Caste-based Differentiation in Sinhalese Society :EPW

Spread the love


Role of Buddhism and Democracy

This paper interrogates as to how caste-based differentiation exists among the Sinhalese in Sri Lanka despite the near absence of the economic structures that initially underpinned the caste system. It looks at the role of Buddhism and democracy, two supposedly egalitarian institutions, in encouraging the perpetuation of caste-based differentiation among the Sinhalese. Such differentiation is not always reproduced in discriminatory terms, but rather—as in the case of democracy—as a bargaining tool of group identity that secures access to patronage goods and resources. In contrast, in the case of Buddhism, caste operates behind a veil of public denial and serves as a “filtering” function in ritualistic gatherings and the organisation of the Buddhist order.

It is hard to find scholarship on Sri Lankan politics and ­society that does not take caste into account in its analysis. Despite this being the case, caste among the Sinhalese in Sri Lanka “increasingly forms a ‘hidden identity’” (Silva 1999: 205) that nonetheless continues to dictate social interaction to a great extent (1990). After multiple encounters with this fluid, ambiguous, elusive and yet omnipresent social institution in our research exercises, we were encouraged to inquire into caste and its prevalence in modern society, particularly in relation to the Sinhalese.

For instance, consider the case of the annual Kandy “Esala Perehara” (a procession that takes place in August). It not only epitomises the practice of caste in Sinhalese society but also, if one looks at it closely, illustrates the nature of the transformation the practice of caste is undergoing at present. The Esala Perahera in Kandy is one of the oldest and grandest (now) Buddhist festivals in Sri Lanka, featuring dancers, jugglers, musicians, fire-breathers, and lavishly decorated elephants. Hundreds of performers take part in this pageant and specific caste groups are designated to perform specific roles. The following description by a Kandyan Govigama (farmer caste) peasant illustrates how caste-based differentiation persists, albeit behind a veil of public denial, despite the gradual weakening of the material existence of caste practices in Sinhalese society.

Nowadays some people of the Govigama caste also perform certain roles in the perahera that were traditionally assigned to people from lower caste groups. For example, some Govigama caste people now carry pandam (torches) in the perahera which was traditionally performed by people of a particular lower caste group. Those lower caste people are no longer willing to carry pandam in the perahera as they have achieved higher socio-economic status now. So now, poor individuals from other caste groups, including ones from the Govigama caste, have come forward to perform such duties for a fee. However, such people cover their faces with a cloth adding a new style to their dress code because of the fear that villagers would recognise them performing duties meant for lower castes.

This description illustrates how caste-based differentiation continues to prevail among the Sinhalese community despite the weakening of its economic bases (de Zoysa 2013: 96), enca­psulating the transformation of caste in general. As scholars such as Ryan (1953) point out, Sinhalese caste was organised according to the occupational functions assigned to various groups in society. Hence, cultivators came to be known as Govigama, washers Dhobidrummers Beravagoldsmiths Navandanna, and so on (Ryan 1953). During precolonial times, the specialised services of various groups were deployed to build water reservoirs, cultivate the land, and/or sustain the Buddhist order by material means (all considered meritorious ­activities in Sinhalese Buddhist ideology) (Gunasinghe 2007: 48). However, as to how and why caste prevails today in Sinhalese society, when labour is free to move where it wants, or in other words, occupation is no longer determined or restricted by birth (in ideal terms), is a question worth exploring. We therefore inquire in this paper as to how, when traditional occ­upations that denote caste status are fast disappearing, caste-based differentiation has managed to survive in present-day Sri Lanka. In doing so, our aim is to understand how group identities, social structures, and social and cultural relations, and practices continue to legitimise and reproduce differences and inequalities. To this end, we hypothesise that both culture and politics—Buddhism and democracy—in two different ways produce the same social outcome: The reproduction of caste as a structure of consciousness and differentiation that gives practical meanings to both institutions.

