The Adolescent in His Group in Its Setting: II. Research Procedures and Findings

Muzafer Sherif and Carolyn W. Sherif

In this chapter, we will outline the procedures and research tools developed for intensive study of natural groups on the basis of the theoretical guidelines presented in Chapter 12. Findings from our research on youth in various settings within urban areas of the southwestern United States are incorporated throughout the chapter.

The procedures and tools in this research program are not entirely distinctive-if there is innovation, it lies in procedures for studying behavior in natural groups of adolescents without arousing awareness of research intent. Nor do we propose that all the findings on adolescent youth are unique to our research. As indicated in the last chapter, we have freely utilized previous research from psychologists and social scientists as bases for hypotheses, concepts, and research procedures. We would be surprised if our findings bore no relation to those previously reported on American youth.

On the contrary, our purpose is to demonstrate the importance of incorporating the many influences on youthful behavior into single research designs. The neglect of some facets has led in the past to emphasis on one particular point to the neglect of others in theories of adolescent behaviors and misbehaviors. We have had ample opportunity to see the fruitlessness of one-facet theories, whether they focus exclusively on early childhood development, intrapsychic conflicts, sexual development, social class or local subculture.

A theory capable of accounting for the behavior of individuals interacting with others in definite habitats-that is, social-psychological theory -must be based on an integration of findings about all the significant influences affecting behavior. These include the motivational dilemma of the adolescent in his society; the person he is at the time, including his skills and his desires to be part of some scheme of human endeavor; the influences from other people who count in his eyes; the properties of his relationships with those people; the values and facilities of his sociocultural setting; and the setting's place in the larger social scene.

( 296)

If a theory is to be adequate, its development must proceed hand in hand with the development of operational tools for research incorporating all the significant variables. Without such tools, no theory can link its concepts together and deal with empirical relationships, no matter how elegant it may sound. And without a theory to integrate the significant variables affecting adolescent behavior, practical measures for preventing the wastage of youthful potentialities can only proceed on a hit-or-miss basis.


Let us sum up the general features of the adolescent period, since the influences that an adequate research design must study follow from these features.

First, the adolescent is a human male or female with at least a decade of life behind him. His body and its functioning are beginning to change toward that of an adult human. As they do, he is expected to alter his behavior toward others and toward his current and future responsibilities. Although ordinarily he wants to do so, these behavioral alterations can be achieved in a consistent way only if he changes his picture of himself relative to others. Thus they demand changes in self or ego-attitudes in various respects.

Second, in modern societies, the motivational problems that come with physical changes and the necessity of changing one's self-image are further compounded by ambiguities in adult definitions of the transition period. The general characteristics of the period as defined, or left undefined, produce strong motivational problems and dilemmas which are, indeed, common to all youth in some degree.

Third, adolescent behavior under the grip of these motivational problems and dilemmas is not entirely unique to the adolescent period. Similarities may be observed among any individuals presented with a motivational problem. A frequent response, when individuals perceive that others share their problem, is to come together in regular association. Since age-mate association is both permitted and encouraged in modern societies, youth do gravitate toward one another, associating more frequently and more intensely than in earlier childhood. The domain of other adolescents becomes a reference set of greater salience than that of adults or younger children (as demonstrated experimentally in Chapter 12).

Fourth, regular associations among individuals of any age with common problems acquire distinctive properties. These include some patterning or organization of interpersonal dealings, and some agreement as to what objects and behaviors are acceptable and which are not, what is to be prized and sought, and what is "ideal." These two properties---patterned interpersonal relations and shared yardsticks for evaluation---are

( 297) the minimums defining a human group. Thus, group formations proliferate during adolescence.

Fifth, what the individual adolescent desires, what he sees as the ideal, what he does in concert with age-mates, the character of their group and its products, are not direct outcomes of the motivational problems which bring him to his fellow group members or which he discovers he shares with them.. Here, we must consider both the immediate sociocultural setting and its facilities, and the values of the larger society of which they are parts.

During adolescence, the character of the general culture and the immediate circumstances of living take on new and added significance. The radius of self-concern is expanding beyond its more limited scope of childhood and extending further into the future. Even in the immediate present, the adolescent is more tuned to sociocultural influences and his society's adult success images. He is more mobile than a child.

The point may be illustrated briefly: In our research, we found that boys and girls in the southwest come to feel, by the age of seventeen or eighteen, that it is their inalienable right to have a car. A car is seen as a necessity. Particular kinds, models, and colors desired are specified to the last detail. There can be no doubt of the motivational press behind these desires, nor of their relevance to other common motivational problems (including notably those related to prestige, contacts with and conquests of the opposite sex). However, it is impossible to understand this phenomenon apart from an understanding of the "car culture" in American life, with its mass salesmanship and broad impact on social life.

The general "car culture," combined with physical characteristics and facilities of southwestern cities, has made possession of a car the ideal in all kinds of neighborhoods, rich and poor. Having or not having a car, as well as the kind of car it is, affects one's status and prestige with those of both sexes. The desire for a car is certainly a "psychological need" as experienced by the individual. But this need and the resulting development of a subculture with stylized patterns of driving around, of joy rides or outright theft, would be bizarre phenomena if a researcher should attempt to study them apart from the widespread importance of a car in American life, and the differential availability of cars in different settings. The "need" for a car, as well as "needs" for thrills, or for defiance, or for "acting out," are meaningless apart from the character of the sociocultural settings in which they have been nourished and shown in action.


If, as in our research, we wish to study adolescents as individuals, we must incorporate into the design at least the major sets of influences that affect their attitudes and behavior. This is equally true whether our problem concerns adolescent accomplishments in socially desirable direc-

( 298) -tions, or their undesirable modes of activity, such as drug usage or car theft. Within the general problem of adolescence as defined in a society, we have seen that the sociocultural setting and its values are major influences, particularly those values prevailing among the reference set of teenagers. Finally, the design would be incomplete without including study of the individual's reference group of age-mates, and his relationship to them and to other groups in his ken (family, school, church groups, etc. ).

As outlined early in the last chapter, the design of research included three main facets to accommodate these major influences. The next sections describe the operational steps in the three facets, with emphasis on the intensive study of behavior in natural groups of youth. Finally, findings in the three facets are summarized to indicate their interdependence.


To be sure, there must be a division of labor in the sciences of man. Study of sociocultural settings is a sociological or anthropological task. If the task has not been done as specifically as necessary for research on particular adolescents, then the investigator must "act like a sociologist" for the time. He can get help from social scientists, borrow their tools, or secure a collaborator, but he cannot ignore the task as being irrelevant to his own interests. Nor can he evade the task by assigning a blanket label to the setting, such as "lower class," "middle class," or "ethnic subculture." This evasion would be analogous, for example, to an experimenter in color vision saying: "I am studying color vision in an illuminated laboratory. But I do not know how much or what kind of illumination it is, because I am not a physicist or electrician. Information of that kind isn't my job."

Like the physical surroundings, the social habitat has regularities in structure, time sequence, and recurrence of events. For example, the range and standards of living in a neighborhood where adolescents live and meet indicate a patterned set of circumstances. These circumstances are pertinent to what youth do, how they spend their time, where they spend it, and even what they consider suitable activities during leisure hours. Boys in favored neighborhoods of high socioeconomic rank own cars, have comfortable homes, and ample spending money. They will think of leisure related to their cars, their visits and parties in each others' homes, and outlays of money for dates and professional entertainment. They are extremely mobile. In fact, the entire city is "their oyster," which makes their study unusually difficult. The contrast in mobility and attitudes is striking in an urban neighborhood where cars are not available, homes poor and crowded, and money lacking.

