Empirical illustration of the hierarchical organisation of social thought: a domino effect?

 

Illustration empirique de l'architecture de la pensée sociale: un effet domino?

 

 

Rafael Pecly Wolter^{1, 2}; Chloé Gurrieri; Estelle Sorribas

Université Paris 5, Descartes, France

 

 

---

ABSTRACT

The "architecture of social thinking", a model presented by Rouquette (1996) suggests that various forms of social thought may be conceived as being part of a hierarchical organisation. Our study aims to show how that variation in a higher level of social thought (ideology) can trigger variations in the lower levels (social representations, attitudes and opinions). Our study compared two ideologically distinct groups (N=81): the first group (n=25) comprised participants who had reported a religious affiliation (Catholicism), while the second (n=56) comprised those who had reported no religious affiliation. We compared participant's representational and attitudinal fields towards a scenario on a charitable activity promoted by the Church. Results indicated that representations and attitudes may depend on higher levels. These findings suggest the hypothesis of a domino effect among the successive levels of the architecture of social thought.

Keywords: Social thought; Ideology; Social representations; Attitudes.

---

RÉSUMÉ

Les différentes formes de la pensée sociale forment une architecture hiérarchisée (Rouquette, 1996). Cette recherche vise à illustrer comment une variation au niveau supérieur de cette architecture (niveau idéologique) peut entraîner des variations des niveaux inférieurs (attitudes et opinions) en passant par le niveau intermédiaire (représentations). Dans cette recherche deux populations (N=81) idéologiquement distinctes ont été comparées: des participants se sentant catholiques (n=25) et des participants sans appartenance religieuse (n=56). Nous avons mesuré leurs représentations et attitudes vis-à-vis d'un texte fictif. Les résultats montrent que le champ représentationnel et le champ attitudinal, se situant à des niveaux inférieurs de l'architecture, sont affectés par la variation idéologique de la population. Cela nous conduit à avancer l'hypothèse d'un effet en cascade (effet domino) entre les différents niveaux de l'architecture.

Mots-clés: Pensée sociale; Idéologie; Représentations sociale; Attitude.

---

 

 

Introduced by Rouquette in 1973, the notion of social thought is often contrasted to scientific thought: the former is apparently being characterised by inadequacy and irrationality, as well as by all sorts of errors and biases, whilst the latter is supposedly productive, "truthful" and effective. Now, these biases and errors are consistent and meaningful. Given that the cognitive processes underlying them do not occur at random, and are linked to each other, their "occurrence regarding a given object is nearly always the mark of a particular social position" (Rouquette, 1998a, p. 33). Social thought could not therefore be reduced to distorted thought, but rather to a "different thought" (Guimelli, 1999, p. 18), with a particular logic.

Guimelli (1999) thus defines social thought as an autonomous and specific thought, independent of rational thought, but cohabiting with it, each one intervening in particular social contexts. This type of thought can be expressed in very different manners such as rumours, opinions, social representations or even ideologies.

Rouquette (1996) suggested an architecture of social thought, where certain forms of this type of thought are organised into a hierarchy depending on their stability (or intra- and inter-individual variability) and their degree of contingency (or level of integration).

Any real or imaginary object is perceived, represented and understood according to our social reference framework. This framework is intrinsically linked to the groups that we belong to; or in the words of Berger and Luckman (1996), we are witnessing a social construction of reality. The same object can be viewed differently, depending on the social construction of the various groups faced with this object. Ideology as a form of social thought would then contribute towards people's shaping of their social construction of the world; it would be one of the available keys to interpretation when faced with an object or event. Within the architecture of social thought, ideology comes into the highest level, in other words like a stable and contingent form of thought, located above social representations.

 

 

Deconchy defined ideology thus: "any organised or potentially organised, and therefore organisable group of representations and explanations of the world, especially of the world of social interactions, where the prime motive is not essentially verification" (Deconchy, 1989, p. 235). Deconchy is not alone in explaining what an ideology is via social representations (Flament & Rouquette, 2003; Ibanez, 1991; Rouquette, 1996; van Dijk, 1998). Ideology and social representations display a great diversity of approaches: according to Vergès (1991, p. 159) "reducing one to the other or confronting one with the other is dependent upon a theoretical stand". Depending on how these two notions are defined, the relations attributed to them will not be the same. Moreover, Jodelet (1991, p. 22) points out that:

in the scientific literature, we have only found one single model that puts representation outside ideology. That of F. Dumont (1974) challenging the overlap between these two notions, in a pluralist approach to the ideologies that relates them to the action.

In view of these considerations, in our approach, ideology would be the instancy that organises social representations, and the latter in turn would organise attitudes. In the words of Flament and Rouquette:

what explains the reason for an attitude or a set of attitudes is a social representation. For example, the communists' condemnation or rejection of psychoanalysis during the 50s (Moscovici, 1961), in other words, their negative attitudes towards the theory and corresponding practices, stemmed from their particular representation of that object . . . In other words, the representation in question created and grouped together a set of attitudes. (2003, p. 17).

