I am facing a problem, the article seems a very important reference that everyone talking about the subjects is citing it. However, looking at these articles, they can't give a reference or link to the cited article. Even I use the volume number, article number to search on the net, I still can't find it.

For example, this article, only the second reference doesn't have a pdf link pointing to it. I think the author may be like me can't find the article, but sees others citing it in a very important position, so he also cites it. I can't exclude the possibility that the author reads a paper-based version.

When you search the citing article, many of them can't give a link to that paper, and I can't find it either. Should I cite it anyway? Because from the inspection, this paper seems an indispensable reference when talking about the subject. On the other hand, I can't find it (even on the official site) and I didn't read the abstract or even the title. I think this is not quite equal to the case where the article is behind the paywall, because behind the paywall you at least see the title and abstract.

  • 27
    I think the author may be like me can't find the article --- I think that because the article was published in 1997 (submitted in November 1996), the author did not look for .pdf files on the internet. I don't know how you were looking for articles in 1996, but except for arXiv preprints and some people's homepages, in 1996 I went to university libraries to find articles, not the internet. Feb 3, 2020 at 7:18
  • 28
    I can't find it --- Step 1: Copy/paste "Sov.+Phys.+JETP" into google (result). Step 2: Click on Journal of Experimental and Theoretical Physics (highest search result for me). Step 3: Under "English version archive (JETP, 1967-1996)", use the menu to select "Vol. 20, 1965". Step 4: Find the issue p. 762 is in. It appears you want issue #3, which is not available. Step 5: Go to a university library and look up Volume 20 #3 (March 1965) of this journal. Feb 3, 2020 at 7:36
  • 19
    I have done all the above steps except 5 --- The Journal of Experimental and Theoretical Physics does not appear to be even remotely obscure, so if by chance your university library did not subscribe to this journal in the mid 1960s (very unlikely for most any not-very-small U.S. university, but of course you may not be in the U.S.), it should be easy to obtain by interlibrary loan. Feb 3, 2020 at 7:45
  • 13
    Why would you cite something you've never read? Isn't that dishonest? If you didn't read that paper, then you've used nothing from it, even if the knowledge you're using was first codified in that original paper, but you learnt about it from the intermediate, transitive papers.
    – Pod
    Feb 3, 2020 at 16:55
  • 9
    If your local university library doesn't have a copy, it's very likely that "interlibrary loan" from elsewhere would be delivered to you in the form of a PDF of a scan of the original, rather than them making you wait to get it in the mail. Feb 3, 2020 at 17:26

9 Answers 9


I would definitely try contacting your librarian at your university library. University libraries collect physical copies of papers and they might have this one. Librarians are also a lot more versed in navigating the various search systems and they will probably be able to find this paper, if only as a hard copy. It is also possible that your university does not have the paper, but they could contact different universities to get it.

I would definitely recommend to no cite what you haven't read, especially if you don't even know the title of the paper.

Edit: As others have also noted, it might be good practice in your field to in fact cite the paper, but with the caveat that you cannot find it, as suggested by Thredolsen's answer. If this is common in your field please do this.

  • 8
    "It is also possible that your university does not have the paper, but they could contact different universities to get it." As far as I know, this is a standard service university libraries provide to researchers, and I'd be very surprised if OP's university didn't offer this.
    – JiK
    Feb 3, 2020 at 18:06
  • 9
    I’m torn between upvoting and downvoting. The first paragraph is excellent advice, and most probably the right advice for the OP. But the last paragraph, answering the question as written, is not right at all — in the rare circumstance that you really can’t access a paper but have good reason to believe it’s relevant, a policy of “just don’t cite it” has obvious bad implications; as Thredolsen’s answer explains, the generally established practice (for fairly clear reasons once you think about it) is “cite it, but with a clear caveat”.
    – PLL
    Feb 4, 2020 at 0:30
  • 3
    Regarding not citing if you can't find the paper: I just don't see how you are doing the field a favor. Anyway, if the established practice is to cite with a caveat, maybe just do this to cover your bases. I have edited my answer to reflect this.
    – pgunnink
    Feb 4, 2020 at 19:50
  • 2
    I'd love to upvote this answer but I strongly believe that the last paragraph is wrong, and is recommending an unethical research practice. Just because it's customary to do something does not make it right. This is how research myths are kept slice. Feb 5, 2020 at 11:35
  • If every scholar followed your advice to "no[t] cite what you haven't read", then academia would be very, very different from how it is today.
    – tparker
    Feb 6, 2020 at 1:49

If you absolutely can't find the original work, a viable solution is to reference the original work "as cited in" the secondary work (e.g. Smith, 1960 as cited in Doe, 2000). Many style guides contain explicit instructions on how to do this.

