Introduction

In May 2021, a significant number of feminist groups on Douban, which functioned as safe spaces like those described by Rosemary Clark-Parsons (2018) on Facebook—invite-only, closed communities with strict membership controls—were deleted masse by the Chinese government. These Douban groups, like the feminist spaces on Facebook, provided members with a sense of security, privacy, and the ability to discuss sensitive issues free from external threats.

On the second anniversary of this mass deletion, online discussions about the 2021 censorship incident resurfaced, prompting me to reflect on how Chinese feminists are now navigating these constraints. This event made me reconsider how they could strategically create new safe spaces under such heavy surveillance and censorship. Unlike the closed, tightly controlled spaces described by Clark-Parsons (2018), Chinese feminists seem to be adapting their approach to develop more fluid, flexible safe spaces that are less easily targeted by authorities. This reflection became the foundation for my research, which examines the evolving characteristics of these new digital safe spaces—spaces that maintain safety but are less fixed and more adaptive to the social-political pressures faced by feminists in China.

This shift from static, closed groups to dynamic, adaptable spaces highlights the resilience of feminist activists, showing that even in the face of aggressive censorship, they continue to find ways to foster collective action and support.

Chinese feminists face a dual challenge of resisting both government censorship and pervasive online misogyny. These two forces interact in ways that amplify the difficulties faced by networked feminists. The Chinese government employs various tactics to censor feminist content, viewing it as a potential threat to social harmony and stability. These include removing social media posts, deleting accounts, and blocking hashtags related to feminist topics, which can indirectly embolden misogynistic narratives by silencing opposition and limiting the visibility of feminist voices. By selectively censoring feminist content while leaving misogynistic voices unchecked, the state creates an environment where patriarchal norms are reinforced, and feminist activism is stifled. Past studies have shown that feminists in China encounter profound sexism and harassment online, often leading to the silencing of their voices through both platform censorship and social backlash (Chen, 2023).

In response to the dual challenges of government censorship and online misogyny, Chinese networked feminists have urgently developed strategies to navigate and resist these obstacles by creating online safe spaces through self-censorship. The self-censorship strategies adopted by Chinese online feminists play a crucial role in carving out these spaces on social media platforms, allowing them to avoid both suppression and misogynistic attacks. Rather than interpreting self-censorship as a passive or negative avoidance strategy, this study frames it as an active and creative tool. By carefully regulating their content—employing coded language, metaphors, and evolving symbols—Chinese feminists have constructed semi-public digital spaces where their discourse can flourish with reduced risk of governmental or societal suppression.

This process resonates with the broader concept of safe spaces in feminist theory, which historically focused on separatist environments free from external threats (Fraser, 1992; Clark-Parsons, 2018). Drawing on Rosemary Clark’s research on feminist safe spaces on Facebook, this study extends Western scholarship by exploring how these spaces are created in the context of Chinese feminist activism. Clark discusses how digital safe spaces, initially formed during the women’s liberation movement in the United States, have evolved into online platforms that provide freedom from harassment and foster collective action (Clark-Parsons, 2018). While Fraser (1992) conceptualized safe spaces as counter public for marginalized voices, this study extends the idea to the Chinese context, where feminists must constantly adapt to changing censorship rules. Unlike in Western contexts, these digital spaces are semi-public, designed to shield activists from immediate suppression while still engaging with broader public discourse.

Understanding the dynamics of feminist activism in China’s digital space is crucial for supporting and advancing these movements. This research contributes to the broader narrative of digital activism, highlighting how feminists adapt to and resist governmental and societal constraints. By examining these tactics, we gain insight into the resilience and ingenuity of feminist activists in China, highlighting their ability to sustain and amplify their discourse despite restrictive conditions (Chen and Whyke, 2022; Hou, 2020).

Misogyny on Chinese social media platforms

While digital media users have made movements more accessible, diverse, and visible, networked activism has also made feminists more susceptible to backlash, repression, and cooptation. Social media platforms have given feminists a more prominent national profile, but they have also facilitated online misogyny (Banet-Weiser and Miltner, 2016). Online misogyny, a form of online gender-based harassment, includes harmful silencing practices such as hate speech, revenge porn, ‘doxing,’ and threats of violence (Jane, 2016).

Online misogyny has long been a focus of scholarly attention both internationally and in China. Scholars attribute the causes of online misogyny to strong resistance against women’s increasing online participation and feminist views (Alichie, 2023), the lenient policies of social media platforms towards misogynistic harassment (Jeong, 2015), as well as the male-biased bias in technological design (Salter, 2018), the fragility of toxic masculinity (Banet-Weiser and Miltner, 2016), as a reaction to the crisis of masculinity (Dafaure, 2022). In any case, widespread online misogyny has opened ‘a new era of the gender wars’ (Banet-Weiser and Miltner, 2016, p. 171). Online misogyny hinders women’s equal participation in public and political life and affects their right to freely express their views (Barker, 2019).

In China, Confucianism guaranteed the existence of widespread misogyny. Confucianism, with its deep-seated cultural and philosophical roots in Chinese society, has long influenced gender norms and expectations, and the maintenance of traditional family roles (Li and Zhu, 2017). Confucianism is centered on the idea that men and women have different divisions of labor and that women should embody virtues such as obedience, humility, and devotion to the family. According to Confucianism, “good women” are virtuous, kind, frugal, and filial (Chang and Bairner 2019). These gendered roles are often seen as part of a harmonious social order, reinforcing patriarchal structures that place men in positions of authority and women in subordinate positions (Han, 2018).

