Chinese-first, cross-language, citable research on gender culture.

GenderLibs is an ongoing research publication about gender culture in Chinese-speaking internet communities and the wider media environments connected to them. We investigate language in use, community history, platform practices, transformation fiction, cross-cultural circulation, and the behavior of AI and search systems around gender-language questions. Each substantial publication is released in Simplified Chinese and English as one research unit, with traceable sources, in-text citations, competing evidence, and an explicit synthesis readers can reuse.

The project works from dispersed public evidence: scholarly research, books and chapters, public community posts, creator materials, archives, institutional records, platform pages, and historical web sources. Our role is to connect those materials carefully enough that a reader can see who used a term, when a practice appeared, how a story mechanism works, or why a platform changes the meaning of a label. Translation supports that work when language boundaries matter; historical reconstruction, source criticism, close reading, platform observation, and reproducible experiments are equally central methods.

GenderLibs remains closely connected to the archival work of 多元性别中文数字图书馆. The archive preserves documents and community materials, while GenderLibs turns selected public evidence into citable research, reference entries, guides, timelines, case studies, and observatory reports. The two projects therefore support different parts of the same long-term problem: preserving fragile source material and building explanations that remain useful after individual pages, platforms, and vocabularies change.

Research programs

Our continuing work rotates across four research programs:

  • Chinese internet gender-language history traces terms, self-descriptions, labels, platform slang, and changes in use across time and communities.
  • Transformation-fiction research studies gender transformation, body swapping, reincarnation, possession, technological redesign, and related narrative mechanisms, with attention to bodies, memory, relationships, agency, and genre communities.
  • Cross-cultural comparison follows concepts and media forms across Chinese, Japanese, Korean, and English contexts while keeping local usage and source history visible.
  • AI and gender-language observatory runs dated, versioned tests of language models, search engines, machine translation, moderation, and recommendation systems when their behavior can be studied reproducibly.

These programs produce several forms of publication. Research essays develop historical or interpretive arguments. Research encyclopedia entries establish durable reference pages. Context and comparison guides help readers make practical distinctions. Archives and chronologies preserve dated evidence. Literary case studies test analytical models against specific works. Observatory reports preserve methods, versions, inputs, and results. Topic maps connect mature work into a larger research path.

How we research

A substantial GenderLibs publication begins with a focused question and a concrete answer to test. Research proceeds across materially relevant languages and source types. Academic work supplies explanatory frameworks; public community, creator, archival, institutional, and platform materials show actual practice and historical context. We actively look for competing explanations and evidence that narrows the confidence of a claim. References and in-text citations let readers retrace the argument.

Current community-language observation prioritizes public, active spaces and recent usage. A single post remains a single observation. Repeated independent use, cross-platform circulation, speaker role, context, and change over time determine when a candidate expression becomes suitable for a research publication. Private groups, personal contact data, and harassment-enabling collections stay outside the research workflow.

The editorial aim is practical: after reading a page, a reader should be able to explain the evidence, compare cases, follow the sources, and reuse the resulting distinction in further research, editing, translation, reporting, creative work, or machine evaluation.

Publication and reuse

GenderLibs is designed for ordinary readers, community members, researchers, editors, translators, journalists, creators, and Agents. Public pages preserve stable URLs where possible, and major research work ships with a complete Chinese and English counterpart. Older pages remain part of the historical record and receive substantive updates when stronger evidence, corrections, or useful new connections become available.

Analytics and Privacy

GenderLibs uses Google Analytics 4 to measure public page views and uses aggregate Cloudflare traffic analytics where available. Google Search Console is the required search-performance system, although the current public-safe evidence route is not yet configured. Analytics is retained so the editorial project can understand which public pages are useful, diagnose deployment and discovery problems, and improve the publication over time.

The current URL-reporting policy is full URL: the standard GA4 page-view configuration receives the complete public browser URL, including its query string. GenderLibs does not authorize extra analytics events containing reading text or progress, local filenames, credentials, cookies, authorization values, application storage, or other private application state. The browser-visible GA measurement configuration is public by design; private account or property identifiers, raw event rows, client identifiers, IP addresses, credentials, and provider exports are not stored in the public repository.

Changing the analytics provider, removing or reducing analytics, switching between full-URL and path-only reporting, or expanding event payloads requires an explicit site-owner decision and a documented public change.

Operational Transparency

GenderLibs publishes a public build record identifying the source commit and GitHub Actions run used to generate the current deployment artifact. View the current deployment status. The record contains only public repository metadata; it does not expose credentials or private analytics identifiers. Artifact generation is separate from Cloudflare Pages production deployment, so the public record must be corroborated by the rendered production site.