A group of researchers work in a problem that no one else works on, hence these researchers are cited only by themselves. They cite other papers, too.

How does the academia consider these cases?

Important update:

From the answers received, I noticed that I have wrongly formulated the question. I don't care if that group of researchers cite other papers. My concern is what if they are cited only by themselves.

  • 6
    What is 'academia' here? There are different researchers, different universities, different funding agencies, and they may all have different views. Commented Sep 15, 2021 at 16:13
  • 3
    I don't think that helps: academia.meta.stackexchange.com/questions/4471/… Commented Sep 15, 2021 at 17:28
  • 10
    There are cases when a group of researchers work in a problem that no one other works on, hence they only citate each other and themselves. - I claim every researcher falls in such a group. Proof: Addend authors of any external citations, and repeat. This process terminates by finiteness.
    – Kimball
    Commented Sep 15, 2021 at 21:59
  • 4
    The question body suggests that the question is not really about self-citations. May I suggest a new title for this question: "If all of your citations came from the same small clique of authors, does that make you irrelevant in academia?" Commented Sep 16, 2021 at 9:35
  • 3
    @Captain Emacs: Number theory used to be exotic and idiosyncratic. --- When was this? Certainly in the 1800s it was one of the most researched areas of mathematics, and it's hard to think of a reasonably well known mathematician then (e.g. someone who might be among the 100 best known mathematicians from the 1800s) who didn't make some kind of contribution to number theory. However, I suppose it was a bit out of fashion from the 1940s to 1970s or so (the mainstreaming of general and abstract approaches to measure theory, functional analysis, lattices, modern algebra, topology, etc.). Commented Sep 17, 2021 at 18:13

10 Answers 10


One can see this from a philosophical perspective ("If a tree falls in the forest and nobody hears it, did it fall?") and one can also approach this from a perspective of "All research is useful, it may just not be useful right now (as @Buffy has in their answer).

But the reality is that if nobody cares about a certain research direction, then that is a statement about the work as well: It is, truly, irrelevant for all practical purposes. Of course, it is possible that at some point in the future someone might start to care, but the likelihood of that happening is rather small even if I am sure that we can come up with examples.

In many parts of academia, one measures the influence of researchers by their number of citations. It is certainly debatable whether that is an accurate metric, but it is a useful metric for sure and if someone's work does not receive any outside citations, that is a meaningful statement as well.

  • 11
    -1 for "it is a useful metric". It is a metric that is used, that doesn't make it useful. Commented Sep 16, 2021 at 9:53
  • 25
    @JackAidley I disagree. It is used and it is useful. Citations counts in a given field are a measure ( maybe not the best ) of the impact of a researcher. However, comparing researcher across different field using citation counts is dangerous since different fields have different citation habits. See this paper ( of which I am a co-author of ) for a quantitative analysis of this : doi.org/10.1016/j.heliyon.2017.e00300
    – BlaB
    Commented Sep 16, 2021 at 10:20
  • 8
    @JackAidley I'd cite Strathern's generalization of Goodhart's law here. Number of citations is a reasonable (if noisy) measure of researcher influence. It's a shitty target.
    – R.M.
    Commented Sep 16, 2021 at 12:24
  • 12
    @JackAidley it is certainly over-used and, in some quarters, even abused but that doesn't make it useless. It is useful, it's just not as useful as it is made out to be.
    – terdon
    Commented Sep 16, 2021 at 12:46
  • 5
    Can we stop the discussion here? Surely everyone here pointing at individual papers (CRISPR or Higgs' paper) understands that the exception does not imply that all papers garnering zero citations will at one point become important. Furthermore, what's the point of looking at history: hiring and promotion decisions need to be made now, and pay raises have to be decided now as well. We don't make these decisions 50 years from now retroactively, and citations are useful pieces of the larger puzzle in making these decisions today. Commented Sep 17, 2021 at 17:29

It is not unusual for some researchers to receive few or no outside citations, but it would be strange for a researcher or research team to give no outside citations. That is because it is always possible to place one's own work in some academic context based on existing published literature. To not give any citations outside one's own group would suggest that the group is not very good at doing a good literature review of their topic area (and it would probably lead their papers to get rejected from journals, so it is unlikely to show up in published literature).

