Facebook 新算法：恶语也分三六九等

一、改算法，辨恶言！

根据《华盛顿邮报》获得的内部文件，Facebook正在改变算法来处理仇恨言论，它将加大力度监管对弱势群体恶意中伤的言语，而不再插手管理那些“针对白人、男性与美国人的批评”，扭转从前“无视种族问题”的审查机制。此前，Facebook很注重在平台上标记、删除有色人种发布的无恶意帖子，同时积极清除攻击白人的话语。

Facebook is changing the way its algorithms handle hate speech. It will spend more time policing actually vile content about underrepresented groups, and less time on comments “against whites, men, and Americans,” according to a series of internal documents obtained by The Washington Post, reversing what the social media platform’s “race-blind” practices. Those practices resulted in the company being more vigilant about removing slurs lobbed against White users while flagging and deleting innocuous posts by people of color on the platform.

文件显示，这个叫做“WoW”（“Worst of the Worst”）的整改项目仍处于早期阶段，它将重新设计Facebook的自控系统，以便更有效地检测和删除“最具危害性”的仇恨语言。针对黑人、穆斯林、多种族群体、LGBTQ人群和犹太人的恶意攻击都在此类言论的范围之内。（LGBTQ：对女同性恋者、男同性恋者、双性恋者和跨性别者等少数群体的总称）

The overhaul, which is known as the WoW Project and is in its early stages, involves re-engineering Facebook’s automated moderation systems to get better at detecting and automatically deleting hateful language that is considered “the worst of the worst,” according to the documents. The “worst of the worst” includes slurs directed at Blacks, Muslims, people of more than one race, the LGBTQ community and Jews, according to the documents.

Facebook根据各类攻击内容的有害程度，设置了不同的分数来评估其严重性。例如，相较于“男人是猪”，系统将优先选择删除“同性恋令人恶心”。该平台仍把轻蔑男性、白人和美国人的评论当成仇恨言论，用户也仍然可以向公司举报此类言论，但算法已将它们视为“低敏感”内容，或者说“有害性较低”的内容，因此它们不会再被自动删除。

As one way to assess severity, Facebook assigned different types of attacks numerical scores weighted based on their perceived harm. For example, the company’s systems would now place a higher priority on automatically removing statements such as “Gay people are disgusting” than “Men are pigs.” Facebook still considers such attacks to be hate speech, and users can still report it to the company. However, the company’s technology now treats them as “low-sensitivity” — or less likely to be harmful — so that they are no longer automatically deleted by the company’s algorithms.

二、“难以言说”的话语窘境

仇恨言论是一种诋毁种族、性别、性取向等个人特征的残暴、不人道言论，Facebook与旗下的INS早就设置了针对仇恨言论的禁令。但在整改前，公司的算法和制度并没有对不同群体加以区分，系统无法判断哪些群体易受仇恨言论的攻击，哪些群体不属于长期边缘化人群。像“白人愚蠢”这样的评论甚至会与反犹太主义和种族歧视并为一谈。(历史上被边缘化的群体(HMG)——妇女、少数种族、少数民族、性少数者以及低收入个人)

Facebook has long banned hate speech — defined as violent or dehumanizing speech — based on race, gender, sexuality and other protected characteristics. It owns Instagram and has the same hate speech policies there. But before the overhaul, the company’s algorithms and policies did not make a distinction between groups that were more likely to be targets of hate speech versus those that have not been historically marginalized. Comments like “White people are stupid” were treated the same as anti-Semitic or racist slurs.

多年来，民权倡导者一直在抨击Facebook大量清除黑人用户帖子的现象，这种情况在黑人描述歧视经历时犹为明显，此次算法修改也是对这一批评的回应。

The shift is a response to years of criticism from civil rights advocates that content from Black users is disproportionately removed, particularly when they use the platform to describe experiences of discrimination.

由于描述歧视经历可能会涉及对白人的批评，Facebook的算法往往会自动删除这些内容，即使是先进的人工智能也会在处理这种微妙话题时“过于热心”。

Because describing experiences of discrimination can involve critiquing White people, Facebook’s algorithms often automatically removed that content, demonstrating the ways in which even advanced artificial intelligence can be overzealous in tackling nuanced topics.

