Twitter abuse: one click to save us all?

A great deal has been said already about the twitter abuse issue – and I suspect a great deal more will be said, because this really is an important issue. The level and nature of the abuse that some people have been receiving – not just now, but pretty much as long as Twitter has existed – has been hideous. Anyone who suggests otherwise, or who suggests that those receiving the abuse, the threats, should get ‘thicker skins’, or shrug it off, is, in my opinion, very much missing the point. I’m lucky enough never to have been a victim of this sort of thing – but as a straight, white, able-bodied man I’m not one of the likely targets of the kind of people that generally perpetrate such abuse. It’s easy, from such a position, to tell others that they should rise above it. Easy, but totally unfair.

The effect of this kind of abuse, this kind of attack, is to stifle speech: to chill speech. That isn’t just bad for Twitter, it’s bad for all of us. There are very good reasons that ‘free expression’ is considered one of the most fundamental of human rights, included in every human rights declaration and pretty much every democratic country’s constitution. It’s crucial for holding the powerful to account – whether they be governments , companies or just powerful individuals.

Free speech, however, does need protection, moderation, if it is to avoid becoming just a shouting match, won by those with the loudest voice and the most powerful friends – so everywhere, even in the US, there are laws and regulations that make some kinds of speech unacceptable. How much speech is unacceptable varies from place to place – direct threats are unacceptable pretty much everywhere, for example, but racism, bullying, ‘hate speech’ and so forth have laws against them in some places, not in others.

In the UK, we have a whole raft of laws – some might say too many – and from what I have seen, a great deal of the kind of abuse that Caroline Criado-Perez, Stella Creasy, Mary Beard and many more have received recently falls foul of those laws. Those laws are likely to be enforced on a few examples – there has already been at least one arrest – but how can you enforce laws like this on thousands of seemingly anonymous online attackers? And should Twitter themselves be taken to task, and asked to do more about this?

That’s the big question, and lots of people have been coming up with ‘solutions’. The trouble with those solutions is that they, in themselves, are likely to have their own chilling effect – and perhaps even more significant consequences.

The Report Abuse Button?

The idea of a ‘report abuse’ button seems to be the most popular – indeed, Twitter have suggested that they’ll implement it – but it has some serious drawbacks. There are parallels with David Cameron’s nightmarish porn filter idea (about which I’ve blogged a number of times, starting here): it could be done ‘automatically’ or ‘manually’. The automatic method would use some kind of algorithmic solutions when a report is made – perhaps the number of reports made in a short time, or the nature of the accounts (number of followers, length of time it has existed etc), or a scan of the tweet that’s reported for key words, or some combination of these factors.

The trouble with these automatic systems is that they’re likely to include some tweets that are not really abusive, and miss others that are. More importantly, they allow for misuse – if you’re a troll, you would report your enemies for abuse, even if they’re innocent, and get your trollish friends and followers to do the same. Twitterstorms get the innocent as well as the guilty – and a Twitterstorm, with a report button and an automatic banning system would mean mob rule: if you’ve got enough of a mob behind you, the torches and pitchforks would have direct effect.

What’s more, the kind of people who orchestrate the sort of attacks suffered by Caroline Criado-Perez, Stella Creasy, Mary Beard and others are likely to be exactly the kind who will be able to ‘game’ an automatic system: work out how it can be triggered, and think it’s ‘fun’ to use it to get people banned. Even a temporary ban while an investigation is going on could be a nightmare.

The alternative to an automated system is to have every report of abuse examined by a real human  being – but given that there are now more than half a billion users on Twitter, this is pretty much guaranteed to fail – it will be slow, clunky and disappointing, and people will make mistakes because they’ll find themselves overwhelmed by the numbers of reports they have to deal with. Twitter, moreover, is a free service (of which more later) and doesn’t really have the resources to deal with this kind of thing. I would like it to remain free, and if it has to pay for a huge ‘abuse report centre’ that’s highly unlikely.

There are other, more subtle technological ideas – @flayman’s idea of a ‘panic mode’ which you can go into if you find yourself under attack, blocking all people from tweeting to you unless you follow them and they follow you has a lot going for it, and could even be combined with some kind of recording system that notes down all the tweets of those attacking you, potentially putting together a report that can be used for later investigation.

