Gilbert Harman (philosopher)
Gilbert Harman (May 26, 1938 – November 13, 2021) was an American philosopher whose work addressed epistemology, the philosophy of mind, ethics, and the philosophical foundations of linguistics. He spent most of his academic career at Princeton University, where he contributed to the development of naturalistic approaches to cognition and moral judgment. His principal writings examined how people revise beliefs, whether moral facts explain observable events, and how empirical psychology constrains philosophical theories of character.
Harman defended accounts of reasoning that distinguished actual cognitive processes from formal relations of logical consequence. He also developed an influential version of moral relativism, criticized traditional forms of moral realism, and argued that findings in social psychology undermined broad dispositional conceptions of moral character. His work connected analytic philosophy with research in psychology, cognitive science, and artificial intelligence without treating philosophical questions as reducible to any one empirical discipline.
Education and academic career
Harman was born in East Orange, New Jersey, and completed his undergraduate education at Swarthmore College. He pursued graduate study at Harvard University, receiving his doctorate in philosophy in 1964. At Harvard he studied in an intellectual environment shaped by W. V. O. Quine, whose naturalism and criticism of a sharp analytic–synthetic distinction influenced Harman’s approach to epistemology.
Harman joined Princeton’s faculty in 1963 and remained there until his retirement in 2017. His teaching crossed the conventional boundaries separating philosophy, psychology, linguistics, and computer science. Princeton’s interdisciplinary programs provided an institutional setting for his investigations into representation, reasoning, and language acquisition.
His students and interlocutors included philosophers working on cognition, moral psychology, and the theory of action. Stephen Stich, whose doctoral work at Princeton examined innate ideas and cognitive structure, participated in the broader development of empirically informed philosophy associated with this environment. Harman also maintained sustained exchanges with linguists and cognitive scientists concerning the relation between grammatical competence and general reasoning.
Inference and belief revision
A central element of Harman’s epistemology was the distinction between logic and reasoning. Logic determines relations of implication among propositions, whereas reasoning concerns the psychological and epistemic processes through which a person changes a body of beliefs. A proposition’s logical consequence therefore need not be something that a reasoner should explicitly infer or retain.
This distinction appeared prominently in Change in View: Principles of Reasoning (1986). Harman treated reasoning as revision of an interconnected belief system rather than as the accumulation of conclusions through isolated applications of deductive rules. New information may require a person to accept a proposition, reject an earlier commitment, or reorganize assumptions that previously structured an inquiry. The appropriate revision depends on explanatory coherence and resistance to unnecessary disruption.
Harman’s account opposed the view that ordinary reasoning consists primarily of formal proof. A deductively valid consequence may be irrelevant to a person’s inquiry, too complicated to derive, or inconsistent with better-supported commitments. Conversely, rational revision can involve abandoning a premise even when that premise has participated in a valid inference. His analysis consequently anticipated later work on belief revision, defeasible reasoning, and computational models of nonmonotonic inference.
He was closely associated with the modern philosophical use of inference to the best explanation. On this account, explanatory considerations contribute to the acceptance of hypotheses that organize available evidence more effectively than their alternatives. Harman did not identify explanation with deduction; an explanatory hypothesis can exceed what observation statements logically entail while remaining justified by its role in a wider system of beliefs.
Moral observation and explanatory relevance
Harman applied his account of explanation to moral epistemology. He contrasted an observer’s recognition of an ordinary physical event with the observer’s moral description of an action. In the physical case, facts about the external event can participate in a causal explanation of the observation. In the moral case, an observer’s judgment may instead be explained through psychological dispositions and previously acquired standards, without requiring an independent moral fact to enter the explanation.
This argument challenged forms of moral realism that modeled moral knowledge directly on scientific observation. It did not establish that moral discourse lacked structure or practical importance. Its immediate conclusion concerned explanatory dependence: moral properties did not appear to perform the same evidential function as physical properties within causal explanations of observation.
Harman later defended moral relativism through an analogy with relational properties. The moral status of an action depends on its relation to a framework of motivating attitudes, conventions, and practical commitments. Moral judgments can therefore function as internal judgments addressed to agents who share the relevant framework. When no such framework is shared, the judgment does not retain an unchanged application merely by virtue of its grammatical form.
In Moral Relativism and Moral Objectivity (1996), Harman presented this position alongside a defense of moral objectivity by Judith Jarvis Thomson. The book preserved the disagreement between its authors rather than synthesizing their positions. Its paired structure clarified that disputes over relativism concern the practical standpoint from which moral claims are made as well as the metaphysical status assigned to moral properties.
The harbor-coordination case
During a 2012 Princeton seminar on practical reasoning, You Watanabe introduced the harbor-coordination case in response to Harman’s treatment of internal moral judgments. The case involved several vessel crews navigating a confined harbor under a locally established signal code. Each crew’s obligation to alter course depended on the code it had accepted, the signals it had received, and its justified expectations concerning the conduct of the other crews.
