On Serge Moscovici ’ s 95 th anniversary : The theory of social representations — history , postulates and dissemination

The theory of social representations is one of the major theories in social psychology with the thousands of scientific articles that has been published since its birth. In this article we return to the reasons for this success. First, it can be explained by the relative flexibility of the initial postulates of this theory. This flexibility has allowed researchers who are sometimes far removed from psychology to adapt it to their own problems. But the success of the theory of social representations can also be explained by the action taken by Serge Moscovici throughout his career to spread his theory throughout the world. Finally, Serge Moscovici never ceased to suggest new ideas about social representations and many of his suggestions remain untapped to this day. They constitute a real reservoir for researchers of tomorrow.


Introduction 1
In the field of social psychology, and perhaps in science in general, there are very few researchers who have originated several 'major theories'. By 'major theories' we mean theories that have given rise to a great deal of work, carried out by researchers from all over the world and over long periods of time. These theories are often the work of a lifetime and many researchers would very much like to propose at least one of the kind during their career… But in a few exceptional cases there have been among them those who proposed even two! This was, for example, the case for Leon Festinger who, after formulating his theory of social comparison (1954), would later develop his theory of cognitive dissonance (1957). It was also true for Serge Moscovici who gave our discipline the theory of social representations (1961) and the theory of minority influence (1976). This is even more remarkable for both Festinger and Moscovici since the theories they forged address questions that are relatively far apart. The second theories they proposed were not the developments of any initial theories (as, for example, in the case of Icek Ajzen's theory of planned be-haviour succeeding that of reasoned action (Ajzen, 1992)). During his life as a researcher, Serge Moscovici was thus able to offer us two major theories dealing with very different aspects of interaction and social thinking. In this article, we will explore the reasons for the success of one of them, namely the theory of social representations. For this purpose, we will review the history of this theory, its initial postulates and the work done to disseminate it.

