Toward Machine Interpreting: Lessons from Human Interpreting Studies

Current speech translation systems, while having achieved impressive accuracies, are rather static in their behavior and do not adapt to realworld situations in ways human interpreters do. In order to improve their practical usefulness and enable interpreting-like experiences, a precise understanding of the nature of human interpreting is crucial. To this end, we discuss human interpreting literature from the perspective of the machine translation field, while considering both operational and qualitative aspects. We identify implications for the development of speech translation systems and argue that there is great potential to adopt many human interpreting principles using recent modeling techniques. We hope that our findings provide inspiration for closing the perceived usability gap, and can motivate progress toward true machine interpreting. Feature Description Example Temporal immediacy Produces interpretation in real-time. Maintains 1-2 seconds ear-voice-span. Spatial immediacy Operates in proximity of speaker & audience. Shares stage with speaker. Multimodality Uses visual or gesture cues when available. Refers to chart while speaker points. Free/diverse actions Dynamically adapts to any situation. Adapts translation approach to content type. Interaction/influence Acts as a independent agent when needed. Requests clarification, improves acoustics. Intent translation Interprets what is meant, not what is said. Interpretation conveys hidden accusations. Interpreter uncertainty Maintains trust by signaling own uncertainty. "Speaker may have said 'revenue'." Speaker errors Indicates or corrects unintentional speaker errors. Corrects "million" to "billion" in context. Adaptation/explanation Adapts or explains culture-specific expressions. "Break a leg!" → "Good luck!". Explicitation Explicitates logic, intent, order, viewpoints. "..., according to X's view." Brevity Keeps sentences short and clear. "The results were strong. More tests needed." Rhetoric quality Delivers exceptionally high rhetoric quality. Adapts style to particular audience. Pleasant experience Works reliably; pleasant voice; eye contact. Avoids hectic speech when falling behind. Cognitive ergonomics Minimizes audience stress and fatigue. Avoids complex language.