ELITE COMMUNICATION AND LEGITIMIZATION OF VIOLENCE DURING INTERGROUP CONFLICTS
By
Sonia Jawaid Shaikh

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Communication—Master of Arts
2014

ABSTRACT
ELITE COMMUNICATION AND LEGITIMIZATION OF VIOLENCE DURING INTERGROUP CONFLICTS
By
Sonia Jawaid Shaikh
This thesis investigates universal themes in the violence-legitimizing communication of elites
during inter-group conflicts. A comparative content analysis was conducted across
communication from elites who have been internationally recognized as having played some
kind of leadership role for in-groups that have committed acts of violence against out-group(s).
Data stretching across two decades from four different countries were examined; including the
former Yugoslavia, Rwanda, Sri Lanka and Myanmar. A thematic coding scheme was
developed, using Susan Benesch’s proposal on dangerous speech and prevention of group
violence and perspectives from Social Identity Theory as the theoretical groundwork. Coding
and analysis revealed that across countries and temporal spans, elites bolster in-group and outgroup distinctions through their language choices, but typically avoid outright derogation.
Moreover, they commonly create non-falsifiable facts about current, past, or future events; and
employ allegations against the intentions, plans and activities of the out-group. Several additional
themes were identified as well. Implications for future research and theory examining
antagonistic public communication are discussed.

Copyright by
SONIA JAWAID SHAIKH
2014

To them. The survivors, victims and true victors of mass violence and genocide.
And
To Pappa, for his insight & Mamma, for her foresight.

iv

ACKNOWLEDGEMENTS

Although copyright laws protect authors and give them sole ownership of their works, but
in essence a manuscript is always an amalgamation of ideas, suggestions, guidance and
challenges put across to a writer by others. I could not have done this research without the help
of so many individuals who impacted my journey so positively. My sincere and deep gratitude
extends to Dr. Timothy R. Levine for being a great teacher, friend, mentor and being there since
my day one at Michigan State University. I am particularly thankful to the committee members
Dr. Stephen Lacy and Dr. Daniel Bergan for putting in their valuable time, commitment and
suggestions during the course of this thesis.
I have great appreciation and gratitude for the Fulbright program which brought me to
Michigan State University for which I am thankful. Here, several teachers within the Department
of Communication and outside have helped me shape my thinking and develop a perspective. I
remain grateful to them. To my friends, dear roommate and buddy Yongrong Shen, and
classmates who lent listening ears and their time to challenge and support me: bohat shukriya!
This thesis required collaboration of translators and I was fortunate to work with Ms.
Vidya Mohan at Less Commonly Taught Languages Program on campus and an anonymous
translator. I am thankful to them for helping me on time and with efficiency on translations and
accuracy checks.
Finally, I wish to extend my immense gratitude and appreciation for my advisor Dr.
Steven McCornack. He has been a tremendous support, the better thinker, certainly the better
editor and above all a true mentor from the start of this project. I owe any finesse in this draft to
him while all the errors and mistakes remain truly mine.

v

TABLE OF CONTENTS

LIST OF TABLES……………………………………………………………………………...viii
LIST OF FIGURES………………………………………………………………………………ix
WORDS AND ACTIONS: AN INTRODUCTION…….…………………………………...……1
LITERATURE REVIEW……...………………………………………………………………….4
Key Terms and Definitions……..……..…………………………....……………………..4
The Four Countries of Analytical Focus…………………………………………………..5
The genocide in former Yugoslavia…………………………………………….6
The genocide in Rwanda………………………………………………………..7
The Tamil separatist movement in Sri Lanka…………………………………..8
The rise of extremism in Myanmar……………………………………………..9
Communication during Inter-group Conflict……………………………………………...10
Myths and symbolism………………………………………………………….10
Salience of identity and name calling………………………………………….11
Media as elites during conflicts……………………………………………..…12
Theoretical and Analytical Frameworks…………………………………………………..13
Benesch’s framework on dangerous speech……………………………………14
The speaker…………………………………………………………14
The audience………………………………………………………..14
The speech act………………………………………………………15
The socio-historical context………………………………………...15
The mode of transmission…………………………………………..15
Social identity theory………………………………………………………...…15
RATIONALE, RESEARCH QUESTIONS AND METHODOLOGY…………………………17
The Chosen Elites……………………………………………………………….18
Ashin Wirathu………………………………………………………..18
Radio Télévision Libre des Mille Collines (RTLM)………………...18
Vellupillai Prabhakaran……………………………………………...18
Slobodan Milosevic………………………………………………….18
Coding and Analysis……………………………………………………………..20
Defining differences and/or conflation of the in-group identity
with the national or supra-ethnic identity……..…………...………..21
Call for duty/exhortation……….……………………………………22
Meta-out-group connections………………………………………...22
Claims of victory – “we win, they lose”…………………………….23
Anecdotes……………………………………………………………23
Fact-making and allegations………………………………………...23
Importance of land/territoriality...…………………………………...24
Name calling…………………….…………………………………..24
vi

Historical/present injustice or victimization……………………..…..25
RESULTS………………………………………………………………………………….….…26
DISCUSSION……………………………………………………………………………,……...28
Limitations and Implications for Future Research………………….…………...34
APPENDICES..………………………………………………………………………………….35
Appendix A………………………………………………………………………..36
Appendix B………………………………………………………………………..38
Appendix C………………………….…………………………………………….46
REFERENCES…....………………………………………………………………………...…...47

vii

LIST OF TABLES

Table 1: Sample………………………………………………………………………………….19
Table 2: Themes by each speaker………………………………………………………………..27
Table 3: Elite communication themes during intra-state, inter-group conflicts from
the literature……………………………………………………………………………………...36
Table 4: List of speeches………………………………………………………………………...46

viii

LIST OF FIGURES

Figure 1. Total number of themes for all speakers………………………………………………26

ix

WORDS AND ACTIONS: AN INTRODUCTION
The relationship between words, contexts, and their impact upon people has long been a
subject of investigation within the social sciences and humanities. In particular, scholarly
attention has focused on the use of speech during conflicts. This domain has received attention
not just because of its practical importance, but because of the legal need related to international
law; that is, specific trials and investigations in which individuals are accused of inciting hatred
and violence against others during conflicts (See Benesch, 2008). International and local
tribunals routinely indict individuals for inciting violence through communication.
The scholarly literature examining the relationship between communication and violence
blossomed following World War II (focusing primarily upon German propaganda). Following
the genocides in Rwanda and Serbia during the early 1990s, an additional body of literature
emerged on the rhetoric that helped incite these genocides. But whereas critical-rhetorical and
communication scholars have sometimes examined particular, individual, political speeches
occurring during these periods, rarely have scholars have focused upon thematic similarities in
the discourse of empowered individuals across speeches, cultures, and temporal spans.
Moreover, careful comparison of speech themes across countries and different instances of
genocide has almost never been done1. A majority of existing literature on communication
during violence does not appear to have a clear systematic analysis of messages (e.g., speeches
or broadcasts), but rather, critical assessments of scattered political narratives. This may be
because many scholars who did such studies herald from disciplines other than communication,
such as History or Political Science (e.g., Kaufman, 2001; Mann, 2005 and Zubero 2012).

1

For an exception see Morus (2000). Her dissertation compared rhetoric from Hitler, Pol Pot and Rwanda.
1

The failure of scholars to examine consistent speech themes during times of genocide is
remarkable, from both theoretical and practical perspectives. In terms of communication theory,
it would seem essential to have a deeper understanding of the particulars of messaging that might
play at least a partial causal role in triggering subsequent violence. From a practical perspective;
speech acts during conflicts may have two consequences: firstly, they might act as precursors to
violence and in extreme cases genocide. Secondly, certain groups may bear psychological or
social harms such as humiliation and fear. This makes it pertinent to understand the dynamics of
such dangerous speech, so that a timely identification can lead to important preventive or
controlling measures, such as inhibiting speakers and/or inoculating audiences (Benesch, 2012).
Further, in an op-ed written by Benesch and Abramowitz in the The Wall Street Journal, the
authors suggest that inflammatory speech can be a “prelude to murder.” This is not hyperbole,
given the evidence of violence in Kenya, the former Yugoslavia, and Rwanda following the
public dissemination of such speech (2013). Thus, it is both theoretically and practically
important to investigate thematic patterns within such speeches.
This thesis investigates the topical themes present within the speech of empowered
individuals during various periods of genocide and mass violence, both past and present. In
doing so, Susan Benesch’s framework on dangerous speech from the realm of international law
(2012), and Tajfel’s Social Identity Theory (1981) will be used to develop a coding and
classification scheme. My focus will be upon public speeches and broadcasts made by elites in
four countries: Myanmar (Burma), Rwanda, former Yugoslavia and Sri Lanka. I examine the
presence of common themes (or lack thereof) which legitimized violence in elite communication
in geo political entities where incidences of intra-state, inter-group conflicts occurred over time.
From this examination, I create a taxonomy of themes prevalent during such conflicts.

2

In the remainder of this thesis, I shall do the following. First, I will detail the four
countries which comprise the focus of this research, and their relevant histories. Second, I
review existing literature focusing upon communication and incitement against out-groups.
Third, I present my rationale and research questions for the thesis. Fourth, I discuss my method.
Fifth, I detail the results of my analysis. Finally, I conclude with a discussion of the implications
of my analysis, along with limitations of this study.

