Home > Positioning > Persons > Harman

Gilbert Harman (1938–2021)

Harman gave moral relativism its sharpest modern statement — not as the loose claim that customs differ, but as a semantic thesis about what moral judgements mean: an “ought” is always implicitly relative to an agreement or framework the speaker shares, so a judgement made from inside one moral framework has no purchase on an agent who stands outside it. He is equally known for the explanatory argument against moral realism — that moral facts, unlike physical ones, do no work in explaining our observations — and for coining “inference to the best explanation.” He taught at Princeton for over half a century.

Gilbert Harman (1938–2021). American philosopher, a student of Quine at Harvard, who spent his entire teaching career at Princeton from 1963. He worked across ethics, epistemology, philosophy of mind, and philosophy of language, and co-founded (with George Miller) the Princeton programme in cognitive science. His most influential books are Thought (1973), The Nature of Morality (1977), and Explaining Value (2000); the 1996 volume Moral Relativism and Moral Objectivity, a debate with Judith Jarvis Thomson, sets his relativism against a defence of objectivity.

Key concepts

Moral relativism as a semantic thesis. In “Moral Relativism Defended” (1975), Harman argues that “ought” judgements of a certain kind — the ones that bear on what an agent has reason to do — are made relative to an implicit agreement among a group. To say a person morally ought to do something is to say so relative to a framework of intentions and agreements the speaker and agent share. Where no such shared agreement exists, the judgement loses its footing: one can say a hardened outsider’s conduct is evil without its being true that he ought, in this inner sense, to act otherwise. The relativity is built into the logical form of the judgement, not added as a sociological caveat.

The explanatory argument against moral facts. In The Nature of Morality, Harman poses the question of whether we ever need to posit moral facts to explain what we observe. A scientist sees a vapour trail and must posit a proton to explain it; a person sees hoodlums set a cat alight and judges it wrong — but the wrongness need play no part in explaining the judgement, which is fully accounted for by the observer’s upbringing and moral sensibility. Moral facts, on this test, are explanatorily idle in a way physical facts are not, and their idleness is a reason to doubt they exist. This became a central argument in the realism/anti-realism debate.

Inference to the best explanation. Harman named and analysed the pattern of reasoning in which a hypothesis is accepted because it would, if true, best explain the evidence. The idea is older, but his formulation made it a standard term in epistemology and philosophy of science, and it underwrites his own explanatory argument in ethics.

Character and situationism. Drawing on social psychology, Harman argued that ordinary talk of robust character traits — honesty, courage as stable dispositions — may rest on a mistake: behaviour is far more sensitive to situation than trait-talk assumes, and the evidence for cross-situationally consistent character is weak. This put him among the philosophers pressing a situationist challenge to virtue ethics.

Where Harman stops

Harman’s relativism is a thesis about one class of “ought” — the reason-implying, agreement-relative kind. He grants that other moral language survives outside any shared framework: we can still call an outsider’s acts evil, dangerous, or to be resisted. Critics press that this concedes much of what the objectivist wanted — that the evaluations which do cross frameworks are carrying the weight the relativised “ought” was meant to bear — and dispute whether the semantic thesis, once these exceptions are admitted, marks as deep a break with objectivity as it announces. The scope of the relativity, and how much it actually withdraws, is contested in his reception.

The explanatory argument has drawn a standing reply that its test is loaded. Critics (Nicholas Sturgeon among them) argue that moral facts do enter our best explanations — that a society’s injustice explains its unrest, or a person’s cruelty their acts — and that Harman’s test, applied evenhandedly, does not single out moral facts as uniquely idle. Whether moral facts are explanatorily dispensable is left as a live dispute, not a settled result, by the argument that raised it.

Key works

See also: Ethics · Quine · Blackburn · Wong