By Catarina Dutilh Novaes

In recent weeks, there has been much discussion on journal editorial practices at a number of philosophy blogs. Daily Nous ran an interesting post where different journal editors described (with varying degrees of detail) their editorial practices; many agree that the triple-anonymous system has a number of advantages and, when possible, should be adopted.* (And please, let us just stop calling it ‘triple-blind’ or ‘double-blind’, given that there is a perfectly suitable alternative!) Jonathan Ichikawa, however, pointed out (based on his experience with Phil Studies) that we must not take it for granted that a journal’s stated editorial policies are always de facto implemented. Jonathan (correctly, to my mind) defends the view that it is not desirable for a journal editor to act as a (let alone the sole) referee for a submission.

With this post, I want to bring up for discussion what I think is one of the main issues with the peer-reviewing system (I’ve expressed other reservations before: here, here, and here), namely the extreme difficulties journal editors encounter at finding competent referees willing to take up new assignments. Until two years ago, my experience with the peer-review system was restricted to the role of author (and I, as everybody else, got very frustrated with the months and months it often took journals to handle my submissions) and the role of referee (and I, as so many others, got very frustrated with the constant outpour of referee requests reaching my inbox). Two years ago I became one of the editors of the Review of Symbolic Logic, and thus acquired a third perspective, that of the journal editor. I can confirm that it is one of the most thankless jobs I’ve ever had.


I am now caught between authors who are (rightly) eager for good referee reports produced in a timely fashion (some of whom are junior colleagues who desperately need the publications), and over-worked referees who simply cannot say ‘yes’ to each referee request they receive. It is not uncommon that I have to contact at least 10 people before securing two referee reports for a paper. As I see it, my main mission as an editor is to ensure that authors get high-quality feedback, such that they will have gotten something out of it even if the paper is rejected. But reliable, thorough referees are a very rare, precious commodity, which must not be overused. Good referees are usually very competent people who already have a lot on their plates. One thing I do to circumvent the issue is to contact more junior but very capable people (say, at post-doc level) to act as referees, the thought being that they will be slightly less swamped with other obligations than more senior colleagues. But often what one needs is an experienced pair of eyes, and then we are back to the over-stretched senior colleagues…

The most frustrating aspect of the whole process are the people who do not even bother to reply to a referee request. Ok, this is one way of saying ‘no’, but besides being rude and uncollegial (after all, we all need the peer-reviewing system to work minimally well!), it holds things back substantially, as I usually wait at least a week or two before moving on to the next possible referee. So, my plea to all of you reading this: if you receive a referee request, please reply within a day or two; if you cannot accept the assignment, please suggest alternative names.

However, getting people to accept a referee request is only the beginning of the story; it remains to be seen whether the referee will then indeed produce a report (many quit without notice), whether the report will be produced timely, and whether it will be a useful, informative report. At times, when it’s been months since the initial submission and there are still no clear prospects of at least two reliable reports being produced, it may well happen that the editor will choose to produce a report herself; it is not ideal, but not uncommon. (Of course, this is something very different from Jonathan Ichikawa’s experience.)

So the point of this post is to reveal yet another aspect in which the peer-review system is very fragile. My impression is that authors, especially the more junior ones, find it hard to understand why it often takes so long before they hear back from editors on their submissions. Although it is definitely the case that some journals simply do not handle matters in an efficient, professional way, hardworking, serious editors face the huge challenge of getting referees to produce helpful reports in a timely fashion. In comments, I welcome suggestions from (more seasoned) journal editors on how to deal with this challenge, including the emotional stress involved in being stuck between eager authors and over-worked referees. (Seriously, sometimes this keeps me awake at night…)

—————————————————-

* Notice however that in many disciplines, for example mathematics and physics, anonymous refereeing is not the norm. Unfortunately, the RSL does not have a policy of mandatory anonymous submissions, so it is not even double-anonymous by default (the author chooses whether to submit anonymously or not).

