Call them demons without calling them demons: How The New York Times legitimises violence in its coverage of 7 October 2023

Abstract

The New York Times’s coverage of Gaza has been extensively studied for media bias, yet its role in legitimising mass violence has been relatively unexamined. Addressing this gap, I analyse the newspaper’s reporting from 7 October 2023, through 7 April 2024 to conceptualise what I term subtle demonisation. Drawing on Hart’s Cognitive Critical Discourse Analysis, I integrate Bakhtin’s chronotope with attention distribution to identify four destructive patterns: (1) chronotopic displacement, (2) agentive asymmetry, (3) semantic bundling, and (4) metaphor-driven reclassification that construes Palestinians as existential threats requiring elimination. I find these patterns collectively construct Palestinians as demons without naming them as such, rendering their lives structurally expendable. In doing so, I extend debates on media bias by showing how the language of journalistic objectivity functions as an instrument of epistemic erasure and a mechanism of complicity in atrocity.

Introduction

From October 2023 to April 2025, over 51,000 Palestinians, primarily women and children, were systematically killed, while more than 10,000 others remain buried and missing beneath the rubble (Al Jazeera, 2025, 24 April). This annihilation must be understood not merely as a humanitarian catastrophe, but as a harrowing testament to the discursive structures that render settler-colonial violence intelligible, permissible, and at times invisible. Despite intermittent ceasefires, ¹ the assault persisted through overwhelming asymmetrical military force and was continually enabled by media coverage that framed the violence in ways that rationalised its continuation. The persistence of this violence cannot be explained without examining the discursive practices that sustained it, above all the demonisation through which Palestinians were constructed as morally expendable (see e.g. Hamad, 2024, 9 April).

By demonisation, I refer to the discursive process through which groups are cast into stark moral oppositions, with one constructed as civilised and rational, and the other as irrational and threatening. The concept resonates with van Dijk’s (1998) account of ideological polarisation yet extends it by construing the out-group as actively malevolent and dangerous. Normand (2016: 1–6) conceptualises demonisation as an accusative speech act that centres on overt declarations, in which enemies are unequivocally labelled using direct lexicons like “Satan” and “evil”, or indirect lexicons like “murderous” or “treacherous”, or figurative tropes like “Devil’s emissaries” that cast them as literal embodiments of evil. This overt rhetoric, which effaces moral personhood by reducing its targets to “nonhuman beings” (Befu, 1999: 26), paves the way for atrocity. The pattern is evident across history. Nazi propaganda portrayed Jews as vermin and existential contaminants, providing the ideological scaffolding for industrialised slaughter (Landry et al., 2022). In Rwanda, state media and politicians labelled Tutsis inyenzi (cockroaches), priming the collective consciousness for genocide (Hefti and Jonas, 2020). Myanmar’s regime framed the Rohingya as “illegal Bengalis” and demographic invaders, a rhetorical manoeuvrer that sanctioned mass expulsion and murder (Egreteau and Myat, 2024). In each case, violence is not solely enacted by force but is first constructed through narrative.

Narratives, as Bakhtin (1981: 84) argues, are fundamentally chronotopic because they situate events and participants within temporal and spatial frames that confer meaning. The “chronotope”, literally “timespace”, captures this inseparability. It acts as a structuring feature that shapes a story’s possibilities, defines its characters, and endows them with specific moral valences. Moving beyond its literary origins and leveraging the central premise of the narrative turn in news scholarship (see e.g. Ørmen and Gregersen, 2019; Zelizer, 2004), I build on Blommaert and De Fina (2016) to extend the chronotope to media reporting. I analyse how it organises events and social actors, legitimises certain subject positions, and casts others as transgressive, thereby furnishing the narrative ground for demonisation. While Blommaert (2015, 2020) resists reading chronotopes as cognitive schemas, he stresses their role in organising action and evaluation, making them a productive site for investigating the meaning-making processes known as construal in Cognitive Critical Discourse Analysis (CCDA). I therefore draw the chronotope into dialogue with CCDA, not to collapse one into the other, but to examine how chronotopic arrangements enact construals that pattern reality, direct attention, and render certain ideological positions socially recognisable. Here, construal refers to the systematically organised patterns of meaning through which a reality is constructed (see Hodge and Kress, 1993: 15; Langacker, 2008: 43–44, 55). Within CCDA, Hart (2025) contends that “any situation or event is subject to construal on at least two levels” (p. 78): the conceptual content we perceive is already shaped by how we choose to construe it, and the content is further mediated through attention, perspective, and values. It is on this latter plane that ideological work becomes most palpable yet least visible, a process exemplified by the discursive production of Palestinians as demonised subjects.

Against this background, in what follows I examine the New York Times (henceforth NYT) reporting from the first 6 months of the genocide against Palestinians, focusing on how construal operates through selective patterns of attention. Specifically, I ask how the NYT’s chronotopic distribution of attention discursively demonises Palestinians and contributes to the legitimisation of ongoing state violence. The NYT is chosen for its unrivalled discursive authority. Long regarded as the “newspaper of record”, it serves as a global benchmark for journalistic credibility and objectivity (Artz, 2014; Gilboa and Sigan, 2024). Its reporting not only shapes public discourse but also influences diplomatic framings and policy decisions in the United States and beyond. This influence is underscored by a Pew Research Center study (Silver et al., 2024), published 5 months after the escalation of 7th October, which found that 58% of Americans believed Israel’s reasons for fighting Hamas were valid, a perception likely informed by media narratives.

