1 The “general problem of pure reason,” along with its two more specific sub-problems, is formulated in § VI of the Introduction to the Critique of Pure Reason at B19–24. Sections V and VI, which culminate in the three questions “How is pure mathematics possible?”, “How is pure natural science possible?”, and “How is metaphysics as a science possible?”, are added to the second (1787) edition of the Critique and clearly follow the structure of the 1783 Prolegomena to Any Future Metaphysics, which was intended to clarify the first (1781) edition. This way of framing the general problem of pure reason also clearly reflects the increasing emphasis on the question of pure natural science found in the Metaphysical Foundations of Natural Science (1786). For an extended discussion of Kant's theory of pure natural science and its relation to Newtonian physics see Friedman, Kant and the Exact Sciences (Cambridge, MA: Harvard University Press, 1992), especially chapters 3 and 4.

2 From the first two paragraphs of § 6, entitled “Empiricism without the Dogmas,” of “Two Dogmas of Empiricism,” Philosophical Review 60 (1951): 20–43; reprinted in From a Logical Point of View (New York: Harper, 1953), pp. 42–43.

3 For extended discussion of Helmholtz and Poincaré see my “Helmholtz's Zeichentheorie and Schlick's Allgemeine Erkenntnislehre, ” Philosophical Topics 25 (1997): 19–50; “Geometry, Construction, and Intuition in Kant and His Successors,” in Gila Scher and Richard Tieszen (eds.), Between Logic and Intuition (Cambridge: Cambridge University Press, 2000); and Reconsidering Logical Positivism (Cambridge: Cambridge University Press, 1999), chapter 4.

4 Reichenbach, Relativitätstheorie und Erkenntnis Apriori (Berlin: Springer, 1920); translated as The Theory of Relativity and a Priori Knowledge (Los Angeles: University of California Press, 1965). The distinction between the two meanings of the Kantian a priori described in the next sentence occurs in chapter 5.

5 Carnap, Logische Syntax der Sprache (Wien: Springer, 1934); translated as The Logical Syntax of Language (London: Kegan Paul, 1937).

6 This distinction is first made explicitly in Carnap, “Empiricism, Semantics, and Ontology,” Revue Internationale de Philosophie 11 (1950): 20–40; reprinted in Meaning and Necessity, 2nd ed. (Chicago: University of Chicago Press, 1956).

7 Carnap explicitly embraces this much of epistemological holism (based on the ideas of Poincaré and Pierre Duhem) in § 82 of Logical Syntax. Quine is therefore extremely misleading when he (in the above-cited passage from § 6 of “Two Dogmas”) simply equates analyticity with unrevisability. He is similarly misleading in § 5 (p. 41) when he asserts that the “dogma of reductionism” (i.e., the denial of Duhemian holism) is “at root identical” with the dogma of analyticity.

8 For an analysis of the principle of equivalence along these lines, including illuminating comparisons with Reichenbach's conception of the need for “coordinating definitions” in physical geometry, see Robert DiSalle, “Spacetime Theory as Physical Geometry,” Erkenntnis 42 (1995): 317–337.

9 Kuhn develops this example in The Structure of Scientific Revolutions, 2nd ed. (Chicago: University of Chicago Press, 1970), chapter 9. There is some irony in the circumstance that Kuhn introduces this example as part of a criticism of what he calls “early logical positivism” (p. 98).

10 Kuhn, “Afterwords,” in Paul Horwich (ed.), World Changes (Cambridge, MA: MIT Press, 1993), pp. 331–332.

11 Kuhn, “Afterwords” (note 10 above), pp. 338–339.

12 See Habermas, Theorie des Kommunikativen Handelns (Frankfurt: Suhrkamp, 1981), vol. 1, chapter 1; translated as The Theory of Communicative Action (Boston: Beacon, 1984).

13 In Kuhn's own discussion of the theory of relativity (see note 9), he explicitly denies that classical mechanics can be logically derived from relativistic mechanics in the limit of small velocities. His main ground for this denial is that “the physical referents” of the terms of the two theories are different (op. cit., pp. 101–2). Here, however, I am merely pointing to a purely mathematical fact about the corresponding mathematical structures.

14 That the convergence in question yields only a purely retrospective reinterpretation of the original theory is a second (and related) point Kuhn makes in the discussion cited in note 13 above, where he points out (p. 101) that the laws derived as special cases in the limit within relativity theory “are not [Newton's] unless those laws are reinterpreted in a way that would have been impossible until after Einstein's work.” I believe that Kuhn is correct in this and, in fact, that it captures a centrally important aspect of what he has called the non-intertranslatability or “incommensurability” of pre-revolutionary and post-revolutionary theories.

15 For a detailed discussion of this case see my “Geometry as a Branch of Physics,” in D. Malament (ed.), Reading Natural Philosophy (Chicago: Open Court, 2002).

16 See again the reference cited in note 15 above.