« What have I been doing these past 8 years? | Main | Frigate bird »

May 03, 2010

How to give a good (professional) talk

Partly because of my year-in-review blog posts (e.g., 2009) and partly because I actively keep track of this kind of stuff (to convince myself at the end of each year that I have actually gotten some things done over the past 12 months), I know that I've now given about 65 professional talks of various kinds over the past 7 years. I cringe to think about what my first few talks must have been like, and I cringe only a little less to think about my current talks. I'm sure that after another 650, I'll still be trying to figure out how to give better talks.

That being said, I do think I can recognize good advice when I see it (even if I have a hard time following it), and John E. McCarthy's advice [1] is pretty darn good. Some of it is specific to mathematics talks, but most of the points are entirely general. Here are the main bullet points (on each of which McCarthy further elaborates):

1. Don't be intimidated by the audience.
2. Don't try to impress the audience with your brilliance.
3. The first 20 minutes should be completely understandable to graduate students.
4. Carry everyone along.
5. Talk about examples.
8. Pay attention to the audience.
9. Don’t introduce too many ideas.
11. Find out in advance how long the colloquium is, and prepare accordingly.
13. You do not have to talk about your own work.

I also found Scott Berkun's book "Confessions of a Public Speaker" to be both entertaining and useful. Berkun is a professional public speaker who works the technology circuit, but a lot of his advice holds just as well for scientific talks. Here's a selection of his advice (some paraphrased, some quoted):

1. Take a strong position in your title.
2. Think carefully about your specific audience.
3. Make your specific points as concise as possible.
4. Know the likely counter arguments from an intelligent, expert audience.
5. You are the entertainment. Do your job.
6. A good talk builds up to a few simple take-home messages.
7. Know what happens next (in your talk).
8. The easiest way to be interesting is to be honest.
9. Do not start with your slides. Start by thinking about and understanding your audience.
10. Practicing your talk will make it much much better.

A couple of years ago, SFI hired a professional "talk" coach. One of her main suggestions, and one shared by Berkun, is that we should each video ourselves giving a talk and then watch it several times to see what exactly we do that we're not aware of. This is a cringe-worthy experience [2] but I can tell you it's highly useful. I think we as speakers are often completely unaware of our nervous tics or obnoxious speaker habits. Watching yourself on video is perhaps the only way to get unbiased feedback about them.

Tip to Cris.


[1] A little Googling suggests that McCarthy's advice originally appeared in Canadian Mathematical Society NOTES in 1999 and was then reprinted by the American Mathematical Society. And, it seems to have a long history of being passed around since then.

[2] "OMG. Do I really sound like that?" and "Woah. Do I really do that when I talk?"

posted May 3, 2010 02:33 PM in Simply Academic | permalink