Marakkala Kolahālaya: Mentalities directing the Pogrom of 1915

Michael Roberts

2a-Moorman Tamby =213This article is a reprint of chapter 8 in Roberts, Exploring Confrontation, Reading: Harwood, 1994 which inturn was re-printed with the above title in ROBERTS, CONFRONTATIONS IN SRI LANKA,  Colombo, Vijitha Yapa Publications, 2009 ISBN 978-955-665-035-8

Introduction: Categorical Clarifications1

In the course of 9-10 days in May-June 1915 segments of the Sinhala population drawn from a wide occupational spectrum systematically attacked the property and at times the person of Mohammedan Moors residing in the south western quadrant of the island—a region containing the majority of Sri Lanka’s population at that point of time. This event has since been referred to in Sinhala as the marakkala kolahālaya and in the English rendering as “the 1915 riots” or “the communal riots of 1915.” Because disputes in front of mosques are known to have been one of the reasons for these “riots”, it has been interpreted as a “religious conflict” between Muslims and Buddhists (Nissan & Stirrat 1990: 31-32; Spencer 1990: 5, 8). By itself, this characterisation is misleading and a corrective is in order.

52-Muslim men prepare for worshipThose whom we refer today in Sri Lankan English as “Muslim” were described till about the 1930s as “Mohammedan.” “Mohammedan” (or Muslim) takes its meaning from its context of usage. In juxtaposition with the categories Burgher, Sinhalese, Tamil, Malay, it is an ethnic label. Where aligned in distinction from Hindus, Buddhists and Christians, it is a religious category. It therefore carries a duality of meaning. This dual-sidedness is accentuated by the Sinhala usage. The Sinhala word, marakkala (Moor), is often used to refer to Mohammedans as well. Though there is ambiguity on this point, marakkala does not, unlike the English word “Mohammedan” (Muslim), usually encompass the (Malays). Indeed, the more erudite Sinhala word for Moors is yon (yona) in distinction from javun, javo, ja.Given such dualities, and the ambiguities attached to the everyday usage of such terms, it is not surprising that one finds the terms “Moor” and “Mohammedan” being used interchangeably in the official literature, sometimes in the same document (eg: Tyrrell 1907). Indeed, it is probably because of such ambiguity, and because the category Mohammedan is both an ethnic and a religious label, that the term “community” has become such an important part of the English vocabulary in Asia.

The significance of these distinctions in any study of the 1915 “riots” is indicated by the fact that the tiny community of Malays, an urban people found in small numbers in several of the towns where the Mohammedan Moors (hereafter ­Moors) were assaulted, were not attacked in 1915 (Jayasekera 1970: 294).The Moors therefore, were targeted as a group, an ethnic-cum-religious body of people.

This line of emphasis is confirmed by the composition of the assailants and the language used by their spokesmen. Though in some parts of “the Catholic belt” extending from the northern suburbs of Colombo along the coast to Chilaw the people are said to have been unsympathetic to the object of the “riots,” in other parts the Moors were attacked.2 Church bells were rung at both Mutwal and Kotahena to assemble people to ward-off alleged attacks from the Moors; while the mosque in Kochchikade (north Colombo) was attacked by a crowd led by Catholics. One of the large crowds from beyond the Kelani River which was prevented from crossing into Colombo at the Victoria Bridge contained both Buddhists and Catholics; and they announced that their intention was to help the Sinhalese in the city (Police Inquiry Commission 1916: 27, 32; Jayasekera 1970: 293).

In the early days of the pogrom, one should note, there was something of a grand peur fanned by rumour, with the marakkalayo (Moors, plural form) cast as the dangerous ogre threatening specific and/or generalised targets. The Buddhist Sinhalese who gathered together to defend themselves in this manner do not appear to have done so as Buddhists alone. They were also acting as Sinhalese. Thus, one of the stirrers who addressed the crowd in the small ‘town’ of Gampola on the evening of the 30th May dwelt on the need for unity among the Sinhalese (Sendanayake 1916). It was this body of people who attacked the Moor properties later on that night. Likewise, at a gathering of religious associations at Attanagalla on the 1st June, the assembly was told that “the Tambies are insulting our nationality and our religion” (below & fn. 57). Thus, those Sinhalese who were moved to attack and pillage the Moors and their properties, whether Protestant (e.g. at Moratuwa), Catholic and Buddhist, were doing so as Sinhalese, while the Buddhists were doubly motivated.

This interpretive clarification must be supported by a conscious act which revises the standard label attached to these events: namely, the term “riots”. This term has long been part of the British legal lexicon and refers to sustained assaults on a segment of the population as well as minor affrays over a few hours involving a handful of people. Numerous scholars, including myself, have slavishly followed this crude terminology in describing events, such as those which occurred in 1915, 1958, 1977 and 1983, which demand a qualitative distinction—that is, a conceptual separation from minor clashes.3

One must follow the lead taken by those writers4 who have described the events of July 1983 as a “pogrom” and apply this label to the clashes in 1915. The innovative use of this label to characterise the attacks on the Tamils in southern Sri Lanka in 1983 has been questioned by K. M. de Silva. His bible is the Oxford Dictionary, which defines pogrom as “an organised massacre, the annihilation of any body or class, especially of Jews; an organised persecution or extermination of an ethnic group, especially of Jews; and especially in Tsarist Russia” (quoted by de Silva). De Silva, therefore, sees the term as specific to Central and Eastern Europe and as involving “officially sanctioned and officially directed persecution or extermination of an ethnic group, a despised group confined to a . . . ghetto” (1988:87).

The Oxford Dictionary (pre 1990s) was “compiled effectively from the 1880s to the 1920s” (O’Connor 1989: 53). De Silva seeks to freeze us in that period of time. He is blissfully unaware that the English word “pogrom” derives from old Russian: it is a word which simply meant “destruction.” Over time this meaning was lost and “the word came specifically to denote the destruction of Jewish life and property” (Kochan 1957: 12).

However, my usage of the term here is not informed by considerations of etymological primordiality. It is situated self-consciously in the 1990s, with all its retrospective advantages. It moves metaphorically beyond sterile etymological positions to extend the term to all contexts in which a dominant segment of a population systematically assails another segment in their midst,5 the systematicity being one of pattern and effect and not implying a planned conspiracy as a defining feature. Likewise, in this usage, the participation of state agents in such pogroms is not a definitional requirement, though they may well be involved in greater or lesser measure.

Thus defined, and standing in the 1990s, a pogrom is treated as something short of the genocidal programmes instituted by the Nazis from 1938-45 and by the Khmer Rouge in the 1970s. The Nazi programme is widely known among the Jews today as “the Shoah,” meaning “catastrophe” in Hebrew. It is seen as a qualitatively distinct phenomenon, something “beyond description.”6

The metaphorical use of the term “pogrom” to identify the so-called “riots” of 1915, therefore, is informed by the knowledge that it is a loaded epithet. Such an emphasis on my part is a mark of my own awakening, one induced by the events in Sri Lanka in the 1980s and a conference on ethnic violence in South Asia held at Kathmandu in early 1987 (see Das 1990). As such, it is a political act of consciousness-raising. Historians are as much in need of this nuance-in-perspective as their readers.

The Pogrom of 1915

The marakkala kolahālaya (disturbances concerning the Moors) began on the night of the 28th May 1915, or, rather in the early hours of 29th May, opposite a mosque at Castle Street, Kandy. During the next nine days or so the clashes and assaults spread through the Central, North Western, Western, Southern and Sabaragamuwa Provinces (note Map 1); and at one point, on 2nd June, affrays were reported to be occurring simultaneously at 116 “centres.”7 On the odd occasion during this violent conflict, Moors shot or killed Sinhalese; and there were massed struggles in one or two spots where the Moors had significant numbers.8 For the most part, however, the pogrom involved an assault on the person and property of Mohammedan Moors by Sinhalese (the latter being joined by some Tamils in localities, for example at Kandy and at Wattegama9)—with the focus in many localities being directed against Moor properties.

