During the past nine years we have seen the rapid growth of the Colombian networked public sphere. As Internet connectivity has spread to the country, and more people have adopted smartphones, the use of Web social software and computer-mediated communication has become part of the everyday life of many youths and adults avid of expressing their opinions and participating in a the new communication environment. According to recent statistics, for instance, there are 6 millions of Twitter users in Colombia. Web platforms and applications like Twitter have become important spaces for public discourse and political debate. From congressmen to taxi drivers to soccer players to ordinary people, Colombians have increasingly started to participate in a vibrant, noisy, and sometimes chaotic arena of discourse in which the political passions of the country seem to be exacerbated. Although the polarization of opinions has characterized the country given its long lasting civil war, in the past six years, I have been noticing an increasing proliferation of hate speech on the Twitter platform. Particularly, I have noticed a consistent attack to the Colombian government, the peace negotiations, and other civic initiatives intended to create discussion and conscientize the population about the complexity of the armed conflict, the victims, and the possibility of a post-conflict scenario. Given the characteristics of those attacks, the language used, the form of the messages, and the pace in which they happen I have started to wonder to a what degree, such proliferation of hate speech in the networked public sphere is animated by twitter bots.
What are twitter bots? basically they are little programs that can produce automated posts on the Twitter platform. They can create any kind of text based on a predefined algorithm, using #hashtags, keyterms, and any other parameters that could include sampling texts from anywhere on the web. They can also follow and reply other users, favorite tweets, and retweets. Although the use of these bots has been usually in the realm of new media arts, spam, and marketing, an emerging field of use is the one of politics. In last year Turkey’s protests and political crisis, for instance, twitter bots were deployed to shape online discussion, and to “troll, hack and corrupt those critical of the Turkish Government.” Since activity in Twitter has started to be correlated with political behavior, socialbots have the potential of helping to shape the information flow and discussion on the networked public sphere, with some possible consequence on the electoral process and the formation of public opinion. In a recent study,researchers have found for instance, that social media attention is correlated with electoral performance in the U.S.
In the context of Colombian current process of peace negotiations, twitterbots could certainly help to polarize the discussion on the networked sphere, especially when they are programed to disseminate hate speech. The fast replies with capital letters, insults, and attacks to other Colombian twitter users have become part of the discursive practices that I have been able to notice during my interactions. Two weeks ago during the public event #vidamarzo8, for instance, we could see for instance, how the discussion on Twitter had several participants that dedicated all their efforts to reply to the participants of a public event with insults and aggressions. The #vidamarzo8 event was a public march, organized by Antanas Mockus, that intended to create awareness about the importance of life in Colombia, and was celebrated across the country and also worldwide. The goal of the event was to reflect about the need to respect life and to imagine it as sacred, as something irreversible. At its core, it was a march for healing, for imagining a country in where people cares about the life of other citizens, even if they think different. During the day of the event, as well as from several days before, one could easily find tweets that were intended to sabotage the event. I by myself received some twitter replies from users who opposed to the event, some of them speaking in CAPITAL LETTERS, trying to attack a twitter post I published that day after having met with other Colombians in front of the Capitol in Austin. The fast replies I received from certain Twitter users that used a #hashtag “nomarchoporlasfarc,” their language, as well as the lack of coherence of their messages made me thought that perhaps twitterbots are already being used for sabotaging deliberation, public discussion, and argumentation. Below are two of the messages I got that day:
@vVvA @AntanasMockus fracaso total marcha por la "vida"no les llego hasta alla la mermelada eso fue #NoMarchoPorLasFarc #NoMarchoPorLasFarc
— Majo (@MariaLacano) March 8, 2015
In this one, the twitter message says that the march was a “total failure” and also makes allusion to the supposedly “mermelada” or payment for the march made by the Colombian government. However, at that time of the tweet the event #vidaMarzo8 was just beginning and growing.
@vVvA @AntanasMockus DIGAN LA NOTICIA COMO ES: POCOS TRABAJADORES DE LA EMBAJADA DE AUSTIN CUMPLIERON CON LA ORDEN DE SANTOS. TRAMPOSOS
— josé rodolfo gomez (@jovamaju) March 8, 2015
In this one, the message talks about a supposedly embassy in Austin, its workers, and an order by the Colombian president Santos. Interestingly, there is not Colombian embassy in Austin, TX, and none of the people that got together for the event had any relationship with the government. We decided to meet just because we thought it was worth to raise awareness about respecting life and we believe in that it is possible to imagine a country without war. It is precisely the president of Colombia, one of the Twitter users who perhaps has to deal with more hate speech. Any post he makes in Twitter is usually replied by several Twitter users that limit their speech to insults and aggressions. Looking at one of his tweets during the day of the “march for live” can serve as example of the kind of “deliberation” that some twitter users want to have. See for instance this reply to one of Santos tweets:
@JuanManSantos CUALES PENDEJO MARCHO POR QUE OBLIGO A MAS DE UNO PERO AMI NO ME OBLIGA GUERRILLERO DE MIERDA, TRAIDOR.
— Opositor de la pas (@guiluga) March 8, 2015
Like this, few minutes after the original post the tweet got more than 20 replies that were full of hate speech, aggression, and offenses. In order to find out if the hate speech replies were done by twitterbots intended to sabotage the event of #vidamarzo8 I run a little test with a software tool developed by the Truthy project and that it let’s you detect to what extent a Twitter user is a bot or not. After using the tool with some of the replies I got to my post, I found out that the aggressive twitter users were not fully bots but had certain characteristics that made them behave as twitterbots (they are in range of 30-45%). They could be human, but the way their interactions happen at the networked public sphere, and more specifically, at the Twittersphere, seems to be shaped by some kind of automated behavior. However, it would be interesting to run the bot-test with other of the users that are constantly shouting hate speech in the Colombian Twittersphere and find out to what extent they are being run by humans or not. To what extent they are just repeating certain kinds of texts and messages in order to sabotage conversations, civic protests, and the discussion of the peace negotiations and a post-conflict scenario.
