Social Representations, Alternative Representations and Semantic Barriers


Alex Gillespie
Department of Psychology
University of Stirling
Stirling FK9 4LA


Social representations research has tended to focus upon the representations that groups have in relation to some object. The present article elaborates the concept of social representations by pointing to the existence of “alternative representations” as sub-components within social representations. Alternative representations are the ideas and images the group has about how other groups represent the given object. Alternative representations are thus representations of other people's representations. The present article uses data from Moscovici's (1976/2008) analysis of the diffusion of psychoanalysis to examine how people engage with alternative representations. It is demonstrated that there can be more or less dialogical relations with alternative representations. The analysis concludes by considering seven “semiotic barriers” which work to neutralise the dialogical potential of alternative representations, thus on the one hand enabling groups to talk about the views of others, while, on the other hand, remaining unchallenged by those views.

In Psychoanalysis: Its Image and Its Public Moscovici (1976/2008) introduces the concept of social representation by discussing the heterogeneous nature of modern society. Durkheim (1898) had used the concept of collective representations to study hegemonic and homogenous representations, such as myth and religion, in traditional societies. The interesting point about Durkheim's concept of collective representations, which Moscovici (p. 3) highlights, is the “mixed” position of this concept between psychology and sociology. On the one hand collective representations are produced by societies, yet on the other hand, they are part of individual psychological functioning to the extent of structuring the minutiae of perception. For example, a person fully socialised in a religion perceives the world through the lens of that religion: events can become the visible manifestations of the will of the given god. The concept of collective representations raises, for social psychology, a profound question: Exactly what is the relation between society in general, or social relations in particular, and psychological function? This question is central to Psychoanalysis and the present article.

Moscovici (2008) makes a case for revitalising the concept of collective representations, but under the new name of “social representations.” The reason for the shift of terminology is because the state of knowledge in late modernity is more pluralistic and fragmented than theorised by Durkheim. Moscovici (2008, p. 5) describes myth, or collective representations, as they existed in previous times as “total” while “social representations are, for modern man, no more than one of many ways of understanding.” The structural differentiation of society has created an increasing number of niche groups each with its own discourse, while globalisation has brought knowledge systems previously separated by geography into close proximity. The conjunction of these two factors means that “we are witnessing the emergence of very heterogeneous political, philosophical, religious, and artistic practices” (Moscovici, 2008, p. 5). One can find examples of the clash of representations in the tensions between secular government and religion and in the relatively recent tensions between common sense and science (Bauer & Gaskell, 2002). In these and other contexts we are witnessing the breakdown of collective, homogenous and “total” knowledge structures and the emergence of a pluralistic field of representations. Accordingly people are now navigating between knowledges and discourses, choosing which is relevant for the given context, and able to defend that choice in relation to possible alternatives.

If, as Moscovici (2008, p. 6) states, “social representations both have an autonomous psychological texture, and are specific to our society and culture,” then what is the relation between the current conditions of plurality and that psychological texture? If we have moved from a sociological condition of collective representation to one of social representation, then what does this shift imply for contemporary psychological function? It might be tempting to assume that an increasing plurality of co-existing representations would lead to an increasing plurality of mind, or at least tolerance for alternative forms of knowledge, but the present article will show that such an assumption is not warranted.

The empirical research reported in Psychoanalysis concerns the movement of knowledge, namely psychoanalysis, from science into common sense. The question Moscovici asks is: how has psychoanalysis, a novel theory produced by a sub-group of scientists, diffused into French society? Methodologically the strategy is to compare how psychoanalysis has been appropriated by different groups, such as Communists, Catholics, students, professionals, and workers. His aim is not to reveal the denigration of scientific knowledge in common sense, but rather to examine the creative re-interpretation and appropriation of the knowledge as it is re-constructed for new ends.

The questions that I want to address in the present article build upon Moscovici's analysis. If there is such a plurality of social representations in contemporary societies, how do social representations themselves deal with this condition of plurality? How do these representations adapt to alternative representations? And specifically, at a psychological level, how do they enable individuals to negotiate the plurality of alternative and potentially competing representations which contemporary society confronts them with? How do they enable those who live by them to engage with alternative, potentially competing and contradictory representations? And, at a more general level, what psychological consequences follow from the contemporary sociological condition of a multiplicity of discourses and representations?