The NYT’s coverage of Palestine and Israel

The role of media in shaping public understanding of the Israeli occupation of Palestine has long been a subject of critical scholarship. From the First Intifada (1987–1993) to the ongoing violence in Gaza, a substantial body of research has problematised how the NYT has discursively mediated the occupation, often reproducing frames that privilege Israeli perspectives while erasing Palestinian experience. As Mearsheimer and Walt (2007: 169) note, the American media’s coverage, including that of the NYT, tends to be strongly biased in Israel’s favour, casting Palestinian resistance as aggression and Israel’s actions as rational and justified. Khalidi (2020) likewise argues that the NYT has systematically obscured the historical and human dimensions of Palestinian suffering, consistently decontextualizing Israeli violence, foregrounding Israeli perspectives, and rendering Palestinians lives contingent and expendable.

These critiques are further substantiated by Jackson’s (2023) computational content analysis of over 33,000 NYT articles on the First and Second Intifadas. She finds persistent anti-Palestinian bias in the disproportionate use of the passive voice to describe violence against Palestinians and in the reliance on more negative, violent rhetoric toward them than toward Israelis. The recurrence of these patterns across both Intifadas, even amid escalating violence in the second, indicates a deepening anti-Palestinian sentiment, as also evidenced in Friel and Falk (2007), who reveal that Palestinian violence, particularly suicide bombings, was consistently condemned, while Israeli violence and lawlessness in the occupied territories were downplayed or omitted. Similarly, Zelizer et al. (2002), in their analysis of 30 days of coverage during the first 10 months of the Second Intifada, agree that a pro-Israeli slant in the NYT is demonstrable, however, they contend it is erratic and embedded in routine journalistic practices, where sourcing, framing and headlines cumulatively yield coverage less critical of Israel. For this reason, they caution against the common tendency to frame such bias as simply pro-Israeli/pro-Palestinian.

This trajectory of institutional bias is further illuminated over a decade later in Artz’s (2014) study of the NYT front-page stories on “Operation Protective Edge” during the 2014 Gaza War. His analysis demonstrates how the newspaper employed a “banal balance” that falsely equated occupier and occupied, selectively legitimised Israeli sources while excluding critical international perspectives, and strategically omitted the context of occupation. He argues that through these practices, NYT did not merely report events but justified them, systematically legitimising Israeli military actions, obscuring facts on the ground and framing the colonial occupation as necessary. In doing so, NYT functioned as a powerful agent within the conflict rather than an impartial observer.

The events following 7 October 2023 reveal that these decades-old patterns have not been disrupted but have instead calcified. A quantitative analysis by The Intercept (Johnson and Ali, 2024, 9 January) of over 1000 articles post-7th October coverage confirms the same discursive asymmetries identified in earlier studies, now operating at scale and speed. Three major U.S. newspapers, including the NYT, unduly emphasise Israeli deaths with highly emotive terms like “slaughter” and “massacre”, while rendering Palestinian casualties in passive, agent-obscuring language. The coverage also minimised the impact on Palestinian children and journalists while devoting significantly more attention to antisemitic acts than to anti-Muslim racism, creating a pronounced narrative imbalance favouring Israeli perspectives. Given this empirically grounded trajectory, the aim here is not to reassert the existence of pro-Israel bias in the NYT coverage but to examine the discursive structures through which language organises attention, encodes agency, and governs the moral legibility of violence.

These discursive patterns warrant closer scrutiny of recent academic interventions that, despite working with similar datasets, arrive at sharply divergent conclusions. Pinker’s (2025) corpus analysis, for instance, finds that “Israel” is referenced more than three times as often as “Hamas”, and that mentions of Hamas combatants killed in Israeli strikes appear in only 10 percent of articles. Yet, he concludes that the NYT overrepresents Palestinian suffering, an inference that conflates limited textual presence with inflated moral emphasis. This paper departs from such interpretations by shifting the focus from surface-level frequency to the deeper mechanics of attentional distribution, that is, how Palestinian lives, even when acknowledged, are grammatically backgrounded, affectively neutralised, and discursively depersonalised. Here, I argue that such strategies do not only distort representation but also sustain a narrative logic that frames Palestinians less as moral patients worthy of empathy than as a political problem or threat to be managed.

A similar epistemic asymmetry underpins Gilboa and Sigan’s (2024) audit of 72 alleged editorial lapses, which locates bias primarily in questions of accuracy. Their analysis frames the NYT’s use of Palestinian sources, particularly in its early coverage of the Al-Ahli Hospital explosion, as a professional failing rather than a constrained act of journalistic necessity. By casting Palestinian testimony as suspect while legitimising Israeli claims without comparable scrutiny, their argument reproduces the very hierarchies it purports to expose. My analysis, by contrast, is less concerned with identifying inaccuracies than with the discursive conditions that make such asymmetries possible in the first place, showing how some agencies are stabilised while others are dissolved into ambiguity, and how this cumulative process makes large-scale violence appear not accidental but necessary.

Chronotopic construal and attention in Cognitive Critical Discourse Analysis

Within Cognitive Grammar, events are bounded portions of experience rather objective slices of reality, expressed in usage events where a shared conceptual representation against a common ground within the current discourse space is coordinated (Langacker, 2001: 144–146). On this view, describing an event always involves construal. Langacker (2008: 55–85) sets out four broad, yet fluid dimensions of construal, which he illustrates through the metaphor of viewing a scene: what comes into view depends on how closely it is examined (specificity), what is selected for observation (focusing), which elements receive the greatest attention (prominence), and the vantage point from which it is observed (perspective). These operations show that events are perspectivally construed in discourse, with agency and causation highlighted or suppressed, reflecting what Cienki and Iriskhanova (2018) describe as the mind’s capacity to “chunk the flow of information and assign meaningful structure to experience” (p. 7). In this way, the internal organisation of events, including participants, processes, and causal relations, is not simply “out there” in the world waiting to be described but emerges through construal.

