Скачать 291.93 Kb.
|GOOD AND BAD FOR SELF AND OTHER: FROM STRUCTURAL TO FUNCTIONAL APPROACHES OF FUNDAMENTAL DIMENSIONS OF SOCIAL JUDGMENT|
Catholic University of Leuven (K.U.Leuven)
Laboratory of Experimental social Psychology
B-3000 Leuven, Belgium
Paper prepared for:
Small Group Meeting on Fundamental Dimensions of Social Judgment:
A View from Different Perspectives.
Namur (Belgium) June 7-9
Trente Années de Psychologie Sociale avec Jean-Léon Beauvois:
Bilan & Perspectives
Paris June 27-29
I do not intend to read the present paper from the first to the last page, at a rate of two pages per minute. Actually the paper is to be regarded as a basket in which I have gathered some fruits grown over the years in my research garden. Some of the fruits may be overripe and have only historical relevance. Others may be fresh and ready for consumption. Some others may still be green, or they are mere seeds awaiting some fertile soil. In my presentation I will pick out one or two, maybe three fruits and try to sell the. Particularly I will point to some unpublished empirical research that may contribute to the unification of the field (e.g.: sections 1.3., 3.3., 3.4., and 4.).
There may be a need to unify the multiple two-dimensional models across the literature of which we all feel that they share some common ground. Do not expect (or fear) that unification be accomplished in this paper. At best I may have built some footbridges, or supplied some building blocks for a larger bridge. Therefore I found it useful to distinguish between structural and functional variants of the model. The prototypical structural variant is established on the basis of some multivariate technique that yields two dimensions as an acceptable model, and the dimensions are more or less arbitrarily labelled on the basis of a subjective essentialist interpretation, or simply by applying the item with the highest loading. The functionalist variants try to account for the model viewing it as a product of the ways individuals and societies function and subsist. In this respect, most attention goes to profitability theory, which is the functionalist theory I am most acquainted with and which has provided the background for my research. Many of the colleagues present will find connections with their own presentations at this meeting. This is most evident for the work of some researchers present in Namur and whom I am already acquainted with, particularly Laurent Cambon, Amy Cuddy, Nicole Dubois, Susan Fiske, Peter Glick, Nicolas Kervyn, Vincent Yzerbyt, Charles Judd, and Bogdan Wojciszke. Also the contributions of Wofgang Scholl and Charles Stangor seem more directly related than many of the other contributions. Nathalie Delacollette and Benoit Dardenne may see some connections with the "power is sexy" finding (2.2.2., and 2.3.1.). Joachim Krueger may find connections in 2, Nicole Tausch in 2.2.1, and Oscar Ybarra in 2.2.2., and 2.2.3.
When reviewing the various models, one may notice two outliers that do not stress two evaluative dimensions. First there's Peabody tetradic model stressing only descriptive dimensions, the evaluative aspect being a matter of optimal versus extreme positions along those dimensions (section 1.3.). Second, there's the one-dimensional evaluative model in which the two dimensions tend to fuse into a single one (section 4). In the oral presentation in Namur we shall look at some unpublished theory and research dealing with the integration of Peabody's tetratic model and the current evaluative two-dimensional models, and at two experiments set up to test the hypothesis that actors may process social information consistent with the two-dimensional model, but observers would tend towards one-dimensional processing. If time is left, I will propose an alternative explanation of mixed stereotype models that directly derives from the empirical validation of the relation pattern model developed to account for the joint cognitive organisation of traits and interpersonal relations (2.2.2).
Finally, it is beyond dispute that the a good deal of this paper relates to some main contributions to our discipline by Jean-Léon Beauvois and his numerous students and collaborators. One collaborator of the first hour is Nicole Dubois who has organised the session "La Connaissance Evaluative et les Composantes de la Valeur Sociale" at the meeting in Paris where I will discuss some of the issues dealt with in the sections on Utility Theory (2.3), Affordance Theory (2.4), and Profitability Theory (2.2, and 3).
The central topic of this paper concerns the nature and implications of a duality of valued concepts (categories, dimensions) underlying social cognitive information processing. They are referred to as communion and agency. However they have been implemented by different theorists working in different contexts emphasizing different aspects, developing different explanations, and using different labels. In the following I will first review some of those implementations arguing that they can be roughly divided into structuralist and functionalist approaches resulting in ditto models. Then I will focus on one of the functionalist approaches. Finally I will deal with some specific issues that concern both approaches, particularly the role of one- versus two-dimensionality about which some empirical research is reported.
