Guest post authored by Shamim Samipour, Middle East Technical University
Lewis Carroll’s (1876) The Hunting of the Snark, follows a hunting party’s quest to catch an elusive, but tasty, snark that instead turns out to be an eldritch boojum. Frank Beach borrowed this image to issue his warning that pursuing what he regarded as the wrong quarry can doom the scientist (as it did the baker in Carroll’s poem) to softly fade away.
Animal learning research once proceeded under a broadly general‑process assumption. Many researchers treated instances of learning as fundamentally similar across species and contexts (a precedent set by Pavlov). This optimism was challenged by work on selective associations and “constraints,” which showed that learning is specific to particular species and motivational situations (e.g., taste–illness learning). As these findings accumulated, they undermined faith in simple, one‑size‑fits‑all accounts of learning. A common reaction was to invoke “biological constraints.”. However, this literature remained largely atheoretical, cataloguing exceptions without offering a positive framework to explain them.
Bowers (2022) revisits the mid‑twentieth‑century ‘biological constraints on learning’ movement and argues that the real problem that caused it to founder was a deep neglect of theory. He uses a broken‑clock analogy to show that merely labeling phenomena as exceptions tells us that something is wrong, but not how to tell the correct time or how the clock works.
Michael Domjan’s work is central to understanding this crisis. He and Bennett Galef (1983) diagnosed the problem not as a clash of biology versus behaviorism, but as an absence of theory to guide methods. Although they endorsed aligning psychology with biology, they criticized much constraints research for superficial evolutionary talk and post hoc appeals to “biological” explanations. Instead, Domjan recommended genuinely biological approaches, especially the comparative method as practiced in biology, and later, behavior systems theory. Bowers reframes this as a challenge to make the field theoretical: to articulate principled bases for connecting questions, methods, and interpretations. This post is a rephrasing of the story presented in Bowers (2022).
Behavior systems approaches begin not with an abstract capacity called “learning,” but with an analysis of the structure of an animal’s behavior in a given context. Learning is change in behavior, and Timberlake (1983) urged researchers to first understand and model the pre‑learning organization of behavior. In a behavior system, actions are organized in clusters that solve recurrent problems such as feeding, reproduction, and predator avoidance. Predatory behavior in rats, for example, can be divided into general search (putting oneself where prey is likely), focal search (locating and tracking prey), and consummatory behaviors (capturing, handling, killing). A comparable three‑phase architecture—general search, focal search, and consummatory/imminent components—also appears in defensive behavior and sexual behavior. Conditioning influences these systems, but it does so by modulating the relations among their components, rather than by attaching responses to stimuli in a simple, decontextualized way.
This theriomorphic, or animal‑centered, perspective has several implications. It shifts attention away from anthropomorphic abstractions and toward the structure of the organism’s behavioral repertoire. It also calls into question some familiar assumptions from standard conditioning paradigms, such as how sharply one can distinguish discrete stimuli and responses in practice, and where to look for learnt changes. A behavior systems approach draws attention to changes in how animals respond to unconditional stimuli—changes that can increase the effectiveness of those stimuli—as part of learning. More generally, behavior systems theory offers a way to connect evolutionary considerations with specificities of learning: instead of talking vaguely about selection for “learning” or “feeding,” the theory encourages more precise statements about the structure of particular systems and interactions.
Behavior systems theory complements the comparative method. A behavior systems model for a given species in a given context functions somewhat like a periodic table for that organism’s behavior, laying out relations among components. Biology, however, has many such “tables,” because each kind of animal has its own structured ways of behaving. Once multiple behavior systems models exist—for example, for reproduction in different fish species or for predator behavior in different rodents—comparative methods can be used to study how these structures evolved, what generalities hold across systems, and what may be basic to motivation and cognition. In this way, these two approaches together address Domjan’s challenge by grounding learning phenomena in both evolutionary history and system‑level organization.
This raises issues concerning education. What theories students encounter strongly influences the views they adopt, and light training in a field with emphasis on empirical “highlights” without theoretical context risks producing an atheoretical generation of psychologists. Introductory presentations and textbooks often misrepresent animal learning, a pattern linked to hyper‑empirical habits and the pretense that data speak for themselves. The value of a framework like behavior systems theory lies less in its metaphors (for example, animals as “control systems”) than in how it directs research toward meaningful structures and relationships.
The peculiarities of learning that once seemed to threaten general-process perspectives in psychology can, when approached with sound theory, become a “prize snark”. The ‘taming’ advocated in ”Taming the boojum” is not about eliminating anomalies, but about keeping focus on making sense of the animal, rather than chasing exotic phenomena. In this view, the ongoing challenge for psychology is not merely to be more biological in language, but to be genuinely theoretical—whether biological or otherwise–in how it studies learning.
References:
Bowers, R. I. (2022). Taming the boojum: Being theoretical about peculiarities of learning. Learning & Behavior, 50(3), 433-440. https://doi.org/10.3758/s13420-022-00535-3
Domjan, M., & Galef, B. G. (1983). Biological constraints on instrumental and classical conditioning: Retrospect and prospect. Animal Learning & Behavior, 11, 151–161.
Timberlake, W. (1983). The functional organization of appetitive behavior: Behavior systems and learning. In M. Zeiler & P. Harzem (Eds.), Advances in analysis of behaviour (Vol. 3, pp. 177–221). John Wiley & Sons.