What are commonly called rumors are those descriptions, forecasts, or explanations of events that are shaped largely in informal encounters and that have not been confirmed by authoritative sources but are, nonetheless, entertained seriou.sly by a substantial portion of an interested collective body. Since the reports are unverified, they may turn out to be either true or false. Although warnings against apocryphal tales can be found in the folklore and literature of many peoples, it is only recently that such tales have become the object of systematic inquiry. Contributions have come from many sources: historians and jurists concerned with the reliability of testimony; psychologists concerned with perception and recall; psychiatrists concerned with the expression of repressed impulses in communicative acts; and sociologists and social psychologists concerned with group problem solving, public opinion, and reactions to disasters. Despite these efforts, comprehension of the phenomenon remains fragmentary, for investigators have faced formidable obstacles in collecting reliable data.
Two conceptions of rumor. Pioneering efforts in any field rest on common sense terms, which are vague in reference and contain implicit value judgments; two somewhat different conceptions underlie current studies of rumor. Some regard it as a message that is passed from person to person, assuming that the first speaker in the chain is an eyewitness whose report is accurate and that distortions are introduced in the course of transmission. The basic unit of analysis is the report, and the central problem is to account for its transformation. Since normal communication is assumed to be accurate, rumor is viewed as pathological. Others conceive of rumor as a summary depiction of an event, a capsule statement that is constructed in the interaction of people who are concerned. For them the unit of analysis is the situation, and the problem is to ascertain how it becomes interpreted. In either case it is useful to distinguish between the product and the process by which it is shaped.
Conditions generating rumor. Investigators are generally agreed that rumors emerge in situations that are inadequately defined—when there has been an unexpected break in the daily routine (the appearance of a stranger), when there is a drastic environmental change (a severe earthquake), when those confronted with alternatives are uncertain of the outcome (an impending election), when people are under sustained tension (a city under siege), or when men performing monotonous tasks become restless but cannot establish the source of their annoyance (rear echelon troops). In each context there is demand for news, although it varies considerably in intensity. When some decision is to be made, any information that might affect the outcome becomes important; once the question has been settled, the sense of urgency wanes, and what had been news becomes history. Ordinarily news is provided by some authoritative source—in modern societies, for example, through the media of mass communication. If demand for news exceeds what is supplied through these channels, interested parties are forced to rely upon one another for information. Thus, any circumstance that hinders the rapid dissemination of news, such as censorship or the physical barriers arising in catastrophes, foments rumors.
Allport and Postman (1947, pp. 33-34) have tried to formalize such observations in their frequently cited formula: R ∼ i x a; “the amount of rumor in circulation [R] will vary with the importance of the subject [i] to the individuals concerned times the ambiguity of the evidence [a] pertaining to the topic at issue.” Although the exact relationship designated has never been demonstrated, the formula does specify the important variables.
In most communities social contacts are limited by barriers of class, ethnic identity, or other group affiliations, and most rumors are nurtured in well-established informal networks. Festinger and his associates (1948) studied a rumor that a community worker trying to establish a nursery school in a housing project was a communist; they found that 62 per cent of the people with children from one to five years of age had heard it, in comparison to only 28 per cent of the other residents. In an experiment conducted at another housing project, Festinger, Schachter, and Back (1950, pp. 114—131) planted a rumor that one of the project’s special activities was to receive nationwide publicity. They found that this information was conveyed for the most part to individuals who had taken a prominent part in the activity. In disasters, however, information is urgently needed by everyone, and new communication channels emerge spontaneously. One week after the disastrous floods of 1955 a rumor that a dam was about to break sent the populace of Port Jervis, New York, streaming out of the town. In their study of this incident, Danzig, Thayer, and Galanter (see Philadelphia …1958) found that a number of individuals took it upon themselves to broadcast warnings to anyone within range, even though most people turned first to their friends and relatives. During the French Revolution, rumors about brigands employed by aristocrats spread panic through the countryside, and Lefebvre (1932) notes that fugitives, traveling monks, postal carriers, servants sent to warn friends, and local officials all participated in disseminating the alarm. Thus, news is transmitted to persons known to be involved and concerned, and the activation of communication channels varies with the requirements of the situation.
