Want to take part in these discussions? Sign in if you have an account, or apply for one below
Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.
This is a split from another thread. The first few comments will simply repost material from there. (If you’re reading this in real time, give me a minute!)
Urs wrote:
Hi everyone,
tomorrow (in a dozen hours from now) I am giving a talk at Quantum Physics and Logic 2011 .
This is required to be a beamer talk. So I have prepared some slides, titled
This should be easy going, with a little bit of fun at the beginning.
I’d be interested in whatever comment anyone here might have!
TobyBartels wrote:
This is required to be a beamer talk.
So it’s required to be bad? Where’s Doron Zeilberger when you need him?
:-P
Urs wrote:
Good morning.
Thanks for all the comments!
I have fixed that floating comma and the “may be/need be”-thing, thanks! I still need to think about what to do with the hookleftarrows (maybe over breakfast, time is getting short now). Thanks for highligting that this causes trouble!
Concerning Zeilberger: yes, I have the same feeling. I had intended this originally as a blackborard talk. But two days ago we received an email which said something like that they expect everybody to use the beamer and warn that the available whiteboard is tiny.
Personally, I find already the step from blackboard to whiteboard a step in the wrong direction. But clearly a majority thinks differently.
Andrew Stacey wrote:
That has to be one of the daftest rants I’ve read from DZ.
I’ve lost count of the number of rambling, inane, crammed, ill-prepared, disorganised chalkboard talks that I’ve been to. I’m no logician, but the argument “So-and-so gave a great chalkboard talk, therefore chalkboard talks are the best” seems … dubious.
Giving a beamer presentation (note: beamer, not powerpoint - I despise powerpoint) means:
I have to actually prepare the entire talk in advance.
How many talks have you been to where it’s clear that the speaker just thought they’d talk about some particular subject and stop roughly when the hour was up? (“roughly” being the operative word!)
I have to think about what the audience can see at any one particular time.
Because you can’t rely on saving bits of the chalkboard, you have to think carefully, “If I need to use this again, I’d better say it again.”. Relying on the audience remembering something that you said 5 minutes previously is dangerous ground, and just waving vaguely at a bit of the chalkboard where you wrote it 10 minutes ago isn’t much better because although the statement is still there, the notation that it relies on is probably gone.
I’m much less likely to stray off-topic.
This is one of my top hates about talks. It just gets interesting (or the opposite: you start to feel that it’s not too long until it’s all over) when Big Name in the front row asks something almost completely irrelevant. Speaker then tries to answer that question (which almost no-one else is actually interested in) and gets completely side-tracked. With a presentation, it’s much harder to get derailed.
I’m much more likely to say something interesting.
Let’s face it, most talks are memorable for only about 5 minutes. There are great talks that really tell you something, but the vast majority are some convoluted complicated explanation of some technical point in an article. They remind me of the graduate talks when I was doing my PhD: they were basically an excuse for the student to tell their supervisor what they’d been doing, which was great for the supervisor but rubbish for everyone else (so we staged a revolution, but that’s another story). Unless it’s one of those Great Talks (and if you’re not sure, then it isn’t), the talk should essentially persuade me to read your paper. That’s it. Leave the complicated stuff for the paper - if I’m sufficiently interested, then I’ll learn it there; if I’m not sufficiently interested then don’t bother telling me about it in the talk!
I’d also much rather listen to a beamer presentation than a chalkboard talk, for very similar reasons.
I gave my first beamer presentation under similar circumstances: a few days before the conference we were told that there was no chalkboard and only a small whiteboard (turned out this was wrong). I made a lot of mistakes, but I still think that I gave a much better talk than I would have done if it had been a chalkboard talk because I was forced to think about the above things.
Yes, there are awful beamer presentations. But I think that the people who give awful beamer presentations are equally likely to give awful chalkboard talks. And also as it’s a fairly new thing (for mathematicians, at any rate) a lot of the “good techniques” haven’t yet gotten into our blood. So we don’t know how to do it properly. All mathematicians should be forced to read the entire beamer manual, then there’d be a dramatic improvement in the standard of talks (to be clear, I don’t necessarily agree with all of the advice that Till Tantau gives, but if someone gives you advice and you decide to do otherwise, at least you’ve been forced to think about it).
Lastly, the statement:
Personally, I find already the step from blackboard to whiteboard a step in the wrong direction.
is actually a damning indictment of the standard of talks that we give to each other. If the difference between a chalkboard and a whiteboard actually makes a difference to the quality of the talk, then it was an awful talk to begin with. Someone who hasn’t learnt how to write clearly and legibly on a whiteboard should not be giving talks using any medium.
I’m not going to apologies for the rant - since someone mentioned DZ’s rant, it seemed appropriate.
As a final word, if this weren’t a beamer presentation, how would the rest of us get to know what you intend to say?
Mike Shulman wrote:
@Andrew #9: Hear, hear! I agree entirely with almost everything you said, with two caveats.
First, I don’t think that beamer is always better than chalk. I think beamer is hands-down better for any talk half an hour or less, because in a short talk all you can hope to convey is ideas and beamer is much better for that. But if you have an hour, then you can hope to go through a medium-length argument and get the audience to follow it, and I think that’s pretty hard to do with beamer because of the small amount of information visible at any one time. Also, chalk is more flexible: you can easily respond to questions by writing more things you didn’t plan on having to say, or rearrange the talk in the middle because of what the audience wants. Obviously that flexibility is rope to hang yourself with, but if you know what you’re doing it can be a big advantage. Also, there is difference in context between conference talks and seminar talks; I think chalk fits better with the seminar atmosphere and beamer better for the conference. Departmental colloquia are somewhere in between.
Secondly, I can’t stand whiteboards. (-: Maybe the quality of the talk shouldn’t be affected by black vs white, but neither should the quality of the talk be affected by, say, loud construction noise outside, or a broken heating system that makes everyone’s teeth chatter – and nevertheless, those would negatively affect my overall experience.
But I don’t understand the statement “I find already the step from blackboard to whiteboard a step in the wrong direction” – there’s no sense I can think of in which that step is in anything like the same direction as the step from black/white-boards to beamer, so what does the rightness or wrongness of one have to do with the other?
