In an effort to pinpoint both the character of culture and the object of anthropology, Clifford Geertz writes that "[t]he culture of a people is an ensemble of texts, themselves ensembles, which the anthropologist strains to read over the shoulders of those to whom they properly belong" (1973a:452). This paper considers the conceptual, ethnographic, ethical and methodological implications of the influential metaphor of culture as “text,” and of fieldwork as “reading.” In Morocco, one of Geertz’s two long-term field sites, the population comprises both native Arabic and Berber speakers, and large segments of the population, especially rural women, are not schooled or literate. Their rich expressive culture is primarily oral. In my ethnographic and linguistic field research beginning in 1995, I have found that given the two factors of multilingualism and low levels of schooling among Soussi Berbers, literacy (usually in Arabic) is more than just one practice among others, or a set of skills to be acquired. Instead, an individual’s engagement with “texts”–-religious or secular–-indexes unevenly distributed social and symbolic capital. This distribution of capital shapes hierarchical relations within the family, neighborhood, and community, according to which, in part, an individual‚s rights and responsibilities are organized. More specifically, reading and writing tasks–-creating and interpreting “texts,” whether the Quran, a school textbook, paperwork for an identity card, or the instructions in a box of medication-–fall to some individuals, who then are relieved of other (usually manual) tasks. When all of culture is text, how does one do ethnography of literacy and oral expressive practices, and of their resulting written and oral texts? Which of the publicly-available practices that Geertz considered to comprise culture constitutes the data available for the anthropological interpretation of such social practices? The paper draws on my own and others‚ fieldwork in Morocco to reconsider central tenants of Geertz‚s contributions for the anthropology of language, and to locate these contributions within the history of ideas shaping ethnographic research more generally.