In our research, residential areas where the youth live are studied through first-hand exploration, mapping of facilities, available municipal statistics, and (in neighborhoods with Spanish-speaking residents) block

( 299) surveys to check the extent of acculturation among residents. Looking for ways of indicating features of the setting, we found that Shevky and Bell had developed a social area analysis based on census tract statistics. A limited number of census measures, combined appropriately, serve to indicate relative ranks of areas in significant respects. These indicators seem to account for the major sources of variance in a host of discrete statistics on urban areas. Each is associated, therefore, with correlated information not included in the analysis.

Bell's chapter in this book (Chapter 11) obviates the need of presenting the three indicators in social area analysis in detail. They are, however, (1) socioeconomic rank (2) degree of urbanization and family conditions, and (3) ethnic status of the population and their concentration in an area. In addition to being reproducible and communicable, these indicators have the advantage of showing gradations in a coordinate system based on standardized scores. For example, it is possible to specify that one area of low socioeconomic rank is not as "low" as another. Or, a low rank area is not as urbanized as one of middle rank in the same city; that is, there are fewer apartments in the low rank area, more children, and fewer mothers who work. Clearly, the ecological conditions are quite different in the two neighborhoods, apart from socioeconomic level. Life in this low rank area, with its shabby two and three-room houses, is likewise different from that in a comparably poor tenement neighborhood of a large eastern city.

Specifying ethnic status of the residents is equally important, though not always sufficient for our purposes. In some areas, with a sizable proportion of Mexican-born citizens, we have conducted block surveys to assess the relative acculturation, on the basis of length of residence and style of living. Indications of relative acculturation to U.S. life were, for instance, the type of pictures on the walls and the cultivation of yards, ranging from the flower-pot culture of the Mexican to the green lawn of the "Anglo." The advantage of such specification of the setting can be illustrated, and may be similarly illuminating in other sociocultural comparisons, for example, between neighborhoods of Negroes raised in urban settings and those peopled by recent migrants from rural areas.

In this instance, we found that a group of adolescent boys in a "Latin" neighborhood had very little interest in organized sports, unlike most of their counterparts in other areas. When we found that their neighborhood was among the least acculturated to American life, they appeared as fairly representative of their immediate setting, although otherwise these boys' lack of interest in sports would have seemed extremely atypical. This information also clarified the frequent observation that whenever one of these boys spoke English to his fellows, he was chastised to remember he was "Mexican." (Their Spanish, however, would not have been altogether understandable in Mexico.)

( 300)

Later in the chapter, the dangers of blanket conclusions about youth in a given socioeconomic level or subculture will be mentioned. We warn against generalizing about youth in any given stratum without sufficient evidence and against viewing any stratum as an isolated phenomenon. We stress both the importance of a comparative approach to youth from different backgrounds and the relationships among different strata in a society. After all, socioeconomic strata are characterized as "low," "middle," or "high" in rank according to social criteria as well as economic needs for physical well-being. The ill fate or good fortune of a particular stratum is, therefore, conditional in part upon the larger social system.


For a particular adolescent, the values and goals of other youth in his ken are salient aspects of his environment. Whether his personal radius for achievement and his potentialities are in line with theirs or not, he knows full well that his actions are gauged relative to their standards, which are revealed to him time and again in episodes of action, appraisal, approval, disapproval, notice and notoriety, or, worst of all, ignominy and personal oblivion.

For such reasons, one facet of the research program assessed conceptions of propriety, of achievement, and of success in the various sociocultural settings. A paper-and-pencil form presented as an opinion survey was administered to representative samples of secondary school students in the study areas. To date, three versions of this form have been used with many identical items in each. Throughout, the aim has been to devise items on significant aspects of adolescent life in forms which will yield a range of conceptions, from the acceptable minimum to the personal ideal. The content areas include conceptions of leisure time and activities, work and future occupations, school and academic goals, financial and material necessities and goals, parental regulation and controls, proper modes of behavior, desirable associates, and personal success and aspiration. The data are analyzed both in terms of the bounds of acceptability and achievement prevailing in, a neighborhood and of the distribution of individual responses within these bounds.

In other words, the data in this facet are sociocultural data on the adolescents themselves. In view of the school context in which the forms are administered, the data are most accurately interpreted as indicating what youth regard as "socially desirable," in Allen Edwards' use of that term (1957). However, in a school setting, young people do not always respond readily in terms of personal experiences or specific social relationships. The records contain overheard statements such as, "Boy! Did I feed them a line." In non-school situations in which the same forms have been administered, observers who have established more rapport with boys than most other adults have encountered resentment to some items, particularly those pertaining to areas where adults are usually

( 301) excluded. Recent administration of the forms in detention and reformatory settings also indicates evidence of attempts at dissimulation, but the data here differ from those obtained in school settings from youth with comparable socioeconomic backgrounds.

The different pattern of responses in different sociocultural settings, and some evidence of dissimulation, support our interpretation that these data express youths' conceptions of what it is socially desirable for them to report. But even these conceptions of "lines" to give adults vary in several important ways, and are interesting data in themselves, revealing socioeconomic and cultural differences.

Since one's self-radius for achievement is, in part, a function of what one conceives as socially desirable, the differences between the bounds prevailing in different sociocultural settings have important psychological implications. Even if not followed, the prevailing bounds in a social setting are significant as stimulating (external) conditions for the individual youth. Their salience can be illustrated briefly.

In a lower middle class area suburban to a large city, one small group of boys studied in our research for nine months had all dropped out of school. Although some of them had been prevented from re-entry by school authorities, the majority had been subjected to persuasive and coercive efforts by their parents to continue in school. One parent had even given a bribe-a new car-as the reward for school attendance, to no avail. By their actions ,and in conversations, these boys deprecated school attendance. They were, however, fully aware that among their age-mates outside their group, high school graduation was the minimum standard for achievement. Their awareness was revealed in repeated instances of withdrawal from the "school crowd" as "snobs," in their unanimous agreement about the impossibility of returning to the local high school, and by statements by all but one member that they intended to finish school in a neighboring town or "in the army" eventually, thereby reaching a par socially and occupationally with their contemporaries.

The data on prevailing self-radius and goals of youth in different urban areas are used in four ways:

First, to detect values and goals upheld by common consent of youth in different areas.

Second, to detect differences between the areas in these respects.

Third, to assess the homogeneity or heterogeneity of standards prevaling among adolescents within an area.

Fourth, to form a basis of comparison for assessing the typicality or deviance of the attitudes and goals expressed by members of small groups studied in an area.


Within each neighborhood of low, middle or high socioeconomic rank, the attitudes and behaviors of boys in informal groups of their own

( 302) choosing are studied intensively over periods of time ranging from six months to a year. This focal facet of the research is carried out in the field -in the natural habitat of boys themselves.

In this aspect, we had to face seriously the problems of research method arising from the context effects discussed at some length in Chapter 12. The very nature of research by people into the current behavior of other persons raises issues which are by no means so obvious in the other sciences. One of these is how to establish research contact with the phenomena under study without fundamentally changing them.

Admittedly, astronomers had the problem of research contact with the heavenly bodies. The technological means for observation (telescopes, electronics, satellites) have not yet substantially affected the movements of heavenly bodies nor their nature.

In the study of human behavior, however, there is growing evidence that technological devices and instruments may be inimical to approaching the phenomena of interest. Their usefulness begins after the investigator has solved the problem of getting close enough to human beings to study them without letting them know they are under observation.