Moreover, this relation supposedly exists only in the representations ® attitudes direction, and not the reverse; Moliner and Tafani (1997) showed in fact that an attitudinal change towards an object had no consequence on the representational level. For Rateau (2000), the representation protects itself from attitudinal changes using rationalisation mechanisms (the "canevasses of reasoning", cf. Guimelli & Rouquette, 1993; Rouquette & Guimelli, 1995). The architecture of social thought is not the only attempt at linking together these three notions (attitudes, representations and ideology): in 1982, Doise suggested that they refer to different levels of analysis, the social representations being a subsystem ensuing from the ideological system. The attitudinal level being more individual, it therefore entails an inter- or intrapersonal level. As far as Rateau (2000, p. 30) is concerned: "this perspective implicitly presupposes the existence of a hierarchical system in which ideology is above and attitudes are below social representations".

We can therefore deduce that if ideology organises a set of social representations and that the latter organise a set of attitudes, then attitudes are not independent of ideology. In other words, depending on our ideology, we will not reject (or subscribe to) a class of objects in the same way. In 2000, Rateau put the theory of this hierarchical relation to the test experimentally, and obtained conclusive results, since the refutation of a representation produces a change of attitude on the part of the subjects. A clarification is perhaps called for at this point: within this framework, we intend to explore the relations between different levels of the architecture of social thought, and not between the actual forms of social thought. This means that the forms of social thought dealt with in this article have only been used as examples, of particular cases, of the level to which they belong within the architecture. We will therefore be dealing more with the "ideological level" from every angle, rather than with ideology in the strict sense of the word, and with the "representational level" rather than with social representations. Indeed, we expect to find that a difference intervening at the ideological level will produce, at representational level, a different acti-vation of cognemes (Codol, 1969) relating to the object. According to Codol, the cogneme would in fact be the basic cognitive element corresponding to a unit of information, and to which a value is attributed. In this sense, the "representation", as we understand it here, would then be equivalent to the set of these cognemes and to their organisation.

In this study, we are interested in the links between the ideological level and the representational and attitudinal levels. From this, it follows that, in order to put forward the hypothesis of a different representational field stemming from a different ideology, we will only be interested in the activation of cognemes (and in their organisation) in response to an inductor. More precisely, we believe that, depending on the participants' ideological insertion, the object will be understood and represented differently, and this will have repercussions on the various levels of the architecture. The objective of this research is not to validate the architecture of social thought, which has already been done by Rateau (2000), nor to supply the proof of any influence whatsoever by ideology over the other levels of the architecture. In response to the preceding objectives, a series of experiments would have to be conducted, manipulating each architectural level, and then measuring the influence of the manipulation on the level immediately below. Our objective here was simply to illustrate this architecture, by adding to it the personal involvement factor (Boudon, 1986; Godelier, 1978; Touraine, 1973). This is not therefore a demonstration in the accepted sense of the word, but rather an exemplification.

We should stress that these various previously given definitions convey a point of view that is not the only feasible one. Indeed, several disciplines are showing an interest in ideology, particularly sociology (Boudon, 1986; Godelier, 1978; Touraine, 1973) and philosophy (Lagueux, 1986); however, our theoretical choice has fallen on the socialpsychology definitions, as we place ourselves exclusively in a sociopsychological approach regarding ideology.

Another frequently used variable in a socio-psycho-logical approach is personal involvement. If we return to the definition of social psychology provided by Moscovici (1984), in which the Ego, the Other and the Object form a triad, the involvement factor would fit into this triangulation (Flament & Rouquette, 2003) as the mediator of the Ego-Object-Other triangle. More particularly, this variable would intervene between the various levels of the architecture of social thought. As defined by Rouquette (1988, 1997, 1998a, 1998b, 1998c; Flament & Rouquette, 2003), personal involvement has three supposedly independent dimensions: personal concern, object valuation, and the perceived capacity to act.

The first dimension, personal concern, relates to the perceived personal concern with the object, which would correspond "to the estimated level of personal adequacy relative to the object" (Flament & Rouquette, 2003). This dimension can generally be measured by a scale ranging from "it concerns me personally" to "it concerns me no more than anyone else".

The second dimension, the object valuation, is tied to the importance that a person or group attaches to the object; in a way, it is a matter of the evaluation of the issue attached to the object. This dimension corresponds to a scale that can range from "it's a matter of life and death" to "it's a matter of no importance".

The last dimension, the perceived capacity to act, conveys the perceived position of the subject: either the latter sees himself as an agent and believes that his action is going to have a certain bearing on events, or, on the other hand, he sees himself as a patient and believes himself incapable of changing anything. This dimension can be measured on a scale that ranges from power-lessness to total control.

