What have philosophers been talking about for the last twenty years? That's the question Kieran Healey attempts to answer in his fascinating analysis of co-citation networks, which Eric reflects on here. I am not quite ready to reflect on Healey's results myself. I have been playing around with a prior question that might interest readers. And I don't feel I have gotten to the bottom of this question yet.
What is a co-citation network? A co-citation edge is a pair of articles that are cited together. Suppose that I write an article that cites both A and B. An A-B edge is thereby formed. Now suppose that B is never cited except together with A—perhaps B is a discussion of A—but A is sometimes cited together with B and sometimes together with C. (A could be a discussion of C; alternatively, A and C could be discussions of the same topic.) The consequence will be two edge sets emanating from A, but only one from B. Thus, B will be less of a hub, or more peripheral, than A.
The idea is illustrated by the most isolated network of all: a two node network connecting Mark Crimmins' 1998 paper, "Hesperus and Phosphorus: Sense, Pretense, and Reference," and Kendall Walton's Mimesis as Make-Believe. Why these two?
Well: Crimmins discusses the following utterance: "Ann is as clever as Holmes and more modest than Watson." When putting forward his central idea, Crimmins writes: "In elaborating this, I will borrow heavily from Kendall Walton's work on make-believe." There you go: you can't cite Crimmins without citing Walton. Neither Walton's work nor Crimmins's is co-cited often enough with any other work for there to be other nodes in this network. (Only four general journals are included in Healey's analysis: I am sure that Walton's book is co-cited with many other works in aesthetics, and is thus an important hub, just not in these journals.)
Crimmins' article is well-cited, but is not a champion (90 cites altogether in GScholar, which is obviously not restricted to the four Healey journals); Walton's work is cited 1372 times. But given that anybody who cites Crimmins has to cite Walton, this connection shows up as an edge.
Now that you have the idea, does it come as a surprise that very few isolated networks that show up in Healey's analysis? In non-value theory fields, the biggest hubs are: Kripke (NN), Lewis (Plurality), Quine (WO), David Chalmers (Conscious Mind), and Tim Williamson (given as TW 2000, but that must be a mistake; KL is 2002). And the overall champion is David Lewis. In other words, these are the works that have to cited in the context of a large number of other works, just as Walton had to be cited in the context of Crimmins.
I still can't figure out what this means. Why is Plurality more of a hub than Naming and Necessity, for instance? Why does a rather strange fragment from Aristotle (apparently Mtp 981a24, about wisdom) occur at all, and then linked to John Locke?
Recent Comments