randombio.com | writing advice
Monday, January 15, 2018

Why are scientists such bad writers?

How to write a good essay, part 3


O ur job as scientists is not, as some people think, to discover stuff, but to write as many papers as possible. Although these papers sometimes get read, that's not their primary purpose. Their purpose, as with patents, is to stake out a piece of intellectual territory. If someone actually reads your paper, it's bad: they might get interested in the topic, and you'd get a new competitor.

The real goal in science writing is to minimize the number of people who read your paper. Unlike in normal writing, the reader is the enemy. Your paper is intended to prove that you got to the goal line first, thereby demoralizing the reader so thoroughly that they return their grant money back to the NIH, drop out of science, and take up gardening.

That's why I was so horrified last week when the peer-review comments came back on one of my papers. One reviewer thought it was “exceptionally well written.” I was mortified. If the reader makes it to the end of your paper, you have failed. What was I doing wrong?

Scientists drinking (Fisher Scientific Co. ad)
Not the reason

To find out, I took a look at Steven Pinker's famous 5,386-word essay discussing why academics stink at writing. It's a perfect example of what I'm trying to achieve. How many readers of this PC-stuffed article actually survived to the end? I didn't: after about 1200 words I realized I had forgotten to unclog my bathtub, and to this day I have no idea how his essay ends.

Everyone says to write about what you know. Well, writing badly is a topic I consider myself an expert on, having practiced it for many years. So here is my advice.

The softer the science, the longer the paper

Biology papers are long; psychology papers are interminable; and sociology doesn't have papers, only dissertations.* Therefore, since the goal of every field is to become physics, we must make our papers as short as possible.

The purpose of a paper The sole purpose of a scientific paper, as with a patent, is not to be read, but to stake out a piece of intellectual territory.

Writing in science is just like in high school, where the first thing the kids demand to know is “How many pages does it have to be?” Reviewers will nail you if you don't get this right. If the grant instructions say 12 pages, you give 'em exactly 12, even if you have to make the figures bigger to fill the space.

If you're a biologist who tells the reader, as Pinker recommends, that Arabidopsis is a flowering mustard plant, you're telling the readers they don't know nothin', and they'll think you just learned it yourself. By leaving it out, you save four words, which means you can fit more data.

How do we reconcile this with Pinker's claim that the academic writers' chief concern is to avoid being convicted of naïveté about their own expertise? The self-conscious style he kvetches about is deliberate. It follows a specific pattern that's designed to let the reader skip over the useless parts. Take this opening line from one of mine:

Recent research has shown that synaptogenesis is not only important during development, but also plays a central role in associative learning and memory.

The reader only sees the expressions “synaptogenesis” and “learning and memory”. That's what it's designed for. We rarely read any more than the first few lines in the introduction; if a scientist has to read the discussion to figure out what the results mean, the article has failed.

Guidelines to writing scientific papers

So, with that in mind, here are some tips you can follow to make your scientific paper less effective.

  1. Use common techniques to make your sentences incomprehensible.
    Use name-year references to break up phrases that might otherwise make sense. Add qualifiers like probably, could be, and on the other hand to demonstrate your grasp of uncertainty. And, of course, polysyllabic lucubrations are always good.

    B-L violation (Luan, Jiang, Shen, Jing, & Bing, 2005) is an unusual (Bovnik, Floomerbam, Cthulhuson, & Guufmaff, 1987) example of broken (Ham, Aham, and Hamaham, 2016a,b) symmetry.

  2. Stuff the narrative full of chemical names.

    First we isolated ELV-N32 ((14Z,17Z,20R,21E,23E,­25Z,27S,29Z)-20,27-di­hydroxydo-triaconta-14,­17,21,23, 25,29-hexaenoic acid) and ELV-N34 ((16Z,19Z,22R,­23E,25E,27Z,­29S,31Z)-22,­29-dihydroxy­tetra-triaconta-16,19,23, 25,27,31-hexaenoic acid).

  3. Make it more concise than possible.
    Writers in organic chemistry are experts at this.

    3HF·Et3N, THF, rt; periodinane, NaHCO3, CH2Cl2, 87% from 9; (i) (Z)-1-bromo-2-ethoxyethene, t-BuLi, Me2Zn, then 10, Et2O, −78°C; (ii) 1M HCl, −78°C → rt, 90%; (DHQD)2PYR, K2OsO2(OH)4, K3Fe(CN)6, K2CO3, 1:1 t-BuOH:H2O, 0°C, 71%, (2.5β: 1α).

    The average reader might only recognize the words “then” and “from,” but every word counts against the page limit, so you should consider: are prepositions and adverbs really necessary? If this author removed them, he'd have space to say “mp 232° (dec).”
  4. Demonstrate your support of the latest political fad.
    Soft sciences can never compete with hard sciences on the basis of finding truth, but you can even the playing field by claiming that there's no such thing as truth. And you can stuff your narrative full of equal numbers of hes and shes to ensure that nobody knows who the heck you're referring to.

    Smith has proven her thesis that culture is fundamentally meaningless. This is, he says, because her premise of his postsemiotic paradigm of discourse is invalid; if she is not paying attention, he can assume that her/his raison d'être of her observer is deconstruction. Therefore, his/her primary theme of his model of her libertarianism is his presemiotic whole.

  5. Make effective use of repetition.
    Readers only remember 10% of what you write, so you must compensate by repeating each fact at least ten times. That's because readers only remember 10% of what you write.

All kidding aside, I think we're getting a pretty good deal here, considering that scientists are writing for no pay. And it would be hard to improve on this paragraph by Steven Weinberg:

Even when the scalar field in a patch of the space was large enough to start a slow roll inflation, quantum fluctuations in smaller regions within that patch would subsequently have driven the inflaton field to even higher values, so that these regions will begin an earlier stage of inflation. In this way, chaotic inflation turns out also to be eternal. [Cosmology, p.217]

Here's another paragraph, from an organic chemistry paper, that would be hard to improve:

The difluoride XIII dissolved readily in concentrated sulfuric acid with evolution of hydrogen fluoride to give a red solution from which desylmethane (XV), identified by comparison with an authentic sample10, was recovered in 86% yield on dilution with water. When the difluoride was heated with a trace of mercuric chloride in air at 190° hydrogen fluoride was briskly evolved and a mixture of compounds XIV and XV was formed. [Bornstein et al., JACS 1963]

It would be nice if all scientific papers and grants could be this concise (and exciting—yikes, it evolves HF!). But it's really not up to us. Don't like it, talk to da boss.


* Proof: look at the word limits. Nature 1500 words, Science 2500 words, Physical Review Letters 3500, J Clin Invest 4000, American Journal of Psychiatry 5000, J. Biol. Chem. average paper 6000, J. Neurochem. 7500, J Health and Social Behavior 10000, Am. J. Sociology 10000, American Sociological Review 12000. Many journals have no specified word limits.


jan 15 2018, 7:07 am

Related Articles

How to write a good essay,
Part 2

In the interest of estab­lis­hing justice and ensuring domestic tran­quil­ity, here is Part 2 of my series of How to Write Good.

How to write a good essay
Part 1

Sometimes it comes down to a choice between writing something and shooting somebody.

On the Internet, no one can tell whether you're a dolphin or a porpoise

back
science
book reviews
home