Citation Practices

by Brian on June 6, 2007

In a recent “post about citing papers on the web”:, Ross Cameron drew the following conclusion.

bq. I’m tempted to think that if you put a paper up on the web, that’s to put it in the public domain, and it’s no more appropriate to place a citation restriction on such a paper than it is on a paper published in a print journal. I’m even tempted to think that conference presentations can be freely cited; i.e.that I shouldn’t have to seek Xs permission to refer in one of my papers to the presentation X gave.

The particular issue here is what to do about papers that the author posts and says at the top “Please don’t quote or cite”. (You occasionally see ‘don’t circulate’ as well, which is a little odd.) I’m not sure how common these notes are outside philosophy, but they are pretty common on philosophy papers posted on people’s websites. Now on the one hand, there is something to be said for following people’s requests like this.

On the other hand, as Ross notes, the requests can lead to annoying situation. One kind of case is where the reader notices an important generalisation of the paper’s argument. Another case is where the conclusion of the paper supplies the missing premise in an interesting argument the reader is developing. Either way, the reader is in a bit of a bind.

I think the main thing to say about these situations is that writers shouldn’t put such requests on their papers.

When you circulate a paper, either informally or by publishing it somewhere, two kinds of good things can happen. First, good things can happen to you, either by people offering you suggestions for how to improve the paper, or increasing their opinion of you because it is such a good paper. Second, good things can happen to the profession, because your paper helps advance the field in certain ways. Given the dynamic nature of research work, that advance consists largely in improvements that we see in other papers that cite the work. Now if you circulate a paper but bar citation of it, you’re basically getting the good consequences for you, without allowing there to be good consequences for the field. (Or, at the very least, you are getting the good consequences now while delaying the good consequences for the field.) This seems, to put it mildly, unjustifiably selfish, and it’s very hard to see a moral justification for it.

It’s also hard to see what exactly the costs of being cited are. It would be annoying to have a journal publish an article critiquing yours before yours came out. But unless you are rather famous, and the paper has already become quite well known, journals aren’t going to publish such articles.

A better reason perhaps might be that if mistakes in the paper are spotted, you want the chance to fix the paper before it goes into print. But other people citing the paper doesn’t prevent that. There isn’t any obligation on you to publish the first version of a paper you post to a website. So if you say p, and someone else writes something that shows you are wrong, but you can say p’ instead which does just as well in the context of the paper, you of course can say just that. It might be a little odd for the citer if your published paper doesn’t make the mistake that they cited it for, but that’s just a risk people take when citing papers off people’s websites.

There is, as was noted in the comments thread over at Ross’s, a rather tricky scope question when someone leaves such a request. Presumably it is OK to quote/cite the paper in some forums, e.g. on an email to a friend, or while txting. In practice, few people would say that you shouldn’t quote or cite it on a blog. (That’s what blogs do, right, they cite stuff that appears on the internet.) What’s really just being ruled out is citing it in print. But it is a little odd that to think that it’s OK to cite a paper on a high-profile blog, but not in a low-profile journal. Some situations in academic life are just odd, so that’s not a reason to ignore the request. But it does make it even stranger why someone would request this.

One last thought. I didn’t understand the ‘even’ in Ross’s comment about conferences. I’ve always been under the impression that presentations at conferences are in every respect public performances. What you say there can be used to establish priority, and so it certainly should be citable. I thought this wasn’t even controversial actually, but maybe the younger generation are thinking of confernces as being something like blog posts; things that shouldn’t be mentioned in formal company.



Kris 06.06.07 at 11:16 pm

One reason to write “please don’t cite without my permission” on a paper posted online is to ensure that you learn who is working on/looking at your stuff, so that you can update that person with a corrected version, or have a conversation with that person. When people have wanted to cite my stuff, I always give permission, but it is nice to know who they are and what there take on the stuff is.

So the thought is: it’s ok to put “don’t cite without permission”, but in practice you should probably always give permission to cite.


