The nature and uses of violence in the Kurdish conflict

Abstract

Turkey, Iran and Iraq, the three modern states among which the vast region known as Kurdistan is divided, have each had their protracted and violent Kurdish conflict. In all three cases the conflict has frequently been described as an ethnic conflict — and rightly so, in the sense that the Kurds’ ethnic identity and political claims based on this identity were central to the conflict. It is, however, important to state at the outset that in none of these countries has the conflict had the form of a confrontation between Kurds and rival ethnic groups. Typically, the conflicts were between separatist political movements and central governments. In the course of these conflicts, the states have often used violence indiscriminately against Kurdish populations but there have not been — but for a few unimportant exceptions — clashes between groups of Turkish, Arab or Persian civilians and their Kurdish neighbours. The armed Kurdish movements have never targeted Turkish or Arab civilians as such; their leaders have proclaimed time and again that their quarrel is not with the Turkish (or Iraqi, or Iranian) people but with the repressive policies of the state. It is true that anti-Kurdish feeling has risen among nationalist Turks in Turkey and among sections of the Arab population in Iraq, but there have never been serious ethnic clashes setting Turks or Arabs against Kurds. (Fears that one day such violence might be triggered by terrorist actions in western Turkey, carried out by Kurdish activists, were not entirely groundless, however.)

Full text

Martin van Bruinessen,
‘The nature and uses of violence in the Kurdish conflict’
Paper presented at the International colloquium "Ethnic construction and
political violence", organised by the Fondazione Giangiacomo Feltrinelli,
Cortona, July 2-3, 1999.
Published in Italian translation as: “La natura e gli usi della violenza nel
conflitto kurdi”, in: Marco Buttino, Maria Cristina Ercolessi, Alessandro
Triulzi (eds), Uomini in armi, Napoli: l’ancora, 2000, pp. 99-113.

Van Bruinessen, Violence in the Kurdish conflict
1
The nature and uses of violence in the Kurdish conflict
Martin van Bruinessen
Utrecht University
Turkey, Iran and Iraq, the three modern states among which the vast region known as
Kurdistan is divided, have each had their protracted and violent Kurdish conflict. In all
three cases the conflict has frequently been described as an ethnic conflict — and rightly
so, in the sense that the Kurds’ ethnic identity and political claims based on this identity
were central to the conflict. It is, however, important to state at the outset that in none of
these countries has the conflict had the form of a confrontation between Kurds and rival
ethnic groups. Typically, the conflicts were between separatist political movements and
central governments. In the course of these conflicts, the states have often used violence
indiscriminately against Kurdish populations but there have not been — but for a few
unimportant exceptions — clashes between groups of Turkish, Arab or Persian civilians
and their Kurdish neighbours. The armed Kurdish movements have never targeted
Turkish or Arab civilians as such; their leaders have proclaimed time and again that their
quarrel is not with the Turkish (or Iraqi, or Iranian) people but with the repressive
policies of the state. It is true that anti-Kurdish feeling has risen among nationalist Turks
in Turkey and among sections of the Arab population in Iraq, but there have never been
serious ethnic clashes setting Turks or Arabs against Kurds. (Fears that one day such
violence might be triggered by terrorist actions in western Turkey, carried out by
Kurdish activists, were not entirely groundless, however.)
Ethnic cleansing of and by Kurds
Such ethnic cleansing of Kurds as has taken place (notably in Iraq) was carried out not
by civilian neighbours but by the military, in a premeditated demographic policy. Since
1970, when an agreement concluded between the Kurdish movement and the central
government of Iraq stipulated that districts with a Kurdish population majority would be
included in an autonomous Kurdish region, strategically important regions were
systematically cleansed of their Kurdish inhabitants. The Kurds have not reciprocated by
forcing Arabs to leave Kurdish-controlled districts.
The closest parallel in Kurdish history to the ethnic purges of ex-Yugoslavia is
probably provided by the expulsion and massacre of the Armenians during and directly
following the First World War. The deportation of Ottoman Armenians from the war
zone in 1915 was planned by the Young Turk regime. These deportations were
accompanied by mass killings, mostly carried out by irregulars, that appear to have been
premeditated. At least some of the Kurds took part in the looting and killing of

