Cultural cartography with word embeddings

Using the frequency of keywords is a classic approach in formal analysis text, but has drawback glossing over relationality word meanings. Word embedding models overcome this problem by constructing standardized and continuous "meaning space" where words are assigned location based on relations similarity to other how they used natural language samples. We show embeddings commensurate with prevailing theories meaning sociology can be put task interpretation via two kinds navigation. First, one hold terms constant measure space moves around them--much like astronomers measured changing celestial bodies seasons. Second, also see documents or authors move relative it--just as ships use stars given night determine their location. empirical case immigration discourse United States, we demonstrate merits these broad strategies for advancing important topics cultural theory, including social marking, media fields, echo chambers, diffusion change more broadly.