Politics of Questions in News: A Mixed-Methods Study of Interrogative Stances as Markers of Voice and Power

Victor Bros, Matilde Barbini, Patrick Gerard, Daniel Gatica-Perez · 2026

Read Politics of Questions in News: A Mixed-Methods Study of Interrogative Stances as Markers of Voice and Power by Victor Bros, Matilde Barbini, Patrick Gerard, Daniel Gatica-Perez (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-21823.txt.

Introduction

Bros and co-authors study the questions that appear inside French digital news stories. Questions in news are rarely counted apart from statements. The team processed over one million articles from 2023 and 2024, detecting questions, typing their role, and finding answers in the text. A hand-coded subcorpus grounds the measures in semantic theory. Most questions open or organize an issue, and most get answered later in the same story, usually in the journalist's own voice. Named people and groups dominate these passages, while broad publics appear far less.

Catalog record

CREATORVictor Bros, Matilde Barbini, Patrick Gerard, Daniel Gatica-Perez
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE95383
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.21823

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Politics of Questions in News: A Mixed-Methods Study of Interrogative Stances as Markers of Voice and Power Victor Bros, Matilde Barbini, Patrick Gerard, Daniel Gatica-Perez Abstract Interrogatives in news discourse have been examined in linguistics and conversation analysis, but mostly in broadcast interviews and relatively small, often English-language corpora, while large-scale computational studies of news rarely distinguish interrogatives from declaratives or differentiate their functions. This paper brings these strands together through a mixed-methods study of the ”Politics of Questions” in contemporary French-language digital news. Using over one million articles published between January 2023 and June 2024, we automatically detect interrogative stances, approximate their functional types, and locate textual answers when present, linking these quantitative measures to a qualitatively annotated subcorpus grounded in semantic and pragmatic theories of questions. Interrogatives are sparse but systematically patterned: they mainly introduce or organize issues, with most remaining cases being information-seeking or echo-like, while explicitly leading or tag questions are rare. Although their density and mix vary across outlets and topics, our heuristic suggests that questions are overwhelmingly taken up within the same article and usually linked to a subsequent answer-like span, most often in the journalist’s narrative voice and less often through quoted speech. Interrogative contexts are densely populated with named individuals, organizations, and places, whereas publics and broad social groups are mentioned much less frequently, suggesting that interrogative discourse tends to foreground already prominent actors and places and thus exhibits strong personalization. We show how interrogative stance, textual uptake, and voice can be operationalized at corpus scale, and argue that combining computational methods with pragmatic and sociological perspectives can help account for how questioning practices structure contemporary news discourse. 1 Introduction The digital transition has reshaped how news is produced, financed, and encountered. Audiences increasingly consume news online and via platforms, while legacy outlets face intensified competition for attention and pressure on business models ( 26 ; 45 ) . In many countries this has contributed to ”news deserts”, where local reporting capacity erodes and communities lose access to robust, professional journalism, with documented democratic harms ( 1 ; 48 ; 39 ; 4 ) . Against this backdrop, questions about who and what is made visible in news, whose problems are problematized, and how publics are addressed acquire renewed democratic significance ( 35 ) . These democratic roles are enacted not only through topic selection and framing, but also through interactional practices such as questioning and challenging. In broadcast settings, conversation-analytic work has shown how question design ranges from deferential information-seeking to highly adversarial, accountability-pressuring formats ( 20 ; 10 ; 25 ) . Linguistics and pragmatics emphasize that interrogatives are context-dependent acts that can request information, highlight issues, embed presuppositions, or function rhetorically ( 3 ; 28 ; 47 ; 7 ) . Most of this work relies on close analysis of relatively small, often English-language corpora. Computational studies of rhetorical questions and stance in online debate ( 37 ; 34 ) show that some pragmatic distinctions can be modeled automatically, but remain limited in genre and scale. Large-scale NLP studies of news, in turn, focus mainly on topics, sentiment, or frames and usually treat all sentences alike, without distinguishing interrogatives from declaratives or separating different interrogative functions. French-language news provides a particularly interesting setting for bridging these strands. It spans multiple media systems and political contexts, including France, Switzerland, Belgium, francophone Canada, Senegal, and pan-African outlets, with distinct histories of press freedom and journalistic cultures ( 24 ; 43 ) . Yet French functions as a shared linguistic resource, enabling partly common repertoires of questioning and address while leaving room for local variation in how power and publics are constructed. This combination makes francophone news a useful testbed for our approach: it moves beyond the Anglophone focus of much prior work while keeping a single language and covering both Global North and Global South media systems. We bring a mixed-methods, computational lens to what we call the Politics of Questions in this fragmented but interconnected media space. Using more than 1.2 million French-language digital news articles (January 2023-June 2024), we automatically identify interrogative stances in context, approximate their functional types, locate textual answers when they are present, and characterize the actors and places that populate questions and answers. These quantitative components are linked to a qualitatively annotated subcorpus, grounded in semantic and pragmatic theories of interrogatives, which we use for both evaluation and discourse analysis. Focusing on interrogative stances allows us to make the politics of visibility and accountability empirically tractable. Questions are a particularly salient subset of journalistic moves: they mark issues as open, allocate epistemic authority, and can be detected and typed from surface cues in ways that most other interactional practices cannot. At the same time, they are normatively charged, because asking or not asking, and how a question is framed, can shape which actors are foregrounded, which problems are rendered discussable, and how accountability is textually staged. This leads to the following research questions: RQ1 : How prevalent are interrogative stances in French-language news, and how do their types vary across outlets, countries, and topics? RQ2 : When journalists pose interrogative stances, to what extent are those questions textually answered, and how often do answers involve quoted external voices versus self-answers by the journalist? RQ3 : Which kinds of actors and social groups are most often foregrounded in interrogative contexts, and how frequently do such contexts include quoted external voices rather than only journalistic narration? RQ4 : How can a triangulated methodology, combining quantitative and qualitative analyses, enhance the interpretation and validity of the quantitative measures? Contributions. Our contributions are threefold: (i) conceptually, we extend semantic and pragmatic work on questions to written news by operationalizing interrogative stances , answerhood, and dialogicity at corpus scale; (ii) empirically, we provide, to our knowledge, the first large-scale, cross-regional description of interrogative practices in French-language digital news, showing how interrogatives structure coverage, how often they are resolved in-text, and which actors and places they center; and (iii) methodologically, we propose and validate a triangulated pipeline in which large language models and fine-tuned neural classifiers are combined with qualitative annotation to yield interpretable proxy measures relevant to debates on gatekeeping, voice, and agenda-setting. 2 Related Work Theories of questions: semantics and pragmatics Linguistics and philosophy of language treat questions not just as syntactic configurations but as discourse moves whose interpretation depends on context. Formal semantics represents questions as abstract objects that partition the space of possible worlds and determine which answers resolve an issue ( 21 ; 8 ) , while contextual models such as Questions Under Discussion and Table-based accounts ( 14 ; 19 ) embed this view in a dynamic picture of interaction where questions guide relevance and shape how common ground is updated. Pragmatic studies emphasize that interrogatives must be classified according to speaker intentions and social situations: they can request information, highlight issues, regulate knowledge display, or indirectly request action ( 27 ; 3 ) . Classic distinctions between information-seeking, rhetorical, leading, and echo questions, and between more or less “conducive” designs, show how presuppositions and turn design can steer respondents toward particular answers or implicatures ( 47 ; 28 ; 16 ; 5 ) .

Read the complete text