Like a lot of people, I periodically comment on the work of others at conferences. At a recent conference, I noticed an odd pattern. At a typical conference, the talks consist of four main parts: A) the main presenter presents her work; B) the commenter presents some comments, criticisms, or suggestions; C) the main presenter responds to the comments; D) the audience members discuss the paper with the main presenter.
The odd pattern was that part (D) often proceeds as if parts (B) and (C) never happened. Nobody asks the commenter any questions. Nobody mentions anything the commenter said in questions addressed to the main presenter.* If somebody confides that he's nervous about giving comments, I tell him not to worry because nobody pays any attention to the comments.† This seems weird, since standard operating procedure is to employ commentators, and almost every conference does it. I started to wonder, though, what the point was. Would we lose anything if we got rid of comments at conferences altogether?
As it happens, I recently attended a conference that did not make use of prepared commentary--that did away with parts (B) and (C) of the standard formula. It seemed to me that the quality of the discussions were not harmed by this fact; the absence of commentators seemed to make no difference in the overall discussion whatsoever. What did seem to be affected was the number of people at the conference. It had the feel of a really small, 1-day conference, even though it was a multi-day several-concurrent-sessions kind of a deal. Maybe that's not bad: there's something to be said for intimacy. But usually you want a lot of people at your session if you're a main presenter. So one nice thing about having commentators is that it doubles the amount of people who attend the conference without doubling the amount of sessions.‡
*There are exceptions. This typically indicates a problem with the main paper. I've seen this occur only in cases in which the main presenter was responding in a specific and narrow way to some other paper, the author of the other paper served as commentator, and the commentator's paper was way better or more interesting than the main presenter's paper.
†This was an exaggeration, but was not hyperbole.
‡Having poster sessions is also an efficient way of increasing attendance. And I suspect that posters, though less efficient, are better than commentary, since posters are intrinsically more interesting. When I did a poster, I had lots of good discussions about it outside of the poster session itself; I've never really had a discussion about some commentary I was giving.