PRIN: PROGETTI DI RICERCA DI RILEVANTE INTERESSE NAZIONALE – Bando 2022 PNRR — Prot. P202293P5P

Countering Online Radicalization and incivility in ITaly: from fringe to mainstream – CORIT

Culture, Creativity and Inclusive Society — Trust in democracy is restored through the expansion of active and inclusive citizenship.

Brief description of the proposal

The use of social media by individuals and organizations to target individuals for political and social change has become increasingly common as the Internet has become a central tool in information diets and mobile devices are increasingly accessible. Coupled with this is a growth of incivility in online public discourse related to political issues and citizenship rights damaging to our democracies. In this context, “fringe platforms” such as Telegram, Gab, Odysee and BitChute have become prime destinations for extremist thought and conspiracy theorists.

Research project CORIT will explore the relationship between toxicity of online public speech and forms of political radicalization of user-citizens and will analyze these processes in digital platforms, looking at how uncivil discourses transit between fringe platforms, mainstream platforms and legacy media.

To achieve these goals, CORIT project will rely on a mixed-methods research design a) investigating online incivility and political radicalization in fringe platform through an ethnographic approach by looking at case studies from Telegram, b) then defining a list of radicalization-related news items to be later tracked within Twitter (via big data approach) – so as to compare whether how discursive forms of incivility about politics are produced at this level – and c) compare the Telegram-Twitter news diffusion with the legacy media agenda – in a perspective of intermedia agenda – in order to more fully explore the influence between the media ecosystem and radicalization discourses. CORIT project has a distinctive impact because it aims to understand the generation of toxicity and not just its circulation, thus focusing on preventing such dynamics.

 

State of the art

The spread of digital media and the process of platformization (Van Dijck, et al., 2018) has reshaped the media and political systems into a hybrid media system (Chadwick, 2017) that has fostered a change in public debate (Bentivegna, Boccia Artieri 2021). This shift in languages and political discursive practices involves a wide range of actors who can “introduce, amplify, and maintain arguments, frames, and speakers” (Jungherr et al., 2019, p. 16). As several scholars have noted, such a change introduced a level of toxicity into the public debate concerning:

a) the growth of incivility as a strategic communicative practice capable of benefiting the political visibility of certain actors (Bentivegna, Rega 2022);

b) the emergence of information disorder in contemporary political communication characterized by increasing misinformation, disinformation, media manipulation, coordinated inauthentic behaviour, and propaganda (Freelon & Wells, 2020). A recent study, in this regard, shows that citizens do not perceive clearly the problematic nature of political behaviour aimed at mystifying facts in order to discredit the opposing party (Bentivegna et al., 2022);

c) communicative polarization, which derives from difficulties denounced by several analyses, in getting in touch with alternative viewpoints. As Yarchi and colleagues point out (2020), this closure occurs through interactional polarization, such as echo chambers and filter bubbles, but also through positional and affective polarization dynamics, emerging from partisan social identity display, which also increases hostility between opposing political fields.

Such processes are enhanced by citizens’ participation in online alternative media, which have been correlated with disinformation ecosystems, the spread of conspiracy myths and the normalization of populist thinking and political extremism (Schwarzenegger, 2022).
More recent literature has shown how there is an interdependent relationship between the rise of alternative online spaces and the growth in far-right forms of extremism (Holt, 2018; Walther & McCoy, 2021), thus stressing a rising concern that “secluded online spaces can function as laboratories that develop extremist talking points that then find entry into the mainstream” (Lewandowsky et al., 2020, p.84). At the same time, these spaces “provide voices to marginalized and disadvantaged communities” (ibid.), so they can also represent new resources for public debate.

In this context, characterized by a growth of toxicity in the hybrid media system, it is only very recently that scholarly research has begun to address how extremist and less visible forms of radicalization find a favourable environment in fringe platforms, such as 4chan, Telegram, Gab, Odysee and BitChute. Moreover, there is still little research on how political discourses produced in these environments influence mainstream social media platforms and the legacy media agenda.

For example, it has been studied how fringe communities on “dark platforms” such as 4chan/pol/, where conspiracy theories, misinformation, and hate speech are normalized (Zeng & Schäfer, 2021), influence the wider digital sphere in terms of the dissemination of (political) news and dangerous narratives. Zannettou and colleagues (2017) showed how fringe communities within 4chan and Reddit have a surprisingly large influence on Twitter since they serve as incubation chambers for much radical and misleading information that eventually ends up circulating in the broader web, thus showing how fringe platforms have become an important piece of the current information ecosystem.

In the Italian context, one independent research (ANPI 2022) has found a network of more than 600 Telegram groups and channels representing a haven for far-right groups that have been banned from Facebook and Twitter. Telegram, in this sense, offers an online space where anonymity and scarce moderation can facilitate the formation of violent, racist, anti-Semitic and subversive discourses.

While much research has focused on the political intermedia agenda dynamics between legacy media and social media agendas (Bentivegna, Boccia Artieri 2019; Bentivegna, Boccia Artieri, Marchetti 2020), there is still little research on the potential agenda setting effects of fringe platforms. Although it is known that these platforms can facilitate the creation of toxic or misleading information, little is known about the communication flows between fringe and mainstream media spaces.

The research project aims to analyze the Italian context of fringe platforms and their impact on mainstream social media platforms and legacy media agendas to identify possible interventions to limit online radicalization and raise awareness of the phenomenon.

General description of methodology

To achieve these goals the project will rely on a mixed methods approach. Fringe platforms and closed groups are indeed resistant to approaches entirely based on big data and computational analysis (Boccia Artieri et al. 2021a). Moreover, an in-depth understanding of radicalising and uncivilised languages cannot do without a qualitative analysis capable of capturing the nuances of meaning and cultural cryptography practices. On the other hand, an entirely qualitative approach based on case studies would risk focusing on too small portions of the phenomenon, missing the overall picture of the Italian networked public sphere. The combination of digital methods with digital ethnography and news content analysis in the investigation of the Italian Telegramsphere constitutes an innovative aspect of the research, thus the project has the potential to develop research methods that will be useful in the future in the analysis of fringe platforms and intermedia agenda setting.

List of research units (RU)

GIOVANNI BOCCIA ARTIERI
Professore Ordinario
Università degli Studi di Urbino Carlo Bo

SARA BENTIVEGNA
Professore Ordinario
Università degli Studi di Roma La Sapienza

ROSSELLA REGA
Ricercatore a t. d. – t. pieno (art 24 c. 3 – b L.240/10)
Università degli Studi di Siena

t

Objectives and research questions

OBJECTIVES

  • trace the network of problematic channels and groups in the Italian Telegramsphere and describe the toxic practices therein;
  • investigate the role of such spaces in the production and diffusion of counter-narratives fostering polarization, incivility and information disorder;
  • identify how toxic discourses and practices move between legacy media, mainstreame social media (as Twitter) and fringe platforms (as Telegram);
  • strengthen the theorisation of the relationship between fringe and mainstream in digital spaces and understand the transformation of the intermedia agenda setting process in relation also to legacy media;
  • providing communication professionals with knowledge and operational guidelines to counter the dynamics of radicalisation and incivility fuelled by fringe platforms.

RESEARCH QUESTION

How does the relationship between fringe and mainstream platforms in Italy participate in the development of online toxicity (polarisation, incivility, information disorder)?

The question starts from the assumption that in order to investigate the impact of fringe platforms on online toxicity, the interplay between fringe and mainstream should be observed. In order to observe this dynamic, the research looks at four constituting processes to which correspond five research sub-questions: RQ1 the type of environment that is cultivated within fringe platforms, RQ2 the flow from mainstream platforms to fringe platforms , RQ3 the flow from fringe platforms to the mainstream platforms, RQ4 the process by which national news are received and framed Telegram problematic networks , and RQ5 the way in which legacy media, mainstream social media and fringe platforms interact in the constitution of an intermedia agenda .

SUB-QUESTIONS

  • r.q.1: how do practices and discourses of toxicity develop within fringe platforms in Italy?
  • r.q.2: how does the migration from mainstream social media to fringe platforms take place?
  • r.q.3: how do toxic practices and discourses developed in fringe platforms impact mainstream social media?
  • r.q.4: how are news produced by legacy media framed and circulated in fringe platforms?
  • r.q.5: how are the agenda setting dynamics between the legacy media agenda, the social media agenda and the fringe platforms agenda?

Research project outline

    The research programme is structured over the two years through seven milestones. 

    l

    Milestone 1: Refining the theoretical and methodological framework to study online radicalization and incivility in the italian context

    DESCRIPTION

    In this phase, the three units will analyze the existing literature on radicalization and incivility in online environments, particularly focusing on fringe platforms, deplatforming processes, and mixed methods approaches to study non-public online spaces. UniUrb Unit will focus on the literature on fringe platforms, while UniRm and UniSi will review the literature on online radicalization, toxicity, and incivility.

    ACTIVITY 1.1

    Review of the academic literature, policy documents and grey literature on fringe platforms (Year 1 Bim 1-2) – UniUrb.

    ACTIVITY 1.2

    Review of the academic literature on online radicalization, toxicity and incivility (Year 1 Bim. 1-2) – UniRm, UniSi.

    DELIVERABLES – D1.1

    Systematic literature review article on the topic of radicalization, incivility and fringe platform to be submitted to an international journal and that will be uploaded on the project website and/or on a pre-print repository (Year 1 Bim 2).

