I shall argue that for cognition to 'happen' it has to be 'coupled' with an external entity in a two-way interaction. In the case of human communication, one can refer to this process as 'social coupling'. Language in this context is the chief means by which action is brought about and is a tool to effect change. In my talk, I shall try to provide a sketch of some research work that has a bearing on this issue. The main themes that I shall address are (1) why and how people shape specific linguistic features of their communicative acts; (2) how are recipients' inferences shaped by specific linguistic features of a message; and finally (3) some of the implications of a social coupling framework for social psychological research. To this end, I shall start with a brief background to this work by distinguishing between (a) linguistic tools (e.g. interpersonal verbs and adjectives) and (b) their use. I shall shortly present the properties of interpersonal language (i.e. Linguistic Category Model) and then describe three different research fields that have examined strategic language use in message production and the impact of such messages upon recipients. These research fields are the 'Question-Answer Paradigm', 'Studies self-regulation processes and strategic language use and finally, the 'Linguistic Intergroup Bias'. In closing, I shall try to point out to some open questions and new directions that we are anticipating to take.