Social Media Speech Police Need Clear Rules

By Cathy Young

Wednesday, August 17, 2016

Did Twitter censor hostile user comments during Q & A sessions with President Obama and television personality Caitlyn Jenner last year? A Buzzfeed report claiming that “hateful” and “abusive” tweets were removed both manually and with an automatic filter on secret orders from then-company CEO Dick Costolo has been flatly denied by Costolo himself and downplayed as containing “inaccuracies” in an official statement from Twitter. But the controversy has revived the ongoing debate about free speech and fairness in the social media. Conservative publications have accused Twitter of censorship and systematic left-wing bias. Meanwhile, Buzzfeed charges that the platform allows rampant “abuse and hate speech,” especially toward women and minorities, while safeguarding a few high-profile users.

Last month, the culture wars around harassment and free speech flared up when Breitbart News writer and right-wing firebrand Milo Yiannopoulos was permanently banned from Twitter for allegedly inciting racist harassment against comedienne and Ghostbusters star Leslie Jones, who is black. While many on the left praised the ban as a step toward curbing abuse, other commentators, conservative and liberal, saw it as a worrying move to muzzle unpopular speech.

Yiannopoulos is a habitual provocateur who almost certainly crossed the line in the attacks on Jones by tweeting fake screenshots of her supposed bigoted tweets, and Breitbart’s defense of him included some dubious assertions — for instance, that Jones herself incited harassment when she urged her followers to “get” a user who had sent her a string of racist insults. Yet some of the site’s claims of bias are hard to dispute.

Twitter did nothing last month when rapper Talib Kweli and his followers attacked black Breitbart reporter Jerome Hudson with racial slurs to mock him as lackey to white supremacy. When female Breitbart contributor Kassy Dillon complained about a user who sent her abusive and sexually harassing tweets and messages, Twitter took no action except to make him delete a tweet suggesting someone should “shoot [Dillon] in the head.”

Is this a pattern? In January, attorney and blogger Marc Randazza wrote that, after tracking actual Twitter users and his own “decoy” handles, he found conservatives were much more likely to be disciplined for offensive messages than “social justice” or feminist accounts. To be sure, this was not a rigorous study. But even some users who doubt deliberate left-wing favoritism by management agree that mostly liberal staffers are probably biased in evaluating complaints.

What’s more, social media’s anti-abuse initiatives are conducted in close collaboration with left-wing activists with a clear agenda. A model anti-harassment code for digital communities that has been used as a basis for policy at Google, Yahoo and Facebook explicitly favors “marginalized people” over the “privileged” and rejects complaints about “reverse” racism or sexism. A roundtable discussion on anti-harassment efforts published by Wired magazine last October, featuring several activists as well Twitter “trust and safety” vice president Del Harvey, reflected an assumption that such efforts should help advocates for progressive causes and prioritize abuse toward “women, people of color and LGBT people.” (In this mindset, conservative women, minorities and gays probably don’t count as “good victims.”)

The same slant is found in coverage of online abuse — such as the recent Buzzfeed article on Twitter harassment, which focuses mainly on female (and feminist) victims of misogynist trolls. Yet Internet harassment cuts across gender, racial, and political lines; sometimes, it targets conservative white men. Public shaming of people accused of racial or gender insensitivity can also turn abusive. Buzzfeed never mentions, for instance, that British comedian Stephen Fry was driven off Twitter by attacks for jokes some saw as offensive to women and transgender people.

The bias problem is compounded by vague rules and non-transparent enforcement. Thus, Twitter does not disclose what specific posts trigger sanctions. In Yiannopoulos’s case, it appears he was blamed for indirectly goading others into harassing Jones. Can people with a large following be accused of incitement simply for criticizing another user?

Private corporations such as Twitter or Facebook have every legal right to restrict speech on their platform. But if they see themselves as socially responsible, that responsibility should include promoting free expression and open debate — especially given their lack of effective competition. In a February interview, Federal Communications Commission member Ajit Pai warned that the “cultural values” undergirding the First Amendment are imperiled when social media abandon “the idea of the marketplace of ideas.”

Harassment and abuse can be bad for open discourse. Anti-abuse policies that fail at fairness and accountability are even worse.