Suggestions for giving talks

We all have to give talks as part of our work from time to time.  In fact giving talks is  very important. Let’s face the truth, people won’t read your papers.  People working on very similar and related things might. Placing a preprint on the arXiv helps,  but generically people won’t get past the title and if you are lucky the abstract. Giving talks at conferences and seminars is the only way to get people to notice your work and importantly YOU.

My general advice (and I have no great insight) is simple,

  1. Select very modest goals and maybe 3 or 4 key points.  Most of the audience will not be experts in your precise field, unless you are at a specialist conference.  For departmental seminars you will have to take great care in what you say and how you say it. You want people to learn something and also see that you are great at communicating.
  2. Open with setting the context of your work.  People need to know how it all fits in. They want to know if what you have done relates to their work.
  3. Be slack with unimportant details.  You can remove terms in equations that play no significant role. Just use words like “+ small terms”.  You can suppress indices.  Don’t be frightened of using analogies or examples to get a point across.  The ethos does  have to be exactly correct,  just close enough to what you are really doing, but you must say this.
  4. Never run overtime. The audience will hate you for this. It is disrespectful to keep going past your allotted time.  It can also show that you do not take talking seriously and have not though out your talk at all.

Geroch way back in 1973 wrote some notes called “Suggestions For Giving Talks”.  He gives plenty of sound advice for giving talks. The notes are available on the arXiv  here.  I suggest anyone giving a scientific talk take a look at it.  I know it has helped me.


Other things online I have found useful and full of good advice include

There is much advice online, so have a quick ” google”.

To improve you talking skills one should do do two things

  1. Give as many talks as possible.
  2. Attend as many talks as possible.

Only with practice will you improve.  You can also take inspiration from good speakers and avoid imitating the bad ones. We have all been to talks by very well respected people, only to be disappointed by their presentation.  Learn from the masters, both good and bad.


Finally, if you have to give a talk soon, good luck.


Leave a Reply

Your email address will not be published. Required fields are marked *