3

LITERATURE REVIEW
This thesis is an attempt to investigate themes which are present in elite communication
during inter-group, intra-state conflicts across cases and decades; with a special emphasis on
legitimization of violence. In this section, I first delineate basic concepts and key terms. Then I
discuss inter-group conflict, and provide brief backgrounds on my four countries of analytical
focus (Myanmar, former Yugoslavia, Rwanda and Sri Lanka) where such mass violence amongst
different identity-based groups has occurred. I then describe findings in the literature on the
nature of elite rhetoric, and the role of media in elite communication. I close with a description
of Susan Benesch's framework titled: Dangerous speech: A proposal to prevent group violence
and concepts from Social Identity Theory which will guide my analysis.
Key Terms and Definitions
In this thesis, I conceptualize “inter-group conflict” as conflict between nations, ethnic
and racial groups. Certainly, conflicts also exist in smaller groups, such as organizations, teams
and clubs; but those types fall outside the domain of my focus (Ashmore, Jussim and Wilder,
2001). In the context of this analysis, only identity-based groups such as religious, ethnic, and/or
linguistic groups within a single geopolitical entity (intra-state) are under investigation and
national or inter-state group conflicts are not considered. For instance, conflict amongst
Bosniaks, Serbs and Croats during the early 1990s in former Yugoslavia is an example of intrastate, inter-group conflict. Because most inter-group conflicts involve identity or ethnicity based
groups, the terms “inter-group” and “inter-ethnic” are used interchangeably in this thesis.
The conflicts that comprise the focus of this investigation have similar characteristics.
First, a hierarchal pattern exists in the conflict, whereby certain elites (e.g., rebel leaders or
presidents) spearhead a certain cause. Second, ordinary people volunteer to commit acts of

4

violence against other groups in large proportions. For instance, many perpetrators during the
genocidal conflict in former Yugoslavia were, in fact, volunteers who were not part of the
government or the armed forces, but who took part in the violence nonetheless (Kirschke, 2010).
Third, the concept of rules and ethics of war are absent, and there is usually systematic
extermination of out-group members, including women, the elderly, and children (Kaufman,
2001).
My specific analytical focus is upon “elite communication” during such conflicts. By
“elite,” I mean leading individuals and organizations which spearheaded or represented some
identity-based groups, and who were accused publicly of involvement in conflicts. By
“communication,” I am focusing upon public forms of communication, such as broadcasts and
speeches that were widely available to audiences during these conflicts.
The Four Countries of Analytical Focus
Myanmar, Rwanda, former Yugoslavia and Sri Lanka are four different countries in
terms of culture, languages, geographical regions, and history. But they have a tragic and
terrible characteristic in common: Violent inter-group conflicts. In the past forty years or so,
these states have had remarkably parallel periods of intra-state, inter-group violence; during
which ordinary citizens belonging to different ethnic groups and under the leadership of certain
elites have taken up arms against each other to kill, loot and destroy – resulting in the deaths and
forced expulsions of millions. These countries were chosen as foci because of the following
reasons: a) instances of inter-group but intra-state conflict b) mass violence, expulsion or
genocide c) identity-based groups involved but not more than four groups within each country d)
accessibility of data and e) conflicts and relevant elites known to publics outside of these
countries.

5



The genocide in former Yugoslavia. Europe’s big tragedy after World War II was the
genocide in the former Yugoslavia where thousands of people were killed and horribly
affected in a civil war that betrayed the continent’s confidence about lasting peace. The
former Yugoslavian federation comprised various linguistic and religious groups who
shared Slavic origins. The federation had geographic prowess in Eastern Europe, with
close ties to the former Soviet Union. The president of the federation, Josep Tito,
somehow kept various groups together, despite the growing nationalism and loss of
power that resulted in the formation of several nation states following World War II.
Not long after Josep Tito's death in 1980, tensions in the Yugoslavian federation started

to emerge, as Croatia and Macedonia began to flex muscles of independence. The election in
1990 gave the nationalists victory which allowed Croatia, Slovenia and Macedonia to declare
independence in 1991. A year later, the Bosnian prime minister also moved for separation; and
Bosnia was recognized by the European Union and the United States. However, the Bosnian
Serb leader Slobodan Milosevic emerged as a leader to Serbs living in Bosnia and in the
federation, making him the biggest opponent to calls for successions.
Beginning in 1992 and ending in 1995, mass violence and genocide plagued the
Yugoslavian region; where an estimated 100,000 people (majority Bosniaks) were killed (Tabeau
& Bijak, 2005). Several thousands were made homeless, expelled or raped. The war engulfed
three main parties – Croats, Serbs and Bosniaks – and Bosniaks suffered the heaviest losses.
The war also was disastrous for western diplomacy. First, the UN-Dutch peace-keeping
mission was criticized severely for failing to keep Bosnian refugees from harm during the
Srebrenica massacre. An estimated 8,000 men and boys were systematically removed by Serb
forces, shot dead, and thrown in mass graves; while peace keepers entrusted with their protection

6

were on duty. What’s more, several rounds of peace talks moderated by Western powers failed,
leading to NATO allied bombing of the Serbian city of Belgrade. Although this bombing ended
both the conflict and Serbian control, it also splintered the ties between those linked with the
former U.S.S.R and the West.


The genocide in Rwanda. While the world reeled from the Bosnian genocide, another
tragedy broke out soon after in Rwanda. Mahmood Mamdani, in his historical account of
the country and the genocide that unfolded in early 1994, suggested that the colonization
in Rwanda created distinct identity patterns for the two main groups – Hutus and Tutsis –
with the former as an "indigenous majority" and the latter as an "alien minority" (pp. 103,
2001). This distinction, in his opinion, is a direct consequence of the colonial Christian
influence which perceived Tutsis as more “European” and “Hammitic” in origin; whereas
the Hutus were considered “subservient” by virtue of their darker skin tones.
Rwanda was originally a German colony and at the start of the twentieth century, the

power was handed over to Belgium. Under the Belgian rule, people in Rwanda were issued
identity cards based on their ethnic group; thus creating a politicized identity in black and white.
Also under the colonial rule, Tutsis remained in power – controlling agriculture and local politics
– while Hutus remained under-privileged. Mamdani tracks Rwanda's transition of
decolonization, where strains occurred within the colonialists and the colonized. In 1959, the first
social revolution started to sweep the country; with an emerging Hutu elite demanding liberation
from both the white power holders and the Tutsi. Although this movement successfully
dismantled Tutsi control over local politics; it also solidified Hutu-Tutsi tensions.
Post-colonization Rwanda saw remarkable shifts in power, as Hutus began to dominate
the local politics and educational systems -roles traditionally occupied by the Tutsis. Moreover,

7

Tutsis started to receive the “alien minority” treatment; becoming the recipients of discrimination
and hostility. In 1994, a plane carrying the Rwandan president (who was Hutu in ethnicity) was
shot down in neighboring Burundi. The Hutu-led government blamed the Tutsis rebels for the
incident, and ultimately considered all Tutsis as responsible. Thousands of volunteers joined
hands to eliminate the “Tutsi problem” – resulting in the deaths of almost 800,000 people (also
among them moderate Hutus). In a just a few days, entire villages were burnt or razed to the
ground; with mass expulsions, killings, rapes and looting which was targeted mostly at the Tutsi
population. The Rwandan genocide is considered the worst event of mass violence in Africa.


The Tamil separatist movement in Sri Lanka. Like its other South Asian neighbors,
Sri Lanka was a British colony from 1833 to 1949. During the British rule, many Tamil
speakers from the Indian south were brought to the Sri Lankan north to work on coffee
and tea plantations. Close proximity to India's southern region, which was also governed
by the British, allowed an influx of missionaries who formed several educational
institutes in the predominantly Tamil north. This subsequently allowed a high number of
Tamils to be employed in the British civil service.
After Sri Lanka's independence in 1949, many Indian Tamils were left disenfranchised

and lost their citizenships, as the country's political landscape began to be colored by Sinhalese
nationalism, including efforts such as making Sinhalese the official language. In 1958, antiTamil riots erupted; killing 200 Tamils and displacing thousands. The ethnic tensions continued
into the late 1960s and early 1970s, when a movement was started by Sinhalese Marxist students
and Sri Lanka was adopted as the official name of the country; further intensifying TamilSinhalese tensions. In 1977, riots left 100 Tamils dead following elections which allowed a
Tamil party to win all of the seats in the north of the country.

8

In 1976, the Liberation Tigers of Tamil Eelam (LTTE) emerged as the rebel movement
against the Sinhalese government from the northern and eastern parts of the country, under the
leadership of Vellupillai Prabhakaran. In 1983, the militia ambushed 13 soldiers which resulted
in the deaths of hundreds of Tamils. This incident is called The First Eelam War. In its efforts
against the state, LTTE used several guerilla warfare tactics, such as suicide bombings and the
use of human shields, in its armed struggle with the government. The leader Prabhakaran became
a wanted man not only in Sri Lanka, but also internationally which resulted in his continued
hiding. However, every year, Prabhakaran would give a Heroes’ Day speech which celebrated
Tamil martyrs and charted the militia's stance on politics.
The civil war in Sri Lanka ended when Prabhakaran was killed by the Sri Lankan
government in 2009 and a ceasefire was announced all over the country. Surprisingly, very little
research has gone into Tamil rhetoric and violence, despite the nature of the conflict being large
scale and terribly taxing on respective groups. The extent of the damage included destruction of
almost half of the Sri Lankan airlines fleet, two heads of state assassinated, mass expulsions, use
of humans as shields, suicide bombings, and killings.


The rise of extremism in Myanmar. A minority group called the Rohingyas live in the
Western part of Myanmar that borders with Bangladesh. This area is called the Rakhine
state, and its people are considered one of the world's most persecuted minority by the
UN (BBC, 2014). The Rohingyas are stateless, and are not accepted by either Bangladesh
or Myanmar. In recent years, they have been severely affected by inter-group conflict
and tension, primarily because of state policies and rhetoric of a renegade Buddhist monk
named Ashin Wirathu who is leading a campaign against them. According to unofficial
figures, thousands of Rohingyas have been displaced and hundreds have been killed.

