The Eastern Regions of Sri Lanka in British Times

Michael Roberts

My D. Phil dissertation at Oxford in the early 1960s centred on British agrarian policy in the mid-nineteenth century and therefore included the British efforts to revive the tank irrigation systems of the Sinhala past. Several British colonial personnel as well as visiting dignitaries were captivated by the ruins of the Anuradhapura/Polonnaruwa periods which they observed during adventure trips. A few saw it as a challenge for their imperial capacity. Some British governors, notably Ward, Gregory and Gordon, took up the prospect.

 Sir Henry Ward and SJV Chelvanyakam

In the 1850s a young AGA in Uva, one Bailey, remarked on the Patipola Aar river and its irrigation potentialities (in effect anticipating the Gal Oya project of the mid-20th century). The Governor, Sir Henry Ward, enthusiastically pursued the idea: he embarked on a journey to the region and thence to Trincomalee District – presumably supported by all the paraphernalia available to an imperial satrap. His detailed account of this journey is contained within a huge tome, entitled Volume of Speeches and Minutes of Sir Henty Ward, which is available in the Peradeniya University Library.

My faint recollections of this text[1] inform my commentary on the political claims entailed in the (Tamil) Federal Party’s concept of “traditional homelands” presented initially in December 1949: see the ITAK manifesto of 19th December 1949 presented in Tamil Person and State. Pictorial (2014, pp. 276-93). Guided thus and deploying common sense, I draw a distinction between the term “eastern regions” and the term “Eastern Province” when dealing with the history of Sri Lanka between the year 1833 and the year 2020.

There was no “Eastern Province” till the British government delimited that space in 1833. Its boundaries were changed in 1886 and 1889 (and more recently in 1961 when Amparai (Ampara) was carved out – uniquely combining coastal areas with interior jungle lands that extended to the foothills of the Central Plateau).

Amparai District from 1961

It is vital to be clear on this point. The concept “Eastern Province” did not exist during the timespan to which the political claims of “traditional homelands” is applied in ardent Tamil political claims today: viz. implicitly from the 13th century onwards. That most of the people occupying the coastal arena of the eastern regions from, say, the 16th century (and perhaps even earlier) were Tamil-speakers of Tamil linguistic/racial origin is reasonably certain – though they were also joined by peoples of Mukkuvar origin or Muslim identity in several coastal localities.

Whatever the ethnicity, from the sixteenth century the peoples of this area were clearly viewed as subjects of the King of Sīhalē, initially at Kotte, then Sitawaka (1580s to 1590s) and then at Kandy (1590s onwards). As practised even in modern times it is probable that a Vädda troupe participated in the broad complex of rituals known as the “Äsala Keliya” (the play in the month of Äsala) at the central politico-religious ceremony held at Mahanuvara (Kandy) which invoked the gods in support of the “galactic polity” known as Sīhalē.[2] At the conclusion of this series of rituals one finds the Väliyak Nätuma (the Dance of the Valiyak). Here, the Vädda appears: he “first fights the Sinhalas and is later reconciled and receives food from [them] Sinhalas” — leading HL Seneviratne to the conclusion that this play-act reflects “the tension and ambivalence characterizing the Sinhala/Vädda relation.”[3]

We require assiduous research to ascertain whether this acceptance of the Sinhala king’s sovereignty in this telling manner was also embodied in the participation of Tamil chiefs from the eastern coastal belt in the annual rite of renewal known as the Äsala Keliya. That the King’s writ was accepted on the eastern coast in the 17th century is quite evident. When the Dutchman Spilbergen visited the Batticaloa coast in 1602 on a diplomatic mission, he was taken to Kandy by the Tamil-speaking ‘chief’ (raja) of that locality. Likewise, when an English ship captained by the two Robert Knoxes (father and son) was forced ashore on the coast in the Trincomalee region in November 1659, the survivors were taken to the Kandyan king by the Tamil chieftains on the coast (and thereafter ‘incarcerated’ in village localities in the highlands).