The present article begins by examining communication and social representation in order to make the case for the representation of difference. In so far as social representations are reflexive and enable communication with alternative representations, it is argued that within representations there must be alternative representations—that is representations of other people's representations. Alternative representations are thus presented as important dialogical sub-parts to certain social representations, enabling those representations adapt to the plurality of representations. The existence of these alternative representations is demonstrated by reanalysing data from Moscovici's Psychoanalysis. Focusing on propaganda, a form of communication one might naively consider as monological (Trey, 1998), it is shown that propaganda does not simply ignore alternative representations. The analysis reveals that propaganda can incorporate alternative representations, and even dialogically orient to those alternatives. However, semantic barriers are used to neutralise the transformative and dialogical potential of these alternative representations. The article concludes by considering how these semantic barriers enable propaganda to persist in a pluralistic world of competing alternative representations.


Social representations have two functions, they enable people to master their material and social worlds and they enable people to communicate (Moscovici, 1973, p. xiii, 2008, p. xxxi; Voelklien & Howarth, 2005). One can conceive of this definition as relating systematically to the different sides of the social psychological (self-other-object) triangle (Moscovici, 1972, p. 52; see also Bauer & Gaskell, THIS SPECIAL ISSUE; Marková, 2003). The instrumental function pertains primarily to the self-object relation, while the communicative function pertains primarily to the self-other relation.

This two-part definition is borne out in the empirical analysis presented in Psychoanalysis. The instrumental function is evident in the way in which people use psychoanalysis to understand themselves and others. Adolescents in particular are shown to use psychoanalysis to make sense of themselves. The communicative function is evident in the analysis of how people talk about psychoanalysis and in the circulation and elaboration of the representation in the mass media. Moscovici (2008) describes representations as being “a medium of exchange” (p. xxxi) and “instruments of exchange” (p. 121). They enable a group to coordinate itself in relation to the given object, and also to establish, or reinforce, the group identity and history. Within the analysis we find that Catholics, for example, have developed a shared language for talking about psychoanalysis in terms of confession. Talking about psychoanalysis in terms of confession gives mastery of psychoanalysis on the one hand while strengthening the position of Catholic institutions and practices on the other.

It is the communicative function of social representations that I want to focus upon. Moscovici (1994) has theorised the relation between social representations and communication in a separate paper. In that paper he suggests that social representations can be both explicit in the content of communication and implicit in the context—“buried under the layers of words and images” (p. 168). Sometimes the content of a representation is talked about and explicitly communicated, but at other times it is only evident if one interprets the assumptions being made within the communication. Van Dijk (1980) makes a useful distinction between the text of an utterance (i.e., what is actually said) and the context (i.e., everything in the background that informs the utterance, including presuppositions and the semantic frame which is assumed to be shared). Using the language of van Dijk, then, we could say that social representations are communicated in both the text and context of an utterance. They are communicated in the context when they are being used as a means of communicating, and they are in the text, when the representation itself becomes a topic of conversation.

The idea that communication is built upon a semantic context which the interlocutors assume they share was illustrated powerfully by Garfinkel (1984) in his analysis of a husband and wife talking about what was done during the day. What is actually spoken is so truncated that it is almost impossible for an outsider to make sense of the communication. Single words invoke whole meaning complexes. The actual communication is rich, although the words exchanged seem to be impoverished. This is possible because of the degree of shared semantic context between the husband and wife. The meaning arises not in the text, but in the relation between the text and the context.

Considering the shared context closely reveals that it is not sufficient to conceive of it simply as a collection of shared and agreed upon meanings. If the context were completely shared it would be questionable whether anything either needed to be said or indeed could be said. The shared context includes difference in the sense of divergent and alternative perspectives. Indeed, the following section argues that such difference, or more precisely the representation of such difference, is in fact central to communication.


While theorising social representations as the context of pragmatic communication, Moscovici (1994) makes an important and insightful point regarding difference. He points out that, although there needs to be a shared context for communication to take place, there also needs to be some difference between the speakers otherwise there would be nothing to communicate about. He writes:

[A] social representation is not completely shared, it is only partially distributed, just as part of the meaning of words is known to some people and unknown to others. Therefore everyone lacks some item of the knowledge that other speakers possess. [new paragraph] I can even add that if all people pictured things to themselves in a similar way, they would be nothing but mirrors engaged in specular conversations. In short, they would be a mass of individuals reproduced in thousands of exemplars, not a real society. In real societies, people routinely understand some statements as agreeing with their social representation and others as conflicting with it (Moscovici, 1994, p. 168).

