Advanced

Imitation, Line-up selection and Semantics

Eriksson, E.; Kügler, F.; Sullivan, K. P. H.; van Doorn, J and Zetterholm, Elisabeth LU (2003)
Abstract
Recent studies have shown that the ability to recognize an imitated voice is affected by topic and familiarity with the person being imitated. These studies have used professional imitations of a well-known politician. In one of the training passages is a political text and in the other a text explaining how to bake a cake. The signal detection methods used in these recent studies gave no specific information about the distribution of non-matching positive responses (false alarms) amongst the voices used as distracters (foils). In the studies, the listeners were instructed to respond “yes” when they heard the training voice and “no” when the voice they heard was not the training voice. This paper considers the distribution of the... (More)
Recent studies have shown that the ability to recognize an imitated voice is affected by topic and familiarity with the person being imitated. These studies have used professional imitations of a well-known politician. In one of the training passages is a political text and in the other a text explaining how to bake a cake. The signal detection methods used in these recent studies gave no specific information about the distribution of non-matching positive responses (false alarms) amongst the voices used as distracters (foils). In the studies, the listeners were instructed to respond “yes” when they heard the training voice and “no” when the voice they heard was not the training voice. This paper considers the distribution of the listeners’ “yes” responses between the imitated target voice, the natural voice of the imitator and four foils. The distribution of responses from four different groups is investigated. Groups One and Two reported knowledge of the target voice, whereas Groups Three and Four reported no such knowledge. Groups One and Three heard the political text as their training passage and the other groups the baking passage. In none of the distributions is the imitator’s natural voice a popular selection. In distribution One, foils 3 and 4 represent 8% and 7% of the yes selections respectively; the imitated voice represented 75% of the yes selections. In distribution two, these foils represented 12% and 23% of the yes selections, in distribution three, 10% and 23%, and in distribution four, 9% and 34%. This paper asks the question why is this foil 4 selected more frequently than the imitator’s own voice. Auditory and acoustic analyses will be reported in an attempt to explain the listeners’ selection behaviour. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Contribution to conference
publication status
unpublished
subject
language
English
LU publication?
yes
id
2933fa7d-ea0f-4893-a10c-f155a6d3a13e (old id 529218)
date added to LUP
2007-09-17 11:58:06
date last changed
2016-04-16 10:43:26
@misc{2933fa7d-ea0f-4893-a10c-f155a6d3a13e,
  abstract     = {Recent studies have shown that the ability to recognize an imitated voice is affected by topic and familiarity with the person being imitated. These studies have used professional imitations of a well-known politician. In one of the training passages is a political text and in the other a text explaining how to bake a cake. The signal detection methods used in these recent studies gave no specific information about the distribution of non-matching positive responses (false alarms) amongst the voices used as distracters (foils). In the studies, the listeners were instructed to respond “yes” when they heard the training voice and “no” when the voice they heard was not the training voice. This paper considers the distribution of the listeners’ “yes” responses between the imitated target voice, the natural voice of the imitator and four foils. The distribution of responses from four different groups is investigated. Groups One and Two reported knowledge of the target voice, whereas Groups Three and Four reported no such knowledge. Groups One and Three heard the political text as their training passage and the other groups the baking passage. In none of the distributions is the imitator’s natural voice a popular selection. In distribution One, foils 3 and 4 represent 8% and 7% of the yes selections respectively; the imitated voice represented 75% of the yes selections. In distribution two, these foils represented 12% and 23% of the yes selections, in distribution three, 10% and 23%, and in distribution four, 9% and 34%. This paper asks the question why is this foil 4 selected more frequently than the imitator’s own voice. Auditory and acoustic analyses will be reported in an attempt to explain the listeners’ selection behaviour.},
  author       = {Eriksson, E. and Kügler, F. and Sullivan, K. P. H. and van Doorn, J and Zetterholm, Elisabeth},
  language     = {eng},
  title        = {Imitation, Line-up selection and Semantics},
  year         = {2003},
}