DESCRIPTIVE VERSUS FUNCTIONAL ACCOUNTS OF PSYCHOLOGICAL EVENTS: CYBERRAT AS A CASE IN POINT

Transcription

1 Behavior and Philosophy, 39/40, (2011/2012) Cambridge Center for Behavioral Studies DESCRIPTIVE VERSUS FUNCTIONAL ACCOUNTS OF PSYCHOLOGICAL EVENTS: CYBERRAT AS A CASE IN POINT Matthew Lewon, Maria I. Munoz Blanco, and Linda J. Hayes University of Nevada, Reno ABSTRACT: While the CyberRat simulation described by Ray (2011/2012) has considerable value as an educational tool, its value also lies in the validation of the descriptive interbehavioral systems analysis (IBSA) approach upon which it was developed. The descriptive IBSA approach differs in important ways from the predominantly functional approach typically adopted by researchers in the experimental analysis of behavior (TEAB), and while TEAB has been successful in identifying many important environment behavior functional relations, a strictly functional approach to the study of behavior can lead to a tendency to overlook subtle but important aspects of the events under scientific investigation. As such, we propose that TEAB may benefit from and expand its scope by incorporating elements of a descriptive research approach. Key words: descriptive analysis, functional analysis, experimental analysis of behavior, simulations Simulations are generally considered to have value in two broad areas: (1) education and scientific validation and (2) prediction (Axelrod, 2007). With regard to the former, CyberRat (Ray, 2011/2012) provides an array of features designed to provide users with an exceptionally lifelike simulation of a rat in an operant chamber setting. CyberRat was developed primarily as an educational tool for student learning in times when live animal laboratories are increasingly rare, and its many features seem likely to provide students with a relatively authentic animal laboratory experience. Although some of the vagaries of the animal laboratory are inevitably missing, and one may wonder about the qualitative experiential differences between training a real rat and a computer representation of one, there can be little doubt that CyberRat has significant value as a training tool, and the adequacy of the simulation is impressive. Beyond its value as an educational tool, Ray argues that CyberRat also serves a scientific purpose. Although CyberRat was admittedly not designed for prediction, Ray contends that it has significant conceptual value in that it validates the descriptive interbehavioral systems analysis (IBSA) research approach upon which it was developed. The descriptive IBSA approach adopted in the creation of CyberRat differs in important ways from the functional approach traditionally espoused by researchers in the experimental analysis of behavior (TEAB). The AUTHORS NOTE: Please address correspondence to Matthew Lewon, Psychology Department, University of Nevada Reno, MS 296, 1664 N. Virginia Street, Reno, NV 89557; matthew.lewon@gmail.com 315

