Confluence of Science and History in the Experimental Analysis of Behavior Course

Abstract

This is a review of content and method for incorporating the history of the experimental analysis of behavior (EAB) into the EAB course, although the material also could be adapted for any course related to the topics of learning and behavior change, or the history of psychology. Six elements associated with establishing a new discipline are considered as a framework for introducing the history of EAB: the intellectual leader/founding scientist(s), early proponents of the new area who advance and elaborate on the founder’s ideas, the cultural context in which the discipline develops, a set of methods, a textbook, and means of communicating with other, similarly inclined scientists. The historical ebb and flow of research and some of the reasons for these shifts are discussed next, with examples of EAB research themes that have shifted over time. Illustrating the history of EAB with specific milestone experiments seems a useful way to both introduce substantive research and its history. To that end, milestone experiments in EAB are discussed. The review ends with considerations about locating historical material within the EAB course.

“The history of science is science itself.”

Johann Wolfgang von Goethe

The content of graduate-level courses in the experimental analysis of behavior (EAB) typically include both empirical and theoretical material describing and interpreting controlling environmental variables related to behavior change. Such content also often includes material related to single-organism research methods and the roles of physiological and phylogenic variables in understanding learned behavior. The research reviewed in support of this content typically involves both human and nonhuman animals. All of this content is an amalgam of experimental investigations and general principles of behavior. The content develops in part from accumulated knowledge over time, but also from contemporary research topics that qualify and challenge established principles. Indeed, in many instances, principles may be best understood by a discussion of the conditions of their discovery or early articulation, giving credence to Goethe’s observation with which this review began. This confluence of past research and the present state of EAB gives particular relevance to including historical material and context in EAB courses otherwise dedicated to contemporary research and theory. The purpose of this review is to (1) examine aspects of EAB research in historical context and (2) suggest how instructors might incorporate this material into EAB courses.

The Birth of a Discipline

There are at least six elements associated with the development of a new discipline. They are the intellectual leader/founding scientist(s), early proponents of the new area who advance and elaborate on the founder’s ideas, the cultural context in which the substantive content of the discipline develops, a set of methods, a textbook, and means of communicating with other, similarly inclined scientists.

Historians of science note the contributions of both outstanding individuals and the Zeitgeist to the development of a scientific discipline’s methods and content. B. F. Skinner’s methods, insights, and research findings that began in the 1930s are the foundation of EAB and therefore hold an important place in any discussion of the history of the discipline. The elaboration and evolution of Skinner’s original ideas by the generations that have followed him constitute the contemporary substance of the EAB course. Discussion of the assumptions, methods, and philosophy of the experimental analysis of behavior is necessarily historical because of the importance of Skinner’s original work to those topics. Skinner’s early experiments, on such topics as conditioning, extinction, and discrimination learning, remain central to the scientific and conceptual content that today defines the EAB course. Although Skinner’s work was the foundation, as time passed, the forces of history and change operated on that early data base to move the science in myriad directions. Skinner’s influence, though still present in many EAB courses today, has been complemented by the work of many others, building on and deviating from those early directions of EAB. The shifting sands of scientific research often are represented historically in the context of individual scientists and their work. And so it is with EAB courses. Early work in areas of research like punishment or stimulus control are closely associated with particular experiments conducted and reported by particular individuals. It is not possible, for example, to discuss contemporary choice research without describing Herrnstein’s (1961, 1970) development of the matching law, or aversive control without acknowledging the work of Sidman (1953) or Azrin (e.g., Azrin & Holz, 1966). Other examples of such historically important work appear in the section labeled “Milestones in the Experimental Analysis of Behavior.”

Individual scientific behavior, however, also is a function of the environments in which it occurs. The topics and problems that comprise the core of the EAB course result from both scientific and cultural influences that often are less obvious than the individuals to whom the research is attributed. The role of more general influences and trends both within and outside the discipline—the Zeitgeist—sometimes can be overlooked in discussions of EAB research. It may be useful to distinguish these as internal and external Zeitgeists. Internal Zeitgeists in EAB might include developments within the broader discipline of psychology, including but not limited to more general topics in the psychology of learning. An example of research influencing and perhaps influenced by such an internal Zeitgeist is autoshaping/automaintenance. Autoshaping first was reported from a laboratory that had been focusing on discrete trial procedures in the study of operant behavior (Jenkins, 1970) and feature-positive/feature-negative discrimination learning (e.g., Jenkins & Sainsbury, 1970), both providing a context for what then developed. The discovery of autoshaping also occurred at a time when discussions of giving greater attention to evolutionary/biological variables and learning were gaining traction (summarized by Seligman, 1970).

An example of an external Zeitgeist influencing research topics and trends in EAB is translational research, which for a number of years has been a topic in many areas of science outside of psychology and behavior analysis. For both scientific and research-funding reasons, translational research has been the impetus for developing both methods and content areas that connect basic behavior analysis to behavioral processes in nonlaboratory settings. In so doing, many new connections and even research problems and areas have been stimulated or enhanced. One example is delay discounting (e.g., Madden & Johnson, 2010, which began primarily with basic questions about delays of reinforcement and choice behavior. Extending those findings more broadly into translational topics not only has enhanced applications of delay discounting principles, but also stimulated new basic research topics.

Three components of the research methods developed in EAB are instrumentation, experimental design, and data analysis. As noted in the preceding section, the history of research methods in EAB, like everything else, began with Skinner. During his graduate school days, Skinner moved his research to the laboratory of W. J. Crozier, an experimental physiologist. There Skinner fell under the spell of the apparatus of physiology and individual-subject experimental designs, designs that behavior analysts continue to use to the present (Michael, 1974; Skinner, 1956; Perone & Hursh, 2013). Skinner developed and adapted instruments capable of accurately recording high rates of discrete responses over extended periods that would have been impossible for a human observer to record with accuracy (e.g., Lattal, 2004). Once created, the instrumentation of EAB was advanced and refined by both behavior analysts and technicians from other disciplines whose instruments were adapted for use in pursuing EAB’s research agenda (Lattal, 2008). As is discussed below in the sections on “The Ebb and Flow of Research in EAB” and “Other Considerations in the Teaching of EAB,” changes in instrumentation also sometimes herald new research problems and questions. The experimental designs and methods for data analysis were sufficiently well-established that by 1960 Sidman elegantly summarized and embellished them in his Tactics of Scientific Research. Tactics remains in print today, over 60 years after its debut. One of the great strengths of Tactics is that it discusses research methods and experimental design at both a practical and conceptual level. Students learn not only how to do things, but the implications of doing them one way as opposed to another.

It is through classroom experiences that most students are introduced to disciplines and research areas. Courses require textbooks that organize and explain the accumulated knowledge of the discipline. Fancher and Rutherford (2016) noted “. . . the important role played by the appearance of a new textbook formally establishing a new academic discipline or subdiscipline." In the case of EAB, the foundational text book was Keller and Schoenfeld (1950). Inspection of the table of contents shown in Table 1 reveals a survey that is eerily contemporary. Although the contemporary content of the chapters might differ from what it was in 1950, Keller and Schoenfeld provided a general topical outline to which most teachers of EAB today easily will relate. Skinner (1953) also wrote an influential introductory text, but it is broader. Although it covers the basic principles of behavior, it does so without reference to specific experimental investigations and it then extrapolates basic principles to a host of individual and cultural issues that go beyond the experimental-analytic scope of Keller and Schoenfeld. One other textbook of particular note in the history of EAB is Honig’s (1966) Operant Behavior: Areas of Research and Application. This formidable edited volume summarized pioneering research of many leading EAB researchers, including Sidman, Azrin, Morse, Bijou and Baer, Catania, and Terrace, to mention just a few of the authors of the 19 chapters comprising the book. The material in Honig’s textbook served to introduce generations of scientists to many of the topics that remain central to EAB.

Table 1.