Content formation. Although rumors are not necessarily false, neither are they direct descriptions of events. The central problem in the study of rumor is ascertaining how content is formed. This has been approached from three standpoints.
Experimental studies. The popular conception of rumor as a message that is passed on from person to person is the basis of experimental studies of serial transmission. The first subject in a series either is asked to describe a picture or is given a predetermined message by the experimenter. He repeats the account to the next subject, who in turn passes on what he has heard. A careful record is made of what happens at each point of contact; additions and omissions are measured. Such studies were first conducted by William Stern in 1902 and have subsequently been repeated by numerous students of legal psychology. The work of Allport and Postman (1947) is the most comprehensive. They summarize their findings in terms of three concepts. Because of leveling, the accounts grow shorter, more concise, and more easily grasped. Sharpening is the selective perception, retention, and reporting of a limited number of details. Assimilation is the tendency of reports to become more coherent and more consistent with the presuppositions and interests of the subjects. Their experiments have been replicated by several others, with substantially the same results; in the course of serial transmission reports become shorter and less accurate. The conclusion is that rumor content becomes what it is because of faulty perception, limitations of memory, and the inability of most men to repeat, verbatim, complex reports that they have heard.
In spite of the frequent and consistent replications of Allport and Postman’s experiments, their conclusions have been challenged by several investigators, who insist that the artificial situation in the experiment differs in important respects from the circumstances under which rumors actually develop. The contention that rumors necessarily become less accurate as they develop has been questioned. In his study of informal communication among American soldiers in the South Pacific during World War n, Caplow (1947) found that many reports became more accurate as they spread. Most men involved in dangerous situations are not gullible; they try to check the accuracy of what they hear, and details that are denied by officials are eliminated. Furthermore, diffusion occurs through a small number of well-defined paths where past rumors had been received with appreciation, and individuals who prove unreliable tend to be eliminated from these networks. Schachter and Burdick (1955) planted a rumor in a girls’ school that four students summarily removed from their classroom by the principal were connected with some missing examinations; they likewise found no distortions. On matters in which there is personal involvement, details are not readily forgotten. The contention that rumors necessarily become shorter has also been challenged. Peterson and Gist (1951), who studied the public excitement over the murder of a 15-year-old girl, insist that numerous details accumulate to elaborate a central theme. The prime suspect was believed to be a man who had employed the victim as a baby sitter; and subsidiary reports developed about his absence from the party he was attending at the time of the crime, his relationship with his wife, and other items presumably suppressed by the police. Festinger et al. (1948) also found that various particulars emerged to support the view that the previously mentioned community worker was a communist. Disagreement on this point arises in part from differing conceptions of rumor; individual items may become more concise, but they tend to be supplemented by additional reports. Several studies, then, support Vergil’s assertion that fama crescit eundo.
Psychoanalytic approaches. Psychoanalysts have developed another approach to the problem of rumor content. Since a person taking part in the growth of a rumor can claim he is only passing on what he has heard, he can speak without assuming full personal responsibility for what he says. This type of activity therefore provides another avenue for releasing repressed impulses. Jung (1910) discusses a rumor, in a girls’ school, of sex relations between a teacher and one of the students. The account became disseminated only because of the active participation of a number of students; the girls had similar erotic interests, and the portrayal tapped something that was already “in the air.” Distortions are also accounted for in this manner; variations are introduced by participants whose personal needs are slightly different. Rumors take shape, then, as individuals entertain and pass on reports that enable them to give vent to anxieties or hostilities they are reluctant to acknowledge. The method used by psychoanalysts is clinical, but support for their position comes from other studies as well. Allport and Lepkin (1945), for example, found that on the American home front during World War n rumors of waste, special privilege, and corruption in government and in the armed forces were more likely to be accepted by individuals who thought rationing to be unfair or unnecessary, who suffered personal inconveniences from the scarcity of goods, or who had no close friends or relatives in combat zones and were therefore not so personally involved in the war effort.
