Toward a Science of Public Opinion

Floyd H. Allport

Since 1924 Dr. Allport (Harvard University, A.B. 1913; Ph.D. 1919) has been Professor of Social and Political Psychology at Syracuse University. From 1921 to 1924 he was acting editor of the Journal of Abnormal and Social Psychology; has served on the governing boards of the Social Science Research Council (1925-27-29-31), and the American Psychological Association (1928-3a). In addition to numerous articles in psychological periodicals he is the author of Social Psychology and Institutional Behavior.

Literature and popular usage with reference to public opinion contain many conceptions which impede clear thinking. These notions are drawn from analogies, personifications, and other figures of speech and are employed for journalistic terseness, for the purpose of arousing vivid imagery, or to conceal the emotional bias of the particular writer. They are so widespread in their use and are regarded with so much respect, even in textbooks of political and social science, that their reexamination is necessary as a first step in formulating a workable, scientific approach.


1. The Personification of Public Opinion. Public opinion, according to this fiction, is thought of as some kind of being which dwells in or above the group, and there expresses its view upon various issues as they arise. The "voice of public opinion," or the "public conscience," are metaphors of this sort. The fiction arises through thinking of an expression given by a "group" at one time and another expression given by the same group at another time, and then assuming a continuity of some sort of soul principle between the two expressions. It might be said, for example, that public opinion in 1830 favored slavery, but in 1930 opposed it; and the daemon of the group is thus thought of as changing its mind. When viewed from the descriptive standpoint of science, this fiction, of course, disappears, and we find only groupings of specific individuals with a certain common agreement among them at one time and a different sort of agreement at another time. Though misleading from the standpoint of research, this fiction may have arisen partly from a wholly genuine situation. A certain psychological continuity

(8) does exist in the fact that there are established in individuals, over a period of time, a number of habitual ideas, traditions, customs, and formulations of past experience, in short, a "reservoir" of accepted beliefs and practices, upon the basis of which many current issues are decided. The error, however, consists in thinking that these habitual, neural dispositions in individuals make up collectively a soul or being called "public opinion," which con-templates and decides upon public issues as they appear.

2. The Personification of the Public. A related fiction is one in which the notion of a collective, super-organic being is applied not to the opinion process itself, but to the public which "holds it." A personified "Public" is spoken of as turning its gaze, now this way, now that, as deciding, and as uttering its opinion. One of the effects of this loose, journalistic manner of writing is that, since "the public" is here not an explicitly denotable reality, but a metaphor, any kind of opinion may be attributed to it without the possibility of checking the assertion.

3. The Group Fallacy of the Public. Somewhat less mystical, but equally uncritical, is the usage of those who renounce the idea of a collective entity or group mind, holding that when they say "the public" they mean individuals; but who, nevertheless, go on employing such phrases as "the public wants so and so," or "the country voted dry." Whether we personify the notion of the public or not, we are likely to commit a fallacy when we use a collective term as the subject of a verb denoting action. For the statement which the verb implies will often be true only of a part of the aggregate concerned. By this sort of terminology, which has also been called the "part-for-the-whole fallacy," one conceals facts concerning minorities which it is the business of research to uncover.

4. The Fallacy of Partial Inclusion in the Use of the Term "Public." Applying the foregoing criticism more specifically the question arises, "What do we mean by a public?" Is it a population defined by geographical, community, political jurisdiction, or other limits; or is it merely the collection of people, within such an area, who have a common interest? In the first instance the term is totally inclusive, that is, it is employed to include dl of each individual in the area, his body, his physiological processes and needs, as well as his various opinions and reactions. This usage, however, is not common because it is too complete; it includes so much that the categories of social scientists and leaders cannot be intelligently used in dealing with it. We cannot speak of the opinion of this public, because it includes ton many alignments of opinion, many of which may be irrelevant or even contradictory.


The second meaning of the term public is usually, therefore, the one intended. This meaning is made up, not of entire individuals, but of an abstraction of a specific interest (or set of interests) common to a certain number within the population. Those who have such a common interest are said to constitute a "public." We may call this usage of the term public one of partial inclusion. Now let us suppose that the individuals having this particular interest (that is, comprising a public from the partially inclusive standpoint) are not also members of some other partially inclusive public. That is, let us suppose their public does not overlap with any other public. If we conceive opinion to go with interest, as is likely on important issues, this public becomes coterminous with the spread of an opinion upon certain issues. The public, in other words, would be defined as the number of people holding a certain opinion, and the people holding that opinion would be identified as those belonging to that public. The definition of the term public would thus he circular. The term public, as a partially inclusive phenomenon, would thus be found superfluous for the purpose of research, and the problem would be reduced directly to the task of discovering where and in what degree these alignments of individuals having similar opinions exist among the population concerned.

