Contemporary societies are awash in rumor. For better and worse, we live in an information surround in which there is simultaneously too much information and too little. Many wish to persuade us of truth claims, or, at the very least, to share them. These claims may have an uncertain provenance, but, under the right circumstances, we incorporate them into our belief system, act upon them, and recall them through collective memory.

Given these claims, the question becomes who, what, where and when do we trust. To the extent that reactions to assertions about society channel our actions, judgment is necessary, even if it is only implicit and tacit. The analysis of rumor belongs to the sociology of action. To produce a response, we must judge both those who communicate and what they communicate. As an everyday practice, we engage in the politics of credibility and the politics of plausibility. These concepts are tied to issues of public trust within civil society.

The terms ‘gossip’ and ‘urban legend’ are often used interchangeably with ‘rumor’ by both naïve laypersons and professional scholars. Consider our students as naïve laypersons: When asked to evaluate the terms ‘rumor’ and ‘gossip’ along certain dimensions, they rate them almost identically. When asked to ‘think of a rumor’, they frequently report a scandalous tidbit of gossip. Similarly, some of our colleagues - professional scholars of rumor, gossip and/or urban legends - also tend to use these terms interchangeably. For example, participants at a recent interdisciplinary conference on rumor and legend came to no consensus over what distinguishes rumor from legend; scholars of rumor and gossip at a recent conference of social psychologists also argued over whether rumor could be differentiated from gossip. This conceptual fuzziness has been noted for some time (Ojha, 1973). And though much progress has been made in sharpening the construct of rumor (Fine, 1985; Rosnow and Georgoudi, 1985; Rosnow and Kimmel, 2000), ambiguities still remain.

Meaningful and important differences do exist, however. In this article we further clarify the concept of rumor and distinguish it from its two cousins, gossip and urban legend. The source of conceptual difficulties has arisen from insufficient attention to the context and function of these genres of informal communication, and conversely, overmuch attention to content.2 We therefore posit a definition of rumor focused upon contexts, functions and contents. We also examine contexts, functions and contents of gossip and urban legends, and explore similarities and differences between these three forms of social discourse. We begin with rumor.

To associate the theory of rumour with the theory of problems may seem incongruous. And yet a rumour may easily be perceived as a solution, one that is quite circumstantial and wholly marked by mental improvisation, to a problem of collective relevance: to explain, for example, why one is right to be worried, to account for whatever reason one might have for showing hostility to an innovation, to show what face we should put on our uncertainties, to argue the distance which separates us from others unlike us and to prove, by this example, how hateful are our oppressors and those who exploit us. But there is more. Explicitly or otherwise, rumour ends up generating practical advice, an injunction to action or to refrain from action (‘don't do this’, ‘don't go there’, ‘don't eat that’, ‘watch out’, ‘check up on that …) which also links it to a concrete solution. Which goes to suggest that, antecedent to the rumour, there existed a need to know or to know how.

Such a way of looking at things permits an advance in parsimony, since there must exist fewer originating ‘problems’ than attested ‘solutions’. On the other hand, this point of view installs rumour in the function of a revelation or symptom of a social state which encompasses it. Finally, this perspective reconnects the production and development of rumours to the set of cognitive mechanisms of the computational type. One rediscovers therein the continuity of processes of social thought, and rumours cease to be confined to their status of monstrous singularities.

We can construct a typology of rumours - defined broadly as unverified news - according to their relation to reality after their degree of veracity has been established, at least in the current state of knowledge, by experts (historians, scientists, police officers, journalists, and so on). If a rumour turns out to be correct it becomes an item of information. If a rumour is untrue it comes into the categories of affirming or denying rumours. Affirming rumours, which are the most common type, state the reality of imaginary facts, for instance the rumour claiming that bananas from Costa Rica pass on to people flesh-eating bacteria that cause general necrosis. Denying rumours, which are rarer, deny the reality of established facts, for example rumours claiming that Elvis Presley is not dead or that Americans have never walked on the moon.

It seems useful to distinguish various terms indicating negation. Rumours will be termed negative when they state facts associated with fear or hostility, whereas positive rumours evoke happy, welcome events. French researchers, who assign colours to rumours, talk about ‘black’ rumours or ‘rose-tinted’ ones. Specialists agree that there are around nine black rumours for every rose-tinted one. Rouquette (1975, 1990) has thoroughly studied this ‘bias towards negativity’, which is one of the main features of rumours. I shall talk about denying rumours and ideas when they deny commonly accepted events. These rumours are not necessarily negative: for instance survival legends often express the wish that a loved person were not dead. Finally, among denying rumours and ideas, we shall reserve the word negationist for revisionist claims that the gas chambers did not exist, or that a Nazi did not plan to exterminate the Jews during the Second World War.

This paper examines the stability of the concept of rumor in the past century. It is suggested that not only do models of explanation change, but rumors themselves also change - not just in content, but perhaps in the way they are believed or disbelieved. Social scientific interest in rumors begins with the birth of modern psychology in the 19 th century, shifts to social psychology and sociology in mid-20th century, prompted by governmental concern over subversion through rumor during the Second World War, and is finally revived by folklorists in more recent decades. Understood variously as a conduit of the unconscious and otherwise unendorsable thoughts, a mundane communication drift, or an intentional form of deception and provocation, many of our rumor model assumptions are drawn from that era and remain basically unchallenged. A central assumption emerged that ambiguous situations create a vacuum which rumor fills. By the late 1960s, despite a decline in social scientific interest in the topic, a handful of significant empirical and theoretical challenges emerge from scattered studies. The discipline of folklore begins to take more interest in contemporary rumor in the 1970s, and by the early 1990s the rubric of the rumor is almost entirely supplanted in English language scholarship by the ‘urban legend’ (Brunvand, 1981).