Through the span of these several centuries the interior districts of the eastern regions – constituting roughly 80-85 percent of the land area – were heavily forested jungle land within which the scattered and isolated villages were old purana hamlets with Sinhala-speakers or Vädda hamlets of ancient lineage. It is possible to conjecture that the elephant population outnumbered the humans. We urbanites of the modern age need to absorb the writings of Dr RL Spittel (1881-1969) as well as more recent endeavours from such personnel as DGB de Silva, Gananath Obeyesekere. and Wilfrid Jayasuriya, in order to comprehend the character of the terrain and the limited density of population in the low-lying interior districts of the eastern coast in the 19th and early 20th centuries — features that can safely be extended backwards to the centuries before the Portuguese appeared on the scene.

Central to any such review is the work on the distribution of population in the late 19th to mid-20th centuries embodied in Gerald Peiris’s seminal paper An Appraisal of the Concept of a Traditional Tamil Homeland in Sri Lanka.” This essay was presented originally at a seminar at the ICES of Colombo in 1985 – an occasion graced by several international scholars. The heads of the Colombo branch of ICES then were Neelan Tiruchelvam and Radhika Coomaraswamy. Apparently surprised by the findings, Neelan and Radhika commissioned another geographer, Yoga Rasanayagam, to review the facts underpinning the Peiris thesis. Yoga did so and could not find any errors. After presenting this verdict to her ‘paymasters,’ she divulged this process to her good friend Gerald.

detailed computation of ethnicity of villages in Battcaloa District a/c to 1921 Census by Gerald Peiris … and below the composion for Trincomalee District 

However, her findings were not made public by anyone. Peiris’ essay eventually entered the printed realm in 1991 and present-day readers can access subsequent versions of his findings in the Ethnic Studies Report of 1991.[4] Central to this set of studies are his two maps. Such work, it seems, is studiously neglected by the political spokespersons and the ‘assiduous’ journalists of the present generation.

Contemporary Tamil Dishonesty in Political Claim

The political demands of the TULF and other Tamil politicians in recent decades after the collapse of the Tamil Tiger forces continue to be underpinned by the idea that they had sovereignty over “traditional homelands” in the north and east. This thesis lurks in insidious fashion within some present-day pronouncements. The elliptical and lurking character inscribed within their pronouncements is a mark of the deceit within their politics.

This deceit and a horrendous reading of Sri Lanka’s history were facets deeply etched within the foundational moment of the leading Tamil party in the early decades of independence, namely, the Ilankai Thamil Arasu Kadchi or ITAK. The ITAK, in fact, had a broader compass-in name: the Federal Freedom Party of Tamil-speaking Peoples (so that it became widely referred to as the “Federal Party”). This founding act occurred at Maradana in Colombo on the 18th December 1949.

A form of Tamil imperialism was inscribed within this foundational step. Chelvanayakam and the other Tamil leaders regarded their foundational act as “fundamental to the cause of the Tamil-speaking nation in Ceylon.” Note the singular “nation.” This formulation underpinned the demand for “a Federal union of two autonomous linguistic states – the Singhalese and the Tamil.”

So, here, the concept Tamil embraced the Muslims of Ceylon within the concept “Tamil.” Thus, the introduction to this document spelt out by EMV Naganathan and V. Navaratnam raised this question: “how will the Muslim section of the Tamil-speaking nation stand to gain under a Federal Act.” While this line of assertion was qualified by a proviso mentioned by Chelvanayakam at the end of his speech,[5] the imperial encompassment of the Muslims is as stark as clear within the latter’s speech as well. This was also signalled by the absence of any attention to the history of the Muslim peoples in the island and the fact that Muslim participation at this event launching the Federal Party (ITAK) seems to have been limited.[6]

Even more striking – and, in fact, alarming – is the bald dishonesty in the reading of the island’s history.

  • In referring to the parlous situation to which “the Tamil-speaking people of Ceylon” had been reduced in the steps towards independence in the 1940s, Chelvanayakam stressed that “[b]efore the advent of the European nations to Ceylon in the 16th century, the people of this Island had their own governments. But a fact that must give rise to deep thought on our minds is that for a number of centuries preceding the advent of the Portuguese, the people of Ceylon had divided themselves into two nations; the Singhalese speaking nation and the Tamil speaking nation” [ Pictorial 2014: p. 279].
  • While noting initially that the “Tamil-speaking people were divided into three groups, Muslims, Indian Tamils and Ceylon Tamils” [ Pictorial 2014: p. 283], Chelvanayakam went on to assert that “the Tamil-speaking people of Batticaloa District are of the same mind as their brethren of Trincomalee and the northern province” [see TPS. Pictorial 2014: p. 289].

These claims were founded on a summary understanding of the island’s ancient history that is (and was) laughable to the degree alarming.

Before the advent of the European nations to Ceylon in the 16th century the people of the island had their own governments. …. they had divided themselves into two nations, the Singhalese speaking nation and the Tamil speaking nation.

Because of geographical and other reasons the population inhabiting the northern parts of the Island must for long have been predominantly Tamil and it must have been difficult for the Singhalese kingdoms in the south to hold sway over the northern areas with its preponderance of Tamil population. Ultimately a natural solution was reached and about the 9th or 10th century the Tamil areas of the north divided themselves and became a separate kingdom, whilst the south of Ceylon remained Singhalese though breaking up at times into two or three kingdoms.

Though fortunes changed at times, this division into Tamil and Singhalese nations remained intact for many centuries till the arrival of the Europeans who first of all destroyed the Tamil kingdom and later, in stages, the Singhalese kingdoms” [TPS. Pictorial, pp. 279-80]

The loose understanding of the concept “nation” and its extension into the past in starkly  evident here. But that is not what I would focus on. Rather, what is astounding is the glossing over of complexities and an omission of any reference to the eastern regions of the island. The English-educated of Chelvanayakam’s vintage would have been schooled by the histories of the island produced by such authors as HW Codrington LE Blaze and GC Mendis[7]. Yet they chose to ignore their findings wholesale.

These early works of history have deficiencies and some authors may have overblown the achievements of the Sinhala civilisations in the past. But GC Mendis had been schooled in London and his Ph. D dissertation appeared as The Early History of Ceylon in 1932 in a publication produced in Colombo. It is unlikely that the educated professional classes in the island would have been unaware of its lineaments.

Set within this background, Chelvanayakam’s mauling of the past is outrageous. The historical picture in his foundation speech is an act of chicanery – chicanery most awesome.

As such, the emphasis on the “traditional homelands” and the implication that the whole of the Eastern Province constituted by the British in the nineteenth century was part of the swathe of land occupied by Tamil-speakers over several centuries from the time when the Polonnaruwa civilisation collapsed was/is a horrendous act of dishonesty and bluff.

Admiral van Spilbergen meeting King Vimaladharamsuriya … & Robert Knox

Significantly, the East is by-passed in these excursions into the pre-European past. This is not surprising. We know little about the migration history and the politics of that region, though the writings of Robert Knox and the early Dutch embassies make it quite clear that the Tamil chieftains on the coast owed allegiance to the Sinhala kings. However, Dennis McGilvray’s intimate and long-standing anthropological research in the Batticaloa region from the 1970s serves up an intriguing picture of some commonalties between the Muslim and Tamil peoples n the Batticaloa region amidst abiding cultural differences that are also of some political import.[8]

The Batticaloa Muslims adhere to matrilocal settlement patterns which enabled some intermarriage with Tamil lineages devoted to matrilocality and especially with the Mukkuvars[9] – though those linkages seem to have been increasingly out of favour in recent decades. Broadly speaking, however, there are strong socio-cultural preferences that underline the distinctiveness of the east-coast Muslims from the Tamils in neighbouring enclaves. The Muslims favour (a) the use of perfumes rather than sandalwood paste; (b) female circumcision; (c) desire their men to shave their armpits and pubic hair every 40 days; and (d) take care to have the front of their latrines facing the direction of Mecca. The Muslim fisherfolk invariably decorate their fishing boats and bullock carts, whereas the Tamils do not.[10]

Quite fundamentally, McGilvray stresses that the Muslims are hostile to the polytheistic practices that are prominently embedded in Saivite forms of worship. To them such lifeways are “demonic.” It is hardly surprising therefore that the considerable intermixing of settlements is marked by distinctive enclaves whose boundaries are clearly etched in local knowledge.

These regional variations underline the gross act of political imperialism etched within the Federal Party’s 1949 manifesto vis a vis the peoples of the east in general and the Tamil-speaking Muslim enclaves of the Batticaloa District in particular.

In parenthesis, it should be noted that the Muslim political spokesmen in the first half the 20th century were located within the southwestern quarter of the island. These leaders had been alienated by Ponnambalam Ramanathan’s assertion in 1888 in a high-profile essay that the Moor were simply Muslim members of the Tamil “race.” Thus, the Muslim political leaders of the 1930s and 1940s were not on the same page as Chelvanayakam, Navaratnam and other ITAK spokespersons of 1949.

Moreover, the dispersal of Muslim peoples in distinct pockets throughout the island encouraged localised political strategies once universal suffrage was in place. One witnessed the fruition of such coalition politics after 1948/49 in the success in which such Muslim leaders as Sinhala Marikar and Badiuddin Mohammed worked with the SLFP bandwagon in the 1950s-to-the-70s; while other Colombo-based leaders stayed with the UNP. Once the Eastern Province Muslims flexed their muscles under MHM Ashraff’s inspiration in the 1980s, the thrusts of the Muslims in the east emerged as a significant political force that doomed the ITAK aspirations completely.

The massacres inflicted by the Tamil Tigers on Muslim men and boys at prayer in mosques at Kattankudy[11]  in August 1990 and the emergence of jihadist elements among the Muslim youth of the Batticaloa-Kalmunai localities in the 20th century has underlined the Muslim-Tamil divide. It is, therefore, evident that the contemporary claims of Tamil political extremists such as Wigneswaran have been cast in some medieval dreamland. These claims are also pockmarked with numerous acts of intellectual deceit.


HAP Abeyawardana 1999 Boundary Divisions of Medieval Sri Lanka, Academy of Sri Lankan Culture, Pulgasovita.

Kannan Arunasalam 2012 “The Kattankudy Mosque Massacre. 22 Years After,” 8 June 2012,

LE Blaze  1933 History of Ceylon … with reprint in 2004 by Asian Educational Services.

HW Codrington 1947 A Short History of Ceylon, London, MacMillan

DGB De Silva 1996 [1998] “New Light on the Vanni chiefs based on historical tradition, palm-leaf manuscripts and official records,” JRAS CB Special Number 1996 vol. LXI, pp. 153-204.

Donald Ferguson 1927 “The Earliest Dutch Visits to Ceylon,The Journal of the Ceylon Branch of the Royal Asiatic Society of Great Britain & Ireland, vol 30/80, pp. 361-409

Robert Knox 1911 [1681]  A Historical Relation of the Island of Ceylon, ed by James Ryan, Glasgow;Maclehose & Co.

Dennis B. McGilvray 1998Arabs, Moors and Muslims: Sri Lankan Muslim Ethnicity in Regional Perspective,” Contributions to Indian Sociology vol.32/2, 433-482.

GC Mendis 1932 The Early History of Ceylon Calcutta, YMCA Publishing House.

GC Mendis 1935 Our Heritage, Colombo Apothecaries Co.

GC Mendis 1944 Ceylon under the British, Colombo Apothecaries Co.

GC Mendis & SA Pakeman 1935 & 1938 Our Heritage, Colombo Apothecaries Co.

