Presenting research: old and new

When giving a paper at a conference, or at a research seminar, should the work presented be complete or in progress? There seems to be no in principle reason why there should be a rule one way or the other. In my own discipline of philosophy, there appears to be a reasonably well established, though implicit, norm to the effect that one should present work to others only when that work is still in progress. (I’m setting aside seminars that are titled “work in progress”, where matters are quite explicit!)

I have been attending philosophy research seminars for more than ten years, having started in 1999, when I began my PhD. I was a sporadic attendee while still a graduate student, but for the last ten years, I must have seen at least 25 papers a year – probably double or triple that in years when I’ve been especially active at conferences. So I estimate that, to date, I have seen at least 300 philosophy papers. Out of that sample, I can think of only three cases in which the speaker explicitly told the audience that the paper being presented was already published. No doubt there are another few that I’ve forgotten. And no doubt there were other cases where the paper presented was in the late stages of publication, and would not be significantly changed between the presentation and publication. But these are, I expect quite rare. They are openly acknowledged only very rarely.

Much more typically, the speaker expresses the wish to receive useful feedback on the paper. Thanks for your comments, they’ve been tremendously useful. Or: I’m very glad to have the opportunity to present this paper, as I’m really struggling with the argument in section 3. You get the idea.

In my own case, I once remember having planned to give a paper at another department, and finding that, in between the invitation to speak, and the occasion to present, the paper was accepted at a journal in surprisingly quick fashion. I did not have time to prepare an alternative paper to present, so I reluctantly went ahead. I felt it was somehow improper to give a paper to an audience that was “fixed”. When people made suggestions or raised criticisms of the paper, I was genuinely interested in the questions, but I felt that the audience might expect that their comments would be reflected in the final version of the paper. But it was too late for that. I kept this from the audience, and felt that the whole affair was, to some small degree, in bad faith.

Over the last few years I have been attending more and more of the seminars in the Economics department at Monash, which has a thriving culture of research seminars, with roughly 3–5 every week. In general, the standard has been very good, and I’m sorry to say that, on average, I find them much more interesting than the average philosophy seminar. Though no doubt a large part of that is due to the sheer novelty – for me – of the methods used and questions addressed.

Although economics seminars – like those in philosophy – tend to involve presentation of work that is not yet published, I believe there is a significant difference in the expectation that the research will be influenced by the feedback received in the presentation.

  • It is common for the speaker to commence by telling the audience about some earlier, published research, which provides some of the background to the current project. In some cases, summarizing the researcher’s earlier published research takes up the overwhelming majority of the allotted time, and this is apparently considered perfectly normal.
  • The research which the speaker is presenting is frequently empirical, and there is a sense in which the experiment is already done, so there is no chance to go back and do it again.
  • Many of the questions raised are about whether various statistical tests have been performed to check for various alternative hypotheses, to rule out confounds and the like. Although the mood is generally friendly, the attitude of the speaker at this point often turns mildly contemptuous as they dismiss the objections with a brusque, “We checked that.” Very rarely, it is conceded that one more regression is worth performing, but one doesn’t get the feeling that it will make it into the published paper.

Having reflected on this, here is a too-strong generalization about the expectations of audiences at philosophy seminars versus economics seminars.

At a philosophy seminar, we expect to help the speaker more than we expect to learn from the paper. At an economics seminar, we expect to learn more from the speaker than we expect to help the speaker.

This can’t really be right, because at a typical philosophy seminar, no more than half the audience will ask questions. Although discussion takes up a large part of the session, and clearly is thought to be very important, it is not the case that everyone attending is expecting to be of any help at all to the speaker. But the generalization is getting at something broadly correct. The attitude to audience feedback clearly differs between the disciplines, in something like the way I have suggested.

Is there reason to prefer one norm or the other? There is certainly something appealing about the altruistic, collaborative aspect of the philosophy seminar norm. By attending a paper, we are hoping to assist the researcher by making a suggestion, by raising an important problem, and so forth. One could caricature the alternative approach as one in which the presentation is merely advertising for the researcher’s program of research: yuck!

But I suspect there is an insidious side effect associated with the philosophy norms: they encourage a degree of self-indulgence on the part of the speaker. At a stretch, one imagines the speaker implicitly thinking: “You’re all here to listen to me, and to help me get over my confusions, so I don’t have any obligation to make myself interesting or clear.” If the presentation is confused, has poor visual aids, is hard to understand, this is somehow forgivable, because we are all attending the presentation to help the poor confused presenter, rather than there to learn from him or her.

No doubt this is unfair: I have been to many philosophy presentations – and I hope this includes all of those which I have given – in which the speaker has conscientiously striven to be clear, informative, and to make the experience as valuable as possible for the audience. But there is something to this concern. Why is philosophy, for instance, the only discipline I am aware of in which it is still reasonably common for the speaker to read the paper verbatim? This is the most appalling habit, given how incomprehensible such papers usually are, and how much better off we would all be simply being given a printed copy and reading it for ourselves. Typically, reading a paper verbatim is likely to reveal minor problems of phrasing and the like, and therefore will be of benefit to the author, but is otherwise a waste of everyone’s time.

Personally, I prefer to go to a seminar to learn something, and I think this does not get sufficient priority in many philosophy talks. I therefore prefer the economics norms. Philosophy presentations could be shorter, could involve presenting already published research as well as work in progress, and could involve much less time for discussion. The discipline may well benefit from such a change.