Social interaction. A third approach to the problem of content formation, found largely in the writings of historians and sociologists, is that men who are involved in inadequately defined situations pool their intellectual resources to form reasoned estimates. Rumor is something that is composed in social interaction. Persons deprived of authoritative news speculate about what is happening. They piece together what information they have. Observations are interpreted in the light of what is taken for granted, and the definition that eventually prevails is the one that appears most plausible. For example, rumors developed among German soldiers invading Belgium during World War I that Catholic priests were inciting the populace to commit murder, that machine guns were located in church towers, and that wounded Germans were being mutilated. According to Langenhove (1916), the invaders, who had expected no resistance, were shocked by the guerrilla tactics of the Belgian Army, and reached these conclusions by interpreting events in terms of their conceptions of the French francs-tireurs of 1870. Furthermore, as Bysow (1928) points out, rumors tend to become more plausible as they develop. Details that appear incongruent are challenged and eliminated, and others thought to be relevant are added to make the over-all picture more convincing.
Patterns of rumor development. After a comparative analysis of a large number of case studies, Shibutani (1965) concluded that there are several patterns of development. When collective excitement is mild, as in some trivial break in routine, communication occurs through well-established informal networks. Reports are evaluated critically, and the interpretations that are eventually accepted are both plausible and reasonably accurate. On the other hand, when excitement is intense, as in a catastrophe or in the heat of mob action, critical ability is relaxed. New communication channels emerge spontaneously. Rumor content then becomes consistent with the prevailing mood—be it fear, anger, or joy—and men act on the basis of reports that they would question under other circumstances. Rumors are shaped in a selective process, and the basis of selection varies with intensity of collective excitement. Another pattern of development is found in situations characterized by widespread boredom. When demand for accurate information is not urgent, reports are often labeled as rumors and passed on for entertainment by those who do not believe them. These rumors frequently become vehicles for expressing suppressed attitudes.
Plausibility and consensus. Before any kind of information can serve as the basis for concerted action, it must be accepted by a substantial portion of the interested public. In many situations people remain skeptical and delay decisions until more authoritative news becomes available; this is especially true of disasters, when many wait until the last possible moment before taking action. Disagreements may arise, and a group may be immobilized. In many cases, however, consensus develops over a rumor. Studies of group problem solving reveal that once a particular view becomes widely accepted, considerable pressure to conform is placed upon those who remain unconvinced. Furthermore, seeing others act on the basis of a rumor tends to make it more credible. Action usually provides an opportunity for reality testing; if there is failure in adjustment, the item is discredited as having been “only a rumor.” But many reports are never tested. Beliefs that develop spontaneously in crises are continually being incorporated into the reservoir of popular knowledge to serve as the basis for subsequent judgments.
Politics and propaganda. Rumors of all kinds —of secret treaties, troop movements, misdeeds of candidates for public office—have always played an important part in politics; and the problem of manipulation and control has attracted the attention of scholars. Publicists have sometimes made extravagant claims of successful rumor campaigns, but their contentions are difficult to assess in the absence of independently collected evidence. The propagation of desired rumors is difficult precisely because content does not always remain fixed. News items may be “leaked” into informal networks, but the ultimate product becomes what is plausible to members of the target group. The exploitation of rumors that are already current or the staging of events that lend credibility to them are techniques that have apparently been more successful. During election campaigns and in time of war, there is concern lest opponents sow inconvenient rumors. Denials are usually ineffective, for a rumor does not develop unless it is plausible. Threats of prosecution and other attempts at suppression have also been unsuccessful; indeed, the known existence of censorship tends to create more demand for unofficial news. Appeals for public participation in combating rumors are usually of little avail, although they tend to make people more conscious of the possibility of being deceived. The one procedure that seems to work is the neutralization of sources, for example, the establishment of a conviction that the enemy has a powerful propaganda machine. Then, any inconvenient item can be attributed to the machine and discounted. The techniques of control that actually work are those that rest on principles, in most cases only intuitively recognized, of natural rumor development.