Andrew Stacey wrote:
Mike, I accept your caveats. Your rudimentary classification of talks is good - I’d quite like to see a more refined classification! I think I’d put seminar talks still in the better-with-beamer category and put seminar series on the other side of the line. That’s partly because the majority of hour-long seminar talks that I’ve been to would have been far better as half-hour talks. I don’t think that an hour is a good length of time for a talk: it’s not enough or it’s too much. (A colleague once said that in an hour-long talk, 20 minutes should be for everyone, 20 minutes for those in your field, and 20 minutes for the experts. At the time, I thought that reasonable. Now, I think it’s appalling.)
Chalk is more flexible.
Tried a graphics tablet? Or an iPad? Seriously, it makes a huge difference.
but if you know what you’re doing
Aye, but thereby hangs the rub. Too few do know what they’re doing.
Secondly, I can’t stand whiteboards.
I actually have a similar reaction, though I think it’s more due to the fact that I’ve seen more awful whiteboard talks than chalkboard talks.
What I’d really like to see in this debate is something beyond “I like” or “I don’t like” or “in an ideal world”. What, actually, is it about a chalkboard talk that means that many people think it better than a beamer talk? The number of good or bad talks is neither here nor there for that. I still, even though I now use beamer for everything, feel that there’s something missing from my beamer presentations that would be present in a chalk talk. But unless I can actually identify what it is, I’m not going to swap back. And if I can identify it then maybe I can figure out a way to include it in the beamer presentations.
So I’d love it if those that prefer chalk to projector could explain exactly what it is that they prefer.
(I had a “technology failure” in a lecture this semester and had to give the lecture using chalk. The response was, “Stick to the projector.”)
Urs wrote:
Hi Andrew,
What I’d really like to see in this debate
How about opening a new thread for your debate?
Todd_Trimble wrote:
I haven’t given a talk in front of mathematicians for 10 years now, but still…
The obvious danger of using slides or beamer or power point is that one tries to say too much. With chalk, you are presenting a flow of ideas and reasoned discourse in “real time”. I’m not sure how eloquently I can put this, so maybe I can just refer to what Rota says about chalk talks in Indiscrete Thoughts?
I sense there is greater chance for spontaneity in chalk talks, that they lend themselves less to being prefabricated. One can make a quick change of plan in response to how it is playing to the audience. There is something a little bit thrilling about the performance of a good chalk talk which can’t be captured in the same way with pre-prepared notes.
Now beginning new material:
I think that the problem is that it’s much easier to give a bad talk with a beamer (that is an overhead projector displaying previously written material, whether this material was prepared by beamer.cls, PowerPoint, or hand-drawn slides) than with a chalkboard. This goes back to Andrew’s point #4, with which Zeilberger would emphatically agree:
Let’s face it, most talks are memorable for only about 5 minutes. There are great talks that really tell you something, but the vast majority are some convoluted complicated explanation of some technical point in an article. They remind me of the graduate talks when I was doing my PhD: they were basically an excuse for the student to tell their supervisor what they’d been doing, which was great for the supervisor but rubbish for everyone else (so we staged a revolution, but that’s another story). Unless it’s one of those Great Talks (and if you’re not sure, then it isn’t), the talk should essentially persuade me to read your paper. That’s it. Leave the complicated stuff for the paper - if I’m sufficiently interested, then I’ll learn it there; if I’m not sufficiently interested then don’t bother telling me about it in the talk!
If you’re doing this wrong, then the beamer will let you do it very wrong, putting all of the complicated stuff (probably in small font size too). If you work at it, you can get your entire paper onto the slides! With a board, you have to slow down; you can’t put anything on display that you don’t write during the talk. So while both boards and beamers can be misused, beamers can be misused more. And Zeilberger is so cranky because he goes to talks where people are misusing them!
Of course, if you use a beamer well, then you won’t have this problem. (Actually using beamer.cls is a good step right off, since it has a reasonable default font size.) But the example of the bad talk (which is both Andrew’s point #4 and Zeilberger’s entire argument) goes to the board, I think. (I agree that Andrew’s arguments #1–3 go to the beamer.)
Oops, didn’t see this new thread. Although theoretically, that is not quite the same topic.
If we can keep the momentum, I think it could be of value to have two threads. In the other, I’m asking mainly about self-evaluation of a talk. Let’s make this one about objective advantages and disadvantages of beamer and chalkboard talks. I’d like the goal to be to make a list of each so that when someone is going to give a talk, they can look at this list and know what to avoid. I think that Toby’s post in #11 is a good start in this, and I’d like it to get more refined. It may be that some of the disadvantages (of either) can be countered, but of course to know that they need to be countered then one needs to know about them.
If this works, we can make a page on the nLab (or nLab meta) containing the list, so we don’t need to be too neat in the discussion.
The other thing to say is that this will, by its nature, be subjective. But by making small statements (like Toby’s warning about putting too much on a slide), we should be able to keep the emotion level down!
I think that the problem is that it’s much easier to give a bad talk with a beamer … than with a chalkboard.
I have heard so many bad chalkboard talks that I am not convinced by this. It’s true that a beamer makes it easy to try to say too much. But a beamer also makes it harder to be disorganized and rambling, and harder to spoil the talk with bad handwriting and bad board usage.
Tried a graphics tablet? Or an iPad? Seriously, it makes a huge difference.
No, I haven’t personally. I have seen one talk where the speaker augmented a beamer presentation with some handwriting via a tablet PC, and honestly the handwritten part was the worst part of the talk. Partly that was due to his handwriting being difficult to read, though.
Regarding the general standard of Beamer talks vs the general standard of board talks, I think it's worth noting that for many mathematicians, Beamer is still a relatively new medium. Most mathematicians have given waaaaay more board talks than Beamer talks. So I think a lot of people are still learning how to give a good Beamer talk, and we can hope for standards to improve.
Of course, there's the possibility that as people get more practised at using Beamer, their talks get worse. Maybe they learn ways of cramming more on to each slide, for instance, or maybe they get too fancy. But you'd hope that most people would get better.
There's a mathematician who I've watched, over a period of years, make the transition from always giving board talks to always using overhead slides. His slide talks have noticeably improved. Admittedly, it was from a low base: his first slides were simply printouts of Latex documents, often with sentences being interrupted at the end of one slide and resuming at the beginning of the next. But you can see the progress.
Recently I went to a week-long conference where all but four of the talks were done in Beamer. One of the exceptions was PowerPoint, two were overhead slides, and one was blackboard. Only one board talk all week! I was amazed. I think part of the reason was that many talks were only 20 minutes; and short talks are surely a reason to choose Beamer.
Another great reason to use beamer is if you have pictures to show, like string diagrams. Micah McCurdy gives excellent beamer talks which consist mainly of one or two proofs showcasing some aspect of string diagrams. The slides are perfect for seeing how the string diagram arguments go; if you set them up well then flipping from one picture to the next is almost like watching a movie of the isotopy relating them. I actually understand my own string diagram proofs better after going through them like that!
I think that the question as to which medium is used in the best talks is one that I’d like to avoid in this thread. I have a suspicion that, as Tom alludes to, there’s a large sample bias due to the fact that beamer is a new medium (for mathematicians, at any rate). Since we’ve all seen many more chalkboard than projector talks, the probability of the best (and most memorable) being chalkboard is higher. Since most mathematicians have given more chalkboard talks, the probability of them knowing how to use it is higher. And so on. I think that this thread will work better if we focus on the little things, such as that which Mike says in his last post.
I’d like to add something for consideration. I haven’t thought this through in detail. I think that a big difference between a projector and a board talk is the order in which stuff happens. Simplifying wildly, consider the way in which one might present a theorem.
Board: You talk a bit about the theorem, write up the conditions, say a bit more, then write up the conclusion. At this point, you pause for people to take notes.
(You do pause, don’t you?)
Presentation: You show the whole theorem at the start, then talk about it.
So the default on a presentation is to be reactionary: you react to what is on the screen. Whereas on a board, I think the default is to be a bit more proactive: what you write on the board tends to summarise what you’ve just said.
Of course, there is no reason why that should be the case; just that that’s an easy way to do it.
Let me also reply to Mike about the graphics tablet. It is important to practise. I think that my use of a tablet is now acceptable, but my writing on the iPad is boderline - it’s newer, and is inherently less precise. I also think that in a seminar talk, one should never plan to use the tablet (lectures are different). It’s there as a backup in case someone asks something that you didn’t think of, but for a seminar then your Plan A should be to give the talk without using it. (Mind you, it can be useful for highlighting bits of a presentation and then I’d say that the above doesn’t apply.)
You can see my lecture annotations on my course wiki. The early ones used a graphics tablet (and xournal). 30th September was my first one with an iPad.
Most mathematicians have given waaaaay more board talks than Beamer talks. So I think a lot of people are still learning how to give a good Beamer talk
Of course, there is no reason why that should be the case; just that that’s an easy way to do it.
Both arguments can be considered as counterarguments: we should give more talks with easier preparation rather than waste a time for talks where in preparation we need to be focused on thinking on technology, or to do a lot of such hard-to-prepare talks to overcome the difficulties of that. The purpose of teaching math is to convey one’s own mastering of mathematical THOUGHT. So one has to think during the talk and convey the thinking. Most often these are elementary steps and one can perfectly write them in real time, then one masters it naturally and has a natural pace, natural reactions, natural voice etc. The technology is needed only when one needs to skim over huge formula, huge diagram or picture whose details are not essential to convey but rather to remind of a shape, nature or something like that. Like reminding of something most people already saw. Here I do not distinguish blackboard from real time writing on some whatever-Pad which real time projects on the screen. i distinguish it from reading prepared notes.
At conferences there is also strong counterargument against prepared talks, namely what is said at the conference. I know speakers who change up to 80 % of the talk during the conference, regarding what other speakers in the area said before them about the subject. That would mean one would need to skip lots of activity if one had to rewrite 80% of a beamer talk.
Another problem is also that it is hard to see at transparencies (there are people like me who have some dioptria but do not wear glasses, what is OK for me for a blackboard talk if I sit in front row or two, but for beamer, transparency tc. talks I have problem in any row) and also that there are technology glitches like problem with files, projector etc. which often waste a lot of times at talks and also before the talks. Our faculty in math dept in Croatia is currently having all mandatory courses broadcasted, so they use smart boards and all that. The students tell me that the natural structure of the lecture is quite destroyed and that often the lecture stalls at some technical problem.
On the other hand, I think we should have software to record electronic pen writing (which is one dimensional trace so it is low bandwidth) together with timing information of the pen and the simultaneous voice. Such prerecorded lectures (made out of many small recording intervals, like with taperecorder) would be excellent for internet. Unlike the filming of whiteboard and blackboard which takes 2D info (not the trace but the whole picture) so it is very bandwidth consuming and still not clear to read (and often out of focus). So the videos at the internet are so hard to read and the picture is mainly irrelevant information like themotion of the figure of the speaker, while one can barely see the handwriting. Maybe I’ll go after writing such a software once my job position expires (what is in two months),
But a beamer also makes it harder to be disorganized and rambling
Nice letters, titles, sections and so on can easily cheat that the talk is not contentwise rambling, and regarding that people in beamer talks get typically less to the step by step reasoning it is even easier to hide. Instead of saying “our geometric theorem from the beginning of the talk can now be applied” one is saying irrelevant things like “on the next slide we will apply the theorem from slide 3”. I mean irrelevant is to focus our mind of flipping the slides and imagining this and not on mathematics. One should remember math by content, meaning and so on, and the blackboard talk should make its INNER symbolics vivid. It is maybe easy to display instead but this is focusing on ephemeral quasicontent.
part of the reason was that many talks were only 20 minutes
Correct, technology is the best to use at short talks. Though if I would come to some decision body, I would certainly decrease the funding to conferences which schedule talks of 20 minutes. One really learns at conferences where one can ask questions and where the focus of the whole conference is in some area. When there are multiple sections, short talks and all the mixture, then one can do lots of things superficially. Such big conferences sound good on CVs, and many people come there just to get business done and not math done. Typically I have seen that at such conferences there are also slips of complete rubbish, what does not happen at small focused conferences where the organizers control whom they will invite, rather than secretaries, fees and business. Also to make coherent the timing of different sections, one has to have the timing very strict, what does not allow honest public discussion. And for discussion one to one, people could do more skype and less pretending that they are interested in talks in a conference with 300 speakers. Of course, there are good exceptions. I am talking in general, I expect that the conference quoted by Tom above was more focused and is more tight in positive sense.
On the other hand, I think we should have software
Once these things are correctly solved and the community accepts the standards I will be happy to nearly stop traveling to conferences. It is usually huge waste of public and personal resources, health, strength, nerves and so on, and production of carbon etc. Millions wasted partly because nobody writes a good kit of software for complete in information, easy to use and low in bandwidth, free math communication.
I was thinking a little about chalkboards and came up with a list:
And on the other side:
Zoran wrote:
Maybe I’ll go after writing such a software once my job position expires (what is in two months),
That would be terrific!
These days, I give talks to undergraduates, on a subject that I don’t choose and spread out over ten weeks, rather than to colleagues, on a subject that I’m researching and summarised into an hour, so my experiences can be very different. But one thing that I’ve largely settled on is to make use of both the projector and the board. I show the basic facts and formulas that the students need from the projector and leave this up during the lecture, while I do examples on the board. The projector material is not very fancy; usually it’s just HTML (which the students can go online to review) shown at a large font size, not even divided into pages. The examples on the board are mostly made up on the spot (or done in response to specific questions from the students); sometimes I get one out of the textbook if I need it go a certain way.
So it’s possible to have a mixture of prepared and spontaneous material, each with its own technology. That is, if the lighting cooperates!
That is, is [sic] the lighting cooperates!
That’s one reason why I generally prefer not to mix these technologies and so use a system where I can write on a blank page in the presentation.
Maybe I’ll go after writing such a software once my job position expires (what is in two months),
Software like this already exists. On the iPad, there’s an app called “Explain Everything” that allows you to “record” a presentation. It records what you write in the way that you describe, and has the ability to record an audio channel as well. I haven’t tried it for its recording facility (and I think that a 2hr recording would stretch it a bit), but that’s what I understand from its description.
That aside, in that which Zoran said was something relevant to building a list of things to consider:
Nice letters, titles, sections and so on can easily cheat that the talk is not contentwise rambling
This is another danger for beamer presentations: to use sectioning as a way of breaking up a presentation. Just because beamer interacts nicely with sections doesn’t mean that that is necessarily a good way to structure a presentation.
This is another danger for beamer presentations: to use sectioning as a way of breaking up a presentation.
Can you elaborate? I have trouble imagining how to structure a presentation in a way that doesn’t involve some sort of grouping of separate parts, which might as well be called “sections”. Of course one doesn’t have to display the sections to the audience, but it seems to me that a little bit of displayed structure like that can help the audience to know where they are. (It helps me, as an audience member, when it is well done.)
That is, if the lighting cooperates!
That’s one reason why I generally prefer not to mix these technologies and so use a system where I can write on a blank page in the presentation.
Where I teach, the lighting for the board and the lighting for the projector are compatible if the projector is displaying the contents of the computer monitor. If instead the projector is showing a photograph of paper or transparencies that I can draw on, then the lighting must be adjusted. So I can either mix prepared HTML or PDF with spontaneous board writing, or I have to do it all by projector from paper or transparencies.
I guess that the point is that if you know where you’ll be speaking and can inspect the facilities, then you know your options.
Further to #25: When I give chalkboard talks, I always outline my notes into sections and subsections, with about half a page to a page of notes for each subsection. I’ve always done this, since long before I learned to use beamer, and I wouldn’t know how to give a coherent talk any other way.
Mike:
Remember that we’re just trying to get a list of common pitfalls in the two media. When properly used, sections (and similar) can be beneficial. But the problem is that their ease of use in a beamer talk means that people can use them without thinking about whether or not they should be used. Perhaps the underlying problem is that because beamer is created using LaTeX, and people get familiar with LaTeX for writing articles, they go into an “article writing” mode when they start writing a beamer talk, so use things like sections in the same way as they do for articles.
So I’m agreeing with your last sentence in #25:
It helps me, as an audience member, when it is well done.
with huge amounts of emphasis on the last bit.
Toby:
I guess that the point is that if you know where you’ll be speaking and can inspect the facilities, then you know your options.
This is crucial! I naively assumed that because my laptop worked in one lecture hall, that it would work in the other one that I’m in this semester. No such luck! Moreover, despite there being a desktop PC physically in the room and connected to the system, it was in a locked cabinet with no keyboard or mouse so I couldn’t use it. I’ve learnt the hard way that at the start of every semester I should go round the lecture rooms that I’ve been assigned to and check that everything works.
Software like this already exists. On the iPad, there’s an app called “Explain Everything” that allows you to “record” a presentation. It records what you write in the way that you describe, and has the ability to record an audio channel as well. I haven’t tried it for its recording facility
It is nice to hear that, but it is not the full story. It seems you did not entirely understand what I described. If you use a graphics application on your computer you can plug ANY electric pen device whatsoever to it. You can draw with a primitive mouse, or you can use a tablet with it. If I say SOFTWARE I mean software. That is you do not need a special single vendor hardware to be able to use software. It should go across platforms. Once I record it I should be able to view it as I view any other movie file. I should be able to SHARE the whiteboard with multiple users online, and record the session. Absolutely important is that I can paste together short records of several seconds at the time. If I do a record for a public movie, I should be able to do it like movies are done: from many perfected pieces. The purpose of the system would bot be to supply the guys at Harvard and Princeton who have all the resources already and are bored if the talk goes over 50 minutes but to those who have little resources, no resources to travel. I want to make presentations for colleagues in India and Pakistan for example. So tthe software should work on their linux.
It seems you did not entirely understand what I described.
Actually, I did. I was giving an example of a case where I already knew that such software existed. As it exists for the iPad, I would be surprised to learn that it doesn’t exist already on other platforms. If not, I can imagine that a fairly simple extension of xournal would fit the bill for Linux: simply record the time at which events occur as well as the events themselves.
The main point is not to invent new hardware. I am happy with electronic tablets attached as peripherals since I had one some 12 years ago or so. The price of an external electronic tablet is of the order of 30 euro and some give information on how close the pencil is to the tablet so one can have thick and thin trace and so on. No need for something new. Just I need to channel it to some of the record formats which could then be canonically converted to one of the free movie formats. Of course, in that last conversion one would loose the bandwidth compression, but one can have the expander into that format at the other end, or use a native viewer.
