Umati: Monitoring Online Dangerous Speech
Umati: Monitoring Online Dangerous Speech
Umati: Monitoring Online Dangerous Speech
Introduction
Hate speech has garnered growing interest in Kenya since the 2007/8 Post Election Violence, in which it seems to have played a role, and because it has been rising again in certain contexts - online for example - in the period leading to our next presidential elections in just over two weeks. In response to this, Umati has conducted a unique, first-ever project to 1) monitor the Kenyan online space for hate speech; 2) analyze the speech for how likely it is to stir violence; 3) find and use non-government ways of countering it. Under Article 13 of the National Cohesion and Integration Act of 2008, a person who uses speech (including words, programs, images or plays) that is threatening, abusive or insulting or involves the use of threatening, abusive or insulting words or behaviour commits an offence if such person intends thereby to stir up ethnic hatred, or having regard to all the circumstances, ethnic hatred is likely to be stirred up. Notably, the Act mentions ethnic hatred only - not hatred based on religion, gender, nationality, sexual preference, or any other group category. Other Kenyan laws also touch on hate speech, in diverse ways. The 2010 Constitution notes that freedom of expression does not extend to hate speech - but does not define that term. Kenyas Code of Conduct for political parties (attached to the Political Parties Act) forbids parties to advocate hatred that constitutes ethnic incitement, vilification of others or incitement to cause harm.
The law is still imprecise, in other words, and there has been an escalating demand from peacebuilding organisations, politicians, government officials and the general public for more detail on how to define, identify, mitigate, report and deal with hate speech. This need motivated the Umati project to facilitate easier identification of hate speech, especially the type of hate speech that has a potential to trigger violence so that the violence can be avoided or diminished. The type of hate speech that has the capacity to
Dangerous speech does not by itself cause violence, but instead has the capacity to promote or inflame violence - even when people are heavily influenced by speech, they are able to resist its power, and are legally and morally responsible if they commit violence.
1.National Cohesion and Integration Act, 2008, Art. 13. http://www.cohesion.or.ke/ images/downloads/national%20cohesion%20and%20integration%20act%20 2008.pdf
Professor Susan Benesch of American University (Washington, DC, USA), an authority on hate speech as a precursor to violence in many countries, defines dangerous speech as speech that has a reasonable possibility of helping to catalyze violence. She has developed a five-point analytical tool for gauging when violence is likely to be stirred up, to borrow language from the National Cohesion and Integration Act, or as Prof. Benesch puts it, for estimating the dangerousness of a particular speech act in the context in which it was made or disseminated (The impact of speech always depends on the context.) These are factors identified by Professor Benesch that make speech more or less powerful: the speaker and his or her influence over an audience (a political, cultural, or religious leader? Someone with a large following of another kind?) the audience and its reasons for taking inflammatory speech seriously (already fearful? receiving information mainly from one source?); the content in the speech that may be taken as inflammatory (serious offense against what is sacrosanct to another community? Referring to humans as pests or vermin?)
the social and historical context of the speech (previous clashes between two groups? Competition over land or other resources? hardship?); the means of spreading the speech, including the language in which it is expressed (mother tongue?) and medium (a radio station, TV network, or blog that, itself, has influence?).
Note that this list does not include the intent of the speaker. Intent must always be considered when defining a crime or building a case for prosecution, but that is not our purpose. Umati aims above all to prevent violence, and we are also strongly dedicated to freedom of speech. Therefore we seek to prevent dangerous speech and violence by mobilizing civil society, not government regulation or prosecution.
The Umati project seeks to identify and understand the use of dangerous speech in the Kenyan online space, in order to find and use non-government ways to reduce its effects of violence on the ground. To this end, we have created NipeUkweli - an outreach effort to debunk inciteful myths and reduce the possible effects of dangerous speech.
It is not the goal of Umati to define the law, or to find and prosecute the perpetrators of dangerous speech. Umati is a civil society project, not a legal or policing body.
While most projects related to hate speech have been looking at mainstream media, we are aware of the influencepositive and negativethat New Media such as the blogosphere and online forums had during the 2007 Post Election Violence in Kenya. Therefore, our flagship project seeks to monitor and report, for the first time, the role New Media plays on a Kenyan election.
*
As of February 2013, we added a sixth monitor from the Somali community.