Not only does there need to be a shared context against which something can be said, there also needs to be a difference between interlocutors which can become the basis of something to say. In the context of social life, difference is rarely neutral and most often manifests as tension and conflict. It is more often disagreements than agreements which keep people talking. Hannah Arendt (1958) made a similar point about communication being possible because of both sharedness and difference. In developing upon Arendt's work, Jovchelovitch phrases the interdependency in the following way:

It is because people are different—and yet the same—that action and speech become necessary: if we were all identical there would be no need to communicate and to act upon an unvarying sameness; if we had nothing in common at all the very process of speech would lose its basis and action would not justify itself (Jovchelovitch, 2007, p. 85).

The possibility of communication is born out of similarity, while the necessity of communication is born out of difference. It is this tension between similarity and difference that characterises the public sphere—the public space in which divergent interests clash and people negotiate coordinated social action. Social representations, as described in the previous section, provide the shared basis for communication. But, they are also the source of difference which makes communication necessary.

Although Arendt, Moscovici and Jovchelovitch use the concept of difference in diverse ways, and with distinct emphasis, they all recognise the importance for communication of not just sameness, but also difference. Focusing upon the nature of this difference, I suggest that communication entails not just difference, but also some representation of that difference. If there was difference without any representation of that difference, without awareness of that difference, then there would be misunderstanding, that is, a failure of communication (Laing, Phillipson & Lee, 1966). Indeed, the need to say something might not even arise. For example, in analysing why Person A asked Person B a question, it is not enough to say that they did this because Person B knew the answer. Person A asked the question because they assumed that Person B might know the answer (regardless of whether Person B actually does know the answer). That is to say, Person A represented a possible difference between themselves and Person B.

Let us consider an example from Moscovici's (2008) Psychoanalysis. Consider the following utterance, made by a liberal professional in response to a question about the popularization of psychoanalysis:

As far as the general public is concerned, these [psychoanalytic] ideas just disturb people and fill their heads with wild imaginings (p. 50).

There are three parts to this utterance. First, there is the question which motivated it, and that question implied a context of possible debate. The question made the assumption that some people would be for the popularization of psychoanalysis, and other people against it. Second, the first part of the utterance adds a contextual consideration for the answer to follow, it is narrowing the focus of popularization on “the general public.” Although the text of the utterance is “as far as the general public is concerned” and thus focuses the conversation on the general public, the need to make salient this focus implies that what follows might not be true for specialist groups, such as doctors or academics. Third, the final part of the utterance, which answers the question, takes a position in the contextual semantic field which is assumed to be shared (i.e., that the conversation is about being for or against the popularization of psychoanalysis to the general public). The position is that psychoanalysis should not be popularized, as it disturbs people by filling their heads with “wild imaginings.” The utterance stakes a position vis-à-vis the possible alternative (i.e., that psychoanalysis fills people's heads with sensible ideas). In stating a position (the text), the context (i.e. the alternative possibility that psychoanalysis fills people's heads with sensible ideas) is invoked. It is the existence of the alternative, as part of the shared context, which makes the utterance worth saying. Without the possibility that psychoanalysis gave people sensible ideas, in a context where psychoanalysis could only be conceived of as “wild imaginings,” then this utterance would have no meaning—it would be like saying psychoanalysis is psychoanalysis. In Billig's (1993) sense, this utterance is rhetorical. The meaning of the utterance lies in its positioning vis a vis what has not been said but assumed. The utterance assumes difference and the speaker, to some extent, is aware of that difference and assumes that any interlocutor is also aware of those differences.

Each of the three aspects of this seemingly simple utterance is based upon difference. The interviewer assumes a difference in asking the question. The speaker assumes that the interviewer may take what is about to be said as pertaining to all groups, thus assuming a different interpretation to the one intended. And finally, the speaker takes a position that is staked against an assumed difference, namely that psychoanalysis fills people's heads with sense. Thus, within an utterance we can see that at each step, when something is said, there is an implicit representation of difference—which is part of the context which is assumed to be shared by the interlocutor.


The representation of difference, which along side a shared context lies at the heart of communication, can also help us to address one of the questions posed at the outset of this article, namely, how representations enable people live amongst a plurality of potentially competing representations. In order to theorise this I suggest a sub-component of social representations which can be called “alternative representations.”

