The famous cancer researcher leaned against the podium, smiling as she fielded questions. For the last hour, she had thrilled us with a truly stellar keynote lecture containing a pleasing mix of historical context and cutting-edge results, and now the audience was showing their appreciation with a slough of thought-provoking queries. Always easygoing, the speaker now seemed entirely relaxed in the aftermath of a job well done. In that mood, it is probably no surprise that she was feeling expansive.
“Do you have any idea,” asked a man in the back row, inevitably, “of the mechanism of action?”
The audience was silent in anticipation. The cancer researcher paused, a conspiratorial twitch to her mouth as she seemed to weigh up options. And then, leaning forward slightly, she said, “Don’t tell my post-doc I told you, but we’re pretty sure it’s down to…” – and she then proceeded to expand a little bit on the biology, not giving everything away, but certainly leading us all in the right direction.
As we filed out of the room towards the reception drinks afterwards, those words kept ringing in my head: Don’t tell my post-doc. I couldn’t help picturing this person, giving him a backstory. On the young end of scale, I speculated, maybe coming up to the end of his fellowship funding. For the past two and a half years he’d slaved over his work, trying to tease out the secret that made his protein orchestrate such intriguing behavior in cells. He’d been to two meetings so far, but both times had decided that to reveal the fact that the switch was governed by that particular mode was too easy to follow up; he didn’t like keeping such interesting news under wraps, but he only had two papers from his PhD, so he needed a really big one to even hope to compete for a career-development fellowship next year.
And a few days before when his boss was preparing her talk, they’d talked it though again and had decided that she wouldn’t mention the mechanism – after all, the field was absolutely saturated (4000 new papers published a year, the cancer research had told us in her introduction) and many of the best and brightest would be in the audience. No, they’d wait until the paper was submitted before spilling the beans.
I imagined how the post-docs might feel when his labmates returned from the meeting with the news that his secret was now out. I’ve been there myself, so it wasn’t difficult: disappointment and anxiety in equal measure. Perhaps he might even have second thoughts about that Christmas break he’d decided to treat himself to.
Now this is obviously all speculative, my novelist’s brain on overtime: I have no idea if the data really were that crucial or whether there had been any sort of pact in this particular case. But I do know these sorts of mini-betrayals happen all the time. Lab heads – especially tenured, well-established ones – have everything to gain by openness, but pretty much nothing to lose. They can afford to see the benefits of sharing information and can participate in the often joyful process of unfettered communication. Most crucially, they can weather the odd scoop. But to the young first author behind the nascent data point, it is another story. One lost paper, these days, can be the catalyst that transforms an up-and-coming researchers into a failure.
I believe that it is very difficult for lab heads to remember this in the heat of the moment, when they are riding the wave of a stimulating meeting and secure in their immutable place in the scientific community. But until the system for rewards and promotions in biological sciences changes, there will always be a clash of priorities.
I have nothing sensible to add but I am sniggering at your tags. Sorry.
That’s ok – I’m glad I’m amusing someone. Especially as I skipped the breakfast hour to blog this!
Now that’s dedication. Thought: I hope you blog when they publish the paper. Essential narrative and all that.
I remember a particularly successful PI using data that was obtained the week before a talk, much the to the consternation of the postdoc who generated the data – once, without confirmation. This particular PI had a great sense for what was “real”, so inevitably he was right, but he could easily have been wrong, or given away the result to a hungry competitor (which once in a while came very close).
I am guilty of the same, but I think instead of a clash of priorities, it is a learning experience for the trainees that they can and should spill their precious beans to an audience. It is their duty as scientists, and they should do so enthusiastically, damn the torpedoes. I doubt that most divulged results have led to a massive scoop; of course it happens, but if one is at the point that the data is worth talking about, I doubt that too many other labs are right up there with the same result ready to roll. There are some fields where this might be the case (limb development in the 1990s, iPS cells now), but I think that the PIs job in those cases is to talk to the competition, figure out where everyone is at (as much as possible) and make arrangements for co-publication, for example. It is no accident that a recent issue of Nature had 5 papers on the same iPS cell result, or that the cover of Nature was shared years ago by two articles on the role of lunatic fringe in chick limb development. There will be some unfortunate situations, but I hate nothing more than going to a conference where a speaker only tells us about their published work; I could have stated home and reread the papers instead.
