Classics in the History of Psychology

An internet resource developed by
Christopher D. Green
York University, Toronto, Ontario

(Return to Classics index)


Kenneth B. Clark and Mamie K. Clark (1939)

First published in Journal of Social Psychology, S.P.S.S.I. Bulletin, 10, 591-599.

Speculations concerning the nature of the self and the development of consciousness of self have long been a significant part of psychology. Little experimental research has been done, however, in an attempt to raise this problem out of the maze of speculative conjectures. Piaget (9), elaborating upon Baldwin's (3) concept of a three stage development of "personal consciousness" (transitions from a "projective" to a "subjective" to an "ejective" sense of personality -- the final "ejective" stage being the social self wherein the child is aware of the fact that other people's bodies have similar experiences to his own), maintains that the child comes to discover himself through a progressive comparison of his own body with other people's bodies. Concerning the psychical qualities of self, he states: "In the same way, with regard to psychical qualities, it is by imitating other peoples' behavior that the child will discover his own."

Lewin (7) emphasizes the formation of the concept of property as a feature of the development of self consciousness .

. . . The "I" or self is only gradually formed, perhaps in the second or third year. Not until then does the concept of property appear, of the belonging of a thing to his own person.

G. W. Allport (1), on the other hand, speaks of an earlier consciousness devoid of self reference, while self consciousness develops only after the age of four or five years.

Until the child has a fairly definite conception of himself as an independent person, he cannot conceptualize his relationship to the surrounding world and hence lacks the subjective nucleus for the development of his own personality . . . . Even at the age of four or five the self, is by no means firmly encapsulated . . . . The advent of self consciousness is gradual, and its growth continuous, but a certain critical stage is reached around the age of two . . . . Its symptom is the period of negativism. [p. 592]

Some experimental work on the problem has been done. Bain (2) made daily observations and records of a child's speech in normal family situations from birth to one and one-half years. From an analysis of the "self" and "others" words of the child he concluded that self is social; it appears and develops rapidly and observably from five months on and begins to be verbalized after about the first year. Bain said of the child:

It is out of his responses to others that his "consciousness of self" arise, together with appropriate verbal symbols for naming it.... The "I" is a social concept. It is quite different from the concept of self as object which arises much later.

Goodenough (4), assuming the use of the first personal pronoun to be evidence of having reached a primitive stage in the development of self-awareness, studied the use of certain specified pronouns by children in free play and control situations. Pronouns of the first person singular were used with far greater frequency during free play with other children than in the controlled situation. The same trend was shown for the possessions my and mine.

Insofar as the use of these pronouns is indicative of something in the nature of an ego-consciousness, it is evident that this feeling is brought to the fore in the more competitive situations of group play far more frequently than is the case during the less socialized conditions of the controlled situation.

Moreno's (8) study is of interest in that it approaches the problem of the development of consciousness of self from the point of view of the dynamics of group development. He used a sociometric test wherein children chose from pictures of various racial groups the boy or girl whom they would like to have sit on either side of them. First and second choices were given. Moreno found that gradually from the first grade on the group develops a more differentiated organization . . . . From about the fifth grade .. . . a greater number of Italian children begin to choose Italian neighbors . . . . a larger number of white children reject colored children . . . . It indicates the beginnings of a racial cleavage.

Attacking the problem of consciousness of self from the point of view of determining the spatial localization of the self, E. L. Horo - [p. 593] witz (5) found that he was unable to present a final statement describing where the self actually is. He concludes that

the localization of the self . . . is not the basic phenomenon one might hope for to ease an analysis of the structure of the self and personality.

More recently, R. E. Horowitz (6) has been concerned with the problem of children's emergent awareness of themselves, with reference to a specific social grouping. Her study dealt with race consciousness conceived as a function of ego-development. Her procedure is described in detail because of its similarity to the one used in the present study. Two picture techniques were used with 24 children from two to five years of age: (a) Choice tests -- one pair of photographs for boys and girls respectivelv, showing a white child and a Negro child; one pair of line drawings showing a white boy and a Negro boy; one set of four-line drawings showing a white boy, a Negro boy, a chicken, and a clown. Boys were asked in each case to identify themselves. The form of the question was "Show me which one is you. Which one is " (using name of the subject). The girls, after having identified themselves in the first item, were asked to identify brothers or cousins in the three boy's items. (b) The Portrait Series -- 10 portrait pictures were exposed, one at a time. Children were asked, "Is this you? Is this ?" (using name of child). This latter technique did not give as satisfactory results as the first. In the Choice Tests Horowitz found that 68.4 per cent children were "correct" and 31.6 per cent were "incorrect." In the second series of line drawings more errors on the whole were made, "but the balance of errors was weighted in the same direction." In the third series of this technique, four of five Negro boys made "correct" identifications and three out of seven white boys made "correct" identifications.