I would like to think that Twitter are looking into these possibilities – but more complex solutions are less likely to be attractive or to be understood and properly used. Most, too, can be ‘gamed’ by people who want to misuse them. They offer a very partial solution at best – and the broadly-specified abuse button, as I noted above, I suspect will have more drawbacks than advantages in practice. What’s more, as a relatively neutral observer of a number of Twitter conflicts – for example between the supporters and opponents of Julian Assange, or between different sides of the complex arguments over intersectional feminism, it’s sometimes hard to see who is the ‘abuser’ and who is the ‘abused’. With the Criado-Perez, Creasy and Beard cases it’s obvious – but that’s not always so. We need to be very careful not to build systems that end up reinforcing power-relationships, helping the powerful to put their enemies in their place.

Real names?

A second idea that has come up is that we should do more against anonymity and pseudonymity – we should make people use their ‘real’ names on Twitter, so that they can’t hide behind masks. That, for me, is even worse – and we should avoid it at all costs. The fact that the Chinese government are key backers of the idea should ring alarm bells – they want to be able to find dissidents, to stifle debate and to control their population. That’s what real names policies do – because if you know someone’s real name, you can find them in the real world.

Dissidents in oppressive regimes are one thing – but whistleblowers and victims of domestic abuse and violent partners need anonymity every bit as much, as do people who want to be able to explore their sexuality, who are concerned with possible medical problems, who are victims of bullying (including cyberbullying) and even people who are just a bit shy. Real names policies will have a chilling effect on all these people – and, disproportionately, on women, as women are more likely to be victims of abuse and violence from partners.

Enforcing real names policies helps the powerful to silence their critics, and reinforces power relationships. It should also be no surprise that the other big proponent of ‘real names’ is Facebook – because they know they can make more money out of you and out of your data if they know your real name. They can ‘fix’ you in the real world, and find ways to sell that information to more and more people. They don’t have your interests at heart – quite the opposite.

Paying for Twitter?

A third idea that has come up is that we should have to pay for twitter – a nominal sum has been mentioned, at least nominal to relatively rich people in countries like ours – but this is another idea that I don’t like at all. The strength of Twitter is its freedom, and the power that it has to encourage debate would be much reduced if it were to require payment. It could easily become a ‘club’ for a certain class of people – well, more of a club than it already is – and lose what makes it such a special place, such a good forum for discussion.

Things like the ‘Spartacus’ campaign against the abysmal actions of our government towards people with disability would be far less likely to happen if Twitter cost money: people on the edge, people without ‘disposable’ income or whose belts have already been tightened as far as they can go would lose their voice. Right now, more than ever, they need that voice.

Dealing with the real issues…

In the short term, I think Criado-Perez had the best idea – we need to do everything we can to ‘stand together’, to support the victims of abuse, to make sure that they know that the vast, vast majority of us are on their side and will do everything we can to support them and to emphasise the ‘good’ side of Twitter. Twitter can be immensely supportive as well as destructive – we need to make sure that, as much as possible, we help provide that support to those who need it.

The longer term problem is far more intractable. At the very least, it’s good that this stuff is getting more publicity – because, as I said, it matters very much. Misogyny and the ‘rape’ culture is real. Very real indeed – and deeply damaging, not just to the victims. What’s more, casual sexism is real – and shouldn’t be brushed off as irrelevant in this context. For me, there’s a connection between what John Inverdale said about Marion Bartoli, and what Boris Johnson said about women only going to universities to find husbands, and the sort of abuse suffered by Criado-Perez,  Creasy, Beard and others. It’s about the way that women are considered in our society – about objectifying women, trivialising women, suggesting women should be put in ‘their’ place.

That’s what we need to address, and to face up to. No ‘report abuse’ button is going to solve that. We also need to stop looking for scapegoats – to blame Twitter for what is a problem with our whole society. There’s also a similarity here with David Cameron’s porn filter. In both situations there’s a real, complex problem that’s deep-rooted in our society, and in both cases we seem to be looking for a quick, easy, one-click solutions.

One click to save us all? It won’t work, and suggesting that it would both trivialises the problem and could distract us from finding real solutions. Those solutions aren’t easy. They won’t be fast. They’ll force us to face up to some very ugly things about ourselves – things that many people don’t want to face up to. In the end, we’ll have to.

34 thoughts on “Twitter abuse: one click to save us all?

  1. “@flayman’s idea of a ‘panic mode’ which you can go into if you find yourself under attack, blocking all people from tweeting to you unless you follow them and they follow you has a lot going for it,”

    How does that differ from a locked account?

