Toward empirical behavior laws: II. Effectiveness of a conditioned reinforcer paired with food and shock

Abstract

a conditioned reinforcer paired with food and shock' Ab8traetA buzzer was paired with food in control rats and with food and shock in experimentals. (1) Baseline bar pressing in experimentals was depressed. (2) Both groups pressed equally often, on the average, to produce the buzzer. (3) Behavior Change was greater in experimentals. The data are consistent with a totalityof-reaction theory of conditioned reinforcement. ProblemConsider the following hypothesis. The strength of a rehlforcer is a function of the totality of reaction to it; as the reaction evoked by a stimulus increases, the reinforcing effect is, first, increasingly positive, and then decreasingly positive, and then increasingly negative; Some data in support of the hypothesis were reviewed in another paper (Segal,1965).Additional support is the following:(1) The opportunity to engage in one behavior may reinforce other behavior. Fighting cocks were reinforced by the sight of another fighting cock; Siamese fighting fish were more reinforced by the sight of their mirror image, or a moving model of another fish in aggressive display, than by the sight of a stationary model. The reinforcingness of the displays was correlated with their tendency to elicit aggressive reactions in S (Thompson, 1963; cf. Breland & Breland, 1961).(2) When a link is added to an operant chain, the quantity of behavior that must occur before primary reinforcement arrives increases. If the totality of reaction evoked by stimuli close to food is about optimal, then adding more reaction by adding more links may move the stimulus toward the decreasing limb of the reinforcing-effect function (cf. Segal, 1965). Thus, the stimulus correlated with the earliest link of a long chain may be nonreinforcing, or aversive (Kelleher & Fry, 1962). Schedule preference studies (e.g., Findley, 1958) also indicate that stimuli associated with large behavior requirements are less positively reinforcing than stimuli associated with smaller behavior requirements. As the requirements progressively rise, the stimuli correlated with the schedule may become aversive (e.g., Appel, 1963;Azrin, 1961).(3) Pliskoff-& Tolliver (1960) transformed a conditioned positive reinforcer to a conditioned negative reinforcer by depriving food-reinforced rats of water. The SD for the food-reinforced operant probably elicited conditioned alimentary respondents which increased in intensity or variety when the SD was presented to rats that were water-as well as food-deprived.Poychon. Sci., 1965, Yolo 2. Evalyn F. Segal SAN DIEGO STATE COLLEGEAs a result, the SD became aversive.(4) A stimulus correlated with a food-reinforcement schedule is more reinforcing when it appears in alternation with a stimulus correlated with extinction, punishment, or a less favorable food-reinforcement schedule (e.g., Reynolds, 1961Reynolds, , 1963Segal, 1964). The alternation of stimuli may elicit SNS reactions that get conditioned to the preferred stimulus. The combination of mil d SNS reactions and alimentary reactions to food may be more re...

Toward empirical behavior laws: I. Conditioned reinforcement

“…These data, and others to be reviewed elsewhere (Segal, 1965), are consistent with several theories of reinforcement. The present hypothesis is worth considering because it embraces a variety of reinforcement phenomena in a simple framework, it is testable, and it leads to some interesting, contrary-to-commonsense, predictions.…”

“…For example, one implication ofthe hypothesis is that, within limits, increasing the reaction to a stimulus in any way whatever, even by pairing it with shock, will increase the positive reinforcingness of the stimulus. This implication was the subject of an experiment reported elsewhere (Segal, 1965).…”