Carly Butler sets out to demonstrate the relevance of Harvey Sacks's (Reference Sacks1992) notion of membership categorization in the production of social action. She presents data from a New Zealand primary school, where she conducted extensive observations of children's school activities. In this book, she focuses on an original and fascinating sort of play, the “fairy club.” The fairy club is not a game, although it can contain various games; it is a sort of institution within the institution, in which children enact their own version of “school” under the guidance of a leading girl who is “teacher” and sometimes “mother” of the other fairy-children.

The first chapters of the book are a very clear and exhaustive, as well as compact, introduction to children's talk and play from an interactional perspective (chap. 1) and of Conversation Analysis and membership categorization analysis (chap. 2), covering, for each section, the seminal works and the most recent developments and positions. Sacks's theoretical work on membership categorization also receives a more in-depth treatment in the findings section, with regard to its relevance for actions such as compliments and complaints and for the organization of play.

I found Butler's description of the research process (chap. 3) exemplary and thoroughly enjoyable. She provides accurate as well as entertaining descriptions of the hazards of her observation, including how she dealt with the lack of consent for some of the pupils, the difficulties in getting useful data out of children's coming together and scattering around, and also her ingenious ways of involving them in the research process (to mention just one, she provided children with audio recorders in bags of different styles, so that they could actively choose the one best suited for their next activity). She reports on various stages of her perceived role as researcher, teachers' helper, or just caregiving adult, showing all the limits of the ethnographer's control over her public persona and her contract with the host institution. Similarly thought-provoking are her reflections about the children as gatekeepers in this kind of study, while typically being excluded from the communication of the research outcomes. Together with the theoretical overview, Butler's account of the fieldwork is a sufficient reason to recommend the book to advanced students who need an introduction to ethnomethodologically informed research, particularly, but not exclusively, for those who wish to undertake a study of children's activities.

At the end of chap. 2, Butler clarifies her intent as that of “foregrounding the notion of membership categorisation device not only as resource for establishing shared understanding … but as a locus for social action,” offering “an analytic foothold into the mutually informing relevance of categories and sequential organization” (p. 42). Ultimately, she wishes to reassert the significance of “shared common-sense knowledge” as revealed by the use of the membership categorization device against more circumscribed approaches to categorical work (notably that of Schegloff Reference Schegloff2007). The way she pursues such an analytical enterprise is by showing, for example, how the category of (fairy-club) “teacher” is relevant to the opening and closing of rounds of speech activities, or how that of “friend” underlies assessments and complaints.

Butler deliberately downplays any implications about children's “learning” and even the children-specificity of what she illustrates. She is concerned with membership categorization as an inherent and constitutive component of social action in general, and, for the specific social group she focuses on, this means revealing the seriousness of play, to frame it in a Goffmanian way. In no way can “fairy-club” can be seen as an imitation of, appropriation of, or preparation for the adults' world. This is one of the major strengths of Butler's contribution: In the course of her account we gradually come to see the children's world as an intricate society, with rules and rules-breaking that needs fixing, with enmities and alliances delicately managed, with duties which are complained about and rights which need reaffirming at every other step. In line with authors who have revolutionized the way we look at children's competences (I am thinking of Bruner Reference Bruner1968 and Ochs & Schieffelin Reference Ochs and Schieffelin1983, in relation to infants' cognitive skills and child language, respectively), Butler calls into question the idea of childhood as a preparatory stage of life and shows how children's everyday reality is the product of a machinery that is as complete and complex as that of adults, ultimately dissolving the notion of “adulthood” as the endpoint of a developmental march.

I found Butler's analysis both convincing and effective in highlighting the role played by membership categories in the sequential organization of children's activities, especially since she rigorously attends to the situated and always emerging nature of the categories themselves. However, in reaching the end of the book, one feels that the selective focus on membership categorization has left out much of the meat of the play, exposing, so the speak, only the bones of fairy-club as a social institution. As a consequence of her orientation to instances of membership categorization, as it happens, most of the episodes she presents have to do with tension over the assigned roles in the play. Most typically, we watch interactions in which the authority of the 6-year-old “teacher,” Polly, is either explicitly enforced by the teacher herself or resisted by the other participants, in fact essentially by the principal antagonist of Polly's authority, Tara, caught either in challenging Polly's instructions or covertly grumbling about them with her fellow club members. It may well be, as Butler remarks, that pretend play needs extra work for the categories to be maintained, to the point that a major part of the play itself may reside in the activity of implementing them. Yet the impression remains that “this much sustained attention to categorizations” (190) is an artifact of the analysis targeting precisely those moments. Although all analyses inevitably result in an artifactual version of their objects, the sieve of membership categorization seems better apt to catch the strain rather than the flow in the work of the social institution observed. Rather than with any flaw in Butler's own work, this has probably to do with the paradoxical nature of any attempt to capture shared commonsense knowledge. As linguistic anthropologists know too well, and in the words of one of them, “We are not going to find in language a faithful representation of everything a community thinks.… All the worse for the external observer who will not find in his magnetic tapes, however much listened and re-listened to, will not find traces of what everybody knows and takes for granted” (Cardona Reference Cardona1985:1–2).