 

Methodology

Population

We questioned two totally comparable populations (students at the same university, attending the same degree course), but who had different ideologies: that is, a group of students who identified themselves as Catholics and a group that reported no religious affiliation. To start off with, this study included a total of 90 participants, who were questioned in May 2006 at the University of Picardie; subsequently, nine participants were excluded from the analysis as they professed to belonging to religions other than Catholicism. The results were therefore calculated at the end on a total of eighty-one participants (average age: 19.5; standard deviation: 1.29), out of whom twenty-five³ reported to belong to the Catholic religion, and fifty-six claimed to have no religious affiliation. This numerical difference between the two groups of participants reflects a wider social phenomenon, which is clearly demonstrated in the national statistics (the number of people professing to be Catholics has fallen drastically in France, particularly with young people⁴).

News Story

On the first page of the questionnaire presented to them, the participants had to read a five-line news story, describing a Church initiative to help the homeless.

Regarding the administration of the message to the subjects, the current theories in the field of information processing, stemming from research by Chaiken (1980), reveal two types of processing: systematic and heuristic. In this study, we will not foster this type of explanatory framework, given that we are trying to show the modes of influence of socio-historic factors, which, by definition, are rooted in and linked to a given society. Chaiken's ahistorical theory would therefore come within another framework.

Lastly, our objective was not to carry out a supplementary study on the difference in understanding a message that depended on which groups the participants belong to. The aim of this research was rather to study how ideology plays a part in the perception of an ambiguous message, via the cognemes activated at representational level, through the participants' adherence to or rejection of this message, and, as we shall see, through the personal involvement factor.

This (false) news story is made to be ambiguous. The message has in fact been constructed by juxtaposing two types of information: firstly, a distribution of free meals to the homeless outside a church, and secondly, an invitation, still intended for these homeless people, to attend a seminar on "Being a Catholic Today". To arrive at this definitive version, 3 different types of message were tested on 100 participants from the parent population. During this pre-test, after reading the message, the participants had to say to what point the initiative outlined seemed "positive" or "negative" to them (two Lickert scales were used for this, ranging from "Not positive at all" to "Totally positive" and from "Not negative at all" to "Totally negative"). For one of the three versions of the message, the median of scores for these two scales only diverged by half a point from the middle of the scale.

This pre-test of the message was a clear indication that in the reference student population (a mixture of all religious denominations), one of the three messages was seen to be ambiguous (half of the students viewed it as positive and the other half as negative). This message was therefore selected, and presented on the first page of the questionnaire as a "Reuters" news story.

Having read the "news story", the participants had to turn the page and answer questions evaluating the dependent variables. At the end of the questionnaire, standard identity-related questions (gender, age, socio-professional class, etc.) were asked.

Explanatory Factor

The only explanatory factor used, created a posteriori, was religious ideology in two forms: Non-religious participants (SAR = Sans Appartenance Religieuse) and Catholic participants (ARC = Appartenance Religieuse Catholique). Religious belief was chosen as it is part of the continuation of a great number of studies on ideological and operational levels, the latter through religion and through Catholicism in particular (Aebischer, 1991; Deconchy, 1987, 1989, 1991).

The decision on how to separate these two groups was based on the participants' answers to the question "Do you have a sense of belonging, or rather belonging, to the... religion?" The multiple choices offered included: "I have no sense of belonging to a religion" and choices corresponding to each of the other religions (only the answers of the participants reporting a sense of belonging to the Catholic faith were used thereafter).

Measurements

The first measurement was attitudinal, composed of three questions to which the subjects responded on a six-point Lickert scale: (a) the first question related to the participants' general attitude towards the initiative outlined in the news article; (b) the second related to their attitude towards the distribution of the meal; (c) the third related to the attitude towards the invitation to take part in the seminar.

The second measurement was that of personal involvement. For the first two measured dimensions of involvement, namely personal concern and valuation, subjects responded to a question by positioning themselves on a six-point Lickert scale. The third dimension of personal involvement, the perceived capacity to act, was assessed by a yes/no binary question asking participants if they were prepared to give up time in helping to carry out an initiative such as the one presented in the "news story"; this type of process has already been used in other studies (Gurrieri, Wolter, & Sorribas, 2007; Wolter & Rouquette, 2006). This question can be considered a good measure of participants' perceived capacity to act, as, by agreeing to help, they show that they feel that they are capable of acting on the object.

The third measurement was designed to reveal the contents of the representational field relative to the initiative presented. After reading the news report, the participants were asked to produce three hierarchically organized evocations. The message presented in this report is not an object of representation as such; it nevertheless refers to standard objects of representation, such as solidarity towards others, or even the Catholic religion [regarding the links between objects of social representation (Abric, 2002)].

In terms of data processing, the associations specific to the two sub-groups of participants were processed with the double criteria of Frequency and Rank in mind (cf. Vergès, 1992, 1994). By crossing these two criteria (Frequency, strong versus weak; Rank, weak versus strong) all the associations may be classified in a Table with four boxes.

Regarding the representation that the participants may have of the initiative presented, the associations given by a large number of people (strong Frequency) and with a rather weak Rank are considered significant and potentially structuring.

It must be pointed out that although we are referring to the theoretical framework of social representations, our object (the initiative described in the pseudo news story) cannot strictly speaking be considered an object of social representation. In fact, the initiative described cannot be considered a socio-cognitively salient object, and therefore cannot entail the emergence of social representations. Nevertheless, the method systematised by Vergès (1992, 1994) seems particularly relevant for separating, according to their degree of importance in the perception and interpretation of the message, the associations produced from a reading of the latter. It must also be pointed out that this process is not new: several studies in the theoretical field of social representations have used texts of news stories to activate the representational field, such as studies focussing on the ideal group (see especially Flament, 1984; Moliner, 1989, 1994; Rateau, 1995, 2000) or on GMOs (Marchand & Burguet, 2004).

Hypotheses

Generally speaking, between the two groups of ideologically-different participants, we expect to find differences intervening at two other levels of the architecture of social thought (Flament & Rouquette, 2003; Rouquette, 1996), namely the representational level and the attitudinal level.

Moreover, we expect that ideology will determine the different degrees of involvement shown by our two groups of participants.

To sum up, the participants belonging to the Catholic faith should be more involved, should have generally more positive attitudes and different representations concerning the initiative described, compared to the participants who had reported no religious affiliation.

 

Results

Attitudes

The scores on the attitude scales showed that, regarding general attitude, the ARC group (Catholic participants) had a more positive general attitude towards the initiative (4.64) that that of the SAR group (non-religious participants), whose attitude is close to the average (3.89; the middle of the scale is at 3.5). This difference is statistically significant (t [79]= -2.25; p=.03).

Regarding the attitude towards the distribution of the meal, the two groups do not differ (SAR group mean score = 5.84; ARC group mean score = 5.64; no significant differences in the t test).

Lastly, regarding the attitude towards the seminar, the ARC group has an attitude close to the middle of the scale (2.96), whereas the SAR group has a rather negative attitude (2.16). This difference is significant (t [79]= -2.5; p=.01).

These results indicate that although the two groups strongly approve and in a similar fashion of the aspect of the initiative connected with the distribution of the meal, they differ regarding the invitation to take part in the seminar and regarding their general attitude towards the initiative.

The fact that our two groups show no distinction regarding the attitude relating to the distribution of the meal could reflect the social desirability that is attached to this type of action (and which is present, regardless of participants' religious affiliation).

To sum up, according to these results, the Catholic group appears, generally speaking, to have attitudes that are more positive overall to the initiative than the non-religious group, as we expected on a theoretical level. Moreover, as we have seen, the difference found between the two groups at the level of the initiative in general is supposedly not due to the distribution of the meal, viewed positively by all the participants, but to the seminar (which polarises the differences between the two groups).

Lastly, the study of the correlations (r of Bravais-Pearson) between these three attitudinal measurements showed that the general attitude of the ARC participants had a correlation of .30 with the attitude towards the meals, and of .38 towards the seminar. This means that the general attitude correlated in a similar fashion with the two attitudes, or, in other words, it was influenced as much by one as by the other.

Conversely, with the non-religious participants, the general attitude correlated strongly (.61) with the attitude towards the seminar, and not at all (.06) with the attitude towards the meal. This showed clearly (cf. above) that the general attitude of the non- religious participants had been influenced mainly by the attitude towards the seminar, and that the attitude towards the meal had not contributed to the general attitude. It should also be pointed out that the correlation between the attitudes towards the meal and towards the seminar was positive (.33) with the Catholic participants, whilst it was non-existent with the non-religious participants.

Personal Involvement

Regarding personal involvement, the "Personal concern" variable shows that the Catholic participants feel closer to the initiative (Mean score = 2.88) than the non-religious participants (Mean score = 2.09). This difference is significant (t [80] = -2.49; p=.015).

Regarding the "initiative valuation" variable, the results show that this is valued in a similar fashion by the Catholic participants and the non-religious participants (test t n.s.): the initiative was viewed overall to be positive by all the participants (ARC Mean = 4.40; SAR Mean = 4.21).

 

 

Regarding the third dimension of personal involvement, that is the "Perceived capacity to act", 84% of the Catholic participants declared themselves prepared to give up some time to helping the initiative described in the text. On the other hand, this proportion fell to 53% with the non-religious participants. It can be seen therefore that the great majority of Catholic participants would be prepared to give up their time to this initiative, compared to just one out of every two non-religious participants. This difference is statistically significant (Khi2, dl=1, with Yates' correction = 6, p=.01).

To conclude, helping to carry out such an initiative appears to be consensual with the Catholic participants, whilst for the non-religious participants, this eventuality is potentially contentious. These results showed that the Catholic participants were more involved than the non-religious participants in the initiative presented in the pseudo-news. This difference in the level of involvement was due to personal concern and to the perceived capacity to act, with both groups attributing a similar value to the initiative.

