Who Speaks in the News? Measuring Voice through Automated Quote Extraction.

Quotes are central to journalism and a valuable resource for media analysis. Yet, they remain underexplored in large-scale empirical research, due to challenges in automatic extraction. We argue that quotes offer valuable opportunities to revisit core questions on the mediated public sphere – provided an extraction strategy exists that measures up to concrete social science research problems. To this end, we introduce QED (Quote Erat Detectendum), a lightweight, open-source, multilingual tool that comprehensively identifies direct, mixed and indirect reported speech triplets – composed of a source, a statement, and a cue – and attributes them to speakers. QED vastly improves on classic techniques and outperforms generative LLMs in flexibility, granularity, and scalability. Applying QED to French political news, we demonstrate its analytical value for contemporary social scientists via three illustrations in political sociology, analyzing the voice of political actors in news content.

Authors

Publication Details

Journal
SocArXiv (OSF Preprints)
Published
2026-11-12
DOI
https://doi.org/10.31235/osf.io/5zpfb
Primary Topic
Computational and Text Analysis Methods
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Who Speaks in the News? Measuring Voice through Automated Quote Extraction.

Étienne Ollion, Guilhem Sicard, Emma Bonutti D'Agostini
SocArXiv (OSF Preprints)
Computational and Text Analysis Methods
preprint

Who Speaks in the News? Measuring Voice through Automated Quote Extraction.

Étienne Ollion, Guilhem Sicard, Emma Bonutti D'Agostini
preprint en

Abstract

Quotes are central to journalism and a valuable resource for media analysis. Yet, they remain underexplored in large-scale empirical research, due to challenges in automatic extraction. We argue that quotes offer valuable opportunities to revisit core questions on the mediated public sphere – provided an extraction strategy exists that measures up to concrete social science research problems. To this end, we introduce QED (Quote Erat Detectendum), a lightweight, open-source, multilingual tool that comprehensively identifies direct, mixed and indirect reported speech triplets – composed of a source, a statement, and a cue – and attributes them to speakers. QED vastly improves on classic techniques and outperforms generative LLMs in flexibility, granularity, and scalability. Applying QED to French political news, we demonstrate its analytical value for contemporary social scientists via three illustrations in political sociology, analyzing the voice of political actors in news content.

SocArXiv (OSF Preprints)
Computational and Text Analysis Methods
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon