Sixty-eight subjects were asked to solve a series of six Wason selecti
on problems by physically turing relevant cards over until the could s
ay whether the rule was correct or false. In the first experiment most
of the subjects misunderstood the rule they were asked to verify, dif
ferent subjects misunderstanding in different ways, but commonly readi
ng (a) ''top/underneath'' for ''one side/other side'' and (b) a dual c
onditional for ''if..., then....'' However, a majority of subjects the
re-after responded in complete logical consistency with the rule as th
ey understood ii. to be. When the misunderstandings revealed in the fi
rst experiment were themselves presented as rules to be verified in a
second experiment, performance was much more logical than is usually r
eported. IL is concluded that the illogicality commonly associated wit
h Wason's selection task is only apparent.