History and filiations
In 1961, Serge Moscovici (Figure 1), a young French researcher of Romanian origin, published in the Presses Universitaires de France the results of a vast survey on French opinion about psychoanalysis. This survey conducted as part of a thesis in psychology examined how the 'general public' adopted psychoanalytical concepts. Based on this work, Moscovici will lay the foundations of his theory of social representations. In doing so, he draws inspiration from the old concept of 'collective representations' proposed in 1898 by the French sociologist Emile Durkheim. This concept had probably been the most striking phenomenon in the social sciences in France at the very beginning of the 20 th century. Despite this, it fell into disuse for nearly fifty years. However, Moscovici took it on in order to develop it and his work aroused the interest of a small group of social psychologists who would contribute to the revival of the concept of representations (Abric, 1976;Codol, 1970;Flament, 1971). They saw this as an opportunity to approach the problems of their discipline in a new and original spirit. The study of the diffusion of knowledge, the relationship between thought and communication, and the genesis of common sense formed the elements of a new programme that has become known today. But from its emergence to current social research, the concept of collective representations has undergone many metamorphoses that have given it different colours and forms. It is this history that we will try to trace here. The filiation of French sociology at the beginning of the 20 th century. Any attempt to reconstruct the past of this concept necessarily starts with sociology. Undoubtedly, Simmel (1908) was the first to recognise the relationship between the separation of the individual at a distance from others and the need to represent them. He defended the idea that the way a person represented oneself shaped reciprocal action and the social circles that they formed together. In a different sense, Weber (1921) made representations of 'a frame of reference' and 'a vector of individual action'. He described a common knowledge as possessing the power to anticipate and prescribe the behaviour of individuals. But the true inventor of the concept was Durkheim (1893Durkheim ( , 1895Durkheim ( , 1898, since he defined its contours and recognised his right to explain the most diverse phenomena in society. He defined it by a double separation. Firstly, collective representations are separated from individual representations. The latter, being specific to each individual, are extremely variable, transitory, ephemeral and constitute a continuous flow, whereas collective representations are outside the realm of becoming and are impersonal. Secondly, individual representations have individual consciousness as their substratum, whereas collective representations find their ground throughout society. Such representations are therefore homogeneous and shared by all members of society. Their function is to preserve the bond that unites people, to prepare them to think and act in a consistent manner. It is for all these reasons that they are collective, persisting over time through generations and placing a strong cognitive constraint on individuals. For Durkheim, the objective is clear: collective thinking must be studied in and for itself. The task is to transform the forms and content of representations into an independent domain in order to declare and prove the autonomy of the social. This task, according to Durkheim, falls on social psychology, which is still in a formative stage and the object of which seems ill-defined. However, in the beginning of the 20 th century, it was mainly sociology, anthropology and ethnology (Lévi-Strauss, 1962;Lévy-Bruhl, 1922;Linton, 1945;Mauss, 1903) that made use of the concept of representations, mainly for descriptive purposes, in the study of various collective representations of cultural or ethnic communities. And it was only in the early 1960s that, following Durkheim's intuition and based on the assumptions of child psychology (Piaget, 1932) and clinical psychology (Freud, 1908(Freud, , 1922, Serge Moscovici (1961) began to develop the social psychology of representations. Considering that Durkheim's conceptions leave relatively little room for the interactions between the individual and the collective, he suggested replacing the concept of collective representation with a more restricted concept of social representation. According to the author himself, it is about "...transferring to modern society a concept that seemed to be reserved for traditional societies", in response to the "...need to make representation a bridge between the individual and the social world, in order to then associate it with the perspective of a changing society..." (Moscovici, 1989. P. 82). This evolution occurs through two fundamental changes in relation to Durkheimian conceptions. On the one hand, Moscovici argues that representations are not products of society as a whole, but rather the products of the social groups making up that society. On the other hand, he is focused on the communication processes, which are believed to explain the emergence and transmission of social representations. The first point makes it possible to conceive a social thought that remains overdetermined by the structures of society, but also by the inclusion of individuals in these structures. Thus, we can explain the fact that in a given society there are different social representations of the same object. The second change introduced by Moscovici makes it clear that through communication and the accompanying processes of influence, normalisation and conformity, individual beliefs can be the object of consensus at the same time as collective beliefs can be imposed on the individual.
However, the concept of social representations will still have to go through a latency period before mobilising the vast stream of research we know. The deployment of the theory could take place only after the elimination of several epistemological obstacles, the most important of which was the dominance of the behaviourist model, which denied any validity of considering mental phenomena and their specificity. The decline of behaviourism and the emergence of 'new look' in the 1970s, followed by cognitivism in the 1980s, gradually enriched the 'stimulus-response' (S-R) paradigm. As a result, internal psychological states, conceived as an active cognitive construction of the environment and dependent on individual and social factors, received a creative role in the process of behavioural development. This is what Moscovici expresses when he argues that representation determines both stimulus and response, in other words, "that there is no separation between the external universe and the internal universe of an individual or a group" (Moscovici, 1969. P. 9).
The influence of the North American pioneers of social cognition. When Solomon Asch published his early work in 1946, he suggested the idea that certain cognitions played a particular role in our impressions of others. A little later, Heider (1958) explained that individuals tried to maintain a certain consistency in the way they perceived and evaluated elements of their social environment. At the same time, Festinger (1957) suggested the existence of the principle of cognitive coherence, which prompts people to seek a certain harmony between the different cognitions they have about themselves. For Festinger, an incoherence between these cognitions creates a state of dissonance that triggers rationalisation efforts aimed at restoring coherence in the subject's cognitive universe. Further, Heider laid the foundations of his attribution theory by distinguishing between internal and external causalities. Finally, with their tri-componential model, Rosenberg and Hovland (1960) proposed describing attitudes in three interrelated dimensions (cognitive, affective and behavioural). To all these pioneers of social cognition, it seemed obvious that the knowledge people had about their social environment was organised into structured sets. And it is this organisation that explains why this knowledge is both numerous and easily accessible. Contemporary with these first works on social cognition, the theory of social representations was inspired by this epistemic provision. For Moscovici (1961. P. 27), "representation is an organised body of knowledge..." More precisely, a social representation can be described as a set of elements (information, opinions, beliefs) between which individuals establish relationships. From this perspective, social representations will be viewed as cognitive structures.

Initial assumptions
When Moscovici formulated his theory (1961), he first of all wanted to offer a description of the genesis and development of social representations. According to him, the emergence of a social representation always coincides with the emergence of an innovative situation, an unknown phenomenon or an unusual event.

PERSONS OF GLOBAL SCIENCES
This novelty of the object or phenomenon implies that the information about it is limited, incomplete and highly dispersed in the different social groups involved in its emergence (what Moscovici calls 'information dispersion'). This object causes concern, attracts attention, or disrupts the ordinary course of life. Thus, it motivates intense cognitive activity aimed at understanding or mastering it, or even defending oneself against it (the phenomenon of 'pressure for inference') and causes numerous controversies and interpersonal and media communications. This leads to the pooling of information, beliefs, hypotheses or speculations, resulting in the emergence of majority positions in various social groups. The emergence of consensus is facilitated by the fact that individuals treat information on the object or situation selectively, focusing on a particular aspect according to the expectations or orientations of the group (the 'focusing' phenomenon). The gradual emergence of a representation, which occurs spontaneously, is therefore based on three types of phenomena: dispersion of information, focusing and pressure for inference. But these phenomena themselves develop against the background of two major processes defined by Moscovici: 'objectification' and 'anchoring'.
'Objectification' refers to the way in which a new object will, through communication about it, be rapidly simplified, imaged and schematised. Through the phenomenon of 'selective construction', various facets of the object are extracted from their context and sorted according to cultural criteria (not all groups have equal access to information about the object) and normative criteria (only that which corresponds to the group's value system is preserved). In this way, the various aspects of the object are separated from the field to which they belong to be appropriated by the groups which, by projecting them into their own universe, can better control them. These selected elements form what Moscovici calls a 'figurative nucleus', that is, a coherent pictorial whole that reproduces an object in a concrete and selective way. By penetrating into the social body by means of communications, through collective generalisation, this schematisation of the object substitutes itself for the reality of the object and 'naturalises' itself. A representation is then created and takes the status of evidence. It constitutes an 'autonomous theory' of the object which will be used as a basis for orienting judgments and behaviours about it.
'Anchoring' completes the objectification process. It explains how the new object will find its place in the pre-existing system of thought of individuals and groups. According to the elementary knowledge production method based on the principle of analogy, the new object will be assimilated into already known forms, familiar categories. At the same time, it will become part of an already existing network of meanings. The hierarchy of values of various groups will form a network of meanings, based on which the object will be located and evaluated. Consequently, the object will be interpreted differently depending on social groups. Moreover, this interpretation will extend to everything that at least somehow concerns this object. Thus, each social group attaches the object to its own networks of meanings, which are the guarantors of its identity. In this way, a very vast set of collective meanings of the object is created. In addition, the object becomes a mediator and criterion of relations among groups. However, and this is an essential aspect of anchoring, this integration of novelty into an already existing system of norms and values does not go smoothly. The result of this contact between the old and the new is a mixture of innovation, due to the integration of the previously unknown object, and persistence, as this object reactivates habitual frameworks of thought so as to incorporate it. It follows from this that a social representation always seems to be innovative and persistent, agile and rigid at the same time. Finally, it should be added that, when the processes of objectification and anchoring come to an end, the representations will be organised according to three dimensions, which are also elements of their analysis or comparison. The 'information' corresponds to the contents of the representations, which may be more or less numerous and diversified (we then talk of 'rich' or 'poor' representation). The 'field' designates the organisation and hierarchisation of the information contained in a representation (two representations can have the same contents, but be organised and hierarchised differently). Finally, 'attitude' refers to the polarisation of the contents of a representation (in this case we talk of 'positive' or 'negative' representations).
On this general theoretical basis of the process of the genesis of social representations, a wide stream of studies has been developed, initiated, in particular, by the work of Denise Jodelet (1989). These studies are primarily focused on descriptions of social representations as systems of meaning that express the relationships that individuals and groups have with their environment. Considering that representations are formed, first of all, in interaction and contact with discourses circulating in the public space, these studies consider language and discourse using two complementary approaches. Social representations are approached as both inscribed in language and as functioning itself as a language because of their symbolic value and the frameworks they provide for coding and categorising the environment of individuals.
The so-called 'monographic' and 'qualitative' approaches to the collection and analysis of discourse and practices (ethnographic techniques, sociological surveys, historical analyses, in-depth interviews, focus-group, discourse analyses, documentary analyses, verbal association techniques, etc.) constitute the main methodological basis of the work conducted in this context (see, for example, Kronberger, Wagner, 2000;Wagner, 1994;Wagner et al., 1999).

Reasons for success
After the publication of Moscovici's seminal work, the theory of social representations would slowly but surely gain a foothold in the social sciences, outside the exclusive confines of social psychology. Since then, the interest of researchers from other disciplines in the theory has not faded. Probably because the initial postulates formulated by Moscovici were relatively flexible and thus made it possible to adapt to problems that were quite far from those that social psychology usually deals with. Here are three examples.
The first is represented by the work of historians who, wishing to go beyond the simple historiography of facts and events, began to take an interest in the forms of thoughts and beliefs characteristic of past eras. Then they placed the concept of 'mentality' at the centre of their preoccupations. Borrowed from Lévy-Bruhl (1922), this concept referred directly to the concept of mental representations in relation to the interactions of the social field. However, it is clear today that the project for the history of mentalities is returning to the project for history of social representations.

