Beyond Sweetness: How Decades of Research at Brooklyn College Illuminate the Foundations of Learning
For decades, the field of learning has operated under a deceptively simple question: how do experiences shape the mind? But answering that question requires navigating a complex interplay of psychology, neuroscience, and increasingly, artificial intelligence. Andrew Delamater, a professor of experimental psychology at Brooklyn College and the CUNY Graduate Center, has dedicated his career to precisely this challenge, meticulously dissecting the mechanisms by which both humans and animals learn from their environments. His work isn’t about proving that learning happens – that’s a given – but about understanding how it happens at every level, from the firing of individual neurons to the complex computations that underpin behavior. What’s often lost in popular science reporting about the brain is the painstaking methodology required to move beyond correlation and towards genuine mechanistic understanding, and that’s where Delamater’s career provides a valuable case study.
Reporting from brooklyn.edu informs this analysis.
Delamater’s journey began in 1994 at Brooklyn College, drawn by the opportunity to build a research program and collaborate with established faculty. He initially approached the question of learning from a purely psychological perspective, intrigued by how expectations could influence perception. He cites the example of anticipating sweetness: does merely thinking about something sweet alter how we perceive its actual taste? The evidence, he notes, suggests a resounding yes. But this initial observation wasn’t an endpoint; it was a launching pad. The crucial shift in his thinking came with the advent of new neurobiological tools. Rather than simply observing behavioral changes, researchers could now directly measure neural activity while an animal anticipated a reward, like sugar water. This allowed for a direct comparison: did the brain activity associated with anticipation mirror the activity triggered by the reward itself? This move towards understanding the biological underpinnings of learning is a hallmark of the field’s evolution, and Delamater’s work exemplifies this trend.
This integration of disciplines – behaviorism, neurobiology, and computational modeling – is central to Delamater’s approach. He’s not simply interested in what animals learn, but in how that knowledge is represented within the brain. This extends beyond simple reward associations. He and his colleagues are now exploring how the brain encodes more abstract concepts like number and time, asking whether language is even necessary for these fundamental cognitive abilities. The question isn’t whether brains can represent these concepts, but how they do so using the relatively simple building blocks of neuronal excitation and inhibition. This pursuit of fundamental mechanisms is a deliberate effort to move beyond descriptive accounts of cognition and towards predictive models of brain function.
Delamater’s perspective, honed during his recent tenure as editor-in-chief of the Journal of Experimental Psychology: Animal Learning and Cognition, highlights a critical tension within the field. While increasingly sophisticated tools allow for deeper investigation of neural mechanisms, measuring these mechanisms across species remains a significant challenge. The field has made strides, but progress is often hampered by the difficulty of isolating and quantifying the underlying processes. This is where computational modeling enters the picture. The resurgence of artificial intelligence, particularly “deep learning” systems, has provided a new avenue for exploring how complex systems can learn and represent information. Researchers are now using AI as a kind of experimental participant, comparing its learning patterns to those of humans and animals. Interestingly, these comparisons reveal that AI often learns differently than biological systems, underscoring the need for interdisciplinary collaboration between psychologists, neuroscientists, and computer scientists to develop more biologically plausible AI models.
The implications of this research extend far beyond the laboratory. Delamater points to the role of associative learning in understanding trauma, where seemingly innocuous stimuli can trigger powerful re-experiencing of past events. This highlights the often-unconscious processes that shape our thoughts, feelings, and behaviors. He emphasizes that humans are often poor narrators of their own internal states, and that a great deal of cognitive processing occurs outside of conscious awareness. This realization has profound implications for fields like mental health and education, suggesting that interventions should focus not only on conscious thought but also on the underlying neurobiological and psychological mechanisms that drive behavior. Looking ahead, Delamater identifies three core questions that will continue to drive the field: the conditions necessary for learning, the content of that learning, and how that learning translates into observable performance.
For students considering a research career, Delamater’s advice is straightforward: find a question that genuinely excites you and then learn the tools to investigate it. But a crucial question remains for the next generation of researchers: as AI systems become increasingly sophisticated, will they serve as valuable models of the brain, or will their fundamentally different learning mechanisms continue to highlight the unique complexities of biological cognition? The answer to that question will likely determine the future direction of learning research for decades to come.
