I was reading a newsletter at my toxicology lab when I noticed the article used as a source an unpublished article. This got me wondering:
–why would someone go to the effort of writing a science article and getting an article reviewed and then not publish them? Is it a sign the article couldn’t make the standards required of a published article?
–where do you find unpublished articles? Is there some obscure database that keeps track of these?
There are a couple things that “unpublished cites” could be. First, it could be an actual paper that hasn’t yet been accepted for publication. Maybe it hasn’t been submitted, maybe it’s in review. Usually these come from colleagues or even the same lab. Also, in my field, posters and talks presented at conferences are never directly cited, even though the abstracts are published. Instead that stuff is cited (with consent of the person being cited) as a “personal communication”.
Yeah, these citations haven’t passed peer review, but they’re usually peripheral to the main point of the paper. If those cites don’t stand up in the end, or if they’re never even published, the paper should still stand OK on its own.
A lot of physics articles are put on arxiv.org and then never published. Some of them are just crackpots, but some well established scientists just post papers there and then don’t bother with the journals. Since much of the community just reads papers on arxiv anyways, and publishing in journals is both expensive and time consuming, there isn’t actually much point if you are in a point in your career where you aren’t concerned with building your publishing count.
Alternatively, an increasing number of gov’t grants require research to be made public. I imagine some results, especially in a filed like toxicology, might require results to be made public even if they don’t merit a paper for whatever reason.
Lastly, sometimes it just takes a long time to get papers published, so preprints are posted and then a year or more passes while the referee process does its thing.
The researchers who wold be most interested in a given paper are usually well aware of it long (years here often) before it’s published. They see working paper versions that are now circulated with the internet and in the old days by mailing Xerox copies or they heard someone give a seminar based on the paper.
Often these are cited by others before they are published. In fact Google Scholar has a way you can merge your working paper with the published version (in case they have slightly different titles) so that proper, or at least better, cite counts can be determined.
Sometimes (actually, quite often in my experience) editors (and peer reviewers) will turn down papers not because there is anything objectively wrong with them, but because they do not think they are “interesting” enough: they think that not enough people will read or cite them, and thus they will not enhance the reputation of the journal. They do not always get this right, and even papers that are not very widely cited may nevertheless prove of interest, and worth citing, to a few.
Negative results (ie., failure to find an expected correlation between two variables) are one example of a type of result that is notoriously hard to get published, even though teh negative finding might be quite significant from some perspectives. However, there are lots of other (good and bad) reasons why an editor or reviewer might think a paper is not suitable for the particular journal to which it has been submitted.
Someone whose paper is turned down for these sorts of reasons may continue to submit it (or a rewritten version) to different, perhaps less prestigious, journals, and it may get published, in some form, somewhere, eventually, but this process can take years. (Generally speaking, there are strict rules against submitting the same paper to more than one journal at a time, and peer review can take months at each journal. I once had a paper tied up in a lengthy, two stage peer review process at one prestigious journal for two years, only, eventually, to have it turned down by the editor against the advice of all but one of multiple referees. It was eventually published in another journal, and has been cited many times.)
In some fields, most papers will probably get turned down several times (for subjective reasons, rather than because of any objective flaw in them) before finding a publisher. Sometimes the author will give up.
In the meantime, the paper will very likely have been circulated as a draft or “preprint” amongst colleagues, some of whom may wish to cite it. If their paper gets accepted more quickly, it may be out before anyone knows where, or whether, the paper they cite will be published. (The internet has made the practice of circulating drafts and preprints even more widespread than it used to be, and, often, these are easier to get hold of than formally published papers, that disappear behind publishers’ expensive paywalls.)
To late for the edit window for the last part of the question:
As Simplicio mentioned, in physics arxiv.org provides a quasi-formalized system for dissemination of unpublished or not-yet-published drafts and preprints. In other fields it is mostly less centralized and formalized: there may well be several competing or overlapping archives, and many authors will make their work available just on their own web site, or their university’s site (or they won’t bother at all). If they know you, or if you hear about something and ask nicely, they may email you stuff (or even snail mail you stuff). Ask someone for a copy of a paper they have published, and they may send you a slew of related, unpublished stuff as well (or stuff that has been published, but in obscure journals or books that you probably would not have heard about). Sometimes you have to poke around on the web quite a bit. Google Scholar is very helpful, but it does not catch everything that is out there.
I don’t understand the use of “published” here.
Why is publishing them on the Internet not considered publishing?
