The Relative Relevance of Conceptual and Acoustical Information in Similarity Judgments of Animate and Inanimate Sounds

Recent studies using behavioral measures of similarity between acoustic events have had interesting results, but have been inconclusive with regards to the basis on which event-class discrimination is made. In the present investigation, we suggest that much of the difficulty has resulted from confusion about the basis on which participants make similarity judgments. Judgments may well be made on the basis of acoustic similarity, but there may also be a strong effect of a priori conceptual knowledge in participants’ assessments of similarity. It is unclear whether these influences can be separated through biasing. Also, it is possible that participants have a default scheme for similarity when no biasing instructions are given. Finally, it has been suggested that different bases underlie the processing of sounds with animate vs. inanimate sources, which might influence behavioral measures of similarity. We have devised an experiment involving similarity judgments of both acoustical and verbal-label stimuli, using animate and inanimate stimuli, in an attempt to answer these questions. Our results suggest that participants are more highly affected by conceptual knowledge when judging animate compared to inanimate stimuli. Also, biasing was shown to increase the difference between acoustic similarity judgments and similarity judgments of verbal stimuli.