9

Recently, the government of Myanmar ordered an international NGO to leave the
Rakhine state; which means that thousands of Rohingyas are now deprived of basic
medication and resources (BBC, 2014).
Wirathu is considered extremely controversial, has been branded as The face of Buddhist
terror by TIME (Beech, 2013). Wirathu has proclaimed himself "The Buddhist bin Laden," and
the Rohingya conflict is still ongoing.
Communication during Intergroup Conflict
The scholarly study of elite communication during inter-group conflict has focused
primarily upon large-scale conflicts which have received global coverage or greater media
attention. The exemplar for this trend is the case of the Nuremburg trials, which indicted several
Nazis including the German Propaganda Minister Josef Goebbels. In the mid 1990s, the study of
conflict communication garnered new interest following the genocides in Rwanda and the former
Yugoslavia. Some of this attention has been in the form of critical analyses focusing upon
individual speeches, and the use of certain rhetorical themes by individual speakers. A bulk of
commentary or any ideas on elite communication during conflicts stems from disciplines such as
Political Science, Anthropology and History.
In the subsequent paragraphs, I will elaborate some common themes found by historians
and communication scholars in elite communication and the role of media as elites.


Myths and symbolism. Historians, critical and rhetorical scholars have suggested several
characteristics that typify the communication of elites during intergroup conflicts, such as
myths - which are modern versions of old stories or stories created a new, ethnic
symbolism, narratives and rhetoric to fuel hatred and recruit volunteers (Bozic-Roberson,
2004; Kaufman, 2001). For instance, Serb nationalists in their rhetoric against other

10

ethnic groups created a myth of a battle fought between Bosniaks and Serbs in the
fourteenth century in Kosovo, which then was used to claim that Serb lands were invaded
by Turks and made them (Serbs) into victims of violence (Mann, 2005). However,
scholars believe that this was a false account used in war rhetoric; designed to create
“politicization of ethnicity” – the efforts made by then Serbian president Milosevic to use
ethnic identity as the basis of argumentation and identification in his political narratives
(Bozic-Roberson, 2004; Mann 2005)
Some work on rhetoric outside of Rwanda and former Yugoslavia has looked into elite
communication during violent conflicts. Roy and Rowland (2003) found in Hindu rhetoric
accusations against Muslims and claims that Muslims are a minority with too much influence.
They also suggested that Hindu rhetoric appeals to religious sentiments, by invoking myths and
the promise of saviors. Zubero (2012), in his critical examination on Rwanda and Germany,
argues that elites make strangers of out-groups. Zubero’s work is not a systematic analysis of
either Rwandan or German communication, and does not specify any particular messages of how
it happens. But he does suggest that in the context of Nazi Germany, Jews were called out as
Jews despite being citizens of Germany, which was an attempt to differentiate them something
other than German.


Salience of identity and name calling. Focusing on role identity and redemption, it has
been suggested that former Serbian leaders rose to power through their use of rhetoric
and politicization of ethnicity well before the conflict started (Bozic-Roberson, 2004).
Morus, in her work on why violence may have been normalized in former Yugoslavia,
argues similarly to Michael Mann; suggesting that Milosevic’s speeches focused on
myths and the exclusiveness of the Serbian people (2007). Politicization of ethnicity or

11

identity salience also was found by Donohue (2012) in his framing study of Radio
Télévision Libre des Mille Collines (RTLM) broadcasts in Rwanda. Donohue also noted
that humiliating an out-group by calling them names such as ‘cockroaches’ was one of
ways RTLM communicated to Hutu audiences.
Studying elite communication during intra-state, inter-group conflict from a comparative
perspective is rare and Morus’ dissertation on rhetoric in genocidal regimes in Rwanda,
Cambodia and Nazi Germany is an example of it. She finds similarities in language, such as
hailing a savior, dehumanization, and de-individuation of an out-group; and portrayals of ingroups as victims of historical injustices that must be addressed to remove any present danger
(2000).
A different approach to understanding hate speech was undertaken by Glaeser, who
developed an economic model of political hatred. He suggests that activities pertaining to
monetary and economic benefits are at top of the political agenda and politicians use hatemongering to attain some kind of material gains. These messages include blaming an out-group
for some economic grievances or lack of resources (2005).


Media as elites during conflicts. The role of elites is not limited to individuals making
public speeches. Instead, the media forms an integral part of elite networks; as evidenced
from work in Africa, particularly Rwanda and former Yugoslavia. In an examination of
Serbian communication during the civil war, Bozic Roberson (2004) and Kissoupoluos
(2007) suggested that Milosevic was very efficient in manipulating media (such as TV
and newspapers) to create ethnic problems and gain power. Li (2004) also suggested that
radio in Rwanda along with other factors was used to pursue Hutu ambitions against
Tutsis. Only Strauss remains an exception, in his argument that the role of the RTLM in

12

the genocide was not as significant as suggested, and that the radio motivated only a few
individuals (2007). However, a powerful analysis made by Yanagizawa-Drott (2010)
shows that number of broadcasts received by users in Rwanda correlated highly with the
number of killings within particular areas. The role of the media in genocides is not only
limited to Rwanda and the former Yugoslavia. Recent work on the role of radio and
violence in Kenya has also come to the surface; suggesting that although the intensity of
rhetoric cannot be compared to RTLM, the radio did play some role in inciting violence
(Beyene, 2012).
Comparative work in understanding the role of the media during inter-group violence is
not prevalent either; but Beyene (2012), Frère & Marthoz (2007) and M’Bayo (2005) are a few
notable exceptions. Beyene, in her dissertation, did a comparative study of the Kenyan and
Rwandan media’s role to investigate how types of media can exacerbate ethnic violence. Her
research suggested that countries with weak or absent institutional regulatory mechanisms have
media affecting ethnic violence. Frère & Marthoz (2007) and M’Bayo (2005), in their works on
conflicts and media in Africa, focus on factors such as the lack of journalistic training, and
political and institutional factors in the media’s role in violence.
Based on the literature review above, themes such as identity salience, out-group
dehumanization, ethnic polarity, myths and stories can be extracted (amongst others) in the
rhetoric of elites in incidences of intra-state but inter-group conflicts2.
Theoretical and Analytical Frameworks
In the following paragraphs, I will delineate an analytical framework and theory of relevance
which can help explain elite communication during conflicts.

2

Appendix A contains a list of studies and the themes extracted from them.
13



Benesch's framework on dangerous speech. Susan Benesch is a scholar who is
associated with the World Policy Institute and who works within areas of conflict
communication. In 2012, she presented an analytical framework to the institute, which is
a recent attempt at systematically identifying what constitutes “dangerous speech”
(2012). By “dangerous speech,” Benesch refers to what is commonly termed “hate
speech” that is, communication "... that denigrates people on basis of their membership in
a group such as an ethnic or religious" (pp.1). She believes that the importance of speech
and its effect on persuasion have been studied previously, but not in the context of
violence. The capacity of a communication acts "to inspire action" in contemporary
times, the international law implications for distinguishing dangerous communication
from free speech, and concern for preventing future violence, all guide her work. Her
proposal is available online on the institute's website.
In Benesch’s framework, A proposal to prevent dangerous speech, she identifies five

key variables, of which one or all may be utilized by a speaker. With these areas, she delineates
questions which help assess the extremity of a speech. Benesch believes that although these areas
are important; not all may be factors in determining consequences. For instance, in some cases
audience vulnerability due to grievances can play a bigger role in triggering violent outcomes.
1. The speaker. This refers to the speaker and his/her position with
respect to the group. Benesch urges the reader to think about his/her identity,
prominence, and influence as relevant factors in dangerous communication.
2. The audience. The audience at the receiving end of such
communication is important to understanding the effects of dangerous speech.
The proposal recommends understanding whether the speaker is communicating

14

with his/her group or all groups at large. If the speech is directed to his/her own
group, then it certainly contains more hazardous implications.
3. The speech act. This variable looks into the speech act itself – that
is, if it contained any humiliating terms for the out-group; a direct call for
violence; elliptical language and any fear-mongering or victimization.
4.

The socio-historical context. In order for a speech act to trigger

violence, the social historical context is also important; such as, previous
contentions between groups, weak governance, and any recent violence following
speech.
5. The mode of transmission. Media outlets which are without
competition, repetition of messages, the use of music, and inaccessibility of
alternative opinions also can intensify a speech.
Benesch’s work is an important addition particularly in international law and genocides.
Yet, to date, her work has not been put to the test on the speeches which were actually made
during inter-group conflicts. For the purpose of this thesis, the variable of “The Speech Act” will
be investigated rigorously through thematic analysis of elite communication.


Social identity theory. This theory suggests that the part of an individual's self-concept
which is derived from group membership is called “social identity” (Tajfel, 1981). An
integral part of this theory is the impact of social identity on an individual's emotional
and behavioral levels. Group membership may be highly relevant to create a sense of
attachment and mutual fate amongst people; thereby urging them to act in certain ways.
This theory has already been put to the test in order to understand the words and actions

of elites and volunteers in several inter-group conflicts; particularly in Rwanda and the former

15

Yugoslavia. For instance, Donohue (2012) and Morus (2000) used aspects on categorization,
social and collective identity from the theory to understand communication in Rwanda and the
former Yugoslavia, respectively. They explicated framing techniques such as "us vs. them;" ingroup and out-group differences, messages of mutual fate and dependence; use of name calling;
and pejorative statements by elites. The theory has been used extensively in the fields of social
psychology and communication, and this makes it suitable to understand arguments made within
contexts of inter-group conflicts.

