Techniques of Neutralization and Identity Work Among Accused Genocide Perpetrators

Following the 1994 Rwandan genocide, many defendants on trial at the International Criminal Tribunal for Rwanda (ICTR) testified on their own behalf. This article analyzes transcripts of their testimonies to learn (1) how defendants discuss the grave crimes of which they are accused, and (2) how their explanatory styles allow them to rationalize their actions and negotiate their tarnished identities. We find that defendants employ Gresham Sykes and David Matza’s (1957) classic techniques of neutralization as a means of rationalization, impression management, and identity negotiation. Nevertheless, these techniques, along with those developed in the decades since, do not capture all aspects of defendants’ accounts. We thus identify additional techniques of neutralization to provide a more comprehensive understanding of how defendants account for their actions. By extending this classic literature, we call attention to the situational context of international trials, the nature of the crime of genocide, the relatively high social status many defendants once occupied, and existing narratives surrounding the legitimacy of the ICTR. In doing so, our analysis contributes to understandings of narratives of violence and accused genocide perpetrators. K E Y W O R D S : genocide; accounts; neutralization; international criminal tribunal; symbolic interaction. In 1994, mass violence claimed up to one million lives in the small East African nation of Rwanda. Despite mounting evidence of atrocities and large-scale systematic killings, the international community was slow to respond. After the violence subsided, however, the United Nations Security Council quickly mandated the creation of the International Criminal Tribunal for Rwanda (ICTR) to help bring those responsible for the genocide to justice (United Nations 1994). The ad-hoc Tribunal issued its first indictment in 1995 and, prior to closing in 2015, tried 75 individuals for planning and executing the violence. Many of these accused perpetrators testified on their own behalf and thus acquired an international stage on which to explain their actions. This article draws upon testimonies of The authors wish to thank Elizabeth Heger Boyle and the Social Problems reviewers for their constructive comments, as well as Meral Aboghaba, Genna Danial, Courtney DeRoche, and Pamela Devan for their assistance. Previous versions of this article benefited from the feedback of those in attendance at BOS-CONN, the University of Wisconsin Colloquium, and ASA and SSSP Annual Meetings. Direct correspondence to: Emily Bryant, Department of Sociology, Boston University, 100 Cummington Mall, Room 260, Boston, MA 02215. E-mail: eabryant@bu.edu. VC The Author 2017. Published by Oxford University Press on behalf of the Society for the Study of Social Problems. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com 1 Social Problems, 2017, 0, 1–19 doi: 10.1093/socpro/spx026 Article Downloaded from https://academic.oup.com/socpro/advance-article-abstract/doi/10.1093/socpro/spx026/4056221 by University Of Minnesota user on 29 December 2017 27 defendants to analyze how accused genocide perpetrators explain and rationalize what are widely considered among the gravest atrocities known. We ask two questions: First, how do defendants discuss the genocidal crimes of which they are accused? Second, how do their explanatory styles allow them to rationalize their actions and negotiate their identities? Sociologists have recently demonstrated the discipline’s potential to inform the study of genocide (Campbell 2009; Hagan and Rymond-Richmond 2009; Savelsberg 2010). Drawing upon several core sociological concepts, including identity maintenance and the presentation of self (Goffman 1959), we demonstrate how those on trial at the ICTR employ techniques of neutralization (Sykes and Matza 1957) as a means of rationalization, impression management, and identity negotiation. We find that those accused of genocide and related atrocities frequently employ a number of these wellestablished techniques not only to account for their actions, but also to assert a positive, socially accepted sense of self in light of identities deeply tarnished by ICTR accusations. We also identify additional techniques particular to the crime of genocide and the situational context of international trials. Beyond testing the scope conditions of neutralization theory, our investigation informs scholarship on the rationalization of genocide, a crime that has taken millions of lives throughout history while destroying cultures and communities. This analysis also has repercussions for historical narratives of violence, as international tribunals help establish historical records of events (Savelsberg and King 2005). These narratives are often powerful, consequential, and fiercely contested (Chiwengo 2008; Del Rosso 2011); as several neutralization techniques invoke claims of denial, the techniques displayed at international tribunals may also contribute to historical revisionism. With the Tribunal having closed its doors, the time has come for systematic investigation of the content and meaning of these narratives of violence. G E N O C I D E P E R P E T R A T O R S A N D T H E S O C I A L P S Y C H O L O G Y O F T H E S E L F The 1994 war and genocide in Rwanda captured global attention due to the devastatingly rapid and immense loss of life, destruction, and displacement (Government of Rwanda 2012; Mullins 2009; Verpoorten 2005). The mass violence began shortly after the plane carrying Rwandan President Juvénal Habyarimana was shot down on April 6, 1994, and subsided after the Rwandan Patriotic Front (RPF), a rebel group of Tutsi refugees, seized control of the Rwandan capital and established a new government that July (Des Forges 1999; Prunier 1995). A number of high-ranking officials within Rwanda, particularly those associated with the late President’s political party, orchestrated the genocide, though ordinary Rwandan citizens perpetrated the majority of the violence (Mamdani 2001). Numerous scholars have analyzed these perpetrators (e.g., Fujii 2009; Straus 2006), contributing to a robust body of literature regarding individuals who commit genocide, otherwise known as génocidaires. Many of these studies (e.g., Adorno 1950; Arendt 1963; Brustein 1996) are grounded in the works of Stanley Milgram (1974) and Philip Zimbardo (2007), whose experiments demonstrate how psychologically “normal” people could commit seemingly unthinkable acts under particular conditions (but see Perry 2013). Their research shows that the performance of cruel and inhumane actions is a learned process, particularly when demanded by those in positions of power, leading scholars to suggest that génocidaires are generally not psychologically unstable but are, in many ways, ordinary 1 Estimates of deaths, rapes, and other forms of victimization vary widely and are contested; issues range from who is included in death tolls and the government’s vested interest in claiming victimization to the underreporting of Tutsis in census data and the difficulty inherent in quantifying victimization during mass violence (see Lemarchand 2013). 2 While we use the term “perpetrator,” we recognize that there is much heterogeneity among perpetrators, ranging from those who plan the genocide to those who execute it in various settings. 3 As our article makes clear, the label “génocidaire” is complex and highly contested. We use the term given its prevalence in genocide studies, though we acknowledge the important distinction between labels applied to individuals and those applied to their actions. 