Norms for writing, research, and citations

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

It is a probably basic question about academic norms. When you write a paper, concerning any train of thought (claims, arguments, etc.), should you simply think about it on your own and write it out as your own plus a reference to similar work if you do find such work, *or* search the literature and try to find if someone else makes similar claims, arguments, etc., and then simply explain this person's idea in your paper, not claiming it to be your own even if you come up with it yourself? I intend this to be a very general question---the choice in question can occur any number of times when writing a paper. Thanks in advance.

This is a great query, and it echoes a set of questions raised by NK in this thread a few months ago:

I wonder if it might be worth starting a more general discussion here about the *point* of citation. I, for one, would be interested to hear others' views on when, and why, one *ought* to cite a given work.

One suggestion made above is that you ought to (try to) cite everything relevant. Another (to which I take Helen to have objected, though the following is an attempt to describe it a bit more charitably) would be that you ought to cite relevant work that you think is good/worth reading and engaging with. These obvious differ in at least one important way: on the latter proposal, citations would be prestige-markers; on the former, they wouldn't.

Are there other options? And what is there to be said for and against each option?

I'm also curious what others think of the following sociological hypothesis: when you cite "low-prestige" work (whatever exactly that means), you thereby lower the perceived prestige of your own (i.e., the citing) work, in the sense that others are going to be less likely to read your work and/or take it seriously (if/when they take note of the character of your citations).

N.K.'s last point (that citing "low-prestige work" can lower the perceived prestige of your own) is one of the main reasons that I think authors are under an obligation to do some basic research--via doing some simple searches for key words in PhilPapers and Google Scholar--before writing, and to cite and/or discuss explicitly up front any relevant precursors (including similar arguments) to the one they intend to write about.

After all, suppose you know that citing low prestige work is likely to undermine the prestige of your work (and, say, your claim to originality), and you operate according to the rule of only citing work that you think is "good" or "worth reading." In that case, here is something you might very well do: you might systematically avoid reading the work of "low prestige scholars" because "their work is not good or worth reading", but then publish ideas similar to theirs (and take credit for those ideas) because you didn't bother to read them. That's messed up, and it is (I think) a very real conflict of interest that researchers can face. We are not just thinkers: we are supposed to be scholars, and sound scholarship requires being aware of and engaging with what has been published on your topic of interest before you publish. For example, imagine that Einstein (a lowly patent clerk) published his 1905 paper on relativity in Annalen Der Physik, but then a more established German physicist didn't bother to read physics journals and published basically the same paper in 1908, taking credit for the discovery. Then suppose that because of their more prominent stature, everyone cited the latter researcher rather than Einstein. I hope we can all agree that this would be pretty messed up! It's something that norms of scholarship and publication should aim to avoid.

Of course, this raises the more difficult question of what sound scholarship requires. It's impossible (and undesirable) to read and cite "everything," as that would result in reference lists of infinite length. So, following N.K.'s query, how much is one obligated to read, be aware of, and cite? Here, I go back to my first rule: at a minimum, you should do simple PhilPapers and Google Scholar searches for key words relating to your argument--and then, if there is anything like a Stanford Encyclopedia of Philosophy entry on the topic, you should be aware of material there too. Beyond that, my attitude is: we should do our best, and give a good faith effort to try to find (and cite) relevant material that predates your own--including if you come across it while writing. If, for example, you find only after writing a rough draft of your paper that someone did defend a similar idea before, you should go back and cite them explicitly in your introduction, giving them credit rather than claiming full originality--and then distinguish your argument from theirs (as arguments are rarely identical in all respects).

Finally, I would just add this. None of us are perfect. We all make errors, and accidental omissions are probably inevitable. Yet, this is, I think, just yet another reason why academic disciplines should move toward an open, online, crowdsourced approach to peer-review. It's hard for a single author to do sound scholarship all on their own--and referees are often helpful in pointing out omissions. The more "referees" a paper has (as in an open, online form of review), the more likely errors of omission are likely to be recognized and corrected before journal publication.

But these are just my thoughts. What are yours?

Comments

[Michel]

I agree entirely.

I'll add that I don't worry about citation conferring prestige to my work--I worry about the quality of the work, and of its venue, doing that (and I've done pretty OK publication-wise!). If I come across written work with obvious citation gaps, that counts against the work's worth and "prestige" as far as I'm concerned. By contrast, I _do_ notice when the work I read points me towards work I've never heard of, and that burnishes the work in my estimation. It also makes me more likely to cite the paper pointing to such work, because it's that much more useful a resource for anyone reading up on the topic.

