How to publish a paper that does not seem to be within the scope of any journal?

Question

I have written a paper on applying a mathematical theory to a linguistic problem. I privately submitted it to several renowned professors (some of them very famous indeed) in the field of linguistics, semantics and so on. They made some suggestions and expressed some criticism on some details, but for the most part they praised the quality of the paper.

Following their suggestion, I submitted it to a leading journal in the field. The answer I got from the journal’s preliminary assessment was that the paper had huge potential, but was too mathematical for them. They advised me to resort to their sister publication, where much more focus on natural language processing and mathematics is to be found. The answer was the mirror image: nice paper, definitely scienficially worthy, but too soft, that is, too linguistic or informal for advanced mathematicians or computer scientists.

Now I am in no man’s land. The paper is indeed interdisciplinary (as corresponds to a linguist using some mathematics as he needs them), but I find it very unfortunate that due to a false sense of the own domain, noone seems to feel responsible for its contents, which I deem valuable. Before trying any further submission, I was entertaining the option of submitting it to the ArXiv, but I do not know whether that will make any sense.

How can I proceed to find a journal that considers my paper to fit within its scope?

^{I posted this question on Math Stack Exchange before and was recommended to repost it here.}

Answer 1

I sense impatience or even desperation in your original question and comments: If you have got serious indications of you are about to be scooped by other scientists, then I will advise you to publish now on ArXiv or likewise. In some other scenarios, you could benefit most of keeping your paper secret forever. More on this later on.

Do you realize that your research from a mathematician's point of view most likely is applied science, but might be basic science in the field of natural language analysis? Scientists with a pure linguistic background might not even be aware of the potential of advanced mathematic theories.

I can hardly imagine a single paper suitable for journals in different fields without heavy editing aimed for each separate field: Mathematicians want to read about implementation details of the theory, and why one method works better than other ones. Too less of this stuff and your work is “too soft”. However, they do not care about the implications of your results in the field of linguistics. Language researchers prefer to know what results you can achieve using advanced mathematics but too many implementation details will look like unnecessary "noise", which never will pass the assessment committee.

Even you do not like the journal's comments I will encourage you to consider that they provides you with useful facts.

Suppose a journal accepted your paper just to make you happy. Language researchers scanning literature for a method to solving a specific problem identical to yours setup, say extract meaning of natural language texts in the style of IBM’s Watson, will not get the point in relation to their fields, because they do not understand larger parts of your paper. Mathematicians looking for examples of appliances of the theory you applied, say bayesian networks, will quickly move on to papers they can read with lesser efforts. Only people with experience from both fields can take full advantage of the paper, which implies a very low number of citations. That is not what you want.

Interdisciplinary research often leads to spectacular new knowledge and if your paper contains revolutionary concepts or eye-catching implications, it could be a candidate for Nature or Science without changes. Also, try reading “call for papers” for conferences. If you match their hot topics, they will probably accept your paper as it is.

Otherwise, it is a matter of finding that journal, which you most easily can make your paper fit into, and which is the best choice for your career: Did you ever ask yourself why you wrote this paper? Academic ambitions? Then I guess only journals of linguistics will count. Do you look for a position as researcher in a private company? Then you should go for applied mathematics.

Based on the comment "definitely scienficially worthy, but too soft" and the professors' general acceptance, it sounds like that the right way for just proper academic recognition could be to rewrite your paper as a book. Wiley, Prentice Hall, etc. tend to listen to recommendations from professors - their actual customers. A submission to ArXiv does not prevent this, and you do not write a single line until you have got a signed contract.

Follow up on just submitting to ArXiv:

A scientist uncovering new areas should always discuss the potential for patents with attorneys before attempting to publish. Publishing prevents patenting. Patenting will worst case only delay publishing.

If your research is useful for software making Siri looking like a toy, I assumes you would like to get along with financial strong companies (Google/Microsoft), which usually keeps their best secrets as secrets and that means no patents/publishing. You deserves economic recognition too.

Good luck!

Answer 2

Interdisciplinary papers have this problem a lot. A - now - very influential paper in my field, I was told by the first author, was in limbo for 4 years before it was published, despite the senior author being world-famous.

Persistence is the key. Try to find someone influential who is knowledgeable and you can convince to help you expanding/adapting/explaining the work better and you could then co-submit with. Or else, you could consider aiming for a general-purpose journal such as PLoS ONE, and whose reviewers are advised to select by novelty and correctness, rather than (subjectively judged) relevance.

Learning to write for the specific audience of a journal can also help.

Answer 3

Here are some suggestions to get your work published, that may or may not apply to your case – you have to judge for yourself.

Ask the journals

When submitting a rejected paper to another journal, it is often recommendable (and some journals even require it) that you detail to which journals you submitted your papers before and why they rejected it. In your specific case, this may lead to editors being more lenient regarding the scope, in particular if we are talking about sister journals and even more so, if they aspire to cover the entirety of a given field.

If you did not do this, I recommend writing to both journals you submitted to so far and inform them that the respective sister journal also rejected the paper for being out of its scope. At the same time, you can ask the respective journals to recommend a suitable target journal.

Another possibility that you have to consider is that declaring the paper off-topic was an easy way out for the editors to deal with a paper they didn’t really know what to do with. So it may be that your paper does have some relevant flaws that were unmentioned. If you write to the journals as suggested above (and if you do not make any accusations in this direction), they may give you some hints as to what you can improve.

Ask your private reviewers

Another inspiration for journal selection may come from the professors who privately reviewed your paper. If it pertains to their field, they should be able to recommend you a journal. At the same time, you can ask them whether there is anything they would change about the paper to adapt it to the journal they suggest.

Split your paper

This is a standard technique used by my interdisciplinary surrounding: Publish a methods paper (or similar) in a more theoretical journal and then publish an application paper in a more applied journal. A major problem of an interdisciplinary paper – even though never said explicitly – may be the journals have no idea who could review it (this problem would even arise with mega journals). Remember that they need to find a number of people who have sufficient knowledge of all relevant fields and they have to accept to review. Splitting your paper may allow each part to be reviewed by experts of the respective field and thus solve this problem.

Answer 4

To go a little further than Captain Emacs's answer: even for people doing non-interdisciplinary fields, it's pretty common for nice papers having trouble finding a venue. One important conjecture in my field (and has motivated a lot of my personal research) was made 30 years ago and remains unpublished because the author had trouble getting it accepted and eventually gave up.

In addition to persistence being important, I want to add: browse journals for similar kinds of papers.

I also had trouble getting one of my papers published because it was on a novel type of problem, which both makes it hard to gauge interest and makes it hard to find appropriate referees. The first journal we submitted to had it refereed by someone in another field who didn't seem so interested in the pure mathematical aspects. Then we tried a couple of other journals (including one recommended by an editor of a previous submission) who said it wasn't a good fit and/or couldn't find referees. After looking around again and browsing a lot of papers in various journals, we finally found an appropriate one. Yay!

Also, in order for editors and likely referees to be able to properly understand it and find it appropriate, you may need to rewrite it for the intended audience of the journal (Cap. Emacs' last point). Again, browsing papers in the journal will help you get a better sense of the intended audience.

Answer 5

Linguistics Ph.D. student here.

I'd say keep looking through the linguistics journals. Several of the big name ones seem to be able to handle quantitative analysis just fine, and some of them welcome it.

What I'm thinking of specifically is that at the recent meeting of the Linguistic Society of America, the award for the best paper that appeared in the prestigious journal Language in 2015 was given to a phylogenetic analysis of various features found in the Indo-European languages, pinpointing the likely geographic origin of the language family. I read this article in its entirety when it first appeared in print. It's fairly math-heavy for a linguistics article, and I certainly didn't understand all of the numbers and statistics in it; regardless, evidently the journal evidently did not object, and the linguistics community in general was impressed. Offhand, I know that Language also once published an innovative article detailing a rather unorthodox use of geometry as applied to the analysis of a language's morphology. What you've come up with probably isn't weirder than that, at a guess.

I get the feeling that even theoretical linguistics is becoming more and more receptive to empirical/quantitative work. The Journal of Semantics was pretty open to interdisciplinary research last time I looked into it. Others that might be worth trying: Linguistic Inquiry, Natural Language and Linguistic Theory, and Natural Language Semantics. (And if you can tie the project to language change at all, Language Variation and Change always expects to see numbers in some form or another.)

Answer 6

My suggestion is to try to present the result in a conference first. In some fields (not sure if that's true for yours) the bar for getting into a conference is lower to some extent; and even if that's not true, a conference organizers may be more inclined to have a "weird stuff / odds and ends / couldn't fit this anywhere else" session. Maybe even try to get it in as a non-refereed paper, i.e. as an invited talk - that would mean contacting the organizers rather than the program committee, or both.

When this happens, you might be able to attract enough attention to increase your chance for a journal publication.

This is not to detract from the fine suggestions in @Wrzlprmft's excellent answer.