Rich Puchalsky 06.06.07 at 11:42 pm

“I’m tempted to think that if you put a paper up on the web, that’s to put it in the public domain”

Whatever the ethics of citation, placing a paper on the web certainly does not place it in the public domain (in the sense of that phrase is most commonly used with regard to papers).


vivian 06.07.07 at 1:03 am

I agree with Kris. When I see that request on a draft, it tells me to contact the author for the latest version before my final response/next development. So far authors (some quite famous)have always thanked me and sent a copy of the latest version, or said “go ahead and quote that.” I usually ask about pre-pub papers even when there is no note on the cover, just to find out if it has been published somewhere, so that I can cite the “canonical” version and make it easier for any of my readers to follow up.

Although I now see Brian’s point, that the web is actually the easiest place to find copies, and thus IS better canonical citation.

Is there a better way to indicate “go ahead and cite me, but please email to make sure you have the most current version” without (what some see as) the weasely connotations?


ben wolfson 06.07.07 at 1:54 am

Don’t such notices mostly appear on works in progress? There’s are good reasons to make such works public (attract comment and criticism, disseminate some ideas) and not to want them to be cited (you might change your mind, be convinced otherwise, in the case of quotation the quoted text might be edited out, etc). I think it would be more than a little odd if someone cited a paper as claiming p and the published paper (which may well be the only available version) did not claim p. Anyway I’ve mostly seen the notice with “without permission” appended, which changes the nature of the case.


Matt McIrvin 06.07.07 at 2:29 am

Right, “public domain” is a legal term having to do with copyright that would not be applicable here. But I don’t think that is what Ross Cameron meant, since it isn’t relevant to citation.


J. Ellenberg 06.07.07 at 2:42 am

1. I’ve put “do not circulate” on papers, though not ones that I posted on web — the situation where you need this is when person X asks you a question, and you’ve written something relevant but very preliminary, and you don’t want person X to write a paper saying you’ve proved Y when you’re really not sure you have.

2. “Do not cite” is for, among other things, the following situation: 1) you write a paper with a lot of explanatory material; 2) the editors say the paper is too long and ask you to cut out everything that doesn’t directly contribute to the argument at hand. 3) you leave the longer version of your paper on your website, because you think the explanatory material may be useful to others, but you attach a note asking people not to cite the web version unless they are referring specifically to material not included in the published version.

2.5. I suppose “do not cite” would also be for preprints whose results weren’t fully checked; I myself wouldn’t put something like that on the web, but might give it to friends and colleagues with a “do not cite (yet)” note. Certainly one might want to prevent people from basing their papers on unchecked assertions, but at the same time publicize the ideas and strategies in the paper which might be useful to others.

3. In math, conferences are decidedly not the same things as publications; people feel much more free to speculate, make guesses, and share opinions in conference talks than they would on paper.


Jacob T. Levy 06.07.07 at 2:48 am

Yes, agreed with Kris and Vivian etc. It’s fair to say “I’d rather you not cite, or especially quote, this version of the paper if I’ve written a more recent version.”

In ye olden days, when you might circulate the conference version of your paper to one person (the discussant) or four (the panelists) or at most a couple dozen (we used to have an official obligation to make 50 copies available at APSA), and it was clear that “published” was the only version that counted as “public,” you could try lots of attempts at getting wording right and arguments right before you riskedliving in infamy as the person who said X stupid thing. The line’s a lot blurrier now, but I do think there should still be a stage of “paper I’m not ready to stake my lifetime reputation on, but I want to present to an audience that migh involve circulating it to a small group.”

And, like everyone else, I’ve typically gotten permission, or else the more recent version, when I asked, and given permission, or the more recent version, when someone asked me.


