《言语感知的联结主义模型》由会员分享，可在线阅读，更多相关《言语感知的联结主义模型（5页珍藏版）》请在金锄头文库上搜索。
1、2009年 9月第 30卷第 5期null null null null null null null null null null null null null null null外语教学Foreign LanguageEducationnull null null null null null null null null null null null null null nullSept. 2009Vo.l 30 No. 5 l 崔null 刚, 杨null 莉(清华大学外语系北京 100084)K null 1 : 言语感知是语言理解的起始阶段, 是指听者把连续的语音辨认为单词的过程b
2、本文主要针对联结主义心理语言学对于言语感知的相关研究进行综述与评价,包括语流的切分a音位的识别和单词的辨认三个部分,最后又对相关研究的发展趋势进行了简要的讨论b1 o M : 言语感知; 联结主义; 模型 m s | : H 030 null null D S M : Anull null c I | : 1000null5544( 2009) 05null0001null05Abstract: A s the in itial period of language com prehension, speech perception refers to the process in wh ic
3、h the listeners recnullognize the words from con tinuous speech. In this paper, the author review s the m ajor connection ist m odels of speech percepnulltion, including the segmentation of speech, the recogn ition and d iscrim ination of phonemes and words. A discussion about thetrends of the relev
4、ant stud ies is also m ade in the final section.K ey words: speech perception; connection ist; m odelsnull null S , P O M V b , 1 M s # M O Z bV 20W 80M , l $ W . ! ! J. S , 1960( 4). 10 null . B 0 s J. , 1958( 3). 11 d l . !3 l y y J. C , 2008( 2):121null130. 12 d l . q M ! Z. S ) . , 2007. 13 null
5、 . !3 E J. , 1954( 4). 14 y . !3 A . q ? E C. 2005. 15 . 1 !3 ! ) J. , 1959( 5). 16 null . a& a J. , 1954( 8 ). 17 null . S C E M . : 6 , 2000. 18 null . q N M . : , 2003. 19 , m . 5 3 !3 ) 5 J. , 1959( 5). 20 f .q E M M . Z : Z , 1959. 21 u . u ( B ) M . : 6 , 2002. :本文为作者承担的国家社科项目语词构块式音义象征关系理论研究(
6、( 07BYY001)和教育部人文社科项目英汉语词构块式音义象征关系对比研究( (教社科司 2006 170号)的部分研究成果b感谢国家社科基金和教育部社科基金对本研究所提供的资助!T e :d l , 2 v S q , p V , p V 3 = , Z _ b收稿日期null null 2008null09null23责任编校null null ( 4: ) 14 Norris, D. & M cQueen, J. M. Shortlist B: A Bayesian model of connulltinuou s speech recogn ition J. Psychologica
7、l Review, 2008, 115( 2):357null395. 15 Norris, D. Shortlist: A connection istm odel of continuous speech recognulln ition J. Cognition, 1994, 52: 189null234. 16 Norris, D. The m erge mode:l Speech p erception is bottomnullup J.Journal of theAcousticalSociety ofAmerica, 1999, 106( 4 ): 613null640. 17
8、 Norris, D. , Perceptual learn ing in speech J. CognitivePsychology,2003, 47: 204null238. 18 Pison,i D. B. & Luce, P. A. A cousticnullphonetic rep resentations inword recogn ition J. Cognition, 1987( 25): 21null52. 19 Protpapas, A. Connection istm odeling of speech perception J. PsynullchologicalBul
9、letin, 1999, 125 ( 4): 410null436. 20 Rum elhart, D. E. , H inton, G. E. & W illiam s, R. J. Learn ing internullnal rep resentations by error propagation A . In J. L. M cC llelland &D. E. Rum elhart ( ed s. ). ParallelDistributedProcessing ( vo,l 1) C.Cam bridge, MA: M IT Press, 1986. 21 Saffran, J.
10、 R. , Aslin, R. N. & New port, E. L. Statistical learn ing by8nullmon thnullold in fan ts J. Science, 1996, 274: 1926null1928. 22 S eebach, B. S. Eviden ce for the developm en t of phonetic prop erty d enulltectors in a neural net w ithout innate know ledge of lingu istic structu re D . Unpublish ed
11、 doctoral d issertation, B rown Un iversity, Provinulldence, R I. 1990. 23 Selk irk, E. O. On the m ajor class featu res and syllable theory A. InM. Aronoff& R. T. Oehrle ( eds). LanguageSound Structure C .Camb ridge, MA: M IT Press, 1984. 24V room en, J., van den Boshch, A. , and deG elder, B. A co
12、nnection istm odel for bootstrap learn ing of syllabic stru cture J. Language andCognitiveProcesses, 1998, 13( 2 /3): 193null220. 25W atrous, R. L. Phonem e d iscrim ination u sing connection ist n etw orks J. Journal ofAcousticSociety of America, 1990, 87( 4): 1753null1772. 26 anull . l J. , 2006( 2).T e :a ,b v q ,p V , p V 3 = , Z _ b ,b v = , V , Z _ b收稿日期null null 2008null12null31责任编校null null null9null
