Wednesday, May 05, 2010


Not long ago I listened to a Famous Scientist give the exact same talk that I had heard him give at a conference a few years ago. The talk was not an invited review in which it would be reasonable to expect a summarizing of previous work. He was speaking in a session in which it was reasonable to expect to hear results of new research, not a rerun.

When is giving the same talk OK and when is it not OK?

From what I've seen, a common scenario for talk repetition is when the talk is first given at a small conference, then repeated at a large conference, especially if the two conferences are not separated by a substantial amount of time (delta t < 1 year?). If you happen to have been at the small conference, you see the talk twice, but most of the people in the audience at the big conference are hearing the talk for the first time. I suppose that is acceptable, although if the speaker anticipates quite a bit of overlap in audience (i.e. most people from the small conference are also at the big conference), it would be good to make the second talk a bit different.

What is less acceptable is to give the exact same talk in different years at the same annual conference. In the most recent example of this that I saw/heard, the Famous Scientist was also the former professor of the conference session convener. So I suppose the decision to let him talk about the same old stuff was based on sentiment and/or on the desire to have a Famous Scientist speak in a session.

If, however, the session convener had reason to know in advance that Famous Scientist would present the same old stuff, would it be so bad to give him a poster and give a talk slot to some up-and-coming scientist with new results? I am sure not every single person in the audience had seen the talk before, but does that justify a rerun talk of research published a few years ago; research to which nothing new had been added?

If there are plenty of talk slots to go around, maybe; if talk slots are in short supply and highly desirable, I'd rather hear something new, even is the speaker is a not-famous professor, postdoc, or student. I would be less harsh on Famous Scientist (who is actually a colleague I admire and like quite a lot) if he had been asked to give a review talk or some kind of overview of his work or a particular topic. It bothered me that he gave a total rerun talk as a regular research talk.

It can be good to have Famous Scientists in a session if it attracts an audience for the less Famous Scientists (as long as there isn't a mass exodus after the FS talk), and of course some (many?) Famous Scientists have interesting new things to say. And some people may want to hear the old classic work in a talk. I suppose I got annoyed in this particular case because there were some interesting early-career scientists who did not get to give talks at that conference.

Sometimes I give similar talks at conferences that are not too far apart in time. Talk 1 will focus on a research project at whatever stage it is at during Conference 1, and Talk 2 will be an update, presenting the newest results of that project. Ideally, there will be substantial new results, but some updates are more significant than others, depending on how the research progresses. In that case, even those of us who intended to give different talks might find ourselves giving talks that are more similar than we wanted.

Some of that is to be expected when you are working on a long-term project in which amazing data and discoveries are not ideally spaced relative to conferences, but typically you can say something new about what you've done and where you're going with the work.

Or at the very least you can change the background designs of your slides.

What do you think? When is it OK to give the same talk at different conferences with similar audiences and when is this proof that you are a slacker-scholar who just wants to justify a desire to jet around to different conferences?

If you have nothing new to say and are invited to give a talk, do you agree to give the talk (assuming that people will be fascinated to hear about your old, published work) or do you decline because you don't have anything new to say?


Anonymous said...

when is this proof that you are a slacker-scholar who just wants to justify a desire to jet around to different conferences?

Here is an example of such proof that I encountered last week at a conference: Famous Scientist (full professor with >> 6 figure salary) comes up to the session chair 10 minutes before his talk, who is busy chatting with a bunch of us grad students and postdocs who are all paying out of our own pockets to attend, and interrupts the conversion to complain to the session chair (right in front of us) that the hotel front desk asked for his credit card (the front desk staff not realizing that the keynote speakers have their room paid directly by the conference). 10 minutes later he gives his recycled keynote speech, gets off the stage, and goes to collect his hotel reimbursement as skillfully as an anti-welfare Republican collects a social security check.

queenrandom said...

I can't say at this point in my career (late stage grad student) I'm giving a lot of talks, so take what I say with a grain of salt :)

I like to think of giving talks (or lectures, for that matter) as not about talking about your research (or subject matter), but talking *to* your audience. It's a subtle difference, but it puts the focus on the audience rather then myself (which also helps with my anxiety :P) I therefore like to tailor my talks to the audience based on say, the majority department, conference or session topic, etc. So, even though I recently gave 2 talks within a month and a half (so the vast majority of the data was the same), the focus and structure were different, in order to capture the attention of the different audiences.

These tactics may change as I become a busier person, but I'd like to think I will continue so that I can keep both my audiences and myself from getting bored.

Kevin said...

I only go to about 1 to 1.5 conferences a year, and I never recycle talks at conferences. I will recycle talks when invited to speak at a university research seminar, though I do try to update talks every time I give them.

Conference hopping with the same talk is a waste of everyone's time and money.

Venkat said...

What does he say at the end of the talk? "The grad student who did this work is now a full professor"?

mixlamalice said...

In my field, it seems that Famous Scientists, especially seniors, tend more than others to give always the same talk, which is usually a long review of what they did in their carreer. With a focus on when they were at their prime, 20-30 years ago, even when the conference is some kind of workshop where you are supposed to talk only about ongoing/unpublished research.

Maybe they have the kind of feeling an old actor has: they still want to run the show and they fear that each time might be the last. And they want to be remembered, so they give their own eulogy.

It is usually nice when you hear them for the first time. Then it quickly becomes boring. And, after a while, sad.

Anonymous said...

I'm still just a grad student, but I gave ~8 talks/posters in the last year. My advisor probably gave 3x that number.

I think it would be a waste of time to write up a new talk for each one of those.

Sometimes you go to the conference not just to spew off about the same old crap, but also to learn new things from others. It's of course a balancing act, but I don't think it necessarily has to be a waste of time and money.

chemprof said...

At the conferences I go to (5-8/year) it seems to be very much related to age. The older (and perhaps more 'famous' you are) to more similar the talk will be, and often it will be exactly the same. Everyone knows it, most complain about it, but the conference organisers seem to still invite the same speakers. The ones who do present new stuff are well known for it, and for all the right reasons too.
As a relatively young (35) associate professor, I can honestly say that I have never given the same talk twice. Similar ones at times, but always with some new data, or with some new twist to suit the audience. This year I decided to not speak about the same work at more than one meeting. It's going well so far, and keeping me more excited about presentations than usual. As well as getting a few people very interested in how much different work is going on in my group.

Anonymous said...

I got a copy of my department head's CV when applying for a grant last year, and found he had given 29 invited talks at conferences/workshops over the previous 12 months, over five continents, with only two given in his country of residence. That's a lot of jet-setting on someone else's money.

I haven't heard him speak in a really long time (I would never bother to go to his talk if we attended the same conference) but I very much doubt every talk was different. First, because I don't believe anyone could generate an entirely new talk every two weeks, and second because if he travels that much, how much time does he have for research, or even with talking to the people doing his research back home, to find out what they are actually working on? It would not surprise me if he gave the same talk, word for word, for months at a time.

Personally I always try to present new results. With invitations to speak often coming 6-12 months ahead one always hopes/expects to have new results, and it doesn't always work as planned...but I would always at least try to emphasize what little is new.

Anonymous said...

I'm really impressed by the grad student who managed to attend 8 conferences last year (Anon 5:56) grad student attended 4 conferences last year plus a workshop, and I'm always amazed she managed to travel that much and still take courses, teach, and do enough research to fill those talks...but she did. Maybe every field is different but if each conference takes about a week, plus at least several days to prepare a talk/poster (more for younger grad students), write abstracts and proceedings manuscripts (another week +), make travel arrangements etc, I think that is seriously productive.

Anonymous said...

My view may be a bit different as I am a mid-stage grad student. At least in my field and geographic location, most grad student do not have the opportunity to go to many conferences that are likely to attract BigName scientists ( I am averaging 1 per year). Therefore, the chance that I have seen BigNameX speak before is extremely slim and I get a thrill out of hearing a BigName talk, regardless of whether their talk is a review of the last 10 years (I actually prefer these sorts of talks sometimes, they can be quite interesting in terms of learning about an academic career).

Maybe I am still at the groupie stage, I can certainly see how this would get annoying if heard multiple times.

mixlamalice said...

To be honest, I have to admit that when I went for a cycle of seminars/interviews in 5 departments in 10 days, I gave the exact same talk... (except for a couple of rearrangements about things that were not that clear, typos etc).

That said, I am a post-doc so I can talk about: my PhD work, my research topic 1 as a post-doc and my research topic 2 as a post-doc. My PhD work was quite broad so I can focus more on one side or the other, but all in all, my options remain limited. Anyway, I don't give that many talks, so this is not a really big deal, people have time to forget they already heard part of the story a couple of years before.

Dave Backus said...

Who tells FS he/she has lost it? Not a job anyone wants.

I think the bigger issue here is who runs the conference. In my field (economics) some of the best conferences are run by young people. They rarely make this mistake.

PLW said...

In econ, it is quite common to give the same talk at multiple occasions. If you present at two conferences a few months apart the overlap of your audience is likely to be less than 5%. Maybe this is because our fields are broader or our conferences are bigger or something? When I'm attending I often have a hard time deciding which sessions to attend, as there might be two (or even three) that interest me all in the same slot.

