Earticle

초록

영어

This study tests whether sentence-level source text syntactic complexity predicts time-normalised disfluency in English–Korean simultaneous interpreting. Complexity is indexed by mean dependency distance (MDD); disfluency comprises silent pauses (≥1.0 s), filled pauses, repetitions, self-repairs, and other events, reported as rates (events·min⁻¹). Six graduate student interpreters produced sentence-aligned outputs, with sentence × interpreter as the observation unit. Analyses used a negative-binomial GLMM with a duration offset, complemented by Kruskal–Wallis tests across MDD tertiles. Results showed that higher MDD predicted higher total disfluency rates (IRR = 1.20, 95% CI [1.09, 1.32], p < .001), with only silent pauses and repetitions exhibiting reliable increases across complexity groups. These findings support treating disfluency as a rate-sensitive indicator of structural load rather than a categorical defect. Methodologically, the study demonstrates a tractable workflow for sentence-level alignment and incidence-rate modelling. Pedagogically, it motivates difficulty calibration by MDD and targeted practice to manage pausing and repetition in student interpreter training.

목차

Abstract
1. Introduction
2. Literature review
2.1 Functional roles and perceptions of disfluency
2.2 Characteristics of student interpreter performance
2.3 Syntactic complexity as a factor in interpreting performance
2.4 Material selection and difficulty calibration in interpreter training
3. Methods
3.1 Data collection
3.2 Data preparation
3.3 Data analysis
4. Results
4.1 Total disfluency rate under equalised time
4.2 Type-specific patterns under equalised time
5. Discussion
6. Conclusion
References

저자

• Eunjin Heo [ Jeju National University / Graduate School of Interpretation and Translation / Lecturer ]

참고문헌