Lexical recognition memory across dialects.

Abstract

Implicit recognition memory for spoken words is more accurate when words are repeated by the same or a similar talker than when they are repeated by a different talker. The current study explored implicit recognition memory for words repeated by the same talker, by a different talker from the same dialect, and by a different talker from a different dialect in a word recognition task in noise. Repetitions produced by the same talker facilitated word recognition performance. However, for target words originally produced by talkers from the Northern dialect of American English, repetitions produced by a talker from the Midland dialect of American English inhibited word recognition performance. No repetition effect was observed for repetitions produced by a different talker from the same dialect or for words repeated by Northern talkers that were originally produced by Midland talkers. These results suggest an asymmetry in how indexical information is stored and activated in lexical processing. The same talker repetition effect was observed for both dialects, but one variety (the Northern dialect of American English) inhibited the repetition effect across dialects and the other variety (the Midland dialect of American English) did not.