This intersection of Confucian values with state policy not only reinforces traditional gender norms but also creates a cultural environment where misogyny can thrive (Li and Zhu, 2017). The revival of Confucian values is evident in the state’s emphasis on traditional gender roles and fertility. This cultural backdrop significantly impacts the persistence of misogynistic attitudes, as it legitimizes the subordination of women within both the family and society (Chang, 2024). State policies that encourage higher fertility rates often frame motherhood as a national duty, aligning with Confucian ideals that see women primarily as bearers of children and caretakers of the home (Han, 2018). This ongoing revival of Confucianism thus plays a critical role in shaping the challenges faced by feminist activists in China, as they navigate a landscape that continues to value patriarchal ideals (Chang, 2024). The resurgence of Confucian ideals in modern China, including the state’s promotion of family values and fertility, exacerbates these challenges, creating a hostile environment for feminist discourse.

Calling oneself a feminist in public broadly challenges the Confucian framework and its core views of sexual ethics, gender norms, and social values (Han, 2018). Scholars have argued that online misogyny is part of a broader ideology in the post-reform period and opening-up of China, which is reflected in the use of misogynistic language by both men and women (Jing-Schmidt and Peng, 2018). Moreover, misogynistic messages that demean and discredit women create a more intensified misogynistic environment with the complicity of the state and digital platforms (Wang and Yuan, 2023). Chinese popular culture is still permeated by an essentialist gender discourse that often blatantly devalues women or only values femininity that conforms to male standards (Wang and Driscoll, 2019). For example, in April 2021, several Chinese feminists had their social media accounts removed after torrents of angry men trolled them. Those trolls often espoused nationalistic views, saying that feminists and extreme feminists were trying to disrupt traditional Chinese society.

The pervasive online misogyny rooted in China’s patriarchal social structure presents a significant challenge for feminist activism on social media platforms (Han, 2018; Huang, 2016; Tan, 2017). In response, Chinese feminists have developed various strategies to navigate these hostile environments. These include forming alliances with mainstream journalists to amplify feminist voices (Wang, 2018), adopting moderate stances to avoid conflict with misogynistic audiences (Chen and Chen, 2024), and employing digital alteration techniques, such as masking or altering images, to evade censorship and harassment (Tan, 2017). Building on these existing strategies, this study integrates Rosemary Clark’s concept of online safe spaces to examine Chinese feminists’ self-censorship tactics across three key areas: navigating platform algorithms, fostering supportive networks, and creatively disguising digital content.

Feminist self-censorship strategies under Chinese internet censorship

In China, feminist activities are subjected to routine government censorship. Some scholars argue that feminists face harsh scrutiny because their advocacy of women’s rights represents a Western ideology that conflicts with China’s political framework (Mao, 2020). Feminist activities could trigger collective action that would challenge the harmonious society advocated by the Chinese Communist Party (Guttenberg, 2021; Fincher, 2021; Mao, 2020; Wallis, 2015). In short, feminist content has become a target of Chinese censorship.

Feminist discussions critical of the government in a directly confrontational way are likely to be censored. In January 2022, after the Xuzhou Fengxian “chained women” incident, feminists used social media to criticise the government’s incompetence in tackling human trafficking and possible corruption within the government. Posts and online articles criticising the government were deleted within 24 h of being posted because of their direct challenge to and criticism of government policies related to the treatment of women in society.

Past studies have discussed how the Chinese government has selectively enclosure feminist issues into its policies, often aligning them with state interests rather than supporting the broader feminist agenda (Dong, 2023; Jiang, 2014; Fincher, 2016). This approach is evident in the government’s adoption of certain Marxist feminist ideas, which are reframed in a manner that supports social stability. For instance, the Chinese Communist Party facilitated women’s entry into the workforce and empowered women’s working rights since 1949 as a means to boost economic productivity.

An example of the Chinese government adopted Marxist feminist ideas is the promotion of women’s participation in the labor force, which is often touted as an achievement of gender equality. In 2021, the State Council published the Outline for the Development of Women (2021–30), which addressed promoting women in STEM (science, technology, engineering, and math) careers and closing the gender pay gap. However, the economic transition has weakened women’s rights, reflecting a utilitarian approach rather than a genuine commitment to feminist ideals (Nowak, 2017).

Additionally, some gender issues have been assimilated into the mainstream discourse and coopted by government policy despite challenging traditional social norms (Dong, 2023). This selective adoption serves to neutralize potential sources of dissent by incorporating a controlled version of feminist discourse into state propaganda, thereby maintaining the appearance of progressiveness without substantial changes to the patriarchal status quo. This adoption function is emphasised in Dong (2023), which seeks to guarantee an increase in fertility by making men partly responsible for child-rearing and guaranteeing women’s rights and interests in assuming motherhood responsibilities. Traditional Confucian gender norms and family labor division persist within the private sphere, limiting the effectiveness of state feminism in addressing gender inequality (Li, 2023; Meng and Huang, 2017). In essence, despite elevating the safeguards for women when they take on domestic labour, it still reinforces essentialist gender roles.

