1. Introduction

Social media has achieved unprecedented prevalence in the digital era, with 5.04 billion users representing 62.3% of the global population [1]. Among global internet users, 70% are male and 65% are female [2]. Online gender discourse exhibits new characteristics. Digitalization both challenges and reinforces traditional gender roles, with gender expression increasingly intertwined with platform characteristics [3], including phenomena such as “unpopular feminism,” which reflects intentional control of feminist visibility [4]. However, new challenges emerged through coordinated state and market efforts that fostered anti-feminist forums and strengthened network effects of anti-feminism [4].

Despite extensive research on online gender issues, significant gaps remain regarding platform governance mechanisms. Studies show algorithmic recommendations are underexplored [5], and platform governance often lacks transparency [6]. Current research lacks integrated frameworks explaining how technology interacts with psychology.

This study explores gender discourse online, its impact on gender relations and the psychological and technological mechanisms shaping it. It explores how social media platforms function as spaces for identity negotiation and symbolic empowerment, particularly for young women in China. These questions are critical for understanding the evolving nature of gender dynamics in digital spaces. Theoretically, this research enriches the social psychological understanding of online gender interactions. Practically, it provides new perspectives for understanding and improving contemporary gender relations in digital contexts.

2. Theoretical framework

Gender identity isn't biological, but rather a social construct shaped by discourse and interaction, as Berger & Luckmann found [7]. Social reality, including gender, is thus constructed through shared knowledge, not pre-determined by nature. This constructivist understanding positions gender as part of our everyday reality that is continuously negotiated through language and social institutions. Hacking's idea of "interactive kinds" highlights the way in which social categories shape the people they classify [8]. His "looping effects" show how people both conform to and reshape the categories that define them, creating a dynamic process that is especially evident in digital environments where gender identities are constantly reshaped through networked interactions. Gendered experience is always situated, so it is necessary to be attentive to intersectionality. Crenshaw's critical framework shows how gender identity must be examined alongside other systems of power (i.e., race, class, and sexuality) [9]. This intersectional approach reveals how individuals experience multiple identities, which shape their social positioning. Within digital contexts, these theoretical insights gain renewed significance. Hayat et al. show how men and women behave differently online [10]. Men tend to start topics and communicate in a more assertive way. Women tend to reply to others and communicate more supportively. Their analysis shows how network structures influence gender power distribution and suggests that digital platforms reproduce traditional gender hierarchies and create new sites for identity negotiation. This evidence underscores how gender construction operates through technological mediation, where algorithmic systems and platform affordances shape the possibilities for gendered self-presentation.

This study uses three concepts from social psychology to understand how gender discourse becomes polarised and adversarial on online platforms. Isenberg's study on group polarization shows how individuals' initial tendencies are reinforced after group discussion [11]. This explains why gender-related debates on social media often become more extreme. Social identity theory explains identity conflicts seen in gender discussions online. According to this framework, A sense of belonging to an ingroup not only fosters stronger emotional ties among its members but can also lead to antagonistic dynamics with outgroups [12]. On social media platforms, users often identify with specific gender-related identity groups (e.g., feminists and traditionalists), leading to increased ingroup cohesion and outgroup hostility, which in turn reinforces divisive discourse patterns. The technical architecture of social media platforms further reinforces these psychological tendencies through algorithmic recommendation systems, creating an echo chamber effect. As Sunstein noted, the Internet facilitates exposure to like-minded opinions, making it simpler for individuals to echo and strengthen their pre-existing beliefs [13]. These personalised algorithms amplify users' gender perspectives by selectively presenting content that aligns with their beliefs, reinforcing gender echo chambers and limiting exposure to diverse viewpoints. Group polarization, social identity, and algorithmic recommendations form a self-reinforcing mechanism that exacerbates the polarization of gender discourse in the digital context. The social psychological processes described show how platforms amplify gender attitudes and provide a theoretical basis for understanding the role of platform algorithms in shaping and constraining gender-related discourse.

3. Literature review

3.1. Research trends

The evolution of digital gender discourse research demonstrates three significant shifts reshaping the scholarly landscape.

3.1.1. Platform shift

The transition from traditional media to social platforms has fundamentally altered how gender politics unfold in digital spaces. As Schneider points out, the feminist perspective clearly illustrates how the digital economy manages to appear transparent and inclusive while consistently evading responsibility [6]. This apparent openness masks embedded infrastructures that shape gender discourse through content moderation and algorithmic curation. Han and Liu argue that the rise of “pink feminism” on Chinese social media reflects an institutionalized alignment between nationalist discourse and gendered narratives, where platform-guided content moderation selectively amplifies voices that merge feminism with patriotic themes [14].

3.1.2. Interactive turn

Recent researchers increasingly move beyond static content analysis toward interactional paradigms that examine discourse-in-use. Wang and Ouyang exemplify the field’s methodological shift by analyzing Weibo’s repost-comment mechanisms [15]. Their study illustrates how interactive affordances on Chinese platforms enable users not only to disseminate feminist content but also to reframe and reinterpret messages dynamically, marking a transition from static textual analysis to situated discourse-in-use research. Li and Wu provide a case of multimodal interaction, showing how gender identity is discursively co-constructed through creator-viewer dynamics, further blurring producer-consumer boundaries in digital gender discourse [16].

3.1.3. Localized approaches

Contemporary research increasingly adopts interdisciplinary and non-Western approaches, bridging feminist theory, media studies, and digital sociology. Raza emphasizes the importance of combining feminist theory, media studies, and digital sociology to examine the ways in which gender narratives are shaped and circulated on social media, while also highlighting the limited scholarly attention given to gender expression in China and the Global South [17]. Building on this shift, recent scholarship in the Chinese context reflects a growing trend toward combining cultural studies and feminist psychology to examine digital gender performance. For instance, researchers explore how users mobilize everyday digital symbols—such as profile images or emotional storytelling—to navigate gendered expectations while retaining cultural legibility [17-18]. Peng illustrates how social media influencer Mimeng’s WeChat posts blend consumerism with gendered emotional appeals, reflecting a neoliberal feminist framework where personal choice and self-branding are equated with empowerment [18]. This commercialized discourse style has since become a template for KOLs shaping gender norms in China.

3.2. Review

Recent research shows that social media platforms' algorithms systematically reinforce gender bias, contributing to polarized discourse. Hall & Ellis indicate in their systematic review of gender bias in AI algorithms that algorithmic systems often amplify preexisting gender biases, thereby polarizing gender narratives online [19]. This is particularly evident in large language models, as Palacios Barea et al. found in their technofeminist analysis of GPT-3 [20]. They noted that the model exacerbates existing biases and reproduces dangerous ideologies like white supremacy and hegemonic masculinity as factual knowledge. Research reveals that gender biases stem from cultural understandings and social constructions embedded in training datasets, which are then amplified through algorithmic outputs. These findings highlight how cultural norms in training data, when processed by algorithms, can amplify existing inequalities and divide gender discussion online into polarized groups.

Social media platforms facilitate the emergence of alternative gender communities by enabling discursive practices that strategically negotiate and reconfigure identity boundaries. Li and Wu go beyond interactional analysis to demonstrate how multimodal discourse practices construct resistant masculinities on Bilibili [16]. Their analysis highlights how identity boundaries are collectively shaped in algorithmically structured comment cultures, where viewers act as co-constructors of gender norms. Building on this interactional foundation, Wang and Ouyang further explore how digital feminist communities on Weibo engage in boundary-making [15]. Through collective reposting practices, users challenge hegemonic narratives, producing oppositional discourses that reshape the boundaries of feminist identity. Their analysis reveals that community norms are not given but negotiated through participatory discourse practices embedded in platform infrastructures.

From a cultural-psychological perspective, Chang, Ren, and Yang highlight how young women’s profile picture editing creates “virtual gender asylum,” fostering empowerment and belonging [21]. These digital self-presentations enable women to maintain social acceptability within Confucian-influenced norms while subtly resisting prescriptive femininity. Together, these studies demonstrate how digital affordances transform gender boundary work into dynamic, interactive processes, simultaneously enabling resistance and rearticulation of collective gendered identities.

Algorithmic infrastructures play a central role in shaping which gender narratives are circulated, suppressed, or amplified across digital platforms. Research reveals that platform algorithms and governance strategies have insufficiently addressed the disproportionate online harassment experienced by women, with women showing significantly higher preferences for content removal and account banning compared to men. This gendered discrepancy reveals a critical mismatch between user experiences and platform governance, emphasizing the need for response systems sensitive to differential harms [5].