Table of Contents from Keller and Schoenfeld (1950)

Chapter	Title
1.	Psychology and the Reflex
2.	Respondent Conditioning
3.	Operant Conditioning
4.	Extinction and Reconditioning
5.	Generalization and Discrimination
6.	Response Variability and Differentiation
7.	Chaining
8.	Secondary Reinforcement
9.	Motivation
10.	Emotion
11.	Social Behavior

The final element of a new research area is the means of communicating scientific findings. Laties (1987, 2008) summarized the history of both the Journal of the Experimental Analysis of Behavior, the flagship journal of EAB, and its governing organization, the Society for the Experimental Analysis of Behavior. Systematically rotating its editorial staff of practicing scientists, a wide range of scientists working in many different facets of EAB have served as editors and editorial board members. Even in the digital age, with its greatly reduced delays in publication times relative to earlier periods, personal contacts and informal communication remains, as in any science, at the core of EAB. To that end, professional societies developed, with their annual conventions and myriad opportunities for both formal and informal communication of data and ideas.

The Ebb and Flow of Research in EAB

To paraphrase a well-known biblical phrase, “for every research program, there is a season.” Lakatos (1970) noted that at any given time in the history of a science, some research programs are progressive and others are degenerating. These extremes are points on a shifting continuum such that sometimes, a research program can be ascendant, attracting scholarly interest and research funding. At other times, that same research program can be descendent with respect to the criteria like those just mentioned, but still with advocates and practitioners, only to ascend again at still another time. Research programs thus ebb and flow, becoming more or less dominant or dormant across the life of a discipline. In contemporary times, examples of research problems that seem to be ebbing include avoidance behavior, behavioral contrast, parameters of reinforcement, punishment (see below), schedules of reinforcement, and stimulus generalization. Table 2 provides a list of what could be called progressive research programs within each decade since the 1950s, when JEAB first appeared. This analysis requires a couple of caveats: (1) Attempts to characterize the relative strengths of the many concurrent research topics being pursued at different times in EAB is admittedly loose and subjective, but some at least general trends of research areas ebbing and flowing can be discerned. (2) This is not to suggest that these progressive research programs were “the only game in town” during the indicated decades. For example, to its credit, JEAB’s editorial staff has encouraged considerable breadth in EAB research. As a result, although the areas listed in Table 2 seemed especially productive in those decades, research in most, if not all, of those areas also appears in other decades (e.g., behavioral economics was first discussed systematically in JEAB in the early 1980s).

Table 2.

Examples of Progressive Research Programs in EAB by Decade

Decade	Program
1950s	Schedules of Reinforcement
1960s	Punishment and Stimulus Control
1970s	Autoshaping; The Matching Law
1980s	Behavioral Momentum; Stimulus Equivalence
1990s	Behavioral Economics
2000s	Delay Discounting
2010s	Translational Research

In developing the history of EAB for students, it may be useful to consider some of the reasons for the ebb and flow of operant research. Novel discoveries in or outside the discipline result in shifting research emphasis, reflecting to a more or less extent Skinner’s (1956) observation that “when you run onto something interesting drop everything else and study it” (p. 223). Examples include schedule-induced polydipsia (Falk, 1961) and autoshaping (Brown & Jenkins, 1968). Shifting research priorities often are tied to broad cultural issues such as military readiness and drug abuse, and it is possible to relate some of the shifting research foci within EAB to funding priorities of agencies supporting behavioral research. Herrnstein et al.’s (1976) experiment on natural concepts with pigeons reportedly grew out of research at least partially funded by the military to help detect troop movements during the Vietnam War era. Funding is not just for programs with direct and immediate application, however. A significant amount of Skinner’s basic research on schedules of reinforcement (e.g., Ferster & Skinner, 1957) was funded by the Office of Naval Research. Funding priorities are, of course, not the only, and certainly not always the primary, force behind shifting research foci. Another such force is new technologies either developed in the discipline or imported from outside it (Lattal, 2008; see also the section on “Other Considerations in Teaching the History of EAB,” below). These technological changes may facilitate the appearance of new research problems and programs (see the discussion in the section below labeled “Other Considerations in Teaching the History of EAB”).

Research areas and problems may suffer when a central player moves on to other problems retires, or, worse, dies. From the late 1950s to the mid-1960s, Azrin and colleagues (e.g., Azrin & Holz, 1966) research on punishment appeared frequently in JEAB, after which it diminished and then disappeared almost completely as a research topic. Punishment becoming a degenerating research area may relate in part to the perception among researchers that it was rather “mined out,” with researchers moving on to other, seemingly more fruitful research programs. Of course, a major factor for its decline was that the primary contributor to the area, Azrin, shifted his research interest to other problems and their applications, leaving the experimental analysis of punishment with a “leadership” void. Especially in the case of punishment, ethical issues concerning its use in practice began to arise in applied behavior analysis. Such issues also may have contributed to the degenerating research program related to punishment.

In the 1970s, Hake and his colleagues (e.g., Hake & Vukelich, 1972, 1973) contributed extensively to research in the area of social operant behavior. When Hake died unexpectedly in 1982, only a few researchers (e.g., Marwell & Schmitt, 1975; Schmitt, 1998) continued to conduct such research. As their research activity diminished over time, no one picked up the gauntlet and the area began a degenerative phase and ultimately disappeared from EAB, only to undergo something of a resurgence in the last few years in both JEAB (e.g. Carvalho et al., 2018) and elsewhere (e.g., Silberberg et al., 2014; Tan & Hackenberg, 2015).

Milestones in the Experimental Analysis of Behavior

There are experiments in any discipline that introduce new behavioral phenomena, change the way a problem is viewed, reshape theory with respect to the immediate problem, and more generally excite a generation of students new and old as they leave lasting contributions along the historical path to the present. The EAB course must focus on contemporary issues, but most contemporary issues can be traced to articles describing classic experiments that have shaped the discipline. This state of affairs suggests both the value of including such historical material and the necessity of distilling it to its essence. What follows is an example of such distillation, with the recognition that milestone research always will be in part a matter of an individual proclivities, orientation, and history in the discipline. Thus, the baker’s-dozen list of milestone experiments that follows is offered as neither prescriptive or definitive, but rather to encourage EAB instructors to consider historically groundbreaking experimentation that they believe provides students a sense of the provenance of, and perspective on, the current research activities of the discipline. The articles appear in chronological order, with a brief explanation of why each was included.

Skinner ( 1948) is one of his most cited articles and is foundational in delineating the importance response-reinforcer temporal contiguity in learning. Skinner’s interpretation has been countered in several important subsequent articles (Staddon & Simmelhag, 1971; Timberlake & Lucas, 1985), but the notion of superstitious behavior and adventitious reinforcement remain important concepts in behavior analysis. For example, the rationale for including the change-over-delay in concurrent schedules is to minimize adventitious chains of behavior.

Sidman ( 1953) analyzed an avoidance procedure that still bears his name. The experiment is important not only because it questioned the dominant two-factor theory of learning based on the learning of fear to a warning stimulus associated with shock delivery, but also because it harkened a new perspective on reinforcement, one involving reinforcement as integrated events over time (see Herrnstein & Hineline, 1966) that is contrasted to the temporal contiguity notions exemplified by Skinner’s (1948) analysis described above. Other free-operant procedures subsequently supplanted Sidman’s (de Villiers, 1972), largely because they eliminated temporal cues as a source of control over responding.

Guttman and Kalish ( 1956) analyzed the formation of stimulus generalization gradients around the positive discriminative stimulus. This was not the first experiment examining stimulus generalization gradients (see Kimble, 1961, for a review of even earlier investigations of stimulus generalization), but Guttman and Kalish showed the development of stimulus generalization as a function of the presence and absence of discrimination training of the operant response.

Blough ( 1958) investigated the time course of dark adaptation of the pigeon’s visual system. Although dark adaptation functions were well-established by the time of this experiment, Blough’s novel approach to the problem of psychophysical relations in nonhuman animals and its incorporation of conditional stimulus control methods took the analysis of stimulus control to a new level of sophistication, in terms of both the instrumentation of the experiment and its methods of analysis.

Azrin and Holz ( 1966) remains the definitive review of punishment (itself a commentary on the decline of research in this area). Furthermore, it is one of the best illustrations of a systematic experimental analysis of behavior in the history of the discipline.

Herrnstein’s ( 1961) experiment was the foundation for the matching law (Herrnstein, 1970; see also Baum, 1973, 1974) and the subsequent development of the quantitative analysis of behavior (e.g., Commons et al., 1987).