2 Bryant, Schimke, Nyseth Brehm, and Uggen Downloaded from https://academic.oup.com/socpro/advance-article-abstract/doi/10.1093/socpro/spx026/4056221 by University Of Minnesota user on 29 December 2017 social actors—a paradox that remains the hallmark of research on perpetrators (Alvarez 1997; Browning 1998; Owens, Su, and Snow 2013; Waller 2007). Scholarship on political elites argues that genocide is the outcome of rational decision making that government officials turn to when their authority is threatened (Midlarsky 2005). Other works address those implementing the violence, analyzing the situational influences of collective action frames (Hagan and Rymond-Richmond 2009) and bureaucratic hierarchies (Bauman 1989). While much of this scholarship relies upon historical records, several studies examine the accounts of perpetrators themselves. Scott Straus (2006) contends that uncertainty, fear, and anger in the context of war can influence participation, as people engage in violence in response to encouragement, intimidation, or coercion. Others likewise argue that peer and familial networks (Fujii 2009; McDoom 2013, 2014) contribute to individual participation in genocide or similar mass violence. Departing from these invaluable studies, we analyze the accounts of genocide perpetrators and build upon work such as Diana Scully and Joseph Marolla’s (1984) study of convicted rapists, which illustrates how excuses and justifications help reveal the cultural repertoire of potentially neutralizing accounts. The social psychological tradition of symbolic interaction theorizes the self as continuously constructed through interaction with other social actors (Cooley 1922; Mead 1934). The looking-glass self (Cooley 1922), for instance, asserts that individuals see themselves through the eyes of others, shaping both identity and behavior in processes of role taking and reflected appraisals. An individual’s actions must, accordingly, be viewed within the context of social relations. Relatedly, Erving Goffman’s (1959) notion of impression management illustrates the techniques individuals employ to influence the perceptions of others and maintain desired impressions. The presence of an “audience”—real or imaged—dictates whether and how individuals work to be viewed in a par

In 1994, mass violence claimed up to one million lives in the small East African nation of Rwanda.Despite mounting evidence of atrocities and large-scale systematic killings, the international community was slow to respond.After the violence subsided, however, the United Nations Security Council quickly mandated the creation of the International Criminal Tribunal for Rwanda (ICTR) to help bring those responsible for the genocide to justice (United Nations 1994).The ad-hoc Tribunal issued its first indictment in 1995 and, prior to closing in 2015, tried 75 individuals for planning and executing the violence.Many of these accused perpetrators testified on their own behalf and thus acquired an international stage on which to explain their actions.This article draws upon testimonies of considered among the gravest atrocities known.We ask two questions: First, how do defendants discuss the genocidal crimes of which they are accused?Second, how do their explanatory styles allow them to rationalize their actions and negotiate their identities?
Sociologists have recently demonstrated the discipline's potential to inform the study of genocide (Campbell 2009;Hagan and Rymond-Richmond 2009;Savelsberg 2010).Drawing upon several core sociological concepts, including identity maintenance and the presentation of self (Goffman 1959), we demonstrate how those on trial at the ICTR employ techniques of neutralization (Sykes and Matza 1957) as a means of rationalization, impression management, and identity negotiation.We find that those accused of genocide and related atrocities frequently employ a number of these wellestablished techniques not only to account for their actions, but also to assert a positive, socially accepted sense of self in light of identities deeply tarnished by ICTR accusations.We also identify additional techniques particular to the crime of genocide and the situational context of international trials.
Beyond testing the scope conditions of neutralization theory, our investigation informs scholarship on the rationalization of genocide, a crime that has taken millions of lives throughout history while destroying cultures and communities.This analysis also has repercussions for historical narratives of violence, as international tribunals help establish historical records of events (Savelsberg and King 2005).These narratives are often powerful, consequential, and fiercely contested (Chiwengo 2008;Del Rosso 2011); as several neutralization techniques invoke claims of denial, the techniques displayed at international tribunals may also contribute to historical revisionism.With the Tribunal having closed its doors, the time has come for systematic investigation of the content and meaning of these narratives of violence.

G E N O C I D E P E R P E T R A T O R S A N D T H E S O C I A L P S Y C H O L O G Y O F T H E S E L F
The 1994 war and genocide in Rwanda captured global attention due to the devastatingly rapid and immense loss of life, destruction, and displacement (Government of Rwanda 2012; Mullins 2009;Verpoorten 2005). 1 The mass violence began shortly after the plane carrying Rwandan President Juve ´nal Habyarimana was shot down on April 6, 1994, and subsided after the Rwandan Patriotic Front (RPF), a rebel group of Tutsi refugees, seized control of the Rwandan capital and established a new government that July (Des Forges 1999;Prunier 1995).A number of high-ranking officials within Rwanda, particularly those associated with the late President's political party, orchestrated the genocide, though ordinary Rwandan citizens perpetrated the majority of the violence (Mamdani 2001).
Numerous scholars have analyzed these perpetrators2 (e.g., Fujii 2009;Straus 2006), contributing to a robust body of literature regarding individuals who commit genocide, otherwise known as ge ´nocidaires. 3Many of these studies (e.g., Adorno 1950;Arendt 1963;Brustein 1996) are grounded in the works of Stanley Milgram (1974) and Philip Zimbardo (2007), whose experiments demonstrate how psychologically "normal" people could commit seemingly unthinkable acts under particular conditions (but see Perry 2013).Their research shows that the performance of cruel and inhumane actions is a learned process, particularly when demanded by those in positions of power, leading scholars to suggest that ge ´nocidaires are generally not psychologically unstable but are, in many ways, ordinary 1 Estimates of deaths, rapes, and other forms of victimization vary widely and are contested; issues range from who is included in death tolls and the government's vested interest in claiming victimization to the underreporting of Tutsis in census data and the difficulty inherent in quantifying victimization during mass violence (see Lemarchand 2013). 2 While we use the term "perpetrator," we recognize that there is much heterogeneity among perpetrators, ranging from those who plan the genocide to those who execute it in various settings.3 As our article makes clear, the label "ge ´nocidaire" is complex and highly contested.We use the term given its prevalence in genocide studies, though we acknowledge the important distinction between labels applied to individuals and those applied to their actions.
social actors-a paradox that remains the hallmark of research on perpetrators (Alvarez 1997;Browning 1998;Owens, Su, and Snow 2013;Waller 2007).Scholarship on political elites argues that genocide is the outcome of rational decision making that government officials turn to when their authority is threatened (Midlarsky 2005).Other works address those implementing the violence, analyzing the situational influences of collective action frames (Hagan and Rymond-Richmond 2009) and bureaucratic hierarchies (Bauman 1989).While much of this scholarship relies upon historical records, several studies examine the accounts of perpetrators themselves.Scott Straus (2006) contends that uncertainty, fear, and anger in the context of war can influence participation, as people engage in violence in response to encouragement, intimidation, or coercion.Others likewise argue that peer and familial networks (Fujii 2009;McDoom 2013McDoom , 2014) ) contribute to individual participation in genocide or similar mass violence.Departing from these invaluable studies, we analyze the accounts of genocide perpetrators and build upon work such as Diana Scully and Joseph Marolla's (1984) study of convicted rapists, which illustrates how excuses and justifications help reveal the cultural repertoire of potentially neutralizing accounts.
The social psychological tradition of symbolic interaction theorizes the self as continuously constructed through interaction with other social actors (Cooley 1922;Mead 1934).The looking-glass self (Cooley 1922), for instance, asserts that individuals see themselves through the eyes of others, shaping both identity and behavior in processes of role taking and reflected appraisals.An individual's actions must, accordingly, be viewed within the context of social relations.Relatedly, Erving Goffman's (1959) notion of impression management illustrates the techniques individuals employ to influence the perceptions of others and maintain desired impressions.The presence of an "audience"-real or imaged-dictates whether and how individuals work to be viewed in a particular way.People often exaggerate actions they believe others will view favorably when they are "front stage," while also concealing or minimizing socially undesirable actions or attitudes that may be detrimental to their identities or interactions.This "stigma management" or selective non-disclosure protects the self from feelings of disgrace or shame (Goffman 1963) that may have resulted from not abiding by others' normative standards or expectations, thus indicating why people minimize or conceal stigmatizing conditions, attitudes, or behaviors in social interactions.
Identity management is particularly salient when people have failed to abide by socially accepted standards of behavior.In such cases, individuals often provide explanations or justifications for their conduct as a means of impression and stigma management and to restore the meanings, interactions, and identities damaged by their words or actions.C. Wright Mills (1940) argues that individuals continuously engage in ongoing processes of negotiation in an attempt to repair damaged identities by employing "vocabularies of motive," or words that are used to make behavior appear acceptable and aligned with mainstream societal norms and values.These vocabularies, which often involve excuses, justifications, or other forms of accounts (Scott and Lyman 1968), are guided by perceptions of others' expected reactions to unacceptable conduct.Identity maintenance efforts-which occur and unfold across a range of settings, including courtrooms (Gathings and Parrotta 2013)-are thus constituted by a process of negotiation during which social actors engage in the continuous mitigation of stigma and mediation of tarnished identities.

T E C H N I Q U E S O F N E U T R A L I Z A T I O N
While these theories pertain to any behavior that may be stigmatizing or otherwise fall outside the bounds of socially accepted norms, several classic applications have focused on identity management and crime.Most notably, Gresham M. Sykes and David Matza (1957) suggest that people use "techniques of neutralization" to neutralize their guilt and justify their participation in activities perceived as criminal.In line with the interactionist tradition, the neutralization framework rests upon the assumption that individuals who engage in deviant acts typically share mainstream societal values, which are suppressed in order to facilitate criminal behavior.
Specifically, Sykes and Matza (1957) identify five techniques of neutralization, which we term the "classic techniques": denial of responsibility-negating personal accountability and breaking the link between oneself and one's actions; denial of injury-claiming that no one was injured by one's actions, thus breaking the link between the act and its consequences; denial of the victim-claiming that the injury inflicted was, in fact, retaliation or punishment, thus transforming the victim into the wrongdoer; condemnation of the condemners-attacking one's critics, claiming that their own condemnation is unjustified; and appeal to higher loyalties-highlighting one's conflicting roles and adherence to higher principles in choosing one action over another.
While these techniques of neutralization were originally theorized in reference to juvenile delinquency, they have since been extended to adults and to the more serious crimes of rape (Bohner et al. 1998;Scully and Marolla 1984), murder (Levi 1981), and sex trafficking (Copley 2014).Indeed, techniques of neutralization are well established across a wide array of groups and behaviors (e.g., Christensen 2010;Eliason and Dodder 2000).This body of work has inspired additional techniques-theorized primarily in the context of property offenses and white-collar crime (Maruna and Copes 2005)-that move beyond the five classic techniques noted above.Two extensions pertinent to the study of genocidal violence include metaphor of the ledger (Klockars 1974), which involves claims that someone also performed good deeds and thus offsets undesirable actions by highlighting those better aligned with normative behavior, and defense of necessity (Minor 1981), which suggests that an individual's survival depended upon a criminal act and consequently challenges its characterization as deviant (Dunford and Kunz 1973;Geis 1967).Like Sykes and Matza's (1957) techniques, these extensions are applicable across a range of criminal activity (e.g., Gruber and Schlegelmilch 2014;Harris and Daunt 2011), though they have not been explored in the context of genocide.
Only two studies apply "classic" neutralization theory to genocide.Alexander Alvarez (1997) employs Sykes and Matza's framework to examine the process by which ordinary German citizens adjusted to participating in the Nazi Holocaust.He contends that their techniques were used to "neutralize or repackage internal values and beliefs that were antithetical to genocide" during the Holocaust (p.169).In doing so, he proposes a new technique-the denial of humanity-and argues that the dehumanization of Jews prior to the genocide helped facilitate the atrocities of the Holocaust.Similarly, Frank Neubacher (2006) illustrates the presence of Sykes and Matza's techniques in Nazi Germany by analyzing Himmler's 1943 speech calling for the extermination of Jews.
