Docstoc

Barbarism and Intimacy Violence in Civil Wars

Document Sample
Barbarism and Intimacy Violence in Civil Wars Powered By Docstoc
					CIVIL WARS THROUGH
HISTORY: THE LOGIC OF
VIOLENCE
Stathis N. Kalyvas
Department of Political Science
Yale University
VIOLENCE IN CIVIL WAR
 A very long preoccupation with civil war
 One reason is their violence

 Two features

 Barbarity
       Arno Meyer: “If war is hell, then civil war belongs to
        hell’s deepest and most infernal regions”
   Intimacy
     Fratricide: Abel and Cain
     Montherlant: civil war is the war of neighbor against
      neighbor, rival against rival, friend against friend
THE QUESTION
   How to explain the barbarity and intimacy that
    characterize the violence of civil wars?
       Note that this is not the same as asking what causes
        civil wars
   What type of theoretical and empirical account
    succeeds in explaining both features at the same
    time?
OUTLINE

 Features of violence in civil war
 Four intellectual traditions

 A theoretical account

 Some empirics
BARBARITY
   Excessive violence; cruelty; atrocity
     The term is culturally specific
     Bombs vs. knives

 Post 1945 armed conflicts have been primarily
  internal with noncombatants as the primary
  target
 Number of victims
     WWII; Chad 1970s
     Civil wars as residual category of warfare

   Identity of victims
   Combatants vs. noncombatants
   Intimacy between victimizers and victims
INTIMACY

 Not all victims of a civil war die from the action
  of “intimates,” but many do
 Yet the perception of barbarity is partly
  motivated by the occurrence of intimate violence
A HIGH SCHOOL CLASS IN NORTHEAST
BOSNIA
CONVENTIONAL UNDERSTANDINGS OF
VIOLENCE
 Many facts from journalistic reports, historical
  case studies, and human rights reports
 …yet, little systematic analysis

 Violence is typically characterized as
     Madness
     Hatred
     Evil

 No real explanations
 It is possible to reconstruct four intellectual
  traditions that have grappled with this question
FOUR INTELLECTUAL TRADITIONS
  Hobbesian tradition: violence in civil wars is
   barbarous because of anarchy
     Recent version: new wars

  Schmittian tradition: violence in civil wars is
   barbarous because of polarization/enmity
     Recent version: ethnic conflict

  Foucauldian tradition: violence in civil wars is
   barbarous because challenging the sovereign is
   transgressive
     Recent version: post September 11 US practices

  Clausewitzian tradition: violence in civil wars is
   barbarous because of specific technological features
   unique to civil wars
     Recent version: guerrilla war as “dirty war”
ALL APPROACHES HAVE MERITS, BUT ALSO
PROBLEMS

    Hobbesian: most civil war contexts do not
     produce anarchy (understood as absence of
     authority) but proliferation of authorities
    Schmittian: Polarization often characterizes
     elites rather than masses; intellectuals rather
     than average individuals; it is often the
     product rather than simply the cause of civil
     war and violence
    Foucauldian: Violence against the sovereign
     can be as barbaric as violence by the sovereign
    Clausewitzian: It is not clear what the relation
     between warfare technology and violence is
 Rather than discuss each of these traditions in
  detail, focus on one the and push it to its
  analytical and empirical limits
 The Clausewitzian one

 How much leverage do we get?

 The intuition in two pictures
  TAL AFAR, IRAQ (SEPTEMBER 2005)




An Iraqi informant in Tall Afar uses a thumbs down to signal that a detainee
should be released.
SOME FACTS
 The majority of civil wars (but not all!) are
  guerrilla or irregular wars
 Frontlines aren’t clear; rebels blend into civilian
  populations
 In other words, there are informational
  asymmetries
 Civilians have information about who is who that
  armed actors lack but desire
THE ARGUMENT

   Informational asymmetries are the causal
    link between civil war, on the one hand, and both
    the barbaric and intimate dimensions of its
    violence, on the other hand
ROADMAP
   A few definitions and concepts
       Civil war; irregular war; sovereignty and control;
        types of violence
 A theory of irregular war
 A theory of violence in irregular war

