People do not say that a barometer "knows'' when it is going to rain; but I doubt if there is any essential difference in this respect between the barometer and the meteorologist who observes it.--Bertrand Russell (1923) "Vagueness"
"[Gender] identity is performatively constituted by the very "expressions" that are said to be its results"--Judith Butler Gender Trouble. (33)
In my spare time I am reading Tim Maudlin's (2007) The Metaphysics Within Physics (recall this) and Judith Butler's (1999 --I have used the 1999 edition which has an additional, fascinating preface]) Gender Trouble: Feminism and the Subversion of Identity in seperate, ongoing reading groups. This has led me to notice some surprising similarities. For example, they are both unabashedly concerned with metaphysics, and while they differ in lots of ways, they have a shared target: the idea that metaphysics ought to be articulated in terms of substance(s) with properties (or attributes/accidents); they both dislike universals. Both attribute the metaphysics of substance to an illegitimate "projection of the structure of language onto the world." (Maudlin, 79; compare that with Butler pp. 25-28 [where she summarizes others] & p. 33 [where she speaks in her own voice]). Maudlin aligns his view with Bertrand Russell's position in "Vagueness" (79-80), while Butler traces (33) her criticism back to Nietzsche's famous passage in which he explains how in our metaphysics we are "a dupe of the tricks of language." Russell, who knew his Nietzsche, may well have gotten this line of argument from him (I'll leave that to the scholars). The idea that our inherited metaphysical categories may be a projection arguably goes back to Spinoza and his famous argument against final causes in the Appendix to Ethics 1. (Perhaps, it is better to say that Protagoras is the grandfather of this whole line of argument?)
Maudlin does better than Russell. For, in addition to some name-calling ("going Spinozistic is really a Pyrrhic victory for a metaphysics of substance and universals"), Maudlin's book (which does not list "Spinoza" in its index) does offer an argument against the "Spinozistic" position:
[I]f there really is only one substance and every physical fact corresponds to a distinct fact of it, then we are getting no structural insight at all into the nature of the world by being informed of which substances exist. One retains the vocabulary of substance and universal, but the cash value is nearly nil. (102; emphasis in original)
Now, it is not entirely clear that Maudlin has a right to this argument. For the whole point of his chapter (which is directed against David Lewis’ metaphysics) is that there are no short-cuts from language or intuition to metaphysical truth. As he puts it while diagnosing his (appropriate) dissatisfaction with Quine, “the lion’s share of real metaphysical work is done when settling on the right translation [into a formal language],” which just “is part and parcel of deciding on the most acceptable theory about the deep structure of certain properties and relations.” (83) That is to say, the cash-value of one’s privileged vocabulary ought to be nil. And this is precisely what Spinoza’s system is designed to do.
One fails to see this if one mistakenly thinks that Spinoza’s metaphysics is supposed to be a guide to physics! (Or so I claim.) (Of course, Maudlin is not worried about the historical Spinoza.) In his defense of (general) final causes, Clarke exploited this aspect of Spinoza's approach brilliantly and used it to shift the burden of proof. He pointed out that Spinozistic "blind metaphysical necessity" might have sufficient reason for everything that exists (by way of E1p16), but could not explain for the surprising features of observed, particular variety. A version of this argument found its way into Newton's General Scholium. The authority of physics successfully promoted final causes for a long time, after all. (Of course, burden shifting arguments do not refute its targets.)
Now, as regular readers may realize, I am very sympathetic to Russell's claim that there is no "essential difference in this respect between the barometer and the meteorologist who observes it." This just is a fundamental commitment of a proper general philosophy of science (GPOS) that can do justice to the practices of the sciences and the epistemologies they give rise to. But we should not mistake this claim for a general metaphysical outlook. In the sciences, 'the' identities of (say) 'the' barometer and 'the' meteorologist get stabilized by (among other things) protocol and calibration in the service of theory-mediated measurement. In fact, I do not wish to deny the significance of mathematics in such stabilizing: part of the reason for the privileged status of gauge theory in physical science is that it facilitates renormalization (and this certainly makes the results of measurement more easily digestable). 
 Thank you to Nick Huggett, Eric Winsberg, and Jody Azzouni for commenting on a draft version of the sentence to which this note is appended. None of them should be held accountable for the claim (one of them thinks it is “not not clear that's true.”)