Monday, March 16, 2015

Scientific Writing

I was informed, about a month after my defense, that my PhD dissertation was both well-written and the writing was not liked by the examining committee. This puzzling combination - good but not enjoyed - was explained as my writing lacking elegance. I was advised to seek a creative writing class when I am in Waterloo, because my technical writing is suitably technical but lacks a certain story-skill. I'm not sure I'm making that clear, which may be a function of my apparent weaknesses as a writer.

As a form of slightly-more-productive-than-nothing procrastination, I've been reading some of the opinion papers that scientific journals publish every once in a while bemoaning the current state of this enterprise of Science. Soil Scientists have contributed their share of these articles, mostly written by senior, established scientists and expressing ideas regarding how to write a good paper (or how papers were better back when they were young, or both - those two concepts can dissolve in each other with the help of some sneering condescension as an emulsifier).

Currently open is a scan of Janzen (1996), a paper with the provocative title "Is the Scientific Paper Obsolete?". After an opening in which Dr. Janzen answers the title question with "no", there are a few sections describing the trend of increasing publication productivity by (soil) scientists, and some speculation on the effects this trend is having on the average quality of papers. This follows closely the paper I have just read, Hartemink et al., (2001), in which the lead author invites contributions to his essay from a range of colleagues, built around a similar quantity-vs-quality argument.

I'm not all the way through Janzen (1996), but mid-way in the essay he is describing a trend of increasing specialization of the literature, and uses as an example the number of papers published on earthworms; the per-year rate of those papers doubled every year from 1930 to 1980 (here Janzen cites Satchell, 1992 - which also has a provocative title. This is an interesting rabbit-hole I'm down). Janzen continues:
"At this rate, 6.5 new papers will be published every week by the year 2000, on earthworms alone."
(emphasis original)
Earlier comments in this paper included a description of the growing presence of scientific publishing on the then-nascent World Wide Web; I have access to a much-expanded Web compared to what Janzen had in 1996, so I ran a quick check on his prediction.

In Web of Science, I ran an advanced search with the following string:
TS=(earthworm or Lumbric* or Moniligastr*) and PY=(2000) 
TS is for topic, or a word that appears anywhere in a paper's title, abstract, or keywords, and PY is for publication year. I chose those terms based on the wikipedia page for earthworms: - Lumbricina and Monligastrida are two suborders that animals commonly called "earthworms" fall into.

Web of Science returned 783 hits (there may be some duplicates or other false positives), or nearly 15 papers for each week of the year 2000. Janzen missed it by more than a factor of 2, which I think only strengthens the argument in these papers about the increased productivity derived from the effectively-universal adoption of computers for both analyzing data and composing papers for publication. Actually, come to think of it I'm having a very hard time imagining writing a paper without a computer, and never mind the statistical analyses.

Literature Cited
Janzen, H.H. 1996. Is the scientific paper obsolete? Canadian Journal of Soil Science 76: 447-451.  
Hartemink AE, Buurman P, Dick WA, McBratney AB, van Cleemput O, Young A. 2001. Publish or Perish (6) - Soil science for pleasure. Bulletin of the International Union of Soil Sciences 100: 50-56.
Satchell J. 1992. Take the money - call the tune. Soil Biology & Biochemistry 24: 1193-1196.

Post-script. I tracked down Satchell, 1992 and Janzen mis-cited that paper - earthworm papers did not double every year from 1930 to 1980, they doubled on average every 13 years.
It's hard to get precise numbers from the rambling introduction to Satchell (1992), but he seems to state that by 1984, 3000 papers on earthworms had been published cumulatively since 1930.
Satchell (1992) states directly the 6.5 per week figure, so I suppose it's not Janzen (1996) that missed the mark.

No comments: