Good citation practices?

In our newest "how can we help you?" thread, a reader asks:

What's the purpose of citations in papers?

I'm having trouble determining the line between over-citing and losing my tone (and risk just saying what other people are saying) but also under-citing (but feel like I'm doing something wrong).

I want to make nods to the literature and earn my way into a conversation through citations - but often, there just aren't people saying exactly what I'm saying and I lose my reader when I try to manipulate citations to fit my project.

Any thoughts would be helpful on people's views of citing other work in philosophy papers!

Good question. I've made my own views on this clear before: I think (A) the purpose of citations is to provide an accurate scholarly record of relevant work that has appeared before yours on a given topic, and (B) under-citation is a bigger problem than over-citation.

Against (A), I've heard some say that they only cite works that actually influenced them. However, I think this clearly misunderstands the purpose of citations, as it presupposes that there are no normative requirements on what constitutes sound scholarship. If one doesn't read the literature and you then publish a paper saying, "I'm the first to give argument X" when other people have given that argument before, but you don't cite any of it because you didn't read it and it didn't influence you, you've done something wrong. You've misrepresented the scholarly record, presenting yourself as providing a novel contribution that isn't novel (you've reinvented the wheel). This is why, in all scientific disciplines, one is expected to begin with a literature review. Yet, in philosophy, I've routinely seen scholars (both junior and senior) give examples on social media of their work not being cited by someone who gives a very similar argument several years later. I've also seen cases as a referee where people haven't cited anything after the year 2000 on a topic, even though there's a ton of work relevant to their argument. Although one obviously cannot read everything, one does have an obligation to be generally aware of what is out there and do due diligence in determining whether someone has given your argument (or something close to it) before. Of course, mistakes may be made, and you may miss something. But if your work cites widely, you've at least shown good-faith effort--whereas if you only cite a small number of sources and a simple PhilPapers search reveals that some of the top results for a keyword search on your topic provide arguments similar to your own that you don't cite, then I think that's a problem.

But these are just my thoughts. What are yours?

Comments

[benji]

I suspect the challenges the poster reports have to do with how their papers are constructed rather than with the number of citations. If the poster feels like some of their papers end up with too much summary of others, the paper may need to be restructured. You can (and usually should) write papers centering your own argument while still citing a lot of other literature. In my own papers, I often have a line in the introduction where I cite, to the best of my ability, everything that has been previously written related to the topic. (My papers tend to have niche topics; obviously in some fields you cannot cite everything but still aim for a representative sampling.) The rest of the paper, however, only discusses some of those cited works in detail, namely those that most directly inspired my own argument or alternative views that I argue against directly. I do philosophy of science so I also cite scientific papers when I make a non-obvious factual claim or present a case study. So my papers end up containing a large number of citations although only a few of them receive extended discussion.

If in contrast you find that you have to "manipulate" citations to be relevant to a paper, then perhaps the work is not sufficiently integrated with existing literature, which is likely to present challenges with getting it published. Consider ways of framing your argument so that it engages with some existing philosophical literature. For instance, can your position be claimed to have implications for an existing philosophical interest? Can it apply a method, insight or formalism from another area of philosophy? If so then you can (should) cite that literature, even if your work is in a different area of philosophy. You can also cite work just to explain what about your paper is new and different. For instance you can say things like, "While (Citations X Y and Z) have written about rabbits and (Citations A B and C) have explained chocolate, few works have accounted for the relation between rabbit and chocolate." Or "(Citations D E and F) have analyzed chocolate but no one has yet accounted for the case of rabbit shaped chocolate as done in this paper." These kinds of statements signal to the reader that you understand the scope of the existing literature and are also helpful for people who want to find more information about the topic.

[Daniel Weltman]

I agree with Marcus and benji. I will add that I think OP is misunderstanding who one ought to cite. You should cite more than just the people who are saying what you are saying. You should also cite people who say things relevant to what you are saying, e.g. people who answer the same question but in different ways, or people who provide a similar answer to different questions, etc.

Similarly, you do not cite in order to earn your way into a conversation or to make nods to the literature. You cite in order to situate your view in the context of what others have said that is relevant to your view.

This is all pretty much just a restatement of Marcus's first point. But, hopefully putting it in a way that is in direct conversation with what OP said makes the point clearer.

