Current scientific databases hold papers that were published over the past few decades (I haven't checked what is the oldest paper available, but I remember seeing papers from the 1960s, there might even be older papers than that which are cataloged in current databases - not talking about the principia...). One can come upon such old papers either when searching through a database or as a reference in a more recent paper.

My question is: how can I tell if a paper is too old such that I had better not rely on its findings while performing a more contemporary research?

(It probably differs between disciplines, so I'm looking for some general advice.)

  • 5
    Student's classical paper is from 1908 - nothing is to old to cite if relevant. Cayley's paper from 1859 is still useful. Etc. etc.
    – user68958
    Commented Jul 17, 2018 at 12:19
  • In my field, I still see people citing Gerhard Gentzen's paper on cut elimination occasionally! (FWIW that one's from the 1930s)
    – xuq01
    Commented Jul 17, 2018 at 15:08
  • Look how many modern manuscripts cite Fick (1855). Such old papers are fun to read and often appropriate in the introduction. Of course, often they don't need to be cited since they've become textbook knowledge.
    – user9482
    Commented Jul 17, 2018 at 15:12
  • 2
    Yesterday I read a computer science paper whose oldest reference was from the 12th century. I regulary cite publications from the 18th and 19th centuries.
    – JeffE
    Commented Jul 19, 2018 at 13:54
  • 1
    @JeffE: Care to share what you were reading yesterday? Commented Jul 19, 2018 at 19:04

3 Answers 3


The age of the paper per se should have no relevance. However, the researcher had better be certain about whether the conclusions of the old paper have been superseded or even invalidated by more recent work. This sort of problem is unlikely in, say Mathematics, but other fields can be different. There are Philosophy papers that, to my knowledge, have invalidated old work. The old work said "X is impossible". The new work said "X is indeed possible and here is an important example." This wasn't mathematical philosophy (logic) of course, but observations about learning (epistemology).

The fact that a paper hasn't been cited in 50 years is of little importance, especially in a narrow subfield of some larger field. The fact that it is no longer relevant is of vast importance. However, there are some studies intentionally done to compare the old with the new and here, citing both is the essence, of course.

One reason for not finding old papers is just the difficulty and expense of properly indexing them in databases. For "popular" things someone is likely to want to go to the trouble, but for arcane things, not so much as there seems to be little payback for the effort.

And yes, you normally don't need to go back before the Principia, but, you know, Euclid was an interesting guy.

Note that another question here asks about finding more recent papers than the one you are reading.

  • Thanks, that question is one of mine as well... ;-)
    – Don_S
    Commented Jul 17, 2018 at 15:27

how can I tell if a paper is too old such that I had better not rely on its findings while performing a more contemporary research?

It may be a little difficult to give a general guideline. However, a few points may come in handy here.

  1. If you look into most contemporary work, you could find some patterns in their references. In other words, may of them would have some very common papers which they cite. It is reasonable to know for what reasons these papers are cited by checking where in the text they are cited.
  2. The references of this contemporary work may also have some old references cited for specific reasons, for this reasons you may also cite these references.
  3. Citing very old references may be very beneficial if you are able to properly trace developments in the area over the years. However this may be a bit difficult as a lot of literature may have to be reviewed.
  4. Another thing is that, sometimes, the author may come up with something absolutely novel which changes a fundamental notion or aspect from those "old references". Such papers will draw the attention of the research community to a completely different finding which may impact on the domain for years to come.

In summary, in my humble opinion, it is not easy to say that papers from a particular year becomes absolutely obsolete for research today. As a general rule, based on my experience, new researchers are usually encouraged to extend their research from a recent (say now till five years back) publication. However, for some of the reasons mentioned above, there are circumstances where new values may emerge from those very old research papers.


Mostly, the age of a publication should not be an argument against citing it, as mentioned by Buffy and Abdulhameed. I agree with their answers but I'd like to add that for certain statements about state-of-the-art or recent developements, you absolutely need up-to-date references.

An example for why this is necessary is that it would make no sense to cite a 20 year old paper to show how widely used a particular software is used, because the information from the paper is outdated. You could, however, look for more recent publications citing this paper. That way, you'd get better evidence for whatever it is you'd like to show.


You must log in to answer this question.

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