Abstract:
The current thesis aimed to test the generality of the finding that language functions as a resource for influence within an intergroup comparative context. A further aim was to investigate the role of language content in the emergence of such an influence hierarchy.
Forty-eight male and female year-one students from several departments at Victoria University of Wellington were selected for the main study on the basis of their responses to a pre-test questionnaire. This questionnaire, which was administered to a sample of 169 students, surveyed views (pro vs. anti) on two ethical issues: capital punishment, and euthanasia. The issue of capital punishment, which was the more polarized of the two issues, was used as the basis for social categorization and intergroup discussion. Eight single-sex groups of six participants (three pro, three anti) discussed their views for 30 minutes before completing a post-discussion questionnaire. This questionnaire elicited social evaluations of social influence, prototypicality, social attraction, similarity and solidarity for both ingroup and outgroup members.
Expanding upon earlier work conducted within intragroup comparative contexts (e.g., Ng, Bell, & Brooke, 1993; Ng, Brooke, & Dunne, 1995), it was found that conversation functions as a resource for influence in an intergroup comparative context; that is, social influence was strongly and positively correlated with conversational participation (Hypothesis 1). Using Self-Categorization Theory (SCT, Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), it was predicted that the resource implications of language articulate with social identity processes through the relative prototypicality of speech acts (as a priori defined with respect to normative consistency). Specifically, interruptions encoded in prototypical speech acts were more strongly and positively correlated with social influence than interruptions encoded in non-prototypical speech acts (Hypothesis 2). However, there was little evidence for a positive relationship between evaluations of prototypicality and social influence (Hypothesis 3). It was further predicted and found that evaluations of prototypicality were more strongly and positively correlated with interruptions encoded in prototypical than non-prototypical speech acts (Hypothesis 4). In examining socio-emotional aspects of intergroup discussion it was found that speakers who were evaluated as more prototypical were perceived to have demonstrated greater solidarity for their ingroup and less solidarity for the outgroup than speakers who were evaluated as less prototypical (Hypothesis 5). However, little evidence was found for the predictions that social attraction would be positively correlated with either evaluations of prototypicality (Hypothesis 6), or prototypicality encoded in normatively consistent speech acts (Hypothesis 7). Answering two research questions, it was found firstly that conversational participation operates as a resource for evaluations of prototypicality; and secondly that social attraction varies as a non-linear function of conversational interruptions.
Although evidence for most of the hypotheses was stronger for male participants, post hoc analyses demonstrated that a number of the lesser correlations for females could be accounted for by three specific groups (2 anti, 1 pro). For these three groups there were strong negative or non-significant correlations between measures of conversational participation and social influence. It is argued that these particular groups withdrew from the argumentative aspect of the discussion task, and reinterpreted the social context in intragroup terms.
The findings are interpreted in light of theory that argues for conversation as a resource, and SCT. It is concluded that both the resource implications of language and social identity processes, as encapsulated in the prototypicality of speech acts, play a joint role in the group influence process. Consequences for the theorization of prototypicality and interruption meaning are discussed.