Obeyesekere, Gananath n. d.Colonial Histories and Vädda Primitivism: An Unorthodox Reading of Kandy Period Texts,”

Gerald H Peiris 1985 An Appraisal of the Concept of a Traditional Tamil Homeland in Sri Lanka,” paper presented at the international conference on ‘Economic Dimensions of Ethnic Conflict in Sri Lanka’, International Centre for Ethnic Studies, August 1985  …………. published in Ethnic Studies Report, IX (1): 1991: 13-39.

Gerald H. Peiris 2009(?): Twilight of the Tigers. Peace Efforts and Power Struggles in Sri Lanka, ….

Gerald H. Peiris 2020The Insidious and False Dimensions of the ‘Traditional Homeland’ Thesis,” 5 September 2020,

Ponnambalam Ramanathan 1888 “The Ethnology of the Moors of Ceylon,” Journal of the Royal Asiatic Society, CB, vol. 10, pp. 234-62.

Roberts, Michael 1966: “The Paddy Lands Irrigation Ordinances and the Revival of Traditional Irrigation Customs, 1856-1871,” Ceylon Journal of Historical and Social Studies, 10, 114-130. Reprinted in Irrigation and Agricultural Development in Asia, ed. by E. Walter Coward Jr., Cornell University Press, 1980, pp. 1186-1202.

Roberts, Michael 2004 Sinhala Consciousness in the Kandyan Period, 1590s to 1815, Colombo: Vijitha Yapa Publications.

Roberts, Michael 2014: Tamil Person and State. Pictorial¸ Colombo: Vijtha Yapa Publications.

 Roberts, Michael 2020: The Demarcation of Provinces in Ceylon under British Rule,” 5 September 2020,…….

HL Seneviratne 1978 Rituals of the Kandyan State, Cambridge University Press.

Dr RL Spittel 1924 Wild Ceylon

Dr. RL Spittel 1933 Far-Off Things

Stanley J. Tambiah, 2006 “The Galactic Polity: The Structure of Traditional Kingdoms in Southeast Asia,” Annals of the New York Academy of Sciences 293(1): 6 -97.

Wikipedia n.d. “Kattankudy Mosque Massacre,” ………………………

MEMO from DENNIS McGILVRAY responding to specific QUERIES, 26 September 2020 after my article was posted:

A. In response to QUERY: “Am I right in speculating that the matrilocality of the Mukkuvars in Batti region enabled and/or encouraged a few marriages with the Muslims [also matrilocal] in the past?”
The legend of the Mukkuvar vs Timilar war says the Muslims (“Pattanis” “Tulukkars”, etc.) were rewarded for their aid to the Mukkuvars by being given the choice of land or Mukkuvar wives.  They chose the latter, thereby gaining both wives AND land because the Mukkuvars are matrilineal and Mukkuvar women often own houses & paddy lands.  It’s a very neat & tidy legend, obviously too perfect to be entirely true. Still, however, there seems to be popular acceptance by all parties, and circumstantial evidence from matrilineal clan names on the Muslim side, that Mukkuvar-Muslim intermarriage once occurred.
B. In response to QUERY ” …. but that this occasional practice declined or came to a full stop with the emergence of Ashraff’s movement and/or after the Kattankudy massacres?”
By the time I arrived in 1970, Muslim intermarriages in Akkaraipattu with any of the Tamil castes had virtually ceased.  I probably learned of only 2-3 such marriages over my two years in Akkaraipattu, always between a Muslim man and a Tamil woman (who subsequently converted). I presume that the Mukkuvar-Muslim intermarriage pattern had dwindled much earlier than that, but I have no specific evidence.  But in any case, Ashraff & KK Massacre had nothing to do with it.
C. In response to QUERY:  “what are the local TALES about the origin and the arrival of the Mukkuvars? What links are claimed with the pearl diving Mukkuvars of the Chilaw-Mannar area?”
Interestingly, apart from the Mattakkalapu Manmiyam document (presumably compiled under the Dutch in the 18th century like the Yalpana Vaipava Malai), the Mukkuvars did not construct an independent origin myth or foundational legend to account for their presence in the Batticaloa region. Except, I should say, that their caste name is really “Muk-kuhar” from the mythological Kuhan boatman in the Ramayana  —  and this may well be a 19-20th century invention. None of the Mukkuvars I knew in Akkaraipattu acknowledged any connection with the Mukkuvars/Mukkiars in Jaffna, or even with Mukkuvars in Puttalam. The Mukkuvar caste status claim in the eastern region is that they are kings & warriors (equivalent of Kshatriyas) who came to control the land in the wake of Magha’s invasion in 1215.  A number of Hindu temple disputes have hinged on whether the Mukkuvars should have the right to administer the temple and to supervise the ritual duties of temple specialists who are Vellalars or Virasaivas. It struck me as quite reminiscent of South Indian temple disputes over whether kings can be bosses over priests.
I never heard a single word about the Mannar pearl fishers from people in Batticaloa.
AN Example of rapid scholarly cooperation that keeps our wheels turning and churning, thanks Dennis


