With this article I would like to propose a phenomenological argument to sustain the idea evoked by several authors about anthropology having an interpretative character. This position may be defined as a set of two theses. The first thesis enunciates that culture can be read as a text (or texts). The second one admits that ethnography is a text (or texts) about texts. The first assumption defines the interpretative character of culture; the second one alludes to the descriptive character of ethnography.