In Response to: Wayne Grennan’s An argument evaluation procedure incorporating arguer credibility

Grennan's discussion opens with an analysis of how the credibility of the arguer can be reliably factored in to the evaluation process. He demonstrates quite clearly how important judgements of credibility are in real world arguments, and describes an approach which manages at once to handle credibility cleanly, to respect logical autonomy, and to be sufficiently straightforward to give to one's students without compunction. Grennan suggests that in an argument with two or more premises, the approach could founder. In an argument demonstrating linked structure, we would expect the argument to fall down completely if one premise were unreliable. But he identifies a problem with convergent argument structure too. Specifically, if one of the premises adduced in support of a conclusion is judged to be absolutely unreliable, then, by laws of probabilistic combination, support provided by all premises is absolutely unreliable. This result, he suggests, is "too paradoxical to be acceptable". It is worth noting that perhaps in some situations this is exactly the result, though for rhetorical rather than argumentation theoretic reasons. If I demonstrate my faults by the fatuousness, inanity or, perhaps, insanity, of one of my premises, then my audience may well as a result utterly discount all the other premises that I may bring to bear