The speaker gave almost 30 minutes of unpublished data – and yes, I assume these were far more robust (and far
closer to publication). What I was questioning was the apparently cavalier breaking of what sounded like an agreement to wait between a lab head and a postdoc. There is a difference between data a few months away from submission, and those that need more time to finalize. What I object to is the idea that young researchers can be treated as pawns to be discarded in aid of an overall advance in knowledge.
I see your point now; but I seriously doubt that most PIs treat their trainees as pawns, and especially in discardable form. We value their contribution way too much, and in fact I view my job in very large part as developing their careers and making sure that they succeed. This is one of my top priorities, and I know that many others share this sense of duty, so sabotaging this would be the antithesis of this role. As you say, you don’t know if there really was an agreement or not; might have been a facetious may of putting it. The system for rewards and promotions at my institution includes my trainees’ success, and I firmly believe that for most this is far more important than giddily sharing a last-minute piece of data at a conference.
Hi Benoit – you sound like a great supervisor and I’m sure in most cases indiscretion does no harm. But I think unless you’ve had first-hand experience with its pitfalls, it’s difficult to appreciate how harmful it can be. Just because it’s rare doesn’t make it insignificant for the afflicted individuals. I think simply that pacts should be honored if the lab head thought It reasonable to establish them in the first place.
I should point out I’m all for presenting well established unpublished work, and indeed have revealed one of my new genes in a podium talk last Friday.
I can attest that Benoit is a great supervisor – my ex-student very much wanted to go to his lab, but he’s too popular! (And in a wonderful city.) However, she did pretty well for herself, so no hard feelings. 😉
Yes, I think Benoit’s comment that this may have been the PI making a humorous aside is quite a likely one.
While one does hear about PI’s who are ferocious sociopaths and drink the blood of their postdocs etc. etc., I suspect it is rare that postdocs go to work for such people in total ignorance. If the PI is a Big Hitter, and the likely reward is career-making papers in Nature/Science/Cell, then it is often more of a Faustian Pact.
PhD students are a bit different, as there is it much less “Faustian Pact” and much more “arranged marriage” – meet the prospective PI twice, with chaperone present, and then sign up for 3-5 yrs. As a postgrad tutor, one of my (unofficial) jobs was to steer students gently away from unsuitable prospective supervisors, or at least to try to ensure that students of less “robust” personality wouldn’t end up with PIs who came from the more Aspergerist /empathy-free end of the spectrum.
Would second what Benoit says about “judge people by how those they mentored did”. I haven’t done all that much in science, or been boss to too many people, but I count the success of those I did help train/mentor as probably the thing I take most pride in.
I love the Fauatian pact vs Arranged marriage analogy – brilliant, Austin.
I think I was a bit unclear about the ‘rewards’ issue. I wasn’t saying that PIs aren’t rewarded when their proteges do well – of course they should be and are. What I was saying was that there needs to be a better system for rewarding the young scientists who do great work but whose publication record doesn’t reflect this through no fault of their own. I also want to point out that I did say this was speculation in the case of yesterday’s keynote speaker – it may very well have been humorous – but I also want to make it clear that I have firsthand experience with a normal, non-blood-sucking, otherwise very supportive and congenial person causing me, personally, serious issues by indiscretions – saying things that we had firmly agreed not to, in the heat of the moment, to detrimental effect. So it’s not just some paranoid fantasy – trust me. It happens, and I think that’s a shame. There is two-way trust in the mentor-protege relationship and I think it’s important to treat each other with respect.
By the way, in a lovely inverse to this situation, in the morning session today, the speaker said – “I’m about to show you some data that really is too preliminary. In fact, I’m a bit nervous because I emailed my postdoc last night to ask ‘Is this still true?’ – and he hasn’t answered yet!”
Heh. Isn’t life lovely when it’s symmetrical!
I do like Faustian pact vs Arranged Marriage; however in the US we get to date first: the graduate student spend three 4-10 week rotations in various labs, to figure out where they’d like to go, and conversely if they’re a good fit for the host lab. I used to think this was a waste of time when I was a student, but it is a very enriching experience, and makes for much better graduate school experiences!
Your last comment, Jennifer, is precisely what I was alluding to; the rushing to conclusions, and that could certainly frighten a student or postdoc, especially if they’re not convinced themselves that the result is fit for human consumption.
