Paradigms of Spatial Statistics as Footnotes to Tobler’s First Law of Geography

Tobler’s First Law of Geography (TFL) is foundational to spatial analysis. Responding to Sui’s (Citation2004) invitation to consider whether geography is just a footnote to TFL, we develop a five-axis framework to compare major spatial statistical paradigms. Four axes cover spatiotemporal interconnectedness, distance and nearness, heterogeneity and scale, and temporal dynamics, with a methodological layer on estimation, inference, and learning. We examine five paradigms in depth: geostatistics and Gaussian random fields, lattice and spatial econometrics, Moran eigenvector spatial filtering, geographically weighted regression, and geospatial artificial intelligence and spatial learning. The framework also extends to additional paradigms. This synthesis supports an expanded TFL in which spatial relatedness varies with distance, connectivity, context, scale, and time, and highlights implications for spatial statistical modeling in the artificial intelligence era.