Rebel Lexicons: How Bilibili’s Fragmented Communities Engineer Slang Diffusion

Rebel Lexicons: How Bilibili’s Fragmented Communities Engineer Slang Diffusion | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Rebel Lexicons: How Bilibili’s Fragmented Communities Engineer Slang Diffusion Ke Jiang, Wei Li This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8026908/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In the context of China’s Internet, the generation and diffusion of online slang are undergoing profound changes as grassroots platforms are asserting an increasingly dominant role in linguistic innovation. This study uses Bilibili as an example to explore the underlying mechanisms through which these emerging centres drive the diffusion of linguistic innovation. Big data analysis is conducted on 66 representative initialisms (e.g., yyds and xswl) circulating in 2021 (source: Tencent). With diversity indexes (Simpson’s indexes) and regression models, this study reveals that (1) Bilibili internal adoption intensity can significantly predict national diffusion scale ( β = .743). (2) gender innovation diversity ( β = .274) and cross-community penetration breadth ( β = .371) are positive drivers of slang diffusion, while interest fragmentation ( β = − .450) hinders diffusion. In summary, Bilibili’s community structure fosters two complementary dynamics: diversity that drives innovation and bridging mechanisms that break down silos, enabling effective bottom-up slang diffusion. The findings challenge the traditional linguistic authority long dominated by official media and provide empirical insight into the shifting distribution of cultural power in the digital age. Social science/Development studies Scientific community and society/Geography Social science/Geography Business and commerce/Information systems and information technology Social science/Science technology and society Social science/Sociology Linguistic innovation online slang diffusion community structure Bilibili Figures Figure 1 Figure 2 Introduction The emergence of language and writing marked a new dawn in human civilisation. In today’s digital world, much of China’s youth culture and broader social dynamics remain opaque to outside observers. The continual proliferation of linguistic innovations online thus offers a crucial lens through which to discern these otherwise hidden transformations. Traditional models of linguistic innovation and diffusion typically describe a process in which a linguistic feature spreads gradually from a relatively narrow geographical area or social group to broader regions and larger populations. Bailey’s (1973) wave model conceptualises this diffusion as a series of ‘concentric’ ripples that advance outward step by step from a central point. Building on this metaphor, Britain (2010a: 148) invokes the ‘ripple effect’, noting that ‘innovations, like the ripple effect caused when a pebble is dropped into a puddle, radiate over time out from a central focal area, reaching physically nearby locations before those at ever greater distances’. Another water-based model is the “cascade diffusion” hypothesis, which Chambers and Trudgill (1980: 192) describe as a process completed ‘by the “jumping” of the innovation from one large town to another, and from these to smaller towns, and so on”. This top–down pattern of dissemination in terms of urban hierarchies is also known as the ‘hierarchical effect’ (Britain 2002: 623). Trudgill’s (1974) study of sound change in Norwegian dialects provides a clear example: the innovation spread progressively from major urban centres to smaller towns, and ultimately to rural areas. Such ‘centre–periphery’ or top–down diffusion models have long been regarded as the dominant pattern of linguistic spread. However, within China’s contemporary online environment, a new bottom–up trajectory of linguistic diffusion has begun to emerge. The widely circulated initialism “yyds” (derived from the pinyin initials of yǒngyuǎn de shén, literally “eternal god”), which rose to prominence in 2021, provides a representative example. The term first gained traction within the youth-oriented cultural community of Bilibili, before rapidly permeating the wider Sinosphere and even appearing repeatedly in official mainstream media, ultimately generating more than 200 million circulation instances. As emblematic linguistic forms of youth subculture, such online initialisms are widely used among Generation Z. They frequently originate in marginal or niche communities at society’s periphery, embodying perspectives and expressive styles that diverge from those of the mainstream. Non-mainstream expressions like “yyds” not only draw the attention and commentary of mainstream media, but also signal the forceful rise of subcultural voices within the digital sphere. Their diffusion is not driven by institutional authority; instead, it is produced within marginal communities and amplified by participatory cultures and platform architectures, enabling bottom up cross-community dissemination. This phenomenon raises two central questions: why have grassroots platforms displaced traditional authorities to become key engines of linguistic innovation? And what underlying social-structural mechanisms—specifically, what structural features of platform-based communities—enable localised innovations to escalate into large-scale, society-wide diffusion? Theoretical framework and research questions This study examines how online linguistic forms diffuse from a grassroots platform to central platforms. On the surface, this concerns a linguistic innovation and diffusion process; at a deeper level, however, it concerns the shifting and restructuring of linguistic authority within the contemporary Chinese Internet. We argue that the emergence and circulation of online slang on Bilibili is not merely a linguistic phenomenon, but the outcome produced by the interaction between the platform’s distinctive participatory cultural ecology and its community structure tension. This process challenges the traditional language authority held predominantly by official media, giving rise to what we describe as ‘rebel lexicons’. To account for this reversal, we develop an integrated theoretical framework and a set of key concepts that link four core theories to our empirical variables. Linguistic innovation on participatory culture platforms: a theoretical lens for observing Bilibili adoption intensity. The theory of participatory culture, proposed by Jenkins (2006), emphasises how media users actively engage in producing meaning by appropriating, remaking, and disseminating existing media content, thereby forming informal communities marked by strong social connectivity and shared norms. This theory disrupts the traditional paradigm of passive consumption by conceptualising media use as a social and creative practice. Participatory culture has also been widely applied in studies of Chinese online communities (Zhang & Mao, 2019). Bilibili is a paradigmatic example of participatory culture in China. Originating as a bullet-comment video community oriented towards animation, comics, and games (ACG) subculture, it has, over more than a decade, expanded into a comprehensive cultural platform dominated by Generation Z and characterised by diverse content genres. According to data from QuestMobile, as of 2021, 82% of Bilibili users belonged to Generation Z, with secondary school and university students forming the core user base. Growing up in a highly digital environment, this cohort is adept at constructing cultural identities through symbolic creativity, secondary dissemination, and community interaction. Bilibili’s ecosystem, including bullet-comment interaction, secondary video production (such as Kichiku and MAD/AMV), and layered interest communities, greatly facilitates imitation, parody, and collaborative creativity among users. The initialisms found on the platform (e.g. ‘yyds’, ‘xswl’) are emblematic linguistic products of this environment: they serve not only as internal ‘codes’ of communication, but also as symbolic carriers through which users express belonging and cultural creativity. Unlike traditional official media that rely on one-way dissemination, Bilibili has developed a multi-layered and highly interactive participation mechanism. Its interactive affordances span not only conventional functions such as commenting and reposting, but also bullet comments, likes, ‘coins’, and other features, together comprising six forms of interaction that significantly increase user willingness to participate as well as the engagement intensity. In this study, we employ Bilibili adoption intensity as a core variable and calculate it using the mean value of these six interaction indicators to quantify the level of participation triggered by each slang term. This variable reflects not only a term’s popularity, but also its degree of embeddedness within the community and its potential for wider social diffusion. Accordingly, we conceptualise Bilibili as a ‘participatory incubator’ in which technological architectures (such as bullet comments and content-partition systems) and cultural practices (such as meme-making and secondary creation) jointly constitute the foundational environment for linguistic innovation. Based on this theoretical mechanism, we propose the following hypothesis: Hypothesis 1: The Bilibili adoption intensity of an online slang term is positively correlated with its national diffusion scale. Theoretical mechanism of information Balkanisation and the information fragmentation index. The concept of ‘Balkanisation’ originates in geopolitical studies, where it refers to the division of a region into multiple small, mutually antagonistic entities (Van Alstyne & Brynjolfsson, 1996, 2005). In the context of information studies, information Balkanisation (Chang, 2008) is used to describe Internet users’ tendency to become fragmented into numerous isolated and highly homogeneous interest groups, such as ‘echo chambers’ or ‘filter bubbles’. Within these clusters, particular viewpoints and preferences are continually reinforced, while cross-group information flows and communication are significantly impeded (Sunstein 2006; Pariser 2011). Bilibili’s content architecture is a representative case of information Balkanisation. Originating in ACG subculture, the platform gradually expanded into a comprehensive multi-interest community. Its professional user-generated content (PUGC) system comprises 16 formal content categories, including life, gaming, animation, knowledge, fashion, Kichiku, and others, each of which is further subdivided into multiple subcategories. These content partitions operate in relative isolation and collectively encompass nearly all interest dimensions of Generation Z users. While this structure enables highly accurate content matching, it also produces unintended information silos that limit communication and visibility across groups (see Annex). In this study, we employ an information fragmentation index to operationalise the Balkanisation effect by measuring the degree of dispersion in a slang term’s distribution across Bilibili’s content categories. A higher value on this index indicates that discussions of the slang item are scattered across multiple subcommunities without forming a concentrated volume of discourse, thereby reflecting structural barriers to diffusion within the platform. Based on this theoretical framework, we propose the following hypothesis: Hypothesis 2: The information fragmentation index of an online slang term is negatively correlated with its Bilibili adoption intensity. Structural holes and the bridging mechanism of cross-community penetration breadth. The structural holes theory, proposed by Burt (1992, 1995), posits that when disconnected clusters exist within a social network, the gaps separating them constitute ‘structural holes’. Actors or nodes who can bridge these holes, often described as bridges, occupy advantageous positions, gaining disproportionate access to information, control over resources, and influence in diffusion processes. In this study, the variable cross-community penetration breadth directly operationalises this bridging mechanism. By measuring the distributional range of a specific slang term across Bilibili’s interest-based content categories (such as