Many women and trans* folks who use the online social network Twitter and those who publicly identify as feminists have had to bear the brunt of misogynistic vitriol online. While it seems hateful and at times violent messages directed at other people online (what’s known as trolling) has increasingly been part of the Internet landscape, so too have the gendered forms of harassment and violence on Twitter.
As a response to this gendered online violence, and to help deal with misogynist trolls, artists and activists from Peng! Collective created Zero Trollerance, an online campaign that, over the course of six days back in April 2015, used 160 bots to tweet out six patronizing and funny “self-help” steps designed to help the misogynist trolls change their sexist online harassing ways.
The Young Feminist Wire spoke with Ada Stolz* from Peng! Collective about the Zero Trollerance campaign.
YFWire: What is Peng! Collective?
Ada Stolz: We are a group of activists, artists and people who care about politics, based in Berlin. We got tired of the same old tactics in campaigning and are really interested in the ways that you can create attention and stir up debate around certain issues by using creative tactics. We use hoaxsters, media hacks, culture jamming and guerilla communications. These techniques have always been on the fringe of art and activism, but are not used by traditional NGO campaigners because they are seen as too edgy, a bit radical, and sometimes border on illegal. That’s the area we work in. We think that this kind of expression can be really useful for reaching new audiences, generating new kinds of support for an issue that you maybe wouldn’t get through your standard online petitions and Facebook pages. We look for tactics that are online and offline. We do many offline actions as well, in Yes Men style, impersonating company representatives and saying that the company will do something that the company would never do.
YFWire: Where did the idea for Zero Trollerance come from?
AS: The idea for Zero Trollerance came out of two meetings I went to last year. The issue of online harassment against women and trans* people just kept coming up. This was riding on the back of Gamer Gate and Anita Sarkissian’s story. Up until these meetings, I hadn’t really realized how widespread the problem really was. Not only was this happening to these really, really high profile western people, [harassment] was happening to your average Indian or Pakistani or South African activist or blogger, any woman or trans* person who is raising attention to their work. If they have a lot of influence, if they are out there on Twitter or generating a following, chances are that they are going to get harassed and attacked.
Through some discussions, I saw that there was a need to create a response that was something other than just talking about the problem. It has obviously been important for the feminist movement in general to document cases of abuse and reveal the gravity of the situation and express that to society through newspaper articles and mapping projects. While that’s good, I also thought that we needed something with a bit more of a fight in it, and that plays the perpetrators game. That’s what I was trying to do with this project. I had a discussion with a few people about the idea of creating bots to target or at least speak back to the trolls.
YFWire: Take us through the process of creating Zero Trollerance. What had to happen to make the Zero Trollerance campaign a reality?
AS: At Peng!, for the longest time, we had this idea of building up a kind of Twitter army. There would be a number of Twitter profiles that we would have control over, and that would all follow each other and could do something, but we didn’t know what yet. So the idea was already there and we discussed it as a group and talked a lot about what to do about trolls and how we could respond to them on Twitter. We came back to the idea of the [anti-sexist] Twitter robot army [and paired it with] the idea of self-help programs because I’m really into satire and parody and I find the style of self-help really funny and patronizing and that’s the response I wanted to go for – to patronize the trolls. I feel that you can’t really reason with them, or that’s what I learned by talking to people who have been harassed. There’s this idea that you shouldn’t feed the trolls, but ignoring them and trying to reason with them doesn’t make sense either. So I thought, ‘What can we do? Well, maybe we can make fun of them.’ So, that’s the patronizing idea of the self-help program. We came up with this idea that there would be troll coaches online and every day [of the campaign], they would message the trolls and give them inspirational messages. To us, the idea that a troll could be harassed back by people who want to help them to get better was a really beautiful idea.
I worked with the hacker Jenny Mainframe who is not in Peng! Collective but really wanted to help with this project. She helped me design the script for the bots. Then, we collaborated with a filmmaker who is also in our network and found an actor [to film the self-help steps].
YFWire: How do the anti-sexist Twitter bots work?
AS: There are a few layers to the bot function. The first is to compile a list of trolls or a database of people that we were going to enroll in the program. Because, the idea is that you would get a message when enrolled in the program on the first day and then you get a message every day, for six days. So, to create that database, part of the bot script was just to scan the Twitter API [application program interface] for a list of key phrases. If you only search for the word bitch, for example, you get lots of false-positives — so people who used the word in a playful way, not actually to harass people. We did a lot of testing to find which phrases and terms would actually return language that trolls use. The script was essentially scanning Twitter for users who were using the phrases and keywords that we had come up with and it was compiling a database. One Twitter profile was doing that and then, the other 159 profiles would work systematically from the database to message those people in the database. And, the bots were programmed in such a way that they tag-teamed. So one bot would only tweet up to 50 messages in a day and then it would ‘sleep’ and come back the next day. We had to do a lot of testing and plan them around Twitter’s spam filters.
