Killing neighbors: Social dimensions of genocide in Rwanda

1
Killing neighbors Social dimensions of genocide
in Rwanda

• Lee Ann Fujii
• PhD Candidate, GWU
• Africanist Summer Fellow, WWICS

2
Roadmap

• Research question
• Methodology and fieldwork
• Characteristics of the violence
• Process model for participation
• Findings
• Conclusion

3
Research question

• How do neighbors become genocidal killers of
  neighbors?
  
• Question applied to Rwanda
• Helps to explain how genocide could have occurred
  at all
  
• No mass participation, no genocide

4
Puzzle of participation

• Not rational for masses to participate
• Benefits accrue mainly to leaders
• Not in any individuals self-interest
• Ties should preclude neighbor violence
• Reciprocal ties social capital
• Social capital greater cooperation
• More difficult to kill intimately
• More difficult to kill someone one knows
• More difficult to kill at close range

5
Features of genocide in Rwanda

• Elite-organized and led but mass implemented
• Killers were ordinary farmers
• Hutu men b/w 30-40
• Married with children
• Many with Tutsi relatives and friends
• No special training in mass killing
• Killers knew their victims

6
Characteristics of violence

• Killers killed at close range
• Machetes, hoes, clubs
• Killers killed in groups
• Groups were large (15-50)
• Killers killed out in the open
• During daylight, in front of others
• Killings took ritualized form
• Chants, costumes, audience
• Hunting, encircling, burying

7
Approach

• Most approaches focus on elites v. masses
• Proposed approach
• Micro-level
• Disaggregates the masses
• Looks at vertical and horizontal relations
• Process-oriented
• Views genocide as process, not event
• Assumes changing, not static, motives and
  identities
  
• Subjective
• Looks at process from inside out

8
Methodology

• Archival research and document search
• 9 months of fieldwork in Rwanda
• 2 research sites
• One village prison in the north (Ruhengeri)
• One village prison in the south (Gitarama)
• 80 respondents
• Purposive sample
• Spectrum of participants/actors killers,
  rescuers, bystanders, resisters et al.
  
• 230 intensive interviews
• Strategy of repeat visits
• With ever smaller sample at each round

9
Main participants
Profiteer Leads and organizes genocide
Collaborators Denounce victims
Joiners Go along with what others are doing
10
Process model for Joiners

• Initiation
• Pull of social ties
• Continuation
• Constitutive power of group acts
• Exit
• Response to external change (e.g.,
  escalation/de-escalation of war)

11
Initiation stage

• In conditions of insecurity or uncertainty
• People go along with what others are doing
• Existing ties pull people in same direction
• People see few options or feel powerless to do
  otherwise
  
• Avoiding the pull requires conscious act of
  refusal

12
Initiation in Ruhengeri

• Rumor circulates that neighbor is hiding ibyitso
• Group of 15 plans day of attack
• At appointed day, group encircles house, throws
  rocks on roof
  
• Man comes out brandishing machete
• One in the group kills man instantly

13
Initiation v. Resistance

• Stefan
• No conscious decision
• Just ended up at neighbors house
• Edouard
• Believed what he heard
• Wanted to protect against the threat
• Gustave
• I refused to do anything that had to do with
  politics. They asked me to do night patrols but I
  refused.

14
Continuation stage

• In group contexts, people engaged in specific
  practices and acts of killing
  
• Whistle, chant, encircle, watch, dig holes
• Acts of killing constituted group as social actor
  called Interahamwe
  
• Interahamwe identity adhered only to groups
• Only in group (killing) contexts
• Killings make groups and groups make killings

15
Killing v. Not killing

• Groups killed, individuals did not
• When in groups, impossible to save
• When alone, killers did not kill
• Olivier
• True Interahamwe
• Involved in nearly every killing
• Accused of having Tutsi wife and harboring Tutsi
• Helps Tutsi boy escape killers

16
Findings

• Participation was socially embedded in set of
  dense ties
  
• Profiteers, Collaborators, Joiners, victims, and
  rescuers came from same families
  
• Joiners and victims were friends
• Joiners worked in groups of friends

17
Findings (cont.)

• Ties could help, but mostly a liability
• While some individual Interahamwe helped Tutsi
  friends
  
• Same person might also kill a stranger
• Rescuers rescued anyone, including strangers
• While some Hutu families helped Tutsi family
  members, Collaborators often came from inside the
  family
  
• Danger of ties
• Source of local knowledge, personal motives
• Source of momentum for joining in
• Very difficult for victims and Joiners to
  escape these

18
Conclusion

• Participation occurs through process of social
  interaction, not instrumental calculation
  
• Social ties not only shape neighborly violence,
  they also facilitate it