Opinion 60: Still Like That Old-Time Blackboard Talk

By Doron Zeilberger

Written: May 20, 2004.

Call me a relic, call me what you will..., but I still like that old-time blackboard talk, the kind of math that stimulates the mind (and soothes the soul!).

I have never been to an understandable power-point talk, and to very few (previously-prepared) transparency talks. Sometimes, in the hands of masterful speakers, these high-tech talks are really entertaining, and if the speaker is really good, he or she can convey the spirit. But even these talks (with the possible execption of Ron Graham and Xavier Viennot) would be even better if they were given with a blackboard. But for less-than-masterly speakers, the quality of a blackboard talk is orders-of-magnitude more understandable, and often also more entertaining than a power-point or even an over-head-projector talk.

The problem with transparencies and especially power-point is the baud-rate. The speakers try to squeeze several months of work (and sometimes a whole life-time of research) into an hour (and often just into twenty or even ten minutes). So in a typical conference or seminar, you just sit there and nod politely, but usually you don't have a clue.

Blackboard talks force the speakers to say the essential, and explain everything clearly. High-tech talk is just a core-dump.

I was inspired to write this editorial by my good friend and colleague, Janos Komlos, himself a brilliant blackboard speaker, whom I invited to attend a talk by my good friend Aviezri Fraenkel. Janos, for some reason, thought that the talk was one day earlier, and went to the lecture room. That day, there was an all-campus power-outage, and Janos wrote me E-mail that he said to himself: `great, now the speaker will be forced to use the board'. Even though his wish did not materialize that date, Janos promised to still come at the scheduled time, taking a chance that the talk will be power-point. The next day, before the talk, Aviezri, I, and two of my computer-whiz students, went before lunch to set-up the power-point interface, and after fiddling for ten minutes, it was all set-up. When we came back after lunch, there was good news (at least to me and Janos), the power-point didn't work, in spite of the competent hands that tried to restore it. So Janos's (and my!) wish came true, and Avizeri was forced to use the (white)board, and guess what? His talk was really excellent, much better than it would have been!

But an even better proof that good-old-time blackboard talks are the best was last week, at the Stat. Mech. meeting organized by Joel Lebowitz. Amongst quite a few mediocre-to-boring over-head and power-point talks, there was one talk that struck out, and that was definitely the best, by Bill Bialek, who talked about `how many bits does the brain use', and discussed how the brain manages to extract the few significant bits from the bulky chaff. Analogously, because of Bill's wonderful presentation, using the blackboard only, as well as intonation and hand motions, people were able to extract the really significant part of his talk, and unlike most of the other talks, that were forgotten as soon as they were over, Bill Bialek's talk will stay in our memories for a very long time.

So Just SAY NO TO POWER-POINT!, and even to over-heads! But, even when using the blackboard, remember to speak SLOWWWWLY and Clearly.


Added May 1, 2009: Yet another proof that blackboard talks, delivered by the right person, far surpass laptop and overhead projector talks, was given today, when Melanie Matchett Wood delivered the very best talk at the Rutgers Mathematics Colloquium. She started at the very beginning, saying that algebraic integers are analogs of usual integers, and GL1 is just a fancy way of saying {-1,1}, then cited a classical theorem of Gauss, from about two hundred years ago, that described the n=1 case, then another one, from about fifty years ago, of Fadeev and others, that described the n=2 case, followed by a far-reaching generalization for all n, that she discovered (presumbly, there was no attribution, even not a "W."). The talk was a true gem on many levels, from the localest local to the globalest global.


Added July 9, 2013: Yet another data point! I just came back from the 24th British Combinatorics Conference. There were quite a few good talks, in spite of the fact that they used laptops, but most of the talks were impossible to follow, since they raced through the slide. However, the best talk by far was the only talk (in addition to my own's) that only used a whiteboard! It was given by Felix Lazebnick. Felix very slowly motivated the problem, described the statement and concepts very lucidly and slowly, and only at the very end indicated the main ideas behind the proof.
Doron Zeilberger's Opinion's Table of Content

Doron Zeilberger's Homepage