There were no attacks on government properties. Barring the rare instance, no European seems to have been the target of attack qua European. Indeed, during the heavy-handed restoration of order by state agencies, Europeans were sometimes treated with great respect; and in the Province of Uva (where there were hardly any “riots”), in separate actions, two European planters seem to have single-handedly squashed any prospects of trouble by their intimidating presence in the bazaar.10

Large crowds were involved in the attacks on the Moors: mobs of over a thousand were reported at Matale, Wattegama, Kadugannawa, Gampola, Rambukkana, Panadura, Godapitiya, and Akuressa; while the crowd at Gewilipitiya-Aranayake was variously estimated at 800 to 4,000.11 At Gampola and Panadura the authorities (both police and British civil servants) were forced to retreat, and in Gampola the missiles directed at them included bottles filled with sand.12 On at least five occasions the crowds, by sheer weight of intimidation or by storming (at Pasyala) a police station, effected a release of individuals who had been arrested; while a “village sergeant” was killed at Kottawa when attempting to quell a riot.13

According to the official estimates, which must be taken as approximate, the Moors suffered the following casualties and damages:14




Houses and boutiques looted




Houses and boutiques fired




Mosques burnt 




Mosques otherwise damaged



Bland statistics cannot convey the effect of these attacks, the sense of terror that coursed through the threatened bodies of Moor people (J. G. Fraser in Police Inquiry Commission 1916: 11. Cf. Kanapathipillai 1990; Das 1990b; and Srinivasan 1990). Some measure of these affective effects can be gleaned by considering three illustrative moments:

(1) At Trincomalee Street, Kandy on the 30th May evening a Moor stabbed several Sinhalese (killing one), thereby provoking a rounded assault on Moors and Moor houses in the vicinity from a raging mob of Sinhalese. When Vaughan, the G.A., Central Province, reached the scene, he recorded the following impression: “I have never seen a crowd so excited—men were crying from excitement, and appeared to be demented.”15

(2) The scene that greeted Browning (A.G.A. Matara) when he reached Godapitiya near Akuressa in the Southern Province on the 4th June:

On arrival I found a terrible scene of ruin, but no sign of the rioters. . . The village was in flames. The mosque, one of the finest in the district, had evidently been blown up with dynamite, while the surrounding coconut trees had been felled so as to fall on the roof. Parts of it were on fire. There was one dead Moorman in the compound, while 8 others (two of whom died on the way to Hospital) were seriously injured. It was most pitiful and one’s sense of helplessness was overwhelming.16

(3) The combination of trauma and anger which permeates the report on the disturbances provided by the Mohammedan Member of the Legislative Council, W.M. Abdul Rahiman; especially the moment when he was called upon to refer to the desecration of some mosques:

With regard to Mosques I would ask Your Excellency to order all such buildings as are badly damaged be rebuilt at the expense of the rioters, and those slightly damaged be repaired with the addition of a new cement floor, and all the walls replastered to a height of 7 feet. Your Excellency may already be aware that the sacred edifices have been polluted in several ways, some of which I can hardly mention (Abdul Rahiman 1915).

Such illustrative moments call attention to the importance of the victim’s voices in any study of a “riot.” Nevertheless, previous studies of these “riots,” including mine, have paid limited attention to the experience of the Moors. The most striking of these omissions has been the failure to dwell on the desecration of the mosques; and particularly Abdul Rahiman’s horrified and apoplectic letter on this point—a letter which is in the CO 54/482 series which virtually every author on the subject has consulted.17 The reasons for this appear to be threefold. In the first place, there is conclusive evidence to demonstrate that some Moors laid false charges against specific Sinhala notables in the aftermath of the riots.18 Secondly, the nationalist climate of opinion in the 1960s seems to have insidiously shaped the selective decoding of data. And, thirdly, this selectivity seems to have been directed by the empiricist quest for “facts,” so that “experience” and “suffering” were not in the scholarly agenda.

In trying to restore the balance in this essay, let me note that the recovery of the victims’ experiences does not preclude the careful sifting out of concoction or exaggeration from the irreducible happenings. Indeed, some of the allegations presented by Abdul Rahiman were disputed in their specifics by British district officers.19 However, other details receive overwhelming corroboration from diverse sources. There is little doubt that a number of mosques were dynamited, fired or otherwise damaged. At least two mosques, at Maningala and Kendagamuwa, were desecrated.20 The acts of defilement in such circumstances would have drawn upon the ‘standard tools’: offal, faeces, blood and pork. That Abdul Rahiman himself was rendered revengeful by his contemplation of the events is not in doubt. It is unlikely that he was the only Moor affected in this manner by the experiences of May-June 1915 (see also Macan Markar 1915 and J. G. Fraser in Police Inquiry Commission 1916: 11).

Given such effects, given the scale and breadth of the attacks on the Moors and given the wide occupational spectrum from which the assailants were drawn (below: 128-129), the so-called “riots of 1915” must be deemed a pogrom.

The Historians and their Themes

The historiography on the pogrom has tended to dwell upon three issues. Firstly, the relative importance of religious and economic factors in directing the activities of those Sinhalese who participated in the “riots” as stirrers or assailants. Secondly, the degree to which there was pre-planning and conspiracy among those Sinhalese involved in the mayhem and assault. Thirdly, the consequences of the events, and the manner in which the extra-legal measures taken by the British in restoring order influenced the on-going nationalist movement of the “Ceylonese” opposed to British imperialism.

This article is concerned with the first of these issues, the causes of the pogrom. Its interest is not only ethnographic. It ventures upon revisionary inter­pretations in ways which scrutinise the assumptions that have guided much of the historiography. In particular it explores the limitations of the empiricist rationality which has dominated the literature on the subject and which has prevented scholars from entertaining the idea that the pogrom was also an elliptical attack on the British. Included within this historiography is one of my own essays, written in 1970 and widely available in mimeographed form as a Ceylon Studies Seminar paper before it was republished (Roberts 1981). This article, therefore, is also an auto­-critique.

A summary review of the literature is an obvious point of departure for this endeavour. One of the prominent lines of emphasis has been informed by Marxist orthodoxy. The work by Kumari Jayawardena and P. V. J. Jayasekera embody this emphasis. Jayawardena’s initial work emerged from her study of the early labour movement and concentrated solely on the “riots” in Colombo. She is quite unequi­vocal in concluding that the rioting in Colombo “had hardly any religious motives” (1970: 229 & 224). The argument is that the rise in the cost of essential goods because of the war led to resentment against profiteering by traders, with the Moor traders becoming a convenient scapegoat. In a subsequent essay Jayawardena extends her economistic interpretation beyond Colombo to a generalised picture of Sinhala traders promoting attacks on their Moor rivals during a wartime situation when shortages and price rises had nourished popular antipathy towards aliens (1984: 12. 23, 133).

The latter work is remarkable for the manner in which it bypasses data in the studies by Jayasekera (1970) and Roberts (1981) which would undermine the thesis. It renders the Sinhala merchants into enormously powerful elites and the mass of the assailants into pliant tools. As Rogers has indicated, moreover, even in Colombo, contrary to Jayawardena’s understanding, five mosques were damaged, Moors were killed and “members of the crowd justified their actions in religious terms” (1987: 190). To discount the influence of Sinhala Buddhist ideology altogether,21 at all levels of the class order, is grossly misleading. It is Marxism in its most vulgar form.

Jayasekera’s dissertation is a veritable mine of information. He is attentive to the “religio-economic conflict” that induced the Sinhala newspapers to mount “a vicious campaign against all ‘foreign’ businessmen,” especially the “Muslims;” and calls attention to the desire among the people to protect their traditional rites (1970: 106 & 338). However, he attaches secondary weight to the force of “communal and cultural bonds.” In this view the “disturbances were essentially the outcome of numerous accumulated grievances, commercialised rivalries and immediate difficulties caused by the rising cost of living;” while the religious conflict was an expression of “basically economic grievances” (1970: 321 & 327, emphasis added). In other words, ideology and religion are epiphenomenal, superstructures not foundations.

Jayawardena and Jayasekera thereby betray their attachment to a line of reasoning that can also be described as utilitarian instrumentalism, or what Sahlins would call “material rationality.”22 In Jayasekera’s study this leaning is linked to the standard thesis of Durkheimian sociology: a picture of social disorganisation in the south western and central districts of British Ceylon. The expansion of a market economy and the growth of plantations are said to have “hastened the disintegration of the rural economy.” In this circumstance, the hostility of the Sinhala villagers was directed against the ubiquitous Moor traders (1970: 299ff., quotation p. 310). Unlike Jayawardena’s thesis, the underclasses are attributed some agential initiative, albeit an initiative that is read in instrumental ways.

Instrumentalist interpretations do not require a Marxist inspiration. The study by George Scott Jr. is suggestive of the material rationality behind American interest group theory. He allows that religion was implicated in the 1915 conflict as a secondary, “epiphenomenal” factor; but concludes that these “riots” were “primarily motivated by economic conflict” (1989: 19). The reasoning is that the procession disputes were “the immediate cause” and “the vehicle” of justification. Being the immediate cause they were, ipso facto, of secondary importance. Eco­nomic competition, in contrast, had been long-standing and therefore provided “the ultimate cause.” In effect, the duration of a factor determines analytical weightage. The conviction within this casuistry is provided by the self-confirming adrenalin of an utilitarian epistemology. Even within these terms, Scott’s ethno­graphy is both second-hand23 and flawed: in “The imperialism of silence” I have demonstrated conclusively that the localised procession disputes between Moors and Buddhist Sinhalese date from 1899, if not earlier, while that at Gampola originated in 1907 (Roberts 1994b).

John Rogers does not fall into the trap of mono-causal determination and cautiously attributes significance to both religious and economic factors in the genesis of the “riots” (1987: 190-91, 199). This is the position taken by Blackton (1970), Fernando (1969); and my earlier work (1981 b: 89). As I saw matters then, the Governor, Chalmers, put matters in a nutshell in stating that the “riots” were due to “a combination of creed and purse” (Command Paper 8167: 35).

Though Ameer Ali and Kannangara also allow for the influence of economic factors, their emphasis is on the religious issue, namely, the “fears and resentments aroused by the procession disputes”—to use Kannangara’s words (1984: 148 and 152-53; Ameer Ali 1981: 4ff.). Informed by the new material on sabda pujā and the festering sores on this issue, I would now endorse their weightage. The Sinhala Buddhist stirrers and assailants were, in the main, seeking to avenge an “insult” to their religion (see below). The religious confrontation and the verdict of the Supreme Court on the Gampola Perahara Case in early February 1915 were critical factors in leading to the outbreak of violent hostility in mid 1915— rather than the years before or in January 1915.

