THE KNOWLEDGE ARGUMENT AND HIGHER-ORDER PROPERTIES Amir Horowitz and Hilla Jacobson-Horowitz Abstract The paper argues that Jackson’s knowledge argument fails to undermine physicalist ontology. First, it is argued that, as this argument stands, it begs the question. Second, it is suggested that, by supplementing the argument (and taking one of its premises for granted), this flaw can be remedied insofar as the argument is taken to be an argument against type-physicalism; however, this flaw cannot be remedied insofar as the argument is taken to be an argument against token-physicalism. The argument cannot be sup- plemented so as to show that experiences have properties which are illegitimate from a physicalist perspective. Frank Jackson’s celebrated (though controversial) ‘knowledge argument’ is meant to undermine all versions of physicalism, ver- sions of type-physicalism as well as versions of token-physicalism. Commentators who attempt to refute this argument standardly do not attach any significance to the distinction between type- physicalism and token-physicalism as far as the argument is con- cerned. We believe, however, that this distinction is essential for evaluating the knowledge argument. We wish to argue that the knowledge argument fails to undermine physicalist ontology. The way we argue for this claim is as follows: first, we argue that, as this argument stands, it begs the question. Second, we suggest that, by supplementing the argument (and taking one of its prem- ises for granted), this flaw can be remedied insofar as the argu- ment is taken to be an argument against type-physicalism; however, this flaw cannot be remedied insofar as the argument is taken to be an argument against token-physicalism. 1 The © Blackwell Publishing Ltd. 2005, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. Ratio (new series) XVIII 1 March 2005 0034–0006 1 As will be shown, our claim that the knowledge argument can be supplemented so as to refute type-physicalism is conditional: its truth depends upon a certain assumption which we do not attempt to defend, namely the assumption that there is a property of experience that cannot be known from the third-person perspective. Our main (and unconditional) purpose is to show that the argument does not (and cannot be supple- mented so as to) undermine token-physicalism, and thus physicalist ontology.