I am citing in Harvard style, so my entries end up something like this:

Widom, H. (1975). Asymptotic inversion of convolution operators, Publ. Math. Inst. Hautes Études Sci. 44: 191–240.

With the above entry, I noticed that the same author cited the year as 1975 from his later papers, but the publisher's website says volume 44 is from 1974. I looked at the actual journal volume and sure enough it has 1974 in big letters on the title page, but the small copyright notice says 1975. It seems that this is the journal volume for 1974, but was actually published in 1975. (According to the stamp on it, my university's library received its copy at the start of April 1975, so my speculation is that they only narrowly missed a deadline to pubish in 1974.)

My question is: how should I cite this? Should I mention both years? If so, how should I fit them in to the citation, and if not, which should I use?


3 Answers 3


With the advent of "online first" and "just accepted" publication streams, this dilemma is becoming even more common. It is now frequently the case that a publication will become available and begin gathering citations well before its "official" publication date.

As noted in the answer by Wrzlprmft and Jim Oldfield, publication dates serve two purposes:

  1. Extra meta-data to help with a search for the reference
  2. Establishing priority between competing researchers

The citation that you write only needs to do the first. Thus, my practice when dealing with multiple publication dates is to write whichever is the "most persistent". For example, I have had citations like "published online June 2012" doi:XXX", then in another later work cited the same paper as something like "34(4), pp 1-10, January 2013, doi:XXX" once it's progressed in the publication queue to actually receive a final volume/issue/date.

Regarding the question of priority: that's not really your job in a mere citation. If order of precedence is actually important to your discussion, you should be discussing it in the text, and not just leaving it to the reader to infer from citation dates, because the real story is often more complicated. That's why in the actual text of many publications, you will see multiple dates in addition to the final publication date. In the most extreme cases, you will see a long string like: "Submitted Date1, Revised Date2, Accepted Date 3, Published Online Date 4."

In short: don't sweat the precedence issue. In the citation, use the date that will be most useful in obtaining the full publication, and let precedence questions be handled separately.


Since the purpose of the citation is to enable others to find the work in question, use whatever number serves this purpose best.

Most likely this will be 1974, as this is the number required by people who want to find the printed journal. People searching online will probably find the paper equally well with either number and even if there is a difference, it will be negligible in comparison to the annoyance experienced by somebody searching for the printed article.

  • 1
    I disagree that this is the only purpose of a citation. Part of it is to establish academic primacy. If I say "... was discovered by Widom (1974)" in the main text, it gives the impression that by the start of 1975 it was public knowledge (in principle) that the fact had been found by him. Oct 10, 2014 at 15:26
  • 2
    @JimOldfield can you provide a reference for your claim about primacy. The APA says the purpose is to credit others and direct people to sources. I cannot find any source that suggests it is about establishing primacy.
    – StrongBad
    Oct 10, 2014 at 15:33
  • In addition: with the review process and journals' backlogs, the date of publication gives only a very rough upper bound on when something was discovered. Quite a few things get credibly discovered independently by multiple people at roughly the same time. Oct 10, 2014 at 15:39
  • Also, there is a small difference between making a discovery and publishing it anyway. In times of preprint servers and online-first publications, the same holds for the first publication and the publication in a journal anyway (I coauthored a paper published in the 2014 issue of a journal, that was almost available online at the end of 2012). Though I am no expert on the academic culture in mathematics in the 1970s, this might have been true even then due to conferences and similar. What is more relevant for academic primacy is the submission date.
    – Wrzlprmft
    Oct 10, 2014 at 15:40
  • 1
    @Wrzlprmft I wasn't using "judge" in the deep meaningful sense, I just meant "interpreted" or similar, which of course is something you must do. The stress was on "the very precise meaning of my words". To be honest, it seems that's what happened again :-) Oct 10, 2014 at 16:47

One way to cite a journal issue with 1974 printed on it but actually published in 1975 is to say that the year is 1974 (1975). (For an example, see the answers to this question.)

You must log in to answer this question.

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