Despite its preternatural clarity and force, my 'New Lilliputian Argument' against Machine Functionalism I has reaped some considerable confusion and misunderstanding. I am grateful to Professor Elugardo 2 for providing a correct restatement of the Argument and yet an opportunity for me to sharpen its moral and to make some further clarificatory comments.
As I have used the term, 'Machine Functionalism' is the view that a correct metaphysical explication of a particular mental state (-type) would take the form, "To be in mental state M is to realize or instantiate machine program P and to be in functional state S relative to P", where a physical organism or system O realizes or instant iates P just in case a one-one mapping holds between some set or other of O's (possible) physical stimuli, inner states and responses and, respectively, the abstract input symbols, computational states and output symbols tabulated in P (cf. the more formal definition Elugardo quotes on p. 000). The intent of the Argument (hereafter 'NLA') was to show that any such explication of a mental state-type would be far too liberal, since if we are thus allowed to select any subset of physical states of O we like, organism-program correlations of the sort just mentioned are trivially easy to come by. In particular, if we attend to what Elugardo calls "certain small and uninteresting parts of Oscar's body", viz, those parts which Oscar shares with Harold, we can see that such a correlation holds trivially and parasitically between Oscar and the program P approriate for thinking that index cards ought to be painted phosphorescently.
Elugardo's own handling of the 'Hoky Stipulation' objection (p. 000) seems to me exactly right (I have made the same reply to Ned Block in conversation), so I shall pass directly on to his criticism of (Def.), which I think is confused as it stands. He questions whether "(Def.) does in fact specify a one-one mapping" (p. 000) and doubts that "(Def.) does in fact show ... that Oscar realizes a Turing Machine program in the way that [I describe] a Turing Machine realization" (p.