On the face of it the place of publication should be a simple matter of fact. But of course, publishing houses, which because they will insist on employing intelligent types full of ideas, are expert at over-complicating things. In the sixties and seventies Cambridge University Press’ title pages would announce the place of publication as Cambridge, London, New York, Melbourne. I guess if you lived in Brisbane, Melbourne would be the place the actual book in your hand came from, unless Aunt Daisy had mailed it to you from Cheltenham as a Christmas present. But of course “publication” as well as “place” are squishy concepts.
In the early days of publishing, before there were such organizations as publishing houses, books were printed and retailed by the same organization. If the printing press was in Chiswell Street, that’s where you’d go to buy the book, and the place of publication would obviously be London. As the publishing side of this operation became predominant, similar relationships between the front and back offices remained. A surviving consequence of this state of affairs is the law requiring that the name of the printer appear in every book printed in Britain — so that the printer can be sued for libel. Eventually however the printing plant and the publishing office grew further apart, culminating in most publishers selling off their printing arm entirely. So by the mid-twentieth century you might have a publishing office in London, and printers in Tiptree, Frome, Aylesbury, Fakenham, Edinburgh, wherever; and you certainly wouldn’t want to be describing the place of publication as the loading dock of your Tiptree printer. Actually things were even more complicated than that, since in those days printers tended to be separate from book binders, so you’d probably be shipping your sheets from Tiptree to Harlow, say. Publishers strove to have their warehouses in London, in order to be close enough to the big bookshops to be able to service the carriage trade, but post-WWII economics tended to force warehouses out of town. Thus your publishing office might be in an elegant building in central London, your warehouse in Northamptonshire, and your printers and binders all over the place. As consolidation and expansion pursued their irresistible march through the industry and many publishers moved their entire operations out of town, the picture naturally became even more complicated: just the sort of thing intelligent chaps with too much time on their hands love to debate.
Then there’s the wrinkle of what exactly “publication” means. Coming out of the copyright law, we have all generally accepted the word as meaning “the offering of copies for sale”. Now at CUP in the sixties, sales, marketing, publicity — everything one might connect with the offering of copies for sale — was run out of London. In Cambridge there was editorial, production and design, as well as the Printing House. So who was doing the “offering for sale”? It doesn’t really matter which answer one gives — the situation was if not complicated, at least not utterly straightforward. Nobody would really deny that the editor of a book had something to do with its publication. Thus Cambridge and all the sales offices were included on the title page. At that time Oxford University Press, which had many more branches around the world and couldn’t include all of them on the title page without having to shoulder something else off, described their (trade) books thus: London, Oxford University Press, New York, Toronto. They would make a distinction between books “published” out of London or another branch (the majority) and books “published” by the Delegates in Oxford. These latter would be labelled: Oxford, at the Clarendon Press. They’d all be sourced from the same warehouse in North London.
At the end of the day, place of publication doesn’t matter to anyone — except for academics compiling their bibliographies and I guess the librarians studying them. I often used to think that referencing books with their UK place of publication was a bit of an affectation among some US scholars. It sort of implied that you’d done all your research in the British Museum Reading Room or the Bodleian. I suppose in some ideal world referencing a book as Cambridge University Press, New York might pick up a distinction between an edition printed in America, as against the original printing done in Britain, which might be referred to as Cambridge University Press, Cambridge, or maybe London. But such nice distinctions, which aren’t agreed to by all parties anyway, may be utterly irrelevant. The two books may be to all intents and purposes identical, differing only in the paper used. That this sort of distinction may be exactly what some future PhD student wants to study doesn’t justify the whole game. What would be so wrong with just abandoning the place of publication in our referencing systems? Nowadays many a book published in Britain may be purchased in America even though there’s an entirely separate edition published here by a different publisher: these are the ones on Amazon with the odd prices — $16.71 or such-like direct conversion from a more rational £ price. [Publishers spend much time checking up on this sort of thing, and usually the overseas edition is quickly removed by Amazon.] And let us not get into any discussion of the place of publication of an e-book; The Cloud?