Regarding the rewards for junior scientists who have not published “through no fault of their own”…. that’s perhaps a different conversation, but sadly the system does not work well in this regard. It is very difficult as a mentor to convince a search committee to hire someone with no paper, or not that stellar one or two. It is possible however, and I have seen it happen, and even experienced it myself. Again, one of our jobs is to ensure the success of our trainees, and boosting them and promoting them at conferences and in reference letters is the best we can do, and it sometimes does the job.
But I’d like to hear what you mean by “through no fault of their own”….
I meant being scooped because their boss just couldn’t keep their mouth shut. But more broadly, I was thinking of a more utopian universe in which we are all judged by more factors than just how many papers we publish.
Heartily second your advocacy of PhD rotations. Definitely saved me from a bad decision!
in which we are all judged by more factors than just how many papers we publish.
must resist…
Agreed on the desirability of PhD rotations / 4 yr programmes (as they have to be to do this in the UK set-up). It would certainly have limited the kind of, er, disputes one sometimes used to have to mediate as a postgrad tutor..! Of course, it does cost more money per grad student trained than the “3 yrs only arranged marriage” model.
Austin, I reckon they get their money’s worth because a good student will start producing papers in years 3 and 4 and this will probably outstrip any sort of output you’d see with 3. Also, rotating students are often given discrete little projects that often lead to authorships – e.g. cloning a construct or testing something. So for the student it’s also beneficial.
I like the thought of rotating students. Don’t they get dizzy?
Jennifer, agree completely. We (the Postgrad Ctte’s in Univs, and other academics) all believed exactly what you say. It was widely perceived as being the funders like the Research Councils (and more accurately the funders’ paymasters in Govt) who would dispute the added (apparent) cost of switching to 4 yr programmes. Another way to put this was “well, you could have that (all 4 yr programmes) but there would be a third less studentships” – which again was widely seen as politically unacceptable to both Govt and Univs.
It is perhaps significant that in the UK the Wellcome Trust started funding 4 yr programmes, and then the Research Councils (to some extent) followed after. Again, the widespread “trade” perception was that the Wellcome always led (or effectively shamed, if you prefer) the Research Councils into changing the way they did things on PhD studentships. The most obvious example of this was student stipends in the early 90s, see various discns on PhD student funding in your post about Brooke Magnanti. But I reckon the same is true with the change to running more 4-yr studentships/programmes.
So did the RC stipends finally catch up with the Wellcome?
We MRC brats (the best of the RCs for stipends) couldn’t afford stones to throw at the posh Wellcome kids.
I’d like to see more care and feeding of students – but coming back to my original point, it would be great if talent could be measured by more than just published papers. I am trying to imagine what science would be like if we didn’t have to worry about publications at all – it’s a pretty astounding vision. My old friend John Cairns, who was doing molecular biology science post-war, has amazing stories of how people didn’t necessarily have to publish – they just circulated manuscripts for comments – and that’s also how you got your next job. People could see you were good from your output, even if there were simultaneous discoveries.
Yeah. That does make a lot of sense. People know who’s good and who’s not in their field. All this crap about Eigenfactors and h-indices and whatnot all come out with the same answer anyway. If you’re good, people will know.
Talking of “mol biol post war”, Francis Crick is often taken as the classic example of someone who was highly rated by everybody round the Cavendish Lab but hadn’t published anything after several years in the lab as a graduate scientist. You can find this view in Jim Watson’s The Double Helix.
Not sure if it is the same, but even somewhat later (e.g. in the 60s) there was no drastic rush to publish unless you had something worth publishing. When Ada Yonath won the Nobel this year, my father told me “guess what, my first ever postdoc just won the Nobel”. When I looked up Pubmed, I saw he and Yonath had never published anything together. He said “Oh yes, we spent a year doing some small angle x-ray stuff (in Pittsburgh in the late 60s), but it was never enough to publish”.
Still later, I seem to recall that at the time of the landmark 1991 olfactory receptors paper that led to the 2004 Nobel. co-author Linda Buck was a postdoc in her mid-30s who hadn’t published anything for several years.
I guess what this last example says, and the set-up today, is that if years of publication silence are followed by a “big discovery/paper”, you will be OK. But as Jenny says, if the big thing doesn’t work out, or you get scooped, it will not pan out nearly so well. So it is a very, very high-risk strategy.
I have the impression that in some of the fancier US labs people will still “gamble” more in this way – on long gaps while they get all the ducks in a row for the big paper. But in the UK it is almost impossible, as the funding for the project would have dried up.
