Perpetrators in multimodal media discourse: a case study of personalization in images from The Telegraph

The study situates itself within media framing theory, viewing framing as a multimodal principle that organizes meaning by making certain discourse elements salient. In news discourse, framing selects and highlights aspects of events and actors to construct causal and evaluative narratives. Prior work shows media representations of social actors are never neutral and often encode ideological and affective stances. Building on van Leeuwen’s sociosemantic inventory, the paper distinguishes inclusion/exclusion strategies and focuses on personalization (representing social actors as identifiable humans) across verbal and visual modes. Visual personalization is operationalized via social distance (shot range), social relation (camera angle), and social interaction (gaze direction as ‘offer’ vs ‘demand’). The paper asks how The Telegraph frames alleged perpetrators visually and via verbal mediation so as to ascribe social identity and agency along a safety–danger axis, and why certain alleged perpetrators are positioned as more or less dangerous. It argues that combining Multimodal Critical Discourse Analysis with Social Semiotics can reveal ideological work in image-text pairings, despite critiques of MCDA’s subjectivity. The study is relevant to understanding how news images may signal danger and legitimize institutional responses (e.g., law enforcement).

The paper draws on framing theory (Entman; Reese; D’Angelo) and multimodal discourse/social semiotics (Kress & van Leeuwen; van Leeuwen; Caple; Kress) to analyze how images and text jointly construct meaning. Van Leeuwen’s model of representing social actors (inclusion/exclusion; personalization vs impersonalization; nomination, functionalization, abstraction, objectivation) underpins the verbal analysis, while visual analysis follows Kress & van Leeuwen’s ‘grammar of visual design’ (social distance via shot range; social relation via vertical/horizontal angle; interaction via gaze as ‘offer’/‘demand’). The review references scholarship on media portrayal of social actors (e.g., Johnston & Noakes; Norris; Haynes & Hennig; Journal of Perpetrator Research) and research on salience and attention in images (e.g., Achanta et al.; Yan et al.) and emotion/gaze (Adams & Kleck; Hakala et al.). It also acknowledges debates on MCDA’s subjectivity (Wodak; Wooffitt) and suggests that integrating Social Semiotics can mitigate this. Discussions of ingroup/outgroup dynamics and evaluation (Pickett & Brewer; Nesdale; Nevala) inform the interpretation of inclusion/exclusion and moral positioning.

Design and corpus: A case study of The Telegraph (the online platform of The Daily Telegraph) analyzed images accompanying news articles about violence (as defined by WHO, 2002) from January 2010. Articles were accessed June 2019–February 2020 (before subscription changes). Selection criteria included presence of an image portraying alleged perpetrators, identifiable via captions or article text; courtroom sketches and digitally manipulated images were included. For articles with multiple images, only the image directly under the headline/standfirst was analyzed. The final corpus comprised 95 images. Analytical framework: Multimodal Critical Discourse Analysis informed by Social Semiotics. The analysis followed a three-step framing procedure (after D’Angelo): (1) identify linguistic labels (e.g., nomination/functionalization) and visual features; (2) examine discourse patterns and image attributes; (3) content-analyze narrative conventions. Visual variables operationalized: (a) gaze direction (‘offer’ images with averted gaze vs ‘demand’ images with direct gaze), (b) social distance via camera range (close-up to long shot), and (c) social relation via vertical and horizontal camera angles. Verbal variables included nomination (use of personal names) and functionalization (roles/labels such as murderer, rapist, terrorist). Contextual features (backgrounds, presence of police/victims, masking) were noted. Data were categorized and counted by these dimensions and illustrated with exemplar cases (e.g., ‘Detroit bomber’ Umar Farouk Abdulmutallab; Osama bin Laden; ‘mercy killers’).

Overall distribution by gaze: Of 95 images, 53 were ‘offer’ (averted gaze) and 42 ‘demand’ (direct gaze). Demand images (n=42): - Social distance: 34 close-ups (face/head-to-upper-shoulder); 8 mid-range (head-to-chest/waist). Close-ups often resembled police or document photos; 18 appeared to be license/booking shots. Backgrounds were frequently absent or blurred (20 without background; 5 blurred), heightening facial salience. - Labelling: Captions and headlines commonly used nomination (full names) and functionalization (e.g., murderer/rapist/terrorist); terms such as murder/murderer (11 uses), death/dead (9), kill/killer (10) reinforced danger. Visual proximization was paired with linguistic proximization. - Angles: Vertical: 36 eye-level, 6 from below; Horizontal: 28 frontal, 14 side. Eye-level close-ups positioned viewers to confront alleged perpetrators, amplifying perceived threat. Some masked faces (e.g., ‘female terrorist’; acid attacker) intensified salience via exposure of eyes/mouth. - Contextual compositions: A few paired portraits juxtaposed perpetrator and victim, typically placing the alleged perpetrator on the left (given) and victim on the right (new), orchestrating attention via conventional layout. Offer images (n=53): - Social distance: Head-to-shoulder/chest (28), head-to-waist/hip (13), head-to-knee (3), full figure (9). Mid/long shots frequently showed groups (e.g., militants, protesters) and/or included law enforcement. Long shots often depicted non–white-European groups with weapons (e.g., Taliban, al-Qaeda), symbolically distant and threatening. - Angles: Vertical: 27 eye-level, 22 from below, 4 from above; Horizontal: 46 side, 7 frontal. Lower-angle group shots visually ascribed power while maintaining distance. Presence of police in frame (often mid-range) mitigated perceived danger by implying control. Patterns and interpretations: - Perceived danger scales with proximity: Individuals presented as imminently dangerous (e.g., alleged murderers, serial rapists, ‘Detroit bomber’) were most often framed in close-up demand images at eye level, inviting confrontation. - Geopolitical distance correlates with visual distance: Foreign terrorist groups were framed with longer-range offer shots, often from lower angles, enhancing power yet maintaining symbolic distance. - Repetition/familiarity: Frequently reported figures (e.g., Osama bin Laden, Mehsud) received closer-range portrayals over time, fostering recognizability. - Ideological inflection: A right-leaning, pro-Christian slant was evident in coverage of ‘mercy killing’ cases: some women (e.g., Kay Gilderdale acquitted) appeared in offer images; others (e.g., Frances Inglis; Dr. Jane Barton) in demand close-ups, with strong negative functionalization. Celebrities accused but not charged often appeared smiling, eliciting more neutral readings versus stern portrayals of convicted offenders. - Naming plus imagery supports exclusion: Combining names with close-up images facilitated social exclusion and public recognition, aiding enforcement but risking lasting stigma if acquitted.

The findings demonstrate that The Telegraph’s multimodal framing strategically calibrates social distance, relation, and interaction to position alleged perpetrators along a safety–danger continuum. Close-up demand images at eye level function to confront readers with the faces of those framed as most threatening, aligning with schemas of danger associated with mug shots and police documentation. Longer-range offer images, frequently of foreign militant groups, simultaneously ascribe power (via low angles) and distance, reinforcing an outgroup threat while maintaining separation from the reader’s ingroup. Verbal labels (nomination and functionalization) work in tandem with visuals to guide interpretation, often unconsciously. Police presence within frames symbolically reassures readers that control is asserted, legitimizing institutional responses and reducing perceived danger of the depicted individual. Repetition and familiarity of certain figures (e.g., bin Laden, Abdulmutallab) entrench associations between names, faces, and threat categories. Ideological positioning shapes agency attribution: a pro-Christian, right-leaning outlook surfaces in the treatment of euthanasia-related ‘mercy killing’ narratives and in the differing visual treatment of celebrities versus convicted offenders. The coupling of naming and imaging can facilitate social exclusion and public vigilance but risks harm when allegations do not result in convictions. Overall, the visual-verbal orchestration acts as a tool for constructing social identities of ‘perpetrators’ and for legitimizing legal-political actions (e.g., the ‘war on terror’).

This case study shows how The Telegraph achieves ‘personalization’ of alleged perpetrators by coupling verbal nomination/functionalization with visual controls over social distance, relation, and interaction. Individuals presented as imminent threats most often appear in close-up demand images at eye level, while distant terrorist groups are framed through longer-range offer images from lower angles. The dual function of such imagery is to warn audiences about danger while legitimizing law-enforcement and broader political actions aimed at public safety. Ideological orientation influences the assignment of agency and evaluative tone, as seen in euthanasia-related cases and celebrity coverage. The study contributes a multimodal account of how framing organizes viewer engagement and social identity construction along a safety–danger axis. Future research should expand beyond alleged perpetrators to include victims, law enforcement, and other actors, and compare across outlets and time periods to test generalizability and ideological variability.

- Scope restricted to one outlet (The Telegraph) and one actor type (alleged perpetrators), excluding systematic comparison with victims, law enforcement, and bystanders or articles without images. - Temporal focus limited to January 2010 images, potentially constraining generalizability across times and news cycles. - Case-study design and MCDA entail interpretive subjectivity; although mitigated by Social Semiotics, coder subjectivity may influence categorizations (e.g., angle, range, gaze). - Selection limited to the first image under headline/standfirst per article may miss alternative framings within the same piece. - Ethical/judicial outcomes vary; reliance on media labels risks conflating allegation with guilt, affecting interpretation of findings.