A central mechanism of construal is the recruitment of image schemas. These are recurring patterns abstracted from bodily experience, by which Johnson (1987) means the very ordinary interactions we have with the world that are so basic we often overlook them. From repeated experiences such as moving through space, placing objects inside containers, or exerting and resisting physical force, the mind distils abstract patterns that provide the conceptual basis for interpreting events (see Johnson, 1987: 29; Langacker, 2008: 32). For example, the Source-Path-Goal schema derives from movement through space and is applied not only to physical journeys but also to abstract pattern processes such as “reaching a goal”. Hart (2025: 49) describes this schematic organisation as structural configuration strategies in CCDA, operations through which discourse schematises events by defining the structural properties of entities, relations and processes. These strategies work across topology (spatial relations), sequence (temporal order), causation (cause-effect relations) and meronimic organisation (part-whole structuring), selecting image-schematic meanings to scaffold event representation.

Drawing on Langacker (2008) and Talmy (2000, 2011), Hart (2025: 51–52) further notes that these configurations may be abstract (e.g. contact/separation, uniplex/multiplex) or archetypal (e.g. motion along a path, force exertion). These configurations are grounded in conceptual archetypes structured by patterns of energy flow and resistance, involving participants or thematic roles that specify the functions assigned to entities across different event-types (see also Hart, 2014, 2020) (Table 1).

Table 1.

Event types, energy dynamics and archetypal roles in CCDA.

Event type	Definition	Energy dynamics	Archetypal roles
Action events	Intentional acts in which one entity brings about change in another	Structured by energy transfer (agent → patient, sometimes via an instrument)	• Agent (the wilful initiator of the action and energy source) • Patient (the entity undergoing change as the energy sink) • Instrument (the means or transmitter through which energy passes)
Motion events	Spatial occurrences in which an entity moves relative to a reference point	No necessary energy transfer; structured by relative positioning and trajectory (figure → ground)	• Figure (the entity that moves, propelled by energy either internally or externally) • Ground (the reference point against which the movement is measured)
Force events	Dynamic interactions in which an entity’s intrinsic tendency is resisted or facilitated by another	Structured by energy interaction (agonist ↔ antagonist)	• Agonist (the entity with an intrinsic tendency, such as to move or to remain at rest) • Antagonist (the opposing or supporting force that applies energy to shape the agonist’s state)

These image schemas, both configurational and archetypal, provide the conceptual basis of linguistic structures from morphemes to clause patterns. However, Hart (2025: 52) also emphasises that the relationship between clause types and image schemas is neither fixed nor exclusive, since the same event may be construed through different schematic domains. This flexibility also extends into figurative thought, where metaphor operates both as a cognitive process of frame projection (see Fauconnier and Turner, 2002; Lakoff and Johnson, 1980) and also as a discursive framing strategy (Hart, 2025). In this capacity, culturally specific source frames (Fillmore, 1982), which are themselves structured by image schemas, are mapped onto target domains, assigning archetypal roles to social actors and importing evaluative entailments.

Crucially, even when events are given a basic shape by image schemas, what portion of that shape comes into view depends on attention. In Cognitive Grammar, attention is seen as a limited yet structurally powerful resource that cannot be spread evenly across the complexity of any event. Langacker (2001: 144–146) contends that attentional distribution is always partial since language users access only a restricted “conceptual field” at any given moment, focusing on some aspects of a scene while backgrounding others. Talmy (2000, 2008) refers to this as the “windowing of attention”, where linguistic forms bring selected segments onstage and leave others offstage. Attention, then, is not simply present or absent but gradient: it varies in salience, shifts flexibly albeit not randomly. Its distribution follows patterns shaped by grammar, cognitive habits and cultural expectations (Langacker, 2001: 161–162; Talmy, 2006, 2011).

Because attentional distributions are patterned rather than free-floating, they presuppose a broader semiotic ground, and in practice, I argue that this anchoring takes the form of chronotopes, that is timespace configurations that stabilise what is made salient and render particular construals socially recognisable. Here, following Blommaert (2020: 58), I do not treat chronotopes as cognitive schemas but as semiotic configurations, that is, active arrangements of time and space that organise action and evaluation. As Lorino and Tricard (2012: 213) emphasise, the chronotope functions as “the matrix of meaning”, materialising the interdependence of time and space while endowing participants with characteristic roles and competences. In this sense, chronotopes not only stabilise attentional patterns but also anchor identities in evaluative frames, legitimising some subject positions while delegitimising others (Blommaert and De Fina, 2016).

Against this background, I suggest chronotopes function as chronotopic construals – configurations that frame time, space, agency, and identity in ways that both enable and constrain ideological readings. Central to my argument is that attention distributions derive their ideological force only when chronotopically anchored, which supplies the socially recognisable ground within which salience becomes meaningful. Consequently, CCDA operationalises how these stabilised attentional patterns are discursively realised. Hart’s (2025: 78–104) identification strategies thus serve as the linguistic mechanisms that enact chronotopic construals, operating within the narrative terrain to manage salience. They can be most productively approached through the heuristic questions below proposed in Table 2.

Table 2.

Identification strategies for attentional distribution and ideological construal.

Strategy	Function in discourse	Heuristic question
Inclusion/exclusion	Regulates visibility to erase entities or redirect accountability.	Who or what is included or omitted, and what is the ideological consequence of this silence or presence?
Foregrounding/backgrounding	Assigns interpretive priority to shape moral and emotional focus.	Which elements are emphasised or suppressed, and how does this guide the reader’s judgment?
Profiling	Directs attention to specific event facets to guide causal and evaluative judgment.	Which specific aspect of the event (e.g. agent, act, outcome) is brought into focus for interpretation?
Trajector/landmark alignment	Configures agency and blame by establishing cognitive prominence or obscurity.	Who is positioned as the primary focal point (trajector) and who is relegated to the background (landmark)?
Scanning	Modulates event perception to intensify or mitigate narrative and moral impact.	Is the event presented as a dynamic, unfolding action or as a static, summarised fact?

The strategies often produce what Hart (2025: 81–86) terms mystification, a discursive effect that obscures agency and diminishes perceived responsibility. For instance, in coverage of the 2018 Gaza protests, NBC reported “Israeli forces have killed more than 100 demonstrators” and The Guardian wrote: “Israeli forces shot and killed 58 Palestinians”, both construing a full-agent patient chain. By contrast, the NYT used the intransitive “28 Palestinians die in protests”, an intransitive construction that leaves agents unexpressed (see more discussion in Hart, 2020). Such patterns contribute to van Dijk’s (1998) “preferred models” of events, legitimising dominant ideologies by concealing or minimising institutional responsibility, as likewise demonstrated in Marín-Arrese (2002), where impersonalisation strategies function to mystify agency and elite accountability while reflecting broader ideological interests in news discourse.

Data

This study analyses a purposive sample of 312 hard news reports published by the NYT between 7 October 2023 and 7 April 2024. To maintain focus on the newspaper’s institutional voice, the dataset excluded opinion pieces, editorials, and feature articles. Articles were collected through targeted keyword searches (e.g. Palestine, Gaza, Hamas, Israel) and filtered by publication date and section (e.g. “Global News”). The dataset was organised using ATLAS.ti25, a qualitative data analysis software, where each article was assigned a unique identifier and full metadata to ensure transparency. Although basic lexical tools such as word frequency counts provided initial orientation, no automated coding or content analysis was performed. The core analysis was conducted manually, with ATLAS.ti25 serving as a qualitative workspace to support iterative coding, memo writing, and the analysis of code co-occurrences. This enabled the identification of consistent and strategically significant discursive patterns in attentional distribution, agentive construal, and moral framing. This approach also allows for a context-sensitive examination of discursive mechanisms, including syntactic profiling, metaphorical reclassification, focusing on their rhetorical function and ideological effects rather than their raw prevalence.

The systematic analysis of the complete dataset was essential for mapping these qualitatively consistent patterns. These patterns provide the context for selecting articles that function as paradigmatic cases (see Pavlich, 2010) for close reading in this paper. These articles were chosen for their capacity to clearly and typically illustrate the broader discursive patterns observed across the NYT coverage. Selection was based on theoretical salience, defined by their clarity, typicality and explanatory power in illuminating the identified patterns, rather than statistical frequency. This aligns with the CCDA principle that ideological mechanisms are best exposed through detailed analysis of exemplar texts, thereby building on but departing from the existing quantitative studies of the 7 October violence (e.g. Johnson and Ali, 2024, 9 January; Pinker, 2025) by revealing how the linguistic and cognitive structures that underlie dominant narratives configure violence as rational and necessary.

Findings

Across the NYT reports, one of the most consequential mechanisms is chronotopic displacement. Narrative time and space are reorganised to isolate Hamas’s actions from their historical and spatial context, thereby foregrounding them as the central point of violence. These attentional asymmetries run throughout the coverage but emerge most clearly in the immediate aftermath of October 7. During the first 48 hours of coverage, the NYT published stylised timelines narrating the events of 7 and 8 October 2023. Critically, in Text 1, antecedent Israeli actions, including repeated military incursions in Jenin and the 2021 raid on Al-Aqsa Mosque, are not entirely absent. However, they are relegated to a retrospective bulleted list under the generic rubric “Here is a summary of some of the main events of the conflict”, a framing that signals informational relevance while stripping these events of narrative weight. Syntactically, although many of these actions appear in active constructions, they are consistently embedded within reactive causal frameworks, establishing from the beginning of the 7 October coverage a logic of Israeli response rather than initiation. Across the list, Israeli violence is framed not as autonomous action (X), but as the consequence of prior Palestinian provocation (X because of Y):

• military raids are linked to the killing of Israeli civilians;

• the killing of 166 Palestinians is justified as occurring after a spate of terrorist attacks;

• kidnappings by Hamas prompt Israeli attacks;

• rocket fire elicits military strikes;

• suicide bombings are cited as grounds for an “incessant war”

These causal relations are realised through a range of syntactic resources – because clauses (causal subordination), after clauses (temporal sequencing with causal implicature), participial resultatives (e.g. prompting), and prepositional adjuncts (e.g. in response to), all of which linguistically encode a discourse of justified retaliation. Even when active voice is employed, the embedding of Israeli actions within these justificatory structures displaces volition and reframes agency as externally compelled. The result is a construal in which past Israeli force is rendered necessary and rational, abstracting responsibility and compressing temporality in ways that do not merely background these actions but structurally remove them from the reader’s evolving simulation of causality. By delimiting what falls within the scope of attention, this narrative logic narrows the conceptual field through which moral evaluation and causal attribution can occur, emphasising an asymmetry introduced at the very start of the timeline.

Text 1: Screenshot from 7 October 2023.

Text 2: Screenshot from 8 October 2023.

In contrast, Hamas’s actions are syntactically prominent, lexically intensified, and saturated with volitional agency. In Text 2, for instance, clauses such as “Hamas launched an attack”, “fighters crossed” and “stormed bases” exhibit high transitivity and telic processes anchored by agentive subjects, foregrounding Hamas as the initiator of violence. Verbs like launched, stormed, seized, and attacked activate a force-dynamic schema associated with immediacy and aggression, constructing a cognitively salient image of deliberate, coordinated assault. This hyper-agential portrayal is reinforced by referential choices, with Hamas is consistently labelled as “militants”, “fighters”, “armed Palestinian group” or “terrorists” – nominative strategies that render it illegitimate and threat-laden. Notably, although the timelines span both 7 and 8 October, the narrative properly begins only in Text 2, reinforcing the illusion that Hamas initiated the violence on 7 October 2023. The entry at 6:35 a.m. reads, “The first sirens warn. . . This is the start. . .”, a construction in which proximal deixis (this) and the definite article (the start) serve as deictic anchors.

These linguistic choices establish a temporal zero-point that narrows the attentional frame, foregrounding Hamas’s actions as temporally immediate, volitional, and morally unanchored, while simultaneously erasing the buildup of Israeli aggression. By positioning this moment as originary, the timeline chronotopically severs the events of 7–8 October from their historical grounding in occupation, repeated raids, and systemic violence. Hamas is thereby construed as erupting into violence spontaneously, disconnected from the structural conditions that precede and provoke it. This schematic narrowing of scope exemplifies attentional gapping, the exclusion of causally relevant information from the mental simulation of events, which in turn distorts both causal attribution and moral evaluation.

Such framing is also consistently reinforced through trajector-landmark realignment, agent suppression, and selective event sequencing, each of which contributes to the systematic obscuring of Israeli agency. By profiling Hamas as the initiating trajector (tr), that is, the figure in cognitive focus, and relegating Israel to a non-salient or omitted landmark (lm), the discourse reconfigures the force-dynamic structure of the event. Responsibility is further obscured by commencing the narrative midstream, bypassing prior Israeli provocations. This reflects a process of initial path profiling, in which the causal ground (G), representing structural conditions such as occupation, blockade, or military aggression, is excluded from the attentional frame. As illustrated in Figure 1, Hamas (A) is positioned as the trajector directing volitional force toward a patient (P), while G is both visually and conceptually omitted, severing the causal chain and framing Palestinian violence as abrupt and unprovoked. Together, these mechanisms produce a rupture effect as Israel’s role as initiating energy source is erased, and Palestinian resistance is recast as excessive, sudden, and self-originating.

Figure 1.

Initial path profiling with omitted causal ground (G).

This epistemic configuration is further amplified through nominal labelling, which imposes divergent construals on otherwise continuous episodes of violence in Gaza. Across the dataset, the same events are variously nominated as war (945 instances), attack (464), conflict (293), assault (97) and retaliation (38). Each of these nouns activates a distinct schematic frame that organises agency, legitimacy, and moral accountability in different ways. What follows is not an analysis of verb argument per se, as in Hart’s (2020, 2025) discussion of transitive versus intransitive construals, but of how nominalisations package events into schematic configurations, selectively profiling portions of an underlying action chain. Consider the following examples:

Text 3: The war continues to take a heavy toll on those gathering the news (4 November 2023).

Text 4: Louisa has acknowledged the deaths of Israeli civilians in the Hamas attack (5 December 2023).

Text 5: The conflict has turned Gaza into a “graveyard for thousands of children” (6 January 2024).

Text 6: Israel had vowed after the deadly Oct. 7 assault to hunt down Hamas leaders wherever they are. . .The Hamas-led assault on Israel killed about 1,200 people, and 240 others were taken back to Gaza as captives (3 January 2024).

Text 7: Much of Gaza has been left in ruins as Israel bombards the territory in retaliation for the attacks on Oct. 7 (14 February 2024).

In war (Text 3), violence is construed through a symmetrical agentive schema, profiling two volitional actors (A¹ and A²) as co-equal participants in a bounded conflict (see Figure 2). This framing abstracts violence from its historical asymmetries, foregrounding mutuality while backgrounding occupation, siege, and structural domination. By directing attention equally to both parties and excluding antecedent causes, war mystifies the ideological imbalance and legitimises disproportionate force under the guise of symmetrical engagement. By contrast, attack in Text 4 functions as a nominal construal of a fully transitive action-event schema in which a volitional agent (A) directs energy to a patient (P), foregrounding causality and moral accountability within a tightly bounded attentional frame (see Figure 3).

Figure 2.

“War” as a nominal construal of a symmetrical agentive event.

Figure 3.

“Attack”/“Assault” as nominal construals of a fully profiled transitive action chain.

The phrase “the Hamas attack” uses an attributive noun modifier to construe Hamas as the source of energy, presupposing deliberate harm while suppressing prior Israeli actions. This construal isolates Palestinian violence from its structural antecedents, rendering it abrupt, hyper-agentive, and morally unprovoked. Assault, as in Text 6, instantiates the same transitivity schema as attack but intensifies the evaluative load, presupposing not just intentionality, but criminality, indiscriminateness, and exceptional brutality. The phrase “Hamas-led assault” functions metonymically, collapsing actor variability into a unified, hyper-agentive entity while erasing political context. Both labels foreground Palestinian agency while defocusing Israeli responsibility, producing a morally asymmetrical event construal in which culpability is selectively assigned.

The noun conflict evokes a reduced process schema resembling an intransitive construal, with a single trajectory (T) is profiled, while no agent or energy source enters the attentional frame (see Figure 4). For instance, in Text 5, conflict occupies the subject position in a causative construction “has turned Gaza into a graveyard”, yet no human actor is profiled. Israeli force, the actual energy source, is entirely mystified. Here, conflict functions as a process noun that schematises only the downstream effects of violence while excluding its origin. This contributes to agent-based mystification while simultaneously presupposing a dyadic structure, subtly implying parity between Israelis and Palestinians. On the other hand, retaliation in Text 7, functions as a nominal construal of a cause-effect schema in which Israeli violence is framed not as initiatory but as compelled response to Hamas’s prior actions (see Figure 5).

Figure 4.

“Conflict” as a nominal construal of an intransitive event with defocused agency.

Figure 5.

“Retaliation” as a nominal construal of a transitive action schema with selective attentional expansion.

