Second International Conference on Spoken Language Processing (ICSLP'92)
Banff, Alberta, Canada
Memory for spoken words and speaker's voices was assessed in two experiments. In the first experiment, continuous recognition memory was superior for words repeated in the same voice as their original presentation than for words presented in one voice and later repeated in another. Tin's same-voice recognition advantage was consistent up to delays of 3-4 minutes between first and second presentations. In the second experiment, discrete recognition memory and implicit memory were examined. The explicit recognition memory test replicated the same-voice advantage observed in continuous recognition memory, but the advantage disappeared over time. The implicit memory test, however, showed a same-voice advantage in perceptual identification that was robust over time, lasting up to a week. Taken together, the results suggest that perceptually detailed, episodic memories of spoken words are retained in memory. These episodes affect conscious recognition decisions in the short-run, and they affect online spoken word recognition for indefinite periods thereafter.
Bibliographic reference. Goldinger, Stephen D. / Palmeri, Thomas J. / Pisoni, David B. (1992): "Words and voices: perceptual details are preserved in lexical representations", In ICSLP-1992, 591-594.