The Sidelines: The trolls beyond the bridge
With Clarissa Lee, researcher
The Sidelines is the supplementary issue to every main edition of The Starting Block. Here you will find the interview transcript and more information about the conversation of the week. The interview is transcribed by Otter.ai and edited for length and clarity. All links provided come from me, and not the guest, unless stated.
Listen to the audio version here.
TRANSCRIPT
TINA CARMILLIA: Hello, my name is Tina Carmillia and this is The Starting Block, a weekly conversation on science and society with an emphasis on disinformation, data, and democracy.
As usual, the transcript and credits for this conversation are available on The Sidelines, the supplement to every main edition of The Starting Block.
Now, in the next lane: Clarissa Lee, an art-science creator and researcher at Universiti Malaya and fellow at the ISEAS-Yusof Ishak Institute, for a sneak peek into her research on Internet trolling and Malay social media.
Ready? Let’s go.
CLARISSA LEE: Okay.
TINA CARMILLIA: So just very briefly I have an idea of your academic work and how it’s multidisciplinary and how it’s interconnected in so many different ways but for me, but I got to learn about your work based on, particularly on your work on troll profile and behaviour. That was what interests me in your work. Maybe we can touch on that—your research interests, and how it has evolved since.
CLARISSA LEE: I got interested in trolls because I was actually looking at political trolling in the US, as well as during the whole Brexit thing. And initially, I didn’t think it was relevant to Malaysia, also because my collaborator for that project is actually British, well, Scottish actually. So we have a lot of conversations on politics and 2016 was when our work started and in the beginning I was collecting a lot of articles and following a lot of postings relating to—actually it wasn’t really about trolling, but political discourse, left and right, in these two countries because there was [Donald] Trump declaring his candidacy. And then there was Brexit that was happening—or not happening, people didn’t know yet.
From there, I noticed there were a lot of articles coming out about trolling. I was curious what does that mean, is it true I knew was the one under the bridge, it was the mythical creature. So I started reading about trolling, as a sort of like an academic digression and also as a stress reliever—me and my friends started to use some of these troll-ish phrases that were used elsewhere, we’d troll each other as well.
Then we were thinking, why don’t we write a paper on trolling, which, by the way, was not even supposed to be trolling in Malaysian. It was just the idea of the philosophy of trolling, which was what we did. We actually look at the philosophy of trolling, we looked at philosophical works on trolling because my collaborator was philosopher, as well.
And from there I got interested in the connection between trolling and hacking and I was basically arguing that trolling is the way of social hacking, as well. And in the age of cybersecurity interest, which, by right, is an interest that has been around but I think it’s only gained mainstream interest—in the sense that people with no technical interests, started becoming interested in cybersecurity when it used to be just people with ethical interests who was interested—and I say that trolling itself has an important social engineering implication on cybersecurity as well. In fact, a lot of hackers for them to gain me information, it’s by trolling/
Trolling is not what we think it is. It’s not just me going to cyberbully you. People mistaken cyberbullying for trolling, which is not the same thing. Trolling actually is a lot more like a puppet master and usually they are invisible. In fact the fight sometimes the cyberbullies are controlled by the trolls. They are not the same as the trolls. The cyber troopers are trolls in the mundane sense—it’s a very mundane term, but if you go back to the real technical meaning of trolls, they're not really trolls, because their strings are being pulled, they’re being told how to behave in a particular way.
When I say pull, it doesn’t mean that, okay, everyone, come listen to me, I have these instructions for you. It’s not like that. It’s more like, for instance, when I’m hacking something, I put a Trojan horse into your system, and then I see what I can gather about you, and then from there I can launch an attack. So I don't actually tell you I’m going to attack you in a very open way, but rather, I let you behave the way you already do and I use that against you. That’s basically how trolls actually are able to manipulate cybertroopers.
That was interesting for me because when we're going through the different versions of the paper that we were writing there were also [Malaysia’s 2018 general elections] coming up, and I was thinking it was interesting because, for the first time, the opposition was launching a major offensive online. They will also be countered by the incumbent [Barisan Nasional government] because back then, I think there was a time when BN didn’t really care about what went on online because they had so much control of the rest of the media. But they also realised that they’re now dealing with a different generation of people who were probably teenagers in their last election, but who are now young adults who are able to vote, and they realised they cannot rely on mainstream media, because most people don’t watch TV unless it’s available on their computer. So they realised they had to attract a different audience and, I think, they also brought in a lot more information officers who are young who are part of that generation, who probably had them change the entire strategy of operation.
And I think that’s when trolling to started become a big thing in Malaysia, in a very explicit way. Of course you know, there was Invoke that was using data from Cambridge Analytica. This was what Rafizi [Ramli] claimed. That got me interested in how far is the reach of Cambridge Analytica for all the world politics.