Cited incidences of hate speech are translated from vernacular to the countrys official language, English. The monitors check blogs written in their vernacular language, blogs in English, Facebook pages and groups, Twitter timelines, online newspapers and video streams of the major media houses in Kenya.
This jigger men will steal votes as the usually do, this time Even Jesus was a Gay. its a matter of personal choice. stylep [tribe]
Tumezoea mawe, tairi, na kelele ni kawaida kwa omena [subtribe] are useless cowards who will never develop
To note here is that some of the statements that fall in the offensive speech category may be very strong and ugly insults, or negative stereotypes that may encourage the audience to hate the target group. Hence, if these statements are repeated by influential speakers, to more vulnerable crowds, they can very easily become extremely dangerous statements, which have the highest potential to ignite violence.
In an effort to avoid fuelling hateful speech, we have deliberately omitted the naming of any tribes, political parties or politicians when writing this report. For example, when we quote statements verbatim from our study, we replace the named tribes with the terms [tribe], [tribe1] etc.
2. Moderately Dangerous Speech Comments in this category are moderately inflammatory and are usually made by speakers with little to moderate influence over their audience. The content of the statements have a mixed effect on the audience; they can be very inflammatory to some, and barely inflammatory to others. Though some can be viewed as very inflammatory, they are grouped in this category because we consider the moderate influence the speaker has over the audience, which is a factor of the little to moderate response the statement received from the audience. Examples: sorry 2 say so [religion] r not peaceful peoples. thy r creatures that should not live near or together with human. look at Asia continent Africa, europ everwhere in world they are trouble makers. something mst be done with this so calld .. Never ever.If anything the [race] who live among us are Worms in our stomachs, Jiggers in our toes and we should not reward them for promoting theft of public resources in the name of fake busineses in the last 50 years WHEN WILL SOME [tribe] STOP CHOOSING TO BE THICK. SHOW US WHY YOU WENT TO SCHOOL YOU FOOLS
In an effort to avoid fuelling hateful speech, we have deliberately omitted the naming of any tribes, political parties or politicians when writing this report. For example, when we quote statements verbatim from our study, we replace the named tribes with the terms [tribe], [tribe1] etc.
a) exacerbating fear in the audience and in this way encouragexample v and vi above);
c) encouraging the audience to harm the targeted group based on often inaccurate beliefs the speaker promotes about the targeted group (example ii above). Comments in the Extremely Dangerous Speech category have the highest potential to catalyse violence as they provide a plan of action that can be well understood and even acted upon by the intended audience (although perhaps not by all readers).
Note that, unlike hate speech, dangerous speech focuses on the speechs effect on the audience and not the the state of mind of the speaker. The effect of speech on an audience cannot be accurately predicted of course, and usually cannot be measured even in retrospect. To identify dangerous speech, we must make a guess about the likely consequences or impact of speech.
Does limiting dangerous speech impede ones freedom of As Umati, we have received several questions concerning hate speech, political speech, dangerous speech and the freedom of speech. This section addresses these questions and others, in order to correctly define what speech is legal, and what speech is not. What is the difference between hate speech and dangerous speech? According to Professor Susan Benesch, hate speech is a large variously defined category of speech that is usually offensive to members of groups it purports to describe, but may not increase the chances of violence being committed against them. Dangerous speech on the other hand is communication that may help catalyze mass violence by moving an audience to condone - or even take part in such violence. What is the difference between political speech and dangerous speech? Political speech and dangerous speech are not mutually exclusive categories. There can be dangerous political speech and responsible political speech. Political speech is dangerous when it calls upon the audience to harm or condone the harm of a political group/class. speech? No. Correctly defining the type of speech that is harmful to the society allows the society to express itself even more freely in the type of speech that is harmless. Due to the potential harm of dangerous speech, limiting it promotes peace in the society and encourages cohesion among various differing groups of people.
10
Does Umati arrest the offenders of dangerous speech? No. Umati is a research project under iHub Research and Ushahidi that tracks the Kenyan online space for hate and dangerous speech.
