In Memoriam: Brian A. Iwata: A Mentor by Proxy

I was deeply saddened when I learned of Brian Iwata's passing, but his death also set the occasion for me to think about and appreciate how lucky I was to have known him and to have learned from him. I never worked directly with or studied under Brian Iwata, though I viewed him as a mentor, and he had a profound influence on my career. I don't know for certain, but I suspect that many behavior analysts who use functional analysis methods to study and treat severe problem behavior also viewed Brian this way.

My first interaction with Brian was in the mid-1980s, after I left the Medical Center Rehabilitation Hospital at the University of North Dakota to become the Director of Psychology at the State Developmental Center at Grafton in Grafton, North Dakota. The developmental center was under a court order from a class-action lawsuit (Association for Retarded Citizens of North Dakota v. Olson, 1982/1983) for the expressed purpose of ensuring that the residents and clients of the center, “may realize the rights to which they are entitled; obtain needed services; investigate complaints, abuse, and neglect; and remove barriers to identified needs.” One of my responsibilities in my new position was to ensure that the residents who displayed severe problem behavior (e.g., aggression, self-injurious behavior [SIB]) received active and appropriate treatment. As I attempted to fulfill that responsibility, I quickly realized that I was in way over my head. Too many of the residents exhibited highly dangerous behavior (e.g., resulting in broken bones or permanent scar tissue), and far too often, the treatments involved chemical or physical restraint (e.g., high doses of haloperidol, five-point restraint).

To better address these challenges, I went to the University of North Dakota library and conducted a literature search on aggression and SIB (because in those days, few people had home computers and none of them connected to the internet). Through the search, I found an article by Brian and his colleagues titled, “Toward a functional analysis of self-injury” (Iwata et al., 1982/1994). It may sound hyperbolic (or alternatively, timeworn) for someone to say that a given event changed the course of one's life, but reading that article certainly changed my career path. Until that point, I thought that I practiced behavior analysis when managing problem behavior, but I did not. I practiced behavior modification—I introduced recycled antecedents and consequences that were effective with prior clients without analyzing and understanding the ones that maintained the current client's problem behavior. And if the first set of antecedents and consequences I introduced didn't work, I would move on to the next set without knowing why the first set failed or whether the next set would fare any better. It was akin to trying to unlock a door by cycling through the keys on a keyring until finding the one that worked; only in this case, the client suffered as I fumbled to find the right key.

Shortly after reading Brian's seminal article on functional analysis, we stumbled through our first attempt at a functional analysis with a young man who displayed severe SIB. The analysis didn't identify a specific function, but it did show that SIB could be reduced to near-zero levels in the toy-play condition. Such results are commonplace now, but they were eye opening at the time. After that, I contacted Brian and asked him to come to our developmental center to present on functional analysis and to consult on our most recalcitrant cases of SIB. He did, and during that visit I realized how much more I needed to learn about behavior in general and SIB in specific, and I wasn't going to acquire what I needed to learn in the position I held at the developmental center.

So, almost five years after obtaining my PhD, I applied for a postdoctoral fellowship at the John F. Kennedy Institute (now the Kennedy Krieger Institute), where Brian directed the SIB Unit, and at the Johns Hopkins University School of Medicine, where Brian was a faculty member. My plan was to study under Brian, but unfortunately for me (and good for him), he was offered and accepted a faculty position at the University of Florida in 1986. Despite his leaving the John F. Kennedy Institute just before I arrived, I still learned an amazing amount about behavior from Brian, by proxy. He was meticulous about collecting, organizing, and analyzing the behavioral data collected on patients of the SIB Unit, and all the behavioral data for each patient treated on that unit was kept in a blue, three-ringed binder called the patient's blue book. For the first few months of my fellowship, I pulled one of the patients' blue books every day or two and studied it. From those blue books, I learned a great deal about single-case designs and how to graph and interpret data. I learned how treatments like extinction and differential reinforcement differed depending on the function of SIB. I learned about demand fading and restraint fading and other stimulus-control procedures that could be used to reduce severe problem behavior. I learned all these things before Brian and his colleagues published the data contained in the blue books in the Journal of Applied Behavior Analysis (JABA) and other journals.

I also learned a great deal from Brian by observing his vocal behavior at meetings and conferences and his textual behavior in his publications and in his comments as a reviewer or editor on papers submitted by others. When I worked on my first several papers submitted to JABA, I used Brian's prior publications in JABA as models. Similarly, I used him as a model when composing my presentations or when I served as a discussant at conference symposia. Finally, while on the faculty of the Kennedy Krieger Institute, I recruited Brian's former students (e.g., Dorothea Lerman, Iser DeLeon), and learned from them and, again by proxy, from Brian.

The seminal work of Brian and his colleagues not only affected my work; it fundamentally changed our field. Prior to the publication of Iwata et al. (1982/1994), the assessment and nosological categorization of problem behavior was almost exclusively based on its structural or topographical characteristics and on the extent to which certain responses co-occurred (e.g., the combination of inattention, impulsivity, and hyperactivity led to the diagnosis of attention-deficit/hyperactivity disorder). In his pioneering research, Brian provided us with an alternative method of assessing and characterizing problem behavior, one in which behavior is evaluated and categorized in terms of its functional properties. This approach has fundamentally changed how we assess and treat problem behavior because it focuses on why the behavior occurs, which, in turn, provides us with a roadmap for intervention. Other fields of science, such as microbiology, have long understood the importance of analyzing both the structure and function of dynamic entities. Thanks to Brian's landmark work, over the past 41 years, behavior analysts have increasingly assessed and categorized problem behavior according to both its structure and its function.

The functional analysis procedures developed by Brian and his colleagues have emerged as the predominant method of prescribing effective behavioral treatments for persons with autism and developmental disabilities who display severe problem behavior (Horner, 1994; Melanson & Fahmie, 2023; Repp, 1994), and its use has expanded markedly for other conditions as well (e.g., Chapman et al., 1993; Northup et al., 1995; Piazza et al., 1996, 1997). Functional analysis provides an empirical method of selecting highly effective behavioral interventions that address the specific environmental variables that occasion and maintain the problem behavior. Many investigations have directly compared behavioral interventions that were and were not based on a functional analysis, and the results have consistently favored the functional analysis method (e.g., Kuhn et al., 1999; Repp et al., 1988; Smith et al., 1993). In addition, results of randomized clinical trials, large-scale meta-analyses, systematic reviews, and consecutive-controlled case series indicate that behavioral treatments tend to be more effective than pharmacological interventions for individuals with autism and developmental disabilities who display severe problem behavior and that behavioral treatments were more effective when they were based on a prior functional analysis (e.g., Didden et al., 1997; Greer et al., 2016; Hassiotis et al., 2009; Iwata, Pace, Dorsey, et al., 1994; Lindgren et al., 2020; Richman et al., 2015).

In addition to its relevance to treatment, there are at least three additional advancements that have resulted from Brian's development of the functional analysis method. First, functional analysis represents a valuable research tool with which to better understand environmental influences on severe behavior disorders. For example, epidemiological investigations using the functional analysis method have largely supported the operant hypotheses about SIB described by Carr (1977, e.g., Derby et al., 1992, Iwata, Pace, Dorsey, et al., 1994). Second, the functional analysis method has resulted in the refinement of existing behavioral interventions. For example, research has shown that extinction procedures differ dramatically depending on the function of the problem behavior (Iwata, Pace, Cowdery, et al., 1994; Kuhn et al., 1999). Third, functional analysis research has led to the development of a variety of innovative and effective interventions not previously available that were uniquely designed to address problem behavior maintained by one or more specific behavioral functions (Carr & Durand, 1985; Horner et al., 1997; Iwata et al., 1990; Owen et al., 2020; Richman et al., 1998; Smith et al., 1993; Tiger et al., 2009; Vollmer et al., 1993). Brian's investigations have been at the forefront of each of these research advancements.

In addition to his landmark research, Brian has been the preeminent mentor in the field of applied behavior analysis for the past 40 years. His current and former students have produced (and continue to produce) some of the best research in our field. Brian has directly mentored or served as a role model and advisor to nine of the last 10 editors and at least 20 associate editors for the JABA. He instilled in all his students a strong zeal to excel and a stalwart commitment to advance the field of behavior analysis through clinical excellence, innovation, and systematic research.

Due to Brian's pioneering work, Skinner's vision of a thoroughgoing functional analysis has been increasingly realized over the past 41 years.

At the start of this century, the American Psychological Association recognized B. F. Skinner as the most influential psychologist of the 20th century. Over the last half century, Brian Iwata has been the most influential applied behavior analyst in the world. He directly mentored many of the leaders in our field, and he served as a role model and a mentor by proxy to many others, like me.