And then of course BN was hiring very obvious cybertrooper leaders who I wouldn’t really consider are trolls because if they are trolls, they’re not very smart, but they still fulfilled their function because technically for you to sway the crowd, all you have to do is create a bystander effect, you don’t really have to be very smart. So from there I also got interested in the bystander effect, which for me is a very important way of understanding how the trolling strategy is operating as well. So I hope this gives you the technical background of my work.
We have to decide, are we going to go with the average person’s understanding of social media to create the literacy, or do you want to go with the ideal person’s understanding … and then impose down on the average people.
TINA CARMILLIA: Yeah, and then you touch a little bit on the relationship between social media use and politics and democracy within the Malaysian context. But, I suppose, specifically because of my research interests as well as what you’re currently working on right now that we’ve discussed prior to this conversation—you have a specific focus now as well, on the Malay social media sphere, which is, which is very interesting. I know that globally, it may sound a little bit niche, but I think it’s what makes it interesting to examine. Maybe you can share this a bit more on what you’re looking at.
CLARISSA LEE: The work is actually still developing. So, I think the driving area of inquiry that’s my entry into this—but I’m planning to expand on it beyond that—is actually understanding so-called ‘minority communities,’ right? Firstly, when you think of the Malay language just on Google Translate it is probably one of the last languages to be included and even then it still doesn’t translate very well.
But if you think about how Google Translate is used on platforms—I was trying to look at some translation, because we have now more Malay people who type in jawi. I can read some jawi but I haven’t been very consistent in practising so I can only read certain alphabets and guess what that word is but it takes a long time, as much as Chinese characters are for me to read the whole sentences.
Sometimes I would kind of like jump cut if I just want to see what the big picture is, I try to use translation, and sometimes they’ll use either Google Translate for Persian, which is actually not even what jawi, or for Arabic, and I get very weird translations. Sometimes the translation just becomes a two-word thing—it was apparently a long sentence, I don’t know why they think it was a two-word thing, maybe Google Translate is also confused.
So thinking with all this confusion—and we know that a lot of technology write off each other, it’s not like they create a whole new technology, especially for language translation which is so complex, it's not possible for one platform to create their own technology to do that, because that will cost them a tonne of money to do it. And that got me thinking about what is the value of these communities for these platforms. And the fact that Facebook and Twitter are in Singapore now, not because they are trying to cater to Singapore, because Singapore is the best place for them to be because they’re actually trying to expand to the Southeast Asian market, as well, because they know that this is the emerging—I mean, a lot in developed countries are already ageing so they are running out of young people. But there are still a lot more young people coming out and being born every day in Southeast Asia.
And if you think about all the minority languages that actually are present in Southeast Asia, there’s more differences. At least even in Europe, even though they have their own languages, but they sometimes do speak another person’s language, but in Southeast Asia, maybe English is a common language but you can’t say that’s the case for every single person in that country. So a lot of times there’s like silos of languages, that’s actually happening.
And that got me thinking about how these people get amplified, given that probably these platforms don’t understand what they’re saying? And I sometimes wonder whether—I mean, of course, they might say that typing in Jawi itself maybe is a nationalistic move. You know, it started from this whole hoo-hah about introducing Jawi back into the school system and people protested against that and we started to see people writing in Jawi. For fun I started writing my name in Jawi just to see whether that changes thing for me as well. And now they actually have Rumi and Jawi going on. That’s like—okay, I mean, is that going to mess up the whole machine learning platform. There’s that.
And I also was interested as to why Malaysians especially Malay speaking audiences think that social media platform is a good way for them to get together, or to build a critical mass for shared belief systems. And this happens a lot on Facebook. In fact, when I told someone once that I was going to look at Facebook and Twitter—and this was when I was just starting up because I wasn’t interested in Malaysian social media back then, I was just interested in the larger concept of trolling, in the broader philosophical sense—I was told that Facebook and Twitter are very different because back then that I didn’t follow pages, unless it’s related to my area, like, hobbies. Facebook was actually more of a fun thing for me to be a troll sometimes, right, but it wasn’t meant to be a serious thing. But I started following a lot of political pages, including Malay troll pages on Facebook, and then I realised, yeah, they were very right.
And also I was once involved with some controversy, I wouldn’t say what it also involved a friend, and he got into some trouble because he was making a comment that was on a religious holiday, and I didn’t know what was going on by but I thought he posted something that was cute so I commented on that cute thing. And for some reason, I got screenshot, together with other comments—I wasn’t even their target because they were a lot of other more important targets in that screenshot that they were going after but I was surprised to see the screenshot with my little tiny inconsequential comment making the rounds. My friend actually got into trouble for that. That’s where I actually learned about how things travel from Twitter to Facebook as well. And I thought that was actually a very interesting thing.
