Yesterday morning as is my custom, I grabbed a cup of coffee, fed my dogs and retreated to my porch to review my newsfeeds on Twitter. One of the first things I saw was that Rose McGowan’s Twitter account had been locked in the midst the Harvey Weinstein scandal. When I first noted it was still dark. By lunchtime, feminist groups were calling for boycotts of Twitter over Rose’s account being locked and after about 100 articles on the interruption to Rose’s account, Twitter finally said why they did it. Apparently, Rose had posted a private phone number, which is and has been a term of service violation since I joined the platform five years ago.
Because of stature and scrutiny, Rose was made aware of the specific thing that resulted in being locked out and Twitter unlocked her account. Quite frankly posting someone else’s private phone number or address without permission should have consequences for the person who does it on any public platform. Do I think putting Rose in a timeout for 12 hours like a toddler who misbehaved is necessary? No. A simple “delete this tweet because…….to begin tweeting again” would seem to be a rational way to deal with adults.
Rose is lucky. Most Twitter users never find out the reason their specific tweet is in violation of Twitter rules when asked to delete. Worse than that, many are never told why their accounts are permanently suspended by the platform after seemingly normal interactions. Here are two examples.
A user who went by the name @orneryyg had his 2100 follower account suspended. He was a regular Twitter user who blogged for Misfit Politics, Halsey News, and The New Americana and would engage in what were sometimes heated Twitter political debates. He admits sometimes they became contentious and he might have used profanity from time to time, but nothing out of the ordinary from what he has seen on the platform. The account had been locked a few times for what he assumes was language, but because Twitter doesn’t offer an explanation, he is not sure. He has appealed his original suspension over a dozen times offering to delete any required tweets and still has no idea why his account can not be returned.
He does know that several accounts have claimed responsibility for getting his original account and subsequent accounts he created suspended by using coordinated reporting mass reporting. The same users posted screenshots celebrating getting the account of popular Twitchy editor Sam Janney (@Politibunny) being locked out. Because Twitter lacks transparency in the administration of it’s TOS many users believe that coordinated reporting from a number of accounts or using script designed for Twitter can affect a suspension or lockout without a significant TOS violation.
Rick Canton who tweeted under the handle @RickCanton beginning in 2012 was a very popular Conservative account with over 70,000 followers. As a political activist in Virginia and new blogger, he used his Twitter account in conjunction with his very popular Facebook account to promote his writing and his causes. He tweeted a tweet critical of the Black Lives Matter narrative. He used the word “retarded”. He admits it and while I may have used a different word it is surely not shocking for Twitter. Another user responded admonishing him for using the word and he pointed out a case where she had used it a tweet herself. When Rick refused to capitulate to her criticism she blocked him, posted a screen cap of the tweet she found offensive and reported Rick for harassment. Yup. She responded to his tweet, but he was harassing her? Rick has appealed over 20 times for the return of his account has only been told he engaged in “targeted harassment” in an auto-reply.
Like @orneryyg, the interaction for which he was permanently suspended was not even notable or unusual until he was unable to log into his account. Both users believe they were suspended for what is commonly called “Tweeting while Conservative”. I believe based on my own experience and observations it is an utter failure in the Twitter algorithm combined with the fact that Twitter Support does not do a personal review of most reports or appeals. They simply can’t. If numbered sequentially, one report I received was nearing 7 million cases.
I have referred to Twitter’s new guidelines as “The 500 Dirty Words” (H/T to George Carlin) that they simply refuse to articulate. Combined with an algorithm that seems to respond to mass reporting, and a seeming 3 strikes and you’re out rule, the perception of political reasons for why some users are locked or suspended more often will persist. When Conservatives see garbage like this flow through our mentions from verified Liberals with no consequences it only reinforces the perception. These tweets are still live.
Maybe I’m simple, but rather than create elaborate algorithms that create a perception of political favoritism in a country already divided, perhaps the Technocrats at Twitter could consider the following:
- While the majority of users endeavor to be respectful even if they are snarky, if there really are words you don’t want us to use on your platform, tell your user base what they are and apply your standard consistently.
- Political debate can get contentious. If emotions run high and someone tweets in anger or makes a prohibited statement that you have clearly said is not allowed, keep it simple. Say “Delete this tweet because (insert reason here) to resume tweeting.
- Stop using the algorithm to infer “targeted harassment”. A few users on the Conservative side have told me they are starting to figure it out. If someone tweets them multiple insulting tweets, they report them for targeted harassment. And they have observed the accounts be affected. This doesn’t need to be a game of three-dimensional chess where everyone is trying to figure out how to negatively impact another user using your reporting function.
- Focus on actual threats of physical harm, posting of personal information, revenge porn and contacting people’s employers. The worst Twitter is the Twitter that stomps into your personal life and tries to ruin you or do you harm. These are far less common and may be able to be reviewed by an actual person.
- Treat your users like adults. I have spent years using both my “Block” and “Mute” buttons with great success. I have only made reports in the most egregious circumstances.
- If scripts and bots are used on your service to harass users, find them, eliminate them and prevent it from happening again.