The general limitations involved in the small number of cases are recognized by the author. She states: "Further work will, of course, have to be done to determine how common it is and within what framework of circumstance it operates."


The present study is an attempt to investigate early levels in the development of consciousness of self in Negro preschool children with special reference to emergent race consciousness. The term [p. 594] consciousness of self may be considered as awareness of self as a distinct person; as distinct from other groups of things or individuals. The term race consciousness is here defined as consciousness of self as belonging to a specific group which is differentiated from other groups by obvious physical characteristics. It is hereby assumed that race consciousness and racial identification are indicative of particularized self consciousness.


A modification of the Horowitz picture technique was used. There were three sets of line drawings as follows: Set A-one white boy, one colored boy, a lion and a dog; Set B -- one white boy, two colored boys and a clown; Set C -- two white boys, one colored boy and a hen. Combining all the line drawings there were four white boys and four colored boys. Each of the four pairs of white and colored boys was alike in every respect save skin color. The same white and colored boy never appeared in any one set of the pictures.

Materials and instructions were presented in the same manner as in Horowitz's investigation. "Show me which one is you. Which one is _______?" (using name of subject); with girls "Show me which one is ________?" (using name of brother, boy cousin or boy playmate). Subjects were examined individually.

Subjects: 150 Negro children in segregated Washington, D. C., nursery schools (75 male and 75 female-50 three-year-old, 50 four year-old, and 50 five-year-old children). These children were taken from five W.P.A. nursery schools, one private nursery school, and one public school kindergarten.


Table 1 presents the choices of subjects on the total picture series. On the entire series of pictures the total group of 150 Negro

[p. 595] children made more choices of the colored boy (50.9%) than of the white boy (44.1%) (CR 2.13). This table, however, is meaningless from a genetic point of view, in that all age groups are combined. A mass presentation of the data completely disguises any factors which may be operative in the dynamics of self consciousness and racial identification.

It is apparent that there is a consistent increase in the differences between choices of colored and white boys with age. The significance of these differences increases thus: -2.7 per cent at the three year level (CR 0.4), to 10.9 per cent at the four-year level (CR 1.87), to 12.1 per cent at the five-year level (CR 2.08). These differences are in favor of the colored boy.

The absolute number of choices of colored boy increase from the three-year level (41.2%) to the four-year level (55.40/o) (CR 2.44) and slightly again at the five-year level (56.00/0) (CR 1.03). The number of choices of the white boy remain approximately the same at the three-, four-, and five-year levels.

Choices of the lion, dog, clown, and hen constitute 14.7 per cent of the total responses at the three-year level, but disappear at the four-year level and do not appear again at the five-year level. The increase in the percentage of choices of colored boy is at the expense of choices of the less relevant pictures of lion, dog, clown, and hen. Thus, beginning at the four-year level, these children cease to identify themselves in terms of the animals or the clown and consistently identify in terms of either the colored or white boys with a trend toward more choices of the colored boy.

The most significant aspect of the results presented in this table (Table 3) is the fact that the choices of the boys show significant trends whereas those of the girls seems to approximate chance. This fact can be best understood if it is remembered that the boys were [p. 596] making identifications of themselves while the girls were identifying brothers, cousins, and in a few instances a boy playmate. Because of this difference in response it would appear that either the technique used in this investigation has greater validity when used with boys than when used with girls, or that the dynamics involved when girls identify someone other than themselves is quite different from the self identification of the boys.

In view of the fact that the reason for this difference is at present unknown it is necessary to compare the choices of the boys alone with the choices of the total group. For choices of the colored boy by the males alone at each age level, there is a consistent increase is these responses with age from 31.5 per cent at the three-year level, to 60.8 per cent at the four-year level, to 63.0 per cent at the five-year level. This is in agreement with the general trend of results found for the total group. For choices of the white boy by the males, there is a consistent decrease in these responses with age from 50.7 per cent at the three-year level, to 39.1 per cent at the four-year level, to 36.9 per cent at the five-year level. For the total group, however, there is no such consistent decrease in the choices of white boy. An examination of Table 2 will show that for the three-, four-, and five-year level the percentages of choices of the white boy are respectively 43.9, 44.5, and 43.9. This stability is obviously due to the non-differential responses of the females.