 An empirical demonstration
A MINIMAL DEFINITION OF CIVIL WAR
       Armed combat taking place within the
        boundaries of a recognized sovereign entity
        between parties subject to a common authority
        at the outset of the hostilities
         “Armed combat” implies a threshold of
          organization and violence
         “Sovereign units” are primarily states, but may
          include a multitude of historical units
         Compatible with any cause
         Includes wars of resistance against occupation and
          colonization
       A more precise term: internal or intrastate war
A THEORY OF IRREGULAR WAR
 Argument applies to irregular war contexts
 Where violence in bilateral and coercive
       e.g. Chechnya, Afghanistan
   Excludes cases of unilateral violence or violence
    used to expel or exterminate
       e.g. Bosnia, Rwanda, Stalin’s purges
   Two actors: incumbents and insurgents (same
    logic for multiple actors)

   No clear frontlines
EL SALVADOR, 1984
KOREA, 1951
SOVEREIGNTY & CONTROL
 Sovereignty is divided; monopolies of violence are
  localized
 Two types of divided sovereignty
     Segmented: two or more distinct “states” within a
      state
     Fragmented: two or more rulers on the same piece of
      territory within a state
   Control: The ability of organizations to control
    access to the population and perform state-like
    functions; varies spatially and temporally
SUPPORT (“COLLABORATION”)
   Why and how people collaborate with armed
    actors?
     Preferences: class, ethnicity, ideology, etc.
     But also, the dynamics of the war itself
     As the war goes on, the will of armed actors
      increasingly shapes the behavior of the civilian
      population—through the use of violence
     Most people most of the time value survival over
      other considerations
   Violence is used by armed actors to induce
    civilians to collaborate with them and not with
    their rivals
THE IDENTIFICATION PROBLEM

   Armed actors can’t tell the enemy combatant (or
    the enemy noncombatant) from the “bystanding”
    (or even friendly) noncombatant

   Two “solutions”
     Collective profiling: target every member of a
      suspected group (i.e. select on visible ascriptive
      characteristic, location, etc.)
     Individualized: ascertain individual guilt
   Violence is

       indiscriminate when targeting is collective(i.e. every
        man in a village)

       selective when targeting is individualized (i.e.
        particular individuals)

   Choice is dictated by efficiency rather than
    morality considerations
STRATEGIC IMPLICATIONS OF VIOLENCE
IN CIVIL WAR

 Because there is more than one armed actor, they
  must take the consequences of their violence into
  account
 Indiscriminate violence tends to be
  counterproductive
       Α U.S. advisor in Vietnam (1960s): “This is a political
        war and it calls for discrimination in killing.”
 Selective violence is more effective but more
  difficult to achieve as it requires a high degree of
  information
 Under certain broad conditions, armed actors opt
  for selective over indiscriminate violence
JOINT PRODUCTION OF VIOLENCE
   Information is asymmetrically distributed
    between political actors and individual civilians

   Selective violence is the outcome of a transaction
    between organizations and civilians through
    denunciation
   Selective violence is open to problems of “moral
    hazard”

   Denunciation is often malicious: denouncers use
    armed actors to achieve their own ends which
    often consist of nothing more than typically
    trivial local and private disputes

   A US officer, Iraq:
       “These people dime each other out like there’s no
        tomorrow; … out of a hundred tips we’ve gotten from
        Iraqi intelligence, one has worked out.
EAMON COLLINS, IRA

After a while, one aspect of my encounters with people and their
complaints began to depress me. I realized that a lot of people,
often not even republicans, would seek the help of Sinn Fein in
order to draw on the threat of IRA muscle – so they hoped – in
solving their disputes. At times I felt as if people were treating me
as a Mafia godfather. One former work colleague asked me if I
could sort out his son-in-law. Apparently the latter was beating
up his wife, my former colleague’s daughter. I said that it was
none of Sinn Fein’s business. Then my former colleague said:
“Yes, it is. That man is never out of the police station. I am sure
he’s an informer.” I said that he was making a very serious
allegation. I said that if the IRA were to investigate it and find it
to be groundless then they would come looking for the person who
made the allegation. Unfortunately, the allegation that so-and-so
was an informer (was ‘never out of the police station’) became
one that I heard regularly from people who wanted extreme
violence done to their neigbours.
SUPPLY OF DENUNCIATIONS ALWAYS
EXCEEDS EXPECTATIONS

