Several notation systems for reinforcement schedules have beendeveloped just after the start of this research field. Some systemsprovided good tools for creating complex schedules and for educationalpurposes, and sometimes they were interpreted to some commoncomputer languages via specific invented languages. The mostsuccessful system among them was the State Notation System(SNS) developed by Snapper and his colleagues and their interpretablelanguage SKED. In this paper, I propose another notation systemfor reinforcement schedule based on these previous efforts, particularlySNS and SKED. This new system, named Reinforcementschedule notation system (RNS), is composed with two subsystems,the Structural schedule and the Constituent schedule, which are mutuallycomplimented their limited scopes. The Structural scheduleconsists of the Scenes and the Conditions which sketch the outline ofreinforcement schedules an organism faces to. On the other hand,the Constituent schedule consists of connected chains of the ScheduleElements which contains three slots of (1) the evaluation of eitherresponse or time counters, (2) the operation of environmentalchanges and system signals, and (3) the jumped addresses dependedon the result of the evaluation slot. Although RNS provides compactdescriptions for the complex structures of reinforcement contingencies and many ideas for concrete operations of reinforcement schedules,it has several weak points, such as no elegant descriptive devicesfor dealing with serial responses （e.g., double alternate learning）and complex configurations of discriminative stimuli （e.g., higherconditional discrimination）. These limiations will be solved bydevelopment of frameworks for analysis of discriminative stimuli infuture.
|
