Exploring the Impact of Task Complexity on Disfluencies and Their Problem Sources in Consecutive Interpreting

This study investigates task complexity’s impact on novice interpreters’ fluency in consecutive interpreting (CI) and further examines the patterns of disfluencies and their underlying problem sources. While fluency is extensively researched in psycholinguistics, its application to CI - a specific context of speech production - remains limited. This study bridges this gap by providing a systematic account of interpreters’ cognitive processes in CI and comprehensively analyzing the interplay between task complexity, disfluencies, and problem sources.

The study draws on theories and models from several fields, including speech production, interpreting, and task complexity, to develop an analytical framework for examining the cognitive processing involved in CI. With this theoretical basis, it addresses three core issues: 1) the impact of task complexity on the cognitive processing in CI, as reflected by variations in the duration of unfilled pauses, and whether an optimal pause threshold can be identified to effectively signal problem activity in CI, 2) the impact of task complexity on the distribution patterns of four types of disfluencies - unfilled pause, filled pause, repetition, and self-correction, and 3) the underlying problem sources associated with these different types of disfluencies.

A total of 55 novice interpreters participated in this study, performing two Chinese-to-English CI tasks with different levels of complexity and taking part in retrospective interviews to report the problems underlying their disfluencies. Both their performances and interviews were recorded, transcribed, and analyzed. Disfluencies were categorized into unfilled pauses, filled pauses, repetitions, and self-corrections, and a taxonomy of problem sources was applied to identify the underlying issues. The collected data were analyzed using descriptive statistics, inferential statistics, and qualitative analysis.

The key findings reveal that task complexity significantly influences the patterns of unfilled pause duration (threshold ≤ 600ms), with the more complex task leading to longer pauses. The study identifies 500ms as the optimal pause threshold for indicating problem activity in CI. Task complexity also influences the distribution of certain types of disfluencies, particularly unfilled and filled pauses, though to a limited extent. Moreover, interpreters exhibit distinct patterns of attentional resource allocation and problem sources across different stages of speech production in CI. Finally, based on an integrative analysis of the core issues above, a model is proposed to provide a structured account of interpreters’ cognitive processes in CI.

This study makes key contributions by systematically examining the impact of threshold settings on disfluency patterns and successfully experimenting with task complexity control through specific metrics. The findings have implications for both research and interpreter training and assessment. Future research could refine methodologies to explore the interaction between indicators of and beyond fluency, task complexity, and interpreting performance. Additionally, this study suggests the potential for developing more comprehensive, standardized assessment criteria for training and assessment purposes.

Keywords: consecutive interpreting, task complexity, disfluency, problem sources