The content of stereotypical traits used to describe groups has been the focus of research on stereotypes since its outset (e.g., Katz &amp; Braly, 1933). More recently, the Stereotype Content Model (SCM; Fiske et al., 2002) has proposed that groups are typically stereotyped along two key dimensions, i.e., as cold vs. warm and as competent vs. incompetent. However, for the SCM, which particular stereotypical traits are used depend on the position of the group in the social structure, meaning that (barring large-scale social change) stereotypical characterizations are seen as quite stable. By contrast, we propose that stereotypes are flexible communicational tools that can be used strategically. Thus, whilst we acknowledge the centrality of the dimensions proposed by the SCM, we argue that stereotypical description of groups along these dimensions may vary as a function of more localized factors pertaining to the specific nature of people's goals and the specific social interaction setting in which people are asked to do these judgments. <p>This idea can be derived from at least 2 important lines of research. On the one hand, research inspired by the Linguistic Category Model (LCM; Semin &amp; Fiedler, 1988) has shown how descriptions of others may vary as a function of group interests.</p><p>Through subtle language variations, communicators can make different inferences available to recipients and thereby influence the way they perceive social reality in ways that may serve to protect ingroup identity (e.g. depicting positive but not negative ingroup behavior as indicative of long-term dispositions, and vice-versa for outgroup behavior; see Maass et al., 1996). What is more, such descriptions may vary as a function of the communicative context, such as the type of audience which is addressed or the specific interaction goal pursued by communicators (e.g., Wigboldus et al., 1999; Semin et al., 2003).</p><p>On the other hand, research on the &lsquo;strategic' component of the Social Identity model of Deindividuation Effects (SIDE; Klein et al., 2007; Postmes et al., 2000) has pointed out how stereotypical descriptions of groups can be used to achieve specific political goals. The idea that stereotypes may fulfill social/political functions was already proposed by Tajfel (1981) when he argued that stereotypes allow for the &lsquo;justification of actions, planned or committed, against the outgroup' (p. 156; see also Jost &amp; Banaji, 1994). It is easy to grasp, for instance, how the paternalistic stereotype of Africans as childlike constituted a useful legitimizing tool for colonization (i.e. they need our help to develop). However, such idea has received little empirical attention even amongst the social identity tradition of research initiated by Tajfel. Moreover, Tajfel essentially viewed this as a macro-social process happening largely outside of people's awareness (Condor, 1990). By contrast, SIDE proposes that stereotypes can also be used intentionally and strategically in social interactions for similar political purposes. That is, by conveying different explanatory beliefs about groups and social reality (McGarty et al., 2002; Yzerbyt et al., 1997), stereotypes can be used to influence others as to how that social reality should be changed or maintained, in ways that are perceived to serve the ingroup's interests. Thus, for instance, Reicher et al. (1997) have shown how the expression of self-stereotypes by Scots (e.g., as communal or entrepreneurial) can be used strategically to promote support or opposition to Scottish independence (e.g., &lsquo;we' need to be independent vs. part of Britain to express our communal vs. entrepreneurial qualities; see also Klein &amp; Licata, 2003; Sindic &amp; Reicher, 2008).</p><p>The purpose of this project is to apply this general reasoning to the issue of immigration and the stereotypical traits used to describe immigrants. That is, we combine an interest in content, exemplified by models such as the SCM, with the strategic and communicational approach to stereotypes proposed by SIDE and the LCM. However, whilst building upon previous research, the present project is also innovative in a number of important ways.</p><p>In relation to research based on the LCM, the novelty of this project is twofold: (a) whilst such research has looked at whether or not people try to induce dispositional inferences and in which circumstances, we propose to focus on the specific nature of the dispositions that are induced (i.e. which specific stereotypical traits are used), and (b) we combine the idea of strategic language use with the broader reading of social identity processes proposed by SIDE, whereby descriptions of social reality can serve specific (group-level) political goals that are not limited to direct ingroup favoritism. </p><p>In relation to social identity research, the main novelty of the project is threefold: (a) it proposes to investigate a crucial but neglected facet of the social identity tradition, i.e. the social/political functions of stereotypes (Tajfel, 1981), but to do so through the lens of the strategic approach proposed by SIDE; (b) it aims to do so with resort to experimental methods that allow inferences about causality, in contrast to the few existing studies on this question which have relied on qualitative methodology (Reicher et al., 1997) or content analysis (Klein &amp; Licata, 2003); and (c) it aims to go beyond the cognitive/strategic divide of the SIDE model by shedding light on the relation between cognitive and strategic processes, that is, by examining whether or not strategic variability in the expression of stereotypes is reflected in changes of stereotypes at the cognitive level for the communicator, and whether such cognitive variability underpins or is the result of strategic variability.</p>