As these studies illustrate, the classic techniques of neutralization and their extensions are typically conceptualized as preceding-and thus facilitating-criminal or deviant acts.For Sykes and Matza, these techniques enable crime and are drawn upon before engaging in criminal behavior.Accordingly, in defining techniques of neutralization, they distinguish between neutralizations preceding criminal activity and rationalizations made after the fact (1957:666-67).A number of scholars similarly emphasize the presence of neutralization techniques prior to criminal activity (Agnew 1994;Alvarez 1997).However, without definitive empirical evidence supporting the presence of neutralization techniques proceeding or following deviant acts, many assert their purpose in both time periods (e.g., Cohen 2001;Hirschi 1969;Maruna and Copes 2005).Indeed, Sykes and Matza's "classification of accounts," as Stanley Cohen (2001) notes, "functions after the act to protect the individual from both self-blame and blame by others, and before the act to weaken social control . . .and make delinquency possible" (p.60).
In the case of genocide, it is plausible to conceptualize neutralizations as both preceding and following genocidal acts: to overcome widely shared norms against murder, rape, and other crimes committed during these atrocities, perpetrators may first neutralize internalized controls that prohibit these behaviors; later, in rationalizing such behaviors to others, they may draw from or expand upon such neutralizations.Though defendants likely utilized neutralization techniques before and during the Rwandan genocide, we specifically address their post-genocide role in impression management and identity negotiation.We thus build upon a line of research that considers whether and how highly stigmatized individuals engage in "identity work" (Einhowner 2006;Futrell and Simi 2004;Snow and Anderson 1987).While it may seem trivializing to apply concepts initially developed to explain delinquency to the context of genocide, our study builds on prior research (Alvarez 1997;Neubacher 2006), the established finding that most ge ´nocidaires are not psychologically abnormal but rather act in accordance with mainstream societal norms, and the contention that "political accounts most often follow the same internal logic and assume the same social function as ordinary deviant accounts" (Cohen 2001:77).

M E T H O D O L O G Y
To analyze the techniques of neutralization employed by accused ge ´nocidaires, we examine official transcripts from the International Criminal Tribunal for Rwanda (ICTR).Established by a United Nations Security Council resolution in November of 1994, the ICTR was created "for the sole purpose of prosecuting persons responsible for the genocide and other serious violations of international humanitarian law," including systematic rape, murder, and extermination, that occurred in Rwanda between January 1, 1994and December 31, 1994(United Nations 1994).Located in Arusha, Tanzania, the ICTR was one of several international criminal tribunals created since the Nuremberg trials following the Holocaust.When it closed in December 2015, it had indicted 93 individuals and completed 75 trials.
Because we are interested in accused ge ´nocidaires' accounts of their actions, we analyze the transcripts from trial days during which defendants testified on their own behalf.We determined these days from each ICTR Summary Judgement and subsequently obtained transcripts from the UN-ICTR Judicial Database or through direct communication with ICTR staff.This focus excludes the testimonies of other witnesses, as well as parts of testimonies in which the defendants were asked to recite material previously spoken or written by either themselves or another individual.These exceptionally rich narrative data illustrate precisely how defendants construct their accounts, negotiate their identities, and employ neutralization techniques in a highly charged public setting.
In total, we identified 47 defendants who testified on their own behalf, and we obtained the transcripts of 41.The remaining six transcripts had yet to be redacted or translated into English at the time of data collection.Due to the sheer quantity of testimonies (some transcripts of defendant testimony contain over 1,000 pages), we sampled two-thirds of these defendants, yielding 27 testimonies spanning 152 days.Each of the 27 defendant's testimonies lasted between 1 and 17 days; because the average day of testimony consists of 80 pages, this generated over 10,000 pages of text.We coded all but the missing portions of testimony. 4Given variation in lengths of testimony, we standardize our findings by calculating the average number of techniques used per day for each defendant, and all figures below reflect this. 5e focus on the neutralization of accusations and thus include both those who were acquitted alongside those who were convicted.We anticipate that standing trial for the "crime of crimes" will in itself tarnish defendants' identities.Nonetheless, in order to examine potential differences, we ensured that the percentage of those found guilty and acquitted in our sample, as well as those whose cases remained on appeal, mirrored these proportions across all ICTR defendants.6At the time of our sampling, roughly 60 percent of defendants had been found guilty, 20 percent had been acquitted, and 20 percent had cases on appeal-both within our sample and across all cases.Since this time, the Tribunal has finished its final cases, finding 61 defendants guilty and acquitting 14.Within our sample, 19 defendants received sentences and 8 were acquitted.
The 27 defendants share a distinctive profile.All are men who were between the ages of 32 and 59 in 1994.Many had held comparatively high-ranking positions in Rwanda, whether as political leaders, military leaders, or wealthy businessmen.Nearly all were indicted for complicity in genocide and either genocide or conspiracy to commit genocide.Many were also charged with direct and public incitement to commit genocide, as well as crimes against humanity, including rape, murder, and extermination.Convicted defendants received sentences ranging from 12 years to life imprisonment, as displayed in Appendix A.
Testimonies were originally transcribed in Kinyarwanda, then translated into French and later into English.To be sure, mistranslations and misinterpretations occurred, as with all international courts or tribunals (Combs 2010; Namakula 2014), due to resource constraints, diverse cultural understandings, and few opportunities to identify mistranslations of Kinyarwanda (Rwandans were not active in the trials, and some words in Kinyarwanda are not easily translated).However, to reduce potential issues, interpreters often asked for clarification of words and meanings, and lawyers frequently instructed defendants to speak slowly.Bearing this in mind, we focused our coding on general ideas in addition to specific words and descriptions.
We coded transcripts in several phases.We initially coded 15 defendants' testimonies inductively, and based on preliminary analysis, developed a deductive coding scheme that included the "classic" techniques (identified by Sykes and Matza) and their extensions (Klockar's metaphor of the ledger, Minor's defense of necessity, and Alvarez's denial of humanity), which we used to code all 27 defendants' testimonies.Other techniques emerged through the coding process, however, and we discuss these in our findings below.The first three authors coded transcripts using Atlas.ti, a qualitative data analysis software. 7We assessed inter-coder reliability throughout the data collection and analysis process and took considerable effort to resolve coder uncertainties and discrepancies by meeting frequently and coding the same transcripts to confirm agreement with codes that had been applied, which remained consistently at or above 70 percent.

