Stereotypes

BIBLIOGRAPHY

The term “stereotype” originated in the technology of printing, where it has a clearly defined meaning. A body of type is set up; then a mold is made from this type, and a solid metal plate is cast in the mold. This metal plate is the stereotype. Its printing surface is precisely equivalent to that of the original type. The major purpose of stereotyping is to produce a printing surface that can be used for thousands and thousands of impressions without needing to be replaced. Thus, the adjective “stereotyped” has come to mean “mechanically repeated” or—in a broader usage—“hackneyed” or “trite.”

During the past forty years the noun “stereotype” has been widely used as a social science concept without ever being precisely defined. Its usage was introduced by the American journalist Walter Lippmann in a book called Public Opinion (1922). The major thesis of this book is that in a modern democracy political leaders and ordinary citizens are required to make decisions about a variety of complicated matters that they do not understand. People believe that their conceptions of German soldiers, Belgian priests, or American Ku Klux Klansmen, for example, are accurate representations of the real members of these classes. Not so! said Lippmann. The conception in most cases is actually a stereotype acquired by the individual from some source other than his direct experience. The situation is not usually improved even by direct experience with the social object corresponding to the stereotype because (according to Lippmann) people see mainly what they expect to see rather than what is really there.

Lippmann's book was much admired by social scientists, and in many of their writings (especially textbooks) the term “stereotype” has continued to have essentially the meaning Lippmann gave it. When a concept is referred to as a stereotype, the implication is that (1) it is simple rather than complex or differentiated; (2) it is erroneous rather than accurate; (3) it has been acquired secondhand rather than through direct experience with the reality it is supposed to represent; and (4) it is resistant to modification by new experience. [See the biography of LIPPMANN.]

It is unfortunate for the development of social science that attention has been focused on the noun “stereotype” rather than the adjective “stereotyped.” When one considers the adjective it becomes obvious that in the realm of concepts (though not in printing technology) stereotypy is a matter of degree. It is also fairly clear that the extent to which an individual's conception of, say, “the German soldier” is stereotyped is not a unidimensional variable. The four characteristics of a stereotype described above can vary independently; as a matter of fact, there is no solid evidence that they are even positively correlated. In these circumstances there would be little scientific value in setting up arbitrary criteria for the extent to which a concept had to have these characteristics in order to be termed a stereotype. It would be more profitable instead to treat the dimensions of stereotypy as quantitative variables and to investigate their variation from individual to individual, from concept to concept, and from situation to situation.

Empirical research . There has actually been very little of such systematic investigation. The only dimension that has been seriously studied is the one of resistance to modification by new experience [for a discussion of this research, see Attitudes, article on Attitude change; Persuasion]. On the other hand, there have been very few attempts to set up criteria for classifying an individual's concepts in a particular area into “stereotypes” and “nonstereotypes.” In empirical research the term “stereotype” has usually been employed simply as a pejorative designation for “group concept.”

This “group concept” usage became established in a classic study by Katz and Braly (1933). A group of 100 white American college students were asked to select from a list of 84 traits those they considered characteristic of each one of ten ethnic groups; then they were asked to choose the five “most typical” traits for each group. An index of definiteness of stereotype was constructed by counting the least number of traits required to include 50 per cent of the 500 choices by all subjects. This index ranged from a minimum of 4.6 for Negroes to a maximum of 15.9 for Turks. The six most frequent characterizations of each group were as follows (figures in parentheses indicate the number of students choosing a trait as one of the five “most typical”)—Negroes: superstitious (84), lazy (75), happy-go-lucky (38), ignorant (38), musical (26), and ostentatious (26); Turks: cruel (47), very religious (26), treacherous (21), sensual (20), ignorant (15), and physically dirty (15). Both Negroes and Turks were viewed very unfavorably by these students, but there was considerably more agreement on the characteristics attributed to Negroes.

The Katz and Braly procedure has been repeated many times, for many different ethnic groups, and in many different countries. In one of the more recent studies, a list of 99 adjectives was submitted to a group of 100 Arab students in Beirut, Lebanon (Prothro & Melikian 1954). These adjectives, in Arabic, had been selected from a longer list developed by other students at the same university to characterize members of various ethnic groups. The Arab students characterized Negroes in a manner similar to American students twenty years earlier, but their ratings of Turks were entirely different. The six traits most frequently chosen as “most typical” of Turks were strong (36), militaristic (33), nationalistic (33), courageous (31), progressive (18), and arrogant (17).