What is the compression for iPad ? I mean how big file would take 50 minutes of lecturing with drawing and voice ?
Also it is big thing that one can share the whiteboard with some voip or alike system, like voipdiscount or skype. Even recording skype voice requires mainly pay software (which can record sometimes for free but if you have only two conversants and the talk is shorter than 10 minutes in the version I tried).
I would also like to have a source code so one can do extensions. Like mathematicians in future would like to combine the real time drawing with some embedded equations generated from LaTeX. I am not sure if iPad has open sourcecode of such utilities so one can extend them for special purposes. I am not likely to be interested in solutions which are destined to die, as they are proprietary, partial and so on. Mathematicians need to use pencil as they do on the paper and share that with voice online, record that in pieces and import other pictures, and they have to be able to do it on free OS, and with low bandwidth. If so, it would revolutionize mathematical communications, especially outside of privileged rich centers.
There is Livescribe which does much of what I want but proprietory, with specific hardware, in windows and the hardware needs also special paper which is not reusable! I had somewhere colected the data from Australian researcher on the educational use of those but I do not remember where did I put it.
I did write about this some time ago in azimuth project:
http://www.azimuthproject.org/azimuth/show/Remote+conferencing
See the reference list. Though I was there more about low baddwidth conferencing than about recording lectures. But it is waste of time if one does conferences and can not record them or to do preparation of lectures if one can not collaborate online on this.
Toby 22 > That would be terrific!
Expiration of my job is not that terrific, from my point of view. Depending on refereeing speed of my works currently stalled in long time refereeing process in journals it could still turn it does not terminate but with memory to past speed of refereeing I really doubt. One of the papers was rejected at some point and the part of explanation was that the referee does not agree with me that the algebraic dual of the ring of polynomials is the ring of formal power series. Instead the referee insists that it is again a ring of polynomials in contradiction with the undergraduate fact which amount to the characterization of fin dim vector spaces as compact objects in the category of vector spaces…But being wrong does not matter if you are a referee…
Andrew #28: I think what I’m asking is, can you give examples of what it means to use sections poorly in a beamer talk? I’ve probably seen talks where the sections weren’t particularly helpful, but I don’t think I remember ever finding them actively detrimental.
Some talking points:
http://zachholman.com/posts/slide-design-for-developers/
http://www.slideshare.net/otikik/how-to-make-awesome-diagrams-for-your-slides
http://lifehacker.com/5810271/how-to-create-presentations-that-dont-suck
…except possibly the last point: narrative structure. Urs’ slides had this, for example, for fun to start with, and then in a serious way, which mirrored the fun bit.
In Wales, for many years we had a more or less weekly MPPM video seminar that was shared between Abersystwyth, Bangor, Cardiff, and Swansea. The technology used was a set of cameras with one pointing at the seated speaker, one at the ‘audience’ and finally one on a overhead projector-type of thing next to the speaker. This resulted in some very good (and some bad) seminars. I think the best was given by a young Japanese mathematician who had prepared neat handwritten sheets (of white paper) but with some of the details left out. She thus talked at a speed that was similar to many who use Beamer style techniques, but then when needing to slow to get a point across would write (very neatly) the details in a more interactive style.
The techniques she used could be well adapted to a beamer plus board presentation, where some slides are very sparsely filled with gaps to be expanded on. Unfortunately this is not always possible as the ‘experts’ who design lecture and seminar rooms all to often use the board space for the projector screen leaving about 2 cm^2 of space for writing on the actual board. (Mathematicians should consistently complain when this is forced on them. I now tend to prepare chalk talks because I never know what the space available will be like.) One could get around this with technology as discussed by Zoran above but this needs to be integrated into the same pdf-file system if it is to work well, The effect of seeing the equations being written is too important to allow fiddling around with the software at the time of presentation.
I have tried to use Beamer with messages to myself such as (We need some pictures of cobordisms so I should draw some!). This changes pace and concentrates attention away from the screen as that can be mesmerising.
Another excellent talk I recall was by Jeremy Gunawadena on the use of the max+ algebra in operational research and systems modelling. Some of his slides (in a ppt format) had just three words on them. The effect was great for breaking into the routine and grabbing attention. Points about overloaded slides have already been made and I won’t repeat them.
(Edit: before someone asks MPPM = mathematical Physics and Physical Mathematics.)
Another question with Beamer which came up in a discussion with a friend last night is the (over)use of \pause. It’s tempting, when you’re new to Beamer, to put in lots of \pauses, especially if you are thinking about a chalkboard talk where you only write things as you say them. However, I think this is usually a mistake — it generally works better to have almost no \pauses, and only use them when the extra emphasis or amusement value of having something new appear at the end of a slide is really worth it.
I think the reason for this is that on a chalkboard, it’s okay if words only appear on the board at the same time as you say them, because they’re going to stay up there for a while. But on a slide, everything is going to disappear when you move to the next slide, so other things being equal, it should be up there for as long as possible—starting as soon as the slide appears. Let the audience read ahead of what you’re saying, unless there’s some particular reason for them not to. Few things are as annoying as having a crucial definition finally appear only at the very end of a slide, then immediately disappear when you move to the next one. (Of course, this also flows into the principle of repeating important definitions and things on later slides as well.)
Formulating this point suggests to me that there are ways in which listening to a beamer presentation is a somewhat different skill from listening to a chalkboard talk—it helps if you’re willing to read ahead down the slide as you listen to the speaker talking about the first part of it.
WIth regard to the subsections, I think that the problem with sectioning is having too many levels. In an article, it can be common to have a structure like:
section
- subsection
- subsection
section
- subsection
- subsection
section
- subsection
- subsection
The problem with that in a presentation is that Section 2.1 probably doesn’t follow on from Section 1.2. I’d say that a presentation should have a linear structure and each piece should follow on from the previous one. Too much sectioning means that bits don’t follow on from immediately preceding bits.
I think you’re right about the pauses. They should be only used when it’s important that the audience doesn’t read ahead. So I think that they are liable to be overused, but they can also be used to good effect to ensure that the audience’s attention is at the right place at the right time.