When I read student work, I check the list of works cited first, because that tells me a lot about the kind of paper they wrote (and whether they followed my instructions). I do the same when I referee, for similar reasons.

I recently refereed a paper with a reference list of just five items--three classic "big books" from the 1960s/70s, one paper from the mid-1990s by a major scholar, and a paper from the early 2000s by another major scholar. That's a _huge_ red flag in and of itself. When I read the paper, it turned out to be on a topic that's received a fair bit of attention in the last ten years or so, including some books and articles which have won major prizes. As a result, the paper was retreading ground already well-covered by recent work, presenting it as novel. You really can't justify ignoring all that. I guessed it was a paper submitted for a grad seminar that focused on the foundational stuff from the 60s and 70s, since that seemed the only way you'd get that kind of reference list for this topic. I recommended rejection.

I've since discovered the identity of the author, through a presentation. In fact, they're a major scholar in the field, and one of those five citations was a self-citation (it was perfectly apt, but it goes to show that the research process was worse than it looked, since that's 20% of the cited material!). IMO, that's inexcusably sloppy. Sometimes, you really are one of just a handful of people who've written on something. But you should take the time to make sure that's the case, rather than simply trusting your gut.

As for the original question, there's a balance to be struck, and it's a matter of finding your professional voice. I'll just add that you have to become adept at identifying which kinds of work and ideas need only be acknowledged (as in a reference), and which need to be engaged with in some fashion (cursorily, substantively, etc.).

[Daniel Weltman]

Philosophers in general cite each other much less than scholars in most other fields cite each other. This is, I think, bad, and to remedy it, most of us should be citing more (although I realize that many of the reasons people don't cite are hard to fix, like word counts in journals - I'm definitely guilty of deleting citations to more peripheral but still useful stuff because I need to get my paper under a certain word count).

We can't do this if we don't find and read the relevant stuff in the first place, which means if we do the first option that the question asker describes, we won't be heading in the right direction. Instead we should do the second option: if you're going to write about something, at some point (it doesn't have to be before you write down your own ideas) you should find out who else has said similar stuff and cite them, either by attributing the idea to them (if it turns out they thought it up first) or by describing how your view and theirs differ (if they have a similar but still distinct view).

In general, most ideas you think up are likely to have been thought up by someone earlier. Being a good researcher entails finding that stuff, or at least making an effort, rather than doing something akin to the process Marcus describes in the Einstein case. I agree with Marcus that it's not always possible to do this perfectly and that to some extent this would be fixed by a crowdsourced peer review process. Sometimes I've had peer reviewers point out relevant literature I missed, and sometimes I've done this during peer review, but having just a few reviewers helping you catch the missing stuff is very unreliable.

[Citation]

That really depends if you’re writing in something you have expertise in or specialized in. I’m assuming if you specialized in that area, you can recall the author(s) who wrote something similar almost instantly. Citations should be easier if you’re an expert in your subfield. But if you’re writing in a topic that is outside of your subfield, then you may struggle a lot since you can’t recall any literature you haven’t read about regarding that topic. In such a case, you might want to read more to avoid getting rejected for not engaging the literature enough.

I often find myself struggling when to cite if I’m working on a paper that is outside of my usual interest or competence. But in my usual one, I can recall relevant authors instantly. Common knowledges do not need to be cited FYI. If you keep your papers narrow, then you also wouldn’t need to worry too much about citations since you should have sufficient knowledge of the literature already.

This is one reason why I like books and papers that gives a broad and impartial literature engagement. It helps the reader to research easier and faster having secondary sources.

[IncentivesMatter]

Great discussion here, one thing I think worth remembering and emphasizing is that, at least in philosophy publishing, the *incentives* are set up by (most) journals to *not* cite widely as well. For reasons beyond my comprehension, many journals still count references in the word count for either a hard or soft word limit. I think this is particularly harmful for philosophers doing interdisciplinary or otherwise empirically-engaged work, e.g. philosophy of psychology, moral psychology, philosophy of physics.

[crowd]

Thanks for the very helpful and intriguing discussion. I want to especially second Marcus's suggestion for a crowsourced peer review. I think we are not only bad in finding citations but also misread work (by obscure authors). I am certainly myself guilty. But I frequently find myself misread by others too (often not in print---my citation counts are still extremely low).