In building our analysis, we primarily draw from fieldwork carried out in Dedigama (in the district of Kegalle, Sabaragamuwa province) and Kelaniya (in the district of Colombo, Western province) in Sri Lanka over three years from 2009–12. The choice of field locations was encouraged by the need to capture rural and urban dynamics related to caste, since caste is more apparently manifest in rural areas whose material conditions partially warrant its prevalence, whereas in urban areas it exists mostly in the ideological terrain. For instance, ancestral settlement patterns still continue in rural areas to a degree, defined by erstwhile caste arrangements, and this in turn encourages or discourages people from buying land in areas where such concentrations are apparent (Silva 1997). We have employed an interpretive approach to capture the material and ideological existence of caste and caste politics in Sinhalese society. Therefore, methods such as in-depth interviews, obs­ervation, and collection of demographical and interpretivist data from secondary sources were used. Our observations are framed within the parameters of Bourdieu’s concepts
of “habitus” and “social practice” that capture the dialectic relationship between structure and agency in creating the social experience of humans.

We first argue that the “social practice” of Buddhism in Sri Lanka has created a “habitus” that feeds, and is fed by, the social consciousness of caste that finds expression in both the cognitive realm of awareness and the material realm of socio-religious customs and practices (Bourdieu 1989). Bourdieu termed this phenomenon “structure structured and structure structuring,” whereby social structures help shape individual cognition and how the individual perceives and interacts with such structures has an impact on those structures in turn. We make this argument based on our field observations where exi­sting caste stratifications are accommodated in the seating arrangements of the bana maduwa,1 and different temples are set up for different caste groups, among other things. These signal not only the internalisation of caste by individuals (that leads to people identifying with their caste categories to ­organise themselves this way), but also the externalisation of it in outside structures that symbolise, and thereby reinforce, caste stratifications.

Second, in terms of democracy, political parties and party competition have given new impetus to the caste discourse in the country. As a consequence, caste-based differentiation has survived despite the egalitarian promise of democracy, transforming itself to a powerful tool with which to negotiate politics. We, therefore, posit that the supposed equalising effect of democracy expected through its “one person, one vote” concept has been subverted precisely by itself, whereby the aggregate ultimately comes down to how many such votes each “group” in society has. This has led to increased intensification of caste-based differentiation, encouraging patronage networks of material benefit distribution to be organised according to caste. It is ironic not only how the political system takes note of caste identity, but individual caste groups themselves demand that such identity be electorally significant. Again, therefore, we demonstrate that the “habitus” created by the interplay of cognitive grasping of caste and the material manifestation of such grasping in segregated socio-economic organisation seem to be dictating the political space of Sri Lanka.

One noteworthy finding stemming out of this study is that caste-based differentiation does not always lend itself to reinforcing hierarchy. In the case of Buddhism, the accommodation of caste in the organisation—that is, nikayas or fraternities—and ritualistic expressions—such as the establishment of caste-based temples, alms rituals preventing members of so-called lower caste groups from offering cooked food, and the division of labour in funeral rites—of the religion necessarily imply ­hierarchy. But in the political sphere, caste has acquired far greater nuance, whereby electoral choice is influenced by caste considerations, but mostly within a patronage framework and in numerical terms.


Buddhism informs the way of life in Sri Lanka to an extent that has well transcended the strictly cultural realm and has spilled over onto politics and the economy as well. As such, it would be useful to study the interaction between this supposedly caste-neutral religion and caste practices in the island’s majo­rity Sinhalese community. In this section, we examine both the philosophical position of Buddhism against caste and how the religion’s material manifestations have nonetheless evolved to endorse this institution. In this connection, we will look at how Buddhism is practised in Sri Lanka in terms of (i) how Buddhist religious rituals reproduce caste hierarchies and thereby caste-based differentiation and (ii) how Buddhist cultural practices are shaped by, and in turn shape, their religious counterparts.

As is popularly known, Buddhism was introduced to Sri Lanka by Bhikku Mahinda, son of emperor Asoka. Shortly ­after, Bhikkuni Sanghamitta, Asoka’s daughter and sister of Bhikku Mahinda, arrived in Sri Lanka, and brought along with her craftsmen belonging to 18 vocations. Since then, Buddhism has become the main religion of the Sinhalese majority of the island, thriving under the patronage first of kings and then even of elected leaders. The occupation-based system of social stratification introduced to the island by Bhikkuni Sanghamitta highlights the initial relation between Buddhism and caste. However, it would be misleading to suggest that this system is what is singularly responsible for giving rise to notions of caste in Sri Lanka. There is evidence, in fact, that the Sri Lankan caste system predates Buddhism.2 In addition, the Buddhist doctrine does not allow space for caste.

Drawing from core Buddhist teachings on personhood, Senaratne (2017) argues that in denying a permanent soul or self, Buddhism “critiques the rationale underlying the caste system” (p 105). His observation stems from the fact that in the absence of a perpetual soul, a social system based on the “high” or “low” trans-mundane origin of that soul is rendered meaningless. In addition, he points to the Buddha’s assertion that nirvana is a state that can be realised by anyone, regardless of their birth, as evidence of the anti-caste disposition of Buddhism. Drawing from the works of scholars such as Kalupahana (2009), Karunadasa (2013), Malalasekera and Jayatilleke (1958), and Harvey (2000), who reproduce many of Buddha’s famous references to the logical deficit and moral repugnance of the caste system, Senaratne concludes that the Buddhist doctrine has assumed a definitive stance against the oppre­ssive institution that is caste. Other scholars have made this point before (De Jong 1988; Weber 2001; Silva, Sivapragasam and Thanges 2009).

On the other hand, scholars such as Ryan (1953) and Chakravarti (2007: 17) observe that Buddhism is perhaps more caste-neutral than actually anti-caste (Ryan 1953; Chakravarti 2007). It tries to provide a means by which individuals—particularly of lower caste—may come to terms with caste inequalities. The Buddha used the karma theory to explain that birth in a low caste is the result of stupidity and foolishness in previous births, which has, in a sense, “enabled Buddhism to accept the world as it exists” (Chakravarti 2007: 16). Though there have been instances where the Buddha has openly criticised the caste system,3 in using the karma theory this way “Buddhism … accepted ideological roots for the established order of social hierarchy” (Ryan 1953: 36).

This is not to deny the emancipatory function of Buddhism, especially in the rigidly casteist context of Hindu India. As Chakrawarti herself has pointed out, the Buddha’s main method of handling social divisions was to “create a parallel world through the sangha [monastic order], a society that does not actually accept social divisions” (2007: 17). She explained that inequalities owed mainly to production and reproduction patterns, and advised opting out of both to break free of the structures of the lay world, which is the mission of a Buddhist monk (Chakravarti 2007). Buddha’s intervention, then, was not meant to reform or transform the lay society, but to present an egalitarian social ontology by offering a “caste-less,” egalitarian and “republicanist” alternative in the Sangha organisation. His Sangha community was an egalitarian utopia outside society, sustained by a personal ethical commitment to equality.

However, social activists like Bhimrao Ambedkar have sought an ­expansion of even this reading such that Buddhism was taken to have a social emancipatory function. In Ambedkar’s view, Buddhism “met the complex requirements of reason and morality” (Queen and King 1996: 46) he was looking for by espousing ideas of liberty, equality, and fraternity, and well before the French Revolution as well (Jaffrelot 2017). In this understanding, Buddhism promoted the same societal values that Ambedkar sought to further through political democracy (Jaffrelot 2017).

Though the Buddhist philosophy is largely accepted as not endorsing caste, there has been great symbiosis between how Buddhism is practised and the caste structure in Sri Lankan society. Earliest evidence of this coexistence can be traced back to the 18 vocations Bhikkuni Sanghamitta brought with her to the island. This coexistence of Buddhism with caste at present can be illustrated on two counts: (i) caste practices among the Sangha, and (ii) caste practices in lay Buddhist ritualistic and cultural expressions.

First and foremost, the institution of the Buddhist clergy ­itself is organised such that “at the island-wide level sangha organisation … reflected caste divisions” (Rogers 2004: 70). Though admittedly Buddhist Nikayas were originally founded upon different “schools of religious opinion” (Bandaranayake 1974: 18; Gunawardena 1993: 7), “… by the 18th century one could un-hesitantly assert that caste had become an integral element” of the institution (Liyanagamage 2008: 206). The play of caste in the Buddhist order has found expression in the formation of various Nikayas based on caste.

The largest and the oldest is the Siyam Nikaya founded in mid-18th century. Its center is in the Central Province of Ceylon, in the up-country districts surrounding Kandy. By far the wealthiest sect, it has retained title to many of the lands given to its predecessors by the Sinhalese kings. It has an estimated 12,000 members or approximately 65% of the Sangha. Higher ordination in this sect is restricted to members of the higher levels of Govigamas. It has been traditionally dominated by the up-country elements who control, through family succession, title to the temples and their lands, although many of the most learned members are said to come from coastal areas. A second sect, the Amarapura, was founded as recently as early 19th century with the intention of purifying the Sangha. One of its principal targets was the caste-based differentiation of the Siyam Nikaya. It is open to members of the Karawa, Salagama, and Durawa. (Wriggins 1960: 191–92)

In fact, prominent members of the Sangha during the colonial period sometimes actively participated in caste debates, and non-Govigama laymen of the south-west of the island ­“often worshipped at temples where the monk was of their own caste” (Rogers 2004: 70). Kannangara (1988) cites Roberts (1982) in recounting one such debate in which Reverend Weligama Sri Sumangala argued for the superiority of Karawe (fisher) caste in a book named Ithihasaya, provoking a strong response from the renowned Hikkaduwe Sri Sumangala, a monk of Govigama origin. Divisions escalated to the point that the three fraternities now run their own temples, and have their own networks of patrons defined by caste (Perera 2018). Kannangara (1988: 149) explains that unlike Kandyan temples that enjoy rich land endowments as a result of royal patronage, low-country temples were very much dependent on the generosity of their patrons since they had no such endowments. The “rich” in the low-country areas coincided with the newly ascending castes of non-Govigama origin, and as such, the ­patron bases of these temples, and through them the respective nikayas to which they belonged, came to be defined by caste, particularly non-Govigama.4

Apart from these internal dynamics of temples, one also sees caste manifesting in the religio-cultural rituals practised by the Buddhist laity. Establishment of temples, practises of various Buddhist rituals, and maintenance of Sangha–laity ­relationships are also heavily coloured by caste-based differentiation in Buddhist lay society. For instance, one sees very clear caste implications in the division of labour in traditional village funeral events, where playing of drums is done by the Berava caste, digging of the grave is done by the Dhobi caste, and so on. Further, those belonging to higher castes do not as a practice dine at the funeral house, though the custom is to partake in the post-burial/cremation meal. Due to the awkwardness these divisions give rise to, people now prefer to have ­funerals organised by members of their own caste group. This way, everybody can participate in every activity of the funeral without segregating. In the field location of Dedigama, multiple “village funeral societies” have been formed in order to accommodate mono-caste groups. What is noticeable about these situations, and an observation that brings back the reference we made at the outset to the Kandy procession, is the muted manner in which these practices are continued. No ­explicit mention is ever made of caste, but differentiation—on the part of both so-called higher and lower castes—prevails nonetheless.

Uyangoda (2012) succinctly illustrates how caste hierarchies and caste-based differentiation come to define the Sangha–lay relationship. He explains how poor people of downtrodden castes in Kurunegala find it difficult to set up a temple for their village and to have a resident monk in it. Illustrating the intensity of caste influences in this regard, he states:

The villagers do not have free access to the Buddhist temple in the ­village to fulfill their religious and cultural means. There are several elements in the religious cultural, and of course social deprivation experienced by villagers in Kohombakanda. Monks in the Buddhist temple do not accept cooked food offered by Kohombakanda villagers. They do not attend funerals in the village to perform Buddhist religious rituals. Monks of the temple do not visit homes of kohombakanda villagers to chant pirith, to accept alms, or for any other religious function of the families. When the villagers have a family funeral, they invite monks from a distant village or lay people to perform Buddhist rituals. (p 47)

Our field experience in Kegalle was another case in point. The chief incumbent of one of the most prominent temples in the area, one that has enjoyed royal patronage during the time of kings, noted thus:

Nowadays we don’t take caste differences that seriously, but up until 10 years ago people from certain caste groups were not allowed to bring cooked alms to the temple. Instead, those groups were asked to bring dry rations and get the temple staff to prepare the meal. (Chief incumbent of a prominent temple, personal communication, Dedigama, 25 May 2009)

Caste, then, defines the temple as an institution, and also the relationship between the temple and its patrons, as evidenced by laymen also inviting priests of their own caste to perform alms rituals in order not to have their caste sensitivities offended (Perera 2018). An ultimate expression of the ­accommodation of caste in the Buddhist clergy–laity relationship is how the bana maduwa in some temples has been organised. Arranged in two levels, the higher castes sit in the upper level, while the lower castes sit in the lower one. The logic is that the patrons of the temple deserve higher seating, but in villages where caste and wealth mostly coincide, the implications of these arrangements are hard to miss. Again, however, no ­explicit references are made to the actual line of demarcation involved here. This shift from overt to covert practising of caste-based differentiation also signals a transition from a ­material to more of a cognitive basis on which such practices are grounded.

Social and Cultural Practices

Traditional peraheras organised annually by temples—on both the national and local levels—constitute another example of caste-based differentiation in Sinhalese society. These processions are usually organised in celebration of either full moon (poya) days (considered religiously significant by Buddhists), the beginning of the rainy season when monks retreat into temples for meditation (katina pinkama), or, in the case of the Temple of the Sacred Tooth Relic, Sri Lanka’s most consecrated Buddhist temple, to pay homage to a relic of the Buddha. These processions consist of many cultural attributes like drumming, whip-dancing, fire-dancing, performing the traditional Sinhalese mode of dancing, etc, all of which are typically assigned to the relevant caste groups traditionally associated with each of these practices. Gombrich and Obeysekera (1990) have dubbed this particular variant of Buddhism “Sinhalese Buddhism,” in the sense that it is a version of the philosophy that has adapted itself to the agrarian social structures of Sinhalese society.

These cases illustrate that caste-based differentiation is at once a subjective (in that it resides in the cognitive structures of the individual whose experience is defined by this knowledge), as well as an objective phenomenon that manifests in long-standing sociocultural practices. It “is a state of mind … which is experienced by the emergence, in various situations, of groups of various orders generally called ‘castes’” (Dumont 1998: 34). As the above examples show, rituals related to ­funerals, alms giving, processions, and even seating arrangements inside Buddhist bana maduwa are manifestations of caste-based differentiation in society. Setting up of different temples for different caste groups is a particularly telling ­example of both the internalisation of caste (in that it shows how people have made caste part of their identity to the extent of organising—and defining—themselves according to it), as well as the externalisation of it in outside, visible objects, constituting a “social practice” as Bourdieu (1989) would have it. This process of subjective phenomena becoming objective realities over time through manifestation in social practices has defined a “habitus” of caste in Sri Lanka that continues to dictate the cultural choices and practices of Buddhists, contributing to the prevalence of caste on both the ­cognitive and material levels (Bourdieu 1989).


During the late colonial era, caste constituted the primary unit of identity that determined competition, differentiation, and rivalry (Uyangoda 1998). “The colonial power … had to use the existing structures of domination and political power” to ensure effective control (Uyangoda 1998: 19). This, on the one hand, reaffirmed the power of the Govigama caste that had enjoyed a dominant position till then, and on the other, incentivised other caste groups to also aspire for colonial favours. It was to this society that universal franchise was introduced in 1931, with the promise of an egalitarian ethos that would transform precolonial feudal hierarchies into more horizontal relations. Despite having experimented with democracy for close to a century, caste-based differentiation among the Sinhalese continues to remain important. In this section we argue that the practice of democracy, among other things, has contri­buted to the reproduction of caste-based differentiation. Political parties and party competition, both direct results of the int­roduction of democracy in 1931, gave new impetus to the caste discourse in the country. As a consequence, caste-based differentiation has survived despite the egalitarian promise of democracy, transforming itself to a powerful tool with which to negotiate politics.

