A method for the objective study of tool-using behavior.

Key pecking for food was shaped in four crows within a conventional operant-conditioning test chamber. When pecking stabilized, a metal screen with openings 2.5 cm high by 1.0 cm wide, was placed over the response key, so that the crow could still see but could no longer peck the key. At the same time, several dozen wooden matchsticks, which could be used to operate the key, were placed in the test chamber. The crows made no use of these during 50 to 75 hr of exposure to this condition. Subsequently, the behavior of two crows was shaped so that they approached the matchsticks, picked one up in their beaks, approached the response key with the matchstick in their beak, and finally operated the response key by poking the matchstick through the screen. This shaping procedure was ineffective with the two other crows. However, these birds were successfully trained through positional fading of the tool. This involved suspending a metal rod from the ceiling so that it hung directly in front of the response key, and the crow had only to peck it to operate the key. Then, the rod was gradually lowered by lengthening its tether until it eventually rested on the floor of the test chamber. The principal advantage of this methodology is the automatic recording of the terminal (tool-using) behavior under study.