    DELIVERABLES – D1.2

    Methodological statement of the project which will clarify the ethical precautions and research techniques to be employed and that will be uploaded on the project website (Year 1 Bim 2).

    TARGETS

    Reach at least 100 national and international scholars with the project’s first deliverables.

    l

    Milestone 2: Mapping of the italian telegramsphere of problematic contents

     

    DESCRIPTION

    This research phase aims to build a network of the main Italian Telegram channels and groups that disseminate problematic content in terms of incivility, polarisation and information disorder. The mapping starts from lists of Telegram channels/groups compiled by debunkers (i.e.the list from BUTAC https://www.butac.it/the-black-list/) as already experimented with in other research (Fletcher et al., 2017). This list is expanded through a snowballing method (Peeters & Willaert, 2022) by adding all linked channels/groups for at least one month until sufficient saturation is reached (when new links appear for a certain time with a frequency below a set threshold). The network composition will use a combination of ethnographic observation and scraping tools such as 4Cat and the Telethon Python library (Urman & Katz, 2020). This mapping will serve to collect material for the next phase and experiment a mixed-method approach to study the Italian Telegramsphere.

    ACTIVITY 2.1

    Collection of Italian problematic Telegram channels/groups (Year 1 Bim. 3-4) – UniUrb.

    DELIVERABLES – D2.1

    Building a dataset with problematic Telegram channels/groups (Year 1 Bim. 4).

    TARGETS

    A quantitative target for the number of channels/groups to be collected is not feasible, as this is dependent on the configuration of the network that will be encountered at that time.

    l

    Milestone 3: Gaining and understanding of toxicity practices on the italian telegramsphere: a “polarisation glossary”

     

    DESCRIPTION

    Drawing on the network of channels/groups collected in the previous stage, in this phase the research will collect and analyze those contents that contribute to increasing the risk of toxicity. Through field notes and scraping tools, we will collect 1) news posted and their comments, 2) the most frequent targets of hostility, 3) recurrent lexicons participating in affective polarization, which will be clustered through the Emotional Text Mining technique (Boccia Artieri et al. 2021b). The collected news items will serve as the basis for the fifth phase of the research, while targets and lexicons will be analyzed to produce a ‘Polarisation Glossary’, namely a report of discursive practices that foster affective polarization in fringe platforms.

    ACTIVITY 3.1

    Activity 3.1: Participant observation in the collected groups and channels; collection of field notes and scraping on vocabularies, representations and practices related to affective polarization; analysis of collected data to construct the Polarisation Glossary (Year 1 Bim 3-6) – UniUrb, UniSi.

    ACTIVITY 3.2

    Collection and dataset organization of news shared or mentioned in the analyzed groups and channels (Year 1 Bim 4-5) – UniUrb, UniSi.

    ACTIVITY 3.3

    Preparing the ‘Polarisation Glossary’ based on materials gathered from participant observation and scraping (Year 1 Bim 5-6) – UniUrb, UniRm, UniSi.

    DELIVERABLES – D3.1

    Polarisation Glossary, based on materials gathered from participant observation and scraping. This document is structured to advise news and digital communication professionals to avoid polarizing counter-reactions. (Year 1 Bim 6).

    DELIVERABLES – D3.2

    Preparation of a paper for submission to an international journal on vocabularies, representations and practices related to affective polarization in Telegram radicalized channels/groups (Year 1 Bim 6).

    TARGETS

    Reaching at least 200 news and digital communication professionals with the Polarisation Glossary, thanks also to the support in the dissemination provided by Parole Ostili, AgCom, Associazione stampa parlamentare, Ordine dei giornalisti, Associazione Carta di Roma.

    l

    Milestone 4: Achieving a comprhension of the upstream and downstream interrelationship between telegram and twitter in Italy

    DESCRIPTION

    This phase is dedicated to the investigation of the relations between the Italian Telegramsphere and Twittersphere, in order to understand upstream and downstream (Gursky et al. 2022) patterns between fringe and mainstream social media. This will be accomplished by observing how links, groups, influential personalities, hoaxes and lexicons from the previous phase are mentioned on Twitter, and whether an increased risk of incivility develops around these interchanges.

    ACTIVITY 4.1

    Searching on Twitter for Telegram accounts, lexicons and links collected in the previous phase, and gathering and analyzing the results (Year 1 Bim 6 – Year 2 Bim 3) –UniRm, UniUrb.

    DELIVERABLES – D4.1

    Report containing maps of Telegram links on Twitter and of the occurrence of toxic lexicons also present on Telegram (Year 2 Bim 3).