This construal selectively widens the attentional frame just enough to legitimise Israeli use of force, while narrowly bracketing antecedent conditions such as occupation, blockade, and repeated military incursions. Crucially, this is not simply a matter of temporal omission but of construal, specifically, the use of semantic bundling through profiling shifts and metonymic contiguity, which collapse distinctions between combatants and civilians. Across the dataset, and especially in the first 2 months when narrative templates solidify and attentional patterns become routinised, these strategies are especially pronounced. Referential conflation and metonymic profiling operate to diffuse moral culpability across the Palestinian population, constructing a representational logic in which civilian subjects are discursively repositioned as legitimate targets. Consider the following texts taken from the first 2 weeks of NYT coverage following 7 October 2023:

Text 8: The crowds of civilians in Gaza cheering the Hamas fighters underscored the extent of anti-Israel hostility among Palestinians (8 October 2023).

Text 9: She did not see a meaningful distinction between supporting Palestinians and supporting Hamas, she said (13 October 2023).

Text 10: “If Hamas kills Jews, and people are partying and celebrating that in Gaza, we hear that you Jews had it coming” (15 October 2023).

Text 11: “Just eliminate them, eradicate them all. . . the Palestinians next door are led by a group using the money they get to spread terror” (15 October 2023).

Here, the phrase “crowds of civilians. . . cheering the Hamas fighters” in Text 8 initiates the metonymic projection as it forms a syntactic and conceptual gestalt in which proximity and simultaneity are reinterpreted as ideological alignment. The verb underscored adds epistemic weight, implying evidentiary confirmation of generalised hostility. The attentional frame projects from the surface profile (Palestinian civilians, A) to an implied target (Hamas, B), enabling a moral transference (A → B) that reframes civilians as ideologically complicit (see Figure 6).

Figure 6.

Metonymic profiling in semantic bundling.

Text 9 escalates this logic through relational conflation, where the reported assertion, “she did not see a meaningful distinction between supporting Palestinians and supporting Hamas” (13 October 2023), forcibly collapses two distinct conceptual domains, that is, ethnonational identity and armed resistance, into a single attentional object. The syntactic structure encodes equivalence rather than comparison, reclassifying expressions of Palestinian solidarity as latent militancy. Though presented as reported speech, its unqualified inclusion in the NYT legitimises the conflation, recalibrating the attentional lens to render Palestinian identity coterminous with terrorist alignment.

This pattern intensifies in Text 10, which introduces affective-indexical profiling. The juxtaposition of “if Hamas kills Jews” with “people are partying” constructs a pseudo-causal schema, mapping affect (celebration) onto intention (killing). Here, celebration is metonymically reframed as moral endorsement, forming a co-profiled event structure in which Gaza’s population becomes emotionally fused with acts of terror. Unlike Text 9, where the conflation is reported indirectly and subtly reinforced through framing, Text 10 presents the logic in direct quotation, amplifying its affective charge and collapsing emotional expression into ideological complicity. Finally, Text 11 culminates in full-scale referential generalisation, where “the Palestinians next door are led by a group. . .” activates a part-whole metonymy that casts Hamas leadership as synecdochic for the entire Palestinian population. The repeated pronoun them, offered without qualification, performs deictic erasure, collapsing intra-group distinctions – civilian, militant, political – into a singular, targetable entity. The imperative “just eliminate them, eradicate them all” completes the semantic bundle, transforming rhetorical slippage into a directive for totalised violence. Across these texts, the cumulative effect is a discursive logic in which co-occurrence becomes complicity, identity becomes ideology, and population becomes target.

Building on the cumulative logic of conflation and deictic erasure, there is a strategic activation of eliminatory conceptual frames through metaphor and event-structure schemas that can be observed across the dataset. This functions as a mode of demonisation not through explicit delegitimation, but by reconfiguring the ontological status of the population. Though not numerically dominant, such instances exert disproportionate attentional force, recalibrating interpretive frameworks to render large-scale violence cognitively coherent and ethically permissible. Rather than merely abstracting or exaggerating, this mechanism shifts the construal of Palestinians from political or civilian actors to existential contaminants requiring elimination. It embeds itself within the discursive-cognitive system through the routinised encoding of eradication as rational necessity. Consider the following examples:

Text 12: The calls for a war like no other to root out Hamas “once and for all”, a consensus in Israel’s national unity government and much of Israeli society, will blunt their voices for now. The peacemakers are in the minority as a devastating invasion of Gaza looms (22 October 2023).

Text 13: Israeli officials point out that Hamas is also active in the West Bank and say that many of those clashes resulted from the military’s efforts to root out militants. Three Israelis have been killed in attacks by Palestinians since Oct. 7 (15 November 2023).

Text 14: Israel says it is rooting out Hamas activity at the medical centers, which it says Hamas has used to hide military operations – accusations it has made about multiple hospitals in Gaza (19 February 2024).

Across Texts 12–14, the repeated use of “root out” enacts a metaphorical construal grounded in the force-removal image schema, which configures the target not as a discrete adversary but as an internally embedded contaminant. This schema, drawing on embodied experience of purging or weeding, profiles Israel as a volitional energy source exerting downward force on a threat imagined as spatially and morally invasive. Crucially, this metaphor does not only intensify the perception of danger, but it also restructures the attentional field, expanding the scope of what may be construed as targetable while displacing the presence of those affected. The internalisation of the threat, its being within rather than opposed to the space of operation, renders the surrounding environment, whether Gaza, the West Bank, or hospitals, not merely collateral but complicit, cognitively repositioned as part of the object of elimination.

