{"title":"Operant Variability and the Power of Reinforcement.","authors":"A. Neuringer","doi":"10.1037/H0100673","DOIUrl":null,"url":null,"abstract":"\"Since Darwin, the central project of evolutionary biology has been to explain the origin of biodiversity--to determine how novel species and their characteristics have evolved\" (Thorton, 2006, p. 157). Operant conditioning can be described in similar terms: Since Skinner, the central project has been to explain how operant behaviors originate and change. To explain biodiversity, on the one hand, and behavioral diversity, on the other, we must consider the variations from which each emerges. This paper is about the causes, consequences, and possible applications of variability, but I begin by noting some parallels in the area of genetics. Continual variation in genetic material provides the bases of all evolved forms of life. Lewis Thomas said this in a more evocative way: \"The capacity to blunder slightly is the real marvel of DNA. Without this special attribute, we would still be anaerobic bacteria and there would be no music\" (quoted in Pennisi, 1998, p. 1131). Genetic variability, due to mutations in DNA, have many causes including errors during normal replication, insults from chemicals or radiation, jumps or transpositions of genetic materials, and other \"spontaneous\" changes. In sexually reproducing organisms, another source of continual variation occurs during gamete formation. When genetic material in sperm and egg cells divide, there is random and independent assortment within individual chromosomes and random crossings between portions of maternal and paternal chromosomes. Mutations, jumps, assortments, and crossings are said to occur \"randomly,\" that is, without regard to the current \"needs\" of an organism or changes that result. However, \"random\" does not mean without influence or boundaries. The processes that permit and maintain genetic variability have themselves evolved under selection pressures. \"Chance favors the prepared genome ... Evolutionary strategies evolve, under the pressure of natural selection; this makes the process of evolution more efficient ... (T)he genome ... (has an) ability to create, focus, tune and regulate genetic variation and thus to play a role in its own evolution\" (Caporale, 1999, pp. 1 & 15). A combination of variation and selection at work within the genome itself may best be described as bounded stochasticity , with mutations, mixings and variations occurring stochastically and unpredictably, but within a confined milieu that has been selected and conserved over evolutionary time. As will be seen, similar bounded stochasticity is an attribute of operant behavior as well. A word about terminology. \"Stochastic\" and \"random\" will be used synonymously in the present paper, both to indicate a sequence of instances from a defined set, with prediction of the next instance not possible at a level greater than the relative frequencies of members of the set. Imagine, for example, a well-mixed tub filled with 200 red balls, 200 blue balls, and 200 green balls. Balls are selected one at a time with replacement and continual mixing. Prediction of the next color will be no better than 1/3 on average, (200/600), the relative frequencies of each color. If there were 100 red, 200 blue, and 300 green balls, then prediction of green would be correct no more than 1/2 the time on average, (300/600), and so on. Similarly, one can imagine a biased pair of die: Although both dice come up sixes more often than expected (say each does so 1/3 of the time), the outcome of each die roll remains stochastic because having just rolled a six neither raises nor lowers the chances of rolling another six in the future. Although selection may occur stochastically, the emergent instances are limited to members of the set and influenced by their relative proportions or \"strengths.\" As will be shown, there are many types of pressures at work to define and influence the set from which operant instances emerge. All behaviors vary, of course, as do all things physical. But variability plays an especially important role for operant behaviors. …","PeriodicalId":88717,"journal":{"name":"The behavior analyst today","volume":"10 1","pages":"319-343"},"PeriodicalIF":0.0000,"publicationDate":"2009-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"26","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"The behavior analyst today","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1037/H0100673","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 26
Abstract
"Since Darwin, the central project of evolutionary biology has been to explain the origin of biodiversity--to determine how novel species and their characteristics have evolved" (Thorton, 2006, p. 157). Operant conditioning can be described in similar terms: Since Skinner, the central project has been to explain how operant behaviors originate and change. To explain biodiversity, on the one hand, and behavioral diversity, on the other, we must consider the variations from which each emerges. This paper is about the causes, consequences, and possible applications of variability, but I begin by noting some parallels in the area of genetics. Continual variation in genetic material provides the bases of all evolved forms of life. Lewis Thomas said this in a more evocative way: "The capacity to blunder slightly is the real marvel of DNA. Without this special attribute, we would still be anaerobic bacteria and there would be no music" (quoted in Pennisi, 1998, p. 1131). Genetic variability, due to mutations in DNA, have many causes including errors during normal replication, insults from chemicals or radiation, jumps or transpositions of genetic materials, and other "spontaneous" changes. In sexually reproducing organisms, another source of continual variation occurs during gamete formation. When genetic material in sperm and egg cells divide, there is random and independent assortment within individual chromosomes and random crossings between portions of maternal and paternal chromosomes. Mutations, jumps, assortments, and crossings are said to occur "randomly," that is, without regard to the current "needs" of an organism or changes that result. However, "random" does not mean without influence or boundaries. The processes that permit and maintain genetic variability have themselves evolved under selection pressures. "Chance favors the prepared genome ... Evolutionary strategies evolve, under the pressure of natural selection; this makes the process of evolution more efficient ... (T)he genome ... (has an) ability to create, focus, tune and regulate genetic variation and thus to play a role in its own evolution" (Caporale, 1999, pp. 1 & 15). A combination of variation and selection at work within the genome itself may best be described as bounded stochasticity , with mutations, mixings and variations occurring stochastically and unpredictably, but within a confined milieu that has been selected and conserved over evolutionary time. As will be seen, similar bounded stochasticity is an attribute of operant behavior as well. A word about terminology. "Stochastic" and "random" will be used synonymously in the present paper, both to indicate a sequence of instances from a defined set, with prediction of the next instance not possible at a level greater than the relative frequencies of members of the set. Imagine, for example, a well-mixed tub filled with 200 red balls, 200 blue balls, and 200 green balls. Balls are selected one at a time with replacement and continual mixing. Prediction of the next color will be no better than 1/3 on average, (200/600), the relative frequencies of each color. If there were 100 red, 200 blue, and 300 green balls, then prediction of green would be correct no more than 1/2 the time on average, (300/600), and so on. Similarly, one can imagine a biased pair of die: Although both dice come up sixes more often than expected (say each does so 1/3 of the time), the outcome of each die roll remains stochastic because having just rolled a six neither raises nor lowers the chances of rolling another six in the future. Although selection may occur stochastically, the emergent instances are limited to members of the set and influenced by their relative proportions or "strengths." As will be shown, there are many types of pressures at work to define and influence the set from which operant instances emerge. All behaviors vary, of course, as do all things physical. But variability plays an especially important role for operant behaviors. …