From ethnic outbidding to ethnic
conﬂict: the institutional bases for Sri
Lanka’s separatist war
Political Science Department, Hartwick College, Oneonta, NY 13820, USA
ABSTRACT. Beginning in the mid-1950s Sri Lanka’s politicians from the majority
Sinhalese community resorted to ethnic outbidding as a means to attain power and in
doing so systematically marginalised the country’s minority Tamils. This article conse-
quently argues that institutional decay, which was produced by the dialectic between
majority rule and ethnic outbidding, was what led to Tamil mobilisation and an ethnic
conﬂict that has killed nearly 70,000 people over the past twenty years. It also analyses
the inﬂuence informal societal pressures exerted on formal state institutions and how
this contributed to institutional decay. Evaluating the relations that ensued between
social organisations and the Sri Lankan state shows how institutions can prescribe
actions and fashion motives even as it will make clear how the island’s varied institutions
generated a deadly political dynamic that eventually unleashed the ongoing civil war.
Beginning in the mid-1950s the majority Sinhalese community’s politicians in
Sri Lanka (then called Ceylon) embraced ethnic outbidding and in doing so
laid the foundation for a gruesome civil war. This essay explains how ‘institu-
which was produced by the dialectic between majority rule
and ethnic outbidding, led the minority Tamils to mobilise politically and
militarily and forge an ethnic conﬂict that has killed nearly 70,000 people over
Racial or ethnic outbidding refers to the auction-like process wherein
politicians create platforms and programmes to ‘outbid’ their opponents on
the anti-minority stance adopted (Rabushka and Shepsle 1972; Horowitz
1985). Yet as Giovanni Sartori has observed, if ‘outbidding becomes the rule
of the game’, then ‘somebody is always prepared to offer more for less, and
the bluff cannot be seen’. What thereafter ensues ‘is no longer a situation
which allows the survival of a political system based on competitive principles.
Beyond certain limits, the politics of over-promising and outbidding is the
very negation of competitive politics’ (1966: 158).
Nations and Nationalism 11 (1), 2005, 141–159. rASEN 2005
Outbidding stems from politicians’ desire and determination to acquire and
maintain power and may be practiced in varied contexts. Yet whenever it
incorporates race or ethnicity it marginalises minority communities, exacerbates
interracial or polyethnic tensions, and undermines the state’s ability to function
dispassionately. When a government in a polyethnic state utterly disregards
minorities’ legitimate preferences and instead cavalierly institutes policies
favouring a majority or other community, which is precisely what ethnic
outbidding engenders, those marginalised lose conﬁdence in the state’s institu-
tions. This could easily promote reactive nationalism among those disfavoured
and create a milieu conducive to ethnic rivalry and conﬂict. As Jack Snyder has
noted, ethnic nationalism is especially acute when extant ‘institutions are not
fulﬁlling people’s basic needs and when satisfactory alternative structures are
not readily available’ (1993: 12). If the marginalised group is territorialised, and
thereby has claims to a historical homeland, they could mobilise to seek a
separate existence. This is indeed the setting for Sri Lanka’s sad ethnic saga.
Sri Lanka is a polyethnic and multi-religious society. Its last all-island
census conducted in 1981 placed the ethnic breakdown as follows: Sinhalese
74 per cent, Ceylon or Sri Lankan Tamils 12.7 per cent, Indian Tamils 7 per
cent, Moors 7 per cent and others 0.6 per cent. The Sinhalese and Tamils had, in
the main, coexisted amicably for over a millennium, including the period
between 1505 and 1948 when the Portuguese, Dutch and British colonised the
island. The country received universal franchise in 1931, though this heightened
fears especially among the Tamils that the island’s Sinhalese could use their
superior numbers to disband the hitherto utilised communal representation
system and dominate the political scene unfettered. Such fears led some Tamil
elites to demand that the Sinhalese and minorities be provided equal representa-
tion, which led to a communal milieu especially during the 1930s and 1940s.
Some ﬂimsy constitutional guarantees coupled with the commendable camar-
aderie enjoyed among Sinhalese and Tamil elites, however, enabled the groups
to join forces and seek independence from the British in 1948. The extant
interethnic elite confraternity notwithstanding, the population disparities, the
lack of substantive minority guarantees and the ﬁrst-past-the-post electoral
system provided a political structure that made outbidding an enticing strategy,
especially for those politicians relegated to the opposition. Thus ethnic out-
bidding was introduced into the political arena less than a decade after
independence. The mechanism used to do so was the Sinhala language.
Sinhalese and Tamils had grouped together to form a swabasha (self
language) movement, which demanded that Sinhala and Tamil replace
English as ofﬁcial languages.
But with Sinhalese politicians controlling
over eighty per cent of the parliamentary seats, many soon realised that there
was nothing preventing Sinhalese elites from instituting Sinhala alone as the
island’s ofﬁcial language. This led the two main Sinhalese parties to outbid
each other on who could provide the better deal for the Sinhalese community.
While it was linguistic nationalism that initially galvanised the outbidding
process, the country’s political parties have continued outbidding each other
142 Neil DeVotta
on various issues so that the practice is now embedded in the island’s political
culture (DeVotta 2003). Enforced over forty years, this outbidding phenom-
enon has severely undermined minority conﬁdence in the country’s institu-
tions and is mainly responsible for Tamil extremists seeking a separate state.
The island’s ferocious ethnic conﬂict notwithstanding, no substantive
attempt has been made to explain the civil war from an institutionalist
standpoint, and this is despite numerous political scientists emphasising
that there exists a strong correlation between impartial institutions and
polyethnic stability (Huntington 1968; Horowitz 1985; Snyder 2000). Other-
wise noted, the more partial a polyethnic state’s institutions are towards a
particular community, the greater the chance for resentment among those
disfavoured and, hence, the more likely there is to be ethnic instability. This
essay thus uses an institutionalist analysis to argue that the Sri Lankan ethnic
conﬂict is best explained by understanding the correlation between institu-
tional decay and Tamil mobilisation. It also seeks to disabuse arguments that
suggest that the island’s discriminatory language policies are unrelated to the
Tamil rebels’ quest to form a separate state (Laitin 2000). Indeed, it was
linguistic nationalism that was used as the mechanism to perpetrate institu-
tional decay. Furthermore, the institutionalist explanation utilised evaluates
the inﬂuence informal societal pressures exerted on formal state institutions
and the negative implications this had for dispassionate governance. Finally,
the essay reinforces how a political culture that eschews conciliation,
accommodation and compromise (and instead tries to scapegoat its minorities
through outbidding practices) can unleash adverse consequences.
