Integrating linguistic and geographic methods in toponymic analysis

Toponyms offer valuable clues regarding historical language geographies, but toponymic analysis has often suffered from methodological shortcomings. In particular, post-hoc semantic claims, unmotivated formal correspondences, and lack of systematic corroboration undermine the reliability of many proposals. This paper outlines two modes of toponymic research — form-to-geography and geography-to-form — and introduces four simple, replicable methods for reducing false positives. These methods draw upon the dual character of toponyms as both geographic and linguistic data as well as the quantitative analysis of large datasets. We apply these methods to two languages of the Central Andes: first Aymara, and then Puquina, a poorly documented language whose toponymic record is uncertain. A clear Puquina footprint is identified around Lake Titicaca. We advocate moving beyond the treatment of toponyms as isolated etymological puzzles and focusing also on their spatial patterning. Such an approach emphasizes reproducibility, scalability, falsifiability, and empirical transparency, while remaining complementary to traditional qualitative methods. 1