¹ It is my sad task to respond to Michael Kac on my own: Jerrold J. Katz died on February 7, 2002, at the age of 69. Jerry and I did not always agree; and he would often surprise me by saying something deeper and more interesting than I would have in defense of views we shared. I can only hope he would have been satisfied with what I say here.

² Not all analytic sentences are true. Consider, for example, the present king of France is male, which is analytic if kings are male is, but which is not true (on either a Fregean or Russellian approach to definite descriptions).

³ Just as, for example, the apparent conflict between the intuitions that both dusting removes dust and dusting applies dust are analytic though remove and apply are antonymous is resolved on the supposition that dust is ambiguous.

⁴ Hence, on our account, plastic flowers are flowers is analytic on one reading of plastic flower ('flower made of plastic') and contradictory on the other ('imitation flower made of plastic'), while plastic flowers are not flowers is contradictory on the former reading and analytic on the latter. (Compare: dusting removes dust is analytic on the 'remove dust from' reading of dust and contradictory on the ‘apply dust to’ reading, while dusting applies dust is contradictory on the former reading and analytic on the latter.)

⁵ It is not obvious in the case of plastic flower because there are no plastic flowers in the conjunctive sense in the actual world.

⁶ As does Kac's application of the apparatus of generalized extensions to temporal adjectives (p. 194) (which, by the way, is essentially the same account Katz and I offer in n. 4 of our article).

⁷ Our other examples are explained in similar ways.

⁸ Or, perhaps, one might hold that expressions have both ungeneralized and generalized extensions in the language, and it is typically utterance context, rather than linguistic context, that disambiguates. As far as I can tell, Kac does not endorse this position.

⁹ On the ungeneralized interpretation of flower, the sentence asserts that flowers that are made of plastic are not flowers; on the generalized interpretation of flower, the sentence asserts that things that look like flowers and are made of plastic are not things that look like flowers.

¹⁰ PE stands for ‘perception is everything’, which Kac glosses as ‘one is entitled, except in certain specially defined circumstances, to take appearances at face value’ (p. 191). He also refers to it as the ‘Duck Principle’ (n. 4)—presumably an allusion to the saying ‘if it looks like a duck and sounds like a duck and ... then it is a duck’. He asserts (n. 7), however, that he does not mean that the principle licenses taking the referents of flower in these circumstances to be flowers. Rather, his claim is ‘only that [PE] sanctions the occasional use of flower as if it had artificial flowers in its extension’. Yet, in n. 5, he claims that ‘PE justifies taking non-flowers that look like flowers to be flowers’ (Kac's emphasis). He can't have it both ways of course.