Van Bruinessen, Violence in the Kurdish conflict
2
Armenians — whereas other Kurds protected Armenians and saved many lives.
1
By the
end of the World War, when the Ottoman Empire had lost, Armenians took revenge and
attempted to expel the Muslims from the region where they hoped to establish, with
American support, an independent Armenia.
2
Kurdish and Turkish chieftains and
notables in this region established “defence committees” to oppose the Armenian
claims, thereby launching what became in retrospect known as Turkey’s War of
Independence. Kurds took active part in this war and in the ethnic purging of the
Armenians — but they did this not yet as ethnic Kurds but as Muslim subjects of the
sultan.
In the past decades, several other religious minorities have virtually disappeared
from Kurdistan. Most members of the Jewish minority left after the establishment of the
state of Israel, in their case without any significant pressure from their Kurdish
neighbours.
3
The Kurdistani Jews still constitute a distinct community in Israel, which
cultivates positive memories of its region of origin.
The departure of most of the Syrian Christians (Süryani, Suryoye) of Mardin province
in Turkish Kurdistan took place under less harmonious conditions. The opportunity of
labour migration to Germany drew away many young men, thereby weakening the
remaining community and making it more vulnerable to pressure from its neighbours.
Kurdish and Arabic-speaking tribesmen forced Christian villagers to give up their
agricultural land and houses. Women were abducted, material property stolen, men
beaten up or killed, causing a rapid exodus from the region.
4
In Iraqi Kurdistan, too, the
Christian minorities (Assyrians and Chaldaeans) have recently been subject to similar
violent pressures from certain Kurdish neighbours.
The Yezidi religious minority has suffered similar but perhaps even more severe forms
of oppression than the Christian communities. Despised by Muslims as “Devil-
worshippers” and not protected by any form of official recognition, they constituted the
1
The prominent Kurdish historian, Kamal Mazhar Ahmad, devotes a chapter in his book on the First
World War, Kurdistan le salekanî sherrî yekemî cîhanda (Stockholm 1990 [first ed. 1975]), to the
Armenian massacres and makes no attempt to exonerate the Kurds, though rightly noting that the Kurds
have often been blamed for brutalities committed by others. The greatest Armenian expert on the
period, Vahakn N. Dadrian, emphasises the role of the Young Turk’s secret service in organising the
massacres (The history of the Armenian genocide, Providence and Oxford: Berghahn, 1995). On
Turkish-Armenian relations see also the excellent article by Hamit Bozarslan, “Remarques sur l’histoire
des relations kurdo-arméniennes”, Journal of Kurdish Studies 1 (1995 [1996]), 55-76.
2
A. Rawlinson, Adventures in the Near East 1918-1922. London: Cape, 1923.
3
Moshe Gat, The Jewish exodus from Iraq, 1948-1951. London: Frank Cass, 1997; Joyce Blau, “Les
relations entre les juifs et musulmans au Kurdistan”, Les Annales de l’Autre Islam no. 5 (1998), 199-224.
4
See my “The Christians of Eastern Turkey, the state and the local power structure”, ICMC Migration
News no. 3-4 (1979), 40-46, reprinted in M. van Bruinessen, Mullas, Sufis and heretics: the role of
religion in Kurdish society. Istanbul: Isis, 2000.

Van Bruinessen, Violence in the Kurdish conflict
3
most vulnerable community. Adult men were forcibly circumcised, their unshaven
moustaches — symbol of their religious identity — cut, their property destroyed and,
inevitably, many of their women abducted, forcibly Islamised and married by Muslim
neighbours. Yezidis persist in a few districts in Iraq and Syria where they are strong
enough to defend themselves. In Turkey, the surviving Yezidi communities were by the
1970s already so weak that the only alternatives that they had were conversion to Islam
or flight abroad.
5
In none of the last three cases was there a deliberate policy of ethnic cleansing. Even the
perpetrators of violent attacks on minorities do not appear to have nurtured ideas about a
more homogeneous Kurdistan or a more purely Muslim land. There was little collective
violence against these minorities, no pogroms, no deliberate group expulsions. The
violence was of a more private nature — although it could only take place in a climate
where neither the local Muslim population nor the local government considered this
violence as a serious offence. The violence was a social phenomenon, however, in that
its increasing occurrence, and the ultimate dwindling of the minorities, was directly
related to the social and economic changes that accompanied the opening up of the
region to trade and modern communications, and to increased geographical mobility.
Nomads settled and needed land. Villagers moved to the district towns, where
competition made minorities vulnerable. Especially among the minorities, the most able
men left the region altogether to work in Istanbul or western Europe, which made the
remaining communities ever less capable of defending themselves.
Better communications, increased mobility, and mass education also contributed to
the emergence of a broadly based Kurdish ethno-national movement. The violence
against members of minorities does not, however, appear to be directly related to the
rising national aspirations of the Kurds. Kurdish nationalist parties and associations have
never condoned violence against minorities, and only one insignificant party ever openly
flaunted ideas of an ethnically pure Kurdistan.
6
Another exception may have to be made
for the violent Kurdish Islamist movement Hizbullah, which not only appears to have
carried out assassinations of personalities considered as pro-PKK but also of prominent
Christian personalities, in both cases in part for ideological reasons.
7
The PKK, on the
5
Robin Schneider (ed.), Die kurdischen Yezidi: Ein Volk auf dem Weg in den Untergang. Göttingen:
Gesellschaft für Bedrohte Völker, 1984; John S. Guest, The Yezidis, a study in survival. London: KPI,
1987; Johannes Düchting & Nuh Ates, Stirbt der Engel Pfau? Geschichte, Religion und Zukunft der
Yezidi-Kurden. Frankfurt am Main: medico international & KOMKAR, 1992.
6
This was the (Iraqi Kurdish) party KAJYK, led by Cemal Nebez.
7
The Kurdish Hizbullah of the Batman-Diyarbakir region emerged in the early 1980s as a serious, and
violent, rival to the secular PKK. One wing of the movement is accused of collaborating with Turkish
police and intelligence in waging a “dirty war” against nationalist Kurds. An article in the February
2000 issue of the Assyrian journal Renyo Hiro (published in Switzerland and distributed throughout

