PNAS: We're only for "lab biologists"
Steve Evans, Ken Wachter, and I submitted our recent paper, developing a new mathematical model of
mutation-selection equilibrium, to the Proceedings of the National Academy
of Sciences (PNAS), a journal which purports
to "span the biological, physical, and social sciences", but which has degenerated
over time into a biology journal. It was reported that the editorial board
desired to raise the journal's profile in applied mathematics. We
submitted the paper to the Applied Mathematics section. It bounced
back to us a week later from the "member-editor", with the lapidary remark
that it was an original piece of work, with potentially useful new insights,
but that it could not be published in PNAS because it was, in a word, mathematical,
hence unlikely to be understood by many of their readers. (Click here for the complete report.)
One imagines this reader -- call him Simple Simon Scientist -- breezing through
a typical 500-page issue, from Agricultural Sciences to Zoology, and suddenly
crashing into our paper, with its shameless exposition of the Feynman-Kac
formula. Simple shoots bolt upright in bed, clutching his head in agony.
Next day, in the office, he dictates a letter, cancelling his subscription.
"Never did I think that the National Academy of Sciences would so assault
me, S. Simon Scientist, by printing a paper that I do not understand."
While people lament "math phobia" as an infirmity of school children, it
also seems to strike grownup scientists. It has, perhaps, gone unrecognized
because of the distinct symptomology: not the open "I hate math", but "That's
too abstract", together with the suggestion that "others" would not understand.
The result is an invidious double standard. Imagine that a team
of experimenters were to develop original techniques to attack a problem that
has already been considered to be of broad interest, or that they were to
adapt techniques that had never before been applied to this sort of problem.
Suppose that their experiments then showed that earlier approaches had been
misleading, and that essential conclusions need to be revised. Would
their paper be rejected, with the comment that "the methods will be familiar
to only a small fraction of readers", so that it would be suited to a more
specialized journal? On the contrary, one might say, this should be
an inducement to more people to learn these techniques. One would think,
from the reactions we have garnered, that we had perversely elected to write
the paper in an obscure dialect of Akkadian.
We appealed against this decision, as being contrary to the journal's official
policy, and they did agree to send it out for review. One month later,
we got two referee reports. One
said that the paper is mathematical, hence unlikely to be understood by the
majority of PNAS readers. "It would go totally over the head of any
lab biologist..." ("Only the purest of pure probabilists would understand..."
This reference to "purity" suggests the mixture of inferiority complex and
prejudice that we're up against.) The other was from a mathematician: Our
formulae, it was claimed, were really anticipated by others in a 1976 paper.
To the native Laputan,
of course, God and the universe are mere special cases of your own favorite
theorem. (Or, as Steven Weinberg remarked in another context, "So might a carpenter,
looking at the moon, suppose that it is made of wood.")
There are good reasons why one might consider this paper unsuited to PNAS.
It is unfortunate, though, that the journal clandestinely adopts a
policy of treating mathematical approaches to science as though they were
rituals of an esoteric cult -- "If we needed stochastic analysis, God would
have given us stochastic fingers to count on" seems to be the motto, if
not "Everything we needed to know about mathematics, we learned as freshmen"
-- rather than actually reading the paper, to decide whether the techniques
might be appropriate to advance the common cause of science.