I’d say that a presentation should have a linear structure and each piece should follow on from the previous one.
That’s ideal, yes, but sometimes there is no escape: you want to say C, which depends on both A and B, neither of which has anything to do with the other; so you need to say one of them first, then switch gears and say the other.
Regardless, I do agree that overuse of subsections is at best useless in a beamer talk. I rarely if ever use subsections at all.
you need to say one of them first, then switch gears and say the other
In a narrative, you have transitional phrases here; if nothing else, say ‘This concludes A; now we must deal with B.’ (and hopefully both A and B were mentioned in a common introductory bit beforehand). This might not need to go on the slides, but it should go into the talk.
on a slide, everything is going to disappear when you move to the next slide
This is one reason why I no longer put my prepared material on slides at all; it’s all just a web page that I scroll down. So I’m usually talking about the stuff that’s summarised in the bottom half of the screen, while earlier material is still on the top half.
if nothing else, say ‘This concludes A; now we must deal with B.’
Yes, obviously. But I don’t think that makes B “follow on” from A in any logical sense.
There’s a difference between the logical sense and the narrative sense. I’d say that the narrative flow is more important. What I’m highlighting is that subsections are part of the logical sense, but need to be made subservient to the narrative sense and that takes concious effort.
subsections are part of the logical sense, but need to be made subservient to the narrative sense and that takes concious effort.
Sure. But not very much. I don’t think this has been a noticeable problem for me with many (or even any) beamer talks that I’ve seen – except possibly those that also had so many other problems that I stopped paying attention quickly.
Here’s a relevant thread on MathOverflow: http://mathoverflow.net/questions/80056/using-slides-in-math-classroom. Most of the answers make me want to scream! I went through leaving little comments but deleting them all before posting them …
Anyway, back on topic. Mike: I agree with the sentiment of your last post. A presentation where there’s lots of disjointed parts probably has more issues than just the level of subsectioning, but maybe if someone is made aware of the need for a decent story through the lecture then they’ll take a critical look at the whole thing, not just whether to use \subsection or not.
The MO thread is about teaching, whereas this thread is I think mostly about giving conference and seminar talks, but there’s some overlap I guess. Your answer there has convinced me that I might, maybe, conceivably, consider one day giving a lecture course with beamer. I still have trouble believing that there is space on beamer slides to explain even a medium-length argument.
I’m curious about your vehement objection to combining a screen and a chalkboard (which is relevant to conferences and seminars too). The lighting depends a bit on the room, but I’ve definitely seen rooms where the blackboard can be lit without washing out the screen display unacceptably. But does it matter anyway, if you don’t expect people to look at the screen at the same time as the blackboard? That doesn’t seem any different from switching to a blank page on your laptop to write on. And I have no idea what you mean by “the wrong mindset” – how is the mindset of watching someone writing on a chalkboard different from the mindset of watching someone writing on a document camera or tablet/iPad?
Firstly, my vehemence on MO should be taken with a considerable amount of salt. It’s partly because of the anti-discussion format of MO and also because this particular area (beamer vs chalk) is so bizarre that I feel that I have to shout just to be heard.
I still have trouble believing that there is space on beamer slides to explain even a medium-length argument.
It’s quite astonishing how ones perception of space changes. The projector screen in my lecture hall is absolutely massive, covering acres more space than the chalk board. And yet, my reaction is the same: how do I get all the steps of, say, the completeness of onto one screen? I could easily do so on the chalk board without erasing it. The wrong answer is to cram more onto a slide, just as we cram loads onto a chalk board (really, how visible is that “medium-length argument” from the back of the hall?). The right answer is to split it up so that it becomes several short arguments and to explain how it splits, explain the pieces, and then explain how the pieces come back together. How many of your students really remember the start of the medium-length argument when you get to the end?
So to the mix-and-match. As I said, my vehemence is deliberately over the top. I’m sure it is possible to have both, though even with acceptable lighting there is a considerable contrast change from going from dark-on-light to light-on-dark. But mainly the problem is deeper: when someone switches back and forth between the two then it’s a sign that they haven’t really thought things through properly. They like bits of one style, and bits of the other, but aren’t prepared to commit. I’m talking really about the kind of talk that is meant to be a beamer presentation, but then the speaker gets a bit scared and suddenly realises that something isn’t explained right, so they’d better write something on the board (the worst is when they leave a gap for them to draw a picture on the board!). If the switch is planned, and is a long switch: maybe 5 minutes of presentation, then half an hour of chalk, then 5 minutes of presentation again as a summary, then my objections are not so strong.
As well as the contrast, the difference between if the speaker switches to a blank page on the laptop as to a board is that I have some hope that I can get a copy from the laptop so I can continue in my mode of concentrating on what the speaker is saying and not have to switch suddenly to rabidly copying down every chalk stroke in fear of missing something.
Ultimately, it boils down to the one thing that I didn’t say, which perhaps I really ought to have put first. A chalk lecture and a presentation lecture are completely different in style. Now, this may be as much down to our perception of them as to anything in reality, but there is a qualitative difference. In a chalk talk, there is a much greater synergy between you and the board. It is you and the board out in front, not just you alone. With a presentation, then it is either you or the presentation, but very hard to have it be both. Some will stand to one side and simply talk about what’s on the screen. Then it’s the presentation out front and you are the supporter. Or you can talk a lot and use the presentation simply as a way of remembering important points. Then it’s you out front. It’s probably not hard to guess that I’m in favour of the latter! I would also say that the companionship of the board is only apparent to the speaker, not to the audience. They may pick up on it, but only second hand. There is no particular relationship between them and the board.
But then if you put me in front of about 100 students then I do turn into a bit of a show-off. I don’t want to share my limelight with a chalk board! But I don’t regard that as a bad thing. I want the students to engage with the maths, and if it helps on the way to that to wave my arms about and tell silly jokes, then I’ll do so. And, to be honest, I feel any different when I go to a conference as when i sat in an undergraduate course, so I do the same in seminar talks (though perhaps not to quite the same degree).
a considerable contrast change from going from dark-on-light to light-on-dark
On a whiteboard, it’s also dark on light. I have very little experience with mixing projectors with a blackboard, but I imagine that it might be annoying.
when someone switches back and forth between the two then it’s a sign that they haven’t really thought things through properly
No, it means that they’ve thought things through differently from how you would have. You basically admit this, saying that you’re really only talking about certain cases. I agree that if the speaker switches to the board because they realise that something isn’t right, then this is because they didn’t think things through, but that’s no great revelation. (And in that case, thank goodness that they at least have the board!)