Representational Field

The technique of free association used here for gathering evocations, which enables hierarchically organized evocations to be analysed (Vergès, 1992), is based on Marbe's Law, which supposes there to be a link between the Frequency of an evocation and how fast it is produced as a response (reaction-time). This verbal-association technique has served here to process the evocations while taking into account the cogneme notion introduced by Codol (1969): the pseudo-news in fact described various cognemes, which function as inductors, and which serve, in the participants, to activate other cognemes induced by this message (Rateau, 1995).

In this context, given that the evocations were made according to the message presented, and immediately after this had been read, it seemed equally important to us, when comparing these two groups, to disregard terms that were identical to those appearing in the text (in parentheses in the Table); indeed, our objective was not to concentrate purely on the cognitive and/or mnesic effects.

In this context, we will focus on the terms scoring a low Rank and high Frequency: these two related characteristics would in fact determine the salience and importance of the term within the population under consideration.

The following are the "Rank versus Frequency" Tables, relating to our two groups' representations of the initiative described.

 

 

A comparison of the evocations of the two groups appears to reveal some possible common points. The idea of Solidarity seems important for both groups, since it appears rapidly (its Rank is low) and frequently. Conversely, the term Religion only appears to be salient for the non-religious participants: this may signify that this group attaches importance to the fact that this initiative is linked to religion.

To conclude, in the quadrant that includes the most important and salient elements, the only notable difference lies in the presence of the term Religion with the non-religious participants (with the members of the other group, only the ideas Helping and Solidarity are present). Moreover, the idea of Help is of prime importance to the Catholic group, but secondary for the non-religious group; meanwhile, when salient, the term Solidarity is present but associated with Religion.

Regarding the terms that only appear with one of the two groups (shown in bold in Table 6 and Table 7), with the Catholic participants, these specific terms are only found in the quadrant corresponding to high Rank and low Frequency: these terms would have little salience and importance on the level of this group's own representation of the initiative. The terms that are specific to the non-religious group are: Poverty, Manipulation, Blackmailing, and Sect. It should be pointed out that the last three terms have a clearly negative connotation. These differences on the level of unique terms would be related to a sort of stigmatisation of the initiative on the part of the non-religious group. These results allow us to put forward the hypothesis that a subgroup may exist for whom Manipulation would be salient in the representation of the initiative described (Vergès, 1992, 1994). From this point of view, we can consider that the representation of the initiative is not altogether the same in the two ideologically distinct populations. The differences between the representational fields need to be nuanced, as the Rank/Frequency method does not allow us to attest with certainty that two fields are distinct. This method simply provides us with indices showing that the same message has not activated the same cognemes in both groups of participants, and that some of the cognemes are more salient (shared, sig-nificant) than others in the organisation of the field.

For a more sound comparison, it would be necessary to make experimental use of the model of basic cognitive schemes (Guimelli & Rouquette, 1992; Rateau, 1995; Rouquette, 1994), which would facilitate a quantitative study of the level of structuring of the two represen-tations.

 

Discussion

This research aimed to illustrate how ideological differences may lead to a differentiation in the participants' representations, attitudes, and personal involvement. If we examine the last factor, it can be seen that ideological differentiation seems to be behind a differentiation of two of its dimensions: perceived capacity to act, and personal concern. The Catholic participants feel closer to the initiative and are more likely to feel they can act on it than the non-religious students. Conversely, there is no differentiation between the two ideologically distinct populations when it comes to the valuation of the initiative, probably owing to issues of social desirability. Nevertheless, the ideological level would play a part in personal involvement, particularly influencing personal concern and perceived capacity to act.

Several empirical studies (Baggio, 2006; Flament & Rouquette, 2003; Gruev-Vintila & Rouquette, 2007) have demonstrated the role of personal involvement in "the implementation and dynami of social represen-tations" (Flament & Rouquette, 2003, p. 122). Since the ideological level organizes the representational level, and personal involvement plays a part in the structuring of representations, the hypothesis that personal involvement would be one of the mediating variables between these two levels seems likely to us. This would produce the following schema: ideology à involvement à representation.

The results concerning the representational field relating to the message show certain differences between the two fashions of conceiving the initiative described in the news article. Indeed, even though both groups seem to see the elements of Solidarity and Help as totally decisive in this initiative, they differ on the term Religion, which is salient for the non-religious participants and not so for the Catholic ones. This result highlights the fact that, for non-religious people, the linking of the event to the Church is paramount, whereas this does not seem important to the Catholic group. The two groups also differ regarding certain less salient terms, with the non-religious participants on several occasions using words with negative connotations such as "Manipulation", "Blackmailing", or even "Sect". This group therefore supposedly focuses on the two components of the message: the distribution of the meal (which would activate terms linked to "Help"), and the seminar (which would activate terms linked to "Manipulation"). The participants in the other group seem only to focus on Helping; in fact, no terms stigmatising the event described in the text were mentioned. The two populations would therefore differ at representational levels both groups see "Helping" as salient, but only the non-religious participants activate "Religion", "Manipulation", and "Blackmailing". It should be pointed out that these last terms would be due to a subgroup of non-religious participants, as they appear in the Low Rank and Low Frequency quadrant. In 1994, Vergès suggested that terms that produce a rapid response and for few people would be characteristic of a sub-population within the population being studied. These results would therefore show that ideologically distinct populations would activate different cognemes in the presence of the same message. As we have previously suggested, this influence could be "modulated" by personal involvement: ideology would lead participants to feel a greater or lesser degree of involvement, which would play a part in activating certain elements at representational level.

According to the architecture of social thought, the representational level would coordinate attitudes (Vergès, 1994). The results of this study would indicate that the two ideologically differing groups have different general attitudes towards the text. The non-religious participants have a rather neutral general attitude (with a score of 3.8, close to 3.5, at the middle of the scale), whereas the Catholic participants have a rather positive general attitude (with a score of 4.64; middle of the scale = 3.5), and, as a whole, are also favourable to the initiative. By examining the two sub-scales measuring the attitude towards two components of the initiative the meal and the seminar it can be observed that the two groups only differ in the attitude to the seminar. Both groups are in fact favourable to the distribution of meals, whilst the non-religious participants are more opposed to the seminar than the Catholic ones. In order to demonstrate this tendency, we examined the corre-lations between the three scales (Vergès, 1994). These scores show that the non-religious subjects' general attitude correlates strongly with their attitude towards the seminar (correlation of .61) and does not correlate at all with the attitude towards the meal (.06). On the other hand, the general attitude of the Catholic parti-cipants correlates as much with their attitude to the meal (.30) as it does towards the seminar (.38).

These results would therefore indicate to us that the two groups with their different ideologies also have different attitudes, differing in their general attitude and their attitude towards the seminar. This would indicate that the upper level of the architecture of social thought (the ideological level) would have an effect on the lower level of the architecture (in our case, the attitudinal level). This causal link, if one follows the model of the architecture of social thought, would moreover not be direct, since it would pass through the representational level. Personal involvement would figure as an inter-level moderating variable, that is, acting at different architectural levels. "It must be understood, more precisely, that personal involvement depends on the ideo-logical level . . . and that it intervenes as a determinant of certain properties of representations" (Gurrieri, 2007, p. 256). Based on these results, we can intuit the presence of a domino effect, since a variation in the ideology produces variations at every level of the architecture of social thought. Such a domino effect would therefore be in some way due to an intervening link between the various levels of social thought, ensuing from a variation at the first ideological level.

The domino effect hypothesis, which has not been put to experimental test here, seems nevertheless to be important for the theoretical advancement of the study of social thought, as it conveys the fact that a differentiation at the upper level of the architecture proposed by Rouquette would engender effects cascading down to the lower levels, passing through every stage: ideological, representational, and attitudinal/opinion related. As a result, one should take this domino effect into account and try to test its scope with a procedure that is as experimental as possible, to get a better understanding of the structure underlying different forms of social thought: the social groups' stances, opinions and representations.

 

 

These results are, of course, only exploratory and do not allow us to demonstrate the presence of a causal link between the different levels; it nevertheless seems to us that the architecture of social thought can effectively serve as a general explanatory framework for phenomena that, although observed at attitudinal or representational level, "are linked to forms of social thought that are ideologically situated . . ." (Gurrieri, 2007, p. 256). To test these theoretical hypotheses would require working in particular with a "standard" object of representation, well known in the literature, as well as with techniques that enable a more reliable delineation of the content and structure of the representations generated: for now, the results obtained lead us to believe that in order to apprehend social reality, it appears necessary to take into account the different levels of social thought, along with how they are linked. In particular, further research with a variety of populations and fields of objects should be carried out, in order to consolidate and generalize the tendencies shown within this framework. This study, we hope, will be able to form part of the continuation of existing studies (Rateau, 2000), and open the way to new research focussing on relations between "levels" and "forms" of social thought.