It is clear cut from the results that a definite delimitation of the self on the part of these children occurs between the three- and four year age levels. The dropping out of irrelevant choices of the lion, dog, clown, and hen indicates the attainment of a developmental level where consciousness of self is in terms of a distinct person. [p. 597]

This is undoubtedly a precursory level of development to the consciousness of belonging to one group as distinct from another. This latter contention appears to be justified by the increase in the number of choices of colored boy over white boy with age on the part of the total group, and is shown even more clearly in results for the males alone. The dynamic aspect in this development of self consciousness is even more apparent if one conceives of this increase in the number of choices of colored boy over white boy with age to be an indication of the emergence of the still higher level of personal racial consciousness.

The fact that definite age trends in increased choices of colored boy were evident from the three-year to the four-year level but, while continuing the trend, were not as definite (statistically significant) from the four-year to the five-year level indicates the probability that this technique is inadequate when used with higher age levels.

An alternative explanation of this finding would assume that it is an indication of the facts as they are; that the greatest (most significant) amount of development in self consciousness and racial identification, occurs between the third and fourth years. After the fourth year there is relatively little development of the mechanisms operative. This explanation would assume that the ceiling had been approached if not actually reached. Obviously there must be a ceiling, but the data and incidental experiences of the investigators do not seem to warrant the assumption that the ceiling had been reached in this study by the technique used. In support of this belief is the fact that a few of the five-year old children refused to identify themselves with any picture, saying "I'm not on there," or "That's not me," or "I don't know them," etc. Some were hesitant, evidently because they thought the same. Some five-year olds said before making identifications on the first set of line drawings: "This is a white boy, this is a colored boy, this is a lion, and this is a dog." These responses gave an inkling of the fact that these five-year-olds were developing ideas of themselves as intrinsic individuals. It appeared to be a conflict with this idea for them to identify themselves with either the white or the colored boy, just as most of the three-year-olds and all of the four-year-olds had not identified themselves with any of the irrelevant drawings.

A more refined technique, which would be as sensitive for the [p. 598] five-year-olds as this one is for the three- and four-year-olds, would undoubtedly yield more valid information concerning the operation of this mechanism in the older children. Identification of self from line drawings seems to be too great an abstraction of an emergent concrete entity for the five-year-old boys. The hesitancy in interpreting self in terms of the line drawings, but rather conceiving of it as an intrinsic, concrete entity, is suggestive of another stage in the development of self consciousness.

The fair degree of significance of the male responses and the seeming chance responses of the females indicate that the line drawings used in this study should have been used exclusively with males for greatest validity. Line drawings of girls should have been used with the girls. The dynamics involved in identification of a brother or cousin on the part of the girls is obviously different from those in identification of one's self. Further data on the problem will appear in a later paper.


In an effort to get some indication of the nature of development of consciousness of self in Negro preschool children, with special reference to emergent race consciousness, 150 Negro children in segregated schools were shown a series of line drawings of white and colored boys, a lion, a dog, a clown, and a hen and asked to identify themselves or others. The results were as follows:

The total group made more choices of the colored boy than of the white boy.

The ratio of choices of the colored boy to choices of the white boy increased with age in favor of the colored boy.

Choices of the lion, dog, clown, and hen were dropped off at the end of the three-year level, indicating a level of development in consciousness of self where identification of one's self is in terms of a distinct person rather than in terms of animals or other characters.

The seeming chance responses of the girls warrants further study of girls making identifications of themselves on similar line drawings of girls.

The fact that the sharpest increase in identifications with the colored boy occurred between the three- and four-year level and failed to increase significantly at the five year level suggests that either this picture technique is not as sensitive when used with five-year-olds [p. 599] as when used with three- and four-year-olds, or that a plateau in the development of this function occurs between the ages of four and five, or that the five-year-olds have reached a stage in self-awareness which approaches a concept of self in terms of a concrete intrinsic self, less capable of abstractions or external representations.


1. Allport, G. W. Personality: A Psychological Interpretation. New York: Holt, 1937. (Pp. 159-166.)

2. Bain, R. The self-and-other words of a child. Amer. J. Sociol., 1936, 41, 767-775.

3. Baldwin, J. M. Social and Ethical Interpretations in Mental Development. New York: Macmillan, 1897. (P. 7.)

4. Goodenough, F. L. The use of pronouns by young children: A note on the development of self-awareness. J. Genet. Psychol., 1938, 52, 333346.

5. Horowitz, E. L. Spatial localization of the self. J. Soc. Psychol., 1935, 6, 379-387.

6. Horowitz, R. E. Racial aspects of self-identification in nursery school children. J. of Psychol., 1939, 7, 91-99.

7. Lewin, K. Dynamic Theory of Personality. New York: McGrawHill, 1935. (Pp. 106.)

8. Moreno, J. L. Who Shall Survive? Wash., D. C.: Nervous and Mental Disease Publishing Co., 1934. (P. 61.)

9. Piaget, J. The Moral Judgment of the Child. New York: Harcourt, Brace, 1932. (P. 393.)