YFWire: What was the response to the Zero Trollerance campaign?
AS: There was a lot of positive response from the feminist community and generally from people who are online and politically active. There was a lot of positivity coming from those people because the people who are working on or are aware of these issues saw it as a useful and powerful response.
Then, from trolls themselves, the response was that we got trolled back. We had a lot of violence images sent to us, and also we got some baiting from trolls who wanted to engage with us. I maintained the Zero Trollerance Twitter profile as if I was a coach and was very patient and understanding, but at the same time really patronizing to them. We had a lot of engagement with trolls who would try to justify their behavior as their right to freedom of expression and they were calling us feminist Nazis, who are trying to force our ideologies on other people and trying to control the space on Twitter.
More generally, we got a lot of media attention for it because it is quite a topical issue and also because Twitter is starting to employ technical solutions to the problem of harassment.
YF Wire: What were some of the lessons learned from the campaign?
AS: First that language is a really difficult concept and there are no technical solutions to this problem. We saw that Twitter was trying to roll out a technical solution but our reaction is that there are no technical solutions to misogyny and sexism online. But then we also in some ways are trying to come up with technical solutions with the bots. In running the campaign, we reached a lot of trolls through the data scrapping and the bots, but there are still lots of false positives in that because you can never program something to find all the trolls on Twitter because there are no recipes for the kind of language they will use. Yes, you can search for rape threats and you can search for death threats and these kinds of things but at the end of the day, the human element of [the problem of trolling] is the most important.
While I think [the Twitter bots were] a really innovative tool to get attention, and that the surprise element of messaging trolls [who use misogynistic language] was on point, we still need the feminist organizing happening around immediate responses on Twitter and those who handle trolls so that the person who is getting trolled can find support. This kind of organizing is really important. As much as I like the bots and I think they are a beautiful idea artistically, they will not solve the issue of systemic online violence against women and trans* people.
YF Wire: Fighting back against trolls can be really time-consuming and draining, but why is it important to fight back?
AS: One of the reasons for creating this project is that I like the idea that if someone is getting trolled, they can send a link of the project to the person who is trolling them. It’s not going to stop the troll necessarily but there is something empowering in knowing that there is a resource out there that one can send to a troll. We need an online feminist response that is visible for trolls to see that we can do something and fight back in a way that’s not just making it seem like we are the victims of the situation.
YF Wire: On the Zero Trollerance site, there are mock quotes like “Thanks Zero Trollerance, I learned how to tweet like a feminist” from “reformed trolls”, did you encounter any reformed trolls?
AS: No, I wouldn’t say there were any reformed trolls during the campaign but I did have one interesting engagement [on Twitter] with someone who wouldn’t term themselves a troll but who was tweeting a lot around the hashtag #StopWebHate. It was the hashtag that the U.S. Congress was using while they were discussing issues of violence online, and it got taken over by the Gamer Gate trolls. He engaged with me and said he doesn’t really think he’s a troll and that he actually makes quite an effort when people are getting targeted and when he thinks it is wrong, to stand up and say something out loud, online. I think that there are a lot of these kinds of people, especially in the techie/gamer communities, who don’t troll but are the silent bystanders, who say they don’t get involved in violent speech, but those are the people who can be empowered to say something, to step up and denounce [the violent behaviour they encounter online]. I wanted this campaign to speak to those people too.
YF Wire: What happens with the Zero Trollerance site now that the campaign is over?
AS: Definitely, it will remain online as a resource. We have a pre-written tweet that says “Having problems with trolls? Just retweet this message” with a link to the website that people can keep using. Even though we can’t keep running the bots, humans can play the role of the bots by being the ones that are monitoring Twitter and responding to trolls with the Zero Trollerance website.
Also, the code for the bots is open source and we will be giving the code to any groups who want to run a similar action, they can mail us on firstname.lastname@example.org, so it can be used again if some other feminist group wants to run a similar campaign and in other languages, they can just change the keywords that they are searching for on Twitter and they can change the messages and share the self-help videos. They can replicate it and make it their own.
*Ada Stolz is the Peng! Collective member’s artist name.
 A bot is a software application that runs automated tasks over the Internet. Typically, bots perform tasks that are both simple and repetitive, at a much higher rate than would be possible for a human alone.