How to Publish a Scholarly Paper

One of the most useful things I’ve attended at this year’s LSA conference (or any conference, really) was a panel discussion on the publishing process for academic papers. The discussion was aimed at graduate students, but I figured there would be good information there for others as well, and there was. On the panel were Sandra Chung of UC Santa Cruz; Greg Carlson of the University of Rochester, and the editor of Language; Monica Macaulay, of the University of Wisconsin-Madison, and the author of Surviving Linguistics; and Chris Potts, of Stanford University, and associate editor of Linguistics and Philosophy. “Not one graduate student!” complained a fellow attendee, who found the tone of the panel paternalistic. That’s when I realized that the ambiguity of the phrase Graduate Student Panel had led to an unfortunate misunderstanding. For my part, I took page after page of notes, and I’m going to synthesize them here.

To begin with, they had a handout with a publishing flowchart, which I’ll translate into pseudocode:

1. Write a publishable paper.
2. Choose an appropriate journal.
3. Submit the paper.
4. If editor decides it’s not appropriate for the journal, return to step 2.
5. If editor sends it to referees…
   1. If it’s accepted, you’re done.
   2. If it’s accepted on condition of revisions being made, make the revisions and you’re done.
   3. If they ask you to revise and resubmit, revise and return to step 3.
   4. If it’s rejected, revise and return to step 2.

Sandra Chung

1. Write a publishable paper. How do you know if it’s publishable? You don’t want to make this decision on your own. While you’re in grad school, this is one of the things your academic advisor is supposed to do, especially if you’ve published little or nothing before. When they give you feedback on a paper you’ve written for a class, they may even suggest on their own that you submit it to a journal. If they don’t, but you think it might be publishable, then ask what they think about the possibility.

Alternatively, or in addition, try submitting the paper to regional conferences. If it’s accepted, not only do you have some indication that you may have something worth publishing, but you will also be presenting to other people who can give you feedback. Regional conferences are an especially good idea if you’re not in an academic environment where you have an advisor or others easily available for advice.

Before you submit it anywhere, get it to a point where, to the best of your knowledge, “it can be revised no more.” Just as you shouldn’t donate blood in order to find out if you have HIV, you shouldn’t submit a paper to find out from the reviewers if it has problems. (The HIV analogy is mine, but the main point is from the panelists.) Although many referees will give a lot of constructive advice, it’s not their job to write the paper for you. (It’s not even their job to read it: Referees are a volunteer force.) Although it takes a lot longer to get the paper to that state, it’s still shorter in the long run, because the worse your paper is, the longer it’ll take the referees to respond to it. (Even so, expect to have to make changes before the paper is published.)

In terms of writing style, the best policy is to be unnoticeable, as if the information is just being piped into your readers’ heads. But also read the journal to get an idea of its style (see below). And at risk of being obvious, check your spelling and punctuation, and make sure your examples and figures are numbered correctly. Not doing so won’t affect the quality of your data or argumentation, but it will make it look like you’re not serious.

Greg Carlson

2. Choose an appropriate journal. So how do you know a journal is appropriate? One good sign is that you find yourself citing papers that were published in that journal. If you are, then you’re probably writing for the same kind of audience that read those papers. In fact, it’s risky to submit to a journal you haven’t read. You won’t have picked up on things like how much background knowledge is assumed, what kind of jargon or abbreviations are acceptable, whether authors use a conversational style, etc. Referees tend to get grumpy when reading papers by authors who don’t know the norms for the forum, just like you get annoyed by newbies to an online community who don’t spend any time lurking or reading the FAQ before jumping in. (Another analogy I threw in there for added value.) You can also go to the journal’s website and read their mission statement, and/or browse the tables of contents from previous issues.