Van Bruinessen, Violence in the Kurdish conflict
4
other hand, has at least verbally come to the defence of the threatened minorities (and in
a few cases “punished” Kurdish oppressors of the Christian minority). Perhaps in order
to distance itself from the assimilationist policies of Kemalist Turkey, the PKK has at
least verbally been a consistent advocate of an ethnically pluralistic Kurdistan and of
minority rights.
Violence exercised by Kurdish political movements against other Kurds
When the Kurdish nationalist movement had recourse to violence, it did so — most
clearly in the case of the PKK — against two distinct “enemies”: the enemy without, i.e.
the state and its coercive apparatus, and the enemy within, i.e. fellow Kurds
collaborating with the state and/or supporting rival organisations. From 1978, the year
when the PKK was formally established, until 1984, when it initiated the guerrilla war
proper, virtually all its violent actions were directed against other Kurds. Even after
1984, the “enemy within” remained high on the PKK's agenda, and violence against
fellow Kurds was never less than that against military targets.
We find this tendency to mete out the most severe violence against members of
one’s own nation in many “liberation movements”. The enforcement of national unity
and discipline (not to mention the imposition of the revolutionary party’s authority over
the nation) often takes precedence over the struggle against external enemies. The
closest parallel to the PKK, in this respect, may be Algeria’s National Liberation Front
(FNL). This organisation, too, directed much of its energies during the liberation war
against collaborators and rival organisations. More Algerians were killed by the FLN
during this war than Frenchmen.
8
Kurdish political organisations have directed politically motivated violence against
various categories of Kurdish targets as a rational strategy. In some cases, violence was
declared legitimate (within the perspective of Kurdish nationalism), in others it had to be
denied or declared away. Most of the examples I give will concern the PKK, which has
had recourse to violence more systematically than other Kurdish parties. Analytically,
four categories may be distinguished (in practice, there may be an overlap between
them).
a. Punishment of “collaborators”
The first category of “legitimate” targets consists of Kurdish “collaborators,” i.e. those
co-operating with the military authorities or intelligence services of the dominant state.
western Europe) documents assaults on Christian personalities by armed bands it identifies as Hizbullah
(“Konter-Hizbullah, der Feind des Assyrer-Suryoye-Volkes”, pp. 10-14).
8
Maria Crenshaw Hutchinson, Revolutionary terrorism: The FLN in Algeria, 1954-1962. Stanford: The
Hoover Institution Press, 1978.

Van Bruinessen, Violence in the Kurdish conflict
5
This violence (often called “exemplary punishment”) is usually intended as a deterrent
and as armed propaganda. At moments when the Kurdish guerrilla movement was very
weak, as happened for instance in Iraqi Kurdistan in 1979, hit-and-run actions against
individual Kurdish “collaborators” or “traitors” were virtually the only activities carried
out at its own initiative (as opposed to defensive armed actions against government
offensives). Such actions served to show the population in general that the movement
still existed, and were a warning to others that betrayal would be punished.
In Turkey, the PKK defined the category of “collaborators” more broadly,
including all members of the traditional elite who denied or under-stated their Kurdish
identities and who were co-opted into the Turkish political system. In its early tracts of
the late 1970s, the party declared such “collaborators” guilty of class and national
oppression and singled them out as appropriate targets of revolutionary violence. The
first spectacular violent PKK action in 1979 was an assault on Mehmet Celal Bucak, a
powerful tribal chieftain who was at the same time a member of parliament for the
conservative Justice Party. The failed raid resulted in a long-lasting feud between the
Bucak tribe and the PKK, which allied itself with a tribe, the Kirvar, that had long been
in conflict with the Bucaks. Later, when the state armed Kurdish militias against the
PKK, the Bucak tribe came to constitute the strongest of those militias.
b. Reprisals and warfare against pro-state militias
This brings us to the second category of targets of violence, the armed militias that were
mobilised by the states once a guerrilla movement had started. In Iraq, the government
initially recruited tribes that were hostile to those that had joined the uprising. Frequent
hostilities had taken place between the district of Barzan, that was an early centre of the
uprising, and the neighbouring Zibari, Bradost and Herki tribes well before the Kurdish
armed rebellion of the 1960s began. In the following decades too, these militias
(nicknamed jash, “donkey foal,” by the Kurdish nationalists) were typically recruited
among tribesmen and they operated under the command of their own tribal chieftains.
(There are a few cases of powerful militia commanders who were not clearly affiliated
with a well-known tribe, but these are exceptions; moreover, their armed followings
over the years began to act ever more as tribes in their own right.) In order to resist these
tribal militias, whose numbers at most times exceeded those of the Kurdish guerrilla
fighters, the Kurdish parties of Iraq saw themselves forced to ally themselves with other
tribes.
9
The effect was that at times of war the Kurdish conflict came to be dominated by
the tribes and the more ideologically motivated urban intellectuals had to take a back
seat. I shall return to this problem below.
In Turkish Kurdistan, too, the government recruited primarily tribesmen as “village
9
The process is described in some detail in: Martin van Bruinessen, “Les Kurds, États et tribus”, Études
kurdes, revue biannuelle de recherches (Paris) no. 1 (février 2000), 9-31.