And one of the things that was also curious—because I have a lot of followers outside of Malaysia, and I find it very interesting that some of them who basically do not know any of these Malaysians started following them [and] liking post by Malaysians, some of whom I don’t actually know because now Twitter will show you on the timeline of you they like. And, of course, most of the time they like post that are in English, not written in Malay but sometimes if you go through the nest, you might find some of the posts are in Malay as well.
So it got me interested in understanding how the heck did this language spread, you know these are so-called ‘minority communities,’ who I thought, are just talking to each other anyway, but I was surprised to see how that actually spread. But I don’t know whether this is a consequence of my location. If I were to move back to the U.S., will I still see the same thing, or will I see it differently, or maybe if I can borrow a friend’s account, or maybe I can change my GPS location and say I’m in Atlanta or like that, would that actually change—or maybe I should create another bot account for me to just be able to track that.
So that got me thinking that maybe Facebook and Twitter are actually slowly changing how the algorithms work to cater to all these minority languages which have inadequate translation ability. Of course, they still amplify a lot of the wrong things, they still translate everything wrong, but the fact that they think that it is valuable and is necessary connects to who they hire because I know Facebook had brought in Malay translators but I don’t think there are enough of them.
And of course, there’s always the conflation of Bahasa Indonesia and Bahasa Melayu, as well. For the longest time, usually people asked me , “Do you have any proficiency in Bahasa Indonesia?” So if I want to get a job as a translator, I have to know Bahasa Indonesia because who give a damn about Bahasa Melayu but now they actually have a distinction between that, so I thought that was an interesting development. So maybe, there is a value in the minority communities that the platform's themselves are seeing that also changes the way they are designing the platforms to appeal to these people and to get more and more and more of them on board.
Another thing that’s interesting for me is the number of people who log into places like—you know, on Facebook they use their real names, of course, they are some who use their pseudonyms, but on Twitter, they have very weird handles that are almost very cryptic. It’s almost like your phone asking you, “Do you want to use a strong password?” It looks like that, and by the time you get a look at it I don’t remember what was this person’s handle because it’s like a random number generator. I will sometimes click on them to see whether they are real people or bots. They look like real people or they are really very smart bots or maybe they are just bots with real people behind it. And I wanted to look at this phenomenon a lot more as well. So that as of now is all I can tell you in terms of my foray into—becasue it’s still very early.
I remembered hearing how people of colour in the U.S. regularly get their account on Facebook suspended because they say something that white people do not like to hear. So I was thinking about something parallel for the Malays, except that Malay people are saying something that other Malay people do not like to hear. So it’s not like it’s me, a Chinese person, going out at them, but it’s actually their own people— which also, in a way, helps you understand the class tensions.
TINA CARMILLIA: I would love to check in again in the near future once you're further into it. But I also think that maybe we can touch a little bit, just generally—especially with your background in more of the socio-philosophical elements of social media as well—there’s certainly a difference in the Malay, that we see for formerly, like, you know, in the news or in the literature that we read, and the Malay that we write and read on social media. It’s significantly different, it's perhaps even more significantly. How does that factor into all of this?
CLARISSA LEE: I think the difference is already present in earlier versions of social media, Web 1.0. If you look at a lot of forums that we posted in. Originally when I used to go to these forums and look at a lot of of Indonesian postings, obviously the slang in Indonesian is very different from slang in Bahasa Baku Indonesia. Sometimes I don’t even understand what they’re saying even though I understand Bahasa Baku Indonesia. And I think the Malays also started following in that form.
Back then there were not a lot but if you look at a lot of blogs you would actually see languages like that. And interestingly, once I was typing a word that a student was using and I got into a sex blog instead and it was Malay sex stories, something like that, I don’t remember the title of it.
So that already existed and we just brought that over onto Twitter. If you want to see a lot of slangs, you have to look at some of the most crass, most disturbing, and also most offensive tweets, especially when they’re targeting women, if they are cyberbullying women. If you look at that, you will see that language coming up very strongly, and a lot of times I actually use that to determine the sex of the person speaking even when they try to mask it because sometimes you can actually tell the masculinity of language, but there are also of course certain female gendered persons who sometimes adopt the language, especially if they are trying to hide who they really are. I haven’t looked at that in great detail yet, so I probably will need to work with the linguist, because I’m not a stylistician, but I think there’s a lot to consider here and I think there are other people who are probably studying versions of it but mostly in English. I haven’t seen a lot of people trying to do the same thing in Malay. Maybe it has to do with expertise and how broad-minded they are as well.