In the Chinese context, Han analyzes the feminist platform "Gender Watch Women's Voice," arguing that instead of driving meaningful social change, digital feminist activism has, in some cases, coincided with a rise in online misogynistic behavior [22]. Her study reveals how state-algorithmic coordination mechanisms suppress radical feminist discourse, transforming platforms into contested spaces where feminist expression is simultaneously enabled and constrained. Extending this structural critique, Li and Wu show how Bilibili’s recommendation algorithms act as both “filters and amplifiers” in shaping the visibility of male beauty vloggers' gender performances [16]. They argue that algorithmic systems do not merely mediate identity expression but actively co-produce discursive hierarchies through their filtering logic.

3.3. Research gaps

Despite growing interest in gendered algorithmic cultures on social media, current literature reveals several critical gaps that warrant further investigation. While recent studies have examined how gender identities are discursively negotiated on Chinese platforms, such as Li and Wu's analysis of male beauty vloggers on Bilibili, this line of research primarily focuses on content creators and user interactions [16]. Most existing studies concentrate on user expression or identity construction, yet neglect the technological infrastructures that shape these discursive spaces [15].

A significant gap exists in empirical fieldwork on platform governance mechanisms. Although Im et al. demonstrate that women perceive greater harm from online harassment [5] and Schneider calls for "governable spaces" with democratic structuring of platform rules [6], few empirical studies offer field-based data on platform policy-making or institutional moderation practices. The literature remains largely theoretical, with limited examination of how different platform norms generate unequal gendered effects across user groups. Additionally, geographical bias toward Global North platforms overlooks localized algorithmic governance in Chinese digital platforms like Bilibili, Weibo, or Xiaohongshu. According to Han, much of the existing scholarship fails to adequately address how online misogyny manifests in culturally distinct ways in China [22], highlighting the need for culturally grounded investigations of platform governance and gender dynamics.

4. Case studies and discussion

This section examines the production and dissemination of counter-discourses within China's digital ecosystem through a detailed analysis of two case studies. This analysis does not merely describe phenomena; rather, it dissects the specific mechanisms through which social psychological processes interact with technological infrastructures to shape contemporary gender discourse in China.

4.1. The GWWV experience

This case study draws primarily on Han's ethnographic study of Gender Watch Women's Voice (GWWV) [22], enhanced by Wang and Ouyang's analysis of the "Jiangshanjiao" incident [15]. The case demonstrates the emergence of feminist counter-discourses through strategic resistance to algorithmically amplified misogyny.

4.1.1. Discourse strategies

As demonstrated in the research conducted by Han, the encounter of GWWV with "toxic masculinity" online resulted in the implementation of a complex counter-discourse strategy [22]. The group's transition from seeking "friendly and supportive" commentary to actively managing hostile attacks reveals three distinct linguistic tactics. Firstly, the deployment of archival resistance: rather than immediate counteraction, GWWV developed a systematic approach of "archiving negative comments, thinking about them, and writing review articles" to present counter-arguments [22]. This strategy transforms individual attacks into collective pedagogical moments.

Second, the utilization of follower networks as discourse multipliers. As Han observes, GWWV relied on supporters who "voluntarily formed networks to fight back against online antagonists [22]," creating what Wang and Ouyang term a "counterpublic" through participatory discourse practices [15]. Thirdly, the innovation of satirical exposure is exemplified in the "Jiangshanjiao" case, where the question "do you have period?" employed satirical expression to expose the "oddness" of patriarchal discourse by making "visible what used to be taken for granted" [15].

4.1.2. Platform mechanisms

The technological infrastructure plays a crucial role in both amplifying misogyny and enabling counter-discourse production. Han's analysis demonstrates how Sina Weibo's algorithmic recommendation system creates what she terms "state-sanctioned misogyny [22]." The platform's practice of regulating feminist content while allowing misogynistic attacks to persist indicates how algorithmic governance reproduces patriarchal hierarchies. However, Wang and Ouyang's study demonstrates how digital affordances can be strategically appropriated, for instance through the use of screenshots to preserve banned content and hashtags to maintain discursive continuity after deletion [15].