Herrnstein and Hineline ( 1966) offered evidence supporting the assertion that shock-frequency reduction reinforced free-operant avoidance behavior. This experiment was among the first of many arguing support for a molar, as opposed to molecular, mechanism for reinforcement (see also Herrnstein’s [1969] subsequent further development of a theory of avoidance behavior).

Rachlin and Green ( 1972) provided both an empirical and conceptual basis for what later developed into delay discounting in their experiment investigating choice between more immediate, smaller rewards and larger, more delayed ones.

Nevin ( 1974) described a series of experiments closely related to his earlier investigations of behavioral contrast (Reynolds, 1961), but in this article he introduced his ideas about resistance to change and response strength, forerunners to a more general account of behavioral momentum (Nevin et al. 1983). Subsequent research by Nevin and his colleagues on these topics have greatly advanced mathematical modeling of behavioral processes and the quantitative analysis of behavior generally.

Herrnstein et al. ( 1976) examined the learning of natural concepts as an instance of stimulus control. This experiment also is one of the pillars of the area today called animal cognition. (The inclusion of three of Herrnstein’s experiments in this list reflects his important role in the history of EAB as an experimenter, a theorist, and a teacher.)

Hursh (1980; see also Lea, 1978) provided systematic analysis of how principles from economics relate to and expand on principles of behavior derived from EAB.

Sidman and Tailby ( 1982) presented and discussed evidence for stimulus equivalence, a topic that became ascendant with this research (see Table 1), ultimately offering a behavior-analytic alternative to cognitive theories of concept learning and to critics who suggested that behavior analysis might not be useful in describing the development of multifaceted verbal relations.

Peele et al. (1984) analyzed the contribution of temporal contiguity to response maintenance in a series of sophisticated, well-designed experiments they interpreted as showing the critical role of molecular variables in the maintenance of operant behavior.

Other experiments worthy of consideration for a more expanded version of the list include Weiner’s (1964) investigations of human behavioral history, which raised questions about the controlling variables of human schedule performance; Galizio’s (1979) analysis of contingency and rule-controlled responding; Segal’s (1972) chapter on the provenance of operants; Morse and Kelleher’s (1977) discussion of the functional nature of reinforcement, bringing home the point about functional definitions of behavioral phenomena; Staddon and Simmelhag’s (1971; see also Timberlake & Lucas, 1985) critique of Skinners (1948) “superstition” experiment and their subsequent interpretation of their results in relation to selectionist theory; and Dews’s (1955) early experimental analysis of drug effects on behavior. Mazur’s (1987) research provided the methods for investigating delay discounting; Brown and Jenkins’s (1968) investigation of autoshaping was commented on previously; Reynolds’s (1961) analysis of behavioral contrast, along with Herrnstein (1961), was one of the first experiments to emphasize reinforcement context in maintaining responding; Catania and Reynolds’s (1968) quantitative analysis of interval schedule performance provided data illustrating the exquisite sensitivity of operant behavior to subtle variations in reinforcer distributions; and Baum’s (1973) analysis of the correlation based law of effect was a next logical step from Herrnstein (1970) into the molar reinforcement framework. These are just examples of the many experimental analyses that shaped the discipline that in turn account for the topics of contemporary EAB courses. The inclusion of such classics as these depends on the course time and depth that can be devoted to the history of EAB. Nonetheless, discussions of contemporary topics in an EAB course might consider at least acknowledging the historical grounding of those topics in articles like the ones noted in this section.

Other Considerations in Teaching the History of EAB

Two further considerations teaching the history of EAB are the roles of women, minorities, and international groups in that history and the role of technology in the development of EAB. Social context in recent years has gained a voice in the history of psychology. Increasingly featured in history of psychology courses are the professional histories of women, minority group members, and members and movements beyond the shores of the United States. All of these groups have played a role in the history of EAB. Among women behavior analysts, however, only the research of Segal (1972) was mentioned in the preceding section. This paucity of women in EAB in its earlier days is reflective of a more general state of affairs in psychology and science generally. Marion Breland (in later years, Breland Bailey) was among the first, if not the first, woman behavior analyst. Early in her career, she moved into the field of animal training with her first spouse, Keller Breeland (e.g., Skinner, 1983, p. 31). The first JEAB Editorial Board female member was Barbara Ray (1970–1972), who studied with Murray Sidman; the first female associate editor (1977–1979) was Patricia Blough, at the time affiliated with Brown University, and Amy Odum, a graduate of the behavior analysis program at West Virginia University, was the first female editor (2015–2019). As women, LGBTQIA, and ethnic minority group members gain representation among EAB researchers, future iterations of articles like this one will reflect the changing research topics and trends that their inclusion has brought to the discipline. The international community has been represented in EAB from its onset, with many prominent international EAB scholars in EAB’s early years obtaining their doctoral or postdoctoral training at American universities. These individuals subsequently returned to their native countries—for example, Brazil, Japan, Mexico, and Norway—to establish graduate training programs, and founded journals and societies promulgating EAB. Many of these same individuals, and subsequent generations trained by them, continued to publish in English-language journals, as well as in their native languages.

A seldom-discussed dimension of the history of EAB is the relation between the basic science and technology. The nature of operant research, with its complicated schedules of reinforcement and often thousands of responses to be recorded per session, required automated control of experiments. Experiments often required the construction of unique, sometimes one-of-a-kind apparatuses (e.g., Azrin et al., 1966, analysis of extinction-induced aggression; the analysis of drugs as reinforcers) and apparatus modification to meet specific experimental needs. These technical demands were met by turning to the technologies of the times. Automating experimental control and data collection, for example, first involved the relay technology of early telephone systems (see, e.g., Dinsmoor, 1990). Relay technology gave way to transistors, and transistors to integrated circuitry and digital computers, and digital computers to ever more portable devices such as smartphones and the like. From the first appearance of digital computers to control operant experiments noted above to adapting smartphones to control an experiment from literally anywhere in the world (Kuroda, 2017), the history of behavior analysis mirrors the history of these and other technologies of their times. Science and technology are symbiotically related in that science creates demands for new technology where it doesn’t already exist and extant technology creates tools that are adapted by science to allow movement into areas of research that otherwise would not be possible (Lattal, 2008).

Integrating Historical Information into EAB courses

Including material on the history of EAB provides students a perspective on and a framework for considering the contemporary state of the science. This inclusion can be done in several ways: with a “Genesis” unit at the beginning of the course, integrating historical introductions to the different substantive modules (e.g., reinforcement, punishment, stimulus control) throughout the course, or, perhaps best of all from a history-teaching perspective, a combination of the two. A Genesis unit could consist of a single or multiple class period(s) near the beginning of the course that places the data and theory to follow in historical context by drawing from material like that presented in this review. As an alternative, this latter material could include historical examples from each of the substantive units comprising the course. Given the inductive nature of EAB research, it seems most fitting to trace the development of research areas in terms of the progressions and “building on” nature of behavioral research.

Conclusion

Contemporary science flows both from what has come from its more distal past and from newer channels cut from the same firmament but that have spread along different paths from the original streambed. Showing how present research has benefitted from its own past and how it has veered in what were at the time not-well-understood directions helps students place the present state of the science in perspective, facilitates questioning the status quo, and recognize the importance of change in EAB’s viability.

There are many paths to the present that one could consider in reconstructing the history of EAB research and its relation to broader historical issues in behavior analysis. Where to begin and end “the past” is a judgement call by the interpreter/teacher of history. Trends in research take time to develop. Some seemingly promising contemporary experiments “fizzle” in terms of stimulating further research, whereas other immediately seemingly less-promising ones with time result in new and exciting research directions. The present review took a long view by tracing current topics commonly taught in EAB classes and looking back to their origins.

In considering the nature of the operant, Skinner (1953, p. 91) observed that ". . . an operant is not something which appears full grown in the behavior of an organism. It is the result of a continuous shaping process." And so it is with EAB. It has been shaped by generations past and will continue to be so shaped by generations to come. Helping students appreciate the events and people responsible for shaping  the science by teaching something of the history of EAB strengthens the present science and helps ensure its future.

Declarations

Conflicts of Interest

I have no conflicts of interest vis-à-vis this article.