2.4 Use of punishment: spare the rod?
One issue, about which there is regular debate, concerns the use of ‘punishment’ to control children's behaviour. Behaviourism might, at first glance, appear to offer support for using punishment to reduce undesirable behaviour. For example, imagine that a father and his daughter are out shopping and the child steals a bar of chocolate and eats some of it whilst her father is distracted. He then sees her and shouts at her. In operant terms the stealing event is followed by an aversive response. This suggests that the stealing behaviour will occur less often in the future. However, one also needs to consider the contingency of the events. (Two events are said to be contingent on one another if the presence of one event immediately results in the occurrence of the other.) Before the father scolded the child, she had already eaten some of the chocolate. Consequently, eating the chocolate is likely to have been a contingent positive reinforcement, and perhaps the child's hunger may have also been reduced (a contingent negative reinforcement). Both of these immediate consequences increase the likelihood that the child will steal again (Huesmann et al., 2003).
Behaviourist research has shown that for punishment to be effective, it must be immediate (contingent), severe and consistently applied (Klein, 1996). However, outside of the laboratory, it is virtually impossible to achieve such aims: adults cannot supervise the behaviour of children continually and be in a position to intervene immediately with appropriate punishment every time a child misbehaves. In the absence of these conditions, punishment as a means of behavioural control is, at best, short lived. Even Skinner (1938) found that punishment can only temporarily suppress a behaviour in a specific context, not eliminate it. Furthermore, in terms of classical conditioning, the child might associate the aversive response (shouting) with the person delivering the response, and show a conditioned fear response to her parent (see Figure 4).
There is also a risk of inappropriate association: the child who is punished for stealing in the supermarket may associate the punishment with the wrong behaviour (e.g. eating the chocolate, or visiting the supermarket).
Punishment has been found to stimulate aggressive behaviour in some circumstances and to suppress it in others, and its long-term effects on behaviour are often not what was initially expected (Huesmann et al., 2003). A range of research studies has shown that, overall, punishment can be used to successfully manage inappropriate behaviour but it also has many negative short- and long-term consequences (Gershoff, 2002). These include increased aggression, decreased quality of relationships with carers, decreased mental health, and a later increased likelihood for antisocial and criminal behaviour. However, timeout procedures have been found to be effective in controlling the extent of behaviours like tantrums in typically developing children and those with learning difficulties (Klein, 1996). Yet, it should be noted that the need for contingency and consistency would also apply to the application of such schemes.
Punishment only teaches a child what response not to make. For behaviour to change, children also need to learn what alternative behaviour is appropriate and then be reinforced for producing it. For these reasons contemporary techniques of behavioural change based on behaviourism do not use punishment, but teach appropriate behaviours and increase their frequency through reinforcement. One example of such an application is known as applied behavioural analysis (ABA).