Tactics in Brazil


Dear blog reader,

I must admit that this was another introduction that I found somewhat difficult to write. Where to start for someone who’s contribution to behavior science and the area of symbolic language and cognition is not only so substantial (both in Brazil and internationally), but who is also so genuinely decent and nurturing to the academic careers of so many, mine included. While of course knowing of his pioneering work on transfer of functions in particular for many years, I only first met Professor Julio de Rose in person at the beginning of March 2019 when I came to visit him at the Federal University of São Carlos in Brazil for a short research stay abroad toward the end of my PhD. We had exchanged a few emails back and forth beforehand but nothing too substantial. A far far way from home and all things familiar, and in a country the language of which I did not yet speak a single word, safe to say I was quite nervous and feeling a bit out of my depth. However, all of that immediately dissipated upon meeting Julio, who kindly collected me from the bus stop with his equally welcoming and kind wife, Ieda. And from then on all nervousness and uncertainty on my part had well and truly gone. I learned a lot from Julio and the many other Brazilian behavior scientists I met during that trip, many of them current or former students of Julio’s, many now friends and colleagues of mine, and all testaments to his approach to science, research and education. In fact, my experience during that month was so impactful, that I applied for post-doctoral research funding to return after completing my PhD, a research post that I am currently enjoying as I write this short introduction from my sitting room here in São Paulo.

Colin and Julio on St Patricks Day this year at the Federal University of São Carlos (UFSCar), Brazil

In terms of his approach to science and research specifically, something that has always particularly impressed and inspired me about Julio is his willingness and excitement to go wherever the data goes, regardless of what it says, regardless of the approach from which the study was born (e.g., RFT, Sidmanian equivalence, Skinnerian VB, etc.), regardless of whether it undermines or supports something he may have said previously. This is an attitude and excitement for the experimental analysis of behavior that I had also found in my PhD mentors, Dermot and Yvonne Barnes-Holmes, but that seemed to me unfortunately not quite as ubiquitous in the field as it perhaps should or could be. This is one of the many reasons that his contribution seems perfect for the current blog series and for the message that Dermot and I are trying to encourage during our time curating the series. And one of the many reasons his contribution seems particularly fitting as we enter the 50th anniversary of ABAI.

Knowing his modesty, I suspect Julio will disagree with some of my praise of him here, and perhaps even wince a little while reading it. But this humbleness is another reason that we all could learn something from Professor de Rose. Anyway, enough from me! Sit back and enjoy another fantastic addition to the series!


About the author:

Julio C. de Rose is a professor of psychology at the Federal University of São Carlos, Brazil. He received his Ph.D. in experimental psychology at the University of São Paulo in 1981, after which he was a Fulbright scholar at the Eunice Kennedy Shriver Center, in Massachusetts, USA, from 1984 to 1986. Back in Brazil, he was one of the founders of the research network on Behavior, Cognition, and Teaching, which later became the National Institute for Science and Technology for Behavior, Cognition, and Teaching. Dr. de Rose is currently the research director of this Institute. In addition, he has published widely in the main behavioral journals in the field, such as the Journal of the Experimental Analysis of Behavior, the Journal of Applied Behavior Analysis, The Psychological Record, and The Analysis of Verbal Behavior, while also having served on the editorial board of JABA, and currently serving on the editorial board of The Psychological Record.

Tactics in Brazil (clique aqui para a tradução em português deste blog)

Behavior Analysis in Brazil started at the University of São Paulo (USP), in its main campus, at the city of São Paulo. Professor Fred Keller went there in his first visit to Brazil, in 1961. This visit shaped the first generation of Brazilian behavior analysts, led by Professors Carolina Bori and Rodolpho Azzi. The Graduate Program in Experimental Psychology at USP was founded a few years later. The program was not exclusively on behavior analysis and had professors with different approaches, such as ethology, perception, and emotion.