2 LEWON, MUNOZ BLANCO & HAYES remainder of the present paper will describe some of these differences and consider ways in which incorporating elements of a descriptive approach might enhance TEAB s predominantly functional approach to investigations of psychological events. The descriptive IBSA approach underlying CyberRat s development is based upon J. R. Kantor s (1958) interbehavioral psychology, in which the description of events takes precedence over experimental manipulation. Ray notes that one of the difficulties inherent in descriptive research of the kind espoused by Kantor is determining what sorts of observations are necessary and sufficient for an adequate description of original events. In other words, how much description is sufficient for a relatively complete account of events? With respect to this issue, the conceptual value of CyberRat lies in its capacity to validate the adequacy of the descriptions produced by the IBSA research approach. The degree to which the performance of CyberRat is indistinguishable from the performance of an actual rat in an operant chamber under certain user-defined conditions provides a measure of the adequacy of the descriptive research methods from which CyberRat was derived. Ray proposes a series of Turing (1950) tests to evaluate the structural, functional, and operational aspects of the simulation and concludes that CyberRat passes all such tests to a reasonable extent. Upon these bases, Ray argues that the IBSA approach to research leading to the development of CyberRat is valid because it allows for authentic reconstructions of original events. The fact that CyberRat convincingly passes the Turing tests highlights some important differences between the descriptive interbehavioral research approach underlying its development and the functional approach typically taken by researchers in TEAB. TEAB s functional approach is exemplified by the work of its founder, B. F. Skinner, who characterized behavior as TEAB s primary dependent variable and rate of response as its basic datum (Skinner, 1953, 1966). Environmental events were regarded as independent variables that influenced rate of response, and TEAB researchers were tasked with the identification of all such variables of which the rate of some class of responses (i.e., the operant) was a function (Skinner, 1966). TEAB s emphasis on the operant and the variables of which the operant s frequency is a function, while successful in identifying a wide range of useful behavior environment relations, has led to a tendency to overlook some of the more subtle aspects of the events TEAB researchers are investigating. As Ray notes, the breadth of phenomena that CyberRat simulates, as well as the naturality of the simulation, could not have been realized by a functional account alone. Functional accounts such as those emphasized by researchers in the TEAB tradition definitely identify useful relations between events, but the emphasis on function can lead to the tendency to overlook related phenomena. The shortcomings of a strictly functional approach to the analysis of psychological events are exemplified by the fact that, despite over 80 years of research utilizing the operant chamber, there is still no recognized systematic procedure for magazine training or response shaping. As Ray puts it, these techniques tend to be passed down to subsequent generations of researchers as a 316

3 DESCRIPTIVE VERSUS FUNCTIONAL ACCOUNTS OF PSYCHOLOGICAL EVENTS sort of laboratory folklore or art (p. 283). Many other issues relevant to preexperimental training that have conceivable effects on experimental performances tend to be similarly overlooked by TEAB researchers. For example, what are the most effective ways to thin schedules of reinforcement from continuous to intermittent reinforcement? If an experiment requires discrimination, what is the most efficient way to establish it? What effects do pre-experimental training procedures have on organisms experimental performances? All basic researchers, at one time or another, have had difficulties training particular individual organisms for whom standard pre-training procedures are ineffective. How do we account for and remediate such difficulties? TEAB research often uses organisms with prior experimental histories (e.g., Ferster & Skinner, 1957). What effects do such histories have on subsequent performances? The descriptive approach used to develop CyberRat involved an attempt to systematically examine at least some of these questions, all of which deserve serious consideration. A descriptive research approach like IBSA has the potential to inform and enhance TEAB research by addressing some of these potentially important but historically overlooked issues. This is not to say that a functional approach is without its merits. Functional accounts are particularly useful because they allow for the identification of factors of event fields which, when manipulated, produce reliable changes in some other part of the field. In TEAB terms, functional analyses allow for the identification of the robust environmental variables of which behavior is a function. Such analyses, however, should not be mistaken for full descriptive accounts of psychological events. Descriptive accounts involve attempts to identify all of the factors participating in a particular event field as well as the relations among those factors. Descriptive accounts do not focus solely on the so-called functional relations with their connotations of causality (Kantor, 1970). A descriptive account of an event is necessarily a functional one in that the relations identified by a functional analysis are included in the description, but functional accounts are not necessarily sufficiently descriptive. Descriptive accounts consider the full field of factors in which events occur and entertain the notion that when one factor within the field changes, all other factors of the field, as well as the totality of the field itself (i.e., emergent properties), change in turn. Functional accounts, on the other hand, abstract two or more portions of the field in which events occur to determine relations between them, relations that may or may not hold when other factors participating in the field are taken into account. As we have noted, this approach has pragmatic utility but can lead scientists to overlook more subtle but equally important factors involved in psychological events. It is worth noting that TEAB has always incorporated some descriptive elements into its predominantly functional analyses. To cite a few examples, Skinner (1957) recognized the importance of multifactor fields (i.e., context) in his discussion of multiple causation, in which a number of functional relations, established independently, combine on a given occasion to produce a novel instance of behavior. TEAB researchers have on occasion shifted their focus from the operant to investigate phenomena including schedule-induced (Staddon, 1977) or adjunctive behavior (Falk, 1971), autoshaping (Brown & Jenkins, 1968), and the 317

