Log in

No account? Create an account
Giving talks - Arvind Narayanan's journal [entries|archive|friends|userinfo]

Giving talks [Apr. 21st, 2006|02:04 pm]
Arvind Narayanan
[Tags|, ]

One of the hardest lessons to learn in giving good talks is to always say "that's a very good question" when asked a question even if it was in fact not such a great question. This is true at all times but especially for questions in the middle of your talk, rather than at the end. It activates the reward system in your audience's brains and encourages them to pay more attention.

Other than obvious things like looking at your audience when you talk rather than the floor (sadly, a large fraction of academic speakers appear oblivious to even this basic piece of oratorical wisdom), it is important to incorporate humor into your talk. There is nothing else that even comes close to keeping audience attention. Have at least one joke prepared for every two slides. It doesn't have to be on the slide, just in your head. You don't even have to use all of them. But you must have them prepared because the moment you observe drooping heads, you should be able to say something funny. Laughter is a key social activity in primates; by stigmatizing the non-participants it disincentivizes sleeping during a talk.

Another thing I've found very useful is to have images on every slide. Language comprehension is in part conscious, having evolved relatively recently, but image processing is not. Since a person can perform only one task consciously, they will either read the text on the screen or listen to what you're saying (both conscious processes). But listening and looking at images can be done at the same time (only one conscious process.)

You must have figured out by now that I like evolutionary psychology. Some people consider it a pseudoscience, but I don't care: it helps me give better talks (and has excellent descriptive and predictive power in general).

[User Picture]From: descartes_rock
2006-04-21 02:49 pm (UTC)
As a journalist, I was trained that when people say "that's a very good questoin" or "I'm glad you asked that" it generally means the very opposite. It's a cue that you have hit a nerve with the person you are interviewing. Now, whenever I hear a speaker say that, I always think they don't want to answer.
(Reply) (Thread)
[User Picture]From: arvindn
2006-04-21 03:33 pm (UTC)
That's very interesting. I think I've observed what you say with politicians and other wishy-washy types. However, there isn't much scope for such an interpretation in academia, because a researcher usually wants the audience to understand as much about their research as possible, rather than the other way round. Further, when we don't have an answer to a question we admit it openly. Trying to fool your audience is a dangerous game that is played rarely, if ever, at least in the higher echelons.

Finally, saying "that's a good question" has concrete semantics in the academic context: it means that the person asking the question understood the content more deeply than is presented in the slides, and as such is completely different from "I'm glad you asked that question". Thus, while "that's a good question" might sometimes be a lie, it is always a complement.
(Reply) (Parent) (Thread)
[User Picture]From: arvindn
2006-04-21 03:38 pm (UTC)
I must qualify my comments above by saying that I was referring only to talks where the audience is of one's peers, i.e also academia. Whatever I said probably doesn't apply when talking to journalists. If I am ever interviewed about my research I will keep in mind not to say "that's a good question" :)
(Reply) (Parent) (Thread)
[User Picture]From: descartes_rock
2006-04-21 04:50 pm (UTC)
Isn't what you're describing an idealized view of the world of science? I mean, my experience with scientists tells me that science is heavily political. What about the researcher who rushes to get his paper published before a competitor, and maybe doesn't have all the angles figured out. Then he's presenting his paper and someone starts probing some of the weaknesses -- maybe some piece of data isn't as solid as it could be or whatever. In my experience, scientists are not entirely collegial.
(Reply) (Parent) (Thread)
[User Picture]From: arvindn
2006-04-21 05:06 pm (UTC)
Maybe you're right, maybe I'm just too green to know it. Maybe it's just that my field is less political than most, because it's heavily mathematical. I have definitely observed examples of dishonesty and hiding the weaknesses in one's research, but I'd always assumed it was limited to the less prominent fora. I might need to reevaluate my opinion.

I think I personally do fit the optimistic picture I painted earlier: when someone once pointed out a flaw in one of my papers right after it was accepted for publication, I admitted it during my conference talk. There are colleagues with whom I hold diametrically opposed views on some issue or the other, but we have completely congenial relationships. We either politely agree to disagree, or agree that one or the other of us is right and we haven't figured out which yet.
(Reply) (Parent) (Thread)
[User Picture]From: descartes_rock
2006-04-21 05:22 pm (UTC)
Well, I'd like to think most scientists are like you. Perhaps they are.
(Reply) (Parent) (Thread)
[User Picture]From: brokentooth
2006-04-21 06:23 pm (UTC)
I totally agree on the humour aspect. I realised how useful this was, after taking three classes with the prof I currently work with - he's actually honed these jokes over time, so he knows the best place to insert a witty comment, and he's always spot on. Definitely an underestimated tactic.

> because a researcher usually wants the audience to understand as much about their research as possible,

For some strange reason, this semester, there were 2-3 talks I went to, where I felt the exact opposite was true. Quite sucky.
(Reply) (Thread)