njnu_hupan
初级会员
想分析研究大学生口语错误，应该使用哪个语料库比较好？请各位高手指点。
[本贴已被 xujiajin 于 2005年09月15日 16时51分09秒 编辑过]
口语错误分析Error analysis in L2 learners" Eng Speech
- 主题发起人 njnu_hupan
- 时间
njnu_hupan
初级会员
thanks a lot! It looks that I have to wait for the publishing of SECCL.
njnu_hupan
初级会员
最近一直在看语料库的文献，所以头脑中有分析口语错误的想法，目前还在入门阶段，当然不能操之过急，希望各位大虾能不吝赐教，指点方向。
michelhere
初级会员
for syntactic errors, i guess, errors of spoken English would be pretty much the same with written English.
for lexical errors, there are two folds. One is inappropriate word choice and the other involves pronunciation.
I have been transcribing English speech of non-native speakers from different countries: e.g. Japan, Korea, China, Turkey, Spain, Italy
It seems to me that I can distinguish their background and native language by listening to their English speech. I guess this has something to do with the L1 transfer on both segmental level and prosodic level.
Also in terms of conjunction, there seems a pattern for non-native speakers, for example, they would stick to conjunctions like 'and' a lot, even where 'and' is necessarily needed.
In a nutshell, I think in terms of syntatic error and inappropiate word choice (which also involves both syntactic factor, morphological factor, semantic factor and pragmatic fator), erros occur in non-native speakers' English speech would more or less resemble those occurring in their written output.
In terms of pronunciation on segmental level and prosodic level, we can see a lot of L1 transfer effect. Take vowel as an example, vowel space in Chinese is different from that in English, learners would, in most cases, substitute an English vowel with a similar vowel in Chinese, e.g. many people cannot pronounce "big" correctly. because to them the only difference between 'i' in sheep and 'i' in big is merely the duration difference, but it is not the case.
for lexical errors, there are two folds. One is inappropriate word choice and the other involves pronunciation.
I have been transcribing English speech of non-native speakers from different countries: e.g. Japan, Korea, China, Turkey, Spain, Italy
It seems to me that I can distinguish their background and native language by listening to their English speech. I guess this has something to do with the L1 transfer on both segmental level and prosodic level.
Also in terms of conjunction, there seems a pattern for non-native speakers, for example, they would stick to conjunctions like 'and' a lot, even where 'and' is necessarily needed.
In a nutshell, I think in terms of syntatic error and inappropiate word choice (which also involves both syntactic factor, morphological factor, semantic factor and pragmatic fator), erros occur in non-native speakers' English speech would more or less resemble those occurring in their written output.
In terms of pronunciation on segmental level and prosodic level, we can see a lot of L1 transfer effect. Take vowel as an example, vowel space in Chinese is different from that in English, learners would, in most cases, substitute an English vowel with a similar vowel in Chinese, e.g. many people cannot pronounce "big" correctly. because to them the only difference between 'i' in sheep and 'i' in big is merely the duration difference, but it is not the case.
回复：口语错误分析Error analysis in L2 learners" Eng Speech
I do appreciate your work and/or perception in identifying the inappropriate inherent duration of some vowels. But, in addition to vowel length, we can understand L2 learners’ speech very well with the extended phonetic context.
In your view what are the most prominent error types of L2 speakers, in terms of phonetic and phonological features, that may cause great understanding problems to native speakers of English?
Some researchers at Nanjing University to my knowledge also use experimental phonetics (annotate certain features in L2 speech with Praat and export the results for statistical analysis or things like that) method to examine phonological problems with L2 learners’ speech.
I do appreciate your work and/or perception in identifying the inappropriate inherent duration of some vowels. But, in addition to vowel length, we can understand L2 learners’ speech very well with the extended phonetic context.
In your view what are the most prominent error types of L2 speakers, in terms of phonetic and phonological features, that may cause great understanding problems to native speakers of English?
Some researchers at Nanjing University to my knowledge also use experimental phonetics (annotate certain features in L2 speech with Praat and export the results for statistical analysis or things like that) method to examine phonological problems with L2 learners’ speech.
michelhere
初级会员
回复：口语错误分析Error analysis in L2 learners" Eng Speech
I do appreciate your work and/or perception in identifying the inappropriate inherent duration of some vowels. But, in addition to vowel length, we can understand L2 learners’ speech very well with the extended phonetic context.
In your view what are the most prominent error types of L2 speakers, in terms of phonetic and phonological features, that may cause great understanding problems to native speakers of English?
Some researchers at Nanjing University to my knowledge also use experimental phonetics (annotate certain features in L2 speech with Praat and export the results for statistical analysis or things like that) method to examine phonological problems with L2 learners’ speech.
As for vowel length, it's only an example of how difficult L2 learners acquire a proper L1 segmental feature. It doesn't necessaily lead to misunderstaning on listener's part.
Besides this famous /i:/ and /i/ difference, there still a lot distance between Chinese vowel space and English vowel space. The point is the inability to aqcuire the poper L1 segmental feature doesn't necessarily lead to inability to communicate with foreigners.
Because this whole process of communication involves two ends: producer and receiver.
On the recerver end, there are a bunch of factors we should consider, e.g. the degree of familiarity with certain foreign accent of a certain native speaker.
As for your question of the factors causing great understanding problems to native speakers of English, I would assume that major erros are more lexical and pragmatic.
As for the most prominent error types of L2 speakers in terms of phonetic and phonological features, it's really hard to quatify the degree of prominence. Like i said before, it involves both production end and perception end. Maybe on the production end, the distance between two sounds are distant, while perception doesn't distinguish them much. (The most relevant example is categorical perception of voice onset time of voiced and voiceless plosives in English) Therefore, the degree of prominence can be understood from both production perspective and perception perspective.
It's also interesting to see how native speaker acquire the ability to understanding a foreign accent. I guess it also involves categorical perception and reassociation between meaning and sound.
I would be extremely interested to see how researchers at Nanjing Univ. explore the segmental features with Praat and do the statitical analsysis.
Could someone from Nanjing Univ. grossly introduce about their reasearch on this.
Many thanks !
I do appreciate your work and/or perception in identifying the inappropriate inherent duration of some vowels. But, in addition to vowel length, we can understand L2 learners’ speech very well with the extended phonetic context.
In your view what are the most prominent error types of L2 speakers, in terms of phonetic and phonological features, that may cause great understanding problems to native speakers of English?
Some researchers at Nanjing University to my knowledge also use experimental phonetics (annotate certain features in L2 speech with Praat and export the results for statistical analysis or things like that) method to examine phonological problems with L2 learners’ speech.
As for vowel length, it's only an example of how difficult L2 learners acquire a proper L1 segmental feature. It doesn't necessaily lead to misunderstaning on listener's part.
Besides this famous /i:/ and /i/ difference, there still a lot distance between Chinese vowel space and English vowel space. The point is the inability to aqcuire the poper L1 segmental feature doesn't necessarily lead to inability to communicate with foreigners.
Because this whole process of communication involves two ends: producer and receiver.
On the recerver end, there are a bunch of factors we should consider, e.g. the degree of familiarity with certain foreign accent of a certain native speaker.
As for your question of the factors causing great understanding problems to native speakers of English, I would assume that major erros are more lexical and pragmatic.
As for the most prominent error types of L2 speakers in terms of phonetic and phonological features, it's really hard to quatify the degree of prominence. Like i said before, it involves both production end and perception end. Maybe on the production end, the distance between two sounds are distant, while perception doesn't distinguish them much. (The most relevant example is categorical perception of voice onset time of voiced and voiceless plosives in English) Therefore, the degree of prominence can be understood from both production perspective and perception perspective.
It's also interesting to see how native speaker acquire the ability to understanding a foreign accent. I guess it also involves categorical perception and reassociation between meaning and sound.
I would be extremely interested to see how researchers at Nanjing Univ. explore the segmental features with Praat and do the statitical analsysis.
Could someone from Nanjing Univ. grossly introduce about their reasearch on this.
Many thanks !
frankliang
普通会员
As error analysis is problematic in that it focuses on learners' errors rather than achievements, SECCL has not been error-tagged. I have no idea whether COLSEC has been error-tagged.
Focusing on errors does not seem to be a good idea.
Focusing on errors does not seem to be a good idea.