Adjunctive Behavior

Light Versus Gravity Views of Reinforcement

Photo by Zac Ong on Unsplash

I was reading the Fixed Interval by Fred, an excellent read that comes out every Sunday on Substack. The topic was ‘scheduled induced behavior’ aka ‘adjunctive behavior’. Afterward, I thought I didn’t know enough about the topic so I update myself on recent research. I came across an article by Peter Killeen and R. Pellon (2013) who propose that adjunctive behavior is not scheduled induced as previously thought but rather an operant (behavior that is affected by consequences) where there is no contingency involved only proximity. In my mind this is a big deal. Am I misunderstanding the issues? I admit the experimental analysis of behavior (EAB) is not my area of expertise, I’m more of an applied (ABA) guy. But I gave it my best shot. I know what I’m saying now may not be clear but read on to discover what I found out.

Historical Foundations: Falk and the Origins of Adjunctive Behavior

The term “adjunctive behavior” was coined by John Falk in the early 1960s. In his seminal study, Falk (1961) observed that rats working for food pellets under a fixed time (FT) 1-minute schedule engaged in excessive drinking—even though they were not water deprived. Water intake was approximately 50% higher during the sessions than in their home cages with unlimited access to water. Falk called this phenomenon “schedule-induced polydipsia” (SIP).

Over the next decade, Falk expanded the concept, describing adjunctive behavior as activity that emerges during the post-reinforcement pause—the time between receiving one reinforcer and the availability of the next—when another reinforcer is being earned on a schedule (Falk, 1971). Importantly, adjunctive behaviors:

1. Occur at high probability during predictable post-reinforcement periods.

2. Are influenced by the level of deprivation for the primary reinforcer.

3. Can be maintained as operants if access to them is made contingent on some behavior.

4. Are often “excessive” relative to the organism’s baseline level.

Falk and others demonstrated schedule-induced phenomena across many behaviors, including drinking, aggression, pica, stereotypy, drug-taking, self-imposed timeouts, and preening. These findings suggested that schedules of reinforcement do not merely shape the target operant (e.g., lever pressing) but can also indirectly shape other, sometimes maladaptive, behaviors.

Debate and Skepticism: Overskeid’s Challenge

In the early 1990s, G. Overskeid (1992) challenged the assumption that human behavior could be truly schedule-induced. He argued that the evidence for human adjunctive behavior was thin, often based on small, uncontrolled studies, and that alternative explanations—such as social contingencies, instructional control, or simple operant reinforcement—were equally plausible. Overskeid’s skepticism forced researchers to clarify what would count as true adjunctive behavior in humans and to test whether the same temporal patterns observed in animals could be reliably reproduced.

Falk responded, both directly and indirectly, by noting that adjunctive patterns in humans are most visible when the environmental context is carefully arranged to parallel animal studies—fixed schedules, clear deprivation states, and minimal competing reinforcers. Under such conditions, one can observe “adjunctive-like” temporal distributions, even in people (Falk, 1996).

Adjunctive Behavior as a Model for Human Problems: The Colotla & Keehn Study

One of the more interesting studies I read came from Colotla and Keehn (1981), who used SIP as a laboratory model for alcoholism. They arranged fixed-time schedules for delivering food to rats, but replaced water with ethanol solutions of varying concentrations. The rats developed excessive, schedule-induced alcohol consumption, especially at certain fixed intervals and ethanol strengths. This study suggested that the temporal structure of reinforcement schedules could help explain certain patterns of human substance use—particularly binge-like episodes—without relying solely on social or cognitive explanations.

---

Human Evidence: What We Know and What We Don’t

Despite strong animal evidence, the human literature remains sparse. Foster (1978) lamented that adjunctive behavior was “under-reported” in applied behavior analysis, a reflection of the gap between basic and applied research. Only a handful of human studies exist, here are three:

• Kachanoff et al. (1973) studied schizophrenic patients under fixed-interval token delivery and observed increased pacing and water consumption—patterns similar to animal SIP.

• Wieseler et al. (1988) found that delivering edible reinforcers on fixed-interval schedules increased stereotypy in developmentally disabled adults.

• Lerman et al. (1994) explored whether severe self-injury could be adjunctive but found no consistent temporal patterns; however, stereotypy remained a possibility.

These studies suggest that adjunctive behavior can occur in humans, but typically in highly controlled settings and with clear deprivation or motivational conditions in place.

---

Schedule Parameters and Rich-to-Lean Transitions

Later research expanded beyond fixed-time and fixed-interval schedules to consider how shifts in reinforcement conditions might trigger adjunctive-like responses. Kupfer, Allen, and Malagodi (2008) examined “Rich-to-Lean” transitions—moving from conditions with dense reinforcement to those with sparse reinforcement—and found these could induce aggression in pigeons. Similar schedule dynamics have been documented in humans, such as increased irritability or pacing during abrupt changes in work-to-reward ratios.

---

A Modern Reframing: Killeen & Pellón’s Operant Competition Model

The Killeen & Pellón (2013) article argues that so-called “adjunctive behaviors” such as schedule-induced polydipsia should not be treated as by-products outside operant conditioning. Traditionally, adjunctive behavior was thought to occur in the pauses between reinforced responses, be unrelated to reinforcement contingencies, and therefore fall outside the operant framework.

Killeen & Pellón review evidence that challenges two core assumptions of that view:

1. Contingency is necessary for conditioning – they show that temporal proximity can be just as important as contingency.

2. Delay-of-reinforcement gradients are uniformly steep – in fact, different behaviors have different “time courses” in relation to reinforcement.

They propose that many adjunctive behaviors are actually operants shaped by the temporal structure of the schedule. Using polydipsia as an example, they show how “slow-course” behaviors (e.g., drinking) dominate early in the interval when the probability of reinforcement is low, while “fast-course” behaviors (e.g., orienting to the food source) dominate near reinforcement delivery. This pattern results from competition among behaviors for different temporal “niches” within each inter-reinforcement interval.

They also address competing explanations:

• Against adventitious reinforcement – the consistent waxing/waning patterns across intervals can’t be explained by chance pairings alone.

• Against purely elicited accounts – only behaviors that can be operantly reinforced appear adjunctively, suggesting reinforcement processes are involved.

Practically, they predict that schedules with irregular or very short inter-reinforcement intervals reduce temporal space for slow-course behaviors, thus reducing adjunctive responding. This reframing integrates adjunctive behaviors into operant theory and supports a more general conditioning framework based on proximity and behavior-specific delay gradients, rather than a strict contingency requirement.

Light versus Gravity Views of Reinforcement

In promoting the causal notion of proximity Killeen & Pellon discuss two theories of reinforcement: light versus gravity.

“An implicit US (unconditioned stimulus) or reinforcer can affect only one thing—condition one CS (conditioned stimulus) or strengthen one response. We call this the light model: If a sunbeam falls on one object, then other objects behind it fade to umbral. Gravity has a different modus operandi; the sun that throws the beam also attracts the object, and it attracts objects behind it equally. There are no gravity shadows. Is the force of a reinforcer more like that of light or gravity?”

They picked gravity. This perspective was new to me.

They write:

“Notwithstanding, the case for adjuncts as maintained by reinforcement is parsimonious of mechanism, productive of models, and superior to alternative explanations of these important phenomena. In outline, the logic seems to us plausible: Proximity between events is manifestly important. If reinforcers can work absent contingency, the primary role of contingency is to arrange proximity. Reinforcers can work absent contingency. Therefore, contingency works, when it does, by arranging proximity. Proximity may extend over dozens of seconds. Reinforceable behaviors that are proximate to reinforcers may be increased by reinforcement. Adjunctive behaviors are proximate to reinforcers and are reinforceable. Therefore, adjunctive behaviors occur at high rates because they are reinforced. Different classes of responses may have different associabilities with reinforcers over delays. Therefore, different classes of responses may emerge at different times before reinforcement and compete for expression. Those that are expressed are reinforced. This generates an unstable dynamic system. Signals of reinforcement may favor perceptual and motor responses with steeper delay gradients over those with shallower gradients and may strengthen already-dominant ones over novel ones, leading to many of the phenomena of classical conditioning. This article constitutes an extensive grounding of these arguments, an adduction of evidence for their premises, and a validation of their conclusions.”

Summary

According to Killeen & Pellon (2013) adjunctive behaviors should be understood as operants competing for time within the inter-reinforcement cycle, not as a third, separate class of behavior. This perspective clarifies when such behaviors will appear and how they may be modified by changes in schedule structure, deprivation, and reinforcer value. They argue that these behaviors are operants shaped by the temporal structure of the reinforcement schedule.

They propose that different behaviors have different “delay gradients”—curves that describe how sensitive a response is to the time between the behavior and reinforcement. Some behaviors are “fast-course,” peaking close to reinforcement (e.g., approaching the food source), while others are “slow-course,” peaking earlier in the interval (e.g., drinking in SIP experiments). Under this view, adjunctive behavior is simply what emerges in the “temporal niche” left open by the schedule. As the next reinforcer approaches, fast-course behaviors crowd out slow-course ones; earlier in the interval, slow-course behaviors dominate.

This reframing unifies adjunctive behavior and operant behavior under the same principles of reinforcement, proximity, and competition for time. It also predicts that making the schedule more variable—shortening some inter-reinforcement intervals—will squeeze out slow-course adjuncts by leaving less time for them to occur.

Adjunctive behavior research is not usually read by applied behavior analysts, even though the research investigates the processes involved in learning. Are we moving from ‘contingency shaped behavior’ to proximity shaped behavior? I don’t think practitioners are going to change their verbal behavior anytime soon, but it is important to understand the role of schedules and consequences on our behavior.

---

References
Colotla, V. A., & Keehn, J. D. (1981). Schedule-induced alcohol drinking: Effects of fixed-time intervals and concentration of ethanol. Pharmacology Biochemistry and Behavior, 14(2), 217–222. https://doi.org/10.1016/0091-3057(81)90221-7

Falk, J. L. (1961). Production of polydipsia in normal rats by an intermittent food schedule. Science, 133(3447), 195–196. https://doi.org/10.1126/science.133.3447.195

Falk, J. L. (1971). The nature and determinants of adjunctive behavior. Physiology & Behavior, 6(5), 577–588. https://doi.org/10.1016/0031-9384(71)90223-2

Falk, J. L. (1996). On creation: A meditation from a behavioral perspective. Behaviorology, 4(1), 3–29.

Foster, W. S. (1978). Adjunctive behavior: An under-reported phenomenon in applied behavior analysis? Journal of Applied Behavior Analysis, 11(4), 545–546. https://doi.org/10.1901/jaba.1978.11-545

Kachanoff, R., Leveille, R., McLelland, J. P., & Wayner, M. J. (1973). Schedule-induced behavior in humans. Physiology & Behavior, 11(3), 395–398. https://doi.org/10.1016/0031-9384(73)90272-3

Killeen, P. R., & Pellón, R. (2013). Adjunctive behaviors are operants. Learning & Behavior, 41(1), 1–24. https://doi.org/10.3758/s13420-012-0095-1

Kupfer, A. S., Allen, R., & Malagodi, E. F. (2008). Induced attack during fixed-ratio and matched-time schedules of food presentation. Journal of the Experimental Analysis of Behavior, 89(1), 31–48. https://doi.org/10.1901/jeab.2008.89-31

Lerman, D. C., Iwata, B. A., Zarcone, J. R., & Ringdahl, J. (1994). Assessment of stereotypic and self-injurious behaviors as adjunctive responses. Journal of Applied Behavior Analysis, 27(4), 715–728. https://doi.org/10.1901/jaba.1994.27-715

Overskeid, G. (1992). Is any human behavior schedule induced? The Psychological Record, 42(3), 379–387. https://doi.org/10.1007/BF03399506

Perone, M., & Courtney, K. (1992). Fixed-ratio pausing: Joint effects of past reinforcer magnitude and stimuli correlated with upcoming magnitude. Journal of the Experimental Analysis of Behavior, 57(1), 33–46. https://doi.org/10.1901/jeab.1992.57-33

Wieseler, N. A., Hanson, R. H., Chamberlain, T. P., & Thompson, T. (1988). Stereotypic behavior of mentally retarded adults adjunctive to a positive reinforcement schedule. Research in Developmental Disabilities, 9(4), 393–403. https://doi.org/10.1016/0891-4222(88)90009-1

---