One of the arguments for rigid style, which is dispositive in my experience, is collaborative writing. When there are multiple authors, contributing to both the writing itself and to the editing of each other’s writing, it is just much easier if everyone converges quickly on one style, and the Schelling point is IMRAD (with some leeway about where the M appears). Everyone writing the paper already knows how to write in IMRAD, and how to edit a paper written IMRAD.
Since the number of authors per paper has also steadily increased over time, this would further consolidate IMRAD as a standard.
I would put it this way: we have cut most of the useful information (inspiration, project history, unknowns, intuitions & opinions,
creative speculation) out of the papers.
Papers do not link to blogs so the web of citations is sparse and misleading, and it will usually lead you nowhere fast.
(As an aside, something that could absolutely revolutionize science: if authors could put up a flag "this is crap please do not follow me down this barren path" after a paper got accepted and cited a couple times.)
Or more generally if you were able to add more free thoughts after acceptance.
One perfect example of "style" is the sentence from the one-page paper in Nature 1953 from Watson/Crick announcing the double helix: “It has not escaped our notice that the structure we have postulated immediately suggests a possible copying mechanism for the genetic material.”
What about the reader's perspective? The increasing burden of knowledge means that researchers in a field need to have familiarity with a lot of existing work, and anecdotally people have developed strategies for getting through that quickly that rely on a standardization like IMRAD ("read 100% of abstracts, 75% of results sections, 5% of methods sections" or whatever). If everyone did their own thing, reading might be more interesting but it would be slower, so it would be even harder to get to the knowledge frontier.
Great usage of turgid….you could argue argue they go hand in hand, its easy not to think about style when you dont have to think about overall structure in any kind of significant way
I was a graduate student in the late 1980s, and I had to present a paper at a journal club. I think the paper concerned finding a gene for some particular disorder (I might be getting the details wrong here). I couldn't figure out how the authors knew to look for the gene on a specific chromosome. I asked one of my professors, and she said that it was sheer good luck that the gene was in the first place they looked. But nothing in the paper admitted to that fact. Seems dishonest not to acknowledge that they were lucky.
Excellent essay :-) I, for one, didn't know that, of all A. Einstein's papers, only one was "peer reviewed"! This would be unthinkable nowadays ;-) Nobel Laureate Roger Tsien liked to say that it baffled people to know that his Nobel prize-winning discovery was "only a JBC paper".
In favor of the rigid, standard style - I believe it levels the field, as it makes it universally understood. Especially for non-native English speaking scientists, like myself. Setting aside considerations as to how it has come to be, or whether it is fair or not... English is the universal language for science, understood as the application of the scientific method. We may then use our own languages for artistic, subjective, individual expression... But for science to continue to be universal, each term must mean the same for everybody.
One of the arguments for rigid style, which is dispositive in my experience, is collaborative writing. When there are multiple authors, contributing to both the writing itself and to the editing of each other’s writing, it is just much easier if everyone converges quickly on one style, and the Schelling point is IMRAD (with some leeway about where the M appears). Everyone writing the paper already knows how to write in IMRAD, and how to edit a paper written IMRAD.
Since the number of authors per paper has also steadily increased over time, this would further consolidate IMRAD as a standard.
It's a great point, one that definitely would have been good to touch on in the essay :)
I would put it this way: we have cut most of the useful information (inspiration, project history, unknowns, intuitions & opinions,
creative speculation) out of the papers.
Papers do not link to blogs so the web of citations is sparse and misleading, and it will usually lead you nowhere fast.
(As an aside, something that could absolutely revolutionize science: if authors could put up a flag "this is crap please do not follow me down this barren path" after a paper got accepted and cited a couple times.)
Or more generally if you were able to add more free thoughts after acceptance.
Put the important stuff back in papers!
well said!
One perfect example of "style" is the sentence from the one-page paper in Nature 1953 from Watson/Crick announcing the double helix: “It has not escaped our notice that the structure we have postulated immediately suggests a possible copying mechanism for the genetic material.”
What about the reader's perspective? The increasing burden of knowledge means that researchers in a field need to have familiarity with a lot of existing work, and anecdotally people have developed strategies for getting through that quickly that rely on a standardization like IMRAD ("read 100% of abstracts, 75% of results sections, 5% of methods sections" or whatever). If everyone did their own thing, reading might be more interesting but it would be slower, so it would be even harder to get to the knowledge frontier.
I don't mind the IMRAD format, it's the turgid unreadable prose that's the killer.
Great usage of turgid….you could argue argue they go hand in hand, its easy not to think about style when you dont have to think about overall structure in any kind of significant way
I was a graduate student in the late 1980s, and I had to present a paper at a journal club. I think the paper concerned finding a gene for some particular disorder (I might be getting the details wrong here). I couldn't figure out how the authors knew to look for the gene on a specific chromosome. I asked one of my professors, and she said that it was sheer good luck that the gene was in the first place they looked. But nothing in the paper admitted to that fact. Seems dishonest not to acknowledge that they were lucky.
I think you need to read some medical papers: https://www.bmj.com/content/347/bmj.f7198
Excellent essay :-) I, for one, didn't know that, of all A. Einstein's papers, only one was "peer reviewed"! This would be unthinkable nowadays ;-) Nobel Laureate Roger Tsien liked to say that it baffled people to know that his Nobel prize-winning discovery was "only a JBC paper".
In favor of the rigid, standard style - I believe it levels the field, as it makes it universally understood. Especially for non-native English speaking scientists, like myself. Setting aside considerations as to how it has come to be, or whether it is fair or not... English is the universal language for science, understood as the application of the scientific method. We may then use our own languages for artistic, subjective, individual expression... But for science to continue to be universal, each term must mean the same for everybody.