Now let us suppose, on the other hand, that the publics overlap, that is, that an individual may belong simultaneously in two or more groupings because of different opinions or interests he possesses on different issues. In such a case if we try to state, or discover by a canvass, the opinion of a certain partially inclusive grouping (a "public"), we might not know where a certain individual should be placed. Since he is in two groupings, he may have attitudes which tend to contradict each other on certain questions. One of these attitudes must be suppressed in favor of the other. If we place him arbitrarily in one of the publics we may be misjudging which attitude is dominant, thus producing a false result. If we place him in both, we count him twice, or perhaps have him cancelling himself, both of which consequences are absurd. With terminology such as this it becomes impossible to define our problem, or to discover our empirical units of study. Opinions are reactions of individuals; they cannot be allocated to publics without, becoming ambiguous and unintelligible for research.[1]


5. The Fiction of an Ideational Entity. Another non-scientific way of speaking about public opinion, sometimes encountered in popular usage and even in the literature, represents the opinion content as a kind of essence which, like a Platonic "idea," is distributed into the minds of all those who endorse it. The expression that a certain opinion is "public" illustrates this usage.

6. The Group-Product, or "Emergent," Theory. We now come to formulations which refer not to personifications or agencies, but to results. Public opinion in this sense is regarded as a new product emerging from integrated discussion in a group, a product of concerted individual thinking which is different both from an average or consensus of views and from the opinion of any particular individual.[2] A variant of this definition is that which de-scribes public opinion as "a step on the way toward social decision, a sort of gathering point of the social will in its organization toward action,"[3] This fiction will he discussed in connection with the one following.

7. The Eulogistic Theory. Those who are inclined to regard public opinion as the emergent result of group discussion usually carry the implications of their theory farther, viewing this result not only as different from the products of minds working individually, but as superior in character. In the process of interaction errors are thought to be weeded out so that the opinion of the more enlightened, improved by discussion, will in the end prevail. Public opinion is thus considered not as a segment of behavior common to the many, but as a single ideational product of interacting and creative personalities.[4]

The criticism of the emergent and eulogistic theories calls for some careful distinctions. It is granted at the outset that when one individual enters into a discussion with others he often reaches conclusions which are different from any conclusion he would have arrived at through solitary reflection. The assumption which we should guard against, as unworkable in scientific methodology, is that this emergent product is something floating out, as it were, in space, and belongs to a group mind rather than to individuals' reactions. Argument A must be related to argument B and argument C in a particular individual's thinking. A cannot be in one individual's mind, B in another's, and so on, and produce any emergent that can be known to human intelligence. The emergent product must he expressed by some

(11) individuals or we cannot know it at ail; and if it is expressed by some individual, it becomes difficult to show just how much the influence of integrated discussion has helped to form it. For no matter what common result individuals have reached through discussion with others, when they put that conclusion forward in avert action, in voting, for example, they are expressing not only what they think, but what they want. So-called "group thinking" may have taken place in individuals, as we have shown; but in the arena of practical affairs it is individuals who do things and nor the integrated product of group thought. It may he that the individuals are acting in accordance with group thinking; but in large alignments of opinion this may be difficult to establish since it is so difficult to know what the content of this emergent opinion is.

We must realize, of course, that the questions which make up the con-tent of public-opinion phenomena are usually not questions of ascertainable fact, but of opinion. There is, in such instances, no way of knowing whether the product of the interaction of individuals is of a higher or a lower order so far as truth, or even value, is concerned. Such interaction does bring out the issues more definitely, and it shows more clearly haw -the individuals are aligning upon different sides. In other words, it gives a clearer picture of what the individuals want. This result, however, does not necessarily constitute an intelligent solution of the problem. Such a solution can be known to have been reached only when time and experiment have given us some basis upon which to judge; when, in other words, the issue has become to some extent a question of fact. When this time arrives it is probable that the emergent product will be the result not of group deliberation alone, but of a considerable amount of overt experimentation as well.[5]

We are not denying the possibility that a superior product of group interaction may exist. We are merely saying that, if there is such an emergent product, we do not know where it is, how it can be discovered, identified or tested, or what the standards are by which its value may be judged. Though not discredited in the realm of possible abstract truth, theories of this sort seem to be blind alleys so far as a scientific treatment of the problem