黑人用户表示，Facebook声称要保护少数群体，但在由白人建立并主导的公司里，这种仇恨言论政策和内容审核系统却很难让他们满意。但他们别无选择, 只能继续使用Facebook和INS，这两大社交媒体已经成为了黑人团体的救生索。

Black users say hate speech policies and content moderation systems formulated by a company built by and dominated by white men fail the very people Facebook claims it's trying to protect. But they have no choice but to stick with Facebook and Instagram, both critical lifelines to their communities. 

借助社交媒体的舆论运动，警察杀害美国黑人的问题以及就业、健康和其他领域的种族差异都被提上了公共议程。

Movements on social media have helped put the deaths of Black Americans by police officers on the public agenda, along with racial disparities in employment, health and other areas, they say.

黑人用户还说，在谈论种族主义时为了躲避禁令，他们会使用白人头像、白云emoji以及 “wypipo”(音似white people )和“huite”(音似white )等俚语，在语音和视觉上代替提及“白人”；他们用假名发帖、设置备份帐户来联系其他黑人和防止信息丢失；他们还开发了一个好友系统，以便有人受到Facebook的制裁时及时提醒朋友和追随者，同时分享处罚的具体情况和那些所谓“违背规定”的帖子内容。

To elude bans when talking about racism, Black users said they resort to using a white avatar, slang such as "wypipo" or "huite" or a cloud emoji as phonetic or visual stand-ins for white people. They operate under aliases and maintain backup accounts to avoid losing content and access to their community. They've developed a buddy system to alert friends and followers when someone has been sent to Facebook jail, sharing the news of the suspension and the posts that put them there.

三、言论新出路在哪里？

一名活动人士正积极推动社交媒体转变规则，呼吁它们在管制言论时考虑不同群体面临的历史遗存压迫。据他透露，民权活动者一直在向Facebook施压，要求它以更进取的姿态处理所谓的“仇恨言论”,同时建立一种“兼顾权力与历史动态的管制体系”。

Civil rights activists have been pressuring Facebook to adopt a more aggressive stance toward so-called “hate speech” on its platform, demanding an “enforcement regime that takes power and historical dynamics into account,” according to one activist involved in pushing the social media network to take historical oppression into account when moderating speech.

一位Facebook发言人对《华盛顿邮报》说，“我们知道针对弱势群体的仇恨言论可能才是最有害的，这就是为什么公司已在集中技术筛查用户和专家所指定的最恶劣言论。在过去的一年里,平台变更了管理措施以捕捉更多隐晦的仇视话语，描述有关“黑脸”的内容，或者表露出犹太人“控制世界”这一刻板印象都在平台限制范围之内，同时我们还禁止用户发表否认大屠杀的观点。”（黑脸：历史上白人演员把脸涂黑扮成黑人表演，显示了对黑人的刻板印象以及种族歧视）

“We know that hate speech targeted towards underrepresented groups can be the most harmful, which is why we have focused our technology on finding the hate speech that users and experts tell us is the most serious,” a Facebook spokesperson told the Post. “Over the past year, we’ve also updated our policies to catch more implicit hate speech, such as content depicting Blackface, stereotypes about Jewish people controlling the world, and banned Holocaust denial.”

一位进步“言论自由专家”告诉《华盛顿邮报》，她赞同Facebook提出的新举措，即使它最终起到审查其他争议性内容的作用，但该举措保留了平台上批评白人的言论，有助于解决“系统性的种族主义”和“结构性白人至上”的问题。（系统性种族主义：指白种人优先的观念上升到了社会系统层面，融入社会治理的方方面面。）

One progressive “free speech expert” told the Post she agrees with Facebook’s proposed policy because it helps in addressing “systemic racism” and structural white supremacy specifically because it allows criticism of white men to remain on the platform, even if it ultimately has the effect of censoring other controversial content.

波士顿法学院教授丹妮尔·西特龙在接受《华尔街日报》采访时说：“不能谈论系统性种族主义就无法对抗它，而批评白人和白人至上理念是探讨种族主义的一项重要内容。如果生硬的算法直接过滤了这些言论，这种对话就无法展开。”

“We can’t combat systemic racism if we can’t talk about it, and challenging white supremacy and White men is an important part of having dialogue about racism,” Danielle Citron, a Boston Law School professor told the outlet. “But you can’t have the conversation if it is being filtered out, bizarrely, by overly blunt hate speech algorithms.”