Conversational Feedback in Scripted versus Spontaneous Dialogues: A Comparative Analysis

TL;DR

This study compares communicative feedback in scripted subtitles and spontaneous dialogues across multiple languages, revealing significant differences in feedback frequency and types, with implications for training conversational NLP models.

Contribution

The paper provides a comprehensive quantitative analysis of feedback phenomena in scripted versus spontaneous dialogues, highlighting key linguistic differences across languages and genres.

Findings

Feedback is less frequent in subtitles than in spontaneous dialogues.

Subtitles contain a higher proportion of negative feedback.

LLM-generated dialogues resemble scripted dialogues more than spontaneous ones.

Abstract

Scripted dialogues such as movie and TV subtitles constitute a widespread source of training data for conversational NLP models. However, there are notable linguistic differences between these dialogues and spontaneous interactions, especially regarding the occurrence of communicative feedback such as backchannels, acknowledgments, or clarification requests. This paper presents a quantitative analysis of such feedback phenomena in both subtitles and spontaneous conversations. Based on conversational data spanning eight languages and multiple genres, we extract lexical statistics, classifications from a dialogue act tagger, expert annotations and labels derived from a fine-tuned Large Language Model (LLM). Our main empirical findings are that (1) communicative feedback is markedly less frequent in subtitles than in spontaneous dialogues and (2) subtitles contain a higher proportion of negative feedback. We also show that dialogues generated by standard LLMs lie much closer to scripted dialogues than spontaneous interactions in terms of communicative feedback.