Beware The Troll

I very much enjoy reading Kyle Baldwin‘s “Apples and Bongo Drums” blog and thought I’d reblog one of his posts here. I was a little spoiled for choice, however. (I recommend you visit Kyle’s blog and read his other posts). After some umming and ahhing, I decided to go with his “Beware The Troll” post from July. Let’s just say it resonated.

Over to Kyle…


Beware the Troll

Kyle Baldwin, July 25 2016

troll.png

Once, Trolls were purely mythical creatures that lived under bridges and ate unsuspecting would-be bridge crossers. Or, if you prefer Tolkien mythos, Trolls were great big dim-witted creatures that can only be defeated by a combination of Hobbit stall tactics and sunlight petrification. Either way, it’s safe to assume that when these myths were being written, the total sum of actual death certificates that read “Cause of death: Troll” was nil.

That’s not quite true anymore.

If you’ve ever wasted a few hours scrolling through the YouTube comments section, and been left wondering where it all went wrong for the human race, there’s a good chance you encountered an internet troll or two. If you haven’t, then here’s a brief outline of what an internet troll is: an awful human being. More specifically though, it is someone who constantly posts in online forums with no purpose other than to disrupt conversations, provoke arguments, or just plain bully. They may hardly sound like the world’s biggest problem right now, and may even sound mostly harmless, but sometimes they’re a more insidious pest than my description does justice.

For example, Jessica Laney, a 16 year old Floridian girl, took her own life after being on the receiving end of online bullying, which included messages telling her to go kill herself. There’s also Charlotte Dawson, 47 year old New Zealander and TV presenter who committed suicide after years of online harassment – harassment carried out under the Twitter banner #diecharlotte. These are just a couple of highly publicised examples, but cyberbullying related deaths are becoming increasingly common, to the point where they no longer make headlines. One study shows that suicide ideation is strongly linked with cyberbullying, and we’ve not even touched on how many cases of clinical anxiety or depression can be linked back to online abuse. The exact numbers for cyberbullying related long-term mental health issues is unclear (unsurprisingly), but it is very well documented that bullying leaves permanent scars, and according to one study, as many as 43% of students will get to enjoy being cyberbullied during their lifetime.

If you thought lone trolls were a nuisance, you clearly don’t know much about 4chan. This is a little dark corner of the internet where trolls like to meet up, hang out, pick a target, and make casual rape threats (amongst other things). It’s a misogynist’s paradise, it’s a mob, and it has a way of achieving its goals through “hacktivism”. Wonderful.

LutherTo be fair, for the most part trolls are just people who get a kick out of causing a bit of mischief by derailing a conversation, or by mildly irritating people with political bait. But there are also those who take it many steps further, and the internet is the perfect platform for all their trolling desires. Further, comments sections seem to bring out the utter worst in people in ways that would never happen in the non-virtual world. Why does this happen, and is there any way to prevent it? … Other than the brand of anti-troll vigilante justice Luther endorsed, of course.

What feeds the Trolls?

In 2004, John Suler coined the term the “online disinhibition effect”, which, in a nutshell, states that people are willing to behave differently online than they would in reality. This isn’t really a single effect, but rather a collection of different factors and psychological effects that add up to make trolling inevitable. List time!

