Since speech is a way to express group membership, people adopt convergence or divergence in communication to "signal a salient group distinctiveness, so as to reinforce a social identity".
Communication accommodation thus, becomes a tool to emphasize group distinctiveness in a positive way, and strengthen the individual's social identity.
In fact, people can both converge at some levels and diverge through others at the same time.
People use convergence based on their perceptions of others, as well as what they are able to infer about them and their backgrounds.
People can converge through many features of communication such as their use of language, their "pronunciation, pause and utterance lengths, vocal intensities, non verbal behaviors, and intimacy of self disclosures"(Giles and Smith, 1979, 46), but they do not necessarily have to converge simultaneously at all of these levels.
The last assumption puts emphasis on social appropriateness and norms. expectations of behaviors that individuals feel should or should not occur in a conversation".
Those expectations give guidance to people's behaviors, helping them to figure out the appropriate way to accommodate.
The communication accommodation theory has broadened this theory to include not only speech but also the "non-verbal and discursive dimensions of social interaction"..
Like speech accommodation theory, communication accommodation theory continues to draw from social psychology, particularly from four main socio-psychology theories: similarity-attraction, social exchange, causal attribution and intergroup distinctiveness.
Search for accomodating person:
Causal attribution process The causal attribution theory "Suggests that we interpret other people's behavior, and evaluate the individual themselves, in terms of the motivations and intentions that we attribute as the cause of their behavior" It applies to convergence in that convergence might be viewed positively or negatively depending on the causes we attribute to it: "Although interpersonal convergence is generally favorably received, and non-convergence generally unfavorably received, the extent to which this holds true will undoubtedly be influenced by the listeners attributions of the speaker's intent." Giles and Smith provide the example of an experiment that they conducted amongst French and English speaking Canadians to illustrate this.