Behaviorism is the view that mind consists in behavior: to be in pain, to believe it will rain, to want coffee just is to behave, and to be disposed to behave, in certain characteristic ways. For roughly the first half of the twentieth century it dominated psychology and much of philosophy, and understanding why it rose and why it fell explains most of what came after it in the philosophy of mind.
The Problem It Was Solving
Descartes left philosophy with a big question: if our mind and body are separate, how do they work together? And how could mental states, inaccessible to third-person observation and unverifiable in principle, ever be studied scientifically? Dualism made psychology look like ghost-hunting.
Behaviorism’s response was radical: eliminate the inner entirely. Mental states aren’t private inner events — they’re patterns of observable behavior and behavioral dispositions. To say someone is in pain isn’t to report an internal event; it’s to say they are disposed to wince, withdraw, say “ouch,” and seek relief. Pain just is that cluster of dispositions. No ghosts, no inner theater, no verification problem.
This was a serious philosophical program, not a naive one. Its main forms:
- Methodological behaviorism (Watson, Skinner): psychology should restrict itself to observable behavior as its subject matter, whatever the metaphysical facts about inner states
- Logical behaviorism (Carnap, Hempel): mental vocabulary can be translated into behavioral vocabulary without remainder — “pain” means nothing more than the relevant behavioral dispositions
- Philosophical behaviorism (Ryle): Cartesian talk of inner mental events rests on a category mistake — what looks like a report of an inner state is actually a description of behavioral competence and disposition
Ryle’s The Concept of Mind (1949) is the philosophical high-water mark. His “ghost in the machine” phrase remains the most famous diagnosis of the Cartesian picture: dualism, Ryle argued, treats the mind as a ghostly engine hidden inside the physical body, causing behavior from the inside.
The Standard Objections
The objection that did the most damage: behaviorism seems unable to distinguish genuine mental states from perfect behavioral mimicry.
Consider pain. A person in genuine pain and a skilled actor pretending have matching behavioral dispositions in the staged circumstances: both wince, withdraw, say “ouch,” seek relief. Yet one hurts and the other doesn’t. Hilary Putnam pressed the point in both directions, imagining “super-Spartans” who feel pain but have trained away every pain-behavior: if behavior can occur without the feeling and the feeling without the behavior, neither can define the other.
Behaviorists had a reply. Dispositions reach wider than any performance: offstage, in unstaged circumstances, the actor’s dispositions diverge from the sufferer’s, and even a super-Spartan stays disposed to avoid what damages him. The critics’ rejoinder: stipulate the pretense or the stoicism total, and a felt difference still seems to remain, which suggests the feeling never consisted in dispositions at all.
A second objection targets the circularity of dispositional analysis. To cash out the disposition to “seek relief,” you need to invoke the belief that relief is available, the desire to obtain it, the intention to act. But beliefs, desires, and intentions are themselves mental states awaiting behavioral analysis, and the circle never closes. Roderick Chisholm and Peter Geach both pressed this point in the 1950s: no dispositional analysis of one mental state completes without invoking others.
What Survived It
Even behaviorism’s harshest critics kept two of its lessons. First, Ryle’s diagnosis outlived the analysis it served: much philosophical confusion about mind does come from treating mental-state vocabulary as if it named hidden inner objects, and “the ghost in the machine” remains the standard name for that picture. Second, mental states need some anchor in the publicly accessible world. Behaviorism chose behavior; later theories chose causal roles or world-involving relations. Different answers to the same demand.
The Transition to Identity Theory
By the late 1950s, the objections had persuaded most philosophers that inner states could not be analyzed away. U. T. Place (1956) and J. J. C. Smart (1959) drew the conclusion: if we need inner states, and dualism is unacceptable, then inner states must be physical — specifically, brain states. Pain is C-fiber firing, as the slogan went. This is identity theory: the inner is real, and it’s neural.
My View
Behaviorism was the overcorrection to dualism. It threw out the ghost, and with it the genuine inner states a theory of mind exists to explain. The actor marks the exact spot where the analysis fails: the difference between pretending and hurting is the inside, and behaviorism had defined the inside away. That failure is not a minor gap — it falls on the most basic case the theory needed to handle, the felt quality of the state itself.
What I keep is Ryle’s diagnosis. The ghost in the machine really is a bad picture, and my own campaign against reified inner objects is a Rylean move; I recruit his diagnosis while declining his cure. Behaviorism’s demand for a public anchor was legitimate too — my answer just runs through world-involving representational relations rather than through behavior. The Chinese Room exposes a behaviorist shortcut that later readers sometimes build into the Turing Test: fluent performance alone does not constitute understanding. Behavior remains defeasible evidence, and the test does not settle what realized organization produces the performance.
The history, as I read it, forms a spine, each position failing in the way that makes the next step necessary:
| Position | Inner states? | Physical? | Where it fails |
|---|---|---|---|
| Dualism | Yes | No | No causal story; no science |
| Behaviorism | No (analyzes them away) | — | Can’t distinguish genuine states from mimicry |
| Identity Theory | Yes | Yes (specific substrate) | Multiple realizability |
| Functionalism | Yes | Not by definition | Narrow, world-bracketed role leaves content open; wide functionalism may converge with the book’s world-involving account |
| Strong representationalism within a physicalist, world-involving account | Yes | Yes, without substrate chauvinism | The book’s view; embodiment is one powerful realization route, not a universal gate |