Logical expertise as a cause of error: A reply to Boolos

He has overlooked two points. First, in these premises, as in all the problems we gave our subjects, each assertion contains a definite article, e.g. “All the boxers are clerks”. Such definite descriptions establish or presuppose the existence of members of the corresponding sets. Second, we explicitly instructed our subjects that each problem concerned three sets of individuals gathered together in a room. We therefore doubly ensured that there was no doubt about the existence of individuals in all three sets. It follows that Boolos’s analysis of “All the boxers are clerks” is wrong since it is consistent with the non-existence of boxers. It also follows that the inference above is, as we claimed, entirely valid. Boolos is neither the first nor, we imagine, the last to err in this way, e.g. Kyburg (1983, p. 266) anticipates him. Indeed, logicians from California to East Anglia have made the same mistake so often that the phenomenon has become worthy of psychological investigation in its own right. We believe that the following conditions are necessary to give rise to it: