But I bet it’s more than I&D would have if it weren’t for Vagueness, Knowledge and Its Limits &c. It’s noticeable how recent many of the citations of I&D are.

And if you need n publications by a certain time, then you need to send out some multiple of n submissions; and it’s not clear how much you can lower the multiple by improving the paper as much as possible. (Very rarely can you lower it to 1.) So I think it really is a tragedy of the commons.

I completely disagree that the reason for exposing oneself to our peers is merely prudential, as you said in the last post. Knowledge is a social product. And submitting to journals is part of the process of hypothesis testing and correction. Getting papers to referees, and then to the community, is indispensable for refining ideas. So here’s a caveat on the idea that you should wait until your papers are really good before submitting. Say (to take a completely hypothetical example) you work on free will in Australia, where few people work on the problem. Suppose when you give papers on free will, the only response is “but we sorted all this out in the 60s”. Then your relevant peers are in the US, and you may find that you can present to them, at most, every 18 months or so. Then the peer review system becomes more important to you. People only have so much tolerance for blog posts, in terms of length (of course this doesn’t excuse trigger happy submitters who don’t wrestle with the tyranny of distance).

Having said that, I don’t think that “Epistemic Modals in Context” is my best work, so I agree that citations aren’t the same as quality. But arguably it is the largest contribution to the literature that I’ve made. Writing a great paper that’s never read isn’t making the world a better place after all.

Also, I have some doubts about the reasons you mention why one should work really hard on making one or two papers really good. Mainly because there’s a very good chance that these will end up amoung the uncited ones. Which papers get widely cited is not that well correlated with quality in my experience. Things just snowball (which explains the order of magnitude gap between the most cited papers and the rest) – once a few people get talking about a paper, more people notice it, and so on. And sometimes the reason for a high citation count is simply that there are lots of easily correctable errors. (One reason, incidentally, why I think citation counts are such a dubious marker of research quality.) One reason people like Lewis and Williamson have lots of multiply-cited papers is that once you get a reputation people start reading everything you publish, and will also go back and read stuff that probably – like most work by most people – sank without trace at the time of its original publication.

That’s not to say one shouldn’t work really hard on one or two papers, if one prefers to work that way. But if one’s motivated by career concerns and professional respect it might well be a bad way to go. It might be better to get as many papers out there as possible, in the hope that some of them will catch the eye of the philosophical world. Then, if you end up with a Lewis/Williamson reputation, hopefully people will go back and read and cite your other papers too.