New APPS

Art, Politics, Philosophy, Science. Mostly Philosophy.

recent posts

about

    On the APA’s method for anonymizing papers for review

    I'm not sure that the APA has this right.

    The APA guidelines for submitting papers state:

    All papers are anonymously reviewed. Author's name, institution, or references pertaining to the identity of the author must be removed from the paper, abstract, notes, and bibliography. Papers containing such identifying references may be rejected.

    There are at least two ways that one might remove one's identity:

    1. One might leave in the references to oneself, but refer to oneself in the third person, e.g., "As Millstein (2009) argues, populations are individuals."
    2. One might delete all references to oneself, e.g., "As I have argued elsewhere (reference deleted), populations are individuals."

    As I understand the APA instructions and the clarification I received when I asked, we are being told to anonymize the second way rather than the first. However, it seems to me that the first way is superior to the second. If the referee is familiar with Millstein's argument that populations are individuals and the second methd is used, then she will know immediately that the author of the paper is Millstein. However, with the first method, she won't know if the author is someone who is endorsing Millstein (2009) or if the author is Millstein herself. So, I think that the first way does a better job of anonymizing than the second way.

    Or am I missing something here?

    Posted in , ,

    16 responses to “On the APA’s method for anonymizing papers for review”

    1. Jonathan Kaplan Avatar
      Jonathan Kaplan

      The first way is clearly superior, so much so that the second way seems entirely misguided by comparison. What would one do, in the second system, with a quote? And if one is big in the field, you’d be glaringly missing, wouldn’t you? I’m almost inclined to think that the person answering hadn’t actually thought this through, or didn’t understand the question, because their answer is so terrible.
      On the second (terrible) system, to actually anonymize things, I recommend redacting all references to all individuals cited or otherwise referred to in the paper, so people can’t tell which one is you by elimination. All references should therefore read [reference deleted] and you should refer to everything passively [it has been argued that]. Then no one will know!
      (Grumble grumble…)

      Like

      Reply
    2. Roberta L. Millstein Avatar
      Roberta L. Millstein

      Ha! Option 3! No references at all!

      Like

      Reply
    3. OUP Avatar
      OUP

      On the other hand, if you are not at all well known, method 2 will not give away much, while, in method 1, your insistence in discussing the views of One Unknown Philosopher (forthcoming) might be a tell-tale sign that you are her…

      Like

      Reply
    4. alan nelson Avatar
      alan nelson

      #2 is silly. Why is it even considered as an option?

      Like

      Reply
    5. Neil Sinhababu Avatar
      Neil Sinhababu

      Using the first person and ‘(reference deleted)’ is an (expletive deleted) way of anonymizing papers.

      Like

      Reply
    6. Roberta L. Millstein Avatar
      Roberta L. Millstein

      That’s true, although method #2 still has the problem with quotes that Jonathan Kaplan points out. Or maybe One Unknown Philosopher is not as unknown as the referee thinks. 🙂

      Like

      Reply
    7. James Avatar
      James

      Supporting OUP, when I read things in the dead-center of my AOS, I feel like notice it immediately when there’s even a minor anomaly in the relative frequency of references, as compared with the “industry standard.” Especially in more expository material. Like if a book keeps supporting its general exposition of some topic by reference to an essay in some edited volume (however legit that source may be) instead of X’s widely admired overview in SEP or a Cambridge Companion or Cambridge History of Philosophy (my AOS is historical).

      Like

      Reply
    8. Roberta L. Millstein Avatar
      Roberta L. Millstein

      Thanks, James, I think you’ve raised two important types of cases here. One is where the relative frequency of citations to one author is higher than the “industry standard.” Here I think the “industry standard” might be relative to AOS. I know that in philosophy of science I have seen large numbers of citations to a particular author who was not the author of the paper in question. Sometimes that is because the oft-cited author is the paper-author’s advisor, but I’ve seen other cases where the paper-author just wanted to use the oft-cited author’s ideas as a springboard.
      The second type of case is where the person is citing obscure Y most would cite well-known X I agree that such citations are a red flag. But perhaps (at least in most cases) they should be. At a minimum, X should be cited in addition to Y 🙂

      Like

      Reply
    9. matt w Avatar
      matt w

      What about this kind of case?
      “This raises questions that go beyond the scope of this paper. [See Weiner, forthcoming, for a full treatment of those questions.]”
      I often wind up with that sort of thing when I have a paper I’ve had to split in half, and I think here option 1 gives away the author of the paper, and option 2 doesn’t. Even if the paper’s appeared in print, I think option 1 is more of a giveaway than option 2; it’s always possible that the referee can’t guess exactly which paper I mean if I don’t supply the reference.
      Maybe the ideal way to do it is a mix of option 1 and option 2. If I say “As I argue elsewhere [reference deleted]” in one place and “For examples of this view, see Bigshot, Rockstar, and Weiner,” that might throw referees off the scent. (I recently participated in a non-philosophical event involving members of a small community writing things under temporary pseudonyms; one participant concealed her identity very effectively by thanking herself in the acknowledgments.)

