On Xing Tian and the Perseverance of Anti-China Sentiment Online

Xinyue Shen, Xinlei He, Michael Backes, Jeremy Blackburn, Savvas Zannettou, Yang Zhang

Abstract

Sinophobia, or anti-Chinese sentiment, has been a persistent element of web discourse for years, but the COVID-19 pandemic and subsequent global lockdowns significantly amplified its presence. This study provides a large-scale longitudinal measurement of Sinophobia from 2016 to 2021, analyzing 8 billion posts from Reddit and 206 million posts from 4chan’s /pol/. We investigate the origins, evolution, and thematic content of Sinophobic rhetoric. Our findings reveal that anti-Chinese content is often triggered by political events not directly related to China and that the COVID-19 pandemic caused a permanent semantic shift, where terms like "China" and "Chinese" became increasingly synonymous with racial slurs.

Introduction

While the surge in anti-Asian hate during the COVID-19 pandemic was widely documented by media and short-term academic studies, there has been a lack of long-term, quantitative understanding regarding the evolution of Sinophobia. This paper addresses this gap by examining five years of data across both mainstream (Reddit) and fringe (4chan) platforms. We explore the triggers for Sinophobic spikes, the semantic evolution of China-related terms, and the specific topics that dominate these toxic discourses.

Methodology

The researchers employed a multi-faceted computational approach to analyze the massive corpus:

  1. Temporal Analysis: Correlating spikes in keyword usage with real-world news events using peak detection.
  2. Slur Discovery: Utilizing word2vec models and vector arithmetic (e.g., "nigger" + "china") to identify community-specific derogatory terms like "chinkland" or "chicoms."
  3. Diachronic Word Embeddings: Training monthly word2vec models over 57 months to track how the "concept space" of words shifted over time.
  4. Toxicity Scoring: Quantifying "Severe Toxicity" and "Identity Attacks" using Google’s Perspective API.
  5. Topic Modeling: Applying Top2vec to cluster toxic posts into thematic categories.

Key Insights and Semantic Evolution

The 2020 Semantic Shift

One of the most significant findings is the "Great Semantic Shift" of early 2020. Before the pandemic, "China" was primarily used as a geographic or political descriptor. However, starting in January 2020, the word's linguistic "neighbors" in vector space shifted drastically toward slurs and pandemic-related insults. Crucially, this shift remained stable even after the initial shock of the pandemic subsided, suggesting a permanent change in how these online communities discuss China.

Political Reactivity

Sinophobia is often used as a reactive tool for broader geopolitical grievances. Spikes in anti-Chinese sentiment were observed during events like the U.S. withdrawal from the Paris Agreement and the 2021 Presidential Inauguration. In these contexts, China is often framed as a geopolitical antagonist or a scapegoat for domestic policy changes.

Platform-Specific Discourse

The "flavor" of Sinophobia varies significantly between platforms:

  • 4chan /pol/: Discourse is overtly obsessed with "ethnics," "genes," and "weapons."
  • Reddit: Sinophobia is more integrated into "daily life" topics. Users express sentiment through complaints about "Chinese cheaters" in video games, food safety, and financial stocks (e.g., Nio, Alibaba).

Quantitative Results

The data illustrates a clear intensification of Sinophobic rhetoric:

  • Volume: On 4chan, the percentage of daily posts mentioning "china" or "chinese" more than doubled post-COVID, rising from 0.50% to 1.23%.
  • Slur Proliferation: The term "chicoms" (Chinese Communists) saw a 5.2x increase on 4chan and a 40% increase on Reddit during the pandemic.
  • Toxicity: China-related posts are significantly more toxic than the platform baselines. On 4chan, 34% of such posts were flagged for "Severe Toxicity," compared to 11% on Reddit.
  • Persistence: The "interest period" for China-related topics—the duration a topic remains relevant after a peak—increased from 2.8 days pre-pandemic to over 8 days during the pandemic.

Moderation and "Leakage"

The study highlights the limitations of platform moderation. Despite Reddit’s 2020 ban on hate speech, Sinophobic slurs continued to proliferate. Users often bypassed filters through intentional misspellings or by adopting emerging terms that automated systems had not yet categorized as hateful.

Conclusion

This longitudinal study demonstrates that Sinophobia is not a static phenomenon but a dynamic and evolving form of online hate. The COVID-19 pandemic acted as a catalyst for a permanent linguistic shift, embedding toxic rhetoric into the standard vocabulary used to discuss China. Understanding these patterns is essential for developing more effective moderation strategies and for recognizing how fringe rhetoric can eventually permeate mainstream discourse.

Summary updated

Abstract

Sinophobia, anti-Chinese sentiment, has existed on the Web for a long time. The outbreak of COVID-19 and the extended quarantine has further amplified it. However, we lack a quantitative understanding of the cause of Sinophobia as well as how it evolves over time. In this paper, we conduct a large-scale longitudinal measurement of Sinophobia, between 2016 and 2021, on two mainstream and fringe Web communities. By analyzing 8B posts from Reddit and 206M posts from 4chan's /pol/, we investigate the origins, evolution, and content of Sinophobia. We find that, anti-Chinese content may be evoked by political events not directly related to China, e.g., the U.S. withdrawal from the Paris Agreement. And during the COVID-19 pandemic, daily usage of Sinophobic slurs has significantly increased even with the hate-speech ban policy. We also show that the semantic meaning of the words "China" and "Chinese" are shifting towards Sinophobic slurs with the rise of COVID-19 and remain the same in the pandemic period. We further use topic modeling to show the topics of Sinophobic discussion are pretty diverse and broad. We find that both Web communities share some common Sinophobic topics like ethnics, economics and commerce, weapons and military, foreign relations, etc. However, compared to 4chan's /pol/, more daily life-related topics including food, game, and stock are found in Reddit. Our finding also reveals that the topics related to COVID-19 and blaming the Chinese government are more prevalent in the pandemic period. To the best of our knowledge, this paper is the longest quantitative measurement of Sinophobia.
BibTeX
@inproceedings{shen2022xing,
  author = {Xinyue Shen and Xinlei He and Michael Backes and Jeremy Blackburn and Savvas Zannettou and Yang Zhang},
  title = {{On Xing Tian and the Perseverance of Anti-China Sentiment Online}},
  booktitle = {ICWSM},
  year = {2022},
  url = {https://openreview.net/forum?id=IcMHeGBG0w},
}
Figure