No more binaries: a case of Pakistan as an anomalistic discourse in American print media (2001–2010)

The study investigates how American media discourse, particularly in the post-9/11 context, constructed and negotiated the identity of Pakistan and its leader General (President) Pervez Musharraf within and beyond the dominant Us vs. Them and friend vs. foe binaries. It argues that media discourse is central to shaping public opinion and reflects underlying US foreign policy ideologies that required Pakistan’s cooperation after 9/11. Focusing on the personalization of politics and the prominence of leaders in international media, the paper examines Musharraf as the key individual actor representing Pakistan to reveal how in-group/out-group boundaries blur and how Pakistan was framed to align with US strategic goals.

The paper situates itself in the corpus-assisted discourse studies (CADS) tradition, integrating Corpus Linguistics (CL) and Critical Discourse Analysis (CDA). It reviews the evolution of this synergy (Baker et al., Partington, Krishnamurthy, Stubbs) and addresses mutual critiques: CDA’s alleged small samples, opacity, and cherry-picking; and CL’s purported neglect of broader sociopolitical context and difficulty addressing absences. The study adopts Sinclair’s model of the lexical item (core/node, collocation, colligation, semantic preference, semantic prosody) and clarifies operational collocation parameters (window span, frequency, statistical validation). For CDA, the analysis draws on Reisigl and Wodak’s nomination and predication strategies, van Leeuwen’s taxonomy of social actor representation, and van Dijk’s ideological square of Us vs. Them to interrogate how media reproduce and manipulate images of actors within power relations and historical contexts.

Data source and sampling: 509 Time magazine articles referencing Pakistan, spanning post-September 11, 2001 through December 2010, were collected from Time archives (Karl Franzens University of Graz). Time was chosen for its large circulation, weekly format, and international reach. Analytical approach: A mixed-method CADS design combined quantitative corpus analysis with qualitative CDA. The node/core term was Musharraf* (capturing variants including possessive forms). Tools used were GraphColl and WordSmith. Collocation parameters: Collocations were identified within a window span L5:R5. Statistical significance was tested with a t-test (p < 0.05; t > 2; frequency ≥ 10) and association strength with MI² (threshold MI² ≥ 10; frequency ≥ 10; MI² ≥ 9 considered strong). Absolute yearly frequencies were computed for diachronic patterns. Semantic categorization: Collocates were restricted to nouns to align with nomination and predication strategies and feasibility. Noun–noun collocates were manually assigned to over 30 semantic categories (e.g., Politics, Multiple Actors, Countries/Nationalities, Military/Intelligence, Temporal, Reputation, Power/Authority, War/Terror). The top eight categories by percentage were analyzed in depth through expanded concordance lines. CDA framework: Qualitative analysis interpreted statistical patterns within broader political-historical context, employing nomination/predication (Reisigl & Wodak), social actor representation (van Leeuwen), and ideological square (van Dijk) to examine identity construction, power negotiation, and friend–foe framing.

• Diachronic pattern: Mentions of Musharraf peaked in 2007 with a sharp decline afterward; early increases in 2001–2002 relate to the War on Terror and Pakistan–India tensions. Coverage declined markedly after Musharraf’s 2008 resignation, reflecting reduced relevance once out of power.
• Collocation significance and strength: Both t-test and MI² results show content collocates including Pervez, president, Pakistani, Pakistan/Pakistan’s, general, military, Bush, and power. ‘President’ ranks far above ‘general,’ indicating media foregrounded a political identity while downplaying military dictatorship. Bush appears as the sole named foreign leader strongly associated with Musharraf, reflecting bi-personal ties and US–Pakistan relations.
• Semantic categories: Politics (≈21.3%) dominates over Military/Intelligence (≈6.9%), with Multiple Actors (Proper Nouns ≈19.4%) and Countries/Nationalities (≈9.9%) also prominent. This distribution corroborates the emphasis on political framing over military identity.
• Reputation framing: Positive labels such as ally, leader, hero occur, while explicitly negative terms like dictator and dictatorship appear infrequently (dictatorship only twice; dictator four times), evidencing a downplaying of the non-democratic aspect when supportive alignment is desired.
• Mixed support and criticism: Musharraf is alternately supported, directed, and criticized by US actors. Personifications like “Washington” are used instead of naming officials, lending collective US agency while reducing Pakistan to Musharraf as a singular actor.
• Double-game narrative: ISI is portrayed negatively (rogue elements), while Musharraf is framed as attempting to ‘tame’ it, creating a dichotomy where the leader appears committed to the War on Terror amid institutional ambiguities.
• War/Terror framing: Musharraf’s decision to join the War on Terror is depicted as a dramatic U-turn; assassination attempts against him are covered as consequences of this alignment. Anonymous US officials question the intensity of Pakistan’s efforts, keeping criticism low-key.
• Friend vs. foe blurring: Coverage allows Musharraf (and by extension Pakistan) to traverse a gray zone between ally and adversary, undermining rigid Us vs. Them binaries and producing an anomalous, ‘dubious ally’ identity aligned with evolving US foreign policy priorities.

The findings demonstrate that US media discourse strategically foregrounded Musharraf’s political identity to accommodate and legitimize US foreign policy objectives post-9/11, while selectively downplaying his military coup origins when alignment was beneficial. When his actions conflicted with US expectations (e.g., judicial crises, leniency toward militants), coverage accentuated military/authoritarian frames to express displeasure. The strong association with President Bush highlights the personalization of international politics and the role of elite ties in shaping national portrayals. Semantic category distributions and concordance evidence show how power and identity were fluidly negotiated: ‘Washington’ is collectivized and empowered, while Pakistan is personalized as Musharraf, reflecting asymmetrical power relations. Overall, the representation of Pakistan and Musharraf deviates from strict friend–foe dichotomies, instead occupying an anomalous middle ground that serves contemporary policy agendas. This addresses the research question by revealing how media discourse constructs and manipulates identities and alliances to reflect and reproduce geopolitical power dynamics.

The study shows that Time magazine’s 2001–2010 coverage constructed an alternate, anomalous identity for Pakistan and Pervez Musharraf, emphasizing a political leadership persona over military dictatorship to align with US foreign policy during the War on Terror. Mentions peaked around major political-security events (notably 2007), and collocation/semantic analyses confirm a consistent downplaying of non-democratic aspects when supportive narratives were required, while critical frames emerged when policy divergences occurred. This nuanced portrayal blurs Us vs. Them binaries, positioning Pakistan/Musharraf as a ‘dubious ally.’ The paper contributes methodologically by integrating CL and CDA to combine statistical robustness with contextual interpretation. Future research could extend to multiple US outlets, compare across leaders/countries, incorporate additional linguistic patterns (beyond noun–noun), and examine post-2010 evolutions to assess continuity or change in anomalous representations.

The study focuses on a single actor (Pervez Musharraf), a specific decade (2001–2010), and a single media outlet (Time magazine). For feasibility, the corpus analysis is restricted to noun–noun collocations and manual semantic categorization, which may omit insights from other parts of speech or automated semantic modeling. These constraints may affect generalizability and the completeness of representational nuances.