Therefore, the intensive study of behavior in natural groups proceeded from the premise that the timing and selection of research procedures should be such that they would have little effect on the group and its members. This is the reason for all of the efforts to keep group members from being aware that they were being observed for research purposes.

In the varied and changing field of social actualities, a single observer will inevitably choose facts selectively. It was our second premise that selectivity could best be minimized if the observer focused on one aspect of the interaction process at a time, and a variety of independent data gathering techniques were introduced at choice points. In appropriate sequence, the observer's reports and ratings of behavior were checked against ratings from an independent observer, from sociometric interviews with group members, from interviews with adults who knew them, and from public records pertinent to their past activities. The use of a combination of techniques in a variety of situations over time is the best way to insure validity of findings, which cannot be guaranteed by sophistication in test design, or in planting recording devices, or in later analysis.

To date, the study of 24 groups of boys (ages 13-18 years) has been completed through the cycle of steps envisaged. The difficulties, and in some cases failures, in completing the study cycle, would make a fascinating chapter in themselves on the rigors of social research. These difficulties range all the way from practical problems of carrying out procedures in a particular setting, to some observers' lack of persistence and interfering preconceptions on the nature of research. Many of the latter preconceptions, formed in other training, must be revised in an attempt,

( 303) such as ours, to fit research procedures to the actualities of the field, rather than vice versa. The real challenge of social research is to secure precise and manageable data that faithfully reflect the actual properties of events.


When an observer begins to hang around in the vicinity of a group, there is always a period when the members observe and scrutinize him. The observer being observed by his subjects is one instance of a general phenomenon of person perception: first impressions are invariably accompanied by efforts to place the person and his presence in our area in some social categories.

For this reason, the prime criterion for placing an observer is his "fit" with the predominant socioeconomic and cultural background of the residents. In appearance, speech, and manner, he must blend into the scene. Even then, the recurring presence of a strange face in familiar surroundings calls forth questions about why he is there and what he intends to do.

In our research, the observer is instructed to develop first a reasonable pretext for his presence that circumstances cannot contradict even before he makes contact with a cluster of boys. Preferably, the pretext is one which will bring the boys to the observer because of their own interest in his activities or possessions (for example, his athletic equipment, or his car).

The age difference between observer and observed is another factor to be reckoned at this stage. Our observers are ordinarily in their early twenties, slightly older than the group members they observe. This has proven necessary for two reasons: First, the observer must be trained and mature enough to follow procedures in step-by-step fashion. Second, an adolescent observing adolescents is in the constant danger of becoming part of their group, competing for status, and seeing the interaction process from his particular role in the structure.

Fortunately, we discovered that adolescent boys in most communities do have contacts with slightly older males and are, in fact, somewhat attracted by the possibility of contact with a person representing what they may become. The age factor, therefore, has not been insurmountable when the observer succeeded in establishing his presence as a sympathetic, possibly helpful young adult-somewhat like an older brother.

Still, for weeks or several months, the observer is observed. He is asked pointblank who he is, his connections, why he is around, and what he is doing. Over the course of time, the boys check up on his replies before they start opening up to him. The observer is still denied the luxury of direct questions to the boys. The procedures he employs are planned in step-by-step fashion to gain data while strengthening rapport.

( 304)

The initial and most difficult period in the intensive field study of natural groups is making contact with the members and overcoming their wall of resistance and secrecy about their more private activities. It challenges the patience and skill of the observer. Depending on these as well as on the group, the study has sometimes not progressed beyond the period of the observer being observed. We suspect that in most such cases, the rigid walls of resistance and secrecy indicate tightly knit groups with stable organizations and binding commitments in activities of an illegal or antisocial nature.


Once contact with group members has been made, the real cycle of procedures in the study of natural groups begins. However, the particular methods used, and their sequence and timing, are planned to coordinate with the development of rapport between observer and group members.

In the early period, no techniques are introduced that require probing questions or other assaults on the privacy of group activities. The observer simply reports on what he sees and rates member behavior after leaving the group. No other techniques are used until the observer is tolerated by members in activities and in discussion of matters considered exclusive and private.

Gauging the degree of rapport is too critical to be merely a matter of intuition. We utilize a combination of indicators in deciding on the timing of the procedures in the study cycle. Indicators of degree of rapport include these signs: the observer's success in finding the various places the group congregates when not in plain public view; the members' tolerance of his presence in these more private places; the degree of intimacy of activities the members discuss freely in his presence; the extent to which members welcome him into activities which they would hide from other adults.

Valid data on group properties and member attitudes can be obtained only if rapport is developed. For this reason, ratings by independent observers, sociometric choices, interviews and other procedures that assume cooperation with the researcher (without dissimulation or fabrication) are kept for near the end of the study cycle. Collection of case history materials is the final step in the cycle. Data obtained from all of these procedures can then be cross-checked against the mass of documentation of member attitude and action which has been collected during interaction episodes.

Developing rapport and adjusting procedures to the changing degree of rapport are continuing processes. Recently one of the observers (Mr. Lauderdale) attempted to conduct individual interviews with items from the Self-Radius and Goals schedules; this followed seven months of observation, during which he had been permitted to learn some of the more

(305) intimate secrets of the group including drinking, sexual activities, and incidents involving the police. The individuals responded freely enough about money, jobs, leisure activities, and the like. But he reported responses ranging "from apparent indifference and withdrawal to outright hostility" when he queried them about school experiences and about their special friends. "Remarks ranged from simply `I don't know' to explanations that `I don't like to answer questions because someone might try to get something on me,' or `You're sure full of questions, aren't you? I don't like snoopy questions from my friends or anyone else."'

Another observer (Eduardo Villarreal ), who returned to a group he had studied about three years earlier in order to check on present membership, reported that new members were withdrawn and suspicious even though the older members introduced him as an old "camerado." The newer boys relaxed only after questioning him themselves to find out exactly what he was up to.


The complete study cycle is accomplished through ten successive sets of instructions to observers, followed as rapport warrants going on to the next step. The instructions are revised regularly on the basis of new research experiences, but they are substantially the same as those presented as the appendix of our book (Sherif and Sherif, 1964). Here, the steps will be summarized briefly along with definitions of the principal measures to be used by observers.

First, a group is singled out for observation on the basis of observed frequency of association among a specifiable cluster of boys between 13-18 years in specific locations (recreation center, pool hall, drive-in, vacant lot, etc.).

Second, after establishing rapport to the point that he is tolerated by the boys in locations and activities other than those of their initial encounter, an observer directs his attention to the status or power relations among the individuals. The prime criterion for status or power in the group is defined by the concept of effective initiative, or the extent to which suggestions made by the various individuals are actually followed at that time or later. Allied observations pertinent to ratings of status or power include concrete examples of deference, effective dominance, or submission.

Although the pattern of power relations, as defined above, is not the only dimension differentiating the attitudes and behaviors of members in a group, it is the most useful dimension for predictive purposes. If one knows which boy has most frequently suggested alternatives that the others actually translate into common decisions and actions, he has found the operational leader of the group. If he can rank other individuals in the same respect, he has a powerful predictive device for what may trans-

( 306) -pire among them. The observers in our research have been able to do this, and the fact that the resulting pictures of group structure are invariably hierarchical reflects something about the nature of social power. Unlike some other attributes, individual power is necessarily limited by the nature of interaction. If members adopt the suggestion of one member, those made by others are bound to be neglected for the time.

Of course, other dimensions contribute to the differentiation of individual roles in a group: popularity, special skills in activities valued by members, and special resources, such as money, a car, or a home for entertaining. However, any one of these dimensions, so frequently taken by responsible adults as the criterion for adolescent standing, is imperfectly correlated with status (power) in their actual interactions. (Cf. Sherif and Sherif, 1964, Chapter 7.)

