Why do conferences have such short page limits?

Question

These days, most of the conferences I attend provide their proceedings on a USB stick, or online only. Yet at the same time they have a strict page limit, which is usually set to something between 4 and 8 pages. I'm curious about the rationale for such a small limit. This is probably something that varies between fields, although I've experienced it in more than one field myself (subfields of artificial intelligence and physics). It doesn't seem particularly limited to big conferences.

I realise that super-long papers would hard on the reviewers, who have to do a lot of work for a large conference. But on the other hand, a very short paper will have to leave out a lot of background and important information, which at best will make it difficult to read, and at worst will make it impossible to judge on its merits. It's also a lot more effort to write a four-page paper than a 10-page one, unless you happen to be presenting a single experimental result using an established methodology (which is rarely if ever the case in artificial intelligence and related fields).

Looking at conference proceedings from the 1950s through to the early 1980s, it seems that proceedings papers were generally much longer. As an extreme example, I know of one that's over 100 pages (it's quite a famous paper and the length seems worth it), but up to 20 pages doesn't seem uncommon. In the '90s the standard seems to have been 10 pages, at least in subfields of computer science. Although there's obviously selection bias involved, these old papers don't seem particularly longer than they need to be, and their length allows them to make rather deeper points than could be made in a modern 4-page paper.

I'd like to know the reasoning behind short page limits at conferences. Is there a school of thought that says these extreme limits lead to better-quality papers, and/or less work for the reviewers? Or is it just a holdover from the recent past, when big conferences were expected to provide a paper proceedings to every participant, who would then have to carry the whole thing home?

Answer 1

There are a number of inter-related reasons for short (and shorter) conference paper lengths:

• The 10-12 page limit is typically so that reviewers don't have to review an inordinately long document.
• Final versions are also limited in length because the fact that proceedings are now mostly electronic doesn't mean that the cost of publication is independent of paper length. Editing costs are per-page (one can of course argue that professional editing does little for a conference paper, but nevertheless that's how it's priced).
• in order to make sure "what you review is what you get", one generally enforces the page limit at submission time as well
• finally, the 4-page limit I suspect is so that a fleshed out version of the work can be published in a journal without questions about the level of original content over and above what was presented at the conference. In this sense, the 4 (or 2 page) conference version is merely a short advertisement.