It is widely believed in UK Univs that one of the key reasons that the UK bioscience Nobels virtually all came out of the Cavendish/LMB was that the people there were not having to re-apply for their salaries and/or research funding every 2-3 years. John Sulston in his autobiography noted that that he didn’t have to wrote a grant application until he was 40+.
Austin, great examples all. (Note to self – better quickly repair that gaping hole in my publication record since my last first-author paper, before the universe implodes.)
It’s a real shame how things have changed. But it feels like a huge, mindless machine – how can we ever change a system that is so entrenched?
how can we ever change a system that is so entrenched?
That’s the crux of the problem isn’t it? There seems to be almost universal recognition that there’s a problem, but I’m yet to hear even a reasonable attempt at a solution. Part of the problem may be that those with the most influence or power to make changes have the least motivation to do so?
A related problem is that there are, pretty well everyone agrees, more people that want permanent jobs in science – especially as PIs, since that is where “permanence” largely resides – than there are such jobs.
So – how to decide who gets the jobs?
The system of points for papers and grants, which is what largely operates now, is imperfect – but the question is, how you would do it otherwise? As search commitees and appointments panels get bigger (a trend over my career), and decisions subject to scrutiny (“what objective criteria did you pick this person on?”), there is inevitably a trend to act largely upon what is deemed to be quantifiable. This means grants got, and papers published, plus how “good” the papers are deemed to be – which often defaults to “what journal they were in “.
In my (admittedly limited) experience, appointments panels and their members are wary of sticking their neck out and saying “this person is much better than their purely spread-sheet based track record says”. And the bigger the panel, the more risk-averse it tends to become.
When people defend this status quo, the argument often goes roughly as follows. In all the various spread-sheet / output measure / citation count / metric-based “grading” exercises, one often hears that “different panels of assessors reach very similar verdicts on the same assessed (people)”. What that tells you is that everyone rating people based on spreadsheets works to broadly the same criteria. Hence the UK Govt’s enthusiasm for metrics.
The problem is that most people also agree that the spreadsheets do not always give you an accurate assessment of a person they represent as a scientist. But there is typically no strategy to deal with this, other than the hope that, if it is a job selection process, the interviewees get a chance to say something and hopefully flesh out the spreadsheet picture. But then, of course, the risk is that you end up scoring the person’s personality, rather than their scientific chops.
Anyway, most people in the system, while recognising it as imperfect, just shrug and say “what can you do”. Or, a bit more sympathetically:
sigh
Looking over the above, I see I haven’t really said anything more in that mini-essay than Darren did in four lines. Oh well.
Well, it was good to see the logic laid out. The real solution might be population control – much as the Chinese can have only one child, perhaps PIs should only be allowed to have one PhD student every four years. That might take care of the postdoc glut…
It works in the medical system (at least in Australia). Numbers of positions in the various specialty training programs are tightly controlled by the professional bodies (colleges); it’s essentially a cartel. Not the best model for ensuring a good supply of doctors, or providing any form of price control – as supply is deliberately restricted, but pretty much assures everyone gets a job if they compete their higher training
I guess the argument against that is that many PhDs go on to have other interesting non-research jobs that require a PhD. And there’s the obvious economic argument – our research machine requires loads of disposable parts to keep churning out the results. Students and postdocs, even ultimately dead-end ones, are essentially cheap labor and to abolish them would severely reduce output. In the States, grad students also do a large part of the teaching of undergrads, so they’d struggle without their sweatshop teaching factory too.
our research machine requires loads of disposable parts to keep churning out the results
That’s kind of depressing. True, but depressing.
I guess I am one of those dead-end post docs. I had my name on 6 publications this year, but they were in the wrong journals of course. “Quality” is king. I’ve helped out a lot of people, neglected my own work, and am presented with a “bill”.
The information flow towards the groupleader is a difficult process for a dependent researcher. The supervision here is relatively non-invasive so you have the option to not show the latest data, an option which is sometimes used.
I also have the feeling that some datasets are pushed too soon, but not mine of course. I decided to only do what interests me in what could be my last year in science. And my interest doesn’t really equal high impact factor journal material. So I am safe.
@Darren: A similar system exists in the Netherlands, only in the medical community. Each year, the amounts of medical students, awarded MDs and specialties are decided upon beforehand. The result is that everybody who goes into medical school will end up with a job. But, and that always is the downside, only 25% of the people who want to (*and* who have the grades to be able to) actually get into medical school.