** Note that this essay has been mostly fashioned in Sri Lanka in early September and then under quarantine in Sydney so that I had no access to my personal libraryi n Adelaide

[1] I read this document in the early 1960s while a member of the History Dept at Peradeniya University and as part of my dissertation work. The broad impression of a deeply forested terrain in the east with few settlements remains firmly embedded in my memory. Also see Roberts 1966 for Ward and Bailey’s role in initiating the British irrigation programme of the late 19th century.

[2] For the concept “galactic polity” coined by SJ Tambiah, see his essay in 2006 and note that it is a theory covering the form of many states in South and Southeast Asia – not just ancient Sri Lanka. Note that I have reworked this concept as “tributary overlordship” in analysing the political form of the Kingdom of Kandy as Sihale (Roberts 2004).

[3] Email note from HL Seneviratne in Virginia, 11 September 2020.  See Seneviratne 1978, pp. 102 and 178 n.11.

[4] Also see

[5] Chelvanayakam: ‘“All our campaign however goes on (sic) the basis that the Muslims should be at complete liberty to decide for themselves whether the areas they occupy should be attached to the Tamil speaking provinces or to the Singhalese speaking provinces. It is for them to decide what they want” [TPS. Pictorial,  ].

[6] In fact, Chelvanyakam referred en passant to the parting of ways in 1945: see TPS. Pictorial, p.

[7] Blaze’s book was published in 1926 and Codrington’s in 1947; while that by Mendis was in 1933.

[8] See McGilvray, “Arabs, Moors and Muslims: Sri Lankan Muslim Ethnicity in Regional Perspective,”Contributions to Indian Sociology vol.32/2, 433-482.

[9] Ameer Ali (in Perth now) has confirmed this: ”the matrilocal kudy system that prevails in Kattankudy is a variation of Mukkuvar kudy system, suggesting intermarriage between Muslims and Mukkuvars in the past” (email 10 September 2020).

[10] All these points of difference except for that marked (d), are based on McGilvray’s 1998 article. The intriguing practice (d) is an ethnographic finding I came across somewhere else in ways that made an indelible impression – but I forget where this practice is marked.

[11]  About 30 Tigers attacked four mosques on the 3rd August 1990 and 147 men and boys were killed according to the Wikipedia (n.d.) account; while Arunasalam (2012) refers to two mosques as the targets.

Leave a comment

Filed under ancient civilisations, British colonialism, caste issues, centre-periphery relations, commoditification, communal relations, cultural transmission, disparagement, economic processes, ethnicity, heritage, historical interpretation, immigration, Islamic fundamentalism, island economy, land policies, landscape wondrous, life stories, LTTE, Muslims in Lanka, politIcal discourse, population, power politics, religiosity, self-reflexivity, Sinhala-Tamil Relations, sri lankan society, Tamil civilians, Tamil Tiger fighters, the imaginary and the real, trauma, truth as casualty of war, unusual people, women in ethnic conflcits, working class conditions, world events & processes

Leave a Reply