Beyond Pavlov's Perfect Student

A recent piece co-authored by Michael Weingarth and me called There Is No Such Thing As “The Science of Learning” has garnered a tremendous response on social media. For those who read, considered, and engaged in its ideas in good faith, we thank you. However, we did feel the need to mount a bit of a response to our loudest critics and elaborate and expand on ideas that didn’t make it into our original piece. One critique even went so far as to blame us and the ideas in the piece for “the educational shitshow we’re in now.” (Not actually having read the piece or considered its ideas appears to be a fairly common theme among the most dismissive responses, so I suppose this may be lost on them as well.) 

While it is not a reality some of our critics want to engage with, it is a true fact that “Cognition is not simply a brain event.” We also understand that schooling doesn't happen to brains separate from bodies, culture, or purpose; separate from the full experience of being human. If cognition is not simply a brain event and schooling happens to embodied humans, it's just not that useful to draw an authoritative box around a handful of practices from cognitive science and call it "The Science of Learning." That's not how learning, or cognition, or schooling – or science for that matter – works. 

This isn’t to suggest that there may not one day be a comprehensive science of learning – a science that investigates how and why humans learn in ways both broad and specific. This also isn’t to debate the “evidence” that says certain practices have improved outcomes. The issue is scale, scope, criticality, operative definitions, the siloing of academic and scientific research, and generalizations – the issue is that until teachers and researchers better understand the science of learning in real-time complex environments like school, they can’t critically engage with or meaningfully implement “The Science of Learning.” 

Acknowledging the existence of scientific evidence in educational practices is crucial. However, the crux of the matter lies in understanding the nuanced limitations and contextual applications of this evidence. It's not that the science itself is flawed or unreal, but rather that its critical understanding and practical implementation presents very real and serious problems. Availability bias, where what is known becomes what is done when facing new problems or contexts, can crowd out educators’ ability to conceive of learners beyond the limits of what “The Science” says, but worse: can paint a student’s very real struggles as some kind of abnormal response to evidence-based practice and therefore likely behavioral or attentional in nature. The varied and dynamic nature of learning environments – too many variables to isolate one out, the way norm-referencing is leveraged to discount outliers, and the lack of applicable research on neurodiverse students – necessitates a more flexible and holistic approach. 

And that brings us to Pavlov, Skinner, and Ebbinghaus. As one critical response read, “The idea that there can’t be a science of learning because we can’t know precisely what is going on in the brain is laughable. Someone tell Pavlov, Skinner, Ebbinghaus, etc.” 

Most textbook diagrams depicting Pavlov’s experiments in classical conditioning don’t show the subjects themselves, preferring instead to use sanitized graphics and cartoon dogs to demonstrate the relationship between stimulus and response. The reality is far more grim, notes Michael Specter, “The dogs may have been irreplaceable, but their treatment would undoubtedly cause an outcry today.”

Pavlov would remove a dog’s esophagus and create an opening, a fistula, in the animal’s throat, so that, no matter how much the dog ate, the food would fall out and never make it to the stomach. By creating additional fistulas along the digestive system and collecting the various secretions, he could measure their quantity and chemical properties in great detail. 

Behold Pavlov’s perfect student. 

‍

BF Skinner’s subjects, on the other hand, were pigeons, rats, and even his own daughter. But Skinner was not content to keep the implications of his new “radical behaviorism” confined to the lab. In his fictional Walden Two he imagined a utopian society entirely guided by the principles of operant conditioning:

“Now that we know how positive reinforcement works and why negative doesn’t,” [Frazier] said at last, "we can be more deliberate, and hence more successful, in our cultural design. We can achieve a sort of control under which the controlled, though they are following a code much more scrupulously than was ever the case under the old system, nevertheless feel free. They are doing what they want to do, not what they are forced to do. That's the source of the tremendous power of positive reinforcement - there's no restraint and no revolt. By a careful cultural design, we control not the final behavior, but the inclination to behave - the motives, the desires, the wishes. The curious thing is that in that case the question of freedom never arises."

(Later in the same dialogue, Frazier, the founder of the utopian community of Walden Two, remarks, “Democracy is the spawn of despotism.” I wonder about the extent to which Skinner himself separated the science of behaviorism and its implementation as the ideal way to organize human societies.)

