Abstract

Some questions are posed concerning issues to investigate when testing the hypothesis that semantic priming contributes to change, especially unidirectionality in change. They include extent of prior context, similarity of semantic primes, and univerbation effects. Ways in which priming could be used to test hypotheses about grammaticalization are suggested in connection with the development of auxiliaries, and the obligatorification of subject markers. Finally, the necessity of distinguishing motivations and mechanisms is outlined.