What is needed, yet all too often lacking, is a way of thinking about parts that are complex, but fall far short of constituting the sort of wholes that social theory posits. When Sigmund Freud tackled the idea of "group psychology," he may have had this goal in mind, but I find his collectives to be suspiciously similar to his conception of individuals. Sociologists and anthropologists show a similar failing, only coming from the opposite direction, writing of sub-categories that seem like their categories rewritten without a sense that in becoming "sub" they undergo qualitative as well as quantitative change. I don't think it should be that hard to conceive of groups that are distinct from both individuals and society as a whole. The challenge is to find a way of setting limits to our all-too-human impulse to convert the flesh of particularity into the air of abstraction.
This is where reflection on relationships proves handy. As is often the case, the way we live complexity is more complex than the way we think it. If we can manage to navigate our way through social networks in which the contrast between points of dense erotic and emotional investment and the diffuse cloud of acquaintance is richly textured, we should be able to conjure that in-between space conceptually through what social theorists, following Kant, term "rational reconstruction." That is, in describing what we already do, we have the potential to restrain our tendency to focus excessively, either on the individual or society as a whole. Perhaps musing on the practice of social networking in the New Media sense might help to advance this project along.