These examples illustrate the complexity of feminist discourse in China, where the feminist movement often clashes with state-driven narratives that seek to use the adoption of some of the slogans of the feminist movement to achieve broader political goals. As a result, Chinese feminists not only face direct censorship, but also the possibility that their message will be diluted or redirected by state influence.

In response to the state’s heavy-handed censorship, Chinese networked feminists frequently employ self-censorship to navigate these constrains while continuing their activism. Self-censorship is a type of free-speech restriction or “opinion expression inhibition” (Hayes et al. 2005:300) that is not carried out by official actors but applied by a person or entity responsible for producing a piece of creative expression to prevent any perceived adverse reaction to that expression. Based on his research on journalists, Bourdieu (1986) argues that self-censorship is a professional habit of journalists that can help them accumulate capital; that is, resources in the form of money, prestige, or influence. In other words, self-censorship is a form of structuring how social life and access to resources are organised (Fedirko 2020).

Past studies of self-censorship have treated it as a negative strategy in speech expression (Hayes 2007; Williams 2002; Walulya and Nassanga 2020; Sleeper et al. 2013), however, unlike traditional views of self-censorship, which treat it as a negative, passive response to external repression (Hayes et al., 2005), Chinese feminists have reappropriated self-censorship as a creative and proactive strategy for navigating both censorship and online misogyny. Rather than retreating, they adapt their content to subvert these constraints and create semi-public safe spaces where feminist discourse can continue to thrive.

This study builds on Foucault’s notion of power and knowledge (Bunn, 2015) and Bourdieu’s understanding of social capital accumulation through professional practices of self-censorship (Reese, 2001), extending these ideas into the realm of feminist activism in China. Bunn (2015) presents a nuanced view of censorship influenced by the theories of Foucault and Bourdieu, proposing that censorship extends beyond the external repressive actions used to control thought. Instead, it involves the reproduction of social relationships and acts as a creative process of knowledge production. Under this new framework, individuals have employed allegory, symbolism, and other creative techniques to express themselves, leading to the emergence of new ideas (Waterson 2011, Favoretto 2010).

By examining self-censorship not as a repressive action but as a creative process, it redefines the role of feminist agency under authoritarian constraints. This aligns with Roestone Collective’s (2014) conceptualization of safe spaces as dynamic, relational entities that must constantly evolve in response to external threats.

The various self-censorship strategies include employing coded language, metaphors, or symbols. Using pinyin, a romanization system for Chinese characters to represent the sound of Chinese words, or mosaics and sharing articles as images have become significant methods for feminist creators to circumvent censorship, as highlighted by many scholars (Meng, 2011; Mina, 2014; Nordin and Richaud, 2014; Zeng, 2019). This strategy is commonly known as “camouflage.” An illustrative example of camouflage usage is from the #MeToo movement in China, where social media users introduced the alternative hashtag #rice bunny after #Metoo became a sensitive term in China. The Chinese pronunciation of “rice bunny” is identical to that of “me too,” and the emojis and match. These emojis, and , have also become symbols of the #MeToo movement in China. Another tactic for evading censorship involves “splitting” Chinese characters. For example, 女权 (feminist) may be written as 女木又 (women; wood; also), since certain words such as “women’s rights,” “homosexuality,” and “government” may be considered sensitive at certain times.

The specific communication strategies employed by feminists resemble a cat-and-mouse dynamic, as described by Fincher: “If one person’s social media account is shut down… they find another person to post something. It’s constantly a cat-and-mouse game with internet censors. Even though there are tight constraints on social media, there’s still room to get your message across.” The ‘cat-and-mouse’ dynamic described by Fincher (2019) illustrates how feminist creators are in a constant state of adaptation. As soon as one form of expression is censored, they find new ways to reintroduce their messages using coded language, satire, or new digital tools, such as image mosaics and evolving metaphors. This dynamic underscores the resilience of Chinese feminists in resisting both censorship and online misogyny as they actively shape their digital activism in response to these threats.

The question is, what self-censorship strategies feminists have used to engage in online safe space building in the face of pervasive censorship? The fact that the Douban groups that met Clark’s definition of a safe space have been massively deleted by the Chinese government in conjunction with the administrators of the Douban platform in 2021 means that the safe spaces practiced on Facebook do not meet the needs of Chinese feminists in their response to censorship. As Roestone Collective (2014) emphasized in 2014, the opposite of safe space is recognizing insecurity. But neither Clark-Parsons’s (2018) emphasis that safe spaces in social media still face insecurity in lieu of information leakage, nor Kanai and McGrane’s (2020) emphasis that safety requires a great deal of sustained labor to maintain, address the ambiguous as well as ubiquitous censorship and misogyny under the influence of state ideology for Chinese networked feminists to bring about unsafety. Therefore, as a creative and productive Chinese feminist self-censorship strategy, it is urgent to explore how to create a safe space that meets the needs of Chinese feminists by proactively modifying content and changing the form of creation.

Methods

This study employs two primary data collection methods: online observation and semi-structured interviews. The data analyzed in this paper form a subset of a larger research project on Chinese networked feminism, which examines feminist digital engagement across multiple social media platforms, including WeChat, Weibo, and XiaoHongShu. This study specifically focuses on XiaoHongShu due to its unique affordances, which allow feminist content creators to strategically navigate censorship and online misogyny. Data collection involved systematic online observation of five prominent feminist accounts and one-hour semi-structured interviews with Chinese feminist content creators actively engaged in digital activism.