Since 1933 the paper-and-pencil questionnaire has been the typical method for investigating stereotypes. Most studies have dealt with beliefs about ethnic groups, but a substantial number have investigated beliefs about occupational groups, social class groups, the two sexes, etc. What conclusions can be drawn from all this research?

One conclusion is that most individuals feel able to make at least a guess about the characteristics of almost any defined social group on the basis of information that a social scientist would consider quite inadequate. Opinions are picked up from other individuals, from the mass media, and—to some extent—from direct personal contact. There are some highly conspicuous national groups— Americans, Russians, French, British, Chinese— about whom the majority of the world's citizens seem now to hold definite opinions.

To what extent are these opinions correct? A great many social scientists accept the “kernel of truth” hypothesis (Klineberg 1950), which asserts that if we could determine objectively and accurately the characteristics of a defined social group and that if we ascertained the beliefs of some other social group about the first, we would find a more than random correspondence between the two sets of characteristics. There will always be some individuals whose beliefs about members of the defined social group are almost completely accurate and others whose beliefs are more wrong than right. However, according to the “kernel of truth” hypothesis, the over-all amount of truth in stereotypes is greater than the amount of error.

This hypothesis is clearly impossible to test in its general form. What can be tested are hypotheses about the accuracy with which specific sets of characteristics are attributed to the members of group X by the members of group Y—provided we can determine the true distribution of these characteristics in group X. For obvious reasons, little research of this sort has been done, and the findings have been highly divergent. In at least one study of attitudes toward an ethnic group (Armenians in California) there was a negative relationship between the actual characteristics of the group and the stereotype of it held by nonmembers (LaPiere 1936).

There are many hypotheses about the circumstances under which stereotypes (in the sense of group concepts) are likely to be accurate or inaccurate. One of the most plausible is that of Roger Brown (1965), who argues that beliefs about members of a particular social group are most likely to be accurate if the group consists of people playing a defined social role—for example, the members of a particular occupation, caste, or sex. In this case, “what is prescribed for the category is ordinarily performed by the category and expected from the category” (ibid., p. 172). Social life would be almost impossible if this were not so.

A very widely held belief is that the stereotypes held by educated people are in general more accurate than those of the uneducated and that the concepts of social scientists are most accurate of all. This is certainly Lippmann's view, and it seems very reasonable; however, it has not been demonstrated.

When members of two different ethnic groups are asked to rate both themselves and each other in the Katz and Braly manner, two things usually occur. One is that the array of characteristics selected by group X as most typical of itself is similar to the array of characteristics selected by group Y as most typical of group X. The other is that socially desirable characteristics are more likely to be emphasized in a group's description of itself, while undesirable characteristics are more likely to be emphasized in the description of a group by members of another group—especially if there has been a recent history of conflict between the two groups. A representative study is that of Reigrotski and Anderson (1959), in which large numbers of Belgians, Dutch, French, and Germans were asked to characterize both themselves and the other three nationality groups. From a list of 12 adjectives the ones most frequently selected by the Germans to describe themselves were (in order): hardworking, brave, intelligent, practical, progressive, and peace-loving. On the other hand, the French thought the Germans were most characteristically: domineering, hardworking, cruel, progressive, brave, and intelligent.

When there is agreement between a group's image of itself and the image that members of a second group have of the first, this is usually taken as evidence for a substantial “kernel of truth” in both sets of stereotypes (e.g., Vinacke 1949). The tendency of members of a group to see themselves more favorably than members of other groups see them is generally considered to be a reflection of ethnocentrism; this tendency has been widely demonstrated. In the case of low-status groups, however, such as the American Negroes, lower-caste Hindus, and people in depressed areas, even their own stereotypes of themselves may be unfavorable (Bayton 1941; Bayton & Byoune 1947; Rath & Sircar I960; Hughes et al. 1960, pp. 244-311).

