in the school: those to which the school as an institution is entitled to react, and those which lie outside of its jurisdiction. To the first set belong matters of scholastic performance and discipline; to the second, matters of sociobiological descent and of individual psychological personality. As we saw earlier, scholastic achievement is considered to be an aspect of the biological endowment of the student, while discipline is about regulating personality choices. We can summarize the above in the following diagram.
|Outside its Domain||Race-, sex, etc.||Life-style|
|Inside its Domain||performance||Discipline|
From this we can see that the school is interested in one way or another in all aspects of human life, but only in a limited fashion. This, of course, reflects the more general idea that the school is a piece of society rather than a microcosm of this society. The school as an institution is thus about two subsystems of the more general systems.
The school, however, is not only an institution whose mission is culturally defined; it is also a bunch of people. These people are not only in a- contractual relationship with the institution and are expected to maintain its image of itself, but they are also human beings entitled in other contexts to behave according to their personality. In other words, teachers, like students, see their realm of action divided into those aspects which are submitted to the institution and those on which they are fully their own.
The difference between teachers and students is, of course, that the former enforce the definitions while the latter submit to them.  On the one hand, the teacher is an agent of the institution and thus asked to rank students constantly. On the other hand, the teacher is a person who may not rank, though who may like or dislike other persons and wish to mingle with them. Outside of school, or with formal equals (e.g., other teachers), this latter right is not problematical: one is not forced to mix with teachers one does not like. Complex patterns of territorial segregation have been developed by teachers to maximize non-teaching time spent with one's clique, and to minimize contacts with teachers to whom one is indifferent, if not hostile.
With students, the problem is more delicate. Of course, one must not rank them or judge them except through grades. This ranking must be accomplished through objective consideration of performance. But one will also, by necessity, be confronted by students whom one "doesn't like," "cannot stand," and so forth; with students whose lifestyles one cannot approve of; students who are persons with whom one would not want to talk in a social situation. In other words, a teacher is allowed to choose the neighborhood to live in, the friends to mingle with, the teachers to lunch with, but not allowed to choose students to teach.
What happens is that the teacher grades the students in the classroom and in official reports to the administration, and judges them as "good" or "bad" in personal relationships with other teachers or members of the administration. The teacher translates the classification of students according to their lifestyle choices, and possibly their biological endowment if self-consciously "racist," into the classification of students according to their grades. It is not only that "students with good grades" (a performance qualification) become "good students" (a lifestyle judgment), but that the pattern of thought, which ranks students according to their grades makes the operator rank them according to their lifestyle. A student with good grades may be a good student for certain teachers and a bad student for others. The process, of course, goes even further than a moral evaluation of students. This judgment is further transformed into an unequal rewarding system in terms of minor privileges which are given by teachers to their "good" students. In his paper, Riffle gives examples of some of the types of privileges which the school is entitled to distribute, and he hints at some of the ways in which these are allocated to the students, sometimes according to formal criteria derived from performance and other times according to criteria which may be considered to derive from prejudice. There are also the somewhat unconscious processes which make teachers "expect" differential performance from students who have been pegged a priori as "good" or "bad," and thus generate it by this very expectation. Such processes have been documented at length in the educational literature.
Structurally, the process is one by which two classification systems, which we can visualize as two axes: one horizontal for a segregating scale and one vertical for a grading scale, are transformed by operators in ambiguous situations into a stratification scale. Distinctions which are very clear in the school charter, regulations, and administration, become much more fuzzy when used by actual people. Carryovers cannot but happen, and so it is recognized by the natives. It is widely admitted that to compartmentalize successfully the different ranking systems so that your choice on one scale does not influence your choice on another scale is very difficult and a feat out of the reach of most individuals. In consequence, complex systems have been developed to ensure that the legal ranking, development of which is the goal of the school, be effectuated as "objectively" as possible. The most important tests in the school life of students, the College Board Tests, are administered by agencies outside the school. Of course, by this time whatever influence the prejudices of the teachers may have had on the intellectual development of the student had already done its damage.