XiaoHongShu is the primary research site in this study; an Instagram-like social media platform in China and highly popular among young women in China. According to QuestMobile (2023), over 67% of its 260 million monthly active users are women, primarily urban, white-collar, and aged 18-34, making XiaoHongShu an ideal space for engaging with feminist discourse. Compared to broader platforms like Weibo, XiaoHongShu focuses on personal stories, lifestyle content, and e-commerce, creating a unique environment where feminist creators can reach their audience effectively while advocating for gender equality (Lian Chen and Zhang, 2021). These characteristics make XiaoHongShu particularly suitable for studying feminist content creation and the navigation of digital spaces under censorship.

Table 1 lists the five XiaoHongShu accounts I selected for this study, along with their observation frequency and periods. These accounts were carefully chosen to reflect a diverse range of feminist activities on the platform, including topics like reproductive rights, gender discrimination, and women’s empowerment. The accounts were selected for their influence, high engagement, and coverage of key feminist issues, making them representative examples of feminist discourse on XiaoHongShu. During online observation, I engaged with posts through likes, comments, and following ongoing discussions to understand how feminist creators interacted with their audience. I prioritized content that discussed censorship, self-censorship strategies, and online misogyny, as these were key themes of the study. My observation period, from May 2023 to December 2023, coincided with the second anniversary of the 2021 Douban feminist group deletions, a significant event that reshaped online feminist spaces in China. The timing of this observation was crucial, as it coincided with the second anniversary of the Douban feminist group deletions. This significant event reshaped the digital feminist landscape, prompting activists to develop innovative strategies for maintaining safe spaces online. This period allowed for an examination of how feminists responded to increasing censorship pressures in real-time. During my fieldwork, I engaged with these accounts as an active user would, regularly checking updates and saving content, which was then compiled into a single document for analysis. Although the study focused on these five accounts, broader observations of the platform’s feminist community also contributed to the research insights.

Table 1 Accounts observed on XiaoHongShu.
Full size table

The second part of the data comes from semi-structured interviews with feminist content creators on social media, particularly those publishing on XiaoHongShu. These participants engage in a variety of activities, including creating text, images, and videos, and are responsible for reviewing and adapting content within their feminist teams to navigate evolving censorship and public opinion. This role gives them a heightened awareness of both censorship practices and the misogynistic environment on social media. During the content review process, these editors can clearly identify which content is most likely to be removed or cause their Official Accounts to be banned. To ensure the comfort of the interviewees, I conducted the interviews in whatever setting and medium they preferred, often using Tencent video or audio chat. In total, I interviewed 15 feminist creators, with each interview lasting between 45 min and 1 h. The semi-structured interviews explored their motivations for creating feminist content, their understanding of censorship, the self-censorship strategies they employ, and their perspectives on creating safe spaces for feminism online.

Finally, I analyzed the content posted on five feminist accounts on XiaoHongShu during the fieldwork period, as well as the hashtags, to understand the discourse practices of feminist digital content creators. In this study, a thematic analysis approach was used to identify, analyze, and interpret key themes related to self-censorship strategies and the creation of feminist safe spaces online. The thematic analysis allowed for the systematic categorization of data into distinct themes, such as the creative use of hashtags, group collaboration for content moderation, and adaptive language strategies. The coding process followed an inductive approach, where initial themes were identified directly from the data without predefined categories. As the coding progressed, key themes were refined through an iterative process, combining inductive and deductive strategies to ensure both emergent and theory-driven themes were captured. These themes were developed through an iterative coding process, wherein both interview transcripts and online content were examined for recurring patterns that reflect how Chinese feminists navigate censorship and online misogyny. The use of MaxQDA facilitated the coding and organization of this qualitative data, ensuring comprehensive coverage of emerging themes.

Given the ethical implications of feminist data studiesLeurs, 2017) and the harm and vulnerability of being a feminist, we followed the ethics of care in feminist research ethics (Franzke et al. 2020). Because Xiaohongshu is a social media platform for publishing personal life content, we have chosen not to publish identifiable data, such as their specific personal life-related account content. In addition, all personal information disclosed by respondents during the interviews was anonymized. These precautions were taken to ensure that all materials in our study remain untraceable. The study was approved by the institutional ethics committee.

The following description of Chinese online feminists’ self-censorship revolves around three main themes, which were identified through my textual analysis: (1) platform: how to use the characteristics of the platform to self-censor and create a safe space; (2) form: how to use self-censorship within the team to create a safe space; (3) content: how to self-censor and change the content to create a safe space. These three themes derived from data analysis show that feminists have adopted a rigorous attitude in the face of ambiguous censorship systems and have used creative self-censorship strategies and collective action to self-censor within the team to establish a safe online space.

Research findings

Throughout interviews and participant observations, Chinese networked feminists described the various ways they navigate censorship and misogyny to build and maintain online safe spaces. Much like the digital spaces described by Clark-Parsons (2018), the feminists in this study emphasized the importance of self-censorship as a key mechanism for preserving these spaces. However, they seldom provided explicit details of the digital practices that underpin the creation of these safe spaces, which often only become visible when their content is censored or reported. Feminists understand these spaces to be flexible and ever-evolving in response to external threats, though the exact methods of maintaining these spaces are often left unspoken.