Van Bruinessen, Violence in the Kurdish conflict
6
guards,” the anti-PKK militias established after 1985. Unlike the Iraqi Kurdish parties,
however, the PKK had a strongly anti-tribal ideology and has attempted to avoid allying
itself with some tribes against others. (It did not always succeed, and in certain areas
some tribes are known as pro-PKK and others as anti-PKK, but the PKK never became
dependent on the tribes as the Iraqi parties did.) The most notoriously brutal acts of
violence by PKK guerrillas were raids on villages of these “village guards,” in which
women and children were killed indiscriminately. The extreme brutality of this violence
can be explained as a deliberate effort to his the tribal group where it hurt most, namely
in its honour, the ability of its men to protect their dependents.
c. Elimination of political rivals
The third category of targets of politically motivated violence consists of the supporters
of rival organisations and of rivals within the same organisation. The attempt to
establish a monopoly of military, political and ideological leadership are a very common
phenomenon in liberation struggles,
10
and there have been many such attempts, some
more violent than others, in the case of the Kurds.
In the Iraqi Kurdish movement, infighting between urban nationalists with a
modern education, based in the Sulaymani region on the one hand, and the charismatic
warrior Barzani was present from the beginning. Both fought for control of the entire
movement, but the conflict was fanned by ideological and cultural differences that were
initially put forward as legitimisations and that came to lead a life of their own. In the
1960s, there were numerous armed clashes between both groups. Barzani, who had the
stronger military force (i.e., was supported by the larger number of tribes), succeeded in
drawing many of his opponents' followers into his own camp.
In Turkey, the PKK declared itself the sole legitimate representative of the Kurdish
people and told other organisations to dissolve themselves or to disappear from
Kurdistan. This led in 1979 to a prolonged armed conflict between the PKK and a rival
organisation, KUK (“National Liberation of Kurdistan”) in the district of Mazidağ
(North Mardin), which much resembled a traditional tribal feud. In later years, the PKK
has been engaged in armed confrontations with most of the other Kurdish organisations
and virtually destroyed their presence on the ground. Moreover, the PKK has almost
from the start, but especially since the early 1980s, been marked by a never-ending
series of purges within the organisation, many of them bloody. (The PKK leadership has
commonly declared its rivals to be “collaborators” or “traitors”, but the distinction with
the targets of the first category is quite clear.) It is hard to say whether it was in spite of
or precisely as a result of this violent attitude towards its rivals that the PKK attracted
10
On the efforts of the Algerian FLN to destroy rival organisations, see Hutchinson, Revolutionary
terrorism. Another well-known case is the civil war between MPLA and UNITA in Angola during the
anti-colonial liberation struggle. Examples are easily amplified.