(12) is concerned. Writers who have stressed them have perhaps been thinking of small, totally inclusive rural or pioneer communities where adjustment to nature and to one's fellow men is direct, and where the common, integrated opinion is practically synonymous with the common life; or else they may have been thinking of discussion groups in which a deliberate attempt is made to reach a result satisfactory to the wishes and judgment of all participating. In our modern vast and growing urban populations, complex in composition and organization, where face-to-face contacts of whole personalities are giving way to occupational and other groupings, it is doubtful how much real integrative effect does take place in an individual's ideas through discussion with others. Some occurs, no doubt; but it is probably mingled with the effects of emotional conditioning, with susceptibility to stereotypes, symbols, and "straddle-terms" of political leaders, as well as an undeviating regard for one's own individual interests. In any case the view that public opinion is a product of group thinking superior to the thinking of individuals and effective as a kind of super-individual group will or judgment is a scientifically sterile notion. This theory, like the others we have discussed, may be motivated by the desire of publicists for the support of a kind of "social providence" for their acts. Though comfortably optimistic, the emergent and eulogistic theories may lull us into a sense of false security in which the need for research and for facts regarding attitudes and control processes is in danger of being forgotten.

8. The Confusion of Public Opinion with the Public Presentation of Opinion. (The journalistic Fallacy.) The preceding discussion has dealt with theories of the nature of public opinion itself. There should be added to these a common fallacy concerning the criterion by which a given opinion content should be regarded as "public" (that is, as widely accepted). This is the illusion that the item which one sees represented in print as "public opinion," or which one hears in speeches or radio broadcasts as "public information" or "public sentiment," really has this character of widespread importance and endorsement. This naive error has been fostered by review and digest journals, and by surveys urging popular or legislative action, in which evidence presented concerning "public opinion" has consisted of news-item and editorial clippings from different sections of the country. The lack of statistical foundation, or of studies relating this material to the actual lay of attitudes in the population, is so obvious that further comment is unnecessary.



Notwithstanding these many futile characterizations of public opinion, there appear certain points of common agreement in the work of various scholars which may prove useful in guiding us past the blind alleys and setting us upon the proper road. These points of agreement the writer ventures to restate in his own way and to add a few other distinctions which, he believes, have value for research. The phenomena to be studied under the term public opinion are essentially instances of behavior of which the following conditions are true.

a. They are behaviors of human individuals.

b. They involve verbalization.

c. They are performed (or the words are expressed) by many individuals.

d. They are stimulated by and directed toward some universally known object or situation.

e. The object or situation they are concerned with is important to many.

f. They represent action or readiness for action in the nature of approach or disapproval of the common object.

g. They are frequently performed with an awareness that others are reacting to the same situation in a similar manner.

h. The attitudes or opinions they involve are expressed or, at least, the individuals are in readiness to express them.

i. The individuals performing these behaviors, or set to perform them, may or may not be in one another's presence. (Public-opinion situation in relation to crowd.)

j. They may involve verbal contents of both permanent and transitory character, constituting "genetic groundwork material" and "present alignment," respectively.

k. They are in the nature of present efforts to oppose or accomplish something, rather than long-standing conformities of behavior. (Public opinion phenomena contrasted with law and custom.)

1. Being efforts toward common objectives, they frequently have the character of conflict between individuals aligned upon opposing sides.

m. They are sufficiently strong and numerous, as common behaviors, to give rise to the probability that they may be effective in attaining their objective.

These points of common agreement require some comment. Item (a), stating that the content of the phenomenon must be conceived as related to the actual behavior of individuals, is self-evident. It cannot be merely an invention, for example, of a journalist purporting to represent actual behaviors of acceptance. As for item (c), "many individuals," the specific number or proportion necessary cannot be stated, since it will vary with the situation. The number required to produce an effect toward the objective (m) must be considered in this connection.


(b) Verbalization. The common stimulating object or situation must be something that can be expressed in words; it must be capable of being immediately and clearly named. There can be no such thing as opinion without stating the content of the opinion in language form. The response of individuals to this common stimulating situation may be either verbal or non-verbal. It may, for example, be a grimace, gesture, or emotional expression. This reaction, however, must be capable of being readily translated into words, such, for example, as expressions of agreement or approval.