 

References

Abric, J.-C. (2002). L'approche structurale des représentations sociales: développements récents. Psychologie & Société, 4, 81-103.         [ Links ]

Aebischer, V. (1991). Vers une anthropologie des médecines naturelles. In V. Aebischer, J. P. Deconchy, & J. M. Lipiansky (Eds.), Idéologies et représentations sociales (pp. 183-204). Cousset, Switzerland: Delval.         [ Links ]

Baggio, S. (2006). Pensée sociale et risques collectifs: effets de l'implication personnelle sur la construction sociale des catastrophes naturelles. Unpublished doctoral dissertation, Ecole doctorale 261, Cognitions, Comportements et conduites Humaines, Université Paris Descartes, Paris.         [ Links ]

Berger, P., & Luckmann, T. (1996). La construction sociale de la réalité. Paris: Armand Colin.         [ Links ]

Boudon, R. (1986). L'idéologie. Paris: Fayard.         [ Links ]

Chaiken, S. (1980). Heuristic vs systematic information processing and the use of source vs message cue in persuasion. Journal of Personality and Social Psychologuy, 39, 752-756.         [ Links ]

Codol, J.-P. (1969). Représentation de soi, d'autrui et de la tâche dans une situation sociale. Psychologie Française, 14, 217-228.         [ Links ]

Deconchy, J.-P. (1987). Filtres idéologiques et anthropologiques, chez des "croyants" et des "non-croyants". In J. Marx (Ed.), Athéïsme et agnosticisme (pp. 99-120). Bruxelles, Belgium: Editions de l'Université de Bruxelles.         [ Links ]

Deconchy, J.-P. (1989). Psychologie sociale, croyances et idéologies. Paris: Méridiens Klinksieck.         [ Links ]

Deconchy, J.-P. (1991). Mécanismes idéologiques et représentations idéographiques. Essai de recherche expérimentale. In V. Aebischer, J. P. Deconchy, & J. M. Lipiansky (Eds.), Idéologies et représentations sociales (pp. 205-234). Cousset, Switzerland: Delval.         [ Links ]

Doise, W. (1982). L'explication en psychologie sociale. Paris: Presses Universitaires de France.         [ Links ]

Flament, C. (1984). From the bias of structural balance to the representation of the group. In R. Farr & S. Moscovici (Eds.), Social representations (pp. 269-285). Cambridge, MA: Cambridge University Press.         [ Links ]

Flament, C., & Rouquette, M.-L. (2003). Anatomie des idées ordinaires. Comment étudier les représentations sociales. Paris: Armand Colin.         [ Links ]

Godelier, M. (1978). La part idéelle du réel. Essai sur l'idéologique. L'Homme, 18(3/4), 155-194.         [ Links ]

Gruev-Vintila, A., & Rouquette, M.-L. (2007). Social thinking about collective risk: How do risk-related practice and personal involvement impact its social representations? Journal of Risk Research, 10(3/4), 555-581.         [ Links ]

Guimelli, C. (1999). La pensée sociale. Paris: Presses Universitaires de France.         [ Links ]

Guimelli, C., & Rouquette, M.-L. (1992). Contributions du modèle associatif des schèmes cognitifs de base à l'analyse structurale des représentations sociales. Bulletin de Psychologie, 45(405), 196-202.         [ Links ]

Guimelli, C., & Rouquette, M.-L. (1993). Note sur la formalisation des schèmes étRankes dans l'étude des représentations sociales. Cahiers Internationaux de Psychologie Sociale, 19, 43-48.         [ Links ]

Gurrieri, C. (2007). Approche structurale des représentations sociales du voyage. Rôle des thêmata et de l'implication personnelle. Unpublished doctoral dissertation, Ecole doctorale 261, Cognitions, Comportements et conduites Humaines, Université Paris Descartes, Paris.         [ Links ]

Gurrieri, C., Wolter, R. P., & Sorribas, E. (2007). A implicação pessoal: Um instrumento psicossocial para compreender a relação população-objeto. Psicologia em Estudo, 12(2), 423-432.         [ Links ]

Ibanez, T. (1991). Représentation des nouvelles technologies et détermination idéologique. In V. Aebischer, J. P. Deconchy, & J. M. Lipiansky (Eds.), Idéologies et représentations sociales (pp. 175-182). Cousset, Switzerland: Delval.         [ Links ]

Jodelet, D. (1991). L'idéologie dans l'étude des représentations sociales. In V. Aebischer, J. P. Deconchy, & J. M. Lipiansky (Eds.), Idéologies et représentations sociales (pp. 15-33). Cousset, Switzerland: Delval.         [ Links ]

Lagueux, M. (1986). L'idéologie et les stratégies de la raison. Philosophiques, 13(1), 179-177.         [ Links ]

Marchand, P., & Burguet, A. (2004). La construction socio-représentationnelle des OGM: bidimensionalité et communication persuasive. In J.-L. Beauvois, R.-V. Joule, & J.-M. Monteil (Eds.), Perspectives cognitives et conduites sociales: Vol. 9 (pp. 81-104). Rennes, France: Presses Universitaires de Rennes.         [ Links ]

Moliner, P. (1989). Validation expérimentale de l'hypothèse du noyau central des représentations sociales. Bulletin de Psychologie, 41, 759-762.         [ Links ]

Moliner, P. (1994). Les méthodes de repérage et d'identification du noyau des représentations sociales. In C. Guimelli (Eds.), Structures et transformations des représentations sociales (pp. 199-232). Lausanne, Switzerland: Delachaux et Niestlé.         [ Links ]