Focusing on outbidding from an ethnic standpoint mandates that due
attention be paid to politics and political elites, which most articles dealing
with ethnonationalism usually tend not to do adequately.
Given that this
essay seeks to provide a causal explanation for Sri Lanka’s ethnic conﬂict, it
focuses on the island’s ethnic politicking in the post-independence era. In
doing so, it adopts a historical institutionalist perspective and comprises three
main sections: The ﬁrst outlines the conditions promoting ethnic outbidding.
The second discusses the theoretical approaches that scholars generally resort
to when explaining ethno-nationalism and argues that the institutionalist
approach is generally more encompassing for this purpose. Here, given space
constraints, I merely introduce the alternative approaches but elaborate on
the institutionalist approach. The third section details how and why out-
bidding was instituted in Sri Lanka, why this undermined Tamils’ conﬁdence
in impartial institutionalism, and the way such institutional decay eventually
contributed toward ethnic conﬂict.
Conditions facilitating ethnic outbidding
Among the four most important conditions facilitating ethnic outbidding,
the political structure is arguably the most important, for the subsequent
Institutional bases for Sri Lanka’s separatist war 143
three conditions – bad leaders, socio-economic grievances and territoriality –
are inﬂuenced by what the political structure enables. Thus, for example,
a political structure that encouraged polyethnic coalitions would most
likely elicit ethnic coexistence, while one that encouraged competition
between ethnic parties would likely engender ethnic outbidding. This is
because the former typically creates divisions that are ideological as
opposed to being ethnic related. It is thus argued that interethnic groups
should resort to vote-pooling (Horowitz 1991), since doing so usually
encourages mutual dependence between polyethnic voters and politicians.
The obverse is to have one group solely create and institute the rules of
the game and thereby entice ‘communally based political entrepreneurs [to]
seek to increase the salience of communal issues and then to outbid the
ambiguous multiethnic coalition’ (Rabushka and Shepsle 1972: 83). What is
even more dangerous is if the respective ethnic groups are represented by
multiple parties so that the ensuing outbidding is both interethnic and
intraethnic, for in such a scenario ‘the centrifugal character of the
competition may so increase the distance between the positions of the groups
as to propel them toward violent outcomes, including secession’ (Horowitz
It is noted that ‘bad leaders are the biggest problem’ (Brown 1996: 575)
encouraging ethnic conﬂict, and it is clear that political entrepreneurs
determined to attain and maintain power often place their personal prefer-
ences above interethnic coexistence or the national interest and promote
ethnic outbidding. Politicians typically do so by resorting to agitprop about
the dangers and threats facing their ethnic group (Lake and Rothchild 1998:
20; Gagnon 1994/95), thereby transforming themselves into saviour ﬁgures
dedicated to protecting the group and its socio-cultural and economic
interests. Ethnic entrepreneurs thus play a signiﬁcant role in justifying and
perpetrating ethnic outbidding.
Territorial status also plays an important role in inﬂuencing ethnic out-
bidding. This is because a territorially concentrated group, as opposed to a
territorially dispersed group, is typically endowed with conspicuous cultural
markers, making it possible for the group to be more easily targeted for
outbidding purposes. Indeed, religion, language, customs and culture asso-
ciated with a particular territory make differentiating along ethnic lines easier,
even as socio-economic conditions and prosperity levels are much more easily
compared between territories.
Finally, grievances, real or imagined, also promote ethnic outbidding.
Economic disparities between groups are especially easily manipulated by
politicians seeking themes to campaign on. When their competitors decide to
join such a bandwagon, ethnic outbidding is quickly introduced into the
political arena. This essay makes clear how institutional decay in Sri Lanka
was enabled by Sinhalese politicians who used linguistic nationalism as a
mechanism to rectify extant economic disparities and how this led to Tamil
mobilisation and civil war.
144 Neil DeVotta
Competing theoretical approaches to explaining ethno-nationalism
Theories seeking to explain ethno-nationalism may be compartmentalised
into four overarching categories as primordialist (both historical and biolo-
gical), constructivist (sociological), instrumentalist (elite and rational choice)
and institutionalist. Primordialist theories argue that the phenotypic and
genotypic markers people inherit determine their societal and historical
positions in the world (Geertz 1963; Shils 1957). Primordialism’s socio-
biological strand goes so far as to argue that blood ties alone ensure
individuals and their respective groups have common interests, so that even
odious practices such as racism and nepotism can be sociobiologically
explained (Van den Berghe 1978 and 1996). Such arguments, however, are
thoroughly deterministic and primordialist explanations, and in general, fail
to explain ethno-nationalism’s protean nature.
Constructivist theories argue that nationalism is a modern phenomenon
and that states’ necessity to mobilise mass armies, the complex processes
related to industrialisation and print capitalism and the subsequent capacity
to imagine and invent communities are what enabled nationalism to ﬂourish
(Anderson 1991; Gellner 1983). While constructivist theories are more useful
than primordialist theories to explain ethno-national movements and ethnic
conﬂicts, it is amply clear that they mainly pertain to Western societies and
thereby disregard many cultures’ ‘nationalistic’ tendencies in the pre-indus-
trial and colonial world (Smith 1986; Armstrong 1982).
Instrumentalist advocates, on the other hand, can be divided into two
groups: those who are branded elite theorists, as they focus on how elite
tactics, pacts and preferences inﬂuence policy (Brass 1991), and rational
choice theorists (Banton 1994; Hechter 1986 and 1995). Rational choice
theorists hold that individual behaviour may be predicted whenever individual
preferences are assumed to be known, are transitively ordered and are
temporarily stable. Though designed to explain individual behaviour, some
rational choice scholars argue that large numbers allow for aggregate out-
comes to also be predicted. Critics argue that rational choice theories over-
generalise, simplify reality and thus disregard the maddening complexities
associated with ethno-national struggles and other political phenomena
(Green and Shapiro 1994; Walt 1999). Elite theories are also criticised for
being tautological, since when not carefully conceptualised these theories
merely end up arguing that outcomes reﬂect elites’ preferences. That noted,
instrumentalist theories can be useful provided requisite attention is paid to
historical context and their insights are situated within the structures facil-
itating opportunities and constraints.
The institutionalist explanation
I refer to institutions as the ofﬁcial establishments that coalesce to create ‘the
state’ – the legislature, bureaucracy, judicial system, public education system,
Institutional bases for Sri Lanka’s separatist war 145
and police and defence forces – as well as those private establishments gaining
legitimacy from and/or providing legitimacy to political elites representing the
state. Institutions provide the requisite networks for individuals to negotiate
and interact with the state, demarcating parameters for strategic behaviour
for/between elites and masses, promoting convergent expectations between
the state and polity (Powell and DiMaggio 1991), and generating predict-
ability. Given that individuals and groups may seek to maximise accruable
beneﬁts, polyethnic societies especially need to design institutions that
promote and provide impartial interactions. Institutions encourage co-opera-
tion amidst uncertainty by legitimising rules of engagement and encouraging
trust between strategic actors regarding standard operating procedures
(March and Olsen 1989: 38; Hall 1986: 19). For ‘institutions are not only
‘‘the rules of the game’’. They also affect what values are established in a
society, that is, what we regard as justice, collective identity, belonging, trust,
and solidarity’ (Rothstein 1996: 138). Consequently, institutional decay will
gradually set in when the state’s rule-making, -applying, -adjudicating and
-enforcing institutions shower preferential treatment on a particular group
while disregarding the legitimate grievances of other groups.
It ought to be noted that there are limits to what institutions can do,
especially when demographics encourage politicians to disregard minority
preferences. Politicians seek to be elected and re-elected and are easily tempted
toward expediency. But such expedient politics undermine impartial interac-
tions between institutions and all members of the polity and delegitimise their
operations. This is because institutions, especially in a polyethnic setting,
operate with the most legitimacy only when all those in the polity, irrespective
of their race and ethnicity, acknowledge that the rules governing institutions
have been fashioned fairly. On the other hand, whenever the state’s most
important institutions egregiously favour one particular group and concur-
rently enable the subjugation of other groups, they would signal that they are
incapable or unwilling to deal dispassionately with all members of the state.
The eventual institutional decay bound to follow would likely be countered by
those being marginalised mobilising politically and militarily against the state.
It is natural to expect that the rules governing relations between state and
polity would be designed to ensure that they beneﬁt those who make the rules.
That noted, polyethnic coexistence is unlikely to be achieved if those who
make the rules egregiously discard minority aspirations, compromise and
conciliation. Thus the biggest challenge facing those crafting rules and laws
for a polyethnic society is to ensure that their group’s preferences are codiﬁed
even while enabling minorities to have a stake in the system. If minorities
believe that the rules of the game allow them to triumph some of the time then
the rules crafted by even a majority community will be deemed to be fair. The
more such feelings are embedded the greater the chance for polyethnic
coexistence to be attained. Conversely, a system of rules designed to margin-
alise, subjugate and humiliate minorities could unleash reactive nationalism
and undermine polyethnic coexistence.
146 Neil DeVotta
Institutional scholars argue that institutions can signiﬁcantly inﬂuence
actors’ strategies and their policy preferences (Steinmo 1989; Immergut 1992),
and it is clear that Sri Lanka’s institutions framed elites’ preferences and
inﬂuenced the policies they championed. Thus while Sri Lanka’s elites
operated instrumentally, there is no denying that their interactions with
various groups as well as the opportunities and constraints the island’s
institutional structure facilitated inﬂuenced their decisions. The fact remains
that despite being cognisant about the dangers of communalism and realising
that ethnic outbidding could unleash intolerance and conﬂict, these elites
cavalierly promoted rules, routines and conventions that created an ethnoc-
racy over time. Their statements at campaign meetings and in the parliamen-
tary record make clear that many suspected ethnic outbidding and the
burgeoning ethnocracy could potentially dismember the island, though their
proclivities for power at any cost merely egged them on to continue to
manipulate ethnic differences and outbid each other.
This analysis does not mean to suggest that institutions determine behav-
iour. They do not. They instead provide the ‘context for action’ that helps us
explain decision makers’ tactics, preferences and strategies. It is thus impor-
tant to recognise that while the institutional structure can discourage, con-
strain and inﬂuence, it does not determine speciﬁc political action. Politicians
can always eliminate, reconﬁgure or create anew institutions that encourage,
or even demand conciliation and compromise among rival ethnic groups. This
then means ‘facing the same set of institutional hurdles, self-reﬂective actors
can make creative decisions about how to proceed’ (Immergut 1998: 26).
If courageous politicians can seek to change the institutional structure in
order to generate positive and stable relations between state and polity,
unprincipled politicians may prefer to operate within a sub-optimal and anti-
minority structure if that furthers their political careers. Indeed, Sri Lanka
represents a classic case of how ruling politicians, having manipulated the
institutional setting to gain power, periodically sought to implement creative
policies to thereafter promote ethnic harmony but were compelled to back off
due to pressures exerted from institutions within and without the government,
thereby also reinforcing arguments about how institutions can lack malle-
ability once their policies become embedded. Thus successive governments
abrogated or failed to implement interethnic agreements that might have
quelled the burgeoning ethnic animosity between the Sinhalese and Tamils.
Understanding the respective actors’ strategising mandates paying due atten-
tion to the context in which ethnic outbidding was perpetrated and perpe-
tuated (Steinmo 1989: 502), which is what the following account tries to do.
The Sri Lankan case
As already indicated, the swabasha movement initially included both Sinhala
and Tamil speakers who demanded that their vernacular languages replace
Institutional bases for Sri Lanka’s separatist war 147
English, which only about ten per cent of the population spoke ﬂuently. When
the call for linguistic parity was jettisoned – partly because over seventy-ﬁve
per cent of the population spoke Sinhala and the Sinhalese saw no reason to
learn another vernacular language, and partly because they realised their
superior numbers allowed them to dictate the outcome – swabasha became
synonymous with Sinhala only and led Sinhalese politicians to try and outbid
their opponents on who could be the most pro-Sinhalese and anti-Tamil.
The British had marginalised the Buddhist religion, which most Sinhalese
adhere to, and also showed a preference for hiring Tamils into the civil service.
The alacrity with which young Tamils embraced the English language was a
major reason for this British preference, though the fact that this also enabled
the British to effectively marginalise the majority community was considered a
strategic bonus (Kearney 1967; Russell 1982). The upshot, however, was that
the Tamils beneﬁted disproportionately under British colonialism. For
example, in 1946, just two years prior to independence, Tamils comprised
thirty-three per cent of the civil service and forty per cent of the judicial service
(De Silva 1984: 116). They also accounted for thirty-one per cent of the
students in the university system. Such disparities had to be rectiﬁed, and
linguistic nationalism in the post-independence era became the symbolic
mechanism by which to do so. The rhetoric utilised in the process partly
indicates how ferocious ethnic outbidding was and the corrosive impact this
had on society and its governing institutions.
Linguistic nationalism and ethnic outbidding
With language being connected to upward mobility in the most fundamental
ways and with many Sinhalese holding that the English language had been
utilised to suppress their socio-economic aspirations, the quest to make
Sinhala the country’s only ofﬁcial language soon gained momentum. S. W.
R. D. Bandaranaike, who was the leader of the Sri Lanka Freedom Party
(SLFP), realised that the language issue could be manipulated to defeat the
governing United National Party (UNP), and he consequently embraced the
Sinhala-only position during the 1956 general election (Manor 1989). Ban-
daranaike’s SLFP, together with all other parties, had embraced linguistic
parity during the 1952 election campaign. Bandaranaike himself had defended
linguistic parity as early as 1944 by arguing that it ‘would be ungenerous on
our part as Sinhalese not to give due recognition to the Tamil language’.
Bandaranaike did a political somersault and embraced fully a ‘Sinhala-only’
stance for the 1956 general election. He thereafter went around claiming that
the Sinhalese were engaged in a ‘life and death struggle’
to preserve their
language and that ‘parity to both Sinhalese and Tamil will only lead to the
deterioration of Sinhalese which may disappear from Ceylon within 25
Bandaranaike was strongly supported by the island’s omnipresent
and inﬂuential Buddhist clergy and Sinhalese nationalist organisations,
almost all of whom only cared about Sinhalese Buddhists’ socio-economic
148 Neil DeVotta
upward mobility. Many among Bandaranaike’s supporters had come to
espouse the position that the minorities lived under the majority community’s
sufferance and that with independence the time had come for Sinhalese
Buddhists to gain ascendance over all communities, especially the principal
The clamour for Sinhala only now forced the ruling UNP to also alter its
position. John Kotelawala, the party’s leader and prime minister, had
repeatedly proclaimed that he and the UNP would not budge from linguistic
parity. For example, in June 1955 Kotelawala had stated that ‘I can assure
you – and I have said it not once but many times – that the UNP, of which I
am the head, have accepted the principle that both Sinhalese and Tamil will be
the languages of this country. We have said so, and we propose to put it into
Kotelawala repeated this sentiment many times over the next few
months. Yet having opted for Sinhala-only, the prime minister thereafter
claimed, ‘I want Sinhalese to be the ofﬁcial language of the country as long as
the sun and moon shall last’.
This belated switch hardly endeared the
anglophile Kotelawala or pro-Western UNP to the electorate. Bandaranaike
had campaigned on the slogan ‘Sinhala only, and in twenty-four hours’, and
he and his party handily won the 1956 election. Two months after becoming
prime minister, in June 1956, the Ofﬁcial Language Act No. 33 of 1956,
making Sinhala the ‘one ofﬁcial language of Ceylon’, passed the House of
The Sinhala-Only Act and its consequences
When the Sinhala-Only Bill was introduced in parliament Tamils resorted to
satyagraha (nonviolent protest) outside parliament and throughout the coun-
try. This in turn led to anti-Tamil riots that saw at least 150 Tamils killed
(Vittachi 1958: 20). Tamils protested passionately against the Sinhala-Only
Bill and the subsequent Act, and many leading politicians from all ethnic back-
grounds indicated that the Sinhala-Only Act was destined to lead to Tamil
separatism unless the government introduced some form of linguistic parity.
The administrative changes required to implement the Act were not adopted
overnight; they were instead authorised via gazette notiﬁcations over the next
few years. But the emotive protests demanding a Sinhala-only policy, the
Act’s subsequent passage amidst severe rioting, and the positive and negative
consequences envisioned by Sinhalese and Tamils, respectively, guaranteed
that signiﬁcant changes along ethnolinguistic lines were in store and that these
changes were bound to poison further relations between the ethnic groups.
At its most fundamental level the Sinhala-Only Act exempliﬁed a radical
change in policy implemented by the Sinhalese for the beneﬁt of the Sinhalese.
There is no disputing that the policy was dictated by using the majoritarian
principle. But it is also indisputable that the Act’s sponsors had cavalierly
demanded that the majority’s will be the will of all. Rules and laws rarely get
enacted with complete consensus, and this is especially true for rules dealing
Institutional bases for Sri Lanka’s separatist war 149
with ethnic issues in a polyethnic society. But if the rules governing the
constitutional structure are seen as fair, then even those deploring the new
rules and regulations are bound to support them with the hope that the
offending legislation can be changed in the near future. The island’s political
structure, however, ensured that it would be impossible to change the Sinhala-
Only Act without support from the very Sinhalese lawmakers who had
enacted it in the ﬁrst place. Some Sinhalese elites realised that a policy that
explicitly recognised Tamil in some form was crucial for interethnic stability
even as they campaigned supporting a future Sinhala-only policy. Indeed, the
events that immediately followed the 1956 election indicated that Bandar-
anaike wanted to institute a ‘Sinhala-only’ law and also accommodate Tamil.
But opposition to even token accommodation by nationalists and extremists
within the Buddhist clergy forced Bandaranaike to back off. What is
important to recognise here is that if rational logic justiﬁed ethnic outbidding
between the country’s major parties on the road to capturing power, it also
justiﬁed accommodating the Tamils even as a ‘Sinhala-only’ policy was
implemented. Having become prime minister, this is what Bandaranaike
now preferred to see happen. But the very political structure that induced
ethnic outbidding now exerted sufﬁcient pressure on him and others so that
they jettisoned their rational preferences for a more extremist stance in order
to continue in power.
With Sinhalese and Tamil political elites realising that the burgeoning
extremism on both sides needed to be defanged, the prime minister and the
Tamils’ Federal Party (FP) leader, S. J. V. Chelvanayakam, met in July 1957
and agreed to the so-called Bandaranaike-Chelvanayakam (B-C) Pact. The
Pact stipulated that the Tamil language would be used for all administrative
purposes in the Northern and Eastern Provinces while Regional Councils
would be created to deal with education, agriculture and colonisation. The
Tamils thus agreed to drop their demand for linguistic parity and also call off
the nonviolent protest campaign (Wilson 1994). Many Sinhalese, however,
feared that compromising on the island’s unitary status by instituting
autonomy would become a ﬁrst step toward dismemberment, and Sinhalese
nationalists and extremist Buddhist monks immediately began blackguarding
the prime minister and the Tamils.
The UNP now realised that it could score politically against the prime
minister, and the party’s leader, Dudley Senanayake, claimed that the B-C
Pact would ensure the ‘majority race is going to be reduced to a minority’
whereby ‘Ceylon would in no time be a state of India’. He further thundered:
‘I am prepared to sacriﬁce my life to prevent the implementation of the
Bandaranaike-Chelvanayakam Agreement, which is a racial division of
Ceylon under the guise of the Regional Council System and is an act of
treachery on the part of the Prime Minister’.
150 Neil DeVotta
In May 1944 future UNP prime minister and president J. R. Jayewardene
had proposed a motion in the State Council to have Sinhala made the
country’s ofﬁcial language, though it was ultimately amended to include
Tamil as well. Then Jayewardene had pleaded that ‘we want the Tamils . . . to
co-operate with us to make Sinhalese and Tamil the ofﬁcial languages in this
country’. Indeed, just two days before the Sinhala-Only Act was debated in
parliament, Jayewardene argued that ‘No Government should and could
make Sinhalese the ofﬁcial language by trampling down the language rights of
over a million of the permanent residents of the country. It cannot thrust to
the wilderness the cherished languages of these people. The doors of the public
services should not be closed to the thousands of youth who did not know
Sinhalese for no fault of their own. Surely that was the way to sow the seeds of
a civil war.’
Yet now Jayewardene organised a protest march from Colombo
to Kandy against the B-C Pact. These ethnic entrepreneurs were well assisted
by extremist Buddhist monks and Sinhalese jingoists who went on hunger
strikes and resorted to rabid rhetoric to protest against the Pact.
The Sinhala-Only Act had inﬂuenced the Minister of Transport and Works
to call for all vehicles to be issued number plates with the Sinhala letter ‘sri’
marked on them. Tamils protested by tarring the Sinhala ‘sri’ and substituting
it with the Tamil letter ‘shri’ in the Northern and Eastern Provinces. Soon the
Sinhalese retaliated by tarring all business and street signs in Tamil. When
amidst this commotion a group of Buddhist monks congregated outside the
prime minister’s residence and demanded that he abrogate the B-C Pact,
Bandaranaike complied supinely. Soon thereafter a train carrying Tamils to a
FP convention was derailed and its Tamil passengers beaten. When anti-
Tamil race riots erupted again, between 300 and 400 Tamils were killed
(Wriggins 1960: 268).
In September 1959 a disgruntled Buddhist monk assassinated Bandara-
naike. The SLFP now recruited Bandaranaike’s widow, Sirimavo, to head the
party, and she surprised everyone by embracing numerous anti-Tamil policies.
Among the most controversial was the Language of the Courts Bill, which
was designed to promote and expand the use of Sinhala in all courthouses,
and the attempt to rigorously implement the ‘Sinhala-only’ policy, which
ofﬁcially took effect on 1 January 1961. This coupled with the government’s
well-calibrated policy to hire Sinhalese into the government service saw Tamil
representation in certain government sectors drop drastically. For example,
while thirty per cent of the Ceylon Administrative Service, ﬁfty per cent of the
clerical service, sixty per cent of engineers and doctors, forty per cent of the
armed forces and forty per cent of the labour force were Tamil in 1956, those
numbers had plummeted to ﬁve per cent, ﬁve per cent, ten per cent, one per
cent and ﬁve per cent respectively by 1970 (Phadnis 1979: 348).
A second attempt at a ﬂimsy devolutionary scheme was made possible
when UNP and FP leaders agreed to the so-called Senanayake-Chelvanaya-
kam (S-C) Pact of 1965 after the UNP returned to power. The SLFP,
however, traduced the agreement by claiming that the UNP and FP had
Institutional bases for Sri Lanka’s separatist war 151
conspired to undo the ‘Sinhala-only’ policy and dismember the country.
Besides the standard support from the Buddhist clergy and nationalist organi-
sations, this time the Muslims and Marxists, who had hitherto advocated
linguistic parity, also joined them. The collective pressure these parties
mustered consequently forced the UNP to back away from the S-C Pact.
Institutional decay and Tamil mobilisation
As already noted, institutional decay in a polyethnic setting ensues when a
state’s rule-making, -enforcing, -applying and -adjudicating institutions are
incapable of acting honourably, impartially and constitutionally. By the time
the Sinhala-Only Act took effect in January 1961 it was amply clear that most
Tamils had lost faith in the state’s capacity to treat them dispassionately. The
events that followed merely marginalised the Tamils further and radicalised
the youth among them. For example, when the FP organised a satyagraha in
February 1961 and thereby brought the general administration in the north-
east to a standstill, the government initially dithered but eventually attacked
the satyagrahis in brutal fashion. The government also imposed emergency
rule, detained the FP’s leaders for over six months, and banned the party for a
year. The emergency rule imposed on the northeast allowed the military to
operate in a ham-ﬁsted fashion and with impunity. The variegated harassment
and procrustean tactics the military resorted to merely solidiﬁed this view.
During subsequent years Tamils in the Northern Province were ordered about
and searched in humiliating fashion, beaten, stoned by soldiers passing by in
military vehicles, and women occasionally raped so that by the mid-1960s the
army especially was seen as a Sinhalese occupation force bent on subjugating
Mrs Bandaranaike’s second government (1970–77) also passed a new
constitution in May 1972 that codiﬁed Sinhala as the country’s only ofﬁcial
language, guaranteed that Buddhism alone would receive the foremost place
(thus demoting Hinduism, Islam and Christianity, which were practiced by
approximately sixteen per cent, eight per cent and eight per cent of the
population respectively), and declared that the regulations passed by a UNP
government in January 1966 (amidst bilious anti-Tamil rhetoric) to imple-
ment the Tamil Language (Special Provisions) Act were ‘subordinate legisla-
tion’. A constitution represents a country’s most important institution, and
the fact that this one was put together without any input from the Tamils
clearly indicated the disregard the government displayed towards them. This
Bandaranaike government further ensured that scarce resources were made
scarcer among the Tamil population, at times going so far as to shelve
internationally sponsored development projects in Tamil areas. In the higher
educational arena the government introduced standardisation and district
quota schemes, including ethnic cut-off policies, that enabled under-qualiﬁed
Sinhalese students to replace the hitherto over-represented Tamils. This
combined assault against Tamil aspirations had made Sri Lanka a veritable
152 Neil DeVotta
ethnocracy masquerading as a liberal democracy by the mid-1970s. It was
thus hardly surprising that the Northern and Eastern Provinces were infested
with disenchanted youth by the time Mrs Bandaranaike was defeated and
J. R. Jayewardene became prime minister in July 1977.
In August 1978 Jayewardene utilised the nearly ﬁve-sixths parliamentary
majority the UNP had obtained in the election and instituted a new
constitution that made Tamil a national language. The constitution also
incorporated a complex electoral system that was supposedly designed to
increase Tamil inﬂuence in electoral politics, though the system mainly sought
to ensure that Jayewardene could rule unfettered and the UNP remained the
dominant party. Most important, the constitution disregarded Tamil claims
for autonomy. Indeed, Tamil politicians were not consulted when designing
the constitution, and this despite thousands of Tamils having voted for
Jayewardene and the UNP and it being clear that the rebellion in the
northeast was getting out of hand. In hindsight it is clear that J. R.
Jayewardene crafted the constitution to satisfy his megalomaniacal ambitions.
What bears emphasising from an ethnic standpoint is that while Jayewardene
changed the institutional structure, he did so not to accommodate the Tamils,
stem their incipient separatist movement or provide a freer and fairer system
of governance; instead, he changed the political structure in order to arrogate
unrestricted powers and decimate any opposition to his party’s rule. This was
clear by the way the UNP used its super-majority to introduce constitutional
amendment after amendment to satisfy Jayewardene’s every whim and fancy.
Having made himself executive president, Jayewardene, rather than ac-
commodate the Tamils’ legitimate grievances, sought to muzzle the burgeon-
ing Tamil rebellion. Consequently, the draconian Prevention of Terrorism Act
was passed in 1979, which allowed the security forces to arrest, imprison and
leave without trial anyone they considered to be a threat to national security.
As innocent and not-so-innocent Tamils were imprisoned, tortured and
otherwise abused, they became further radicalised and contributed to the
various rebel groups’ numbers (Wilson 2000). The country had experienced
anti-Tamil riots in August 1977 and July–August 1981, but when Tamil rebels
ambushed and killed thirteen soldiers in July 1983, the worst ever anti-Tamil
violence ensued. Between 400 and 2,000 Tamils were killed and thousands of
businesses and houses burned. As the Economist reported, ‘the majority
Singalese observed that more than half of ‘‘their’’ new industries were Tamil-
owned – and, cutting the nose to spite the face, burnt down the Tamil
The orgiastic violence Sinhalese hoodlums resorted to on this
occasion targeted Tamil persons as well and was the most horrifying: ‘Cars
were stopped and this time if Tamils were in the cars they were burned inside
them, petrol was poured over people and they were set alight, people were also
burned in their houses, and were hacked to death’ (Hyndman 1988: 26).
Government vehicles transported the rioters, and in some instances Buddhist
monks and at least one cabinet minister led the mob. While the armed forces
had worked to control and halt past riots, this time they watched passively or
Institutional bases for Sri Lanka’s separatist war 153
egged on the rioters. A stunned and pusillanimous Jayewardene did not
impose a curfew or address the nation until three days after the riots. And
when he did so he offered the hagridden Tamils not the slightest sympathy;
instead his lachrymose comments seemed to justify the violence given the
growing Tamil radicalism in the Northern Province. The 1983 riots thus saw
every major institution in the country fail to live up to its obligations and
responsibilities to protect its minority citizens; on the contrary, those institu-
tions representing the state apparatus coalesced to attack Tamils. Thousands
of Tamils consequently ﬂed the island and thereby formed the Tamil diaspora
that would fund the burgeoning Tamil separatist movement, while thousands
of others ﬂed to the Northern Province and joined the rebels ﬁghting for
separatism. There were only a few dozen Tamil rebels demanding secession in
the late 1970s and early 1980s, and the vast majority of Tamils preferred to
repose faith in moderate Tamil politicians seeking more autonomy for the
northeast. But the July 1983 anti-Tamil riots, which indisputably proved that
not a single institution representing the Sri Lankan state was capable of
treating the Tamils as equal citizens, saw the rebels’ numbers grow by the
thousands. It was the beginning of Sri Lanka’s civil war. In the years that
followed the Liberation Tigers of Tamil Eelam (LTTE), currently proscribed
as a terrorist group in a number of countries, took up arms and engaged in
one of the twentieth century’s most gruesome separatist conﬂicts.
The Sri Lankan case makes clear that outbidding and ethnocentrism can be-
come embedded and path dependent. Indeed, many Sinhalese beneﬁted from
the ethnocentric practices successive governments pursued and they now pro-
test against any proposal that promotes devolution or dispassionate govern-
ance. The fact, however, is that while these ethnocentric practices have
beneﬁted the majority community, they have led to an illiberal democracy
and inﬂuenced the principal minority community to seek a separate state. The
majoritarian principle does not justify minority domination, and liberal
democracies are thus designed to eschew the tyranny of the majority. A
proper majoritarian system, especially in polyethnic settings, thus seeks to
encourage consensus politics as much as possible and thereby allay minority
fears. This is why institutions are of paramount importance and also why an
institutionalist approach is better suited to explain Sri Lanka’s ethnic conﬂict.
This is also to say that while it was the language issue and its ethnocentric
legislation that initially undermined minority conﬁdence in the state’s institu-
tions, the Sinhala-Only Act’s impact was not immediate. In fact, many had
believed that the Sinhala-Only Act would revolutionalise their socio-economic
status quickly and were disappointed when this was not to be. But this also
meant that successive governments, in seeking to appease their Sinhalese
constituents, pursued additional ethnocentric policies that further under-
154 Neil DeVotta
mined minority conﬁdence in the state’s institutions. Consequently, each suc-
cessive policy that was implemented to legitimise the Sinhala-Only Act merely
re-emphasised Tamils’ marginalisation and deepened the institutional decay
facing their community. The Act may have ultimately uplifted the Sinhalese
socio-economically, an outcome that was bound to take place in any case as
facilities in Sinhalese educational institutions improved, but it also introduced
ethnic outbidding, legitimated ethnocentrism and inﬂuenced a reactive Tamil
nationalism that reached its logical conclusion seeking separatism.
From a theoretical standpoint, primordialism does provide some historical
context for appreciating the Sinhalese–Tamil rivalry, but it lacks robustness
when explaining the conﬂict itself. There are numerous myths of origins on
both the Sinhalese and Tamil sides, and these accounts seek to craft an
identity for each group vis-a
`-vis the other. That recognised, the fact remains
that both groups coexisted more or less peacefully for over a millennium.
Ancient ethnic hatreds and congenital identities stemming from phenotypic
and genotypic characteristics thus fail to explain why violent ethnic relations
ensued only since the mid-1950s.
The constructivist arguments are sounder, because there is ample evidence
that colonialism and nationalism did emphasise and inﬂuence more compe-
titive notions of ethnic identity. Yet, the constructivist project in Sri Lanka,
especially among the Sinhalese, had been ongoing for some time and does not
explain the post-1956 interethnic violence. Certainly modernisation and the
competition for scarce resources inﬂuenced ethno-nationalism, and this would
be part of the constructivist explanation. However, evaluating the competi-
tion for scarce resources without analysing the institutional structure provides
for a limited account.
Likewise, rational choice accounts, while useful when describing elite
machinations, fail to recognise the socio-cultural and historical inﬂuences
motivating elites. Ethnic entrepreneurs, like all actors, seek to maximise their
preferences; yet what often get overlooked are the suboptimal outcomes
ethnic entrepreneurs manipulating ethno-national issues are responsible for.
Indeed, there is no guarantee that elites and masses would react rationally
when ethnic and nationality issues are involved, because people often react
viscerally and capriciously at such times. Some may argue that doing so
bolsters the primordialist approach. This is hardly the case because such acts
represent discrete, as opposed to continuous, events and such discrete events
cannot explain an outcome fashioned and inﬂuenced over nearly half a
century. Thus, some Sinhalese who worked for Tamil employers and were
incensed by the Tamil rebels’ anti-government activities in the late 1970s and
early 1980s joined the rampaging herds and burned their workplaces during
the 1983 ethnic riots. They may have, by doing so, displayed indisputably
their love and passion for their nation and state, but their actions were hardly
rational given that they eradicated their very livelihoods in the process. It was
not uncommon to see some among these Sinhalese rue their actions once the
economic consequences of the 1983 riots became evident, which indicates how
Institutional bases for Sri Lanka’s separatist war 155
the ethno-psychological attachment to their nation temporarily overrode and
masked their long-term rational preferences. It is likewise hard to see how the
LTTE suicide cadre’s actions could be credibly explained using a rational
choice explanation. At least the Islamic fundamentalists who resort to suicide
bombings believe their sacriﬁce would ensure heavenly rewards; LTTE cadres,
on the other hand, merely blow themselves up on orders from their leadership.
In fact, all evidence suggests that the LTTE’s suicide bombers carry out their
missions out of allegiance to their organisation and their fearsome leader,
Vellupillai Prabhakaran, as opposed to any appreciable cost–beneﬁt calcula-
tions (Swamy 2003). Consequently, a greater appreciation for the historical
and social contexts may further the rational choice approach by helping to
explain where actors’ preferences come from, which is something rational
choice scholarship is typically inept at explaining. Within this context, the
institutionalist explanation provides a more encompassing explanation for Sri
Lanka’s civil war.
The LTTE has vowed to create a separate Tamil state, but most believe the
best solution to ending the civil war is a federal set-up that provides the
Tamils in the northeast with widespread autonomy. Yet the vast majority of
Sinhalese oppose federalism. They instead champion the extant unitary
structure. The civil war notwithstanding, this situation has amply encouraged
ethnic outbidding. Thus parties in power seek to promote dubious conﬂict
resolution only to be checkmated by the respective opposition, which typically
claims that the proposed solutions are bound to eventually dismember the
island. This has especially been evident in the past decade as the Bandar-
anaikes’ daughter, President Chandrika Kumaratunga, has tried to accom-
modate some Tamil demands when the coalition she leads has had a majority
in parliament but then sings a different tune when her opponents control
parliament. Such unprincipled politics merely contributes to the political
decay in the island and reiterates that a lasting peace is unlikely until Sri
Lanka’s leaders can craft the requisite institutions that would treat all citizens
1 This essay consists of a few excerpts from Blowback: Linguistic Nationalism, Institutional
Decay, and Ethnic Conﬂict in Sri Lanka (Stanford: Stanford University Press, 2004) which have
been reproduced with the publisher’s permission.
2 By ‘institutional decay’ I refer to a situation where the individuals and institutions representing
the state function in a corrupt, partial and violent manner whereby they jettison the norms, values
and practices that ensure liberal democracy, operate with impunity, engender anomie and
undermine citizens’ conﬁdence in the state. In the Sri Lankan case institutional decay ﬁrst
impacted the Tamils. Over time, however, all citizens have had to deal with its pernicious effects
3 Many academics, politicians and journalists cavalierly alternate between ‘Sinhalese’ and
‘Sinhala’ when referring to the majority community’s ethnicity and language, I speciﬁcally use
Sinhalese and Sinhala to refer to the community’s ethnicity and language, respectively.
156 Neil DeVotta
4 The swabasha movement was initially led by the English-speaking provincial elite. There were
periodic demands even before independence for the vernacular languages to be recognised, but
such demands reached a crescendo around the mid-1950s.
5 Recent exceptions to this criticism would include Brown 1998; Gagnon 1994/95; Snyder and
6 Ceylon, State Council, Debates, 25 May 1944, pp. 810–11.
7 See Ceylon Daily News (Hereafter CDN), ‘Press for a ‘‘Sinhalese Only’’ Pronouncement –
SWRD’, 28 November 1955, p. 6.
8CDN, ‘Sinhalese Fade-Out With Parity Predicted’, 4 January 1956, p. 5.
9 Ceylon, House, Debates, Vol. 21, col. 485.
10 CDN, ‘Sinhalese Only – If the UNP Gets 68 Seats or Less’, 15 March 1956, p. 5.
11 The Communist Party’s Pieter Keuneman noted it best, when he claimed that while the
Sinhala-Only Act had already led to communal riots, ‘Ten years from now it will be several times
worse. This Bill is heading straight for the division of the country . . . Every order and regulation
under it will be a cause for further strife.’ Quoted in House, Debates, Vol. 24, col. 1711.
12 CDN, ‘Pact a Racial Division of Ceylon, Says Dudley’, 12 August 1957, p. 7.
13 CDN, ‘J. R. Jayewardene On Sinhale [sic] Bill’, 4 June 1956, p. 5. (italics added).
14 Economist, ‘Ergophobia’, 13 August 1983, p. 29.
Anderson, Benedict. 1991. Imagined Communities: Reﬂections on the Origin and Spread of
Nationalism (2nd rev. edn.). London: Verso.
Armstrong, John A. 1982. Nations Before Nationalism. Chapel Hill, NC: University of North
Banton, Michael. 1994. ‘Modeling ethnic and national relations’, Ethnic and Racial Studies 17(1):
Brass, Paul R. 1991. Ethnicity and Nationalism: Theory and Comparison. New Delhi: Sage
Brown, David. 1998. ‘Why is the nation-state so vulnerable to ethnic nationalism?’, Nations and
Nationalism 4(1): 1–15.
Brown, Michael E. 1996. ‘The causes and regional dimensions of internal conﬂict’ in
Michael E. Brown (ed.), The International Dimensions of Internal Conﬂict. Cambridge, MA:
MIT Press, 571–601.
De Silva, C. R. 1984. ‘Sinhala–Tamil ethnic rivalry: the background’ in Robert B. Goldman and
Jeyaratnam A. Wilson (eds.), From Independence to Statehood: Managing Ethnic Conﬂict in Five
African and Asian States. New York: St. Martin’s Press, 111–24.
DeVotta, Neil. 2003. ‘Sri Lanka’s political decay: analysing the October 2000 and December 2001
parliamentary elections’, Commonwealth and Comparative Politics 41(2): 115–42.
Gagnon, V. P. 1994/95. ‘Ethnic nationalism and international conﬂict: the case of Serbia’,
International Security 19(3): 130–66.
Geertz, Clifford (ed.). 1963. Old Societies and New States: the Quest for Modernity in Asia and
Africa. New York: Free Press.
Gellner, Ernest. 1983. Nations and Nationalism. Oxford: Basil Blackwell.
Green, Donald P. and Ian Shapiro. 1994. Pathologies of Rational Choice Theory: a Critique of
Applications in Political Science. New Haven, CT: Yale University Press.
Hall, Peter A. 1986. Governing the Economy: the Politics of State Intervention in Britain and
France. New York: Oxford University Press.
Hechter, Michael. 1986. ‘Rational choice theory and the study of race and ethnic relations’, in
David Mason and John Rex (eds.), Theories of Race and Ethnic Relations. Cambridge:
Cambridge University Press, 264–79.
Hechter, Michael. 1995. ‘Explaining nationalist violence’, Nations and Nationalism 1(1): 53–68.
Institutional bases for Sri Lanka’s separatist war 157
Horowitz, Donald L. 1985. Ethnic Groups in Conﬂict. Berkeley, CA: University of California
Horowitz, Donald L. 1991. A Democratic South Africa?: Constitutional Engineering in a Divided
Society. Berkeley, CA: University of California Press.
Huntington, Samuel P. 1968. Political Order in Changing Societies. New Haven, CT: Yale
Hyndman, Patricia. 1988. Sri Lanka: Serendipity Under Siege. Nottingham: Spokesman.
Immergut, Ellen M. 1992. Health Politics: Interests and Institutions in Western Europe. New York:
Cambridge University Press.
Immergut, Ellen M. 1998. ‘The theoretical core of the new institutionalisms’, Politics and Society
Kearney, Robert N. 1967. Communalism and Language in the Politics of Ceylon. Durham, NC:
Duke University Press.
Laitin, David. 2000. ‘Language conﬂict and violence: the straw that strengthens the Camel’s
back’, European Journal of Sociology XLI(1): 97–137.
Lake, David A. and Donald Rothchild. 1998. ‘Spreading fear: the genesis of transnational ethnic
conﬂict’ in David A. Lake and Donald Rothchild (eds.), The International Spread of Ethnic
Conﬂict: Fear, Diffusion, and Escalation. Princeton, NJ: Princeton University Press, 3–32.
Manor, James. 1989. The Expedient Utopian: Bandaranaike and Ceylon. Durham, NC: Duke
March, James G. and Johan P. Olsen. 1989. Rediscovering Institutions: the Organizational Basis of
Politics. New York: The Free Press.
Phadnis, Urmila. 1979. ‘Ethnicity and nation-building in South Asia: a case study of Sri Lanka’,
India Quarterly 35(3): 329–50.
Powell, Walter and Paul DiMaggio (eds.). 1991. The New Institutionalism in Organizational
Analysis. Chicago, IL: University of Chicago Press.
Rabushka, Alvin and Kenneth A. Shepsle. 1972. Politics in Plural Societies: a Theory of
Democratic Instability. Columbus, OH: Charles E. Merrill Publishing Company.
Rothstein, Bo. 1996. ‘Political institutions: an overview’ in Robert E. Goodin and Hans-Dieter
Klingemann (eds.), A New Handbook of Political Science. New York: Oxford University Press,
Russell, Jane. 1982. ‘Communal Politics Under the Donoughmore Constitution, 1931–47’, Ceylon
Historical Journal 26. Dehiwala: Tisara Prakasakayo.
Sartori, Giovanni. 1966. ‘European political parties: the case of polarized pluralism’ in Joseph
LaPalombara and Myron Weiner (eds.), Political Parties and Political Development. Princeton,
NJ: Princeton University Press, 137–76.
Shils, Edward. 1957. ‘Primordial, personal, sacred and civil ties: Some particular observations on
the relationships of sociological research and theory’, British Journal of Sociology 8(2): 130–45.
Smith, Anthony D. 1986. The Ethnic Origins of Nations. Oxford: Blackwell Publishers.
Snyder, Jack. 1993. ‘Nationalism and the crisis of the post-Soviet state’, Survival 35(1): 5–26.
Snyder, Jack. 2000. From Voting to Violence: Democratization and Nationalist Conﬂict. New York:
W. W. Norton & Company.
Snyder, Jack and Karen Ballentine. 1996. ‘Nationalism and the marketplace of ideas’, Interna-
tional Security 21(2): 5–40.
Steinmo, Sven. 1989. ‘Political institutions and tax policy in the United States, Sweden, and
Britain’, World Politics 41(4): 500–35.
Swamy, M. R. Narayan. 2003. Inside an Illusive Mind: Prabhakaran: the First Proﬁle of the
World’s Most Ruthless Guerrilla Leader. Colombo: Vijitha Yapa Publications.
Van den Berghe, P. L. 1978. Race and Racism: a Comparative Perspective. New York: Wiley.
Van den Berghe, P. L. 1996. ‘Does race matter?’ In John Hutchinson and Anthony D. Smith
(eds.), Ethnicity. Oxford: Oxford University Press, 57–63.
Vittachi, Tarzie. 1958. Emergency ’58: the Story of the Ceylon Race Riots. London: Andre
158 Neil DeVotta
Walt, Stephen M. 1999. ‘Rigor or rigor mortis? Rational choice and security studies’, International
Security 23(4): 5–48.
Wilson, A. Jeyaratnam. 2000. Sri Lankan Tamil Nationalism: its Origins and Development in the
Nineteenth and Twentieth Centuries. Vancouver: University of British Columbia Press.
Wilson, A. Jeyaratnam. 1994. S. J. V. Chelvanayakam and the Crisis of Sri Lankan Tamil
Nationalism, 1947–77. Honolulu, HI: University of Hawaii Press.
Wriggins, Howard. 1960. Ceylon: Dilemmas of a New Nation. Princeton, NJ: Princeton University
Institutional bases for Sri Lanka’s separatist war 159