F I N D I N G S
In all courtroom settings, defendants have particular incentives to portray themselves in a positive light by rationalizing past actions, salvaging tarnished identities, and attempting to avoid or minimize punishment.Many of the defendants who stood trial at the ICTR were once respected individuals of relatively high social status.Accordingly, we find that they broadly employ vocabularies of motive (Mills 1940) through their use of particular techniques of neutralization (Sykes and Matza 1957).Evident in their testimonies, these techniques serve to manage and preserve defendants' social identities while also legitimizing their behavior.Like social actors in other settings, ICTR defendants engage in identity work to create and sustain social acceptance.This identity work may reflect defendants' psyches and inward attempts at rationalizing any alleged behavior (e.g., Snow and Anderson 1987); however, we focus on its function for identity maintenance.We first assess the frequency and nature of the classic neutralization techniques and their extensions, followed by a discussion of the additional genocide neutralization techniques we observe in defendant testimony.Finally, we consider the factors that influence the frequency and nature of the techniques employed.

Classic Neutralization Techniques and Their Extensions
All defendants in our sample employ Sykes and Matza's (1957) neutralization techniques, but, as previous studies have found (Copes 2003;Hollinger 1991;Mitchell and Dodder 1983), there is great diversity in the amount and types of techniques they employ.Once standardized to account for varying lengths of testimony, some defendants employ one or two with great frequency, while others select a different technique (or set of techniques) on which to rely.All defendants utilize between 1 and 14 techniques per day, with an average of 5.As seen in Figure 1, clear patterns in technique use are evident.
In particular, we find frequent use of the condemnation of the condemners technique.This technique allows defendants to shift blame away from themselves and onto their accusers while denouncing both the accusations against them and those levying the allegations.Condemnations frequently focus on the international community, genocide survivors and their families, and the Rwandan Patriotic Front (RPF), which ended the genocide and remains the country's dominant political party under President Paul Kagame.Defendants often express frustration with these actors and, in doing so, also question the legitimacy of the ICTR.Some defendants are critical of the Tribunal's founding, noting that while the international community failed to directly intervene in the violence, it wasted no time creating a criminal tribunal in its aftermath.Others contend that some individuals unfairly evaded prosecution because the ICTR was conceived to prosecute both former Rwandan political elites and members of the RPF but ultimately did not pursue RPF crimes.Notably, such condemnations reflect existing narratives surrounding the ICTR, which has been criticized by academics and policymakers for these very reasons (Cruvellier 2010;Peskin 2005;Popovski 2012).
A retired Lieutenant Colonel exemplifies this condemnation, claiming, "Counsel, as a matter of fact, the charges against me . . .have been made up to fulfill a political scheme, that of the current Rwandan authorities" (Simba ICTR-01-76).Similarly, another former Lieutenant Colonel contends: I have always rejected this simplistic approach of creating disorder by seeking scapegoats.The issue of the Rwandan disaster, or massacre, and the responsibility for that massacre is known . . .And yet, you are looking for scapegoats . . .that is how politics is (Bizimungu, A. ICTR-00-56).
In these instances, defendants are able to capitalize on critical narratives of the ICTR by expressing their suspicion of its legitimacy and the intentions of those involved.Former military chaplain Rukundo demonstrates this, stating, "When somebody has chosen to accuse you, they can fabricate any kind of story against you.Once again, these are words that have been spoken, which may never be verified . . .Those utterances are fabrications designed to incriminate the poor me" (ICTR-01-70).Levying a critique of the international community, the former Minister of Foreign Affairs contends, "The international community seemed to be concerned with the repatriation of foreigners that  Identity Work Among Accused Genocide Perpetrators 7 were resident in Rwanda . . .everybody was trying to save his skin at the time . . . the country was being abandoned to itself" (Bicamumpaka ICTR-99-50).These and other defendants attempt to locate fault within their accusing parties and challenge the condemnation they face, thus diverting attention from their alleged crimes by shifting blame elsewhere.Defendants also frequently employ the denial of responsibility technique.The ICTR purposefully selected defendants in accordance with the international community's expectation that the architects of the genocide be brought to justice.This included people who "planned, instigated, ordered, committed or otherwise aided and abetted in the planning, preparation, or execution of a crime" (United Nations 1994, Article 6).While some defendants arguably did not orchestrate the violence, all defendants were positioned as powerful leaders, regardless of their actual level of authority.Additionally, most of the defendants previously occupied respected positions of authority, ranging from government ministers and military commanders to local governors, religious leaders, and successful businessmen.This implies their responsibility for and ability to direct the actions of their subordinates.Accordingly, many defendants attempt to break the link between themselves and acts of violence by denying their authority (Cohen 2001).
As former mayor Kajelijeli contends, "If I am one of the citizens of Ruhengeri and you ask me what I should have done, tell me in what capacity I would have done it as an ordinary citizen . . .This is all I could do in my position as a citizen who had no other capacity, other power or other authority under the law" (ICTR-98-44A).A former physician and hospital director similarly asserts, "As concerns those who came here to tell you that . . .I was supposed to defend them.But I did not have enough means to assist those who were in distress.On that basis, therefore, I cannot be held accountable for all the damage that was done to the victims" (Ntakirutimana 1: ICTR-96-10; 2: ICTR-96-17).Likewise, former Rwandan Army commander Ntabakuze states, "I agree that people were killed, but . . .neither myself nor the men under me were responsible for those killings" (ICTR 98-41).In each case, the defendants claim a lack of authority and assert their inability to stop the violence.
While condemnation of the condemners and denial of responsibility are quite prevalent, we find far fewer instances of appeals to higher loyalty, denial of injury, and denial of the victim.The appeals to higher loyalty technique allows individuals to attribute their actions to forces or values greater than themselves.This is exemplified when Ntakirutimana states, "In my country we don't question the forces of law and order much, so if you're told something like that, well, you don't argue" (1: ICTR-96-10; 2: ICTR-96-17), and when Akayesu, a former mayor, recalls, "But since the orders came from above, since they were broadcast over the radio, the population was asked to erect roadblocks and so . . .we erected the roadblocks" (ICTR-96-4).Yet, while parallel to denial of responsibility, appeals to higher loyalty largely depend upon defendants confessing to those very actions-an infrequent scenario in trial settings.
