I haven’t been blogging much as I am a) recovering from the semester and b) writing technical stuff 24/7. So I am a little tapped out. As I am thinking about writing, it’s fitting that I write about writing. Perhaps I should go full-meta and write about writing about writing… For now, I give you a few technical writing and publishing vignettes.
The wimpy paper
The paper is competently written, correct, and boring as hell. Why? Because there is no story. Each figure is clear, pretty even. The problem? The text pertinent to each figure is banal — just stating what the figure shows, what each symbol or line means (needlessly duplicating the caption), and trivially reading off trends. For example, all you get is
” For parameter C>C0, A increases slowly with increasing B. For C≤C0, A is independent of B.”
Yeah, I can see that from the figure, so what? Tell me why it’s important! What does it mean about the system at hand? Give me some nontrivial insight that the figure corroborates. For instance, “This dependence has been predicted based on the Orthodox Theory, but it was never experimentally measured before. The measurements presented in Fig. X confirms that Orthodox Theory accurately describes the underlying physics.” Alternatively, “The dependence of A on B presented in Fig. X is in contrast with the Orthodox Theory, which predicts A to be a monotonically decreasing function of B for all values of C.” Then go on to say what you think happens and why, ideally support with a different set of experiments or new theory theory.
I hate the wimpy, non-committal papers where the authors don’t state any conclusions or make any strong statements. Science isn’t stamp collecting, it requires you to understand and interpret data; the understanding and interpretation are what’s included in the body of knowledge.
The “I will make you sweat” paper
In my field (and many others, I am sure), there are comprehensive papers and there are letter papers that are typically 4 pages long. A well-written letter should be readable and understandable with <50% of the focus on the part of the reader. You should nudge me towards what I should think, with figures and text playing off and reinforcing one another. Don’t make me sweat like a constipated buffalo, reading two terse, cryptic paragraphs over and over again, trying to figure out what on earth you are talking about and how any of it has anything to do with the data in the figure (or common sense, for that matter).
This is mainly for professors, but others can play as well, and I know the variation will be drastic among fields.
How many papers can you conceivably write in a year?
I mean, the papers on which you are the lead author or do a significant amount of writing? I work almost exclusively with trainees and each paper is a lot of work: I can’t just sit and write it, we have to do the back-and-forths, the edits and the teaching. It takes a lot of time. I think my upper limit for papers where I am really involved would be 1 paper per month on average (more during the summer, less during academic year); usually, it’s less than that, roughly 1 paper per group member per year (senior ones can do 2, junior ones have 0, but it comes out roughly about 1 per group member per year). Full disclosure, I had 8 papers last year; I suppose if I had more than 12 people in the group we’d run into my own bandwidth limitation.
I see these people with 20+ papers per year from their group and I wonder — how? I suppose it depends on the length of the paper (a 4-page letter versus a 12-page comprehensive paper), how many senior people are on the paper (which can be a blessing, as they hopefully know what they are doing, and a curse, since they can be super busy and make it hard to get the edits done with), if you have good postdocs or research scientists… But still, how do you produce so many papers per unit time? I suppose there are these elevated planes of productivity that are like Mount Olympus: only a select few ever reach them.
Which brings me to a related question: There is definitely a need to edit and revise papers before submission. At some point, the gains from successive edits become insufficient to justify the extra time. This issue is particularly important when attempting to publish in high-profile journals, as an insane amount of polishing goes on just to get a chance to pass the editorial desk and go out to review. A recent paper of mine with collaborators took years of hard work of several people, and about 2 years of writing and editing, and eventually went into a Glam Offshoot. Unfortunately, it seems we may have missed the wave of interest or whatever, because, after all that time and effort and the battles with reviewers, it doesn’t seem to be attracting much attention. Oh well.
So how do you decide when to pull the trigger? When it’s perfect? (Never.) When you are sick of it? When the student reaches the point of sending you two emails per day, begging you to submit already? When your grants are up for renewal? When you’ve done the three back-and-forths with student, edited the final version for two weeks and it’s as good as it’s going to get in the near future?
Update: This week’s PhD comic