In the scientific and academic world, being published means being published by a recognized journal (or in a formally edited book from a reputable publishing company), and this means that your work has been certified as competent and worthwhile by peer reviewers and an editor. The process is far from perfect, but it is an indispensable way of indicating scientific and scholarly quality. If something has not passed through that process, it is considered “unpublished”. Anyone can “self-publish” any old shit on the internet, and the vast majority of it is, very rightly, ignored by serious researchers (and, indeed, by almost everybody else). Life is far to short to read every crank’s crazy claims or ideas. The “unpublished” papers we have been discussing in this thread are usually by people who have proved that they are not fools or cranks, and have demonstrated their competence and scholarly integrity, by having previously had stuff formally published through the accepted channels.
I would just add that it is much easier to get an interesting result published than its refutation. You have a contract with a drug company to study some drug. You find an effect. Maybe there was or maybe it is one of the 5% of times that the effect is by chance. You publish it. Some repeats your study and finds no effect. Ho hum, the journal is not interested. Happens all the time.
The gold standard is supposed to be publication in peer-reviewed journals. Well, referees are not perfect, they have their prejudices, some of them have agendas, etc. Nowadays you can write a paper and post it on your web site. If you have tenure and don’t need a research grant, then there is little incentive to go through the publication hassle. But that doesn’t make your results less interesting.
It is not uncommon to write a fairly detailed article that begins life as an internal note among a group of researchers. Years later, it might become relevant to a broader community and get a citation, and as long as the citing article’s publisher is okay with citing it, all’s good.
An example would be a write-up of the technical details of a particular method, algorithm, instrument, etc., that is in use by only a handful. For example, we have a custom piece of equipment that tests widgets. If someone else finds themselves needing to test similar widgets, they might track us down and ask how we test our widgets. We can send them a technical article that describes the whole apparatus, but that article doesn’t really belong in any wide-audience journal.
On large scientific collaborations, there will be literally thousands of internal documents (some very rough-and-ready, some in nearly publishable form) that never move beyond the confines of those scientists due to the specific nature of the documents’ subjects.
Technical and scientific write-ups are a mainstay communication tool of scientists. When you see “unpublished article”, don’t think that that’s a year’s worth of toil that never made it to a journal. It might just be an afternoon’s worth of writing up something that the author was dealing with over the past week and needed to describe to someone else.
Regarding an up-thread comment about people eventually reaching a point where they give up on journals: not in my experience. Unless they’ve checked out of the game entirely (very rare), publication in peer-reviewed journals remains important well after tenure, and even grey-haired scientists push hard on publications. Funding to keep one’s research going and internal promotions (read: pay raise) are both tied to one’s success and reputation in the field, and these are still measured (in part) by one’s publication record. Someone who just throws a couple of articles up on the arxiv once in a while is going to wither quickly.
People most certainly do give up.
I was talking about particular papers before. I have given up on certain papers, and I know of others who have given up on lots of them. Sometimes (in some fields anyway) it can look worse on your CV to have a paper in a notoriously crappy journal than not to have it published at all. Some fields just do not have that many journals, especially as, as you work your way down the food chain, the journals tend to become more and more specialized. As I and others have pointed out, most negative results are unlikely to find a publisher anywhere (although things may be changing, with open access, pay to publish journals like PLoS One and the various online journals of negative results now appearing).
You, however, seem to be talking about people giving up on publishing altogether. Damn right that happens! Lots of people who have embarked on a scientific or other academic career give up and stop publishing. In fact, they are the very large majority. It is only the lucky (yes, and talented and industrious) few who get into tenured or otherwise secure jobs that will last them until retirement, and where they have the luxury of, if they want to, keeping on revising and submitting that rejected paper for ever until it finally gets accepted somewhere. Even than, they may conclude that their energy is better spent on new work for a new paper, rather than on gussying up an older one until some editor is persuaded to like it enough.
Lots of potential reasons. For example, a field tends to produce a lot of “folklore” — results that everybody knows are true but was never published for various reasons. There comes a point when somebody decides that all this folklore needs to be written down somewhere so that the knowledge isn’t lost to the next generation. Of course, as “everybody” already knows this stuff, it’s essentially unpublishable.
Another example: the French until recently only published in French (they still have French-only journals even in new fields like CS). Some of these French language articles are extremely significant (especially in my field, where a lot of the theory behind dependent type systems was developed in France), so will be translated into English by other academics. The English-language article will often be cited alongside the French original despite it never being published.
Obviously that is not the group I meant. Clearly anyone who is no longer on that career path will have no need to publish. Someone mentioned tenured, successful academics who felt they didn’t need to bother with journals. That’s what I was saying doesn’t happen.
Did someone mention that? I was the one who mentioned the possibility of giving up on publishing, but I was talking about someone (at any stage of their career) giving up on an individual paper, not stopping trying to publish altogether.
The quote I was responding to was this one, regarding the arXiv: