Exploring linguistic features, ideologies, and critical thinking in Chinese news comments

The study addresses a gap in research that has traditionally focused on how authors frame news, by instead examining news comments—short, real-time reader responses—to understand how linguistic forms and critical thinking reveal ideological beliefs. Building on van Dijk’s premise that news discourse is inherently ideological, the authors investigate how critical thinking interacts with linguistic features to inform commenters’ expressed ideologies in the Chinese context. The research aims to identify typical linguistic forms, ideologies, and critical thinking evident in news comments, and to analyze how these constructs inform one another. The paper reconceptualizes critical discourse analysis (CDA) from the audience/commenter perspective to reveal readers’ acceptance or rejection of discursive practices and the power relations behind them.

Under the umbrella of CDA, the paper reviews three major approaches: Fairclough’s Critical Approach (language as socially conditioned discourse/text production and interpretation), Wodak’s Discourse-Historical Approach (macro-level, strategy-oriented, historically informed analysis, often applied to national identity and media), and van Dijk’s Sociocognitive Approach (SCA), which mediates between text and society via social cognition and emphasizes how discourse reproduces social inequalities such as racism. The authors argue that no single approach fully captures their constructs, and therefore adopt SCA as the primary framework while incorporating elements from Fairclough’s three-dimensional model (textual analysis) and Wodak’s focus on social/national ideologies. They further reinterpret CDA from the audience/commenter perspective, positing that commenters, as recipients rather than initiators of discourse, reveal ideological positions and critical thinking through their responses.

Following the hermeneutic orientation of CDA (Wodak & Meyer, 2015), the study operationalized its design by (1) selecting major Chinese social media platforms—WeChat (1.04 billion users), Baidu search (700 million users), and Sina vlog (392 million users); (2) sampling Baidu News columns. Twelve columns were screened; ideologically loaded or highly gendered columns (military/army, gaming, beauty) were adjusted, synthesizing to six: domestic/international affairs, finance, entertainment, sports, technology, and game. Twelve news items with more than 30 comments were finalized (two per topic where possible). (3) Participants included public social media commenters (from Baidu, Sina Weibo, The Paper/Surging News) and 19 undergraduates from a research-based university in Northeast China (3 freshmen, 14 sophomores, 2 juniors; 9 male/10 female) across eight schools. (4) One-on-one, ~20-minute, conversational interviews elicited immediate reactions to sampled news, guided by prompts; sessions were recorded and transcribed, with many students using WeChat text/audio to respond. An initial pool of 25 interviewees (≈50 pieces of information) was narrowed to 19 after removing outliers. Data were integrated into four analytic sections: news content, typical responses, interview responses, and analyses. Analytic procedures drew on SCA’s multi-level steps (semantic macrostructures/topics; local meanings; subtle formal structures; global/local discourse forms; linguistic realizations; context), applying figure-of-speech analysis, lexical/connotative assessments, and contextual interpretation to link language structures with cognition (critical thinking) and ideologies.

• Linguistic features: News comments frequently employed idioms (e.g., “hai ren zhi xin bu ke you, fang ren zhi xin bu ke wu,” “yan zhen yi dai,” “jing dai hua kai,” “zuo wu xu xi,” “qi er yi ju”), signaling shared sociocultural knowledge and ingroup identity. Lexical choices carried connotative meanings and metonymy (e.g., “backyard”), often reflecting regional or dialectal nuances and ingroup–outgroup distinctions. Syntactically, comments showed complexity via parallel structures, metaphors, and repetition, serving to emphasize points and create emotional momentum (e.g., enumerations of sports victories; “Ball, the ball, is just a ball; the country is always behind you”).
• Ideologies and critical thinking: A strong ingroup affiliation and patriotic pride were evident, especially in responses celebrating China’s economic, political, and technological achievements (e.g., support for domestic products; praise for the women’s volleyball team). At the same time, commenters displayed rational, multi-perspective critical thinking, balancing patriotism with pragmatism (e.g., acknowledging the value of learning advanced foreign technologies while supporting domestic products; recognizing that sports are culturally borderless yet taking national stands when politics intrude).
• Topic effects on language and ideology: Politics, economy/finance, and technology topics elicited more formal, structured, and logically organized commentary (e.g., numbered reasoning about 5G and retail market dynamics). Entertainment topics produced more informal language and divergent opinions (e.g., on celebrities’ private behavior or dress). Topics such as technology, sports, and international/domestic affairs tended to evoke patriotic ideology and affirmative stances; entertainment and some domestic topics stimulated more explicit critical thinking and diverse viewpoints.
• Model of relations: Critical thinking operates as an analytical/filtering mechanism shaping personal and social ideologies; these may converge or diverge depending on topic and framing. Critical thinking also influences the selection of linguistic forms deemed appropriate for expressing those ideologies. News framing and topic cues can stimulate a sense of ingroup belonging and modulate both ideological positions and linguistic choices. The study proposes a sociocognitive model linking linguistic features, ideologies, and critical thinking evidenced in news comments.

The findings align with CDA’s view of language as social practice: comments reflect attitudes and enact discursive/social actions (e.g., condemning public smoking, opposing cyber-violence, supporting national sports teams), evidencing social participation and shared norms. Certain statements perform social functions (e.g., discouraging public smoking, advocating independent judgment) and demonstrate attention to national development in science and technology, reinforcing patriotic sentiment. Among younger commenters, generational slang/buzzwords (e.g., “liang liang,” “sha diao,” “zhong cao,” “di biao zui qiang,” “shen xian da jia”) index identity and distinctive ideological stances. Interviews revealed emphasis on social justice, objectivity, and fairness: students respected individual choice (e.g., attendance/refunds around NBA games) while upholding respect for national symbols; they critiqued inappropriate actions (e.g., signing the national flag) and analyzed celebrity-related issues from broader social perspectives. Overall, critical thinking mediates how commenters position themselves ideologically and select linguistic resources, with topic framing shaping the degree of convergence/divergence between personal and social ideologies.

The study provides an integrated account of how linguistic features (lexical, syntactic, figurative), ideological beliefs, and critical thinking manifest in Chinese news comments. Comments function as vehicles for value propositions around power, ideology, inequality, and social justice, and showcase generational identities through language. Topic domains significantly shape linguistic formality, ideological tenor, and the presence of critical, multi-perspective analysis, producing divergences across comments. A sociocognitive model is proposed to explain how critical thinking filters and shapes ideologies and linguistic choices. Acknowledging critiques of CDA’s methodological vagueness, the authors emphasize careful, systematic analysis with a separation of description and interpretation. Future research should further examine how critical thinking interacts with personal and social ideologies and explore effects of other cognitive attributes on language and ideology within the SCA framework.

The authors note a common critique of CDA: methodological simplicity or vagueness and the absence of an accepted canon of data collection, which can affect interpretability and generalizability. While addressing this through systematic analysis and clear separation of description and interpretation, the study remains subject to these broader methodological constraints in CDA.