The Psychology of 'Attitudes':[1] Part I

Muzafer Sherif and Hadley Cantril[2]
Princeton University


During the past two decades the problem of attitudes has become central in social psychology. Thus, G. W. Allport writes, "The concept of attitudes is probably the most distinctive and indispensable concept in contemporary American psychology. No other term appears more frequently in experimental and theoretical literature" (2, p. 798). Murphy, Murphy and Newcomb, in their monumental volume summarizing the state of social psychology in 1937, again emphasize the point: "Perhaps no single concept within the whole realm of social psychology occupies a more nearly central position than that of attitudes" (97, p. 889). We need not multiply these representative statements from other sources to demonstrate the important position the concept of attitudes holds in contemporary social psychology.

It is significant that sociologists feel as much at home in the study of attitudes as do psychologists, so that the study of attitudes is by no means simply the concern of psychologists alone. Some sociologists have gone so far as to equate social psychology with the study of attitudes (14, 51, 123). In the first decade of this century, attitudes also became a focal problem in experimental psychology. After the failure of the introspective analysis of the higher mental processes at Würzburg, descriptions were given in 'attitudinal' terms such as 'Einstellung' and 'Bewusstseinslage' (105, 134). With this prominent start in experimental psychology, attitudes came to stay as an important concept in that field.

In this paper we are proceeding with the conviction that experimental and social studies on attitudes have much to contribute to one another in bringing about a unified psychology of attitudes. The psychologist's task is to give an adequate account of the psychological mechanisms involved in the formation of an attitude in any individual. The formation of a social attitude in the individual should he essentially the same as the formation of any attitude if the explanation of this formation is to be psychology at all. We are here groping in this direction. As

( 296) we shall see, attitude studies at present do not give us a unified picture. In fact the problem of attitudes is in a very confused state. Perhaps the con fusion has increased in proportion to the wealth of literature: different investigators have had different interests and points of view in approaching the problem.


When we look at any society, whether primitive or highly developed, whether simple or complicated, we observe conformities of behavior, within the limits of variations due to individual differences, on the part of the individual members of any society as they carry on the daily business of living-for example in regulating instinctive activity, dress, likes or dislikes of other groups, or responses to events which have social significance. When the established conformity of any particular individual member is taken for investigation in any of these matters, that is to say, when analysis and explanation are carried to the psychological level, the problem becomes primarily one of the psychology of attitudes.[3]

We have used the term ‘established conformity’ very deliberately. For these established conformities result from conformity to social standards or norms which have come into existence in a deterministic way as a consequence of the interaction of individuals in the all important business of living. Individual members within a society come to acquire these established conformities of experience and behavior within the limits of their individual differences.[4] We may, in fact, say in a summary way that the socialization which occurs when an individual becomes a member of a group consists mainly in the achievement of conformity in experience and behavior to social values, standards, or norms already established. And the process of achieving conformity is, if we analyze it closely, nothing more nor less than the formation of appropriate attitudes in relation to these socially standardized values or norms or other criteria of conduct.


The approach to the psychology of attitude has been many-sided. The early work of the Würzburg laboratory gave the historical setting to more recent work on Aufgabe. In 1918 Thomas and Znaniecki stimulated other sociologists to analyze the concept of attitude further as a useful tool in the explanation of the phenomena with which they dealt. The work of cultural anthropologists such as Boas, Malinowski and Sapir, and of many contemporary ethnologists, highlights the variation in cultural norms and the consequent differences in the attitude of individuals living under different social systems.

Psychologists beginning with Thurstone and continuing with Likert, Droba and others worked out procedures for measuring attitudes. In recent years the attitudes of all kinds of people toward almost all conceivable subjects have been measured. Other investigators such as G. W. and F. H. Allport, Katz, Bain, Faris and Lasker, concerned themselves with the genesis and

( 297) nature of attitudes. More recently, the interest in measuring attitudes has given way to the measurement of public opinion by the use of stratified samples of different populations. While the techniques for measuring attitudes and opinion can be usefully employed to gather valuable information on attitude determinants and properties, and while the data obtained from such measurement often have a high practical, strategic, or systematic value, the great bulk of quantitative research has been designed for the primary purpose of measurement alone, not systematic understanding.

After an exhaustive review of the literature on the psychology of attitudes, Nelson reports his impression, shared by others, "of the wide variety of meanings which are ascribed to this term" (100). He cites 23 rather distinct characterizations given the term 'attitude' by psychologists or social scientists up to 1939, ranging from ‘organic drives,' ‘neural sets,' or ‘trial responses,' to ‘ways of conceiving objects,' or sum totals of ‘inclinations, feelings, notions, ideas, fears, prejudices, threats, and convictions about any specific topic.' In a recent paper concerned with the definition and use of ‘attitude' in social psychology, Strauss points out that "the concept, despite its key position, is marked by considerable confusion" (122, p. 329). He notes that much of the research on attitudes has little or nothing to do with attitude theory and that the use of attitude as a ‘common sense explanation' rather than as a 'genuine causal explanation' (p. 334) has retarded systematic understanding. It is therefore not surprising that some strict experimentalists--dissatisfied with what they may regard as the ‘practical' or unsystematic nature of attitude research in social psychology--look down on it and on social psychology in general as having little to do with pure science. They may take it as an example of why the pure scientist must follow Titchener's dictum and stand "apart from the great majority of his fellow men," disavow interpretation as something entirely ‘foreign to him' and move "in the domain of bare existence" (124, pp. 69 f.).

The laboratory and everyday life.---Before proceeding with a discussion of the psychology of attitudes, it is necessary to clear up the relationship between pure experimental research and the systematic study of attitudes in everyday life situations. A psychological construct-if it is to prove valid and adequate-must be as valid and adequate in handling the stuff of ordinary human affairs as in handling the controlled variables of the laboratory experiment. Various writers have fallen into the error of making a distinction for one reason or another between 'psychology' and ‘social psychology.’ Klineberg, for example, makes such a distinction "for purposes of convenience" (78, p. 4), and puts his emphasis on a demonstration of "the wealth of social patterning" rather than on the "constants of human behavior" (p. 8).

Science consists of a set of conceptual constructs which have high predictive value. If a psychologist is to make any claims that will have scientific validity, he must obviously be as objective as possible. This does not mean, however, as Titchener felt, that a psychologist must be ‘disinterested and impersonal.' The highbrow superciliousness of those who hold that a true ‘scientific' psychologist should not and cannot deal with the concrete realities of men in their social life comes from their confusion of the fact that scientific objectivity derives from its methods and not from the aloofness of its observers (26). William James long ago pointed out that psychology is scientific to the extent that it uses methods which make veri-

( 298) -fication possible, irrespective f the theories, biases, or prejudices of the experimenter. "The most useful investigator, because the most sensitive observer," said James, "is always he whose eager interest in one side of the question is balanced by an equally keen nervousness lest he become deceived" (70).

Since the scientist's objectivity derives from his methods rather than from his own interests or from the type of material with which he deals, and since the psychologist's goal is to understand the thought and behavior of men in real life situations, the scientific psychologist becomes obligated, rather than forbidden, to formulate concepts that will stand up both in the laboratory and in everyday life (21). His task is, of course, particularly difficult since the determinants of thought and action in real life are so complicated, the context within which they function so varied and changing. Nevertheless, the search for adequate constructs must proceed if psychology is to become anything but the nasty little discipline James once dubbed it.

If the psychologist becomes fearful of the consequences of crossing a borderline which he feels separates his scientific laboratory research from its application to real life problems, he should remember that the concepts of chemistry, physics, and biology hale been developed by frequent crossing back and forth. Theoretical concepts arrived at in the laboratory hale been tested for their adequacy in normal, non-laboratory contexts. In the history of the physical sciences, real life situations hale forced modifications and revision, stimulated further research which has produced more adequate constructs. Julian Huxley, for example, has pointed out the interdependence of application and theory in contributing to the advances of the physical sciences (68). Giving examples of research on high voltage tensions and its use in the transmission of electricity, he notes that "sometimes it is not very clear what is pure and what is applied, or in which direction the current is flowing."

