The Quirks of Scientific Public Speaking

As previously noted, I spent most of last week at the 2013 DAMOP meeting, where I listened to a whole bunch of talks. At some point, I was listening to a talk, and said "I bet this guy hasn't given a lot of these before." What was the give-away? The fact that he almost never said "Um."

To the dismay of many students entering science majors, public speaking is a very significant part of being a professional scientist. Scientists are expected to give talks of a variety of different lengths-- 10-15 minute "contributed" talks at big meetings, 25-30 minute "invited" conference talks, 45-60 minute seminars and "job talks." Oral presentations are one of the most important ways in which we communicate scientific results to other scientists.

And yet, the way scientists do public speaking is... very odd. We place a great deal of importance on public speaking, and most professional scientific talks are meticulously prepared. And then they're delivered in a way that attempts to make them seem like off-the-cuff presentations.

The factor that made me think the guy giving that DAMOP talk was new at this was that he had very clearly memorized a prepared text. It wasn't quite a rote recitation-- he managed to vary the inflection of his voice in the ways appropriate for normal human speech-- but he clearly knew exactly what word came next in each sentence. And while that's a fairly effective way to deal with giving a talk when you're not comfortable doing it, it's extremely unusual in scientific presentations, particularly in physics. I've been going to scientific talks for better than twenty years, now, starting with colloquia at Williams when I was an undergrad, and I only hear clearly memorized text a couple of times a year.

What's much more standard is a talk where the speaker has put a great deal of thought into their slides, and the key ideas that need to be expressed on each, but where they try to make up the exact wording of the talk as they go. Which means that, when you hear talks from scientists with a lot of practice giving talks, you tend to hear "Um" a lot. And also verbal filler like frequent sentences starting with "So, we see that..."

It's kind of a weird business in that respect. To some extent, of course, this is just what good public speakers do-- during the last Presidential election, there was a fascinating comparison of Bill Clinton's DNC speech in 2012 with the prepared text he had written, and it's chock full of brilliant ad-libs, and also the occasional bit of folksy verbal filler ("Now, folks..."). But even Clinton was working off a prepared text projected in front of him on a Teleprompter. Which is something basically no professional scientist does.

Instead, we have a very strange relationship with our PowerPoint slides. The slide in the "featured image" at the top of this post is the title slide from one of my research talks, listing my collaborator on the original proposal, and the students and funding agencies who have contributed. A typical slide from the middle of the text looks like this:

A slide from my research talk A slide from my research talk

That's a fairly typical sort of slide for a scientific talk--if anything, it's probably a little lighter on text than most of the slides I saw at DAMOP. The slide has a schematic diagram-- in this case, a cartoon version of a proposed astrophysical detector using liquid neon as a scintillation medium. There's also a list of important points regarding the proposal over on the left side; in the actual presentation, these are minimally animated, so I reveal each as I get to it-- you only see the full thing at the end of the spiel, just before I change to the next slide.

The list-of-points thing is a little risky-- people can generally read faster than you can talk, which is part of the reason for the minimal animation. Having the text present serves two purposes, though: one is as an aid to anyone who's taking notes-- which really does happen at conferences-- but the most important function is as a reminder to the speaker. It's considered bad form to explicitly speak from notes, as is common in the humanities, and so scientists use prepared slides with text to cue them as to the next key point. Which leads to a lot of "Um" and "Uh" as the speaker does the mental reconstruction of the text.

(There are occasional hints that this is a recent-ish development-- if you look at some of the many videos of Feynman lecturing-- this one, say-- he's frequently referring to notes, not using slides or an overhead projector. By the time I entered physics in the early 1990's, though, everything was overheads, and a chalk-talk from notes was very rare. these days, it's all PowerPoint or the equivalent.)

This is, as I said, a very strange situation when you stop to think about it. Professional scientists spend a great deal of time speaking in public, but we do it in a manner that is designed to maintain the pretense of spontaneity. And because of that, even experienced speakers frequently appear sort of amateurish compared to speakers in politics and business. Students are often specifically coached not to memorize a specific text, just the key points, and to invent the exact words as they go. And people who are clearly working from a carefully crafted script stand out as unusual in the context of a week-long meeting.

Why it is that we do this is one of the enduring social mysteries of science. It's very deeply ingrained, though-- even when I do public lectures, where I very deliberately eliminate almost all the text from the slides, I still don't go from memorized text, but use the images on the slides to cue myself as to what to say next. The exact words are invented on the spot. Which allows for fewer really clever turns of phrase, and creates a lot of space for verbal filler.