As it happens, we are quite well aware of the origin of a group belief. For instance, the history of baseball in the USA is a kind of contemporary myth whose origin, however, is not mysterious. In the US there is a place called the Hall of Fame dedicated to the great figures in baseball history. The spot can be found in Cooperstown, a small American town in the middle of New York state, that is otherwise totally unremarkable. Why was a building put up there to celebrate the sport that is so emblematic of the United States? Simply because the famous baseball is supposed to have been invented there by one Abner Doubleday in 1839. The date is precise but the myth of origins associated with it is no less so. In the early 19th century Doubleday is alleged to have interrupted some children playing marbles behind the shop belonging to the town's tailor. Then he is supposed to have started to teach them the rules of a new, more exciting game which he had just invented (if we adhere to this myth of origins) and which he proposed to call ‘baseball’. So he marked out a small-scale field on the ground: the first game of this typically American sport could now begin.

Studies of the phenomenon of rumour can no longer avoid paying particular attention to recent transformations of structures in the public sphere. On the one hand, the central role of media in the classification of items of information as ‘rumours’ and in their subsequent publicization and dissemination has become entrenched. On the other, rumours have come to be treated not just as simple pieces of information but rather as a dimension of the political or secular interchange that goes on within larger or smaller groups. These two paradigmatic hypotheses have been recently advanced notably by Pascal Froissart (2002) and Philippe Aldrin (2005). Froissart has shown that our awareness of rumours is very largely moulded by the major place accorded to them in the media, and that many so-called rumours said to have been spontaneously generated from within the body of society in fact find their origin and their channels of propagation effectively within the media itself. For his part, Aldrin would break with the standard interpretation of rumour content as often reflective of the shifting nature of collective psychology, substituting for this, notably for rumours of a political type, a ‘transactional’ approach, by which the rumour is analysed as a form of information that enters circulation in the course of everyday interactions, or ones linked to potential political outcomes.

The apparent erosion of public trust has become a defining feature of our age (Fukuyama, 1996; Sztompka, 2000). Collapse of trust in the political process is the subject of intellectual concern and reaction (Warren, 1999; Bentley, 2005). Another important focus for analysis and response to mistrust is science and technology risk management. Sociology is at the forefront of problematizing science and technology risk management, proposing a renegotiation of the relationship between science, citizens and policy (Irwin, 1995; Irwin and Wynne, 1996; Wynne, 2000; Stilgoe, Wilsdon and Wynne, 2005). Science is admonished in the now influential sociological critique for failing to communicate the uncertainty that is central to their endeavours, instead continuing to insist upon a ‘top-down’ process of science communication. Science is presented as having unreflectively brought about a scientific and technological revolution without regard for the consequences that the resulting ‘risk society’ is now struggling to come to terms with (Beck, 1992, 1995). An influential review of specific examples of 20th-century scientific and technological innovations argues the ‘late lessons’ that can be drawn are the failure to heed ‘early warnings’ that could have been detected about potentially damaging impacts (Harremoës et al., 2002). A new, precautionary approach is urged - one that can avoid the mistakes of the past.

Bits and pieces of a fragile, fragmented discourse. Messages that traveled from cell to cell, cellblock to cellblock, even from one prison to another. Scrupulously and copiously analyzed, discussed and, on occasion, transformed during recreation periods and family visits, they were capable of inspiring both hope and fear, depending on the circumstances. But most importantly they were vehicles - spontaneous, improvised vehicles - for combating misinformation and uncertainty. Political prisoners had a name for them: bembas. Often the sole topic of conversation and argument among political prisoners, bembas always elicited a broad array of opinions and reactions: skepticism, blind faith, humor, meticulous examination, scorn and, most often, expectant, painstaking analysis.

As an addition to the papers brought together in this special issue here is an informative dossier, which we have tried to make comprehensive, on recent research work (published since 2001) connected with the field of rumors and urban legends.

This dossier includes reviews of eight French publications.

In 2002 came the second edition of Jean-Bruno Renard's Que Sais-je volume Rumeurs et légendes urbaines; the book by Véronique Campion-Vincent and Jean-Bruno Renard De source sûre; and Pascal Froissart's study La Rumeur, histoire et fantasmes.

In 2003 there was Gérald Bronner's L'Empire des croyances.

The year 2004 saw the publication of a more marginal work by Yves-Marie Bercé, A la recherche des trésors cachés, du XVIe siècle à nos jours, which highlighted a whole complex of rumors and legends around this universal theme.

In 2005 the following were published: Philippe Aldrin's Sociologie politique des rumeurs; Véronique Campion-Vincent's La Société parano. Théories du complot, menaces et incertitudes; and finally Frédéric Dumerchat and Philippe Veniel's Forêt de Chizé: attention puma! focusing on the group of legends about mystery big cats.