This section identifies three major dimensions of feminist self-censorship central to the construction and ongoing adaptation of these networked safe spaces: (1) how self-censorship strategies and platform affordances are used to set up safe spaces, (2) how collective group creativity helps maintain these spaces through mutual protection and content regulation, and (3) how feminists engage in fluid, creative techniques such as evolving satirical language to confront censorship while ensuring the continuity of feminist discourse. These strategies, while essential for creating online safe spaces, also highlight the tension between maintaining safety and ensuring the visibility of feminist activism in broader public spheres. As these spaces are networked across multiple platforms and communities, the balance between protection and engagement becomes a dynamic and ongoing process of negotiation, making them resilient yet constantly in flux.

Setting up the online safe space—creative use of hashtags

Audience feedback emerged as a reason for self-censorship in the feminist creative process. The main concern was that misogynistic audiences would report feminist content to the social media platform’s censors, resulting in the content being blocked or the account being deleted. This challenge was emphasized by the feminist creators interviewed, one creator said, “[Content] being removed is predictable, and people are constantly reporting my content [to the platform] in addition to the official censorship going on.” Another interviewee reported that his social media account, which he had been using for eight years, was subjected to focused reporting by feminist haters, leading to deletion from the platform in the space of a single day: “In just one afternoon, I could not log in to my account.”

The suppression of feminism by misogynistic audience reporting and censorship systems has forced feminist creators to try to find safe spaces on the internet to maintain the resilience of feminism. Unrelated hashtags have become popular in the Chinese social media outlet XiaoHongShu to, according to one interviewee, “avoid being swiped by male users.” Hashtags are based on user experience and guesswork and are based on big data “push algorithms.” These hashtags are also creatively used by feminists in online content creation to target their audience better and to avoid being restricted by social media platform administrators. On Weibo, hashtags organize trending topics and facilitate public discussions, making content widely discoverable. In contrast, Xiaohongshu’s hashtags focus on categorizing lifestyle content, like product reviews and travel, helping tailor personalized recommendations rather than fostering public debates (36Kr, 2023).

The use of hashtags in this networked environment is not merely a tactic for content distribution but also a form of self-censorship. Feminists adapt their language and tagging strategies to maintain their presence on platforms like Xiaohongshu while minimizing the risk of censorship. Self-censorship in this context does not involve silencing one’s voice but rather engaging in a creative and tactical modification of content to navigate platform restrictions. This aligns with Foucault’s notion of power dynamics, where resistance emerges in response to repression (Bunn, 2015). By selecting hashtags that are unlikely to attract the attention of male users or censors, feminists preserve a fluid, semi-public safe space that is difficult to penetrate by those seeking to suppress feminist discourse.

Existing studies have highlighted the role of hashtags in sparking social movements. For instance, Clark-Parsons (2018) explores how hashtags like #YesAllWomen and #NotAllMen have been used in Western contexts to foster solidarity and mobilize action around feminist causes. Similarly, Freelon, McIlwain, and Clark-Parsons (2018) discuss the use of hashtags in the Black Lives Matter movement, illustrating how digital activism can transcend platform limitations through networked online communities. In the Chinese context, Meng (2011) and Zeng (2019) examine how feminist activists utilize hashtags to evade censorship and mobilize digital activism in China. These studies demonstrate how hashtags function both as tools of visibility and as shields against censorship—concepts that are highly relevant to the strategies discussed in this paper.

In Chinese popular culture, #BabySolidFood refers to content related to parenting and infant care, which makes it appear benign and unlikely to attract attention from misogynistic users or censors. However, feminist creators use this hashtag to filter out male users who might report their content, while ensuring it reaches an audience that understands the satirical and coded feminist message.

The brilliance of this strategy lies in its use of everyday language that resonates with the platform’s algorithmic recommendation system. The term “baby” is also closely associated with the Chinese government’s encouragement of childbirth under its pro-natal policies. By embedding their messages within such seemingly innocuous terms, feminists avoid immediate detection while simultaneously critiquing the patriarchal structures that underpin Chinese society. For example, the hashtag #MaleSterilization sarcastically critiques notions of male superiority under Confucian ideals, highlighting the denial of men’s reproductive rights is tantamount to questioning masculinity.

This strategy gains further relevance in the context of China’s current pro-natalist policies, which emphasize women’s roles as mothers and caregivers. The state encourages childbirth, reinforcing traditional Confucian gender roles that assign domestic responsibilities to women. Feminists, aware of these socio-political dynamics, cleverly manipulate these Confucian norms by using hashtags that emphasize childcare or family duties. The assumption that women are focused on their families and men on their careers is subverted to create an online filter bubble. This bubble effectively excludes misogynistic men from engaging with feminist content, as these hashtags appear irrelevant or uninteresting to male users. In this way, feminists strategically exploit both state-driven rhetoric and cultural gender norms to shield their spaces from unwanted attention while preserving feminist discourse within these seemingly innocuous topics.