The remainder of my answer pertains to the more common case where a researcher receives no outside citations. One would ideally wish that the relevance of an academic paper is determined purely by the topic, substance and quality of the work. However, as a practical matter, literature that has fewer citations (particularly fewer external citations) is harder to find/notice when searching for papers, and literature that only has self-citations does not show up in outside papers by other authors in the literature review. This makes it less likely you will find the paper with a search engine, or using a "track-back" search where you start from existing known papers and then expand out via the citations. Consequently, certain papers can get caught in the academic equivalent of a black-hole.

There is of course nothing inherently wrong with self-citations, and they are a natural part of what happens when researchers write multiple papers with overlapping topics. A lack of outside citations means that a work has not generated interest from other publishing researchers, and this is unfortunate. It can have the practical effect that a paper essentially gets forgotten, particularly if the journal is not highly visible. Some high-quality papers do not get outside citations because people just don't notice them initially, and they end up being "forgotten" (i.e., not showing up easily in literature searches) and become more remote as the journal issue they were published in ages.

To respond to your title question, while it is possible that papers may essentially be forgotten if they fail to generate outside citations (and therefore become irrelevant in terms of influencing the profession), that does not mean that the researcher is "irrelevant". Academics perform teaching and service duties to the profession beyond our research, which are valuable professional outputs. We are also people with lives, families, etc., making us highly valuable and relevant to many people. If an academic is not successful at generating outside interest in their research, they are not "irrelevant".


Actually, it depends more on the work itself than the fact that it is only self-cited. But every truly new topic could have that characteristic for a while. If others are interested in the future, that will change.

But if a small subfield over a relatively long (whatever that means) time period fails to attract interest it will just disappear. Someone may resurrect it, of course.

Thinking about mathematics in particular, if someone came up with a new idea that depended only on "common knowledge", say, as published in text books, then there would be no need for citation to early work.

If no one cites you, it is, perhaps, harder to find your work since chasing backward citation links is an important way to discover things. If you are publishing in good (highly visible) journals, that may not be an issue in the short term.

But even things that make a clean break with the past (Einstein/Relativity) tend to cite things pointing out the differences and why the new thing is correct.

  • 12
    "If no one cites you," a lot of this is (research) cultural. A lot of papers on "Deep Learning" fail to cite relevant works on neural networks published in the 1980s and 1990s and there is a lot of reinventing of wheels. I suspect a lot of that is because it isn't on ArXiv. Similarly there is a lot of "common knowledge" that has been lost because of the way research in that field is done these days. Research can also be cyclical and things get lost between cycles. Commented Sep 15, 2021 at 18:53
  • 8
    @DikranMarsupial Not only in Deep Learning. It;s always amusing to find a paper on computational mechanics where the authors have rediscovered something that was in standard textbooks published in the 1950s.
    – alephzero
    Commented Sep 16, 2021 at 0:22
  • 10
    @DikranMarsupial Perhaps someone should work on an AI to index research on AI. Although I suppose there's probably already been work done on it, I'm just not aware of it. Commented Sep 16, 2021 at 2:30
  • 17
    This isn't anything new by the way, the same thing happened in the "old days", the classic "back propagation" algorithm for neural nets was independently discovered at least half a dozen times over the space of 15-20 years before it was discovered at the right moment. Back then it was more difficult to find papers because they were not easily searchable, these days it is difficult because there is too much to search. Commented Sep 16, 2021 at 6:31

I am of the opinion that no problem really exists in isolation.

For every problem, one can identify at least three areas of interaction:

  • The problem itself
  • The methods that are being used on the problem
  • The potential areas impacted by solving the problem (intellectual or practical)

Even if only a small community cares about the problem per se, the tools available to address the problem are changing, as is the context of the environment in which the problem is being solved.