Approaches called "structuralist" have focused on establishing the dimensional structure rather than on the psychological interpretation. The latter has often being limited to attempts to grasp the essence the dimensions' contents and to render them in strong essentialist labels. In this way one dimension has been characterized using labels such as communion, evaluation, social evaluation, likeableness, solidarity, morality, warmth, and so forth. The other dimension has been characterized as agency, dynamism, intellectual evaluation, power, status, dominance, competence, and so forth. Similar essentialist characterizations provide useful cues about the contents represented by the dimensions, indeed, but they do not explain yet why those dimensions would underlie the organization of social cognitions. Explaining the dimensions has been a primary concern of functionalist approaches that have focused on how the given dimensional organization of social cognition may contribute to the optimal functioning of the individual and society.
To a certain extent the structuralist/functionalist distinction is arbitrary. Like structuralist models also functionalist models characterize dimensions using labels that may not attempt to grasp the essence of a set of empirically established characteristics such as adjectives loading high on particular factors, indeed, but that attempt to grasp the essence of the dimensions' roles in individual or societal functioning. In this way, functionalist theories have assigned labels such as "other-profitability, other-relevance, and social desirability" to the mentioned "communion" dimension and "self-profitability, self-relevance, and social utility" to the mentioned "agency" dimension. Rather than defining neatly separated categories, the structuralist/functionalist distinction sets two reference points that help to organize the field comparing and interconnecting various approaches and related models. This is illustrated by the following brief and incomplete review of some main approaches/models.
1. Some (Predominantly) Structuralist Approaches
1.1 From Trait-Inference Models to BIAS Map
From Trait to Trait: Rosenberg's Implicit Personality Theory (IPT) Model. The IPT model concerns inferences from trait to trait. Using multivariate techniques, Rosenberg, Nelson, & Vivekanathan (1969) established their influential 2-dimensional model of IPT that accounted for inferences from trait to trait underlying the formation of impressions of personality in Asch's classic warm-cold experiments (Asch, 1946). The psychological interpretation of the dimensions was clearly an essentialist attempt to grasp dimensional contents, one dimension being labelled "social evaluation" the other "intellectual evaluation". Close examination of the reported outcomes leaves us with the feeling that "social evaluation" may characterize quite well the general flavour of the traits representing the first dimension, but that "intellectual evaluation" does only partial justice to the second dimension. For instance, the latter could just as well be labelled "self-confidence."
From Behaviour to Trait: Implicational Schemata and Cue Diagnosticity. Trait inferences from perceived behaviour have been investigated proceeding from Reeder & Brewer's schematic model of dispositional attribution (Reeder & Brewer, 1979) and the related cue diagnosticity model (Skowronski & Carlston, 1989). Numerous studies have confirmed that rules underlying those inferences vary between two inference content classes, usually referred to as "morality" and "competence" (for a recent review, see: Reeder, 2006). Once more, "morality" and "competence" are essentialist labels used to characterize large domains that seem to correspond to Rosenberg's social and intellectual evaluation.
Stereotype Content and BIAS Map Model. For a long time social psychologists seemed more interested in stereotype processes than stereotype contents. When about 15 years ago I submitted a paper on stereotype content structure (Peeters, 1993), the focus on content rather than process was a stumbling block for acceptance. I have been told that Phalet & Poppe (1997) experienced the same problem until they met an editor (me) who was highly interested in content. The breakthrough was achieved by Fiske and her team (e.g., Fiske, Cuddy, Glick, & Xu, 2002). A common stand is that stereotype content organisation involves the same communion- and agency-related dimensions, referred to in many studies as "warmth" and "competence", which suggests essentialist labelling connected with the structural approach. However, students of stereotype content structure have stressed functional aspects beyond the mere structural aspects. For instance, the studies mentioned have yielded convergent evidence associating manifest conflict the communion-related dimension (warmth). In this respect, a marvellous synthesis is the BIAS map model (Cuddy, Fiske, & Glick, 2007) in which the warmth and competence dimensions are connected with emotional and behavioural propensities.
1.2. From Semantic Differential to Relativistic Evaluative Meaning Model
Osgood and collaborators (e.g., Osgood, C.E., Suci, G.J., & Tannenbaum, P., 1957) established the well-known semantic differential, a structural model of connotative and/or affective meaning involving the three dimensions E(valuation), P(otency) and A(ctivity). P and A may collapse into a single dimension D(ynamism), resulting into a two-dimensional model in which E seems related to communion and D to agency. Although the semantic differential was not designed for person perception but to map basic structures of meaning in general, it may represent basically person perception, impersonal objects being assumed to be metaphorically dealt with as persons (Osgood, 1969; Peeters, 1986).