Alternative representations refer to the representation of a potentially competing representation from within a social representation. They are evident whenever we hear the phrases “they think” or “they claim” or “they say.” Alternative representations only exist within social representations. They are that part of a social representation that orients to the social representations held by others. They are representations in the sense of being coherent theories. They have a logic and an internal coherence. They may, if very elaborate, even have iconic cores which can be generative. However, they tend to be quite shallow and are popularly called “straw men.” That is to say, they usually simplify and stereotype the alternative. Alternative representations only exist as dialogical shadows within polemical or emancipated representations. They are shadows in the sense of reflecting, usually in a distorted and very simplified form, the social representation of other groups. They are also shadows in the sense that they are firmly attached to what the speaker wants to say, and are a sort of inverse of what the speaker wants to say. Alternative representations are “alter” in the sense of being attributed to other people and in the sense of being foreign objects within the given representation. Alternative representations are the Alter within the given social representation.

Alternative representations are peculiar because they can both destabilize and stabilize the given social representation of which they are a part. In so far as the alternative representation represents a real alternative to the main representation, then it can be destabilizing—posing a challenge to the core. Yet alternative representations can also protect the main representation from the challenge of alternatives. One could conceive of alternative representations as elements in the periphery of the representation that protect the core from counter arguments1. That is to say, they enable those using the given representation to avoid seeming naïve, to be aware of alternatives, and even to dialogically resist alternative arguments. In some cases, however, the core of the given representation is actually shaped by the alternative in the sense of it being first and foremost a position taken vis-à-vis the alternative.

Alternative representations are not representations of others, but the ideas which are attributed to real or imagined others. Of course there is a close relation between how we represent others and the ideas we attribute to them, but the focus of the concept of alternative representations is on the latter. It is because of this that they are alternatives to the main representation—because they have the same object as the main representation. In this sense, alternative representations are always second-hand. They are always attributed to someone else and thus are always bracketed, held in a state of disbelief, open to questioning and critique. In contemporary society, the very heterogeneity that Moscovici highlights in the opening of Psychoanalysis ensures that we cannot converse only with people who represent the world in a similar way to ourselves. Alternative representations enable communication between groups.

Examples of alternative representations abound in Psychoanalysis. In the interview excerpts, one can read interviewees talking about what other individuals and groups think about psychoanalysis. Alternative representations are attributed to family members, priests, the Pope, authors, fictional characters, Communists, Christians, Americans, French, Germans, Nazis, the mentally ill, neurotics, wealthy women, and people with strong personalities, not to mention psychoanalysts themselves. The point is that the context in which the interviewees are speaking is a heterogeneous field of simplified and usually denigrated alternatives.

By way of summary, alternative representations can be characterised in the following way. Alternative representations existing within social representations are a necessary outcome of the existing plurality of representations. They enable people subscribing to different representations to communicate with each other, even if only in a partial way. Alternative representations are potentially both destabilizing and protective. They can destabilize because they allow the alternative within, but they are protective because they stereotype that alternative and as such can aid ingroup members in fending off challenging arguments and even immunise them against conversion.


In order to illustrate the utility of the concept of alternative representations, it is useful to draw upon Moscovici's (1988, p. 221) distinction between hegemonic, polemical and emancipated representations. The aim of the present section is to characterise the manifestation of alternative representations within hegemonic, polemical and emancipated representations and to present illustrations from Psychoanalysis.

First, hegemonic representations are described by Moscovici (1988) as being contemporary variants of collective representations: they are coercive, uniform, and unchallenged. An example of a contemporary collective, or hegemonic, representation is individualism (Farr, 1998). It is a pervasive assumption which is often unquestioned in contemporary society. For example, it is embedded in our meritocratic institutions such as school and university, in which it is the taken for granted assumption that the unit of grading is the individual. Of course, however, contemporary society is also more concerned with rising individualism than ever before, and as such, it is a moot point whether even this representation is genuinely collective or hegemonic. This only underscores the questionable nature of collective representations, and underscores the need for the concept of social representations. In any case, we can characterise hegemonic representations as completely devoid of alternative representations. In this sense, hegemonic representations are completely egocentric, and they orient to and dialogue with no alternative perspectives.

Second, emancipated representations circulate with a degree of autonomy in society. Emancipated representations are forged in the context of a field of diverse alternatives. The field of health contains several emancipated representations, such as those originating with doctors, traditional healers, and patients (Wagner, Duveen, Themel & Verma, 1999; Foster, 2003).

