by Bill Wadge
Academics love to talk, talk, talk … and to give “talks”. I was no exception.
Sometimes they went well, sometimes not so well … and sometimes they went weird. Here are some outstanding ones in various categories.
First real talk. Of course I made presentations as a grad student but my first real talk was what a friend calls an “I-want-a-job-talk” (IWAJT) for my first job. This was at the fabled “University Far Far Away” (UFFA).
My main problem was that although I was applying to a Computer Science department, I didn’t know much about computer science. All my education was in math, but the math job market had collapsed.
So what could I talk about? Luckily, my PhD involved infinite games, and winning strategies are basically algorithms. I’d talk about them! I found quite a few, some involving players putting marbles in buckets or filling in grids with X’s and O’s.
But not just talk, that would have been boring. I hit on the idea of having the audience participate as one player in some sample games. They joined in enthusiastically with the grids and the buckets, and thought I was great (for the usual six months).
Worst title. Soon after that I gave a research talk at a conference, about a result from my PhD. The result was OK, but not the title: The number of quantifier alterations is absolute. I got teased about it for weeks.
IWAJT near disaster. As I’ve explained in a previous post, my career at the UFFA (University Far Far Away) was severely limited, i.e. I got fired. Just in time my friend David was hosting his friend Mike visiting from the University of Warwick in the UK. They were looking for a postdoc and David recommended me. We arranged for me to meet Mike and tell him about my research.
Which I did – declarative languages and dataflow, how imperative languages are obsolete etc. The beginnings of the Lucid work. Mike listened patiently, then went back to David … and told him he thought I was crazy.
Mike was into algorithms and like a lot of algorithms people thought programming was just coding; basically trivial. In fact a lot of my problems in academia were caused by the fact that I had zero interest in algorithmic complexity.
Anyway I was clueless as usual and was about to head out the door when the phone rang. It was David. Could I meet again with Mike and tell him about my other research? Ahhh … OK send him down. So I hauled out the marbles and the grids and the buckets and this went over much better. The second IWAJT (I-Want-A-Job-Talk) was a success and I started at Warwick the next year.
Dualin’ speakers. I loved the UK and one of the reasons was distances were so small you can go anywhere in at most a few hours by train. I got to attend lots of workshops with interesting people. At one I was up at the blackboard, extolling dataflow as usual, when one of the audience members objected.
The dissident got up, went to the left side of the blackboard, and began using it to justify his claim. I realized it wasn’t going to be easy to put him in his place …
… It was Edsgar Dijkstra himself.
So what I did was move to the right hand side of the blackboard and continue my talk. In parallel. Allan Kay said that in computer science arrogance is measured in nano-Dijkstras. I got a lot of sympathy.
Best visual. Another time I was giving yet another talk about dataflow and I had old fashioned transparency – one showed a graph. I remarked that almost any topic in computing can be illustrated with a graph. Nods. Inspired, I said you can make it even more abstract … and lifted the slide so that the image became blurry.
Try that with power point!
Worst audience feedback. I was in California and was persuaded to give a talk based on the categories in my infamous ‘Cowboys’ section of the Lucid book. I did some nice cartoons, e.g. proponents of structured programming as Preachers.
At one point a well known researcher, let’s call him “Joe”, raises his hand. “Is there any technical content in this talk?”
I wish I could report a snappy comeback. Maybe I should have said “it’s more profound than that” but I didn’t. I mumbled something that didn’t satisfy Joe.
He ostentatiously stood up and walked out.
At least he didn’t go to the black board!
Best snappy comeback. I eventually left the UK for Canada, ending up at the University of Victoria. My IWAJT went well, but at one point a well-known Dave asked about some of my terminology. “Is this something I should know, or is it your own terminology?”
I answered, “both!”.
We all got a kick out of this and in fact Dave and I always got along after that.
Worst technical fiasco. I went to Amsterdam to give a talk in a logic conference. I don’t like powerpoint and prefer using a blackboard. Unfortunately they didn’t have one and set up a small easel with a sort of pad of sheets. That was already a disaster but even worse they couldn’t turn off the powerpoint projector, which was shining in my eyes and blinding me. Eventually someone crawled up into a sort of balcony and managed to shut it down.
Pity, it would have been a good talk. Maybe not marbles and buckets, but good.
Best technical fiasco. I was at Schloss Dagstuhl, which I like because they have plenty of blackboards – that are even black.
However some speakers are determined to use modern technology (powerpoint) with the usual consequences – technical problems with the temperamental projectors. More than once we saw the screen display the dreaded message “KEIN SIGNAL” (no signal).
When it was my turn I apologized for using old technology but assured them that I could reproduce any powerpoint effect. To illustrate, I went to a blank blackboard, picked up a piece of chalk, fiddled with it, then wrote KEIN SIGNAL on the board.
Then I rebooted. I put the chalk back on the chalk tray, picked it up again, and proceeded to draw buckets and grids.
Best overall. In my last year at the UFFA some colleagues and I were griping about how hard it was to explain matrix eigenvalues. We need an accessible example … and came up with the idea of a mini-ecosystem whose step-by-step evolution is governed by a matrix. Then behaviour of the ecosystem can be explained in terms of the eigenvalues of the matrix.
We came up with a system with grass, which grows at a constant rate; rabbits, that eat the grass; hawks, that eat the rabbits; insects, that eat grass but give a fatal disease to the hawks; and little birds, that eat the insects. can’t remember and have long since lost all my notes and slides :(.
Anyway we fiddled with the matrix until we got eigenvalues that would make for drama. We called our ecosystem the Eigenjungle and decided to make a multimedia presentation.
We even had a soundtrack: some well known processional music as we entered. Also Sprach Zarathustra (2001 theme) as the sun rises over the Eigenjungle; Flight of the Bumblebees when the insects are introduced; I‘m a Lover not Fighter for the bunnies. The eigenvalues of norm less than one were populations that slowly die out – cue the funeral music. One of the eigenvalues was 1 and that gave a stable population. But one had norm slightly greater than 1, and that was ominous.
To illustrate the ominosity we showed what happened when a single bunny comes and joins the stable eigenjungle population. I remember the music played was happy and innocent, from the pastoral symphony. But the happiness is short lived because the single bunny has a component in the norm>1 eigenvector.
The extra bunny causes the grass to get over grazed and the bunny population to shrink. This cuts back on the hawks and as the grass makes a comeback the bunnies prosper. But then the hawks increase, the bunnies are eaten etc etc. The populations starts to oscillate more and more violently and I recall one slide showing the sky dark with hawks. The sound track was Flight of the Valkyries – truly an apocalyptic vision.
It was popular with students and even many colleagues and was performed a couple of more times. But as you can imagine this contributed exactly nothing to our careers at the UFFA: it wasn’t a grant, or a refereed conference publication, or an archive-quality journal paper. It was just a talk – one I’m still proud of.
I didn’t like the UFFA’s eigenvalues and was very happy to emigrate to the UK.