1.2.1. The Evaluative Nature of D
An important problem regarding matching E and D to communion and agency is that communion and agency are both evaluative dimensions, while in the semantic differential only E is assumed to be evaluative. Indeed, defining "evaluation" in terms of good versus bad, only the E-dimension deserved the essentialist label "evaluative" in that "good" and "bad" were found representative for that dimension only. A related problem is that the contents of the semantic differential dimensions seemed unstable making, for instance, that in some studies parts of D were absorbed by E. Those problems have been dealt with using a "relativistic evaluative meaning model" (Peeters, 1986). The model defines good and bad as agreement respectively disagreement with an evaluative standard. The standard may vary across contexts and evaluating subjects. In this way, "cold" may be good when applied to ice cream, and bad when applied to a cup of coffee, at least for most people who want their coffee hot. Some adjectives, such as good and bad, convey agreement or disagreement with the evaluative standard directly (direct evaluative meaning). Thus "good" may be defined as "high agreement with the evaluative standard." Other adjectives such as cold and hot carry meaning contents that are essentially descriptive and evaluatively neutral. However, when they are used to describe an object they can get a positive or negative indirect evaluative meaning according as they fit, respectively violate, the evaluative standard that applies to the object--e.g. "cold, sweet and creamy" for ice cream.
The semantic differential was established by factoranalysis of correlations among a wide variety of adjectives obtained from participants using the adjectives (scales) to describe a wide variety of common objects (concepts) including persons (e.g., mother). By this procedure, directly evaluative adjectives, such as good and bad, should load high on a factor together with adjectives that carry the same indirect evaluative meaning across a wide variety of contexts. For instance "pleasant" may fit the evaluative standard of almost any object possible, although there may be exceptions, e.g., a hard drug may be considered more dangerous and so more negative if it is described as "pleasant" than if it is described as "unpleasant." However, the impact of similar exceptions may have been limited in the construction of the semantic differential, which has resulted in the well-know E dimension represented by direct evaluative adjectives (good/bad) on the one hand and particular descriptive adjectives (pleasant/unpleasant) of which the indirect evaluative meaning hardly varies across contexts. Descriptive adjectives of which the evaluative meaning varies across contexts load on different factors such as P and A, or their combination D. The point is that also D is not evaluatively neutral but its valence varies across contexts. So "strong" and "active" may be positive when used to describe oneself or one's in-group, but negative when used to describe a rival or hostile outgroup. In this way the relativistic model explains why the compositions of the dimensions of the SD is not stable across contexts: when correlations among adjectives are not computed across a wide variety of objects (concepts), positive and negative values of adjectives belonging to D may not neutralize each other, but in some contexts the positive value may stand out, in others the negative value (e.g., strong-weak loading high on E in descriptions of athletes).
1.2.2. Evaluative Meaning as a composite of Direct and Indirect Meaning Components
The relativistic model may highlight the evaluative character of D and in that way the affinity between D and agency. A problem left, however, concerns the valence of D, and of agency by extension, when D (agency) is considered "in vitro" beyond any context. D, as well as agency, seems to involve fixed evaluative poles. For instance, "strong" is spontaneously viewed "positive" or "good", and "weak" is "negative" or "bad". This was explained as a positivity bias defined as a tendency to assume a positive context when no information about context would be available. Hence, the evaluative meaning of "strong" in isolation would reflect the indirect evaluative meaning of strong in "strong friend" rather than in "strong enemy." This explanation was empirically tested and confirmed, but it could not account for some of the outcomes (Peeters, 1992). In order to fully account for the data, it was to be assumed that adjectives carry indirect evaluative meanings that vary across contexts and stand out in adjectives beyond context, indeed, but in addition they often carry direct evaluative meaning that linearly combines with the indirect evaluative meaning. In this way, clever and cunning share the same neutral descriptive meaning that may be indirectly positive in the context "friend" and negative in the context "enemy." In addition, they carry different direct evaluative meanings, the one being rather positive, and the other negative. In this way the directly negative evaluation carried by "cunning" may neutralize the indirect positive evaluation endorsed by "cunning" in the expression "cunning friend."
Урок представляет собой практикум по говорению и аудированию по теме «Везение и суеверия» (Unit 16 Good luck, bad luck). Место урока...