The representations of psychoanalysis which Moscovici describes as circulating in the liberal press can be characterised as emancipated. The Catholic representation of psychoanalysis circulates side-by-side with a range of alternatives. Indeed, most of the newspapers analysed by Moscovici (p. 213) contain a diversity of views towards psychoanalysis—none is hegemonic. Rather each view is forced to come into contact with alternatives. And the use of humour and irony (e.g., p. 253 & p. 325) clearly demonstrate that alternative representations are in play. Emancipated representations, and their playful relation to alternative representations, are also evident in the interview data that Moscovici analyses. Consider the following excerpt from a mother's interview.

I've never tried to apply psychoanalysis in order to understand myself or other people. . . . People around me do, young people (my sons in particular take something of an interest in it), they've read lots of books on the question and they try to use psychoanalysis to understand the way their little sister behaves. And they criticize me for not bringing them up properly, for giving them complexes. I think all mothers should be given a little handbook explaining the basics of psychoanalysis, if only to stop their children from criticizing them in that way (Moscovici, 2008, p. 118, underlining added to indicate the invocation of the alternative representation).

This mother takes up a position outside of psychoanalysis, which, she says, she has never tried to use either to understand herself or anyone else. The relevant point for the present analysis is that although she does not use psychoanalysis, she is able to attribute a working model of psychoanalysis to her children, and, through them she is in effect using the theory. She is using the theory to understand what her children think of her parenting. She reveals the alternative representation when she says: “they criticize me for not bringing them up properly, for giving them complexes”. The idea that not brining children up properly leads to complexes is a common interpretation of psychoanalysis, and in relation to this mother, it is the alternative to her rejection of psychoanalysis. It is not necessarily a representation which this mother subscribes to, but it is a representation that she uses to understand her children's views on her.

Third, polemical representations, which are evident in Moscovici's analysis of propaganda, are elaborated in the context of an inter-group ideological conflict. Polemical representations tend to have one major alternative representation, which is not a real alternative, but rather a rhetorical counter point. The Communists’ representation of psychoanalysis is a clear example. The Communists were engaged in an ideological struggle with capitalism. Within this context, psychoanalysis was elaborated in a polarized manner, as a tool of capitalism, and thus as something to be resisted. In a polemical representation one would expect to find one alternative, or shadow, representation. Moreover, one would expect this alternative to be a caricature, a straw man, which serves only to reinforce the ingroup's representation. The following quotation from a Communist interviewed for Psychoanalysis illustrates the way in which an alternative representation manifests within a polemical representation.

Psychoanalysis claims to be a therapy, or even a worldview (amongst other things, it claims to be able to explain the origin and development of society in terms of conflict with the libido). It is in fact a falsifying tool that uses so-called complexes to cover up social conflicts. The use that is being made of it today, especially in the USA, is the best proof of that (Moscovici, 2008, p. 76, underlining added to indicate the alternative representation).

The direct representation of psychoanalysis, evident in this Communist perspective, is that it is a capitalist ideology which conceals social conflicts as problems within individuals which can be dealt with by interventions at the level of the individual. The alternative representation is presented in the first sentence which describes psychoanalysis as “claiming” to explain the origin and development of society in terms of a conflict between the libido and reality. The alternative representation indicates a working model of psychoanalysis, even if it is a caricature. And the alternative is evident despite it being rejected outright. The alternative is presented, but held at a distance: the alternative “claims to be able to explain.” This phrase works hard to separate the alternative from reality. The phrase “claims to explain” or even just “explains” would perhaps be insufficient to contain, and neutralise, the alternative.

In Moscovici's analysis of the way in which the Communist representation of Psychoanalysis is built upon the alternative (Moscovici, 2008, figure 15.1) we can see just how elaborate the alternative representation is. Yet, despite being built upon an opposition it is striking how little credence is given to the alternative, and how rigid the representation of the alternative is. The result is that despite the ever-present existence of the alternative representation, the main representation is not compromised and the alternative is simply dismissed. In short, there is no dialogical interchange between the main representation and the alternative. Strictly speaking, the coexistence of the main and alternative representation means that we are dealing with a case of cognitive polyphasia (Moscovici, 2008, p. 190), yet, these coexisting representations do not seem to be in any creative or productive dialogical relation. The dialogical relation is completely asymmetrical, and although ever present, the alternative representation is never given real voice. The question we need to ask, is: why is the alternative representation within a polemical representation resistant to dialogue?