Third, the observer focuses on the prevailing customs, procedures and norms of the group. He looks for collective products of their interaction-procedures typically followed in frequent activities, common jokes and sayings, signs of dress or decoration, names for "us" and "not us," and expressions or deeds denoting what is considered acceptable, unacceptable, desirable, or unsuitable enough to demand reprimand or punishment. These normative properties are assessed in three ways:

1. Observation of distinctive similarities in expressed attitude and behavior among members that differ from those in other circles or groups.

2. Members' reactions to the usually acceptable range of behavior, including rewards and approval for praiseworthy actions and various forms of punishment or disapproval, indicating deviation.

3. A new member's conformity over time with established procedures, modes of attitude and action.

Next, the observer's ratings and reports are checked through independent techniques, including ratings of a significant group activity by an independent observer, sociometric ratings obtained from the members individually in informal interviews, and situational tests devised especially to create forced choices between association with the group and other activities.

Finally, through interviews with the boys, parents, teachers and other adults, as well as any available records in school or community, a "natural history" of the group and the backgrounds of particular members are reconstructed. The observer also maps the spread of the dwellings of the members, which indicates the effort they make to associate, especially when they do not have transportation available. It is noteworthy, perhaps, that only one of the groups studied thus far has been clustered within the same block. Even "neighborhood" groups are scattered over considerable areas, which shows something about the mobility of urban life for adolescents, but also reveals a selective process in the formation and membership criteria in these groups.

( 307)


Through these procedures, we have specified the essential properties of adolescent groups and their changes over time. These group properties, and the individual's place in the group, provide a powerful basis for predicting adolescent behavior, whether desirable or not. The extent to which behavior is regulated by group membership is related to a person's standing in the group. The degree to which group norms are binding for an individual is related to the stability of his relationships with other members (as demonstrated experimentally in the last chapter).

Being part of recurrent interactions among contemporaries means, for the individual, that he will to an extent regulate his behavior relative to others. Hence, the stabilized patterns of interaction in a group and the individual's location in the pattern affect his treatment of others, their treatments of him, and how closely he adheres to group norms.


In our earlier studies of group formation and intergroup relations in summer camps (Sherif and Sherif, 1953; Sherif et al., 1961), we observed certain regularities in the formation of group structure. Invariably, the first signs of structure were the stabilization of the top and bottom levels of status or power. An observer watching the groups in a variety of activities could consistently rate the persons highest and lowest in effective initiative long before he could differentiate intermediate positions. The boys themselves also agreed more consistently in choosing who was most and least effective in initiating their activities and getting things done.

It is striking, therefore, that the end-anchoring perceived by group members (indicated by their sociometric choices) is similar to the judgments of status by an observer. On the basis of the ratings made by observers of 24 groups, we have constructed a theoretical diagram showing the end-anchoring of an observer's judgments over several blocks of observation periods. Actual data necessarily deviate in detail from the theoretical curves, depending on the number of individuals in a group, changes in membership, or shifts in activities which affect the group structure. Nevertheless, the phenomenon in question was found for each of the groups studied.

Figure 18 represents the average variations in observer's ratings of effective initiative of group members over a substantial block (10) of observation periods, each an hour or longer. Each point on the graph represents the average changes, from one observation period to the next in a given block of observations, in his ratings for members in different segments of different status (base line). In other words, the base line represents the average rating for members during a block of observations;


Figure 18.
Figure 18. End-anchoring in observer's judgments of status in natural groups over time

the ordinate indicates average shifts from one period to the next. For example, if the observer changed his ratings on an average of one status rank for each of ten periods, a point would be located on the ordinate at .9. Or, if he changed his rating by five ranks twice throughout the block, the average change would be 1.00. Such large shifts usually reflect a process of change in the status structure.

Each curve in the figure represents a different block of observations, lasting around a month, from Block a, during the early weeks of observation, to Block n, when the observer has been with a group several months. End-anchoring is revealed in the observer's greater consistency in rating persons in the highest (leader) and lowest (bottom) positions than in the intermediate ranks. The slightly lower variations for high status ranks

( 309) just below the leader and for low ranks just above the bottom represent empirical findings. The intermediate ranks are typically most subject to change.

In our data, the introduction of a new member increases the variation in ratings for other persons near the status level where he first enters the group. If he is a friend of a low status member, variability increases in the low status ranks. If, on the other hand, he immediately shows some highly prized skill or possession, his presence increases variability in ratings of high status ranks. Changes in group structure for other reasons (for example, significant changes in activities, or departure of a member) are similarly reflected in sharp increases in variability of the observer's ratings.

Figure 19 represents the observers' average ratings of confidence in how they ranked status from time a (bottom) to time n (top), when they had become closely familiar with a group. Since the confidence ratings

Figure 19. End-anchoring revealed in confidence of observers in judgments of status in natural groups
Figure 19. End-anchoring revealed in confidence of observers in judgments of status in natural groups

(310) were introduced more recently, this theoretical figure is based on the ratings of fewer groups. Each time he ranks members according to effective initiative, supporting each rating with concrete observations of behavior, the observer indicates how confident he is of each rating, from "altogether uncertain" (0) through "wavering between certainty and uncertainty" (3) to "altogether certain" (6). Figure 19 shows that early in observation (time a) observers are generally uncertain of their ratings, and that their confidence increases over time (to time n). The end-anchoring effect also shows here, in the consistently higher confidence for the high and lowest positions.

Is the end-anchoring effect in group structure solely a perception of the observer, or does it reflect also a reality of group life? Independent raters have agreed closely with the observer when given sufficient opportunity to observe the group, and group members themselves agree more closely in sociometric choices at the highest and lowest levels. End-anchoring seems to be a general phenomenon in perceiving interaction patterns in a group. Furthermore, the ratings by observers can be reconstructed by a content analysis of the concrete events in which members show varying degrees of effective initiative.

We have been asked if the more dominant anchoring position of the leader indicates that groups possibly form through clustering around one individual. On the basis of our research evidence, we feel that this explanation is not sufficient. In only two or three groups have we been able to ascertain that the boys came together initially around a given individual whose initiative was accepted from the beginning. These were all fairly unstable groups. But in the majority of cases, we have found that the leader position has evolved during day-to-day interaction, rather than being the cause for interaction. Our observations include cases of leadership changes, reflected in temporarily increased variability of the observer's ratings of the high status levels, and then followed by renewed end-anchoring on the new leader.

In short, the predictability of individual behavior regarding effective initiative in interactions with others is related to status in the group, in a curvilinear fashion. The curvilinear end-anchoring effect is strikingly similar to that reported in the psychological laboratory for judgments of any series with well-defined end stimuli (for example, series of weights, lines, or sound intensities).

This finding has both practical significance for the prediction of adolescent behavior in natural groups, and theoretical significance in the study of human behavior. We see here the operation of a well-established principle of psychophysical judgment, namely that variations in judgment are smallest for well-defined extremes in a series and greatest for intermediate values. Our conceptualization underscores the operation of this general principle in field observation. The differences between judgments

( 311) of series of objects and judgments of individuals interacting as members of a social unit indicate that the operation of the general principle should be analyzed discerningly.

In both the psychophysical laboratory and the field ratings of status, the extreme items (objects or persons) serve as anchors, as revealed by reduced variability in judgments. However, lower variability in judging the top position than the bottom is a peculiarity of the human organization, not found in judgments of physical series. One might conclude from this that the general principle is not, after all, general. But another conclusion is that any general principle of human behavior must take account of the context and the particular properties of the stimulus situation. This is our conclusion.

