Umm... With Transformers? Insights from Filled Pause Use across Four Slavic Parliaments

2026-07-07

AI Summary

This study addresses the limitations of prior research on filled pauses, which has largely relied on small-scale monolingual corpora and lacks systematic cross-linguistic analysis—particularly for Slavic parliamentary discourse. Leveraging approximately 4,000 hours of parliamentary speech data across four Slavic languages, the work combines Transformer-based automatic detection with Generalized Estimating Equations (GEE) incorporating Mundlak correction to disentangle within- and between-speaker effects. It investigates how speaker age, gender, speech rate, emotion, political affiliation, and power status influence filled pause rates. The findings reveal, for the first time in a multilingual Slavic parliamentary context, a positive association between filled pause frequency and emotional intensity, a language-specific gender effect that contradicts prior literature, and consistently lower filled pause rates among opposition members compared to government coalition speakers, with the moderating role of political identity varying across parliaments.

Abstract

Filled pauses (FPs) are a universal feature of spontaneous speech, yet most studies rely on small, single-language corpora, limiting the generalisability of their findings. We analyse ~4,000 hours of parliamentary speech across four related Slavic languages (Croatian, Czech, Polish, Serbian). FP occurrence is obtained via transformer-based automatic detection, while FP rate is modelled using Generalised Estimating Equations (GEE) with Mundlak correction to distinguish within- from between- speaker effects. We replicate a negative association of age and speech rate with FP rate, but find that gender effects are language-specific and directionally opposite to most prior literature. Novel analyses of sentiment, political orientation, and power status reveal a consistent positive association between sentiment and FP rate, alongside parliament-specific modulation by orientation and power status, with opposition speakers tending toward lower FP rates than governing coalition speakers.

Problem

• filled pauses
• spontaneous speech
• cross-linguistic
• parliamentary speech
• generalisability

Innovation

• Transformer-based detection
• filled pauses
• multilingual parliamentary speech
• Generalised Estimating Equations
• sociolinguistic variation