Posted in

18 responses to “Reliable referees: a rare commodity”

  1. Dave Ripley Avatar
    Dave Ripley

    Amen!
    As an editor for the AJL, I have had a truly staggering percentage (over 80%) of my referee requests simply go unanswered, and this is the single biggest source of delay in the papers I’ve handled there. Having just taken up editing at the RSL as well, where there will be a higher volume of papers on my plate, I’m keen to have some ideas for how to find referees, and so I’m looking forward to this thread.
    I believe, from a variety of conversations over the years, that there are a number of people who’d make great referees who simply aren’t being asked. My difficulty is in finding these folks: my limited experience suggests that most philosophers don’t respond at all to referee requests. (For me, like for you, each unanswered email usually costs about another week.)
    It’s recently occurred to me that http://philwiki.net is potentially a good resource for coming up with people to ask who might not spring to mind otherwise; I haven’t had a chance to put this into practice yet.

    Like

  2. Sara L. Uckelman Avatar

    Indeed, you sure do get a different perspective whether you’re sitting in the author, referee, or editor’s chairs (I rotate through all three, depending on the day of the week). Last spring I pondered why referee reports take so long, and came to the conclusion that, for me, they shouldn’t. Reviewing my actual practice and the timeline that I take in writing a report made me realize that, for me, most of the time taken in writing a report is procrastinating doing that. So I now make a special effort not to do that, and I’ve substantially cut my turn around time for reports. (This doesn’t hold for last term, though, as you know well why, Catarina. 🙂 )

    Like

  3. Dan Hicks Avatar

    It seems like there’s a compelling case to be made for including “how to be a good peer reviewer” in grad student professional development (along with “how to give a good talk,” “how to be a good teacher,” and standard things like “how to write a good paper”).

    Like

  4. Anonymous Avatar
    Anonymous

    I suspect the situation would be much improved if there were a system for people to volunteer to referee papers in their areas of specialization. I have a Leiter-top-5 PhD and multiple publications, and I have only been asked to referee once, ever, which I did on time with a very thorough multi-page report that the editor appeared to appreciate very much. I even sent my CV to a high-volume journal a few years ago offering to be a referee, and it has resulted in zero requests so far.
    IMO, exclusionary practices are very much to blame here. It’s not like editors can’t ask junior people–even ones they don’t know–to referee while also pursuing senior referees, just in case the junior people happen to produce timely and high-quality reports.

    Like

  5. Eddy Nahmias Avatar

    There is so little incentive to accept (and then do) referee reports. I try to do 6-12 a year motivated by the good of the profession. But I think research universities will have to find a way to give more credit for this professional service.
    For my grad seminars (and advanced undergrad seminars) I have students do (graded) referee reports on peer’s papers and it is very successful in helping them learn how to do them (and understand that it’s part of the profession), help each other improve their papers, and see problems in their own papers by seeing them in other’s. With their peer reviews and extensive comments from me, I find students’ papers typically improve a full letter grade or so.

    Like

  6. Brian Weatherson Avatar

    Michigan is starting to do something like that, though how well we do it remains to be seen. But it is certainly true that being a good reviewer, commenter etc is an important part of being a professional these days, and it is something that could perhaps be taught.

    Like

  7. NYU Grad Student Avatar
    NYU Grad Student

    I have only anecdotal evidence for this, but my impression is that graduate students are a largely untapped resource. I’m a late-stage grad student, and I’ve only been asked to referee once (and I suspect that I was recommended to the journal in question by a friend).
    There may be concerns that grad students (particularly those who haven’t published much or anything) aren’t experienced enough to be responsible choices as referees. Two responses to this concern: (1) I think it’s probably overstated: there’s often really not much difference in philosophical experience between, say, late-stage grad students and post-docs; (2) there may well simply be a trade off here between the time it takes to find referees and the level of experience demanded. And (in light of the insane review times at many journals) it’s not at all obvious that the current equilibrium is optimal.

    Like

  8. Matt Avatar

    On one of the Daily Nous threads about journals some time ago, a few people who are (from their CVs, which I looked at) pretty clearly highly competent philosophers said that they had never been asked to referee. They were not senior, but had been teaching for a few years at least and had some good publications. It was a surprise to me, as I get asked a pretty fair amount. (A fair number of the requests I get are not strictly in philosophy, though, but from other fields that interact with some of the areas I write in, and sometimes publish philosophical work.) Now, I suspect that it’s harder to find competent referees in logic than in fields like ethics or epistemology, but it does seem likely that there are un-tapped populations available. Perhaps asking for volunteers as suggested by anonymous 5 above, would help. (It does seem very odd to not reply back to a request at all. The only think that I can imagine is that I know I sometimes lose track of emails if they slide off the front of my email system, so if people don’t reply quickly, but get lots of emails, that might be happening.)