(By the way, if you want to see the full presentation those slides above came from, to verify that the examples are fairly typical, here it is on SlideShare:

(That might not be the exact version I grabbed the screenshots from, but it's very close.)

More like this

To be fair, Feynman is giving an hour+ long blackboard talk. That's more like teaching a class than giving a modern-day talk, where the allotted time is hardly enough to even explain what the question the speaker is addressing is.

By Evan Berkowitz (not verified) on 11 Jun 2013 #permalink

Public speaking always has to be a vividly told story to work. I've certainly sat through a lot of these deadening presentations. Maybe we could have more TED style, with a bit more backup?

Feynman's doing a long seminar, true, but these day those are also generally given in the same manner. The slides I embedded are for a 50-minute seminar talk, and my public lecture runs about the same.

Also, regarding the uninspiring nature of many talks, there's this story I was pointed to yesterday, about a study showing that Smooth Lectures Foster Only the Illusion of Learning. So maybe those talks are boring for your own good...

(I.I. Rabi, according to Norman Ramsey and Martin Perl, was a famously terrible lecturer. Rabi apparently claimed it was deliberate, because watching him figure out his mistakes was good for the students...)

Heh. My first thought on seeing that liquid neon slide was "What's with all those words?!" Now that I'm a corporate scientist, I find that I have to create two different types of slides. For talks to scientists, I tend to have minimal words, on the assumption that I'd rather the audience pay attention to what I'm saying, not what I've written. For managers, though, who won't read the ten page report I've written, but will look at the one-slide summary, I have to have a slide with a too-small font and too-small pictures that has enough content to be an archival summary.

By Grant Goodyear (not verified) on 11 Jun 2013 #permalink

"Fiducial Volume" would make a great name for a rock band.

Hmmmm. I find umms usually indicate some lack of experience, perhaps not in general but in that specific talk. Fixed speechifying is rare but indicates either inexperience with extreme preparation or so much experience it only amounts to repetition. One thing often said is to pay attention to the audience which I find next to impossible unless fairly well prepared but this is more a fluid speech pattern rather than a rigid one which requires extreme attention to the speech itself. It isn't that difficult to eliminate umms but remain relaxed though probably much more difficult to eliminate filler which assists in establishment of cadence and flow as does vocabulary choice. If comfortable with the material, extemporizing is much much easier than fixed speech but this requires less attention to the speech itself, which self consciousness of umms can interfere with though many remain unconscious of them .

Bad slides and presentations are a cultural artifact. The style is expected and in some cases enforced by the generation that chairs and organizes events.

Even if a young scientist wanted to present in a different style there'd be great social pressure not to stand out and be different from the others, a risk, in academic circles, few people are willing to take.

See Why Do People Make Bad Slides? http://t.co/kWQUzqo2pp

By Scott Berkun (not verified) on 11 Jun 2013 #permalink

Depends greatly on the field. I once was at a social science conference and half of the speakers were reading their "talk" off their computer or paper.

One key thing about most talks is that interruptions are expected. Memorizing the whole talk would be counterproductive if the speaker has to keep stopping to answer questions.

Another thought occurred to me, and it is the key one, I think.

Anyone can memorize a talk without understanding the material. Giving a talk without memorization, however, requires knowing the material. The bizarre tendency to give presentations off of the cuff is a way to communicating the speakers understanding.

The other factor that hasn't been mentioned yet is that, at least for conference talks, there's frequently not all that much time to prepare them. My advisor used to write his talks on the flight to the conference, and many people do the final tweaking of their slides the night before the talk. It's easier to do improvise-from-cues in that situation than to work out and memorize the full text. Not only are you signaling mastery of the material, you're also signaling that it's up-to-the-minute stuff.

many people do the final tweaking of their slides the night before the talk

This is definitely a big part of it. Back in the days of 35 mm slides, you had to have your slides fully prepared before you got on the plane. With overhead transparencies, and to some extent with PowerPoint (although the major conferences I go to these days try to discourage this), you can continue tweaking your talk up until the start of the session, if not the moment the session chair calls your name. And there is no social penalty for acquiring a reputation for doing this. Scientific research encourages workaholism, so there is some pressure to give the appearance that you have been working on your talk at least in the 24 hours prior to giving it.

It's also a way of signaling that you are fully fluent in English. In my experience, people who speak English as a second (or third, fourth, etc.) language but are not fluent are more likely to write out and rehearse their talks than people who are fully fluent in English. (Only about half of the people at many conferences I go to, even in the US, are native English speakers.) To be fully fluent in a language, you have to be able to think in that language, and not everyone is capable of acquiring that skill in a language learned after childhood.

By Eric Lund (not verified) on 12 Jun 2013 #permalink