  1. I am no one. Probably the most obvious factor is that commenting is often completely anonymous, and anonymity gives a sense of security against any reprisals. You might be able to find a mister B. Kaldwin and give him a piece of your mind, but Prince_RobotIV? Who even is that guy?
  2. I am invisible. This is similar to the above, but also adds the fact that the form of communication – text – does not deliver tone, facial expressions or emotion. They are just words on a screen that could be interpreted a hundred ways, and concern over appearance is lost entirely. Further, a troll can pretend to be any age, sex, race or species he/she/it wants if it serves their purpose better. Not only do you not know who or where I am, but you don’t know what I am. And that gives me power.
  3. LOL BYE. The fact that conversations online are asynchronous – one can leave and come back to a conversation at any time – lowers inhibitions by allowing a person to go away and take all the time they need to think of the perfect cutting words. It also gives the troll the opportunity to blurt out something and not worry about a reply until they’re ready to log back in. Very unlike real life. This also means you can throw in an inflammatory remark on a popular YouTube video, grab a box of popcorn, and watch the ensuing mayhem.
  4. You’re who I say you are! This is a slightly abstract aspect, and is the most subconsciously active of the list yet. When you read a comment, although you don’t see, hear or smell the person on the other end, your brain automatically assigns characteristics to them. Usually, for some reason, people imagine that the commenter is male, white, and less intelligent than you. This is particularly the case in political arguments, where as soon as a disagreement begins, you imagine the other person as the pure stereotype of your political nemesis, and slowly, but surely, you begin to feel vindicated in displaying your ire.
  5. Want to play a game? Suler observed (alongside criminal lawyer Emily Finch studying online identity theft) that many people see the online realm as a form of escapism, and interacting with other commenters is nothing more than a game. Games have no real consequences, so what’s the worst that could happen? That girl won’t really think I want her to kill herself! That would be mental!
  6. You’re in my world now. This is kind of obvious when you think about it, but surprising to the uninitiated – the hierarchy of the online world follows its own rules. Even if you know the status of a person in the real world, it has little bearing on their perceived status in online communities. Instead, in the online realm, their online following is what gives them authority in chat rooms. To put it simply, a vocal twitter user who has many online followers but no power in reality, has more “cyber power”, and is feared as an authority figure in cyberspace over a prominent politician who only occasionally tweets (and especially if the tweets are of his own name – I’m looking at you Ed Balls).

Add all these effects together, and you get a worrying reduction in social inhibitions online. Politicians are still debating how to grapple with this surprising consequence of giving everyone a cyber-voice, and we’re left with the question:

Do we just accept them?

troll2.png

The good news is that research suggests that trolls are in the vast minority. The bad news is that YouTube comment algorithms create the perfect troll feeding grounds. When you leave a comment, there are two ways in which it will find its way to the top of the pile – lots of thumbs up, and lots of replies. It is far more heavily weighted in favour of number of replies, though, and this is the crucial factor that makes trolls so visible – people can’t help but to reply to a comment they strongly disagree with.

The simplest solution would be to just learn when people are trying to get a rise out of you, and ignore … but that’s really difficult. Replying is reactive (as opposed to the far more passive ‘thumbs up’ option), and trolls, almost by definition, exist purely to illicit reactions in others. They’ve found their craft, honed their skills, and have become very adept at it. No matter how many times I hear the phrase “don’t feed the trolls”, when I read something outrageous, I just can’t help but want to send a snarky “Oh YEAH? Well let me tell you, sir/madam, that I think your specific politics are WRONG!”.

Trolls get the most replies, their comments find their way to the top, and suddenly they seem like they are the majority – we’ll call this troll-bias. Recently, a few YouTubers have begun requesting that their viewers leave a “+” reply to all comments that they agree with, to offset this troll-bias. It’s a clever response, and it does indeed seem to already be making comment sections far more pleasant places.

With regards to other forms of cyber-bullying, there aren’t easy answers, sadly. Forum moderators can take comments down, but usually they act after the harm has been done, and as Jon Ronson’s book So You’ve Been Publicly Shamed highlights, no number of moderators could hold back the tide of hateful comments that come with the Twitter mob.

Stricter laws should be implemented, I think. It’s not okay to tell someone to “go kill yourself” – no amount of free speech rights should leave the door open for that kind of behaviour. The online world though is multi-national, so how do we ensure safe spaces when laws aren’t always universal? It’s also, as I pointed out above, almost entirely anonymous! The ever-watchful, and infinitely creepy NSA might be able to automatically respond to a person openly asking Facebook how to build a dirty bomb (not that they ever do!), but a local police force is not going to be able to do much about a string of hateful comments coming from 4chan/b/.

School bullies have existed for as long as there have been schools, and, so far, no one has figured out how to stop them from growing like weeds. Now, our school bullies have the internet at their fingertips, and the online disinhibition effect in their heads, and it’s a bit scary. So what do we do when a troll doesn’t even recognise his victim is a person? Teach them.

Lindy West, an online activist and vocal feminist, was accustomed to online abuse – anyone who writes about feminism is an easy target for the more misogynistic trolls out there. But then one day, a specific troll took a different tactic than the usual rape/death threats, and decided to impersonate her dead father. Sickening, right? Usually, she would follow her peers’ advice and ignore the trolls, but this time and she decided to write a piece about it. The troll read the piece, realised there was real person on the other end of his abuse, apologised to her directly, and changed his ways. The troll decided he’d had enough and came out from under the bridge.

The answer, then, has to be in education. First educate police forces to take online abuse seriously (because they are often woefully inadequate at this so far), and educate youths to understand that not only is bullying a crime, but the people on the other end of comments are just that – people. They’re not tropes of your political nemeses, this isn’t a game, and you are not absolved of all responsibility by your anonymity.

And if that doesn’t work … Well, there’s always the Luther approach.

Note: the author of this blog in no way endorses vigilante justice.

Trollface image attribution: By Azzy10 (Own work) [CC BY-SA 4.0 (http://creativecommons.org/licenses/by-sa/4.0)%5D, via Wikimedia Commons, https://upload.wikimedia.org/wikipedia/commons/7/73/Trollface.png.

Luther (Idris Elba) image attribution: By DFID – UK Department for International Development – https://www.flickr.com/photos/dfid/15418120205/, CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=35853981

We are anonymous. We are legion. We are (mostly) harmful.

This revelation appeared in my Twitter timeline earlier this week:

On the same day, Nature News published a fascinating interview with Brandon Stell, the founder of PubPeer who has revealed his identity to the world:

I’ve waxed lyrical about PubPeer a number of times before, going so far as to say that its post-publication peer review (PPPR) ethos has to be the future of scientific publishing. (I now also try to include mention of PubPeer in every conference presentation/seminar I give). I’ll be gutted if PPPR of the type pioneered by PubPeer does not become de rigueur for the next generation of scientists; our conventional peer review system is, from so many perspectives, archaic and outdated. I agree entirely with Stell’s comments in that interview for Nature News:

Post-publication peer review has the potential to completely change the way that science is conducted. I think PubPeer could help us to move towards an ideal scenario where we can immediately disseminate our findings and set up a different way of evaluating significant research than the current system.

But one major bone of contention that I’ve always had with PubPeer’s approach, and that has been the subject of a couple of amicable ‘tweet-spats’ with the @PubPeer Twitter feed, is the issue of anonymity. I was disappointed that not only was it just Stell who revealed his identity — his two PubPeer co-founders remain anonymous — but that there are plans (or at least aspirations) to “shore up”, as Stell puts it, the anonymity of PubPeer commenters.

I am not a fan of internet anonymity. At all. I understand entirely the arguments regularly made by PubPeer (and many others) in favour of anonymous commenting. In particular, I am intensely aware of the major power imbalance that exists between, for example, a 1st year PhD student commenting on a paper at PubPeer and the world-leading, award-winning, scientifically decorated and oh-so-prestigious scientist whose group carried out the work that is being critiqued/attacked. Similarly, and in common with Peter Coles, I personally know bloggers who write important, challenging, and influential posts while remaining anonymous.

I also fully realise that there are are extreme cases when it might not only be career-theatening, but life-threatening for a blogger to reveal their identity. However, those are exactly that: extreme cases. It’s statistically rather improbable that all of those pseudonymously venting their spleen under articles at, say, The Guardian, Telegraph, or, forgive me, Daily Mail website are writing in fear of their life. (Although, and as this wonderful Twitter account highlights so well, many Daily Mail readers certainly feel as if their entire culture, identity, and belief system are under constant attack from the ranked hordes of migrants/PC lefties/benefit claimants/gypsies/BBC executives [delete as appropriate] swamping the country).