As for invited talks, the audiences are even less likely to overlap, and you tend to only be working on a couple things at once, so it would be very difficult not to repeat. Also, I tend to like to get a lot of feedback from different quarters when I'm working on something, so I want to present it to multiple audiences in different places.

Presenting published work happens, but rarely. Generally, what happens is that you submit something to a conference around the same time as you've submitted it to a journal. Since you never know if it will be accepted and if it's not the feedback from the conferences (and the exposure) might be useful for a resubmission. Every now and then, the paper will hit before the conference. All of this is magnified by the LOOONG publication revision cycle in economics... 1 year minimum, often more.

John Vidale said...

I only mind recycled talks at small conferences in which there is no poster session or other talks to flee.

Talks in meetings can also be just review from already seeing the work in proposals, recently published papers, manuscripts to review, or talking with the speakers.

Seeing a good talk the second time is better than being trapped at a talk or poster that has little interest.

anon @ 12:24

It'd be nice to have a >> 6 figure salary, even if it were only in the low 7 figures. 8 or 9 figures would be grand.

Anonymous said...

mixlamalice: right on, right on. I have encountered exactly this progression with a particularly prominent Famous Scientist in my field. I guess the question for me is, is there anything you can say to these people (assuming you're not a complete stranger, and that you do it in a tactful way) to shake them out of their funereal march??

Anonymous said...

I tend to follow FSP's model - if the talks are in the same year, there will be overlap but also updates.. and like queenrandom, I try to tailor background/which studies I talk about to the audience.

I have seen one SUPER FAMOUS scientist talk 3 times - twice as an invited seminar speaker, and once as a keynote at a major conference. Slide-for-slide it was a rehash of the evolution of his career since the 70's (I'm in biology, that would be the equivalent of the dark ages), with the last ~5 slides devoted to new stuff (which did at least change). The annoying thing is this guy's lab has >30 people at any time, so there should be no need for any recycling.. and many of them are future professors, so seeing the talk from the big guy might help them (of course, just having authorships with him helps them a ton). I now know to not bother showing up to his seminars. In contrast, I just saw other FAMOUS SCIENTIST come in for a seminar, and while some of it he may repeat from talk-to-talk within the year, it was all at least less than 5 years old. And, he made a point to repeatedly say the names of the grad students/post docs who did the work.

Thinkerbell said...

I wish I could pipet fast enough to completely re-vamp my talk every few months. However, in my area progress is sloooow and I am just one person by myself. I think it's a different story when you are a PI with multiple projects in the lab -then there's a beter chance of having new data every time. Even then, I am always pleasantly surprised by FamousScientist who always has at least 10-15 minutes new data every time I her him, no matter how close together the talks are in time. In some fields the fear of being scooped I think prevents people from presenting fresh data. Something I really dislike, but a reality nonetheless.

Anonymous said...

One interesting thing are talks by nobel prize awardees (NPAs). The Medicine or Biochemistry awards are often given to fairly young scientists nowadays, and I've had the priviledge now of seeing three different NPAs give talks.

It seems like the expectation is that the NPA should talk about the work that won them the prize, not about what they work on now. I think it might actually be a requirement of the award?

However, I know several of the NPAs I've seen still have active research groups - so when do they get to talk about the new stuff going on in their lab? Seems there might be a downside of winning the Nobel after all - you have to talk about work you did 10 years ago for the rest of your career...

John Vidale said...

I'm not sure placing such an emphasis on giving good talks is the best approach. As scientists, higher priorities are doing good research and writing good papers. People with big groups are also hard pressed to spend enough time on collaborative research with their students and post-docs.

Further, some people naturally lay out material in an intuitive way that aids understanding and speak well, but many scientists don't.

I'd argue for leaving the senior scientists in those commemorative and overview sessions (that I tend to avoid) and sending the people in the trenches doing the research to talk in the hot sessions.

One should work to be understood, but repeated talks by people past their research prime are more a curiosity than a hazard.

anon @ 7:57

In my experience, more senior organizers do better than younger ones from (1) having heard many of the potential speakers multiple times, (2) not being so intimidated by the big names, (3) knowing better what's considered hot, (4) having the coercive power to get who they want. They do tend to sometimes get ossified, however.

Alex said...

There's a research group in my subfield that pioneered a certain technique. Any time a student or postdoc from that group talks, the talk goes something like this:

1) Here's how our group first solved a major problem.
2) Here are several things that our group has since done to improve on that solution.
3) Finally, here is my own specific contribution.

I know that not everybody in the audience is equally well-versed in the work of this group, so maybe point 1 does need some emphasis. However, point 2 is more like an ad for the group than for the speaker. I want more of 3, because some day that student or postdoc will be job hunting, and the rest of the community needs to know what this person has done.