(d) Common Stimulating Object. The object or situation toward which the individuals' responses are directed must be dearly understood, and within the experience of all. It must be sufficiently limited to be related to a definite proposal for action. It could not be, for example, the general subject of taxation; but it might be the proposal of some particular tax law. Properly speaking, public opinion does not exist about the nature of the deity, though it might well exist with regard to spoken violations of accepted theological creeds.

(e) The common stimulating situation must not only be well known; it must be a matter of universal importance. Mere interest is not enough; the situation must touch upon fundamental needs or desires. The hazards of a man ascending in a stratosphere balloon arouse widespread interest, but they could not ordinarily be called matters of public opinion, since they are not important to many. A government policy of building military aircraft for "national defense," however, might well become a matter of public opinion.

(f) Readiness for Approval or Disapproval. The responses aroused or prepared in the individuals must be in the nature of active liking or disliking, of support or opposition. For example, the common knowledge of the various methods by which the sale of alcoholic liquors may be controlled, and of the relative advantages of these methods, does not belong in the category of public opinion unless such knowledge is connected with the widespread favoring or opposing of some particular method.

(g) Awareness of Others Reacting. A number of writers have maintained that public-opinion phenomena involve a "consciousness of kind" in the individuals holding or expressing the accepted view. It may make a considerable difference in one's behavior, in supporting or opposing a particular measure, if he is aware, or even if he imagines, that others are reacting in the same manner.[6] Although this "impression of universality" is an impor-

(15) -tant part of the opinion process, it is perhaps best not to require it as an essential element in every opinion alignment to be studied. Otherwise important phases of the problem may be overlooked, such, for example, as the distribution of opinions existing at the first moment the common proposal or stimulating object appears, and before people have had a chance to become aware of, or concerned about, the reactions of others.

(h) Opinions Expressed. If item (g), the effect of the opinions of others, be accepted as an important phase of public-opinion phenomena, the corollary follows that the individual's opinions must be outwardly expressed, or at least capable of being readily elicited. As shown by the work of Dr. Richard Schanck, it makes a decided difference in how one feels or thinks, whether the opinion is one that the individual readily expresses or acknowledges to others, or is his own personal and private view. Dr. Schanck has called these two types of reaction "public" and "private attitudes," respectively.[7] To a publicist, the unexpressed opinion is usually unimportant since it does not represent a recognizable alignment. It is not his concern what the reasons of different personalities for holding or not holding certain common opinions may be. The fact of common acceptance or rejection is alone significant. From the scientific standpoint, however, although we recognize that a public-opinion phenomenon requires expression of opinions, we cannot neglect the field of private attitudes. In the long run, the existence of a widespread similarity of unexpressed private attitudes may be highly important, and should be discovered and measured by our techniques. Consider, for example, the potential importance of the opinion which great numbers of Germans or Italians may have about their rulers, but do not dare reveal,

(i) Relation to Presence or Absence of Others. A number of writers have discussed the difference between a public and a crowd.[8] They seem in general to agree, however, that the phenomena which we call public opinion can occur in either situation. The condition of partial inclusion which we have previously cited as characteristic of the usual definition of a public is recognized by implication in the general agreement that an individual can be in a number of publics at one time, but in only one crowd. Another way to state the matter is to recognize that in either case we have a situation comprising many individuals reacting to a common object or situation, but

(16) under different conditions of association, proximity, stimulation, and response. Where individuals are separated, for example, in their own homes, there is not the possibility of visual, touch, and olfactory sensations from the other individuals which obtains in a crowd situation. Modern radio, however, has brought auditory stimulation from others into this segregated domain, as when we "listen in" to the applause of an audience in a political address. This limitation of sensory modes probably has an effect in the lessening of facilitation, or reenforcement, of the responses characteristic of the crowd situation; but it probably does not abolish such reenforcement. In the main, where individuals are reacting in one another's presence, motor responses often have the possibility of being more expressive, overt, vigorous, and direct in their action, In cases where the individuals are separated, the reactions are likely to be more implicit, and can usually become effective only through some symbolic or representative mechanism, or indirect political process, such as voting. For the most part, however, the distinction between crowd-action and public-opinion phenomena seems to be one of degree rather than one of kind.