16

RATIONALE, RESEARCH QUESTIONS AND METHODOLOGY
The importance of understanding what elites say in times of conflicts is not merely a
scholarly exercise, because its implications are far-reaching; such as in areas of international law.
After the genocides in Rwanda and the former Yugoslavia, several people were put on trial for
incitement to genocide. Media elites such as the founders of the RTLM Ferdinand Nahimana and
Jean-Bosco Barayagwizan, and Hassan Ngeze – the editor of Kangura (wake others up!) – were
found guilty of those charges during the trials at the International Criminal Tribunal for Rwanda
(Kagan, 2008). It seems pertinent to study the argumentation forms that are pervasive in elite
communication during conflicts in a comparative fashion, so that macro structures can be
identified.
Most of the existing literature on communication during conflicts describes and
comments on elite communication that is employed to attain certain goals during conflicts. But
the literature does not examine universality in the themes of elite communication during intergroup, intra-state conflicts. Instead, there is a tendency to: a) focus on only one conflict at a
time; b) rely solely on critical interpretation to analyze the content of speeches or broadcasts; c)
use very small samples such as between one to three speeches/broadcasts at a time; d) investigate
mostly infamous and well-known conflicts such as Rwandan or Bosnian genocides in isolation;
and e) compare elite communication in terms of critical theory rather than speech or
communication analyses (for e.g. Kaufman, 2001; Zubero, 2012). In light of this, it becomes
pertinent to inquire if there is any universality to elite communication:
RQ1: What themes are common to elite communication during intra-state and
inter-group conflicts across different cases and decades?

17

The Chosen Elites
Based on the literature and involvement of key figures, the following elites have been
identified as important and relevant to this thesis.
1. Ashin Wirathu: He is a renegade Buddhist monk in Myanmar who has explicitly called
for removal of the Rohingya Muslims and initiated a campaign that encourages followers
only to trade with certain Buddhist factions in the country. Wirathu is currently active in
his anti-Rohingya campaign and his speeches are available online in Burmese.
2. Radio Télévision Libre des Mille Collines (RTLM): Radio broadcasts in Rwanda were
one of the most successful means of communicating with militia and identifying Tutsi
targets. New research indicates that the extent to which radio waves were received in an
area is strongly correlated with the number of people that were killed in that area
(Yanagizawa-Drott, 2012). During the conflict, RTLM was controlled by Hutu
sympathizers and labeled Tutsis as inyenzi or ‘cockroaches’ (Donohue, 2012). Several
broadcasts by the RTLM were made in French and Kinyarwanda.
3. Vellupillai Prabhakaran. Prabhakaran was the leader of the Liberation Tigers of Tamil
Eelam or LTTE; a Tamil militia that claimed to fight for the rights for Tamils from the
majority Sinhalese population in Sri Lanka. For nearly four decades, the Sri Lankan state
was at war with the Tamil rebel groups. Prabhakaran was killed in 2009. His speeches
were made in the Tamil language.
4.

Slobodan Milosevic: Milosevic was the president of former Yugoslavia and gained
power after the death of Josep Tito. Milosevic identified himself as a Serb and purported
to claim the rights of Serbs when Bosnia and Croatia sought independence. Milosevic

18

was later apprehended and accused of crimes against humanity. His public addresses
have been recorded in Serbian.
The content under review will be speeches made by the above-mentioned elites, and these
speeches will form the units of analyses. The availability of speeches or broadcasts varies greatly
from one conflict to another. For instance, the number of recorded and available broadcasts is
much higher for the RTLM than for Ashin Wirathu's speeches. I think that this could be due to
several reasons: a) the speeches were censored b) they were never recorded c) criminal or
judicial inquiry or time lapse may have made more records publicly available d) the speeches
may only be distributed through local media, and e) the use of local language may hinder
accessibility in English.
The sampling of speeches for this thesis, therefore, depended heavily on the availability and the
translation of speeches
Table 1. Sample.3

Source/Speaker

Number of speeches

Languages

Ashin Wirathu

Two

English (originally in
Burmese)

RTLM

Six

English (originally in French
and Kinyarwanda)

3

Slobodan Milosevic

Three

English (originally in Serbian)

Vellulipai Prabhakaran

Six

English (originally in Tamil)

Total

Fifteen

See Appendix C for details on dates of speeches/broadcasts and their media.
19

In this comparative analysis, I expect to find the following themes: Evidence of in-group
vs. out-group argumentation such as the speaker addressing to his group. Moreover, I believe
that reduction of an out-group to sub-human or evil forms may be found. I think that speakers
may also present "facts" which may be difficult for audiences to verify or cross-check and may
appear to be correct on the surface. These "facts" may be about wrongdoings attributed to the
out-group, some historical accounts, or the fight put up by the in-group against alleged injustices.
Speakers may also resort to suggesting the danger that the in-group faces in the future
(particularly in relation to others), or how much they have already suffered. The elites can also
be praising the sacrifices of the in-group, exhorting others to follow the suit, or offering new
instructions and/or showing some effort for group solidarity.
Coding and Analysis
As described previously, I expected certain themes to be present in the discourse of
elite’s communication during intergroup conflicts, based on the work of Susan Benesch and
Social Identity Theory. These included in-group/out-group argumentation or logic and/or
referencing, use of derogatory words for the out-group, calling for violence and historical/present
injustices. But I also wanted the data to “speak for themselves.” Consequently, I reviewed all of
the speeches and broadcasts available, and made extensive notes of apparent themes. From this
preliminary assessment, a set of eleven tentative themes was extracted and defined; including
guidelines for coding discourse excerpts into each category. To count as a “theme,” a thematic
point of emphasis had to be present in the communication of at least two different elite sources.
This initial coding scheme (see Appendix B) was then tested, by having an additional
coder independently assess a set of thirty discourse excerpts. Inter-coder agreement on this
initial assessment was low: specifically, coders agreed on only 19 of 30 excerpts. The coding

20

scheme was then revised; specifically, the initially-proposed themes titled “Anger or sadness as
justification of riots” and “Veneration of the martyrs or the fallen” were dropped, and the criteria
for coding the other themes were refined.4 Inter-coder agreement was then tested anew, within a
different sample of 20 discourse excerpts. Inter-coder agreement for this set was 19 out of 20;
producing a Cohen’s Kappa of 0.94. I now turn to a detailed description of the final, reliable
coding scheme; along with some of the discourse excerpts from my data that instantiate these
categories and were used to test them.
1. Defining differences and/or conflation of the in-group identity with the national
or supra-ethnic identity
This theme derives in part from Social Identity Theory and refers to speakers'
assumptions that the in-group is the national group, while the out-group is external to the nation
and thus foreign. This also includes speakers’ claims to a group's ethnic identity connected to
larger regional identity.
In instantiating this theme within their discourse, speakers may use certain words to
define, extrapolate, advocate, and/or reiterate ethnic differences between their groups and
targeted others. They also may talk about in-group defining characteristics, such as cognitive,
moral and physical traits. They may also use words which clearly refer to an out-group and can
also use the opportunity to give historical evidence as to why these differences exist. They may
refute any ideas which suggest that differences are of little or no importance.
(Note: This thematic category does not include derogatory words used for an out-group.)
RTLM: We have already confirmed that ethnic groups exist and even tried to see their
characteristics over the years in the history of Rwanda.

4

See Appendix B for drafts used during the development of the coding scheme and relevant excerpts and details.
21

Milosevic: If we fail to create and strengthen this front, comrades, there will be no
Kosovo, there will be no Serbia and there will be no Yugoslavia.
2. Call for duty/exhortation. This refers to speakers encouraging members of their ingroup to protect themselves, take certain actions, do certain things and/or ensure their survival.
This may be in form of a general call to duty, or detailed actions such as boycotts, taking up
arms, patrolling, staying put, passing information, keeping an eye on others, defending against
attack, and communicating in certain ways. The speakers may also use the words or equivalents
to requests. The speakers can also use exhortation or consequential warnings to those who are
scared or reluctant to comply.
Milosevic: Therefore we the Communists, must do everything in our power to eliminate
the consequences of the nationalist and counterrevolutionary forces both in Kosovo and other
parts of the country.
RTLM: The ministry of defense asks the population who live in areas where the war is
taking place and those who live in Kigali to keep on being vigilant so that no enemy can
infiltrate.
3. Meta-out-group connections. The speakers may connect the out-group or the conflict
to some other out-group of ethnic or political identity/power/ideology/race/religion/nationality;
including international countries, groups, organizations or persons. These “meta-out-groups”
(i.e., out-groups associated with the targeted out-group) may be considered conspiratorial; siding
with or controlling the out-group, negligent or critical of the speaker's group, richer, powerful,
all-knowing, rulers and in some way intangible. However, enemies on the ground may be
considered an embodiment of such meta-out-groups.

22

Wirathu: If we lose now, they would form groups and by the support of Saudi Arabian
oil money, they would compete against our businesses.
Prabhakaran: The Indian military occupation of Tamil Eelam posed a major challenge to
our determination.
4. Claims of victory – “we win, they lose”. During the conflict, the in-group elites often
refer to instances in which they may be winning the war or taking control of the situation. This
also includes abasing the enemy and attempts to demoralize the out-group as losers.
RTLM: People are asking: "What has become of the Inkotanyi in Kigali City? The score
of the Inkotanyi is indeed sealed; the fate of the Inkotanyi in Kigali city was sealed at about 2
PM, yesterday, Monday when the inhabitants killed at least nine Inkotanyi at the Sun City Hotel
in Nyamirambo.
5. Anecdotes. The speaker may inform the audience of specific examples of an outgroup's activities or events that purportedly have happened, through personally witnessed
accounts, or by claiming credible information/source/witness. First person or third person
narratives are most likely to be employed – but importantly, the anecdotes typically are
impossible to verify.
Wirathu: Let me tell you one thing. I visited Myot Hla Township. Local monks
complained that they could not perform Buddhist ritual ceremony on Martyrs day.
6. Fact-making and allegations. Akin to “anecdotes” (i.e., Theme 5, above), this theme
involves non-falsifiable claims made by the speaker, informing the public of some facts, actions,
events, plans or strategy of the out-group and/or any other group/situation/event/historical
incident. This also includes allegations or false claims aimed at an out-groups' strategy, plans,
beliefs or intentions. These allegations may appear to be damaging to the speaker's in-group or to

23

peace in general. This also includes any historical or mythical stories presented as facts. The
distinction between this theme and “anecdotes” is that the facts presented are not framed as
personal (i.e., first- or third-person experiences).
Wirathu: At first their plan was to invade those three towns on the anniversary of 8.8.88.
RTLM: Still, RPF Inkotanyi (reference to Tutsi rebel army), has never really agreed to
stop the combats.
RTLM: About a thousand persons cross the border daily, fleeing from the Inyenzi, those
Inyenzi who continue to target intellectuals, as well as those they accuse of being interahamwe.
They actually check identity cards and do not spare any Hutu.
RTLM: Still, RPF Inkotanyi (reference to Tutsi rebel army), has never really agreed to
stop the combats.
7. Importance of land/territoriality. This refers to claims of ownership/possession of a
region or land and that being a symbol of the ethnic group. This also includes referring to a
particular area, to an entire country, or region in order to protect and/or defend it. Speakers may
also suggest what areas or borderlines are problematic and a source of future trouble for the ingroup or national solidarity.
Prabhakaran: Today the war of aggression against our land by the enemy has reached a
phenomenal scale.
Milosevic: Yugoslavia and Serbia will not give Kosovo away.
8. Name calling. The speakers may use racist, derogatory slurs and/or swear words
against the out-group; including reducing them to sub-human levels such as animals, insects such
as cockroaches, and/or condemned religious or mythical beings such as devils. They may also

24

include character-defining yet negative attributive words such as: wicked, enemy, chauvinistic,
and so forth.
RTLM: When you tell them: "You Inkotanyi, you kill people, you kill people atrociously,
you indulge in torture", they resort to flattery with their smooth tongue, pride and
arrogance.
RTLM: Inyenzi…
9. Historical/present injustice or victimization. The speakers may suggest to the
audiences that they are victims of gross historical or present injustices which makes they are
victims of gross historical or present injustices which makes them victims. They may also
suggest that their group "deserves" power but they have been denied any resources which are
either rightfully theirs or a function of prejudice.
Prabhakaran: This aggressive war that has been launched in the guise of a ”war for
peace” and as a “war for the liberation of the Tamils” has seriously disrupted the peace of the
Tamils, reduced them to refugees, as subjugated people, destroyed their social and economic
existence and brought them intolerable suffering.

25

RESULTS
Following figure and table present the results. The results were calculated by doing frequency
counts of all the themes across all speakers. It was found that theme one “Defining differences
and/or conflation of the in-group identity with the national or supra-ethnic identity” and theme
six “Fact making and allegations” were the most commonly used themes.
Figure 1. uses a column chart where total frequencies were calculated by aggregating all the
frequency of themes across all the speakers. Table 2. separates themes by speakers and the reader
should be able to see how themes occur by frequency by each speaker.
Figure 1. Total number of themes across all speakers.

350
300
250
200
150
100
50
0

26

Table 2. Themes by each speaker.
Milosevic

RTLM

Prabhakaran

Wirathu

Differences

35

209

64

18

Call for duty

16

72

7

7

Meta-out-groups

2

17

7

7

Claiming victory

1

35

11

0

Anecdotes

4

9

0

3

Fact making &
allegations

34

126

46

28

Territoriality

7

1

47

4

Name calling

1

87

20

5

10

28

0

Historical/present 14
injustice

27

DISCUSSION
Previous scholars across various traditions have examined violence-inciting “hate speech.”
Yet most analyses have narrowly focused upon particular texts or speakers. In this thesis, I have
identified and analyzed the presence of consistent themes existing across speakers, cultures, and
temporal spans. In doing so, nine consistent themes were identified that occur within the speech
of elites during periods of intergroup violence.
However, these results also suggest (albeit in a preliminary fashion) that these themes are not
proportionally distributed within the speech of elites. Instead, my data suggest that while all of
these themes are routinely utilized, themes pertaining to in-group/out-group differences and factmaking/allegations comprise the most common forms.
Past scholars also have focused their attention heavily on ideas that immediately seem extremely
negative or brutal such as threats or calling names like Inyenzi (cockroaches). Moreover, there
has been extraordinary emphasis on theme of historical injustice or victimization of the in-group
by scholars. The data analyzed in this thesis validates the presence of such themes. However, a
speech is more than some names and ideas of historical injustice. It is a combination of several
themes; and in my data, in-group out group referencing and fact making and allegations are the
most commonly used ones.
Social Identity Theory had informed the literature review of this thesis on how ingroup/out-group referencing and categorization can be used to mark groups and create identities.
Consistent with the literature and theory’s ideas, one of the dominant themes across speakers and
speeches involved the use of words or sentences that define, categorize or create differences
between groups. However, an interesting observation is the fact that speakers also conflate ingroup identity with the national identity. For instance, a speaker on RTLM suggested that they

28

(i.e., the Tutsi) want to slap Rwandans. This implies that Hutus and ethnic groups other than the
Tutsis are Rwandans – but the Tutsi’s are not. This is akin to Zubero’s idea regarding German
propaganda against Jewish people; in which the rhetoric of Nazi leaders constantly referred to
the Jews as Jews – despite the fact that they were German nationals – thus rendering them
foreigners in their own land. The usage of this theme is persistent across the RTLM, Wirathu and
Milosevic.
At the same time, this latter theme was largely absent within the speech of Prabhakaran.
This raises the question of, why? My take on this is that it reflects the nature of the particular
conflict. Prabhakaran was a minority rebel separatist who was fighting against the state.
Therefore, his claim for Tamil identity cannot be conflated with Sri Lankan identity – unlike the
other speakers who were embroiled in conflicts where they represented the majority and the
minorities were being persecuted.
The single most common theme within my data, however, was fact-making and
allegations. This is defined as when speakers create information or relay news either about the
present or the past. This includes making allegations on part of the out-group, challenging their
motives and using different kinds of facts or even myths to bolster their claims. The practical and
theoretical significance of this finding is noteworthy. Past scholars have focused all too often
upon specific, egregious speech acts, such as making threats, calling for violence or name
calling. But this does not appear (from my data) to be the dominant theme used to potentially
elicit violence and create division. Instead, elites during intergroup conflicts pepper their speech
with non-falsifiable allegations and statements, designed to “stir up” or “convince” their in-group
– all the while appearing as if all they’re doing is “sharing information.” For examples, speakers
may claim that Canada is taking care of Tutsi orphans, or that the Rohingyas plan to invade

29

towns in the Rakhine state. In saying things like this, elites are (in effect) making political or
social arguments. And critical to this “fact sharing” is that the information in question is nonfalsifiable: it would be hard (if not impossible) for an ordinary listener in his or her house in a
small village, town or even a city to verify these allegations. One may argue that the quality of
arguments could be a concern and reasonable audiences may be able to understand them.
However, borrowing from Levine’s Truth Default Theory, I would suggest that people will have
a “truth default” which makes them susceptible to believing in things they hear rather than
challenging or doubting it immediately (2014) .In addition, given that these “facts” are shared
within the broader cultural context of conflict, it is likely that they not only would be believed,
but they would incite anger. Violence may not necessarily be legitimized by the use of words
such as “cockroaches” or “evil,” but when “facts” are presented in conjunction and in bulk that
cast the out-group in a harshly negative light, it’s not especially surprising that violence may
result.
Whereas many scholars in the past have focused their attention upon the use of
derogatory words such as ‘cockroaches’ against out-groups (Benesch, 2012; Benesch &
Abramowitz, 2013; Donohue, 2012); my data suggest that such incidences are not proportionally
dominant. That being said, they did prevail within the RTLM data. One broadcast contained the
word Inyenzi (cockroach) 67 times! The use of negative words or slurs does exist, but it is not the
most important thing in elite communication and does not solely depend on usage of extreme
labels such as vermin or insects. Besides the RTLM, most speakers used negative attributes to
define their enemies rather than outright name calling, including such terms as: wicked, evil,
arrogant and chauvinistic. The prevalent idea – posited in past research and in Benesch’s

30

proposal – that so-called “dangerous speeches” contain words that reduce humans to vermin may
be heavily influenced by RTLM data.
Benesch, in her treatise, encourages her readers to look for calls for violence and
elliptical language in communication. In my theme of “Call for duty/Exhortation,” I coded for
violent and non-violent ideas, whereby the speakers encouraged the listeners that they “must”
and “should” do certain things. Akin to “Name calling,” the RTLM data turned out to be the only
one which contained very specific commands such as asking the population to train, or be
vigilant. The other speaker who called for action with certain commands was Wirathu, who
asked his audience to economically boycott the products of the Rohingyas. Other speakers such
as Milosevic and Prabhakaran did not necessarily give explicit calls for violence but did exhort
for certain actions such as joining Communist forces and LTTE respectively. Again, RTLM
proved to be an outlier in the intensity of their arguments and specifications. This is not to say
that elites do not ask for actions; but rather, that the language they use is not always specific or as
frequent.
Another interesting aspect of elite communication during conflicts is their idea of
claiming victory over the out-group, or at least making progress under duress or holding their
ground. Surprisingly, none of the research in the past has looked into this. In my data, speakers
consistently refer to how well they are doing and how poorly the out-group is performing. From
a social psychological point of view, this is simply a morale booster for the in-group and
humiliating for the out-group. A speaker like Prabhakaran, who was involved in a militaristic
conflict against the Sri Lankan state, consistently boasted of his group’s prowess and skill in
defeating the state army. However, his speeches did not contain the intensity of language

