Papers hiding and being seen

Mon 18 Dec 2006 12:41 PM

James Beebe posts at the group blog Certain Doubts regarding double-blind peer review and posting preprints on the web. As he notes, putting a preprint of a paper on your website before it has been accepted at a journal makes it possible for referees to search the web, find the draft, and identify you as the author. This would break the protocol of the peer review.

Many people note that it would be irresponsible of referees to search the web to divine the author of a piece, unless perhaps they did so after deciding whether or not the piece should be accepted for publication. The caveat to this, of course, is that referees may or may not do the responsible thing.

The following is my attempt to distill the discussion to its essentials. Alas, it has grown rather long-winded.

Is this a problem?

People seem to agree: Having referees know the identity of the author will help established, big-name philosophers and hurt neophyte unknowns.

I am not sure that this is true. A reviewer might accept a paper by Famous McX at Prestigious University as a shoe-in for publication or might expect more out of it because McX can clearly do better work. Conversely, a reviewer might discount a paper by Cletus O'Nobody at Podunk State or might be impressed that O'Nobody has managed to string two thousand words together. Since all sorts of crap gets published, a referee might accept the effort as probably the epitome of O'Nobody's deadend career.

Regardless of how it works in practice, I suspect that people are just wielding anecdotes and intuitions when they speculate on the effects of this.

Many people have noted that the problem is not unique to the internet. Conferences give referees a similar opportunity to break the seal of confidentiality. Since many conferences put schedules and papers on the web, a paper that is presented before it is accepted for publication may be on the internet anyway. (The paper I presented at one of the first conferences I ever attended is still on the web. A successor to it was eventually published, but parts of the original would be best lost in the past.)

Moreover, referees who search for the title of the paper will be able to connect you with it if you so much as have a list of works in progress on your website. Avoiding any on-line association with a paper would require some serious austerity.

In the comments over at Certain Doubts, James Beebe concludes, "Although it is not clear how the institutions of our profession should respond to this problem, I've received a very clear answer on how I should respond: I have taken all of my articles that are works in progress or under review off of my website." This might avoid problems, but at a definite cost.

The benefits of being online

Any disadvantage must be weighed against the advantages of having drafts on the web:

A. Publication is most important to junior philosophers who need it to get a job or to get tenure. While a manuscript is sitting in the inbox of a referee, the job market and tenure clocks are ticking. Having a preprint on your website is a way of publicizing the work so that employers can know about it now. (This point has been made many times by Brian Weatherson.)

B. Having a preprint on the web can result in useful feeback. This has happened for me a few times.

C. Having a preprint on the web establishes priority. Even if your article is not published until later or, with a backlogged journal, much later, the first post establishes a date at which you publicly advocated the ideas or arguments therein.

Priority led me to release forall x a few months earlier than I would have otherwise. Once I had come up with the title, I wanted to claim it.

I really should be more careful about putting the date of the first web version of a paper in the footnotes of later web versions.

What is one to do?

Given these considerations, there are several options.

Be seen: Put papers online and let referees find them. This is just to face any negative consequences with 21st-century, information-wants-to-be-free bravado.

Hide from everybody: This is the solution Beebe seems to have chosen. For reasons given above, I think that it is too high a price and not likely to work anyway.

Hide from robots: Later in the comments at Certain Doubts, Richard Zach recommends telling search engines not to index your paper. This seems like an odd half-measure, since it nullfies most of the advantages of having the paper online at all. Moreover, mentions of the paper that link to your page will still be indexed.

Be seen, but wear a fake moustache: Brit Brogaard and Robbie Williams advocate submitting a paper to journals with a different title than its online preprint counterpart. This means that referees will not be able to suss you out just by searching for the title of the submission. More determined referees might still find your paper, but this is probably true regardless.

I think I may adopt this latter strategy, because it does not seem to have any real downside. I often end up changing the title of a paper once or twice between first draft and publication anyway.