I argue this position in association with a critical caveat. The Moors were said to be “insulting our nationality” as well as “our religion,” (below). This is but one piece of evidence for the more general proposition that the distinctions between “religion,” “economics” and “politics” are based upon a modern, secular Western epistemology; and are therefore problematic categories for the Indian world (cf. Devane 1915). Dharmasāstra and Arthasāstra implicate each other, religion and politics interpenetrate each other in the vamsa traditions. The writings of the Sinhala ideologues from the late nineteenth century onwards reveal a deep concern with the condition of the Sinhala jātiya (race, nation). In this view the Sinhalese were in a weakened state because of the expan­sion of Christianity, the spread of a Western life style and the commercial activities of intruding foreigners, whether British, Coast Moors, Moors, Chetties or Kocci (Malayalam-speakers from the Cochin region).24 In such writings the explicit economic nationalism was part of a broad programme to rehabilitate the political, strength and cultural purity of the Sinhala people (see below). Whether one dissects the sayings of a Piyadāsa Sirisena, a Gnānāloka Thera, a John de Silva or an Anagārika Dharmapala, their outpourings—and they were deeply felt outpourings—must be read as a totality, with the distinct strands merging into a syncretist whole in the hearts and minds of men possessed.

Kannangara’s study is also significant for the emphasis placed on two other dimensions. Seizing on evidence that specific depressed caste villages had been in the vanguard of the assaults on the Moors in three or four localities, he contends that the resentments among castes seeking to break free from the hierarchical constraints of yesteryear were an important factor in promoting the “riots”. Secondly, he attaches significance to the social disorganisation attending commer­cialisation and the movement of people to the bazaar towns (1984: 148,152-53,131). Where Jayasekera sees disintegration as a generalised feature of the south western districts, Kannangara narrows this theory to the bazaar towns. Migration and urb­anisation, in other words, spell greater scope for aggression.

In the manner favoured by orthodox historians, Kannangara is not self­-conscious about the theoretical underpinnings supporting such an argument. But it must be read as a particular instance of relative deprivation theories of riot and rebellion (e.g. Gurr 1970; Adas 1979). Such theories, as we know, are informed by structural functionalist interpretations of Durkheim and by the psychological twist attached to this interpretation by Malinowski. Persons disturbed from their place in society by drastic change are seen to have greater potential for violence as they seek outlets for their frustration (deprivation) as a compensatory mechanism.

Such theories remain immensely popular today, despite penetrating ques­tioning (e.g. Taussig 1980: 14ff.; and Skocpol 1979: 34, 9-10 & 25). The thesis has even been proposed as one explanation for the anti-Tamil pogroms of recent years (Tambiah 1986: chap. 4; Gombrich & Obeyesekere 1988: 100-01 & 67-70). In Gombrich and Obeyesekere’s reading, paradoxically, the disintegration of the “traditional village community” is a process that has occurred in the post-war (1940) era rather than the period of British supremacy.

To address the limitations of this thesis would require an essay in its own right.25 For the moment, I would endorse Rogers’ questioning of Kannangara’s version of the social disorganisation theory on the ground that in 1915 the market towns of the interior were composed of small populations and “the imper­sonal forces of urbanisation had not proceeded very far” (1987: 199). The statistical data in Tables 1 and 2 bear this out. Even more critically, the ethnographic evidence undermines Kannangara’s emphasis. The data on the assailants in some localities in Kegalle District (Roberts 1981: 117-120) and the range of personnel engaged in the stirring and assaulting (see details below) simply destroys the disintegration thesis. It is manifest that the attacks on the Moors occurred in distinctly rural areas and isolated spots as well as the bazaar towns; and that many Govigama villages were party to the pogrom.26 In the vicinity of Gampola, for instance, on the day after the “riots” had broken out in Kandy some 25 miles away, tom-tom beaters went round the surrounding villages advertising a pinkama (a merit-making, religious procession) in the town that evening. Villagers were among the crowd that assembled in Gampola that evening of the 29th May to hear stirring speeches as Sinhala activists girded their loins to teach the Moors a lesson.27

The Assailants in their Socio-Political Context

My ongoing investigations28 permit me to identify the occupational background of those Sinhalese who were activists in the pogrom, whether as stirrers and/or assailants and/or looters. These, in a greater or lesser measure, included:

       landed proprietors and superintendents of plantations;

       businessmen (mudalālis)

       boutique keepers and small mudalālis

       headmen of various sorts, generally middle and lower echelon;



       other government functionaries;


       plumbago (graphite) mineworkers;


       smallholder cultivators;29

       road and rail repair personnel;

       estate conductors and estate “coolies”;


       railway workers in Colombo;

       stevedores in Colombo;

       such amorphous categories as “villagers” and “workers”.

Needless to say, there was regional variation in the pattern of participation. The survey here cannot be fully attentive to such diversity.

In order to comprehend the implication of the categories delineated above, one needs some understanding of the socio-economic backdrop, preferably one that is elucidated further by a conceptual framework. Sri Lanka as it stood in 1915 had been a British colony for some 120 years. The British moved fairly quickly to establish the institutional framework of capitalism: through statutory and administrative provisions which secured property rights, assisted conveyancing and the rights of contract, including those between master and servant. A particularly significant measure in this regard was the abolition of rajakāriya (corvee labour) 1832. They also engineered a revolution in communications which stitched together the island, especially its south western cum central quadrant (see Map 2).

These developments encouraged the rapid growth of coffee plantations in the plateaus and hills in the south centre of the island from about the year 1837; while an enormous expansion of coconut plantations commenced in the Western and North Western Provinces from about the 1850s. In combination, all these developments accelerated the monetisation and the commoditisation of the economy. The commodity form was not unknown in pre-British times and money was in circulation even in the Kingdom of Kandy (1590s-1815) in the interior, but numerous institutional practices, especially rajakāriya, had restricted the commoditisation of all the factors of production. Both the spatial extension of the commodity form and the depth of capitalist practice was perhaps taken several notches further when tea replaced coffee as the principal export crop from the late 1880s and when rubber plantations commenced in a big way, especially in the Kelani Valley districts and the mid-country foothills of the south west, from around 1900. The mining of graphite, usually in open-cast mines, also provided an additional boost to capitalist growth from the 1870s through to the 1910s.

The benefit from these economic developments did not flow only to the British planters and merchants who invested capital in the island. A non-British bourgeoisie emerged, including Asian migrants such as the Parsees, Memons, Borahs, Indian Moors, Chettiyars and Sindhis as well as indigenous capitalists from the Sinhalese, Moor, Tamil and Burgher communities. The liquor industry, transport contracts, contracts to provide labour and raw materials, coopering, graphite mining, the gem grade and the import-export trade were among the most profitable lines of investment which spawned and reproduced the indigenous capitalist class. Their investments in such lines of distribution and/or production were often comple­mented by investments in plantations: coconut and rubber in particular, but also coffee and tea in some instances (Roberts 1979b & 1982: 98-130, 304-19).

The vast majority of the land under coconut plantations in British Ceylon (perhaps 400,000 acres), it should be stressed, was held by Ceylonese, not British or Asian migrants. Among the Ceylonese, the Sinhalese bourgeoisie easily sur­passed the rest, both in numbers and in wealth held. In some individual instances the wealth of families was enormous. One can speak metaphorically of merchant princes and plantation kings.

Invariably, those who emerged from the ranks of the population to become arriviste merchant princes had to consolidate their social advance by ensuring that their children received a good education in English. Fluency in English and the trappings of Westernisation were critical elements in the social scene by the late nineteenth century. The British government required “natives” for the middle and lower echelons of the government service. They had promoted an English education with the aid of the missionary schools. In consequence, education provided another avenue of social mobility. Doctors, lawyers, surveyors and high-echelon government servants quickly secured a station at the apex of the indigenous social hierarchy. In this scheme of superiority:subordination, property ownership was not the­ distinguishing mark. Cultured ways, a fluency in English and a Westernised life style were modalities of domination. Those marked out by these attributes can be described as a middle class, with the indigenous aristocracy of yesteryear being gradually absorbed into its folds or occasionally being pushed down into the ranks of the “others.” These others were sometimes described as “hoi polloi.”

In this scheme of things, it was possible to be a member of the middle class without being a capitalist, that is, a property owner who deployed labour in distri­bution or production. Contrawise, there were many indigenous capitalists who had limited fluency in English and few of the valued Western social graces. They were, therefore, not part of the middle class as I have defined the term.

In effect, this analytical framework distinguishes the indigenous population along two dimensions: that of the bourgeoisie: proletariat on the one hand and that of the middle class: hoi polloi on the other. The bourgeoisie and middle class, therefore, were (are) overlapping, but not synonymous categories. So too were (are) the proletariat and hoi polloi. The two-dimensional scheme must be extended further by providing space for marginal and in-between categories of people who inhabited the space between the dominant elements (bourgeoisie and/or middle class) and those dominated. These were the petty bourgeoisie and other inter­mediary elements. In opposition to the Marxist vision of the petty bourgeoisie as a “transition class” (Marx 1970: 122 & 173-75), my conceptual scheme emphasises their strength and continued reproduction as an intercalary social force of con­siderable significance for Sri Lankan politics (also see Roberts 1974, 1989a).

The capacities of these intermediary classes were revealed in the temperance agitation of the year 1904, agitation which lasted from April to November only, and which was something of a mushroom development. The movement was initiated by local elites in the Southern Province. It was not till June 1904 that the campaign was hijacked by members of the middle class (rather like the pattern identified in some instances in India by the Subaltern Studies). This partial takeover was literally invited by the intermediary classes in so far as they asked middle class notables from Galle or Colombo to address their gatherings. The popular support for this campaign was considerable and at one stage there were about 600 temperance associations, “with perhaps 200,000 members or one-fifth of the adult male population of the island” (Rogers 1989: 327).

The temperance movement of the 1910s was also a popular affair. It also had a solid organisational grid in the south western and central districts, being “minutely organised” in the words of a British civil servant (Bertram 1920). The command centre was the Colombo Total Abstinence Central Union, a body in which the middle class and/or bourgeoisie had a controlling interest. By 1914-15 they had speakers fanning out into the rural areas on a regular basis, addressing the meetings of local associations. They usually travelled by car and were often greeted with great pomp and pageantry, the receptions sometimes outdoing those accorded to British district officers on tour.30 However, the popularity of this campaign could not have been sustained without the active support of the intermediary classes at the local level.          

The expansion of the commodity economy in British times also involved the influx of Low-Country Sinhalese traders, carters, workers, government functionaries and others to the interior Kandyan Sinhalese districts. This influx began at an early date (the 1820s) and some Low-Country Sinhalese became settlers in these new localities. The transformative character of the market society in these Kandyan areas is revealed by the composition of the population in several of the market towns in the region where the Kandyan population was as low as 12 to 35 per cent in 1911 (see Table 2).

This background permits one to place in sharper perspective the information on the various socio-occupational categories who were active participants in the pogrom. It is evident that all three tiers in the status/power order were implicated in the assault on the Moors. However, the number of activists from the middle class and/or bourgeoisie about whom one can entertain suspicions of participation, after sifting through the allegations and tenuous evidence, is relatively small. This said, there are indications that several members of the middle class, including a Police Magistrate, other government functionaries and police officers, were in sympathy with the punishments being meted out to the Moors (Rogers 1987: 191; Brayne 1915: 265-69).

The number of activists from the margins of the middle class and from the intermediary elites is considerable. In virtually every locality one can identify a number of individuals from categories c, d, e, f, g and h who were active participants in the mobilisation against the Moors, often as “ringleaders.” Thus, small businessmen and boutique-keepers appear to have been activists in the pogrom in virtually every locality.31 So, too, were the middle-to-lower echelon headmen, both in the Kandyan and the Low-Country areas (with the North Western Province being exceptional).32 Other government functionaries, such as registrars, police officers, deputy coroners, fiscals and peons, were also involved in several localities. In a few instances, the evidence indicts “sons of” a specific headman or registrar.33 Notaries and school teachers, usually those who operated in Sinhala, were among the leading activists in a few localities; as were clerks, especially in Colombo; while bhikkhus were identified as stirrers on at least three occasions.34 Further investigations are certain to extend the pool of activists drawn from all these categories.

In several localities in the mid-country belt of foothills the graphite mine workers and carters were identified by officials as being among the most energetic activists. In the Kandyan regions many of these proletarian workers were said to be Low-Country Sinhalese. Likewise, in Sabaragamuwa and the North Western Province the kankānamas (plantation foremen) and “estate coolies” who were identified as activists were said to be of Low-Country origin.35

The high preponderance of Low-Country Sinhalese, whether proletarian, hoi polloi, intermediary elite, bourgeoisie or middle class, among the ringleaders and other accused in the Kandyan districts led many officials to the contention that the Kandyan villagers were catspaws in the hands of designing individuals from the mobile, modern order.36 I follow Jayasekera in questioning this thesis (1970: 293-96). It is more a product of the romanticist perception of the villager as a simple innocent, a pastoral vision that was widespread among British officials alienated by the challenges of the upstart middle class, than an accurate depiction of the villager. In other words, I surmise that the villager (Kandyan or otherwise) was not the rustic simpleton that he is held out to be. “The villager” (itself a working abstraction coined by officials) was as alive as anyone else to the opportunity afforded by the generalised assault on the Moors to accumulate loot. This is evidenced from the pillage of isolated boutiques in the backwoods of Ratnapura district.37 Secondly, the British officials at the same time pinpointed specific villages in rural areas which were “notorious” or “rowdy.”38 This data is amplified by John Rogers’ researches on cattle stealing and criminality in British Ceylon: this study reveals pockets and networks of crime across a wide range of rural localities (1987). The pastoral myth has to be laid to rest.

The villager, nevertheless, was not everyman in the modern, Western, secular, rational sense. Several Sinhalese, whether villagers, intermediary elites, or middle class, engaged in rites and practices which were moulded by Sinhala Buddhist cosmological notions. Given such modalities of being in the world, they were extremely disturbed by the British policy against the sabda pujā of traditional Buddhist processions in front of mosques. Likewise, their notions of astro-chronological junctions and omens from the gods may have been activated by the contingencies that preceded May-June 1915, such as the floods that hit the town of Gampola (see below: 141). At this distance in time it is not feasible even to speculate about the proportion of the population in the Kandyan and other areas who would have been susceptible to such modalities of reasoning.      

But we do know that by 1915 there were a significant number of Sinhalese who had received a range of messages depicting the deplorable conditions of the Sinhala jātiya (race, nation) in their time. These messages had been conveyed from the 1870s, if not earlier, by written, oral and theatrical means. By the early 1900s, a number of newspapers (usually bi-weekly), many pamphlets, the occasional novel, associational gatherings, the sermons (bana) of militant monks, the gramophone records of John de Silva’s plays and the rhetoric of itinerant activists had ensured that the thinking of the Sinhala ideologues had been disseminated along the thoroughfares and by­ways in the south western and central parts of the island. The ability of the propagandists to reach a wide audience is attested by their choice of pilgrim sites in peak season as locations for their work: in March 1915 a “street preacher” was reported to be active in the Adam’s Peak area by local Moors who found his rhetoric “offensive” and “scurrilous”; and he had in his possession pamphlets in verse form printed at Dharmapala’s printing press (Thaine 1915: 251). The latter, kavi kola (broadsides in verse), have been an influential mode of political dissemination in the twentieth century because they are memorisable and often satirical.

The Degradation of the Sinhalese : a Profound Sense of Doom

There is a substantial body of literature that has elaborated upon the nationalist movement in Sri Lanka dating from the mid-nineteenth century and revealed its multi-faceted nature (e.g. Amunugama 1978 & 1979; Wickremeratne 1969; Roberts 1979c and 1989a; Rogers 1987). Our interest here is in the overlapping, but not synonymous, currents of Sinhala Buddhist and Sinhala nationalism. The Sinhala activists were extremely disturbed by their colonial situation and the “inferiority complex” — in Sirisena’s words – that this condition had implanted in the attitudes of their people. The Sinhala degeneration was regarded as both cultural and economic. The reliance on Western imports was adversely remarked upon. The widespread adoption of Westernised lifeways and the diffusion of Christianity among the Sinhala people were seen as marks of their degeneration as well as instruments which furthered this process—undermining their gunadharma (religious virtues), kulacaritra (traditional customs) and bhāshava (language).39

The picture of decline painted by the Sinhala ideologies was a dire one, reminiscent of the stage of history known as the kāli yugaya (epoch of evil and chaos). The scenario was depicted with vehemence, in “what appears to outsiders [as] ‘exaggerated’ language” (personal communication from John Rogers). As Rogers recognises, his characterization is an outsider’s etic reading that is probably removed from the felt experiences of the Sinhala spokesmen. Their earnestness of belief is a message in itself. As Rosaldo observes, “social critics remain grounded in the local cultures to which they direct their exhortations and invectives;” while their “narratives about their own conduct merit attention as forms of social analysis” (1989: 182 & 143).

It was from deep-felt concerns that many a Sinhala ideologue called on the Sinhalese to arrest their decline by developing their economic strength. In short, by becoming a successful entrepreneur a Sinhala person was rendering a national service. Bourgeoisie ideals were rendered into nationalist goals.40

The tone of the articles, pamphlets, novels and plays which exhorted the Sinhalese varied from the didactic abstruse to the biting satire of the zealot. An index of the convictions that drove these ideologues is provided by the consistency with which they birched the Sinhalese themselves—indeed to such a degree that one can speak of self-flagellation. Perhaps the sharpest diatribes were directed against those Sinhalese who were aping the Westerner. In Piyadāsa Sirisena’s writings such Sinhalese are even rendered into a distinct ethnic category: the samkara (mixed) and/or the tuppahi (low and mixed).41

They were not the only target for diatribe. Abuse was also heaped on the ultimate source of threat, the paradēsakkāra (low and vile foreigners). These foreigners included the British, the kocci (Malayālis), the hamba (Indian Moors), the marakkala (all Moors), the hetti (Chettiyars), the javo (Malays), the bhai (Borahs), and the para demala (low and vile Tamils).42 In one of Anagārika Dharmapala’s essays in 1911 there is even a diatribe directed against the kocci demalā.43 As Obeyesekere notes, in the thinking of such zealots as Anagārika Dharmapala there was no place in Sri Lanka for anyone other than the Sinhalese (1979: 304). It is noticeable that in many of his writings the term “Ceylonese” shades off imperceptibly and insidiously into “Sinhalese” (Dharmapala 1965: 501-­44).

The Moors were so widely dispersed in all localities in the south western quadrant that they were one of the favourite targets of criticism among those espousing economic nationalism. In the typical metaphor both the hambayā (Indian Moor) and the marakkalayā (Moor) was depicted as a bloodsucker or viper who was debilitating the Sinhala people.44

At the moment of confrontation in May-June 1915, such lines of economic hostility were clearly bolstered by economic motives of a specific instrumental nature: on the one hand, the attacks on their immediate competitors by Sinhala merchants and boutique keepers, and, on the other, the desire for loot among so many. In sheer quantitative terms, at the individual level, it is likely that the pillaging motive was the most widespread of causes.

Our focus, however, should be qualitative rather than quantitative. There were a number of occasions when the goods of the Moors were piled up and burnt,45 while in several localities, especially in the Western Province, the trees and crops on the small plantation properties held by Moors were systematically devastated.46 Significantly, on several occasions, the incendiary action around the bonfire of piled up goods belonging to Moors was accompanied by cries of religious exhilaration: “sādhu! sādhu” (Jayasekera 1970: 288; evidence in Eastern 128: 244). Here, then, was a discursive practise of symbolic punishment which, more than any other, generated a sense of communion among those assembled—marking a situation where each individual in the crowd could be said to lose his/herself in the wholeness of a worthy action and become a particle in a wider configuration, a whole.

These symbolic moments are clues to the ways in which the economic and religio-cultural threat from the Moors shaded into one another in the thinking of several militants. This said, the fact remains that the focus on the Moors as a set of bloodsuckers was a refrain which had been harped upon for several decades before 1915— accompanied, more often than not, by suggestions that an economic boycott of Moor shops should be organised.47 The very fact that such clarion calls were raised intermittently from the 1880s onwards suggests that boycotts never got off the ground or were rarely attempted. For the same reason, too, the economic factor alone can hardly explain why the attacks on the Moors took place in mid 1915 and not at some other moment. The Gampola Perahara Case and the immediate repercussions of the Supreme Court’s verdict, namely, further Moor demands in the town of Kandy, emerge as a critical catalyst.

In order to comprehend the force of this issue one has to modify one’s secularised Western code of rationality. One needs to appreciate the Sinhala Buddhist world view. Then, within this world view one has to concentrate on what Valentine Daniel (1990: 227-30) describes as the mythic attitude of being in the world in participatory awe with and through the gods—gods who are, like the yakku (demons), perētayo (ancestor spirits), jealous neighbours and sorcerers, truly part of the everyday world.

Noise as Cultural Struggle

From the year 1898 the British authorities chose to tighten the implementation of the Police Ordinance. This meant restrictions on the traditional pursuit of sabda pūjā in a number of urban centres (Roberts, 1994b: 160ff). A number of localised confrontations between Moors and Buddhists developed (Roberts, 1994b: 160ff; Jayasekera 1970: 116; Blackton 1970: 238; Woolf 1962: 241-42). By the 1910s several of these were taking place in the regions of the former Kingdom of Kandy. Here, successive G.A.’s in the Central Province, namely, Lewis, Saxton and Vaughan, adhered to the principles of the Police Ordinance in a rigid manner. In rejecting the appeals of the Buddhist authorities of the Wallahagoda Dēvālaya of Gampola in 1912, Saxton noted:

I personally do not see any hardship to the Buddhists in their procession being required to stop their tom-toms for a short distance, in order to show consideration for the religious feeling of other people.48

Here, then, in Saxton’s statement we have the voice of the Enlightened and the vocabulary of practical reason rendering the position of Wallahagoda Buddhists into something unreasonable. In this view, enlightened tolerance lay with the British. Not surprisingly, the position taken by the Buddhist authorities associated with the Wallahagoda Dēvālaya was in diametric opposition to that adopted by Saxton and his government. They stated quite explicitly that they were “deeply hurt” by the “obstruction” placed in the path of their annual Äsala Perahära. Insofar as one letter refers to the “redress” of their grievances, it is evident that they considered the state policy to be quite unjust. Not only is their discourse permeated by an aggrieved tone, it also reveals the different world they occupied and a desperate effort to come to terms with the imposed world view of the British without losing their own anchors. Their letters also display the force of tradition in their thinking. To them, time was a measure of value: so that the phrases “ancient times” and “time immemorial” repeatedly buttress their arguments (Nugawela 1912a, b and c).

The depth of the Kandyan Buddhist sentiments is seen in correspondence in late 1915 and 1916, after the pogrom, when the aristocrats in charge of the pitisara dēvāla49 refused to subject themselves to the potentially restrictive conditions of a procession license even though their processions did not, at that stage in time, pass a mosque. They observed that the requirement for a license at that point of time would have serious implications should a new mosque arise along the traditional route at some future stage. They deemed the government stipulation a “hardship” (amāruvak) and decided to conduct their water-cutting rites in ponds on their temple lands rather than submit to this condition.50

Likewise, earlier in 1912, the spokesmen for the Wallahagoda Dēvālaya dug their heels in and refused to abide by the demands of the British government. They initially decided to take the Äsala Perahära on another route. Faced with protests from their workers and supporters—a highly significant development this—they changed their minds on this point. They decided not to have any perahära; and filed a legal case against the Crown on the 23rd July 1913.51 This was the origin of the famous Gampola Perahära Case.

Underlying the position of the Wallahagoda Buddhists, therefore, was a different world view, the mythic mentality which Val Daniel has outlined (1990: 227-30; also Kapferer 1988). In this perception, past actualities become contemporaneous and the power of the gods is a means of ordering time, fertility and prosperity as well as the various elements (mahabhuta) that constitute the world of being. In this view, too, religious rites are pragmatic acts that order the everyday world. Given such perspectives, the failure to hold the Äsala Perahära was a frightening event—for the god of Wallahagoda Dēvālaya, god Kataragama, was an exacting master. And as the ancients prescribed, disaster struck, that is Kataragama struck. A kapurāla (dēvāla officiant) died suddenly; while Gampola town was hit by floods on three occasions in 1913.52

In June 1914, the District Court decided the issue in the favour of Wallahagoda Dēvālaya. But, then, this decision was overturned in appeal by the Supreme Court on the 2nd February 1915. The leading Buddhist activists in the island immediately set up a Buddhist Defence Fund to collect monies to take this issue to the Privy Council. At a gathering of Sinhala activists on the 2nd March 1915 one speaker is alleged to have proclaimed: “we will wait the decision of the appeal and then give medicine according to the sickness” (Dowbiggin 1915b: 4). In the meanwhile, the Moors of the mosque at Castle Street, Kandy—a Hanafiite mosque with many Indian Moor and Afghan devotees—were encouraged by this decision to demand that “music” should cease in front of their mosque. As the Vesak season approached in late May, both parties, the Moors of Kandy and the Buddhist activists of Kandy, jostled for advantage. By the 28th May tension had accumulated, trouble was expected and both sides were prepared for battle.53

Given the several confrontations between Moors and Buddhists in other localities in previous years (see Roberts 1994b and Rogers 1987) the tension in Kandy must be seen as one moment in a long and escalating history of confrontation. This evidence not only demonstrates that the religious conflict was a major cause of the pogrom, but was critical in bringing Sinhala Buddhist hostility to the boil in mid-1915. The aggrieved resentment that courses through the correspondence of the Wallahagoda authorities in the period 1912-16 (above: 139-40;  Roberts 1994b) was not confined to them. W. A. de Si1va in 1918 and Akbar in 1927 explicitly called attention to such sentiments among the Sinhalese; and Akbar linked them to the pogrom:

Buddhist processions, especially the tom-tomings, are an essential part of the Buddhist religion, and the activities of the Police in stopping tom-toming in the vicinity of the places of worship have been always resented by the Buddhist people. This was the direct cause of the riots in 1915 (Akbar 1927).

This interpretation gains in value from the fact that Akbar (1890-1944) himself was a Muslim, albeit a Malay rather than a Moor.

This viewpoint gains support from another non-partisan source, an unlikely source emanating from Simla in India: namely, the General of the Salvation Army for the Indian region, Booth Tucker. In August 1915 he observed that “the animosities [between the Moors and the Sinhalese had] been aggravated of late by the offensive and pugnacious activities of the Moormen [who] have put up mosques in prominent situations and have sought to hinder the Buddhists from the exercise of their immemorial processions and prescriptive right.”54 There can be little doubt that Tucker was conveying the views of his Sinhalese workers and soldiers in British Ceylon—for the letter noted that they had “over 100 Sinhalese officers” in Ceylon.

Perhaps the most meaningful assessment of the pogrom comes from the schoolchildren in the village of Patboriya in Kegalle District. When the Assistant Government Agent, Burden, was on circuit a few months after the event and “tested the children on matters of general and local knowledge,” he slipped in a question on “the riots.” He was informed that they had “happened because the Moors insulted the Buddhist religion.”55 Here then, in the lucid voice of children was a distillation of attitudes expressed by the elders who surrounded them.      

Support for this position is provided by two striking pieces of evidence during or immediately before the violent assault on the Moors. At Walgama in the Western Province, on the 2nd June 1915, when rumours were rife and violence was being perpetrated around them, the Moormen of the locality had been approached by the local headman, a bhikkhu and a crowd of villagers: they were asked “to sign a document agreeing to the Buddhists beating tom-toms when passing the mosque for a period of 99 or 100 years” (Brayne 1916: 31).56

The other incident was part of the build-up for the pogrom and is also from the Western Province.

Evidence was obtained at Attanagalla, in the Western Province, as to the manner in which the attack on the Moors was committed. At Attanagalla there was a samagama or village society. Sixty or seventy villages belonged to this particularly society; four representatives from each village attended meetings, which were held at Attanagalla. Every village got a notice to hold a general meeting on May 28. On that day people were ordered to collect at the Attanagalla junction, when they heard the bell ring ten times in ten minutes. The bell is hung in the Buddhist temple. On June 1 the bell was rung. All the principal men of the samagama were present and all those who assembled were addressed as follows:

“The Tambies are insulting our nationality and our religion. We must harass the Tambies, and they must be driven out of Ceylon. We have received letters to the effect that Tambies are being harassed all over the place, and ordered (sic) to break their boutiques and mosques. We will give our lives for this. Go without fear, and break the churches and boutiques of the Moors and loot their property.”

Those assembled were divided into parties under leaders and moved off in different directions, looting the boutiques of the Moors. They were provided with food and arrack free of charge, and armed themselves with clubs, guns, swords, dynamite and bottles filled with sand.57

In summary, then, in the view of the Sinhala activists, the “Tambies [were] insulting [their] nationality and [their] religion.” They were disposed to teach the Moors a lesson for this presumption. There was an imperative in their grammar.

This imperative may have been all the more imperative for those Sinhalese who paid attention to astro-chronological junctures or turning points. The 2nd March 1915 was a hundred years to the day on which the Kandyan Convention, which ceded the Kingdom of Kandy to the British, was signed. This moment was highlighted in 1915 by public gatherings organised by nationalists, while a special issue of the Dinamina, replete with lion flag, sold out in no time and was quickly reprinted (Blackton 1970: 236).

The British were not inattentive to the importance of the occasion. On the 2nd March 1915 the Governor ceremoniously donated a pintāliya (a water container for pilgrims) to the people of Kandy.58 But this goodwill offering was undermined by the gods: a swarm of bees bombinated the pintāliya (Denham 1915: 9). One of the forms taken by Huniyam Dēvatāva or Huniyam Yakā, a godling and demon, is that of the bee (personal communication from Chandra Vitarana, 21 April 1991). To those Sinhalese attuned to the mythic wavelengths, this may well have been an omen of the gods, conveyed through Huniyam’s displeasure . . . and thus a sign of an impending change in the political order.

The British Raj Endeth?

The suggestion that the pogrom of May-June 1915 was a little more than an assault on the Moors, at least in the minds of a few activists, is raised by the content of their cries. Thus, Dowbiggin (1915: 8) alleged that among the frequent shouts raised during the disturbances were:

       The Western Province belongs to the Sinhalese.

       There is no more British Government.

       There is no more British flag.

       Sinhala Nation.

Likewise, if we can follow Denham, during the pogrom some headmen even announced that there was no more English government: “ingrīsi anduwa näta!” (1915: 8, 20). More finely-grained regional studies are required before one can fully comprehend the significance of such practices. The type of dispersed data which one has to sift through is indicated by the following illustrations: (i) during the court martial of alleged ringleaders of the attacks at Yatihena in the Western Province, Moor witnesses alleged that the first accused had a whistle, the second a flag and the third a gun, while the rioters shouted: “There is no English Government. This is the day of the Buddhists. This is our flag;”59 (ii) in the Dumbara region of the Central Province in “several instances” the cry, “British rule endeth on the 27th; Buddhist rule is reviving,” was raised (Wodeman in Eastern 128: 36); (iii) likewise, this fabulous expectation seems to have been voiced in other localities in the Western Province as well (e.g. Izat 1915: 122).60

Despite Izat’s scepticism as to the veracity of these reports, the incidence of such claims from a number of places suggests that he was in error. At Walgama, in fact, when a Moor trader had fronted up to a crowd and offered his goods, one of the ringleaders (a gurunnānse or schoolteacher) seems thereupon to have announced to the world-at-large: “There is no longer English Government. Now our Buddhist Government. Dynamite the buildings and set fire to the cadjan buildings.”61 The crisp, foreshortened language embodied in the second sentence is characteristic of the vocabulary deployed in moments of action (cf. Guha 1983). Among the stories in circulation and arguments deployed by stirrers seeking to rouse the crowds in the Western Province was one of the effect that: “British rule [had] been in existence for a century and was now about to cease, and there was a prophecy that a Sinhalese king would ascend the throne. [While another] version [contended] that the British had taken a lease of the country for a hundred years and the term had now expired!” (Brayne 1915: 266).

The seemingly bizarre brew of utilitarian manipulation, fabulous anticipation and credulity which is embodied in these eddies and currents of talk during a moment of high tension is something that we cannot easily grasp from our cloistered, rational chambers. That the brew proved to be combustible we know full well. As did many a Moor victim.

The symbolic use of flags is also significant. The flag would have been the Buddhist flag, a recent innovation inspired by Olcott and dating from the 1880s. As such it was an emblem of Buddhist resurgence. It has been one of the marks of the Vesak celebrations since the late nineteenth century. Late May 1915 was a Vesak period. It is likely that flags were on display in houses. But its also evident that some of the marauding crowds and a few of the itinerant cars promoting the violence carried flags. Indeed, one car in the environs of Colombo which an official, on subsequent reflection, surmised to have been involved in acts of incitement, had a roll of flags (Brayne 1916: 29); while at one place in the Western Province a flag was “borne by a man on an elephant” and at another spot a flag “was hoisted on to a tree amidst the acclamations of the people” (Brayne 1915: 266). It would seem that, to the stirrers and assailants, the Buddhist flag was an emblem of a new order.62 Challenging the British Raj meant displacement of Union Jack..

Given the impact of the British policy on the procession disputes, the swelling resentment against the insult inflicted upon the Sinhalese and their Dhamma and the possibility that a karmic change in the political order was anti­cipated by a few individuals, we need to treat these expressions seriously. Clearly, the implications attached to such practices are qualified by the fact that govern­ment buildings and Europeans qua Europeans were not attacked. Clearly, too, the principal target of hostility was the Moor community. The, question, then, is whether the disturbances were not merely a pogrom, whether in some localities they also contained a nationalist thrust against the British order—not only in effect, but also in intent.

Jayasekera has been the only scholar so far to go beyond the standard focus on procession disputes and to consider the possibility that millenarian hopes may have influenced some assailants. However, he qualifies this line of interpretation by noting that the fantasising slogans “expressed a belief in a happier state of life under a Buddhist and Sinhala government rather than a desire to overthrow British Rule” (1970: 242-43; and 391). Jayasekera’s unwillingness to pursue this argument appears to have been influenced by Hobsbawm’s thesis in Primitive Rebels: that millenarian protests are pre-industrial and pre-political. Since Jayasekera wrote, Ranajit Guha has run a coach and four horses through Hobsbawm’s thesis (1983: chapter 1). We do not need to be bound by Hobsbawm’s ethnocentric and restricted notions of the political.

I would argue that these fabulous expressions during the heat and excitement of action against the Moors embodied a lucid political insight: viz. that the economic activities of the Moors and their ability to interfere with time-honoured and critical religious acts were made possible by the British Raj. The Gampola Perahära case, after all, was instituted against the crown.

By the 1910s, the Ceylonese and Sinhalese activists were vociferously articulating a whole litany of grievances about British rule—from the govern­ment’s alleged expropriation of waste lands from the peasantry to the manner in which the British had encouraged the consumption of alcohol. The ardent Buddhist activists took up the latter cause through the temperance move­ment and added it to their pool of complaints (Rogers 1987, 1989; Roberts 1989a). In this view, moreover, the British government’s hostility to the Buddhist cause was amply demonstrated in the manner in which they handled the bo-tree affair at Kalutara in 1896 and the manner in which they continued to treat sacred space at Anuradhapura (Rogers 1986 & 1987: 183-87; Harischandra 1904; and Nissan 1989).

Thus, in the period 1890s-1915, the Buddhist Sinhalese were arraigned in opposition to both the British Raj and the Mohammedan Moors. The imperialism of silence initiated by the British through the Police Ordinance was seen to advantage the Moors (as indeed it did on the procession question). In attacking the Moors, then, Sinhala militants were attacking those whom they believed to be a rapacious, arrogant minority, and outsiders (paradēsakkāra) favoured by the British imperialists who were also paradēsakkāra. Their chauvinist violence was a political act, an indictment of the British Raj. The exultant cries that “the English state is no more” must be interpreted in these terms.

These political thrusts were constrained by caution. Unlike the Santals in 1855 and the Saya San rebels in Burma in 1930,63 those Sinhala activists who were moved by a touch of the millennium did not take on the British guns. Their fantastic hopes were kept within limits by the pragmatism of their peers: an awareness of British military power (evident every day in news about the war in Europe).

In this sense, the 1915 pogrom was not only a chauvinist operation, it was also a nationalist exercise in ellipsis. In taking this position, I have, quite irrevocably, abandoned one of the lines of interpretation I took when writing on this subject in 1970 (Roberts 1981). On that occasion I dismissed the idea that the pogrom was a protest against British rule. The reason for this stance was simple: I thought it ridiculous that anyone would attack the Moors in order to attack the British.

This was the reasoning of empirical-logical rationality. Such rationality, I now claim, is ill-equipped to understand either the mentality of those people whose perspectives are mythical in the sense clarified by Val Daniel, or the mentality of zealots such as the Anagārika Dharmapala. Dharmapala’s outpourings are parti­cularly illuminating and worthy of more in-depth studies.64 Not for him the step­-by-step progression of a lawyer’s brief, the coherent arguments presented by a D. B. Jayatilaka, W. A. de Silva, Marcus Fernando or James Peiris (leading national­ists of the early twentieth century). His angry letters shift gears in bewildering frequency. They literally ferment with turbulent criticism. But within the ferment, often, lies essence, insight. So, too, did the Sinhala Buddhist activists behind the chauvinist pogrom 1915 grasp the intimate connection between the British policy on sabda pūjā and the capacity of the Moors to impose injustices (ayuttak rendered plural) and hardships (amāruvak rendered plural) on them.


1. This is a revised version of a paper presented at the 12th Conference of the International Association of Historians of Asia, at Hong Kong, June 1991.

2.     Brayne 1915: 263 & 265; Jayasekera 1970: 267-68; and Diary of the A.G.A., Chilaw for June 1915 (rep. in Eastern 128: 83-86).

3.     I would extend this contention to India, where the use of the term “riots” in the social science literature can be called into question in the case of massed assaults such as those in Delhi in Oct-Nov. 1984 and Bhagalpur in 1989. Likewise, in the historiography of U.S.A. the prevalence of the phrase “race riots” would seem to require modification in those instances where the dominant White peoples systematically attacked the Blacks in their locality, e.g. in Atlanta in 1906 (Crowe 1969).

4.     Among them, Newton Gunasinghe and Shelton Kodikara in newspaper articles and the pseudonymous Piyadasa 1984.

5.     Thus, in this usage, it could even be used to describe Israeli Jewish assaults on Arabs in their midst. Note: as the story of Palestine reveals, dominance is often a majoritarian dominance, but need not necessarily be so.

6.     A phrase used during the discussion of Saul Friedlander’s seminar paper at the Common­wealth Center for Literary and Cultural Change, University of Virginia, mid-November, 1915. Also see B. Lang (ed.) Writing and the Holocaust, Holmes: Meier, 1988.

7.     Dowbiggin in Police Inquiry Commission 1916: 5.

8.     Tranchell 1916: 85; Diary entries by Vaughan (G.A., Central Province), 29 & 30 May 1915, DNA 18/46; Fraser and Festing in Police Inquiry Commission 1916: 16 & 66; Hellings 1915: 68; Rogerson in Eastern 128: 8-9; Diary entry by Festing, 2 June 1915, DNA 35/28.

9.     Jayasekera 1970: 265-66; Roberts 1981: 115, 117 and Ceylon Observer, 1 July 1915.

10.     Diary entry by Bartlett (G.A., Uva), 10 June 1915, DNA 57/77.

11.   Jayasekera 1970: 275; Diary entries by Cumberland (G.A., North Western Province) June, 11 August and 10 Sept. 1915, DNA 38/21; and Diary entries by Browning (A.G.A., Matara), 4 and 5 June 1915, DNA 26/173.

12.   Sendanayake 1915; Packeer Ally 1915; and Diary entry by Festing 1(A.G.A., Kalutara), 21 June 1915, DNA 35/28.

13.   Sendanayake 1915; Tranchell 1916: 83; Izat 1915: 122; Diary entry by Cumberland, 11 August 1915, DNA 38/21; and Ceylon Observer, 1 July 1915.

14.   Police Inquiry Commission 1916: i and Ameer Ali 1981: 2.

15.   Diary entry by Vaughan (G.A., Central Province), 30 May 1915, DNA 18/46. Also see Tranchell 1916: 85.

16.   Diary entry by Browning (A.G.A., Matara), 4 June 1915, DNA 26/173.

17.   A passing reference is made to the desecration of mosques by Blackton (1970: 243) and Ameer Ali (1981: 2). Blackton does draw attention to the “vengeful note” at the end of Rahiman’s letter.

18.   Sinhalese Memorial, 1915 and its appendices; Diary entry by Browning (A.G.A., Matara), 23 August 1915, DNA 26/173; and by Thaine (G.A., Sabaragamuwa), 15 July 1915, DNA 45/336.

19.   See minute by Col. Office clerk re-enclosures 4 & 5 in Chalmers to Sec. of State for the Colonies, 16 August 1915 in CO 54/783; Hellings 1915: 72; and Thaine 1915: A257. Cf. “[Rahiman’s report] I consider very fair and wonderfully restrained” (Festing, Special Commissioner, Kalutara to Col. Sec., 12 July 1915 in Eastern 128: 39).

20.   Diary entry by Codrington rep. in Eastern 128: 84; and Diary entry by Thaine, G.A., Sabaragamuwa, 8 June 1915, DNA 45/336.

21.   Jayawardena’s first essay does attend to the backdrop provided by the temperance movement and Sinhala communalism in the decades prior to 1915 (1970: 224-25), but this is glossed over during her analysis of the riots in Colombo. Her second essay (1984) is of great value in illustrating (in critical vein) the expressions of Sinhala chauvinism during the course of the twentieth century. There is a paradox here. Note the manner in which Metcalfe, following Sartre, clarifies his “dissatisfaction with the forms of class analysis… that explain phenomena by situating them in a pre-existing classificatory model or historical projection” (1988: 126, his emphasis).

22.   See a report by J. Thomas on a conference organised by the Humanities Research Centre, Australian National University, in the HRC Bulletin, 1992.

23.   No primary sources have been consulted in this essay—which is a comparative study that also embraces a Sinhala-Muslim clash in Galle in 1982.

24.   Roberts 1994b; Wickremeratne 1969; Amunugama 1978 and 1979; Roberts et al. 1989a: 10-21; A. P. Gnānālōka Thera in the Hitavādi, 18 Feb & 13 May 1914; and K. P. Sumanasāra Thera in the Sinhala Bauddhayā, 21 Nov. 1914.

25.   For a brief criticism of this thesis on ethnographic grounds with reference to the anti-Tamil violence of the 1970s and 1980s, see Roberts 1989b: 69.

26.   See the diaries kept by Thaine (G.A., Sabaragamuwa) and Browning (A.G.A., Matara), DNA 45/336 and 26/173 respectively; Devane 1915; Festing 1915; Izat 1915.

27.   See the evidence provided by several Sinhala villagers and by H. D. Keppitipola in DNA 65/242; and Walters in Eastern 128: 11-12.

28.   Thus far, I have only examined one part of the data in my own hands.

29.   The category of “cultivator,” which figures in the official correspondence, is nebulous and could include modestly comfortable individuals, such as Loku Bandā Aturupana (Roberts 1981: 120). Here, I have narrowed its range to smallholders, whether tenants or proprietors.

30.   Dowbiggin 1915b, quoted above in text; Jayasekera 1970: 116 & 189-246; and The Ceylonese, 1 Jan. 1915.

31.   Jayasekera 1970: 283ff. & 325ff.; Roberts 1981: 113-22; Diary entries by Burden (A.G.A., Kegalle), 16 July 1915 in DNA 36/30 and by Thaine (G.A., Sabaragamuwa, 8 July 1915, DNA 45/336; Ceylon Observer, 7 July 1915.

32.   Diary entries by Thaine, 5 Aug. 1915 (DNA 45/386), by Burden, 1 Sept. 1915 (DNA 36/30), by Festing, 25 June 1915, (DNA 35/28) and Browning, 6 & 9 June 1915 (DNA 27/173); Abdul Rahiman 1915; Markar 1915; and Police Inquiry Commission, 1916: passim. Re the North Western Province, see diary entry by Cumberland, 25 July 1915 (DNA 38/21).

33.   Roberts 1981: 115, 116, 119 & 120-122; Jayasekera 1970: 281-82 & 331ff.; Ceylon Observer, 1 July 1915 and Diary entries by Browning, 6 June 1915 (DNA 26/173); by Vaughan, 7 June 1915 (DNA 18/46); by Cumberland 13 Sept. 1915 (DNA 38/21). “The local registrar, notary, schoolmaster, whenever such was a Buddhist, was almost invari­ably found to be working behind the scenes” (Devane 1915: 61).

34.   Jayasekera 1970: 271ff. and 281ff.; Roberts 1981: 114,120 & 121; Abdul Rahiman 1915; Ceylon Observer, 3 July 1915; Diary entry by Thaine, 2 June 1915, DNA 45/336; Blackton 1970: 240-41; and evidence re the bhikkhu at a Wahumpura village, Dumbara, in Ratnapura District collected through fieldwork by Chandra Vitarana (April 1991).

35.   Diary entries by Thaine (G.A., Sabaragamuwa), 4 June and 15 Sept. 1915, DNA 45/336; Jayasekera 1970: 276-77; Kannangara 1984: 146-47; and the evidence in DNA 65/242.