    Like

  9. Dave Ripley Avatar
    Dave Ripley

    Just to follow up on the last suggestion: I’m hesitant to simultaneously ask more potential referees than I need for reports, for two reasons.
    First, on the (very slim) chance that I actually get more reports than I need from doing this, I’ll have wasted what is a very scarce resource: people who write referee reports.
    Second, if this were to become accepted practice, I think an even smaller percentage of those asked would respond. (“Why bother answering this? They must’ve asked more people at the same time.”)
    But I absolutely agree that it would be great to give untapped potential referees a place to volunteer.

    Like

  10. Sara L. Uckelman Avatar

    There are also a few things that people can do to raise their profile, and thus the chances that they’ll be found and asked to referee:
    * Make sure you have a webpage with an up to date CV and list of research interests, and current email. It’s amazing how many people (a) don’t have websites or (b) don’t have an easily findable current email on it.
    * Make sure your philpapers profile is up to date, again with current papers, current research interests, current email.
    * If you don’t have an academia.edu page, consider making one and populating it with your research interests.
    When I am searching for referees, my selection is generally drawn from a combination of people I know, people whose work is closely related to the submitted work on the basis of the bibliography, people I find via google, people writing on the same topic found via philpapers, and people who list that the relevant area as an area of research on academia.edu.
    As a referee, I’ve also noticed a strong correlation between submitting to a journal and being asked to referee for it within the next 6 months or so. So, another way to raise your profile as a potential referee for a journal is to submit to it!

    Like

  11. Charles Pigden Avatar

    Isn’t there is a problem with Sara’s policy? Promptness (not to mention thoroughness) as a referee, leads to procrastination wrt to other tasks. I have two (long-delayed) papers to complete within the next fortnight, a course proposal that needs my comments, plus duties associated with the new teaching year (classes start in less than a fortnight.) Since the new year I have refereed four papers, mostly with a rapid turn-around time, in two cases within three days of the original request. In each case I gave feedback, the shortest report being 443 words long, and the longest nearly 1800, with three of the four reports being multi-page affairs. One took me a day and the others two days. Consequence: I am further behind than ever.
    On the substantive issue I have blogged on New APPS before about the structural reasons why the pool of referees is rather restricted, being largely confined to well-cited philosophers who get asked to do a lot more refereeing than their own work tends to generate. You tend not to be asked if you don’t get cited, most professional philosophers are relatively uncited, ergo …
    (http://www.newappsblog.com/2014/05/the-many-problems-with-peer-review-yet-again-and-some-proposed-solutions.html?cid=6a00d8341ef41d53ef01a3fd14cf23970b#comment-6a00d8341ef41d53ef01a3fd14cf23970b.)
    The fact that a minority of the profession is getting asked to do a majority of the work, leads to burnout and to bad behavior such as not answering emails. (The conscientious but burnt-out referee feels that she ought to referee that paper, hence she does not refuse, but she can’t bring herself to actually referee it, hence she does not reply in the affirmative. Since she feels she shouldn’t say ‘No’ and since she can’t bring herself to say ‘Yes’ she ends up not replying at all. Is this bad? Yes.
    Is it humanly understandable? Very much so. )
    One of the suggestions being mooted is that over-extended referees should delegate to their juniors, including post-docs and even graduate students. But you can’t shoulder–tap your juniors if you don’t know enough of the right shoulders to tap. And that’s pretty much my situation. In many cases I would have had no idea who to recommend as a possible a substitute if I felt too overwhelmed to undertake the task. The suggestion, to my mind, presupposes a certain model of what an in-demand referee is likely to be like. She will work in a large department near other large departments, she will have graduate students and/or postdocs working in her areas and she will be a regular conference-goer with a wide circle of acquaintance amongst those who share her research interests. None of this is true of me. You you can be an in-demand referee whilst living at some distance from the main metropolitan centers, without being much of a conference-goer, whilst being a member of a small department and without having many graduate students of your own. I referee for thirty-five journals, and I get asked to referee papers on the following topics: imperative logic, general meta-ethics, especially the error theory, Hume’s ethics generally, Is and Ought (sometimes wrt Hume, sometimes not so much), conspiracy theories, truthmaker theory and negative facts, modal realism, applied epistemology and renaissance conceptions of the will (that one was a bit of surprise!). In the last five years I have had only two graduate students who I would have thought competent to comment on any of the papers I have been sent, not because I haven’t had good students (there haven’t been many of them but they have been good!) but because, for the most part, they have not been working in the right areas.
    Now, perhaps I am an outlier statistically (as well as an outlier geographically), and most overworked referees are much more like the paradigm that the ‘go for graduate students’ suggestion presupposes. But if I am not, sheer ignorance may often be an obstacle to suggesting a suitable substitute.

