ed from any kind of wider integument. Symbolic interactionists and ethnomethodologists, in particular, have no explicit conceptualization of the supra-situational, of social structure or culture, as societal phenomena, even less of inter-societal cross-cultural or world-systemic relations the proper macrocontext, rather than the 'nation-state'.8 In dealing with the internal ordering of society, familiar intermediate categories like 'institution', 'class', 'level', 'domain', 'primary' and 'secondary' notions widely used by 'members' are not conceptualized explicitly, though they are often smuggled in because life has a habit of spilling over the edges of inadequate theoretical boxes. Nor is there any workedout temporal framework: development, history, and evolution become nonproblems, because history has to be the history of some substantive entity, such as a society, and evolution the evolution of types of society in successive epochs . To work with a social theory that mirabile dietu has no concept of society as an entity (however much we need to avoid 'systems' theory assumptions that underplay internal inconsistency and overemphasize the liminality of external boundaries) makes it impossible to produce a sociology which can answer to most of the major problems of understanding social life. Only a decade after Homans had had to argue for bringing men back into sociology, we have to plead for bringing society back in. This is by no means to denigrate whole modes of doing sociology. Anyone who fails to respond to the fine-grained and sensitive illumination of the everyday and the interpersonal provided by Erving GofFman would be a poor human being as well as a poor intellectual, for to read him is to experience a deepening of perception. It has been a mind-stretching experience for a whole generation whom he has taught, like Brecht, to look at the familiar in new ways, and the exotic as common human practice. Yet the exploration of the devices people use to manipulate others in face-toface encounters, to present images of themselves, to detect meanings informing others' behaviour in clues and cues provided or given off, is, in the end, to work within the confines of formal sociology. The usual riposte to that kind of criticism is to assert that relations are between men, and that it is only alienated thinking that persuades us, fetishistically, that we have relations with reified 'things' or 'forces'. A second, more substantive, riposte is that this kind of analysis can, with equal effectiveness, be applied to 'high level' groups, and that the Cabinet or the Politbureau is, after all, simply a small group. This content downloaded from 40.77.167.104 on Sat, 09 Apr 2016 05:57:46 UTC All use subject to http://about.jstor.org/terms