An appraisal of the concept of a traditional Tamil homeland in Sri Lanka

Gerald H. Peiris, reprint from Ethnic Studies Report, Vol.IX, No.1, January1991

 GERRY 11Among the various exemplifications of Tamil nationalism in Sri Lanka, those that relate to claims over territory have acquired increasing prominence during the recent past. These claims are based upon the perception that certain parts of the country belong exclusively to the Sri Lankan Tamils – a constituent ethnic group of the multi-ethnic Sri Lankan nation – in the sense that such areas constitute their ‘traditional homeland’. The present study is an attempt to place this perception under critical scrutiny.

In the current Sri Lankan ethnic conflict diverse claims and counterclaims are being made on the ‘traditional’ rights of the different ethnic groups over land and territory, ‘traditional’ invariably carrying the connotation of persistence over a long period of the past. Hence, the contending viewpoints are often based on interpretations of ancient and medieval history. The approach adopted in this study is somewhat different, at least in emphasis. Our focus is on the modern period and on spatial rather than temporal aspects. 


To understand the nature of this concept, we have to rely largely on statements that have been made from time to time by the political parties that have drawn support from a large segment of the Sri Lankan Tamil community. There are, of course, a few observations and comments made by scholars1  which correspond to interpretations of history upon which the ‘homeland’ claim is largely based. For example, in a research paper titled ‘The Origin of the Tamil Vanni Chieftaincies of Ceylon’, it is stated: “The traditions in Tamil chronicles refer to a time when invaders and settlers from South India, including the Vanniyar,occupied the present Northern and Eastern Provinces…”. Again, there is the introductory observation in a book titled The Kingdom of, Jaffna according to which: “They (the Sri Lankan Tamils) have been concentrated from medieval times in the areas that correspond to the Northern and Eastern Provinces and in the western littoral from the north up to Chilaw” (in both quotations, emphasis added).

 Turning from these to the more frequently found political statements on this issue, we find, from about the early 1950s, a crystallization of views on what is perceived to be the ‘traditional Tamil homeland’ in Sri Lanka. The territorial dimensions of the Homeland concept were placed on a formal political plane when the inaugural convention of the Illankai Thamil Arasi Kachchi (ITAK, or the ‘Federal Party’) held in April 1951 resolved that: “Inasmuch as the Tamil speaking people have an inalienable right to the territories which they have been traditionally occupying, this first national convention of the ITAK condemns the deliberately planned policy of action of the government in colonizing the land under the Gal-Oya reservoir and other such areas with purely Sinhalese people as an infringement of their fundamental rights and as a calculated blow aimed at the very existence of the Tamil speaking nation in Ceylon”.

These same ideas were reiterated at subsequent conventions of the ITAK which, since the mid-1950s, remained the foremost political party representing Tamil interests in the national parliament. For instance, at its 4th Annual Convention held in August 1956, the party resolved that: “…the colonization policies pursued by successive governments since 1947 of planting Sinhalese populations in the traditional homeland of the Tamil speaking people is calculated to overwhelm and crush the Tamil speaking people in their own national areas…”

The escalation of the Sinhalese-Tamil conflict in the years that followed was accompanied by an increasing vehemence in the ITAK’s Opposition to the expansion of Sinhalese settlements in those areas of the country which the party regarded as the ‘traditional homeland’ of the Tamils. Thus, in 1976, when the ITAK spearheaded the formation of the Tamil United Liberation Front (TULF), committed to the creation of an independent Tamil State (Eelam) within the island of Sri Lanka, the newly formed front spelt out the spatial dimension of the ‘Homeland of the Tamils’ in the following terms: “Whereas throughout the centuries from the dawn of history, the Sinhalese and Tamil nations have divided between them the possession of Ceylon, the Sinhalese inhabiting the interior parts of the country in its southern and western parts from the river Walave to that of Chilaw and the Tamils possessing the northern and eastern districts … (the TULF resolves that) … Tamil Eelam shall consist of the people of the Northern and eastern Provinces (of Sri Lanka)”.

The Manifesto of the TULF published on the eve of the General Parliamentary Elections of 1977 contained the following elaboration of the concept of the ‘Traditional Tamil Homeland’: “Even before the Christian era the entire island of Ceylon was ruled by Tamil kings. (The manifesto makes references here to certain Tamil rulers who held sway over parts of the northern Dry Zone in ancient times). From this background of alternating fortunes (of the Sinhalese and Tamil kings of ancient Sri Lanka) emerged at the beginning of the 13th century, a clear and stable political fact. At this time the territory stretching in the western sea-board from Chilaw through Puttlam to Mannar and thence to the Northern Region, and in the east, Trincomaiee and also the Batticaloa Regions that extend southward up to Kumana or the northern banks of the river Kumbukkan Oya were firmly established as the exclusive homeland of the Tamils. This is the territory of Tamil Eelam”.

 The foregoing statements of the ITAIC and the TULF are only a few among many of similar vein which when considered collectively enable the recognition of certain historical and geographical perceptions that are basic to the notion of an exclusive Tamil homeland within Sri Lanka. There is, first, the perception that from ancient times the Tamils of Sri Lanka formed a distinct nation with its own territory stretching over the northwestern northern, northeastern and eastern parts of the island. In this perspective the Tamil nation lost its territorial identity only after the conquest of these areas by the Portuguese at the beginning of the 1 7th century, and the formation of the unitary nation state of Sri Lanka (Ceylon) was a 19th century phenomenon – a consequence of the arbitrary territorial unification of two nation states after the establishment of British rule over the island. Thirdly, it is held that, following the withdrawal of the British from Sri Lanka in 1948, through a process of State-aided colonization and State-encouraged illegal occupation, the Sinhalese have encroached into areas that traditionally belonged exclusively to the Tamil nation. A corollary of this latter perception is that while colonization and encroachment have benefited the Sinhalese to a disproportionate extent, they represent one of the forms of state discrimination against the Tamils.

 In examining the validity of these perceptions, however, we encounter a series of problems that arise from certain ambiguities and vaguenesses of content in the statements from which they are drawn. For instance, while the TULE Resolution of 1976 declares that the Northern and Eastern Provinces of Sri Lanka were possessed by the Tamils from the “dawn of history”, the Manifesto of 1977 refers to the emergence of a Tamil kingdom “at the beginning of the 13th century”. This difference is associated with a lack of clarity on whether the claimed pre-13th century possession also implied political control of the areas as a national entity. A further element of confusion arises from the difference between the Resolution and the Manifesto on the demarcation of the territory of Eelam. According to the former, “the Tamil Eelam shall consist of the people of the Northern and Eastern Provinces”. The latter declares that an area stretching (clockwise) from Chilaw on the western seaboard to Kumana in the southeast is the territory of Eelam, which of course, is a much larger area than the Northern and Eastern Provinces. More significantly, perhaps, while those sections of the various resolutions adopted by the ITAK which carry territorial connotations refer invariably to a “traditional homeland of the Tamil speaking people’”, in the Resolution of 1976 and the Manifesto of 1977 the relevant sections omit the words ‘speaking’ and refer repeatedly to the ‘Tamils’, thus introducing a further element of ambiguity on whether the claimed homeland is meant exclusively for the Tamils. The ambiguity here is perhaps not unintentional on the part of those in the vanguard of the Eelam movement. Their occasional references to the Tamil speaking people were obviously meant to add weight to their claims. However, any implied or explicitly stated assertion that the agitation for a separate Tamil State within the island receives support from all Tamil speaking ethnic groups of the country can hardly be sustained. It is well known that the Muslims have persistently rejected not only the secessionist demand but also the right arrogated by the TULF to represent all Tamil speaking people of Sri Lanka. Much the same can be said of the so-called ‘Indian Tamils’ of Sri Lanka whose political leadership has formally dissociated itself from the Eelam demand. In fact, in the political sphere, the main focus of its attention has been that of promoting the fuller involvement of the Indian Tamils in the prevailing unitary political system of the country.

  Since there are many obscure aspects in the concept of the Tamil homeland, for the purpose of the present scrutiny of the related perceptions, we make the assumption that the territory being claimed as the exclusive Tamil homeland consists of the Northern and Eastern Provinces of modern Sri Lanka and that this claim itself is founded on the notion that the Sri Lankan Tamils should have greater rights over the entirety of this territory than the other ethnic groups of Sri Lanka.