Fortunately for psychology, most of those now working in the field are becoming increasingly aware of the challenge which the explanation of everyday life situations poses to their scientific interest. We mention the problem here because it is essential to our argument that any final conceptualization of the nature of attitudes must make quite explicit the mutual dependence of data obtained from the laboratory and data obtained from real life situations. In other words, the nature of attitudes is not two problems--one for the laboratory and one for everyday life-but one single problem.

The concepts or variables thus formulated will eventually make it possible to construct a psychology of attitudes which can be used to study any kind of attitude, social or non-social, in this or any other social system. For in spite f cultural diversities, psychological laws ought to be the same for individuals in any social system. Otherwise, logically speaking, we should be arguing-even if we do not mean it---some sort of ‘cultural racism,' similar to biological racism. If psychological laws are not the same for all individuals, then the German Kultur apologists were right in arguing that the members of their unique Kultur could be understood-not explained-only by the peculiar logic of their Kultur.[5]

( 299)

Evaluation of norms. We must make one more prefatory remark before discussing psychological concepts themselves: our position concerning the evaluation of standards, social values, or norms, since this position has important ideological implications. We repeat that standards, social values, and norms are the product of human interaction in the process of living. There is no finality about them. They are not absolute. In spite of their inertia after they once come into existence, and in spite of the efforts of those who for one reason or another are vitally interested in preserving the established order, we observe in the course f history that social standards, values or norms do change as a consequence of new modes of human interaction which are brought about by changes in technological and economic conditions (22, 23).

Any given individual is confronted with the social standards, values, or norms of his environment. In short, social norms are first external to the individual or on the stimulus side with respect to him. As such, social standards, values, or norms are first of all the data of the social sciences, not of psychology. Strictly speaking, then, the task of the psychologist is the study of the formation of attitudes as a consequence of contact with these social norms through other individuals or groups or through various products of the social and economic environment. In spite of the diversity and variations of social standards, values, or norms in different societies, human beings do by and large form attitudes in conformity to their group. The psychologist and the social scientist, however, cannot complacently stop at this point. They can go further and show the consequences of the experience and behavior which are regulated by attitudes developed and prescribed by existing social norms. On the basis of such studies, they can reach conclusions as to whether or not the attitudes formed are conducive to a harmonious and well-adjusted personality or to a contradictory and unadjusted personality, whether or not the attitudes formed bring about social solidarity or friction. In short, the psychologist and the social scientist can reach conclusions concerning the extent to which attitudes formed from prescribed social norms conform to or are contradictory to the objective conditions existing at any given period in a given social system.

Our present task. We hale seen that in spite of the wealth of material in the field of attitudes, we still do not hale a unified and established psychology of attitudes. The variety of phenomena covered by the concept ‘attitude' vary to a large degree in their specificity and range. On the one hand we hale strict experimental laboratory investigators who are interested in the attitude or set to a very limited and precise stimulus situation in a laboratory set-up; on the other hand we find social psychologists and social scientists working on attitudes of individuals or groups to a whole nation or race, or to values and concepts which hale wide extensions such as the concepts of fascism and democracy. It is no wonder, therefore, that investigators working on one level of generality often use concepts which hale very little meaning or relevance for investigators working on a different level of generality.

It is our methodological conviction that attitudes in strict laboratory situations and attitudes in the most complicated social situation hale, essentially, the same psychological mechanism at bottom, that the basic psychological

( 300) substrata functioning in both instances are the same in nature. This is, of course, not in the least a denial of the rich and complicated motivational, affective, and cognitive factors which take part in social or inter-personal attitudes. But in order to develop concepts which have scientific generality of the kind found in the natural sciences, we must give up the purely descriptive concepts used by social psychologists and sociologists and see what functional variables have been well established and verified in experimental psychology and other controlled investigation. Once equipped with these well established variables or concepts, we can then proceed to see if we can extend these to the more complicated cases of attitudes which lead the individual to react in characteristic ways in actual social interaction.

Our specific task then becomes self-evident. We shall proceed to find the most essential criteria which can be detected in any attitude-from the relatively simple case of the laboratory set-up to the more complicated situations in actual social life. We shall proceed with a minimum of assumptions and our analysis may seem to be extremely simple and self-evident.


Attitudes are among those components of the psychological make-up of the individual which determine that he shall react not in a passive or neutral way, but in a selective and characteristic way especially in relation to certain specific stimulus situations. Attitudes are not, of course, the only psychological components or states that determine that an individual will react to the environment in a selective or characteristic way. When the individual is hungry, thirsty, or sexually aroused or in some other emotional state, or has been recently stimulated by a functional change in the receptor organ or in the organism at large, he reacts in selective or characteristic ways to the environment. Attitudes, then, are among the various psychological factors which determine the individual's selective reaction to his environment. In our opinion, Woodworth has rendered psychology a great service by accentuating the fact of the selective reaction of the organism at the very beginning of the fourth edition of his general Psychology.

In all the representative definitions or characterizations of attitudes, one feature is common to them all: that an attitude, whatever else it may be, denotes a functional state of readiness which determines the organism to react in a characteristic way to certain stimuli or stimulus situations. A glance at some representative definitions of attitudes, imposes on us the fact that their essential feature is a functional state of readiness.

"The ‘attitude’ is primarily a way of being ‘set’ toward or against certain things" (Murphy and Murphy, 96). "[Attitude] is readiness for attention, or action, of a definite sort" (Baldwin, 6).

"Attitude---the specific mental disposition toward an incoming (or arising) experience, whereby that experience is modified, or, a condition of readiness for a certain type of activity; . . ." (Dictionary of Psychology, Warren, 132).

"An attitude is a more or less permanently enduring state f readiness of mental organization which predisposes an individual to react in â characteristic way to any object or situation with which it is related" (Cantril, 20).

After reviewing these and other representative characterizations of attitudes, G. W. Allport reaches the conclu-

( 301) -sion that "the essential feature of attitude is a preparation or readiness for response." Allport gives his own definition in which this functional state of readiness is essential.

"An attitude is a mental and neural state of readiness, organized through experience, exerting a directive and dynamic influence upon the individual's response to all objects and situations with which it is related" (2).


Whatever other features attitudes may have (and the features of attitudes do vary according to the degree of complexity), it is certain that all attitudes have a state of readiness in common. However, every state of readiness of the organism is not an attitude. There are numerous states of readiness which cannot be called attitudes. For example, a child of two or three is hardly ever neutral or passive to his environment. He is apt to be extremely partial one way or another to the things or persons surrounding him. He is constantly after all sorts of satisfaction. All these momentary tendencies imply states of readiness. But in spite of all this, we can hardly say that a child of three is full of attitudes. In fact he has very few (if any) established or stable attitudes. He may have developed some attitudes toward certain persons like his mother or certain objects such as types of food. But a child at this age will even react in a very negative way to his mother if she proves to be an obstacle in his effort to gain some satisfaction at the moment. We need not elaborate here other cases of readiness which cannot properly be labelled as attitudes.

We must conclude therefore that cases of readiness are not exhausted by all cases of attitudes. The state of readiness of the organism is a more general term. Attitudes constitute specific cases of readiness. Therefore, we must have some concrete criteria which single out cases of readiness as attitudes. It seems to us that the following five criteria are found in cases of readiness which are labelled as attitudes.

1. Attitudes always imply a subject-object relationship.—Attitudes are always related to definite stimuli or stimulus situations. These may be objects such as home, automobile, souvenir, some particular eating place; persons such as one's own body, mother, father, brother, some friend, rival, teacher, sweetheart, wife; or groups of people such as classmates, playmates, Negroes, the community; institutions such as the school, college, church, club; or socially established and standardized concepts, values, or norms such as the flag, the Constitution, democracy. These subject-object relationships are not innate, are not biologically given. The items or objects toward which the subjectobject relationship is developed are always first on the stimulus side for the individual. Only after contact with these outside stimuli, does any relationship develop between them and the individual.

In various definitions of attitude, the content of an attitude is often mentioned in such terms as ‘objects,' ‘social stimuli,' ‘social objects,' etc. (cf. 100). But the content of attitudes will depend upon the particular nature of the subject-object relationship established. The contents of attitudes are as numerous and as different as the stimulus situations to which attitudes are related.