Even for highly theoretical mathematics, the problem does not exist in a vacuum: something caused the researchers to become interested, they are working on it in a world full of evolving potentially relevant methods, and something is causing them to stay interested despite the fact that they haven't solved it yet.

As such, I would be highly suspicious of a group of researchers who only engage in self-citation. They may have come to that point honestly, but it would still indicate a severe level of intellectual detachment from the rest of the scientific community.

Bottom line: even if nobody else cares about their work, they should care about other people's work enough to cite them.

  • Thank you for this answer. You have no idea how important this answer is to me. I have been working on a problem for decades. I have partial solution, not a full one yet. I want to get to the bottom of it, but my health is degrading (I am getting old). I kind of deciding to give up my research altogether. After reading this answer, I changed my mind. I'll take my doctor's advice and wait until I can recover to a point so I can continue my research. Thanks again.
    – Nobody
    Commented Sep 22, 2021 at 9:07

To reiterate an important point already made in other answers, but, I think, deserving emphasis: it's not so bad to cite oneself, considering that (duh!) one's prior work could easily be background for one's current work, but to not cite anyone else is crazy... probably wildly inaccurate, etc. At the very least, setting the context by mention of prior art is a significant part of the interest of a new idea for me. To implicitly claim that there was no prior art is implausible, doubt-generating, and, at least, unhelpful.

I do confess that I did not understand any necessity of acknowledging prior art in my earlier years... Somehow the idea I'd gotten was to pretend to have the least indebtedness possible to other people. Well, this turns out to be ridiculous, in many ways.

So, again, the most silly/stupid aspect of only self-references is that it's surely just not accurate at all. Then, beyond that, ...

(So if I see a paper/preprint with all references to the authors or a limited circle of buddies, I am immediately suspicious...)


These days, when people judge researchers (especially hiring/grant committees, tenure boards, etc.) it is common to place much emphasis on impact. Impact is measured in a couple different ways, but all of them share the common idea that more people citing you is better.

If you only ever cite yourself, your citation count is 1. That is only one step up from the worst possible count, 0. So the debate of you citing your papers vs. the one other guy in the world who cares, is kind of dwarfed by the elephant in the room - namely that you're at the bottom. What difference does it make if you finish 99th or 100th, when only top 3 get a medal?

But given the obvious perverse incentive, usually self citations are not even counted as a full citation. Obviously you citing yourself is less interesting than a stranger citing you. In most modern citation metrics, self-citations are either scaled back with a factor, or disregarded entirely. When humans calculate an "informal impact factor" by Googling your papers, their brain will naturally apply such a factor as well. So it's like not only did you finish 99th, but you're tied for it as well.

In sum, I don't think anyone will care about whether you were cited by yourself or not, they will have already stopped caring when there is only 1 person citing you.

Incidentally, if you are the only one talking about your work, then you truly are irrelevant in the literal sense - nobody is finding you relevant to what they do. But I feel like that's putting it a bit harshly. Just because nobody has found your research relevant yet, doesn't necessarily imply that your research is inherently worthless and cannot possibly have bearing on anything else. It's just a heuristic after all. But that said, the academic community appears to be invested heavily in citation metrics for the time being, and as a result, having only self-citations will not impress many strangers.

There are some exceptional cases where I think a situation like you describe might be understandable:

  • A very junior researcher for whom just publishing a paper at all is an accomplishment
  • The self-cited paper is in a very prestigious journal (although such journals tend to try very hard to eliminate papers that aren't likely to get cited)
  • The paper itself has extraordinarily impressive content, despite the few citations
  • The person evaluating you is himself in such a "thin" field, himself has few cites that are non-self, and himself appreciates the difficulty of having a high citation count in your field
  • The paper is very recent (<a few months) so it is plausible people are planning to cite it, but are just taking a while

But in my opinion, the effective solution here is not to split hairs, but instead to either broaden your work to have more relevance to people outside your narrow field, or start thinking about moving on to an allied field with broader appeal. That, or just accept that the world is not yet ready for your ideas, I suppose... Such is life.