The top rank position in a human group is a more important anchor than the lowest position because of the properties of human organization. These properties are not present in a series of weights, lines, or sound intensities. So the general principle is valid in both psychophysical and psychosocial judgments if the special characteristics of the stimulus situation are taken into account: in this case, the fairly equal importance of end items for ordering a series of physical stimuli, and the inequality of the leader and lowest positions in a human group for ordering those intermediate in position. This demonstrates one way of bridging the appalling gaps still prevailing between field studies and the more precise laboratory approach to the study of human behavior.


The individual's position in his group structure provides still another basis for prediction-the extent to which his behavior will fall within a range of variation defined as acceptable by other group members. This latitude of acceptance varies with the significance or importance of the activity for the group. Its limits are defined operationally by behavior indicating common disapproval, threats, or actual punishments.

In every group we have observed, the latitude for acceptable variation in individual behavior has been smallest for matters affecting the maintenance of the group and its solidarity in the face of outside threat (from other groups of age-mates, parents, school authorities, police, etc.). This narrowing of the limits on behavior in the face of outside threat has long been noted in "gangs," especially those involved in illegal activities. It is equally true for groups in activities which are not defined as socially deviant.

For example, one group observed in our research made a great point of "good sportsmanship" in encounters with other groups of age-mates. Nothing called forth disapproval, scolding, or threat of isolation so quickly as unnecessary roughness in a game with other boys or being a

( 312) "poor loser," despite the fact that the same behavior was accepted when these boys played among themselves. In several groups, the common desire to do things together conflicted with parental requests to do homework. The member who dared tell his parents that they were not actually studying together, but playing records, was treated as a child at least and in some cases as a traitor. Similarly, in a high rank area, the reputation of members as "smooth" but proper young men was zealously guarded for the benefit of the attractive "nice" girls and their mothers. Members who found a younger or less cautious girl willing to engage in sexual contacts were severely chastised if they made it a public matter, outside of their private circle.

In the groups we observed who engaged in definitely illegal activities, the boy who welched out through fear or who "squealed" was treated either to a "good beating" or to ostracism. In these cases, the maintenance of the group as a unit, unsullied by adult or police surveillance, made the norm of privacy in group activities essential. "Getting caught" at an illegal act (carrying a weapon or stealing, in these cases) was a reason for group censure, since it endangered other members. Two group leaders who were caught were both chastised severely by others for being foolhardy and taking risky actions.

Figure 20 shows the relationship found between the range of individual variations permitted and the relative importance of the activity for the group. The latitude of acceptance (ordinate) is narrowest in matters of greatest importance for maintaining the group and its standing as a unit. The latitude is broadest for those matters of least importance, particularly those involving activities strictly within group confines.

The relationship depicted in Figure 20 is further complicated, however, by the relative status of members in the group. Figure 21 shows two curves of the relationship between the latitude of acceptance (ordinate) and the relative status of group members (base line). In matters of minor importance to the group (top curve), the latitude of acceptance is greater for high status members, especially the leader, and smaller for low status members. In other words, especially within the group, the leader and his high status cohorts typically have considerable leeway in regulating behavior relative to others; it is the low man on the totem pole who is continually being nagged, laughed at, or barred from play.

On the other hand, in important matters (lower curve) the leader is expected to be exemplary; even slight deviations on his part arouse questioning, criticism, or scolding from other members. For example, in several groups where money is scarce, it was observed that the leader consistently pitched in with all of his funds or with more than the others, and that this was expected of him. If he indicated that he planned to spend it in other ways, he ran the risk of being called stingy. Similarly, the high status

( 313)

Figure 20
Figure 20. Range of individual variations acceptable in natural groups as a function of importance of activity

member with a car was expected to have it available for group activities, and conflicting plans on his part brought forth cries of "he thinks he's too good for us." In the same activities, however, a low status member could beg off without anyone expecting any more of him. He was saving money for shoes, or his sister needed the car. "What do you expect of him, anyway?"

To sum up: In observations of natural groups of adolescents, we have found a relationship between the person's status and the expectations others have for his behavior in various respects, as well as his expressed

( 314)

Figure 21
Figure 21. Conformity in natural groups as a function of member status and importance of activity.

attitude and actions. These relationships hold promise as a base for predicting adolescent behavior. Relative effectiveness of an individual in initiating activities for other members is most predictable at the top of the status structure (most effective) and at the bottom (least effective). Adherence to group norms (expressed as the range of variation in individual behavior which does not call forth sanctions from others) is most strict for all members in matters of importance to the group. However, the acceptable latitude varies with the person's status. The more significant the activity for the identity and continued maintenance of the group as a unit, the narrower the latitude of acceptance for all members, the latitude for the leader being smallest. The more incidental the activity, the broader the range of individual variations without arousal of sanctions, the latitude for the leader being greatest.



The larger social scene, the neighborhood with its prevailing values and facilities, and the informal social relationships with age-mates do not act independently in influencing adolescent attitude and behavior. The rest of the chapter summarizes generalizations from our findings, with illustrations, that show the interdependence of these influences. No one of these, considered by itself, is more important than another in every case. Our main generalization is that it is impossible to determine the more significant source unless all are included in the analysis.

Many of our generalizations have concerned differences between different settings. Therefore, we begin on a contrasting note: there are unmistakable similarities between youth in widely different settings in the United States.


Contrary to some popular theories about youth of different classes, we found that certain values and ambitions unmistakably marked the adolescent in the United States. Though differing in appearance, following different fashions in dress, sometimes speaking different languages, youth in all areas cherished an image of individual success as an adult, as it is spelled out through the magic symbols purveyed by television, advertising, movies, magazines, and popular books. This was true of members of groups studied intensively. The ingredients of the image were desired by nearly a hundred per cent of every sample of high school students studied: cars, comfortable homes, attractive clothing, appliances, telephones, radios, TV sets, money for entertainment, including movies. Regardless of what they have or what is available, youth in settings of low, middle and high rank want the "good things of life," as defined in mass salesmanship so characteristic of this country today.

Another finding common to youth in all areas was their orientation toward age-mates as the reference set. This was strikingly apparent in observations in every setting. The school samples, for example, asked why "school is fun"-three-fourths of the replies pertain specifically to the opportunities that school gave to interact with age-mates.

It should also come as no surprise to find, as we have, a common and intense interest in the opposite sex at this period. Observations of the small groups of boys are peppered with references to girls. In the surveys of school students, the interest is more hidden because of the tendency to give "socially desirable" answers (a problem mentioned earlier). However, it was readily revealed in the great majority of youth by their answers to the question, "What do you and your friends talk about?" (After starting to write, then erasing, the word "sex," one boy substituted "girls."

( 316) He was overheard to say "I don't want those people to think I'm a sex fiend.")

Another similarity was the ability to distinguish what adult authorities consider "right" and "wrong." Given a list of forty-five actions, youngsters in very different settings rated their acceptability-unacceptability in much the same way. These were teen-agers attending school, not dropouts. Therefore, it was gratifying to read a recent report of similar findings by Short and his co-workers (1963), who compared nondelinquent and officially delinquent youth on a written questionnaire. Of course, some youth in both samples may have violated the precepts in question, but it seems that violation because of ignorance of what society deems acceptable, or of what deviation is considered serious, is more rare than some theories of lower class youth or of psychopathology would lead us to believe.

In the small groups studied intensively in different neighborhoods, the finding most common to all was the boys' insistent desire to do things on "our own," without adult programming or supervision. Since these expressions and this active search for "freedom" are manifested in every area, the ecology and rank of each neighborhood become very important in determining whether the "freedom" is found in a friend's house, in one's car, behind a pool room, or in the streets.


Youth living in areas differentiated by socioeconomic level and cultural composition are well aware of these differences. As reported earlier

(Sherif and Sherif, 1964), we have found that those living in low rank areas freely express their desire to move to a high rank area, while those in high rank areas express contentment with their present circumstances and derogate the schools and type of persons found in lower class settings.

Differences between neighborhoods are apparent to the naked eye, of course. What is not so apparent is the effect that the differences in facilities have on a youth's actual mobility, his access to adult programs designed to improve his social mobility, and his ease in entering any situation peopled by those whose appearance and manner identify them as residents of an area differing substantially in rank.

Like other investigators, we found that youth in different neighborhoods set levels for personal goals according to the socioeconomic level of the neighborhood. However, we also found that the differences in goals are accompanied by differing conceptions of achievement. Figure 22, for example, presents recent data from one city, collected in areas which are upper middle, lower middle, and low according to the ShevkyBell index of socioeconomic rank. The low rank area is largely Spanishspeaking; the lower middle is mixed (Spanish-speaking with some Negroes); the upper middle area is mixed ("Anglos" with some Spanishspeaking persons).


Figure 22
Figure 22. Conceptions of subsistence, "average income," comfort, and personal goals for future income as a function of socioeconomic rank of neighborhood


The triangles in the figure represent the median values of personal goals for estimated income (per week) when one is adult and married. The personal goals vary, on the average, with the socioeconomic level of the area.

The figure also shows the median estimates obtained from the youth of the amount needed per week for subsistence (bottom bars), for comfort (circles), and the estimated income of the "average American family" (X). As the figure shows, youth in all three areas are "ambitious" according to their own conceptions of achievement, that is, according to their conceptions of enough to be "really well off" (circles). However, the upper middle rank conception of comfort is higher than the goals set, on the average, in the other areas.

These, and similar data on success in school and occupation, seem to belie the notion that middle or upper class youth are necessarily more ambitious or achievement-oriented than those in lower ranks. Those from areas of lower rank are ambitious relative to their own ideas of achievement, and even more ambitious relative to their own parents' achievements (cf. Sherif and Sherif, 1964).

The data on samples of youth in school do not represent completely the values and goals Of all youth in an area, because varying proportions have withdrawn or are barred from public school. A further finding on school youth in the various areas will help us in analyzing the social situation of those who withdraw from school.


In analyzing data obtained from youth in schools in the areas where our small groups were found, we were not entirely prepared to find that the least diversity of individual values and goals was found consistently in schools serving areas of higher socioeconomic rank. This discovery of so many like-minded students is a comment on the residential arrangements in large cities and on the cultural situation of youth in high rank areas. In these high rank areas, students had the least individual variation in personal goals, or conception of success in education and use of leisure time.

In both middle and low rank areas, greater heterogeneity was found in the values and goals of the student body, with actual cleavages into distinct social strata being characteristic of the middle rank schools (so classified on the basis of modal parental occupation and education). Even in schools serving areas of low rank, only in matters which are strictly financial (income, spending money, and the like) did the like-mindedness equal that of students in high rank areas.

Figure 23 represents the personal goals for education in three schools, in terms of years of schooling desired. The ordinate represents


Figure 23. Personal goals for education in neighborhoods of different socioeconomic status
Figure 23. Personal goals for education in neighborhoods of different socioeconomic status

percentage of the samples, and the curves give the distribution of goals in a school of low rank, lower middle rank, and upper middle rank in a city where free transfer is permitted. (These are the same three schools whose financial goals are represented in Figure 22.)

This figure represents one example of the greater diversity in values found in the lower rank areas. Ninety per cent of those in the high rank

( 320)

Figure 24. Personal goals for education in low, middle and high socioeconomic rank in city with median educational level
Figure 24. Personal goals for education in low, middle and high socioeconomic rank in city with median educational level

school (dash line) desire to complete college (16 years) or more advanced degrees (18 or more years). Greater diversity is evident in the low rank area (solid line), although the mode is for high school diploma. The distribution of responses in the lower middle rank area is strikingly bimodal (broken line).

( 321)

Figure 24 presents comparable data from a different city in which transfers are seldom granted, and the median educational level is substantially higher. The middle rank school (broken line) serves a large and diverse section of the city; but the low rank school is located in the oldest portion of the city, where most residential areas are of low socioeconomic rank, with a substantial Indian-Mexican-Negro minority. The distributions for both low and middle rank areas are clearly bimodal, while that for the upper rank area (dash line) again represents greatest homogeneity.

These findings have led us to suspect that some theorists on lower class life may have overlooked the actual diversity within lower class settings, perhaps because of their preoccupation with specific social problems. We suggest that generalizations about the characteristics and values of a class or ethnic grouping must specify carefully what values are involved, and how widely they are shared, before referring to distinct "subcultures" in different settings.

Specification of the homogeneity or diversity of the sociocultural setting proves helpful in understanding the nature of adolescent groups in that setting.


Let us start with a picture of what went on in the groups studied intensively in our research. What people do and talk about when they get together regularly of their own accord can tell us a great deal about them. The boys in our groups engaged in a great range and variety of activities. As noted previously, these invariably included girls, cars, and sports, in all but the least acculturated neighborhoods. During a considerable portion of the observation time, nothing seemed to happen: youth spend a lot of time hanging around, driving around if they can, while rehashing the past, planning an evening or a future scheme. The apparent aimlessness of much of the activity conceals active evaluations of things past and planning of things important to them. The discussions and plans revealed the boys' overwhelming desire to know themselves and each other as (near) adults, and to do things they wanted, on their own, without programming or intervention by adults.

Frequently, doing things on their own as adults involved activities deemed improper for the age level, even immoral or illegal. This was true in middle and high rank neighborhoods as well as low rank, even though most of the groups and their members were not labeled as "delinquent" by police.

All of the boys were concerned about their "manliness" (though this meant somewhat different things in different settings), with clothing fads, and with other cultural trends available in the larger setting. Each boy

( 322) translated these trends meaningfully for himself in relation to the others in his group. For example, the discussion and consumption of alcohol were so frequent that it seemed that drinking must be considered a sign of adult-ness for a young male-at least, the ability to talk about drinking.

We have not yet found a group which specialized in one sort of forbidden activity. On the contrary, groups emphasizing drinking not only tolerated non-drinkers but had many other activities. There were groups which engaged in theft (usually in twos or threes), and others which tolerated such activities by only two or three members. Observations of drug usage were remarkably similar to those involving alcohol. Although not all members participated, it was invariably a social affair.

The observations in these respects are difficult to reconcile with the sociological classification of groups into types on the basis of their activities. Particularly obscure, in terms of our observational material, is the classification of drug usage or alcohol as "retreatist." If one may infer any centralized purpose, our data imply that these activities are, on the contrary, ordinarily social-a means to enhance attraction and prestige (even though they may be carried to the point where the individual becomes "withdrawn" into unconsciousness or hallucinatory behavior, as in paint solvent or glue sniffing).

Activities labeled delinquent if detected by adult authorities were not distinctive to groups in any setting, nor the exclusive specialization of any one group. Nevertheless, frequent "delinquent" activities had decided consequences for the character of the group. Roughly proportional to the frequency of forbidden activities (not necessarily illegal ones), the groups became highly secretive. The frequency of activities involving such violations of adult dictums was, in turn, a function of the amount of time spent together; the wall of secrecy shielding members was the most impenetrable for groups which associated most frequently.