2. This means that attitudes are formed and formed in relation to objects, persons, and values which may or may not have motivational appeal at first. Almost any food may satisfy hunger, but we may develop a special liking for a special food, even for a special restaurant, and even a special table in that restaurant. When these

( 302) particular likes or dislikes are more or less fixated, we have formed attitudes in relation to these particular objects. Almost any average member of the opposite sex will satisfy sexual need, but when it is fixated, with all its affective overtones, it becomes an attitude-an attitude toward a particular person. George Bernard Shaw has aptly defined a lover as a man who exaggerates the difference between one woman and another.

Since attitudes are not innate states of readiness, since they are formed in relation to particular objects, persons, institutions, and values or norms, the individual has first to come into contact with them. And coming into contact is a perceptual situation. This means that the primary stage in the formation of attitudes is a perceptual stage. The presence of a perceptual stage is of the utmost psychological importance. For certain basic facts about perceptual situations provide the starting point for the formation of attitudes, as we shall see in the next section.

The perceptual stage in the formation of attitudes is especially important in cases of attitudes which do not have a motivational basis.[6] As the accumulating investigations in the field f attitudes show, many social attitudes are formed through verbal judgments of adults. Indeed the most directive and important social attitudes which determine status, social distance, and the like seem to be formed through verbalized short-cut dictums or value judgments and through situations which outwardly do not have any momentary motivational appearance.

The fact that attitudes are not innate but are formed as a result of the individual's contact with his environment means, of course, that attitudes are learned or conditioned. Just what the psychological or physiological mechanisms of this learning may be are irrelevant to the present discussion. We are concerned here with demonstrable psychological properties and characteristics of attitudes. Obviously, the more adequate the psychology .of learning or conditioning becomes, the better will we understand the processes involved in attitude formation. But it is almost inconceivable that any final adequate account of learning developed in the future would ever negate the statement that attitudes are formed.

3. Attitudes have affective properties of varying degrees.---Established attitudes are charged with affective or value properties in varying degrees. The affective property of attitudes may be due to motivational (instinctual) origins such as hunger and sex (as exemplified in cases of attitudes toward a certain food, a certain restaurant, a sweetheart or wife) or may be due to non-motivational sources (non-instinctual).

The affective property of attitudes with motivational origins is self-explanatory. The affective property f attitudes with non-motivational sources is due to the fact that these attitudes are formed in relation to social values or norms which in themselves are standardized affective fixations. They are usually verbalized, short-cut value judgments such as "the home is a sacred institution." Value judgments are always given in adjectival form. And all judgments given in terms of adjectives certainly have affective properties. The social values presented in short-cut dictums, or in other ways accompanied by praise or blame, naturally are affectively charged. The individual is forced to respect and uphold the values of the family, school, church, or other institu-

( 303) -tions he is a member or would-be member of. If he does not respect and glory in his flag, he is compelled to. The very fact of membership and participation in group activity or ceremony makes certain standardized values or practices sacred, justifiable, right, honorable, or dutiful in the individual's eyes. Consequently, the attitudes an individual forms in relation to such activities or practices become affectively charged.

Another important reason why attitudes are affectively charged is the fact that many attitudes prescribe the individual's relationship, status, or rôle with respect to other individuals or groups (such as teacher, worker, boss, minister, assistant, etc.). And experiences connected with status are affectively charged.[7]

4. Attitudes are more or less enduring states of readiness.---There are states of readiness which are more or less momentary, depending on the state of the organism and the situation at the time. For example, we may be very hungry and snatch a loaf of bread. After eating enough and becoming satisfied the loaf may then be pushed aside. At the time of sexual tension, a person toward whom there is no established attitude but who can satisfy the sexual need may be passionately seized, but after the need is satisfied so the tension is resolved, one may never look at the person again. In these cases the state of readiness dissolves as the satiation point is reached, at least for the time being. But not so with attitudes. They are more or less enduring states of readiness. Thus a wounded soldier tries to show his respect to his superior officer who has really shared the hardships with him. A very much preferred food may be the subject of praise after the point of satiation has been reached. A sweetheart still has the sentimental halo in the eyes of .her lover even after sexual satisfaction. A dear friend is still liked even during moments of minor friction. Some people in India actually starve to death rather than eat meat which is 'forbidden' by the very rigid norms established and accepted by Hindus. These examples are sufficient to illustrate the fact that attitudes, once formed, are more or less enduring states of readiness, quite independent, within limits, of the momentary states of the organism.

An attitude becomes a more or less enduring state of readiness because of the cognitive components in its formation. We saw above that attitudes are not innate entities, that they are formed as a consequence of contact with objects, persons, or situations to which they are related. We referred to these contacts as the perceptual stage in the formation f an attitude. It is this perceptual stage which begins to give an attitude its cognitive component in the process of formation. Since the first stage in the formation of an attitude is a perceptual stage, we can begin at once to utilize the concepts developed to account for perceptual situations. Here we are on relatively safe ground. The fact that attitudes are more or less enduring indicates, further, that attitudes are learned. And the more adequate the account of learning becomes, the better will we be able to understand the basis for an attitude's enduring quality,

( 304) just as we will better understand the basis of its formation. And it is also almost inconceivable that any final explanation of learning would alter the statement that attitudes are more or less enduring.

We should be very clear on one point, however, so as to avoid misunderstanding. We have said that attitudes are more or less enduring states of readiness to stimuli, objects, persons, groups, values or norms, in relation to which they are formed and which determine the individual to react in a characteristic way in relation to them. But attitudes are not absolute, fixed states of readiness. Since they are formed as a consequence of contact with objects, values or norms to which they are related, they may change, disintegrate. For example, good friends may become deadly enemies, a religious person may become an atheist, a conservative may become radical as a result of contact with new facts and events.

5. Attitudes range in the number and variety of stimuli to which they are referred.---Since attitudes develop as a consequence f experience and since attitudes involve a cognitive component, the extent or range of stimuli to which an individual will relate an attitude will vary according to the nature of the source of the attitude and according to the relationship the individual makes between an attitude and the stimulus situations he confronts. Although certain attitudes-especially some of those created in laboratory situations or those of children-may only be evoked by the situation under which they originally developed, the more usual process is that an attitude, once established, will be related by the individual to a variety of objects or situations that have not necessarily been active in its original establishment.

The wide range exhibited by some attitudes is possible because of the fact that the stimulus situation toward which the attitude has developed is itself extensive, that is to say, possible of representation or expression in many different specific contexts. When we noted as our first characteristic of attitude that it always implied a subject-object relationship, we pointed out that we could label institutions, concepts, social values, or norms as 'objects' in this sense. Social attitudes toward such stimulus situations are, we have said, often derived from the verbal judgments of others. For example, once an individual has accepted the value-judgment of his group that 'Negroes are inferior' and should occupy a lower status in society, he can and does easily relate his acquired attitude to innumerable specific situations. As we shall see, a considerable body of research in the past two decades has shown that attitudes have a directive effect in specific situations, that a very general attitude will reveal itself in a wide variety of ways .[8]


This approach to the problem of attitudes renders meaningless various attempts to classify them. G. W. Allport has summarized the major varieties of classification: positive and negative, specific and general, public and private, common and individual attitudes (2). Since the characteristics of any attitudes in any individual will vary according to the situation or circum-

( 305) -stances under which the attitude has developed and the function the attitude serves for the individual, any classification of attitudes becomes almost as nebulous as any classification of stimulus situations (including objects, persons, groups, values and norms) or of personal and societal relationships. Simple, dichotomous classifications especially distort and falsify the problem. Take, for example, the distinction made by Thomas and Znaniecki, between 'common' and 'individual' attitudes. Their use of the word 'common' refers to attitudes 'common to all conscious beings,' their use of 'individual' refers to attitudes 'peculiar to only one individual member of the group' (123, p. 18). Such a distinction makes no place for the attitudes of members of small or large groups, or the attitudes characteristic of a given social system, not to mention the elaborate and tenuous classification that would be necessary to place an attitude according to the subtle variations it might have for the individual with reference to the attitudes of other members f the same group or social system.