In Text 12, the phrase “a war like no other to root out Hamas once and for all” activates a telic and totalising frame. Hamas is referentially indeterminate, no leadership, combatants, or specific locations are identified, enabling metonymic projection across the entire population of Gaza. The modifier “once and for all” encodes cognitive finality, framing the intended violence not as tactical or limited, but as morally conclusive. This is reinforced by the clause “a devastating invasion of Gaza looms”, which enacts patient suppression by reducing Gaza to a depopulated spatial target. Palestinian civilians are erased from the attentional frame, and the monovalent process noun invasion profiles only the event itself, omitting both agent and patient. This exemplifies upstream action-chain truncation: the agent (Israeli political and military actors, A) is profiled as the trajector (tr), while the patient (Palestinian civilians, P) is excluded from view (see Figure 7).

Figure 7.

Profiling in patientless active construction.

This is not passive omission but radical exclusion, a deliberate discursive strategy that truncates the action chain by profiling the agent and the action while omitting the patient, thereby foregrounding Israeli agency (e.g. “national unity government”) as unified and volitional, while displacing Palestinian suffering to the conceptual periphery. Linguistically, this construal renders Gaza as an abstract zone of military intervention, stripped of social and human dimensions.

In Text 13, the same metaphor is redeployed within a retrospective, justificatory construction: “clashes resulted from the military’s efforts to root out militants”. The nominalisation “efforts” performs agent defocusing, while the low-transitivity structure (“clashes resulted from”) suppresses causal direction and consequence. Unlike “Hamas”, which specifies a distinct political entity with identifiable agency and declared objectives, the substitution with “militants” creates strategic vagueness. The term is semantically broader, potentially encompassing any armed individual, and spatially diffuse, lacking clear affiliation. This vagueness is pivotal because it permits the ‘rooting out” metaphor to persist without anchoring responsibility to a specific accountable group. As in Text 12, the patient of the military action is not specified, there is no reference to civilians, casualties, or those affected by the clashes. However, in this instance, suppression is achieved through abstraction rather than outright erasure, the grammatical choices displace moral consequence not by excluding it altogether, but by rendering it cognitively inaccessible. The metaphor thus functions less as a call to action, as in Text 12, and more as a post hoc rationalisation of state violence, reframing it as the natural outcome of defensive necessity.

Text 14 extends the metaphor into the domain of civilian infrastructure: “rooting out Hamas activity at the medical centers”. Here, the construal undergoes semantic reprofiling. Hospitals, normally framed as neutral and protected, are discursively transformed into operational sites of embedded threat. The force-removal metaphor remains intact but is now mapped onto institutions of care, collapsing functional boundaries between humanitarian space and military target. This framing narrows attention to the presumed presence of Hamas while suppressing the visibility of those who occupy these spaces, that is, doctors, patients, displaced families. The metaphor thereby legitimises the extension of state violence into civilian zones by reframing them as inseparable from threat.

Discussion

The NYT’s coverage in the 6 months following 7 October 2023, lacks the overt rhetoric that has historically foreshadowed mass atrocities (see e.g. Landry et al., 2022), nor does it exhibit the accusative speech acts that, as Normand (2016) argues, brand enemies as literal incarnations of evil whether through direct or indirect lexicons or figurative tropes. This absence is analytically crucial, revealing not lack of demonisation but its reconstitution into a more structurally insidious form, which I term subtle demonisation. As the findings demonstrate, this mechanism unfolds through nuanced discursive processes embedded within the routine journalistic practices that present themselves as objective. It works through systematic patterns of construal that manipulate the distribution of attention, cumulatively constructing a malign reality in which Palestinians are reclassified not merely as adversaries but as existential threats. Here, the accusatory force is never declared but enacted through the narrative structure itself, advancing through implied menace rather than explicit accusation. Under the guise of impartiality, the coverage demonises Palestinians, without naming them as such, repositioning them as targetable entity and a profound danger to ontological security.

Crucially, the distinction between demonisation and its subtle variant here is one of form, not consequence, for the latter still directly manufactures the stark ideological polarisation central to van Dijk’s (1998) work by bifurcating the social world into a virtuous “us” and a monstrous “them”. Such polarisation, in turn, enables the decisive shift that Barker (2007: 119) describes, where rationale moves decisively from containment and exclusion to expulsion or destruction. That outcome fulfils the core function of demon construction, producing a target “that can be exterminated without regret” (Keen, 1987). ² The apotheosis of this logic is laid bare in the “preferred model” of violence it installs (see van Dijk, 1998). As El-Akkad (2025) articulates with devastating clarity, this model dictates that those in power must reason as follows:

Yes, this is tragic, but necessary, because the alternative is barbarism. The alternative to the countless killed and maimed and orphaned and left without home, without school, without hospital, and the screaming from under the rubble and the corpses disposed of by vultures and dogs and the days-old babies left to scream and starve, is barbarism (p. 5).

Consequently, the obfuscation of agency and causality must be understood not simply as mystification (see Hart, 2014, 2020), but as the core operative mechanism of this demonisation. It repositions systemic violence from a contingent event into a necessary and justified imperative. This represents a significant departure from analyses that frame the NYT’s coverage primarily in terms of media bias (e.g. Jackson, 2023; Zelizer et al., 2002 among others). Where such studies identify a failure of objectivity, my analysis uncovers a successful ideological project, the production of a social reality in which atrocity becomes logically inevitable. Such a project, enacted through nuanced linguistic construals, necessitates the cognitive-linguistic approach advanced here, as quantitative methods alone cannot capture this qualitative restructuring of salience.

Central to this performative is the manipulation of the chronotope. As discussed earlier, chronotopes are not mere backdrops but semiotic matrices that organise moral legitimacy and responsibility (see Blommaert and De Fina, 2016). The NYT’s persistent resetting of the narrative clock to 7 October 2023, constitutes a profound chronotopic rupture. This is a powerful discursive act that displaces the longue durée of settler-colonial occupation, blockade, and asymmetrical violence – a context where violence aims not merely at territorial conquest but at existential erasure, targeting history, culture, and connection to the land (Al Jazeera, 2025, 23 August). By constituting October 7 as the temporal ground-zero, the coverage actively produces a narrative world in which Hamas’s violence erupts ex nihilo, stripped of antecedent cause and thus of political intelligibility.