First a caveat: yes, you should generally do your best to find and read the article that you want to reference, including by getting help tracking it down (for example, by asking a librarian). This is especially important if there's a risk that the original work has been cited incorrectly.

However, in some cases, this might not be possible, for various reasons. When that happens, you can reference the original work as cited in the secondary work that you have access to.

For example, here is what the APA says on the subject:

How do you cite a source that you found in another source?

Use secondary sources sparingly, for instance, when the original work is out of print, unavailable through usual sources, or not available in English. Give the secondary source in the reference list; in text, name the original work and give a citation for the secondary source. For example, if Allport's work is cited in Nicholson and you did not read Allport's work, list the Nicholson reference in the reference list. In the text, use the following citation: Allport's diary (as cited in Nicholson, 2003).

Similar conventions appear in other style guides, such as MLA and Chicago.

  • 6
    the idea is that you still preserve some amount of information for future, rather than avoid citing at all? Sort of "I haven't read the paper, but [Doe 2000] says that [Smith 1960] supports claim that E=mc^2" Feb 4, 2020 at 0:11
  • 6
    Yes, exactly. While it's of course preferable to read a paper before citing it, there are going to be rare cases where that's not an option, but that doesn't mean that you should simply not cite that paper. The important thing is to be transparent about not having read it, and about your alternative source.
    – Thredolsen
    Feb 4, 2020 at 1:23
  • 1
    The quotation implies that Allport should be mentioned in the text, but not in the reference list. Feb 4, 2020 at 14:33
  • 4
    Best answer. It's not uncommon for papers to misrepresent the source they're citing, often in error but occasionally even deliberately. Citing without checking the original and without acknowledging the second-hand nature of the cite risks adding weight to the deception. In my PhD days I encountered a cluster of half a dozen papers all reporting the same baseless figure (which was incorrect by a factor of about 2000%) apparently due to second-hand citation through a paper that misrepresented the original.
    – G_B
    Feb 4, 2020 at 21:41
  • 1
    A version of this is incredibly common in the legal scholarship and related fields. We normally list the source we are citing and then put in parenthesis "citing XXXX" to give credit to the earlier source while relying on the source actually read. Feb 5, 2020 at 22:53

Don't cite it without reading it, but get help finding it. You can consider asking a librarian for this kind of help.

  • 58
    Do not underestimate librarians! They can do things, you can't do.
    – Dirk
    Feb 3, 2020 at 13:01
  • 10
    @anoffercan'trefuse why do you think they would have to find it online? They might be able to pull out the magazine from the depot. Also there is a thing called 'interlibrary load' so they can request a copy from a different library.
    – lalala
    Feb 3, 2020 at 14:31
  • 15
    I once needed a book, published in Russia by a publisher that seems not to exist any more. I couldn't find a copy on line, and my university's library didn't have it. Nevertheless, a librarian here quickly found it at another university (one that wouldn't have occurred to me as a likely source) and got me a scanned pdf of the whole book. Feb 3, 2020 at 17:57
  • 13
    @anoffercan'trefuse. Librarians do magic. Talk to them.
    – TRiG
    Feb 4, 2020 at 0:14
  • 4
    @anoffercan'trefuse Librarians spend literally years learning how to organise and retrieve knowledge, including knowledge that never made it to the web. Don't assume they can't find it until you've asked them.
    – G_B
    Feb 4, 2020 at 21:47

Apply the principle: "When in doubt, explain your choice."

You'll need to devote a sentence or two to say where you've actually found the point you're citing; where it cited from; and the fact that it the indirect source is difficult to obtain.

If you do that - everyone will understand:

  1. What the reality is.
  2. What you did.
  3. Why you did it.

... so there's no misrepresentation and nobody can fault you.

If you can't spare the space, then @Thredolsen's suggestion is also fine, in my opinion.


Do you know any of the authors citing the papers?

  • You can email them and ask where to get the paper or whether they can sent it to you.
  • If you have an adviser, ask them. Maybe they know about the paper or one of the authors citing the paper.
  • If you know from current or former members of your group citing the paper, you might find it in some repository or shared paper storage place.

As others mentioned, the library is very useful and helpful.


I would suggest pursuing an Interlibrary Loan to get the article from another library. These can take a couple of weeks to come through. As others have noted, librarians are remarkably capable of hunting down copies of sources that are tricky to obtain.


I recommend the following actions:

  1. Make a reasonable quest to find the article, including contacting a library. However, I do not feel hard-to-find papers deserve too much tracing effort, unless your purpose is to be exhaustive.

  2. If the paper seems to be very important/relevant, cite it in the text by explicitly referencing the source of the reference, e.g. a popular handbook and/or survey article and explicitly stating "not read at the time this article is going to press."

  3. If the paper seems to be of some minor relevance, as with point 2 above, but in a footnote rather than in the main text, again stating you haven't read it.

  4. If the paper seems of minor importance, don't cite it.


I can't exclude the possibility that the author reads a paper-based version.

It's absolutely this one, for sure.

I cite a 1948 paper in the Journal of Navigation in many articles. It is not found online anywhere, not even in abstract form. So in those articles you see a citation like the one you point to, the original cite with no PDF or URL.

In case you're wondering, I contacted the author via email after poking about on FB and he photocopied it and mailed it to me.


I'm going to go against the grain here a little bit. I will preface my answer by saying I don't know anything about your field, so everything I've found is based on the information you provided.

Yes, cite it, even if you can't find it

If it were a relatively recent paper (~5 years old) or is controversial, you should absolute not cite it if you have not read it yourself. However, the paper you're looking for (or at least, the examples you provided) is an English translation from 1965 of a paper published the previous year, entitled "Inhomogeneous State of Superconductors" (or "Nonuniform state of superconductors"). It appears to be in an issue not available on the journal's website, but many more from the same authors are available on the site to give you a scope of their work adjacent to this apparent seminal work. It also appears that the work in that paper was among the first to describe a now-validated component of superconductor physics; this, I would imagine, makes it something that would appear in textbooks, which generally classifies information that doesn't require citation. You can get the gist of what's in the article by what others who cite the article say about it (you can determine that from Google Scholar), or by the article by other researchers that found the same results.

So why cite the seminal work if you haven't read it and you got the information about it from another source, especially if the information may not need to be cited at all? Sometimes citations are simply a nod of understanding or recognition to major works in our fields. You would be ethically safe not citing these authors at all (unless you're addressing something very specific about their paper), but it's sort of a sign of respect to cite them anyway.

  • 4
    (repost of a 30 August 2016 comment) For what it's worth, a 1919 paper by Besicovitch --- Sur deux questions d'intégrabilité des fonctions --- has been cited numerous times (even by Terence Tao), but I suspect no one alive has seen a copy of this paper. Feb 4, 2020 at 18:35
  • 5
    Citing a work attests to its contents. If you haven't read it - you cannot thus attest, and must not cite, except if you make the circumstances clear.
    – einpoklum
    Feb 4, 2020 at 18:49
  • 1
    @einpoklum That seems like a very black-and-white view of citations, especially when the work in question seems excessively laborious to acquire. You can pretty confidently attest to the contents of a paper you haven't read if those contents are common-knowledge within the field.
    – RexJ
    Feb 4, 2020 at 18:55
  • 6
    @John: If you haven't read the work, you can't attest to anything about it. If something is "commonly-known" then it doesn't need citation, or alternatively, you can cite wherever it is you learned the commonly-known information. In my opinion it is a borderline ethics violation to make this kind of citation.
    – einpoklum
    Feb 4, 2020 at 19:37
  • @einpoklum Again, while you're correct about what a citation usually means, it's not necessarily true in practice that you must read something to attest to its contents. Consider the example Dave referred to; are all the authors citing that article guilty of ethics violations? Are they wrong to cite something when the contents of the paper are widely known just because they haven't read precisely how those contents are worded? In practice, a very strict citation criteria that requires common references to be read (and at that, how in-depth must they be read?) before citing is a hindrance.
    – RexJ
    Feb 5, 2020 at 20:20

You must log in to answer this question.

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