Functionalism and behaviorism offer fundamentally different answers to the same question: how should we study and explain the mind? Behaviorism says we should stick to what we can observe, meaning stimuli going in and behavior coming out, and treat the mind’s interior as either off-limits or irrelevant. Functionalism says the interior matters enormously, but what matters about it is not its physical makeup but what it does, the roles mental states play in causing other mental states and producing behavior. The split between them shaped twentieth-century psychology and philosophy of mind, and the reverberations still affect how researchers think about consciousness, artificial intelligence, and animal cognition today.
What Behaviorism Actually Claims
Behaviorism is not one position but a family of related commitments. The version most people have in mind when they hear the word is methodological behaviorism, which holds that the terms and concepts used in psychological theories should be grounded in observable stimuli and behavior. On this view, psychology earns its status as a science by restricting itself to things that can be measured from the outside: what an organism does in response to what happens to it. Internal states like beliefs, desires, and feelings are either redefined in terms of behavior or set aside as scientifically untouchable.1PubMed Central. Methodological Behaviorism from the Standpoint of a Radical Behaviorist
B. F. Skinner’s radical behaviorism parts company with methodological behaviorism in an important way. Skinner did not deny the existence of private events like thoughts and feelings. He acknowledged them but insisted they should be understood through the same principles that govern publicly observable behavior, as products of reinforcement histories and environmental contingencies rather than as evidence of an inner mental realm operating by its own rules. Where methodological behaviorism treats inner life as scientifically invisible, radical behaviorism treats it as more behavior, just behavior that happens to be private.2PubMed Central. Methodological Behaviorism from the Standpoint of a Radical Behaviorist
This distinction matters more than it might seem. Critics who dismiss behaviorism as denying that people have inner lives are usually attacking a caricature of the methodological variety. Skinner’s version is subtler. It does not deny your pain or your daydreams. It denies that explaining those experiences requires postulating a special mental dimension that works differently from the rest of the natural world.
Why Behaviorism Lost Its Grip
For roughly three decades in the mid-twentieth century, behaviorism dominated American psychology. Its influence began to weaken in the late 1950s and 1960s, and the most famous single blow came from the linguist Noam Chomsky. His 1959 review of Skinner’s book on language has been called one of the most influential documents in the history of psychology.3PubMed Central. The Case Against B. F. Skinner 45 years Later: An Encounter with N. Chomsky Chomsky argued that Skinner’s attempt to explain language as learned behavior shaped by reinforcement could not account for the creativity and structure of human speech. Children produce sentences they have never heard before, and they do it in ways that follow complex grammatical rules no one explicitly taught them. That looked far more like the output of an internal system with its own logic than like a set of conditioned responses.
Chomsky’s critique arrived at a moment when other pressures were building. The rise of computers provided a new metaphor for the mind, one that made internal processing seem not only legitimate but essential. If a computer could take inputs, run them through internal operations, and produce outputs, then maybe the interesting action was happening in the middle, not just at the edges. Psychologists increasingly felt that ignoring the interior of the system meant ignoring most of what needed explaining.
What Functionalism Claims
Functionalism emerged as a way to take the mind’s interior seriously without falling back on the old idea that mental states are just brain states in disguise. Its core claim is straightforward: mental states should be understood in terms of their functional relationships to sensory inputs, behavioral outputs, and other mental states, rather than in terms of whatever physical material happens to implement them.4Cognitive Science. Parallel computation and the mind-body problem
Think of it this way. Pain, on a functionalist account, is whatever state is typically caused by tissue damage, typically causes wincing and avoidance behavior, and typically produces a desire for the sensation to stop. If something plays that role in your nervous system, it counts as pain regardless of whether it is implemented by neurons, silicon chips, or something else entirely. The job description defines the state, not the stuff doing the job.
This idea drew direct inspiration from computer science and from Alan Turing’s foundational work on computation. Turing showed that what matters about a computing process is the function it realizes, not the physical hardware that carries it out.5Journal of Logic and Computation. Turing is among us The philosopher Hilary Putnam made the connection explicit in the 1960s, arguing that mental states could be described in terms of the kind of abstract state-transition tables used to characterize Turing machines. Different physical systems could run the same program, so different brains could realize the same mental states.6Studies in History and Philosophy of Science Part A. Functionalism, computationalism, and mental states
The Core Disagreement in Plain Terms
Strip away the history and the technical vocabulary, and the disagreement between behaviorism and functionalism comes down to whether the mind’s internal organization is a legitimate part of scientific explanation. Behaviorism says no, or at least not directly. You study inputs and outputs. If you posit internal states at all, they earn their keep only by connecting back to observable behavior. Functionalism says yes, emphatically. Internal states are real, causally powerful, and scientifically describable. But functionalism also insists that what makes them the states they are is not their physical composition but their functional role.