Van Bruinessen, Violence in the Kurdish conflict
7
many former members of rival organisation into its own ranks. In spite of its brutality,
the method appears to have had an undeniable success in imposing unity of purpose on
the Kurds.
d. Violence against neutral civilians
The final category frequently subjected to brutal violence, whose victimisation has been
most difficult to legitimise, has been the neutral peasant population. There have been
many cases in which violence was deliberately used against innocent villagers in order
to force them to take sides in the conflict between the Kurdish movement and the state.
Especially the PKK has deliberately used violence — both the violence that it exerted
itself and the state repression that it deliberately invited — as a means of making the
peasantry aware of its Kurdishness and of its inherently conflicting relations with a
nation state based on another ethnic identity. In many districts, villages were visited by
the guerrillas at night and by the military by day, both putting heavy pressure on them to
co-operate. Both the PKK and the armed forces have forced unwilling villagers to leave
the area and have destroyed their houses.
11
“Traditional” tribal violence and modern political violence
From the cases of intra-Kurdish violence mentioned, it is tempting to conclude that the
present Kurdish conflict is to a large extent a continuation of “traditional” tribal violence
in a wider arena. The states concerned have often adduced the conflict-ridden nature of
Kurdish tribal society to explain the need for their heavy-handed efforts towards
pacification and detribalisation.
12
But also observers sympathetic to the Kurds have
often drawn attention to the “traditional” tribal conflicts that appear to prevent them
from pursuing their common, “national” interests. In fact, in their papers for this
conference both McDowall and Bozarslan refer to the persistence or resurgence of
tribalism as a relevant factor in the particular forms of (intra-Kurdish) violence
encountered in the Kurdish movement. Some reflections on tribalism and tribal violence
in Kurdish society appear called for.
13
11
Gunnar Wiessner, “Nicht nur die PKK: soziale und politische Grundstrukturen des aktuellen
Militärkonflikts in Ostanatolien. Ebenhausen: Stiftung Wissenschaft und Politik, 1996.
12
Thus for instance the Dersim rebellion was attributed to the primitive and violent habits of the Dersim
tribes, and the repression of the rebellion, in which more than 10% of the population were killed and
following which many of the survivors were deported, was described as a civilising measure. See
Martin van Bruinessen, “Genocide in Kurdistan? The suppression of the Dersim rebellion in Turkey
(1937-38) and the chemical war against the Iraqi Kurds (1988)”, in: George J. Andreopoulos (ed),
Conceptual and historical dimensions of genocide. University of Pennsylvania Press, 1994, pp. 141-
170.
13
Cf. Van Bruinessen, “Les Kurdes, États, tribus”; M. van Bruinessen, Agha, shaikh and state: the
social and political structures of Kurdistan. London: Zed Books, 1992, pp. 50-202.

Van Bruinessen, Violence in the Kurdish conflict
8
Kurdish tribes are named social groups consisting of several hundred to several tens
of thousands of families with a segmentary structure. Only a few tribes are nomadic;
most are settled and have been so for a considerable time. Each tribe consists of a
number of sub-groups (clans), which in turn consist of several smaller units, and so on.
The members of these groups believe that they are closely related, and the lower-level
sub-groups are in most cases real patrilineal descent groups. Even units that were
initially formed as purely political alliances turn into descent groups in a few
generations, because of a strong tendency towards endogamy, with a preference for
marriage with the father’s brother’s daughter. Self-defence and mutual aid are important
functions of the tribe and its segments. The segmentary structure of the tribe becomes
most clearly visible in the case of a blood feud within the tribe, when closest relatives
join forces against more distant relatives.
14
The smaller tribal units tend to be egalitarian but the larger tribes show a distinct
social hierarchy, with one or more leading families vying for control over the clans that
together make up the tribe. In this competition, which may turn quite violent, rival
chieftains of a tribe commonly conclude political alliances with significant outsiders,
such as neighbouring tribes or officials of one of the relevant states. An ambitious social
climber wishing to establish himself as a paramount chieftain will not shy from having
his armed retinue raid villages of his own tribe in order to terrorise them into
obedience.
15
It is not uncommon to find within a single tribe one chieftain who co-
operates closely with the state and a rival who is a “bandit”, a rebel or a collaborator
with a neighbour state. Similarly, the Kurdish movements often found “patriots” and
“traitors” among the tribal elite; the rivals of a “patriot” had often little choice but
becoming “traitors.”
This means that by seeking tribal allies — which they often found necessary for
military reasons — Kurdish political parties also imported pre-existing tribal conflicts
into their relationships with the state (or with rival movements). The state did the same
when it recruited tribal militias and thereby alienated most of the “traditional” rivals of
the “loyal” chieftains. Before there was an organised Kurdish movement that acted as an
ideological and military alternative to the state, rivals of chieftains allied with the state
often ended up as rebels or bandits. The presence of a Kurdish movement presented a
new type of possibilities, which had the effect of politicising tribal conflicts. Because of
the Kurdish conflict, rival chieftains found far greater external resources (in terms of
14
Van Bruinessen, Agha, shaikh and state, pp. 64-80. There exists an extensive literature about
segmentary organisation and blood feud in the Middle East, see e.g. Ernest Gellner, Muslim society.
Cambridge University Press, 1981, pp. 36-41 and the discussion in Dale F. Eickelman, The Middle
East, an anthropological approach. Englewood Cliffs, NJ: Prentice-Hall, 1989 (2nd ed.), pp. 131-138.
15
See, for instance, the description of Hajo’s rise to leadership of the powerful Hevêrkan tribe in the
first decades of the 20th century, in Bruinessen, Agha, shaikh and state, pp. 101-105.

