Communities of practice: 
Legitimacy not choice

Communities of practice has emerged as an alternative to other current sociolinguistic models such as speech communities and social networks, particularly in the area of language and gender. The valorization of non‐linguistic behaviours as adding further explanatory power to sociolinguistic models is timely: it has often been implicit in linguistic study (through ethnography) but rarely been given recognition. However, the types of self‐constituting communities of interest to sociolinguists are not the same as the communities of learning studied by Lave and Wenger (1991) and Wenger (1998) . If this construct is to be useful to sociolinguists, then the mechanisms by which it models access, gate‐keeping and its internal hierarchy need development. Using Eckert's (2000) Belten High data, and other work on adolescent talk, it is argued that gaining legitimate peripheral participation is a matter of sanction from within the hierarchy. Individuals do not have open access to communities based solely on their desire to be part of that community and to take part in its practices. While practices may define the community, the community determines who has access to that practice.