Modeling the audience's perception of security in media discourse

The study investigates how U.S. media discourse shapes audience perceptions of an unconventional security issue, focusing on Confucius Institutes. Building on post-9/11 U.S. security discourses characterized by a Self–Other dichotomy, continuous threat construction, and emotive resonance, the paper argues that more empirical work is needed to show how linguistic configurations influence public thinking. The research combines securitization theory with discourse analysis to examine the nexus between media practices and audience affective and cognitive processes. It addresses underexplored unconventional security issues (culture and education), asking how media texts dialogically and discursively construe Confucius Institutes as a threat and to what extent audiences are cognitively modeled as sharing these beliefs. The purpose is to provide a comprehensive understanding of how such issues are framed in media and their potential impact on public perception.

The paper situates itself within critical security studies and discourse analysis. It reviews the Copenhagen School’s securitization theory, emphasizing the speech act and the evolution toward processual, bottom-up understandings that highlight institutionalization, routinization, audience, and context. It introduces dialogism and intersubjective stance resources from systemic functional linguistics (ENGAGEMENT and ATTITUDE; Martin, White, White PRR) to analyze alignment and positioning. It incorporates proximization theory (Cap) to explain how linguistic deictic choices cognitively legitimize preventive measures by portraying an external Other (ODC) closing in on an in-group (IDC) across spatial, temporal, and axiological dimensions. The U.S. security discourse literature underscores dualistic structures and emotive perceptions of enmity, informing the study’s combined framework for analyzing how media normalize and routinize security definitions.

Design: A combined discourse analysis informed by securitization theory, dialogic engagement and attitude resources, and cognitive proximization. A process-oriented framework examines how securitization emerges through discourse over time, focusing on media actors and audiences as subjects-in-process. Case context: Confucius Institutes in the United States are traced across three stages: positive reception (2004–2013), cognitive transformation (2014–2020), and post-closure (2021–present). Key events and actors include university closures (Chicago, Pennsylvania State), academic organizations (AAUP, NAS), and policy moves (NDAA FY2019; designation of CI U.S. Center as a Foreign Mission in 2020). Data collection: Using ProQuest, the study collected coverage of Confucius Institutes in The New York Times and The Wall Street Journal from 2004 to 2022. After screening, 124 texts were compiled (Size=145,437; Tokens=125,289; Types=12,022). Publication trends peak in 2014 and 2020, mirroring socio-political events. Sample for qualitative coding: Thirteen texts were selected for in-depth analysis (5 NYT, 8 WSJ) centered on pivotal events (2014 AAUP pushback, 2020 foreign mission designation). Selection required the term “Confucius Institutes” in the article title. The sample totaled approximately 9,437 words. Analysis: Texts were coded for ENGAGEMENT categories (DISCLAIM, PROCLAIM, ENTERTAIN, ATTRIBUTE) using BFSU Qualitative Coder 1.2, and co-occurring ATTITUDE realizations (AFFECT, JUDGMENT, APPRECIATION) were noted to assess stance-taking and alignment. Corpus-based indicators supported proximization analysis, including key lemma frequencies, collocations, and concordances (e.g., MI and log-likelihood for co-occurrence of “Confucius Institutes” with “Beijing”).

• Media engagement and quoting: The 13-text sample contains 137 quotes (about 1.45 per 100 words). On average there were 13 instances of intersubjective stance resources per text, amounting to 1.79 occurrences per 100 words. There is no significant difference between The New York Times and The Wall Street Journal in stance resources (log-likelihood=0.1607, p=0.6885).
• ENGAGEMENT distribution: ATTRIBUTE dominates stance resources (count=96; 56.8%), followed by PROCLAIM (count=60; 35.5%), with less frequent DISCLAIM (count=10; 5.9%) and ENTERTAIN (count=3; 1.8%). This pattern reflects reliance on external voices to project authorial stance while preserving journalistic objectivity.
• Attitudinal language: ATTITUDE resources co-occur with engagement, especially AFFECT and JUDGMENT terms conveying evaluative stances toward Confucius Institutes (e.g., bullying, threatening, heavy-handed; frustrated, displeasure; unacceptable). These evaluations help construct CI as morally problematic and emotionally salient.
• Identity framing and securitizing narrative: Media texts amplify external judgments (e.g., AAUP statements) that portray CIs as an arm of the Chinese state infringing academic freedom, normalizing a shift from cultural-education framing to political-security framing.
• Cognitive proximization: The discourse constructs an IDC–ODC dichotomy with the U.S. academic and political sphere at the center and CIs aligned with China/CCP at the periphery, approaching across spatial, temporal, and axiological dimensions.
  • Spatial and institutional proximity: Key lemmas indicate ODC components (Confucius Institutes/Classrooms, Chinese government, CCP, China, Beijing, Hanban; total frequency 3.28%) and IDC components (Americans, America, universities, colleges; 1.47%). Actions and consequences are encoded via lemmas such as interfere, disseminate, steal, bully, expand (0.13%), and influence, expansion, control, challenge, threat (0.32%), linked to academic freedom and security (0.24%), and soft power, propaganda, contracts (0.17%). Extreme negative axiological markers include authoritarian, genocide, theft, massacre (0.06%). “Confucius Institutes” collocates with “Beijing” (MI plus log-likelihood=2.53, p<0.05).
  • Temporal proximity: Past incidents (e.g., censorship at conferences) are invoked to project imminent and future risks in U.S. contexts, justifying preventive measures.
  • Axiological proximity: The discourse contrasts democracy versus authoritarianism, positioning CI as carrying alien, incompatible ideology encroaching on U.S. values (e.g., research theft, speech censorship).
• Bottom-up securitizing dynamics: Media reproduce and amplify authoritative external voices (academia, officials), aligning public perception with a securitizing narrative that influenced broader societal and policy responses (e.g., university closures, legislative actions).

The findings show how media linguistically model audience perceptions to support securitization of an unconventional issue. By foregrounding external attributions (ATTRIBUTE) and authorial endorsements (PROCLAIM), journalists construct a heteroglossic space that nonetheless aligns readers with a particular evaluative stance toward Confucius Institutes. The recurrent co-occurrence of engagement and attitudinal resources translates external judgments into shared, seemingly consensual knowledge, shaping moral and affective orientations among audiences. Cognitive proximization across spatial, temporal, and axiological axes legitimizes urgent and exceptional responses by depicting CI as an approaching, ideologically incompatible actor linked to the Chinese state. This bottom-up accumulation of routinized discourses contributes to audience uptake and normalization of a security framing, addressing the research questions about how media texts dialogically construct threat and model audience cognition. The broader relevance lies in demonstrating that unconventional security issues can be securitized through iterative, mediated practices that mobilize intersubjective stance and proximization, complementing speech-act-centered accounts in securitization theory.

The study proposes and applies a combined discourse-analytic framework integrating securitization theory, intersubjective stance (ENGAGEMENT and ATTITUDE), and cognitive proximization to examine U.S. media coverage of Confucius Institutes. It finds that American media enact bottom-up securitizing moves that construct CIs as an existential threat through strategic sourcing and evaluative language, aligning audiences affectively and cognitively. Proximization mechanisms across spatial, temporal, and axiological dimensions legitimize urgency and exceptional measures. The work contributes an empirically grounded, process-oriented account of how language shapes perception and belief in unconventional security contexts and underscores that stances and attitudes are social positions operating within current socio-political environments.