There is strong evidence for syntactic priming in language production (Bock
, 1986), but little evidence about the time course of such effects. We repo
rt an experiment that examined the circumstances under which syntactic prim
ing decays in written language production. Participants completed sentence
fragments that allowed completions with one of two syntactic forms (Pickeri
ng & Branigan, 1998). They tended to produce the same syntactic form for im
mediately consecutive fragments, even though the two fragments described di
fferent events. However, when the experimental fragments were separated by
other fragments with unrelated syntactic forms, this tendency rapidly dimin
ished. The results suggest that priming effects in written production decay
rapidly when other structures are subsequently produced. We discuss the im
plications for the application of syntactic information during production.