69

Does anyone have any good advice for obfuscating one's writing style when reviewing papers? I have a fear that I would inadvertently use the same turn of phrase in a referee report as in some published work, thus revealing my identity, and I'm curious if there is a good general practice to follow to mitigate this as much as possible.

  • 31
    One thing to note is that if you're not intending to say anything you wouldn't say to the authors in person, then you have no reason to hide your identity anyway - in that case it really doesn't matter whether they know or suspect that it's you, and I've heard that at least in some fields, some reviewers even sign their reviews. (In other fields, open reviews are the norm, with the reviews being published under the reviewers' names alongside the article.) It might be that in your particular situation you have a reason to want to protect your anonymity, but this seemed worth mentioning anyway. – Nathaniel May 6 '16 at 9:54
  • 8
    There is also a technical aspect to this question. At least in math, referee reports are often passed along as latexed pdf files straight from the reviewer. Without care, the pdf sometimes contains identifying information: for example, my pdf files show that they were compiled on a not-too-common linux distro, which would give me away to at least a few colleagues. Sometimes the default language shows too, depending on the setup. Anyway, I usually compile on a different computer just to be safe. – Mark May 6 '16 at 14:33
  • 31
    The best way to reveal your identity in reviewing a paper: ask the author to cite one (or more) of your papers. – GEdgar May 6 '16 at 17:18
  • 8
    @GEdgar Thus, the best way to obfuscate your identity is to ask the authors to cite someone else's papers :) I often do point out 3rd party papers which I think are missing and sometimes I do wonder whether the authors conclude I'm the uncited author. – luispedro May 7 '16 at 14:10
  • 9
    @HonzaZidek Quite on the contrary. Staying anonymous can help staying honest. That's a fact and we all know it from normal life and ignoring it just because one feels it should be different doesn't make it go away. A truth is independent from its source. To assume one needs to know the identity of the author of a review just to take the review into account appears weird to me, and I'm a software developer as well. It's not the field. – Alfe May 8 '16 at 23:33

11 Answers 11

85

It's not worth it. Few waste time trying to unblind their reviewers to take retribution for bad reviews. It's just not worth it. The average journal submitter doesn't have enough power to do damage to a random reviewer.

If you need a strategy, trying passing your text through Google Translate to Spanish and back to English. It'll probably be garbage afterwards, but it'll definitely lose any idiomatic turns of phrase that might identify you.

  • 44
    The average journal submitter doesn't have enough power to do damage to a random reviewer. I am working in a very specific field with at most 10 main contributors, most of whom know each other quite well. I have a somewhat distinct style as compared to other guys in the field (at least I believe so), which makes me identifiable. If I write a negative review and I am traced, I risk getting a certain reputation within that group. Subsequently, I may expect some form of revenge (even if subconscious) when it is their turn to review my manuscript. I am not sure I can afford that. – Richard Hardy May 6 '16 at 8:36
  • 8
    @RichardHardy He did in fact provide you with a workable method to obfuscate your writing style. – March Ho May 6 '16 at 8:45
  • 8
    I'd like to also agree with the "It's not worth it" answer, but would like to point out that even for a young PhD student, it may make sense to try to figure out who wrote the review. This is the case if the review is rejecting due to the topic not being interesting enough, the results not being strong enough, etc. This is somewhat subjective, and the student may want to avoid re-submitting the paper to venues where the same reviewer is likely to be taken. In CS, the researcher may want to save time by identifying conference PC members that should not see the paper again. – DCTLib May 6 '16 at 8:47
  • 13
    I had a reputation with one editor for awhile of being a good reviewer to turn to when he had a borderline paper or a paper with mixed reviews that needed a strong review to steer the authors in the right direction. I don't think that ever came to the knowledge of any authors, and I didn't really care. If your field really is that tiny, then you should cultivate the reputation for being a strong reviewer not just a negative one. Thoughtful, precise, and well backed up reviews are better than ones that simply slam the authors and their work. – Bill Barth May 6 '16 at 12:47
  • 3
    “Few waste time trying to unblind their reviewers to take retribution for bad reviews.” — “Retribution” is a strong word. But in my field it’s common to try guess at the reviewers’ identities (especially in the case of a negative review). And conscious or not, that information is definitely taken into account when judging your peers and choosing who to collaborate with, how to interact at conferences and, yes, how to review their papers in turn. I wouldn’t even condemn this; it’s simply human. – Konrad Rudolph May 9 '16 at 8:49
69

Assuming that you really want to do this, here are some strategies that should not strongly diminish the readability of your review:

  • If you are from an English-speaking country: Use the spelling conventions of another English-speaking country (e.g., British instead of American English).
  • Switch to a strongly different punctuation convention, e.g., the French one:

    On page 2, the authors write : « We could not find any evidence for this. »

  • If you typically tend to write long sentences, split sentences up as much as possible. If you tend to write short sentences, make them slightly longer than you’re comfortable with.

  • Try to use certain words particularly often, e.g., particularly. On the other hand, if you are particularly fond of certain words, avoid them entirely (of course, the main challenge is to become aware of this in the first place).

  • If your field has two competing notations for something, use the other one.

  • Unless this is a feature of your native language, drop articles, in particular if this does not inhibit understanding your review. On the other hand, if your native language is not particularly fond of articles or does not have them at all, use an article wherever possible.

  • After you wrote your review, try to replace every rarer, non-technical word by a synonym found in a thesaurus.

  • Exclusively use either the simple past or the present perfect. Or: Exclusively use either progressive forms or non-progressive ones.

  • To make you seem German, capitalise some arbitrary nouns, but do not capitalise adjectives derived from proper names:

    I disagree with the use of bayesian Statistics.

  • 33
    Very funny suggestions, and very good ones - to what lengths will one go to - even legitimately - hide ones presence :-) But, in any case, "nullum inultum remanebit" - write your reviews (also your negative ones), as if you will be, at some point, found out. Always. No exception, even if you are angry at the author (for stupidity, intransigence, or not citing you). – Captain Emacs May 6 '16 at 11:49
  • 12
    Dropping articles and excessively using articles are both symptoms of speakers whose native languages do not have articles. I’m afraid the only effective way of concealing such origin of the reviewer is to learn how to use articles properly, which may be quite difficult. – Emil Jeřábek May 6 '16 at 16:26
  • 4
    The typical American trying to sound English will sound like a Dickens character. Which I guess is a pretty good way to conceal your usual writing style. I will say though that no even marginally competent German writer would capitalise nouns - that's a trivial rule that everybody internalises quickly. You recognize German speakers by how they structure their sentences - that's really hard to hide (extra points if you manage to make whole paragraphs single sentences :-) ). – Voo May 8 '16 at 15:44
  • 2
    @Voo: Being German and a moderator on German language, I see quite a lot of English writing by Germans and capitalising the occasional noun is quite a common mistake – just that you know how it’s done correctly, does not mean that you will avoid the mistake all the time. I even notice it happening to me occasionally. – Wrzlprmft May 8 '16 at 16:49
  • 4
    You forgot to mention: From all of these suggestions above, pick any four to avoid to be traceable by being absolutely untraceable ;-) – Alfe May 8 '16 at 23:40
21

To add to Bill's answer, most scientists would agree that it is not productive to try figuring out who a reviewer may be, and consciously avoid thinking about it too hard. I make that a point when talking to my students and postdocs. I'm not sure most are that explicit about it, but I've never spoken to anyone who actively tried to find out who a reviewer might be, but I've spoken to many who agree that it's the wrong thing to do.

The whole point being: nothing good can come of it if you know who your reviewers are. But you can violate the spirit of peer review and alienate colleagues if you try too hard. So just let it go.

For you, this means: don't try too hard to obscure your identity.

  • 1
    Would you have any response to my comment under Bill Barth's answer? – Richard Hardy May 6 '16 at 8:40
  • 2
    Yes. You are doing your professional duty when writing a review, and everyone understands that. So, when you write a negative review, then presumably because you had good reasons. You will get a bad reputation if you're confusing being an objective reviewer with being a princess who wants it her way over style issues. But if you have good arguments to reject an article, or to suggest a major revision, then nobody can really argue about that, or fault you for that. Even luminaries understand that they don't have a right to get their papers published the way they were initially submitted. – Wolfgang Bangerth May 6 '16 at 11:34
  • 5
    @WolfgangBangerth you are assuming a lot about a small sample size that the author is closely acquainted with when you say "everyone understands that" and 'nobody can really argue about that". I've been in groups doing math and had people get angry at me because I pointed out holes in their arguments. Holes which would have cost them points on their homework and made them look dumb. I was doing it completely professionally and politely, but they were reacting in an irrational manner, because sometimes humans do that. (I grant that in the abstract, you are correct, just to be clear.) – msouth May 7 '16 at 21:52
  • 1
    @WolfgangBangerth, in the perfect world this would be true. But then we wouldn't need reviewers, since there wouldn't be any mistakes in any of the research. – user24098 May 9 '16 at 7:20
  • @dan1111 -- that's not a very pragmatic point of view. Papers become better by peer review, simply because authors get to see another point of view. This is not a question of "perfect" or "imperfect" worlds, just the reality of writing. – Wolfgang Bangerth May 10 '16 at 2:44
18

I would suggest that you forget your current train of thought. Openly sign your review and write it as if you would tell it to the author's face. If your colleague wrote a bad piece, then help him by pointing out how it could be improved, in a way that does not make him lose his face. If you are working in a falsifiable field and he made an objective error, then explain that error in an objective way (stating facts etc.).

Your new problem is then to figure out how to write your reviews like that, and there are plenty of techniques for that (there should be plenty of resources out there about how to put out criticism without the hurt factor). This is much more healthy and probably also easier than obfuscating your identity.

  • "Write your review as if you would tell it to the author's face". This should be upvoted 10 times. – Erel Segal-Halevi May 8 '16 at 12:21
10

If you are really serious about maintaining your anonymity, you should keep in mind that not just the style of your writing, but also the content of what you write, contain major clues as to your personality and hence (given how small your field is) to your identity.

I therefore suggest choosing a random subset of your recommendations to the author(s) and flipping them: e.g., if you were going to suggest making the paper longer, tell them to make it shorter instead; if you thought of suggesting to collect more data, tell them they have too much, etc.

Finally, for extra safety, also flip the accept/reject bit. With this technique, your anonymity will be virtually guaranteed.

  • 8
    Upvoting under the premise that this is a humorous answer that wants to contradict OPs question. – Turion May 7 '16 at 9:41
3

As others have said, it really isn't worth trying too hard. But an easy measure to take might be to switch from American -> British English (or vice-versa). Of course, this will only work if your review happens to contain enough words that will indicate your dialect.

  • 1
    No. It is far too easy to inadvertently say something you don't mean. 'This paper is quite good.' Will you remember which side of the Atlantic you are meant to be from? Are you even aware of the difference in meaning? And this is just one example. – cfr May 8 '16 at 3:17
  • I always mention "shrimp on the barbie" in all of my reviewer comments, for precisely this reason. No one has ever figured out who I am. – user24098 May 9 '16 at 16:03
2

If you really want to do this, become a better writer. Seriously. Take writer training classes, including the ones aimed at fiction writers. Read books on writing style, non-fiction as well as fiction. Study writing styles. When you read something, take some time to note the phrases, grammar, etc., not just the content. Understand how words and grammar bring meaning across.

Also try some role playing. Go to impro theater classes, or play a RPG. Every now and then, imagine being somebody else, and talking and writing like them.

Once you have developed a sufficient understanding of prose to identify how exactly your style differs from others, you will be able to create one or more personas with a distinct writing style. Write your reviews as this persona.

This requires quite a bit of commitment and you will need to learn considerable skills in that area. On the plus side, you will not only be able to hide your identity, but you will also gain a lot of mastery in writing, which is a very valuable skill for a person judged on the quality of his written publications.

It is up to you to decide if the considerable effort is worth it.

1

To review in a less personal way, I avoid the "I", and use the third person writing. I thus write "the reviewer suggests", "to the reviewer". Accordingly, I write "in the paper", "The authors". This helps me take a more distant look at the work under review.

I learned this from a colleague, received a review in this style. Thus, at least three reviewers use this technique (I won't be detected so easily).

  • 2
    The comment above appears to suggest that using the third person and other such indirect references will make the review less personalized. This is a technique generally to be encouraged. Author of said comment is to be commended and should receive the adulation of the community. :) – msouth May 7 '16 at 22:17
  • 1
    The reviewer conveys interesting insights and the author wishes to acknowledge his useful contributions to the final version of the post – Laurent Duval May 9 '16 at 18:41
  • Based on the general reaction one would expect (on the assumption that said reviewer is human), it is to be believed that the review would appreciate the author's acknowledgement. – msouth May 13 '16 at 1:31
1

Anonymity is ultimately about separating yourself from your speech. Use a standard and concise style that avoids all biases, irrelevance to the purpose of the work under review, and unprofessional language or conduct.

Doing so will additionally elevate any intellectual endeavor and need not be done for privacy purposes alone.

1

You could use an automated paraphrasing tool such as Spinbot.

The text will need a little bit of editing afterwards... here's what it made of your question:

Benefits anybody have in any way guidance for jumbling one's written work style when assessing papers? I have an apprehension that I would accidentally utilize the same turn of expression in a ref report as in some distributed work, consequently uncovering my character, and I'm interested if there is a decent broad practice to take after to alleviate this however much as could reasonably be expected.

1

Many years ago my colleague was asked to review an article written by a well known person in our field of research, who was familiar with the writing style of my colleague.

There were serious issues with the paper and my colleague intended to write a very negative review but wanted to maintain anonymity.

We discussed the issues and I wrote the review. Problem solved.

The article was rejected, but later published in another journal.

protected by eykanal May 10 '16 at 13:16

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.