April 28, 2008 · Lessig
Many have asked me about my Keynote (it is not PowerPoint) presentation style. I honestly don’t have much to say about it, as I’ve not thought it through. But Chris Tunnell, a researcher on the SNO neutrino physics experiment has, and he sent me his thoughts about how and why (and whether) the style works based on his own experience using it for physics presentations. Read about it in the extended entry.
This is a reply to a request that I explain why I’ve shifted to gradient backgrounds (reprinted with permission):
Before I describe my experience with the technique, first let me briefly tell you about the audience. In doing so, hopefully you’ll understand some of the choices I’ve made, but please do bare in mind that I have a negatively biased view of physicists since we’re grumpy and impatient.
I was talking to a group of somewhere between 50-100 physicists who either knew me and my work, or at very least knew of me and my work, so somewhat knew what to expect. This was apart of an all-day meeting that happens annually and everybody’s work is closely related to everybody else’s work; another way of phrasing this is that we know enough to be bored of each talk since we’ve heard it all before, but we know too little to follow the talk given that we’re tired of listening to the hours of talks before. I generally go into these meetings with the mentality of, “Most of the audience is annoyed that I’ve dimmed the lights and am talking to them while they try to check their e-mail”. You see, with the dawn of laptops, regardless of the size of the physics crowd only four people will be listening to you at any given point: one person who did something similar a while ago, two people who are doing something similar now, and one person who had their laptop battery die. My goal of the talk was to not only convince them that I was worthy of listening to, but also teaching them what I did and convincing them that what I did was fairly important (yet neat).
I decided to try a method similar to your’s after I noticed my speaking style slowly leaning towards minimalism (which was actually a result of me needing glasses at first and wishing people used bigger fonts on slides!). This style is contrary to the normal physics slide-style which generally contains four bullet points, one plot, and exactly a minute’s worth of talking while the slide is presented. The reason I like not having much information on any given slide is I find that people don’t listen to you until they’ve fully read your slide, so having single words like you do encourages people to pay attention to the speaker while visually emphasising key points during the talk. Also the control of information on the screen helps the speaker prevent the audience from running ahead and getting lost.
The exact things I’ve stolen from you are as follows:
Minimal text: I try to use the least amounts of words, and letters, as possible in each slide (that isn’t a plot). I find people can’t remember or process more than a few things at any given time, so whether it be text or an equation, it’s dangerous to display too much information. I had to work hard to keep equations as simple as possible in order to make sure that people paid attention to the information of interest. This is generally a good idea in physics and mathematics, and I’ll use a facetious quote from a friend mathematician to demonstrate it: “It’s theoretically impossible to verify more than a line’s worth of algebra”. So the majority of my slides were a word or two, after your technique.
XML tags: I like using XML tags to give the talk a tree-like structure, so accordingly my talk was XML parsable. I find it helps the audience know where they are within the talk. Commonly, I find that I have no clue where the speaker is going with a talk, and specifically where one thought ends and another begins. It also helps the audience realize when they should try to rejoin the band-wagon if they’ve gotten lost. This may be a bigger issue with physics talks than copyright or corruption talks. This also lets the older physicists know when they should wake up if you happen to have a section on something they’re particularly interested in.
Reusing images: I’ve noticed that you like to reuse images to remind people of what you talked about before and to emotionally connect the current part of your talk to previous points. I did the same thing. I reused slides frequently — even if just to flash the slide before them — in order to remind them of what they’ve seen and to draw connections to previous points. I did this because nobody remembers anything ever, so relying on people remembering a previous point — for which they were probably looking at their watch rather than paying attention to you! — is a sure way to lose people and make them hate you. I found that reusing images was a nice way to help people draw connections between what they knew from my introduction to current topics.
Reusing text: This is somewhat similar to the previous point of reusing images, but applied to text. This point is best explained with an example. I was guiding my audience through the physics of my work, and at a point during my talk, I demonstrated a problem I had faced with two possible solution paths. I had a slide which outlined both: “A or B”. I then had a few slides about why method ‘B’ was better than ‘A’ in this case, after which I redisplayed my slide of “A or B” but with the ‘A’ struck-through to indicate that was the bad choice. People like seeing things they remember because it gives them the sense that I didn’t waste their time for the first part of the talk and instead taught them something useful.
Knowing the next slide: Most physicists rely on their slides reminding them what they were planning on talking about. What I’ve noticed in your slides is that you knew what you wanted to say before you said it, which is a foreign concept to our community. For instance, you frequently have slides with a single word corresponding to the word you happen to be saying through the microphone. The advantage of this is that it makes it feel like the slides are an extension of the speaker, rather than having them seem disjoint. The disadvantage of this is that people complain about the talk feeling “rehearsed” since the audience wants to feel special and like they’re getting the inside scoop. I think coinciding slides with words spoken is a great way to emphasize points, but should be used infrequently as emphasis. In my particular case, this allowed me to speak much faster than I normally do, which means at a normal-person pace; I have a slow and lulling natural voice.
Blank slides: I have never seen somebody use a blank slide in a talk before I saw a talk of your’s. Edward Tufte would role over in his grave if he saw that! I made one slide in my talk blank just because it shocked people so much that they really paid attention to the words that came out of my mouth. I find that blank slides are a great attention grabber because the audience is left frantically trying to recalibrate themselves since they have no information to go on. I would have never thought of this without seeing a talk by you, and it’s a useful teaching device.
I may be forgetting something, but those are the key points I remember from giving the talk. I gave the talk a few months ago (I think? It was sometime this year I believe…), so I may be forgetting some points.
Let’s quickly talk about the public reception. While I was giving the talk, over half the audience was listening. This may sound trivial to you, but I was the most widely listened to speaker at the entire meeting by a land-slide (I think nearly everybody was paying attention!). These are record numbers! There were two motivating factors to this: the train-wreck factor and the quick slide attention-deficit-disorder factor. The train-wreck factor is that people see something new and are waiting to see you fail, crash then burn because it will make for good dinner conversation. The quick slide factor is that the higher slide rate (slides per minute) of a minimalist style helps the attention-deficit-disorder physics pay attention. While I was giving the talk, it really seemed like people were much more interested in paying attention because minimalism is a better teaching device since it allows the speaker more control.
After my talk, there was much gossip about how my talk went. In general, I was amazed by how well the public reception was. For instance, a senior colleague of mine walked up to me right after the talk to tell me how the talk went. He told me that this was the best talk he’d seen in years! That was rather flattering. I received quite a bit of compliments from a wide range of people, so the public reception was phenomenal.
It really did seem like people were able to follow the talk and also learn from the talk, which in my mind makes it a great success. I do admit that a good half of everybody who listened to me was more interested in the talking style than the actual information I was trying to portray, but I certainly think that this technique will influence my future talks given how well it can teach information. But it really was amazing how much a new talking style shook up the community since they saw a new way of presenting information.
I thank you for putting your talks for free online because otherwise I would have never been able to give such a successful physics talk.
Summary: the technique I stole from you works surprisingly well on physicists due to the high rate of information flow and due to the control it gives the speaker over meandering physicist minds.
As to your question — no idea. I’ve only ever been guided by what feels
right. I don’t know exactly why black and white seems less useful today.
Well that’s somewhat unclimatic. It’s wonderful you have a natural instinct in presenting then.
The reason I asked is because you sometimes use background color to grab attention (ie. changing to a white background to grab attention).