Ethnicity and population distribution:  

The ethnic composition of the population of Sri Lanka as recorded at the Census of 1981 is as follows:

                                                                    Table 1: Numerical Size of the Ethnic Groups



% of Total




Sri Lankan Tamils



Moors and Malays(Muslims)



Indian Tamils









Though many parts of Sri Lanka, especially in the urban sector, are areas of mixed population, ethnically homogeneous concentrations do occur in certain areas of the country. The Sinhalese invariably constitute the majority in the southern, western, central and north-central parts. While in the southern lowlands, this group accounts for over 90% of the population, elsewhere in these parts their share ranges from about 70% to 90%, except in the main tea plantation areas where it drops to about 40% or less. The foremost concentration of the Sri Lankan Tamils is found in the Jaffna Peninsula (which is almost exclusively Tamil) and in the adjacent parts of the northern lowlands. These areas account for about 50% of the total Sri Lankan Tamil population in the country. Smaller Sri Lankan Tamil agglomerations are also found along the eastern littoral where the Tamil settlements are juxtaposed with those of the Muslims and, less frequently, with Sinhalese settlements. Approximately 27.4% of the Sri Lankan Tamils live outside the northern and eastern areas, usually in urban settlements. On a regional scale, the principal concentrations of the Muslims are encountered in the eastern lowlands. In many other areas, however, they form an important segment of the urban and suburban populations. Such pockets of Muslim population are found along the west coast and the Kandyan areas of the interior. The Indian Tamils, the vast majority of whom are engaged in plantation crop production, live in large numbers in the higher areas of the Central Highlands.

The overall population density in the country is high (about 200 per square km), but the density varies from area to area over a wide range. Rural densities that exceed 1,000 per square km are encountered both in the predominantly Sinhalese lowlands of the southwest as well as in the Tamil and Muslim concentrations of the north and east. The interior parts of the north, northwest and the east usually have low population densities.

 The pattern of population distribution in Sri Lanka does not facilitate a clear classification of provinces and districts on an ethnic basis. The system of dividing the country into provinces and districts originated during the early British period when, initially, the island was divided into 5 provinces. In the course of the 19th century, as population and economic activity expanded, the number of provinces was increased to 9. In demarcating provinces, the attention paid to the ethnic aspects of population distribution was always scant, which meant that most provinces of British Ceylon were areas of mixed population. The majority of districts into which the provinces were sub-divided also had, for the same reason, the same characteristic of ethnic heterogeneity. Among the noteworthy exceptions to this were the Jaffna (including Kilinochchi) and Mullaitivu Districts of the north and the Galle, Matara and Hambantota Districts of the south.

Ethnic patterns of the ancient period: an overview:2 The nature of historical ties that bind the different ethnic groups to their respective areas of agglomeration — a subject of some controversy — has an obvious bearing upon the concept of ethnic homelands within Sri Lanka. The popular Sinhalese viewpoint in this regard usually places emphasis upon the continuity of their historiographic traditions which date back over 2,500 years and on the fact that remnants of their ancient civilization are found in all parts of the island including certain areas that are now inhabited largely or exclusively by other ethnic groups.

 There is, indeed, a mass of evidence which shows that up to about the 13th century, the more powerful Sinhalese rulers did exercise sovereignty over the entire island and that, at most times, the north-central parts of the island – the so-called Rajarata – from where they ruled was the heartland of the ancient Sinhalese civilization. In a summary drawn from this evidence, S. Arasaratnam, a leading Sri Lankan historian, states:3

“This (Sinhalese nationalism) was translated into political terms as an aspiration for all-island soverignty. The former (i.e. pre-2nd century BC) loose relationship between three semi-independent kingdoms in the north, southwest and southeast gave place to a more rigorous assertion of the control of Anuradhapura (the northern capital) over the rest of the country. The city now became the capital of the island in a real sense. No doubt, the extent of its control over the other regions depended on the personality of the ruler and the strength of his own kingdom. But gradually an administrative machinery was evolved embracing the whole country. Anuradhapura remained the capital until the eleventh century when the capital was shifted to Polonnaruwa a few miles to the east. In many respects this was the classical age of Sinhalese power, when, in spite of many political vicissitudes, great achievements were recorded in various fields of culture”.

Explaining the nature of political power exercised by the Sinhalese kings of this period, Arasaratnam states:4 “The Sinhalese monarchy was despotic in character, and the king was, in both theory and practice, the source of all authority. His powers were comparable with those of the rulers of contemporary Hindu kingdoms of South India, with the difference that owing to the compactness of the country and the homogeneity of their subjects, the Sinhalese kings were able to exert more personal authority”.

 Tamil connections with Sri Lanka, though somewhat less prominently displayed than those of the Sinhalese in the records of the past, are nevertheless probably as old as those of the Sinhalese. Throughout the ancient period of the country’s history the Dravidian kingdoms and empires of South India exerted powerful influences on the internal politics of Sri Lanka, and (to cite Arasaratnam once again) “Tamil rulers had usurped power periodically and ruled the island, and pockets of Tamil settlements made their appearance from the early centuries of the Christian era”. Till about the end of the 12th century, such periods of Tamil political dominance were brief and ephemeral. But even during periods of undisputed Sinhalese hegemony, Tamils and South Indian connections were invariably important elements in the Rajarata politics. Moreover, at least from the time Polonnaruwa became the capital, Tamil settlements appear to have gravitated towards certain localities becoming nuclei for expansion in the centuries that followed.

The settlement of the country by the Muslims can be traced back over several centuries to the heyday of Arab maritime activity in early medieval times. Confined initially to certain coastal localities where they were mainly engaged in trade, their subsequent spread to various other parts of the country appears to have been directed largely by the nature of their relationships with the different centres of political power in the country and by the course of development of local trade and commerce.

 Our observations on the ethnic aspects of Sri Lanka’s history up to about the end of the 12th century have to remain brief. They are, in fact, repetitions in summary form of what has been stated in a large volume of historical writings. Yet, our sketch is adequate to show the lack of substance in the assertion5 that “Throughout the centuries from the dawn of history the Sinhalese and Tamil nations have divided between them the possession of Ceylon”. Quite clearly, such a contention does not conform to any of the scholarly interpretations of known facts of history that relate to the nature of links that bind Sri Lanka’s main ethnic groups to the island as a whole and to its different parts.

 ‘Medieval’ transformations:  Discarding, then, the claimed primordial origin and continuity of the dichotomous possession of Sri Lanka by two national groups as a distortion of the country’s ancient history, we can now devote attention to the idea that, following the collapse of Sinhalese political power in the Rajarata around the 13th century, the northern and eastern parts of the country became an exclusive domain of the Tamils.

This perception which, as already shown, has received considerable emphasis in the agitation for a Tamil Homeland, is based on certain interwoven claims concerning, first, the pattern of political control which prevailed and, secondly, the nature of demographic changes that occurred, in the northern and eastern areas of Sri Lanka after the 13th century.

 The interpretations of political history that provide the basis for the foregoing claims are that an independent Tamil kingdom which emerged in the 13th century with its base in the Jaffna Peninsula exercised control over the north and east of Sri Lanka, and, as corollary, that the contemporary local rulers of these areas who (allegedly) were largely of Dravidian stock owed allegiance to the Jaffna kings.

That a Tamil kingdom did originate in the northern areas of Sri Lanka during the 13th century and existed thereafter for about 400 years until its conquest by the Portuguese is an undisputed fact of history. Further, there is seldom any dispute on the fact that this ‘Jaffna Kingdom’ not only stood distinct from the other contemporary centres of political power (which, by about the 14th century, had gravitated towards the southwestern parts of the island) but also had sufficient strength to challenge the authority of the southwestern kingdoms over the areas which were peripheral both to the latter kingdoms as well as to its own core territory. That at certain times the Jaffna Kingdom became a vassal state of the empires of South India, and that for a brief spell during the early 15th century it was subjugated by the Sinhalese kingdom of Kotte may be ignored in generalizing on the long-term pattern.

What may not be so overlooked, however, is that the nature and extent of political authority which the kingdom of Jaffna wielded outside its core territory (and, perhaps, the areas adjacent to it further south) varied from time to time. If the Portuguese and the Dutch accounts written after the fall of the Jaffna Kingdom are to be relied upon, it appears that under its more powerful rulers (e.g. Cankili, 1519-60) the kingdom commanded allegiance from the local rulers of the areas as far south along the coast as Trincomalee in the east and Kalpitiya on the west. But this is far too fragile a base for the thesis that the Jaffna Kingdom during its 400 years of existence, covered the northern and eastern lowlands of the island.6

Several authoritative studies of the period after the 13th century suggest that at least up to the late 17th century, the northern and eastern parts of Sri Lanka, in relation to the main bases of political power in the country, were a no-man’s land in which clusters of population and settlement were ruled by local chieftains (Vanniayars) – both Sinhalese and Tamil – who enjoyed a high degree of autonomy from both the Sinhalese kingdoms as well as from the Kingdom of Jaffna.

 After the fall of the Jaffna Kingdom when the Portuguese began to hold sway over certain maritime areas of Sri Lanka (Figure I), the independence of the northern Vanni chieftaincies appears to have increased, the Vanni as a whole remaining at most times outside the mainstreams of conflict between the Portuguese and the Sinhalese kingdoms of the southwest. This pattern appears to have changed once again, when, during the 17th century, the Kingdom of Kandy became the citadel of Sinhalese political power (Figure II). At most times thereafter, until the latter half of the 18th century, the extent of control which the Kandyan kings had over the eastern lowlands appears to have increased, and the Vanni areas of the east as well as some located in the northern interior were regarded parts of the Kandyan domain.7

             Figure I   Figure 1

  Figure II(1)           Figure 2

What the foregoing summary of political history shows is that, once again, the facts relating to the political patterns which prevailed in the northern and eastern parts of the country after the 13th century do not lend support to the notion that they constituted the national territory of a distinct Tamil state.

 Population changes in the northern and eastern areas of Sri Lanka after the 13th century are perhaps of greater salience to the concept of a ‘traditional Tamil homeland’ than what the concept could draw from political history. There is, indeed, a persuasive body of evidence which shows that this period was featured by an increase of the Tamil population in the north and east of Sri Lanka, a process paralleled by a gradual decline of the Sinhalese population in the areas concerned. Furthermore, both in the north as well as in certain coastal localities of the east, Tamil cultural elements gained prominence, as evidenced by features such as the near-universal adoption of the Tamil language by the non-Tamil ethnic groups of these areas and the widespread replacement of earlier Sinhala place-names with those of Tamil. Thus, on a broad regional perspective, the contention that the Tamil population increased and a process of ‘Tamilization’ occurred in the northern and eastern parts of Sri Lanka over these centuries can be accepted as valid.

  Nevertheless, the crucial issue here relates to the spatial dimension of the above processes – one on which there has never been any clarity. Quite obviously, ‘north’ and ‘east’ are not defined territorial entities. Hence, to evaluate a claim over territory, it becomes necessary to seek a more precise understanding of population patterns, frontiers of settlements and the geographical limits of cultural transformations.

‘Modern’ ethnic frontiers: 

The evidence obtainable from historical sources on the period before the establishment of British rule over Sri Lanka in the early 19th century on the spatial dimensions of the processes referred to above is scanty. In fact, it is only after the commencement of detailed and systematic census enumerations during the closing decades of the 19th century that we find a body of information adequate in depth and comprehensiveness of coverage for a precise study of the ‘geography’ of ethnicity in Sri Lanka.

 Apart from the statistical data recorded in the Census Reports, however, for the 19th century as a whole, there are numerous sources of information, mainly in the form of reports prepared by provincial administrators,8 which shed light on various aspects of demographic trends and patterns in the northern and eastern parts of Sri Lanka. In general, what these sources indicate is that throughout the century there was a continuing expansion of the non-Sinhalese population in several areas of the Dry Zone including its northern and eastern parts. If this was a temporal extension of earlier trends, as it appears to have been, it certainly gained momentum from certain changes ushered in by the British. For instance, in some coastal localities (e.g. Batticaloa, Trincomalee and Mannar) where urbanization acquired new vigour, a large share of the employment generated in tertiary activities was taken up by those of the Sri Lankan Tamil community. Again, the sporadic restoration of irrigation works in the 19th century in the Districts of Batticaloa, Trincomalee, Mullaitivu and Mannar (Rugam, Vakaneri, Irrakamam and Districts of Batticaloa, Trincomalee, Mullaitivu and Mannar (Kantalai and Allai tanks of Trincomalee; Rugam, Vakaneri, Irrakamam and Unichchai of Batticaloa; and the Giant’s Tank of Mannar were the larger ones among them) was reported to have had the effect of attracting Tamils and Muslims from the coastal areas to the interior. For the Sinhalese in these areas, the 19th century was a period of continuing recession, as evidenced by many recorded instances of evacuation of Sinhalese village settlements; depopulation of settlements caused by famines, epidemics and droughts; and cultural assimilation by other ethnic groups. That the transformations initiated by the British contributed in some measure to this latter process is borne out by the following extracts from a report on the Trincomalee District of the 1890s.9

“This part of the district (i.e. Kaddukulam Pattu West) is inhabited by Sinhalese villagers of Kandyan descent (who), I fear, are dying out or becoming effaced. This district is most interesting, being dotted over by numerous village tanks, some of which are restored and others abandoned. The villagers retain many of their primitive Kandyan customs, but they are rapidly becoming ‘Tamilized’ … They intermarry with the Tamils and many of them speak Tamil as well as they speak Sinhalese. Even the government schoolmaster is Tamil, and only that language is taught in the only school, and unfortunately, in some cases the Sinhalese villages have been bought out by Tamils, who now own all the paddy land in some villages … and even the names of villages are assuming a Tamil dress. This process is not to be wondered at when the interpreters of the court and the Kachcheri, the petition-drawers, and all through whom the villagers have access to Government officials can speak nothing but Tamil”.

The practice of conducting regular census of population was begun in Sri Lanka in 1871. With the passage of time, these enumerations generated detailed sets of statistical data on various demographic phenomena, including the ethnic composition of the population of each village and town. This latter data can be plotted on large-scale maps for the purpose of identifying the exact location of the settlements belonging to the different ethnic groups. The pattern of settlement distribution identified in this manner facilitates not only the recognition of the exact areas which were being occupied by the different communities but also a more precise determination of their respective ‘settlement frontiers’ than is possible with the type of information available on pre-census times.

  In accordance with the foregoing lines of reasoning, I have prepared a series of maps (Figure III to V) based upon the 1921 population data on villages and towns in the Eastern Province. The method adopted in the compilation of the maps was as follows:

         (a) Preparation of a classification of settlements on the basis of ethnicity– The larger settlements of the province invariably had two or more units of enumeration. In such instances, each unit was regarded a single ‘settlement’ Each settlement (village, town or part of town) was then categorized as ‘Sinhalese’, ‘Tamil’ or ‘Muslim’ on the basis of the ethnicity of its majority group. About 58% of the total number of settlements in the province were ethnically homogeneous. In 88% of the total, a settlement placed in a given ethnic category had 80% or more of its population belonging to that ethnic group.

           (b) Identification and mapping of the location of settlements…..About 30% of all Eastern Province settlements of 1 921 are marked and named on the Map of Ceylon prepared to a scale of 1”:4 miles and published in 1940. This map was used as our ‘base map’. The location of all other settlements had to be identified on the Ordinance Survey Maps (1”: 1 mile) published by the Surveyor General’s Department between 1901 and ‘21, and then transferred to the base map using the standard co-ordinates and the settlements already marked on the base map as guides, thus achieving (as we can confidently claim) a high degree of cartographic accuracy.

My cartographic compilations are not intended to show the spatial differences of population density in the areas to which they relate. In preparing them, the variations among the settlements in respect of population size have been ignored. The settlements, each represented by a dot, range from those with less than 10 persons to large agglomerations of several thousands. The sole purpose of the maps, it needs to be stressed, is that of identifying the territories occupied by the different ethnic groups in 1921.

 For the issues with which the present study is concerned, the feature of crucial relevance borne out by these maps is that, in the Eastern Province as a whole, in 1921, almost all Tamil settlements were confined to a coastal strip barely extending even 10 miles to the interior. The Sinhalese settlements, on the other hand, though comparatively few in number, were scattered over extensive areas of the interior, covering the entirety of the administrative Divisions of Bintenna Pattu; Uda Palatha, Yati Palatha and Meda Palatha of Wevgam Pattu (these were partly in the Divisions of Akkarai Pattu and Sammanturai Pattu of that time); and almost the whole of Panama Pattu. In the northern parts of the then Batticaloa District and in the Trincomalee District, extensive tracts of territory in the interior were either uninhabited or were the venues of scattered Sinhalese settlements This must be taken in the context of the fact that the Sinhalese names of numerous abandoned village tanks marked on my source maps in the uninhabited tracts bear testimony to earlier processes of de-population. The maps show further that the only non-Sinhalese population clusters that were located (in 1921) even a few miles to the interior of the seaboard were those associated with the irrigation works restored in the preceding decades. We can also observe here that the Tamil settlement nearest to the claimed ‘southern boundary’ of the Tamil Homeland (Kumbukkan Oya) was a coastal township of mixed population located about 35 miles north of that ‘boundary’.

