Abstract
Normal hearing listeners judged loudness differences between two complex speech sounds, one consisting of "n" consonant-vowel (CV) pairs each spoken by a different talker and one consisting of "2n" CV pairs. When n was less than four, listeners' judgments of loudness differences between the two sounds was based on the level of the individual CVs within each sound, not the overall level of the sounds. When n was four or more, listeners' judgments of loudness differences between the two sounds was based on the overall level of the two sounds consisting of n or 2n CVs.