Carolina Bori. Also of note, Carolina was the first woman to be president of the Brazilian Society for the Advancement of Science. The building at the Federal University of São Carlos in which Julio’s lab (among others) is located, is called Edifício Carolina Bori, in her honor.

In the graduate program at USP the students were not required to take a course on statistics. There were no mandatory courses, but virtually all students interested in behavior analysis took Carolina Bori’s course on Tactics of Scientific Research (Sidman, 1960). Many of our professors at the program were brilliant lecturers. However, as far as I recall, they seldom lectured at the graduate courses. Carolina Bori, for instance, had collaborated with Fred Keller in the first course with the Personalized System of Instruction (PSI) and she did not believe that students learn much with lectures. Perhaps some influence of Professor Keller extended also to the professors who were not behavior analysts. As an example, I took a fantastic course on observation of animal behavior with Professor Walter Cunha, who required us to complete a hundred hours observing ants in his laboratory. He was an ethologist, not very fond of behavior analysis, but had a favorable impression of Professor Keller. In his course, students did the lab work at their own pace and Professor Cunha carefully shaped our behavior of observing and reporting, with thoughtful and nearly immediate feedback on our reports.

To understand how profound was the influence of Tactics on Brazilian behavior analysis, it is important to know that it was in the Graduate Program at USP that the first generation of Brazilian behavior analysts shaped the following ones. Carolina’s course on Tactics had a huge influence on how these following generations conceived research. Since many of them later became professors, they would also extend this influence to their students.

But let’s leave USP for a while and move to Brasília, then the new capital of the country, located in central Brazil. The new University of Brasilia attracted scholars of different fields who wanted to make profound changes in Brazilian science and education. The University was open to new methods, and it was for its new Department of Psychology that Fred Keller, Carolina Bori, Rodolpho Azzi and Gil Sherman developed a PSI course on behavior analysis in the early sixties.

It was also at the University of Brasilia that, some years later, I obtained my undergraduate degree in Psychology. I studied there in the dark times of the Brazilian military government, which had suppressed the innovative experiences in Brasilia and had driven out most of the university’s professors. Nevertheless, all the department’s faculty were behavior analysts and the whole undergraduate program was based on behavior analysis. I had the opportunity to do laboratory courses and also to do some research, with my colleague Olavo Galvão, who later would do a post-doc with Professor Sidman. We did research just because we were interested in it: it was not required to obtain our degree. But the laboratory was there, it had equipment which could be used, there was a good supply of white rats, and our teachers encouraged us to do research.

Olavo and I learned to use the electromechanical equipment of the time and did an experiment to investigate behavior under a schedule of reinforcement never studied before: an interlocking schedule in which an FR requirement increased as time elapsed. We did not have any particular hypotheses for our research. We were just curious to see how the bar pressing response would change under our schedule.

Photo by Aleksandr Gusev on Unsplash

However, we had not yet read Tactics, and our Sidmanian approach didn’t go further than our desire to ask “a question to nature”. We believed that the effects of the schedule should manifest only in the rate of bar pressing. And yes, our schedule did affect bar pressing. All our rats stopped pressing the bar mid-session and began to do something else, session after session. We thought the research had failed, even though our results were very consistent: the rate of bar pressing went to zero and all the rats started to emit the same alternative behavior. Since we were not prepared to consider behaviors other than bar pressing, we could not make any sense of our data. After I read Tactics, and with a little more basic knowledge of behavior analysis, I would do better in my graduate research in the Graduate Program at USP, where I went to obtain my master and doctoral degrees, under the supervision of Professor Carolina Bori.

Virtually all graduate students interested in behavior analysis took Carolina’s Tactics course. The main activity in the course was to read Tactics, usually a chapter each week, and then discuss the chapter in the class. Several courses that I took in the graduate program contributed much to my research career, including a remarkable course on stimulus control taught by Professor Maria Amelia Matos. But I believe the one that contributed most has been Carolina’s course on Tactics.