There are many ways in which representations protect themselves from the potential change implied by dialogue with alternatives. Perhaps the most common way is through manipulation of who has access to the debate. Excluding interested parties violates the ideals of the public sphere (Habermas, 1989; Jovchelovitch, 1995), but is a very effective way of maintaining the symbolic order. However, I do not want to dwell on the ways in which the symbolic order is sustained through manipulations of the public sphere. Rather I want to remain at the semantic level, and consider the ways in which certain semantic structures inhibit dialogue with alternative representations. That is to say I want to focus upon the way in which particular meaning complexes can prevent dialogical engagement with alternative representations. Moscovici (2008, p. 334) refers to such meaning complexes as “semantic barriers” and in the present section I want to draw out the two main semantic barriers that he identifies, and then add to this a consideration of five additional semantic barriers.

1) Rigid Oppositions

In his detailed analysis of propaganda, and how its logic of representation operates, Moscovici (2008) identifies rigid opposition as a central semantic barrier. “Negation is,” he writes, “a natural feature of the cognitive structure of propaganda” (p. 315). The communist representation of psychoanalysis is built upon not just one, but many rigid oppositions: Marxist psychology is scientific while psychoanalysis is pseudo science, Marxist psychology enhances workers’ interests while psychoanalysis undermines their interests, and so on (see Moscovici, 2008, Figure 15.1). Although there are many distinct oppositions they are all part of a larger opposition, namely communism vs capitalism, and at a moral level, good vs bad. These oppositions are rigid in the sense that they do not allow rapprochements: each pole of the opposition demands “total support from the subject, or total rejection” (Moscovici, 2008, p. 325). The rigidity of this network of oppositions inhibits dialogical relations between the core representation and the alternative, because the relation between these representations is fixed a priori.

2) Transfer of Meaning

The second semantic barrier that Moscovici identifies concerns the transfer of emotion from core oppositions to secondary oppositions. For example, in the Communist press, the term psychoanalysis is linked to the term American in the often repeated phrase “American psychoanalysis.” This linkage, Moscovici (2008, p. 330ff) argues, transfers the richly elaborated and emotional opposition of American capitalism to American psychoanalysis. This categorisation of psychoanalysis as American is an oversimplification, of course, but in itself, this categorisation avoids the need for subtle thought—the known “facts” about America are transferred to psychoanalysis. This political frame for conceptualising psychoanalysis is loaded, none of the evidence needs to be considered, because the broader opposition to Americanism applies. This frame then acts as a semantic barrier inhibiting any potentially challenging dialogue with psychoanalysis.

3) Prohibited Thoughts

Moving beyond Moscovici's analysis of the semantic barriers, we can identify a third barrier in the form of negatively sanctioned or prohibited thoughts. This particular semantic barrier works to limit the extent to which people work to enrich their understanding of the other's representation. Examples include utterances such as “psychoanalysis is the work of the devil” (p. 135) and “[e]ncouraging people to look at themselves too closely is a very bad idea. It's the best way to create misfits” (p. 50). These meanings mark out the alternative representation as dangerous. To entertain or elaborate the alternative representation would be to either side with the devil or become a misfit. This mechanism inhibits elaboration of the alternative representation in part through fear. The phenomenon of heresy is an institutional mechanism operating to create the same kind of semantic barrier. This semantic mechanism has been described by Valsiner (2005) in his analysis of “promoter signs” which either promote or prohibit certain lines of thought. This semantic barrier gains its power from a logical extrapolation. Given the major premise that psychoanalysis creates misfits, and the minor premise that one does not want to become a misfit, it follows logically that one should keep away from psychoanalysis.

4) Separation

An alternative to prohibiting the elaboration of the alternative representation is separating it from the main representation. This semantic mechanism allows elaboration of the alternative, precisely because it ceases to be threatening to the core representation. The semantic mechanism of separation is particularly evident amongst Catholics who are keen to elaborate psychoanalysis, but need to ensure that it does not conflict with their faith. One Catholic says: “Faith is a different domain to science; there's no conflict” (p. 134). Another says “faith is grace and opens up the soul; certain problems to do with faith may come within the remit of psychoanalysis, but there's no absolute relationship” (p. 134). This latter quotation insists upon a separation between faith and psychoanalysis, yet also concedes a degree of relationship—arguably the semantic barrier is only partially successful. The interesting point about separation as a semantic barrier is that by virtue of separating the core from the alternative representation, and thus blocking dialogical interchange, this particular semantic block opens up the possibility of elaborating the alternative representation. These Catholics can proceed to find out about psychoanalysis and even engage in psychoanalysis confident in the assertion that it will not interfere with their faith. Exactly how they will respond when they read about Freud's interpretation of religion is unclear. Presumably this type of semantic barrier works in conjunction with a selective appropriation and elaboration of the alternative representation to ensure that no conflicts do arise.