[Hannah]

I think we have to also recognize that some secondaries are really bad and insignificant. You could cite them with the comment they are bad, but then you would have to decisively prove they are bad. If it's just a matter of a very poor reading of a source, it feels unnecessarily mean to slice them to pieces, plus also a waste of your precious space in an article trying to advance your field in more constructive ways. Also, you risk starting a polemic with the bad source, which wasn't very important before and now becomes enshrined in the canon, and suddenly the whole secondary field will be citing a major "debate" over work which is frankly not worth reading. If the bad article very popular or written by someone very influential, perhaps it's worth doing. But I would say that barring those cases, it is not unreasonable to allow certain work to be phased out as more appropriate resources become available, instead of valorizing its badness by arguing with it.

Thus: one should absolutely cite the secondary literature, but it is appropriate to set aside articles that do not match a certain standard of scholarship or spirit of critical, careful analysis. This may not be as much of a problem for more mainstream philosophical areas, but in sparsely populated subfields the refereeing is not as strict and this may be more of a risk.

[Literature situation]

Also site to situate your work in the literature. One person says this, another this, yet another this, but I say distinct about related point… against this argument and that…

[an opposing view]

FWIW, I think most people cite way too much. I also think "situating oneself within the literature" is a fairly easy route to a paper that takes way too long to get interesting.

If you borrow from or build upon the work of someone else, then of course you need to credit them (simply writing on the same general topic that someone else has previously written on doesn't count here). And if someone has some idea that is important enough that you wouldn't be doing right by your reader if you didn't talk about it, then talk about it (and a mere citation doesn't count as talking about it). Similarly, if your paper doesn't make sense without the work of others, then use that work when framing your paper (again, a mere citation doesn't satisfy this, either).

Beyond that, write your own paper and don't clog things up with barely-relevant citations.

[Daniel Weltman]

"an opposing view" says that there are some ideas that are "important enough that you wouldn't be doing right by your reader if you didn't talk about" them. This is correct.

What "an opposing view" also thinks is that there are not as many ideas that are important as most people seem to think (because most people cite too much). I think this is incorrect.

I think "an opposing view" has too high a bar for importance. I don't know what criteria you're using for importance, "an opposing view," but I think it's better to have a bar low enough that you think most people in philosophy either cite enough or not enough, rather than too much.

Possibly some disagreement here is being driven by "an opposing view's" view that "a mere citation doesn't count as talking about it," such that if you can't engage in substantive discussion, there's not any point in citing at all, and so if you cite something you sign yourself up for extensive discussion of it. If this were a real requirement then one would want to have a high enough bar for importance such that one's paper does not get bogged down in too much discussion of others.

This might also explain why "an opposing view" thinks situating oneself within the literature tends to make a paper take way too long to get interesting. Certainly it might do that if every paper you discuss requires somewhere between a paragraph and a page (or more) of extra writing.

But, you don't need extensive discussion of everyone you cite. Sometimes, especially when situating yourself in the literature, you can do this pretty quickly in a few sentences that cite many people, as discussed by benji and "Literature situation."

So, one cannot raise against the "cite a lot" view the worry that this bog down a paper. You can cite 30 papers in a couple sentences, if 15 of them endorse one view and 15 endorse another.

[Ten-Herng Lai]

I wrote something here a few years ago.

https://philosopherscocoon.typepad.com/blog/2021/07/dealing-with-undercitation-in-philosophy-guest-post-by-ten-herng-lai.html

The comments are also really good.

I'm still a huge fan of citing, and only cut citations due to word limit. That being said, I might have done it too much occasionally. Once a reviewer recommended that I put a block of citations in a footnote.

[Ten-Herng Lai]

I think this paper is a good contribution:

Lundgren, B. (2024). Is Lack of Literature Engagement a Reason for Rejecting a Paper in Philosophy?. Res Publica, 30(3), 609-616.

The paper talks about not citing per se isn't a huge problem and shouldn't lead to a rejection, but there are very good reasons to engage with and cite relevant literature.

[an opposing view]

@daniel weltman. What do you take yourself to be conveying when you cite 30 people in a couple of sentences? How do the 30 relate to what you're doing in the paper?

[stickler]

Citing practices vary by sub-speciality. I publish in philosophy of science, and some papers may only cite 20 papers and books. But I also publish in the history of philosophy and those papers usually cite 40+ papers and books. One key principle: DO NOT CITE something you have not read. So the bar is high if you plan on doing a comprehensive literature review - you have to read all the papers you cite.

