4

I am an independent researcher and have spent the past year writing a fairly extensive (60+ pages) research paper on a topic in probability/statistics. The paper is almost complete but there are a couple minor loose ends that I am having issues with proving. Furthermore, given the length of the paper I am not sure of the appropriate journal to submit for publication and thus am considering contacting relevant professors in academia for help/guidance. My plan was to offer co-authorship in exchange for help addressing these loose ends and guidance with the publication process. That said, my impression is that research faculty are inundated with these sorts of requests and so I was not sure if this approach was likely to get a response. Is this practice frowned upon in academia? If not, are there general best practices for reaching out with these sorts of requests?


Update (2/8/22): The now 122-page paper is available on ArXiV here. I was not able to find someone interested in helping due in part to the length of the paper. I am still faced with the challenge of how/where to publish this and most importantly receive feedback. Having stared at this for so long I find that I am unable to see better ways to organize the paper and make it more readable. I did receive some useful feedback over on MathOverflow. In particular, it was suggested that I have done too much of the work to ethically bring in someone as a collaborator. This puts me in a strange place because I am really struggling to improve the readability of the paper without feedback and at the same time cannot get anyone to read it because it is so long. I was told I should move chapter 2 to appendices and condense chapter 3, which is where the "meat" of the paper is. Some suggested publishing this as a monograph but I'm not sure. I still am not sure about best practices for publishing works of this length. How many journals accept papers with 50 pages of appendices? Guidance on how to proceed with getting this to an appropriate journal would be greatly appreciated.

9
  • 1
    What journals are the papers you cite published in? That will guide you on which journal it is appropriate to submit to.
    – astronat
    Sep 14, 2020 at 17:14
  • 1
    @astronat Thanks for the comment. This is typically how I would search for journals to publish in. The problem is that this paper took an applied problem in optics and approached it from a purely mathematical "theorem-proof" approach and so the content of the paper is too math heavy for the optics journals most of the related literature is published in. Sep 14, 2020 at 17:49
  • 1
    Is there any way to reduce the length of the paper (by splitting it into multiple papers, or breaking into appendices, or what have you)? My concern is that getting a stranger to read your paper can be tough, but a ~5 page paper is much more enticing than a 60 page paper.
    – cag51
    Sep 15, 2020 at 14:19
  • @cag51 I agree that the length is a problem if I want someone to read it. Ultimately I think the paper will have to be split up into multiple parts for publication. However, how to go about doing this in the proper way is one of the things I am seeking guidance on. Perhaps I could select the main theorems of the paper without the proofs and write them up in a letter (~5 pages)? That way someone could read the letter and decide if they wish to take a closer look. Sep 15, 2020 at 15:01
  • 1
    @RodrigodeAzevedo I have done this with limited success. Here is one example. Feb 8 at 16:46

6 Answers 6

6

This is not my field, so I can't assess the technical results (and this is not our role in any case). But from looking through quickly, my impression is that you have a mix of paper and manuscript/textbook here. Consider:

  • Chapter 1 begins with three pages of "mathematical preliminaries", which defines terms such as the gamma function and the binomial coefficient. This makes sense for a textbook, but in a research paper, you would normally not define "common knowledge" like this.
  • Your later chapters (from the titles, anyway) seem to similarly spend many pages summarizing known results. Indeed, section 3.1 is a "brief review of photon transfer theory" -- four pages. Normally, the entire discussion of "textbook results" in a research paper would be less than a page (though conventions in your field may be different).
  • You then have some lengthy theoretical results. I don't have the background to determine whether these results are "exciting" / interesting (and should be a separate manuscript), or if you're just slogging through the work needed to define your algorithm (in which case, you will need to summarize the procedure / results without showing the derivation line-by-line).
  • Then, 58 pages in, you define your estimator and set up an experiment to test it. You seem to conclude that it works well. I'm concerned you don't have a state-of-the-art benchmark to compare it against, so how do you know it's better than what exists already? But I don't know your field, so fine, maybe my concern is misguided.

Based on this, I would suggest that you should split this two or three ways:

  1. A paper based on your estimator. This will < 10 pages in which you describe the problem, describe your algorithm, and show the experimental "proof" that your algorithm works well. If there are any other "really nice" properties of your estimator, you can state those as well if you can squeeze them into your page limit. Keep in mind your audience: I assume your audience is algorithm users who are considering adopting your method -- so, you'll want to motivate them by showing that it works and telling them how to do it. Or maybe your audience is mathematicians, who will have different motivations. But no one will be motivated by a 122 page paper :-)
  2. A manuscript intended for those new to this field. You seem to have quite a bit of this already, though you'll want to remove the "novel" bits that are not well-established results (or use your novel bits as examples).
  3. If any of your theoretical work is an "exciting result" in itself, a separate paper there may be appropriate.