What does Umati do with the hate speech statements it collects? Umati forwards instances of extremely dangerous speech as well as any calls for action that require intervention to Uchaguzi (uchaguzi. co.ke). Uchaguzi is a multi-stakeholder initiative coordinated through an ICT platform built by Ushahidi (ushahidi.com), which enables Kenyans to keep an eye on the vote and provides avenues through which they can report, with any technology available to them, any incidences significant to the election. Furthermore, Umati qualitatively analyses the data it collects in order to contribute to research in the areas of machine learning, human monitoring, education on ethnic diversity, influence of religion on speech and other research areas.
11
12
Its important to note that an ugly or critical comment about an individual - a politician, for example - is not hate speech unless it targets that person as a member of a group. Hate speech is directed at a group, or at a person as part of a group: a tribe, religion, etc. During election periods, it is not uncommon for negative statements to be made against politicians and other influential personalities. This is a normal part of the political process, as long as the statements do not constitute defamation, threats, hate speech, or dangerous speech.
13
14
ii. Suggest that some people are spoiling the purity or integrity of the group Of the five ethnic communities we are monitoring, all are known to possess certain characteristics and/or perform certain socio-cultural activities that define them, e.g. Luos fish, Kikuyus run businesses, Kalenjins are pastoralists, Somalis trade and Luhyas are farmers. However, in our research we found that commenters use these stereotypes negatively or state other negative stereotypes as truths against these communities. This category contains comments that use historically negative stereotypes to suggest that the targeted community is impure to the audience. Also, comments that we found from our research exhibited calls to remove these impure groups from the society. Examples follow: ..wats wrong wid ths community? God wat r u waitin for wid ths evil,heartles,assasins in kenya? please clear 4 us this whole Gomorra and sodom of kenya we are tired kindly! m3 c mkabila but u [tribe] should b cast uota dis wald,,,,,,,U R SHITTT STUCK IN OUR BUTTS EVRYDAY We dont want madoadoa including [...] COUNTY!! Examples from our data follow, Killing all [political party] leaders is the only way to prevent further loss of innocent lives!!! it wil b either you kil or get killed. iii. Suggest that the audience faces a serious threat or violence from another group Another indicator of a statement has the potential to promote violence is when the statement suggests that the audience should equip themselves because another group will attack them. Often, these comments are not based on truth but are instead intended to invoke fear in the audience so that they can defend themselves against the claimed violence. These statements often promote mistruths against the targeted community so that the audience can move to act against that community in the name of defending themselves. we are beginning to think that this [subtribe] shouldint have been grouped in our comunity they resembles the [tribe] even their language doesnt match ours
*
Picture source: Afromusing on Flickr
15
Moderately dangerous speech, as well as offensive speech, are of concern due to the fact that they fuel negative sentiment that already exists in the audience. Instead of encouraging the audience to live in harmony with other groups, they provide fodder to those that already discriminate the targeted group. This in the long run can result in the audience acting out violently due to these negative sentiments that have been accumulating over time.
16
17
Calls to Forcefully Evict Mixing [tribe1] and [tribe2] will not work,lazima IDPS watoke Rift valley. My dream is to see a [tribe]less kenya ,no aids,no stones no uhiis, kila mtu arudi kwao wageni wengine apana wanatuaribia boma [tribe] WAHAME NYANZA waende... mko wengi na hamna makao shame on you. If mzalendo kibunja wants me, let him arrest me but i think we push this [tribe] out of easleigh and kenya if the government cannot do it for us. We cannot live in fear on our land, and the cause of the fear is a refugee. Calls to Kill They deserve Twa! Twa! Twa! Twa! Twa! Twa! Twa! Twa! wachomwe!wachinjwe! i will get out from it when all $ i mean all [tribe] will be in the grave!!! Bullets need 2 be used in [town1] and [town2] for these goons 2 straighten-up ..... KENYA SI YA [tribe]... this time try you will see we will slaughter you.... A bullet should be put into the skull of this dog called [tribe]....
18
The Umati team encourages the media to steer clear of sensationalist reporting especially around highly volatile national events such as referenda and general elections.
nyatta becomes President The highest contributor to online dangerous speech is the upcoming 2013 general elections and the politics and campaigns surrounding it.
19
Between November and January, Umati saw a declining number of extremely dangerous speech statements (category 3 statements). The drop in December can possibly be attributed to the Christmas holiday break. Additional data from the month of February will need to be analysed in order to conclude that there is a declining rate of dangerous speech as we near elections.
