Donahoe J W, Burgos J E
Psychology Department, University of Massachusetts at Amherst, 01003, USA.
J Exp Anal Behav. 2000 Nov;74(3):331-46. doi: 10.1901/jeab.2000.74-331.
Revaluation refers to phenomena in which the strength of an operant is altered by reinforcer-related manipulations that take place outside the conditioning situation in which the operant was selected. As an example, if lever pressing is acquired using food as a reinforcer and food is later paired with an aversive stimulus, the frequency of lever pressing decreases when subsequently tested. Associationist psychology infers from such findings that conditioning produces a response-outcome (i.e., reinforcer) association and that the operant decreased in strength because pairing the reinforcer with the aversive stimulus changed the value of the outcome. Here, we present an approach to the interpretation of these and related findings that employs neural network simulations grounded in the experimental analysis of behavior and neuroscience. In so doing, we address some general issues regarding the relations among behavior analysis, neuroscience, and associationism.
在操作性行为被选择的条件作用情境之外,通过与强化物相关的操作改变了该操作性行为的强度。例如,如果通过使用食物作为强化物来习得按压杠杆行为,之后食物又与厌恶性刺激配对,那么随后测试时按压杠杆的频率就会降低。联想主义心理学从这些发现中推断,条件作用产生了反应-结果(即强化物)的关联,并且操作性行为强度降低是因为将强化物与厌恶性刺激配对改变了结果的价值。在此,我们提出一种解释这些及相关发现的方法,该方法采用基于行为实验分析和神经科学的神经网络模拟。这样做时,我们解决了一些关于行为分析、神经科学和联想主义之间关系的一般性问题。