Voice-specific information and the 20-second delayed-suffix effect.

Two experiments were conducted to investigate the nature of the delayed-suffix effect reported by Watkins and Todres (1980). In both experiments, subjects were presented lists of digits for serial recall. At the end of each list either a tone or a voice reading aloud the word go was presented. The voice was either in the same voice that read the digits or in a different voice. Past research has indicated that the tone control produces the least interference, followed by the different-voice suffix, which in turn produces less interference than the same-voice suffix. The results of both experiments indicated that when subjects were tested on immediate recall for the lists, the typical ordering of tone control, different-voice suffix, and same-voice suffix on recall at the last serial position was found. However, when there was a 20-s filled interval between the last list item and the suffix, there was only a difference between the tone control and the same- and different-voice suffixes with no difference between the latter two conditions. In addition, Experiment 2 failed to support a simple attentional account of the differential influence of voice in the immediate and delayed-recall conditions. The results are viewed as supporting a perceptual tuning mechanism in which perceptual specificity decreases with the passage of time.