Misogyny and Twitter: Why that study you saw was bad

Misogyny and Twitter: Why that study you saw was bad

Online misogyny is neither a new or shocking problem, but it's something women have been publicly talking about more frequently. Check the replies on anything tweeted by women with over a few hundred followers; there's a good chance you'll see threats and slurs tossed about as though no other form of conversation could possibly exist. That's not to say women with smaller accounts will be spared. I've seen women with barely 100 followers receive 30 replies to a single tweet expressing the most inoffensive feminist ideology. Ghostbusters is also a huge attractor for serial abusers. Simply mentioning anything positive about the reboot can lead to a flood of angry men in a woman's mentions.

Not a new or shocking problem, but certainly one lacking much attention from Twitter's staff. There's also a lack of research into the shape, causes, and reactions to this misogyny.

Demos, a UK-based think-tank whose About page mentions "ground breaking social media analysis,"attempted to fill this research hole with a study released May 26th. Although, it should be said that 'released' is being used somewhat liberally here; one blogger tried to dig deeper into the results, but was told there was no written report for public analysis. What does that mean? Well, it means we have explanations of only some of the methodology used in this study–with no way of verification.

First, let's look at what Demos says they found.

Over one million tweets containing the words 'slut' or 'whore' were analyzed to make this report. Fifty-four percent were thrown out due to coming from accounts which were advertising pornography. The remaining 650,000 tweets were divided up into aggressive, self-identifying, and other; aggressive tweets were those which targeted people by using 'you', included other slurs, or included commands such as 'shut up'; self-identifying tweets were people referring to themselves as a 'slut' or 'whore'; other contained any tweets that didn't fit into these two categories.

Thirty-three percent of the 650,000 tweets fell into the aggressive category, or 213,000 tweets over a 23 day period. Demos states that 50% of these tweets were sent by women, 40% by men, and 10% by accounts whose gender could not be determined or organizations. These findings are used to state that misogyny is clearly internalized by a lot of women, showing the pervasiveness of misogyny in society as a whole. Further, Demos feels misogynistic content should not be censored, but instead challenged and the purveyors educated. At the end of their release, they state:

By allowing this freedom of speech to be exercised, misogyny can be openly confronted and discussed instead of being swept under the carpet of censorship.

What's the problem?

The selection of only tweets containing 'slut' or 'whore' doesn't really reflect the language of misogyny in online communications. Yes, those two words will be used to attack or shame a woman, but misogynistic attacks go beyond just two keywords. Women receive rape threats, death threats, and other violent tweets that might not even contain a slur. Sometimes, these attacks will be incredibly vivid in ways that wouldn't trigger any kind of word filter. Studying only tweets containing these two words also leaves out a lot of misogyny targeted transgender individuals.

Demos gives no explanation for why these two words were chosen. It can be assumed they were chosen due to Demos' previous study of this nature using those two words, which still doesn't give us much information.

Another issue concerns the findings of which gender was sending misogynistic tweets. Demos does not explain how they determined the gender of a user–important information that shouldn't be left out of a methodology declaration. Profile pictures have no requirements beyond basic obscenity rules, meaning that anyone can use a profile picture of anyone or anything. Few people list their preferred pronouns in the biography section of their account. Without any disclosure of how gender was determined, we really have no way of trusting any findings about who was sending these tweets.

Further, the study repeatedly refers to misogynistic abuse as 'trolling'. This should go without saying, but trolling is a deflective term which negates the effects abusive communication can have on the recipient. Nearly every discussion of the most toxic social media platforms will undoubtedly have someone declare it's only trolling, and is just how the internet operates. But repeated, unwavering abuse with a clear intention to drive a target into distress is vastly different from the mostly harmless act of trolling, where there may not be an individual target and the purpose is meant to be humorous.

That's not to say trolling is always humorous or even something to be encouraged; only that there's a difference between a group of users spreading a fake image claiming you can dry your iPhone in a microwave, and several people–organized or independent–directly attacking a woman with violent language.

Finally, as mentioned above, the biggest problem of the study itself is the lack of a written report. Nobody is able to confirm or confront the findings based on actual data as no data is publicly available. The only methodology provided by Demos regards the Natural Language Processing Algorithms they used and how accurately the algorithms sorted tweets into the three categories used by Demos.

A positive reception

After the study was released by Demos, several news outlets reported on it. Some news outlets, such as the BBC and Metro, ran headlines focusing on the findings that 50% of misogynistic tweets came from women. Neither outlet included Demos' warning that "this study cannot claim to be exhaustive, for better or for worse." The only major outlet I've found discussing the problems of the study was Fusion.

BBC's article shows in a Google News search as being highly cited. This is not surprising, considering how quickly certain groups associated with long-term misogynistic abuse took pride in being told how women are worse than men at abusing women. The article was celebrated by GamerGate's forum on Reddit, alongside several other subreddits dedicated to distributing misogyny or decrying feminism. The only subreddit geared towards women to discuss it was TwoXChromosomes, which has become a place for men to tell women what they think women should care about ever since it became one of the forums every new account is automatically subscribed to. Twitter measured up in the same way.

Was there any value?

As the study says, there is indeed a problem with women attacking other women using language similar to that of the most misogynistic men. You certainly won't fail to find women who accuse Amber Heard of lying about being assaulted by Johnny Depp, nor is there a lack of women who will call other women a 'slut' or 'whore' because of how they present themselves online.

This is often referred to as internalized misogyny, and has been known since probably the dawn of feminism. For some women, there may be a point in which they recognize that they are limiting themselves due to accepting the expectations traditionally given to them by men. Others may live their entire life fulfilling only the roles society has constructed for them.

And the sheer number of tweets containing 'slut' or 'whore' being directed at others in this study does illustrate just how commonplace these slurs have become. Adding in other words and phrases commonly directed at women would only further show just how unwelcoming online communities and platforms can be when you're a woman.

Demos says we need better education to fight misogyny, and that's certainly true. Misogyny isn't a disease to be contracted or a sudden revelation in which you decide to start hating women; it's a learned trait that starts with your environment at an early age. Better education means we may get to an era where all genders can agree on the definition of consent, the harm of discrimination, and why women do not owe men their time and attention.

Where Demos doesn't seem to understand abuse–beyond all the other problems of the study–is in stating that policies disallowing misogyny are a harmful form of censorship. This belief implies that abusers deserve to be heard, that every woman being abused has a duty to educate her abuser and answer to them. But that's just not realistic. When a woman is targeted, she may face dozens to thousands of abusers seeking a reaction. And, obviously, nobody should be expected to address their abusers, regardless of the importance awarded to 'free speech'.

If policies disallowing misogynistic abuse were instituted and enforced when the social media giants first started up, we may not be in a situation where communities founded on abuse celebrate a flawed study before engaging in more misogyny. We also might not have a situation where the only way to make a platform usable is to utilize a service which automatically blocks thousands of accounts.

Please help support Amala Network with a small donation

Follow us on Twitter @amalanet.