5) Stigma

A fifth semantic barrier to dialogical relations between the main and alternative representations is stigmatisation. The stigmatisation here is of people who subscribe to psychoanalysis, especially the patients. People who go to psychoanalysts are thought to be “fashionable women” who “don't know what to do with themselves” (p. 64). Patients are described as “pseudo-hysterical women,” and sometimes there is the suggestion of sexual relations with the psychoanalyst. The patients, assumed to be women, are said to “lie on a bed, tell lots of stories about their lives and sometimes sleep with the psychoanalyst” (p. 64). The point to make about this stigmatisation of the patients is that it operates as a semantic barrier, not only discouraging people from becoming patients, but also from even beginning to take the patient role by, for example, engaging in self-analysis. The stigmatisation marks the alternative representation as clearly “for other people” and “not for self” and accordingly blocks any movement towards or even elaboration of the alternative representation. Stigma is similar to prohibition, but it works in a more subtle way. The stigmatising representation, in so far as it is a social representation, is shared by the ingroup, and as such, each individual within the ingroup knows how other ingroup members will react if they are seen to entertain the alternative representation. That is to say, this representation creates a structure of recognition within the ingroup which promises to withhold the recognition of the entire group to any one member who is seen to engage with the alternative.

6) Undermining the Motive

The sixth semantic barrier entails undermining the motives of the psychoanalysts. Moscovici (pp. 100–101) describes the representation of psychoanalysts as charlatans, conmen and sex maniacs. In response to the question: “What is your image of the psychoanalyst?” the responses were “a man who wants to make money,”“a charlatan who's rolling in money” and “a maniac who takes an interest in the sexuality of others because he is obsessed with his own.” The semantic content of this representation attacks the motives of psychoanalysts and importantly provides an excuse to discount the significance of what they say. If the psychoanalyst says that someone has a neurosis or complex, then it can be dismissed because it is in the interests of the psychoanalyst (as a greedy conman) to say this. Equally, if the psychoanalyst gives an explanation in terms of sexuality, this can also be dismissed as originating in the obsessions of the psychoanalyst. Thus again the alternative representation, in this case the views of the psychoanalyst, are blocked for entering into dialogical relation with the core representation. The mechanism of the barrier in this case, however, is particular: rather than prohibiting, separating or stigmatising, undermining the motive of psychoanalysts is particular to the content of psychoanalysis and provides ingroup members with widely applicable ad hominem arguments for dismissing the alternative representation.

7) Bracketing

The seventh semantic barrier operates at a textual level, by bracketing and holding in question alternative representations. Consider the following utterance from one of Moscovici's (Moscovici, 2008, p. 76) interviewees: “Psychoanalysis . . . claims to be able to explain the origin and development of society in terms of conflict with the libido.” This utterance holds psychoanalysis as a representation belonging to others in suspense and at a distance. Instead of saying that psychoanalysis “reveals” the origin of society to be conflicts with the libido, the speaker uses the term “explain.” While there can be only one revelation, there can be many explanations. Thus the use of this particular term places psychoanalysis as one explanation among many. But more than this, the speaker does not simply say “psychoanalysis explains” or indeed that “it claims to explain” instead the speaker goes for a triple bracketing with “claims to be able to explain.” Thus the speaker holds the alternative representation apart from the status of reality. Other linguistic phrases which serve this same semantic function include: “psychoanalysis claims that,”“so-called complexes,”“they think that,”“they say,”“they believe that,”“they try to explain” and so on. The terms “claim,”“so-called,”“think,”“say,”“believe,” and so on are used to manage the multiplicity of representations. These terms encourage one to take a certain representation as taken-for-granted, while putting alternatives into question. In short, these terms police the boundary of what is accepted as real.

The above analysis has identified seven semantic barriers that enable people to treat alternative representations as foreign bodies, isolated and quarantined from dialogue. “Dialogicality,” according to Marková (2003, p. 85), is the “capacity of the human mind to conceive, create and communicate about social realities in terms of the Alter.” The basic idea presented above, is that semantic barriers can hinder dialogicality, that is they hinder our capacity to creatively engage with the Alter in general, and the alternative representation in particular.

