I served on a steering committee for a programming unit of the American Academy of Religion for several years, and I have helped plan a seminar for the American Comparative Literature Association national conference this year. I have also proposed paper sessions, “wildcard” sessions, and individual proposals to AAR and SBL groups, with varying levels of success. Out of all this experience, I have derived the following general non-tips for aspiring conference participants.
The first thing to note is that, as with so many things in academia, the most important factors in determining your success are beyond your control. As long as your proposal passes the initial “laugh test” for relevance and competence, you are in the running — but after that point, it’s more a matter of putting together semi-coherent panels than rewarding the “best” proposals. If there’s a really great proposal that stands out, it might wind up contributing to the success of less great proposals that happen to be on the same topic. By the same token, a great proposal might not find its way onto the program because there’s nothing to group it with.
Hence, while one might have success with a proposal that doesn’t fit any of the specified topics in the call for papers, the odds are low — not because they’ll be pissed off you didn’t follow directions, but because the very fact that they asked for certain topics means more people will likely submit on those topics. By contrast, your totally left-field proposal, no matter how brilliant in itself, is unlikely to fit into any plausible grouping.
In general, I’ve had more success when submitting something as part of a “package deal” than when submitting individual paper proposals. This is not only because the package deal saves the programming unit some work — it’s more because it guarantees a level of coherence that a post-hoc grouping of individual proposals can never achieve. This is where a topic outside of the call for papers has more chance for success, because it can take up the “open session” slot that each AAR programming unit typically leaves unspecified.
All that being said, stressing out over the proposal itself and trying to jam pack as much into it as possible seems to me to be a waste of effort and emotional energy. Don’t just phone it in, of course, but don’t promise the moon, either — in my role as a steering committee member, I’ve sometimes expressed skepticism of good ideas that seemed too big to do responsibly in a conference paper. The maximum word count should not be taken as synonymous with the desired word count.
There are of course special factors at work with other major disciplinary conferences, but I would be surprised if these broad principles didn’t apply. Of course, that’s what commenters are for!