Re: STM Publisher Briefing on Institution Repository Deposit Mandates

[Stevan Harnad <amsciforum@gmail.com>]

On Mon, Jan 19, 2009 at 3:14 PM, Sandy Thatcher <sgt3@psu.edu> 
wrote:

> The only statement in Stevan's commentary that I find 
> surprising and questionable--because it is stated with such 
> certainty of its truth, with no reference to any empirical 
> backing, which is unusual for Stevan--is the claim that it is 
> "exceedingly rare" (Stevan's emphasis) for copyediting "to 
> detect any substantive errors" in articles. I have no evidence 
> to disprove this claim that is based on systematic 
> investigation of my own, but in all the years I spent as a 
> copyeditor myself, it does not ring true, and was not 
> consistent with my own experience in editing scholarly work in 
> the humanities and social sciences.

But Sandy, you were copy-editing books, and I was talking about 
journal articles (OA's target content)!

And during those years you were copy-editing at Princeton, I was 
editing (a journal) at Princeton. My only evidence is from those 
25 years: Lots of substantive errors were caught by the editor 
(me!), but that was part of the peer review, the editor being a 
super-peer. Negligibly few were ever caught by the 
copy-editors...

> Are the sciences any different? Not according to one editor who 
> has worked on thousands of scientific articles, who commented 
> on a draft of my article on "The Value Added by Copyediting" 
> (Against the Grain, September 2008). Among other things, he 
> testified that "even in highly technical articles 'the 
> equations are usually accompanied by thickets of impenetrable 
> prose,' and a lot of his work 'involves making sure that the 
> text and the equations say the same thing.' He also adds that 
> he checks 'the basic math in tables, since it's amazing how 
> often scientists get the sums and averages wrong.'"

There's a lot of awfully bad writing in science, alas, and the 
copy-editing is usually so light that it doesn't make the writing 
much better. But I said *substantive* errors, and the 
responsibility for catching those is the referees' (and 
editor's), not the copy-editor's.

> A study by Malcolm Wright and J. Scott Armstrong titled "Fawlty 
> Towers of Knowledge" in the March/April 2008 issue of 
> Interfaces also found high rates of errors in citations and 
> quotations, partly because researchers relied on preprints and 
> never bothered to check the accuracy of citations and 
> quotations from those preprints. I would consider these 
> "substantive errors," since they are not simply matters of 
> style or grammar. So, I would ask Stevan whence his high degree 
> of confidence in this claim derives. Nothing in my experience, 
> or that of other editors I have asked, bears it out.

Sandy and I clearly mean something different by "substantive 
errors": I wouldn't consider citation errors substantive (though 
it's certainly useful to correct them).  I think citations and 
even quotations will be increasingly checked by software, online, 
as everything is made OA. But I agree that only the future will 
decide how much copy-editing service author/institutions will be 
prepared to pay for, if and when journal publishing downsizes to 
just peer-review (plus copy-editing) alone.

Stevan Harnad