I am a PhD student in pure mathematics about 4 months into my program. After consulting with my supervisor and some peers, I have signed up to give a 20 minute talk at an online conference, which has invited final year undergraduates writing a bachelor's thesis, masters students, PhD students and postdocs to sign up and optionally give a talk.

My supervisor and peers have told me that it is a good idea to sign up as it's really good for my CV. At this point in my program however, I am mainly learning about my field, and haven't produced very much original research. I have however come up an (in my opinion) interesting question that would make for a good talk if I had a partial answer, but at this stage I have not had much time to explore the answer.

The conference is in about a month's time, and I think it would be a great exercise to write a talk giving an introduction to the field and how it came about, but at this stage this is all I could do, and even this would be quite a task to do in a month!

The sign up page for the conference does encourage works in progress however, but I am not sure how much I can do in a month.

My questions are:

  • Was it a mistake to sign up for this talk?

  • If, in a week's time say, I feel that I can't present any partial progress, can I ask to be removed from the schedule, or is this unacceptable?

  • 3
    Is there anything about the nature of this conference that suggests new research results are expected? I've seen many talks that were basically surveys of a problem and what others (and often, but not always, the presenter) have done toward solving the problem. In fact, my Ph.D. advisor gave such a talk (I didn't hear this particular talk, however). May 10, 2021 at 6:05
  • 26
    Did you ask your supervisor these questions? May 10, 2021 at 10:20
  • 3
    I would say it's a bit unusual to give a survey talk instead of a research talk if survey talks weren't specifically solicited, but it happens sometimes, and as long you can do a decent job, it should be a good experience and opportunity to introduce yourself to the community.
    – Kimball
    May 10, 2021 at 15:02
  • 3
    In twenty minutes, presenting a new interesting question without answers (but maybe with some cool examples) is perfectly fine for a talk in pure math. May 10, 2021 at 20:43

7 Answers 7


Twenty minutes is not long for a talk. Moreover, most students drastically overestimate how much material you can put in a short talk. Giving a short introduction to the field and how it came about will be more than adequate to fill that amount of time, and would be a perfectly reasonable topic for an incoming PhD student with only four months in the program. Time permitting, it might also be nice to mention your research question and the general avenues of inquiry you are going to use to solve it. You should also leave a few minutes for questions and comments, particularly since this will give researchers in the audience an opportunity to offer suggestions on your research idea.

So no --- you didn't make a mistake signing up for this talk. Don't bother trying to cram out new research results for it --- just stick with your plan to give a short introduction to your field and a brief outline of your research question. That should be enough to hold people's interest and solicit some questions, and the twenty minutes will go by in a flash. I doubt a short talk like this is really going to help your CV all that much, but it will be good practice for giving conference talks, and if you are lucky, it might precipitate a longer conversation on your research idea with one or more of the researchers in the audience.

(And as Buffy points out in the comments below, no one will be upset with you if you don't use all your time --- running over time is bad, but having time to spare is fine.)

  • 44
    And, no one will hate on you if you don't use all the time. Running over is a different story, of course. Good advice here.
    – Buffy
    May 10, 2021 at 12:37
  • @Buffy comment only? seems humble of you. there's a flag to convert answer to comment. i want an anti-flag: to convert comment to an answer!
    – BCLC
    May 10, 2021 at 15:47
  • 4
    @BCLC, no, just an addendum to writer Ben's good advice. But thanks for the thought.
    – Buffy
    May 10, 2021 at 16:00
  • 4
    @BCLC: Comment folded into answer --- consider it anti-flagged!
    – Ben
    May 10, 2021 at 22:25
  • Ben and @Buffy : nice
    – BCLC
    May 11, 2021 at 6:03

For the past 30+ years, I've been giving talks at conferences, both academic and non-academic. Many, many times by the time I start writing the talk, I start to think "how am I ever going to fill the allotted time". Then, when I have my first draft ready, it's more "OMG, I have 60 slides, and only 20 minutes".

20 minutes is not a lot. If that includes time for questions, and an introduction, time will fly.

Was it a mistake to sign up for this talk?

That seems unlikely. Unless you show up without any slides and no preparation because you planned to write the slides the night before, but got drunk in a bar instead (I've seen that happen).

If, in a week's time say, I feel that I can't present any partial progress, can I ask to be removed from the schedule, or is this unacceptable?

It happens frequently that people ask to be removed from the schedule, for many different reasons. For conference organizers, this is part of the game. I was in charge of the program of a conference last year, and I had to make last minute changes because a few speakers cancelled. Conference organizers cannot force people to show up and speak. But you would do yourself a disservice.