  2. I am afraid its not just a one gender issue. I have been on other boards attacked for a reasoned evidenced argument I never swore or shouted, but because of this i was stalked by a woman who would visit every site i posted on and made nasty comments about me but when they said stuff about my friends and family thats ok i am a big boy but not about me and mine she wanted me to react and get banned from those sites I just ingored her as she wanted attention. But Abuse is abuse there are laws to stop it already, to have a block a report would only create as you said a mass blocking or dissenting opinions. Forget the fact the abuse part for the moment. This is yet another way for people to stop valid stories that arent the “PARTY LINE”. There is a line between so called trolling and abuse those that talk about raping them thats abuse.. but trolling can be considered by some to be people with different views to others. Have to be very cautious if you bring in “trolling” rules and the abuse buttons.

    • No, it’s not a one gender issue, though the misogynistic aspects have been the ones brought to the fore this time. People can be targeted for any number of reasons, and sometimes for no reason at all. It’s not good at all.

      • I always say if it descends into insults, that means they have no other answer, I am seeing a lot of misandry coming to the forefront due to this. These events can be useful in highlighting problems but they can also cause more problems. I think in part personal responsibility and and societal change need to occur.. Violence against anyone is vile, rape is violence against people there is no excuse for threats of violence against anyone would have thought we would have evolved beyond it. But some people love the reactions they cause, and in some cases (NOT in this case) provocation occurs to get this very reaction. these people are equally as bad.

        as i said i am against ALL violence against anyone, I hate ALL violence (i am not excusing anyone or anything like that have to use that disclaimer as you have said before on here anti filter doesn’t necessarily mean you are pro but people have a tendency to read their own prejudices into comments)

        The answer, well in my personal life i have a 3 strikes rule, once ok maybe an accident, second time possibly an accident but they should have learned its not right, and the 3rd time shows they are unwilling/unable to change. It does mean i have cut some people out of my life and my life is better for getting those toxic people out of my life

      • The problem is very deep – to some extent we have to develop survival techniques. It’s rarely easy, and there are consequences to whatever we do.

  3. I agree with most of your points but don’t understand your comment “Even a temporary ban while an investigation is going on could be a nightmare.” Twitter is useful but I wouldn’t have thought it was critical to anyone’s business (or personal) life? Maybe I’m missing something?

    • A couple of reasons. Firstly, for some people, twitter IS pretty critical to their business or personal life: journalists, for example, or campaigners, or even politicians like Stella Creasy. Secondly, once you get one temporary ban, you can get another – I’ve been on forums before where this kind of system works, and people play games of seeing how quickly they can get another temporary ban after someone’s account is reinstated. Temporary bans become effectively permanent – and perhaps even more annoying. The trolls and abusers all think this is ‘fun’, and enjoy the challenge of ‘gaming’ the system. It’s not good at all.

  4. Thanks for the reference to my panic mode idea, Paul. On the wider issues, we should bear in mind that misogyny, racism, and hatefulness have not become worse problems with the advent of social media tools. The tools have simply acted as a lightning rod or a magnifying glass. Those problems cannot be solved with a tool and they cannot be solved in a day. They can however be relegated to their former prominence.

    • The potential ‘upside’ is that the magnifying glass can force us to face up to the problems rather than sweep them under the carpet. They’re pretty ugly…..

      • As with all things though, better to leave that as a choice. The panic mode idea needn’t be limited to abuse. It could as easily handle the occasional flare-ups where well intentioned people get the wrong end of something and all pile on. Or it could just be something to try out for a little while. It is less of a brake on the platform than protected mode.

  5. I wish more people would look at the current Twitter reporting procedures ( see here).
    Without denying that a “report abuse” button certainly would be misused, the current abuse procedures offer very little by way of remedy and impose very significant difficulties in applying them. For example, it is not possible to report a crime being committed on twitter unless one is the victim of it and it’s reported as “abuse” and the report form simply directs the person reporting back to local law enforcement.

    Things get worse with people posting personal information: a report that personal information is being abused on line can only be made if the information has not been made public anywhere (something which someone complaining about the combination of an incitement to rape accompanied by their private address would not be able to give, assuming them to be on the electoral roll) AND Twitter reserve the right to pass the complaint including additional personal information onto the alleged abuser I think there are real issues about the effective ability to make complaints which result in any action.

    • Yup, they do need to streamline things – but also to be aware of the risks of misuse. There are other ways to report things though – you could report straight to the police, and I can’t see how the police could refuse to look at it. That, of course, relies on the police understanding what is going on…..

      • See below on what the police will have to do to get user information out of Twitter.
        Some forces may well be set up and resourced to go through the hoops, but others will not, and there will obviously be concerns about limitation periods (the Protection from Harassment summary offences require an information to be laid within six months of the last incident, which if things get tied up in US process isn’t long).

