How can I find an old, obscure, or otherwise inaccessible paper when the usual methods fail?

Question

I am trying and failing to find a particular paper. My university's library catalogue, Google Scholar, and Google all came up empty. I even tried the catalogue of the paper's institutional affiliation, but they didn't have it either. It's possible the paper really has been lost forever, but I find this unlikely; it is a significant paper that still gets citations. What other avenues exist for finding this paper?

Note to readers: this is a frequently-asked question and so we have compiled this "canonical" post to amalgamate some of our best answers on this topic (per this meta discussion). Questions requesting individualized help finding a particular source will be closed as a duplicate of this: we hope these "hints" will be helpful, but tracking down individual papers is not our role.

Answer 1

Your university librarians are MADE for this! Don't feel ashamed to ask -- there are librarians more into archiving and cataloging, and some more into education and service.

Give as much information as you currently have, and they'll find it for you AND/OR show you how to find it yourself in the future (if it's accessible via a resource like some of those mentioned above.)

Answer 2

• Search the web for it: open archives, search engine, authors webpages, whatever (you've probably done that already…)
• Ask the contact author for a copy, which she'll usually be very happy to provide. I regularly get such requests from other researchers in my field, and it feels very good to know that people are actually interested in your work.
• If the contact author doesn't respond, try the senior author, then other authors.
• Try your local library. Look into their database, and also ask the librarian there if the document might be available through loan from other libraries. I've never had much success doing that, but you never know…
• Ask a few colleagues at different institutions if they can get it for free (see this question). If you have friends at large/famous/well-funded US universities (Princeton, Harvard, …), they probably have a more comprehensive access than you.
• If it's a really important paper, pay for it!

Whatever happens, make sure you make a copy for others in your lab/group, and archive it. When I started my PhD, there was a folder (the heavy paper type, not the computer type) labeled “important but hard to find papers” that the group had accumulated along the years. It was the most treasured object in the whole lab.

Answer 3

Why donʼt you contact the library of the institution who published the paper for advice? Iʼm sure that the library has a copy of it, and they might have translations as well.

Answer 4

When doing internet searches, consider using all variants of the author's name. In [one specific case someone asked about], I was able to find the paper by excluding the author's given name; this expands the search to include repositories that only track given names by the first initial.

Answer 5

Assuming your school participates (and most do), you can get scanned copies of articles through interlibrary loan. You should be able to get access to just about any article ever published in a journal that way.

More generally, talk to your librarian. It's what they're for.

Answer 6

This is a use-case for Sci-Hub, an open collection of papers from various journals. Note that using Sci-Hub to access copyrighted materials which they do not have permission to give you is illegal in certain jurisdictions. Otherwise, it's legal to download.

Answer 7

When looking for a paper I first look on line to see if it is available. If it is available for free I download it, if not I then look in my library's on-line catalog. If it is available on-line, but not in my library's catalog I request an inter-library loan (ILL). I have never been unable to obtain an article via ILL that was available on-line, but if I wasn't able, I would just buy the article. If it is in my library's catalog I go to the library and photocopy it.

If the article is not available on-line, I look in my library's on-line catalog. If it is available in the library catalog, I go to the library and photocopy it. If it is not available in the catalog I file an inter-library loan request. The success rate of these requests is lower than I would like.

If the article is not available on-line, in my library, and ILL failed then things get difficult. The first step is to email the authors. The second step is to talk to the reference librarian and see if you can find a library which is not part of the ILL service which has the paper. If you find one, talk to the librarian about how to obtain it.

My field also has a number of email lists where hard to find paper requests are not uncommon, it might be worth trying these. Similarly, you can ask colleagues directly if they have a copy. It might be worth doing a reverse citation lookup to find colleagues who have cited the paper (most people try and read what they cite).

Answer 8

I have had some luck on The Internet Archive (https://archive.org/). If you're familiar with the Wayback Machine (which keeps old web pages), that's one part of the Internet Archive. Another part is a repository of many old books, and some libraries do upload scans of old journals that are out of copyright. I've had roughly a 50-50 hit rate looking for old papers, perhaps a little higher - at least when given several attempted searches in each case, e.g. to allow for issues with different ways of recording journal names and so forth.

[Note that you want the lower of the two search bars -- the top one is for the Wayback Machine. They're clearly labelled but it's easy to miss that you need the lower one. More than once I've caught myself typing a search in the wrong place.]

The quality may be rather mixed, sometimes excellent, sometimes poor -- but you may also get multiple libraries uploading scans of different parts of a journal's run so with a little luck you may have several scans of a given paper to choose from.

Answer 9

Some countries have library networks with a common catalogue that you can search to find out which libraries have it. This is what your favorite librarian uses to get the paper for you. However, it could also tell you, which friend or friendly lab you could ask.

• E.g. the German and Austrian libraries' journal data base will tell you which 6 libraries have the 1930 "Sprawozdania z posiedzeń Towarzystwa Naukowego Warszawskiego" where Łukasiewicz and Tarski published the "Untersuchungen über den Aussagenkalkül" and whether it is available via inter-library loan.

• A similar catalogue exists for books

(feel free to edit and add more such catalogue links)

Answer 10

The way that is most likely to succeed is to ask a librarian. This means actually asking a librarian, not just searching a library's online catalog yourself.

But sometimes you can figure things out by looking harder on google yourself. For example, google just one element of the citation information, and find where else it appears, or find other citations that include that phrase.

Here is an example from another question on this site:

I'm trying to find the following paper:

R.M. Wilson, “Decomposition of a complete graph into subgraphs isomorphic to a given graph.“ Congressus Numerantium XV (1975), 647-659. MR0396347 (53:214)

Also seen cited as:

R.M. Wilson, “Decomposition of a complete graph into subgraphs isomorphic to a given graph.“ Utilitas Math, Winnipeg (1976)

I cannot find it anywhere, online or offline. How am I supposed to progress here?

And here is my answer (it may be useful to look at all the linked documents in detail):

Based on similar references elsewhere, I think it is in "Proceedings of the Fifth British Combinatorial Conference". https://lccn.loc.gov/77361732

See reference 4 in https://hal.archives-ouvertes.fr/hal-02321683/document

or reference 1 in https://www.sciencedirect.com/science/article/pii/S016750600870654X

So now you just have to get this proceedings from a library.

I found this by googling "Congressus Numerantium XV" to find out what that was.

Here is the original answer: https://academia.stackexchange.com/a/185701/

Answer 11

Kindly ask all the authors that recently cited that paper ... at least one of them must have a printed copy or a pdf that can be shared privately with you.