Text analysis is the use of systematic methods for studying texts. Early versions go back thousands of years to the study of religious texts, but modern techniques — often called natural language processing, or NLP — bring computational methods to bear on language so that meaning itself can be quantified and measured at scale. Several broad approaches do this work. Topic modeling, most commonly Latent Dirichlet allocation (LDA), extracts any number of topics from a corpus; its strength is flexibility, its weakness that you rarely know in advance how many topics you should pull, though parameter sweeps and coherence scores help find an optimum. Frequentist feature analytics count words against hand-built dictionaries to measure how often an idea recurs, and more advanced forms employ ontologies that also encode how minds think about those terms — the approach behind the 📝Pythia ontology I use at 📝CulturePulse. Semantic network analysis represents words, concepts, and their relationships as complex networks, a lineage running from Quillian in the 1960s through Carley in the 1980s, and one I have integrated with 📝multi-agent AI to study the emergence of culture. Narrative and database analysis, which I defined in the 2018 Encyclopedia of Anthropology, quantifies cultural data longitudinally and straddles the boundary between qualitative and quantitative work — though quantification can sacrifice the rich semantic meaning it sets out to capture.