> JHC's contribution reminds me of an exchange between a typesetter
> of an august academic company and an up and coming young
> mathematician.
>
> When confronted with the written "co-ordinate" of the proofs, the
> young mathematician asked that it be returned to his original
> coordinate. Upon receiving dissent (house style?) he begrudgingly
> agreed but only on condition that "triangle" be written as
> "tri-angle".
>
> He won his point!
>
> [By the way, this was in the [19]60's.]
>
> John MCKAY
... and of course, as JMcK knows full well, that young mathematician
was JHC. The word "coordinate" appeared lots of times in my text, and
all of them had been carefully changed to "co-ordinate", which often
subtly changed the stress. The proofs were accompanied by a strong
warning that changes at this stage were expensive and should be kept
to a minimum. This wasn't the worst one, which happened to many
sentences in the book that had the form
"There are at most [so many] solutions to [such-and-such]"
which had all been systematically changed to
"There are, at most, [so many] solutions ... "
which throws the whole emphasis on the unimportant "at most" instead
of the significant "[so many]".
To someone who often spends hours discussing fine details of wording
and punctuation with his co-authors, it can be really annoying to find
that some junior employee of a publisher has made a million changes
obviously based on some silly little stylebook, even though it's
equally obvious they haven't a clue what the words they're changing
mean.
That's not to say that subeditors don't perform a useful function -
I often agree with their proposed changes, and they are very clearly
needed, for authors who haven't full idiomatic control of the language
they're writing in, for instance. But for authors who clearly have,
the proper course (which I welcome) is to recommend or suggest changes
rather than just silently make them.
When I started writing for them, the Royal Society was very good
in this respect. But a decade or so ago, they obviously appointed
a new editor, who changed the house style for things like tables,
figures and their captions to what I considered an illiterate one,
and their subeditorial practice to a more rule-driven one. As our
"Low-Dimensional Lattice" series of papers went on, Neil and I had
more and more trouble undoing all the stupid changes this caused
to the very carefully written originals we'd sent them, and we
eventually gave up in despair.
The worst and funniest change of this type that I recall (not
from the RS) was in a paper in which the phrase "numbers whose
binary expansion ends in 10 or 11" became "numbers whose
binary expansion ends in ten or eleven" (the rule here, of course,
being that one should use the names "one" through "twelve";
but the digital forms from 13 onwards!). I was very lucky that
this one didn't get into print - I only noticed it just as I was
about to mail the final proofs off!
John Conway