animation, gaming, fashion, and knowledge), this indicator quantifies the extent to which a slang item transcends its original community and reaches diverse user groups. A slang term that penetrates multiple communities indicates successful adoption and circulation by bridge users (for example, content creators who participate in both animation and music communities) or bridge content, thereby filling structural holes between otherwise fragmented communities. These bridging nodes constitute the crucial channels for cross-boundary information flow, effectively overcoming the diffusion barriers imposed by information Balkanisation. Cross-community penetration breadth therefore reflects not only a slang term’s inherent diffusion capacity, but also the mobilisation potential of connective structures within Bilibili’s community network. Based on the structural holes theory, we propose the following hypothesis: Hypothesis 3: The cross-community penetration breadth of an online slang term is positively correlated with its Bilibili adoption intensity. Innovation diffusion and the initiating role of gender innovation diversity. The diffusion of innovations theory, introduced by Rogers (1962), explains how new ideas, practices, or artefacts spread among members of a social system through specific channels temporally. The theory emphasises four key dimensions: inherent attributes of the innovation, communication channels, temporal processes, and social structure, that jointly shape the trajectory and pace of diffusion. In this study, initialisms are conceptualised as a form of cultural innovation. Rogers argues that the success of diffusion depends heavily on whether different categories of adopters within a social system (e.g. innovators, early adopters, early majority) embrace the innovation, and that diversity among adopter groups is a central structural factor influencing diffusion outcomes. The variable ‘gender innovation diversity’ operationalises the concept of diversity along gender lines. By measuring the extent to which users identifying as male, female or ‘undisclosed’ on Bilibili co-create and adopt a given slang term, this indicator captures the heterogeneity of adopter composition at the origin of diffusion. Higher gender diversity implies a richer collision and fusion of cognitive perspectives, social experiences, and communicative styles, significantly broadening linguistic innovation’s semantic resource pool and expressive pathways. Such diversity provides a wider ‘gene pool’ for the emergence and evolution of cultural memes. Accordingly, gender innovation diversity acts as an innovation initiator within the diffusion process: it ensures that Bilibili, as a ‘participatory incubator’, sustains a diverse and vibrant pool of innovative raw material, thereby increasing the likelihood that highly resonant and adaptable cultural innovations will emerge. Based on the diffusion of innovations theory, we propose the following hypothesis: Hypothesis 4: The gender innovation diversity of an online slang term is positively correlated with its Bilibili adoption intensity. Research methods and implementation This study adopts a multi-step, data-driven methodological design to investigate the origins and diffusion mechanisms of Chinese Internet slang, specifically initialisms. The research process consists of three main stages: (1) Lexicon construction and measurement of national diffusion scale; (2) Quantification of Bilibili adoption intensity; (3) Computation of community-structure variables (gender diversity and cross-community penetration breadth). The procedures are as follows. Lexicon compilation and national diffusion scale measurement. Using the Internet language database published by Peking University as an authoritative source, we selected 66 representative initialisms (such as ‘YYDS’ and ‘XSWL’) as the objects of analysis, thereby constructing the core Lexicon for this study. To measure the nationwide popularity of these initialisms, we collaborated with the Tencent Big Data Platform. Tencent’s dataset aggregates mainstream content platforms in China, including WeChat, Weibo, QQ, Zhihu, and Douyin, and provides statistical access to publicly available content with more than 100,000 views. For each initialism in the Lexicon, we retrieved its frequency of appearance throughout 2021 across all cooperating platforms, including its occurrences in article texts, posts, video titles, and tags. We then summed the total number of appearances for each item across platforms to compute the key dependent variable, the national diffusion scale, operationalised as the total number of mentions of a given initialism on the Chinese Internet in 2021. Bilibili adoption intensity measurement. To verify the hypothesised role of Bilibili as a primary source of online slang innovation, we conducted systematic data extraction for all 66 initialisms on the Bilibili platform. Using Python-based web-scraping tools, we first searched for all videos published in 2021 that contained a target term as a keyword. For instance, in the case of ‘YYDS’, we identified 368 core videos featuring the term. To move beyond simple occurrence counts and more precisely capture each term’s level of engagement, we collected six key interaction indicators for every relevant video: view count, bullet comment count, comment count, like count, coin count, and share count. For each initialism in the Lexicon, we then calculated the mean value of these six indicators. This composite measure served as the Bilibili adoption intensity index, reflecting the overall level of engagement and resonance generated by a slang item within the Bilibili community. Bilibili adoption intensity was calculated as the mean value of these six indicators. Note: View count X1a: the total number of times a video is viewed; a single user ID may have multiple views for a given video. Bullet comment count X1b: the total number of bullet comments a video receives during the period in which it can be played; a single user ID may have multiple bullet comments for a given video, each counted once. Comment count X1c: the total number of posted comments and replies during the period in which it can be played; a single user ID may have multiple comments for a given video. Like count X1d: the total number of likes a video receives during the period in which it can be played; each user ID may only have one like for a given video (likes may be withdrawn). Coin count X1e: the total number of coins received by a video; each user ID may have up to two coins per video (coins may be withdrawn). Share count X1f: the total number of times a video is shared; each user ID may only have one opportunity to share a video. Operationalisation of community structure variables. Gender innovation diversity. Bilibili provides users with three gender-identification options: male, female, and ‘undisclosed’. The availability of this non-binary ‘undisclosed’ category offers a unique opportunity for examining gender diversity within the user base and, to some extent, reflects more pluralistic attitudes toward gender expression among China’s younger online communities. For each initialism, we calculated the gender distribution of all relevant content creators, based on the percentage of users who self-identified as male, female, or undisclosed. We then applied Simpson’s (1949) diversity index to operationalise and measure this dimension of diversity. Simpson’s index (Dz) is a widely used metric for dual-concept diversity (McDonald & Dimmick, 2003), capturing both the presence of multiple actor categories and the overall evenness of their distribution. To minimise distortions caused by the number of categories, we adopted the standardised form of Simpson’s index, which allows direct comparison across distributions with different category counts. Simpson’s standardised index ranges from 0 (no diversity; a highly skewed distribution) to 1 (maximum diversity; a perfectly even distribution) and is calculated as follows: where pi represents the proportion of each gender category (male, female, and undisclosed). Higher index values indicate greater gender diversity within the creator group. Cross-community penetration breadth. To assess the extent to which each initialism permeates different interest-based communities, we divided Bilibili’s content ecosystem into eight major categories: fandom, gaming, daily life, animation, technology, fashion, sports, and other. For each term in the Lexicon, we analysed the distribution of related videos across these eight categories. Operationalising cross-community penetration breadth is straightforward yet effective: it is defined as the number of distinct categories in which videos containing a given initialism appear. For example, if an initialism appears only within the ‘gaming’ category, its penetration breadth would be 1; if it appears across ‘gaming’, ‘animation’, and ‘daily life’, its value would be 3. The variable therefore ranges from 1 to 8, with higher values indicating that a term is less confined to a single interest community and possesses stronger cross-community penetration capacity. Information fragmentation index. To quantify the effect of information Balkanisation caused by interest-based partition within Bilibili, we constructed an information fragmentation index. This index is designed to measure the degree of dispersion in each initialism’s distribution across different content categories. The core logic is as follows: the more evenly a slang term’s discourse is dispersed across unrelated interest communities, the less likely it is to generate concentrated visibility or consensus within any single community, and the greater the structural barriers it will encounter in its diffusion process. The calculation proceeds in three steps. First, for each initialism under investigation, we identify its total frequency of occurrence (measured by number of videos) across the eight major categories defined earlier (fandom, gaming, daily life, animation, technology, fashion, sports, and other), and compute each category’s proportional distribution. This yields an eight-dimensional vector of proportional values. Next, we apply Simpson’s diversity index to calculate the dispersion degree in the distribution. Values of the index closer to 0 indicate that the term’s diffusion is highly concentrated within one or a small number of categories (for example, a purely gaming-related term that appears 90% of the time in the ‘gaming’ category). In this case, a strong community-level consensus is present and fragmentation is low. By contrast, values closer to 1 indicate that the term is evenly dispersed across all eight categories (each representing approximately 12.5%). This reflects a ‘highly fragmented’ pattern in which the term is loosely associated with many communities but fails to gain deep resonance with any of them. Such a distribution suggests that the term is hindered by structural barriers that limit the formation of diffusion momentum. Therefore, a higher index score reflects a stronger information Balkanisation effect and greater difficulty in achieving successful platform-wide diffusion. Data analysis. After computing all variables described above（see Table 1）, we conducted statistical analyses using SPSS. The multiple linear regression model was employed to test the following relationships: The predictive effect of Bilibili adoption intensity on the national diffusion scale. The predictive effects of gender innovation diversity and cross-community penetration breadth on Bilibili adoption intensity. All analyses were conducted to ensure the findings’ robustness. --Table 1 about here-- Research findings Using multiple regression analysis, we systematically examined the relationship between Bilibili’s community structure and online slang’s diffusion effectiveness (see Figure 1). The results reveal several core findings that, taken together, clearly illustrate the diffusion pathways of linguistic innovation on decentralised platforms. Bilibili as the core source: Platform adoption intensity directly drives nationwide popularity. The regression results strongly support the study’s core hypotheses. Our analysis shows that Bilibili adoption intensity has a highly significant and robust positive predictive effect on the national diffusion scale ( β = .743, p < .001) (see Figure 1, P–P plot). --Figure 