Here, it is precisely this chronotopic construction that renders de-demonisation proposed in Normand (2016: 181), not merely inadequate but also analytically untenable in a settler-colonial context. Her framework is articulated through a peacemaking lens that presumes procedural symmetry between Israeli and Palestinian elites. This presumption directs analytic attention to hostile epithets while erasing the structural asymmetry that organises the relation between occupier and occupied. In doing so, it displaces the foundational issue of discursive construals that recast occupation and apartheid practices as a balanced dispute, a structure the UN Special Rapporteur insists must be brought “to a complete end” (Lynk, 2022, 25 March). This conceptual conflation risks mistaking a discursive effect for a neutral baseline, relocating the problem from one of domination and dispossession to one of civility and dialogue. In this case, the chronotope functions as the master frame that governs all subsequent attentional distributions, guiding the reader’s cognitive focus toward Hamas/Palestinian agency and away from the structural aggression that defines the colonial present.

The consequences of this chronotopic displacement emerge in the asymmetrical construal of agency. The initial timelines consistently embed Israeli violence within reactive causal frameworks (e.g. “raids after killings”, “strikes in response to rockets”), linguistically encoding a logic of justified retaliation. Hamas’s actions, by contrast, are profiled through high-transitivity, agentive clauses that foreground volition and initiatory force. This attentional asymmetry, initial path profiling, omits the causal ground (G) of occupation and effects a discursive reordering of moral time. Hamas is thus locked into the chronotope of the “perpetual aggressor”, suspended in a present of inherent threat, while Israeli violence is compressed into the urgent temporality of the “necessary now”, a grammar of security that naturalises retaliatory force as the only logical response to inexplicable evil. This asymmetry is systematically reinforced through nominal labelling, where subtle demonisation operates at the schematic level. Terms like war (symmetrical schema), attack/assault (hyper-agentive transitivity), conflict (agentless intransitivity), and retaliation (reactive causation) are not neutral variations but distinct image-schematic configurations that govern blame assignment. These nominalisations routinise a reality where immense violence appears not as disproportionate assault but as inevitable within a framed “conflict” or justified “retaliation”. Thus, demonisation proceeds by implication, casting Hamas as the archetypal initiatory antagonist whose existence necessitates perpetual war.

The most pernicious effect lies in its facilitation of referential conflation as the central mechanism in the demonisation process. The findings demonstrate how metonymic profiling (e.g. “crowds of civilians cheering”) the reported erasure of any “distinction between supporting Palestinians and supporting Hamas” and the affective indexing of “people are partying” diffuses agency and moral culpability across the population as a whole. This discursive logic transforms co-occurrence into complicity and recasts ethnonational identity as ideological alignment with militancy. It thus creates the conceptual conditions for what Mbembe (2003: 38–40) describes as a “death-world” where civilians are no longer recognised as a protected category but absorbed metonymically into the threat itself, rendered morally indistinguishable from combatants and therefore expendable. This conflation culminates in the eliminatory dimension of subtle demonisation through the activation of metaphors such as “root out” for instance, which resonate with the wider counterterrorism discourse institutionalised in the post-9/11 era. Within this force-removal schema, the target is not construed as a political adversary on a battlefield but as an invasive contaminant embedded within a host body, whether Gaza, a hospital, or the broader civilian population. The metaphor is performative, re-ontologising its object and cognitively restructuring the attentional field so that entire spaces and populations become targetable. Coupled with patient-suppressing syntax that truncates the action chain, this framing enacts profound dehumanisation. Palestinian lives are not simply backgrounded as they are also conceptually erased from the scene of action, their suffering rendered narratively invisible and thus morally unintelligible. This is the logical endpoint of demonisation, where the target is no longer an enemy to be defeated but a demon to be destroyed.

Conclusion

This paper has shown that the NYT’s reporting in the 6 months following 7 October 2023 did not simply recount violence but actively configured a discursive world in which overwhelming atrocity was rendered rational and necessary. What becomes intelligible in such coverage depends on how attention is distributed, how agency is encoded, and how causality is constructed. Drawing on Hart’s (2025) framework of Cognitive Critical Discourse Analysis, the analysis revealed how chronotopic displacement reset history to a “day zero”, how Palestinian agency was amplified while Israeli volition was suppressed, and how civilians were conflated with combatants through metonymic profiling. Together, these mechanisms enacted a process of subtle demonisation, a discursive reordering that did not overtly vilify Palestinians but instead recast them as morally expendable through the seemingly objective routines of journalistic narration.

The contribution of this study is threefold. Conceptually, it advances work on demonisation by theorising how it can operate covertly, not through hostile epithets but through ordinary narrative choices of sequencing, perspectivisation, and labelling. Analytically, it demonstrates the value of bringing Hart’s CCDA into dialogue with Bakhtin’s (1981) chronotope, showing how time-space configurations underpin asymmetries of agency and causal attribution. Normatively, it highlights the ethical stakes for journalism, which must be measured not only by sourcing practices or factual accuracy but also by the discursive logics that decide whose lives are narratable, grievable, and worthy of protection.

The stakes of this critique extend beyond academia into the realm of ongoing consequence. This study suggests that the institutional practice of objectivity, when grounded in unexamined chronotopes, can function as a dangerous fiction, providing moral cover for atrocity while maintaining the appearance of impartiality. History, from Nazi Germany to Rwanda, demonstrates how discourse prepares the ground for violence. In the present, journalism’s narrative choices risk contributing to this dynamic, not merely failing to document atrocity but framing it in ways that render it thinkable, and even necessary.