There are doubtless additional semantic barriers at work and which have yet to be identified. There are certainly other semantic barriers that operate in other fields of content, but which are not applicable to the study of the representation of psychoanalysis. For example, in cases of intergroup conflict it is common to dismiss the perspective of the other, and the alternative representation they espouse, by claiming that they are crazy, senseless and/or hell-bent upon destruction, and as such, there is no rational alternative representation to be elaborated.

The present analysis has focused upon semantic barriers to dialogue within and between people and groups. However, there could also be a corresponding analysis of “semantic promoters,” namely, the meanings which promote dialogue with alternative representations. For example, the absence of any of the seven identified semantic barriers could be described as promoting dialogue. If, for example, instead of prohibiting or stigmatising alternative representations they were promoted then there would be an incentive to elaborate alternative representations. Equally, if instead of making rigid oppositions there were less rigid oppositions not clearly aligned with good and bad, then again, potentially dialogicality would be stimulated. One could also add to this list the dynamics of irony and humour, which Moscovici (2008, p. 325) points towards as enabling a free play of representation.


One of the main questions in Psychoanalysis is the relation between the social and the cognitive. That is the question of how social relations and the plurality of representations in circulation shape our cognitive processes. Articulation of this relation has been a central goal for many sociocultural psychologists. Moscovici's contemporary contribution is to demonstrate that it is too simplistic to equate the increasing heterogeneity in society with an increasing heterogeneity of thought. The structural differentiation of society and globalisation are putting more heterogeneous representations into contact, and increasingly we have to learn to live in a world of alternative representations. Yet this does not mean that people become more open to alternative representations, more able to reflect upon their own representations, more tolerant, or more decentred in their thinking. The concept of “semantic barriers” is important because it enables us to see that there are additional factors at work which can constrain thought processes.

Research has shown that so-called traditional societies can be heterogeneous and stimulate considerable self-critique, irony and questioning (e.g., Pigg, 1996; Gillespie, 2006). And Moscovici's analysis of the Communist representation of psychoanalysis demonstrates that in contemporary society, a rigid representation can adapt to a context of plurality, can have embedded within it alternative representations, and yet this does not necessarily lead to self-questioning, let alone self-critique. Propagandistic forms of representation can still thrive in the context of a plurality of representations.

If propaganda were able to operate without consideration of alternatives, it would surely prefer this mode of operating. However in contemporary societies it is afforded no such luxury. Contemporary conditions, outlined by Moscovici in the opening of Psychoanalysis, mean that few social representations exist in isolation. In contemporary societies, representations cultivated in diverse domains and geographical regions are brought together. Accordingly there are invariably alternatives which cannot be ignored. This is why contemporary propaganda needs more subtle means of maintaining its privileged position. It needs to be able to incorporate alternative representations, thus enabling a semblance of rational dialogue and an ability to orient to the perspectives of others, yet the existence of an elaborate system of semantic barriers ensures that the potential for dialogical transformation is minimised. These semantic barriers are subtle. They do not simply dismiss the alternative. The semantic barriers enable the elaboration of an alternative representation. The Communists, for example, are able to operate with psychoanalytic concepts, to have some image of how psychoanalysts think, and thus would be able to hold a dinner table conversation with a psychoanalyst. The semantic barrier enables this whilst also neutralising the transformative dialogical potential.

Moscovici's (2008) analysis of the diffusion of psychoanalysis shows us not only that there is cognitive polyphasia, but also that there are varieties of cognitive polyphasia. Just because there are co-existing representations does not necessitate dialogical tension or transformation. The present analysis suggests that we need to conceive of a continuum between emancipated and polemical representations in terms of the degree of dialogue or tension between the main representation and the alternative representation. At the emancipated end, there is the potential for cross contamination and transformative dialogue between the main representation and the alternative representations. At the polemical end of the continuum, the alternative representation tends to be locked, by semantic barriers, into a rigid and often un-dialogical and uncreative relation to the main representation. These semantic barriers are one of the subtle means employed by polemical representations to adapt to contexts of a plurality of potentially competing representations. These barriers stand between the sociological and psychological levels, enabling intolerance to exist at a psychological level while plurality increases at the sociological level.


I would like to thank Ivana Marková, Sandra Jovchelovitch, Emmilie Aveling and Flora Cornish for their close reading and constructive commenting on an earlier draft of this manuscript.


  • 1

    I am grateful to George Gaskell for making me aware of this interpretation of alternative representations.