1 here-- This finding holds two layers of significance. First, the extremely high level of significance ( p < .001) indicates that the likelihood of this correlation being a product of random error is less than one in a thousand, thus providing strong statistical confidence. Second, the effect size is substantial: a standardised coefficient ( β ) of .743 represents a large effect. In practical terms, after controlling for all other variables, a one–standard-deviation increase in Bilibili adoption intensity is associated with an estimated increase of .743 standard deviations in the nationwide diffusion scale. Taken together, these results provide strong empirical evidence that Bilibili functions as the core source and diffusion hub of online slang in China. The duality of community structure: A double engine and a key barrier (local consensus and diverse adoption). Our analysis further reveals a dual structural mechanism within Bilibili’s community network: one key barrier that suppresses diffusion, and two accelerating engines that promote it (see Table 2). (1) Key barrier: the inhibiting effect of information fragmentation. Results indicate a significant negative correlation between the information fragmentation index and Bilibili adoption intensity ( β = –.450, p < .001). A higher fragmentation score reflects a more dispersed distribution of discourse across unrelated interest communities, preventing the formation of concentrated visibility. This dynamic confirms the prediction of information Balkanisation theory: highly segmented community structures impede the formation of shared cultural meme, constituting a major structural barrier to its diffusion. (2) Accelerating engines: the dual drivers of gender diversity and cross-community penetration. In contrast to the suppressing effect of fragmentation, two factors are found to actively break community barriers and facilitate diffusion: Gender innovation diversity, which shows a significant positive correlation with Bilibili adoption intensity ( β = .274, p < .05). This suggests that when a slang term is created and adopted by a gender-diverse group (including male, female, and undisclosed users), its expression benefits from a wider range of perspectives and communicative styles, resulting in greater expressive vitality and diffusion potential. Cross-community penetration breadth, which is also positively correlated with Bilibili adoption intensity ( β = .371, p < .01). This finding closely aligns with the structural holes theory. Greater penetration breadth indicates that a slang term has functioned as a ‘bridge’, successfully connecting communities such as animation, gaming, and fashion, thereby expanding its potential audience base. This mechanism is empirically demonstrated as the most effective antidote to information Balkanisation. Among the three predictors of Bilibili adoption intensity, the inhibitory power of fragmentation ( β = –.450) is the strongest, more than offsetting the positive effect of gender diversity ( β = .274). This highlights the severity of fragmentation as the primary barrier to diffusion. However, cross-community penetration breadth ( β = .371), as an active connective mechanism, exerts a sufficiently strong positive influence to partially counteract fragmentation, together with gender diversity, forming the “double engine” driving the spread of linguistic innovation. --Table 2 about here-- In summary, the empirical pattern demonstrates what Centola and Macy (2007) describe as a process of ‘complex contagion’ for the diffusion of linguistic innovations (particularly slang that embodies subcultural identities), rather than ‘simple contagion’. The diffusion of linguistic innovations, especially those embedded in subcultural identities, requires reinforcement and validation from multiple independent sources. Connectivity alone is not sufficient. Successful diffusion requires both local consensus (low fragmentation) and diverse adoption (gender diversity) to generate the social reinforcement necessary for complex contagion to take hold. Discussion The Structural Engine of a Bottom–Up Linguistic Revolution. This study demonstrates that the proliferation of alphabet acronyms on Bilibili is not a random phenomenon but the product of a distinct socio-structural engine within the platform. Our findings provide a clear answer to why grassroots platforms have surpassed traditional authorities as the core source of linguistic innovation: it is rooted in the participatory culture of platforms like Bilibili, which provides a fertile ground for identity construction among Chinese youth. As highlighted in prior research, identity is dynamically constituted through discursive practices. The deliberate creation and use of distinctive acronyms foster a potent sense of group belonging, creating a clear ‘we’ versus ‘others’ dichotomy that reinforces collective identity and social boundaries. This identity work fundamentally subverts traditional top-down communication models. Unlike language changes propagated by institutional elites, these acronyms exemplify a purely bottom–up pattern. They originate organically within youth communities, driven by the need for efficient, affective, and community-specific communication. Our regression analysis confirms this by revealing the structural drivers behind their diffusion: Gender innovation diversity (β = .274) acts as a source of creativity, while cross-community penetration breadth (β = .371) enables local innovations to bridge fragmented interest groups. The fact that leading acronyms achieved national propagation volumes, directly predicted by their Bilibili adoption intensity (β = .743), is powerful evidence of this new, structurally-enabled mode of language change, where users are not passive consumers but active architects of their linguistic landscape. Navigating Cohesion and Fragmentation. The findings of this study vividly illustrate the dualistic nature of alphabet acronyms as a linguistic innovation, serving as a tool for in-group solidarity while simultaneously acting as a mechanism for social fragmentation. This duality is precisely quantified by our Interest Fragmentation Index (β = − .450), which shows that when discussion of a slang term is overly dispersed across isolated communities without forming a consensus in any, its diffusion is significantly hindered. This linguistic exclusion transcends mere misunderstanding. The opacity of these acronyms functions as a sophisticated cultural filter; while it efficiently verifies membership and fosters intimacy among initiates, it systematically excludes outsiders, potentially exacerbating intergenerational divides and creating communication chasms between different digital subcultures—a clear manifestation of information balkanization. Therefore, alphabet acronyms are not merely passive reflections of youth culture but are active agents in shaping social dynamics. They embody a central paradox of digital communication: the same innovative practices that empower marginalized voices and create vibrant communities also possess the inherent capacity to foster insularity and reinforce echo chambers. The challenge lies in navigating this delicate balance—appreciating the creative and affiliative functions of these linguistic innovations while mitigating their divisive potential. Glocalization and Nationalism. The widespread adoption of these acronyms is further explained by their potent cultural logic, which balances global forms with local meanings. The coexistence of English-based abbreviations (e.g., LOL, BGM) and Pinyin-based acronyms (e.g., yyds, nsdd) exemplifies glocalization. English-derived terms facilitate participation in global digital culture, particularly in gaming and entertainment, signaling membership in a cosmopolitan generation. Conversely, Pinyin-based abbreviations represent a process of cultural translation. They appropriate the global format of alphabetization to express distinctly local meanings and sentiments. Critically, this "global" packaging of "local" content can, in some contexts, acquire nationalist connotations. When mobilized for collective action or defense of national icons, as has been observed with the use of 'yyds' to praise national achievements, these acronyms become linguistic markers of a distinctly Chinese digital identity. This strategic alignment with mainstream values creates a certain tolerance from official media, allowing a bottom-up linguistic form to be co-opted into a top-down narrative, thereby facilitating its unprecedented propagation scale. Conclusion Through quantitative analysis, this study reveals the pivotal role of Bilibili as a core incubator of Chinese Internet slang, as well as the complex structural mechanisms within its community ecology that drive or inhibit the diffusion of linguistic innovations. Our findings present a diffusion model fundamentally different from traditional top–down paradigms: a bottom–up linguistic revolution initiated by grassroots communities, empowered by platform architecture, and ultimately reshaping the linguistic landscape of the Chinese Internet. The findings point to a clear dynamic model (see Figure 2). First, Bilibili acts as the engine of the entire diffusion process ( β = .743 *** ), with its level of adoption directly determining a slang term’s nationwide influence. Second, information fragmentation emerges as the primary structural barrier to diffusion ( β = –.450 *** ). Meanwhile, gender innovation diversity ( β = .274 * ) and cross-community penetration breadth ( β = .371 ** ) together form a dual engine that overcomes this barrier and drives diffusion. Diversity supplies the initial fuel and resonance necessary for innovation, while cross-community penetration transports the innovation across community boundaries, dismantling cultural silos. Ultimately, only slang that succeeds internally on Bilibili, by activating these dual drivers and overcoming fragmentation, can accumulate sufficient participatory energy to break through platform boundaries and trigger widespread nationwide diffusion. In effect, the outcome produced within Bilibili (Bilibili adoption intensity) determines the scale of diffusion at the national level, thereby enacting a bottom–up “rebellion” against traditional language authority. —Figure 2 here-- From a theoretical perspective, this study integrates participatory culture, structural holes theory, and diffusion of innovations theory, offering a robust framework for understanding cultural production and circulation on decentralised platforms. Beyond validating these theories’ applicability to online linguistic innovation, we advance their analytical precision by operationalising key variables such as gender innovation diversity, information fragmentation, and cross-community penetration breadth. From a practical perspective, the study offers important implications for platform governance, cultural research, and communication strategy: Meaningful engagement with online culture requires attention to the internal architecture of communities—specifically, to the bridges that link otherwise disconnected groups and to the cultivation of diverse innovation environments—rather than a narrow focus on content alone. In sum, this study demonstrates that linguistic innovation in the digital era is no longer governed by top-down norms, but emerges through bottom–up dynamics. The case of Bilibili illustrates that when communities possess both diverse sources of innovation and effective connective structures, even the most marginal and fragmented ‘micro-tribes’ can converge into a sweeping cultural tide, one capable of redefining how we communicate in a networked age. Declarations Author Contribution Ke JIANG: Conceptualization, data curation, formal analysis, methodology and writing.Wei LI: Funding acquisition and supervision. Data Availability Raw data were generated at Peking University. Derived data supporting the findings of this study are available from the corresponding author. Funding Details This study was funded by Peking University’s 2021 Language and Text Research Program: “Monitoring Language and Text in the Field of New Media” (NO. 7121300014). References Bailey CJN (1973) Variation and linguistic theory. Center for Applied Linguistics. Britain D (2002) Space and spatial diffusion. In J. K. Chambers, P. Trudgill, & N. Schilling-Estes (Eds.), The handbook of language variation and change . Blackwell Publishing, pp. 603–637. Britain D (2010a) Language and space: The variationist approach. In P. Auer & J. E. Schmidt (Eds.), Language and space: An international handbook of linguistic variation, Vol. 1: Theories and methods De Gruyter Mouton, pp. 142–162. Chambers, JK, Trudgill P (1980) Dialectology. Cambridge University Press, Cambridge. Trudgill P (1974) Linguistic change and diffusion: Description and explanation in sociolinguistic dialect geography. Language in Society, 3(2), 215–246. https://doi.org/10.1017/S0047404500004358 Jenkins H (2006) Convergence culture: Where old and new media collide. New York University Press, New York. QuestMobile (2021) Bilibili user profile report 2021. QuestMobile Research Institute. Retrieved from https://www.questmobile.com.cn Zhang W, Mao C (2019) Fan activism sustained and challenged: Participatory culture in Chinese online translation communities. New Media & Society, 21(3), 748–765. https://doi.org/10.1177/1461444818801511 Van Alstyne M, Brynjolfsson E (2005) Global village or cyber-Balkans? Modeling and measuring the integration of electronic communities. Management Science, *51*(6), 851–868. Van Alstyne M, Brynjolfsson E (1996) Could the Internet Balkanize science? Science, *274*(5292), 1479–1480. Sunstein CR (2006) Infotopia: How many minds produce knowledge. Oxford University Press, Oxford. Pariser E (2011) The filter bubble: What the Internet is hiding from you. Penguin Press, New York. Chang WY (2008) The Cyber Balkanization and structural transformation of the public sphere in Korea. Journal of Contemporary Eastern Asia, *7*(2), 29–48. Burt RS (1992) Structural holes: The social structure of competition. Harvard University Press, Cambridge. Burt RS (1995) Structural holes: The social structure of competition. Harvard University Press, Cambridge. Centola D, Macy M (2007) Complex contagions and the weakness of long ties. American Journal of Sociology, 113(3), 702-734. Tables Table 1. Variables. Variable Type English Name Chinese Name Operational Definition Dependent Variable National Diffusion Scale 全网传播规模 Total frequency of diffusion of the 66 keywords across major Chinese platforms Dependent Variable Bilibili Adoption Intensity B站采纳强度 Composite index of six interaction indicators for relevant Bilibili videos (views, bullet comments, etc.) Independent Variable Gender Innovation Diversity 性别创新多样性 Simpson diversity index based on the gender distribution (male/female/undisclosed) of content creators Independent Variable Cross-Community Penetration Breadth 跨圈层渗透广度 Number of Bilibili content categories (1–8) in which each initialism appears Independent Variable Information Fragmentation Index 信息碎片化指数 Simpson diversity index based on the distribution of each term across eight Bilibili categories Table 2. Regression model summary and ANOVA for predictors of Bilibili adoption intensity. Predictor Unstandardized Coefficient (B) Standardized Coefficient (β) t -value p -value (Constant) 8.611 5.041 0.000 Gender Innovation Diversity 5.13 0.274 2.364 0.021 Information Fragmentation Index -4.295 -0.45 -4.127 0.000 Cross-Community Penetration 0.566 0.371 3.022 0.004 Note : Dependent variable: Bilibili adoption intensity. R ² = 0.357, Adjusted R ² = 0.326, F = 11.458， p < .001. Additional Declarations No competing interests reported. Supplementary Files RebelLexicons.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8026908","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":560391791,"identity":"471bd701-2bb0-47d0-8dd4-d18e6bfa4627","order_by":0,"name":"Ke Jiang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1klEQVRIie3LMQuCQBjG8TeCWhTXE8m+gtHYl3lDyKVWpxKncylaD/oSjY0nB7qIc0ODLW1BbTUEebanbkH3X+5eeH4AKtUPRgCQl6/9OXstyLgVqZqGjYkZhWf+PKy8LXMLuPkCjF34nVgax3iTpQt2nDkdlgsgJ/6d2ASR6zRZ7MkcujoV4BCsIcMC4xdNPEeSVxNiEUCh0yVWpNOEmGtEMaB8xLKLE69zTyPHGkLSzL1faTA0IvdcPPyJbbAaAqDJhai+XJ51+7K+HAYNhiqVSvW3vQHO8UX2LiL0KgAAAABJRU5ErkJggg==","orcid":"","institution":"Communication University of China","correspondingAuthor":true,"prefix":"","firstName":"Ke","middleName":"","lastName":"Jiang","suffix":""},{"id":560391793,"identity":"886af2a7-1ff1-4d7b-bc21-a74a7052f2a5","order_by":1,"name":"Wei Li","email":"","orcid":"","institution":"Peking University","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2025-11-04 09:08:40","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8026908/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8026908/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98501181,"identity":"0bfe4992-c6fb-4abd-aabd-49c7b02470db","added_by":"auto","created_at":"2025-12-18 09:40:54","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":651350,"visible":true,"origin":"","legend":"","description":"","filename":"RebelLexiconsHowBilibilisFragmentedTribesEngineerSlangDiffusionThroughCommunityStructureAnonymised.docx","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/b3b8ed46a9a82a4ca64c5770.docx"},{"id":98501187,"identity":"44463e20-e876-4fbc-bbeb-1036fdad5d3d","added_by":"auto","created_at":"2025-12-18 09:40:59","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36117,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/ec4f5a9a2c5c194912ad5c30.docx"},{"id":98501184,"identity":"b4fa6a42-7c1e-4cdf-b2c6-7b0acfe9eec2","added_by":"auto","created_at":"2025-12-18 09:40:58","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70174,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/d250add3cad3ea3a4b296f3f.docx"},{"id":98501171,"identity":"b85bad11-a615-4c25-b4e5-7920ec44734a","added_by":"auto","created_at":"2025-12-18 09:40:46","extension":"json","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3975,"visible":true,"origin":"","legend":"","description":"","filename":"95bb738df25d405f8f3d3d2e3650f5ca.json","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/d5d95a3b79f35f18d80bd6a5.json"},{"id":98501174,"identity":"f50ac925-674e-4695-819d-89a083afb08c","added_by":"auto","created_at":"2025-12-18 09:40:49","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":195765,"visible":true,"origin":"","legend":"","description":"","filename":"95bb738df25d405f8f3d3d2e3650f5ca1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/5e8008e2388c7b344de20ba9.xml"},{"id":98501166,"identity":"9ac7ec63-a3f7-449a-baa4-172e3ac4e350","added_by":"auto","created_at":"2025-12-18 09:40:45","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":295167,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/53c0260bbb0f5670874d7169.png"},{"id":98501179,"identity":"da156e8c-d089-4d70-aca7-3c237cdb2fe4","added_by":"auto","created_at":"2025-12-18 09:40:52","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":279260,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/402b4a1c635648e55b4f80a0.png"},{"id":98501191,"identity":"3573f388-c042-4e83-b9d2-b0b73206c0fa","added_by":"auto","created_at":"2025-12-18 09:41:00","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21775,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/3ca96ca7d880fa2555274298.png"},{"id":98501176,"identity":"264b30da-afa8-4ce4-b764-e220f8b8966b","added_by":"auto","created_at":"2025-12-18 09:40:50","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":55664,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/d53a53ea1fc541954d7d6c6f.png"},{"id":98501178,"identity":"c5944d0f-4788-442d-87c9-e90c7c613955","added_by":"auto","created_at":"2025-12-18 09:40:51","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":54929,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/4c93aa650635c8ab390d870b.png"},{"id":98501170,"identity":"7c3c9868-843a-460b-a83c-dc8dacfa4fa4","added_by":"auto","created_at":"2025-12-18 09:40:46","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":47789,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/ca2c0a7ecb8f117853d4dcc1.png"},{"id":98501186,"identity":"6ac60616-bec7-49eb-97a5-c3d2ac2ef4d8","added_by":"auto","created_at":"2025-12-18 09:40:59","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9749,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/cf56dd3256cfda2b345cf8ad.png"},{"id":98501175,"identity":"e1699f8b-2447-4504-87be-87f838cabafa","added_by":"auto","created_at":"2025-12-18 09:40:49","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20415,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/1dd16b3a8a8deacbc395d457.png"},{"id":98501177,"identity":"3587bd16-3ca9-4d40-bc85-e83ffa4d9fee","added_by":"auto","created_at":"2025-12-18 09:40:50","extension":"xml","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":194312,"visible":true,"origin":"","legend":"","description":"","filename":"95bb738df25d405f8f3d3d2e3650f5ca1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/3899a076a1f43ce08a3458c4.xml"},{"id":98501164,"identity":"d6de96f6-4a1f-4a1a-803e-b66f29a79739","added_by":"auto","created_at":"2025-12-18 09:40:44","extension":"html","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":194953,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/91e3f6d895fdc8fe933dc958.html"},{"id":98501192,"identity":"52ca1de5-8360-448d-8818-4f85fd63a426","added_by":"auto","created_at":"2025-12-18 09:41:01","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":27721,"visible":true,"origin":"","legend":"\u003cp\u003eP–P plot.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/c4a056ed0797977ff2a79508.png"},{"id":98501172,"identity":"5245029d-70d8-4dc7-bb4e-e8d6996dbec2","added_by":"auto","created_at":"2025-12-18 09:40:47","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":55664,"visible":true,"origin":"","legend":"\u003cp\u003eStructural model.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/d7858204d4d2a2c762df71e7.png"},{"id":99862008,"identity":"131fe9f7-3d41-47cb-8374-dbf3496a2d31","added_by":"auto","created_at":"2026-01-09 07:10:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1192087,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/e0bbd97c-094d-46f9-9c5c-12dbd8a95e6f.pdf"},{"id":98501190,"identity":"a0fe83b0-76b3-4f57-99e8-5dba582d8884","added_by":"auto","created_at":"2025-12-18 09:41:00","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":623787,"visible":true,"origin":"","legend":"","description":"","filename":"RebelLexicons.docx","url":"https://assets-eu.researchsquare.com/files/rs-8026908/v1/594df10bc21d8d063c8e698c.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Rebel Lexicons: How Bilibili’s Fragmented Communities Engineer Slang Diffusion","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe emergence of language and writing marked a new dawn in human civilisation. In today\u0026rsquo;s digital world, much of China\u0026rsquo;s youth culture and broader social dynamics remain opaque to outside observers. The continual proliferation of linguistic innovations online thus offers a crucial lens through which to discern these otherwise hidden transformations. Traditional models of linguistic innovation and diffusion typically describe a process in which a linguistic feature spreads gradually from a relatively narrow geographical area or social group to broader regions and larger populations. Bailey\u0026rsquo;s (1973) wave model conceptualises this diffusion as a series of \u0026lsquo;concentric\u0026rsquo; ripples that advance outward step by step from a central point. Building on this metaphor, Britain (2010a: 148) invokes the \u0026lsquo;ripple effect\u0026rsquo;, noting that \u0026lsquo;innovations, like the ripple effect caused when a pebble is dropped into a puddle, radiate over time out from a central focal area, reaching physically nearby locations before those at ever greater distances\u0026rsquo;. Another water-based model is the \u0026ldquo;cascade diffusion\u0026rdquo; hypothesis, which Chambers and Trudgill (1980: 192) describe as a process completed \u0026lsquo;by the \u0026ldquo;jumping\u0026rdquo; of the innovation from one large town to another, and from these to smaller towns, and so on\u0026rdquo;. This top\u0026ndash;down pattern of dissemination in terms of urban hierarchies is also known as the \u0026lsquo;hierarchical effect\u0026rsquo; (Britain 2002: 623). Trudgill\u0026rsquo;s (1974) study of sound change in Norwegian dialects provides a clear example: the innovation spread progressively from major urban centres to smaller towns, and ultimately to rural areas.