Effectively, it means that selection takes place at a different age. The upside is that these students, at age 18 or 19, still have ample opportunity to pursue other, very worthwhile careers.
@Jennifer & Richard: The system uses a lot of disposable parts at the moment. As with everything easily disposable, it will not be sustainable. What goes for ecosystems goes for organisations as well. Most first-world universities need to scavenge people from the rest of the world to work as PhD students (read: single-use-employee).
I am curious, Bart – do you really think it’s not sustainable? Is that because the non-first-world parts will dry up as countries develop and no longer need to go to America or elsewhere to do great science? Or do you think there might be some sort of (dare I say it) Revolution?
I don’t know…I think the system was chewing up and spitting out lots of indigenous students and postdocs long before so many foreign people were recruited. But that’s a bit before my time so I can’t be sure.
Mark, I’m sorry to hear about your plight but am glad to hear that you are enjoying yourself nonetheless. Have you made any alternative plans yet?
Austin: I have the impression that in some of the fancier US labs people will still “gamble” more in this way – on long gaps while they get all the ducks in a row for the big paper
I would agree on this, that some labs have the money to “wait” and assemble the data under a longer time point and make a bigger publication. I also think that maybe the bigger labs can put more people on one crucial part of the experiments and make it quicker than the smaller ones… but most of all, I wonder if it isn’t a bit more like Richard G wrote on his blog about “bigger fish knowing when smaller ones publish one part of the big picture?”
Jenny: I am not sure on the population control… since not all post docs want to go on to Academia, although I guess it might be better if it wasn’t 1/10 going on to his/her own lab but rather 3/5? And yes, there is a lot of teaching by grad students and post docs. I wonder if it is mainly because the professors have to write grant applications all year round instead of focusing on research and teaching? I have to say as an undergraduate, I really liked those classes where the professor taught classes since they’ve been along for a longer time and could tell more of the overall picture and some more history of the subject….. ahh… the wonderful times of undergrad 🙂
Grad student lecturers are like Russian roulette. Some are perfectly decent; others can barely speak English. I’m not convinced it’s a great system but it’s very well established in the States. I don’t think it happens here in the UK though, so there must be other ways of sustaining it.
I can attest that Benoit is a great supervisor – my ex-student very much wanted to go to his lab, but he’s too popular! (And in a wonderful city.)
At one time, he was in another pretty great city, you know… 😉
Grad student lecturers are like Russian roulette. Some are perfectly decent; others can barely speak English.
I’m thinking hard, but the analogy is eluding me. Unless I’m playing Russian Roulette incorrectly… 😛
BANG
no, you’re good.
I admit that simile wasn’t my best work. My point is that teaching is a skill and being a graduate student doesn’t automatically mean that you are any good at it. Many grad students teach in their first year when they are overwhelmed already.
The general “formula” for grad student teaching in the UK science Depts I have worked in has been for them to start off as lab demonstrators – not in charge of classes but helping out a more senior person (usually a junior to mid-level academic) who runs the class. Usually we only let people start this in the second half of their first yr as a PhD student.
I think grad students in the UK teach less now then when I started, as UK Univs are now hyper-sensitive about perceived “teaching quality”. As a first year PhD student at UCL in the mid 80s I ran lab classes solo, did tutorials (not terribly well) and even did a lecture… but like I say, nowadays they would only be “assisting” in labs, and some places don’t even let students do that until PhD yr 2. The pay is much better now, though – the current rate where I work is £ 14.50 / hr for lab class demonstrating, and marking (if required) is paid separately on top of that.
@Jennifer: A large part of this is intuitive (thus: speculation). I do notice that more postdocs and phd-students (in those countries where they actually are employees) are forming unions. However, I doubt that this will lead to revolution.
In other fields going on to get a PhD is only reserved for those who apparently are not enough to make it in the real world (law, business) and it would be a shame if the “disposable parts-model” would result in a similar dynamic in science. If the best students decide that a PhD is not worth the slave labour it requires, science will suffer.
The higher education systems which offer opportunities and security to/respect/acknowledge/support their juniors more would then eventually end on top. The ones who do not, require cheap labour from elsewhere. But China’s, India’s, Pakistan’s, Malaysia’s and Indonesia’s universities are getting a lot better really quickly and will start to compete for their own people more.