4 LEWON, MUNOZ BLANCO & HAYES intrusion of phylogenetic behavior into operant performances (Breland & Breland, 1961). TEAB is increasingly incorporating motivational variables into its analyses via the motivating operation construct (Laraway, Snycerski, Michael, & Poling, 2003; Michael, 1982). These developments represent TEAB s implicit acknowledgement of the need to look beyond the operant in isolation and incorporate field factors into analyses of psychological events. Incorporating elements of a descriptive research approach would only enhance the admirable work being done in TEAB. In closing, we must note again that the functional approach adopted by TEAB has been, by all measures, incredibly successful. By adopting this approach, TEAB has been successful in accounting for a large portion of human behavior. Functional analyses, however, should not be confused with full descriptive accounts of psychological events. The shortcomings of TEAB s functional approach often make themselves apparent in the difficulties inherent in applying the approach to so-called complex human behaviors, especially those deemed private (Hayes & Fryling, 2009; Parrott, 1983). These phenomena have proven to be somewhat intractable to TEAB s primarily functional approach, and any useful account of them will require analyses of the entire fields of factors of which they are a part. CyberRat demonstrates the value of incorporating elements of a descriptive research approach into the analysis of behavior, and TEAB would likewise benefit from and expand its scope by incorporating elements of this approach. References Axelrod, R. (2007). Simulation in the social sciences. In Rennard, J. P. (Ed.), Handbook of research on nature inspired computing for economics and management (pp ). Hershey, PA: Idea Group. Breland, K., & Breland, M. (1961). The misbehavior of organisms. American Psychologist, 16, Brown, P. L., & Jenkins, H. M. (1968). Auto-shaping of the pigeon s key peck. Journal of the Experimental Analysis of Behavior, 11, 1-8. Falk, J. L. (1971). The nature and determinants of adjunctive behavior. Physiology and Behavior, 6, Ferster, C. B., & Skinner, B. F. (1957). Schedules of reinforcement. New York: Appleton- Century-Crofts. Hayes, L. J., & Fryling, M. J. (2009). Overcoming the pseudo-problem of private events in the analysis of behavior. Behavior and Philosophy, 37, Kantor, J. R. (1958). Interbehavioral psychology. Chicago: Principia. Kantor, J. R. (1970). An analysis of the experimental analysis of behavior (TEAB). Journal of the Experimental Analysis of Behavior, 13, Michael, J. (1982). Distinguishing between discriminative and motivational functions of stimuli. Journal of the Experimental Analysis of Behavior, 37, Skinner, B. F. (1953). Science and human behavior. New York: Macmillan. Skinner, B. F. (1957). Verbal behavior. Acton, MA: Copley Publishing. 318

5 DESCRIPTIVE VERSUS FUNCTIONAL ACCOUNTS OF PSYCHOLOGICAL EVENTS Skinner, B. F. (1966). What is the experimental analysis of behavior? Journal of the Experimental Analysis of Behavior, 9, Laraway, S., Snycerski, S., Michael, J., & Poling, A. (2003). Motivating operations and terms to describe them: Some further refinements. Journal of Applied Behavior Analysis, 36, Parrott, L. J. (1983). Systemic foundations for the concept of private events. In N. W. Smith, P. T. Mountjoy, & D. H. Reuben (Eds.), Reassessment in psychology: The interbehavioral alternative (pp ). Lanham, MD: University Press of America. Ray, R. D. (2011/2012). CyberRat, interbehavioral systems and analysis, and a Turing test trilogy. Behavior and Philosophy, 39/40, Staddon, J. E. R. (1977). Schedule-induced behavior. In W. K. Honig & J. E. R. Staddon (Eds.), Handbook of operant behavior (pp ). Englewood Cliffs, NJ: Prentice Hall. Turing, A. M. (1950). Computing machinery and intelligence. Mind, 59,