The total amount of time spent together, the secrecy of the group, and the absorption of individuals in their group activities to the exclusion of others were related to the homogeneity or diversity of the settings. By far the most homogeneous of the settings studied in our research were the least acculturated, low rank areas populated by persons of Mexican origin, and another low rank area populated by poor whites at considerable distance from the city center and from schools.

The one group of Negro boys studied in the Southwest were in a much more diverse setting, by virtue of proximity and contact with school youth and recreational agencies, which reflected the hope from the nationwide movement for equal opportunities. For example, Negro youth in school set their goals for education on a par with those of school students in the highest rank areas of their city, although not half of their parents had finished high school. Of course, these findings depend on circumstances in the cities in question, and cannot be generalized to other regions.

( 323)

Whether labeled delinquent or not, the groups in the most homogeneous settings of low rank were also the most solid and stable. The boys' perception of great differences between themselves and anyone outside "my neighborhood," inadequate and crowded homes, difficulties in achieving in school, real deprivations of physical needs-all of these circumstances combine to make the adolescent group the central place for each boy as an individual, where he can find pleasure and entertainment and, in some cases, necessities not provided elsewhere. These boys were the least mobile. Leaving the area involved considerable planning and effort, and often subjected them to the suspicious eyes of police.

Given these circumstances, the ecology of the immediate neighborhood and its facilities become weighty matters. Interference in using the facilities is eliminated only when the group defines a territory which is its own, which it is known to claim, and, hence, which represents the group's investment in prestige. When the boys went outside the territory, or when others came in, there were occasions of gang fighting, complete with weapons. Such incidents were recalled later with excitement and pride by members. They looked for accounts of them in the papers.

To assess this love of violence and bravado, it is essential to bear in mind that it was not directed to fellow group members, even though adolescent boys are not habitually gentle with one another. Nor was violence directed to members of those other groups with whom relations were friendly. The violence occurred in conflict with individuals defined by the boys as enemies, a capsule formula used to justify the actions of one's group in conflict, whether the opponent be another informal group, a club, the police, or a nation.

We cannot share the cynical view that violence is a prime value for adolescent boys, and that their efforts to avoid it are superficial. The difficulties of avoiding violence between conflicting groups are well known in adult life today. There is little reason to expect small groups of adolescents to succeed better, even when they genuinely desire not to get into trouble.

In fact, there were repeated observations that groups involved in conflict had made efforts to avoid it. Avoiding trouble was frequently discussed; precautions against it were quite as observable as the violent actions, though less dramatic. They "cased" a place for the presence and location of antagonists before entering. They avoided going into places where trouble might start. Joel Garza, one of the observers, recorded a half-hour episode during the intermission of a drive-in-movie, to which he had driven a group by no means noted for shrinking from conflict. A rival group was spotted between their car and the rest room; with considerable self-control, the members laid a strategy to use the facilities without encountering the rivals.

Much attention has been devoted to the pleasure, or even "thrills," which adolescent youth say they derive from violent encounters with an

( 324) enemy, from a forbidden joy-ride, or theft with one's fellows. We suspect from our observational data that these "thrills" have something to do with the effect of interaction among individuals who feel they "belong" together, who are open with one another, and who derive satisfaction from concerted efforts not available to them in the formal arrangements of living. Such exhilaration and "thrills" have also been reported by members of groups engaging in concerted activities toward desirable ends. In such a group situation, it is thrilling to win a ballgame, to elect a member to the student council, or to hold a party.


Most of the groups studied intensively in our research came from more diverse settings, which were characterized not only by greater variations in values but more opportunities for mobility and choice. In such settings, we find groups composed of individuals with differing backgrounds and aspirations; groups whose activities place them in opposition to the predominant values in their setting; and groups of boys directed outside of their immediate settings by the pull of opportunities seen elsewhere.

Using the prevailing values of the immediate neighborhood as a yardstick, we find that one group of boys in a lower rank neighborhood is quite atypical in several respects. Because of success in grade school athletics, these boys were imbued with the goal of "making" the high school team. Unlike the local "toughs" on their streets, they not only continued into high school, but chose a school far across the city because it had the best teams. Dubbed "Escuelantes" (school boys) by others in their neighborhood, they have not performed well academically and have riot made the school team. Most of their leisure time is spent together, warming up, practicing, talking about sports and glorifying the heroes of their school teams. It would be gratifying to report that the aspirations of this group reflected their parents' efforts, but their personal histories offer no such reflection. Three of the boys are illegitimate sons of servicemen, and none of the families provides the "solid, middle class" concept of achievement.

After being with this group for several months, the observer concluded that their chances of making the school teams were slim, for they were smaller than most varsity athletes. He reported that the intensity of their concentration on the school team and their own skills was so great that, at times, it seemed (in his eyes) to be almost fantasy. But the group process has kept the boys in school, and he predicts they will complete it.
By contrast, another group, composed of lower middle class boys in a predominantly middle rank area, had all dropped out or been expelled from school. In terms of their future hopes for jobs and income, the members proved to be quite typical of school students in their community, with their goals in these respects clustering around the median.

( 325) But their own initial encounters and continuing activities have focused on the pursuit of pleasure in cars, pool halls, and with girls. This kind of group activity required considerable outlays of spending money (over ten dollars a week), more money than 99 per cent of the students in their former school feel is needed. Together, these boys have their "fun" and secure spending money to support it by sporadic odd jobs. Meanwhile, the joint pursuit of such fun has resulted in several members being sent to "training schools" by the authorities. None of these boys is "mentally handicapped" in terms of standardized tests, and the parents of some have exerted strong but unsuccessful pressures on their sons to return to school and keep away from "the boys."

Still another group, which met regularly at a public center in a low rank area, was primarily recreational. In many respects, the members were diverse in their attitudes and aspirations. For example, two boys dropped out of school in 10th and 11th grade, two intended to graduate, one planned to go to further training, and one wanted to finish college. Although this was not a very solid group, its pull on members is so strong that the parents of the boy with college ambitions moved out of the neighborhood, where they had lived many years, in an attempt to break the association.

These examples are sufficient to indicate the variety of adolescent groups to be found in areas providing some diversity and, especially, some mobility to members. The immediate neighborhood, the family, the school, the images of success in society are not clearly "to blame" for the directions taken by any of these groups, yet they are all involved. To some extent the directions taken hinge upon the selective process of interaction among adolescents who, for the time, find some significant psychological identity within the group circle. The directional power of such groups must be included in the picture in order to understand why the Escuelantes are so directed outside their neighborhood, why the dropouts have resisted persuasive and corrective efforts to bring them into line, and why a recreational group seems threatening to the parents of a boy who wants to go to college.

Ordinarily, most of these groups last only a few years, until members leave the area, get jobs, go to school or military services, or marry. Their transitory nature makes their little "organizations" and their little "cultures" no less real, nor does it lessen their impact on individual members as long as they last.


The reality of the little "cultures" of adolescent groups may be seen more clearly if one could see what happened when at least some of the members "stayed put," did not leave the neighborhood, or did not enter into formal social relationships fundamentally different from those of adolescence. Such an opportunity arose in the research project in a neigh-

( 326) -borhood of low socioeconomic rank with a population of Mexican origin. The transmission of a group identity and its culture (norms) through three generations of adolescents was traced.

Early in the research project, a group of adolescent boys called "Los Apaches,"[1] was observed for several months by Eduardo Villarreal. Subsequently, he has carried through the observation cycle on an "older generation" of former members, ranging in age from 23 to 36 years, still living in the same area. Then he returned to find the current membership of Los Apaches, which had changed in the interim. Thus, it is possible to compare the norms and status relationships in three generations of members during a period when the neighborhood itself was essentially the same, but life in the city underwent considerable alteration (roughly, since the postwar period).

When first observed, Los Apaches consisted of twelve boys, with other occasional hangers-on. (See Sherif and Sherif, 1964.) Most of its members were known to police and detention home officials for stealing, carrying weapons, and engaging in fights with a rival group-the Lakesiders. Currently, nearly three years later, those who were members at that time have apparently escaped serious convictions. Some of them have jobs, are thinking of getting married, and are not active as Los Apaches members.

The observer could find very little change in the customs or norms of Los Apaches during this time. A "lieutenant" took over the leadership position when the old leader got a job and a steady girl whom he is thinking of marrying. Several new members have been added, some at high status levels. The rivalry with the Lakesiders continues, and seems to be even more important in group discussions. Occasionally, the members steal (always in twos or threes) ; a member of the older generation remarked, "These boys have more money than we used to get." Beer, solvent-sniffing, and marijuana are included at group get-togethers; there is always as much beer as money and ingenuity permit.

The young adults (now the older generation) meet regularly at favorite beer joints, a member's house, or night spots. They all quit school in the elementary grades, and all have unskilled jobs, with the exception of an auto mechanic and a shoe repairman. With very few exceptions, the observer reports, they seek sexual relations frequently while they are out together, even though most of them are married.

The members of the older generation belonged to Los Apaches of different times, when each was in his teens. Although the name Los Apaches was adopted when one of the younger adults was a member, all the men refer to the age-mate group by this name. They all still refer to to the Lakeside as the rival group, even though few of them now en-

( 327) -counter adults who were Lakesiders when they were Los Apaches. Those who did belong to the group at the same time reveal closer attachments than those who did not, as revealed by patterns and frequency of visits, lending money, or borrowing a car. A former leader (El Apache) is among the highest in status in this older generation group.

By observations over a considerable period of time, and at an arranged picnic for older and younger generations, it was ascertained that the older generation of young adults has very little contact with the recent or present membership of Los Apaches, except in the parental home of one member of each of the three generations. One other possible exception is that the present generation has, on occasion, sought help or a car from the older generation during conflict with the Lakesiders, which the older generation refused. The former leader (second generation, now almost twenty) and other older adolescents make moves to be accepted by the older generation. But the present membership (13 to 18 years old) preserves a respectful distance from their elders, in line with cultural expectations.

Older generation adults say that the present generation is not as tough as they were and does not get into as much trouble, chiefly because they have more money from parents and can borrow cars. But, they say, Los Apaches and Lakeside will "always" be rivals, and the younger members agree. All ages hold an implacable dislike for "perros" (police), but the older generation says the police are also less tough now than they used to be.

These are some changes in the group as seen through the eyes of former members. The observer reports other differences traceable to the changing times: The younger generation places a much higher value on education, although none has gone further than the ninth grade. The younger generation uses slang identical to that of the older generation (Spanish), but speaks English more frequently. Unlike the adults, who strongly prefer Mexican popular music, the younger generation both prefers and sings the current popular songs in English.

In short, as a miniature culture, Los Apaches has endured with minor changes, but has reflected the impact of broader culture change. The impact of the group may still be seen on the young adults who no longer see themselves as members, but in fact continue to interact in a group with a broader age range.


The groups studied in this research, formed by individuals of their own choice, are important to their individual members. Our evidence shows many instances of individual preference for group activities and values over those of family, church, school, girl friends, and even reasonable considerations of personal safety.

( 328)

It is also undeniably true that the informal groups of age-mates were of greater import for boys whose past history and present circumstances offered few other sources for a stable definition of self in relation to other human beings. These were not just boys from broken homes, for some had stable homes and were devoted to their parents. They were not only the poverty-stricken youth, for some had more than the necessities of life. They were not just school failures, for some were progressing in school. They were not all members of subordinated cultural minorities. When the entire set of such unfavorable circumstances prevailed in extreme degrees for some boys, they produced the strongest attachment to an adolescent group. However, these circumstances did not guarantee that a boy would belong or would be attached to his group.

Attachment to a group meant that the individual also had to be accepted by the others, live up to them, and prove himself a reliable and worthy member. Many boys who had unfortunate social circumstances in home, school, and community did not belong to a group because they could not achieve this necessary acceptance. Some had been expelled from groups.

The boy who was accepted, could live up, could prove himself reliable over a time, was usually among the more able, the more sociable, the more responsible to his fellows. Such personal attributes were necessary to gain status in any of the groups, regardless of other prerequisites for particular groups (for example, smoothness, or toughness, or daring). Having a secure place, especially a high place, among one's fellow-members further enhanced the importance of the association for the individual.

Thus, we have an apparent contradiction: Those to whom the group became most important as individuals, those most devoted to its tenets and outlook (whether socially desirable or undesirable) are likely to be individuals who display, within their groups, certain qualities usually prized by society. These qualities include loyalty, responsibility, and consistency in dealing with one's fellows. These very qualities may contribute to the individual's performance of antisocial actions directed toward outsiders or the larger community.

This generalization in no way condones those activities and norms of a group which are socially undesirable. If these are malicious, undesirable, or destructive, they should be denounced. But generalization does imply that adolescent attitudes and behaviors which are traceable to interaction in their reference groups cannot be defined within the traditional lexicon of psychopathology, or by new labels for the same pigeonholes.


In reporting on the research program, it has been our aim to emphasize the multifaceted nature of influences shaping social attitude and

( 329) behavior of individuals, for good or for evil. We have focused on participation in reference groups of age-mates because of a conviction that many aspects of the immediate and larger sociocultural setting are filtered to the individual adolescent through this highly salient medium.

The findings have shown that adolescent reference groups are patterned affairs, with power relations and "cultures" (norms) of their own. Attitude and behavior of particular individuals vary in terms of their places in these informal social schemes. The larger social scene, the immediate setting, and the age-mate group are interdependent influences in the life of particular adolescents. Especially when the immediate setting offers diverse alternatives, the interaction patterns and goals of the groups may become decisive in choosing among these alternatives.

For the single individual, participation in a reference group of age-mates may be beneficial for his future, or it may have disastrous consequences. In either case, an adequate theory of his attitude and behavior cannot rest with assigning labels of social approval or disapproval, mental "health" or "illness," to the individual. It is necessary to develop an adequate theory of individual behavior in close alignment with empirical findings on personal background, on associations with other humans who count in his eyes, and on the sociocultural setting. The social conditions in which he acts do not merely offer opportunities or hindrances to his strivings as a human individual; they also offer success pictures which affect what he will strive toward.


Edwards, A. L., 1957. The Social Desirability Variable in Personality Assessment and Research. (New York: Dryden Press.)

Sherif, M., O. J. Harvey, B. J. White, W. R. Hood, and Carolyn W. Sherif, 1961. Intergroup conflict and cooperation. The Robbers Cave experiment. (Norman, Oklahoma: The University Book Exchange.)

Sherif, M. and Carolyn W. Sherif, 1953. Groups in harmony and tension. (New York: Harper.)

———, 1964, Reference Groups: Exploration into conformity and deviation of adolescents. (New York: Harper, Row.)

Short, J. F., R. A. Tennyson, and K. L. Howard, 1963. Behavior dimensions of gang delinquency. Amer. Sociol. Rev., 28, 411-428.


  1. Like other identifications in published reports of this research, this name is fictitious.



Valid HTML 4.01 Strict Valid CSS2