In conclusion, the innovative use of hashtags by Chinese feminists represents a strategic adaptation to both censorship and male-dominated backlash on social media platforms. By creatively employing hashtags like #BabySolidFood and #MaleSterilisation, feminists tailor their messages to their intended audience while constructing virtual safe spaces that exclude potential disruptors. These spaces, while not immune to infiltration, provide a fluid and adaptable environment that aligns with Clark-Parsons’s (2018) theory of safe spaces in the digital age. Just as networked feminist communities move between platforms and adapt their strategies, their digital spaces are also fluid and dynamic, designed to preserve feminist discourse while evading censorship and misogyny.

Maintaining the online space—group creativity

Unity within a group is instrumental in addressing these challenges by fostering greater self-censorship among members. This collective self-censorship, far from being a passive form of repression, is an active and creative process within Chinese networked feminist groups. These groups work collaboratively to maintain their presence on platforms like XiaoHongShu by ensuring content adheres to censorship guidelines while simultaneously safeguarding members’ identities from potential online backlash.

One of the key strategies for sustaining feminist discourse under censorship is the group structure of many feminist social media accounts. These accounts are not run by individuals but by teams that advocate feminist views and issues. These groups consist of an editor-in-chief, editors, and writers. Writers submit their content to editors, who then perform internal self-censorship by carefully managing sensitive topics and keywords to avoid triggering censorship. This process is crucial to maintaining a feminist presence on Chinese social media platforms, as one interviewee explained: “When editing an article criticizing the Chinese Communist Youth League’s use of the word ‘cancer’ to describe feminists, we asked ourselves, ‘Is it possible for us to change the words Communist Party of China into a pinyin abbreviation and then mosaic the image?’ “Such internal self-censorship illustrates how these groups engage in a form of content moderation that preserves their critical stance while mitigating the risks of censorship. By working together, these feminist networks can continue their activism while adhering to the platform’s restrictive guidelines. This collective approach aligns with Foucault’s concept of power, where power is not merely repressive but productive, enabling these groups to create safe spaces for feminist discourse through collaboration (Foucault and Deleuze, 1977).

The primary purpose of this internal self-censorship within the group is twofold: to protect creators and to sustain their motivation to produce content. Editors and editors-in-chief play a crucial role in screening audience feedback, which shields writers from negative comments and cyber-violence. This protection allows writers to focus on their creative work without the emotional toll of dealing with harassment. One interviewee stated: “I feel like personal information is more strongly protected in this form of group creation. “In the case of group authorship, individual identities remain invisible, and it is this collective invisibility that provides security. The group shares a common identity as “Author of the team” or “Editor of the team,” which reduces the likelihood of personal attacks on individual members. By sharing the responsibility of content creation, these feminist groups can shield their members from the targeted violence that often accompanies feminist activism online. This group dynamic creates a sense of solidarity and shared responsibility, reinforcing the idea of a safe space where feminist voices can be expressed without fear of personal retaliation.

However, those who work independently and lack the protective layer of group anonymity face greater risks. One independent feminist writer shared her experience of facing online misogyny and censorship: “Naysayers would mail letters to my father’s workplace accusing me of posting ‘inappropriate comments’ on social media. They would even find my selfies on the social media platform to abuse me for being a fat woman. “This stark contrast between group and individual experiences highlights the importance of collective action in building safe spaces for feminist discourse. The shared identity of the group offers a layer of protection that individual authorship does not, demonstrating the value of networked feminism in an oppressive environment.

Drawing on Clark-Parsons’s (2018) and Fraser’s (1992) conceptualizations of feminist safe spaces, the collective self-censorship practiced by Chinese feminists demonstrates how safe spaces can be dynamically maintained through networked collaboration. Unlike static safe spaces that rely on strict membership controls, such as those Clark observed on Facebook, the safe spaces created by Chinese feminists on platforms like XiaoHongShu are fluid and adaptive. These spaces rely on continuous cooperation, where feminist creators collaborate to manage content, filter out potential disruptors, and engage with audiences.

This group-driven approach highlights the resilience of Chinese feminists, who are not simply reacting to censorship but actively working to subvert it by using the affordances of the platform to their advantage. Through internal self-censorship and group coordination, they can navigate the constraints imposed by the state and misogynistic users, while still maintaining a critical and activist stance. This networked form of activism illustrates the adaptive nature of safe spaces in a hostile environment, where feminist creators must constantly modify their strategies to ensure their voices are heard.

The resilience of Chinese feminist networks lies in their ability to engage in a collective process of self-censorship that allows them to sustain their activism in the face of state censorship and online misogyny. By working together as a network, these feminists can pool their resources, share strategies, and protect one another from both censorship and personal attacks. The network itself becomes a safe space, providing a platform for collective action and mutual support, in line with Fraser’s (1992) theory of subaltern counterpublics. These feminists form their own counterpublic by creating a discursive space that challenges the dominant patriarchal discourse while shielding individual members from the consequences of speaking out.

Fluid the online safe space–creative techniques

Rather than passively avoiding censorship in their activism, Chinese networked feminists actively confront power dynamics in the virtual realm. During their process of creative self-censorship, they employ creative techniques like satire to actively resist censorship, rather than simply avoiding forbidden topics. Moreover, feminist creators use satirical discourse iteratively rather than in a static way. When they find that a satirical term has been listed as a focus of censors, they will promptly exercise self-censorship of the content and use other satirical terms to circumvent censorship.