(j) Transitory and Permanent Aspects. In the treatment of public-opinion phenomena writers of one school have stressed the stable and rational character of the content and the aspect of its universal acceptance, while other writers have represented the opinion content as unstable, emotional, opportunistic, subject to propaganda, and divided upon controversial issues.[9] This disagreement can be resolved if we view the phenomenon as a process with a time dimension, in which the older content becomes the stabilized and universal portion, while the more recent content represents the present ever-shifting alignment. We have referred above to what we have figuratively called a "reservoir" of common beliefs, attitudes, and knowledge, which forms a part of the sociologists' "culture pattern." More specifically, these mores of thinking and feeling are merely reactions that can be predicted to occur with greater certainty, both now and in the future, than can other types of reactions. Some of these long-standing behaviors may be of a rational character; or they may be the product of trial-and-error experience on a large scale, as, for example, the doctrine of American isolation or the avoidance of inflation. Others may be equally long-standing and predictable, but more emotional in character, such, for instance, as race prejudices. Now in the process of forming the new align-

(17) merits, publicity agents employ these universal and long-standing attitudes to secure their immediate ends, their method being that of transferring the old reaction to a new stimulus by the familiar method of the conditioned response. The old response of approaching, withdrawing, rejecting, or struggling is evoked by the old stimulus term, and white it is occurring the new stimulus to which it is desired to transfer it is introduced. The result is the association of the old response, in the future, with the new stimulus.

We now have the suggestion of a solution of the disagreement regarding public-opinion content. The old responses, stable and universally accepted as following upon their original and "rightful" stimuli, still exist in the background. They are the universal, tried, and stable aspect of the opinion. But the fact of their transfer by conditioning to a new stimulus is something new, unstable, opportunistic, and effective among certain portions of the population (who are more biased, more gullible, or more heavily propagandized) but not among others. Hence we have here an explanation of the shifting, irrational, and divided aspect of the public-opinion process.

To take an example of the conditioning process above described, Let us consider the doctrine that "All men are created equal." This idea has long been accepted as a part of American mores. Now such an established attitude alone does not satisfy our criteria for public-opinion phenomena, since it does not, of itself, suggest definitive action toward some objective. Nevertheless, it is one of the psychological foundations upon which opinion alignments, which do satisfy our criteria, can be built. In 1776 support for the war against George III was elicited by conditioning the responses of approval aroused by this formula to the proposals for revolutionary action. Thus the older maxim of individual equality was the stable, enduring, and unanimously accepted phase of the phenomenon. Its transfer to the revolutionary cause was the new, opportunistic, and, at first, highly controversial aspect. Between 1830 and 1861 the same reaction of individual equality and liberty was increasingly connected with the argument against slavery; and after the Civil War the reaction against slavery also became a part of the basic mores. In later years the same doctrine (with aversion to slavery added) has been employed to help align individuals toward abolishing compulsory prostitution (white slavery), child labor (child slavery), and undesirable working conditions (wage slavery). In a similar manner (to take another example) a nation-wide inveterate pride in race and culture, combined with a long-standing prejudice against Jews,

(18) are being employed by Hitler as an instrument with which to unify his followers in support of the measures of the Nazi regime.

We may calk this body of long-standing, common attitudes which are conditioned to newer situations the genetic groundwork responses of public opinion; and in contradistinction we may speak of the consensus of many individuals, induced by transferring these earlier reactions to new stimuli, as the present alignment. One of the important problems of research is to discover the groundwork materials of real or potential importance for opinion in a population, and to determine their relation to alignments existing at present or in the process of formation.

(k) Action toward Present Objective. The distinction between genetic groundwork and present alignment suggests a further contrast between public-opinion phenomena and another set of long-standing behaviors, namely, those which constitute laws, customs, and traditions. These latter phenomena are perhaps special cases of the genetic groundwork upon which opinion alignments may be built. They differ, however, from the other groundwork in the existence of a steeper mode of conformity resulting from the more vigorous coercion of punishment and public disapproval for those who fail to conform. Usually, however, the opinion phenomenon does not represent a conditioning of the legalized response to a new stimulus, but is a widespread struggle reaction against individuals or proposals which go against the customary or legally prescribed practice. Thus we do not say that a law requiring a householder to shovel the snow from his sidewalk is itself a part of a public-opinion phenomenon, so long as everyone obeys it. It is simply a common and expected practice of citizens. If, however, certain individuals in a neighborhood persistently fail to remove snow from their sidewalks, causing inconvenience and danger to their neighbors, there may arise an alignment of expressed opinion against them. In order to make such an alignment effective the existence of the common practice expected and prescribed as Law is Likely to be cited. Laws protecting property are not, in themselves, public-opinion situations; but should numerous unpunished burglaries occur in a community within a short time, a condition fulfilling all the criteria of public opinion might speedily arise. Public-opinion phenomena arise when non-conformists openly refuse to treat the national flag with respect, to wear clothes, or to conform with other customs. With regard to laws not established but in prospect, the situation is reversed. It is not now a case of public-opinion phenomena arising against those who violate expected or legal practice, but of the new law being championed or opposed according as it conforms or does violence to previously existing

(19) groundwork (or can be made to appear to do so). An example of this relationship is afforded by the passing of legislation to prohibit Negroes from teaching in white schools in localities where they were likely to be appointed. Here the genetic groundwork of race prejudice was the response to which the newly proposed law became the conditioning stimulus.