They like bits of one style, and bits of the other, but aren’t prepared to commit.
As a bisexual person, I have basically zero tolerance for this argument. (I have also been known to vote for both a Democrat and a Republican in the same election. Sometimes I eat broccoli and cauliflower in the same meal.) You have to establish that the two styles are incompatible before you can argue that they shouldn’t be mixed. This is what you have failed to do, in the face of people who acknowledge the differences and tell you that they use both for different purposes in the same lecture.
rabidly copying down every chalk stroke in fear of missing something
This I take as an argument against board talks, full stop.
when someone switches back and forth between the two then it’s a sign that they haven’t really thought things through properly
No, it means that they’ve thought things through differently from how you would have.
Hmm, I could have phrased that better. I was meaning “a sign” to mean “evidence towards the conclusion” not “definitive proof of”.
I guess that what I would really like to see in a talk is evidence that the speaker has thought things through at all. And by “things” I mean the “everything-but-the-maths” things: the room, the technology, what to do if you suddenly remember that you haven’t defined something, whether to give notes. All these things are important. Yes, it’s important to have something to say. But once you have it, how you say it becomes the next thing to decide.
So you should interpret all of my remarks as saying, “Think!”. Now the best way I’ve found to get people to think about these things is to tell them what I think, and do it in such a way that they are forced to decide for themselves whether or not they agree with me. Sometimes I get a bit enthusiastic. Sorry. Particularly with regard to:
They like bits of one style, and bits of the other, but aren’t prepared to commit.
As a bisexual person, I have basically zero tolerance for this argument.
You’re right. It’s a lousy argument. But to go on to your broccoli and cauliflower analogy, as a parent I have basically zero tolerance for people who say “I don’t like broccoli” simply because it is green. The kind of person that I had in mind when saying this is the kind that says, “Oooh, shiny projector! I like it! And I’ll hold on to a piece of chalk as a security blanket.”.
This is the slightly surreal part of debating this with you and Mike. Clearly both of you have thought seriously about how you give your lectures and talks. So I’m not that bothered about trying to convince you to do anything differently. Nonetheless, I think that there is value in discussing it with you both (and anyone else who cares to join in) and I’ll try to tone down the vehemence and remember that I’m in the nForum (which is a library, after all) and not in MO (bit more like a cafeteria).
So, the challenge:
You have to establish that the two styles are incompatible before you can argue that they shouldn’t be mixed. This is what you have failed to do
To be clear, it isn’t that I consider that the styles are incompatible within the same lecture but that I consider them to be sufficiently different that the change has to be clearly marked and not done lightly or for only a short time. Not of all these will apply every time, but here’s my starting list:
Lighting. (As I said, not all will apply always.) Particularly when the board is lit from the roof, it can make the slides indistinct.
Positioning. In many halls, the projector screens obscure a significant amount, if not all, of the boards.
Contrast. Even when the lighting is okay, there is a difference in contrast between going from dark-on-light projected text to light-on-dark board text.
Size. Board sizes vary considerably, and particularly with sliding boards, but it may be that the board and screen are positioned such that a good place for seeing one is not a good place for seeing the other.
Audience activity. The default activity with a board talk is to copy down everything on the board. What to do with the contents of the screen is much more up to the individual. Knowing whether or not to switch is important.
Audience focus. The order of audience focus is often different. With a presentation, often the material is revealed first and then talked about. With a board, often it is the other way around: the material is talked about and then written up on the board.
That’s what comes to mind. No doubt I’ll think of more right after hitting “Add your comments”.
rabidly copying down every chalk stroke in fear of missing something
This I take as an argument against board talks, full stop.
Please do. It is meant as such. I’m amazed that I used to think it a good use of my time to effectively dictate stuff for the kids to write down. This is my one chance to actually interact with the students and get them hooked on maths.
The projector screen in my lecture hall is absolutely massive… The wrong answer is to cram more onto a slide
Yes, as you are obviously aware, the physical size of the projector screen is irrelevant; what matters is for how long a given slide will be visible.
How many of your students really remember the start of the medium-length argument when you get to the end?
Well, splitting up an argument into bite-size pieces is good advice regardless of the presentation mechanism. But the advantage of having it all up on the chalkboard is that they can still see it, and I can still point to it.
I guess that what I would really like to see in a talk is evidence that the speaker has thought things through at all.
But why is switching between beamer and board any evidence that they haven’t? Couldn’t they have just come to the conclusion that switching is what they will do if they realize they haven’t defined something?
I must say I don’t really understand the “contrast” argument. Can you point to studies showing that it is difficult for people to switch contrast? I have a key combination on my Emacs to switch it between white-on-black and black-on-white (and I do, occasionally), and when I switch it it takes me less than a second to adapt.
have to switch suddenly to rabidly copying down every chalk stroke in fear of missing something
I’ve long ago mostly given up trying to take notes in a serious way on conference and seminar talks, unless I’ve agreed with the speaker in advance to be a notetaker. I want to spend my mental energy trying to follow the talk, for as long as it looks like it might be interesting, and then if it turns out not to be interesting I don’t want to waste my time. The most that I’ll do now is if the talk does turn out to be interesting, I’ll jot down a few main ideas that I want to think more about later, or to remind myself to look up the corresponding preprint or get the slides when they’re posted online.
Plus, I’ve found that taking notes on a beamer talk, no matter how well-designed it is, is basically doomed. Taking notes (for me) is always a process of jumping ahead and behind, sometimes writing down what the speaker is saying rather than what is written down (when it’s important and interesting), which sometimes gets me ahead of the game if he then writes down what he just said, and sometimes listening to the speaker while writing down what was just written, or even what was written a while ago. I can never stay exactly on pace with writing down what is shown on the board or projector, so even if the slides are displayed for long enough that I could, in theory, have written them down, I always get off by a few seconds at some point and then I’ve lost when the slide changes. Chalkboard talks are much more forgiving in terms of timing, if you’re trying to take notes.