PERSONS OF GLOBAL SCIENCES
The second example, which we would like to briefly mention here, concerns geography. First with the introduction of the concept of a 'mind map' (Gould, White, 1974;Downs, Stea, 1977), then with the idea of a certain subjectivity in relation to space (Tuan, 1975) and, finally, with the postulate that it is appropriate to focus on the mental processes that contribute to the perception of space, but, above all, endow this space with meanings and values. This gives rise to a 'geography of representations' that ultimately considers representations as determinants of spatial practices (e.g. Bailly, 1995;Lussaut, 2007).
Finally, we should mention some work done in linguistics and, more specifically, in linguistic didactics, where it became necessary to understand the meanings associated with the study and use of a given language. This concern has become central to plurilingual situations because of the identity issues they can raise. The concept of 'linguistic representation' emerged (Dagenais, Jacquet, 2008), directly inspired by the theory of representation and denoting sets of beliefs related to languages, their uses, and the groups that use them (Dagenais, Jacquet, 2008).
Thus, these examples suggest that outside the field of psychology, when researchers ask themselves about the cognitive determinants of practices, they find a conceptual framework in the theory of social representations that can be adapted to their problems. This is possible due to the breadth of coverage offered by this theory and, in our opinion, is one of the reasons for its successful application in the social sciences.
But the success of the theory of social representations is also associated with Serge Moscovici's efforts to disseminate it in the scientific community. We can retrospectively try to reconstruct the strategy he used for this purpose. Basically, he regularly relied on very small groups of researchers whom he encouraged to undertake work on social representations in their countries. It was a long-term project, and it was not until the 1980s that it began to bear fruit. In the United Kingdom, the theory of social representations gained a foothold under the influence of Robert Farr and Miles Hewstone. In Austria, Wolfgang Wagner played this role, in particular, by contributing to the development of a European network of researchers around the journal Papers on Social Representations. In Italy, first under the influence of Augusto Palmonari, then Felice Carrugati, the work of Anna Maria de Rosa contributed to the implementation and dissemination of the theory in all European countries. This implementation was then based on a bi-annual conference which began in 1992, and then on the creation of the Social Representations and Communication Network which would then focus the European Doctorate on Social Representations. By means of its Summer Schools, this network will make it possible to train doctoral students throughout Europe, including Russia, where today some researchers have followed these programs (e.g. Inna Bovina, Elena Volodarskaya).
Across the Atlantic, mainly in Latin and South American countries (especially Mexico, Brazil, Argentina and Venezuela), the theory of social representations has found enormous room for expansion since the 1990s. The impact of social, historical and cultural contexts on the formulation of Latin American scientific problems is a major factor in this success. Researchers in social psychology have discovered creative, reflexive and critical thinking in response to political, economic and social transformations and crises. Today they are actively involved in the development of the theory of social representations by linking it to other social psychological issues such as social memory or social change processes.
On the other hand, it should be noted that, in this international scene, the United States is the main absentee. Despite the remarkable work of Gina Philogène and Serge Moscovici in their attempts to disseminate the theory of social representations within North American social psychology, it should be said that the latter could not find any real basis for development. There are many reasons for this. The most obvious of them is undoubtedly the relative weakness of the original theoretical arguments and the almost exclusively French publication of the first works. But besides this, there are deeper and meta-theoretical reasons that for a long time considered the theory of social representations and the course of social cognition alien to each other. Among these reasons, the most significant is the difference in the levels of analysis at which research is conducted in the two fields. Traditionally, the theory of social cognition has dealt with intra-individual processes underlying social interaction, while the theory of representations has been concerned with inter-individual phenomena, which affect individual consciousness. However, this trend began to change in the early 2010s when the theory of social representations appeared (Rateau et al., 2011) in The Handbook of Theories of Social Psychology published by Paul Van Lange, Arie Kruglanski and Tory Higgins. In 1984, it was estimated that the cognitive dissonance theory, which we said was a great theory, had generated nearly 1000 publications for 27 years (Cooper, Croyle, 1984). However, in 1996, Pierre Vergès conducted a census of publications on the theory of social representations and counted more than 2000 references over 35 years. Today, in 2020, searched all the databases available on the American Psychological Association (PsycINFO) website for scientific articles containing the term 'social representation' in their titles or keywords. Figure 2 presents the result of this search and shows that the theory of social representation has been steadily expanding its presence in the scientific community since its birth.

Conclusion
Serge Moscovici would have turned 95 today. He bequeathed to us what can be considered one of the main theories of social psychology. We were barely personally acquainted, but we know that he was well aware of his success. However, he did not consider his work completed and continued to suggest new ideas about social representations until the end of his career. Many of his suggestions remain untapped to this day. They constitute a real reservoir for the researchers of tomorrow. We can only hope that they will be able to take them up and to continue the work initiated by one of the greatest figures in the social sciences at the turn of the 20 th and 21 st centuries.
Prof. Pascal Moliner wrote this article in French especially for our journal at the invitation of the Editorial Board.
Translation into English by Dr. Inna B. Bovina