Abstract

Available from UMI in association with The British Library. Requires signed TDF. ;An attempt is made to find the proper response to paradoxes of the sorites-type. In PART I, a chain argument is given with 'tall' in which it seems that a false conclusion is derived by repeated steps of MPP from true premises. In particular, each of the conditional premises is derived by UE from a so-called inductive premise sustained by familiar claims that 'tall' is both vague and observational. A variety of tokens is then given of what seems to be the same paradox-type; some particularly interesting examines are yielded by a common form of essentialism. In PART II, it is argued that the original paradox with 'tall' is not plausibly solved by taking borderline sentences as either true or false. The 'gap' account on which borderline sentences are neither true nor false provides no leverage in itself, nor when combined with the distinctively anti-realist theses introduced in PART III. In PART IV, the force of the paradoxes is also shown to survive the proposal that borderline sentences should be assigned one of many truth-values between true and false. The problem of interpretation is compounded by the same difficulties faced by some of the more sedate accounts considered earlier. This leaves the conclusion in PART V that our original appraisal of the chain argument with 'tall' is mistaken only in the perverse sense that it is a mistake to suppose that some mistake must have been made. From this it follows that there is a sense in which our language is inconsistent, but much of the hostility aroused by this claim is shown to dissipate once its content is explained in terms of the rule-conflict thesis