Technology, Conferences and Community

  • Jonathan Grudin

Communications of the ACM | , Vol 54(2)

Commentaries in the January, May, and December 2009 issues of Communications address a perceived crisis in Systems and other Computer Science fields [1, 2, 3, 4]. The discussion covers our shift from a traditional emphasis on journals to the current focus on conferences, and the challenges of conference reviewing, but at its heart is our sense of community. One commentary speaks of a “death spiral” of diminishing participation [3]. This Viewpoint considers how and why these changes occurred.

In a nutshell, the commentaries note that a focus on conference publication has led to deadline-driven short-term research at the expense of journal publication, a reviewing burden that can drive off prominent researchers, and high rejection rates that favor cautious incremental results over innovative work. Some of the commentators suggest that we return to emphasizing journal publication. To understand whether this is possible, and I doubt it is, we must understand why Computer Science in the United States shifted to conference publication in the first place. It was not simply that Computer Science requires quick dissemination of results: Conferences did not become prominent in Europe or Asia, or in other competitive, quickly-evolving fields such as Neuroscience or Physics.