Self-monitoring And Self-repair Patterns In Consecutive Interpreting

This dissertation attempts at a systematic study of self-monitoring and self-repairpatterns in consecutive interpreting (CI) with the theoretic framework ofpsycholinguistics, in particular language production theory. Drawing on theself-monitoring theories of L1and L2speech production, this dissertation proposes anintegrated model of CI processing, which is the first of its kind to incorporate theself-monitoring mechanism into a CI processing model. An in-depth analysis of theoperation of the self-monitoring mechanism in CI processing is made in relation withother components that constitute the entire CI process, namely, comprehension,memory, note-taking, production and coordination. The dissertation goes on toinvestigate the overt manifestation of the self-monitoring mechanism in CI, viz.self-repair patterns by analyzing selected data that contains1004speech errors and222self-repairs identified in a students’ CI examination performance corpus. Theself-repair patterns analyzed in students’ CI examination performance are comparedto those in L1and L2speech, which sheds light on the different cognitive processesinvolved in CI and lends support to some hypotheses related to the proposal of theintegrated model of CI processing.Research on interpreting processing indicates that expertise calls forself-monitoring skills and self-awareness and maintaining a well-balanced distributionof mental resources among the different processes in interpreting is one of the mainhurdles to overcome when learning how to interpret. Compared with abundantresearch on self-monitoring and self-repair patterns in L1and L2speech, there ishardly any study on self-monitoring and self-repair in CI, resulting in a long neglectedaspect of the interpreting studies. Based on Levelt’s L1speech production model andKormos’s adapted L2speech production model, combined with a full account of thecomponents characteristic of CI processing, an integrated model of CI processing isproposed in this study and summarized as follows:First, the preverbal message for CI production originates from the sense that has been comprehended, retained, and retrieved by the interpreter. The conceptualizerresponsible for generating the preverbal message in L1and L2speech is replaced bythe interaction of the working memory modules and the comprehension module.Second, the long term memory module proposed in Kormos’s L2speech modelis retained in the integrated CI processing model since CI also involves L2at work.The subcomponent of encyclopedic knowledge is included in the long term memory,as interpreting requires the mobilization of existing and prepared theme knowledgethat helps realize the comprehension and production of the non-linguistic aspects.Third, the proposed model places the monitor in LTWM.Fourth, the proposed model includes three monitoring loops—the preverbal loop,the internal loop and the external loop. The preverbal loop checks the preverbalmessage against the sense comprehended from the LTWM. The internal loopcompares the planned interpreting with the sense comprehended from the LTWM andthe language standard. The external loop compares the uttered interpreted speech withthe sense comprehended from the LTWM and the language standard.The corpus analysis of self-repair patterns in students’ CI examinationperformance yields the following results:First, nearly all types of self-repairs classified in L1and L2speech productionare identified in students’ CI examination performance. Repair of error of information(EI) is newly added to the taxonomy. This categorization of types lends support to thehypothesized three monitoring loops in the proposed integrated model of CIprocessing–preverbal loop, the internal loop and the external loop.Second, the much lower correction rate of errors and inappropriacies by thestudents partially testifies to the hypothesis that the more cognitively complicated taskof CI and the yet-to-be automatized use of their ‘B’ language result in fiercercompetition of attention, which leads to less attentional resource for self-monitoring.Third, the students’ most frequently repaired errors and inappropriacies areerrors of lexical choice, information, phonological realization, and inappropriacies oflexical choice and information, while those of lowest correction rates are errors oftense, syntax, and inappropriacies of ambiguity. This distribution pattern suggests that students weigh accuracy of semantic meaning over form.Fourth, the students tend to interrupt erroneous or inappropriate words earlierthan L1speakers probably due to1) less automatized and more controlled use of theirtarget language in L2and2) specially allocated attention to accuracy of delivery.Fifth, a much lower proportion of self-repairs signaled by editing terms has beenfound in students’ CI examination performance than in L1speech, which is explainedby the finding that students in the CI performance do not solely use editing termsbefore self-repairs, but also silent pauses. The reason is probably the conscious effortby the students to lower the disfluencies that the use of editing terms might cause. Themost frequent editing term in students’ CI examination performance turns out to be‘Uh’, which is the same with Levelt’s finding in L1speech repair.Sixth, the finding that instant repairs and anticipatory retracings are the two mostfrequent ways of restarting for error and inappropriateness repairs in students’ CIexamination performance. The higher frequency of these two ways of restarting in thestudents’ performance indicates a higher degree of conservativeness of error andinappropriateness repairs in students’ CI examination performance.This study of self-monitoring and self-repair patterns fills the gap of the lack of acomprehensive model of CI that addresses the intrinsic difficulties of thelanguage-switching task, which also complements the existing theories ofself-monitoring and self-repair theories in L1and L2speech production. Thepedagogical implications of this study for CI training is that by drawing the students’attention to the self-monitoring process and the three phases of self-repairs, namely,detection and interruption of errors, using editing terms, and restarts of the repairproper, the students are given the opportunity to raise their awareness of the processesthey are carrying out while interpreting, recognize their weak aspects reflected bytheir self-repair patterns during interpreting and engage in specific training.