Reading Tactics was essential for my graduate research, and for the remainder of my career. I was still interested in reinforcement schedules with rats and in my master’s research I investigated behavior under Fixed Interval (FI) schedules. And, again, my rats stopped pressing the bar. But by then I had already read and discussed Tactics. I knew that when reinforcement is programmed at fixed intervals and rats don’t respond with a scalloped pattern, it is likely that extraneous variables are interfering. My experiment required a stable FI baseline, with a scalloped pattern. If I had no scallops, this should be due to poor experimental control. Looking at the rats’ behavior I noticed that all of them eventually started to lick a small gap between the bar and the wall of the chamber. At that time, I already knew something about adjunctive behavior. The operant chamber I used was enclosed in a picknick icebox to isolate the rats from extraneous stimuli. Ventilation was provided by a fan located just behind the bar. So, a stream of air flew between the bar and the wall. FI schedules often produce adjunctive behavior, and this was happening with my rats: they were licking this jet of air. To block that air current, I just inserted a shield between the fan and the outer wall of the box. As there was no longer an air current for the rats to lick, they started to press the bar again.

In Tactics and elsewhere, Sidman insists that we can’t take our stimuli and reinforcers for granted. It is possible that animals or humans may respond to features of the stimuli different than those specified by the researchers. And we need also to calibrate our reinforcement variables. Otherwise, our reinforcers may not be strong enough. Without a strong reinforcer, rats possibly won’t show scallops. Sidman, in Tactics, stressed the importance of calibrating the reinforcement variables and noted that many researchers used to calibrate reinforcement variables based on the curvature of the FI scallop, even when their research was not about FI. So, I tested different deprivation schedules and different concentrations of powdered milk for the reinforcer, until I got a pronounced scallop.

But more important than specific advice, Tactics taught me a general approach to research: a tactical approach, in which daily decisions are based on the orderliness of data rather than on rules or predetermined designs. By the way, the orderly data that I obtained along my graduate research led me to a serendipitous finding that made me change the objectives of the research. I can’t say that may graduate research was groundbreaking, but the data were orderly and with some originality (see de Rose, 1986).

I write here something about my research story only to illustrate how profound was Sidman’s influence on the initial generations of Brazilian behavior analysts, those who, in the majority, learned Tactics in Carolina’s graduate course. As proof of this influence, Sidman was invited by Brazilian researchers for a celebration of the 25th anniversary of the publication of Tactics, in 1985. Sidman accepted the invitation and he and his wife Rita made their first trip to Brazil in that year. I don’t know whether there was any celebration of Tactics’ 25thanniversary in the United States, but it seems that the Sidmans enjoyed the Brazilian celebration.

I could not attend this celebration. I had finished my PhD in 1981 and, in ‘82, I had the mind-blowing experience of reading the article that Sidman and Tailby had just published in the Journal of the Experimental Analysis of Behavior (JEAB). This article changed my research interests. I was hesitant to continue with my animal research, but I kept the interest in basic research. But now I wanted to do basic research with a more direct impact on applications. Sidman and Tailby showed me the way. When I had the opportunity to go for the first time to an Association of Behavior Analysis (ABA) Convention in 1983, I introduced myself to Professor Sidman and asked advice about a post-doc. He advised me to do a post-doc with Larry Stoddard at the Shriver Center in the Boston area. Professor Stoddard accepted me for a post-doc in a lab that had been set up and led by Sidman. I got a Fulbright scholarship and was able to spend two years at the Shriver Center. There I had the privilege to interact and work not only with Larry but with researchers such as Bill McIlvane, Bill Dube and Harry Mackay. Murray Sidman then was no longer at the Shriver Center, but was close by, at Northeastern University in Boston, and he used to participate in the lab meetings at Shriver. Professor Sidman told me about the invitation to go to Brazil for the anniversary of Tactics and how happy he was about it.