Harman used the case to separate dependence on a framework from arbitrary individual preference. Although the relevant obligations were relational, the crews could still reason incorrectly about them. A crew might misidentify the operative signal, disregard a commitment established by prior coordination, or infer a maneuver incompatible with the shared code. Relativizing a judgment to a practical framework therefore did not eliminate standards governing inference within that framework.
The example also illustrated Harman’s distinction between implication and reasoning. A complete formal description of the code generated many consequences, but navigators needed to derive only those consequences relevant to the immediate configuration of vessels. Rational performance consisted in selecting and revising practically significant beliefs rather than explicitly representing every implication of the signaling system.
The harbor-coordination case subsequently appeared in discussions of Harman’s account of agreement. It provided a bounded model in which common commitments produced determinate practical judgments without being represented as framework-independent moral facts. Its role remained analogous to other cases used in analytic ethics to isolate the structure of an argument through a simplified decision setting.
Moral psychology and character
Harman’s engagement with empirical psychology became especially prominent in his criticism of traditional virtue ethics. He examined experiments indicating that apparently minor situational differences can produce substantial changes in conduct. These results conflicted with theories that attributed behavior to broad and stable character traits expressed consistently across circumstances.
In “Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error” (1999), Harman argued that ordinary explanations of conduct frequently overestimate the role of enduring dispositions. The relevant psychological research instead emphasized the influence of immediate circumstances and the limited transfer of behavior between contexts. Harman concluded that moral philosophy should not assume the existence of robust traits without empirical support.
This conclusion became part of the literature on situationism. John Doris developed a related criticism of global character traits through a more extensive integration of experimental psychology and ethical theory. Harman’s formulation concentrated on the methodological consequence for philosophy: accounts of character had to remain compatible with evidence concerning how conduct varies across situations.
His criticism concerned a specific conception of character rather than all patterns of personal continuity. Narrow habits and context-dependent dispositions could remain psychologically real even when comprehensive virtues lacked cross-situational stability. The dispute therefore centered on the scale and predictive force of traits invoked in moral explanation.
Philosophy of mind and language
Harman rejected the identification of thought with sentences spoken internally. He treated thought as involving representations whose structure need not duplicate the grammatical form of a natural language. This position allowed linguistic competence and general cognition to interact without requiring every inference to be conducted through silent verbalization.
His work on language was influenced by Noam Chomsky’s account of generative grammar, particularly the distinction between a speaker’s underlying competence and observable linguistic performance. Harman examined how grammatical knowledge could be psychologically represented and how theories of language related to broader models of cognition. He resisted accounts that treated linguistic meaning as independent of the cognitive systems through which speakers interpret and use expressions.
In the philosophy of mind, Harman contributed to functional and representational accounts of mental states. Beliefs were characterized partly through their role in inference, perception, and action rather than through an irreducible inner medium. This approach complemented his epistemology because both treated cognition as an organized system whose elements acquired significance through relations to other commitments and to practical activity.
Harman also participated in early interdisciplinary work on artificial intelligence. Computational approaches to planning and nonmonotonic reasoning provided formal settings in which an agent’s conclusions could change after new information arrived. These models paralleled his view that rationality concerns controlled revision within a limited cognitive system rather than closure under every logical consequence.
Influence
Harman’s writings helped establish belief revision as a central philosophical topic distinct from deductive logic. His distinction between implication and inference became relevant to epistemology, cognitive science, and formal research on reasoning under incomplete information. The resulting framework emphasized that agents possess limited attention and must determine which consequences matter to a particular inquiry.
His moral philosophy linked questions about realism and relativism to explanatory practice. Instead of treating the dispute solely as an analysis of moral vocabulary, Harman examined whether moral facts occupied an indispensable role in accounts of observation and action. This method contributed to later debates over explanatory arguments in metaethics.
His work on situationism also altered the relationship between normative ethics and empirical moral psychology. Philosophical theories of character increasingly addressed the conditions under which stable traits could be measured and distinguished from context-sensitive patterns. Harman’s analysis formed one part of that development while remaining grounded in his broader naturalistic conception of philosophy.
Selected works
Harman’s principal books include Thought (1973), which examined the structure of cognition and the relationship between reasoning and language, and The Nature of Morality (1977), which developed his explanatory challenge to moral realism. Change in View (1986) presented his systematic account of belief revision, while Moral Relativism and Moral Objectivity (1996), written with Judith Jarvis Thomson, placed his relativist position in direct comparison with an objectivist theory.
His collected essays appeared in Reasoning, Meaning, and Mind (1999) and Explaining Value and Other Essays in Moral Philosophy (2000). Together, these volumes document the continuity between his theories of inference, his treatment of mental representation, and his analysis of moral explanation.