For most of the twentieth century, psychology studied behavior. Observable actions, measurable responses, lawful relationships between stimulus and response. Then something fundamental shifted. By the 1960s, a new generation of researchers began insisting that psychology's real subject matter wasn't behavior at all—it was the hidden machinery that produced behavior. Mental representations. Internal symbols. Computational processes operating beneath the surface of observable action.

This wasn't merely a change in method or emphasis. It was a transformation in what psychology believed it was trying to explain. The cognitive revolution didn't just add new tools to the psychologist's toolkit. It redefined the explanatory target itself. Suddenly, a complete account of human action required something more than describing behavioral regularities—it required positing internal states that represented the world and computed over those representations.

Whether this shift constituted genuine scientific progress remains surprisingly contested. Did cognitive science discover that psychology had been studying the wrong thing all along? Or did it simply change the subject, replacing one legitimate domain of inquiry with another? The answer matters not just historically but for understanding what psychological explanation can and should accomplish. The representational commitment that defines cognitive science carries profound implications—implications that emerging theoretical alternatives are now calling into question.

When cognitive psychology emerged, it brought with it a specific metaphysical claim: minds traffic in representations. This wasn't an innocent methodological assumption. To commit to mental representation is to assert that between stimulus and response lies a structured internal world—symbols, images, propositions, or some other format that stands for things beyond itself. The mind becomes a realm of content-bearing states that can be accurate or inaccurate, appropriate or misguided.

This commitment transformed what counted as psychological explanation. Behaviorist accounts could remain agnostic about internal states—they described input-output relationships without speculating about mediating mechanisms. Representational psychology abandoned this parsimony. It claimed that genuine understanding required knowing what the organism represented and how those representations were transformed through processing. A behavior wasn't explained until you'd characterized the representational states that produced it.

The theoretical weight here is considerable. Representations require content—they must be about something. They require format—some vehicle that carries that content. They require rules—principles governing how representations interact and transform. Each of these requirements generates its own research program and its own philosophical puzzles. What grounds the content of a mental state? What's the representational format of visual imagery versus linguistic thought? What makes a transformation count as rational inference versus mere causal succession?

This explanatory framework shaped everything that followed. Memory became the storage and retrieval of representations. Perception became the construction of internal models. Reasoning became computation over symbol structures. Cognitive psychology didn't just study mental processes—it studied mental processes conceived in representational terms. The framework determined which phenomena needed explaining and what would count as adequate explanation.

Critics have argued that this represents a category shift rather than theoretical progress. The behaviorists weren't failing to explain representation—they were deliberately studying something else. Whether representational psychology advanced beyond behaviorism or simply replaced it with a different research program depends on whether you think representation was always the true subject matter waiting to be discovered, or merely one legitimate focus among several possible ones.

Takeaway

Adopting mental representation as psychology's subject matter wasn't a neutral methodological choice—it was a substantive claim about what minds fundamentally are and do, with cascading consequences for every subsequent theoretical decision.

The representational commitment needed an account of how representations did anything—how they produced behavior, influenced each other, enabled cognition. The answer came from computer science: minds are computational systems. They manipulate symbols according to rules. Cognitive processes are algorithms operating over representational data structures. This metaphor proved extraordinarily productive, but its explanatory scope has always been narrower than its proponents sometimes claimed.

Computational explanation excels at characterizing formal relationships—the structural patterns that determine how one mental state leads to another. If you represent the premises of a syllogism, computation specifies how you derive the conclusion. If you encode a visual scene, computation describes the transformations yielding object recognition. The power lies in precision. Computational theories can be implemented, tested, and refined with unusual rigor. They make psychology look more like the exact sciences.

But computational approaches struggle with several persistent problems. First, the frame problem: how does a computational system know which representations are relevant to a given situation? Humans effortlessly filter vast amounts of stored knowledge to focus on what matters now. Computational accounts have difficulty specifying how this filtering works without invoking the very cognitive capacities they're meant to explain. Second, embodiment: computational models tend to treat cognition as something that could occur in any physical substrate. Yet human cognition seems deeply entangled with bodily states, sensorimotor capacities, and environmental structures in ways that pure computation doesn't capture.

Third, and perhaps most fundamentally, computation describes syntax—formal symbol manipulation—but cognition involves semantics—meaning, content, aboutness. The relationship between these levels remains philosophically murky. A computer program can manipulate Chinese characters according to perfect syntactic rules without understanding Chinese. Does the brain's computation over neural symbols similarly lack genuine understanding? Or does biological computation achieve something that artificial computation cannot?

These limitations don't invalidate computational approaches. They do suggest that computation might be one explanatory framework among several rather than the master key to cognition. The metaphor illuminates certain aspects of mental life while leaving others in shadow. Recognizing where the metaphor breaks down is essential for understanding what cognitive science has actually achieved and where it needs supplementation.

Takeaway

Computational explanation provides unprecedented precision for describing formal cognitive operations, but its limits reveal that processing information and understanding meaning may require different kinds of theoretical account.

The past three decades have witnessed sustained challenges to the representational-computational framework. Embodied, embedded, enacted, and extended cognition—the so-called 4E approaches—question whether mental representation is as central to cognition as the cognitive revolution assumed. Dynamical systems theory proposes that cognitive processes are better understood as continuous, time-evolving patterns rather than discrete symbol manipulations. Predictive processing frameworks reconceive perception and action as unified processes of prediction error minimization rather than separate input-output stages.

These alternatives don't uniformly reject representation. Some argue for more minimal, action-oriented representations rather than rich internal models. Others suggest that representational talk is explanatorily useful for some phenomena but dispensable for others. Still others maintain that what traditional cognitive science called representation is really something quite different—anticipatory patterns, embodied skills, dynamic couplings between organism and environment.

What these approaches share is skepticism about the necessity of the framework that the cognitive revolution established. If cognition is fundamentally about skillful bodily engagement rather than internal modeling, then the explanatory target shifts again. If minds are constituted partly by environmental structures rather than contained within skulls, then the boundaries of psychological explanation expand beyond what traditional cognitive science permitted. Each alternative proposal implies different research priorities, different methodological commitments, and different criteria for explanatory success.

The theoretical situation resembles what Thomas Kuhn described for pre-paradigmatic sciences: multiple frameworks compete without clear criteria for adjudicating between them. Perhaps this signals genuine scientific immaturity. Or perhaps it indicates that mind is too multifaceted for any single framework to capture comprehensively. Theoretical pluralism—maintaining multiple explanatory approaches simultaneously—might be not a failure of scientific maturity but an appropriate response to complexity.

What's certain is that the cognitive revolution's framing can no longer be taken for granted. Its representational commitments, its computational metaphors, its assumptions about cognitive boundaries—all are now legitimate objects of theoretical scrutiny. The question isn't just how the mind works but what kind of thing the mind is and what kind of explanation its nature permits. These are philosophical questions that empirical research alone cannot settle.

Takeaway

Emerging alternatives to classical cognitive science suggest that the representational framework may capture only certain aspects of mentality—raising the possibility that psychological understanding requires multiple, irreducible theoretical perspectives.

The cognitive revolution transformed psychology by changing what it thought required explanation. Mental representation replaced behavior as the primary explanatory target. Computation provided a framework for understanding how representations produced cognition. This framework enabled remarkable empirical progress and established cognitive science as a serious theoretical enterprise.

Yet the framework's foundations remain contested. Whether representation is essential to all cognition, whether computation captures mental processes adequately, whether alternative approaches might prove more illuminating—these questions remain open. The cognitive revolution may have been genuine progress, a change in subject matter, or perhaps both simultaneously.

What persists is the deeper lesson: psychology's explanatory targets aren't given in nature. They're theoretical choices with theoretical consequences. How we conceive the mind shapes what we discover about it. Recognizing this reflexive relationship between framework and finding is essential for understanding where psychological knowledge has come from and where it might still go.