Abstract
Four pigeons were exposed to a discrete-trial schedule in which only responses spaced by at least 6 sec were reinforced. After 45, fifty-trial sessions, they failed to meet the spacing requirement in over 90% of the trials. When an alternative, non-contingent key (pecks on which had no consequence) was illuminated concurrently with the first key, the spacing performance of the three pigeons that pecked the non-contingent key improved so that they were obtaining 75% of the possible reinforcers. These data demonstrated the importance of collateral behavior in mediating spaced performance. It was suggested that pigeons may successfully refrain from responding on the spacing procedure only when another stimulus correlated with reinforcement is available for pecking, and that the form that collateral behavior takes may, in general, be non-arbitrary, and species dependent.