An animal can be said to teach if, at some cost or at least without ob
taining an immediate benefit for itself, it modifies its behaviour in
the presence of a naive observer, such that the observer acquires a sk
ill more efficiently than it might otherwise do. An important theoreti
cal issue is the extent to which the demonstrator's behaviour is sensi
tive to the improving skill level of the observer. This study determin
ed whether domestic hens, Gallus gallus domesticus, were sensitive to
feeding errors by their chicks. The feeding displays of 12 hens were s
tudied in two contexts: (1) in the presence of chicks that fed on a fo
od colour that was palatable to the hen, and hence made no apparent fe
eding errors and (2) in the presence of chicks that fed on a food colo
ur that was unpalatable to the hen, and hence made apparent feeding er
rors. Both groups of chicks, in reality, fed on palatable food. Hens r
esponded more intensely to chicks that made apparent feeding errors. H
ens increased the rate of ground pecking and scratching when they had
no food available, and increased the rate of food pecking (although no
t food ingestion) when they were given food. It is suggested that hens
are sensitive to perceived chick error, and that the costs of increas
ed display intensity are balanced by the benefits of attracting chicks
to more profitable food items. (C) 1996 The Association for the Study
of Animal Behaviour