Matt Kuzma 06.07.07 at 3:12 am

I agree with you that people shouldn’t request that you not cite their papers. If they have, I think the proper response is to not read their papers. That way, you’ll never be tempted to use their ideas in such a way that would demand a citation.


aa 06.07.07 at 3:29 am

Please do not read or respond to this comment without prior permission of the commenter.


a 06.07.07 at 5:30 am

My guess is that Cameron probably also thinks it’s okay to go ahead and make copies of something written on the web and freely distribute it, without asking the author’s permission. After all, it’s already there on the web.


James Wimberley 06.07.07 at 10:59 am

Scripta manent, even on the Web. Even if Heraclitus rewrites his page, Parmenides can locate the first version in Wayback. But he shouldn’t rely on that page as representing Heraclitus’ current views. (Can he ever catch up with them, as with Achilles and the tortoise?)
In practice, isn’t it enough for authors just to write “draft” or “provisional” and expect anyone referring to it to note that character? You could also use “beta”, which has the handy connotation “no guarantees, probably has bugs, use at your own risk”.


John Ellis 06.07.07 at 12:50 pm

Yeah, this seems a no-brainer to me. The point of a conference paper for many scholars is to present very tentative work in search of feedback for revision. A conference presentation is not a form of publication. It is a “making public,” but it doesn’t count as a publication for tenure purposes.

It would be bad scholarship to, as one writer above put it, cite a conference paper as “proving X” when all it did was begin to think about how to prove X. Likewise, it’s disingenuous to use conference paper arguments as straw-men to beat up on, without contacting the scholar to find out if s/he has a more recent or developed version of the argument. (For a particularly egregious example of this, look at Erin O’Connor’s article on “Post Post Colonial Criticism,” which cites and criticizes a work-in-progress presentation by one of her former U of Penn colleagues, Elaine Freedgood — a work in progress that was presented *only* to members of the Penn English department.)


Mike3550 06.07.07 at 1:08 pm

I imagine that, as a practice, people in the academy could start forming the practice of either titling documents available on the web “DRAFT: [Insert title here]” and require that any citations include the word “DRAFT” in them. In this way, it would signal to readers of the citation that: 1. This is probably not the final version – go look up the final version; and 2. still provide the benefits to the community that Brian cites.

Also, following up on j.ellenburg’s comment – in many disciplines (mine is sociology), conference papers are often preliminary conclusions. This means that the conclusions could be based on models that are not fully vetted or the dataset is so new that there has not been time to properly clean the variables. Now, in reality, it would be bad practice to present results that you suspect might be wildly incorrect (for the embarrassment of publishing something opposite of what you claimed in a conference), but there could be strengths of associations that increase or decrease or fall into or out of significance in the final paper.


harry b 06.07.07 at 1:11 pm

I’m very inclined to think there is a difference between a paper put up on the web and a conference presentation. In a conference presentation, or a paper presented at a department, you are clearly just expressing thoughts which range from early conversational thoughts to almost-ready-to-go thoughts, and it is not fair of the audience to make any assumptions about what stage the presenter thinks the thoughts are at. Basically, more like a conversation than a paper. Citing conversations should only be done with permission, unless the conversational point is very clearly in line with published work.

But posting a paper on the web indicates that it is ready enough for public consumption, and as such I’m inclined to think it is fair game (though, I would not cite anything which told me to ask permission without asking permission!).

Its not clear that bneing cited doesn’t also bring benefits to the author. I just the other day noticed that an unpublished paper of mine is currently being cited in a number of publications, some approvingly and others critically; I’m delighted of course, and one of the approving citations helped me to understand better what the paper was doing — without the citation I wouldn’t have gotten that benefit, and the published version (if it ever is published) will be better for it.


clyde mnestra 06.07.07 at 3:29 pm

I don’t quite understand the hostility toward such notices. Like some others, I have always interpreted them to mean “contact me if you want to do something with this, and I’ll tell you if I’ve got a more recent version, or if I now think this paper stinks to high heaven.” Being cited does benefit the author, but it’s requesting the courtesy of retaining some vestige of control. I’d be less sympathetic if the author used this power to discriminate among would-be citers, but it’s hard to imagine that happening, save insofar as personal relations influence the willingness to petition the author.