Having written all that, I realize it’s not that relevant to the discussion, since you’re obviously not proposing that anyone should be trying to take notes on a beamer presentation! But here I think I get a little into the “don’t base your lectures on what worked best for you” problem, in a slightly different incarnation. Namely, I have never been taught a lecture course that used slides in an effective way. I think this is one reason why it’s very difficult for me to imagine how the students will learn. Perhaps this is just a psychological block I need to get over, but it would probably help if I could at least observe someone else teaching well with slides, not just read you writing about it.
A related point is that on the few occasions when I’ve been in a lecture and was given a printout of the slides in advance, I didn’t feel very successful in taking “additional” notes. Generally, for most slides there isn’t much I want to write down, but then on some there is a lot, more than there is space for on the printed slide or on any itty-bitty “note-taking” spot that was placed next to it on the printout. The only thing to do is write “slide 17: …” and then take notes on some separate sheet, but this breaks the continuity of having all my records of the talk in one place.
But why is switching between beamer and board any evidence that they haven’t? Couldn’t they have just come to the conclusion that switching is what they will do if they realize they haven’t defined something?
Yes, but then they’ll do it in a different way to the examples that I have in my mind. That’s the problem with trying to extract generalities from specific examples - none of the rest of you have seen the examples that I’ve seen. And all of the ones that I’ve seen have managed this switching very badly.
So, I concede that it can be done well. But the speaker needs to have thought about it beforehand and thought carefully about what circumstances will require him or her to switch.
Can you point to studies showing that it is difficult for people to switch contrast?
No. I’d like to be able to, but I don’t know of any. However, the counterexample of the computer screen is even weaker! I know that I find it difficult to adjust, but then I’m an old dinosaur …
But here I think I get a little into the “don’t base your lectures on what worked best for you” problem, in a slightly different incarnation. Namely, I have never been taught a lecture course that used slides in an effective way. I think this is one reason why it’s very difficult for me to imagine how the students will learn. Perhaps this is just a psychological block I need to get over, but it would probably help if I could at least observe someone else teaching well with slides, not just read you writing about it.
This, I think, is the crux of the problem. Since we’re never actually taught how to teach (okay, I’m generalising again), we’re never taught about other methods and - more importantly - we’re never shown these other methods. And never shown them in our field.
My wife is a teacher. She’s been telling me for years how to teach, but I’ve always found it hard to envision how what she says can fit into the constraints of a lecture system. It’s only when I actually tried it that I realised how naturally it worked. I’m no where near the standard that she’s at, of course. But I feel it would be so much easier if I just had good examples of lots of different methods of teaching mathematics at the university level.
I have to agree with Mike that blackboard talks are much more forgiving. A blackboard holds about three slides worth. Or, if you have sliding boards, you get the equivalent of about six slides, which is great if the talk is thoughtfully sectioned. With slides, I almost always want to cry out at some point, “wait! I’m not done thinking about that slide!” (But maybe this has more to do with following arguments on slides?)
I guess I didn’t understand what Toby does back in #40. How much of the entire talk can be visible at any one point?
This might be a point where seminars and lectures diverge. In a lecture, I don’t want them thinking “ahead and behind” too much. I want them focussed on what I’m telling them right now.
Lest I seem like I’m very dogmatic, I’m also intrigued by Toby’s approach and I do appreciate that this “back and forth” nature is quite important. With the old transparencies the trick was to have two projectors and to rotate the slides so that the old one was always on the “second” screen. I often have two projectors in the room, but have never thought of using them together. Maybe having a running summary on the second projector (I think that having it so that the previous slide was shown wouldn’t work, it would have to be the previous frame, and even then the method of swapping slides would be awkward).
Todd, how much would a handout version ameliorate your need to see other slides?
A blackboard holds about three slides worth.
Unless you’re at Oxford and lecturing in the University Museum. Then a blackboard holds about 30 slides worth.
Andrew, probably it would ameliorate it a lot, but the problem is that my line of sight might have to shift between straight ahead to the projector as you’re talking and down to the copy on my desk, which is not optimal. Or, I could keep my eyes glued to the copy on my desk, I guess, and just listen to your voice, but maybe you like to gesticulate and be a mixed-media package.
Re thinking ahead and behind: there is also the issue of whether one conceals what is coming up on a slide, say an old-fashioned overhead slide. Doing so means you’re keeping an exact focus on what you’re saying right now, in the moment. But, I gather that a lot of people don’t like that. I think I’ve heard both Rota and John Baez make fun of the practice (“doing a pathetic little striptease” or words to that effect). They would rather you lay bare the entire slide all at once.
However, the counterexample of the computer screen is even weaker!
Weaker than what? Would it be stronger if I said “I have no trouble switching between looking at a screen and looking at a blackboard”?
In a lecture, I don’t want them thinking “ahead and behind” too much. I want them focussed on what I’m telling them right now.
I react instinctively against any philosophy that wants to micromanage the student’s mind. Different people learn in different ways. If a student finds whatever you’re talking about right now obvious, but is confused about something you said a minute ago, why not let them figure that out so that they don’t progressively get loster and loster as the lecture goes on?
there is also the issue of whether one conceals what is coming up on a slide, say an old-fashioned overhead slide
Yes, I think I brought up the same issue in #36. (I know it’s hard to keep track of everything that’s been said in this discussion!)
I guess I didn’t understand what Toby does back in #40. How much of the entire talk can be visible at any one point?
Try this example. This is fairly well polished, but you can see the downsides of using HTML. (Of course, HTML is not really the point of the continuous scroll, but it becomes possible, and then it’s so convenient. I probably would do better making a PDF on a long page or something, although I don’t want students to complain that they can’t print it, because you just know that somebody will try.) Hit Ctrl-Shift-+ (or whatever does zoom on your browser) until you get something that would be a good size; I guess that one good thing about HTML is that you can actually adjust this. I use a serif font, which is bad on the screen at normal size but seems right when things are big.
Any time that you can imagine taking a pause to do an example of what was just said, I probably do it; that goes on the board, and I usually just wing it.
Thank you, Toby. That looks very nice. And it looks like you could even get a lot of that on one screen at one time.
1 to 56 of 56