Analogy-watch
I saw this amusing paper in the latest issue of EMBO reports.
The article's title is "Six senses in the literature. The bleak sensory landscape of biomedical texts", and its authors -- Raul Rodriguez-Esteban and Andrey Rzhetsky -- argue that "scientific texts have a reputation for being factual, rational and 'dry" in contrast to other prose that is designed to evoke emotional responses". They therefore set out to obtain evidence that "the sensory-deprived writing style that dominates the biomedical literature impedes text comprehension and numbs the reader's senses and mind".
To that end, they measured the frequency of use of "sensory" words (terms related to the perception of color, smell, taste, touch, sound and time) in 250,000 articles from 78 biomedical journals and compared it to their frequency in news reports from Reuters, in articles in Wikipedia, and in the complete collected works of Poe, Shakespeare and Whitman. They found (surprise, surprise!) that articles and news reports were a lot "bleaker" in comparison to the works of literature. The figure below is some sort of homunculus that gives you an idea of the "sensory" characteristics of the different works they looked at. It shows the balance of "sensory" terms in the different bodies of work they compared, together with the average -- the face in the center.
The authors speculate that reading texts with the characteristic of a biomedical article "is similar to the effect of a long journey through a colourless flat terrain devoid of prominent features: a numbing of the senses", and suggest that "cognitively bleak biomedical texts can and should be transformed into perceptually richer prose".
All of this is well and good, but I must confess that, when I read scientific papers, I have very little patience for metaphors and analogies, which some authors, alas, love to use. Some of them are so gratuitous that it makes you wonder if their use of imagery and metaphor is itself, in fact, an attempt to "numb the reader's senses and mind".
Back when I was the Editor of Nature Reviews Neuroscience, I remember that a couple of my colleagues and I created something we dubbed "Analogy-watch" to record the most ridiculous analogies that we could find in scientific texts. (I regret to say that many of them were penned by some of our own in-house Editors.) Maybe as a result of that experience I've always tried to be careful and separate literature from science. I hope that prospective Nature Medicine authors don't rush to follow the authors' advice and instead use "sensory words" in moderation. Trust me, you don't wanna end up being part of our Analogy-watch file.
To close, I cannot help but pointing out that Calvin (of Calvin & Hobbes fame) had already observed an effect somewhat related to the one reported by Rodriguez-Esteban and Rzhetsky. I didn't see Calvin cited in the reference list, which is why I thought I would include the relevant "report" down here. (Note, you may have to save the image on your desktop, as it appears truncated in the blog page.)

Comments
I have to admit that I have read this amusing blog with somewhat mixed feelings.
Yes, I do agree that a MEDIOCRE metaphor can hurt a scientific article. However, a GREAT metaphor can make an otherwise impenetrable subject crystal-clear. A mathematical isomorphism (think of Kurt Goedel’s mapping of number-theoretic truths to meta-mathematical statements) is an extreme form of a GREAT metaphor/analogy.
Ironically, the text by Mr. Lopez contains a self-referential loop -- what is the purpose of showing in the blog a newspaper cartoon (well, except attacking the authors of the reviewed article)? Isn't it exactly the kind of metaphor/illustration that Mr. Lopez detests? The Calvin & Hobbes cartoon does not illuminate the subject at all.
Both professional journalism and scientific writing, in my humble opinion, should involve clarity of writing and balance in discussion of competing opinions/views of data. (I don't see such balance in the piece by Mr. Lopez.)
REPLY FROM JCL:
1) I think metaphors and illustrations are very different things. In any case, I agree that there are great analogies that can clarify even the most arcane concepts. That being said, most of the analogies one sees tend to be rather gratuitous. Calvin & Hobbes do not offer an analogy; the cartoon is instead a great parody of how some people communicate science.
2) I wouldn't call my comments on the EMBO reports paper an attack.
3) I agree that professional journalism and scientific writing need to be balanced. However, I wouldn't glorify my entries on this blog by calling them "professional journalism" or "scientific writing". Instead, they are meant to express my personal point of view, not even the Nature Medicine's.
Thanks for your comments!
JCL
Posted by: Andrey Rzhetsky | June 23, 2008 11:28 PM
As news and views editor I would like to log my plea against bad analogies. Please immunologists, no more references to medieval battles and klingon war ships!!
COMMENT FROM JCL: I'LL DRINK TO THAT!
Posted by: Charlotte Schubert | March 11, 2008 11:22 PM