This pattern of settlement distribution assumes significance from several points of view. In the first place, considered in the light of my earlier observations on the trends of demographic change in the preceding centuries, the pattern as it prevailed in 1921 represents what may be regarded as the culmination of a long drawn out historical process featured, on the one hand, by territorial advances of the Tamil population and, on the other, retreat and recession of the Sinhalese population This, in turn, implies that the process of ‘Tamilization’ in the eastern lowlands of Sri Lanka had not penetrated significantly into the interior of these districts even at its most extensive territorial spread.

 The second point of significance arises from the fact that, often, the ‘Tamil Homeland’ is being defined with reference to modern administrative units. Given the spatial patterns of ethnicity borne out by our maps, the demand by one ethnic group for exclusive proprietary rights over Provinces and Districts which encompass extensive tracts of territory which it had never occupied (and much of it, in every sense, the traditional homeland of people belonging to other ethnic groups) appears in its true light as one which lacks a rational basis.

 mr fig 4         Figure 3

    mr fig 5      Figure 4

  mr fig 5      Figure 5

In a functional sense, the sparsely settled interior of the eastern lowlands of Sri Lanka was not a hinterland of the settlement clusters of the littoral. Both the interior as well as the littoral were very largely rural. Hence, a ‘core-periphery concept’ is of no relevance to the issue. More specifically, there is no empirical basis for a theoretical assertion that because there was a numerical preponderance of the Tamils in the coastal areas, the inland areas, regardless of the traditional rights of other ethnic groups, should form a ‘traditional hinterland’ of the Tamil areas.

 The fact that the eastern littoral itself is not ethnically homogeneous (well known but frequently overlooked) is also illustrated by our maps with a fair degree of clarity. The littoral is as much the traditional homeland of the Muslims as it is of the Sri Lankan Tamils. And, in several areas, the former group constituted the numerical majority which, as already stated, has no common cause on this issue with the latter.



A notion which has found repeated expression in the demand for an exclusive Tamil Homeland within Sri Lanka is that throughout the recent past (especially in the period after independence) Sinhalese immigration and settlement in the northern and eastern parts of the country have had the effect of changing the ethnic composition of the population of these areas, in certain instances, making the Tamils a minority in areas that were populated largely or exclusively by them before this process began. The process itself has been attributed to state-aided colonization and occupation of Crown land with state connivance.

Population change at district-level: 

In order to examine the validity of this notion, it is necessary, first, to focus on certain sets of information on the relevant processes of demographic change in the Administrative Districts in which the Sri Lankan Tamils have constituted a large segment of the population and where there have been noteworthy changes in the ethnic composition of the population over the past few decades. Among these are the districts of Mannar, Vavuniya, Mullaitivu and Trincomalee in which, between 1921 and ‘81, the Sri Lankan Tamil share of the population has dropped, respectively, from 58 to 51%, 70 to 57%, 89 to 76% and 53 to 34%. This has been paralleled by an increase in the population ratios accounted for by other ethnic groups. For instance, in Mannar, Mullaitivu and Trincomalee, the Sinhalese population has increased from 2 to 8% and 3 to 6%, and 5 to 34% of the respective district totals. Similarly, in Mannar, Mullaitivu and Vavuniya, the Indian Tamil percentages have increased, respectively, from 8 to 1 3, 5 to 1 9 and 1 to 1 4. In the areas that constitute the eastern district of Ampara, concurrent with a large increase of the Sinhalese population from 8 to 38%, there has been a decrease of the Muslim population from 56 to 42% and the Sri Lankan Tamil population from 31 to 20%.

 The foregoing facts confirm the view that in some of the Administrative Districts under scrutiny the recent increase of the Sinhalese population has been an important cause for the decrease of the Sri Lankan Tamil share of the population. This had clearly occurred in Ampara, Trincomalee and Mannar. On Mannar District, however, two qualifications must be made. The Sinhalese population in the district still amounts to only 8% of the district total. Secondly, the increase of the Sinhalese population in this district (2 to 8% from 1921 to ‘81) is comparable in magnitude to the concurrent increases recorded by Sri Lankan Tamils in certain other parts of the country such as the districts of Nuwara Eliya (1 to 14%), Matale (1 to 6%) Badulla (2 to 6%) and Colombo (3 to 7%). What these latter observations imply is that immigration and settlement leading allegedly to significant changes in the composition of population at district-level have, in fact, occurred only in two administrative districts of the country namely, Ampara and Trincomalee.

 The Ampara district, demarcated as a separate administrative entity in 1961,is made up of the Administrative Divisions of Panama (Lahugala and Potuvil), Akkarai Pattu, Nintavur, Karaivaku and Sammanturai, and a large part of the former Bintenna Division (now Wevgam Pattu), all of which were parts of the former Batticaloa District. In 1921, these areas had, in aggregate, a population of 68,643 of which 56.5%, 34.5% and 8.0% were accounted for, respectively, by Muslims, Sri Lankan Tamils and Sinhalese. As in most parts of the Dry Zone at that time, here too the population was largely confined to coastal localities. In fact, the Division of Nintavur and Karaivakku, located astride the sand bar of Batticaloa lagoon, and the coastal townships further south (Attalachenai, Kurunkottitivu, Sammanturai, Tirukkovil, and Potuvil) accounted for some 82% of the population aggregate of these divisions. These were inhabited almost exclusively by Tamils and Muslims. As already shown, in the interior, made up largely of scattered Sinhalese settlements, the only population concentrations were those associated with the Cal-aar Scheme of the Gal Oya Valley (one of the earliest ‘colonial’ irrigation restorations) and a few other smaller 19th century irrigation works, occupied largely by Tamils and Muslims. Thus, as Table 3 illustrates, the Sinhalese, though constituting only a tiny segment of the population of these areas at that time, accounted for over 80% of the population in the Divisions of Panawa, Wevgampattu and Bintenna which, on a rough estimate, cover about 75% of the ‘district’ area.

Table 2: The Ethnic Composition of District Populations, 1921 and 1981

(Sinhalese and Sri Lankan Tamils only)


Northern & Eastern Districts


Sri Lankan Tamils








































                            Other Districts

Colombo & Gampaha           



































Nuwara  Eliya                        


















































Batticaloa & Ampara, Badulla & Monaragala, and Anuradhapura & Polonnaruwa were single administrative entities in 1921. Puttalam and Chilaw were separate districts in 1921 but constituted a single administrative entity in 1981. Adjustments and desegregations necessary for purposes of comparison of 1921 data with those of 1981 have been made on the basis of sub-district level data. No adjustments have been made for minor boundary changes of districts between 1921 and 1981 (such as the annexation of Ambagamuwa Korale, formerly in Kandy District with Nuwara Eliya District).

         Table 3: Population Change in Ampara District, 1921 and 1981



% of the Division Population

























Akkarai Pattu




























Maha-Oya &Padiyatalawa(Bintenna)







Wevgam Pattuwa







Note:  Less than 0.05 is indicated as zero.

In compiling this table, adjustments and disaggregations have been made for the following changes:

a )   The Vidane Divisions Nos. 145 and 146 which in 1921 were in Bintenna Division are now in the Eravur Division of Batticaloa District. The former Bintenna Division (less 145 and 146) now constitute the Divisions of Maha Oya and Padiyatalawa. 

b)   Wevgampattu Division was carved out of the 1921 Divisions of Sammanturai, Akkarai Pattua and Bintenna.

c)   Lahugala and Potuvil Divisions of 1981 formed the Panawa Division of 1981.

The tabulation omits the other ethnic groups (Burghers, Europeans, Indian Tamils and Veddhas) all of which were numerically very small.

  This population pattern of 1921 appears to have remained more or less unchanged until the early 1950s. On a rough estimate based on the 1946 Census data, during the preceding 25-year period, the aggregate population of areas that came to constitute Ampara District increased by a mere 20,000. Then, with the establishment of the Gal-Oya Colonization Scheme, there began a massive influx of people into the interior of Ampara. By 1981 the population of the district had reached 388,786. Accompanying this expansion, there was a drastic change in the ethnic composition of the population with the Sinhalese share rising to 38% of the total.