31

displayed by the RTLM. The RTLM radio broadcasters went to extreme lengths to humiliate
Tutsis. For instance, one broadcaster remarked:
“The score of the Inkotanyi is indeed sealed; the fate of the Inkotanyi in Kigali city was sealed at
about 2 PM, yesteday, Monday when the inhabitants killed at least nine Inkotanyi at the Sun City
Hotel in Nyamirambo. When they were flushed out, one of them, who was very wily, managed to
escape. But that makes no difference... he can run but he cannot hide. The inhabitants killed nine
of their people who did not know what to do, as they had run out of ammunition, food and were in
rags.”
There were several other explicit references to how Tutsis were resorting to eating unripe
coffee because they are losing the war and thus facing suffering.
My data also clearly suggest that the prevalence of certain themes is dependent upon the
specific cultural and conflict context in which the speech is occurring. For example, the theme
of territoriality (i.e., focusing on the importance of land or region) was used most often by
Prabhakaran, followed by Milosevic. Why? In Prabhakaran’s case, territory was extremely
focal, considering that the Tamils were situated in the North of the country and that was the
region fighting for independence. Similarly, the conflict in former Yugoslavia was largely
territorial, and various groups were claiming independence which was against Serbian president
Milosevic’s ideas. Future scholars interested in examining such speech should always consider
the particular cultural and conflict-related contexts in which themes arise.
One of the most interesting and novel themes within my analysis is the emergence of
what I have described as “meta-out-groups.” All of the speakers within my data suggested that
some other religious, national, or local groups are siding with their enemies, thus rendering those
third parties as some kind of meta-out-group. Thus, it may be logical to conclude that any

32

success against the out-group is also a victory against the meta-out-group or the third party. This
theme has also not been touched upon by previous researchers.
The particular content of speeches also must be viewed through the lens of the mediums
being used to present that content. Specifically, I have little doubt that one reason why the
RTLM stands out as the most vicious and extremist in its communication is because the medium
of communication was mediated. The RTLM was a private entity run by broadcasters and
journalists whose profession gives them at least some prerogative of free speech in many
respects across the globe. Moreover, radio broadcasts can last from few minutes to several hours
where speakers can communicate unhindered particularly given the circumstances. Such
broadcasts also were culturally legitimated by others in positions of power: the ruling Hutu party
and the government gave its blessings to the radio because of the presence of several officials
such as the Prime Minister and the President, and several prefects making appearances on its
shows. Moreover, one broadcast examined in this thesis was made a few months before the
conflict actually began. The broadcast contains explicit reference to out-groups and historical
explanations to presence of different ethnic groups in Rwanda and even uses the words Inyenzi
once. This should encourage researchers to see if the conflict was in the pipeline months before it
actually led to genocide.
This thesis is an exploratory attempt in the subject of elite communication but it does not
strictly adhere to the rules of quantitative content analysis. For such an undertaking, random
sampling, independent coders and other relevant measures have to be taken. However, this thesis
tried to develop a coding scheme which can facilitate such endeavors in the future.

33

Limitations and Implications for Future Research
Understanding communication during conflicts may be a well-intentioned exercise, but it
is still a difficult one. Conflicts by their very nature impede recordings and transmission of
important data and the dynamics of research times frames do not help much either. This study is
limited in its sample size and due to a comparative nature, does not account for several other
differences amongst the countries of focus. The motivation behind such a study was to explore
elite communication within the depths of available data and most importantly within the given
time frame and constraints of the research design. Further research can look into the types of
facts and allegations and how different themes are used in combination with each other. I think
that the future looks bright as this and other studies can help move the theory of conflict
communication forward, develop sensitive scales and thematic coding schemes and be applied to
other domains besides genocide. It is in hope that such measures can be used to predict conflicts
and their nature in hope that they can be forestalled.

34

APPENDICES

35

Appendix A
Table 3. Elite communication themes during intra-state, inter-group conflicts from the
literature.
Source

Conflict types

Themes

Bozic-Anderson, 1992;

Ethnic: former Yugoslavia

Identity, ethnic polarization

Ethnic: Serbia and India

Identity salience, use of

Bozic Roberson, 2007
Desrosiers, 2010

myths
Donohue, 2012

Ethnic: Rwanda

Identity, us vs. them

Donohue, 2012

Ethnic: Rwanda

Out-group dehumanization

Glaeser, 2005

Inter-group

Economic reasoning

Kaufman, 2001

Ethnic and inter-group:

Myths, symbolism

former Yugoslavia, Nazi
Germany
Kissoupoulous, 2007

Ethnic: Serbia and India

Ethnic polarization, identity

Mann, 2005

Ethnic: former Yugoslavia

Victimhood, false
accounts/myths

Mann, 2005

State sanctioned: Nazi
Germany

36

Economic reasoning

Table 3 (cont’d)
Morus, 2000

Ethnic and Inter-group:

Dehumanization, fear of the

Cambodia, Rwanda and

enemy and injustice.

Germany
Morus, 2007

Ethnic: former Yugoslavia

Myths, false stories/accounts

Morus, 2007

Ethnic former Yugoslavia

Exclusiveness of in-group

Roy and Rowland, 2003

Communal: India

Victimization, minority vs.
majority

Roy and Rowland, 2003

Communal: India

Myths, redemption

Zubero, 2012

Inter-group: Rwanda and

Identity, social construction

Germany

of stranger

37

Appendix B
A preliminary examination of texts made for themes in Draft 1 and its examples. This
draft was then tested with new excerpts from the data. To improve inter-coder reliability, some
improvements and changes were made (including dropping theme 2) which led to Draft II. A
new test was employed which improved reliability, dropped theme Veneration of martyrs/fallen
and helped finalize Draft II whose improved version is already presented above in the Coding &
Analysis section. Draft II below contains excerpts from first and second inter-coder reliability
tests.
Themes draft I
Following are the commonly appearing themes from a preliminary investigation of the
data. Each theme is defined and is followed by an example. Note that some themes may be
overlapping with each other.
1. Conflation of the in-group identity with the national or supra-ethnic identity
This refers to speakers' assumptions that the in-group is THE national group while the
out-group is external to the nation and thus foreign. This also includes speaker's claims to a
group's ethnic identity connected to larger regional identity. (Note: This is akin to Zubero's
classification of making strangers out of out-groups)
Wirathu: From Mawlamyine to Yangon, the coach is empty and their people use it
but from Yangon to Mawlamyine the coach is still full because our golden
Burmese use it. So they (Rohingyas) reduced the price. At first it was 6000 kyat
but then they reduced to 5000 kyat. Still no Burmese used the service. Then they
reduced up to 3500 kyat. Now, all are using it again. If we lose now, they would
form groups and by the support of Saudi Arabian oil money, they would compete
against our businesses.
2. Anger/Sadness as justification for riots
The refers to speakers using a grievous incident as a cause of anger/sadness which
subsequently produced unruly behaviors such as rioting, looting or killing. A speech may be
delivered after a grievous or sad incident and can contain references to the said incident.
RTLM (Rwandan President Sindikubwabo): Peace and security were already
undermined in the town of Kigali due principally to anger and sorrow that
Rwandans got because of their enemies who made them orphans.
3. Call for duty/exhortation. This refers to speakers encouraging the audiences to
protect themselves and/or ensure their survival. This may be in form of call or duties to detailed
mechanisms or strategies such doing certain actions, boycott, take up arms, patrol, stay put,
ignore, encourage others, pass information, keep an eye on others, defend, and communicate.
The speakers may also use the words or equivalents to requests. The speakers can also
use exhortation or consequential warnings to those who are scared or reluctant to comply.
Prabhakaran. This is crucial for the survival of the Tamil race. Our Nation has
been forced into a necessity to struggle for survival. The young generation of
Tamils cannot escape from this national duty, from this call of history. Any delay
in this task will endanger the existence of our Nation. Therefore, I call upon the
younger generation to join our liberation movement without delay. The earlier the

38

young men and women join our movement the quicker we can achieve the
objectives of our struggle.
Wirathu: If all our people travelling from both side support only travelling on
both sides and support only our fellow Buddhist even in nationalist sense and
sacrifice the amount 1500 kyat , we golden Burmese would win this fight.
RTLM - prefect of Kigali 14/04/1994: As you know that citizens have their own
means to ensure security in their localities. I want them to remain courageous as
they showed it these last days (sic), to conduct night patrols, to prevent the enemy
from inflitrating because as it is evident, the war in town is different from the war
in the forest
4. The third party connection. The speakers may connect the out-group to some other
out-group of ethnic or political identity/power/ideology/race/nationality including international
countries, groups, organizations or persons. These meta-outgroups may considered
conspiratorial, siding with or controlling the out-group, negligent or critical of the speaker's
group, richer, powerful, all-knowing, rulers and in some way intangible. However, enemies on
the ground may be considered an embodiment of such meta-outgroups.
RTLM: Today certain white people, especially Americans, Canadians and Belgians
believe that the villain in this country is the machete-wielding Hutu, whereas the Hutu
is trying to ensure the Hutus are not annihilated by the descendants of Gatutsi.
RTLM: The United States has a little known phenomenon called lobbying. This is
donate money to leading candidates in return for favors once they are elected. This is
a clever tactic. It is now used, for instance, by Museveni through certain wily
individuals who defend his interests before American members of congress. Then the
congressmen bring pressure to bear on the President of the United States. That is
how this kind of decision is arrived at. A Hima from Uganda uses American crooks
and they tell the President to "kill the Hutus".
RTLM: Thus as we were saying, it is the Tutsi MUSVENI group that the American's
are assisting. Er... as I indicated earlier, the aid brought in by the Americans is for
the Inkotanyi to go and study in their country, in places like Arkansas, Bill Clinton's
home state. Such aid has bow increased from $150,000 to $400,000 US dollars.
Besides, they have disbursed $50,000 or rather $500,000 US dollars to fund the
demobilization of soldiers of MUSEVINI's army. Nevertheless, there has been no
reduction in the army strength as most of the Ugandan troops concerned gave been
sent here. Part of the money was used to purchase arms, which is the reason we are
living under the threat of annihilation.
RTLM: I agree with you with what you have said about the white men... those white
men who sold us, those Tutsis who sold us abroad., Henceforth, I know that there are
some white Canadians here who are doing everything to have the surviving Tutsi
orphans atGisimba's and in other places to be sent abroad.
Wirathu: If we lose now, they would form groups and by the support of Saudi
Arabian oil money, they would compete against our businesses.
Prabhakaran: The Indian military occupation of Tamil Eelam posed a
major challenge to our determination. At that historical conjuncture when we were
hard pressed by the military supremacy of a world power, we fought with fierce
determination without giving up the cause. During that dangerous situation,