Van Bruinessen, Violence in the Kurdish conflict
9
money and arms) available that they could use for their own purposes. The participation
of tribes in the Kurdish political and military movements not only changed the nature of
the Kurdish movement — the parties were typically established and initially led by
urban educated men and women, who frowned upon tribal values — but also the nature
of tribal conflicts, which became even more violent. As a result of the Kurdish conflict,
tribalism has over the past decades become ever more pervasive in Iraqi Kurdistan, and
a similar process is perceptible in Turkish Kurdistan.
Some of the most conspicuous violence in the Kurdish conflict appears to be directly
related to the prominent role the tribes have come to play in it. Even the party that has
been most explicitly opposed to tribalism and to the position of tribal chieftains, the
PKK, has repeatedly been dragged into tribal feuds. Its most notorious acts of violence,
the wholesale killing of women, children and old people in korucu (“village guard”)
villages, may even appear as an apotheosis of tribal violence, attacking the tribe’s
honour in its core. This type of violence gave the PKK a bad press,
16
but it probably was
a rational policy of dissuasion against tribal opponents, for whom the loss of honour was
worse than loss of life. Although related to tribal values of honour and shame, such
violence can hardly be called traditional. I am not aware of such killings of non-
combatants in earlier tribal conflicts, and it is indeed widely believed among the Kurds
that in tribal wars of the past women could freely bring their husbands and brothers food
and ammunition because no one would harm them.
In discussing the role of the tribes in the Kurdish uprisings, a distinction should be
made between those of the 1960s and later on the one hand and those of the 1920s and
1930s on the other. The latter were to a large extent the responses of previously largely
self-ruling communities to the imposition of central control by would-be nation states.
Tribes played an all-dominant role in them, but they were not just traditional rebellions,
for there was a degree of planning and organisation by a nationalist-minded elite. In the
uprisings since the 1960, however, political mobilisation and various forms of
negotiation with the central state preceded the outbreak of violence. The Kurdish parties
and associations that shaped the Kurdish movement all engaged in political propaganda
and had agendas that involved the transformation of Kurdish society rather than efforts
to keep it as it was. The violent phase in these movements was to a large extent defined
by the attitudes of the governments concerned. Even where the initiative of armed
confrontation has come from the Kurdish side (as in the case of the PKK in Turkey), the
amount of violence carried out by the state has exceeded by far that of the Kurdish
movement.
16
In fact, after the first massacre of a korucu village, which had obviously been carried out by a PKK
guerrilla unit, several others occurred for which the PKK vehemently denied responsibility, attributing
them to counter-insurgency forces.

Van Bruinessen, Violence in the Kurdish conflict
10
The transition from political struggle to armed struggle
It has often been remarked (not least by Kurdish nationalist politicians themselves, but
also by analysts, including Bozarslan in his paper for this seminar) that violence has
been the principal means of expression of the Kurdish movement because other
political, civic means of expression were not available to them (or because they lacked
the appropriate skills). Although it is no doubt true that especially Turkey allows little
space for legally expressing specifically Kurdish demands, I believe that this observation
overstates the role of violence in the Kurdish question. The vast majority of the Kurdish
elite has always preferred inserting itself in a system of patronage-based politics,
allowing itself to be co-opted and downplaying its Kurdish identity. Even the major
rebellions of 1925, 1929-30, 1937-38 were relatively marginal affairs in which only a
fraction of the Kurdish elite (and their followers) participated.
With the growth of a broader stratum of well-educated Kurds by the middle of the
twentieth century, informal and formal Kurdish associations emerged that did not
directly challenge the state but while wishing to preserve their members’ privileged
positions at once cautiously re-asserted their ethnic or “national” identities. In Turkey,
the movement’s protagonists cautiously attempted to keep their discourse within the
limits of what was acceptable, while making efforts to gradually expand those limits. In
spite of frequent clamp-downs by the authorities, it was not until the late 1970s that the
first violent expressions of Kurdish political aspirations erupted. In Iraqi Kurdistan, too,
there were some twenty years between the emergence of (underground) modern political
associations and the beginning of the armed struggle with which the major party was
associated (in 1961). With the outbreak of violence, the leadership of the movement —
or at least the initiative in many activities — shifted from the urban, educated elite to
other social groups: tribal elements in the Iraqi Kurdish case, lower class village people
in that of the PKK.
In his comparative study of ethnic violence in Northern Ireland, the Basque country
and Quebec, the German political scientist Peter Waldmann makes the observation that
the transition of an ethno-national movement to a phase of violent action is frequently
correlated with a shift of the dominance within the movement to another social
stratum.
17
He sketches a situation of ethnic nationalism rooted in the middle stratum of
the ethnic group, the majority of which tends to be accommodative towards the
government, but which also includes a radical minority preaching justice and violence. If
the minority succeeds in reaching out to lower strata (which are, in Waldmann's view,
better capable of doing the actual fighting) the movement may enter a violent phase, in
which the leadership is likely to be taken over by a new type of leader (for whom the
revolutionary violence at the same time means social mobility).
17
Peter Waldmann, Ethnischer Radikalismus: Ursachen und Folgen gewaltsamer
Minderheitenkonflikte. Opladen: Westdeutscher Verlag, 1989.

