“Dean Dad” recently wrote a post Confessions of a Community College Dean: When Public Speaking Works Best, in which he talked about the advantages of improvised talks over highly prepared ones:
My best moments as a speaker have consisted of a layer of improvisation on top of a prepared framework. The words were substantially ad-libbed, but in a context that had been thought through in advance. Having the safety net of a clear framework, the knowledge of where I was going, and the security of knowing that the worst that could happen wouldn’t do permanent damage, made it possible to follow the muse of the moment. I could improvise knowing the direction I wanted to go, and having faith that I’d get there one way or the other.
That is my usual modus operandi for giving talks or class lectures. I start by figuring out what I want to cover and (sometimes) in what order, and make sure I understand the material thoroughly. There have been a couple of times when I’ve had to give lectures on material I’m not completely comfortable with, and the results are not really satisfactory. I know that there are people who can give scripted lectures from prepared PowerPoint slides on stuff they don’t really understand, but I can’t—I have to have the stuff really solid in my head. (Which is not to say that I never pass on mis-information—I have sometimes realized after further study that I’ve been teaching a simplification that is incorrect.)
My best classes usually have no more than about 5 words of lecture notes, reminding me of the topic of the day—the entire performance is improvised off of those notes, together with lots of “audience participation”—getting the students to ask questions and come up with partial solutions. Such talks do not use prepared slides, but blackboard/whiteboard or live coding (for programs like gnuplot, where the concepts really rely on seeing what the program does with various scripts). I also get a lot of digressions in the best classes, when students ask about what really interests them, rather than what I have prepared. If the digressions are valuable and I know enough to go in that direction, I’ll take them. If I don’t know enough, I’ll usually put the students off until the next lecture, so that I have time to do some reading.
I have, once, given a talk with a carefully written-out script (see Video of Designing Courses talk), when I had a very short time slot to present a large amount of material. The results were OK, but not as good as the longer, slower improvisational presentation I use in classes.