Over on the Oikos blog, Jeremy Fox shared some interesting thoughts on whether or not a talk should be tweetable. He did not think they should be because he likes to give dense talks that are very rewarding to those that can follow them.
I’m guessing that would be far fewer people than he imagines. Our minds wander during talks, so an effective talk needs to constantly give us a helping hand so we can get back to following it after our thoughts stray. This is hard to do in a dense talk.
Information communicated in a talk is similar to that in a paper in many ways. Both start with a scientific question and some background information. Both tell you something about how the researcher did the work, the methods section. Both have some results that address the main question. Both summarize those results and reiterate how they address the main question. Both usually point out what remains to be done, or what problems arose.
But how talks and papers go about those tasks is very different. A talk needs to tell a story the way we tell stories, with a puzzle, a build-up, a struggle, a solution, a conclusion. It will not have all the details, though it will have a few figures. The figures should be simple and clear. I do not like the recent trend of putting four or five different graphs all on the same slide. One is plenty.
Do I think a great talk should be tweetable? Yes, I do. Maybe 5 tweets will capture the essence of it. Here are the tweets I could write from a great talk I heard yesterday by Lauryn Benedict at Northern Colorado University, on canyon wren song and territoriality at the Animal Behavior meetings.
Tweet 1, 132 characters. Within species, lower pitched songs indicate larger body size. A bird may lower its song to scare an intruder, said Lauryn Benedict.
Tweet 2, 138 characters. Canyon wren males are threatened by playing a recording of a different male. Do they respond by lowering their song, or adding low sounds?
Tweet 3, 140 characters. 13 recordings 1 hour before & after a 1 minute intruder’s song found resident males lowered their song & added a harsh, low note at the end.
Tweet 4, 136 characters. To compensate for the increased cost of low songs, the male canyon wrens decreased both the duration and syllable number of their songs.
Tweet 5, 137 characters. This canyon wren study lends support to Morton’s motivation structure hypothesis, published in American Naturalist 111: 855-869, in 1977.
Lauryn Benedict with a canyon wren.
Was it a challenge to write those tweets? Yes. I did them from my notes, taking care on the counts. Could I have done as good a job tweeting completely live? Maybe, or I could have taken 5 minutes right after the talk and done it.
The thing about learning is that active learning is the best. If you are thinking about how to tweet a talk you are engaging with the material, thinking about it, actively sorting the details from the core information. You will remember the talk better. You will notice the weaknesses, or the truly creative areas. For you, the talk will have been more worthwhile.
Will someone else, your peeps, get anything out of your tweets? Maybe. Notice I tried to make each tweet stand somewhat alone by mentioning the species or the author, though I could not say we were at the Animal Behavior meetings with each one, or give Lauryn’s name repeatedly.
The main point is that a talk should be tweetable and actually thinking about how you would tweet a talk can help your focus, your understanding, and your retention. A talk should also be exciting, tell a story without giving away the fun at the beginning with a boring old outline. And please, please, do not use a laser pointer. When you stir the slides with a bead of light, we watch the migraine-inducing dot and forget about the talk. We are like cats.
Sociobiology 
Thanks for the interesting comments Joan.
Re: the density of my talks, I probably wasn’t as clear about this in the post as I should’ve been, and probably made a bigger deal of it than I should have. I certainly agree that good talks tell a story and omit a lot of technical details. Further, I agree that an important part of telling a good scientific story is appropriate pacing and structure so that the audience can keep up. Certainly I don’t try to jam a bunch of figures onto one slide or anything silly like that! Frankly, I suspect if you graded the density or pace of my talks you’d probably find them only slightly more dense or slightly faster paced than is typical. Whether that’s a (slightly) good thing or a (slightly) bad thing is another question, I suppose (I still do think it’s good).
Re: the importance of active learning, I totally agree. I like to think I’m a very active listener–I’m always thinking about and evaluating what the speaker says as it’s being said. Personally, I don’t need to take notes or tweet to do that, but I certainly recognize that lots of people find that doing so helps them listen actively.
Re: making your talk “tweetable”, I think I see what you mean, but if I understand correctly a “tweetable” talk is just the same as a good talk (meaning, a talk we would’ve considered to be good before Twitter was invented).
Re: laser pointers, I know what you mean. I try to address this by being careful about how I use the pointer (I move it slowly rather than waving it around like a lightsaber, and I also try to structure my slides so I won’t need it that much). But really, an old-school pointer like a yardstick would probably be better. Or maybe even just walking over to the screen and using your arms, although the problem with that is that your body interferes with the projection. Would be interested to hear your thoughts on alternatives to using laser pointers.
Great thoughts. Laser dots can trigger migraines for me. I think a slide should be simple enough that we can use our words to point to things, saying things like “response” variable, or “Y axis,” or even upper left. At Wash U biology, they humor me and have a stick. At Rice we had a bamboo stick.
Another approach to thinking hard about a talk is to always have a question ready. During the talk you can always have a question in mind, rejecting one for the next as you go. Actually ask the question if it is burning important, or if others are not asking sufficient questions. No speaker should reach the end of their talk and face an audience with no questions #ABS2012.
Another way to keep a slide simple enough to be understood without a pointer is to have different elements appear one at a time, when you click the mouse. I do this a lot. It’s an effective way to force the audience to pay attention to only that bit of the slide you want them to pay attention to–because the other bits aren’t there yet. Same thing for slide with bullet points. Make them appear one at a time, so the audience listens to what you have to say about each point rather than trying to read ahead.
I love that you say peeps! Love, love, love it. Wish I was there at #ABS2012 to co-tweet with you!
I really miss you! Hope you can make it to Denver! In the meantime, keep up all your good work!
Thank you for this post! Nice to know that you are also tweeting from the meeting (and using the #ABS2012 hashtag now! :-)) – hopefully that will get more people into it also. Its my live connection to the meeting which I very much wanted to attend, but couldn’t. The live-tweeting from this meeting is still at a low volume with only a handful of you posting regularly. I haven’t even managed to get my students (who are presenting posters on birdsong with me as a co-author) to get on twitter yet! May you inspire them and others.
I’ll try to tweet more, with the hashtag. I did a few posters, and took papers from all that had them so I can tweet some of them. There were a lot of interesting posters.
Pingback: Against live-tweeting talks (UPDATEDx2) « Oikos Blog