We have said above that attitudes always imply a subject-object relationship and that attitudes are affectively charged. Hence the stimulus or object or group to which an attitude is related is reacted to affectively. In other words, the stimulus, object, person, group or norm in relation to which an attitude is formed has value-either positive or negative and in different degrees for that individual. We are using the word 'value' to denote these affective qualities. Therefore, the problem of value as an affective quality is part of the psychology of attitude.

We do not mean to say that all attitudes are social attitudes, or that all attitudes are related to social values. An attitude may be developed toward certain woods in which a person has taken solitary walks or an attitude may be formed in relation to almost anything peculiar to the surroundings and experiences of the individual. And we might mention in passing the abnormal fixations sometimes made by individuals on queer and unusual objects.

In spite of this fact, however, the attitudes most important in daily life are social attitudes-attitudes formed in relation to other individuals, groups, institutions, tools and technology, standardized values or norms. These are the attitudes that really determine an individual's reaction to other people, other groups and that map out for him the main boundaries of his experience and taste. Most of these social attitudes are transmitted by short-cut verbal value-judgments. Words are the most common medium for both the formation and the expression of social attitudes. And it is probably because of this fact that some psychologists have characterized attitudes as verbalized dispositions. Although we know the social attitudes of others largely from the words they use, most of us do have many quite personal attitudes, not related to social values, which we may seldom, if ever, express-attitudes toward some loved one, someone we strongly dislike, some personal keepsake, some house. To be sure, if we are probed concerning these things, we will be able to express our attitude. But many attitudes we may have which we may never verbalize are just as much 'attitudes' as those we do verbalize spontaneously. The point is that all attitudes-whether social or non-social, whether verbalized or non-verbalized---function essentially according to the same psychological principles, even though there may be differences of content, richness, compellingness, or endurance.

( 306)

Although the psychology of values is involved in the psychology of attitudes, it should be borne in mint that social values, being first on the stimulus site in relation to the individual, are essentially the data of . the social sciences. Social psychologists become interested ant concerned with social values simply because social values are part, ant an important part, of the stimuli that surround man ant, through stimulation, influence him.


Various terms such as ‘set,’ ‘stereotype,’ ‘prejudice,’ and ‘opinion’ may all be regarded as attitudes with particular characteristics which have been given certain labels by common use. The psychological states described by all of these terms are developed in relation to certain stimuli in the identical way that attitudes are developed. All are affectively charged in relation to the stimuli. All are more or less lasting, all are acquired states of readiness determining the individual's characteristic reactions to the stimuli to-which they are related. There are, of course, differences in the range of stimuli to which each of these forms of attitudes can be related, ant there are differences in the intensity with which these different types of attitudes are displayed. There are also differences due to the conformity of these attitudes to objective conditions, as, for example, the difference between the attitude of a biologist to the facts of biological evolution ant the attitude of an uneducated Baptist to the same facts. We repeat, however, that all of these forms of attitudes follow essentially the same pattern in their development.

As commonly used the term ‘set' is applied to a relatively restricted, temporary attitude, or to momentary states of readiness; the word ‘stereotype' is applied to an intense ant rigid attitude, while the word ‘prejudice' applies to an attitude still more rigid ant intense ant one generally based on false information. The term ‘opinion' is generally used to describe an attitude that is or has been expressed ant that is based more on objective conditions than a ‘stereotype' or ‘prejudice.'

We are mentioning the fact that these terms are special cases of attitudes for a methodological reason. For some investigators use one particular term ant teal with the problems surrounding it as if they were separate problems unrelated to the characteristics of attitude or the conditions which develop attitudes. We strongly believe that we shall gain much if we unify the concepts which can be really unified ant if we to not use different concepts for the same problem. Following the rule of scientific parsimony, we shall try to restrict ourselves to as few concepts as necessary.


Attitudes are inferred from the reactions (verbal or non-verbal) of man. When an individual reacts repeatedly in a characteristic way (positive or negative) in relation to a certain stimulus object, we infer that he has an established attitude toward that stimulus. When a group of individuals react repeatedly in a characteristic way to a stimulus situation, we infer that the members of the group have an established social attitude in relation to it. This characteristic reaction of groups of people is sometimes called ‘conforming behavior.' These conformities are discriminatory or selective, as all attitudes are. This means that all attitudinal reactions are judgmental activities. It is, therefore, not a mere coincidence that social value-judgments reveal themselves in the psychology of the individual as established attitudes.

( 307) Whether these discriminative activities revealing attitudes are verbally expressed as short-cut judgments f opinion or value, as in logic, or are expressed only in behavior, toes not matter a bit psychologically.

As we have seen before, attitudes always imply a subject-object (stimulus-organism) relationship. Attitudes always are related to some object, person, group or standardized norm. This relationship is not innate, it is formed. In order to be formed, the individual first has to come into contact with the object (person, institution, norm). This is a perceptual situation. Therefore the first stage in the actual formation of an attitude is a perceptual stage, with the internal factors of the organism and external (objective) factors of the stimulus situation coming into play.

The term ‘attitude' (used in everyday language to denote an established state of readiness) does not express any specific psychological mechanism. It is a composite term, especially useful to denote in an empirical way an important common ground between psychologists ant sociologists. When characterized psychologically and traced from the point of view of its formation, it becomes evident that the psychology of attitudes is intimately related to the psychology of perception and judgment. This is true, no matter what the motivational basis or diversity of content of an attitude may be. In the present approach towards the psychology of attitudes, we shall put our emphasis on perceptual and judgmental processes and see how far it can carry us. Here we to not need to take sites in favor of any learning theory. We have already pointed out that any final adequate psychology of attitude will someday be linked closely with the psychology of learning or conditioning, especially in accounting for the more or less enduring character and for the range of attitudes.

General selectivity of perception.---The objective world around us---rivers, hills, trees, buildings, etc.---are, of course, not affected by our perception of them. They are there objectively, determined by physical laws, whether or not we perceive or experience them. On the other hand, what we perceive of the natural and social world around us is highly selective, determined by biological and psychological laws. Because of this deterministic selectivity, men perceive or experience different aspects of the world around them. For example, in the Dark Ages, because of the fact that men were so confined to themselves due to the prevailing social and cultural system in which they were living, they paid little attention to nature (37). It needed, among others, a Rousseau, the rebellious child of the bourgeois system rising against the decaying feudal aristocracy ant clergy, to call attention to his fellow-men that nature was around them in all its glory. Today, in India, for example, many people are so much pre-occupied with their mystic ways, that they simply are not concerned ant, therefore, to not notice the beauties of nature. We to not have to resort to any typologies to account for these facts, as Jung toes, for example (71). At the basis of this selectivity is the bipolar nature of perception-the determination of perceptions by external ant internal factors following certain laws, ant the formation of certain enduring states of readiness.

The selectivity of perception is a universal human phenomenon not confined to any special cases. As Woodworth puts it in developing what he terms 'situation-set,' "the individual is not an unbiased registering instrument" (141). Anyone can easily cite numerous cases from daily experience that what he ob-

( 308) -serves and notices around him is a selective matter: what he sees in a strange city, what he reads in his newspaper, etc. It would be very useful to make longitudinal studies of perception also and find out through a period of time what actual items a person does perceive in his surroundings. Such longitudinal studies would certainly reveal the highly selective nature of our daily perceptions as determined by external and internal conditions.

Experiments in the psychological laboratory require subjects, through carefully formulated ‘instructions,' to observe precisely certain aspects of the stimulus presented and usually made focal for them in a controlled way. Already some representative laboratory studies furnish evidence of the general selectivity of perception. Here we refer to the long line of investigation started over forty years ago by Külpe when he ventured to study the psychology of the so-called higher mental processes against the vigorous protests of his former teacher, Wundt, who argued for the completion first of the ‘mental chemistry' of sensory processes. In Külpe's laboratory it became increasingly evident that the ‘set' assumed by the observer as a result of ‘instructions,' was playing an important rôle in determining response to the experimental situation. Kulpe himself specifically undertook the study of this problem (85). In his experiments he tachistoscopically presented different stimuli, such as printed syllables, about which different aspects or ‘dimensions' could be reported, e.g., the number of letters involved, the locations of the colors, or the total pattern composed by them. Külpe found that more items were noted and more correct judgments were made by the observer about that aspect f the stimuli which had been emphasized by the ‘set' produced as a consequence of the initial instructions; subjects noticed more fully and in more detail those aspects of the stimulus field they were set to observe. Two decades later Yokoyama and then Chapman verified Kulpe's results (15, 38).

