The paper is mentioned over at the Monkey Cage, and available here. I present it without comment, other than to express agreement with the main points. From the introduction:

How should political scientists study the Internet’s influence on politics? Political science can surely help improve current public arguments about the Internet, which center around a few very general questions. Does the Internet exacerbate political polarization? Does the Internet empower ordinary citizens vis-a-vis political elites? Can the Internet help activists to topple dictators?

However, political science has paid little attention to the Internet until quite recently. This is changing. Scholars are beginning to uncover specific ways in which the Internet may affect politics, and to explore these relationships using both qualitative and quantitative data.

Thinking about the Internet in this way has some important implications. First—and most significantly—it suggests that one should not study the Internet as such. Instead, one should disaggregate it into more discrete phenomena, allowing scholars to ask research questions that they have some hope, however faint, of answering. One might do this in various ways. For example, different Internet-based technologies have different architectures, encouraging or discouraging different kinds of behavior (Lessig 1999). Thus, for example, one might plausibly study differences in linking practices across blogs, Twitter, and Facebook, exploring how variations in architecture and other factors lead to different outcomes. One could build on this to compare the consequences of these technologies with other means of communication, e.g., by comparing the diffusive force of audio-cassettes in the Iranian revolution and specific social media in the Arab Spring (Ritter & Trechsel 2011). Finally, one could begin to make clearer arguments about interactions between the Internet and other media where such media are important.