\u003c/p\u003e\n\u003cp\u003eSuch\u0026nbsp;\u0026lsquo;centre\u0026ndash;periphery\u0026rsquo;\u0026nbsp;or top\u0026ndash;down diffusion models have long been regarded as the dominant pattern of linguistic spread. However, within China\u0026rsquo;s contemporary online environment, a new bottom\u0026ndash;up trajectory of linguistic diffusion has begun to emerge. The widely circulated initialism\u0026nbsp;\u0026ldquo;yyds\u0026rdquo;\u0026nbsp;(derived from the pinyin initials of yǒngyuǎn de sh\u0026eacute;n, literally\u0026nbsp;\u0026ldquo;eternal god\u0026rdquo;), which rose to prominence in 2021, provides a representative example. The term first gained traction within the youth-oriented cultural community of Bilibili, before rapidly permeating the wider Sinosphere and even appearing repeatedly in official mainstream media, ultimately generating more than 200 million circulation instances. As emblematic linguistic forms of youth subculture, such online initialisms are widely used among Generation Z. They frequently originate in marginal or niche communities at society\u0026rsquo;s periphery, embodying perspectives and expressive styles that diverge from those of the mainstream. Non-mainstream expressions like \u0026ldquo;yyds\u0026rdquo; not only draw the attention and commentary of mainstream media, but also signal the forceful rise of subcultural voices within the digital sphere. Their diffusion is not driven by institutional authority; instead, it is produced within marginal communities and amplified by participatory cultures and platform architectures, enabling bottom up cross-community dissemination.\u003c/p\u003e\n\u003cp\u003eThis phenomenon raises two central questions: why have grassroots platforms displaced traditional authorities to become key engines of linguistic innovation? And what underlying social-structural mechanisms\u0026mdash;specifically, what structural features of platform-based communities\u0026mdash;enable localised innovations to escalate into large-scale, society-wide diffusion?\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTheoretical framework and research questions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study examines how online linguistic forms diffuse from a grassroots platform to central platforms. On the surface, this concerns a linguistic innovation and diffusion process; at a deeper level, however, it concerns the shifting and restructuring of linguistic authority within the contemporary Chinese Internet. We argue that the emergence and circulation of online slang on Bilibili is not merely a linguistic phenomenon, but the outcome produced by the interaction between the platform\u0026rsquo;s distinctive participatory cultural ecology and its community structure tension. This process challenges the traditional language authority held predominantly by official media, giving rise to what we describe as \u0026lsquo;rebel lexicons\u0026rsquo;. To account for this reversal, we develop an integrated theoretical framework and a set of key concepts that link four core theories to our empirical variables.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLinguistic innovation on participatory culture platforms: a theoretical lens for observing Bilibili adoption intensity.\u0026nbsp;\u003c/strong\u003eThe theory of participatory culture, proposed by Jenkins (2006), emphasises how media users actively engage in producing meaning by appropriating, remaking, and disseminating existing media content, thereby forming informal communities marked by strong social connectivity and shared norms. This theory disrupts the traditional paradigm of passive consumption by conceptualising media use as a social and creative practice. Participatory culture has also been widely applied in studies of Chinese online communities (Zhang \u0026amp; Mao, 2019).\u003c/p\u003e\n\u003cp\u003eBilibili is a paradigmatic example of participatory culture in China. Originating as a bullet-comment video community oriented towards animation, comics, and games (ACG) subculture, it has, over more than a decade, expanded into a comprehensive cultural platform dominated by Generation Z and characterised by diverse content genres. According to data from QuestMobile, as of 2021, 82% of Bilibili users belonged to Generation Z, with secondary school and university students forming the core user base. Growing up in a highly digital environment, this cohort is adept at constructing cultural identities through symbolic creativity, secondary dissemination, and community interaction. Bilibili\u0026rsquo;s ecosystem, including bullet-comment interaction, secondary video production (such as Kichiku and MAD/AMV), and layered interest communities, greatly facilitates imitation, parody, and collaborative creativity among users. The initialisms found on the platform (e.g. \u0026lsquo;yyds\u0026rsquo;, \u0026lsquo;xswl\u0026rsquo;) are emblematic linguistic products of this environment: they serve not only as internal \u0026lsquo;codes\u0026rsquo; of communication, but also as symbolic carriers through which users express belonging and cultural creativity.\u003c/p\u003e\n\u003cp\u003eUnlike traditional official media that rely on one-way dissemination, Bilibili has developed a multi-layered and highly interactive participation mechanism. Its interactive affordances span not only conventional functions such as commenting and reposting, but also bullet comments, likes, \u0026lsquo;coins\u0026rsquo;, and other features, together comprising six forms of interaction that significantly increase user willingness to participate as well as the engagement intensity. In this study, we employ Bilibili adoption intensity as a core variable and calculate it using the mean value of these six interaction indicators to quantify the level of participation triggered by each slang term. This variable reflects not only a term\u0026rsquo;s popularity, but also its degree of embeddedness within the community and its potential for wider social diffusion. Accordingly, we conceptualise Bilibili as a \u0026lsquo;participatory incubator\u0026rsquo; in which technological architectures (such as bullet comments and content-partition systems) and cultural practices (such as meme-making and secondary creation) jointly constitute the foundational environment for linguistic innovation. Based on this theoretical mechanism, we propose the following hypothesis:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypothesis 1: The Bilibili adoption intensity of an online slang term is positively correlated with its national diffusion scale.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTheoretical mechanism of information Balkanisation and the information fragmentation index.\u0026nbsp;\u003c/strong\u003eThe concept of \u0026lsquo;Balkanisation\u0026rsquo; originates in geopolitical studies, where it refers to the division of a region into multiple small, mutually antagonistic entities (Van Alstyne \u0026amp; Brynjolfsson, 1996, 2005). In the context of information studies, information Balkanisation (Chang, 2008) is used to describe Internet users\u0026rsquo; tendency to become fragmented into numerous isolated and highly homogeneous interest groups, such as \u0026lsquo;echo chambers\u0026rsquo; or \u0026lsquo;filter bubbles\u0026rsquo;. Within these clusters, particular viewpoints and preferences are continually reinforced, while cross-group information flows and communication are significantly impeded (Sunstein 2006; Pariser 2011).\u003c/p\u003e\n\u003cp\u003eBilibili\u0026rsquo;s content architecture is a representative case of information Balkanisation. Originating in ACG subculture, the platform gradually expanded into a comprehensive multi-interest community. Its professional user-generated content (PUGC) system comprises 16 formal content categories, including life, gaming, animation, knowledge, fashion, Kichiku, and others, each of which is further subdivided into multiple subcategories. These content partitions operate in relative isolation and collectively encompass nearly all interest dimensions of Generation Z users. While this structure enables highly accurate content matching, it also produces unintended information silos that limit communication and visibility across groups (see Annex).\u003c/p\u003e\n\u003cp\u003eIn this study, we employ an information fragmentation index to operationalise the Balkanisation effect by measuring the degree of dispersion in a slang term\u0026rsquo;s distribution across Bilibili\u0026rsquo;s content categories. A higher value on this index indicates that discussions of the slang item are scattered across multiple subcommunities without forming a concentrated volume of discourse, thereby reflecting structural barriers to diffusion within the platform. Based on this theoretical framework, we propose the following hypothesis:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypothesis 2: The information fragmentation index of an online slang term is negatively correlated with its Bilibili adoption intensity.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStructural holes and the bridging mechanism of cross-community penetration breadth.\u0026nbsp;\u003c/strong\u003eThe structural holes theory, proposed by Burt (1992, 1995), posits that when disconnected clusters exist within a social network, the gaps separating them constitute \u0026lsquo;structural holes\u0026rsquo;. Actors or nodes who can bridge these holes, often described as bridges, occupy advantageous positions, gaining disproportionate access to information, control over resources, and influence in diffusion processes.\u003c/p\u003e\n\u003cp\u003eIn this study, the variable cross-community penetration breadth directly operationalises this bridging mechanism. By measuring the distributional range of a specific slang term across Bilibili\u0026rsquo;s interest-based content categories (such as animation, gaming, fashion, and knowledge), this indicator quantifies the extent to which a slang item transcends its original community and reaches diverse user groups. A slang term that penetrates multiple communities indicates successful adoption and circulation by bridge users (for example, content creators who participate in both animation and music communities) or bridge content, thereby filling structural holes between otherwise fragmented communities. These bridging nodes constitute the crucial channels for cross-boundary information flow, effectively overcoming the diffusion barriers imposed by information Balkanisation. Cross-community penetration breadth therefore reflects not only a slang term\u0026rsquo;s inherent diffusion capacity, but also the mobilisation potential of connective structures within Bilibili\u0026rsquo;s community network. Based on the structural holes theory, we propose the following hypothesis:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypothesis 3: The cross-community penetration breadth of an online slang term is positively correlated with its Bilibili adoption intensity.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInnovation diffusion and the initiating role of gender innovation diversity.\u0026nbsp;\u003c/strong\u003eThe diffusion of innovations theory, introduced by Rogers (1962), explains how new ideas, practices, or artefacts spread among members of a social system through specific channels temporally. The theory emphasises four key dimensions: inherent attributes of the innovation, communication channels, temporal processes, and social structure, that jointly shape the trajectory and pace of diffusion. In this study, initialisms are conceptualised as a form of cultural innovation. Rogers argues that the success of diffusion depends heavily on whether different categories of adopters within a social system (e.g. innovators, early adopters, early majority) embrace the innovation, and that diversity among adopter groups is a central structural factor influencing diffusion outcomes.\u003c/p\u003e\n\u003cp\u003eThe variable \u0026lsquo;gender innovation diversity\u0026rsquo; operationalises the concept of diversity along gender lines. By measuring the extent to which users identifying as male, female or \u0026lsquo;undisclosed\u0026rsquo; on Bilibili co-create and adopt a given slang term, this indicator captures the heterogeneity of adopter composition at the origin of diffusion. Higher gender diversity implies a richer collision and fusion of cognitive perspectives, social experiences, and communicative styles, significantly broadening linguistic innovation\u0026rsquo;s semantic resource pool and expressive pathways. Such diversity provides a wider \u0026lsquo;gene pool\u0026rsquo; for the emergence and evolution of cultural memes. Accordingly, gender innovation diversity acts as an innovation initiator within the diffusion process: it ensures that Bilibili, as a \u0026lsquo;participatory incubator\u0026rsquo;, sustains a diverse and vibrant pool of innovative raw material, thereby increasing the likelihood that highly resonant and adaptable cultural innovations will emerge. Based on the diffusion of innovations theory, we propose the following hypothesis:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypothesis 4: The gender innovation diversity of an online slang term is positively correlated with its Bilibili adoption intensity.\u003c/strong\u003e\u003c/p\u003e"},{"header":"Research methods and implementation","content":"\u003cp\u003eThis study adopts a multi-step, data-driven methodological design to investigate the origins and diffusion mechanisms of Chinese Internet slang, specifically initialisms. The research process consists of three main stages: (1) Lexicon construction and measurement of national diffusion scale; (2) Quantification of Bilibili adoption intensity; (3) Computation of community-structure variables (gender diversity and cross-community penetration breadth). The procedures are as follows.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLexicon compilation and national diffusion scale measurement.\u0026nbsp;\u003c/strong\u003eUsing the Internet language database published by Peking University as an authoritative source, we selected 66 representative initialisms (such as \u0026lsquo;YYDS\u0026rsquo; and \u0026lsquo;XSWL\u0026rsquo;) as the objects of analysis, thereby constructing the core Lexicon for this study. To measure the nationwide popularity of these initialisms, we collaborated with the Tencent Big Data Platform. Tencent\u0026rsquo;s dataset aggregates mainstream content platforms in China, including WeChat, Weibo, QQ, Zhihu, and Douyin, and provides statistical access to publicly available content with more than 100,000 views. For each initialism in the Lexicon, we retrieved its frequency of appearance throughout 2021 across all cooperating platforms, including its occurrences in article texts, posts, video titles, and tags. We then summed the total number of appearances for each item across platforms to compute the key dependent variable, the national diffusion scale, operationalised as the total number of mentions of a given initialism on the Chinese Internet in 2021.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBilibili adoption intensity measurement.\u0026nbsp;\u003c/strong\u003eTo verify the hypothesised role of Bilibili as a primary source of online slang innovation, we conducted systematic data extraction for all 66 initialisms on the Bilibili platform. Using Python-based web-scraping tools, we first searched for all videos published in 2021 that contained a target term as a keyword. For instance, in the case of \u0026lsquo;YYDS\u0026rsquo;, we identified 368 core videos featuring the term. To move beyond simple occurrence counts and more precisely capture each term\u0026rsquo;s level of engagement, we collected six key interaction indicators for every relevant video: view count, bullet comment count, comment count, like count, coin count, and share count. For each initialism in the Lexicon, we then calculated the mean value of these six indicators. This composite measure served as the Bilibili adoption intensity index, reflecting the overall level of engagement and resonance generated by a slang item within the Bilibili community.\u003c/p\u003e\n\u003cp\u003eBilibili adoption intensity was calculated as the mean value of these six indicators.\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1766050222.png\" width=\"684\" height=\"64\"\u003e\u003c/p\u003e\n\u003cp\u003eNote:\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eView count X1a: the total number of times a video is viewed; a single user ID may have multiple views for a given video.\u003c/p\u003e\n\u003cp\u003eBullet comment count X1b: the total number of bullet comments a video receives during the period in which it can be played; a single user ID may have multiple bullet comments for a given video, each counted once.\u003c/p\u003e\n\u003cp\u003eComment count X1c: the total number of posted comments and replies during the period in which it can be played; a single user ID may have multiple comments for a given video.\u003c/p\u003e\n\u003cp\u003eLike count X1d: the total number of likes a video receives during the period in which it can be played; each user ID may only have one like for a given video (likes may be withdrawn).\u003c/p\u003e\n\u003cp\u003eCoin count X1e: the total number of coins received by a video; each user ID may have up to two coins per video (coins may be withdrawn).\u003c/p\u003e\n\u003cp\u003eShare count X1f: the total number of times a video is shared; each user ID may only have one opportunity to share a video.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOperationalisation of community structure variables.\u0026nbsp;\u003c/strong\u003e\u003cem\u003eGender innovation diversity.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/em\u003eBilibili provides users with three gender-identification options: male, female, and \u0026lsquo;undisclosed\u0026rsquo;. The availability of this non-binary \u0026lsquo;undisclosed\u0026rsquo; category offers a unique opportunity for examining gender diversity within the user base and, to some extent, reflects more pluralistic attitudes toward gender expression among China\u0026rsquo;s younger online communities.\u003c/p\u003e\n\u003cp\u003eFor each initialism, we calculated the gender distribution of all relevant content creators, based on the percentage of users who self-identified as male, female, or undisclosed. We then applied Simpson\u0026rsquo;s (1949) diversity index to operationalise and measure this dimension of diversity. Simpson\u0026rsquo;s index (Dz) is a widely used metric for dual-concept diversity (McDonald \u0026amp; Dimmick, 2003), capturing both the presence of multiple actor categories and the overall evenness of their distribution. To minimise distortions caused by the number of categories, we adopted the standardised form of Simpson\u0026rsquo;s index, which allows direct comparison across distributions with different category counts. Simpson\u0026rsquo;s standardised index ranges from 0 (no diversity; a highly skewed distribution) to 1 (maximum diversity; a perfectly even distribution) and is calculated as follows:\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"https://myfiles.space/user_files/58895_8739fc6c57c1c19a/58895_custom_files/img1766050254.png\" width=\"478\" height=\"159\"\u003e\u003c/p\u003e\n\u003cp\u003ewhere pi represents the proportion of each gender category (male, female, and undisclosed). Higher index values indicate greater gender diversity within the creator group.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCross-community penetration breadth.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/em\u003eTo assess the extent to which each initialism permeates different interest-based communities, we divided Bilibili\u0026rsquo;s content ecosystem into eight major categories: fandom, gaming, daily life, animation, technology, fashion, sports, and other. For each term in the Lexicon, we analysed the distribution of related videos across these eight categories. Operationalising cross-community penetration breadth is straightforward yet effective: it is defined as the number of distinct categories in which videos containing a given initialism appear. For example, if an initialism appears only within the \u0026lsquo;gaming\u0026rsquo; category, its penetration breadth would be 1; if it appears across \u0026lsquo;gaming\u0026rsquo;, \u0026lsquo;animation\u0026rsquo;, and \u0026lsquo;daily life\u0026rsquo;, its value would be 3. The variable therefore ranges from 1 to 8, with higher values indicating that a term is less confined to a single interest community and possesses stronger cross-community penetration capacity.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eInformation fragmentation index.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/em\u003eTo quantify the effect of information Balkanisation caused by interest-based partition within Bilibili, we constructed an information fragmentation index. This index is designed to measure the degree of dispersion in each initialism\u0026rsquo;s distribution across different content categories. The core logic is as follows: the more evenly a slang term\u0026rsquo;s discourse is dispersed across unrelated interest communities, the less likely it is to generate concentrated visibility or consensus within any single community, and the greater the structural barriers it will encounter in its diffusion process.\u003c/p\u003e\n\u003cp\u003eThe calculation proceeds in three steps. First, for each initialism under investigation, we identify its total frequency of occurrence (measured by number of videos) across the eight major categories defined earlier (fandom, gaming, daily life, animation, technology, fashion, sports, and other), and compute each category\u0026rsquo;s proportional distribution. This yields an eight-dimensional vector of proportional values. Next, we apply Simpson\u0026rsquo;s diversity index to calculate the dispersion degree in the distribution. Values of the index closer to 0 indicate that the term\u0026rsquo;s diffusion is highly concentrated within one or a small number of categories (for example, a purely gaming-related term that appears 90% of the time in the \u0026lsquo;gaming\u0026rsquo; category). In this case, a strong community-level consensus is present and fragmentation is low. By contrast, values closer to 1 indicate that the term is evenly dispersed across all eight categories (each representing approximately 12.5%). This reflects a \u0026lsquo;highly fragmented\u0026rsquo; pattern in which the term is loosely associated with many communities but fails to gain deep resonance with any of them. Such a distribution suggests that the term is hindered by structural barriers that limit the formation of diffusion momentum. Therefore, a higher index score reflects a stronger information Balkanisation effect and greater difficulty in achieving successful platform-wide diffusion.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eData analysis.\u0026nbsp;\u003c/em\u003eAfter computing all variables described above（see Table 1）, we conducted statistical analyses using SPSS. The multiple linear regression model was employed to test the following relationships:\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003eThe predictive effect of Bilibili adoption intensity on the national diffusion scale.\u003c/li\u003e\n \u003cli\u003eThe predictive effects of gender innovation diversity and cross-community penetration breadth on Bilibili adoption intensity. All analyses were conducted to ensure the findings\u0026rsquo; robustness.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cstrong\u003e--Table 1 about here--\u003c/strong\u003e\u003c/p\u003e"},{"header":"Research findings","content":"\u003cp\u003eUsing multiple regression analysis, we systematically examined the relationship between Bilibili\u0026rsquo;s community structure and online slang\u0026rsquo;s diffusion effectiveness (see Figure 1). The results reveal several core findings that, taken together, clearly illustrate the diffusion pathways of linguistic innovation on decentralised platforms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBilibili as the core source: Platform adoption intensity directly drives nationwide popularity.\u0026nbsp;\u003c/strong\u003eThe regression results strongly support the study\u0026rsquo;s core hypotheses. Our analysis shows that Bilibili adoption intensity has a highly significant and robust positive predictive effect on the national diffusion scale (\u003cem\u003e\u0026beta;\u003c/em\u003e = .743, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001) (see Figure 1, P\u0026ndash;P plot).