39

determination became our ultimate weapon. It was by that moral power we were able
to confront a great military power of the world.
Prabhakaran: We are now facing a new challenge, a new war of aggression.
Our historical enemy, Sinhala Buddhist Chauvinism, has taken institutional form in
the guise of Chandrika’s regime and has been conducting a genocidal war against the
Tamil race.
5. We win, they lose
During the conflict, the in-group elites may refer to instances where they may be
winning the war or taking control of the situation. This also includes abasing the enemy and
attempts to demoralize the out-group.
RTLM: People are asking: " What has become of the Inkotanyi in Kigali City? The
score of the Inkotanyi is indeed sealed; the fate of the Inkotanyi in Kigali city was
sealed at about 2 PM, yesteday, Monday when the inhabitants killed at least nine
Inkotanyi at the Sun City Hotel in Nyamirambo. When they were flushed out, one of
them, who was very wily, managed to escape. But that makes no difference... he can
run but he cannot hide. The inhabitants killed nine of their people who did not know
what to do, as they had run out of ammunition, food and were in rags.
6. Anecdotes. The speaker may inform the audience of an out-group's activities through a
personally witnessed account or by claiming credible information/source/witness.
Wirathu: Let me tell you one thing. I visited Myot Hla Township. Local monks
complained that they could not perform Buddhist ritual ceremony on Martyrs day. I
was blown off. What a rude governor of the Township I said. I thought the authority
was so rude. How come they refused to perform Buddhist rituals that day General
Aung San was assassinated. But the local monks explained me that the property
owner is a 'Kalar' (slur for Rohingya Muslims) and he does not allow us to perform.
7. Fact-making and allegations. This refers to speaker informing the public of some
facts, actions, events, plans or strategy of the out-group. This also includes allegations or false
claims aimed at an out-groups' strategy, plans, beliefs or intentions. These allegations may
appear to be damaging to the speaker's in-group or to peace in general.
This also includes any historical or mythical stories presented as facts.
Wirathu: At first their plan was to invade those three towns on the anniversary of
8.8.88. But at the time of the Taung-goke incident, they were confident because they
had had prior military trainings and enough weapons. They waged a war and tried to
invade three towns in May in an untimely fashion i.e., before their actual plan. But
their war wasn't successful, thus as a revenge they are attempting again now.
Wirathu: They have joined the political parties not because they are interested in
politics, not because they respect human right but for the sake of their people. Their
religion is the most abusive of human rights. They are seriously abusing freedom of
belief.
RTLM: About a thousand persons cross the border daily, fleeing from the Inyenzi,
those Inyenzi who continue to target intellectuals, as well as those they accuse of
being interahamwe. They actually check identity cards and do not spare any Hutu.
RTLM: Thus as we were saying, it is the Tutsi MUSVENI group that the American's
are assisting. Er... as I indicated earlier, the aid brought in by the Americans is for
40

the Inkotanyi to go and study in their country, in places like Arkansas, Bill Clinton's
home state. Such aid has bow increased from $150,000 to $400,000 US dollars.
Besides, they have disbursed $50,000 or rather $500,000 US dollars to fund the
demobilization of soldiers of MUSEVINI's army. Nevertheless, there has been no
reduction in the army strength as most of the Ugandan troops concerned gave been
sent here. Part of the money was used to purchase arms, which is the reason we are
living under the threat of annihilation.
8. Veneration of martyrs/fallen. The speakers may mention and/or praise the fallen and
suggest to the audiences that their sacrifices must never be forgotten.
Prabhakaran: Today is the day of remembrance of our country’s brave martyrs. Today
we pray our respect to those heroes who valued the freedom of their country more than
their life and gave their life for it. We light the lamp of gratitude in our hearts and honor
them.
9. Importance of land. This refers to claims ownership/possession of a region or land
and that being a symbol of the ethnic group. This also includes referring to a particular area, to
an entire country, region in order to protect and/and or defend it. Speakers may also suggest what
areas or border lines are problematic and a source of future trouble for the in-group or national
solidarity.
Prabhakaran: Today the war of aggression against our land by the enemy has reached a
phenomenal scale. Having mobilized all its military power and having utilized all its
national resources, the enemy has launched a massive invasion on the Jaffna soil. Our
traditional land of ancient historical glory is being systematically destroyed by the
enemy’s firepower. The intense shelling that rains down unabatingly has wiped out the
face of Jaffna. The fundamental objective of this war of aggression is to destroy the
economic resources and the cultural heritage of Jaffna thereby uprooting the national life
of the people.
10. Defining differences and name calling. The speakers may use the speech to define,
extrapolate, advocate, reiterate ethnic differences between their group and the rest. They can also
talk about in-group/out-group defining characteristics such as cognitive, moral and physical
traits. They may also use words which clearly refer to an out-group and can also use the
opportunity to give historical evidence as to why these differences exist. They may refute any
ideas which suggest that differences are of little or no importance.
Moreover, in order to stress differences, the speakers may use racist, or derogatory or
identifying slurs. They may also include character defining yet negative attributive words such
as: devil, wicked, enemy, chauvinistic.
RTLM: When you tell them: "You Inkotanyi, you kill people, you kill people
atrociously, you indulge in torture", they resort to flattery with their smooth
tongue, pride and arrogance.
11. Historical or present victimization/injustice.
The speakers will suggest to the audiences that they are victims of gross historical or
present injustice which makes them victims. They may also suggest that their group "deserves"

41

power but they have been denied any resources which are either rightfully theirs or a function of
prejudice.
Prabhakaran: This aggressive war that has been launched in the guise of a ”war
for peace” and as a “war for the liberation of the Tamils” has seriously disrupted
the peace of the Tamils, reduced them to refugees, as subjugated people,
destroyed their social and economic existence and brought them
intolerable suffering

42

Themes draft II
1. Conflation of the in-group identity with the national or supra-ethnic
identity/Defining differences
This refers to speakers' assumptions that the in-group is THE national group while the
out-group is external to the nation and thus foreign. This also includes speaker's claims to a
group's ethnic identity connected to larger regional identity. The speakers may use the speech to
define, extrapolate, advocate, reiterate ethnic differences between their group and the rest. They
can also talk about in-group/out-group defining characteristics such as cognitive, moral and
physical traits. They may also use words which clearly refer to an out-group and can also use the
opportunity to give historical evidence as to why these differences exist. They may refute any
ideas which suggest that differences are of little or no importance.
(Note: This is akin to Zubero's classification of making strangers out of out-groups)
RTLM: …another item of news about the Inkotanyi
RTLM: We have already confirmed that ethnic groups exist and even tried to see
their characteristics over the years in the history of Rwanda
2. Call for duty/exhortation. This refers to speakers encouraging the audiences to
protect themselves and/or ensure their survival. This may be in form of call or duties to detailed
mechanisms or strategies such doing certain actions, boycott, take up arms, patrol, stay put,
ignore, encourage others, pass information, keep an eye on others, defend, and communicate.
The speakers may also use the words or equivalents to requests. The speakers can also use
exhortation or consequential warnings to those who are scared or reluctant to comply.
Milosevic: Therefore we the Communists, must do everything in our power to
eliminate the consequences of the nationalist and counterrevolutionary forces
both in Kosovo and other parts of the country.
Milosevic: The first thing, I want to tell you comrades in connection with this goal
is to stay here.
RTLM: We have already confirmed that ethnic groups exist and even tried to see
their characteristics over the years in the history of Rwanda.
3. The third party connection. The speakers may connect the out-group to some other
out-group of ethnic or political identity/power/ideology/race/nationality including international
countries, groups, organizations or persons. These meta-out-groups may considered
conspiratorial, siding with or controlling the out-group, negligent or critical of the speaker's
group, richer, powerful, all-knowing, rulers and in some way intangible. However, enemies on
the ground may be considered an embodiment of such meta-out-groups.
RTLM: For a long time, the white man has harboured the erroneous conception
that Tutsi are good people.
Prabhakaran: Our historical enemy, Sinhala Buddhist Chauvinism, has
taken institutional form in the guise of Chandrika’s regime and has been
conducting a genocidal war against the Tamilrace.
RTLM: You must have understood that Belgian policy has suffered a setback, as
France has assumed the leading role in Rwandan politics.

43

4. We win, they lose
During the conflict, the in-group elites may refer to instances where they may be
winning the war or taking control of the situation. This also includes abasing the enemy and
attempts to demoralize the out-group.
RTLM: Why are you running to eat our flour in Magerwa? While in CND, were not you
eating pleasant food? Your relatives were not bringing you milk there? Now your
arrogance made them to be exterminated.
5. Anecdotes. The speaker may inform the audience of an out-group's actvties through a
personally witnessed account or by claiming credible information/source/witness. first person or
third person
RTLM: In all corners, we don't see any Inkotanyi.
Wirathu: When I visited Meikhtilar, a person said, the ostrich has become Muslim
(reference to NLD party becoming a Muslim party).
RTLM: A woman living in Nyamirambo asked me to salute a young man, her son, Army
Seargeant Asumani, in Cyamingu.
6. Fact-Making and allegations. This refers to speaker informing the public of some
facts, actions, events, plans or strategy of the out-group. This also includes allegations or false
claims aimed at an out-groups' strategy, plans, beliefs or intentions. These allegations may
appear to be damaging to the speaker's in-group or to peace in general. This also includes any
historical or mythical stories presented as facts.
Wirathu: At first their plan was to invade those three towns on the anniversary of 8.8.88.
RTLM: Still, RPF Inkotanyi (reference to Tutsi rebel army), has never really agreed to
stop the combats
RTLM: You must have understood that Belgian policy has suffered a setback, as France
has assumed the leading role in Rwandan politics.
RTLM: So we found that the attack of the Inkotanyis was the major factor which
undermined our security in Kigali.
7. Importance of land/territoriality. This refers to claims ownership/possession of a
region or land and that being a symbol of the ethnic group. This also includes referring to a
particular area, to an entire country, region in order to protect and/and or defend it. Speakers may
also suggest what areas or border lines are problematic and a source of future trouble for the ingroup or national solidarity.
Prabhakran: This war is being waged for liberation of our land
8. Name calling. The speakers may use the speech to define, extrapolate, advocate,
reiterate ethnic differences between their group and the rest. They can also talk about
ingroup/outgroup defining characteristics such as cognitive, moral and physical traits. They may
also use words which clearly refer to an out-group and can also use the opportunity to give
historical evidence as to why these differences exist. They may refute any ideas which suggest
that differences are of little or no importance.
Moreover, in order to stress differences, the speakers may use racist, or derogatory or
identifying slurs. They may also include character defining yet negative attributive words such
as: devil, wicked, enemy and chauvinistic.

44

RTLM: When you tell them: "You Inkotanyi, you kill people, you kill people
atrociously, you indulge in torture", they resort to flattery with their smooth
tongue, pride and arrogance.
RTLM: ...Inyenzi
9. Historical or present victimization/injustice.
The speakers will suggest to the audiences that they are victims of gross historical or
present injustice which makes them victims. They may also suggest that their group "deserves"
power but they have been denied any resources which are either rightfully theirs or a function of
prejudice.
Prabhakaran: The five year rule of Chandrika has been a curse on Tamil people.
The monumental tragedy that our people encountered in the form of war,
violence, death, destruction, displacement, hunger and starvation was the worst
form of Tyranny ever suffered by the Tamils
Milosevic: It is not right that some get to educate their children abroad while
others have to escort their children around the corner to so that they do not get
beaten up or raped on this short journey.

45

Appendix C

Table 4. List of Speeches
Date of speech/broadcast

Medium

Milosevic


April 24, 1987

Public address



October 17, 1988

Public address



November 19, 1988

Public address

Prabhakaran


November 27, 1993

Radio



November 27, 1995

Radio



November 27, 1996

Radio



November 27, 1999

Radio



November 27, 2000

Radio



November 27, 2001

Radio



December 12, 1993

Radio



April 11, 1994

Radio



April 14, 1994

Radio



May 17-18, 1994

Radio



May 28, 1994

Radio



June 09, 1994

Radio

RTLM

Wirathu


Available on youtube.com

Exact date undetermined. Possibly
2012



Available on youtube.com

Exact data undetermined. Possibly end
of 2012 and early 2013

46

REFERENCES

47

REFERENCES

Ashmore, R., Jussim, L. J., & Wilder, D. (1982). Introduction. In R. Ashmore, L. J. Jussim, & D.
Wilder (Eds.), Social identity, intergroup conflict, and conflict reduction (pp. 42-68).
Oxford: Oxford University Press.
BBC. (2014). Q&A: Communal violence in Burma. Retrieved from
http://www.bbc.com/news/world-asia-18395788
BBC. (2014). Medicins San Frontiers’s shock at Myanmar suspension. Retrieved from
http://www.bbc.com/news/world-asia-26379804
Beech, H. (2013, July 1). The face of Buddhist terror. Time.
Benesch, S. (2008). Vile Crime or Inalienable Right: Defining Incitement to Genocide. Virginia
Journal of International Law, 48 (3).
Benesch, S (2012). Dangerous speech: A proposal to prevent group violence. Retrieved from
http://www.worldpolicy.org/sites/default/files/Dangerous%20Speech%20Guidelines%20
Benesch%20January%202012.pdf
Benesch, S. & Abramowitz, M. (2013). A prelude to Murder: Calling humans vermin. The Wall
Street Journal. Retrieved from
http://online.wsj.com/news/articles/SB10001424052702303497804579241983104982434
Beyene, Z. (2013). The role of media in ethnic violence during political transition in Africa: The
case of Rwanda and Kenya.(Order No. AAI3522072, Dissertation Abstracts
International Section A: Humanities and Social Sciences, Retrieved from
http://ezproxy.msu.edu/login?url=http://search.proquest.com/docview/1413430380?accou
ntid=12598. (1413430380; 2013-99110-377)
Bozic, A. (1992). The rhetoric of Slobodan Milosevic and war on the territory of
Yugoslavia. (Order No. 1348417, Western Michigan University). ProQuest Dissertations
and Theses. Retrieved from
http://ezproxy.msu.edu/login?url=http://search.proquest.com/docview/304015404?accoun
tid=12598. (304015404).
Bozic-Roberson, A. (2004). Words before the war: Milosevic’s use of mass media and rhetoric
to provoke ethnopolitical conflict in former Yugoslavia. East European Quarterly, 38(4),
395-408.
Desrosiers, M. (2010). Opposition, politicization and simplification: Social and psychological
mechanisms of elite-led mobilization. Unpublished PhD dissertation. Retrieved from

48

https://tspace.library.utoronto.ca/bitstream/1807/11197/1/Desrosiers_MarieEve_200806_
PhD_Thesis.pdf
Donohue, W. A. (2012). The identity trap: The language of genocide. Journal of Language and
Social Psychology, 31, 13-29. doi: 10.1177/0261927X11425033
Eriksen, T.H. (1982). Ethnic identity, national identity and intergroup conflict. In R. Ashmore, L.
J. Jussim, & D. Wilder (Eds.), Social identity, intergroup conflict, and conflict reduction
(pp. 42-68). Oxford: Oxford University Press.
Frère, M.-S., Marthoz, J. P., & Institut Panos. (2007). The media and conflicts in Central Africa.
Boulder, Colo: Lynne Reinner Publishers
Glaeser, E. L. (2005). The political economy of hatred. Quarterly Journal of Economics, 120,
45-86. doi:10.1162/0033553053327434
Kagan, S. (2008). The “Media case” before the Rwanda Tribunal: The Nahimana et al. Appeal
Judgement. Retrieved from
http://www.haguejusticeportal.net/index.php?id=9166http://www.haguejusticeportal.net/i
ndex.php?id=9166
Kaufman, S. J. (2001). Modern hatreds: The symbolic politics of ethnic war. New York: Cornell
University Press.
Kirschke, L. M. (2010). Playing the ethnic card: Party networks and violence in the Balkans
(Doctoral dissertation). Retrieved from ProQuest Dissertations & Theses database. (AAT
No. 3401577)
Levine, T. R. (2014). Truth-default Theory (TDT): A Theory of Human Deception and
Deception Detection. Forthcoming.
Li, D. (2004). Echoes of violence: Considerations on radio and genocide in Rwanda. Journal of
Genocide Research, 6(1), 9-27. doi:http://dx.doi.org/10.1080/1462352042000194683
M’Bayo, R. T. (2005). Liberia, Rwanda & Sierra Leone: The public face of public violence.
African Journalism Studies 25 (6). 21-32
Mamdani, M. (2001). When victims become killers: Colonialism, nativism, and the genocide in
Rwanda. Princeton, N.J: Princeton University Press.
Mann, M. (2005). The dark side of democracy: Explaining ethnic cleansing. New York:
Cambridge University Press.

49

Morus, C. M. (2000). Language and power in genocidal regimes: An analysis of the rhetorical
function of language in Pol Pot's Cambodia, Rwanda and Hitler’s Germany. Unpublished
thesis. Retrieved from Proquest Dissertations and Theses
Morus, C. M. (2007). Slobo the redeemer: The rhetoric of slobodan milosevic and the
construction of the serbian "people". The Southern Communication Journal, 72(1), 1.
Roy, A., & Rowland, R. C. (2003). The rhetoric of hindu nationalism: A narrative of mythic
redefinition. Western Journal of Communication, 67(3), 225-248.
Straus, S. (2007). What Is The Relationship Between Hate Radio And Violence? Rethinking
Rwanda's "Radio Machete". Politics & Society, 35, 609-637. doi
10.1177/0032329207308181
Tabeau, E., & Bijak, J. (2005). War-related deaths in the 1992-1995 armed conflicts in bosnia
and herzegovina: A critique of previous estimates and recent results. European Journal of
Population, 21(2-3), 187-215. doi:http://dx.doi.org/10.1007/s10680-005-6852-5
Tajfel, H. (1981). Human groups and social categories: Studies in social psychology.
Cambridge: Cambridge University Press.
Yanagizawa-Drott, D. (2010). Propaganda and conflict: Theory and evidence from the Rwandan
genocide. Retrieved from http://ipl.econ.duke.edu/bread/papers/0511conf/YanagizawaDrott.pdf
Zubero, I (2012). The construction of the stranger and social violence. In Communication and
Violence (eds.) Mingolarra, J. A, Arrena C, and Sabaris, R. M. Current Research Series 8,
Centre of Basque Studies, University of Nevada. Retrieved from
http://basque.unr.edu/docs/CR8.pdf

50