ESHS/HSS Annual Meeting

RAND's Linguistic Drive: Computerizing Anthropology in 1960s Chiapas

• In Event: Cybernetic Anthropologies: Midcentury Redesigns of a Technomodern Field

English Abstract

This paper examines how the RAND Corporation – a militarily-minded, para-state nonprofit organization – became involved in linguistic and anthropological research in Chiapas, Mexico in the early 1960s. Across the previous decade, RAND had developed an elaborate, multi-campus complex of research projects on mainframe computers' capacities for "machine translation" (MT). The focus of early MT projects was to automate translations of Russian for Cold War ends, addressing scientist Warren Weaver's post-World War II call to turn computers into what historian Jacqueline Léon calls “powerful war machines.” But, by 1960, the grand ambitions of MT were dissolving into a more variegated field of projects in "computational linguistics." A core group of MT researchers working at the RAND campus in Santa Monica began to turn their attention to seemingly-esoteric or theoretical projects. Among these projects was linguistic and anthropological research on Tzeltal-speakers in highland Chiapas. Anthropologists Terrence Kaufman and Duane Metzger trained a key ethnographic informant to conduct, tape-record, and transcribe Tzeltal-speakers' recent experiences of illness. RAND researchers key-punched and processed this data free of charge, generating "concordances" defining Tzeltal linguistic patterns. The success of this project motivated anthropologists at Harvard University and the University of Chicago to intensify emerging computer-based research, an early cornerstone of scientifically-minded cognitive anthropology. In turn, RAND walked away from its support of anthropological fieldwork with a database of 500,000 running words in Tzeltal. I ask how this linguistic drive served and shaped the interests of RAND and, by extension, the US state.

Author

• Matthew Watson, Mount Holyoke College