This chapter sets the stage for future empirical work on the linguistic relativity hypothesis. First, the various lines of research reviewed in the first six chapters are briefly characterized as to their overall orientation and most significant problems and contributions. Second, in light of this past research, an evaluation is made of the most promising available approaches and research directions. This evaluation provides the rationale for the design of the specific empirical investigation described in Grammatical categories and cognition.

Overview of past empirical research

Whorf's research served as the historical point of departure for our review, since most contemporary empirical research on the linguistic relativity hypothesis has been stimulated by it. His work can also serve as the substantive point of departure, since it still represents the most adequate empirical approach to the issue among the studies reviewed. Subsequent research efforts have characteristically omitted or altered crucial elements of his approach even as they introduced improvements in some areas. Many of these changes stem more from traditional disciplinary preferences than from any substantive necessity.

Boas, Sapir, and Whorf

Boas put forward the basic claim that languages implicitly classify experience in diverse ways for the purposes of speech. Sapir developed these ideas further by suggesting that these language categories, organized as a coherent system, could shape a person's view of reality. Whorf, in turn, clarified the operation and significance of the systematic interrelations among language categories and provided the first empirical evidence of effects on thought.