Written: Jan. 20, 2010
I have recently commented on the poor quality, and poor attendance, of the invited talks at a sectional meeting. I am sad to report that things are even worse at the annual meetings. What is so frustrating is the fact that with minimal effort on the part of the organizers, the quality of the talks, and consequently (hopefully) the attendance level, could improve dramatically. So please, my two good friends (and collaborators), Presidents George Andrews and Dave Bressoud, don't just say that you "are concerned about the quality of the invited talks" (as George Andrews told me in reply to my above-mentioned opinion) but DO SOMETHING! . And that something is trivial to do!
Let me first comment that the attendance at the major talks was dismal! The website of the meeting proudly announced that they were 5500 participants. A picture of the audience at the main lecture room (that has about 2200 seats), would give a good example of a random sparse 0-1 matrix (with ≤ 200 ones). I counted about 500 (an upper bound!) for the Gibbs lecture (less than ten percent!), and that is supposed to be the center-piece. Richard Stanley's first Colloquium talk (the other center-piece) had about 400 people, and the other invited talks (including the second and third colloquium talks) ranged from 100 to 200 people.
The (obvious!) reason why so many talks were so bad is FONT SIZE. Peter Shor's (after dinner!) Gibbs lecture would have been much better if he would have borrowed the font macros from his MIT colleague, Richard Stanley (whose three Colloquium talks were lucid, engaging, and stimulating, in addition to having a beautiful and large font). Peter Shor's small font was not the only problem. His talk was way too technical, and assumed too much background. He apologized at the beginning for not covering enough, and going too slow, so that people could understand. Maybe I am way-below-average, but in spite of Peter Shor's sincere effort at lowering the level, I barely understood a word. Peter, I wish that you would have iterated the mapping "cut in half" several more times, and focused on the basics.
Another disaster was Joe Harris' talk. He was introduced by David Eisenbud who said that Joe was a great teacher. I believe that he is, but his talk was a big flop, since he used a single-spaced, regular-size font, on an overhead projector, that was barely visible from the first row. Even though in general I don't like laptop talks, in such a big room it makes more sense to use a laptop, but if one is going to use an overhead projector, it is better to use hand-written ones, and lots of colors. If one is using printed text, then please, large font, and double- (or even triple-) space.
Manjul Bhargava did give a delightful and lucid talk, but he too used black printed text on an overhead projector, that was a little hard to read, since the font was small. The slides were much better than Joe Harris' slides, since they were double-spaced, and Manjul is such a natural teacher, that it ended up being a very good talk, but it would have been even better with larger font, and with color.
The font-size in Carolyn Gordon's Emmy Noether's talk was acceptable, but the talk was terrible (for a general audience), it was way too technical. I am sure that Gordon has the potential to give accessible general talks, and a little practice, and feedback from the organizers, could have made it much better.
In addition to Richard Stanley's three beautiful Colloquium talks, I also very much enjoyed Amie Wilkinson's talk, that was accessible and fun, and Brian White's, who used hand-written, and hand-drawn, over-head transparencies, with beautiful colorful pictures, explaining very well, to the mathematical masses, how to evolve curves and surfaces. These talks are existence proofs that invited hour-talks could be excellent.
I also liked Glen Van Brummelen's history talk about ancient trigonometry, in spite of a few annoying corny jokes ("buy my book!"). The slides, and the story, were first-rate. Richard Kenyon's talk was OK (the slides were good), but a bit too technical. Olga Holtz started out clearly and lucidly, but later gave too many details, but her talk was still acceptable.
Another surprising thing is that the AMS (and MAA) does not videotape the talks. It would have been nice to have at least the Colloquium talks on the web (the slides are available for the first, second, and third talk, but it would have been nice to have the movie). Perhaps if Peter Shor would have known that he is being filmed, he would have done a greater effort to make his talk accessible.
I was also disappointed that the second and third Colloquium talks conflicted with many Special Sessions (including, ironically, one on Enumerative Combinatorics!). The time-slot between 11:00am and noon, on Thurs. and Fri. had no special sessions, and Stanley's second and third talks could have been scheduled then.
I realize that most people do not go to conferences in order to listen to talks. The real reason is to wine, dine, meet old friends, and attend endless committee meetings, that are just excuses to socialize. But for those few weirdos, like myself, who do like to learn new mathematics, outside of their own narrow specialty, please, make at least the minimal effort of testing the slides beforehand, and have at least one dress-rehearsal.