\u003c/p\u003e\n\u003cp\u003e--Figure 1 here--\u003c/p\u003e\n\u003cp\u003eThis finding holds two layers of significance. First, the extremely high level of significance (\u003cem\u003ep\u003c/em\u003e \u0026lt; .001) indicates that the likelihood of this correlation being a product of random error is less than one in a thousand, thus providing strong statistical confidence. Second, the effect size is substantial: a standardised coefficient (\u003cem\u003e\u0026beta;\u003c/em\u003e) of .743 represents a large effect. In practical terms, after controlling for all other variables, a one\u0026ndash;standard-deviation increase in Bilibili adoption intensity is associated with an estimated increase of .743 standard deviations in the nationwide diffusion scale. Taken together, these results provide strong empirical evidence that Bilibili functions as the core source and diffusion hub of online slang in China.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eThe duality of community structure: A double engine and a key barrier (local consensus and diverse adoption).\u0026nbsp;\u003c/strong\u003eOur analysis further reveals a dual structural mechanism within Bilibili\u0026rsquo;s community network: one key barrier that suppresses diffusion, and two accelerating engines that promote it (see Table 2).\u003c/p\u003e\n\u003cp\u003e(1) Key barrier: the inhibiting effect of information fragmentation. Results indicate a significant negative correlation between the information fragmentation index and Bilibili adoption intensity (\u003cem\u003e\u0026beta;\u003c/em\u003e = \u0026ndash;.450, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001). A higher fragmentation score reflects a more dispersed distribution of discourse across unrelated interest communities, preventing the formation of concentrated visibility. This dynamic confirms the prediction of information Balkanisation theory: highly segmented community structures impede the formation of shared cultural meme, constituting a major structural barrier to its diffusion.\u003c/p\u003e\n\u003cp\u003e(2) Accelerating engines: the dual drivers of gender diversity and cross-community penetration. In contrast to the suppressing effect of fragmentation, two factors are found to actively break community barriers and facilitate diffusion: Gender innovation diversity, which shows a significant positive correlation with Bilibili adoption intensity (\u003cem\u003e\u0026beta;\u003c/em\u003e = .274, \u003cem\u003ep\u003c/em\u003e \u0026lt; .05). This suggests that when a slang term is created and adopted by a gender-diverse group (including male, female, and undisclosed users), its expression benefits from a wider range of perspectives and communicative styles, resulting in greater expressive vitality and diffusion potential. Cross-community penetration breadth, which is also positively correlated with Bilibili adoption intensity (\u003cem\u003e\u0026beta;\u003c/em\u003e = .371, \u003cem\u003ep\u003c/em\u003e \u0026lt; .01). This finding closely aligns with the structural holes theory. Greater penetration breadth indicates that a slang term has functioned as a \u0026lsquo;bridge\u0026rsquo;, successfully connecting communities such as animation, gaming, and fashion, thereby expanding its potential audience base. This mechanism is empirically demonstrated as the most effective antidote to information Balkanisation.\u003c/p\u003e\n\u003cp\u003eAmong the three predictors of Bilibili adoption intensity, the inhibitory power of fragmentation (\u003cem\u003e\u0026beta;\u003c/em\u003e = \u0026ndash;.450) is the strongest, more than offsetting the positive effect of gender diversity (\u003cem\u003e\u0026beta;\u003c/em\u003e = .274). This highlights the severity of fragmentation as the primary barrier to diffusion. However, cross-community penetration breadth (\u003cem\u003e\u0026beta;\u003c/em\u003e = .371), as an active connective mechanism, exerts a sufficiently strong positive influence to partially counteract fragmentation, together with gender diversity, forming the \u0026ldquo;double engine\u0026rdquo; driving the spread of linguistic innovation.\u003c/p\u003e\n\u003cp\u003e--Table 2 about here--\u003c/p\u003e\n\u003cp\u003eIn summary, the empirical pattern demonstrates what Centola and Macy (2007) describe as a process of \u0026lsquo;complex contagion\u0026rsquo; for the diffusion of linguistic innovations (particularly slang that embodies subcultural identities), rather than \u0026lsquo;simple contagion\u0026rsquo;. The diffusion of linguistic innovations, especially those embedded in subcultural identities, requires reinforcement and validation from multiple independent sources. Connectivity alone is not sufficient. Successful diffusion requires both local consensus (low fragmentation) and diverse adoption (gender diversity) to generate the social reinforcement necessary for complex contagion to take hold.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e \u003cb\u003eThe Structural Engine of a Bottom\u0026ndash;Up Linguistic Revolution.\u003c/b\u003e This study demonstrates that the proliferation of alphabet acronyms on Bilibili is not a random phenomenon but the product of a distinct socio-structural engine within the platform. Our findings provide a clear answer to why grassroots platforms have surpassed traditional authorities as the core source of linguistic innovation: it is rooted in the participatory culture of platforms like Bilibili, which provides a fertile ground for identity construction among Chinese youth. As highlighted in prior research, identity is dynamically constituted through discursive practices. The deliberate creation and use of distinctive acronyms foster a potent sense of group belonging, creating a clear \u0026lsquo;we\u0026rsquo; versus \u0026lsquo;others\u0026rsquo; dichotomy that reinforces collective identity and social boundaries.\u003c/p\u003e \u003cp\u003eThis identity work fundamentally subverts traditional top-down communication models. Unlike language changes propagated by institutional elites, these acronyms exemplify a purely bottom\u0026ndash;up pattern. They originate organically within youth communities, driven by the need for efficient, affective, and community-specific communication. Our regression analysis confirms this by revealing the structural drivers behind their diffusion: Gender innovation diversity (β\u0026thinsp;=\u0026thinsp;.274) acts as a source of creativity, while cross-community penetration breadth (β\u0026thinsp;=\u0026thinsp;.371) enables local innovations to bridge fragmented interest groups. The fact that leading acronyms achieved national propagation volumes, directly predicted by their Bilibili adoption intensity (β\u0026thinsp;=\u0026thinsp;.743), is powerful evidence of this new, structurally-enabled mode of language change, where users are not passive consumers but active architects of their linguistic landscape.\u003c/p\u003e \u003cp\u003e \u003cb\u003eNavigating Cohesion and Fragmentation.\u003c/b\u003e The findings of this study vividly illustrate the dualistic nature of alphabet acronyms as a linguistic innovation, serving as a tool for in-group solidarity while simultaneously acting as a mechanism for social fragmentation. This duality is precisely quantified by our Interest Fragmentation Index (β = \u0026minus;\u0026thinsp;.450), which shows that when discussion of a slang term is overly dispersed across isolated communities without forming a consensus in any, its diffusion is significantly hindered.\u003c/p\u003e \u003cp\u003eThis linguistic exclusion transcends mere misunderstanding. The opacity of these acronyms functions as a sophisticated cultural filter; while it efficiently verifies membership and fosters intimacy among initiates, it systematically excludes outsiders, potentially exacerbating intergenerational divides and creating communication chasms between different digital subcultures\u0026mdash;a clear manifestation of information balkanization. Therefore, alphabet acronyms are not merely passive reflections of youth culture but are active agents in shaping social dynamics. They embody a central paradox of digital communication: the same innovative practices that empower marginalized voices and create vibrant communities also possess the inherent capacity to foster insularity and reinforce echo chambers. The challenge lies in navigating this delicate balance\u0026mdash;appreciating the creative and affiliative functions of these linguistic innovations while mitigating their divisive potential.\u003c/p\u003e \u003cp\u003e \u003cb\u003eGlocalization and Nationalism.\u003c/b\u003e The widespread adoption of these acronyms is further explained by their potent cultural logic, which balances global forms with local meanings. The coexistence of English-based abbreviations (e.g., LOL, BGM) and Pinyin-based acronyms (e.g., yyds, nsdd) exemplifies glocalization. English-derived terms facilitate participation in global digital culture, particularly in gaming and entertainment, signaling membership in a cosmopolitan generation. Conversely, Pinyin-based abbreviations represent a process of cultural translation. They appropriate the global format of alphabetization to express distinctly local meanings and sentiments. Critically, this \"global\" packaging of \"local\" content can, in some contexts, acquire nationalist connotations. When mobilized for collective action or defense of national icons, as has been observed with the use of 'yyds' to praise national achievements, these acronyms become linguistic markers of a distinctly Chinese digital identity. This strategic alignment with mainstream values creates a certain tolerance from official media, allowing a bottom-up linguistic form to be co-opted into a top-down narrative, thereby facilitating its unprecedented propagation scale.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThrough quantitative analysis, this study reveals the pivotal role of Bilibili as a core incubator of Chinese Internet slang, as well as the complex structural mechanisms within its community ecology that drive or inhibit the diffusion of linguistic innovations. Our findings present a diffusion model fundamentally different from traditional top\u0026ndash;down paradigms: a bottom\u0026ndash;up linguistic revolution initiated by grassroots communities, empowered by platform architecture, and ultimately reshaping the linguistic landscape of the Chinese Internet.\u003c/p\u003e\n\u003cp\u003eThe findings point to a clear dynamic model (see Figure 2). First, Bilibili acts as the engine of the entire diffusion process (\u003cem\u003e\u0026beta;\u003c/em\u003e = .743\u003csup\u003e***\u003c/sup\u003e), with its level of adoption directly determining a slang term\u0026rsquo;s nationwide influence. Second, information fragmentation emerges as the primary structural barrier to diffusion (\u003cem\u003e\u0026beta;\u003c/em\u003e = \u0026ndash;.450\u003csup\u003e***\u003c/sup\u003e). Meanwhile, gender innovation diversity (\u003cem\u003e\u0026beta;\u003c/em\u003e = .274\u003csup\u003e*\u003c/sup\u003e) and cross-community penetration breadth (\u003cem\u003e\u0026beta;\u003c/em\u003e = .371\u003csup\u003e**\u003c/sup\u003e) together form a dual engine that overcomes this barrier and drives diffusion. Diversity supplies the initial fuel and resonance necessary for innovation, while cross-community penetration transports the innovation across community boundaries, dismantling cultural silos. Ultimately, only slang that succeeds internally on Bilibili, by activating these dual drivers and overcoming fragmentation, can accumulate sufficient participatory energy to break through platform boundaries and trigger widespread nationwide diffusion. In effect, the outcome produced within Bilibili (Bilibili adoption intensity) determines the scale of diffusion at the national level, thereby enacting a bottom\u0026ndash;up \u0026ldquo;rebellion\u0026rdquo; against traditional language authority.