I have been to a few conferences that worked based on precirculated papers. That is, everyone submits their papers in advance and the assumption is that conference goers who come to a given talk have read the paper in advance. In principle this seems ideal, we can use that face-to-face time to discuss work instead of just listening to people read their papers (this people reading their paper at a conference thing happens all the time in the humanities). In what circumstances should conferences precirculate papers? Is it something that is best for small conferences, or are there good examples of large conferences that work this way too?
I've seen some workshops in the CS field (more specifically in a subfield of Software Engineering, e.g. VaMoS - see workshop format -, in which there is a kind of "precirculation of papers", before the event take place. The workshop format includes a kind of "discussion session", in which, before attending the event, one paper is sent to an attendee other than the paper's (set of) author(s), so that he/she is in charge of reading the paper and preparing some discussion slides. During the event, after the paper's presentation, the discussant will provide attendance with his/her point of view on the paper, thus promoting the actual "discussion session".
Indeed, it is a small-scale precirculation of papers, in a sense that a paper is sent previously to only a few people (usually an author of another paper). However IMHO such a format provides event's participants with an interactive environment (in the worst case, at least you, as a paper author, will be sure that at least one another peer has read you paper... lol).
Quoting the mentioned workshop website:
Each session will be organized such that discussions among presenters of papers, discussants and other participants are stimulated. Typically, after a paper is presented, it is immediately discussed by pre-assigned discussants, after which a free discussion involving all participants follows. Each session is closed by a general discussion of all papers presented in the session.
To be very honest, I guess it's not feasible to do such a thing in big conferences, due to time constraints, but for small (and focused) events, like workshops, I guess this idea is very welcome.
I can't see this happening in any sort of large-scale conference. This would involve too much work and logistical planning on the part of everyone involved to be successful. It's hard enough to get abstracts for many conferences—let alone finished papers far enough in advance that people have time to read them!
In addition, this last point is another major obstacle: people don't have a lot of time to read all the papers that we're supposed to, let alone a bunch for a particular conference. The main reason why I would undertake that much work was if it were for a relatively specialized workshop in my personal field of endeavor.