Although in the great majority of studies the term “stereotype” has been used to mean “group concept,” with no attempt to determine either the adequacy of the concept in representing its object or the quality of thinking of the person using the concept, there has been some research in which the investigator tried to study stereotypes in Lippmann's sense of the term. Most of these studies have dealt with ethnic stereotypes, and the major conclusion has been that the more unfavorable an individual's attitude is toward members of a particular group, the more likely are his concepts of such group members to be stereotypes in Lippmann's sense (Bettelheim & Janowitz 1950; Adorno et al. 1950; Secord 1959).

The confusion in the meaning assigned to the term “stereotype” by different authors, especially the contrast in usage by the majority of empirical workers and the majority of textbook writers, has led to a good deal of discussion of the basic concept or concepts involved in the term. Two of the best articles on this subject are those of Fishman (1956) and Vinacke (1957). In the opinion of the present writer, the broad and undiscriminating usage of the noun “stereotype” is now too well established to be dislodged.

The extreme diversity of research on stereotypes has inhibited authors from attempting to review the entire field. (A review and bibliography of research on ethnic stereotypes can be found in Klineberg 1950 and in Harding et al. 1954.)

John Harding

[Directly related are the entriesIdeology; Prejudice. Other relevant material may be found inAttitudes; Concept formation; Ethnic groups; Race; Race relations.]

BIBLIOGRAPHY

Adorno, T. W. et al. 1950 The Authoritarian Personality. American Jewish Committee, Social Studies Series, No. 3. New York: Harper.

Bayton, James A. 1941 The Racial Stereotypes of Negro College Students. Journal of Abnormal and Social Psychology 36:97-102.

Bayton, James A.; and Byoune, Ethel 1947 Racio-National Stereotypes Held by Negroes. Journal of Negro Education 16:49-56.

Bettelheim, Bruno; and Janowitz, Morris (1950) 1964 Social Change and Prejudice, Including Dynamics of Prejudice. New York: Free Press. ? A reprinting of the authors' Dynamics of Prejudice, with a reassessment of its findings.

Brown, Roger W. 1965 Social Psychology. New York: Free Press.

Fishman, Joshua A. 1956 An Examination of the Process and Function of Social Stereotyping. Journal of Social Psychology 43:27-64.

Harding, John S. et al. 1954 Prejudice and Ethnic Relations. Volume 2, pages 1021-1061 in Gardner Lindzey (editor), Handbook of Social Psychology. Cambridge, Mass.: Addison-Wesley.

Hughes, Charles C. et al. 1960 People of Cove and Woodlot: Communities From the Viewpoint of Social Psychiatry. The Stirling County Study of Psychiatric Disorder and Socio-cultural Environment, Vol. 2. New York: Basic Books.

Katz, Daniel; and Braly, Kenneth 1933 Racial Stereotypes of One Hundred College Students. Journal of Abnormal and Social Psychology 28:280-290.

Klineberg, Otto 1950 Tensions Affecting International Understanding: A Survey of Research. Social Science Research Council, Bulletin No. 62. New York: The Council.

Lapiere, Richard T. 1936 Type-rationalizations of Group Antipathy. Social Forces 15:232-237.

Lippmann, Walter (1922) 1944 Public Opinion. New York: Macmillan. ? A paperback edition was published in 1965 by the Free Press.

Prothro, EdwinT.; and Melikian, Levon H. 1954 Studies in Stereotypes: III. Arab Students in the Near East. Journal of Social Psychology 40:237-243.

Rath, R._; and Sircar, N. C. 1960 The Mental Pictures of Six Hindu Caste Groups About Each Other as Reflected in Verbal Stereotypes. Journal of Social Psychology 51:277-293.

Reigrotski, Erich; and Anderson, Nels 1959 National Stereotypes and Foreign Contacts. Public Opinion Quarterly 23:515-528.

Secord, Paul F. 1959 Stereotyping and Favorableness in the Perception of Negro Faces. Journal of Abnormal and Social Psychology 59:309-314.

Vinacke, W. Edgar 1949 Stereotyping Among National-Racial Groups in Hawaii: A Study in Ethnocentrism. Journal of Social Psychology 30:265-291.

Vinacke, W. Edgar 1957 Stereotypes as Social Concepts. Journal of Social Psychology 46:229-243.