[Daniel Weltman]

@an opposing view: I am conveying whatever I say about the 30 in the sentence. If for instance 30 people endorse Xism because they take Xism to be parsimonious, and I say as much in that sentence and follow it with a citation to 30 people, then I am conveying in that sentence and citation that these 30 works contain arguments for why we should endorse Xism due to reasons of parsimony.

@stickler: is reading 40+ papers and books really a high bar? If you read a paper a week and a book every month or two, then every year you read more than enough to write a paper that cites nothing you've ever cited before. Given that it's quite common to write papers on topics similar to what you've written before, 40+ papers and books per paper you write sounds quite reasonable, assuming there's that much written on the topic.

[conciliatory]

I feel that the "two camps" above are not contradictory, since they are both describing widespread problems with citations. Indeed, I feel that the problems are two sides of the same coin.

There is the widespread problem of under citation. Many people do not try hard enough to dig out all the relevant literature, and sometimes even fail to see the relevance of an entire literature, e.g. metaphysicians of time ignoring philosophers of physics on the same topic and vice versa. I also frequently found myself not cited by papers with the same topics while they cite people who make no more relevant arguments.

There is the widespread problem of over citation. Many people cite just because the title/abstract look vaguely relevant & famous people & people in their circle & potential referees, without actually reading the papers and judging the quality and relevance. I often found myself cited In irrelevant places with zero discussion, and there would be zero loss without the citations.

Sorry that the above comment turns out more of a personal anecdote than anything else:)

[Tim]

If your citation dump is less helpful than just putting the sentence before it into google scholar, I’d prefer your paper not include it.

[an opposing view]

What Tim said.

[Daniel Weltman]

Tim, I think you must have a deep misunderstanding of how Google Scholar works. It is not a magic tool that just finds all the relevant work. It works in part via checking who cites who. If you do not cite certain things this will have effects in terms of what is returned by a Google Scholar search. A decade's worth of nobody citing anything that would be turned up by searching Google Scholar for the relevant sentence will produce a decade's worth of literature that might not turn up when you search other relevant sentences. And then Google Scholar will not be a reliable tool anymore.

Moreover, I think your approach is a really dangerous one. Google (the plain version) used to be a useful search engine. It no longer is. It now inserts AI junk into the results. There are ways around this, but Google may phase those out eventually, and insert even more AI junk, etc. Google Scholar may not always return the results it returns. They may change the algorithm in the future.

The entire edifice of intellectual progress in academia once rested on people citing each other at least conscientiously enough for someone to be able to follow the breadcrumbs and figure stuff out. Methods were developed to make this easier: indexes were compiled and published, bibliographies mimeographed, etc. This has all been rendered more or less obsolete by the Internet in its current form, but we have no reason to expect the Internet in its current form will exist forever. For the sake of people, say, 5 years down the line when Google decides to let AI find your scholar results and it spits out papers that don't exist, please cite stuff.

[an opposing view]

@Daneil Weltman... I think you have a different conception of what citation is for. I don't see myself in writing a paper on X as also providing a historical record of all the discussion of X. I don't see myself doing that *at all.* And given I don't see that as one of my jobs, I don't see any reason to cite in the way you describe.

As for your claim about the what the edifice of intellectual progress in academia once rested upon ........ I think you're quite wrong about that.

[Bibble]

1) Cite things that give building blocks to your current argument.
2) Cite presently-accepted things that your current argument will put in question.
3) Cite things that explicitly or implicitly offer challenges to your current argument's new ideas - then deal with those challenges.
4) Optionally, cite things that can be extended or improved by combination with the novelties of your current argument, to demonstrate its usefulness.

Done properly, this should require a lot of citation - much more than most humanities work currently does.
But I take my four above to be an exhaustive list of the reasons to cite anything. I don't see what argumentative purpose "cite everything previously written on the topic" is supposed to serve.

[Citations as Recommended Readings]

There may be many, many papers about one kind of take in topic X but I will not try to cite every single paper there unless there are only a small handful because I suspect non-specialists, if they’re interested in looking at this more, will want to find a good one quick when looking at footnotes. I think that philosophy compass and SEP articles of course should be more comprehensive but for original articles, I think that that should not be part of the burden.