Conventional conditioning models cannot make sense of drug behavior because they circumvent the psychological, environmental, and social nexus of which drug use is a part. One branch of conditioning theory, social-learning theory (Bandura 1977), has opened itself to the subjective elements of reinforcement. For example, Bandura described how a psychotic who continued his delusional behavior in order to ward off invisible terrors was acting in line with a reinforcement schedule that was efficacious despite its existing solely in the individual's mind. The essential insight that reinforcers gain meaning only from a given human context enables us to understand (1) why different people react differently to the same drugs, (2) how people can modify these reactions through their own efforts, and (3) how people's relationships with their environments determine drug reactions rather than vice versa.
Social-learning theorists have been especially active in alcoholism, where they have analyzed how alcoholics' expectations and beliefs about what alcohol will do for them influence the rewards and behaviors associated with drinking (Marlatt 1978; Wilson 1981). Yet it has also been social-learning theorists who have launched the alcohol-dependence syndrome and who seem to feel subjective interpretation is far less important than the pharmacological effects of alcohol in causing drinking problems (Hodgson et al. 1978, 1979). This lacuna in their theorizing is most noticeable in the inability of modem social-learning theorists to make sense out of cultural variations in drinking styles and experiences (Shaw 1979). Whereas McClelland et al. (1972) offered an experiential bridge between individual and cultural conceptions about alcohol (see chapter 5), behaviorists have regularly rejected this kind of synthesis in favor of direct observations and objective measurements of alcoholic