In the Külpe line of experiments, the ‘set' is produced by the instructions f the experimenter. But the internal factors ('set' in this case) need not always be due to instructional set. In more natural settings some motivational stress, some social pressure, or some established norm in the individual may and does produce the ‘set' or ‘attitude' with which the stimulus field is observed. Take, for example, the case of a hungry man looking for bread or the case of a lover waiting in a crowd for his sweetheart.

In his impressive series of experiments carried out over a period of years (1913-1932), Bartlett obtained a well integrated series of results on remembering (7). This classical work breaks down the artificial classifactory boundaries imposed between the functions of perceiving, remembering and even imagining. Bartlett unanswerably demonstrated that the perceiving, imagining and remembering of even relatively simple objects are selective-influenced in an important way by the interests, attitudes, temperament, etc., of the observer. He makes his experiments on perceiving the starting point of experiments on remembering, indicating a functional continuity between perception and memory. In the experiments on perceiving (7, pp. 14-33) he presented to his observers simple designs and patterns, and complex pictorial material. His instructions were neutral, i.e., he did not try to produce a ‘set' or ‘attitude' in the observers in relation to any given aspect of the stimulus field; in fact, he was especially on his guard "against the use of suggestion other than that conveyed by the material itself" (pp. 17-18). His results and conclu-

( 309) -sions are clear-cut. In his own words, "Very rarely indeed did a subject thus differentiate clearly between a sensory image set up by the stimulating object and his interpretation of the object itself. But, though they did not realize it, the observers were throughout constantly utilizing an imaged setting or background, for their perceptual reactions" (p. 30, italics ours). Again "... to perceive anything is one of the simplest and most immediate, as it is one of the most fundamental, of all human cognitive reactions. Yet obviously, in a psychological sense, it is exceedingly complex, and this is widely recognized. Inextricably mingled with it are imaging, valuing, and those beginnings of judging which are involved in the response to plan, order of arrangement and construction of presented material. It is directed by interest and by feeling, and may be dominated by certain crucial features of the objects and scenes dealt with" (p. 31, italics ours).

The more unstructured and vague the stimulus field is, the more important is the rôle of set and other factors not inherent in the stimulus itself. In one of a series of experiments utilizing the autokinetic phenomenon, Sherif instructed his subjects that the light (which was physically stationary, of course) would move to the right or to the left as the case may be (113). The subjects with the ‘set' produced by the instructions gave results generally in harmony with the instructed direction. Recently Haggard and Rose, utilizing the autokinetic technique, but using a device of reward or punishment instead of direct instruction of movement, obtained similar results (58).

From the facts, of which the above are typical examples, it is plausible to conclude in a general way that the selectivity of attitudes is basically imbedded in the selectivity of perception. But this gives us only our general orientation. We must proceed further and try to single out more precisely the processes involved in the formation of an attitude.

A frame of reference is involved in perceptual and judgmental activity.---It is an established fact in psychology that stimuli do not have an absolute stimulating value. A stimulus is experienced, perceived, judged and reacted to in relation to other stimuli, present or past, to which it is functionally related. In perception this relative character of a stimulus emerges from its relationship in the organized whole (81); in judgments from its relationship to other stimuli (present and past) which are operative at the moment. The concept of 'member-character' expresses the relative nature of the properties of any stimulus the organism is responding to at the moment and can be conveniently used to denote the relative character of any stimulating agent with respect to simultaneous or preceding stimuli with which it is functionally related.

The term ‘frame of reference' is simply used to denote the functionally related factors (present and past) which operate at the moment to determine the particular properties of a psychological phenomenon (perception, judgment, affectivity, etc.). In 1935 and 1936 Sherif brought together a good many experimental facts from various major psychological phenomena (perception, judgment, affectivity, memory, personality, etc.) indicating the way in which a ‘frame of reference' is involved in each of them (113, 114). The scale of magnitudes against which subsequent stimuli of a similar kind are judged, the organized perceptual whole which determines the particular relative properties of its parts, the established social status in relation to which responses to other individuals and groups are

( 310) shaped, etc., are all specific cases of frames of reference. Unless we have good reasons to argue that frames of reference involved in various psychological phenomena have nothing to do with one another, but express psychological tendencies quite different in nature, such a general concept is needed to denote this whole background of factors which together determine the relative nature of response. Because of the fact that frames f reference are involved in all major phenomena, a more precise definition of frame of reference cannot be formulated until we know more about the frames of reference in various specific cases.

Specific cases of frames of reference are most extensively elaborated in the fields of perceptions and judgement. Gestalt psychologists worked out in detail the psychology of perceptual frames. In fact, the whole emphasis of frames of reference in psychology is mainly derived from them. All our time and space localization's, magnitude perceptions, form perceptions, perceptions of melody and harmony are referential affairs. When we say ‘up' we mean ‘up' in relation to something below, when we say ‘far' or ‘near' we say so in relation to a starting point we have in mind.

With the shifts of the reference frames or points, perceptual relationships are altered. Thus "a book is small and a man is large. But if a house is large, then a man is small. And if a book is small and a house is large, then a man is of medium size" (107, p. 43). Years ago Wertheimer demonstrated that a line is experienced as horizontal or vertical in reference to the position of other things in the field of stimulation: if the observer's visual field was slanted by means of a mirror, a similarly slanted objective line tended to appear vertical, indicating that the position-of an object is perceived in its relation to the whole organized field (137).

In our opinion, the psychology of these perceptual frames are basically related to the psychology of attitudes. The forms, proportions and magnitudes of things, buildings, tools, etc.., in the village, town or city we live in become for us in time established anchorages. When subsequently, in some other town or country we face different proportions, forms and magnitudes of things, we perceive them against the whole background f our established frames or anchorages. Consequently they seem to us to be too large, too small, queer or disproportionate as the case may be.

Extensive evidence of the referential nature of judgmental activities has accumulated during the past decade. As we shall see later, the implications of this important finding can be carried on to more or less complicated social problems. Chapman and Volkmann, for example, in a significant study on ‘A Social Determinant of the Level of Aspiration' start with the general principle that "the conditions which govern the setting of a level of aspiration (Anspruchsniveau), in the sense of an estimate of one's future performance in a given task, may be regarded as a special case of the effect upon a judgment of the frame of reference. within which it is executed" (39, p. 225). They draw attention to the "general fact that all judgmental activities take place within such referential frameworks" (p. 225). Convenient and concise summaries and discussions of the studies on judgment are found in the reports of Long (90), Rogers (109), and McGarvey (93). Starting with Hollingworth's work on judgment, it became increasingly evident that judgment to a stimulus in a series is based on the whole background of previous judgments belonging to the series and affected by them in definite ways (63,

( 311) 64). For our present problem, the important fact is that judgments of stimuli shift according to the background furnished by a related series of stimuli, the particular directions of these shifts need not concern our position in this paper. Here it will be sufficient for us to call attention briefly to certain aspects of Long's and Roger's representative studies as they bear on our present problem.

Long studied the effect of preceding stimuli upon the judgments of succeeding stimuli, using series of auditory intensities. Among other things he found that "under certain conditions, stimuli oppose each other in such a way that a weak stimulus preceded by a strong one is judged weaker than it actually is, and vice versa. This is referred to as contrast and its presence has been found in experiments employing a variety of stimuli: namely, tones and weights in the usual psychophysical experiments; and colors,. tones, and odors in experiments on hedonic tones. Thus the reality of the phenomenon cannot be doubted, but an explanation of why contrast operates or what processes (either psychological or physiological) underlie it, must be postponed until the conditions under which it occurs, or does not occur, are better known" (p. 55).

The studies on judgment with the method of single stimuli that have accumulated during the past fifteen years furnish unequivocal evidence in support of the relativity of judgment. In an increasing number of studies in different sense modalities, it has been shown that the use of a standard stimulus is not necessary for the observer to give a judgment about any stimulus in the series. After a few rounds of presentation of the series, observers spontaneously establish a scale without being instructed to do so. Any stimulus in the series is judged or placed in its relative position in the scale. These experiments in which no standard stimulus is used yield a distribution of frequencies of judgment very similar to that obtained by orthodox psychophysical methods using a formal standard (108).

Here again we see the same referential nature of response observed in cases of perception. With the shifts of the reference frame or scale to which any stimulus is related, a corresponding shift in judging that stimulus results. Wever and Zener (138) gave 'an observer a ‘light' series of weights and after this series had become an 'established' scale for the observer, they suddenly introduced a ‘heavy' series. "The effect of the first series on the judgments of the second was quite evident for 20 or 25 presentations, i.e.; for four or five rounds judgments of the ‘heavy' predominated for all the stimuli; from this point on, however, the judgments showed a redistribution conforming to the second stimulus series." On the basis of these facts, we may say that the members of a series of stimuli manifest ‘member-character, just as parts of an organized perceptual field manifest ‘member-character,' revealing 'supralocal' qualities from the functional relationship in which they are found. The member-character of the stimuli in a series is clearly shown in a keen observation of Wedell in his study of pitch discriminations. "It has been said that no judgment is ‘absolute' unless a long time, say 12 hours, has intervened since the last hearing of a tone. Following this line of reasoning in the present experiment, the only absolute judgment would be the first one each day. It must be admitted that there is some basis for this assertion, because some of the subjects seemed to compare the notes with one another deliberately. In addition to this, a subject would sometimes correct a previous judgment

( 312) by saying 'Oh, that other one must have been (so and so)"' (135, p. 497, italics ours).

Psychophysical studies dispensing with the presentation of a formal standard stimulus with every variable stimulus in the series were first started by Wever and Zener (138). Because of the fact that no standard was used, and each stimulus was presented singly, the method was designated as the method of ‘single' or ‘absolute stimuli' and the series of stimulus values as ‘absolute series.' Now it is often referred to as ‘absolute scale' and the judgments obtained as ‘absolute' judgments. It is clear that the term ‘absolute' refers to the method alone, because stimuli are singly presented without a standard stimulus. ‘Absolute' does not refer to judgments or their distribution. The real psychological fact shown in all these experiments is that judgments to subsequent stimuli are relative to preceding stimuli of a similar nature. Referring to the generality of this psychological tendency, McGarvey states in her survey of literature in 1943 that "The relativity of judgments of lifted weights, tonal pitches, visual inclinations, etc., is paralleled by the relativity of affective judgments" (93, p. 14).

It is an unfortunate accident to have the term 'absolute' connected with a whole array of facts which definitely reveal the relativity of judgment, perception and, in a word, experience in general. As the term 'absolute' is used in a contradictory way to designate the relative nature of a general psychological tendency, there is bound to appear a contradiction of terms. Thus Cohen, in a study following Beebe-Center's research showing the relativity of hedonic judgments, uses the contradictory title 'The Relativity f Absolute judgments' (41). The same contradiction of terms is found in a more recent study by Postman and Miller showing the relative nature of temporal judgments. The authors state "An ‘absolute' judgment of magnitude is, of course, not strictly absolute, but is formed in relation to other magnitudes that lie within the immediate universe of attention" (107, p. 43, single quotes in original). Again, "Such subjective scales have been called absolute scales and perhaps the term is no more contradictory than the term absolute judgment. Absolute, in these contexts, means relative to the comparable magnitudes that form the immediate context of the judgment" (p. 43, the last emphasis is ours). To designate a tendency by a contradictory term does not help to clarify but only to confuse the issues in question. It may astound people to see that psychologists designate the same thing as both ‘absolute' and ‘relative.'

It seems to us that nothing is gained by stretching the term ‘absolute judgment' and ‘absolute scale' (already used in a contradictory way by the mere accident of naming a method) to study the relative nature of response in perceptual and social fields. The relative nature of response in sensory judgments is only a specific case of a more general tendency. For example, contrast and assimilation effects which necessarily appear in judgments, are not effects peculiar to judgments alone. Assimilation and contrast effects appear by simultaneous stimulation (perception) as well as by successive stimulation, as can be easily found in any psychology textbook. Not only are contrast and assimilation effects simultaneous as well as successive---they are different according to the place they occupy in the figure or ground of a perceptual relationship (11). In view of these facts, appropriate specific terms should be used to designate specific cases of reference frames appearing in other major fields of psychology. For example, 'aspiration level' and 'ego-level' may be

( 313) used as such referential terms in cases dealing with the experiences of success, failure and status as Hoppe (65), Frank (52), Gould (56) and others have done.

In this connection, Roger's significant study dealing with the anchoring effect of a preceding stimulus of constant value on a scale of stimuli is pertinent. His formulation is aptly offered "as an experimental contribution to the understanding of the frame of reference" (109). His study is based on a previous investigation reported by Volkmann in 1936 (128). Volkmann obtained judgments to a series of visual inclinations, using the method of single stimuli. In the first experiment an 'instructed' position (horizontal) was used as a reference point. As a consequence, the scale shifted and extended considerably in the direction of the horizontal. The second experiment demonstrated that in a similar manner any value which the observer selected and held in mind could exert an appreciable shift in the scale.

Rogers devotes the first part of his study to a discussion of frame of reference and 'absolute scale.' He comes close to characterizing the 'mental formation,' labelled as 'absolute scale' in a contradictory way, as a special case of frame of reference. ('Mental formation' was originally used by Wever and Zener to designate the psychological scale [frame] formed after some rounds of presentation of the stimulus series in these psychophysical experiments.) Rogers states "while it might be argued that the absolute scale, in this event, constitutes a frame of reference accessible to experimental manipulation and investigation, a slightly different view is to be taken here. The frame of reference,-it would seem, may most reasonably be seen_ as the product of a number of influences, notably the influences of stimuli which cannot all be present at any one time" (p. 6). This is true in a general way, but not quite, because the influence of past stimuli, not present at the time, may be operative on an 'absolute' scale as well as on any other specific case of a reference frame. For example, in building up a mental scale ('absolute scale') using the method of single stimuli a previously established scale or anchorage even from daily life may come in to modify present judgments. The operation of a frame of reference or a mental scale as a specific case of a frame of reference is detected from the observable determinations or modifications of reported judgments and perceptions. This fact is clearly noted by Rogers, "The frame of reference, then, is inaccessible except in terms of overt responses which it governs" (p. 6).

In the experiments which Rogers conducted, he used the method of single stimuli. His stimulus series consisted of a scale of visual inclinations (in the first experiment) and a scale of weights (in the second experiment). In Rogers' first sessions the usual distributions of judgments were obtained and the usual scales (or frames) were established psychologically. In subsequent sessions, "an anchoring stimulus" was presented just prior to each presentation of a stimulus to be judged and was designated as the top category of the series. The anchoring stimulus, at first the same as the highest stimulus of the range, "was moved progressively further above the stimulus range, remaining always the same throughout any single session." This means that the most frequently presented stimulus was the anchoring stimulus: to be precise, the anchoring stimulus was presented as many times as the total number of presentations of all the stimuli of the series. In a few cases, the values f the anchoring stimuli used were within the upper part of the stimulus range. Rogers' results obtained from both kinds of stim-

( 314) -uli (visual inclinations and weights) are similar. The anchoring or reference point, experimentally introduced, produces changes in the scale and in the category thresholds within the scale. As the anchoring point moves further from the range, it expands the scale to a certain point, and as it is carried down into the range it causes the scale to shrink. Briefly stated, the scale (frame) expands, to a certain point, or shrinks, but in both cases it is assimilated to the shifts of the anchoring stimulus.

Rogers' investigation constitutes a significant contribution to the systematic experimental study of frames and points of reference and to the nature of their mutual interdependence. Our daily perceptions, experiences, relationships with other individuals are structured or altered to an important degree by the conscious or unconscious use of intruded anchorages of a social or nonsocial nature. Thus when we say ‘early,' ‘late,' we say them in relation to certain reference points, e.g., the time of the departure of a train, an appointment, lunch time, etc. In social life, we shift our judgments, decisions and human relationships by self-imposed or socially given intrusions of value-judgments Which serve us as anchorages. In his work on remembering, Bartlett gives vivid illustrations of how the names f things used at the moment serve as anchorages in relation to which perceptions and memories are structured or altered.