    Like

  12. Matt Avatar

    Looking through the Daily Nous thread again, a common theme seems to be editors who treat referees poorly – not following their suggestions, editing the comments in unacceptable ways (at least once supposedly putting words into the referee’s mouth calling for the opposite conclusion of the referee), and so on. If I knew that this sort of thing happened at a journal I’d been asked to referee for, I’d certainly be very hesitant to referee for it, and if it had happened to me personally as a referee, I wouldn’t serve again and would encourage others to not do so, too. Now, I don’t know how common this is, and I certainly have no reason to think that Catarina engages in any of these bad practices, but I do wonder if they are a contributing factor in the difficulty of finding referees.

    Like

  13. Charles Pigden Avatar

    Having been, perhaps unhelpfully, explaining why the problem exists and why it is difficult to fix, let me make a constructive suggestion (though I suspect that this is something that some editors are doing already). One way to find suitable referees is to use PhilPapers which routinely lists publications on the ‘same’ topic under every posted paper. This is obviously done automatically and the program that selects the papers to list isn’t all that smart, but a clever editor could make creative use of a relatively dumb mechanism. Check the topic of the submission, scan (or consult your memory) for a related paper then go to the relevant page on Philpapers and work you way down, looking for the authors of OTHER related books or articles. That way you can use Philpapers to widen you search from the first article that springs to mind, which will usually be by the overburdened and possibly unresponsive Professor Mutch-Cyted.

    Like

  14. Catarina Dutilh Novaes Avatar

    I’m also shocked by these stories at Daily Nous. I doubt that this is causally related to referees not responding to requests though, as the referees I’m talking about have generally never refereed for me before.
    However, I must say that an editor will at times overrule the advice of a referee, in particular if she/he received conflicting reports. Sad to say, but it’s not uncommon to receive overly nasty, unhelpful reports, and a good editor should be able to look through it. (In the other direction, it hasn’t happened to me yet to overrule a positive report.)

    Like

  15. Catarina Dutilh Novaes Avatar

    And thanks all for the helpful comments! I agree that it would be wonderful to be able to find untapped resources, and this is precisely the reason why I often ask post-docs (and they are often the ones who write the best reports). However, there is often a component of trust involved (as an editor, I often handle papers on highly technical topics, which my expertise does not really cover), and so I still prefer to turn to the advice of people who I know to be competent. In the case of more junior people, this will mean people I’ve seen presenting papers, or with whom I have some kind of institutional connection (e.g. the people at MCMP, where I am an external member), and that of course narrows down my options quite a bit.

    Like

  16. anon Avatar
    anon

    Just before seeing this thread, I posted something relevant about this on a philosophers’ cocoon thread about the possibility of developing a comments exchange on works-in-progress. It seems relevant here too, so I’ll go ahead and cross-post.
    This is a slate article by Schuman — she argues that the lag times in peer review would be fixed by requiring authors to referee one paper for every paper they submit (i’m hoping/presuming double-anonymity would somehow be preserved). Maybe there are downsides I’m not seeing, but I don’t think it’s a bad idea at all(!)
    http://www.slate.com/articles/life/education/2014/07/the_easy_way_to_fix_peer_review_require_submitters_to_review_first.html

    Like

  17. a postdoc Avatar
    a postdoc

    If you think postdocs write the best reports, you haven’t seen mine.
    Rather than one group being less busy than the other, perhaps junior and senior academics just spend their time on different things, like applying for new jobs and doing (almost entirely thankless) refereeing versus hiring new colleagues and editing journals.

    Like

  18. Anonymous Avatar
    Anonymous

    Just to emphasize a theme: it occurs to me that all resources would be rare commodities if people only looked in places where they already knew the resource could be found.

    Like

Leave a reply to Catarina Dutilh Novaes Cancel reply