36.   Diary entry by Cumberland (G.A., North Western Province), 25 August 1915, DNA 38/21; Eastern 128: 110, 156, 267, 269 & 273; and Blackton 1970: 252.

37.   Diary entries by Thaine (G.A., Sabaragamuwa), 16 & 29 June and 13 & 14 July 1915, DNA 45/336. Also see Festing 1915.

38. Thus Thaine described Dēvālagama in Ratnapura as “one of the rowdiest villages in the district” (10 July 1915, DNA 45/336). This village has a large Govigama population (fieldwork information from Chandra Vitarana).

39. E.g. see Roberts et al, 1989: 10-21, 80-81.

40. Wickremeratne 1969; Roberts 1979c: 225-29; Amunugama 1979; and Jayasekera 1970: 86­-87 & 106-11. For direct evidence, see Dharmapala 1965: 501-44; the articles on Sinhala jātiya pirihīma (Decline of the Sinhala nation) by K. P. Sumanasāra Thera in the Sinhala Bauddhayā, 21 Nov. 1914 et seq. and W. P. Perera’s letter in the Dinamina, 2 March 1915.

41. Sirisena, Apata Vecca Dē, 1954: 9ff. and Sucaricādarsaya 1958: 126 & 130. Also see Roberts et al. 1989: 10-21.

42. See “Ratē tibena ävul, apatama ve tävul” in Sinhala Jātiya, 1 June 1913. Sinhala Jātiya 30 March 1915: Sinhala Bauddhayā, 2 Jan 1915: translation of article by W. D. A. Gunatilaka in the Sinhala Jātiya, March 1915; in Dowbiggin 1915b and Roberts et al. 1989: 10-21.

43. Kocci Demalā (Malayālam Tamil) is the title of his piece too (Sinhala Bauddhayā),14 Jan. 1910.            

44. Jayasekera 1970: 106-11; Wickremeratne 1969; Sinhala Bauddhayā, 2 Jan. 1915; the series of articles serialised in the Hitavādi from 11 February 1914 as well as the article on “marakkala kada” (Moor shops) in Hitavādi, 20 July 1914.

45. Denham 1915: 2; Tranchell 1916: 85; Rogers 1987: 190 and Police Inquiry Commission 1916: 64, 129.

46. Jayasekera 1970: 293 and Eastern 128: 276, 272, 70.

47. Wickremeratne 1969; Jayasekera 1970: 106-11 and Hitavādi, 20 July 1914.

48. Letter from Saxton (G.A., Central Province) to P. B. Nugawela, 27 August 1912 in DNA 18/3461.

49. The “shrines to the gods located in the outer district”; namely, those at the temples of Lankātilaka, Gadalādeniya, Embekke, Wegiriya, Dodanwela, Pasgama and Ganēgoda.

50. Correspondence from T. B. Keppitipola et al., espe. letters dated 11 and 22 Sept. 1915 in DNA 18/3462.

51. Correspondence cited above; Letter from Goonewardene and Wijegoonewardene (proctors) to Attorney General, Colombo, 23 July 1913 in DNA 18/3461. Re the protests from the tenants opposed to an alternative route, see Letter from the Basnayake Nilame, Wallahāgoda Dēvālaya, 11 Oct. 1912, DNA 18/3461.

52. Administration Reports 1913, G.A., Central Province, pp. Bi and B7; and a summary of the views expressed by the tenants contained within de Sampayo’s judgment (NLR 1915, 18: 203) and Paul E. Pieris’s judgment (Thalgodapitiya 1963: 57).

5. Evidence from F. R. Dias and H. P. R. Jayawardene before the Police Inquiry Commission 1916: 95 and 100 re the presence in Kandy on the 28th May night of Wanawāhala toughs, and Tranchell 1916: 85-86 re the advice given by some Sinhalese to certain planters not to visit Kandy that night.

54. Letter from F. Booth Tucker to Lord Grey, 6 August 1815 in CO 64/791.

55. Diary entry, A.G.A., Kegalle, 25 Nov. 1915, DNA 30/36.

56. The magic figure of 99 years is an indication of the power exercised by the British legal code—for the figure appears to have denoted “in perpetuity.” Cf. how the Santal insurgents in India in the 1850s showed a parchment of writing as proof of their god’s command to resist—even though they could not read; a practice which Guha brilliantly interprets as an appropriation of the sign of authority (1983: 54-55, 187).. The Moors of Walgama, it should be added, did not save themselves from attack. The next day (?) they were attacked “by another party of Sinhalese” (Brayne 1916: 31). Walgama appears to have been “besieged by hundreds of Sinhalese from the surrounding country” on the 2 and 3 June, and was the scene of a “pitched battle” (Brayne 1915: 261, 265).

57. Admin Reports 1915, Police, by H Dowbiggin, 20 May 1916, Part III, B6. This evidence was collected by a British civil servant named Izat (see Izat 1915). One of the associations involved in these gatherings was the Paranagama Association, a welfare association controlled by the Karunapedigē family of the Batgama caste, one of whom was a police duraya, i.e. minor headman (see Petition of Horatel Pedigē Jambuwa and W. Babaya, and Report of the I.G.P., 30 Oct. 1915, on this petition, encl. in Chalmers to Sec. of State, no. 720, 4 Nov. 1915, CO 54/785).

58. Diary entry by Vaughan, G.A., Central Province, 2 March 1915, DNA 18/46.

59. Evidence from M. L. M. Sameem, M. L. Saldeen and K. M. Ali at the court-martial of Girigoris et al., encl. in Chalmers to Sec. of State, 7 Oct. 1915, CO 54/785.

60. “There is no English Government now, we have a Sinhalese Government” and “Our people are in the Kachcheries and Government offices” are reported to have been expressed in the “many reports of seditious speeches” which Izat received. “Many of these reports are probably false,” said Izat (1915: 122).

61. Evidence of M. Salim at a court martial at Hanwella, 16 July 1915, in Eastern 128: 139.

62. Here, too, Guha’s exposition (1983: passim) of the inversions and emulations of the tribal peasants challenging the British state and/or Hindu supremacy provides suggestive analogues. Contrast the manner in which Jayasekera interprets the flags as “an attempt to give some organization and crusading appearance to the attacks,” and a phenomenon without any political motives (1970: 344).

63. For the Santal uprising, see Guha 1983: 10, 26, 54, 64ff, 95-100, 128-35, 176-88, 205-09; and for the Saya San rebellion, see Solomon 1969; Adas 1979: 38-40, 47-48, 56-58, 72-77, 99-102, 127-28, 141-42 and Sarkisyanz 1965: 160ff. Several of these Burmese peasants believed in the protective power of medicinal potions and tattoos—that they would gain immunity from injury in battle.

64. An outstanding example is provided by the letter of protest he sent to London in 1902, while in Los Angeles, in criticism of the government’s activities in Anuradhapura (Dharmapala to Sec. of State for the Colonies, 3 Oct. 1902, in CO 54/680).

BIBLIOGRAPHY:  for the majority of the citations here see Full Biblio is in Confrontations in Sri Lanka pp. 412-16.


Filed under communal relations, economic processes, historical interpretation, life stories, Muslims in Lanka, nationalism, politIcal discourse, power politics, sri lankan society

7 responses to “Marakkala Kolahālaya: Mentalities directing the Pogrom of 1915

  1. Trevor LaBrooy

    Very well written and presented.

  2. Pingback: Virulent Shameless People! Cry OUR Beloved Country! | Thuppahi's Blog

  3. Pingback: Caste in modern Sri Lankan politics: A repetition of an old excursion from early 2010 | Thuppahi's Blog

  4. Pingback: The Induction Oath of Tamil Tiger Fighters at their Passing-Out Ceremony | Thuppahi's Blog

  5. Pingback: Victimized Muslims of Aluthgama-Beruwela express Grief and Legitimate Resentment | Thuppahi's Blog

  6. Pingback: The War in Sri Lanka and Propaganda Debates | Thuppahi's Blog

  7. anoma abeyerwardene

    .: All this is most informative but apart from astrological movements, totally ignores the huge external influences.
    * Muslims have live amicably with the Sinhalese for centuries (Colombo for instance, was a Muslim port) and as they were interested in trade, not conquest, took pains not to tread on toes.
    *The North Indian Muslims who came with the British were far more aggressive. For instance they objected to Buddhist processions passing their new mosques, while the far older Ceylon Moor mosques on the same routes, had never objected. Throwing stones at peraheras and then shooting those who attacked their houses and mosques, didn’t exactly help matters.
    *Meanwhile the British were extremely skittish. They were at war with the Ottoman Empire and the last thing the British and French Empires needed, was the Turms declaring jihad on them, given their huge numbers of Muslim troop, let alone te civil unrest that would follow.
    To make matters worse, the Emden was rumoured to be recoaling near off Hambantota with local connivance.
    They brought in an Indian Muslim regiment.
    *The Governor Chalmers was an orientelist who got on well with the mahanayakas.But his IGP was a from an Anglo-Ceylonese family who, like the Kenyan planters took a dim view of the natives. He stampeded the Governor into virtually handing over military powers to him and the British Colonel incharge of the aforementioned regiment.
    In the furore that followed, Chalmers was recalled in disgrace while the real villain of the piece Dowbiggin,, appeared to have save the colony from destruction like Brigadier Dyer of Jalianwallah Bagh fame [sic]. .

Leave a Reply