Studies of community undertaken over a period of some forty-five years by the author and his colleagues are re-considered to show how the significance of « communities-in-the-mind » has been inadequately appreciated. The distinction made by Clifford Geertz between « experience-near » and « experience-distant » is used to sharpen up certain assumptions and approaches of community sociologists, including Frankenberg. Some possible explanations for the decline in the perceived importance of community studies from the late 1960s are discussed in the context of the growing centrality of social class in sociological analysis in the 1970s and 1980s. Reference is made to recent research on personal communities by Liz Spencer and the author to illustrate how an ellision between « experience-near » and « experience-distant » approaches may be achieved. It is concluded that the imputed community-on-the-ground, often based on materialistic assumptions, should not be conceptually privileged over the real community-in-the-mind.