The rationale seems to be that “Ebil CCP wants to censor anti-gov’t protestors while at the same time, homogenizing their national identity”

If anyone close to China can confirm or debunk it, it’d be good

On the other hand, I’m skeptical of this idea

I’m pretty sure Cantonese is still living as the lingua franca of 85 million people out of 1,500 million Chinese people worldwide.), which makes it over 5% of the population’s native tongue able to speak it, mostly in the region itself, though I suppose 60 million of them might be outside China

Meanwhile, languages like Occitan in France and Celtic languages of Britain and Ireland are little to match like that, let alone fucking United States of America, with its Native American languages

In fact here’s an example of this sort of shit…

I still wonder why…

  • qwename
    link
    fedilink
    arrow-up
    5
    ·
    edit-2
    11 months ago

    The article from 2022 shown at the beginning of the video: https://fortune.com/2022/10/06/douyin-bytedance-cantonese-livestreamers-cut-off-china-moderation/.

    It links to two SCMP articles:

    From what I can find, this is limited to the platform Douyin, and probably includes other Chinese dialects. The speculation as to why this happens is that content that is not in Putonghua (Mandarin) is harder to review.

    All content posted on social media in China has to be regulated by the companies who provide such platforms, and live streams are harder to regulate as they occur in real-time. There would need to be a reasonable number of content reviewers who understand Cantonese or other dialects, and/or better speech recognition tools that can detect those dialects reliably to lessen the load on manual review.

    • deathtoredditOP
      link
      fedilink
      arrow-up
      2
      ·
      11 months ago

      All content posted on social media in China has to be regulated by the companies who provide such platforms, and live streams are harder to regulate as they occur in real-time. There would need to be a reasonable number of content reviewers who understand Cantonese or other dialects, and/or better speech recognition tools that can detect those dialects reliably to lessen the load on manual review.

      The thing is, the captions did manage capture what the Canto-speaking person was saying, at least according to the narrator of the video…