Abstract
The paper aims to develop an interactional account of illocutionary practice, which results from integrating elements of Millikan's biological model of language within the framework of Austin's theory of speech acts. The proposed account rests on the assumption that the force of an act depends on what counts as its interactional effect or, in other words, on the response that it conventionally invites or attempts to elicit. The discussion is divided into two parts. The first one reconsiders Austin's and Millikan's contributions to the study of linguistic practice. The second part presents the main tenets of the interactional account. In particular, it draws a distinction between primary and secondary conventional patterns of interaction and argues that they make up coherent systems representing different language games or activity types; it is also argued that the proposed account is not subject to the massive ambiguity problem.