  6. Twitter’s Law Enforcement guidelines are particularly worth noting :

    Requests From Non-U.S. Law Enforcement

    U.S. law authorizes Twitter to respond to requests for user information from foreign law enforcement agencies that are issued via U.S. court either by way of a mutual legal assistance treaty (“MLAT”) or a letter rogatory. It is our policy to respond to such U.S. court ordered requests when properly served.

    Non-U.S. law enforcement authorities may also submit requests for emergency disclosure under exigent circumstances, as outlined in the section titled “How to Make an Emergency Disclosure Request,” above.

    and

    Requests for User Information

    Twitter, Inc. is located in San Francisco, California and will only respond to valid legal process in compliance with U.S. law.

    • They kind of have to do it that way – to protect users’ privacy from spurious requests. It’s a tricky balance, and they’ve tended to lean towards privacy in general…

      • Whose privacy, though? I first came across the arcana of the reports procedure when I was attempting to report people who were tweeting the name of the Ched Evans rape victim, a clear breach of UK criminal law and one where the North Wales police had appealed to the public to assist in reporting on-line offenders. Twitter at that point refused to accept third party reports of tweets containing the information in question. It’s hard to see that as “protection of privacy”.

      • Agreed – but when they fought tooth and nail not to give all the details of people associated with Wikileaks to the US authorities, many of us applauded them. They get it wrong a fair amount, but better than most.

  7. Agreed about Wikileaks but there’s a distinction between handing over details of the account which is committing a criminal offence and allowing it to continue to commit that offence without at least suspending it pending an investigation.

  8. what we need are reputation services, that can be pseudonymous; that way the reputation of the person reporting abuse can be used as part of the analysis without needing a real name. no automatic system will get it right all the time, but they are very good at seeing patterns – think of fraud detection systems for credit cards. take the pattern of one person who is already verified by twitter, who normally has a level of interaction X who is suddenly getting hundreds of messages directed at them, who were not previously in their social graph, many from users who have only recently joined the service or who have engaged in previous mass interactions with similar patterns…. it’s never going to be impossible to game, but Twitter and other services could invest a great deal more in reporting & analysis tools that could make the ‘report abuse’ button a useful tool rather than just another method of abuse. As with anything else, technology is an amplifier of human behaviour; we can make it more than a straightforward megaphone while we also work on improving behaviour at the root of this and making it socially unacceptable to make these threats. (Note of concern; discussions of the idiotic comments by Claire Perry on Guido’s blog very quickly started to move from the technological ignorance of someone making ludicrous pronouncements to the same threats of sexual violence against Perry – and few people are up in arms about those threats. Condemnation of these threats mustn’t be a kind of popularity contest for victims.)

    • Reputation services are a complex issue: they can quickly also become a ‘gaming’ issue. There are already ‘reputation management services’, particularly in the US, following the lines of search engine optimisation services, designed to let (rich) people improve their reputations online, wiping out the bad stuff and adding (sometimes fake) good stuff. Having said that, something that can allow a relatively neutral assessment of the validity of a report is at least worth investigating. Even so, I’m wary, as it could create another kind of digital divide, and allow those with ‘good’ reputations and lots of followers too much power.

      On the last point I agree completely! No-one, no matter how ‘disliked’ they might be, deserves that kind of thing.

  9. I congratulate you on your excellent article Paul,well balanced has usual the world is frightened & hitting out,with freedom comes responsibility much that is said on twitter would result in a affray if said in a public house,sat at home protected by a few miles distance it is easy to find Dutch courage, that even a skin full of ale,could not bring to the fore,maybe it is the feeling of releasing the frustrations of life today,but no excuses if i heard what i read on twitter in goading & down right rudeness i would hope they would get them arrested for incitement ,the lack of responsibility integrity & accountability undermines all our freedoms

  10. A quick thought, how about a convention that alerts all your friends? They can see the tweets and judge the nature of the offence, and get involved by making it their problem and alerting their friends etc. Trolling, like bullying, succeeds only in isolation. If we as a community wish to set standards of acceptable behavior, then we all need to act collectively in each other’s defence. This avoids the celebrity issue, as the community is larger than any one individual’s followers. We have the power to police this ourselves, and to do so will give the community greater credibility, and support greater independence from external intervention.
    We’d need a way to make such a tweet visible beyond just a mention and need to protect that method from spammers. Hope someone smarter can figure out that wrinkle.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s