(1) Relation to Issue and Conflict. Public-opinion phenomena, as we have seen, are those which involve readiness for action toward some present unattained objective. The common stimulating situation toward which the responses are directed is a plan or policy through which many individuals are trying to get what they want. This being true, situations will often arise in which the individuals are aligned in special-interest groups, members of each side trying to get what they want in opposition to individuals aligned in an opposing group. Opinions upon the two sides in this case are only aspects or symptoms of a more profound and general struggle. They may be only a rationalization of this struggle to secure favor with neutrals or stronger loyalty from adherents in the drive toward the real objective, which is often more biological or prepotent than the formulated opinions of its supporters would suggest. The doctrine of States' Rights, for example, has been used as a rallying symbol for individuals with strong economic interests of various sorts.

We enter here the field of public opinion in relation to pressure politics, class and labor struggles, and social conflicts of every type. It becomes necessary here to transcend the view of the publicist who is usually interested only in one side of the controversy; for the alignment, or piling up in a J-curve of attitude distribution upon one side, is intelligible only in the light of a corresponding steepening upon the other side. In a two-party system of politics each party alignment has its full significance only in view of the opposing party alignment. The entire distribution becomes U-shaped. Strong communistic developments are contemporaneous with strong capitalistic and fascistic alignments; and the one grouping seems to derive its meaning in contrast with the other. The popular notion that these various "isms" arise as political philosophies gaining momentum through indoctrination as they spread is inadequate. These philosophies represent rationalizations of the more powerful factors which lie beneath. They are the verbal aspects of the total concerted struggle behaviors of individuals aligned upon the two sides. They are the verbal part of the techniques which the individuals are using to get what they want in the struggle. In international conflicts, similarly, we should take our public-opinion field as broader than the limits of one country alone. We should think of a

(20) U-shaped distribution of the population of both countries combined; for the shifts of attitude distribution in one of the countries bears a definite, predictable relationship to the shift in the other.

(m) Probability of Effect. Our final criterion, that of a probable degree of effectiveness is, from the standpoint of control, the most important of all. In the entire field of the population sampled there will probably be found consensuses of individuals favoring or disfavoring all sorts of common objects, in all ranges of number, intensity of conviction, and effort put forth. A thorough program of research would include the charting of all these consensuses. From a more practical standpoint, however, we shall probably have to choose from all this array the particular alignments in which we are most interested. And in this choice the criterion of selecting those which promise to be in some degree effective will probably be found the most useful and natural to employ. In making such a choice the mistake is sometimes made of selecting the alignment which seems to be the largest from the standpoint of numbers of adherents. A careful consideration of the probable effect of a given alignment, in which other factors besides number are taken into account, will help us to make a better selection. There may be many cases in which a large proportion of the people favor some action, but that does not necessarily argue the highest probability of that action being taken. The variable of intensity, that is the degree of feeling, or the strenuousness of the effort which individuals will make toward the common objective, must also be considered. For example, a recent nationwide sample poll on birth control has revealed that a substantial majority of the people are in favor of it. Yet legislative action supporting it has not been generally forthcoming, probably because the desire for it was not sufficiently intense. That is, the need and desire for contraceptive information and help that cannot now be gained by the individual himself is not felt acutely enough by the members of this majority to press for organized action in opposition to a minority who have a very intense feeling upon the other side. Collective results are brought about by enough people holding and expressing opinions, and by their expressing them strongly enough, or acting upon them. The situation must ensure that enough people are intensely enough affected.

Other influences must, of course, be recognized in predicting or understanding the production of effects. The existence of some type of organization for bringing collective action about, and the facility of using such organization, are important. The presence of individuals of outstanding influence and ability to direct the undertaking is another factor. A

(21) third factor is the degree of reenforcement received by each individual through feeling that others have the same attitude as he; and this, in turn, depends upon the ease, quickness, and freedom of communication among the individuals. The channels through which citizens can make known their wishes to authorities must also be taken into account. We must remember also that the process through which the alignment becomes effective is complicated by a circularity of reenforcement. When, for example, an editor pretends in his columns that he is expressing "public opinion," he thereby influences authorities on the one hand and strengthens the alignment among the people on the other. The latter influence increases the popular manifestation of the attitude, with the effect of still further increasing the editor's confidence and aggressiveness in putting forth his editorials as "public opinion."

