On Rejection (Part 1: The Descent)

Two months ago you submitted a paper to TopCon, the top conference in your branch of computer science. This paper is some of the best work you’ve ever done. As soon as it was submitted, you posted it on your website, so that your peers — who undoubtedly check your website every five minutes to find out what you have been up to — can download and print out your latest masterwork and read it on the beach during their vacation.

Periodically, over the next few months, when you are in the middle of doing nothing, you visit your own website and download the masterwork. You want to get a sense of what the experience is like to read it for the first time. You glance longingly at the title, you smell the abstract. That introduction which is so gentle, so economical in its summary dismissal of prior work, so clearly structured to guide the reader to the inevitable conclusion that what you are doing is important and groundbreaking — it’s guaranteed to make any reader swoon, even if you did write it at the very last minute before the submission deadline. Admittedly, the bulk of the paper is taken up with a technical section that gets a bit hairy, with a couple pages in there that even you can’t quite follow — you let your student write them — but this is not kids’ stuff we’re talking about here, this is cutting-edge research, so hey, at some point you’ve got to lose the reader. And anyway, that related work section at the end of the paper, also written at the last minute and unintentionally omitting a few key citations, brings everything into perfect perspective. No doubt.

At some point, you realize you have been downloading your own paper and admiring it a bit too often. Even by your own obsessive standards, you are acting a bit obsessed. You start to wonder if this is maybe a subtle sign that you are secretly worried about one niggling detail: the paper has not yet been accepted.

No, no, you’re not really worried about this — your paper is a clear accept — but there’s that fly in the ointment: the paper must be reviewed by a jury of your peers. And have you checked out your peers lately? They write terrible papers. When you review their papers, you think that the vast majority of them don’t seem to know what they’re talking about, or maybe they know what they’re talking about, but you certainly don’t know what they’re talking about. You look at the program committee of TopCon this year, and your worst fears are confirmed. The only person on the PC who properly understands your work is a frequent collaborator of yours — conflicted out of reviewing your paper — and the rest have probably not read your last 5 papers, a detailed understanding of which is necessary in order to truly appreciate the brilliance of your latest masterwork.

But wait a sec! Aren’t these the same peers you were hoping would download and read your paper on their vacation? You respect these people, in theory at least. Calm down. Your introduction is so clear, your abstract so fragrant, your related work so related, and of course your technical section so damn impressive…that there is nothing to worry about. The brilliance of the paper will shine through, even to the most Neanderthal reviewer.

Then you receive the reviews.

Reviewer A:

Wow, this is a BIG paper. It is tackling a hard problem, and throwing every trick in the book at it. The pieces all seem to fit together, but it is not clear exactly how or what the reader is supposed to get out of it, or how this work is to be differentiated from the many other recent papers in this line of work. Offhand, the approach taken here seems awfully complex, suggesting that the authors have not yet hit upon the right solution to the problem.

Reviewer B:

The paper presents an interesting thesis.  However, it seems to me the problem it is attacking is already largely solved by <insert reviewer’s name here> in their <noname conference> <last year> paper.

Reviewer C:

I think this is a solid paper, but I am not very knowledgeable in the topic.  The results seem good but I cannot verify correctness. The appendix submitted with the paper is very large, which suggests that maybe the paper would be better suited for journal publication.

Reviewer C was your only (weak) “accept”, and they gave themselves low confidence. Your masterwork has been rejected from TopCon — there will be blood.

Next in this series: On Rejection (Part 2: Valley of the Voodoo Dolls)