The shifts and other effects brought about by the introduction of anchorages into a situation (anchorages lying within or without the structure) are demonstrated by facts accumulating almost since the beginnings of experimental psychology. Thus, Henri, working on skin localizations in 1892-1897, found that shifts of localizations always ‘are committed in the direction of the points of reference (points de repère)' and corresponding shifts took place

with shifts of reference points (61). What Henri called reference points (= points de repère) in 1895 were designated 'anchoring points' by Koffka in 1922 (80), ‘anchoring agents' by Volkmann in 1936 (128), and ‘anchoring points' by Rogers in 1941 (109) in studying the same effects.

Before proceeding to our next step---the experimental formation of an attitude-we shall bring together the main points reached so far. The first stage in the formation of an attitude is a perceptual stage. Because of this and because of the discriminative nature of attitudes they are closely linked to the psychology of perception and judgment. The laboratory studies on one hand, and historical and empirical facts of everyday life on the other, reveal that perceptions are selective. Perceptual and judgmental activities take place in referential frameworks. As a consequence of facing repeatedly the proportions, forms, or perceptual objects, scales of magnitudes (both in a physical and a social sense), these scales and magnitudes form frames of reference in the individual which serve as bases by which subsequent situations are perceived and judged. They need not be consciously formed, deliberately instructed or imposed by others. Once formed they act as anchorages to determine or alter an individual's reactions to subsequent situations. In this fact is imbedded the basic psychology f an attitude.

Frames of reference in relation to structured and unstructured stimulus situations.---We shall first consider frames of reference in relation to structured stimulus situations, beginning with the most clear-cut cases of perception. When the stimulus field is well-structured, the grouping or organization that follows gives rise to perceptions of

( 315) forms, magnitudes, melodies, rhythms, proportions, relationships, localization's, etc. that correspond, in general, to the properties of the objective situations: the perception of a circle, a square, an appropriately grouped succession of tones typify these situations. In such cases the structure of the psychological frame will correspond closely to the structure of the external field of stimulation; the figure-ground relationship being determined by the compelling features and salient reference points of the objective situation. The properties of the different parts are determine y their functional relationship in the structure. This is expressed as the member-character of the parts. Factors in the objective field of stimulation (such as proximity, similarity, etc.) with such compelling features have been studied extensively since the outstanding work of Werthimer (136). This member-character relationship holds true with successive stimulation as is the case in the perception of melody well as with simultaneous stimulation. As we have seen, this seems to be true in the case of judgments as well as of perceptions.

The above mentioned facts furnish experimental demonstration in terms of well-controlled variables and measurable quantities of the similar effects well-structured objects or magnitudes have in our daily life reactions. In actual life, the individual, wherever he may be, is surrounded by buildings, tools, furniture, magnitudes, time-tables, schedules or innumerable other types of well-structured stimulus situations of one kind or another. He is stimulated by them repeatedly. As a consequence the particular structures, magnitudes, and relationships become the established scales or frames in him. Thus, to a person coming from a backward farming district, the ten-story buildings of a medium-sized city which he later visits may look disproportionately tall. To an individual brought up in New York City, buildings of ten or fifteen stories in a different city will seem small. The individual who travels in America a distance of, say, three hundred miles in five hours or so, will be bored to death by travelling the same distance in ten or twelve hours in an industrially less developed country, although the same trip may be considered fast by the native population. And, in turn, those accustomed to travelling by airplane must be bored by riding on the fastest train. In a highly industrialized country where time schedules are important affairs, a difference of fifteen minutes may be a matter of serious concern, whereas a difference of an hour or two may be taken for granted in a country still living under feudal conditions. Such cases could be multiplied many times.

This objective determination, by existing magnitudes, scales, relationships, etc., of lasting scales or frames in individuals implies the establishment of lasting norms. These norms, verbalized or non-verbalized in explicit judgments, are fundamental in shaping the mentalities of individuals living in any social system. They are, in fact, at least as important in shaping the mentality of members of a society as value-judgments, beliefs, or the whole superstructure of culture.[9]

Now we shall consider frames of reference in relation to unstructured stimulus situations. We have ample evidence that in cases where the stimulus situation is not well-structured, the resulting psychological experience is by no means

( 316) always chaos or an inconsistent hodgepodge of reactions. It seems that a tendency to organize and group stimuli is a primary psychological fact based, of course, on underlying properties of the nervous system which competent physiologists will no doubt someday explain. Even in cases where the stimulus field is not well-structured and does not have the properties necessary to impose objectively clear cut, non-reversible figure-ground- relationships, there is usually some sort of organization. For example, campers in a forest on a dark night are apt to see or hear different things around them as determined by their individual attitudes or pre-occupations. But in such cases, internal factors are important in determining the properties of the resulting organizations.

The margin of possibilities for the contribution of internal factors allowed by unstructured situations is at the basis of many studies dealing with the problem of individual peculiarities, personality differences, abnormal tendencies, etc. We find here the basis of the projective devices currently flourishing. The use of unstructured ink-blots for detecting characteristics of the individual was suggested and used by G. V. Dearborn back in the end of the 19th century (43, 44). Others soon followed him. Bartlett used a series of ink-blots in his studies on remembering. The recent extensive systematic uses of projective methods based on the Rorschach inkblots or Murray's thematic appreception test are well-known (77, 99). Voth recently studied personality differences ‘as expressed through various forms and amount of autokinetic perception' using autokinesis as an unstructured, unstable stimulus (129). In more recent studies, Voth has used the autokinetic phenomenon as an index to pathological tendencies and has found fairly high correlation between indices thus obtained from his patients and medical diagnosis (130).

Gradations of structure in the stimulus field.---We have considered cases of structured and unstructured stimulus situations: in the former cases the resulting psychological outcome is compellingly determined by the objective situation, in the latter cases a variety of internal or subjective factors come into play to shape response. Actually, of course, there are all kinds of gradations between these two extremes of structuration.

Recently Luchins used gradations of structure or ambiguity in a series of studies on the social influences involved in the perception of complex drawings (92). The conclusion that can be drawn from these studies is that the effects of social influence (the various devices of suggestion used in the experiment) vary with the degree of ambiguity of stimuli presented. The greater the ambiguity of the stimulus, the greater is the effect of attempted social influence. In a previous study in which the stimulus gradations were too few and too abrupt to allow the possibility of graded comparisons, Luchins seems to reach a similar conclusion: "Whether or not subjects were influenced by A's judgment ['A' being the influencing subject in the experiment] seemed to depend on the obviousness of the correct answer, i.e., the clarity of the judgment-situation, on the truth or falsity of A's judgment, and also on the subject's attitudes to and interpretations of their task and the experimental situation" (91, p. 110, italics ours).

Coffin designed an experiment to study the relationship of suggestibility to the ambiguity of a stimulus situation. The tonal attributes of pitch, volume, and a fictitious attribute created for the experiment and labelled 'ortho-sonority' were used as the three tonal attributes varying in ambiguity. Subjects were

( 317) given a tonal stimulus and then, after each tonal dimension had been defined, were told to equate the succeeding tone heard through their head phones with the original stimulus by turning the appropriate dial which was ostentatiously labelled. Subjects were divided into experimental and controlled groups. Results showed that the least ambiguous tonal attribute, pitch, was in most cases not subject to change by suggestion. Volume, on the other hand, could be reversed by suggestion with most observers, while judgments of 'ortho-sonority' invariably followed the experimenter's suggestion. In other words suggestibility to these attributes increases with their ambiguity (40).

