I was not aware of Lance Fortnow’s viewpoint article on journal vs. conference publishing in Computer Science until Neil Ernst brought it up recently, and I’m glad I’m not the only one that thinks the primacy of conferences in our field is a problem.
For a while, during my first few years as a grad student, I even assumed that the conference model for dissemination of results was the norm across the sciences, and had found it quite strange: one submits a year’s worth of work, or more, to the premier conference in one’s field, to be accepted or rejected on what often seems to be an arbitrary basis and without even the ability to retort to the reviewers’ comments (ICSE, the main conference in my field, has a reputation of being a “crap shoot”; at least it now allows for author responses). Later I learned that the rest of the scientific world doesn’t work like that, and I still haven’t found a convincing reason for this eccentricity in Computer Science.
I’ve been lucky with my submissions so far, so I don’t complain about the conference model out of spite (maybe out of fear, as I know my luck is bound to change at some point). But it’s weird that I’m now on the final stretch of my Ph.D., with what I think is a decent publication record, and haven’t even attempted publishing in a journal yet. To be honest I don’t know why I haven’t done it —journals don’t have annoyingly hard page limits or deadlines, and they have better and more careful peer review processes. Conferences do allow us to meet and catch up with the people in our community, and we certainly need some, but we don’t need them to be as crucial to everyone’s careers, nor do we need so many of them. Not that I am yet in a position to do anything about it but to focus on publishing more on journals than on conferences myself.