This means behaviorism and functionalism are not just debating tactics. They are debating what counts as an explanation in psychology. A behaviorist explanation of why you pulled your hand off a hot stove would describe the stimulus (heat), the response (withdrawal), and the reinforcement history that shaped the connection between them. A functionalist explanation would describe an internal state (pain) defined by its characteristic causes, effects, and connections to other states (the belief that the stove is hot, the desire to avoid injury), and would treat the behavior as an output of that internal network.
Multiple Realizability and the Identity Problem
One of functionalism’s most powerful arguments is what philosophers call multiple realizability. The idea is that the same mental state could be physically implemented in very different ways across different kinds of creatures. This creates a serious problem for the view that mental states just are brain states of a specific physical type.
Consider pain again. If pain in humans involves one kind of neural firing and pain in dolphins involves a completely different kind, then pain cannot simply be identical to the human neural pattern, because that would leave dolphin pain unexplained. We feel sorry for a dolphin that shows every sign of suffering even though we do not know the fine details of its neurobiology. Functionalism captures that intuition: what makes something pain is the role it plays, not the neural hardware carrying it out. The identity theorist, meanwhile, is stuck with the awkward conclusion that pain equals one physical state in one species and a completely different physical state in another, which by the logic of identity would absurdly imply those two physical states are the same thing.7Routledge Encyclopedia of Philosophy. Mind, identity theory of
This argument gave functionalism a clear edge over both behaviorism and the brain-identity theories that were popular in the mid-twentieth century. It said: you can take mental states seriously, you can give them causal roles in explaining behavior, and you do not have to tie them to one specific kind of biological tissue.
Behaviorism Did Not Disappear
It would be a mistake to think behaviorism simply lost and was replaced by functionalism. In academic philosophy of mind, functionalism did largely win the debate. But in psychology, especially in clinical practice, behaviorist principles remain deeply embedded. Applied behavior analysis, the field most directly descended from Skinner’s work, is one of the most widely used approaches for treating autism spectrum disorder and other developmental conditions. Cognitive-behavioral therapy, the dominant form of psychotherapy for anxiety and depression, is a hybrid that grafts cognitive elements onto a behaviorist framework.
Within the philosophy of behavioral science itself, behaviorism has diversified considerably. Scholars working in the tradition have extended and modified its principles in ways that sometimes blur the boundary between behaviorism and other frameworks. The result is that people both inside and outside the field are often uncertain about what exactly behaviorism means in its current form.8PubMed Central. Current Diversification of Behaviorism Some contemporary behaviorists accept a richer account of private events than Skinner did, while others push in the opposite direction, insisting on stricter observability requirements. The label covers more ground now than it did fifty years ago.
Where Functionalism Runs Into Trouble
Functionalism’s strength is its flexibility: it does not care what a mental state is made of, only what it does. But that same flexibility creates problems that have dogged the theory since the 1970s.
The most persistent objection involves subjective experience. Imagine two people who are functionally identical in every respect: the same inputs produce the same internal state transitions and the same outputs. Now suppose one of them experiences red where the other experiences green, but neither of them ever shows any behavioral or functional difference because their internal wiring is simply “flipped” in a consistent way. If such a scenario is even conceivable, then functional role does not fully capture what it is like to be in a mental state. There is something left over, something about the qualitative feel of experience, that functionalism leaves out. Philosophers call these qualitative properties “qualia,” and the worry that functionalism cannot account for them has generated decades of heated debate.
A related challenge comes from thought experiments about systems that seem to replicate functional organization without producing genuine understanding. If you could build a system that matched the functional profile of a mind state-for-state but clearly did not understand anything, that would suggest functional organization alone is not sufficient for mentality. These thought experiments have not settled the question, but they have kept functionalists busy refining their accounts to address whether function alone can carry the full weight of mental life.
Teleofunctionalism and the Role of History
One influential refinement of functionalism tries to solve certain problems by adding biological history to the picture. Teleofunctionalism says that a mental state is not defined just by the functional role it currently plays but also by the evolutionary function it was selected to perform. A frog’s visual state that causes it to snap at flies counts as a “seeing-flies” state partly because evolution shaped it for that purpose, not just because it happens to be triggered by small dark moving objects.
This move helps with some classification problems, because it gives functionalism a principled way to distinguish states that play similar current roles but have different evolutionary origins. However, teleofunctionalism has faced criticism for struggling to explain how mental content actually causes behavior. One analysis argues that the theory inadvertently makes the content of mental states irrelevant to explaining why someone acts the way they do, because the causal work is done by the physical properties of the state rather than its historically defined content.9Pacific Philosophical Quarterly. Teleofunctionalism and Psychological Explanation If the evolutionary story about what a state is “for” does no real causal work, then adding it to functionalism may not buy as much as it promises.
When Biology Pushes Back Against Functionalism
Functionalism’s claim that mental states are independent of their physical implementation has always been somewhat in tension with neuroscience, which keeps finding that specific physical mechanisms matter a lot. A striking recent example involves octopuses. Octopuses have nervous systems that are radically different from ours: most of their neurons are in their arms, not a centralized brain, and their evolutionary lineage diverged from ours hundreds of millions of years ago. Yet when researchers gave octopuses MDMA (the drug commonly called ecstasy), the animals became markedly more social, just as humans do on the same substance. The reason appears to be that octopuses have serotonin transport systems nearly identical to human ones.10Synthese. Why prosocial octopuses challenge the autonomy of psychological states
This is a headache for strict functionalism. If the same psychological state (prosocial behavior driven by serotonin) shows up in species with wildly different overall neural architecture but very similar neurochemistry, it suggests that sometimes the physical implementation does matter. The conserved serotonin mechanism, not just the abstract functional role, appears to be doing the heavy lifting. Standard functionalism claims mental states are autonomous from neural states, meaning they can be realized in many different physical ways. The octopus case suggests that at least some psychological states are not as free-floating as functionalism assumes; they are tethered to specific biochemical machinery that evolution has preserved across vastly different body plans.11Synthese. Why prosocial octopuses challenge the autonomy of psychological states
Consciousness and the Gap Both Frameworks Leave
Neither behaviorism nor classical functionalism has a fully satisfying account of consciousness. Behaviorism sidesteps the issue by declining to make consciousness a central theoretical concept. Functionalism embraces it in principle but struggles with the qualitative dimension described earlier.
Neuroscience has pushed the conversation in directions neither framework anticipated. Research into the neural correlates of consciousness suggests that conscious experience is associated with synchronized activity across networks in the brain’s temporo-parietal and occipital regions, with attention, supported by different fronto-parietal networks, entering the process to produce focused awareness of specific features.12PubMed Central. The Neural Correlates of Consciousness and Attention: Two Sister Processes of the Brain These findings are more fine-grained than anything functionalism’s abstract state descriptions typically capture, and they suggest that understanding consciousness will require paying close attention to the specific physical mechanisms involved, not just the functional roles they play.
At the same time, the neuroscience findings do not vindicate old-school behaviorism either. The whole point of correlating brain activity with conscious experience is that conscious experience is treated as real and scientifically investigable, exactly the move behaviorism was reluctant to make.
Enactivism and Newer Alternatives
Some researchers have concluded that both behaviorism and functionalism miss something fundamental about how minds work. A growing movement called enactivism argues that cognition is not the manipulation of internal representations (as classical functionalism and cognitive science assume) and not just input-output pairings (as behaviorism proposes). Instead, cognition arises from an organism’s active engagement with its environment, from the dynamic loop of perceiving and acting rather than from any internal model or program.13Revista Colombiana de FilosofÃa de la Ciencia. Mente devuelta a la vida: Superar las limitaciones representacionalistas y funcionalistas a través de paradigmas enactivistas
Enactivists criticize functionalism for treating the mind as a kind of software that could in principle run on any hardware, arguing that this strips away the role of the body and the environment in constituting mental life. They criticize behaviorism for the opposite reason: by focusing exclusively on observable behavior, it ignores the rich internal dynamics that are part of what it means to be a living, sensing creature. Whether enactivism will mature into a full replacement for either framework or will instead be absorbed as a corrective to their weaknesses remains an open question. But its popularity signals a widespread sense that neither of the twentieth century’s dominant approaches fully captures what is going on when a mind engages with the world.
How the Debate Shapes Artificial Intelligence
The functionalism-versus-behaviorism divide is not purely academic. It has real consequences for how we think about artificial intelligence. If functionalism is right and what matters about a mental state is its functional role rather than its physical substrate, then there is no principled reason why a sufficiently sophisticated computer program could not have genuine mental states. A machine that replicated the full functional organization of a human mind would, on this account, genuinely think, feel, and understand. This is why functionalism has been so influential in AI research and in philosophical debates about machine consciousness.
Behaviorism, interestingly, offers a different route to roughly the same destination. If the only scientifically meaningful facts about minds are facts about input-output relations, then a machine that produces human-like outputs in response to the same inputs is, for all practical purposes, exhibiting the same psychology. The Turing test, which evaluates machine intelligence purely on the basis of behavioral indistinguishability from a human, is essentially a behaviorist proposal.
But the challenges to both frameworks complicate the picture. If qualia and subjective experience matter, and if they are not captured by functional organization alone, then a functionally perfect AI might still lack something essential. And if specific biochemistry turns out to be necessary for certain psychological states, as the octopus evidence hints, then running a program on silicon may not be enough no matter how sophisticated the functional mapping. These are not settled questions. They are live debates that draw directly on the philosophical foundations laid down by the behaviorism-functionalism argument decades ago.