  That the areas which constitute the District of Ampara were never an ‘exclusive homeland’ of a single ethnic group is already abundantly clear from our earlier discussions and from our pre-Gal Oya data (Table 3). Nevertheless, the issue which requires more specific probing is whether the Gal Oya Scheme had the effect of ‘encroachment’ by one ethnic group of land traditionally occupied by another, not for centuries but at least for a few decades before the establishment of the scheme. For this, we have to turn to some facts about the Gal Oya Scheme. The scheme lies in the districts of Monaragala (into which only Senanayake Samudra, its main reservoir falls), Ampara and Batticaloa. In relation to the administrative divisions of this area, the peasant settlements of the scheme are confined almost entirely to the Divisions of Wevgam Pattu, and the interiors of Sammanturai Pattua and Akkarai Pattua of Ampara District and the Manmunai West Division of Batticaloa District.

  The 42 Settlement Units into which the Gal Oya Scheme is divided belong to three distinct irrigation systems (Left Bank, Right Bank and River Division), all fed by Senanayake Samudra and some smaller storage tanks in its command area. The upper and middle portions of the Left Bank System are located largely in the Wevgam Pattu Division of Ampara District which, as already shown, was an exclusively Sinhalese area in pre-Gal Oya times. While the majority of settlements in these parts of the scheme are now occupied by Sinhalese colonists, some have an ethnically mixed population, and some a Muslim population. The lower parts of the Left Bank System extend into the Manmunai West Division of Batticaloa District where 9 out of the total of 42 settlement units are located. These are occupied exclusively by Tamil settlers. The Right Bank System runs through the Wevgam Pattu, once again an area of exclusively Sinhalese village settlements prior to the establishment of the Gal Oya Scheme, and extends into the Akkarai Pattua Division in which the Muslims constituted about 60% of the pre-scheme population. The upper reaches of the Right Bank System are now occupied by sugar plantations and settlement units inhabited by the Sinhalese. The Muslims and Tamils who account for 52% and 16%, respectively, of the present population of the Right Bank System predominate its lower and middle parts. Before the Gal Oya Scheme began, the areas over which the River Division extends were of mixed population with the Sri Lankan Tamils and the Muslims forming the overwhelming majority. Though some reallocation of land did occur when these areas were absorbed into the Gal Oya Scheme, the tiny share of the present River Division population (2%) which Sinhalese colonists account for today is substantially less than what it was before the commencement of the Gal Oya Scheme.

Table 4: Distribution of Ethnic Groups in the Gal Oya Scheme Area – 1981 (percentages in parenthesis) 

  Sinhalese Tamils Muslims Others Total
Left Bank System

64,451 (51.5)

42,114 (33.7)

18,200 (14.5)

327 (0.3)

125,092 (100.0)

Right Bank System

12,067 (32.1)

5,975 (15.9)

19,436 (57.7)

99 (0.3)

37,594   (100.0)

River Division

3,228 (2.1)

41,085 (26.5)

110,119 (71.0)

690 (0.4)

155,122   (100.0)

 Source:  Economic Analysis of Productivity and Equity: Issues in Irrigation Distribution, mimeographed, March 1984, (A report prepared for the USAID-Government of Sri Lanka Water Management Programme, available at the Agrarian Research & Training Institute, Colombo). The data are based on the Population Census of 1981.

The demographic changes in Trincomalee district with which we are presently concerned were similar to but somewhat more complex than those of Ampara district. The complexity arises from the relatively more pronounced periodic variations of trends.

 The period up to about the 1870s was featured by a slow expansion of the district population – an expansion which was due entirely to the emergence of Trincomalee town as an urban centre and to the increasing agglomeration of people in the settlements located along the southern periphery of the Trincomalee harbour. The Trincomalee town was largely inhabited by Tamils, while settlements to the south of the harbour were both Tamil and Muslim. The interior of the district was largely uninhabited except in its northern parts where there were many scattered Sinhalese hamlets. In aggregate these accounted for less than 5% of the district population of that time.

 The period from about 1870 to 1900 forms the second distinctive phase in the recent history of demographic change in the area, one which was featured by an acceleration of the rate of population growth, an expansion of its spread and a change in its ethnic composition. Trincomalee town which continued to dominate the demographic scene attracted people from other districts. Several district administrators and compilers of census reports of this time have observed that while Jaffna district was the foremost source of migrants into Trincomalee town and to the district as a whole, the settlements located along the southern periphery of the harbour were attracting Tamil and Muslim settlers from Batticaloa district.The concurrent extension of the frontiers of population distribution towards the interior of the district was caused mainly by the restoration of Allai Tank (1876), Kantalai Tank (1877) and about 22 village-level irrigation works between 1870 and 1905.10 The Kantalai scheme, the largest among these restorations, attracted a fair number of Tamils and Muslims to the interior of Tambalagama Division which up to that time had been almost entirely uninhabited. The opening up of new land under agriculture here, which was concentrated around the emerging township of Tambalagama was reported to have been undertaken largely by Tamil entrepreneurs and speculators from Jaffna, including a firm called the Jaffna & Batticaloa Agricultural Company which had purchased a large block of land under the restored reservoirs.11 The Allai scheme is reported to have had the same effects on the areas south of the Trincomalee harbour where, immediately after the restoration of the reservoir, land was “quickly taken up by the Tamils and Moormen from Batticaloa district”. These developments were paralleled by an on-going process of decay in the Sinhalese settlement areas of the interior (Kaddikulam Pattu West) where the district administrators of the time repeatedly observed an exceptionally high rate of mortality, frequent abandonment of settlements and various other processes of unmitigated impoverishment.12 The overall impact of these changes on the ethnic composition of the area was that the Sri Lankan Tamil share of the district population which was 30.0% in 1871 had increased to 60.1% by 1901.

Table 5: Population Change in Trincomalee District, 1921 & 1981                                                          


% of Division Total

Division (as in 1981)  










Trincomalee Town & Gravets   

















































Moraweva & Gomarankadawala







All divisions         







Note:     This table is prepared on the basis of the 1981 Divisional framework which was totally different to that of 1921. For example, the Tambalagama Pattu of 1921, now falls into the divisions of Kantalai and Tambalagama. A part of it also extends into the Town & Gravets Division. The old Koddiyar Pattu now forms the Divisions of Muttur and Seruwila. The present Moraweva and Gomarakadawala Divisions constituted the old Kaddikulam Pattu West Division. Adjustments necessary for purposes of comparison have been made on the basis of ‘sub-division’ (i.e. Village Headmen’s Division) data. The possibility of there being small errors in the 1921 data shown here cannot be ruled out.

 The third phase of population change in the Trincomalee District which spans the early decades of the 20th century was characterized by a slowing down (and in some instances a reversal) of earlier trends. For instance, between 1901 and 1931, the district population increased by only 31% as compared with the 45% increase recorded during the preceding 30-year period. In the absence of fresh stimuli for growth of the interior, the pattern of population distribution remained constant. Regarding the ethnic composition, there was a perceptible decline of the Tamil population ratio from the peak it had reached in 1901 to 52.2% by 1921 and to 46.6% by 1931.

 The inter-censul period 1931 to 1946 was an unusual phase in the demographic history of the district on account of the effects of the Second World War. The most noteworthy feature of this phase was the phenomenal increase of population recorded in the Trincomalee town and its suburbs (247%), an increase which is attributable to the crucial role which Trincomalee played in the British war effort as a naval base.

  The foregoing account of population change in Trincomalee District during the British period should provide the backdrop to the study of changes which have occurred during the more recent past. Briefly stated, the salient facts are, first, that throughout the period before independence, the Tamil and Muslim population concentrations along the seaboard constituted the overwhelming majority of the district population; secondly, that the spread of the Tamil and Muslim populations to certain parts of the interior (Tambalagama and Koddiyar divisions) was a phenomenon of relatively recent origin; thirdly, that the change in the ethnic composition of the population in Trincomalee District during the closing decades of the British period was caused largely by processes of tertiary development in the Trincomalee town and along the periphery of the Trincomalee harbour; fourthly, that the decline of the Sri Lankan Tamil share of the population was a trend which commenced at the turn of the century; and finally, that even at the time of independence, while certain parts of the district interior remained undeveloped and uninhabited, in other parts (especially those which constitute the present Divisions of Moraweva and Gomarankadawala) many purāna (ancient) Sinhalese settlements had continued to survive.