\u003c/p\u003e\n\u003cp\u003e\u0026mdash;Figure 2 here--\u003c/p\u003e\n\u003cp\u003eFrom a theoretical perspective, this study integrates participatory culture, structural holes theory, and diffusion of innovations theory, offering a robust framework for understanding cultural production and circulation on decentralised platforms. Beyond validating these theories\u0026rsquo; applicability to online linguistic innovation, we advance their analytical precision by operationalising key variables such as gender innovation diversity, information fragmentation, and cross-community penetration breadth. From a practical perspective, the study offers important implications for platform governance, cultural research, and communication strategy: Meaningful engagement with online culture requires attention to the internal architecture of communities\u0026mdash;specifically, to the bridges that link otherwise disconnected groups and to the cultivation of diverse innovation environments\u0026mdash;rather than a narrow focus on content alone.\u003c/p\u003e\n\u003cp\u003eIn sum, this study demonstrates that linguistic innovation in the digital era is no longer governed by top-down norms, but emerges through bottom\u0026ndash;up dynamics. The case of Bilibili illustrates that when communities possess both diverse sources of innovation and effective connective structures, even the most marginal and fragmented \u0026lsquo;micro-tribes\u0026rsquo; can converge into a sweeping cultural tide, one capable of redefining how we communicate in a networked age.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eKe JIANG: Conceptualization, data curation, formal analysis, methodology and writing.Wei LI: Funding acquisition and supervision.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eRaw data were generated at Peking University. Derived data supporting the findings of this study are available from the corresponding author.\u003c/p\u003e\u003ch2\u003eFunding Details\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eThis study was funded by Peking University\u0026rsquo;s 2021 Language and Text Research Program: \u0026ldquo;Monitoring Language and Text in the Field of New Media\u0026rdquo; (NO. 7121300014). \u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eBailey CJN (1973)\u0026nbsp;Variation and linguistic theory. Center for Applied Linguistics.\u003c/li\u003e\n \u003cli\u003eBritain D (2002) Space and spatial diffusion. In J. K. Chambers, P. Trudgill, \u0026amp; N. Schilling-Estes (Eds.),\u0026nbsp;The handbook of language variation and change\u0026nbsp;. Blackwell Publishing, pp. 603\u0026ndash;637.\u003c/li\u003e\n \u003cli\u003eBritain D (2010a) Language and space: The variationist approach. In P. Auer \u0026amp; J. E. Schmidt (Eds.),\u0026nbsp;Language and space: An international handbook of linguistic variation, Vol. 1: Theories and methods\u0026nbsp;De Gruyter Mouton, pp. 142\u0026ndash;162.\u003c/li\u003e\n \u003cli\u003eChambers, JK, Trudgill P (1980)\u0026nbsp;Dialectology. Cambridge University Press, Cambridge.\u003c/li\u003e\n \u003cli\u003eTrudgill P (1974) Linguistic change and diffusion: Description and explanation in sociolinguistic dialect geography. Language in Society, 3(2), 215\u0026ndash;246. https://doi.org/10.1017/S0047404500004358\u003c/li\u003e\n \u003cli\u003eJenkins H (2006)\u0026nbsp;Convergence culture: Where old and new media collide. New York University Press, New York.\u003c/li\u003e\n \u003cli\u003eQuestMobile (2021) Bilibili user profile report 2021. QuestMobile Research Institute. Retrieved from https://www.questmobile.com.cn\u003c/li\u003e\n \u003cli\u003eZhang W, Mao C (2019) Fan activism sustained and challenged: Participatory culture in Chinese online translation communities. New Media \u0026amp; Society, 21(3), 748\u0026ndash;765. https://doi.org/10.1177/1461444818801511\u003c/li\u003e\n \u003cli\u003eVan Alstyne M, Brynjolfsson E (2005) Global village or cyber-Balkans? Modeling and measuring the integration of electronic communities.\u0026nbsp;Management Science,\u0026nbsp;*51*(6), 851\u0026ndash;868.\u003c/li\u003e\n \u003cli\u003eVan Alstyne M, Brynjolfsson E (1996) Could the Internet Balkanize science?\u0026nbsp;Science,\u0026nbsp;*274*(5292), 1479\u0026ndash;1480.\u003c/li\u003e\n \u003cli\u003eSunstein CR (2006)\u0026nbsp;Infotopia: How many minds produce knowledge. Oxford University Press, Oxford.\u003c/li\u003e\n \u003cli\u003ePariser E (2011)\u0026nbsp;The filter bubble: What the Internet is hiding from you. Penguin Press, New York.\u003c/li\u003e\n \u003cli\u003eChang WY (2008) The Cyber Balkanization and structural transformation of the public sphere in Korea.\u0026nbsp;Journal of Contemporary Eastern Asia,\u0026nbsp;*7*(2), 29\u0026ndash;48.\u003c/li\u003e\n \u003cli\u003eBurt RS (1992)\u0026nbsp;Structural holes: The social structure of competition. Harvard University Press, Cambridge.\u003c/li\u003e\n \u003cli\u003eBurt RS (1995)\u0026nbsp;Structural holes: The social structure of competition. Harvard University Press, Cambridge.\u003c/li\u003e\n \u003cli\u003eCentola D, Macy M (2007) Complex contagions and the weakness of long ties. American Journal of Sociology, 113(3), 702-734.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1. Variables.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"554\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable Type\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEnglish Name\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 79px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChinese Name\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 238px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eOperational Definition\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDependent Variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eNational Diffusion Scale\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 79px;\"\u003e\n \u003cp\u003e全网传播规模\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 238px;\"\u003e\n \u003cp\u003eTotal frequency of diffusion of the 66 keywords across major Chinese platforms\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDependent Variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eBilibili Adoption Intensity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 79px;\"\u003e\n \u003cp\u003eB站采纳强度\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 238px;\"\u003e\n \u003cp\u003eComposite index of six interaction indicators for relevant Bilibili videos (views, bullet comments, etc.)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eIndependent Variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eGender Innovation Diversity\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 79px;\"\u003e\n \u003cp\u003e性别创新多样性\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 238px;\"\u003e\n \u003cp\u003eSimpson diversity index based on the gender distribution (male/female/undisclosed) of content creators\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eIndependent Variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCross-Community Penetration Breadth\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 79px;\"\u003e\n \u003cp\u003e跨圈层渗透广度\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 238px;\"\u003e\n \u003cp\u003eNumber of Bilibili content categories (1\u0026ndash;8) in which each initialism appears\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 100px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eIndependent Variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 138px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eInformation Fragmentation Index\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 79px;\"\u003e\n \u003cp\u003e信息碎片化指数\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 238px;\"\u003e\n \u003cp\u003eSimpson diversity index based on the distribution of each term across eight Bilibili categories\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. Regression model summary and ANOVA for predictors of Bilibili adoption intensity.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"549\" class=\"fr-table-selection-hover\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003ePredictor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003eUnstandardized Coefficient (B)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003eStandardized Coefficient (\u0026beta;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 67px;\"\u003e\n \u003cp\u003e\u003cem\u003et\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003e(Constant)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e8.611\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 67px;\"\u003e\n \u003cp\u003e5.041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003eGender Innovation Diversity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e5.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e0.274\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 67px;\"\u003e\n \u003cp\u003e2.364\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003eInformation Fragmentation Index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e-4.295\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e-0.45\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 67px;\"\u003e\n \u003cp\u003e-4.127\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 180px;\"\u003e\n \u003cp\u003eCross-Community Penetration\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 126px;\"\u003e\n \u003cp\u003e0.566\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 110px;\"\u003e\n \u003cp\u003e0.371\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 67px;\"\u003e\n \u003cp\u003e3.022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote\u003c/em\u003e: Dependent variable: Bilibili adoption intensity. \u003cem\u003eR\u003c/em\u003e\u0026sup2; = 0.357, Adjusted \u003cem\u003eR\u003c/em\u003e\u0026sup2; = 0.326, \u003cem\u003eF\u003c/em\u003e = 11.458， \u003cem\u003ep\u003c/em\u003e \u0026lt; .001.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Linguistic innovation, online slang diffusion, community structure, Bilibili","lastPublishedDoi":"10.21203/rs.3.rs-8026908/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8026908/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn the context of China\u0026rsquo;s Internet, the generation and diffusion of online slang are undergoing profound changes as grassroots platforms are asserting an increasingly dominant role in linguistic innovation. This study uses Bilibili as an example to explore the underlying mechanisms through which these emerging centres drive the diffusion of linguistic innovation. Big data analysis is conducted on 66 representative initialisms (e.g., yyds and xswl) circulating in 2021 (source: Tencent). With diversity indexes (Simpson\u0026rsquo;s indexes) and regression models, this study reveals that (1) Bilibili internal adoption intensity can significantly predict national diffusion scale (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.743). (2) gender innovation diversity (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.274) and cross-community penetration breadth (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.371) are positive drivers of slang diffusion, while interest fragmentation (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;\u0026thinsp;.450) hinders diffusion. In summary, Bilibili\u0026rsquo;s community structure fosters two complementary dynamics: diversity that drives innovation and bridging mechanisms that break down silos, enabling effective bottom-up slang diffusion. The findings challenge the traditional linguistic authority long dominated by official media and provide empirical insight into the shifting distribution of cultural power in the digital age.\u003c/p\u003e","manuscriptTitle":"Rebel Lexicons: How Bilibili’s Fragmented Communities Engineer Slang Diffusion","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-18 09:38:54","doi":"10.21203/rs.3.rs-8026908/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"484e8ae5-0a22-4d42-ad86-6d34be9613e6","owner":[],"postedDate":"December 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":59649883,"name":"Social science/Development studies"},{"id":59649884,"name":"Scientific community and society/Geography"},{"id":59649885,"name":"Social science/Geography"},{"id":59649886,"name":"Business and commerce/Information systems and information technology"},{"id":59649887,"name":"Social science/Science technology and society"},{"id":59649888,"name":"Social science/Sociology"}],"tags":[],"updatedAt":"2026-01-09T07:09:22+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-18 09:38:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8026908","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8026908","identity":"rs-8026908","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}