Skinner (e.g., Holland & Skinner, 1961; Skinner, 1974) offers us another variety of behaviourist theories of emotions. Skinner discusses emotions within an operant conditioning framework. Within this model emotions serve to put the organism into states in which different sets of event contingencies define the reinforcers:
Under different emotional conditions, different events serve as reinforcers, and different groups of operants increase in probability of emission. By these predispositions we can (do) define a specific emotion. (Holland & Skinner, 1961, p. 213)
So in emotion state 1, set of events A will be reinforcing for set of operants A, whereas in emotion state 2, set of events B will be reinforcing for set of operants B. The emotion is defined by the sets of operants and reinforcers that are optimised. In addition, when set of events B is reinforcing for set of operants B, set of events A and set of operants A will be incompatible. For example: "Even when deprived of food, an anxious person may not eat. The responses which increase in probability during anxiety are incompatible with eating" (Holland & Skinner, 1961, p. 216).
There is a dangerous circularity lurking in here somewhere as Lyons (1980) has pointed out. If we, once more, consider Susan's lucky escape from the bear then, according to Skinner, Susan is defined as being afraid because she is running away and because the running away is an escape from the bear. However, it is difficult to see how we could be sure that Susan was not afraid if she stood and smiled at the bear (indeed, the US Park Service recommends standing still in such situations) or scratched her nose or performed any other behaviour. Skinner is forced to argue that Susan's behaviour is only fear behaviour if it occurs in the presence of the correct (fearful) event; that is, in the presence of a bear. However, by what means can we decide that the bear is a fearful event (the event problem)? Skinner must resort to saying that it is a fearful event because it gives rise to fear, and therein lies the circularity of the Skinnerian theory of emotions.
As well as this circularity that faces the behaviourist, Skinner's ideas invite an extra criticism over and above those directed at Watson. As Lyons notes:
. . . many instances of some emotions, and most instances of others, exhibit little or no operant behaviour. Grief, especially when it is about something irretrievably lost or dead, does not lead to much, if any, operant behaviour, because no behaviour can bring about any desired results. (1980, p. 22)
Skinner could retort that grief behaviour brings about sympathy from others, but he is on a slippery slope and there is no real answer to Lyons' arguments.
Psychological behaviourism, in our view, seems to offer us little as a framework within which to investigate emotions. Does philosophical behaviourism provide anything substantially different?
Was this article helpful?