Van Bruinessen, Violence in the Kurdish conflict
11
The rise of the PKK in Turkey appears to be a clear illustration of this thesis. The
Kurdish movement in Turkey of the 1960s and the early 1970s was dominated by the
educated urban middle class and elements of the tribal elite. Its demands were cultural
and economic and it initially did not challenge the incorporation of northern Kurdistan
in Turkey. Due to the state’s fierce repression of this movement, many of its moderate
leaders became more cautious and moved to the background, while a younger generation
of students, many of them of more modest backgrounds and inclined to adopt more
radical positions, took their place. A large number of rival organisations emerged, all of
them gradually radicalising and dominated by young men. The PKK was initially
relatively insignificant among them and only became known because it was the most
violent. Its founders were young men of lower class background studying in Ankara in
the mid-1970s. They combined the discourse of anti-colonialism and national liberation
struggle with that of proletarian revolution, in which revolutionary violence had a central
place. Moreover, unlike the other organisations, they recruited their followers especially
among the poor youth of the villages and towns of Kurdistan. The PKK followers that
were brought to court in the mass trials following the military coup d’état of 1980
clearly represented another, lower stratum than the defendants belonging to other
organisations.
18
It should be added, however, that apart from the mentioned exemplary
attack on a prominent Kurdish “collaborator” and fights with rival organisations for
hegemony over the Kurdish movement, much of the PKK’s violence prior to 1984 was
of a defensive nature. Again, Turkish state repression was instrumental in the transition
to systematic anti-state violence. It was because Abdullah Öcalan fled from Turkey to
Syria in 1979 that he came into contact with Palestinian circles and found facilities for
guerrilla training in Lebanon.
19
In the case of Iraqi Kurdistan, where periods of armed conflict have alternated with
periods of accommodation between the central government and the Kurdish movement,
those periods have been characterised by the domination of different social forces within
the movement. Before the guerrilla war, and at times of peace, when the government
sought accommodation with the Kurdish movement, it was the civilian party leaders
who played the key roles. As soon as armed conflict broke out, however, the tribes came
18
Cf. Martin van Bruinessen, "Between guerrilla war and political murder: The Workers' Party of
Kurdistan", MERIP Middle East Report no. 153 (July-August 1988), 40-46. The party has published a
number of books with brief biographies of its “martyrs” (members who were killed), which provide an
insight in the social composition of its following. By the 1990s this appears to reflect a broader
spectrum of Kurdish society than in the first years.
19
In one of last long interviews before his arrest in 1999, Öcalan gives an interesting account of these
events that appears to confirm that the rise and later virtual hegemony of the PKK was a consequence of
government repression: “Wenn du leben willst, dann lebe in Freiheit”, in: Namo Aziz, Kurdistan und
die Probleme um Öcalan. München: Gallas, 1999, pp. 137-196. Other Kurdish organisations have long
suspected the Turkish intelligence services of an even more active role in the rise of the PKK.

Van Bruinessen, Violence in the Kurdish conflict
12
to play a more prominent role in the political arena. Both the government and the
Kurdish movement recruited tribes, who were believed to be the best fighters in a
mountain guerrilla war. The split of the Kurdish movement into two rival parties
(presently KDP and PUK), which often clashed with each other, made the leadership of
both parties increasingly dependent on tribal support. The largest of the tribes that
served as pro-government militias during the 1980s and thereby became rich and strong
now hold the balance in the Kurdish-controlled northern part of Iraq, and both parties
are highly dependent on their alliance with former militia tribes, making demilitarisation
almost impossible.
20
The difficult transition from violent confrontation to non-violent political struggle
The Kurdish movements, in Iraq, Turkey and in Iran, have at most times been eager to
negotiate with the central governments of their countries, even if under unfavourable
conditions. With the exception of the PKK in the first decade of its existence, the
Kurdish parties used violence as a means of bringing the government to the negotiating
table, and were willing to stop fighting once the government showed a readiness to talk
with them. Since the Kurdish parties had little more to offer to the governments than the
promise to give up violence, however, they had to maintain a credible level of capacity
for violence even during negotiations.
The transition is complicated by a number of factors. Both on the side of the
government and among the Kurds there are groups that have a vested interest in
continuing the war and that may lose power and income when a peaceful solution is
achieved. Most of the upper and middle classes among the Kurds favour a negotiated
solution; they are the ones who have most to lose in a prolonged violent confrontation,
and they are loath to see leadership in society definitively pass to other classes. When
the PKK began signalling, in the early 1990s, its readiness to give up the armed struggle
in exchange for concessions by the government, the Kurdish middle class —
businessmen, professionals, civil servants — responded enthusiastically, expecting that
peace would bring prosperity and that in the political process towards peace they would
automatically come to play a central role as representatives of the Kurdish people. Some
guerrilla commanders understandably were less enthusiastic, and the first unilateral
cease-fire declared by the PKK (in 1993) was deliberately broken when one of these
20
The Swiss anthropologist, Andreas Wimmer, who visited Iraqi Kurdistan in the early 1990s, takes his
analysis of the prominence of the tribes too far, however, when he describes the two major political
parties, KDP and PUK, as confederacies of tribes (“Stämme für den Staat: tribale Politik und die
kurdische Nationalbewegung im Irak”, Kölner Zeitschrift für Soziologie und Sozialpsychologie 47
(1995), 95-113) and attributes a central role to the tribes in the process of state formation in northern
Iraq (“Stammespolitik und die kurdische Nationalbewegung im Irak”, in: Carsten Borck et al. (ed.),
Ethnizität, Nationalismus, religion und Politik in Kurdistan. Münster: Lit, 1997, pp. 11-43).