Satire is a common form of political expression (Qaiwer and Shatha 2020). It often involves speakers using words in stark contrast to their actual meaning, relying on deliberate distortion rather than face-value interpretation (Fubara 2020). The shift of Chinese networked feminists towards satire and active self-censorship may enable them to evade algorithm-based censorship more effectively. Chinese censorship often centres on sensitive words, which may only target specific words and phrases without addressing the underlying emotions or meanings. Chinese feminists once used the term “蝈蝻”(“men as bugs“) to describe certain Chinese men. This term, utilising the word for “insect,” carries a derogatory connotation in Chinese culture and was used to satirise misogynistic or arrogant behaviour. However, when “蝈蝻/men as bugs” became flagged as a sensitive word by censorship algorithms, feminists adeptly shifted to using “男宝” (“male baby”).

The brilliance of the term “male baby” lies in its dual meaning and fluid use in feminist discourse. On one hand, it cleverly mocks the perceived immaturity of these men, suggesting that they possess fragile self-esteem, require constant attention, and lack a sense of responsibility—traits that do not align with traditional adult masculinity. On the other hand, “male baby” carries an ironic undertone that critiques the deeply embedded concept of male superiority within Confucian values, which underpin China’s patriarchal structure. Moreover, the term is difficult to classify as a sensitive word for censorship because it resonates with the Chinese government’s pro-natalist policies that encourage pregnant women and new mothers to share their lives, thus appearing to be part of acceptable discourse related to childbirth and family values.

This dynamic use of terms like “male baby” highlights the adaptability and wit of Chinese feminists in evading censorship. The linguistic shift from “men as bugs” to “male baby” is not simply a tactical retreat; it is a strategic evolution in feminist language that ensures the critique of societal norms continues, albeit in a different form. By utilizing terms that can initially bypass censorship filters, feminist creators sustain their critique of misogyny while remaining one step ahead of internet censors.

The use of satire in this manner exemplifies a form of creative self-censorship that transforms censorship from a repressive force into an opportunity for subversive dialogue. Feminist creators continuously adapt their discourse by altering sensitive terminology, thereby ensuring that their messages remain vibrant and visible in the face of algorithmic restrictions. The term “male baby” exemplifies this strategic evolution of language, serving as both a critique of male immaturity and a subtle subversion of state discourse on family and reproduction.

Moreover, these feminists extend the lifespan of their messages through additional creative techniques. For instance, by posting sensitive content in images or reversed text, they make it harder for algorithms to detect and censor their messages. This combination of evolving language and multimedia strategies allows Chinese feminists to create a flowing, adaptive online space that is harder for trolls and censors to target or disrupt. These dynamic techniques help maintain the fluidity of feminist discourse, making it challenging for opponents to pin down and shut down these conversations.

Chinese networked feminists are not merely avoiding censorship—they are actively shaping and manipulating their online spaces in a constant state of flow. Their iterative use of satire, such as shifting from “men as bugs” to “male baby,” and their adoption of creative multimedia strategies, exemplify a vibrant form of resistance. By continuously evolving their language and techniques, they transform censorship into an opportunity for creative expression, ensuring that their critiques of misogyny and patriarchy remain dynamic, visible, and resilient in the face of both online trolls and government regulation.

At the same time, this fluid, safe space allows women to intuitively grasp the demeaning undertones of terms like “male baby” towards men, reinforcing the subversive nature of the satire. The inherent understanding of these coded words within feminist circles allows for continued critique of patriarchal norms while evading the attention of censors. This self-censorship strategy effectively creates an adaptable, fluid online safe space—one that evolves in real-time to avoid censorship. By continuously shifting their language and using subtle satire, Chinese feminists maintain spaces where they can express their discontent with gender inequalities without drawing the ire of censorship algorithms or online misogynists.

Moreover, this fluid safe space aligns with Clark’s theory of feminist safe spaces by fostering an environment where women intuitively understand the demeaning undertones of satirical terms like “male baby” without the need for explicit explanation. This shared understanding among feminist creators and followers creates a semi-public yet shielded space for discourse. As Clark-Parsons (2018) suggested, safe spaces do not have to be rigid or closed-off; they can be flexible and dynamic, responding to external threats like censorship. In this case, the fluidity of language used by Chinese feminists is key to maintaining a safe space. The strategic and evolving use of satire and coded language allows these spaces to remain hard to detect by censors and trolls, thereby preserving the safety and integrity of feminist discourse.

By constantly adapting their vocabulary—shifting from terms like “men as bugs” to “male baby”—feminists engage in a form of creative self-censorship that is both empowering and subversive. This fluid safe space reflects the kind of dynamic, relational entity that Clark and Roestone Collective described, where the boundaries are not fixed but constantly shifting in response to external pressures (Clark-Parsons, 2018; Roestone Collective, 2014; Kanai and McGrane, 2020). Through these evolving strategies, feminists are not only evading censorship but also fostering a space that continues to critique patriarchy in new and innovative ways, while intuitively understood by those within the community. Thus, this fluidity of self-censorship serves both as a means of resistance and a way to ensure that the feminist safe space remains active and resilient against external threats.

