Notices of the American Mathematical Society

Welcome to the Notices of the American Mathematical Society.
With support from AMS membership, we are pleased to share the journal with the global mathematical community.

PDFLINK

Less is More: 30-Minute Talks

Robert J. Berman

Yanir A. Rubinstein

In 2022, perhaps partly as a post-pandemic rethinking of old habits, we decided it was time to experiment with new formats for mathematics conferences. After years of dreading the hour-long talks and the typical stagnation they bring at math conferences, we decided to try something different: divide that same hour between the speaker and the audience. We tried out this new format in a workshop that we organized in Cetraro in Fall 2022.⁠Footnote¹ All speakers (junior or senior) were allotted exactly 30 minutes to present their mathematics. This was followed by an open discussion (for up to 30 minutes, typically 15 minutes) guided by questions/remarks/comments/suggestions from the audience. The schedule looked like this:

“Convex and Complex: Perspectives on Positivity in Geometry,” October 31–November 4, 2022, Cetraro, Italy, https://sites.google.com/view/convex-and-complex-2022/

✖

Table 1.

The Cetraro workshop schedule.

$\renewcommand{\arraystretch}{1} \setlength{\unitlength}{1.0pt} \setlength{\extrarowheight}{0.0pt}\begin{SVG} \begin{calendar}{8in} \day{}{ Talk 1 \daysep\textbf{9:30am-10:00am} \\[3pt] Talk 2 \daysep\textbf{10:30am-11:00am} \\[3pt] Talk 3 \daysep\textbf{11:30am-12:00pm} \\[3pt] lunch and collaboration \daysep\textbf{12:30pm-4:00pm} \\[3pt] Talk 4 \daysep\textbf{4:00pm-4:30pm} \\[3pt] Talk 5 \daysep\textbf{5:00pm-5:30pm} \\[3pt] } \par\day{}{ Talk 6 \daysep\textbf{9:30am-10:00am} \\[3pt] Talk 7 \daysep\textbf{10:30am-11:00am} \\[3pt] Talk 8 \daysep\textbf{11:30am-12:00pm} \\[3pt] lunch and collaboration \daysep\textbf{12:30pm-4:00pm} \\[3pt] Talk 9 \daysep\textbf{4:00pm-4:30pm} \\[3pt] Talk 10 \daysep\textbf{5:00pm-5:30pm} \\[3pt] } \par\day{}{ Talk 11 \daysep\textbf{9:30am-10:00am} \\[3pt] Talk 12 \daysep\textbf{10:30am-11:00am} \\[3pt] Banquet \daysep\textbf{7:30pm} \\[3pt] } \par\day{}{ Talk 13 \daysep\textbf{9:30am-10:00am} \\[3pt] Talk 14 \daysep\textbf{10:30am-11:00am} \\[3pt] Talk 15 \daysep\textbf{11:30am-12:00pm} \\[3pt] lunch and collaboration \daysep\textbf{12:30pm-4:00pm} \\[3pt] Talk 16 \daysep\textbf{4:00pm-4:30pm} \\[3pt] Talk 17 \daysep\textbf{5:00pm-5:30pm} \\[3pt] } \par\day{}{ Talk 18 \daysep\textbf{9:30am-10:00am} \\[3pt] Talk 19 \daysep\textbf{10:30am-11:00am} \\[3pt] } \par\finishCalendar\end{calendar} \end{SVG}$

Why Are Most Talks an Hour Long Nowadays?

Let’s consider the bigger context encompassing not just conference talks but also departmental seminars.

When we were graduate students, we recall our department elders recounting how when they were graduate students there was just one department seminar: the colloquium. That seems hard to fathom; already in the early 2000’s departments seemed to have many seminars every week in addition to the colloquium. A similar statement could be made about conferences, with the rather few regional meetings of societies (e.g., AMS and ICM) historically playing a much more central rôle due to the scarcity of other workshops. Historically, it seems plausible that seminars stuck to the hour-long format since that was the existing format for the colloquium.

