we can just demand that he tell us what it is to be in state A, and once told, his case fails, for state A then has no unexhausted content. This can be seen at its clearest when Putnam considers the possibility that, since (i) is said to be a 'synthetic proposition' for the machine, ' ... . the machine might be in state A and its sense organs might report that flip flop 36 was not on. In such a case the machine would have to make a methodological "choice"-namely, to give up (i) or to conclude that it had made an "observational error" . .'1 But the error is Putnam's: talk of flip flop 36 shows we are talking of a real TM, and so the 'deciding it is in state A' has no sense here apart from whatever procedure or definition has already been chosen to exhaust the notion 'state A', and no such problem can arise. If such an assignment has not been made, the machine is still abstract and talk of flip flop 36 is out of place. Putnam's claim2 that 'I am in state A' and 'flip flop 36 is on' are 'clearly non-synonymous in the machine's language by any test.., .the machine has to use different "methods of verification" . . ..' is simply and straightforwardly false for any real TM. 4 PUTNAM'S COUNTER ATTACK Putnam anticipates to some extent this attack on his use of the abstract TM. He correctly notes 1 Putnam [1964], p. 139. 2 Putnam [1964] p. I153. This content downloaded from 157.55.39.255 on Mon, 01 Aug 2016 05:43:10 UTC All use subject to http://about.jstor.org/terms