The other model (Ryan & Giles, 1982) highlighted two important dimen- sions of context,
namely a situation's status- and solidarity-stressing features (see suggested prototypes of these in
Figure 9.2) and its person and group cen- teredness, and the kinds of trait attributions that would
be triggered by them. Person-centered conversations are likely to focus on the individuals'
idiosyn- cratic experiences, impressions, and ideas while, in contrast, group-centered interactions
might include people who belong to the same ingroup discussing how to respond to an external
threat or challenge from a relevant outgroup (see Tajfel & Turner, 1986). Ryan and Giles
claimed this model was useful given that most studies about language attitudes have been
conducted in environ- ments that are classified as status-stressing, such as schools. They
suggested that research should be expanded to account for language attitudes and prac- tices in
other locations that would be considered more collective, includingcommunity centers, family
discussions, and so forth (see Giles et al., 2006). Interestingly, not only have these models not
been the subject of empirical scrutiny or further refinement, but the dimensions therein have not
resurfaced in other scholars' subsequent models (as will be seen). Nonetheless, we con- tend that
these dimensions warrant theoretical attention.
The need for and value of these organizational frameworks aside, Giles and Ryan (1982) were
not implying that theoretical models were not already in existence, as Lambert's (e.g., 1967)
initial development of the MGT was built on the following theoretical premise: language features
→ social cat- egorization trait attributions (see also, Robinson, 1972). Around the same time as
the clarion call for more theoretical development, Gallois, Callan, and Johnstone (1984)
speculated about which social categorizations would be the ones assuming salience in context.
They argued that when the social distance between prevailing ethnic groups in the community
was small or when listen- ers' ethnic affiliations were peripheral to their identities (Giles &
Johnson, 1986), other speech markers than ethnolinguistic ones (e.g., gender-related cues) might
assume prominence in dictating social categorizations and attend- ing trait attributions.
For their part, Berger and Bradac (1982), invoking their own uncertainty reduction theory,
proposed that there were other variables (e.g., perceived sim- ilarity) beyond mere social
categorization mediating the language-attribution relationship. In particular, they focused on the
ways in which listeners used language varieties as socially diagnostic cues to lend (oftentimes
stereotypi- cally) the speaker's characteristics more predictable (see also, Gudykunst & Ting-
Toomey, 1990). Importantly, Berger and Bradac suggested that listen- ers' use of language (and
other) features to reduce uncertainty about speakers allowed the former more relevant
information as to how to communicate with the latter more appropriately. In this same vein,
Giles and Powesland's (1975) early rendition of speech accommodation theory had
acknowledged the role of language attitudes in determining speakers' communicative choices.
More specifically, they argued that speakers' convergent or divergent inclinations toward or
away from each other was, at least in part, predicated on prior lan- guage attitudes they had
formed of their partners. Of course, the relationship between interpersonal accommodation and
language attitudes is more of a symbiotic process in that being the recipient of perceived
accommodative acts is part and parcel of listeners' social evaluative processes (Street & Hopper,
1982), and particularly in terms of whether such moves adhere or not to con- textual norms.
Bradac (1990) proposed that listeners respond positively (i.e., attributed high competence and
social attractiveness) to convergence, which conforms to established norms and the converse
when divergence violates them.
The strength of these early models lay in providing not only conceptual schemas in which to
embed language attitudes (e.g., contextual and normative dimensions), but also began to unpack
the ways in which listeners came about the language attitudes they reported. Thereafter, language
attitudes began to take on a more vital role in models of interpersonal communication (e.g., Giles
& Street, 1985) as well as miscommunication (Hewstone & Giles, 1986) more generally. All this
was manifest in terms of highlighting (language attitude) schemas as being prime resources for
constructing self- as well as couple- presentations in the company of others (Giles & Fitzpatrick,
1984). In tan- dem with these developments, more processual models of language began to
emerge that took into account interactions between the nature of the message, mediating
cognitions and affect, the immediate and larger-scale contexts and their histories, and wider
communicative outcomes and social consequences; it is to these we move next. While being
insightful, and sometimes spring- boarding off each other, processual frameworks seem to have
highlighted dif- ferent phenomena and processes and have not always, arguably, built upon each
other's insights to sufficient measure. Our integrative model ultimately attempts to resolve these
issues. While space precludes more details of the models to be overviewed next, their unique
features are highlighted as well as the relationships between them