Rarely do defendants employ denial of injury to claim that their actions did not affect anyone; instead, this technique is used sparingly yet purposefully.For example, Akayesu notes, "I'm certain that rape never took place within the premises of the bureau communal or on the land belonging to the bureau communal" (ICTR-96-4).The limited or partial use of this neutralization technique enables defendants to construct narratives conveying their innocence at particular times and in particular locations without denying the violence taking place on a wider scale.
Indeed, no defendant denies the occurrence of mass violence, the reality of which makes denial difficult.Many do not label the violence as genocide, however, and while we do not consider this denial of injury, the labels they do use are nonetheless informative.Defendants characterize the genocide as war, discussing the violence as a "resumption of hostilities" with the RPF stemming from conflict in the early 1990s (see Figure 2).As state-sanctioned violence, "war" connotes legitimate violence and stands in stark contrast to "genocide."Discussing the violence as war consequently neutralizes what would otherwise be deemed heinous crimes.As former prosecutor Nchamihigo contends, "As to whether there was genocide in Cyangugu, now, I would say there were massacres.But then to move from there and say there was genocide, I do not share that view of yours" (ICTR-2001-63).
Such instances, as well as those in which defendants deny particular instances of violence, do not fully constitute denial of injury (i.e., denying the occurrence of genocidal violence outright8 ) by our coding scheme but rather represent the minimization of injury by virtue of reframing "the interpretative framework placed on these events" (Cohen 2001:105-06).9This minimization or reframing distances defendants from accusations of direct involvement in criminal activity and allows them to both call into question the meaning of "genocide" and deny that one occurred in 1994, even after the ICTR had taken judicial notice of the genocide in 2006 (Mamiya 2006).
Denial of the victim, in which defendants transform the real victim into a wrongdoer, is also used infrequently.Ntagerura, a former government minister, provides one example: "This wasn't violence directed against Tutsi, it was directed against people who had attacked and others who were supposed accomplices of those assailants" .While this statement serves to distance the general Tutsi civilian population from the Tutsi RPF, we still find that defendants do not explicitly suggest that any Tutsis deserved the violence inflicted.Despite framing the violence as war, defendants rarely went so far as to suggest that those who were killed were wrongdoers themselves.This may be linked to the scale of the violence and the inability to deny its occurrence, such that half of our sample (14 defendants) completely eschews both the denial of injury and denial of the victim techniques.
To account for extensions of the classic techniques, we also coded for the metaphor of the ledger (Klockars 1974) and defense of necessity (Minor 1981) techniques (included in Figure 1 despite their infrequent use), as well as Alvarez's (1997) denial of humanity technique (which we find no instances of and thus exclude from our figures).We document negligible use of these techniques, likely because they rely upon admissions of wrongdoing.
Overall, of the five classic techniques and their extensions, defendants employ denial of responsibility and condemnation of the condemners with greatest frequency.While the use of these techniques is partially attributable to the trial setting, it is not only defendants' assumed desire for a particular case outcome that accounts for their use and non-use of specific techniques.Many of their  comments convey the assumption of unfair trials or a likely guilty verdict regardless of their testimonies.Nonetheless, defendants employ techniques of neutralization to maintain a valued social identity despite strong evidence of having broken international norms.Thus, maintenance of identity and presentation of self (Goffman 1959) represent two salient driving forces behind the use of these techniques, regardless of whether defendants ultimately received convictions or acquittals.

New Techniques to Neutralize Genocide
Because the respective classic neutralization techniques (Sykes and Matza 1957) and their extensions (Alvarez 1997;Klockars 1974;Minor 1981) do not take full account of the means by which defendants rationalized their actions, we propose two related techniques of genocide neutralization to comprehensively account for defendants' narratives: victimization and appeals to good character.The victimization technique represents statements in which defendants portray themselves as having suffered losses of some kind (e.g., family, friends, or property) or discuss their own victimization or that of their ethnic group.For example, former mayor Kajelijeli claims, "I felt it was possible for me to die because I had been under permanent threat.I was being persecuted" (ICTR-98-44A).
Many defendants claimed that Hutus were also killed during the genocide, suggesting that an emphasis on targeted violence against Tutsis has overshadowed the victimization of Hutus.To be clear, the violence did target Tutsis, though Hutus were killed for multiple reasons, whether because they posed a threat to hardliners, refused to participate, were mistaken for Tutsis, or were killed as the RPF re-initiated the civil war (Des Forges 1999).Relying upon these and other examples, defendants invoke Hutu losses by claiming to have also faced harassment and oppression."And let me point out," a former Minister states, "that many Hutu were also killed under the same circumstances" (Bicamumpaka ICTR-99-50).A former mayor similarly asserts, "There were even Hutus who were killed because they were mistaken for Tutsis" (Bagilishema ICTR-95-1A).These defendants contend that they or other Hutus suffered some form of loss or were victims of attacks, harassment, and condemnation.Such claims are enabled, in part, by the mass nature of genocidal violence; and while they are similar to condemnation of the condemners-in that defendants shift the blame from themselves-these claims do not include the necessary component of accusing condemners.
The other genocide neutralization technique we identify-appeals to good character-reflects statements in which defendants assert their good deeds or admirable character attributes that they contend render them incapable of committing genocidal crimes.These good deeds include saving others (particularly Tutsis), attempting to stop violence, and calling for peace.As former sub-prefect Ntawukulilyayo states of the massacres in and around his commune, "I was saddened by that news as well as frightened . . .I did not have enough means to act in that situation.However, I did not fold my arms.I did what I had to do and what I could do" (ICTR-05-82).Similarly, former mayor Mpambara notes, "I tried to explain to the population that they must live harmoniously as brothers and sisters . . .and be above ethnic groups" (ICTR-01-65).Defendants often claim to have exerted all available options to stall the ongoing violence.Sagahutu, a former military commander, states, "Everywhere we did everything possible to ensure that [the] civilian population would suffer the least from the consequences of the war" (ICTR-00-56).Each of these statements depicts well-intentioned defendants who were incapable of committing genocidal atrocities and who attempted to halt the violence.
In addition to asserting their roles as rescuers or saviors, defendants highlight other positive character traits by denying any personal ethnic animus against Tutsis, illustrating why we find no instances of Alvarez's (1997) denial of humanity technique.In his testimony, wealthy businessman Munyakazi notes, "I had Tutsis that were in my house, I mean Tutsis that I was hiding.I therefore could not leave the Tutsis in the house and go and kill other Tutsis . . .I had nothing against the Tutsi" (ICTR-97-36A).And Ntawukulilyayo asserts, "I never mistreated the Tutsi.I never said that the Tutsi were not full-fledged human beings" (ICTR-05-82).
Others illustrate their good character by expressing remorse for the violence.Ntakirutimana illustrates such remorse, stating: What I would like to tell the Court is that we were extremely affected by what happened in Rwanda during this period of April to July 1994 . . .I would like to seize this opportunity to ask all persons of goodwill to work for peace-to work for peace before any conflicts erupt (1: ICTR-96-10; 2: ICTR-96-17).
While these accounts parallel the metaphor of the ledger technique, ICTR defendants stop short of admitting guilt when cataloging their virtuous acts.Beyond this, metaphor of the ledger focuses explicitly on good deeds, while defendants in this setting also emphasize their good character.Being accused of genocide may be perceived as an attack on one's character, and defending this character against the stigma associated with genocide participation is important for the management of one's presentation of self.As in other social settings, the courtroom creates the social expectation that defendants portray themselves in a positive light and express their acceptance of and support for collective norms and values.
Paralleling their use of the classic techniques and their extensions, defendants employ the victimization and appeals to good character techniques to varying degrees: over one third of defendants use these techniques between 1 and 12 times per day of testimony,10 though 5 never utilize either.While one third of defendants employ Sykes and Matza's (1957) classic techniques slightly more often than the additional genocide neutralization techniques, Figure 3 illustrates that, in the aggregate, defendants employ these new techniques with far greater frequency.The appeals to good character technique, in fact, is used more often than all the classic techniques combined.
The overall variation we observe in type and frequency of technique utilization is not tied to case outcome, although three of the four defendants who employed classic techniques with greatest frequency (at least 12 times per day) were convicted, and two of the three defendants most frequently using the additional genocide neutralization techniques (at least 19 times per day) received acquittals.We also do not find that defendants' defense attorneys shape their technique use, but there does appear to be some association with defendants' levels of authority, such that those of comparatively moderate authority (e.g., mayors of communes) employ neutralization techniques more often than those with lower or higher levels of authority (see Appendix B).Finally, the only temporal pattern we identify is more frequent utilization of our proposed techniques in earlier testimonies.While ICTR judges cited "assistance to victims"-a crucial component of the appeals to good character technique-as the most common mitigating factor in sentencing (Hola, Smeulers, and Bijleveld 2011) 11 , we observe this technique amongst the defendants testifying in earlier years, thus suggesting an emphasis on managing their identities in addition to or beyond vying for a reduced sentence.
Ultimately, these techniques serve to maintain defendants' identities as they portrayed themselves as socially respectable individuals capable of adhering to and embodying accepted behaviors and values.To be sure, we recognize a distinction between these techniques and defendants' other claims of innocence present in court settings; while the two are not necessarily mutually exclusive, we attend to the identity work at play, which may be-but is not always-enmeshed in conceptions of guilt.As such, our proposed genocide neutralization techniques capture behavior beyond claims of innocence, and consideration of these techniques may thus aid in developing a more comprehensive understanding of accused ge ´nocidaires' identity work.

D I S C U S S I O N
Two primary patterns emerge in our analysis of defendant testimony.First, we find limited use of the appeals to higher loyalty, denial of injury, and denial of the victim techniques-all of which we consider among the "classic" neutralization techniques-as well as some of their relevant extensions.Second, we document frequent reliance on two "classic" techniques (denial of responsibility and condemnation of the condemners) and two proposed "genocide neutralization" techniques (victimization and appeals to good character).We contend that these findings result from four interrelated characteristics of the context in which defendants' testimonies took place: the trial setting and related international media attention; the nature and political salience of genocide; the relatively high social status that defendants selected for prosecution once occupied; and existing narratives surrounding the legitimacy of the ICTR.Against the backdrop of each factor are ongoing processes of impression management and the imperative to advance a positive presentation of self (Goffman 1959).
Defendants' appearance on the world stage of an international trial heightens the salience of the internationally accepted norms and values-institutionalized through mechanisms such as the UN Convention on the Prevention and Punishment of the Crime of Genocide-with which their testimonies are expected to align.While the ICTR did not receive media attention on a daily basis, verdicts were nevertheless frequently published in major news outlets.It is likely that defendants assumed there would be international media coverage given that it was an international tribunal (indeed, a LexisNexis search identifies nearly 1,000 articles or television mentions).In contrast to previous studies of accused or convicted ge ´nocidaires, in which interviews are conducted after rendered verdicts and decided sentences (Fujii 2009;Hatzfeld 2005;Straus 2006), the stakes are high for defendants at international tribunals, and their positions as accused genocide perpetrators influences their reliance on neutralization techniques.The frequent utilization of the appeals to good character technique and the infrequent use of denial of injury, denial of the victim, and the classic techniques' extensions (Alvarez 1997;Klockars 1974;Minor 1981) reflect the influence of this situational context.Simply put, denying the genocidal violence or the humanity of Tutsis would be deleterious to both their cases and their public image (although it may have elevated their position amongst others, such as certain Hutu exiles).
In genocide, as elsewhere, the precise neutralization techniques used by wrongdoers are offense specific (Maruna and Copes 2005).Unable to deny the 1994 genocide due to the nature and evidence of the crime (Cohen 2001), defendants instead deny personal involvement in mass killings and atrocities.Consequently, denial of injury is used sparingly, but minimization of injury is commonplace.This enables defendants to acknowledge that violence had indeed taken place, but only in the context of specific dates and locations.They thus construct testimonial narratives to convey their innocence without denying the broader occurrence of violence.The political salience of genocide likewise influences the techniques employed.Given the horror, death, and devastation wrought by genocide, defendants heavily rely on appeals to good character to neutralize their association with this crime.They also avoid using or, at times, unequivocally contest the term "genocide," instead referencing war and killing as a means of reframing their alleged actions.