      Like

      Reply
    10. Roberta L. Millstein Avatar
      Roberta L. Millstein

      I’ve also been starting to wonder if a mixed strategy wouldn’t make sense, not necessarily to “throw referees off the scent” but to let the author exercise her judgment about which anonymizing method would be more revealing for a particular citation. I agree that the sort of case you’re describing is pretty revealing of the author’s identity.

      Like

      Reply
    11. philosopher Avatar
      philosopher

      I think you are right to recognize that the practice of eliminating self-citations in a manuscript under review can serve multiple purposes. We do not want some well known philosopher getting a free pass through refereeing because no one would dare to reject their paper. It leads to lousy papers being published. And we do not want someone who benefits in no way from their academic identity having their identity known and being punished for their obscurity. “Blind” reviewing is meant to ensure that papers are judged on their merits, not their authors’ identities.
      I think in general authors should avoid references to other works of their own, like the example Matt raises:
      FROM MATT: “This raises questions that go beyond the scope of this paper. [See Weiner, forthcoming, for a full treatment of those questions.]”
      That really is irrelevant to the paper under review. It seems such citations are merely being used to drum up attention and citations to other works by the author.
      One could imagine (and I have heard of) a case where a referee says “the author is completely ignorant of so-an-so’s work which is relevant to this topic, and must be addressed,” where so-and-so is the author of the manuscript under review. This may lead the referee to recommend rejecting the paper. But this is where a good editor uses their judgment.
      Incidentally, I have generally used Method 2, in the original post, but I was recently explicitly told by a journal reviewing a paper, that I must use Method 1.

      Like

      Reply
    12. Sergio Tenenbaum Avatar
      Sergio Tenenbaum

      There is another case like Matt’s: when you’re simply acknowledging overlap with a previously published (or worse, forthcoming) paper or you’re using some apparatus/framework/etc. developed in another paper that no one else is likely to use. I greatly favour the mixed method; I think it is often quite clear to the author which method at each point would most likely reveal her identity.

      Like

      Reply
    13. Jennifer Nagel Avatar
      Jennifer Nagel

      I agree Method 1 is best, not least because it makes it invisible exactly what your own track record is. Established scholars don’t get to say “as I have said before, in my many publications… “. (In my case, saving themselves from irritating me.) I agree self-references to one’s own unavailable forthcoming work should also be cut; where work is available at philpapers it’s fine for anyone to cite it third-personally. Also: whatever APA might have told you when you got in touch with them, it’s worth remembering that the APA committee members who are evaluating your anonymous paper won’t even be able to tell if you use Method 1.

      Like

      Reply
    14. Sara L. Uckelman Avatar
      Sara L. Uckelman

      I also prefer #1 to #2 in most circumstances (others have pointed out circumstances where #2 is useful); but I’ll note that the one time I was asked to do #2 and replied by pointing out I’d done #1, the response I got was favorable, and indicated that the editor hadn’t really realized #1 could be done.
      So, if you’re asked to use the second method, it’s always worth finding out if the first, or a mixed, method is acceptable.

      Like

      Reply
    15. Dominic Lopes Avatar
      Dominic Lopes

      The Pacific Division doesn’t have a policy preferring one of these methods. Both seem to have been used and have worked well enough for us to get the job done. Probably each has circumstances where it’s somewhat more effective, but in our experience both are satisfactory.
      I should also say that if an author has taken reasonable, customary steps to anonymize their paper and a reader nevertheless recognizes the author, then we’d simply reassign it to a new reader (we wouldn’t decline it!) This happens with about one or two in a thousand papers.

      Like

      Reply
    16. Roberta L. Millstein Avatar
      Roberta L. Millstein

      Dominic, thank you very much for clarifying the Pacific Division’s policy. The wording is identical to that of the APA more generally, so perhaps some rewording of the instructions in the future might be helpful.
      And thanks to everyone for the thoughtful discussion. I think I am now convinced that the author is in the best position to choose #1 or #2 or a combination. Maybe some journals will take up that policy more explicitly instead of requiring one or the other.

      Like

      Reply

    Leave a reply to Neil Sinhababu Cancel reply