Forgotten Heroes of Behavior Analysis: Bill Buskist and the Coming of Age of Human Operant Research

This is part of an occasional series celebrating people who made a difference but who may be little discussed in contemporary behavior analysis. If you are interested in penning a guest post about your own Forgotten Hero, contact Tom Critchfield (

What we’re celebrating:

Bill Buskist played a pivotal role in helping human operant research, which once was marginalized, become a central part of the experimental analysis of behavior.

When I entered behavior analysis around 1980, the scientific value of human operant research was very much in question. Early failures to replicate basic findings with animals  (for instance, some reinforcement schedules seemed not to work the same way with humans) led some to believe that conducting basic human research was a waste of time. Cognitively-oriented critics from outside behavior analysis used these failures to suggest that operant processes don’t apply to humans.

Another 10 or 20 years would pass before human operant research appeared routinely in behavioral journals, and even longer before it gained acceptance as an equal partner to animal studies. It may be no coincidence, therefore, that early researchers who dabbled in human studies also typically maintained an animal lab that presumably was a more reliable source of reinforcers. Almost nobody doing basic research at the time dared put all of their eggs in the human operant basket.

Bill Buskist

One exception was my Forgotten Hero, Bill Buskist. Bill was a creative investigator who was all in on human operant research, but his most noteworthy contribution may be the effort he invested in building intellectual infrastructure in which human operant research would grow and eventually thrive. I will add, in an aside to which I cannot do justice in a short essay, that Bill is one of the kindest, most gracious, and most helpful people I have met professionally. I was lucky to get to know him when I was just starting out.

Bill had a great pedigree. He trained at Brigham Young University under the tutelage of Hal Miller, which makes him an academic grandson of the Harvard pigeon lab. Afterward he spent his entire career at Auburn University, where I had the privilege of working with him for seven years.

Bill did a human matching law dissertation, and he left graduate school with human laboratory research as his sole focus. His subsequent work touched on some complex issues that still have been studied too little, such as rule governance and social behaviors like cooperation and competition. He was also concerned with determining the most effective types of reinforcement to use in human operant studies (see below), something that was far from standardized at the time.

Human Operant Research Program

To me, two things stand out about Bill’s approach to science. First, he worked harder at it than most people. Bill never confused native intelligence with scientific skill, and he didn’t shy away from developing new ways of doing research that suited his interests. For example, in one set of studies (see below), he painstakingly built a novel technology for how to record operant responses and deliver food reinforcers to humans. Second, unlike many other scientists I have known, Bill never let his ego and his science co-mingle. This often worked to his detriment — Bill wasn’t much of a self-promoter, which is one reason why he’s little known today. He also encouraged his students to pursue their own research interests, even if those didn’t line up conveniently with his own. As every researcher knows, switching among research topics comes at a productivity cost.

As a researcher, Bill lived in the inductive spirit of Skinner’s “A case history in scientific method,” so he was happy to talk about his miscalculations (in the early days of human operant research, these happened often!). This was part of his persona as a gifted teacher — he wanted everyone, especially students like me in the early 1980s who might join the next generation of investigators, to understand that this unfamiliar thing called human operant research rarely follows a straight path. One of my favorites among his papers is a 1988 Behavior Analyst article, co-authored with Mark Galizio, that shared some of the untold stories of figuring out how to do this kind of work. Among the stories were these two colorful cases:

In one study, Hal Miller and I were interested in assessing preferences for different snack items using a concurrent operant procedure. Female subjects pulled on one of three doors of a modified vending machine for bite-sized snack items (cupcakes, donuts, and coffee cake that had been sliced into tenths, individually wrapped in cellophane, and placed in the vending machine.) Each door operated according to a different variable interval schedule. Upon opening the door the subject removed the snack item, unwrapped it, and ate it. Much to our surprise, subjects did not behave according to our expectations: Subjects overwhelmingly preferred the snack item placed on the leanest schedule. During debriefing each subject informed us that she was concerned about keeping her figure slim — a factor that we should have taken into consideration while selecting subjects for this experiment.

In a different study, Miller and I first gave subjects their choice of food reinforcers to be placed behind the vending machine door: chocolate chips, dried fruit, or mixed nuts. Several days after the experiment had begun, I happened to be looking out of a window in the room immediately adjacent to that in which subjects were running, and noticed tiny dark objects falling from the window next door. A casual inspection of the ground below the window revealed an assortment of “reinforcers” scattered about. [Then] I looked underneath the vending machine, and found a similar array of discarded goodies. I asked each subject if they were consuming their snack items. In each case, they gave me their word that they were!

When I was a student, I was fortunate enough to interact with Bill at conferences, and I found that his approach to science encouraged me to take chances on things that others hadn’t yet tried and to make informative mistakes along the way. However, I believe that Bill’s openness about the uneven progress of his science also led some contemporaries — especially those accustomed to animal methods that had been rigorously standardized over decades — to underestimate his competence. This is unfortunate, because crazy things happen in research all the time. But scientists who are obsessed with looking good don’t usually talk about them.

Building Infrastucture

When Bill came onto the scene in EAB, there was really no comfortable home for human operant work in behavior analysis journals. As Don Hake famously observed, “A common lament is that the basic journals consider human operant research too applied and suggest an applied journal as a more appropriate outlet. Similarly, applied journals consider such work as not applied enough and suggest a basic journal.” In those days, being a “pure” human operant researcher was a lonely quest. One of Bill’s major contributions was to do something about this problem.

In order to create community among the scattered individuals specializing in human research, while still a graduate student Bill helped to found the Experimental Analysis of Human Behavior Special Interest Group of ABAI, through which research problems and ideas could be conveniently discussed with like-minded colleagues. In 1983, the SIG began publishing the Experimental Analysis of Human Behavior Bulletin — still published today — and Bill was its first Editor. Over the years the Bulletin has published some stimulating methodological papers and pilot studies that might not have found a home in other journals.

In 1987, Bill persuaded The Psychological Record (TPR) to publish the first-ever special issue devoted to human operant research, which he edited. At the time that journal was quite eclectic, but for many years after the special issue it was nevertheless more friendly to human operant work than Journal of the Experimental Analysis of Behavior (JEAB). A lot of interesting papers that JEAB rejected, sometimes on questionable grounds in my opinion, found a home in TPR. One of my favorite examples was an early report by the great Alan Baron on using then-novel speech recognition technology to record vocal responses in operant experiments.

Together, the Bulletin and TPR were instrumental in building up a critical mass of work that helped to eventually establish human operant research as an equal partner with animal research in our field’s basic-science wing.

One of the reasons Bill is little known today is that, during his laboratory days, a great deal of his time and effort was siphoned off into service projects like the Bulletin. Although only a handful of other investigators published as much human operant research as he did during roughly 1980 to 1995, had Bill been less civic minded he could have done much more.

A Career Pivot

Another reason Bill is not well remembered is that at mid-career he simply burned out on the laboratory and moved on to other things. I think this happened because of, well, ironically enough given his dissertation topic, the matching law.

On the one hand, reinforcers for doing human operant research were pretty sparse. Steady state human operant research is a slow and exhausting endeavor. Finding and retaining participants can be tough and expensive, and if your participants are college students they are unavailable for large chunks of the year (holidays, summers, and often the starting and ending weeks of each academic term). As a human operant researcher, therefore, you have to work harder to get less accomplished than do people who run animal labs. Like other human researchers of his generation, Bill often got a frosty editorial reception for his work from reviewers who knew only about animal methods. Bill gradually tired of fighting the same battles with critics who, to be honest, sometimes didn’t know what they were talking about where human methods and research questions were concerned.

On the other hand, Bill was a gifted teacher, and in teaching he found a rich schedule of immediate reinforcement. In the end, as do all wise people, he chased his reinforcers, which meant committing fully to teaching and the scholarship of teaching and learning. Bill published many articles on teaching, authored or edited at least 18 books, and for a time helped to stage the National Institute on the Teaching of Psychology. For efforts like these he was much decorated, receiving the Fred S. Keller Behavioral Education Award as well as high-level recognition from Society of the Teaching of Psychology and the American Psychological Foundation. Auburn University, where Bill spent his academic career, was inspired by his contributions to create both a distinguished teaching professorship and a center for the study and promotion of teaching.


Today Bill is happily retired in the mountains of Colorado, where he bikes, hikes, skis, and spends lots of quality time with his family. I have rarely encountered a behavior analyst under 60 years of age who remembers him, but I will be forever grateful that he left the world of basic operant research more friendly to people like me than it was to him. Anyone who has done basic or translational research with humans owes Bill a debt of gratitude. Our careers might not have been possible without his.


As a side note, it’s interesting to contemplate what Bill’s career trajectory says about the sustainability of basic human operant research as it was originally conceived. Ask yourself: Today how many investigators place their primary focus on single-case, steady-state human experiments? I believe it’s not many — perhaps no more than circa 1990. This suggests that although human operant studies enjoy a level of acceptance that didn’t exist during Bill’s research days, challenges of implementation (cost, scheduling constraints, volume of research that can be completed per unit time) remain prominent. I shut down my own lab mainly because of cost: At one point I realized that I had invested over $20,000 in personal funds (mostly for participant payment) that probably should go toward other things, like my kids’ tuition. Perhaps because of such considerations, much of the contemporary energy in human research has shifted from free-operant methods to other ways of conducting studies, such as group-comparison designs employing brief, hypothetical procedures. I am all for grounding a discipline in mixed methods and for finding quicker and easier ways to get data. But different methods should be chosen to scratch different scientific itches, and I’m not sure how confident I am that this is how things have evolved. Though I love to rail about people who reject new methods simply because they violate some cherished tradition, I’m just as concerned that research topics sometimes get chosen because the methods are expedient. Neither is a good way to shape a research agenda, and in the spirit of Bill’s interest in laboratory lore, maybe it is time for a disciplinary conversation about what the experimental analysis of human behavior has become, and should be.

5 Papers to Check Out

Buskist, W. F., Barry, A., Morgan, D., & Rossi, M. (1984). Competitive fixed interval performance in humans: Role of “orienting” instructions. The Psychological Record34, 241-257.

Buskist, W., & Degrandpre, R. J. (1995). Schedule-controlled responding of two persons under a single schedule of reinforcement. The Psychological Record45, 183-205.

Buskist, W. F., Miller, H. L., & Bennett, R. H. (1980). Fixed-interval performance in humans: Sensitivity to temporal parameters when food is the reinforcer. The Psychological Record30, 111-121.

DeGrandpre, R. J., & Buskist, W. F. (1991). Effects of accuracy of instructions on human behavior: Correspondence with reinforcement contingencies matters. The Psychological Record41, 371-384.

Galizio, M., & Buskist, W. (1988). Laboratory lore and research practices in the experimental analysis of human behavior: Selecting reinforcers and arranging contingencies. The Behavior Analyst11(1), 65-69.