  In the period after independence, following a brief post-war lull, there was once again an enhanced rate of population growth in Trincomalee District. This was largely due to further urban development in and around the town of Trincomalee and to an influx of people into the newly launched peasant settlement schemes of the interior.

  New employment opportunities generated by the increasing importance of the port of Trincomalee and the location of certain manufacturing industries around the harbour made the town and its suburbs more ethnically heterogeneous than it was at the time of British rule, leading, indeed, to a gradual lowering of the Tamil ratio of the urban population. In this respect, Trincomalee town has by no means been unique. In fact, in the majority of the larger towns in Sri Lanka (Jaffna town itself forming a conspicuous exception) there has been a perceptible decline in the earlier numerical preponderance of the respective majority ethnic groups.

  Among the peasant settlement schemes of Trincomalee District, in respect of demographic impact, the expanded Kantalai scheme where a new programme of land development and alienation commenced in 1953, and the Moraweva scheme launched in 1958, have been the foremost. Several other smaller schemes (e.g. Galmetiyawa, Mahadivul Weva, Wan Ela) which, in aggregate, have involved the settlement of some 1,500 families, have also contributed marginally to the recent demographic changes in the interior parts of the district.

 The Moreweva scheme and the smaller schemes referred to above are located in the administrative divisions of Gomarankadawala and Moraweva which, as already shown, have always been almost exclusively Sinhalese areas with that group accounting for over 90% of the population. At present, however, an estimated 35% of the settlers in these schemes are non-Sinhalese. In contrast, the Kantalai scheme, the largest of its type in the district, has resulted in an increase of the Sinhalese population in an area where there was a non-Sinhalese majority up to the early 1950s. This is of course, a generalization about an area of extremely sparse population which stretches northwestward over a linear distance of some 15 miles from the Kantalai reservoir. When the pattern of land alienation in the scheme is examined in detail, it becomes evident that land has not been allocated to Sinhalese settlers in the Tambalagama and Vendrasanpura units of the scheme, which, in fact, include the areas settled by Tamils and Muslims following the earlier restoration of the Kantalai scheme in 1877. Thus, today, in the Kantalai scheme as a whole, Sinhalese settlers constitute only about 60% of the total number of families settled. In so far as it is possible to ascertain from a comprehensive body of statistical and cartographic evidence presently at our disposal, prior to the initiation of the expand Kantalai scheme, there were only two small village settlements (both predominantly Tamil) in those parts of the scheme (Kantale, Akbopura, Raja Ela, Gantalawa and Mullipotana units) that are now occupied largely by the Sinhalese.


The peasant settlement schemes of Gal Oya and Kantalai which we have commented upon are among some 100 ‘Major Colonization Schemes’ established in the drier parts of Sri Lanka since the mid-1930s. The emphasis placed on ‘colonization’ of the Dry Zone in the development strategies pursued by successive governments over the past few decades is attributable fundamentally to the objectives of increasing food production and providing relief to the acute ‘land hunger’ prevailing among the peasantry in the densely populated parts of the country. Since the large majority of these schemes were based on restorations of irrigation works constructed during the time of theRajarata kings,13they have also had a symbolic significance for efforts in national resurgence.

  The sites of the major colonization schemes were invariably areas of very sparse population located within the flood-plains of rivers that traverse the dry zone. As in Gal Oya and Kantalai, such areas have had large tracts of land potentially suitable for intensive forms of agriculture and for the re-settlement of people from the densely populated rural areas of the country. In many instances, at least up to about the mid-1970s, the exact locating of colonization schemes has also been guided by the availability of restorable hydraulic works of the past. The main source areas from which the settlers have been drawn are either in the southwestern and central parts of Sri Lanka or in the coastal areas of the north and the east.

  Over the past 50 years, about 100,000 peasant families have been settled in these colonization schemes (excluding the Mahaveli Settlement Systems). Making allowance for increases in family-size after initial settlement, it could be estimated that by the early 1980s, the settlers in the ‘Major Colonization Schemes’ were accounting for about 8% of the total population of the country.

  In the late-1960s, a second category of colonization schemes – namely, Youth Settlement Schemes – was launched in Sri Lanka, representing a special focus on generating employment for the younger segments of the labour force. By the early 1980s there were 40 schemes of this type having an aggregate of 7,150 settlers, some of whom have been drawn from outside the districts in which the respective schemes are located.14

  Implicit in the assertion that the state-sponsored settlement schemes that are located in the northern and eastern parts of the country have had the effect of Sinhalese ‘encroachment’ upon areas traditionally inhabited by other ethnic groups in former times is the notion that the Sinhalese have received a disproportionate share of the benefits of these schemes. Hence, a brief examination of the extent to which this notion has substance is of relevance to the wider issues with which this study is concerned.

            Table 6: Settlement and Land Development in Major Colonization Schemes (as at the end of 1980) 


Number of Schemes

Number of   Allottees

Extent of Land paddy

Developed Highland











Nuwara Eliya                                      










Uda Walave Scheme                        











































































All districts                                 






Sources: Land Commissioner’s Department

Uda Walave Scheme data, from the Agricultural Research Station atAngunukolapelessa

Notes   1. Hambantota data exclude those on the Uda Walave Scheme

              2.  Kurunegala data include the Left Bank unit of the Rajangana Scheme which has  been reckoned as 1 scheme.

               3.  Anuradhapura data include the Right Bank unit of the Rajangana Scheme which has been reckoned as 1 scheme (distinct from the LB unit)

              4.  Parts of the Gal Oya Scheme located in Batticaloa District have been regarded as constituting 1 scheme and added to the district total.

              5.  Ampara data refer to those portions of the Gal Oya Scheme located in that district.

             6.  There are no ‘Major Colonization Schemes’ in the Districts of Colombo Gampaha, Kalutara, Galle, Matara and Kegalle.

  According to records of the Land Commissioner’s Department, the number of Sinhalese family units settled in the ‘Major Colonization Schemes’ (excluding Uda Walave Scheme and settlements under the current Mahaveli Programme) represents 76.4% of the total number of settler families. The inclusion of Sinhalese settlers of the Uda-Walave Scheme (for which we have only a rough estimate) would increase the Sinhalese share of the total to about 81%. This latter figure is clearly higher than the Sinhalese share of the country’s total population, a fact which needs to be looked at in the light of several qualifying observations. Perhaps the most important among them is that, from the viewpoint of economic function, the Sinhalese constitute a proportionately larger segment of the peasantry than the other ethnic groups, in the sense that subsistence agriculture, which the colonies of the Dry Zone were primarily intended to promote, has always been largely a Sinhalese segment of the economy. Moreover, since land settlement in Sri Lanka has hitherto by-passed the Indian Tamils (who have been largely confined to the plantation sector), the benefits of peasant settlement schemes available for distribution among other ethnic groups have been larger than the aggregate of their respective proportions of the national population. Yet another qualification is that, if data on youth settlement schemes (in which the Sinhalese account for only 43.4% of the total number of allotees) are included in our estimates, the Sinhalese share would drop to about 78% of the total. Finally, as the following data (Table 7) show, the peasant settlement schemes that have been established in the Northern Province of Sri Lanka have a small minority of Sinhalese settlers and no Sinhalese peasants have been settled in the schemes located in the districts of Jaffna, Mullaitivu, Batticaloa and Mannar.

Table 7: Allocation of Land in Peasant Settlement Schemes of the Northern and Eastern Provinces


Total number of allotees


(% of total)






















                                     *Ampara data exclude those on the River Division of the Gal Oya Scheme in which the Sinhalese form only a small minority.

Source: Land Commissioner’s Department

  The illegal occupation and use of Crown land has been a widespread phenomenon in Sri Lanka throughout the recent decades. Indeed, in many parts of the country attempts that have been made by successive governments for well over 100 years to check this process of ‘Crown Land Encroachment’ have only had meagre success. The extent to which this has constituted a process of Sinhalese intrusion into areas traditionally inhabited by other ethnic groups cannot be ascertained with the data currently at our disposal. Nor can we verify in quantitative terms the alleged government connivance in this process.