In a study dealing with social determinants of the level of aspiration Chapman and Volkmann produced changes in the standards of their subjects by introducing into the situation a hierachy of standards the truth or falsity of which could not be objectively tested by the subjects. But in a subsequent experiment, after the subjects had considerable experience with the task at hand, the introduction of new standards did not shift the standards of the subjects. With such results in mind, Chapman and Volkmann state that "the ]ability of the judgment, for example, varies inversely with the determinateness of the frame of reference" (39, p. 225). As we shall see later, Asch and his associates get similar results. Cantril's studies of a panic situation and various social movements substantiate the implications of these results on a highly complicated level. While research is still scanty which deals directly with the comparative effects on perception and judgment of gradations of structuration in the stimulus field, from the meagre results obtained thus far we may formulate as a working hypothesis that, all other things being equal, the role played by internal and social factors decreases with the stability, clarity or structuredness of the stimulus situation and with the strength of frames or points of reference already established.

As we have remarked above, material objects, technological and other products of human labor, have a compelling effect in producing corresponding frames in the psychology of men. If the actual truth of all man's individual and social relationships were compellingly imposed on him in daily life, then he would not have any lasting and organized false attitudes. But he actually does have false attitudes about things, especially those he has never seen for himself or closely scrutinized. He has definite attitudes about gods, about life hereafter, and imposed attitudes concerning peoples about whom he really knows nothing.

As formulations and standards have been achieved in science through the laborious and not infrequently persecuted labors of scientific workers such as Galileo, science has changed and corrected many obsolete, ‘survival' social norms and corresponding attitudes. False norms and practices connected with natural events have been corrected and have yielded their place to scientific formulations in the modern world, although there are exceptions such as the norms of Christian Science. However, in social and economic relationships and in the whole superstructure of norms determined by them, many survivals remain. Until social systems base their premises and practices on strictly scientific grounds, ‘survival' norms and attitudes filtering down from different historical periods and outmoded systems will continue to function. The social system that erects itself on a scientific basis is still the exception rather than the rule. In view of this fact, the psychologist who renders a real service is the one who takes a realistic stand and studies human nature and social sys-

( 318) -tems as they really are, not the one who bases his arguments on the inherent goodness or evil of man or the will to truth as Hobbes and Rousseau did in their times and as the advocates of an unchanging 'human nature' are doing today. The more various interest groups strive to perpetuate obsolete superstructures of norms, the more important it is to study social organization objectively. Many an historian and social scientist has observed that unstable, ambiguous and critical situations are those which provide especially fertile soil for the formation and inculcation of social norms.[10]

In the following paragraphs we shall try to typify experimentally the basic psychological processes involved in the formation of frames in unstructured situations. Admittedly, the studies are carried on in artificial and miniature laboratory situations. In presenting them we claim only that they are a starting point for conceptualization, the validity of which will be tested out with more concrete social material in actual situations.

Formation of a frame in an unstructured situation.[11]---In line with the above discussion, the psychologist must find an unstable, unstructured stimulus situation, present it repeatedly to the individual to see if he will structure it somehow to a framework of response---proceeding with the hypothesis that psychological organization or grouping is a primary fact in unstructured as well as in structured situations. And if stimuli are organized by the individual, then the further question is, will they be organized by the members of a group collectively, will members of a group eventually form a frame peculiar to the group? This query lies at the basis of the psychology of what sociologists such as Durkheim, drawing a sharp dichotomy between individual psychology and social psychology, argue so strongly concerning the supra-local, suis generic character of collective values or 'representations' which emerge only as a consequence of collective behavior (46).

The autokinetic phenomenon is one of various experimental possibilities that conveniently lends itself to a test of frame formation when the stimulus is unstructured. In a pitch-dark space a single point f light seems to move and to move in different directions. It seems to move because there is no frame of reference to give it a stable localization. With the introduction of other visible points or objects the point gains stability relative to these points, since all psychological localizations are relative affairs. Even the introduction of sounds in the vicinity of the point seems to affect its stability.

The first experiment in the series studied the formation of a frame (or scale) in the individual alone, thus starting with the general psychology of frame formation. The stimulus light was presented briefly and successively one hundred times in each experimental session. The time of exposure after the perceived movement started was the same in all presentations. This was true for all sessions, individual and group. The observer was asked to report the extent of the perceived movement. The results unequivocally indi-

( 319) -cate that even a scale and reference points are lacking in the objective situation, in the course of the experimental session individuals spontaneously a frame and a central tendency (standard) which may differ from individual to individual in the absence of a compelling objective range of stimuli. In other words, in the absence of an objective scale (frame) and objective standard (reference point) each individual builds up a scale of his own and a standard within that scale. The range and reference point established by each individual is peculiar to himself when facing the situation alone.

In the second part of this first experiment it was found that once a scale is established there is a tendency for the individual to preserve this scale in subsequent sessions (within a week in these experiments). The introspective data obtained furnish further evidence of the formation of a frame. Most typical examples of such introspection are: 'Compared with previous distance'; 'Judgments are all relative'; 'Compared successive judgments'; 'First estimate as standard.' Although the subjects do form frames of reference of their own spontaneously without being instructed to do so, the lack of an objective frame of reference is experienced. The following introspections are typical: 'Darkness left no guide for distance'; 'Lack of visible neighboring objects'; 'No fixed point from which to judge distance.'

(Part II, which contains the list of references for the entire article, will appear in the January issue.)


  1. We have put the word attitude in quotation marks since 'attitude,' in spite of its wide use in psychology, is essentially a term of everyday speech, not primarily a concept of systematic psychology. In this discussion we are indicating what seems to us the proper meaning of the term for psychology and how this meaning is derived from established psychological principles.
  2. Dr. Sherif, on leave from the University of Ankara, is at Princeton on a State Department fellowship.
  3. The question of whether or not the whole of social psychology consists of a psychology of attitudes is not our problem here. We may only say in passing that it is our opinion that social psychology does not consist only of the psychology of attitudes.
  4. Without in any way denying the enormous temperamental differences between individuals, a systematic awareness of the fact that temperamental differences always function within a given social context and set of norms would undercut and bring into proper perspective typologies such as those advanced by Jung, Spranger, and Sheldon which attempt to explain social behavior by categorizing personality (71, 112, 117).
  5. In the field of anthropology an adequate synthesis of the psychology of attitude should be a guide to more adequate generalizations on the anthropologist's own level. Anthropological concepts which try to fit cultures into types or which imply the uniqueness of cultures are for a psychologist quite incomprehensible. From our point of view, the facts and relationships of different economic and social organizations studied by the anthropologist are social norms and values which act as social stimuli and, as such, have characteristics important for the anthropologist and the social psychologist to study.
  6. We are using the term 'motivation' here to denote only drives, needs, or instincts, which have a definite origin in the organism of the individual.
  7. Without making a special issue of it at this point, we may note in passing that attitudes related to rôle or status are ego-involved. And ego-involved experiences and responses have affective properties. The ego-structure is not innate, not instinctive, but a genetic formation the main constituents of which are established attitudes, starting with the delimitation of the physical body with appropriate attitudes formed around it and with the learning of personal names, various adjectives and other fixations. The problem of ego-structure and its bearing on the psychology of motivation is much too important and complicated to be treated briefly in the present discussion.
  8. In this connection, the experimental work on meaning and association time is significant, in view of the fact that the upshot of the controversy was that meaning is more immediate than specific association. (See the experiments of Moore, Tolman, Weld, and Cantril.) This being the case, the immediate 'meaning' reaction determines the range rather than specific associations. This is illustrated later in this paper in the references, for example, to Kay's study of personal and general frames (76) and Horowitz's study on the development of race attitudes (66).
  9. Spiritually inclined culture apologists have tended to ignore or minimize the effects of this basic field of determination. It is too important a field to be done justice in this connection and will be treated in a forthcoming publication under the general heading of technology and mentality.
  10. It might be pointed out here that no realistic purpose is served by conducting experiments of the type in which the attempt is made-by introducing various social influences-to have a subject perceive a perfect square as a circle. In such obvious cases, no problem is involved and nothing is gained by declaring as a conclusion in the face of such a compelling stimulus determination, "Ah ha, people cannot be influenced to perceive a square as a circle, thus they reveal their inherent will to truth."
  11. When not otherwise specified, the results to be summarized here refer to Sherif (113, 114). These references may be consulted for the detailed description of the experimental situation, procedure, subjects, instructions, etc.

Valid HTML 4.01 Strict Valid CSS2