Journal of Language and Social Psychology

A method for describing patterns of language use that establish social group uniqueness is proposed. Language use as an aspect of social identity is described in terms of (1) culturally specific vocabulary, (2) context-sensitive topics and (3) shared attitudes. As people become members of a social group and learn to share that group's identity, they acquire this ‘group talk’. An understanding of group talk analysed on such a three-part model discoverable via participant observation may also be useful in leading toward an understanding of the process of social categorisation as it plays a role in intergroup communication. Examples of establishing social identity are drawn primarily from observations in an American Indian community with illustrations also from familial, professional, and other subcultural situations.

This article presents a test of Clément's model of second language proficiency using the causal modelling technique. Data pertaining to attitudes, aptitude, motivation and second language proficiency were collected from 1,180 francophone students enrolled in grades 7, 9 and 11 in schools of eastern Québec (Canada). The results of the analysis using the LISREL V computer package support the hypothetical constructs and causal sequence posited by Clément. Furthermore, the generality of the model is supported by the cross-grade replicability of the obtained statistical solution. Finally, additional causal links and correlations obtained through refinement of the model suggest avenues for future study. These results are discussed within the context of current social psychological models of second language proficiency.

Factors influencing the ability of kindergarten and first grade children to detect inconsistencies in communications were investigated in two experiments. The first experiment revealed that detection was better when the inconsistency occurred between contiguous rather than interspersed statements but that performance was not influenced by the particular source of the story (an adult versus a puppet). The second experiment manipulated set (whether or not children were explicitly told that some passages would not make sense), passage length (short versus long) and passage content (animal versus non-animal). Main effects were obtained for each variable. The results suggest that children as young as five years of age can detect inconsistencies provided the communications presented to them do not exceed their processing capacities or concern topics which prompt a suspension of these capacities.