An Analysis Of EVS Unit In Simultaneous Interpreting

The present study was designed to explore the EVS in SI,or the lag time between comprehension of the ST and reformulation of the TT.Since simultaneous interpreters store the received information in their STBS which is subject to duration and capacity limits,determining when to start TL rendition and how to segment the source input becomes important.Short lags can reduce cognitive load,but may lead to errors in output.Long lags presumably aid working memory as they afford larger chunks of input to operate on,but can lead to displacement of items from storage and result in omissions in output,because of assumed limitations in capacity of the buffer store.Grasping an ideal EVS unit before rendition enables interpreters to limit the output chunk to a size smaller than the input chunks already received,and thus to avoid errors and omissions.The paper reported three small-scale experiments on EVS unit during E-C SI and C-E SI.Three second-year MTI students and one interpreting teacher in Hubei University participated in the experiments,acting as research subjects and the speaker respectively.By recording the SI practice of student interpreters and collecting the corpus,the thesis aimed at investigating into the length,syntactic structure and segmentation strategies of the EVS units in C-E and E-C SI.After adjusting the starting points of the wave files in each experiment to be the same,they were transcribed through the software Adobe Audition,with time calibrated whenever pauses occur.On the basis of major constituent and pauses,ST and TT can be split up into numerous segments.The length of each EVS unit was then calculated by subtracting the timing when target segment occurred from the timing when the corresponding source segment occurred,followed by a descriptive analysis by SPSS 13.0.By matching the source segment and its corresponding target segment,every EVS unit could be found out,followed by a qualitative analysis of its syntactic structure and segmentation strategies(if the EVS unit contains one source segment,the strategy used will be identity.If it includes less than or more than one source segment,the strategy used will be fission or fusion).Dozens of examples are also given to illustrate how the above analysis has been carried out,and how to use Gerver's Information Processing Model to account for errors and omissions in TT from the perspective of the memory load imposed by source segments.The research findings show that:Firstly,the EVS was found to be longer in C-E SI than in E-C SI,which was attributable to "articulatory suppression" effect.In other words,most TL output coincides with SL input,preventing the rehearsal of information already stored in STBS.The ensuing negative impacts are greater in SI from L2 to L1 than the other way around.Secondly,the higher frequency of EVS unit with a SVO structure in E-C SI than in C-E SI is because in Chinese,there are many long left-branching NPs which act as subject,and long PPs that are interpolated between subject and verb.Interpreters' memory capacity will probably be taxed if they wait till the verb occurs.Thirdly,fusion was used more frequently in C-E SI than in E-C SI,because Chinese input chunks are shorter and each one contains less information than its English counterpart does.If the information value is low,interpreters can maintain a greater EVS or fuse several input chunks,so as to gather the information needed to start interpreting,such as the keyword.Fission was used more frequently in E-C SI than in C-E SI,because there are lots of long embedded clauses in English,and each segment contains more information than Chinese segment does.If interpreters wait too long for a complete processing of one segment,lagging far behind the speaker,the processing of the following sentences might be impeded.The experimental results may be helpful in interpreter training.Trainees should be aware of the possible consequences of long and short EVS,take appropriate strategies to regulate the EVS,and engage in more memory training and attention-sharing activities.