  • Note that, in research quality evaluation in the UK, the word "impact" means the beneficial effect of academia research outside academia, which no-one tries to measure using citation counts. Commented Sep 17, 2021 at 12:47

In our tech-enabled, Alice-through-the-Looking-Glass Red Queen running faster and faster world, we keep on focusing on "impact" within a shorter and shorter timeframe. How many downloads did my paper have last week? How many citations did we get in the year since our last annual evaluation?

It's easy for genuine knowledge accomplishments to take their own sweet time, to sit on the shelf for quite a while before someone else integrates them or applies them to something else. Conversely, other ideas may flash in the pan, have tons of citations straight away, and then decades later be passé or just obviated since some other train of thought is more useful or broader in applicability. My dissertation in the 1990s made integral use of a result from the 1950s that had only one other citation, of the throwaway dead-end form "Somewhat similarly, X has shown that ..." (that nevertheless led me to it). And then my dissertation result was obviated by a more general approach by someone else a decade later.

So I would make no value judgment on a body of research that is ping-ponging in a very small enclosed room, even among its authors only, for even a number of years. I would make a value judgment on researchers who can't articulate why it could/should be of any interest to anyone at some yet-unknown point in the future -- even though I respect some purists would argue academic independence should allow "but I [researcher] find it interesting" should be reason enough.

  • This answer cuts to the heart of the matter. In mathematics, it is not uncommon for theorems to go unused and uncited for generations, till practical uses are found for them after the fact. I know of several occasions where physicists have developed new hypotheses that were validated by the discovery of old, obscure math theorems (I believe Yang-Mills theory was one, but don't quote me, that's not my field). Having ready-made bodies of research waiting on a shelf before the uses come along is not entirely wasteful, by any means. Commented Sep 18, 2021 at 15:56
  • 1
    Also, entire subfields are sometimes consciously ignored by because they throw monkey wrenches into dominant paradigms; scholars are people too and sometimes their shortsightedness (or even outright prejudices) retard academic research. A recent case in point was the initial suppression of immunotherapy research; there have been plenty of other instances in scientific history, in keeping with the pattern defined by Thomas Kuhn in The Structure of Scientific Revolutions. Commented Sep 18, 2021 at 15:58

The Nihilist Take

Your question seems to imply that relevance is absolute. It's meaningless to ask whether something is relevant without asking to whom it's relevant.

To answer your question directly, an isolated group of researchers whose work is never cited by anyone outside their circle is irrelevant to the rest of academia. But this is just a microcosm of academia as a whole. To non-academics, academia is irrelevant until their research has some practical application. On a cosmic scale, the endeavors of the entire human species are irrelevant. Ultimately, nothing matters. Look inward for justification and meaning, not outward.


Definitely not. The fact remains that, as long as you cite yourself and the citations are all relevant, you are not just ok, but you are progressing.

And it also happens that you may be doing research on a topic where not many people are active, except of course you yourself. In that case, it is but natural that your work will be full of self-citations.


Citation counts and origins are a poor measure of relevance.

If you are the only academic working on the history of basket weaving in northern elvish communities, you’re unlikely to be cited by academics working on dwarvish basketweaving, and unlikely to get lots of citation for that work, even if it’s highly relevant to northern elvish communities.

Contrarywise, if you publish a controversial theory - say old elvish basketweaving resulted in basket of greater strengths than dwarvish or hobbit baskets - you might get cited quite a bit, even if your work is irrelevant and possibly even wrong.

There is also the problem of timescale. The work of John Bell (of the inequality fame) was treated as “irrelevant” for a long while before it’s importance was recongnized.

  • It actually took quite a while for Einstein's work on relativity to be recognized, since the luminaries of the day dominated the scene with (eventually) failed ideas.
    – Buffy
    Commented Sep 18, 2021 at 16:49

You must log in to answer this question.

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