Improving CMS meetings, a call to session organizers
Driving back with my students from a somewhat recent CMS Meeting, I realized how demoralizing the basic setup of these meetings can be to our graduate students. Many of us go to these meetings to attend one or two sessions on areas of our interest and potentially stop by a plenary talk or two. The sessions consist of four blocks of five or six 20-minute talks, often given by graduate students and postdocs in the hopes of impressing some faculty members with their results and landing a job offer. The talks are therefore highly specialized and aimed at one or two potential employers. Of course, once this tone is established, others, not necessarily looking for jobs, follow.
The problem is that everyone else in the audience suffers, and no-one suffers more than graduate students who, because of relatively low exposure to topics outside of their research area, have the least chance of understanding what is actually going on in each talk. At best, this leads to disengagement, but it is probably more accurate to call this process straight up demoralizing.
I do not blame the CMS for this setup, as it is quite common in our field and in other areas across the sciences. The Joint Mathematics Meetings in the U.S. follow pretty much the same format, and their early January timing makes them even more prone to have job seekers give a highly technical talk. Besides, what would be the alternative? I also think that the CMS does a fantastic job organizing these events and it is up to us, faculty in attendance, to help the CMS make the most out of these meetings.
With four or so blocks of five-to-six 20-minute slots, what can I do as an organizer to make the event more welcoming and engaging? I thought about this question a lot over the last couple of months and, while I do not claim to have solutions, I would like to share with you some of these thoughts. Hardly any of them are mine, most being offered to me by colleagues in response to my complaining about the problem.
To start, I want to highlight something positive that I saw during the session “An invitation to low-dimensional topology” organized by Adam Clay and Patrick Naylor during the most recent Winter Meeting. Each of the blocks in the session ended not with a talk, but instead with “office hours,” during which the speakers were available to answer questions in a bit more detail than a typical 5-minute Q&A would permit after the talk. Plus, they offer a bit more privacy, allowing for questions to be asked without a whole room listening. These “office hours” can take a little effort to make work; for example, just announcing them might lead to speakers talking to each other instead of engaging with the audience. It is just like your regular office hours: when your calculus students come, you want them to see you ready to answer any questions that they might have, not chatting with your colleagues. The same principle applies here; the session organizers need to create an environment in which students want to attend these “office hours.”
Of course, office hours is only one of the potential uses of the time slots within each session, and the possibilities are almost endless. Two math institutes that regularly host meetings provide lots of inspiration here. The Banff International Research Station (BIRS) and the Simons-Laufer Mathematical Sciences Institute (SLMath) in Berkeley, Calif., have for years looked for different formats to engage the audience with mini-courses, panel discussions, open problem sessions, and so on. Of course, BIRS and SLMath have much more permissive formats, so session organizers need to think about how to incorporate these alternatives into our 20-minute-per-talk setup. Perhaps we could have a 2-part mini-course given by an expert on a recent (not their) result or technique? Or an ask-me-anything session on a specific topic? Or, even better, a “what didn’t work” session where 1-2 speakers could explain an approach to a problem that did not pan out.
It is perhaps unsurprising that if you ask for highly technical talks, you are going to get highly technical talks. During the CMS Winter Meeting 2025, Hans Boden and I decided to ask for the opposite. For context, Hans is a low-dimensional topologist and I am a homotopy theorist. While we are both topologists, we often find that the kind of topology each of us is doing has more to do with other areas than it does with the other kind. But during the recent meeting, we tried to organize a broad and inclusive session, called simply “Topology.” On several occasions throughout the process, we asked our speakers to make their talks accessible by including ample background and motivation sections in their presentations, understanding that the audience will be very broad. The result? While there is still obviously lots of room for improvement, I learned a fair bit of low-dimensional topology, its central problems, techniques, and connections to other areas. And, pretty much for the first time during these meetings, I was able to follow all the talks in my own area of homotopy theory. It was not just me though; the Q&A’s after the talks were far livelier than I ever remember them being. As it turns out, accessible talks tend to engage the audience more than highly technical ones.
All of what I said so far pertains mostly to the scientific content of a session; the organization of a session does not end there. It also has to do with how the organizers run the event. One issue I found quite appalling during the infamous CMS Meeting that made me rethink the whole setup was the quite arbitrary application of time limits. Let me explain. The session I attended was organized by three organizers, two of them working in a somewhat niche area, call it X, and the third, possibly added later to increase the chances of the session being accepted. Correspondingly, while time limits were strictly enforced on all of us not working on X, it felt that speakers on X-related topics were not subject to any time constraints. I remember leaving the meeting feeling like a token of scientific diversity and not an actual invitee, and I was not alone in that.
Imposing time limits on one’s friends and colleagues can be difficult. It is a confrontational situation, and the confrontation often includes a person that the chair has a complex relationship with (for example, if it is their former Ph.D. supervisor). But many (most?) participants, especially graduate students, do not see this complex relationship; they only see clear favoritism. To make the session inclusive, one needs to apply time limits consistently, regardless of the speaker’s seniority, the quality of their talk, and the session chair’s particular interest in the topic. I found two ways of doing this in a less confrontational way. The first comes from a love of soccer (or the game that, unlike football, you play with your feet). I bring with me yellow and red cards. Once 20 minutes pass, I show the speaker the yellow card; this is a sign for them that they are now “eating into” their Q&A time. After 25 minutes, I show them the red card, which indicates that the audience should join me in applause, making it impossible for the speaker to continue. A more low-key version is to simply stand up after 20 minutes, then initiate the applause after 25 minutes. The cards, however, typically add some humor to it, thus helping defuse a potentially awkward and confrontational situation.
If you are anything like me, you like to occasionally go to talks in different sessions, hoping to learn something about an area you are not necessarily an expert in. The worst thing that can happen then is when the speaker decides to rush over the first couple of definitions because “we all saw them already in such-and-such’s talk.” To me, the whole point of the strict schedule and each talk starting at the full or half hour is to allow participants to move between different sessions. If I have to attend such-and-such’s talk earlier to understand yours, then what am I really getting out of it? I would suggest that speakers just go over the same definition again, maybe putting their own spin on them if necessary. In the end, in my experience, mathematicians are like children – they most like stories they already know.
Continuing on the previous point, I would also discourage inside jokes during talks. Making a reference to the CMS Meeting 2013 “which we all remember,” or the annual conference in some specific research area that maybe a third of the audience attends will likely lead to all graduate students and most other audience members feeling excluded.
The list above is far from exhaustive; it is merely meant as a conversation starter for how we can all, as a community, make the CMS meetings more inviting and accessible to our graduate students. There are, of course, many other issues to consider, like the registration fees. My understanding is that the CMS currently runs these meetings at a small loss, and yet the registration fees for graduate students are prohibitively high. Is there a way for the CMS to take bigger losses at every meeting, looking for savings elsewhere, but lower the registration fees for graduate students? Furthermore, the idea of a “weekend conference” is slowly becoming outdated with more and more institutions recognizing the necessity of protecting the work-life balance. Having a conference run past lunch on a Friday is now a faux pas in some circles. How do the CMS meetings adapt to this changing reality? These questions are well above my pay grade; readers like me can focus instead on all changes big and small to make our sessions more inclusive. I welcome any feedback on the ideas discussed here and those that should have been but were not. My email is kkapulki@uwo.ca.
