ABSTRACT

If all our actions were determined solely by condition-action links we would not be able to choose which action we wanted to make to a particular stimulus, because the strongest condition-action link would always capture the control of behaviour. We would behave like the conditioned rats of the behaviourist tradition. Clearly there is more to it. In everyday life we are continually making a series of actions to objects which are inviting, or afford a variety of appropriate responses. Usually, we make these actions, in a goaldirected sequence. For example, when making a cup of tea the sugar bowl, the milk jug and the cup are all containers into which we can pour things. In one part of the teamaking sequence we have to be sure to pour tea into the cup and milk into the jug, not vice versa. Later in the sequence it is appropriate to pour milk into the cup. When distracted, we may make a mistake such as pouring tea into the milk jug. Such “slips of action” have been studied and interpreted as failures of control (Reason, 1979; Norman, 1981). While it is well appreciated that complex behaviour requires some kind of control process to coordinate and organise it, there is to date no clear idea of exactly how this is achieved. However, if we ask an experimental subject to do one task rather than another, respond to one aspect of a stimulus and ignore all others, the subject is able to do it. Somehow the cognitive system can be configured to do one task at one time and another task at another time on the basis of intentions. Thus a major question psychologists have to address is: how is behaviour controlled by internal intentional states (endogenously) rather than by external perceptual states (exogenously)? Until recently little experimentation had been done on the internal control of tasks but this work is beginning and we shall examine some of it later in this chapter.