I just got the page proofs of an article I wrote that is to appear. In it I cited a book, like this:
Author, Title in Italics, Cambridge University Press, 1891.
In the page proof it looks like this:
Author, Title in Italics, Cambridge University Press, Cambridge, MA, 1891.
Obviously that's wrong and I will so notify them.
But I wonder: Does the renown of this particular publisher justify omitting to identify the city after the publisher's name when the city's name is actually a part of the publisher's name and everybody knows which "Cambridge" it is? (This came from a "production company" that contracts with the journal, so maybe "everybody" doesn't include "production companies"). If the city must be named separately I'd be tempted to write
Author, Title in Italics, Cambridge University Press, Cambridge, 1891.
which still relies on "everybody" to know this.
"MA" is an abbreviation promulgated by the United States Postal Service. I had the impression that in England, postal addresses typically put the name of the county after the name of the city, separated by a comma, but that that probably wouldn't be done with Cambridge. If I did that, then "Cambridge" would appear three times in one line. An alternative would be what I would do if I were sending a letter the old-fashioned way from America to someone in Cambridge and put "England" or "United Kingdom" after the comma (although for a letter I'd probably put the postal code after the city's name on the same line and "England" would be on the next line). (Somehow I feel more comfortable with "England" than "United Kingdom" since it's not only an entity created by politicians but also a geographic name that will persist even if the UK becomes defunct. Or as someone once said "There'll always be an England." But maybe that's just personal taste.) I suppose "England" would be harmless and probably no one would notice, but I would feel silly thinking someone would have to be told that's where Cambridge is.
What do people normally do when citing stuff from this particular renowned publisher?
There's really no harm in adding clarifying information. Obviously, we need a city name if we get "University of Georgia Press" because there are two very different University of Georgias in the world. One located in Athens (Georgia, USA — not Greece!) and the other located in Tbilisi.
Likewise, if we see a publisher we're not familiar with, and the citation is only "Athens", it's nice to know at a quick glance whether we're talking about the American Georgian city, or the Greek one. The topic of the paper make dictate which one is the "expected" one — if we're talking Atlanta politics, we can assume Georgia, unless otherwise specified, and if we're talking Plato, we can assume Greece.
Ultimately, though, it depends on the house style. If the house style includes the city, they'll do it for all publications, no matter how obvious one would think. In your specific example, it's undoubtedly an overzealous copy-editor who is probably accustomed to having "Harvard University Press, Cambridge" and fixing it to "Cambridge, MA" to clarify.