Disease and geography are related domains f o r Tojolabal-Maya. Using multidimensional methods, we compare two domains: (1) individual cognitive "maps" f r o m duease terms and (2) hand-drawn maps, both with one another and with an official topographic map. Multivariate study of individual informant data demonstrates correspondence of the axes of maps. Least squares f i t ling of dimensional representations using a method specqically modijed f o r ethnosemantic data allows meaningful comparisons both among and within informants, and with an aggregate f r o m a related survey of 33 informants as well. These multivariate operations help integrate individual data, sampled simultaneously f o r several domains, tasks, and occasions, with aggregate data. For semantic domains, we achieved rapprochement bet ween psychological and anthropological approaches. [disease, folk theories, ethnosemantics, cognition, multivariate, Tojolabal-Maya]DISEASE AND GEOGRAPHY ARE RELATED DOMAINS for the Tojolabal. a Mayan Indian group of about 30,000 living in the southern part of Chiapas, Mexico.' Using multidimensional scaling, we study the two domains. We compare individual cognitive "maps" from disease terms and hand-drawn geographic maps, both with one another and with an official topographic map.In studying the cognition of disease among the Tojolabal, our concern has been to learn the meanings of disease terms to the native. This problem holds both practical and theoretical interest: practical because modem health care delivery systems ate now being extended to the Tojolabal zone, and their acceptance is sometimes problematic; theoretical because this work bears on a classic strand of investigation in cognitive anthropology, testing for psychological reality.Cognitive anthropologists have often proceeded along noncomparative lines. On the other hand, cross-cultural psychologists have been explicitly comparative, usually searching for Western principles among non-Western societies, using written instruments, and often finding inferior performance among non-Western peoples (Ember 1977). On occasion, psychologists have worked with inadequate understanding both of the Western bias in their tests and of the larger cultural setting beyond the test situation. Cognitive anthropologists, on the other hand, more often collect data orally rather than in the written form favored by psychologists. Anthropologists emphasize domains well known to the informant; psychologists focus more often on learning novel tasks. Further, cognitive an-LOUANNA FURBEE is Arrociate Profemor. Department of Anthropolcgy, 210 Switrler Hall, University of M i i u r i , Columbia.