Hybrid spaces of politics: the 2013 general elections in Italy, between talk shows and Twitter

Summary

This article maps the hybrid entanglement of television and Twitter during the run-up to Italy’s 2013 general elections, treating second-screen activity around political talk shows as a privileged site for studying contemporary political communication. Combining a near-complete capture of tweets bearing official talk-show hashtags with scene-anchored content analysis of peak moments around the three main parties (PD, PDL, M5S), the authors argue that while volumes appear impressive, access remains narrow, broadcasters and politicians rarely engage with the connected audience, and tweeting takes the form of opinion and demands for interaction rather than genuine agenda-shaping. Twitter, in this Italian configuration, amplifies TV logic more than it disrupts it.

Key Contributions

• One of the first large-scale empirical mappings of TV–Twitter hybrid practices in an Italian national election cycle.
• Operationalizes Carpentier, Dahlgren, and Pasquali’s distinction between participation in and through the media for second-screen Twitter practices, refining Wohn and Na’s typology.
• Demonstrates a reproducible mixed-method workflow integrating firehose-level Twitter capture, per-minute metrics, peak detection, stratified sampling, and content analysis of tweets paired with the corresponding on-air scene.
• Releases an open dataset of episodes and metrics to support replication.
• Advances hybrid-media-systems theory by showing how broadcast logic continues to structure networked political talk.

Methods

A practice-based, hybrid-media design. The authors collected 2,489,669 tweets (Aug 2012–Jun 2013) containing official hashtags of 11 Italian political talk shows via DiscoverText/GNIP, filtered 1,889,281 of them to airtime windows, and computed per-minute Twitter metrics following Bruns and Stieglitz. Keyword attribution assigned tweets to PD, PDL, and M5S streams; the Marcus et al. peak-detection algorithm identified 530 peaks, from which stratified sampling yielded 23 peaks (8,031 original tweets). Each tweet was coded jointly with its corresponding TV scene along orientation (political/media), form (opinion, information/streaming, request for interaction, new issues/angles, plus tone markers) and content categories, with double-coding (final Krippendorff’s α < 0.81).

Findings

• Only 187,031 unique users tweeted during talk-show airtime—a thin sliver of Italy’s ~4 million Twitter users—signalling narrow access.
• Just 3 of 11 talk shows displayed audience tweets on air; official party/leader accounts contributed only 969 tweets, and Grillo/M5S were almost wholly absent from the conversation.
• Tweet volumes across the three parties were comparable (PDL 11%, PD 10%, M5S 8%) and tracked the broadcast and political event calendar.
• Of the coded sample, 77.3% was politics-oriented; opinion/comment (79.8%) and requests for interaction (28.6%) dominated, while introduction of new issues/angles/sources reached only 2.4%.
• Party-specific signatures emerged: media-oriented opinions clustered around Berlusconi (PDL), politics-oriented live streaming around PD, and interaction requests on political issues around M5S.
• Requests for interaction were disproportionately laced with irony (35%) and violent/vulgar language (11%).

Connections

This paper anchors a long-running Italian-elections research line on how broadcast and social-platform logics interlock; it can be read alongside later work tracing the same hybrid arena into Facebook, problematic information, and coordinated behaviour in subsequent Italian campaigns, including Giglietto2019-882f1900, Giglietto2020-9d8acdd7, Giglietto2023-fa71a001, Marino2024-2fbc690f, Giglietto2025-1765bb4f, and Giglietto2025-1e9a0917. Together they chart a trajectory from spontaneous second-screen commentary to more strategically engineered cross-platform political communication.