1 How Much Can Understudied Languages Really Tell Us

The issue I want to raise today is not often publicly discussed, but I see it as central to future developments in our field. Let me introduce it by way of a naive question about syntax, my area of specialization. How much can understudied languages really tell us about how syntax, in general, works? For much of its history, generative syntactic theory has been guided by the assumption that universal grammar is very tightly constrained. On some views, it is so tightly constrained that all syntactic differences among languages can be reduced to differences in their lexical items. If one adopts this perspective, one might expect languages to differ little, if at all, in what they can tell us about syntactic theory: all languages should be extraordinarily revealing, and equally revealing. Close investigation of any one of them should provide a reasonably direct pathway to an explanatory theory. And if all languages are equal in what they can reveal about syntactic theory, one might expect research on a broad spectrum of languages to have contributed significantly and productively to the theory’s development. In fact, research on syntax has not exactly proceeded along this path. The reality i