A Corpus-based Study On Patterns Of Spoken English

Based on the ICE-GB corpus,this thesis aims at analyzing the patterns of spoken English of ICE-GB.The study first investigates the sentence length of both the spoken and written components of ICE-GB,and then moves to the analysis and interpretation of the major sentence patterns,particularly from the perspectives of high frequency sentence patterns,inversion patterns,reduced sentence patterns and so on.Finally,the similarities and differences between the spoken and written components of ICE-GB are analyzed.Through a systematic study,the following conclusions are drawn.Firstly, sentences in spoken language are shorter than written language.The mean sentence length of the spoken component of ICE-GB is 10.71,which is shorter than that of the written component of ICE-GB - 17.70.Then,the number of sentence patterns of the spoken component of ICE-GB is larger than that of the written component;the compositions of the top 40 sentence patterns are almost the same.In addition,in the spoken component of ICE-GB,SVO is used most often;whereas in the written component of ICE-GB,SVA is used most often.It shows that spoken English is more direct and informative.Different from written language,spoken language does not prefer certain patterns;it uses various types of sentence patterns and each of them occupies a small percentage of the total number of sentence patterns.There are 71 types of shared sentence patterns between the spoken and written components of ICE-GB,and the shared sentence patterns occupy a large proportion of the total sentence patterns respectively.Further,there are spoken-only and written-only sentence patterns,but they both occupy very small percentages of the total sentence patterns respectively.