Catania (1973), in discussing the concept of the operant suggested responses should not be classified based on arbitrary criteria such as response function or topography alone, and rather should focus on the continuous distribution of responding. Catania (1973) hypothesised that response distributions would resemble a normal curve, even when reinforcement was provided for only those responses that occurred between two limits. Six domestic hens were used in a series of experiments in order to investigate Catania’s (1973) suggestion. In the first experiment the hens pecked at a flat screen with an infra-red device was mounted in front the screen to detect response location. Pecks in defined regions of the screen resulted in 3-s access to wheat. Over conditions, different quadrants of the screen were active. The whole screen remained white throughout each session except when the reinforcement was being delivered. The hens responded on a fixed-ration 5 schedule for 40 reinforcers per session. Over sessions the distribution of the locations of responses shifted to being centred over the active quadrant. Proportions of correct responses increased across sessions but incorrect responses persisted throughout. In Experiment 2 the screen was divided into ten 30-mm regions of which one was active in each condition while the whole screen remained white. The active region was moved in steps from the left to the right and from the right to the left of the screen over eleven conditions with eight sessions in each conditions and a fixed-ratio 5 schedule. In general, in each condition the proportion of correct responses increased to around 0.4. In each condition the distribution of response location shifted to being centred over the active region, resembling a normal distribution. In most cases, response rates on the active region were higher than for the inactive regions, hens usually responded on several regions at a moderate rate. In Condition 12 the active region was coloured red and this resulted in a narrower distribution of response locations, and higher proportions of correct response that other conditions for most hens. A large number of incorrect response persisted throughout this condition. These findings provide some evidence that supports Catania’s (1973) hypothetical response distributions. The argument is made that Catania’s (1973) concept of the operant could be a way to organise responses into a class and also that is provides a means of classifying response while maintaining the continuity of behaviour and, therefore, avoids the use of arbitrary criterion.