It is true that these various factors are at present difficult to isolate and measure. To separate them and study the contribution of each to the total effect is one of the problems of the new science of public opinion. For the present we must rely, in the absence of more definite knowledge, upon a practical. familiarity with these complex situations. In applying the criterion of effectiveness it is, however, unnecessary to wait until the effect has already been produced. If we waited until that point, we should miss important aspects of the phenomenon as they were taking place. Nor is it necessary to he certain that the effect will occur, and that the opinion-alignment we are considering will play a definite part in producing it. It is sufficient that, when we survey the whole situation, there seems to be a probability above chance that this will occur. This, in fact, is the very method which political. leaders use in gauging the potential. importance for their programs of current opinion-movements in their communities. And although they have only this subjective weighing of the probabilities to count on, nevertheless, if they accept a certain opinion-alignment and act as if it were going to be effective, the responses of citizens adhering to that alignment will probably tend to become effective or more effective than they were before. Important as the original lay of attitude of individuals may be, we must consider also the entire control situation, with the numerous influencing factors we have cited, as a configuration in a multi-individual field. This phase of the problem cannot be overlooked if we are to be able to predict or even to understand effects. In the language of the new topological psychology we seem to be dealing here with vectors of force operating in a social field.[10]



Our discussion of the fictions and blind alleys of method have shown us where the major futilities lie. When we try to find an object corresponding to the term public opinion, that is, when we regard it as an entity or a con-tent to be discovered and then studied or analyzed, our efforts will meet with scant success. But when we distinguish by this term a multi-individual situation, or some of the relationships in such a situation, and then enter this situation and begin to study the explicit materials which it affords, some valuable results may be gained.

The question now arises as to the nature of this "public-opinion situation" and how its characteristic relationships may be recognized. And the answer to this question is to be found in the points of common agreement which we have previously discussed. We are to deal with situations involving word reactions or reactions to words on the part of many individuals, which are directed toward common stimulating situations important to many, these reactions showing readiness to act favorably or unfavorably toward the situation, to be influenced by the awareness of others reacting, to associate older attitudes with present issues, to be directed toward an objective different from the status quo, to be frequently related to concerted conflict, and to suggest the likelihood of being effective. Through the use of these criteria we thus find reality and use for the notion of public opinion, while discarding those earlier attempts at formulation which led us off upon the wrong track. We have retained and identified public-opinion phenomena, while at the same time keeping our hands upon the explicitly denotable realities before us, upon behaviors of individuals which can be measured and recorded in the form of statistical distributions. The whole argument may be summarized by the following condensed and somewhat formal statement:

The term public opinion is given its meaning with reference to a multi-individual situation in which individuals are expressing themselves, or can be called upon to express themselves, as favoring or supporting (or else disfavoring or opposing) some definite condition, person, or proposal of widespread importance, in such a proportion of number, intensity, and constancy, as to give rise to the probability of affecting action, directly or indirectly, toward the object concerned.


  1. Professor Dewey has recognized this confusion in the notion of a public and its result in connection with our difficulties in dealing with public problems. (See Dewey, The Public and Its Problems.) For further discussion of this point and of Professor Dewey 's position, see F. H. Allport, Institutional Behavior, Essay V. For a further reference to "total and partial inclusion" in their bearing upon method in social science, see F. H. Allport, "Group and Institution as Concepts in a Natural Science of Social Phenomena," Publications of the American Sociological Society, Vol. XXIII, pp. 83-99.
  2. See Krueger and Reckless, Social Psychology, pp. 127, 266, and Gault, R. H., Social Psychology, pp. 176-77.
  3. See the definition cited by J. K. Folsom, Social Psychology, p. 446, and Chapter IX of that work.
  4. See R. H. Gault, op. cit.
  5. For an account of some of the effects of group discussion, upon accuracy of opinion, and of the theory involved, see an account by the present writer in Achilles, P. S., Psychology at Work, pp. 214-18; also Jenness, Arthur, "Social Influences in the Change of Opinion," Jn. of Abn. and Soc. Psych., 1932, 27, 29-34, and "The Role of Discussion in Changing Opinion Regarding a Matter of Fact," Jn. of Abn. and Soc. Psych., 1932, 27, 279-96.
  6. For a discussion of this phenomenon see F. H. Allport, Social Psychology, pp. 305-7.
  7. Schanck, Richard L., "A Study of a Community and Its Groups and Institutions Conceived of as Behaviors of Individuals," Psychol. Monographs, Vol. XLIII, No. 2 {whole No. 195), 1932.
  8. See Krueger and Reckless, Social Psychology, p. 266.
  9. One writer has taken note of this contrast by making a distinction between "public" opinion and "preponderant" opinion.
  10. It might be argued that, if art alignment of individuals of given numbers and feeling intensity is thus to be regarded as a force in a social situation, the treatment of public opinion as an entity, a formulation which we have rejected as sterile, becomes valid, and even necessary, as a working principle of research. The attempt to isolate and measure forces in a social field is, however, too new for us to make a final decision about this matter. In addition to the intangible character of the units we must handle in measuring the force of an alignment, there is the further baffling problem of the circular increase in effect. In figuring the stresses upon the pier of a bridge an engineer does not need to figure the pull of gravitation and the force of the current as augmenting one another in a circular manner. If this were true it would render his task of snaking calculations wellnigh impossible. The situation, however, with human organisms is different. A public official's attitude and program, let us say, are under the influence of the combined "forces" of newspaper publicity and the opinions manifested by individuals at large. But the calculation of these two forces is not an easy problem, because the newspaper appeal, as we have shown, may affect the attitudes of the citizens, and the change thus produced may in turn reaffect the newspaper editor, this change again further influencing the citizens. And so on ad infinitum. The present writer is inclined to believe that "force" calculations will fail in the multi-individual behavior field, and that we shall have to employ some other type of measuring continua, such as the telic or teleonomic, which have been described elsewhere. In such methods, however, we shall he obliged, as we have indicated, to take into account a field of reciprocally acting units, whether these units be conceived as collective entities or as single human organisms.
    But granting that indices of collective alignments of "public opinion" in a social field could be developed (and all efforts in that direction should certainly he encouraged) there is still good reason for us to continue our emphasis upon measurement at the more elementary level, namely, the behaviors of individuals. We must remember that the societal patterns, or gestalten, are far less stable than force fields in the physical or biological sciences, by analogy with which the societal configurations are suggested. Voluntaristic action, such as we have in public affairs, is by its very nature subject to unpredictable changes which may alter the functioning of the entire pattern, and the force which we assign to its various components. The position of iron particles in a magnetic- field, the contributions of various bodily organs and tissues to the entire body metabolism, the role of sensory units in the perception of space,-ill these are phenomena having a high degree of stability. The pattern of relationships is not affected by any likelihood that these component atoms, molecules, tissues, or sensations might think or feel in a certain way, or connect their old, stable ways of thinking to new issues, or communicate their reactions to the other elements of the pattern, or that it would make any difference in the patterns of magnetism, metabolism, or perception if they did. In the human field, however, the mere fact of individuals thinking or feeling in certain ways, or knowing certain facts about the "pattern" of their actions, profoundly affects that pattern itself. Let us suppose, for example, that it should suddenly become known that, in their own private feelings, the great majority of citizens had long been wishing for a law to he passed depriving the "Supreme Court" of the power of declaring enactments of "Congress" unconstitutional. What a marked readjustment this would bring about in our power field, and how it would alter the "force index" of "nine venerable gentlemen"! In contrast with this instability of the collective force field, we come hack to the relative dependability of attitudes in individuals, For this instability of the collective scene is not due to shifts of personal attitudes in the citizens, at least to shifts in the groundwork-reactions of opinion, but to the control methods, political structures, sudden blockings or facilitation of communication, dispelling of "pluralistic ignorance," and the exploiting of group and institutional symbols for shifting purposes. Attitudes of course are not absolutely permanent, not even the groundwork-reactions which have been longer established, They are far more stable, however, than the force indices of the various components in a societal pattern. It is difficult to predict the power which, in coming decades, will he accorded to spokesmen either of the working class or of the government-controlled capitalists, or the authority which will he vested in the President, the members of Congress, or the heads of government bureaus. But we can feel fairly certain that the opinion alignments which will be most broadly held and most effective in this country will be those of protecting the interests of the "common man," of maintaining homes, of liquidating public debts, of demanding security for the future, of keeping out of European politics, and of maintaining an adequate national defense.

Valid HTML 4.01 Strict Valid CSS2