Parallel-integrated Semantic Processing Expetise In Simultaneous Interpreting

It is observed that there are many differences in simultaneous interpreting(SI)performance between professional and student interpreters.The difference in the approaches to semantic processing on the discourse level is further highlighted after a careful literature review.Indicating respectively the semantic routes and semantic processing strategies,the concept of “parallel-integrated semantic processing expertise” is put forward.“Parallel processing” here refers to not only the parallel activation of both “de-verbalization” and “transcodage” as semantic processing routes,but also to the holistic,interactive and distributed activation among factors of influence in the decision-making process.“Integrated” processing here refers to the integrated strategies in SI semantic processing and the coordination among cognitive resources.Centered on the SI parallel-integrated semantic processing expertise,research questions are raised,following relevant theoretical and empirical exploration.There are four research questions.Firstly,how to find an appropriate theoretical framework to explain the phenomena of parallel-integrated semantic processing in SI? Secondly,what are the major factors that influence SI semantic processing? Thirdly,what indices and strategies can be adopted to indicate parallel-integrated semantic processing? And finally,what are the significant differences that distinguish professional interpreters from novice interpreters and why? The first three research questions are addressed through theoretical construction by adopting such research methods as literature analysis and inductive reasoning,while the fourth research question is solved through experiments assisted with questionnaire and qualitative analysis.In the experiment,there are three independent variables: professional vs.student interpreters,directionality and the delivery speed in the source discourse,and dependent variables include the ratio of realization of total and essential information units,parallel semantic route indices and integrated semantic strategies.Parallel-Distributed Processing Model(Rummelhart & McClelland,1986)is referenced as the major theoretical support,on which SI Parallel-integrated Semantic Processing Model is built.The model emphasized the parallel activation of the both “de-verbalization” and transcodage routes and the strategic decision making process: under the interactive internal and external factors,the interpreter makes integrated and coordinated strategic decisions,such as semantic reconstruction,integration and identifying essential and secondary information.With SI Parallel-Integrated Semantic Processing Model as the theoretical framework along with the theoretical support from Revised Hierarchical Model(Kroll & Stewart,1994)and Long-Term Working Memory(Ericsson & Kintsch,1995),the paper further examines the major factors that influence SI semantic processing—working language competence,directionality,language combination,long-term working memory and source text difficulty and establishes indices for “parallel” semantic processing: source language structural interference,lexical interference,semantic reconstruction,and indices for “integrated” semantic strategies: semantic reconstruction,integration,generalization,approximation,summary,coherent addition and explanatory addition.Based on the solutions to the first three research questions,the fourth research question is addressed through empirical study.Under the expert vs.novice expertise paradigm,experiments consisting of C-E and E-C simultaneous interpreting with normal and fast delivery speeds were conducted over eight professional interpreters,all of whom had received formal conference training,and seven second-year student interpreters who were about to graduate from a conference interpreting program.After the experiments,a questionnaire about the language acquisition background,training or practice of SI into B and difficulty rating about the EC and CE experiments was sent to each interpreter.The empirical study yields the following the results.First of all,there are significant differences between professional and student interpreters in terms of the ratio of the realization of both total and essential information units,but no significant differences in the choice or the frequency rate of semantic strategies.The data further show that it is not the choice or the frequency rate of semantic strategies that counts,but the extent of automation of strategies,the coherent logical construction on the holistic discourse level and the related coordination of cognitive resource that matters.In addition,professional interpreters are more likely to adopt parallel semantic processing routes and produce more structural interferences than student interpreters,as runs counter to the ordinary conception that increasing expertise leads to higher degree of de-verbalization.Secondly,the influence of directionality finds expression in the differences between A and B working language competence.“Translation Asymmetry”(Kroll & Stewart,1994)is proved to be valid in our experiments: in SI into A,structural and lexical interferences occur significantly more frequent than that into B;while in SI into B,interpreters adopt more “meaning-based” semantic route and strategies,including semantic reconstruction,integration,explanatory addition,etc.,not to say better coherence in holistic discourse meaning and coordination of cognitive resources.Thirdly,it is shown that speed of delivery,as an important index of source text difficulty,has no significant effect on parallel semantic route indices,which is also against the statement that “the more difficult the source text,the more de-verbalized the products in SI” concluded by researchers like Dam(2001).Finally,regardless of the group differences between professional and student interpreters,relevant parallel route indices and integrated semantic processing strategies are shown to be significantly relevant to the ratio of the realization of total and essential information units,thus supporting the SI Parallel-Integrated Semantic Processing Model constructed in this study on the whole.