A Propositional Analysis Approach To The Assessment Of Information Fidelity In English-Chinese Consecutive Interpreting

Information fidelity assessment has been a central concern of both interpreting studies and interpreter education,of which scholars have exerted great efforts in scholarly quest.Despite the enthusiasm and endeavors thrown in the assessment of information fidelity,it is still subject to controversies and subjectivity,partly due to a lack of standard assessment unit.The present investigation,adopting a proposition-analysis(PA)approach,draws on references from Semantics,Language Testing and Interpreting Studies to validate,both theoretically and empirically,PA serving as an assessment tool of information fidelity in English-Chinese consecutive interpreting and to explicitate its applications and usefulness in interpreter education.Based on a comprehensive review of literature on interpreting quality assessment,the dissertation takes on PA in Semantics and proposes an analytical framework which offers theoretical constructs for descriptive and explanatory purposes in the present study.In addition,the dissertation discusses the potential use of PA in interpreter training and orients further research with regard to the inquiries into interpreting assessment and interpreter training,in an attempt to identify its significance and prospects in the field of interpreting studies.To validate PA in the assessment of information fidelity,the author conducts an experiment in which 44 MTI trainees take part in a simulated E-C consecutive interpreting test.With that done,three raters are hired to score examinees'performance of the message transfer twice,one with the PA-based criterion and the other with a conventional criterion.In addition to construct validity that is involved in the discussion of theoretical and analytical framework,validation procedure of the PA-based assessment also involves collecting empirical evidence from the perspectives of criterion validity and rater reliability.To that end,intra-rater and inter-rater reliability estimates,rater consistency and rater effects of the PA-hased criterion and the conventional criterion are all accounted for by means of SPSS analysis and multi-faceted Rasch analysis.The major findings are:Firstly,proposition proves to be a feasible and practical information assessment unit in interpreting.And PA is valid and reliable in fidelity assessment.Statistical analysis reveals that PA-based scores highly correlate to scores based on conventional assessment sheet.Furthermore,PA performs better in terms of scoring behavior.When assessing with PA,raters are probable to achieve better outcomes in some key indicators of criterion validity and rater reliability,for instance,intra-rater reliability,inter-rater reliability,rater consistency and rater effects.The better outcomes could be partly due to the standard PA procedure and partly due to the stringent rater training.Secondly,a macro analysis of propositional correspondence in the database shows that the top three types in frequency are Complete Match,Omission and Meaning-Alteration,which illustrates a general picture of examinees' performance in message transfer.Thirdly,correlation analysis reveals that Complete Match and Omission are the two key indicators of information fidelity in interpreting performance,as their frequencies closely correlate to fidelity scores.CM positively correlates to fidelity scores,indicating the stage of competence development.Omission,on the contrary,highly correlates to fidelity scores in a negative way,implying the most serious drawbacks in interpreters' performance.Triangulated with examinees' responses,it is found that Omission is,to a large extent,due to serious failure in information reception and processing,an outcome of deficiencies in bilingual knowledge and procedural knowledge.It is proposed that instructors take the problem seriously and throw concentrative efforts and intensited instructions into enhancing examinees'language proficiencyFourthly.PA is conducive to detecting competence trajectory of examinees.It is discovered that upper-end examinees and lower-end examinees use omissions in different ways.Upper-end examinees use omissions as a conscious strategy,which implies better development of interpreting competence.On the other side,lower-end examinees use omission passively,mainly due to information loss at reception and processing stages.Apart from enhancing language proficiency,instructors can help by increasing trainees' meta-cognition awareness.Finally,PA has the potential to be incorporated into the instructions of information processing and peer review.Instructors can incorporate PA into their instructions of information processing,whereby improving trainees' awareness of active listening and note taking.In addition,when incorporated into peer review,PA could facilitate accurate diagnosis and constructive feedbacks.In addition to its advantages,some disadvantages of the PA-based criterion have also been detected,such as complicated procedure that exhausts raters after many hours of rating and its over-concentration on micro-information.However,these demerits could be mitigated or alleviated if we adopt appropriate measures.The present study reveals that PA is a feasible and valuable resource conducive to interpreting performance assessment and interpreter education.By presenting a systemic description of assessment device to raters or instructors,this dissertation could contribute to better understanding and operation of interpreting performance assessment in information fidelity,both theoretically and empirically,which could be a stepping stone towards machine assessment in the future.Besides,findings and discussions of this investigation could offer insights for interpreting practices,didactics and interpreting studies.