Identifying Dr Schneider's voice: An adventure in forensic speaker identification. Künzel, H. J Forensic Linguistics, 3(1):146-154.
abstract   bibtex   
Difficulties associated with accurate speaker identification are explored in a case study involving the arrest of Juergen & Claudia Schneider, who had been wanted in Germany for 13 months, & the identification of two audiotaped voices. Two phoneticians working independently agreed that the audiotaped male voice matched a voice in reference samples from TV interviews, although the recording situations were very different; one expert also noticed a particular type of slight lisp present in both unidentified & reference samples. Recognition by people (N = 4) familiar with the Schneiders was more problematic; assertions ranged from positive identification of the Schneiders to certainty that the speakers were other people. After the Schneiders' arrests, their voices on the tape were confirmed. Factors contributing to the possible bias in the lay testimony include Ss' emotional states & personal attitudes, & semantic content & type of speech material. Implications for forensic voice identification emphasize strategic use of witnesses, distinguishing linguistically useful from unhelpful voice descriptions, preparing voice lineups whenever time permits, & using effective technical analysis.
@article{kunzel_identifying_1996,
	Author = {Künzel, Hermann J},
	Date = {1996},
	Date-Modified = {2016-09-24 18:56:08 +0000},
	Journal = {Forensic Linguistics},
	Keywords = {forensic, forensic phonetics},
	Number = {1},
	Pages = {146-154},
	Title = {Identifying Dr Schneider's voice: An adventure in forensic speaker identification},
	Volume = {3},
	Abstract = {Difficulties associated with accurate speaker identification are explored in a case study involving the arrest of Juergen \& Claudia Schneider, who had been wanted in Germany for 13 months, \& the identification of two audiotaped voices. Two phoneticians working independently agreed that the audiotaped male voice matched a voice in reference samples from TV interviews, although the recording situations were very different; one expert also noticed a particular type of slight lisp present in both unidentified \& reference samples. Recognition by people (N = 4) familiar with the Schneiders was more problematic; assertions ranged from positive identification of the Schneiders to certainty that the speakers were other people. After the Schneiders' arrests, their voices on the tape were confirmed. Factors contributing to the possible bias in the lay testimony include Ss' emotional states \& personal attitudes, \& semantic content \& type of speech material. Implications for forensic voice identification emphasize strategic use of witnesses, distinguishing linguistically useful from unhelpful voice descriptions, preparing voice lineups whenever time permits, \& using effective technical analysis.}}
Downloads: 0