One of the reasons I view these more charitably is because I think they indirectly respect the contributions of others — in practice, they allow one to update papers to reflect helpful comments made in the interim, when it matters. True, this could be done by continually replacing posted versions, and particularizing the standard request depending on whether it’s a conference version or something later, but isn’t this getting nit-picky?


MC 06.07.07 at 5:17 pm

Being cited does brings some value to the author and–like page rank statistics– the more citations to any one paper, the more valuable that paper. If the citations to a particular work are being split between two or more versions, then that work loses (some) value.

This is even more clear from the perspective of the journal which loses out on potential impact factor gains if the non published version is cited.

For any one citation, of course, the effect is probably unmeasurable, but it could be substantial in the aggregate.


Alan Bostick 06.07.07 at 7:10 pm

Kris @ 1: One reason to write “please don’t cite without my permission” on a paper posted online is to ensure that you learn who is working on/looking at your stuff, so that you can update that person with a corrected version, or have a conversation with that person.

Except that, once you post something on the Web, at least 70% of the people who view it will have found it via search, including such off-the-wall searches as kieran+healy+nude+jpeg or hot+philospher+porn, completely outside the scholarly community and with no attachment whatsoever to academic values of fair play. It’s sort of like putting an open sack of gold coins on your front stoop along with a sign that says “Please don’t steal.”


Clyde Mnestra 06.07.07 at 7:45 pm

alan bostick: It’s sort of like putting an open sack of gold coins on your front stoop along with a sign that says “Please don’t steal.”

Ummm . . . 90% of that 70% won’t do anything the author doesn’t want done; reading is okay, only citation and the like are not. So maybe it’s like putting an open sack of subway tokens on your front stoop, when the only place the subway goes is the library.

In any case, I can’t see why the risks the author takes is a sound reason for ignoring her or his preferences.


Douglas Knight 06.07.07 at 7:52 pm

I’m not from a field that has such notices, and I’ve always assumed that they meant “only quote in a positive manner.” Phrased this way, it’s obviously selfish, but it allows the benefits to the field that were suggested in the post!

Another theory I’ve had is that they actually mean “cite but don’t quote.” I’ve always been absolutely certain that they weren’t to be taken literally, but had some well-defined meaning inside each field that used them. Brian and Ross seem to deny the existence of a convention and take the line literally. In particular, Ross’s discussion of the hypothetical that one would not be given permission to cite is, I think, the problem with the discussion. The denial is simply unacceptable, at least in the conventions (perhaps from other fields) I read in the comments here.


dsquared 06.07.07 at 8:52 pm

IIRC, Heath, Jarrow and Morton were like about third or fourth to publish on the HJM interest rate model, but got the credit because it was widely acknowledged that their manuscript was the original source – because of journal publishing being what it was in finance, a couple of extensions to the model got published before the actual model did!


Robbie Williams 06.07.07 at 8:53 pm

Really interesting thread. My feeling is that, at least in philosophy, there’s a lot of confusion about what this kind of note actually means. Conventions are pretty defective, if a bunch of people in the profession aren’t aware of them, and that seems to me to be the current situation.

I do think that a bunch of grad students/early career researchers are worried that (a) they’ll be missing a trick if they don’t put work up online; and (b) that doing so risks the best ideas being taken up in such a way as to damage their chances of getting the paper published. I personally feel (b) isn’t something to worry about so much, but I do think that people have heartfelt concerns over this. And so it’d be natural if they interpreted the notes we’re talking about — perhaps wrongly — as addressing their concerns.

I like some of the suggestions given above: e.g. putting something into the title of work in progress to indicate this, and if one wishes, to put a note in inviting people to contact you for updated versions. Those sort of proposals just seem less open to misinterpretation than the one under discussion.