Problems with Hour-Long Talks

Colloquium talks target diverse mathematical audiences and it is understandably challenging to fit a presentation within the allotted 50–60 minutes. For workshop and seminar talks, naturally more technical, there is no particular reason to stretch the talk over a whole hour. After all, what is the point of the speaker transmitting more information than the audience can absorb? In our mind, the prime goal of a speaker should be to enthuse the audience about the topic of the talk. Putting too much and too detailed information into the talk may, in fact, have the opposite effect. By analogy, imagine inviting friends over for a dinner. Even if the food is delicious, its main effect on your friends—if served in far too large portions—is nausea.

Perhaps one more reason that math talks tend to be around one hour is that the speaker feels an urge to answer all potential questions—before they have actually been raised by the audience. Why not instead leave ample time for questions after (or during) the talk?

Another problem with many workshops is that, while one of their declared goals is to foster discussions and collaborations, in practice there is relatively little time remaining in the packed schedule, aside from numerous talks and seemingly ever shorter meal breaks. Perhaps this is related to the ubiquity of conferences: most faculty will not attend a conference unless they are an invited speaker, which results in squeezing more talks into the schedule.

Experimenting with a New Format

With the previous remarks in mind, our goal in the Cetraro workshop in 2022 was to experiment with a different format. Our first idea was to limit both the number and the length of the talks, while simultaneously increasing the length and number of breaks. We insisted that speakers prepare a 30-minute talk, a rather drastic departure from the traditional model. We asked speakers to focus on the main ideas, challenges, and problems, and not to dwell on technical details of proofs. We also allotted 30-minutes between talks, another crucial departure. In practice what happened is similar to the well-known “flipped classroom” model; there were lively discussions that went on for about 10–20 minutes after the talk ended, and speakers engaged with participants (or, perhaps, participants engaged with speakers!) in a way that was absolutely inspiring, in effect transforming the (typically passive) audience into (active) participants. In that workshop we also limited to three talks in the morning and two talks in the afternoon, allowing a three-and-a-half-hour break in the middle of the day. In addition, on Wednesday and Friday we had only a morning session of two talks. As a result, there were several useful blocks of collaboration time as well as much needed rest time (a side effect was that participants did not have to skip talks in order to find time to collaborate or rest, and this in turn benefited the lively discussions following the talks). This resulted in participants being less fatigued in the afternoon session and more productive in collaboration during breaks.

The Cetraro format was a great success according to testimonials by many participants. Many participants either told us or wrote to us saying this was the best conference they ever attended. But perhaps this was just a fluke?

To further test this format, we repeated this experiment in two more meetings, each slightly different from the first. The first workshop in 2022 concerned the interactions between two nearby fields and was rather small in size (about 40 participants). The second meeting, in Summer 2024, was a major conference encompassing a rather vast field with around 120 participants.⁠Footnote² The third workshop, also in Summer 2024, was an even more focused workshop than the first, and with even smaller attendance (about 30 participants).⁠Footnote³

“Frontiers of Geometric Analysis,” June 3–7, 2024, Santa Cruz, California, USA, https://sites.google.com/umd.edu/frontiers2024/

✖

“Workshop on algebraic and differential geometries related to Arakelov theory,” June 10–14, 2024, Göteborg, Sweden, https://sites.google.com/view/freixit2024

✖

What transpired from these three meetings was that regardless of the size of the meeting and its focus, the format seemed to work extremely well, and in each setting different benefits of it were highlighted.

To sum up, the main advantages of shorter talks that our experiments at Cetraro, Santa Cruz, and Göteborg exhibited are:

1.: Increased Engagement: Shorter talks tend to maintain audience attention more effectively, reducing fatigue and loss of focus.
2.: Concise Communication: Speakers are encouraged to distill their message to its most essential points, making the content more accessible and easier to digest.
3.: Facilitates Discussion: Allocating additional time for audience questions and interactions fosters a more dynamic and participatory environment, enhancing overall understanding and engagement.
4.: Flexibility: Shorter sessions allow for more breaks and unstructured time, which can be used for informal discussions, networking, or collaboration.