The only statistical data presently available on Crown land encroachment in the country as a whole are those generated through a survey conducted by the Land Commissioner’s Department in 1979. Given the nature of the process of encroachment and the methodology adopted in this survey, the results of the survey (on which Table 8 is based) can be treated as no more than rough estimates. Our tabulation shows, inter alia, that at the time of the survey close upon a million acres of land in various parts of Sri Lanka (about 3.5% of the country’s area) were being illegally occupied by people, and that the spatial distribution of this phenomenon does not conform to a pattern explainable with reference to a uniform set of factors, except perhaps the factor of availability of land for people to encroach upon.

 In many areas of the dry zone, including its north and east, it appears that encroachment has been very largely a product of the spill-over of excess population from settlement schemes coupled often with the phenomenon of converting land under shifting cultivation (chēna) into permanent forms of agriculture. There is evidence for this especially from the older ‘colonization schemes’ where all reserve land within the schemes and large tracts of land surrounding them have been illegaly occupied. This is probably why relatively high levels of encroachment have been reported from districts like Kurunegala, Trincomalee, Puttalam, Anuradhapura and Batticaloa (Table 8).


In recognition of the fact that illegal occupation of Crown land is often a response to genuine economic pressures, it has been government policy to regularize encroachment, either when encroachment is fait accompli (and eviction can cause more harm than good) – an old practice followed even under the Raj – and/or where encroachment does not come into serious conflict with considerations of environmental conservation and land-use planning. Accordingly, under a recently launched programme of legalizing encroachments, legal sanction has been granted (up to May 1985) for 518,459 acres of illegally occupied land, which extent represents about 55% of the total extent of encroachment recorded at the survey of 1979. From the district data available on this programme, there seems to be no grounds for a charge of discrimination against the minority ethnic groups in the implementation of the programme. In fact, the only known large-scale government operation against encroachers in the recent past has been widely interpreted as an act of discrimination against the majority community.15


The concept of the ‘Traditional Tamil Homeland’ as promulgated by its exponents is based on the notion that from the distant past the island of Sri Lanka comprised the territories of two distinct Nation States which were arbitrarily unified in the formation of British Ceylon. Through a brief survey which draws from authoritative writings we have argued that such a notion does not conform to known facts and unbiased interpretations of the country’s history. E.B. Denham was probably highlighting the legacies of the Rajarataperiod when he asserted that “…among the races (sic) that are most numerous in Ceylon … only one race can regard Ceylon as the home of the nation and the shrine of its national traditions”. But such an assertion is irrelevant to the themes of our study.

 The present study acknowledges that, following the collapse of the Rajarata civilization in which the Sri Lankan Tamils constituted a vital element, there was a long drawn out process of ‘Tamilization’ in the north and in certain localities of the eastern littoral. However, the population pattern which prevailed at the culmination of this process in the early decades of the present century, considered against the backdrop of earlier trends, drives us to the conclusion that this process was scarcely felt in the interior, at least of the eastern parts of the island which, indeed, remained throughout a ‘traditional homeland’ of the Sinhalese peasantry. Accordingly, we find no basis whatever for the assertion that the ‘Tamilized’ areas correspond to the Northern and Eastern Provinces of today.

Our probe into the demographic impact of recent land settlement schemes shows that the alleged Sinhalese ‘intrusion’ into the ‘Traditional Tamil Homeland’ is a myth. State-sponsorship has admittedly been a vital element in land settlement schemes, for, the schemes were meant for the poorest segment of the population -the landless peasantry. But neither in this nor in state responses to Crown land encroachment do we find any evidence of discrimination against the Sri Lankan Tamils.

The Northern and Eastern Provinces which constitute about 29% of the total area of Sri Lanka are inhabited by 72.6% of the Sri Lankan Tamil population. This latter is equivalent to approximately 9% of the total population of the country. In a densely peopled country like Sri Lanka, where the prevailing pressure of population on land is intense, 9% of its population claiming exclusive rights over 29% of its territory is in itself somewhat unfair, Moreover, the acute scarcity of resources for agriculture from which Sri Lanka suffers implies that the country cannot afford to have uninhabited ‘buffer zones’ between concentrations of different ethnic groups. Nor can such uninhabited or underutilized tracts of territory be reserved untouched as future lebenssraumfor any one ethnic group of the country.

 Author’s note: This paper, written in 1985, was presented at the ICES seminar on ‘The Economic Dimensions of Ethnic conflict of Sri Lanka’ held in August that year. Since then the paper has been referred to and commented upon in several widely circulated monographs and articles, some of which contain misrepresentations of both its general theme as well as its specific factual content. Since the seminar paper has hitherto remained unpublished, the ICES has agreed to carry it in the present issue of its journal. Certain sets of statistical data used in this paper are based on unpublished sources made available by the Land Commissioner’s Department, for which we record our gratitude. The extracts from political statements cited in the paper are from a collection of documents belonging to Professor K.M. de Silva, from whose book(History of Sri Lanka) we have also borrowed Figures I & II.

Editor’s Note: In 1991/92 the Colombo Directors of ICES asked a Tamil geographer to cross-check the Peiris data for accuracy/inaccuracy. No report followed.

G.H. Peiris is Professor of Geography, University of Peradeniya, Peradeniya, Sri Lanka. 


 1.    See, Indrapala, 1970, 135-36 and Pathmanathan, 1972, from which the quotations that appear in this paragraph have been taken.

 2.    This brief account leans heavily on the major writings relating to the history of Sri Lanka, especially, University of Ceylon 1959-60, Nicholas & Paranavitana 1961; Arasaratnam 1964 and de Silva 1981.

 3.    Arasaratnam, 1964, p.52.

4.  Arasaratnam, 1964. p.55.

 5.  The statement cited here has an interesting history of its own. It was initially made, more or less in the same verbal form, in the so-called ‘Cleghorn Minute’, a brief account of Sri Lanka written (according to Ralph Peiris) in 1799. It took Fredrick North, the first Governor of British Ceylon, to realize that “the greater part of what he (Cleghorn) wrote was absurd” (Nadarajah, 1953). Long years later, in the     1930s, C. Sundaralingam, a prominent politician of that time, is reported to have made more or less the identical statement in his then lonely campaign for the creation of a Tamil State of Eelom (sic). Whether the TULF in drafting its famous resolution of 1976 borrowed from Cleghorn or from Sundaralingam is not quite clear.

 6.  For a contrary view that the Jaffna Kingdom did exercise control over the northern and eastern parts of Sri Lanka, see, Pathmanethan (1978, specially chapter 9). where the author has based himself largely on passages from Queyroz, the Dutch Governor Van Goens, and the Tamil writingsVaiyapatalKailayamalai, and Vaippana Vaipavamalai. Queyroz and Van Goens were, of course, writing on what they were claiming was the former size of a kingdom which had fallen under their control, and thus had reason to exaggerate.

 7.  The foregoing account is based on many writings, among which special mention must be made of Arasaratnam 1964 and 1971; Indrapala 1970; Pathmanathan 1972 and 1978; de Silva 1981; and Devaraja 1972.

 8.  For evidence on the processes referred to here, see Rutherford 1871; Keane 1905, pp. 14 and 38; Denham 1912. pp.89-93; Turner 1922, pp.116-118; and the Administrative reports on the Eastern Province for the years 1867, 1877, 1878, 1879, 1887, 1889, 1901, and 1915. That these processes were in operation in the interior of the Northern Province is evidenced by the accounts in Boake 1888, Lewis 1896 and Denham 1912.

 9.  Lushington, p.171898.

10. See, Administrative Reports on Trincomalee district, 1878, p.120; 1879, p.175;1887, p.171;1891,    p.168; Sessional Paper 45 of 1905, p.14-15; Denham 1912. p.108.

 11. Administrative Report on Trincomalee District, 1887, p.171.

 12. Administrative Reports of the Eastern Province, 1987, p.100; 1898, p.181; 1901, p.114; 1915, p.15.

 13. Brohier, 1933.

 14. The other categories of state-aided settlement schemes – the so-called Village Expansion Schemes and Highland Colonization Schemes – are omitted from this discussion because they seldom involved long distance migration of settlers. Under the former type, since 1935, about 940,000 acres have been alienated to some 710,000 family units. Of these, 207,698 acres, allocated to 113,587 families, are in the districts of Jaffna, Mannar, Vauniya, Mullaitivu, Batticaloa and Trincomalee. Of this latter total, Sinhalese allotees account for only 7.1% and are confined to a few schemes in Trincomalee, Batticaloa and Vavuniya. There ere only 2,121 settler families in all Highland Colonization schemes of Sri Lanka among whom the Sinhalese account for 79.3%.