Explanatory theories. How are these observations to be explained? For those who regard rumor as pathological, the answer is relatively simple: Accurate communication is precluded by shortcomings of intellect and character of those who make up the links in serial transmission. Most of the other explanatory theories rest on a principle akin to the gestalt law of Pragnanz. Prasad (1935, p. 7) writes that as long as a change in the environment remains unaccounted for, there is a distracting sense of incompleteness; this elicits an attempt to understand the new circumstances by completing the incomplete. Airport and Postman (1947, pp. 36-38) declare that a report enjoys dissemination if it permits a “good closure”—if it allows relief from tension, justifies emotions that are unacceptable, or makes the world more intelligible. Festinger (1957) has attempted a more formal statement in his theory of cognitive dissonance: the existence of inconsistency is uncomfortable, and a person will (a) try to reduce it and (b) actively avoid situations and information that are likely to increase it. Rumors, then, are neither the product of faulty memory nor of efforts to defraud; most of them represent honest attempts to comprehend ambiguous situations.
Although many have been attracted to the study of rumor by concern over accurate communication, the significance of the subject matter extends far beyond this. It adds to an understanding of the local diffusion of materials presented on the media of mass communication. It also reveals the limitations of formal social control. Whenever men distrust official sources, they rely upon one another for news; all authoritarian regimes are plagued by rumors. Furthermore, some of the ways in which popular beliefs develop are disclosed. Those who believe that rational conduct always rests on verified knowledge can acquire a more realistic view of the premises underlying our daily rounds. Since rumor is a form of group problem solving, it is also of importance to students of social change. Far from being pathological, rumor is an integral part of the social process, an important aspect of the continuing efforts of men to cope with the vicissitudes of life.
Allport, Floyd H.; and Lepkin, Milton 1945 Wartime Rumors of Waste and Special Privilege: Why Some People Believe Them. Journal of Abnormal and Social Psychology 40:3-36.
Allport, Gordon W.; and Postman, Leo 1947 The Psychology of Rumor. New York: Holt.
Bysow, L. A. 1928 Geriichte. Kolner Vierteljahrshefte fur Soziologie 7:301-308; 416-426.
Caplow, Theodore 1947 Rumors in War. Social Forces 25:298-302.
Festinger, Leon 1957 A Theory of Cognitive Dissonance. Evanston, 111.: Row, Peterson.
Festinger, Leon et al. 1948 A Study of a Rumor: Its Origin and Spread. Human Relations 1:464-485.
Festinger, Leon; Schachter, Stanley; and Back, Kurt (1950) 1963 Social Pressures in Informal Groups: A Study of Human Factors in Housing. Stanford (Calif.) Univ. Press.
Jung, Carl G. (1910) 1916 A Contribution to the Psychology of Rumor. Pages 176-190 in Carl G. Jung, Analytical Psychology. Authorized translation edited by Constance E. Long. New York: Dodd.
Langenhove, Fernand Van 1916 The Growth of a Legend: A Study Based Upon the German Accounts of Francs-tireurs and “Atrocities” in Belgium. New York: Putnam. → Simultaneously published as Comment nait un cyc de lèUgendes.
Lefebvre, Georges 1932 La grandepeurde 1789. Paris: Colin.
Peterson, Warren; and Gist, Noel P. 1951 Rumor and Public Opinion. American Journal of Sociology 57:159-167.
Philadelphia, Institute For Research In Human Relations 1958 The Effects of Threatening Rumor on a Disaster-stricken Community. By Elliott R. Danzig, Paul W. Thayer, and Lila R. Galanter. Disaster Research Group, Disaster Study No. 10. Washington: National Academy of Sciences-National Research Council.
Prasad, Jamuna 1935 The Psychology of Rumour: A Study Relating to the Great Indian Earthquake of 1934. British Journal of Psychology 26:1-15.
Schachter, Stanley; and Burdick, Harvey 1955 A Field Experiment on Rumor Transmission and Distortion. Journal of Abnormal and Social Psychology 50:363-371.
Shibutani, Tamotsu 1965 Improvised News: A Sociological Study of Rumor. Unpublished manuscript.
ru·mor / ˈroōmər/ (Brit. ru·mour) • n. a currently circulating story or report of uncertain or doubtful truth: they were investigating rumors of a massacre | rumor has it that he will take a year off. • v. (be rumored) be circulated as an unverified account: it's rumored that he lives on a houseboat | she is rumored to have gone into hiding.