Mark van Roojen 06.07.07 at 11:14 pm

I’m pretty well in agreement with HarryB’s comments above, but I thought it might be worth asking whether it is relevant that putting something on the web counts as publication for securing copyright? As I understand copyright law, once expression is put in a “tangible medium” it is regarded as copyrighted and electronic publication counts as such a medium. (I found a reference to this doctrine at, among other places.)

It seems to me that this is relevant. One point of both citation conventions and of copyright law is to give people credit where it is due. But in both cases you don’t want the relevant rules to stifle further work. Once I have seen an idea presented by someone else I would be doing something wrong if I wrote something based on it without giving credit. But if I hadn’t seen that idea I could and perhaps even would have come up with it myself. If we had a practice where on the one hand we require citation of sources when they are sources, but on the other hand allow those who web-publish to restrain citation of their ideas we would be keeping people from further creative work of the sort we would all like to see. (I believe considerations of this sort went into the formulation of patent law and also copyright law which includes the doctrine of fair use.)

These considerations lead me to think that a practice which favors asking those who request that permission be granted for such permission to cite is OK, but only if the permissible responses at that point will either be (1) yes, go ahead cite me, or (2) don’t cite me, but you may use the ideas without attribution. There should not the a third option of don’t cite me, and don’t publish anything based on what I said until I publish my paper in some other way. That last option would stifle research we may well want to promote. We all know of some people who are just too perfectionist to ever get around to publishing their ideas in final form and we should not allow them to keep us from talking about stuff just because they thought of it first.

These remarks come from someone whose discipline is philosophy and I may therefore be missing points relevant to other fields because of my background.


Jacob T. Levy 06.08.07 at 12:53 am

Harry says: I’m very inclined to think there is a difference between a paper put up on the web and a conference presentation.

But the difficulty is that papers presented at conferences are now often expected to be posted online.

I think there’s some room for discrimination among online sources. Something that’s put on SSRN or equivalent is, like papers in official “working paper” series, game for citation as-is. But something posted to the website of a big conference might just be a conference presentation, and there ought to be space for conference presentations to be experimental or incomplete or just wrong.

Except that, once you post something on the Web, at least 70% of the people who view it will have found it via search, including such off-the-wall searches as kieran+healy+nude+jpeg or hot+philospher+porn, completely outside the scholarly community and with no attachment whatsoever to academic values of fair play.

Apart from the entertainment value of what Alan just added to google, I think this seems wrong. Such web-surfers who are entirely “outside the scholarly community” aren’t going to have any itnerest in even clicking through to most online working papers, even if the papers happen to turn up in their searches for hot philosopher porn.


vivian 06.08.07 at 2:04 am

Why does it seem like a burden to email the author? Can’t you think of it as the old-world version of ‘google alerts’? I can’t think of a case of someone refusing permission – can anyone else?


Ross Cameron 06.08.07 at 9:07 am

Wow – I’ve created a monster!

I would really like to know whether Douglas Knight is right and that there is a convention that this request isn’t to be taken literally. I’ve certainly been reading it as such. So if there is such a convention, it’s perhaps not widely understood that there is such a convention.

The request has certainly caused confusion. I have set (with the author’s permission) works in progress with that request as reading for some of my classes, and I’ve had students worried about whether or not they can cite the paper in their exam!

Sorry for the confusion over the term ‘public domain’. I certainly didn’t mean it in any legal sense. I only meant that if you put something on the web that seems to me to be making it available for public consumption – as such, it seems odd to place a restriction on it being cited.

The case of an author changing their mind is an interesting one. I know of a case of an author writing a very interesting paper on a topic but changing his mind on the whole approach to that topic, and so the original paper remains unpublished. But some of us think his original idea was better than where he’s ended up. What can we do? He doesn’t want that paper cited, because he thinks it’s a completely wrong approach. We want to build on that earlier paper and give him credit for the groundwork, but he doesn’t *want* credit for the groundwork, because he thinks it’s mistaken. To present it as if we came up with the initial ideas ourselves would be terrible, so basically we’re stuck with these ideas that we can’t do anything with. That doesn’t seem quite right to me.


bill wringe 06.08.07 at 9:46 am

‘Citing conversations should only be done with permission, unless the conversational point is very clearly in line with published work.’