Van Bruinessen, Violence in the Kurdish conflict
13
commanders attacked a convoy of unprepared soldiers.
On the government side, there are more factors preventing an easy transition. None
of the governments concerned has been willing to consider the Kurdish question as a
legitimate political question to be solved through political procedures. Kurdish political
actors were only allowed to play a role in the political system when they kept their
Kurdish identity a private and not a public matter. Kurdish parties and associations were
not seen as actors to be integrated into the system but as enemies to be destroyed,
politically if not physically.
Negotiations with the Kurds have been unthinkable in Turkey; both Iran and Iraq
have at times entered negotiations with Kurdish leaders, but not, it appears, with the
intention of making any concessions. Iranian Kurdish leader Ghassemlou and two of his
associates were murdered by an Iranian hit squad even when they were sitting at the
negotiating table with Iranian government representatives.
21
Iraq’s Ba`th regime has
since the 1980s systematically used negotiations to incite one Kurdish party against the
other (promising them concessions on condition that they could deliver all of Kurdistan,
and offering each party support to vanquish its rival). The end of the Iran-Iraq war,
during which large parts of Kurdistan had come under the control of the Kurdish parties,
enabled Iraq to attempt its own final solution and destroy what it did not control,
systematically killing at least 50,000 Kurdish villagers, and possibly double that number,
in the process.
22
The struggle against the PKK has allowed the Turkish armed forces to acquire and
retain an unprecedented control of the state, capable of overruling any democratically
taken decision, and it is unlikely that they are willing to relinquish this control. The army
has blocked all attempts to seek a peaceful solution to the Kurdish question. Kurdish
nationalist circles of various political persuasions have often expressed their desire for a
dialogue with the state; the search for a peaceful solution preceded the military
weakening of the PKK.
23
Even now that the PKK no longer represents a credible
military threat, the armed forces — and obediently following them, Turkey’s leading
politicians — continue to oppose concessions to Kurdish cultural and political demands.
This attitude can in the long run hardly result in anything but a resumption of the violent
struggle.
21
See the careful reconstruction of Ghassemlou’s assassination in Vienna by French journalist Marc
Kravetz in Libération, August 7, 1989.
22
Van Bruinessen, “Genocide in Kurdistan?”; Human Rights Watch/ Middle East, Iraq’s crime of
genocide: The Anfal campaign against the Kurds. New Haven: Yale University Press, 1995.
23
See, for an overview of Kurdish attempts to engage in civilian politics, Martin van Bruinessen, "Die
Türkei, Europa und die Kurden nach der Festnahme von Abdullah Öcalan", INAMO,
Informationsprojekt Naher und Mittlerer Osten, Nummer 18 (Sommer 1999), 9-15.

Van Bruinessen, Violence in the Kurdish conflict
14
Conclusion
Unlike many other ethnic conflicts, the Kurdish conflict has not been rooted in
competition between ethnic communities or community leaders for the control of
economic resources and political hegemony. In the three countries concerned, the
conflict was born of the unwillingness of these would-be nation states to accommodate
ethnic diversity and allow a political expression to a distinct Kurdish identity on the one
hand, and the emergence of a large urbanised, educated middle class among the Kurds
that demanded recognition of its ethnic identity on the other. When the conflict became
violent, this educated urban elite to some extent lost control of the political dynamics to
other segments of society — tribal elites in the case of Iraq, radicalised sections of the
lower middle classes in that of Turkey. It has been the middle strata that have most
strongly attempted to bring the conflict back into the political arena and use political
procedures to represent Kurdish interests in the political and economic institutions of the
states concerned, but their efforts have been rejected time and again by the state
apparatus. Whether the Kurdish movement will be able to make a transition from armed
struggle to political representation will depend more on the attitude adopted by the
political elites of the states concerned (including the military) than on the internal
dynamics of Kurdish society.