You can even come right out and contact the editor, and ask whether the kind of paper you’ve written might be appropriate for that journal. After all, they’re busy, and they’d be happy to save themselves and you some time by saying right upfront if something just isn’t right at all for their journal. There seemed to be some level of surprise at this suggestion, like it might be some kind of cheating, or trying to gain an unfair advantage. I was surprised at their surprisal (?), because in the non-academic publishing industry this is standard operating procedure. You don’t send anything unsolicited (at least if you want it to be taken seriously): You contact the agent or editor first and ask if they might want to consider it. (Not accept it sight unseen, but accept the manuscript for consideration.) There’s even a publishing-specific term for that kind of letter: a query. To be fair, though, I never knew about queries until I started reading up on the publishing industry. It never occurred to me to write a query for the articles I submitted to academic journals, but I’ll start doing it now.

Monica Macaulay

3. Submit the paper. These days, journals are moving to online systems for manuscript submissions, and when you submit yours, it can feel like it’s going into an impersonal machine. Remember that there are human editors behind the system, and they do want to help. There’s no reason, after doing an electronic submission, not to contact the editor personally and let them know if you think a particular associate editor would be the best person to read the paper. As with the querying, it’s not an ethical problem, and it may save the referees and editors some time in evaluating your work.

I understand that different academic fields differ on accepted practice here, but the panelists agreed that for linguistics papers, you should not submit one to more than one journal at a time! Not only is it considered a breach of etiquette, one which could make it difficult for you to ever publish any other academic papers, but you cheat yourself of a chance to incorporate valuable feedback before you submit it to another journal.

Fine, but what if you submit your paper and it disappears into a refereeing black hole, and you never get an answer? Surely it would be OK to submit to another journal in that case? (And yes, I’m calling you surely.) Not yet. If you’ve been waiting a long time for an answer, it’s OK to send the editor a message asking about your paper. Who knows? Telling the referees that the author is curious about the paper’s status might be just the tool the editor needs to get the referees to hurry up. As stated above, they’re volunteers, so direct orders to finish the reviews don’t go over so well.

How long is a long time? The referees looked at each other briefly, and then agreed, “A long time.” They couldn’t agree on anything more specific after that, but one offered that you should wait “longer than you’d want to wait.” Yes, I realize you can get into an infinitely recursive loop if you follow this advice too closely, so you’ll just have to set your own cutoff point on this one.

Chris Potts

4. If editor decides it’s not appropriate for the journal, return to step 2. Not much to add here.

5A. If the editor sends it to referees and it’s accepted, you’re done. Not much to add here, either. But before moving to the other possible outcomes, just get used to the fact that getting referee reports is stressful, more so if you put as much effort into Step 1 as they recommend. The panelists themselves still hate reading referee reports on their own work. Let it sit a while, they advise, before you try to decide anything about what changes to make. With that said…

5B. If editor sends it to referees and it’s accepted on condition of revisions being made, make the revisions and you’re done. You’ll have suggested revisions from two or three reviewers, which are almost certain to disagree on various points. The editor should provide a cover letter giving their judgment and advice based on the reviews, and if they don’t, then ask for one. Some reviewers are, in fact, unfair, and it’s the editor’s job to temper the unfairness, to create a kind of “weighted harmonic mean”, and give you an idea how much priority to give to the overall areas for revision. Also, it sometimes happens that there’s some kind of discrepancy in the paper, or an organizational problem which surfaces as very different reactions from different referees, and it’s the editor who has to figure out what it was in the paper that sent the different referees in different directions, and give his best advice on what the real problem is that they’re reacting to.

Which of the suggested revisions should you make? You should do something with every single one of them. For one thing, it’s a form of mental discipline. Furthermore, when referees see that an author has blown off one or more of their carefully considered suggestions, written with an opportunity cost of lost reading, research, sleeping, or fun time, they tend to get grumpy. So with that said…

For the clearly good suggestions, of course make the appropriate revisions. For the suggestions that you don’t think are really needed, make the changes anyway! The referees have more objectivity about this than you do, so go with it and you’ll probably agree months or years later that the referees were right. The panelists were speaking not as editors or referees at this point, but from personal experiences involving their own papers from years back.

And what about the suggestions that are so completely misguided that following them would ruin the paper? If it’s clear that the referee has completely missed the point of what you were saying, in that case you should be even more sure to revise the relevant sections. If the referee managed to be led so completely astray, despite being one of the more knowledgeable and careful readers, then what hope does the ordinary journal reader have? The problem is probably with you.

