As a follow up question to my previous one, under which circumstances is it reasonable to include the name and/or the address of a publisher in a citation?

In my opinion the most important characteristic of a citation is that it uniquely identifies a given publication and makes it as easy as possible to access it. This, however, is usually already given by title, authors and, if applicable, the title of the collection (proceedings, journal...). If available, a DOI alone would be sufficient for that.

As secondary feature, further details provide more information about the properties of a publication without the need to look them up explicitly. For example, the year is usually very interesting to assess the timeliness of the related work.

Neither the name nor the address of the publisher seems to fall in one of these categories, even though it might be reasonable for example for some exotic books that are otherwise hard to find*. For this reason I usually omit the publisher's name and address in citations, partly to keep them brief, partly since I am not aware who the publisher actually is, partly because others in my area of research (CS) do the same, partly because I am too lazy to put work into something that seems superfluous to me. Sometimes, however, the publishers and their addresses are added during copy editing, so there seem to be at least some people with a different opinion than me.

  • Is it considered bad academic practice to omit the publisher and the address?
  • What are other advantages or disadvantages?
  • Does it depend on the type of the cited publication? Books might be different from articles in journals or conference proceedings.
  • Does it depend on the type of the publication I write? In a short paper, brevity might be more important than in a PhD thesis where accuracy is essential. What about a publication list in a CV?
  • Does it depend on the area of research? Maybe we computer scientists are just lazy?

*Maybe someone wants to drive to Heidelberg to ring the bell at the Springer office just to be told that it a given publication was actually published by SpringerOpen in London...

  • 2
    When referencing a journal, one does not usually include publisher information - the journal alone suffices. When referencing book one usually does, not least because the same titled publication in different countries can have differences in basic things like page numbering.
    – Jon Custer
    Commented Sep 4, 2018 at 14:57
  • 1
    I think some of the norms here depend on the discipline. In my area (pure mathematics) the answer to your title question is "probably not very much"; but I have the impression from posts/comments elsewhere by those in the humanities that such things remain a bigger deal for them, if only because of tradition or institutionalized peer-pressure
    – Yemon Choi
    Commented Sep 4, 2018 at 16:33

4 Answers 4


It's always important to know the venue for a scientific publication, not just in case you have to track it down, but also in order to understand what sort of review or other vetting process (if any) the publication will have gone through.

For a conference or a journal article, that's the name of the conference or the journal. Thus, I wouldn't typically bother including the publisher---although many such effectively include the publisher in their title (e.g., IEEE Signal Processing, ACS Nano, Nature Methods).

For a book, however, the publisher is the venue, as they decide what merits publication and how to try to ensure validity. Knowing that a book is published by the Oxford University Press means something, just like knowing that a conference is sponsored by the IEEE. It's not a guarantee of correctness any more than journal peer review, but it does imply some basic degree of quality control.

As for address: while it might theoretically be useful for disambiguation, these days most large publishing organizations have multiple sites and are really anchored by their online location rather than their physical location. Thus, I find address completely useless in a citation of recent work and will omit it unless forced by a publisher.

Finally, note that all of this does change if you're talking about works that are not accessible electronically (e.g., much of the literature before the 1970s or so), for which as many hints as possible are useful if one must track it down.

  • "before the 1970s or so" --- I would add "and after the 1910s or so", since other than university library "special manuscript collections" and such, I've found that virtually any publication you can find in a university library published before 1910 or so is freely available on the internet. For example, in this answer of mine, which reformats a reference list of mine from April 2008, every single one of those mostly obscure items are freely available on the internet. Commented Sep 4, 2018 at 18:21
  • Publisher location can still matter quite a lot for books, which are regularly edited and amended in various ways for publication overseas (even when the parent publishing company is the same).
    – 1006a
    Commented Sep 6, 2018 at 18:38
  • 1
    @1006a While that may be the case, I still don't know how that interacts with multi-location publishing houses, especially given modern distribution systems. Is the Canadian edition produced in New Jersey or London? If the publisher is listed as having offices in New York, Chicago, and Berlin, how do I know which one produced the edition that had the chapter whose PDF I accessed? If my citation manager tells me it was published in London, is it right? One can certainly attempt to track these things down, but I'm dubious about value vs. difficulty when the information isn't readily available.
    – jakebeal
    Commented Sep 6, 2018 at 20:25
  • 1
    I make no claims for disembodied PDFs, but if you are using a physical book (many, many scholars do still do this, though perhaps not so much in computer science) then it is still worth using the information found in the front of the book. This works for most scholarly eBooks, too. I am a strong advocate for citation managers and other tech shortcuts to citation, but they aren't a substitute for human judgment. (This point was primarily meant as an aside for future visitors who might be in fields where books matter.)
    – 1006a
    Commented Sep 6, 2018 at 20:53

Also: Books (unlike journal/proceedings papers) may not be peer-reviewed. So mentioning the publisher (showing it is a well-knonw publisher) gives a certain credibility to the book. As opposed to self-published books, which nowadays are easy to do.


Journal citations rarely (if ever) include the publisher's information. The reason is simple: journals can change publishers. I have at least one example in mind (cahiers de topologie et géométrie différentielle catégoriques, a math journal). Moreover journal names are usually trademarks, which prevents duplication (and editorial boards from packing up and starting a new journal with the same name when the publisher's bad behavior would require it). So listing the publisher is pointless, because the journal's name (+ other info such as volume, issue...) is sufficient; and even actively bad, because if the journal changes publishers later, people will wonder what they are looking at.

For books, the story is different. GEdgar already said that this allows distinguishing bad publishers from good publishers, though even renowned publishers aren't immune from publishing... subpar texts. Moreover, an edition of a book will always be published by the same publisher, because it's something that has already happened and it's difficult to change the past. If the publisher changes, it's a new edition and will be listed as such.


Is it considered bad academic practice to omit the publisher and the address?

I'd consider omitting the publisher bad practice, because it is sometimes necessary to perform a reverse search to find a paper, e.g., find the publisher, find the proceedings, find the right year, and find the paper.

Does it depend on the type of the cited publication?

I suppose it depends on the infrastructure that's available, in particular, where will manuscripts be archived. When they are archived by the publisher, then identifying the publisher is crucial.

  • "When they are archived by the publisher, then identifying the publisher is crucial." That might be true for small publishers, but publications by the large ones (e.g. Springer, Elsevier, IEEE...) are well indexed in specialized and general-purpose search engines so that it is irrelevant in practice.
    – koalo
    Commented Sep 5, 2018 at 11:17
  • @koalo Try looking up an old manuscript, they are far more difficult to access.
    – user2768
    Commented Sep 5, 2018 at 11:25
  • I agree, so I extend my statement to "That might be true for small publishers or old manuscripts"
    – koalo
    Commented Sep 5, 2018 at 11:30
  • @koalo Which will eventually be true for every manuscript... You could perhaps replace "old manuscripts" with "manuscripts written before X" (I'm not sure what value X should take), but that assumes our archiving mechanisms are future-proof (that is, anything cited today will be accessible in the future), which I'm unwilling to believe.
    – user2768
    Commented Sep 5, 2018 at 11:49
  • Fair point and an argument to provide as many information as possible in a citation, but I do not think that my research is interesting for anyone after the Internet stopped working ;-) Of course, it might be the case that Springer shuts down all servers, but then it also does not help me to know that I could have received something from Springer (except from the fact that it is useless to continue the search).
    – koalo
    Commented Sep 5, 2018 at 13:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .