While Goode's commentary has ably demonstrated the utility of the concept of the role, it has not come to terms with Whitam's remarks (1977) about the importance of sexual Whitam gave several examples of males who clearly had a preference for sexual relations with other males but who did not have knowledge of or contact with a gay community (1977, pp. 5; 7). I agree with Whitam that such males should be conceptualized as having a homosexual preference. Although the gay community does not cause the sexual preference in the Elrst place for such cases, the social norms distinct to the gay community, including the expectations surrounding the role, certainly do have effects on the behavior of those who frequent gay places. As mentioned previously (Omark, 1978), Whitam (1977) did not distinguish between sexual object preference and sexual orientation identity. He subsequently asserted (1978, p. 275) that self-identity as a gay male could emerge along with the preference, without any contact with the gay community. His examples do not show this, and his assertion is absurd. In the role theory perspective, a social identity (such as gay male) is directly connected with the corresponding social role. One may have many identities (e.g., gay male, insurance salesman, golfer, son, etc.), each of which is limited in salience by the frequency of participation in places where the role can appropriately be enacted and where a suitable audience can be found which is willing to respond to the role enactment by conElrming the presented identity (Goffman, 1959; cf. William James, 1890, p. 294). A gay community in a large metropolis which has a territorially delimited neighborhood and many gay social institutions and gay places can more easily encapsulate a gay member's activities and