From Bill: “to say that computers are tools for the analysis of networks is completely ignoring the study of sociology, which developed the idea of a social network long before TCP/IP was invented.”
This is certainly correct. However, we should also keep in mind that computer networks allow for very different interactions between people than traditional sociology was used to (they are certainly catching up). Ubiquitous computing, as Howard Rheingold has written copiously about, changes everything about social networking.
I would also argue, from a philosophical point of view, that it is entirely possible that there are properties that will arise from ubiquitous computing and always on networks that we do not, as yet, have a grasp of, and that may be completely seperate from the study of the people USING the network. The network ITSELF maybe have emergent properties, and sociology is poorly placed in the academy to talk intelligently about communication theory outside of that done by people.