Two experiments are reported that challenge the generalisability of th
e specificity hypothesis. The first experiment in which subjects (N =
24) were required to make one-handed catches under the condition that
they could only see the ball, in an otherwise completely dark room, pr
oduced results in line with the hypothesis. However, when the same sub
jects, after a 3-4 day training period, under the latter condition, we
re transferred to a condition in which they could see both hand and ba
ll they demonstrated, contrary to what would be predicted by the hypot
hesis, significant increases in catching performance. In a second expe
riment subjects (N = 32) were required to catch balls under a variety
of conditions in which viewing the hand only, or viewing the hand and
ball, were interspersed in a number of ways. The results suggested tha
t subjects who have experience of performing in the presence of both k
inds of information (ball only or ball and hand) adopt the particular
perception/action coupling appropriate to the situation in which they
find themselves and, to this extent, treat the two conditions as diffe
rent skills. This lead to a reinterpretation of the specificity hypoth
esis which takes account of the flexibility in performance under chang
ing environmental conditions characteristic of the skilled performer.