|
| Many Faces of Reinforcement, Punishment, and Extinction |
| Sunday, September 29, 2019 |
| 10:30 AM–11:20 AM |
| Stockholm Waterfront Congress Centre, Level 2, Meeting Room 24/25 |
| Area: PCH/EAB; Domain: Translational |
| Chair: Iver H. Iversen (University of North Florida) |
| Abstract: The behavioral effects of reinforcement, punishment, and extinction are the cornerstones of behavior analysis. Yet, critiques of behavior analysis often point to problems with these basic effects even to the extent of stating that “reinforcement doesn’t work” or that the effects are due to other embedded variables (such as intrusion of natural behavior). In reality, the effects are complex and highly dependent on procedural variables. For example, reinforcement can strengthen behavior even at the level of a single reinforcer, yet after several sessions of acquisition, reinforcement no longer strengthens but instead maintains behavior. Effects also depend on levels of analysis, as seen when immediate effects of reinforcement, punishment, and extinction are very different from long-term effects. Stimuli routinely associated with reinforcement, punishment, and extinction can acquire discriminative properties for ongoing behavior with both excitatory and inhibiting effects. The presentations will outline various effects of reinforcement, punishment, and extinction in both laboratory and clinic, using examples from research to illustrate the problems with interpretation and terminology. The overall purpose is to alert behavior analysts that discourse about behavior becomes complex and unproductive when basic terms refer to different effects. |
| Instruction Level: Intermediate |
| Keyword(s): extinction-definition, punishment-definition, reinforcement-definition |
| |
| The Many Faces of Reinforcement |
| (Theory) |
| IVER H. IVERSEN (University of North Florida) |
| Abstract: That reinforcement strengthens and maintains behavior is well established. Yet, critiques of the strengthening effects of reinforcement rest on the idea that the reinforcer is also an eliciting stimulus for natural behavior other than the operant. Recent views in behavior analysis suggest that reinforcement delivery only serves as a discriminative stimulus. The presentation will outline what reinforcement does to behavior at various levels of analysis. Even a single reinforcer can strengthen behavior. Schedules of reinforcement establish stable patterning of operant behavior, and reinforcers no longer strengthen but instead maintain behavior. Reinforcer delivery serves as an immediate S-delta as it stops the operant promptly as the subject collects the reinforcer. Reinforcer delivery can also have other discriminative properties. Examples of strengthening, maintenance, and discriminative properties of reinforcement will be illustrated with examples from experiments. Contingencies of reinforcement as a controlling variable is often not emphasized outside of behavior analysis. The presentation will emphasize that in discourse about reinforcement and operant behavior, the strengthening, maintaining, and discriminative functions of reinforcers as well as contingencies are often mixed up or not fully articulated. Highlighting just one effect can lead to misunderstandings and a lack of appreciation of additional effects of reinforcement. |
| |
| The Many Faces of Punishment |
| (Theory) |
| PER HOLTH (OsloMet -- Oslo Metropolitan University) |
| Abstract: The standard definition of reinforcement requires that: (1) responses have consequences, (2) the response rate increases, and (3) the rate increases because of the response—consequence contingency. A standard definition of punishment is similar, except for a reduction in response rate. Response consequences can consist of the presentation or the removal of a stimulus. In either case, the stimulus change can have functions in addition to punishing the preceding response. First, the stimulus change can function as a discriminative stimulus for punishment, predicting punishment and thereby occasioning a reduced rate of responding. Second, the punishing stimulus change can function as a discriminative stimulus, signaling reinforcement and thereby occasioning an increased response rate. Third, the stimulus change may elicit responses that may be similar to, or incompatible with, the punished responses. When elicited responses are incompatible with the punished responses, the response rate may drop more than what is the direct effect of punishment. On the other hand, when elicited responses are similar to the punished responses, the punishment may be misinterpreted as reinforcement. Reducing, suppressing, discriminative, and eliciting effects of stimuli confuse discussions of the effects of punishment. Different definitions of punishment within behavior analysis have added to the confusion. |
| |
| The Many Faces of Extinction |
| (Theory) |
| MONICA VANDBAKK (Norwegian Association for Behavior Analysis/Oslo and Akershus University College) |
| Abstract: The procedure of withholding reinforcement for a previously reinforced response is called extinction. Extinction is customarily demonstrated in laboratory settings but has not been documented clearly in applied settings. This could be because extinction typically produces a gradual reduction in behavior. The extinction procedure can both eliminate and generate behavior. When reinforcement is removed abruptly, numerous unreinforced responses can follow and this is commonly known as an extinction burst. Other generative effects of extinction are extinction-induced variability, extinction-induced responding (aggression and attack), resurgence, and spontaneous recovery. These effects are typically transitory and disappear. However, stimuli associated with extinction become S-delta and can have immediate suppressive effects, that are permanent (turning on an S-delta can stop an operant immediately). Effects of extinction are of great importance in applied settings and should therefore be made more familiar to those who propose and practice extinction-based procedures in their clinical work. Because extinction procedures have different effects on behavior depending on previous history and time window of observation, communications about extinction effects should specify the particular methods and effects being discussed. |
|
| |