Discussion

In addition to its contributions to understanding feminist activism under censorship, this study deepens the theoretical exploration of safe spaces in the context of Chinese feminists navigating online platforms. The concept of the safe space has been central to feminist theory and practice, originally emerging from physical spaces where marginalized groups could gather free from harassment or judgment (Fraser, 1992). Rosemary Clark-Parsons’s (2018) analysis of feminist digital safe spaces on Facebook emphasized the need for protective, enclosed environments that shield women from external threats, creating an atmosphere conducive to open discourse. This study extends that understanding by demonstrating that, in the case of Chinese feminists, safe spaces are not fixed or enclosed but fluid, dynamic, and networked, requiring constant adaptation to both state censorship and misogynistic backlash.

Drawing on Fraser’s (1992) theory of subaltern counterpublics, which posits that marginalized groups create parallel discursive arenas for the development of oppositional ideas, this study situates Chinese feminists’ self-censorship practices within the broader project of creating alternative public spheres. The networked feminist spaces that emerge on platforms like XiaoHongShu and Weibo serve as digital counterpublics, offering a temporary refuge from mainstream, patriarchal discourses. However, unlike the separatist safe spaces Fraser described, these digital counterpublics are semi-public, navigating a complex balance between visibility and security. Feminists within these spaces must engage with broader public discourse while simultaneously protecting themselves from censorship and misogyny. This dual function of engagement and protection is central to the theory of safe spaces, as it highlights the tension between creating inclusivity for marginalized voices and avoiding external hostility.

The fluidity of these safe spaces, as discussed in the findings, complicates traditional notions of bounded or enclosed feminist spaces (Clark-Parsons, 2018). In the Chinese context, the external threats posed by the state and online misogynists force feminists to rethink the very structure of their safe spaces, making them adaptable and transient rather than rigid and fixed. This is a significant departure from the models of feminist safe spaces in the West, which tend to emphasize fixed membership and clearly defined boundaries to protect against external threats. The networked nature of Chinese feminists’ safe spaces requires constant movement and shifting tactics, as they migrate across platforms, creatively evade censorship, and maintain flexible boundaries to ensure the continuity of feminist discourse.

One key theoretical contribution of this study is the reframing of self-censorship as an active tool for constructing safe spaces. Whereas traditional notions of safe spaces suggest that they are maintained through strict control and exclusion, this study reveals that self-censorship—rather than limiting feminist expression—enables the creation of fluid safe spaces that remain resilient in the face of censorship. Chinese feminists engage in a collaborative process of content moderation, which resembles the boundary work that Clark describes as central to the formation of safe spaces (Clark-Parsons, 2018). This group creativity ensures that content remains within the platform’s guidelines while preserving the critical edge of feminist activism, allowing feminists to continue to operate within highly censored digital spaces without sacrificing their collective voice.

The iterative and adaptive use of satire is another important strategy through which feminists maintain their safe spaces. By shifting from terms like “men as bugs” to “male baby,” feminists ensure that their critique of misogyny persists while avoiding detection by censors. This linguistic fluidity reflects the dynamic nature of safe spaces, where boundaries are constantly being redrawn to accommodate external pressures. Feminist discourse within these spaces becomes adaptive and responsive, transforming censorship from a repressive force into a tool for creative expression. As Clark-Parsons (2018) suggests, safe spaces do not need to be static or clearly delineated to offer protection; instead, they can be dynamic, relational, and constantly in flux. Chinese feminists’ evolving use of coded language and multimedia strategies exemplifies this fluid safe space, where resilience is built into the very structure of their communication practices.

Moreover, the concept of networked safe spaces reinforces the idea that safe spaces are not purely defensive but also productive, fostering collective action and social change. Feminists in China, by operating in networked teams, collectively engage in self-censorship as a way to protect themselves while continuing their activism. This practice parallels Fraser’s notion of counter-publics as spaces for oppositional discourse but adds the layer of networked collaboration, where feminists collectively negotiate the tensions between visibility and safety. By working together in teams, as seen in feminist accounts like Orange Umbrella and Discover Feminism, Chinese feminists create a shared identity that shields individual members from backlash while reinforcing the group’s collective voice.

This study makes a significant contribution to the theory of resilience in feminist movements. The fluidity and adaptability of safe spaces in China underscore the resilience of feminist activists, who continuously adjust their tactics to maintain their presence online. These networked feminists are not merely reacting to external threats; they are actively shaping the space around them, ensuring that their activism continues despite the multiple layers of censorship and misogyny they face. This dynamic process of adapting self-censorship strategies, evolving language, and leveraging platform affordances allows feminists to create spaces of resistance that remain resilient in the face of authoritarian repression.

However, this study has limitations. The exclusive focus on XiaoHongShu restricts the findings to one platform, limiting broader applicability to Chinese feminist activism across different digital spaces. While XiaoHongShu offers distinct affordances for feminist self-censorship, other platforms, such as WeChat and Weibo, have different regulatory mechanisms and audience dynamics. Future research should adopt a cross-platform approach to better understand how feminist content creators adapt their strategies in varying online environments.

In conclusion, despite these limitations, the self-censorship strategies of Chinese networked feminists reveal new insights into the theory of safe spaces. Far from being enclosed or static, these safe spaces are fluid, dynamic, and networked, responding to external threats through collaboration, creative language use, and the strategic manipulation of platform affordances. By reframing self-censorship as an active tool for sustaining feminist discourse, this study demonstrates the resilience of feminist activism in the face of authoritarianism, offering theoretical and practical contributions to our understanding of feminist safe spaces and digital activism under conditions of repression.