By denying their responsibility and any personal ethnic animus, championing their efforts to stop violence, and expressing remorse for such events, defendants distance themselves from the stigma and consequences associated with accusations of direct participation in genocide.These efforts are crucial given many defendants' relatively high social status prior to and during the genocide.Because the ICTR purposefully targeted those it deemed most responsible for the mass violence, it presented defendants as holding the necessary authority to orchestrate and execute the genocide, which provided a powerful impetus for defendants to deny this possibility.
The prevalence of the condemnation of the condemners technique speaks to broader issues surrounding the ICTR concerning the legitimacy of its origins and practices.Numerous scholars and policymakers have condemned the Tribunal for not only failing to prosecute all actors involved in the violence, but also not indicting a single RPF actor (Cruvellier 2010;Haskell and Waldorf 2011;Peskin 2011).As the RPF has dominated Rwanda's government since 1994, many suggest that Rwandan Government opposition curbed such indictments, resulting in what many term "victor's justice" (Peskin 2011).These legitimacy issues readily facilitate defendants' challenges to their accusers, extending beyond the Tribunal's authority to the international community, the RPF, and survivors of the genocide.The extent to which defendants personally share these criticisms matters less than their ability to strategically employ this technique to shift blame from themselves and onto their accusers.
While we analyze defendants' neutralization techniques as a form of rationalization used in identity work following the genocide, we suspect that the techniques defendants employ in their testimonies may have been present both before and during the genocide.Our data limit our ability to assess precisely when techniques emerged for defendants or how they may have changed over time, so we do not address this here.Nevertheless, neutralization techniques appear to play an important role in the construction of narratives of violence and in processes of identity negotiation and impression management for accused ge ´nocidaires.Indeed, the frequency with which defendants seek to convey their positive, socially acceptable behavior during the genocide illustrates the effort exerted to maintain or rebuild their pre-genocide identities.These efforts reflect both instrumental and expressive motives; in defending their innocence, defendants struggle to preserve their social self.Through their frequent use of neutralization techniques, defendants engage in impression management and the negotiation of expressive control to avoid stigma and achieve social acceptance despite their tarnished identities (e.g., Presser 2004), regardless of whether this use reflects personal rationalizations to themselves.While mainstream societal norms and expectations vary, international norms condemn genocide and related crimes in both legal and moral terms.In the context of these pervasive norms guiding the ICTR, defendants employ neutralization techniques to manage the disconnect between the atrocities they are accused of committing and the legal and moral standards to which they are held.In doing so, they attempt to legitimate their behavior and reconstruct a socially accepted identity.
The use of neutralization techniques not only facilitates this identity work but may also serve as a means of subtle genocide denial, and defendants' contestation over terminology and statistics reflects broader struggles to construct historical narratives.Again, while defendants never explicitly deny the genocide-though they question whether the events of 1994 constitute it (which, to some, equates to outright denial; see Cohen 2001)-many of these techniques allow them to do so in more indirect ways.Several of the 12 genocide denial techniques identified by Gregory Stanton (2005) parallel the classic and proposed techniques we observe among ICTR defendants, such as "attack the motivations of the truth-tellers"-which bears resemblance to the condemnation of the condemner's techniqueand "blame 'out of control' forces for committing the killings," which parallels denial of responsibility.By suggesting that the classic and emerging neutralization techniques facilitate subtler forms of denial, our analysis informs research examining the denial of international crimes by state officials and authorities (Cohen 2001;Del Rosso 2014;Stanton 2005).
Our analysis also expands the scope conditions of Sykes and Matza's (1957) neutralization techniques, which were originally employed to explain juvenile delinquency and minor deviance.We thus establish a basic connection between what are arguably the least serious and most serious forms of criminality, while illustrating the relevance of neutralization techniques to the crime of genocide (Alvarez 1997;Neubacher 2006).Moreover, we highlight the politically consequential relationship between micro-level processes of identity negotiation and macro-level outcomes for contested historical narratives of violence.

C O N C L U S I O N
After the 1994 Rwandan genocide, the UN-mandated International Criminal Tribunal for Rwanda quickly began indicting those deemed most responsible for the atrocities.Many of these accused genocide perpetrators testified on their own behalf, and these testimonies illuminate their efforts to neutralize their alleged actions.Like others attempting to manage a tarnished identity, they sought to rationalize their behavior through identity work.Analyses of these ICTR defendants' testimonies indicate widespread use of certain techniques of neutralization (Sykes and Matza 1957), particularly denial of responsibility and condemnation of the condemners.Nevertheless, these techniques fail to encompass the full range of defendants' rationalizations, and we propose two genocide neutralization techniques-victimization and appeals to good character-the latter of which emerges more often than all classic techniques combined.Consideration of both sets of techniques more accurately represents how defendants selectively disclose or exaggerate information as they distance themselves from the genocidal crimes for which they stand trial.
The core social psychological concepts of identity maintenance and the presentation of self (Goffman 1959;Mills 1940) retain considerable explanatory power in understanding how defendants construct narratives of genocide that may be implicated in broader political struggles and restore a positive self-concept and social identity.Our analysis of neutralization techniques in this context provides a deeper understanding of how and why accused genocide perpetrators engage in rationalization, impression management, and identity maintenance.In expanding the scope conditions and applications of neutralization theory, we also suggest that these techniques may facilitate subtle genocide denial, a finding that carries important repercussions for historical narratives of the genocide.Additional research is needed, however, particularly regarding the timing of neutralizations in genocidal crimes, their use in legal proceedings, and the co-constitution of testimony alongside wider historical narratives.In light of the tremendous social toll wrought by genocide-and the social imperative to do justice-further investigation of perpetrators is surely merited.

Figure 1 .
Figure 1.Standardized Frequency of Classic Neutralization Techniques and Their Extensions Note: Values along the Y-axis reflect the sum of instances each technique is used divided by the total adjusted days of all defendant testimony.

Figure 2 .
Figure 2. Standardized Frequency of 1994 Genocide Characterizations Note: Values along the Y-axis reflect the sum of instances each word is used divided by the total adjusted days of all defendant testimony.

Figure 3 .
Figure 3. Standardized Frequency of Classic (with Extensions) and Additional Genocide Neutralization Techniques Note: Values along the Y-axis reflect the sum of instances each technique is used divided by the total adjusted days of all defendant testimony.