Abstract
In the paper I examine (Section I) the best defense for the claim that methodologies shouldnot function heuristically (thesis-LW) as it appears in John Worrall. I then evaluate (Section II) his proposal of a criterion* M which is offered as a criterion for evaluating competing methodologies such as falsificationism, conventionalism, methodology of research programmes. etc. Finally, I consider (Section III) the consequences of arguments presented earlier (Section I and II) as they bear on the problem of selecting a historiographical model.I argue, among other things, (I) that thesis-LW is defended on some very dubious assumptions; (II) that Worrall's criterion* M falters under three clear cases two of which at least*M should accomodate, and that part of* M's failure can be linked to its being hooked to thesis-LW. By arguments analogous to the ones which serve* M, I show thatcontra John Worrall and John Watkins, thesis-LW is testable; finally, (III) if we accept arguments for thesis-LW and* M we are left with a skeptical conclusion with respect to the choice of a historiographical model which Worrall by parity of reasoning should accept, but does not.