Potential Drawbacks

Every format has drawbacks. Let us turn to our advocatus diaboli for the three key ones:

i.: Too short/condensed: The most common objection to shorter talks is that the reduced time frame may not allow for an in-depth exploration of complex topics during the talk. Some speakers may also feel they need to condense an hour-long talk into the 30-minute slot.
ii.: Increased preparation time: The second most common objection is that it requires considerably more time to prepare a shorter talk. Or in the words of Blaise Pascal (under the pseudonym Louis de Montalte), “Je n’ai fait celle-ci plus longue que parce que je n’ai pas eu le loisir de la faire plus courte” (Les Provinciales, lettre 16, 1657).
iii.: Too few/too many talks: Incorporating large slots of time for collaboration, and reducing the overall number of talks can limit the number of speakers. On the flip side, the temptation to squeeze together many 30-minute talks while reducing collaboration time, can result in too many talks.
iv.: Perception of prestige: Being given only 30-minutes as opposed to a full hour could be perceived as less prestigious by some.

Let us consider each of these potential drawbacks more closely:

i. To clarify, in the lion’s share of cases, “too short” refers to a drawback the speaker might perceive (the audience more often than not does not mind brevity!). To that end, including a 30-minute slot after the talk provides ample opportunity for in-depth explanations and detailed discussions post-talk, guided by audience questions. This approach not only addresses the depth concern but also stimulates audience engagement (as described in point 3 above). It is worth pointing out that there were surprisingly few talks that went overtime (just one or two in the first and third meetings and none in the second).

“Too condensed” refers to some speakers’ instinct to try to give a 60-minute talk in a 30-minute slot by speaking rapidly and/or flipping through slides prepared for an hour-long talk. This is of course counterproductive to the audience. The next point discusses tips for preparation of the talk and from our experience can help address this issue.

Admittedly, there are settings (we believe rather special) where a longer talk is preferable. For instance, when a lecture is aimed at a broad audience, e.g., a colloquium on a particularly sweeping topic or other generalist presentation aimed at a broad audience, the traditional 50- or 60-minute talk may be more appropriate.

ii. One of the common complaints we received from speakers was “I had to squeeze the slides for a talk I had already given.” Indeed, given the existing prevalence of hour-long talks, it is important for organizers to ask speakers ahead of time not to try to squeeze slides for a previous talk into the 30-minute slot but rather to prepare a new talk that gives more of an announcement of results with examples and room for discussion. As we found out, multiple reminders might be necessary ahead of the meeting. Another piece of feedback from some speakers was that distilling the essence into a shorter time slot required considerably more preparation. Again, this can be flipped around as an advantage: preparing a shorter talk brings to the fore crucial choices that must be made. Arguably, this process provokes some in-depth reflection of the value different aspects of a topic present to a given audience. Instead of a one-size-fits-all buffet-type lunch, the speaker is asked to serve à la carte, keeping in mind their audience. More often than not, even the speakers who initially complained, acknowledged the value of their extra effort—whether for themselves, the conference as a whole, or their audience. Many also noted that the audience was larger than usual, likely due to the reasons mentioned earlier, such as reduced “talk-skipping.” Finally, it is worth mentioning that some speakers prepared several “optional slides” that turned out useful for answering questions during the discussion period.

iii. The Cetraro workshop included 19 talks (see Table 1). This seems to be a reasonable number of talks, though some organizers might like to accommodate more or fewer speakers. To address this, in the Santa Cruz conference we reduced the lunch/collaboration break by an hour and added a third talk to both of the “half-days” (Wednesday and Friday). This allowed us to accommodate 23 talks. In the other direction, one may also easily reduce the number of talks to 16 by having only 2 talks in the morning sessions of Monday/Tuesday/Thursday.