On the other hand, what if the referee has understood you but disagrees, and there’s really no way to make the suggested change without destroying the paper? Again, here you should be even more sure to make a revision, along the lines of “An anonymous referee has suggested ….”, followed by a logical presentation of how this suggestion just won’t work. Again, if a referee has a problem with some part of your paper, then probably a good proportion of your readers will, too, so this is a good time to address their objections.

In response to an audience question about common problems that led to rejection, it was divulged that there was one “special category of rejections” for authors who refuse to make changes suggested not (only) by the referees, but by the editor themself.”

When you resubmit the paper, include a cover letter that highlights what you’ve done, on what pages, for each and every suggestion. It makes it easier for the referees who now have the task of reading your paper again from the beginning, and it shows them right away that you took them seriously. If you’re lucky, they’ll just jump to the parts you mention, say, “OK, that looks all right,” and give you a green light.

5C. If it’s rejected, revise and return to step 2. Why not just return directly to step 2? Two reasons. First, as covered in part 5B, the referees will have given some good suggestions, no matter how stupid, picky, or perverse they seem at first. Second, even though you’re submitting to a different journal, there’s still a pretty good chance that your paper be sent to at least one of the referees that it was sent to the first time, and the above remarks on grumpiness still hold. But this time, instead of one or two ignored suggestions, the referees will see that you’ve ignored many, maybe even all of their suggestions. You’ll be like the sneaky kids who get a no from one parent and instead of just sucking it up, try to go and get a yes from the other one. (That’s another analogy I’ve thrown in for free.)

Other advice that came in response to audience questions didn’t fit into the main outline.

Although simultaneous submissions are not OK, simultaneous queries are not a problem. Just don’t take a “throwing spaghetti at the wall” approach. Make sure each query is personalized to the editor you’re querying. (This, too, is standard advice from the non-academic publishing industry.)

What about length? First of all, if the journal has length guidelines, then follow them! Second, look at what’s been published in that journal. If the typical length is 30 pages or less, then your 50-page paper had better be damned good. As for short, there is no such thing a paper that’s too short, provided that it is long enough to present its content.

To get a perspective of how things work on the other side of the editor’s or referee’s desk, volunteer for a university’s working papers publication, or serve on an abstract committee. One referee credited doing that with teaching her how to write, and not write, abstracts. And if nothing else, you’ll get a firsthand look at how much arbitrariness there can be when you see how much good stuff gets rejected for dumb reasons. There can be a small comfort in knowing it can happen to better papers than yours.

The suggestion about submitting papers to conferences generated a few questions that led to some other advice. First of all, if your paper is accepted to a conference, think carefully before you agree to have it published in the proceedings. There could be copyright issues if you later want to submit it to a journal, and even if there aren’t, the journal might not accept papers published in conference proceedings. Be honest, and ask the editor. Ideally, the journal version of your conference paper should be viewable as a new paper, with a substantial amount of new material — at least half the paper. I agree with this point: It will also save people who cite your papers some trouble. On several occasions I’ve worked to track down two papers by one author, and found that they were basically the same, which made me feel silly for the time I spent getting both.

In fact, the panelists said, you should think of the conference version of the paper as a draft, which you might well do something further with one day. For that matter, the same goes for papers that have been published in journals. According to the panelists, it has even been said, somewhat jokingly, that graduate students shouldn’t publish on a given topic, because then they quit working on it! This led to a more general point: Whenever you produce anything, you should ask yourself whether it’s worthy of further development. Assume neither that you’re done with the topic for good, nor that you’re not. Grad students feel pressure to produce stuff, earlier and earlier in their careers, but in spite of that (or maybe because of it), it’s especially important to choose your battles, and decide which projects are the best use of your time and effort.

After an hour, I had to leave, but the questions were still going strong when I did. If other people who were there until the end have some other advice from the panel that belongs in this summary, send it along and I’ll put it in.