In addition to these speaker-general results reported by McGarr (1983), several of the studies that we have discussed here found evidence for speaker-specific learning effects, with improvements primarily or exclusively when test items were spoken by the same talker who had been heard in the training phase.