ABSTRACT
Behavioral models maintain that substance use and abuse are learned behaviors. The most widely researched behavioral model is the tension-reduction hypothesis (Conger, 1956), which proposes that states of tension (anxiety, fear, conflict, or frustration) are aversive motivational states and drug consumption is a rewarding activity because it reduces such states (Cappell & Greeley, 1987). The tension-reduction hypothesis includes two central postulates. First, alcohol and other drugs are capable of reducing tension. Second, substance-induced tension reduction serves to increase the likelihood of future drug use through the operant learning mechanism of negative reinforcement (Cappell & Greeley, 1987). Due to inconsistencies in the experimental support for each of these postulates, reviews have been quite pessimistic about the utility of the tension-reduction hypothesis as a global explanation for all drug use behavior and substance disorder development (Cappell & Greeley, 1987; Poherecky, 1991; Wilson, 1988). Criticisms leveled at the traditional tension-reduction hypothesis include its: (a) failure to consider motivations for drug use other than tension reduction, (b) failure to recognize that different drugs have different effects, (c) relatively narrow definition of tension, (d) failure to consider the situational context in which drug use occurs (e.g., relaxed vs. stressed state; in anticipation of, or following, stress), and (e) failure to consider relevant cognitive and individual difference variables (Cappell & Greeley, 1987; W. Cox, 1987; Poherecky, 1991; Wilson, 1988).