Does this include citing someone’s response to a point you’ve made when presenting work in a public forum (eg in a departmental seminar or at a conference).

The reason I ask is that a) this has happened to me a couple of times b)on the occasions that it has happened I only found out about it circumstantially c) on both occasions I was pretty chuffed (pleased) about it and d) I don’t suppose it would have happened if the person referring to the point I had made had felt the need to email me about it. (and e), the points made had very little to do with my published work
– such as it is.)

So I’d be less happy than I am if this sort of convention was observed. But it seems odd to have conventions about stuff posted on the web that protect it more than points made in the seminar room.


harry b 06.08.07 at 3:00 pm

RE #21: I think that the fear of having your ideas stolen is quite real and common, but its worth saying that putting the request not to cite or quote without asking permission will almost certainly have no effect at all on whether they are stolen or not. Some ideas are stolen deliberately and consciously, and others are stolen unconsciously (A recent conversation with a friend revealed that we both believe ideas of ours have been “stolen” by an eminent person, but we also both think that person is completely unaware of having done so). Conference presentations, private conversations, blog discussions, you name it, they all put you at risk of having your ideas stolen, and there’s bugger all you can do about it. Sue? Not worth the effort. They’re only ideas, after all.

In fact, putting something up on the web is probably the best protection possible, because as long as you date it, if you care you can always point to the fact that your instantiation of the idea predates other appearances, if it really is the case that you are first with it. And, if you are too shy or too dignified to do that, others will notice anyway, if you are lucky.


mollymooly 06.08.07 at 3:48 pm

I’m not an academic and I’m confused. Does this “do not cite” thingy refer only to citations in published articles? Maybe there is a hierarchy: you can cite journal articles in web articles, web articles in conference papers, conference papers in blog postings, and blog postings in the common room; but not conversely.


don't quote me on this 06.08.07 at 6:44 pm

“Please don’t quote or cite”? What kind of request/command is that? That’s like a politician submitting to an interview, and then at the end saying, “By the way, this is all off the record.” Too late, bub! Likewise, if you don’t want to be quoted or cited, don’t publish your words. Geez.


Mark Eli Kalderon 06.10.07 at 1:58 pm

Stability is a precondition for the possibility of citation. Part of the reason for citation is to allow your readers to check your sources for themselves. Having the cited material change undermines this. Say, you quoted a line that was subsequently deleted. When your reader looks for it, it is not longer there. When publishing was confined to print, this was not a problem—material on the web can be changed in a way that ink on a page cannot.

To bring this out, consider the following case. A journal has an online version that gets published prior to the printed version. After a paper has been published online but before it has been printed, the author wishes to change something before it hits print. It is wrong to have different versions of the paper online and in print, so there are two choices, change the online version per the author’s wishes and have these changes in the print version, or change nothing. The problem with the first choice is that people might have already read and cited the changed material. And if the journal does make a change it undermines the integrity of the online version as a source of citation.

Authors publishing drafts online is similar. Given that they are drafts, they are subject to change and so not citable. The usual qualification ‘without permission’ allows the author the ability to exercise judgment about particular cases. This seems like a reasonable practice.


jge 06.11.07 at 9:55 am

Perhaps it helps to distinguish between different types of quotation. The first type is the authority-quotation: One cites person X because he has said that q is the case (and one has no other argument supporting q). In this case one should obviously in one’s own interest seek (permission and) the newest version of the paper.
The second type is creation-quotation: X has this cool new idea or argument that has helped one’s own ideas on the way, and one wants to give him credit for it. It seems to me a duty of intellectual honesty to give the credits publicly (with permission or without). But it may be done without actually quote anything.
The third type is further reading-quotation: In this case it may be enough to say: check out the website of X …

Comments on this entry are closed.