Why are universals of grammar easier to explain than geographical patterns?

Since the 1920s, geographical patterns of grammatical properties have been discussed occasionally, and since the 1990s, the discussion of large-scale areality has become an important part of comparative linguistics (“areal typology”). This development came a few decades after universals research took off (Greenberg 1963), and nowadays one gets the impression that geographical patterns are studied even more widely than universal patterns. However, linguists seem to lack a good explanatory framework, and there are many puzzles. For example, why are some features strongly areal (e.g. tone, word order), while others show little areality (e.g. gender, future tense)? Geographical patterns are explained by earlier contacts, but the histories of most areal patterns seem obscure. By contrast, grammatical universals are often explainable by functional factors, especially universals of asymmetric coding and universals of word order. Language universals have often been said to be due to diachronic change and thus history as well (e.g. Bybee 1988; Anderson 2016), but in this talk, I will suggest that universals are easier to explain because of their ahistoricity.