I am firmly of the opinion that the advantages of anonymity are far outweighed by the difficulties associated with fostering an online culture where comments and critique — and, at worst, vicious abuse — are posted under cover of a pseudonym. For one thing, there’s the strong possibility of sockpuppets being exploited to distort debate. Julian Stirling, an alumnus of the Nanoscience Group here at Nottingham and now a research fellow at NIST, has described the irritations and frustrations of the sockpuppetry we’ve experienced as part of our critique of a series of papers spanning a decade’s worth of research. In the later stages of this tussle, the line separating sockpuppetry from outright identity theft was crossed. You might suggest, like PubPeer, that this type of behaviour is not the norm. Perhaps. But our experience shows just how bad it can get.

Even in the absence of sockpuppetry, I’ve got to come clean and admit that I’m really not entirely comfortable with communicating with someone online who is not willing to reveal their identity. I’ve been trying to get to grips with just what it is about anonymous/ pseudonymous comments that rankles with me so much, having been involved in quite a number of online ‘debates’ where the vast majority of those commenting have used pseudonyms. When challenged on the use of a pseudonym, and asked for some information about their background, the response is generally aggressively defensive. Their standard rebuttal is to ask why I should care about who they are because isn’t it the strength of the argument, not the identity of the person making the comments, that really matters?

In principle, yes. But the online world is often not very principled.

Ultimately, I think that my deep irritation with pseudonyms stems from two key factors. The first of these is the fundamental lack of fairness due to the ‘asymmetry’ in communication. Dorothy Bishop wrote a characteristically considered, thoughtful, and thought-provoking piece on this issue of communication asymmetry a number of years back (in the context of the debate regarding the burqa ban). The asymmetry that’s established via anonymous commenting means that those who are critiquing an author’s (or a group’s) work are free to comment with impunity; they can say whatever they like in the clear knowledge that there’s a negligible chance of their comments ever being traced back to them.

Stell and his PubPeer co-founders claim that this is actually a key advantage of anonymity — those who comment are not constrained by concerns that they’ll be identified. But if their arguments are sound, and expressed in a polite, if critical, manner, then why the heck should they be concerned?  After all, it’s long been the case that “old school” journals — the APS’ Physical Review family of titles being a notable example — publish critiques of papers that have previously appeared in their pages. Those formal critiques are published with the names of the authors listed for all the world, or at least the readership of the journal, to see.

We should aim to change the culture so that critiquing other scientists’ work is seen as part-and-parcel of the scientific process, i.e. something for which researchers, at any career level, should be proud to take credit. Instead, the ease of commenting online from behind cover of a pseudonym or avatar is encouraging a secretive, and, let’s be honest, a rather grubby, approach to scientific criticism. I was therefore particularly encouraged by this announcement from The Winnower yesterday. It’s a fantastic idea to publish reviews of papers from journal club discussions and it’ll help to move the critique of published science to a rather more open, and thus much healthier, place.

The second, although closely related, aspect of anonymity that winds me up is that it essentially (further) depersonalises online communication. This helps to normalise a culture in which those commenting don’t ever take responsibility for what they say, or, in the worst cases — if we consider online communication in a broader context than just the scientific community — fail to appreciate just how hurtful their abuse might be. I’ve often seen comments along the lines of “It’s all just pixels on a screen. They should toughen up”. These comments are invariably made by those hiding behind a pseudonym.

As ever, xkcd has the perfect riposte…

This splendid poem also makes the point rather well. “Cuts and bruises now have healed, it’s words that I remember.

Anonymity contributes to a basic lack of online respect and too often can represent a lack of intellectual courage. When we criticise, critique, lambaste, or vilify others online let’s have the courage of our convictions and put our name to our comments.