    TARGETS

    Reaching at least 200 news and digital communication professionals with the report, thanks also to the support in the dissemination provided by Parole Ostili, AgCom, Associazione stampa parlamentare, Ordine dei giornalisti, Associazione Carta di Roma.

    l

    Milestone 5: Developing an interpretative model of the relationship between legacy media, mainstream social media and fringe platforms media agenda

     

    DESCRIPTION

    This phase is dedicated to comparing the legacy media agenda, mainstream social media agenda and the media agenda elaborated in the Telegram network investigated. With the help of the news monitoring tool Volopress, the main news items in three months will be compared with the corresponding news posted and commented on Telegram channels/groups in the same period (taking into account possible latency between the two agendas). This analysis will assess which news items the Telegram network focuses on, from which sources and with which framing. Through the comparison with data from the Twitter analysis in the previous phase, the research will also attempt to determine whether fringe networks influence the visibility of certain news and topics (for example through hashtag hijacking or the visibility of ‘fringe influencers’ in the mainstream media).

    ACTIVITY 5.1

    Definition of the legacy media agenda from Google News and Volopress (Year 2 Bim 2) – UniUrb, UniSi.

     ACTIVITY 5.2

    Comparison of news items and their frame in Telegram channels and groups, Twitter (collected in Activity 4.1) and the legacy media agenda looking for intermedia agenda dynamics (Year 2 Bim 3-4)- UniSi, UniUrb, UniRm.

     

    DELIVERABLES – D5.1

    Preparation of a paper for submission to an international journal on an interpretative model of the relationship between legacy media, mainstream social media and fringe platform agenda (Year 2 Bim 4).

    TARGETS

    3 months of news comparison between legacy media, mainstream social media and fringe platform agenda.

    l

    Milestone 6: Building and sharing training tools for information and communication professionals to navigate fringe platforms cultures, practices and affordances

    DESCRIPTION

    This phase aims to engage the audience of professionals, providing them with research-based training tools and using their insider perspective to best calibrate the final outputs of the project. In particular, workshops will be organized to present and discuss the results, and a book will be prepared to disseminate data and guidelines on how to counter radicalisation and incivility between fringe and mainstream platforms.

    ACTIVITY 6.1

    Organisation of three workshops (one for each research unit) in which the research results are presented to professionals targeted by the project (journalists, politicians, social media managers, fact-checkers, science communicators) which will be invited as discussant; the feedback gathered from the professionals will be integrated in the final publications of the project (Year 2 Bim 2-4) – UniUrb, UniRM, UniSi.

    ACTIVITY 6.2

    Elaboration of the final research report for professionals containing data and training tools to counter radicalization and of a book aimed at a more general public on radicalization and incivility between fringe and mainstream platform (Year 2 Bim 4-6) – UniUrb, UniRM, UniSi.

    DELIVERABLES – D6.1

    Three workshops with news and digital communication professionals (Year 2 Bim 2-4).

    DELIVERABLES – D6.2

    Final report of the research (Year 2 Bim 6).

    TARGETS

    200 news and digital communication professionals engaged through the workshops and reaching at least 1000 professionals and 100 national and international scholars with the research report.

    l

    Milestone 7: Communicating the project effectively to scholars and professionals 

    DESCRIPTION

    Ensure efficient communication of the different phases of the project and set up tools to support impact and scientific dissemination.

    ACTIVITY 7.1

    Planning and writing press releases of the project’s activities and results (Year 1 Bim 1 – Year 2 Bim 6) – Uniurb.

     

    ACTIVITY 7.2

    Organizing events to present the project (Year 1 Bim 1 – 2) – Uniurb.

    ACTIVITY 7.3

    Design and development of the project website with the support of Uniurb communication Office (Year 1 Bim 1 – 2) – Uniurb.

    ACTIVITY 7.4

    Update the website with news and research results (Year 1 Bim 1 – Year 2 Bim 6) – Uniurb, UniRm, UniSi.

    ACTIVITY 7.5

    Organizing a final international conference in Rome with the presentation of research findings involving researchers from academia, the professionals previously engaged, policy makers and politicians (Year 2 Bim 3-6).

    DELIVERABLES – D7.1

    Kick-off meeting (Year 1 Bim 1).

    DELIVERABLES – D7.2

    Kick-off public conference for the presentation of the project (Year 1 Bim 2)

    DELIVERABLES – D7.3

    Project’s website online (Year 1 Bim 2).

    DELIVERABLES – D7.4

    Participation in 4 national and international conferences (such as those organized by The Association of Internet Researchers, European Communication Research and Education Association, International Communication Association, American Political Science Association, etc.).

    DELIVERABLES – D7.5

    Final scientific conference in Rome (Year 2 Bim 6).

    TARGETS

    Present the project to at least 1000 national and international scholars and professionals. Involve an audience of at least 200 participants for the final conference.