So someone submits a manuscript to a journal, and the editor sends me an email asking me if I’d be willing to review it. Once I agree to review the manuscript, I can then obtain a copy, usually through a secure website. In most cases, my review will be anonymous, unless I wish to reveal my identity to the author. When I receive the manuscript, I am then expected to comment on its substance and exposition, and usually recommend one of four possible courses of action: accept as-is; accept with minor revision; accept with major revision; reject.
Acceptance as-is is rare in my experience. In the course of my career I’ve had two articles accepted as written. Most people I’ve spoken to say the same. Outright rejection isn’t all that common either, but it does happen. I’ve had a few of those. I’ve also survived a few attempted muggings, in which one review said only “This is worthless,” and the second reviewer recommended publication with minor revision. Usually the editor will call for a tiebreaker in those cases.
I wasn’t going to write about this subject just yet, but then I saw this post that suggested that reviewers sign their work rather than remaining anonymous. I commented on that post, as did this blogger that defended the practice of remaining anonymous.
Publication in the peer reviewed literature is one of the most important tasks required of a working research scientist. The peer review process is far from perfect, but once an article appears in the peer reviewed literature, one expects that it has been examined carefully, and probably anonymously, by at least one, and probably two or three pairs of eyes that don’t belong to the author, or even to the author’s institution. There are other sources of scientific information that do not go through a rigorous review process. Such things as technical reports and conference proceedings are known collectively as the “gray literature.” To be sure, some very important stuff has appeared in the gray literature, but most folks see the gray literature as being less trustworthy. Committees that consider academic scientists for hiring or promotion place a great deal of importance on a candidate’s publication record, specifically the number of publications in the peer reviewed literature. A lot is at stake here, especially for early career scientists.
The reviewer’s job is not to act as gatekeeper, though many seem to view it that way. The reviewer’s job is to contribute to the process of making the submitted manuscript the best article possible. A few manuscripts arrive on the editor’s desk with enough polish to appear as is. Most can be improved. The trick is to write a review that will lead the author(s) to improve on their original manuscript. Sometimes even the most conscientious and constructive review can result in a paper that is worse than the original, especially in cases in which reviewers disagree.
It’s hard not to take a negative review personally, even if it recommends eventual publication. It’s fine to write reviews as if you were sitting across a table from the author, as the Grumpy Geophysicist suggests. That would actually be easier than submitting either a signed or an anonymous review, because you’d have a chance to explain that a particular comment or word choice did not reflect on the author as a scientist or a human being. If the author takes one of your written comments the wrong way, you won’t be there to clear things up.
When I write a review I try to write in the most straightforward fashion that I possibly can. I take care not to be witty or glib. I take pains not to come across as a smartass. I very rarely ask that the next version of a manuscript include a citation to my work.
I am prepared to stand behind every review that I write. I never write anything that I would not be willing to say to the author, the editor or the other reviewers. Still, I usually remain anonymous, at least on paper. It’s too easy to be misunderstood, however careful I might be, and the stakes are too high. I will not take a chance on making an enemy of someone who should be my friend. If I really like a manuscript, I may tell the author so at our next meeting.
Last thoughts on the matter:
Some journals are double-blinded, i.e., the author is not identified to the reviewer. It’s not a bad idea. I wrote a review for such a journal once. It wasn’t a journal I usually read, so I never found out who the author was. I live in a very small academic world. I suspect that most of the reviewers an editor would choose to review a manuscript of mine could identify me as the author without being told. I imagine that folks who think all reviews should be signed would not be in favor of making the process more anonymous rather than less.
When I was a grad student, one of my mentors told me that an article could be rejected for being wrong or not being new. Otherwise it should be published eventually, in some form. I’ve seen some dreadful papers that weren’t exactly wrong and presented new results. The reviewer’s job is to help make the manuscript the best paper it can be.
Once I reviewed a paper by a peer, a scientist at the same career level that I was at the time. I didn’t particularly like the paper. I thought it could have been much better, but I would have liked to see a whole new set of experiments performed, and I couldn’t ask for that. It was not my paper after all. I wrote the most constructive review I could, recommending publication with some fairly major, but entirely doable revisions. When it appeared, my colleague from the next office walked in with the journal folded to the article. Journals appeared only on paper at the time. “Can you believe this piece of crap that actually appeared in what we all thought was a good journal? Doesn’t anybody review these things?” I didn’t ‘fess up. Call me a coward.