What Is the Problem of Other Minds?

Direct Answer

The problem of other minds is the question of how we can know, or even reasonably believe, that other people have minds — conscious experiences, thoughts, and feelings — at all. The problem arises from a fundamental asymmetry. I have direct, first-person access to my own mind: I know what it is like to be me. But I have no such access to yours. All I can observe is your behavior: your words, your facial expressions, your actions. From these I infer an inner life, but the inference itself seems to need justification. The problem is not about knowing what you think (that is difficult but not impossible); it is about knowing that you think — that there is a mind behind the behavior at all.

The classic response is the argument from analogy, traced to John Stuart Mill and implicit in Descartes. I know from my own case that certain behaviors are caused by certain mental states: when I am in pain, I wince; when I am sad, I cry. So when I see you wincing, I infer by analogy that you are in pain. The inference is analogical: your behavior resembles mine, so your inner states probably resemble mine. The argument's weakness is its inductive base: it is an inference from a single case — my own — to all cases, which is notoriously weak evidence.

The key distinction in the modern debate is between two ways of understanding what the problem is. On the epistemological reading, the problem is a skeptical puzzle: justify the inference to other minds in the face of the "evil demon" or the "brain in a vat" possibility that other bodies are empty automata. On the conceptual reading, defended by Ludwig Wittgenstein, the problem rests on a misunderstanding: mental terms are learned and used through public criteria, so the very idea of a "private" inner realm that only I can know is confused. If "pain" means what is publicly expressed in pain behavior, then there is no inference to justify; other minds are not a hypothesis but a presupposition of our language.

Historical Context

The problem of other minds emerged with the Cartesian picture of the mind as a private inner realm. If minds are essentially private — each of us locked in our own theater of consciousness — then the existence of other theaters is a hypothesis, and skepticism threatens. Descartes himself dodged the problem by appeal to God's veracity: God would not deceive us into believing others have minds if they did not. The problem became acute in the eighteenth and nineteenth centuries: Mill's Examination of Sir William Hamilton's Philosophy (1865) offered the argument from analogy; the behaviorist tradition of the early twentieth century dissolved the problem by denying the inner realm altogether.

In the twentieth century, the problem took two new directions. Wittgenstein's private language argument (in Philosophical Investigations) challenged the very coherence of a private realm, while the logical behaviorists (Ryle) and later functionalists (Putnam) reconceived mental states so that their ascription to others could be grounded in observable criteria. The debate was then transformed by technology: Alan Turing's 1950 paper asked whether machines could think and proposed the imitation game — a behavioral criterion for mind — and Hilary Putnam's "super-spartan" argument showed that no finite set of behavioral dispositions can settle whether a being has inner experience. The problem of other minds became the problem of artificial minds.

Key Arguments & Debates

The argument from analogy remains the intuitive starting point: I generalize from my own case. Its defenders argue that the analogy is strengthened by complexity — the more detailed and specific the behavioral similarity, the stronger the inference — and by the best-explanation inference: the hypothesis that you have a mind explains your behavior far better than the hypothesis that you are a mindless automaton. Why would a machine behave exactly like a person in pain, unless it were in pain? The inference to the best explanation avoids the single-case weakness of the analogy.

The objections are powerful. The skeptical objection: no inference from my single case can license a universal conclusion, and the "zombie" possibility — a being behaviorally identical to me but with no inner experience — is at least conceivable, so the inference cannot be deductive. The behaviorist objection: if inner states cannot be known, and behavior is all we have, then mental ascription is just behavior ascription. The functionalist objection, from Putnam and the AI tradition: whether a system has a mind depends on its internal organization, not on whether it resembles me — which makes the question empirical and opens it to machines. The debate now turns on whether first-person experience can be given a third-person criterion at all.

Contemporary Relevance

The problem of other minds has become the central philosophical question of the age of AI. When a chatbot converses fluently, a robot expresses distress, or a large language model produces empathetic text, the question "does it have a mind?" is exactly the problem of other minds extended to machines. The Turing test is a behaviorist solution to the problem; the Chinese room argument is a refusal to accept behavioral criteria; philosophical zombie arguments question whether any behavioral or functional criterion can capture inner experience. The same problem structures animal ethics (do nonhuman animals have minds?), disability ethics, and the treatment of patients with locked-in syndrome or coma. And in epistemology, the problem connects to the deepest questions about knowledge: if I cannot know that other minds exist, can I know anything about the world at all?

Related Concepts

• Solipsism: the skeptical position that only my own mind exists.
• The Turing Test: the behavioral criterion for machine minds.
• The Chinese Room Argument: the objection to behavioral criteria.
• The Brain in a Vat: the skeptical scenario behind the problem.
• Philosophical Zombie: the being behaviorally identical but internally empty.
• Consciousness: the property whose ascription is at issue.

Further Learning

• Stanford Encyclopedia of Philosophy: Other Minds
• Anita Avramides, Other Minds (2001).
• Alan Turing, "Computing Machinery and Intelligence," Mind 59 (1950).