The "Jiangshanjiao" case illustrates platform mechanics as both constraint and opportunity. Although the original post was subject to censorship within 12 hours, the hashtag #JiangshanjiaoDoYouHavePeriod enabled what Wang and Ouyang term "rejuvenation," allowing the counter-discourse to persist and evolve beyond its initial form [15]. This demonstrates how platform affordances can be tactically deployed to circumvent censorship while maintaining discursive momentum.

4.1.3. Social psychology

The psychological processes underlying counter-discourse formation operate through three interrelated mechanisms. Initially, the process of identity crystallization must be considered. Han's interviews reveal how encounters with online misogyny solidify feminist identification, thereby transforming individual grievances into collective consciousness. One coordinator notes that experiencing anger, hatred, and abuse is a common reality for Chinese feminist social media accounts, highlighting how such hostility paradoxically reinforces group boundaries [22].

Secondly, polarization acceleration: consistent with social identity theory, exposure to hostile attacks enhances ingroup cohesion while intensifying outgroup differentiation. The evolution from GWWV's initial strategy of avoidance to active counter-discourse production exemplifies how group polarization emerges through sustained intergroup conflict.

Thirdly, the development of collective efficacy. Wang and Ouyang's analysis shows how the "Jiangshanjiao" incident generated "the will and agency to spread feminist thinking" through successful resistance to censorship [15]. This demonstrates how effective counter-discourse production can enhance collective confidence in challenging dominant narratives.

4.2. Neoliberal feminism and the construction discourse

This case study explores the ways in which commercial feminist discourse gives rise to new forms of gender relations. The analysis is based on Mimeng's WeChat content, drawing on Peng's [18] critical discourse analysis and contextualizing it within Liao's [4] framework of "unpopular feminism."

4.2.1. Discourse strategies

Peng's analysis shows how Mimeng uses neoliberal feminist rhetoric to construct a "feminized male ideal [18]." Her strategic use of second-person pronouns has been argued to create an "engaging communicative style" that individualizes feminist messages while obscuring structural inequalities. The discourse functions through three primary mechanisms: firstly, the commodification of care, where male attentiveness becomes a consumer product ("he will swipe his credit card for me with no hesitation"); secondly, the domestication of power, where traditional gender hierarchies are repackaged as feminine choice; and thirdly, the sexualization of compliance, where women's domestic management is framed as sexual empowerment.

Liao's concept of "unpopular feminism" provides a valuable context for understanding this discourse strategy [4]. Mimeng's approach can be seen as a type of feminism that "divides yet simultaneously reimagines and rebuilds various forms of communities [4]." This appeals to women's desire for empowerment but also reinforces patriarchal structures through consumer capitalism.

4.2.2. Platform mechanisms

The technological capabilities of WeChat facilitate the dissemination of this commercialized feminist discourse through multiple mechanisms. Peng's research indicates that Mimeng's account amassed a following of over 8 million by leveraging platform features that prioritize engagement over critical analysis [18]. The subscription-based model of WeChat accounts creates what can be termed algorithmic intimacy: the illusion of personal communication that masks commercial manipulation.

The platform's content delivery system has been found to amplify neoliberal feminist messages by targeting women experiencing what Liao identifies as the "intentional control of feminist visibility [4]." WeChat's relative freedom from censorship encourages gender discourse that is depoliticized and consumer-oriented. This illustrates how differing platform architectures engender discrete possibilities for feminist expression, with commercial platforms providing more space for neoliberal variants while restricting radical alternatives.

4.2.3. Social psychology

The psychological appeal of Mimeng's discourse is attributable to the mechanism of aspirational identification. Peng's analysis explores how the "feminized male ideal" aligns with Chinese women's aspirations for both tradition and modernity by proposing a male domestication without demanding systemic transformation. This results in what can be termed pseudo-empowerment, defined as the psychological satisfaction of imagined control without the concomitant redistribution of power.

The discourse's efficacy is attributable to its exploitation of cognitive dissonance surrounding gender equality. By conceptualizing women's management of "interdependent gender relationships" as empowerment, Mimeng's content resolves the tension between feminist aspirations and patriarchal realities through individual solutions. This psychological mechanism helps to explain the popularity of neoliberal feminism among Chinese women seeking agency within constrained circumstances.