 Germany, 1934
 Bureaucrats expressed their surprise about the
  quantity of denunciations, especially false
  charges, noting that they had reached
  “altogether unacceptable proportions”
 The minister of the Interior asked local
  authorities to take steps to curb the rapid
  expansion of all denunciations, “too many of
  which were based merely on conflicts with
  neighbors”
 Arrest for political reasons could not be used as
  grounds for divorce
 Hitler complained: “We are living at present in a
  sea of denunciations and human meanness”
THE WALL STREET JOURNAL,
21 NOVEMBER 2001
THE WALL STREET JOURNAL,
21 NOVEMBER 2001
 Implication: intimate violence is often the result
  of the privatization of politics rather than the
  politicization of private life
 …But can we analyze the process of violence
  more systematically?
AN OPERATIONAL MEASURE OF CONTROL
   Geographical space may
    be divided into five zones
    of control, from 1
    (highest incumbent
    control) to 5 (highest
    insurgent control); zone
    3 is a zone of equal
    control


                                 1   2   3   4   5
COLLABORATION
 Collaboration is a
  function of control
 Individuals collaborate
  with the stronger actor                     k(c)ins                                         k(c)inc



       in zones 1 and 2 mainly
        with incumbents; in zones
        4 and 5 mainly with
        insurgents.
   Zone 3 is the epicenter of
    conflict: collaboration
                                    Control   1         2             3                   4             5




    with both sides
                                                        Defection toward the insurgents
                                                        Defection toward the incumbents
DENUNCIATION
 … is a function of the likelihood of counter-
  denunciation (high in villages where everything
  is visible)
 This means that individuals will denounce only
  where one side has control and can deter
  retaliation: zones 1 & 2 for incumbents, 4 & 5 for
  insurgents
 Surprising prediction: no denunciations in zone 3
PREDICTIONS ABOUT VIOLENCE
   Indiscriminate violence will be likely where
    information and control are minimal
     Zone 5 for incumbents
     Zone 1 for insurgents

   Selective violence will be likely where control is
    hegemonic but not total
     Zone 2 for incumbents
     Zone 4 for insurgents

   No violence in zone 3
AN EMPIRICAL TEST
   The Argolid region in
    Southern Greece
   62 villages; population: 45,086
    (1940)
   Civil war during German
    occupation, 1943-44
   “Ideological” war; pro-
    communist resistance vs.
    right-wing collaborationist
    militia
   No ethnic, religious, or class
    cleavage (5,360 families; 5,090
    farms)
CONTROL AND VIOLENCE, JUNE 1944
INDISCRIMINATE VIOLENCE
SELECTIVE VIOLENCE
DISTRIBUTION OF SELECTIVE VIOLENCE
BY ZONE OF CONTROL
                             Selective Violence By Zone of Control and Actor

                    35




                    30




                    25
 Violent Villages




                    20
                                                                                   Incumbents
                                                                                   Insurgents
                    15




                    10




                    5




                    0
                         1    2                  3                  4          5
                                           Zone of Control
DISTRIBUTION OF SELECTIVE VIOLENCE
BY ZONE OF CONTROL
                                                                 Selective Violence, Argolid Sept. 1943-Sept. 1944

                                                  1.00


                                                  0.90


                                                  0.80
    Proportion of Villages in a Zone of Control
         Experiencing Selective Violence




                                                  0.70


                                                  0.60

                                                                                                                         Incumbent Violence
                                                  0.50
                                                                                                                         Insurgent Violence

                                                  0.40


                                                  0.30


                                                  0.20


                                                  0.10


                                                  0.00
                                                         1   2                     3                    4            5
                                                                              Control Zone
A WAR WITH NO FRONTS
MEKONG DELTA, 1969
VIETCONG VIOLENCE, 1969
(N=61,701 VILLAGES/BIMONTHLY)

                                             1969
                              0.09


                              0.08


                              0.07


                              0.06
   Percent violent villages




                              0.05


                              0.04


                              0.03


                              0.02


                              0.01


                              0.00
                                     1   2          3           4   5
                                             Zones of Control
   COLOMBIA
   WOULD YOU SAY THAT WHEN A GROUP WAS THE
   STRONGEST IN A ZONE, OR HAD MORE CONTROL THAN
   OTHERS, WERE THERE MORE DEATHS OR FEWER DEATHS?
                      100%


                                       More Deaths   Fewer Deaths
                      90%



                      80%



                      70%
Percent Respondents




                      60%



                      50%



                      40%



                      30%



                      20%



                      10%



                       0%
                             CP   IP                            IFARC   IELN
DENUNCIATION IN CONTEMPORARY
COLOMBIA
   ACUSO A UN VECINO DE SER MILICIANO DE LA GUERRILLA PERO NO ERA CIERTO
    PARECIA QUE TENIAN PROBLEMAS PERSONALES
   LAS MUJERES SE ACONSTUMBRABAN A DENUNCIAR A LOS HOMBRES QUE LES
    PONIAN CACHOS Y VISEVERSA
   INFORMACION QUE UN MUCHACHO ERA PARAMILITAR, Y LA INFORMACION ERA
    FALSA. LAS FARC, LOS MATARAN Y LUEGO SUPIERAN QUE ERA UN ERROR. EL QUE
    ODIO LA INFORMACION LE TENIA BRONCA AL MUCHACHO
   PROBLEMAS POR UNOS ANIMALES QUE DAÑARON UN CULTIVO PRO PASARSE DE
    UNA FINCA A OTRA
   SE PRESENTABAN MUCHOS CASOS A DIARIO, QUE POR ROBOS, GANADO, ETC
   2 HERMANOS, UNO GUERRILLERO Y OTRO CIVIL, EL GUERRILLERO ACUSO AL
    HERMANO CIVIL DE SER GUERRILLERO Y LO MATARON
   UN TIO DENUNCIÓ A UN PRIMO
   ALGUNAS VECES OCURRIA POR BRONCAS ENTRE VECINOS
   POR COSAS COMO VIOLENCIA DOMESTICA, PRESTAMOS DE PLATA
   RIÑAS PERSONALES, DINERO, ENVIDIA
   POR PELEAS ENTRE CIVILES O CHISMES DE LAS MUJERES, DENUNCIABAN A LA
    PERSONA Y SI LA ENCONTRABAN CULPABLE LA CASTIGABAN
   POR ALGUN PROBLEMA PÈRSONAL, SOBRE TODO POR PELEAS CUANDO ESTABAN
    BORRACHOS
   CASOS EN QUE CIVILES DENUNCIABAN PORQUE OTROS SE METIERON CON LAS
    HIJAS
   RIÑAS PERSONALES, CELOS, PROBLEMAS AMOROSOS
WHAT HAVE I SHOWN?
 “There is method in madness”
 It is possible to analyze micro-level patterns of
  violence systematically
 These patterns conform to a logic that is
  primarily strategic and corresponds to a context
  of incentives and constraints generated by
  irregular wars
 Though the logic is instrumental, the outcomes
  are not optimal for the actors involved
     Armed actors tend not to use violence where they
      need it most (zone 3)
     Civilians fail to settle their accounts where it is
      safest to denounce (zones 1 and 5)
MORE…

   This context explains a great deal of the
    barbarity of civil wars
     Dynamics caused by informational asymmetries
     Large case atrocities (i.e. indiscriminate violence)
      emerge when and where there is lack of information
      (“cheap counterinsurgency”)
     The interplay of selective and indiscriminate violence
      is consistent with the varying patterns of violence in
      most civil wars
   This context also helps make sense of the
    intimate character of violence in civil wars
       It is not just killers who are involved in killing
FINALLY…
 A“ Clausewitzian” perspective helps make sense
  of both the barbarity and intimacy of violence in
  civil war
 The game on the ground is not a simple reflection
  of the game at the top—the local is not the
  national writ small but is qualitatively different
 Key implication: ideology/identity could be
  epiphenomenal to the violence in civil war in the
  sense that violence that looks purely ideological
  or identity-based may aggregate a variety of
  other processes
 Strategies of intervention ought to be designed
  accordingly

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:8/27/2012
language:English
pages:52