Many Internet commenters have criticized MIRI for not producing enough research, relative to their budget – not writing enough papers, or not getting those papers peer-reviewed, or not getting enough citations. However, MIRI’s specialty is math and computer science, which might have lower citation counts than experiment-heavy fields like chemistry or biology. For a quick sanity check, I looked up a few non-MIRI mathematicians as points of comparison.
Grigory Margulis is probably the most accomplished mathematician I’ve personally met. He’s a Fields Medalist, won the Wolf Prize in 2005, and is doing pure math as a Yale professor full-time, so it seems reasonable to assume he’s in the upper quantiles of productivity. A Google Scholar search for the last five years turned up eight papers that he’s co-authored; by my count, those eight papers (combined) have 35 citations, of which nine are self-citations. All of those papers had multiple authors, so it took well over five person-years of total effort to produce them.
But of course, a Fields Medalist isn’t a representative math researcher. One friend of mine recently got a math Ph.D. from an elite university; as a grad student, they spent years doing math research full-time, and they also did a lot of part-time research in undergrad. They wrote several papers while in grad school, plus (of course) a dissertation, but these don’t appear to be on Google Scholar; presumably they’re still awaiting publication, or they weren’t published in a place Google indexes. They also published two papers before grad school, of which only one was peer-reviewed; these two papers have 13 total citations, of which five are self-citations.
Another friend of mine got a math Ph.D. some years ago, from a less elite university. They wrote four papers which appear in a Google Scholar search. Of those four, one wasn’t a math paper, and was published long after they graduated; one was their dissertation; one was posted on arXiv, but doesn’t seem to have been formally published; and one was published as a conference paper. Excluding the non-math paper, the remaining three papers have eight total citations.
Another friend of mine just got a math Ph.D. from an elite school, and is taking an academic job after graduating. They’ve written a number of papers, given talks, etc.; but again, a lot of these don’t appear on Google Scholar. Three of their papers are on Google Scholar, but all three appear to be arXiv papers that haven’t been formally published, and the three papers have three total citations.
But all that might just be selection bias in who I know. Using random.org to pick two math postdocs – one from Stanford, one from Berkeley – their CVs list a combined total of eighteen papers, of which seven have been formally published, four are listed as “accepted” but not yet published, and the remaining seven are on arXiv or self-hosted. Of these 18 papers, the most cited one had a total of 13 citations, of which four were self-citations.
(Disclaimer: I’m not a math academic; comments/corrections from people who are appreciated.)
The first couple papers on Google Scholar show citation 100+ citations for several Margulis papers. This is from simply entering ‘grigory margulis’ in the Google Scholar search. Alternate spellings are Gregori and Gregory. Not sure where you got your number from — this took a few seconds to search.
Also, in the math field citations don’t happen as fast as in CS. You’d probably weight importance by several other factors, like proving open conjectures, novelty within the field, etc. Basically a bunch of subjective things that relied on other people who are expert in the field judging the results. In CS, on the other hand, citations seem to come much faster because of the sheer number of papers being published.
Per the post, this was only for papers published in the last five years (2011 through 2015).