TINA CARMILLIA: Sso, I guess, in an ideal world what does good social media literacy look like to you especially as someone who you know examines this day to day basis for work.
CLARISSA LEE: You can actually ask me that probably three months from now because I’m actually collecting data for this fellowship, to really learn about what that means. I may have a lot of conjectures right now but they may be completely useless by the end of the three months because these are based on my understanding which may not work with a lot of people. And also because we have to decide, are we going to go with the average person’s understanding of social media to create the literacy, or do you want to go with the ideal person’s understanding of literacy and then impose down on the average people.
So, this is just a thought experiment and we are just making my own assumptions but this is definitely something that I’m working on which is why I took up this project. It’s not just to study Malay social media because that’s not that useful, but what can we learn from that and how we can create something that is empowering and improve agency of especially marginalised groups on social media.
TINA CARMILLIA: And, just for a wrap up: What trends should we be looking out for in terms of the relationship between social media and, and our democracy, I guess, in the general sense, especially now in the prolonged period of lockdown because of the pandemic, we are relying heavily on social media for our communication, for our news for everything almost, right?
CLARISSA LEE: Well, instead of telling you what trends to look out for, I’ll ask you to look at behavioural changes because that will actually determine the current trends, right? For instance, we used to say, you know, Asians are happy to just share their private information with anyone. They are will to dox themselves voluntarily, you don’t even need to dox them. But I’m starting to see a new generation of people who are actually very attentive to their privacy and to their own security. I think that’s why there’s also a growing number of people who go online, anonymous. I mean they don’t try to do that random number generator anonymous handle but they still go anonymous but their handle has some meaning. It’s also a way to mask their identity, because they know they’ll be prosecuted if they’re known, they’ll be subject to trolling and worse.
And I think this is actually more common among Malay people than I think other races, it’s that they like to report people to the police. So if you say something they don’t like on Twitter or Facebook they’ll report you. And sometimes in the process to get your account suspended—I don't know whether there’s a connection between the police and the—maybe the police report to the Malaysian Communications and Multimedia Commission and the MCMC report to the platform and they get suspended, because I remembered hearing how people of colour in the U.S. regularly get their account on Facebook suspended because they say something that white people do not like to hear. So I was thinking about something parallel for the Malays, except that Malay people are saying something that other Malay people do not like to hear. So it’s not like it’s me, a Chinese person, going out at them, but it’s actually their own people—which also, in a way, helps you understand the class tensions.
I’m not sure whether religious polarity is something that needs to be considered, it’s actually not even my expertise, so I wouldn’t go into that but I definitely would like to look at how other people are studying, let’s say, political Islam. And then, for me, taking their work and see how that actually is really being played out on social media. Is it playing out in the same way? Because political Islam actually came about with the rise of social media in Malaysia before even Web 2.0, during Web 1.0, when they were a lot of blogs. Then in fact Twitter was actually already present but what they called ‘microblogging’ so they were people who’d do short blogging, you know like, people go to parties or go to some nightclubs or whatever they would blog with photographs, right? So this proto-Twitter-like behaviour actually already existed.
But Malaysia used to have this site called SoPo—I think maybe it disappeared already but, it’s called Social Political Sentral. It’s like a hub where they collect all the blogs that say anything political, basically, or anything on social-cultural interests. My old blog that is now defunct used to be part of that as well. That’s how I discovered that site and found so many political bloggers who by the way, are just citizen—people refer to them as citizen journalists because they were supposedly doing what journalists failed to do, because back then there was only Malaysiakini and even then it was still developing, so it wasn’t really what it is today. And a lot of the other sites, even though they have a website presence, they were not really using the digital mode for publication and you still have to go and buy the hardcopy like even Aliran, you still had to buy the magazine, you can’t read it online, compared to what it is today.
So, yeah, I will say that it has changed a lot, but it’s also in a way, a reiteration, with some new ornamentation and dressing of what actually was already present for, I would say, two decades now.
TINA CARMILLIA: Exciting space to be in social media, especially for work like this to be done so I’m really interested to continue following your work and I hope you'll keep me in the loop.
CLARISSA LEE: I’ll let you know when some of my writings come out so, yeah, you will actually see it.
TINA CARMILLIA: That’s Clarissa Lee, a researcher at Universiti Malaya and fellow at the ISEAS-Yusof Ishak Institute on The Starting Block.
If you would like to join me on the show for conversations like this, get in touch here.
Don’t forget to subscribe, if you haven’t, and if you enjoyed this episode, consider sharing it with someone. ‘Til the next one, goodbye for now.
CREDITS
Guest: Clarissa Lee, PhD
Producer/Host: Tina Carmillia
Music: Theme for Harold (var. 3) by Kevin MacLeod, licensed under CC BY 4.0.