15. For a detailed account of the eviction of encroachers from the System B area of the Mahaveli Development Project, see, Gunaratna (1983), pp. 82-211.


Administration Reports for Eastern Province, Batticaloa District and Trincomalee District (as indicated in the text).

 Arasaratnam, S. (1964), Cay/on, Englewood Cliffs, NJ.

 Arasaratnam, S. (1971), ‘The Indigenous Ruling Class in Dutch Maritime Ceylon’Indian Economic and Social History Review, 8(i). 57-71.

Boake, W.J.S. (1888), Manna, :A Monograph, Government Printer, Colombo.

 Brohier. R.L. (1933-35), Ancient Irrigation Works of Ceylon. 3 vols., Government Printer, Colombo.

Canagaratnam, S.O. (1921), Monograph of Batticaloe, Government Printer, Colombo.

Census Reports of 1881, 1901, 1911, 1921, 1931, 1946, 1971, and 1981 published by the Government of Ceylon, Sri Lanka. Denham, E.B. (1912), Ceylon at the Census of 1917, Government Printer, Colombo.

 de Silva, K.M. (1981), A History of Sri Lanka, OUP, London.

Dewaraja, L.S. (1972), The Kandyan Kingdom, 1707-7760, Lake House Investments. Colombo.

 Gunaretna, Melinge H. (1988), For a Sovereign State, Sarvodaya, Ratmalena.

 Indrapala, K. (1970), ‘The Origin of the Temil Vanni Chieftaincies of Ceylon’The Ceylon Journel of Humanities 1(2), pp.111-140.

Keane, John (1905), ‘Report on Irrigation in Ceylon’, Sessional Paper XL V-1905, Government Printer, Colombo.

Lewis, J. P. (1896), Manuel of the Vanni District, Government Printer, Colombo. 

Ludowyk, E.F.C. (1962), The Story of Ceylon, Faber and Faber Ltd., London.

 Lushington, C.M. (1898), Administration Report on Trincomalee District for 1898, Government Printer, Colombo.

 Nadarajah, K. (1960), ‘New Light on Cleghorn’s Minute’, Journal of Royal Asian Studies (Ceylon Branch) JRAS (CB) Vol. X.

Nicholas, C.W. & S. Paranavitana (1961), A Concise History of Ceylon,Colombo.Pathmanathan, S. (1972), ‘Feudal Polity in Medieval Ceylon: An Examination of the Chieftaincies of the Vanni’, CJHSS ns, 2(2), 118-13.

 Pathmanathan, S. (1972), ‘Feudal Polity in Medieval Ceylon: An Examination of the Chieftaincies of the Vanni’, CJHSS ns, 2(2), 118-13.

Pathmanathan, S. (1978), The Kingdom of Jeffna, Colombo.

 Pieris, Ralph (1954), ‘Administration of Justice and of Revenue in the Island of Ceylon under the Dutch Government, (Cleghorn Minute)’, JRAS(CB) III(2), pp.125-152.

 Rutherford, S. (1871), Administration Report on Trincomalee District for 1871,Government Printer, Colombo.

 Sessional PapersGovernmant of Ceylon, 45 of 1905.

 Turner, L. (1922), Population Census of Ceylon 1927, Government Printer, Colombo.

 University of Ceylon, History of Ceylon, Vols. I and II, Colombo.


Filed under economic processes, historical interpretation, politIcal discourse, population, power politics, Sinhala-Tamil Relations, slanted reportage, sri lankan society, truth as casualty of war

14 responses to “An appraisal of the concept of a traditional Tamil homeland in Sri Lanka

  1. chandre DW

    Glad to see this article in its entirety and in a conveniently accessible place.

    Gerry has limited himself to the `spatial’ aspect of the Topic.
    However, if we go back in time, during the early days (up to mid 1920s) Lanka was ruled (by the British) with the turbaned first-class upper-caste Tamil families having virtual hegemony over both Tamil and Sinhala matters. The opposition to the colonial administration come from the Ceylon National Congress where the Sinhalese and Tamils were united. At that time the issue of how to divide Lanka between the Tamils and the Sinhalese never came up strongly, with Tamil leaders in the driving seat. This `pan-tamil’ period lasted about three decades, and ended in the late 1920s. The sinhala-tamil unity under the turbaned aristocracy was broken by the Governor’s strategy of holding out a `secure seat’ for Arunachalam in Colombo. Here the back-room promises were scuttled by a clever governor (and a misguided Arunachalam who applied personal pressure on the governor when he visited Jaffna). The governor successfully achieved the divide and rule he searched for, splitting the unity between the Tamils and the Sinhalese, and Arunachalam leaving politics in the hands of the much less liberal-minded Ramanathan.

    Then came the Donoghmore report, when it clearly dawned on the Tamil leadership that universal franchise implied that the power they had taken for granted was no longer safe; every one was equal irrespective of caste, race or gender!

    Ramanathan’s attempts to inscribe the cast hegemony (while also denying the vote to women) in the constitution failed. The truly progressive young leaders in Jaffna went on their now infamous `Jaffna boycott’ demanding not just universal franchise, but also `Sawraj’. Here G. G. Ponnambalam seized the opportunity and converted the caste-hegemony politics into a communal confrontation. GGP succeeded in capturing power into his hands, and pushing out the turbaned aristocracy who were forced to align with the Senanayakas to remain a privileged class. It seems to me that Ponnambalam had already sowed the seeds of the `Tamil homeland concept’ in the 1930s, and it was his political demagogy that led to the first Sinhala-Tamil riot in 1939 at Navalapitiya. S.J.V.Chelva split off from the Ponnambalam group and took a more hard-line Tamil nationalist position in 1949 in forming the ITAK. The Tamil language manifesto of the ITAK issued in 1949 is said to have clearly spelled out the Exclusive-Tamil Homeland doctrine already in 1949, long before `Sinhala Only’ came up (i.e, SWRD coming to power). The 1956 revolution was initially an anti-colonial nationalist revolution rather than an anti-Tamil revolution. If SJVC had the political flexibility to actually offer to join the SWRD’s government claiming a place for `Tamil as Swabasha’ (the term used often by SWRD) and proposing a genuine federalist constitution, I believe that SWRD would have agreed to it, and his government would have been strengthened by numbers, ethnically, as well as with capable people. However, the ITAK was already, even in 1949, a quasi-Eelamist party with a strongly militant program.

    It would be historically very valuable for someone to collect the Tamil-language material put out from the Maradana meeting of the ITAK in 1949 and translate them to English for the benefit of many Colombo and Western-based scholars who do not read Tamil.

  2. Reblogged this on Sri Lanka: A Research Project and commented:
    Thuppahi has uploaded a really interesting article written in 1991 by G. H. Peris, about the concept of having a separate Tamil state in Sri Lanka. It’s got a very interesting map showing settlements in Ampara in 1921, and I may well draw on this to compare ‘then’ and ‘now’, in terms of ethnic distributions across the district. I’m going to be collecting so much spatial ethnic data during this fieldwork; it’s going to be fascinating to see what I can do with it all once I’m back in the UK in front of a computer! Arcmap here I come…

  3. Pingback: Tributary Overlordship and Cakravarti Figures in Pre-British Lanka | Thuppahi's Blog

  4. Pingback: Tributary Overlordship And Cakravarti Figures In Pre-British Lanka | Colombo Telegraph

  5. Pingback: Deciphering The Vanniyas; A People Out Of The Box | Colombo Telegraph

  6. Pingback: Deciphering The Vanniyas; A People Out Of The Box

  7. Pingback: Ratnawalli

  8. Pingback: Primordialist Strands in Contemporary Sinhalese Nationalism: Urumaya as Ur | Thuppahi's Blog

  9. Pingback: Onward Tamil Soldiers! Pirapāharan’s Inspiration remains Potent | Thuppahi's Blog

  10. Pingback: Challenging Ratnawalli’s Imperial Sinhala Position | Thuppahi's Blog

  11. Pingback: Facing Chand Wije, An Aggressive Sinhala Chauvinist | Thuppahi's Blog

  12. Pingback: Internet Assassins: Deciphering Their World | Thuppahi's Blog

  13. Pingback: A Comparative Examination of Sinhala, Tamil and Muslim Demographic rends in Sri Lanka | Thuppahi's Blog

  14. Pingback: A Comparative Examination of Sinhala, Tamil and Muslim Demographic Trends in Sri Lanka | Thuppahi's Blog

Leave a Reply