Getting the 1-2 short papers published should be doable, and the manuscript being on the arXiv is probably fine. You can cite it there...and you may find that you want to turn it into a full length textbook in the future, in which case not having this part formally published may be a good thing.

1
  • 1
    This answer is very good as it proposes a solution based on your seemingly main issue: Not knowing what your target audience is. I scrolled through the work for a bit and it seems like you wrote for students, applied scientists, and theoretical scientists at the same time. Feb 16 at 12:44
7

It sounds like your question is: "My paper is way too long. How do I publish it in an optics journal?"

  1. Identify the most valuable result.
  2. Go through your work word-by-word and remove what is not necessary to support that result.
  3. Find that it is still way too long.
  4. Write a four page, illustrated summary that shows the value of your result. This should have a minimal amount of math.
  5. Submit the summary to an optics journal with the document that is way too long as "supplementary information." The summary should emphasize the value of the result, and only briefly summarize the method.

This is not guaranteed to work if nobody's willing to peer review it.

0
5

Somewhere in the comments you mention "It just took a lot of work (lots of many nontrivial calculations)".

Unfortunately not all the work is relevant to be published, in published paper you may find that what is summarized in a very concise sentence (for example "the parameter X of the hyperbolic PDE of order 2.5 has been estimated by analytically solving for the kernel of the lagrangian in a complex hermitian basic solution") which can be understood by the specific community interested in the results of the paper (in the example I gave the comunity would be the imaginary lagrangian chemistry modellers).

There are papers very dense in their results, that are condensed in 10 pages, but they requires a lot of work of many people, however their work is just added as a very fat appendix.

It seems to me it is the case of your publication: you can easily go from 100 pages to 10 pages + an appendix of 1-2 section, each of 40 pages: please note that the appendix can be much less descriptive and it can be limited to the bare minimum (with references and so on, but with minimum descriptions), so it may be squeezed down to maybe 25 pages each.

My external observation is based on the fact that I can follow most of your publication without being in your sector (it's not a good thing :) ) and I found sentences like pag.11

Given the rich theory of Laplace transforms one might hope that published tables of transform pairs will provide the necessary result to invert this equation. Indeed, inversion w.r.t. β2 is achieved via [19, Eq. 5.4.9] yielding an unbiased estimator for τ when κ1 is known (see Appendix B). However, the final inversion w.r.t. β1 leads to an intractable problem; suggesting the possibility that the estimator T does not exist.

Which can be reduced to a much more dense (I am cutting a lot of corners, however the meat is that Estimator T may not exist, right?)

Estimator T could not be found via inversion (see Appendix B)

the preamble is useless: the reader does not want/need to be surprised, or teased, by the reading of your paper.

The results inside must trigger the interest of the reader, not the way you write it (one can even say that the paper can be boring, to a certain extent).

4

Yes, finding collaborators late in the game is hard. Too late now, but it is much easier to establish these early on where you can share ideas. And a late collaboration on a very large work is even harder.

But, I fear that you are letting the perfect be the enemy of the good. While I think it will be hard to find any publisher willing to take a stab at such a long paper, I think your best (caveat below) option, if you can find such a publisher is to submit what you have and look at what feedback you get.

I've assumed above that what you have yet to "prove" isn't consequential to the overall result. If it is, then you have to admit that you might be wrong, fundamentally wrong. In that case you have to deal with the "choke points" and examine the available evidence.

Alternatively, and maybe even better, is to mine what you have already done for a series of shorter papers that you can submit more widely and for which you can provide appropriate proof.

Combining the above ideas, if you can separate out those choke points into separate "papers" you might be able to get collaboration on the smaller issues by asking an academic or two for feedback. It might even be different people for different issues.

1
  • Thank you for the feedback. Indeed, I am past the choke points and was able to iron out the loose ends I had at the writing of the original post. The problem now is of organization, making it more readable, and highlighting the key innovations. Does it all make sense to me? Yes. But some of the feedback I got on MO was that it was difficult to extract the essential parts of the paper. To do this would require a lot of reorganization, and without a clear goal in mind, i.e. where to publish and in what format, makes this reorganization effort much more difficult. Feb 8 at 15:03
1

It is common practise to ask domain experts for help and, where appropriate, to offer co-authorship. As professors are busy, you might want to seek more junior researchers. (I'll refrain from explaining how to write an email, since you can find answers on this website.)

-2
  1. Keep the current manuscript to support your theses if you receive inquiries about them.
  2. Make a goal of reading daily 5 pages of it and converting them in 1 page.
  3. Put together the new manuscript with the single pages and re-read to check if it makes sense. Repeat the entire process once again (if it’s still too long).
  4. Submit for publication

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .