Louis Dumont has the unique advantage of combining field experience, large ideas, and a basic knowledge of Sanskrit. The patience to master a difficult archaic language, no longer the spoken vehicle of current thoughts in any defined society, is very rare in a scholar devoted to analysis of actual social groups, real people. Dumont is the only person to acquaint himself with traditional societies in India and with the cultural continuity within which they, and their neighbors, are intelligible as a unit. The large and daunting literate past of India is open to him; and there are few who can compete with this combination of skills. Nevertheless his study of the relationship between king and brahmin in traditional and (so far as is relevant) modern Indian society—a study found principally in one essay—impressed me in two ways, which, in association, did not sustain admiration. His ideas need to be placed alongside another interpretation of the facts, in order that a clearer perspective may be obtained. It is not the case that his reading of archaic and ambiguous materials is wrong; we do not yet know enough about the history of India to say that. But there is a real chance that the talent to generalize from the particular (which is one of the strengths of anthropologists) has here run a little ahead of itself. My aim is not to discourage the use of Sanskritic materials, still less to imply that there are “real experts” in competition with whom the anthropologist who emulates Dumont's efforts must remain silent. I value the stimulus Dumont's ideas have offered; and I know they will continue to excite response and encourage further research, even when they arouse suspicion. Let us attempt a new evaluation of the facts, and set that beside Dumont's findings.