The chronicler Garcilaso de la Vega, it may be recalled, was concerned with how Christian ideas might be "adequately" expressed to Peruvian Indians. Garcilaso had an interest in good translation. Translation is always motivated. In Garcilaso's case, there was not only the hope of benefiting the souls of the Indians, but also concern for minimizing stresses stemming from the incorporation of the conquered into a new order. Garcilaso himself traced roots to both the Indians and the Spaniards. His father was a conquistador, and his mother was an Inka noble, a second cousin to Atahuallpa.

When Spanish interpreters of "these times," Garcilaso wrote, wish to express Christian ideas adequately, "they have to seek new words or phrases, or use with great care suitably dignified expressions in the old language or else lay hands on the many words the cultured and scholarly Indians have taken from Spanish and introduced into their own languages, adapting them to their own ways of speech" (1988: 682).

Garcilaso's solution is to coin new terms and expressions, to use with great care possible correspondences (glosses) across languages, and to adapt loan words. These are conventional instruments of translation in a narrow sense, and they can be productive where the translator is sensitive and skilled.

Although the solution endorsed by Garcilaso may serve for purposes of religious conversion (or, at any rate, the appearance of religious conversion), it is inadequate for anthropological purposes. If by "translation" we mean translation in the narrow sense sketched above, then the anthropologist, I think, should subsume translation in explication. Indeed, in my opinion the task of the ethnographer is explication rather than "translation" in the narrow sense of glossing expressions in one language with terms from another or with freshly minted neologisms. Such explication involves the examination of contexts in which targeted expressions occur and the analysis of any encountered polysemy. Our intellectual grasp and appreciation of key terms will be enhanced by an understanding of the domains with which they are associated in native usages.

A well-known example of explication is found in Evans-Pritchard's (1956) discussion of the Nuer term kwoth. I am aware that some anthropologists suggest that Evans-Pritchard's explication of that term is biased by his personal religious proclivities. I do not know enough about the case, however, to evaluate that claim. Regardless of possible inaccuracies or other deficiencies in the contents of what he writes, however, the form of his explication deserves admiration on two counts. First, Evans-Pritchard examines how the term is used in different contexts, and even though his examination may not be exhaustive, he demonstrates the polysemy of kwoth, and he makes efforts to deal with it in ways that we, his readers, may comprehend. Second, his explication is alive to the significance of tropes; and although his fame in that regard largely rests on his analysis of the "twins are birds" metaphor, he alerts us generally to how a sensitivity to tropes might expand our understandings of religious terms and expressions, and so enlarge and potentially improve the translation task.