Materialist said...

Much like a fancy-school degree, I can see how a conference session organizer would want to make sure there are some FS on the schedule to give it some gravitas, even if it's known that they'll be presenting "Old Reliable."
Although the overlapped audience members can be annoyed, they can also arrange ahead of time to miss that talk, leaving more room for the grad students who want to see a FS talk.

Anonymous said...

Personally I fall in the recycle but adapt crowd. For instance last year I started with presenting a talk on new results at my regional meeting then during the year presented versions of that talk at a couple of other bigger conferences (adapting for audience and improving based on feedback). When the time rolled around for the next regional meeting I started off another cycle with a completely different talk.

Having said this, from the audience end how I feel about recycled talks varies. I've got a lot out of seeing the same talk twice particularly if it is complicated work. On the other hand I have seen the same result presented 6 times last year in multiple versions by my supervisor and his collaborator. I could almost give that talk now!

Madscientistgirl said...

I am a post doc. If I am invited to give a talk, I gratefully accept and I give the best talk I can, even if I don't think I'm the best person for the job (unless it's wholly inappropriate because I had nothing to do with the work and someone else who did can give the talk.) I need all the help I can get staying employed. I can't afford to turn down any invited talk. (OK, I did turn down an invited talk in a backwater town in Belarus the week after I moved 800 miles to a new position, and that turned out to be an excellent decision.) I do try to make each talk different and at least try to find a new way of presenting the data and I try to make the talk appropriate for the venue. But I also think it's very different for established scientists in a secure job to turn down an invited talk.

That said, when organizing conferences and workshops, it is sometimes nice to know roughly what someone will talk about - with a new twist, some new results, but enough of a clue to plan a conference. Sometimes what you want is actually more of the same. Some conferences are devoted to the hottest, newest stuff in the field, but not all conferences and meetings have this purpose. I've organized regional meetings where the purpose is to expose younger students to material more senior people know already and meetings where we are really trying to show off to the funding agencies how cool our field is. Sometimes the senior professors in the field are not the part of the audience the conference is geared towards.

And whenever a senior person gets up and gives a talk about his or her students' and postdocs' work, I always wonder why the senior person is giving the talk instead of the junior person. I understand overview talks and talks giving some perspective on the field being given by senior people. I also understand risky talks or (high profile) talks where the student may be a bit young yet to answer questions well being given by senior people. But I'm always annoyed when senior (tenured) people give talks that should be given by their students and post docs.

Unbalanced Reaction said...

The first commenter (Anon 12:24) made me snort my pinot, shooting it all over my monitor, causing me to forget my (surely) witty original comment.

The Boss used to recycle talks All. The Time. That is, until someone older *totally* called T.B. on it. During the Q&A session. Too bad I wasn't there.

EuropeanFemaleScienceProfessor said...

My research is in an area that is not generally to be found at most universities, although there are often people who vaguely dabble at it. I am often invited to give a talk at various universities around Europe on the topic.

I generally start off with the same basic set of slides, but will tailor some slides to the place, or will extend or cut the talk to fit what it was that the inviters really wanted to hear.

The downside of this is that I remember that I made this really awesome slide one afternoon and I don't remember which of my gazillion slides it was in and spend lots of time looking for it again when preparing a talk for a similar group ;)

AnonEngineeringProf said...

Recycling talks is fine when the overlap among audiences is likely to be minimal. (e.g., talks to different university departments) But when a significant fraction of the talk is likely to have seen the talk before, recycling a talk is just lame. For instance, giving the same talk at two major conferences is lame.

That said, don't just blame the famous scientist who recycles his/her talk. Blame the conference organizer who invited the speaker who was a dud. It's their job to find people who will give fresh, engaging talks. If they're inviting people based upon "the big name" rather than based upon quality and novelty of the content of the talk, then I'd argue they're not doing their job.

Anonymous said...

In this day and age of web/video conferencing, downloadable podcasts, etc., we should reduce the number of jet setting faculty. Especially those that like to give the same talk in multiple venues. In this day and age with modern technology, talks can be given in real time via web conferencing, including Q&A from the audience without leaving your office - after all even when you are physically on stage giving your talk it is still your powerpoint slides that are the focus of your talk, not your physical body. (or do you do a song and dance too?) I do think that faculty use talks - especially recycled talks - as an excuse to jet set on someone else's money. Let's take away that excuse so they will spend the money more productively (like on their students or for research) rather than to finance their own mini vacations. We should make it a rule that whenever a talk is recycled, the faculty has to finance the trip out of their own pocket.