And in Ebbinghaus’s lab, studies of memory were conducted (on himself, no less) with lists of three-letter nonsense words to deliberately untangle retention from emotion, interest, and prior knowledge. Of this approach, psycholinguist Frank Smith, wrote, "This was Ebbinghaus's world-changing revelation: If you want to study how people learn without the involvement of interest and past experience — study how they learn nonsense."

To add necessary context is not to say these men are evil and bad and wrong, rather that the proper context is exactly what we need to correctly view these ideas – and their limits – in their implementation in schools. Our students are not surgically modified dogs nor are they pigeons in operant conditioning chambers attempting to learn nonsense words. No child enters a classroom devoid of emotion, interest, or prior knowledge. Owing to the key distinctions between the controlled laboratory and the living classroom, there simply may be no connection between what is taught and what is learned; or between the educational intervention and the desired outcome. This is why, in pedagogies centered on instruction drawn from the narrow view of “The Science of Learning,” behaviorism is a complexity control meant to reduce the number of possible variables between instruction and assessment; to better reproduce the uncomplicated relationship between variables in the Skinner Box. We know from listening to students themselves that there has been a persistent crisis in schools, even before COVID: students ask fewer questions the longer they remain in school, engagement plummets alongside mental health, and absenteeism surges. Ultimately, any science of learning matters far less than its implementation. Maintaining fidelity to what happened in, say, Pavlov’s lab matters significantly less if the practices derived from his work contribute to stress, anxiety, and alienation in students.

If the perfect education system requires that you dehumanize the people in it — adults and kids alike — that's not a system that "works" by most metrics worth caring about. The kids in our schools have to be viewed as more than behaviorist subjects to be acted upon. If we at least admit that much, then the business of teaching gets far more complicated. Suddenly there are a number of other factors we must tend to that matter a great deal. I’ll quote again from apparent “pseudoscientist” Mary Helen Immordino-Yang, “As human beings, feeling alive means feeling alive in a body but also feeling alive in a society, in a culture; being loved, being part of a group, being accepted, and feeling purposeful." These are self-evident truths that we are finally beginning to explore the neurobiological basis for in ways that shatter many previous models of the brain that still hold cultural sway.

If anything, the incredible volume of work being done on neuroimaging at present offers all of us new information about how the mechanisms of consciousness work, how neural pathways and structures get built between anatomical components, how we may uncover the physical basis of memory, and how the interconnections between emotion, cognition, perception, and memory are still being actively explored with tremendous implications across many fields and domains. 

Teachers and educators must navigate these complexities, understanding that evidence-based practices, while valuable, are not a one-size-fits-all solution, may have very real flaws, and can be context dependent. The effective translation of scientific findings into successful classroom strategies requires a deep and nuanced understanding of both the science and its real-world implications, and very real science, with lots of evidence, has been ignored for years about the benefits of small-class sizes, or the deleterious impacts of environmental pollution on physical development, or the simple fact that brains don’t “think” in the absence of other systems. We need to prepare anyone using any science of learning to bridge the gaps between theory and practice then equip them to bridge those in a way that respects the diversity and individuality of learning experiences. And we can’t do that if the debate lives at the level of a label of a field of science rather than the very obvious implications of student and teacher experiences. 

As Luis Pessoa, neuroscientist and director of the Maryland Neuroimaging Center, offers a challenge to the causal, Newtonian perception of how the brain works in The Entangled Brain.“...this way of thinking,” he writes, “which has been very productive in the history of science, is too impoverished when complex systems — the brain for one — are considered…”‍

Ultimately, to explain the cognitive- emotional brain, we need to dissolve boundaries within the brain — perception, cognition, action, emotion, motivation — as well as outside the brain, as we bring down the walls between biology, psychology, ecology, mathematics, computer science, philosophy, and so on. Only then we will be on the right track.

To dismiss all of this off-handedly; to call new fields and findings across affective neuroscience, systems-level neuroscience, and embodied cognition a "pseudoscience" — and by association those who study them "pseudoscientists" — is to engage in a pseudoscience itself that prematurely decided we already knew everything about how kids learn, how school works, and how kids learn in the ecosystem of school. Labeling your perceived opponents “pseudoscientists” and “science deniers” is an incredibly effective way to squash promising lines of inquiry and discourse in education. An idea made even more troubling when prominent movement figures determine that a science of learning begins and ends with them. When these figures communicate to their audience that ideas are disagreeable simply because they reflexively disagree with them, it may help score points for their team, but it doesn’t help the cause of education. To say that we all figured this out long ago and that new evidence is actually pulling us astray, let's call that out for what it is: That's ideology, not science.