However, as Liao observes, this approach ultimately reinforces "patriarchal nationalism, capitalist exploitation, male supremacy" by channeling feminist energy into consumption rather than structural challenge [4]. These narratives provide a psychological comfort that paradoxically undermines collective action by offering individual escape routes from systemic problems.

4.3. Integrated analysis and discussion

4.3.1. Discourse features

The comparative analysis reveals two fundamental characteristics of Chinese digital gender discourse. First, algorithmic mediation shapes both the production and circulation of gender-related content through platform-specific mechanisms. While the political sensitivity of the Weibo environment imposes limitations on radical feminist expression, the commercial orientation of WeChat fosters the growth of depoliticised, consumer-focused gender discourse. Technology actively fosters feminist expression, not just as neutral conduits.

Secondly, psychological amplification occurs through the interaction between social psychological processes and digital affordances. These cases demonstrate how platform features exploit group polarization and social identity dynamics to intensify gender-related discourse. However, the direction of this amplification is contingent on the specific combination of technological affordances, regulatory constraints, and commercial incentives that operate within each platform ecosystem.

4.3.2. Contemporary gender framework

This analysis proposes a dialectical model of gender discourse in China, in which technological possibilities interact with political constraints and commercial imperatives to produce distinctive forms of feminist expression. The GWWV case demonstrates how platform censorship can inadvertently strengthen feminist counter-discourses by creating martyrdom effects and solidarity responses. Conversely, the Mimeng case demonstrates how commercial success can incorporate feminist language while diminishing its transformative capacity.

The implications extend beyond China to illuminate broader patterns in global digital feminism. The Chinese experience reveals how authoritarian contexts create unique pressures that shape the evolution of online gender discourse, producing innovations in resistance tactics while also generating novel forms of co-optation. It is imperative to comprehend these dynamics in order to gain profound insights into the relationship between technology, politics and gender in other contexts where analogous constraints are in operation.

This framework suggests that the impact of digital platforms on gender relations cannot be assessed independently of the broader political economy within which they operate. The Chinese case shows how technology can be used to both resist and be adopted, depending on the specific mix of state power, market forces and social dynamics.

5. Conclusion

This study examined online gender discourse and its impact on contemporary gender relations. Social psychological analysis was applied to understand the mechanisms shaping digital gender interactions in China. By conducting a case study of counter-discourse production on Chinese platforms, this research reveals how psychological processes interact with technological infrastructures to create distinctive patterns of feminist expression.

The analysis demonstrates that social psychological mechanisms, such as group polarization, social identity formation, and collective efficacy building, operate within algorithmically mediated environments to shape gender discourse production. The GWWV case illustrates how platform censorship strengthens feminist counter-discourse through solidarity and innovation. The Mimeng case shows how platforms exploit identification to create narratives that co-opt feminist language while reinforcing patriarchal structures. The findings suggest that psychological processes must be understood within the context of platform architectures. Algorithmic recommendation systems, content moderation policies and commercial imperatives actively co-produce the conditions in which group polarization and identity formation occur, creating "algorithmic-psychological feedback loops".

This research contributes to the advancement of digital gender studies by demonstrating the potential of social psychological theories to illuminate platform governance mechanisms. The dialectical model of digital gender discourse provides a useful analytical framework to understand how identical technological features can produce divergent outcomes across different political and commercial contexts. Methodologically, the study demonstrates the value of combining discourse analysis with social psychological theory, moving beyond purely textual or purely technological analyses.

The findings have significant implications for platform governance reforms, which must consider how algorithmic systems exploit psychological mechanisms. In particular, the study shows how commercial feminism undermines collective action while providing individual psychological satisfaction. However, limitations in the study must be acknowledged. The focus on Chinese platforms limits the study's ability to make cross-cultural generalizations. The relatively short timeframe may not capture long-term discourse evolution trends, and the methodology relies primarily on secondary literature analysis rather than empirical fieldwork data.

Future research should employ international comparative perspectives through cross-platform studies and longitudinal analyses of discourse evolution and conduct empirical examinations of platform governance decision-making processes. As digital platforms become increasingly primary sites for gender identity negotiation and political expression, understanding the psychological and technological interactions is crucial for developing comprehensive theories of contemporary digital gender dynamics.