Now, I understand you're nervous. I still am before giving a talk -- that will never go away. But whatever you're talking about, you are the expert. You know more about it than the audience. Presenting at conferences are a good way for people to get to know (of) you.

  • 5
    I would like to highlight one fragment: “I still am before giving a talk -- that will never go away.” — This is true for some people, but others learn to not be nervous at all over time. May 10, 2021 at 17:39
  • Was it a mistake to sign up for this talk?

No, you will be soon in exactly the same situation you describe, but the talk will be in front of the heavy weight of your topic, not BSc, PhDs and the likes.

Why do I say so? because a talk will and must be showing a work in progress, at different stages of completion. If the work is done, then you are presenting a paper you completed, which is relatively "boring". I can read paper, I prefer to read a paper rather than attending a talk with an introduction by the author to the paper (off-topic: same reasoning applies to the fact I do not like most of the travel guides, where you are suggested with which mood you should enjoy a certain view/building/landmark).

  • If, in a week's time say, I feel that I can't present any partial progress, can I ask to be removed from the schedule, or is this unacceptable?

You are facing a big climb: preparing your first presentation. In the next years you will use and reuse this material, as the core building block of your talk. Or at least, you think so. In reality, you will discover that you will be basically rewriting everything and even having only marginally acceptable partial progress, this will translate in a humongous amount of work in the last week. This first time experience, however, will help you survive for the future iterations of this process.

Jump in the unknown. Present your problem/goal (2 minutes). Present the literature you are based on (3-4 papers, 5 minutes). Present literature open points (3 minutes) and your questions (3 minutes). Present your approach (2 minutes), without any partial progress (0 minutes). Present your partial progress (2 minutes, if any).

One slide = one minute (or more).

Good luck!

  • 2
    ps: people will be quite annoyed if you will take more than 20 minutes, but they will be happy if you leave them 5 minutes or so to interact with you with questions and remarks. This apply also to the big conferences: leave 15-20% of the total allocated time for questions ...
    – EarlGrey
    May 10, 2021 at 14:28

In my years of giving and listening to talks, I've never heard someone complain about a talk that was too short. If you have ten minutes of material for a twenty minute talk, that's perfectly fine; there will be a few minutes of introduction, then your ten minutes of talk, then a bit of time for questions perhaps, then everyone will have a few extra minutes to get to their next talk - it will be refreshing for folks to not have to rush to the next talk after a 28 minute talk, as often happens.

And of course, don't be surprised if your talk does go longer than you expect. They usually do!


I will try to provide general answer that might help you later too.

You have to ask yourself: what is the function of your talk and who is your audience?

I have signed up to give a 20 minute talk at an online conference, which has invited final year undergraduates writing a bachelor's thesis, masters students, PhD students and postdocs

This suggests to me very general audience, and function of this event is more education than presentation of research results. You can choose other function/audience but you risk to have a mismatch.

Assuming these function and audience, 20 minutes is plenty to (for example):

  1. educate your audience on importance of your field, and how it connects to other famous fields
  2. present "an example of research-level question and process to answer it"
  3. educate undergrads/other PhD students on your process. Undergrads will want to learn how math PhD looks like in practice

As soon as you understand the function of your talk, it shouldn't be hard to see that you will provide value/be useful if you just teach people about your field and process. It is still hard though and require polish (short talks especially important to polish).

PS: still you can decide that the function of your talk is to present Fields medal-worthy work, and aim to fulfill that function. But should you?


Do you have any MSc project work or the like? I pretty much had the same experience and chose to talk about the project I had done before starting the PhD (at that point this could have been a basis for the PhD but ultimately this turned out differently). It wasn't super original and I felt slightly bad about it when preparing, but it turned out to be a good experience.


Those types of conferences (including students and PhD candidates) are often designed specifically, or at least partially, for young scholars to test their findings on training grounds. This is more than likely that you do not have enough material, or that it's not original, but that's okay--this is a great opportunity for you to talk about your problem/field of study:

  1. By preparing a talk you learn how to rethink and summarize the issues you tackle, which has a potential of deepening your understanding.
  2. Feedback from your peers and superiors will help you improve your research--by pointing out flaws, suggesting new venues for exploration etc.
  3. Chances are that to many among the audience those issues are new and they can learn something--this is inevitable in all domains too broad to be mastered by one person, and math is indubitably one of those.

There is a great value in being able to define and conceptualize problems concisely and simply for a variety of reasons, and this is an opportunity for you to learn how to do so (incrementally, of course).

You must log in to answer this question.