iv. This drawback is primarily for organizers. Given how some conferences distinguish between “junior” and “senior” speakers, allotting the former shorter slots, we were a bit concerned at the outset that our experiment might fail. To our delight, the acceptance rate of our invitations was seemingly unaffected by the 30-minute requirement. Moreover, even very senior speakers seemed to have no problem entertaining the 30-minute format. In this context, it is perhaps worth pointing out that we initially contemplated 40-minute slots for more senior speakers and 20-minute slots for junior speakers. We were glad we did not go that route. By having the 30-minute requirement for all speakers it seems the prestige issue disappeared. Additionally, a useful trick we found effective is to carefully choose the lineup for the first day to consist of researchers known for their excellent presentations. Their well-prepared talks, which concluded on time, along with the stimulating discussions that followed, set the tone for the rest of the meeting.

Other Short Talk Formats

Of course, there are at least a handful of even shorter talk formats out there, and it is worth pointing them out to emphasize that for some purposes they might be ideal.

• Lightning talks have become increasingly popular in recent years and sometimes are an alternative to the more traditional poster session. For instance, three-minute (or alternatively, five- or even ten-minute) talks can be very effective in getting graduate students attending a conference to network. It also allows such students to become “presenters” which is sometimes formally necessary for their department to fund their travel. This typically allows for just one or two slides.⁠Footnote⁴

⁴

See, e.g., University of Waterloo’s “Three Minute Thesis Competition” which was replaced in 2020 by a 1-minute video competition, https://uwaterloo.ca/math/current-graduate-students/three-minute-thesis-3mt-competition

✖

• American Mathematical Society (AMS) and Mathematical Association of America (MAA) meetings often feature 10- and 20-minute talks. The 10-minute talks are typically found in special sessions showcasing undergraduate research, such as those at MathFest. These shorter talks are ideal for students who may not have extensive material to present but still benefit from the valuable career experience of giving a talk. On the other hand, the 20-minute talks are common at AMS sectional meetings. They allow for the presentation of more substantial content and are well-suited for these meetings where a large number of talks occur simultaneously across many sessions.

• Traditional 30-minute talks have certainly been around for some time, by which we mean 30-minute talks without a built-in 30-minute break/discussion buffer slot between talks. While these do not quite offer the full advantages of the Cetraro format described above, they still can be very successful in many settings compared to hour-long talks.

Finally, it is entirely possible to “mix-and-match” by including several formats in the same conference. For instance, a recent conference in Banff⁠Footnote⁵ featured a mixture of hour-long, 30-minute, and 5-minute talks.

⁵

E.g., “On the Interface of Geometric Measure Theory and Harmonic Analysis,” June 9–14, 2024, Banff, Canada, https://www.birs.ca/events/2024/5-day-workshops/24w5264

✖

What About Seminars?

To bring our discussion full circle, let us return to the topic of seminars with which we began this article. Perhaps the Cetraro format’s success depended on having a group of like-minded researchers in a weeklong conference setting? With that thought in mind, we experimented with the Cetraro format in a regular seminar setting during the academic year 2023–2024. In order to facilitate the experiment, each co-organizer was given the freedom to decide on the format of the talk for a given week. Some weeks the talks were in the 30-minute format (i.e., a 30-minute talk followed by a nonmandatory open discussion of up to 30 minutes), other weeks in the usual 50–60-minute format, and still other weeks in an extended 75-minute format that encouraged speakers to incorporate plenty of background material aimed at graduate students. This experiment revealed that the 30-minute format could also be successful in the seminar setting. It also uncovered advantages and disadvantages particular to that setting. The main extra advantage was participatory: as seminars are typically scheduled in the afternoon to avoid teaching conflicts, often participants are unable to attend due to childcare or other family obligations. Testimonies by participants revealed that the 30-minute format allowed them to attend some talks that they otherwise would not be able to. The main extra disadvantage was that organizers were cautious that some speakers might not accept an invitation for “only” a 30-minute talk. To our delight, however, that never happened. On one memorable occasion the talk ended after exactly 30-minutes followed by a Gelfand-seminar type discussion that lasted well over an hour, very informal in style, with plenty of examples done on the board.

We truly hope that this format will become popular in our profession.