Self-control is a difficult concept for behaviourists. The implication that we are exercising control over ourselves, rather than being controlled by our environment, appears to run counter to strict behaviourist doctrine. The spectre of mind-body dualism with homunculi sitting inside our head looms large on the horizon. The concept of the self as initiator and controller of our behaviour appears to break all the rules in the behaviourist book. Self-control, in this sense, is not only ‘mentalistic’ and ‘inferential’, it is also in danger of becoming ‘explanatory fiction’. Like the man who gives up smoking because of willpower, we can only infer his willpower from the behaviour it is designed to explain. Skinner (1953) has pointed out the dangers of using inferential concepts such as willpower and self-control. He and many other ‘radical behaviourists’ have strongly argued that there is no place for these concepts in scientific psychology.