What is interesting for us here is what this recourse -to a non-personal arbitrator implies. It implies a desire for objectivity about certain things. It implies a definition of those things about which one can and need be objective. It implies a recognition that there may be situations when attitudes which cannot be objective may carry over into judgments that should be. A person cannot be a machine, even though a formal situation may indeed demand machine-like behavior such as absolute neutrality. It is as a person that a. teacher says "this kid is terrible" or "ads kid is a good kid."
We must go even further in this attempt to understand the actual processes of stratification. Teachers may be prejudiced, often violently, against certain students. They frequently have little knowledge of the friendship or clique relationships which exist among the students. In other words, they see students as individuals rather than as members of groups. This might not be evident in a school with strong -ethnic groups. In Sheffield it was particularly clear that the teachers, most of whom did not live in town and did not know the parents, have very little understanding of the social life of the students, even those who were most popular with the teachers.. The teachers knew that kids had friends-they were anxious that the kids have some--but they not generally, see their kids as members of little groups or cliques. Kids were "good" or "bad" according to a certain pattern of behavior which they as individuals had.
This should remind us of the individual nature of the process of scholastic grading and ranking. If groups of kids are created according to performance, it is a secondary process that is not relevant to the structure itself. The processes which we have identified are of a transformational nature in which informants are confronted by two sets of rules for behavior adapted to two predefined social contexts. These rules, however, leave the operators in the dark in certain situations which can be interpreted as pertaining to both contexts. This. allows the operator to use whatever set of rules prefered psychologically, or pressured to use sociologically. The result of the process is a surface structure which may include a set of more or less well-defined "classes." But these exist only at' products of a process; superficial manifestations of a deeper structure. To consider them as possessing a reality of their own is misleading. It is particularly misleading when the supposed existence of cliques or classes is considered evidence that a social situation in America is not as individualistic or as democratic as people consider it to be. In fact, it may be precisely the individualistic and democratic nature of the processes which make them produce the empirical results we as observers are confronted with.
Let us now summarize my argument up to here. I have argued that the field of interaction of an American is divided into two areas. In one of them what I shall call segregative processes-those which derive from statements of the "I Iike/don't like this-person" type-are at work in a fully legitimate manner. In the other field the person is asked to rank other persons and to reward them differentially, but only within a limited domain which is clearly defined in a mutually understood contract. On some occasions the two fields are clearly differentiated. The rater of a test, for example, does not have to worry whether the person being rated is "good" or "bad"; he can be completely detached. Indeed, the rater can be replaced by a machine, considered the outer limit of objectivity.
In contrast, the student who goes to sit at one table rather than .at another because "that's where his friends are" is not rating, he is just segregating. He does not have to assume that the others are bad, just that he does not like them. Indeed, cliques abound in the school, but there is no competition between them as such. Territories are assigned to certain cliques, and the other cliques would not openly challenge their right to that territory. The process is the same among teachers. They,too, have their cliques and their territories, and relationships of tense equality are: maintained between the cliques. When a clique, either among the teachers or among the students, comes to desire something another clique possesses, it can only get it through an appeal to a member of the administrative level above itself, using as a tool whatever it is that this higher level is entitled to judge and to reward differentially, be it creative teaching or academic performance. One cannot claim a privilege by arguing one's lifestyle but only by demonstrating that one will use the privilege in a manner that is a more complete performance of the things which that level is supposed to do. Cliques, or rather their members, may be unequal among themselves in relationship to certain types of gratification; they may be hierarchically related to an institution in a specific domain; but they are certainly not in a- hierarchical relationship towards one another. The relationship between them is one of avoidance and ignorance; it involves no exchange, it is not reciprocal. It is a relationship which involves separate, substantive entities which are inherently "the same," and thus equal.
The equality I am talking about here is, of come, of a different nature from the equality which some critics of the United States social system call for. In the school, the units, individuals, or cliques are not equal in relation to a scale, but in relation to themselves as they get ready to be ranked. It is not an equality of relation-since in fact all the unities are ranked-but an equality of substance-all the units are the same and are ranked according to the same principles.
We are rediscovering here certain choices which America- as a whole has made over and over again. Equality in front of the law, "one man-one vote" laws, all imply the same definitions of humanity and social action as those we have seen emerging from our analyses Of processes of ranking and classification in a high school. The basic choice of America Is that "men have been created equal," not that they should or will remain so from a social point of view all through their lives. The notion of competition, which is another aspect of the notion of ranking, is central to many aspects of American culture-and-well demonstrates the limits which are put to "equality."
But, in fact, as one, reads between the lines of what many critics of American life write, one sees that what they are objecting to is not so much that people, or kids, are actually ranked, but that the ranking process is in one way or another skewed in favor of a certain group and against another one. What those critics are really saying is that there is not a place in America for encompassing groups. The ideal to which they compare the statistical reality which they observe is a state in which rank is filled by people from all categories of Americans—they are talking here about the "objective" categories which are the same as those I delineated earlier—in a truly random fashion. This is the same ideal which the people in the school attempted to reach, except that it was, of course, more difficult for them to reach it than the critics say it is since principal, teachers, and students are confronted by a total social situation rather than by a small aspect artificially taken outside of its context. Ranking is never found by itself, it is always associated with freedom of choice and our responsibility as social scientists is to investigate the dialectics of the two.
To say that the critics have misdirected their attacks, that in fact they are simply restating the myths. rather than attacking them, is not to say that the present American system is ideal. As Merleau-Ponty once wrote:
Nous savons aujourd'hui que l'egalite formelle des droits et la liberte politique masquent les rapports de force plutot qu'elles ne les suppriment. (1966, p. 180) 
' This indictment of the system is much more radical than that of people like Hollingshead or Warner. It is a denial of the value of the myth rather than a complaint that the myth of democracy has not been fulfilled. Merleau-Ponty was, of course, arguing in favor of an idealized Marxist society. From this vantage point, he could see the inescapable tension that exists within a classical democratic society. In the same way as we believe as anthropologists' that our model of foreign societies is more adequate than the native ones because of our distance from them, we have to create a distance between us as social scientists and us as natives when we are talking about societies in which we have been born and reared.
It could be argued that the models of American society proposed by people like Warner and Hollingshead, all the models which attempt to picture this society in terms of well organized classes, are but the creation of another false consciousness, more insidious because it appears grounded in science, one of our most sacred domains. Social classes, we saw, do not explain behavior in Sheffield High School. The apparition of something which looks like social classes. either._ statistically or in the speech of informants, must be considered to be an epiphenomenon, the product ofmore general processes which are not organized on class principles. Indeed, it is the very democratic ideal which allows the individual to choose his social partners—an ideal profoundly valued in an institution like marriage—which in certain situations produces states of being which go against some aspects of the ideal. For an American to fail to recognize this is to delude himself into believing that those unequal results are not the product of something he wishes to continue to value.
1. "The image of imagination is not submitted to empirical verification" (my translation). -
2. I began this work with a rather programmatic paper on "Culture and Stratification in an Equalitarian Civilization" (1974) where I dealt with some of the theoretical problems raised by the cultural approach to the study of human behavior and with some of the empirical problems raised by a traditional approach to stratification in the United States.
3. The foremost influence on Warner's early work had been Radcliff-Brown's. But even though this one was interested in "real relations," an interest picked up by Warner, he also believed that the social relationships which he analyzes as being central to the nature of the system would be recognizable in terminological classificatory systems which the natives used (1968). For him, what the natives say may often be difficult to interpret, but it is never irrelevant nor is it ever "false."
