Associative concept learning, stimulus equivalence, and relational frame theory: working out the similarities and differences between human and nonhuman behavior.
Animal studies look like human equivalence, but we still need proof that the two processes work the same way.
01Research in Context
What this study did
Hughes et al. (2014) wrote a narrative review. They compared three lines of work: animal associative learning, human stimulus equivalence, and Relational Frame Theory.
The authors asked: Do animals show the same derived relations we see in people? They pulled together cross-species lab data to find out.
What they found
The team concluded that pigeon and monkey data look like human equivalence, but gaps remain. We still lack proof that the two kinds of learning serve the same function.
They list open questions. For example, does an animal's emergent bidirectional respond really work like a child's new word?
How this fits with other research
Zentall (2025) extends the debate. He shows a pigeon model that meets equivalence tests (reflexivity, symmetry, transitivity). The bird data add muscle to Sean's idea that equivalence can live without language.
Barnes-Holmes et al. (2018) act as a successor. They use the same equivalence/RFT base but aim it squarely at human language. Their review treats derived relations as the core of verbal behavior, while Sean left the cross-species link unsettled.
Tovar et al. (2023) fill another gap Sean noted. They catalog computer models that build equivalence classes in code. No animals or humans required. The models give researchers a new tool to test what variables matter.
Why it matters
If you teach stimulus equivalence to clients, know the animal work backs your procedure. It shows emergent relations are not just a language trick. Yet Sean warns us not to assume pigeon data equal child data. Use Zentall's pigeon criteria as a benchmark, but keep testing for human-only features like analogy. And if you like tech, try the free software in Tovar's list to preview which stimulus sets might form classes fastest.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick one equivalence set you plan to run; check that it meets reflexivity, symmetry, and transitivity tests like Zentall's pigeon model.
02At a glance
03Original abstract
In their review article Zentall et al. propose that nonhumans can come to relate stimuli based on their physical properties (perceptual concept learning) or the relationship established between or among physically related stimuli (relational concept learning). At the same time, they draw upon findings from within the animal learning literature in order to argue that nonhumans can also derive untrained yet predictable relations between stimuli in the absence of direct training (associative concept learning). We are both intrigued and excited by the body of work contained in this paper and believe that it may accelerate our understanding of animal as well as human cognition in several ways. Nevertheless, a number of important questions still need to be addressed before we can conclude that associative concept learning in nonhumans is functionally similar to that observed in our own species.
Journal of the experimental analysis of behavior, 2014 · doi:10.1002/jeab.60