20
Surprisingly, the highest use of dangerous speech from the Kenyan online space that Umati is monitoring is by identifiable commenters. Identifiable commenters are online users who leave comments in response to a Facebook post, an online news article, a forum or blog post. They are identifiable in that they use their own name or a pseudo name. The lack of caution when speaking online suggests that the speakers are not considering the negative impact their statements could have,nor are they worried about being associated with the dangerous statements they make.
21
Constantly across each month, from October to January, the most frequent call to action is the call to discriminate other groups.
22
Also, if we look at the same data minus the calls to discriminate, calls to kill stand out prominently more than other calls to violence, as shown below.
This highlights an interesting behaviour in dangerous speech conversations; calls to discriminate which are the least harmful amongst the calls to action, are the most frequent in the online space, while calls to kill, which are the most harmful calls to action, are the second most frequent calls to action.
23
24
25
26
What can YOU do? III. Empower the audience to be immune to incitement
The Umati team favours this approach as it gives power to the audience. By defining what dangerous speech is, two goals are met: - the mwananchi (citizen) is able to identify which comments/ statements are dangerous and is then able to react responsibly to these statements. - by educating the public on exactly what kind of speech has harmful effects on the community, the public is then able to freely engage in the speech that is not harmful; defining and empowering the public to correctly identify dangerous speech, which further increases their freedom of expression. Umati does this by conducting outreach events with online content creators and the general public in order to promote education on dangerous speech. There are several ways one can choose to react to dangerous speech: - the audience can choose to ignore the statement; - the audience can choose not to react to the statement; - the audience can in turn educate the speaker to engage in speech that is not dangerous; - the audience can offer correct information so the others are encouraged to react peacefully. Such responsible online activity was exemplified during the Mombasa violence that followed the death of Muslim cleric Sheikh Aboud Rogo, when inflammatory tweets were being spread that stated that a Mombasa church was being burned. A responsible social media user took a tweetpic of the church (which was not burning) and stated, Stop the lies!. This responsible action helped to quell the propogation of such inflamatory lies on social media. Such spreading of correct information has the potential to lead the audience to react peacefully to an utterance intended to incite them to violence. NipeUkweli (Give me Truth) is an outreach initiative calling for communities (bloggers, community radio, traditional media, and grassroots groups) to help discredit the lies and rumors expressed both online and offline.
Act now!
Notes
Discrimination: Discrimination is understood as any distinction, exclusion or restriction made on the basis of race. colour, descent, national or ethnic origin, nationality, gender, sexual orientation, language, religion, political or other opinion, age, economic position, property, marital status, disability, or any other status, that has the effect or purpose of impairing or nullifying the recognition or exercise, on an equal footing, of all human rights and fundamental freedoms in the political, economic, social, cultural, civil or any other field of public life. Source: La Rue, F., (2012, September 7). Report of the Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression, UN Doc A/67/357, p12. Dangerous speech: This is a term coined by Professor Susan Benesch to describe incitement to collective violence that has a reasonable chance of succeeding, in other words speech that may help to catalyse violence, due to its content and also the context in which it is made or disseminated. This possibility can be gauged by studying five criteria that may contribute to the dangerousness of speech in context: the speaker (and his/her degree of influence over the audience most likely to react, the audience (and its susceptibility to inflammatory speech), the speech act itself, the historical and social context, and the means of dissemination (which may give greater influence or force to the speech). [2] Constitution of Kenya, Art 33(2)(c). [3] Political Parties Act of 2011, First Schedule. SOURCES
Identifiable Commenter: A person who responds to an online article, blog post or Facebook post who can be identified by a name, regardless of whether the name is real or fake.
[1] National Cohesion and Integration Act, 2008, Art. 13. http:// and%20integration%20act%202008.pdf
27
www.cohesion.or.ke/images/downloads/national%20cohesion%20
[4] Mitchell, N. First we call them insects: the prelude to horror 26th April 2012. Available at http://www.abc.net.au/news/2012-04-26/ mitchell-first-we-call-them-insects/3969192 For further information and articles on the hallmarks and on Dangerous Speech generally, see www.voicesthatpoison.org
Umati Project Team iHub Research Nairobi, Kenya [email protected] www.research.ihub.co.ke | Twitter: @iHubResearch