Jorge Botana, Guillermo deOlmos Albacete, RicardoMartínez Mingo, AlejandroOlivas Osuna, José JavierMartínez Huertas, José Ángel2024-05-202024-05-202023-12-132474-736Xhttps://doi.org/10.1080/2474736X.2023.2287036https://hdl.handle.net/20.500.14468/12586This article presents a quantitative method for mapping semantic spaces and tracing political frames’ trajectories, that facilitate the analysis of the connections between changes in ideas and socio-political phenomena. We test our approach in Spain, where the Catalan conflict fostered a competition in terms of decontestation of meanings of key political concepts. Using unsupervised machine learning, we track the salience, level of semantic fragmentation and fluctuations in meanings of 216 frames in the two largest Spanish newspapers, El País and El Mundo, throughout 8 years. This is achieved via the extraction, vectorization, and comparison of over 70,000 words. We apply Latent Semantic Analysis, an innovative methodology for the alignment of semantic spaces, and new institutional theory. Our exploratory study suggests that the evolution of many nationalism-related frames resembles a punctuated equilibrium model, and that political events in Catalonia, acted as critical junctures, altering the meanings reflected in the Spanish press.enAtribución-NoComercial-SinDerivadas 4.0 Internacionalinfo:eu-repo/semantics/openAccessartículoPolitical framesnew institutionalismmethodologyCatalan political conflictLatent Semantic Analysis