Photographed at the 3rd International Institute on Verbal Relations, 1989, held at the resort of Aguas de Lindoia, Brazil. Among many distinguished Brazilian and international attendees, is Murray Sidman standing in the left of the photo in red with his hands on the shoulders of Deisy de Souza, seated with Julio to her left and Linda Hayes to her right. Olavo Galvão, mentioned earlier in this blog, is seated in the front row, second from the right (with the big beard).

The change in the frequency of the Sidmans’ trips to Brazil indicates that those trips had reinforcing consequences, strengthening an operant with a high response cost: a trip from the USA to Brazil would last for at least 15 hours, including airport connections. It was not only the Sidmans’ behavior of accepting invitations that was reinforced: the inviting behavior of Brazilians must have been reinforced too, because they made many more invitations. This is easy to understand: the Sidmans accepted the invitations, and their visits were always successful and productive.

A new phase in Professor Sidman’s influence on Brazilian behavior analysis had begun. Before the first visit, Sidman’s influence was already very strong. After the first visit, followed by many others, Professor Sidman had a more direct influence. His earlier influence was mostly based on the Tactics book. Some researchers also knew about Sidman’s research on avoidance behavior and some, led by João Claudio Todorov, also did work on avoidance, inspired by Sidman. Equivalence research was then new and had no direct connections with themes that then interested Brazilian behavior analysts. Few knew about it. Also, little known was Sidman’s research on stimulus control and his concept of stimulus control topographies. In one of his first visits, Sidman taught a course on stimulus control at USP (1987) which was a landmark. I was already back in Brazil and could attend that course together with other Brazilian students and some colleagues.

Sidman’s influence on behavior analysis in Brazil then became somewhat similar to that of Fred Keller in earlier days. I am not sure exactly how many times Rita and Murray Sidman went to Brazil, but I would say that they went there every one or two years until the long trip became too heavy for them. They visited the main behavior analysis centers in Brazil at that time. Many behavior analysts became interested in research on stimulus equivalence, and a few went to the Boston area for post-doctoral or doctoral work with Sidman or at the Shriver Center.

The Sarasota Symposium, 2014. Murray Sidman, 90 years old at the time, can be seen in front on the right beside João Claudio Todorov. Standing in the back from left to right are Julio, Deisy de Souza, and Caio Miguel.

Stimulus equivalence attracted many Brazilian researchers and became the main focus of a research network in whose creation I participated, together with several Brazilian colleagues and with the help of Bill McIlvane. This network evolved into what is now the Brazilian National Institute for Behavior, Cognition, and Teaching, chaired by my colleague Deisy de Souza. The Institute has researchers from 8 different Brazilian Universities and several international collaborators. The members of the Institute have published many articles in international journals and also in Brazilian ones. Other Brazilian researchers who are not part of the Institute have also contributed with important work in the area.

The central role of stimulus equivalence in Brazilian behavior analysis, both in the Institute and further afield, shows that Sidman’s influence remains strong. However, I have the impression that Sidman’s current influence is based almost solely on equivalence. As far as I know, graduate students in Brazil no longer read Tactics as they did in Carolina’s time. Of course, science must move forward, but I believe that, in Brazil (and probably everywhere), we would have much to gain by bringing Tactics back to the training of researchers in behavior analysis. In making this case, it should be appreciated that the book was written more than 60 years ago and perhaps some of the issues it raised may not be as relevant now as they were at that time. So the book should certainly not be taken dogmatically. However, the foundational lesson that the behavior of a researcher should not be governed by rules, but rather shaped by the collection of orderly data (a point that had already been stressed by Skinner) may be even more relevant today.


de Rose, J. C. C. (1986). Behavioral contrast in fixed-interval components: Effects of extinction-component duration. Journal of the Experimental Analysis of Behavior, 45, 175-188.

Sidman, M. (1960). Tactics of Scientific Research: Evaluating Experimental Data in Psychology. Basic Books, Inc.

Sidman, M. & Tailby, W. (1982). Conditional discrimination vs. matching to sample: An expansion of the testing paradigm. Journal of the Experimental Analysis of Behavior, 17(1), 5-22.