Difference between revisions of "Rise of the 140 character paper"
m (1 revision imported)
Latest revision as of 12:28, 4 July 2018
By Ilya, 11:03, 23 June 2015 (EDT).
Lately I've been thinking a lot about how science changed in the 25 years I've been doing it (of which 20 have been here in the U.S.) We have heard a lot about positive things, like the arXiv or Open Access journals, which revolutionized dissemination. We've heard a lot about negative things as well, such as the grant cycle that is getting more and more vicious with every passing year, dearth of academic positions, politicians getting into deciding which is the "right" science. Personally, of all of these, I am the most scared about the general change of public attitude to science: one (generally) wouldn't want surgeries done or airplanes designed by laymen. And yet it seems totally OK in the current discourse to say "I am not a scientist, but I believe that the scientists are wrong."
In any case, this post is not about scientists and the society, but rather about scientists and scientists.
I look back at the first papers I wrote during my PhD work and soon after. These were deep papers, which took on a subject and finished it. There was little left in that corner for the others to explore. My opus magnum, a major part of my PhD thesis, was over 60 pages long. And if I look at what we are publishing now, I see bite-size papers, partitioned into smallest publishable units. I do not think we are alone -- I see the same general trend everywhere I look. We barely scratch the surface and move on, letting ourselves and the other that follow come back and publish many more papers that figure out the details that were left out by the first article. Yes, this increases the number of papers and the citations counts, but something is lost along the way. And this something is what I think attracted me to science as a kid: deep thinking! (Not surprisingly, people lament nowadays that "Deep sequencing has replaced deep thinking." BTW, this phrase, when someone mentioned it to me, was attributed to Sydney Brenner, but I couldn't find any data to support this attribution. Does anyone know the source?) Another thing that is being lost is what I felt when I just started doing my research: that powerful, satisfying feeling of being the only one in the world who really knows the answer to a certain question. Just think about it -- for a few hours (or days, or weeks), before you submitted an article to a journal, there was some mystery of Nature, the answer to which was only known to you! Now this is gone, since we don't really know the answer (just scratched it), and others are only a tiny bit away in their knowledge.
Another thing has happened in the last 20 years. Math, the language in which the book of Nature is supposed to be written according to Galileo, was rightfully viewed with reverence. This has changed. In our push to educate biologists about the importance and the power of math, we have actually moved a lot closer to them than we had expected. We don't view math as lingua franca anymore, but as a necessary evil, as something that must be avoided because every equation in a paper cuts the readership by a half . For me, the first indication of what's coming was in about 2005, when one of the reviewers of my theoretical paper said that s/he couldn't be sure that I was right because I had only proven the result, but had not demonstrated it in a simulation or an experiment. Over the years, I've got many reviews where the only real objections to our work were that the paper was "just a theory" paper (and I am talking about physics-style journals, not about the colored magazines like Nature or Science). Those of you who know my science understand that I don't shy away from experimental data. But if my group is able to solve some experimental problems better or faster than the other theorists, it's because we've already solved these problems years earlier when they used to be "just a theory" work. I value toy models, and the entire history of science shows that some times theory is a step ahead of experiments, and some times it's the other way around. I've been trying to argue this to journal editors, but such arguments have become less and less successful as the time goes by. Just recently, a review of our paper started with a praise to the quality of a substantial analytical work that went into the paper, and then immediately followed with a concern that not all of the regimes we considered were known as immediately biologically relevant, and with an advice to move the math to the Supplementary Information.
Math and theoretical models is the way we "dream" in science, the way in which we fantasize how things can be. Einstein once said "When I examined myself and my methods of thought, I came to the conclusion that the gift of fantasy has meant more to me than my talent for absorbing positive knowledge." When math becomes a necessary evil, and papers are enslaved by what has been observed in the latest experiments, we loose another thing that attracted me to science: an ability to dream, to create.
So, with all of this, I wonder: are we moving towards a 140 character paper, where all the thinking and fantasizing will be relegated to the online supplementary materials?
And the other question is: what should we do about this? I am toying with ideas of not publishing in journals where math is put into the basement, or where purely theoretical papers are considered a faux pa. However, this is not an easy decision since I have graduate students and postdocs, and their employment chances are directly correlated with publishing in such journals. So I am still thinking.