4th International Conference on Spoken Language Processing
Philadelphia, PA, USA
Native and non-native listeners identified English words and sentences in six different listening conditions. When they heard speech mixed with noise or when they had to use linguistic knowledge to respond, non-native listeners suffered greater performance decrements than native listeners. Their performance appears to be data-driven ('bottom-up') requiring full specification of the acoustic-phonetic information relevant for selecting a particular word.
Bibliographic reference. Bond, Z. S. / Moore, Thomas J. / Gable, Beverley (1996): "Listening in a second language", In ICSLP-1996, 2510-2513.