Elizabeth G. Brunner
 

In the justice system, language is often classified as evidence. This is especially true when anonymous audio recordings are involved; in these circumstances, forensic linguist are consulted to provide expert opinions on the evidence. The forensic linguist must then scrutinize the recording(s), and is often asked to determine if the voice of the unknown speaker is the same as the voice of a suspect. The difficulties of identifying an unknown speaker are compounded when that speaker deliberately disguised their voice in the original recording. One way that criminals can disguise their voice is by faking a different accent. The present study examines the role of accents in disguise. Speakers in the Philadelphia, Pennsylvania area are presented with an artificial extortion threat and are recorded reading it in their normal voice, in a disguised voice, and with two accents. Initial results show that approximately half of the speakers choose to imitate a Southern accent when disguising their voice. Very few speakers choose to imitate accents that are from geographically closer regions, such as New York.
            A phonetic analysis of the faked Southern accents will assess both disguise success (the extent to which the speaker changes and presumably conceals their typical linguistic features) and imitation accuracy (the extent to which the speaker employs phonetic features of a Southern accent). The investigation of disguise success may reveal specific phonetic features that can be used in determining which aspects of an unknown speaker’s voice are disguise attempts and which features reflect the speaker’s normal voice. This knowledge could be of great value to forensic linguists who are involved in casework with disguised voices. Imitation accuracy will explore what non-Southern speakers regard as Southern phonetic features and consider whether these features are results of the speakers’ perceptual experiences or stereotypes of Southern speech.