Scientist skeptics, please help innovation instead of stifling it.

Lenny Teytelman Jun 17, 2015

An integral part of becoming a scientist is learning to be skeptical. We often do not believe our own results and we hardly ever believe the published papers of others. If I'm not mistaken, there is a verb in Hebrew which means "to journal club" - that is to shred someone's research presentation. Skepticism is healthy, but we often forget that not everything proposed by someone else is stupid and impossible. We carry over the distrust of lab meetings, qual exams, papers and grant proposals into everything else that comes our way.

As a result, an all-too-common response to attempts to innovate in science communication is "No one is doing/using this now, so it won't work" or "It was tried before and failed because it can't work." As scientists, we should know better.

1. No one is doing it, so no one will.

This dissmissal is by far the most common. Literally everywhere.

  • Preprint can't work in biology because no one in my field has published a single paper on bioRxiv.
  • Post-publication peer review won't work because most scientists are not commenting.
  • Scientists won't use mobile in the lab because I have never seen anyone do so in my lab.

What we forget is that new efforts take time. Three years ago, preprint in biology was virtually unheard of. The explanation for many was "Yes, arXiv works for physics but biomedical is different and it's impossible here." No, that's just because there was no good infrustructure for biologists. Now we have bioRxiv and it's gaining momentum beautifully. So those same folks are saying today, "Yes, but it's only working for genomics and no one in neuroscience is submitting there. It can't work for neuroscience." That's incorrect. Three years ago no one in genomics was using bioRxiv either. It takes time.

Identical story with post-publication discussion and review. Dismissed by many because "scientists don't have time and are afraid of discussing openly, hence no one is doing it today." No, that's because there was no PubPeer and PubMed Commons. And I heard "scientists won't use mobile in the lab" so often, I had to write a rebuttal (by the way, thousands are now using our mobile apps in lab on daily basis).

2. Tried before, failed, so can't work in your case.

This one is rarer but more damaging objection. 

  • Nature tried preprints in 2007, failed, shut it down in 2012. That proves CSHL's bioRxiv can't work.
  • PLOS tried comments, no uptake, so it won't work on PubPeer either.
  • tried to create a central repository in 2000. Failed. OpenWetWare lost traction. Clearly, no way can succeed.

Correlation is not causation. Yes, NaturePrecedings effort failed. But did it fail to get submissions or fail to make money for Nature? Did it fail because biologists won't do preprint or because Nature was the wrong organization to do it? Similarly for post-publication discussion, it's true PLOS comments never really took off. But maybe that's because scientists don't want to comment on a thousand different journal websites? Maybe it was the implementation and user interface?

For our, I know what we are doing is crazy and hard. I know many tried to do it and failed. But did they fail because it's impossible or because they were too early? Chris Yoo, co-founder of BioProtocol is on our advisor board. I have talked to the cofounder of OpenWetWare extensively. The failed efforts of yesterday aren't proof of impossibility - they are lessons for getting it right in the future.


Innovation is hard and requires resources, passion, time and luck. Most efforts to innovate will indeed fail for a host of different reasons, but few will fail because what they are trying to do is impossible. And when you come across good ideas like preprints, don't fall into the skeptic trap. Instead of the natural and pleasant exercise of tweeting reasons it will fail, think of how you can help it succeed. Do the journals in your field not accept preprints? Write to the editorial board and ask them to change the policy!

If too many let the skeptic inside win, changes that we need and crave will take longer and some of the innovative efforts will fail due to lack of support. PLOS One had the hardest time gaining traction and reaching profitability; PLOS nearly went bankrupt. CSHL built bioRxiv and we created, and if you submit your preprints and share your protocols, you infinitely increase the odds of these efforts succeeding.