4. Lévi-Strauss has argued in several instances about the Bororo social structure, about the myth of Asdiwal, that the model of his society which an informant may first give, or which he elaborates in his myths, may not correspond at all with an actual state of affairs (1958, 1960). He also argues that this does not make the mythical model less interesting to analyze and, more importantly, it leaves us with the task of finding how the two models are interrelated.
5. The strongest statement in favor of seeing stratification systems differentially, from the point of view of their internal organization rather than as a simple matter of inequality, has been made by Dumont (1961). He argues there in favor of seeing the Indian caste system on the one hand, and the American so-called "class system" on the other, as two qualitatively different systems that must be understood primarily in relation to themselves and their cultural contexts rather than in relation to an abstract scale of relative deprivation of certain people vis a vis other people.
6. I want to take the occasion here to thank the other members of the team who worked with me in Sheffield and to whose work I am deeply indebted, Patricia Caesar and Rodney Riffle. I want also to express my gratitude to those who have listened to my early formulations—James Boon, Carlos Dabezies, Beth Haggens, and especially my wife Susan.
7. I am not saying here that when human beings classify they use objective criteria of the thing or person classified. But, following Lévi-Strauss (1962a, 1962b), I am saying that in order to express the "differential gaps" (ecarts differentiels) which men have to create among themselves to understand their society and speak about it, they will always relate the categories thus formed to other categories that have been made in the nonhuman world. But "nonhuman" does not mean necessarily "natural." Some societies organize human relations according to the social functions which it recognizes. There is nothing theoretically surprising in finding cultures in which men's characterizations are symbolically expressed in relation to psychological tendencies or biological items.
8. It should be evident that for me all these events are interesting insofar as they are used in a cultural system for cultural purposes. I am not saying anything as to their "real" value within their own systems, social, psychological, or biological.
9. It might be argued that I am overly limiting the function of the school. Some informants will say on certain occasions that its function is also to teach the kid "how to be a member of society." This is, of course, quite different from teaching mathematics and French. In fact, the institution as a whole is not really geared to "teach" the very abstract and nonobjective "subject" of "being social." It appears mostly to be considered a side benefit of learning math and French. And yet, if we were to analyze what is meant by "being a member of society," we would find the same organization of elements that we have outlined, though at a more general level, in which the items of being and doing are less sharply defined.
10. We must note here that teachers are like students in their relationships with the administration insofar as they are regularly rated by it on their performance as teachers within the restricted definition of what a teacher's role is. I do not have the time here to expand on this parallelism.
11. "We know today that formal equality of rights and political freedom hide rather than supppress power relationships" (my translation).
Bachelard, G. La Poetique de l'Espace. Paris: Presses Universitaires de France, 1957.
Dumont, L. • "Caste, racism and `stratification'." Contributions to Indian Sociology, October 1980 (1961).
Hollingshead, A. B. Elmtown's Youth: The Impact of Social Classes on Adolescents. New York: Wiley, 1949.
Lévi-Strauss, C. Anthropologie Structurale. Faris: Plon;
-- Le Totémisme Aujourd'hui. Paris: Presses Universitaires de France, 1965 (1962a).
-- La Pensée Sauvage. Paris: Plon, 1962b.
-- "The story of Asdiwal." In The Structural Study of Myth Leach (ed.) ASA. Monograph No. 5, 1967 (1960).
Merleau-Ponty, M. Sens et Non-Sens. Paris: Editions Nagel, 1966.
Radcliffe-Brown, A. R. Structure and Function in Primitive Society. New York: Free Press, 1958 (1952).
Varenne, H. "Culture and Stratification in an equalitarian Civilization." Proceedings of the IXth International Congress of Ethnological and Anthropological Sciences, 1974
Warner, W. L and Lunt, P. The Social Life of a Modern Community. Yankee City Series Vol. 1 . New Haven: Yale Univ. Press, 1941.