Moliner, P., & Tafani, E. (1997). Attitudes and social representations: A theoretical and experimental approach. European Journal of Social Psychology, 27, 687-702.         [ Links ]

Moscovici, S. (1984). Introduction. Le domaine de la psychologie sociale. In S. Moscovici (Ed.), Psychologie sociale (pp. 5-22). Paris: Presses Universitaires de France.         [ Links ]

Rateau, P. (1995). Dimensions descriptive, fonctionnelle et évaluative des représentations sociales. Une étude exploratoire. Papers on Social Representations, 4(2), 133-146.         [ Links ]

Rateau, P. (2000). Idéologie, représentation sociale et attitude: étude expérimentale de leur hiérarchie. Revue Internationale de Psychologie Sociale, 13(1), 29-57.         [ Links ]

Rouquette, M.-L. (1973). La pensée sociale. In S. Moscovici (Ed.), Introduction à la psychologie sociale: Vol. 2 (pp. 299-327). Paris: Larousse.         [ Links ]

Rouquette, M.-L. (1988). La psychologie politique. Paris: Presses Universitaires de France.         [ Links ]

Rouquette, M.-L. (1994). Une classe de modèles pour l'analyse des relations entre cognèmes. In C. Guimelli (Ed.), Structures et transformations des représentations sociales (pp. 152-170). Neuchâtel, Switzerland: Delachaux & Niestlé.         [ Links ]

Rouquette, M.-L. (1996). Représentations et idéologie. In J.-C. Deschamps & J.-L. Beauvois (Eds.), Des attitudes aux attributions (pp. 163-173). Grenoble, France: Presses Universitaires de Grenoble.         [ Links ]

Rouquette, M.-L. (1997). La chasse à l'immigré. Violence, mémoire et représentations. Sprimont, Belgium: Mardaga.         [ Links ]

Rouquette, M.-L. (1998a). La communication sociale. Paris: Dunod.         [ Links ]

Rouquette, M.-L. (1998b). Les communications de masse. In S. Moscovici (Ed.), La psychologie sociale (7th ed., pp. 499-516). Paris: Presses Universitaires de France.         [ Links ]

Rouquette, M.-L. (1998c). Sur la construction des mondes politiques. Bulletin de Psychologie 51(1), 41-43.         [ Links ]

Rouquette, M.-L., & Guimelli, C. (1995). Les "canevas de raisonnement" consécutifs à la mise en cause d'une représentation sociale: essai de formalisation et étude expérimentale. Cahiers Internationaux de Psychologie Sociale, 28, 32-43.         [ Links ]

Touraine, A. (1973). Les mouvements sociaux. In A. Touraine, Production de la societé (pp. 335-365). Paris: Seuil.         [ Links ]

van Dijk, T. A. (1998). Ideology: A multidisciplinary approach. London: Sage.         [ Links ]

Vergès, P. (1991). Représentation des nouvelles technologies et détermination idéologique. In V. Aebischer, J. P. Deconchy, & J. M. Lipiansky (Eds.), Idéologies et représentations sociales (pp. 159-174). Cousset, Switzerland: Delval.         [ Links ]

Vergès, P., (1992). L'évocation de l'argent: une méthode pour la définition du noyau central d'une représentation. Bulletin de Psychologie, 45(405), 203-209.         [ Links ]

Vergès, P. (1994). Approche du noyau central: propriétés quantitatives et structurales. In C. Guimelli (Ed), Structures et transformations des représentations sociales (pp. 233-254). Neuchâtel, Switzerland: Delachaux et Niestlé.         [ Links ]

Wolter, R. P., & Rouquette, M.-L. (2006). A influência de certos termos socialmente salientes (nexus) e da imagem sobre a percepção de um objeto social. Revista de Educação Pública, 15(29), 79-89.         [ Links ]

 

 

Received 06/06/2007
Accepted 15/09/2007

 

 

Rafael Pecly Wolter. Mestre em Individu Social et Environnement. Professor da Université Paris 5, Descartes, Institut de Psychologie, Centre Henri Piéron, Laboratoire de Psychologie Environnementale, Centre National de la Recherche Scientifique-(CNRS).
Chloé Gurrieri. D.E.A. en Psychologie Sociale et Environnementale. Professor da Université Paris 5, Descartes. Institut de Psychologie, Centre Henri Piéron, Laboratoire de Psychologie Environnementale, CNRS.
Estelle Sorribas. Université Paris 5, Descartes, France.
¹ Address: Institut de Psychologie, 71 avenue Edouard Vaillant, 92100 Boulogne-Billancourt, France. E-mail: rafaelpeclywolter@gmail.com
² Text translated by Andreea Ernst-Vintila.
³ The limited number of subjects certainly restricts the range of results related to the representational field.
⁴ Cf. Le Monde des Religions, Editorial "Vers un catholicisme minoritaire" (The world of religions, editorial, "Towards a minority Catholicism"), January-February 2007.