I recently attended Building Bridges 4, an automorphic forms summer school and workshop. A major goal of the conference is to foster communication and relationships between researchers from North America and Europe, especially junior researchers and graduate students.
It was a great conference, and definitely one of the better conferences that I’ve attended. What made it so good? For one thing, it was in Budapest, and I love Budapest. Many of the main topics were close to my heart, which is a big plus.
But what I think really set it apart was that there were lots of relatively short talks, and almost everyone attended almost every talk.1
The amount of time allotted to a talk carries extreme power in deciding what sort of talk it will be. A typical hour-long seminar talk is long enough to give context, describe a line of research leading to a set of results, discuss how these results fit into the literature, and even to give a non-rushed description of how something is proved. Sometimes a good speaker will even distill a few major ideas and discuss how they are related. A long talk can have multiple major ideas (although just one presented very well can make a good talk too).
In comparison, 50, 40, and 30 minute talks require much more discipline. As the amount of time decreases, the number of ideas that can be inserted into a talk decreases. And this relationship is not linear! Thirty minutes is just about long enough to describe one idea pretty well, and to do anything more is very hard.2
Something interesting happens for shorter talks. For 20 minute, 15 minute, and 10 minute talks, the limitation almost serves as a source of inspiration.3 Being forced to focus on what’s important is a powerful organizing force.
The median talk length was 20 minutes, which is a very comfortable number. This is long enough to state a result and give context. It’s also long enough to tempt speakers into describing methodology behind a proof, but not long enough to effectively teach someone how the proof works.
An extraordinary aspect of a 20 minute talk is also that it’s short enough to pay attention to, even if it’s only an okay talk. It is perhaps not a surprise to most conference goers that most talks are not so great. To be a skilled orator is to be exceptional.
At Building Bridges, I was introduced to math speed talks. These are two minute talks. I’ve seen many programming lightning talks (often used to plug a particular product or solution to a common programming problem), but these math speed talks were different.
People used their two minutes to introduce an idea, or a result. And they either chose to give the broadest possible context, or a singular idea in the proof.
People were talking about real mathematics in two minutes. And I loved it.
Simply having a task where you distill some real mathematics into a two minute coherent description is worthwhile. What’s important? What do you really want to say? Why?
Two minutes is so short that it feels silly. And silly means that it doesn’t feel dangerous or scary, and thus many people felt willing to give it a try. At Building Bridges, the organizers gamified the speed talks, so that getting the closest to 2 minutes was rewarded with applause and going over two minutes resulted in a buzzer going off. It was a game, and it was fun. It was encouraging.
I firmly support any activity that encourages people who normally don’t speak so much, especially students and junior researchers. You learn a lot by giving a talk, even if it’s only a two minute talk.4
This conference had 19 (I think) speed talks over a three day stretch. They were given in clumps after the last regular talk each day. Since people were there for the big talk, everyone attended the speed talks. This is also important! In conferences like the Joint Math Meetings, where there might even be something like speed talks, it’s essentially impossible to pay attention since there are too many people in too many places and you never can step in the same river twice. Here, speed talks were given on the same stage as long talks, to the same audience, and with the same equipment.
Every conference should have speed talks. And they should be treated as first-class talks, with the exception that they are irrefutably silly.
Go forth and spread the speed talk gospel.