In order to examine the nature of democracy’s interactions with Sinhalese caste, we will discuss the expansion of demo­cracy over the past eight decades in three phases: (i) early years of electoral competition under parties of “notables,” (ii) electoral competition under two-party hegemony, and (iii) electoral competition under the proportional representation system. In these three phases, the politicisation of caste identity at national as well as local-level electoral competition contributed to the reproduction of caste-based differentiation, while at the same time renegotiating caste hierarchies in the country.

The two decades immediately following the introduction of universal adult franchise were marked by politics of “notables” (Woodward 1979), or traditional patrons at the village level. Although for the first time the masses received an opportunity to contest for leadership and become political leaders, or at least participate in selecting their leaders, the elites of the country managed to successfully manipulate the electoral process to establish and perpetuate their grip over national and local politics (Woodward 1969; Kearney 1973; Jupp 1978). During this initial stage, the majority of the rural electorate rallied around local patrons—landlords and compadre capitalists—instead of choosing candidates based on ideological and policy positions, as democratic theory ideally expected. Calvin Woodward (1969) rightly called those parties as parties of notables. As a result, Karava, Durawa, and Salagama elites of the western and southern coastal areas whose elite status was the result of colonial capitalist fortunes, and the elites of the traditionally superior Govigama and Vellala (Tamil farmer caste) castes, managed to consolidate their class gains by slicing local electorates along caste lines. During the first two decades after the expansion of the franchise, therefore, leaders of the traditional and emerging upper caste groups were able to hegemonise national politics, forming an exclusive upper class and upper caste bloc. Instead of acting as a deterrent, then, early stages of the democratisation process demonstrated a somewhat reconciliatory attitude towards ­existing caste-based social hierarchies. Groups situated further down in the caste hierarchy benefited out of the arrangement through patronage and economic security provided by these rich notables. Reproduction of existing caste hierarchies during these initial stages of democracy were at the discretion of the elites.

Later, however, the democratic space opened up opportunities for the lower caste groups to resist the continuation of the hegemony of notables. As already mentioned, in addition to caste superiority, these notables also belonged to rich classes that were either of the Christian faith or at least had English education in missionary schools. The first to challenge the ­hegemony of these political notables were the Buddhist revivalist and swabasha (vernacular) movements as a backlash against the cultural humiliation experienced by the locally oriented elite class at the hands of the Westernised political bloc. This long brewing agony of Sinhalese vernacular groups towards the ruling elite found expression in the 1956 coalition called the Mahajana Eksath Peramuna (MEP) made up of what was attractively labelled as the Pancha Maha Balavegaya (Five Great Forces) comprising monks, teachers, indigenous doctors, peasants, and labourers.

The MEP formed a government overthrowing the United National Party (UNP) that was considered the party of the rich and upper classes. Interestingly, although the MEP mobilised marginalised economic and caste groups and claimed that the 1956 victory was also one of poor and marginalised caste groups, the leadership of the MEP came neither from the poor classes nor downtrodden caste groups (Woodward 1969; Jupp 1978). Instead, and also in contrast to the UNP leadership, the MEP leadership comprised politicians from rich and intermediary classes. Even though the MEP managed to mobilise oppressed caste groups who suffered at the hands of upper-caste landlords and state officials who associated with the UNP, it was mainly peasants belonging to the traditionally “superior” Govigama caste who came to the legislature to represent both marginalised caste and class groups.

The 1956 election marked the beginning of two-party electoral competition, which changed the electoral landscape— particularly electoral mobilising—radically. Both main parties, the Sri Lanka Freedom Party (SLFP) and the UNP, under intense electoral competition, started to reach out to the masses irrespective of class, caste, ethnic, religious, and cultural identities of individuals in order to maximise their vote count in the election. During this second phase of electoral politics, both main parties opted for catch-all style party positions by thinning their ideological and policy differences and canvassing for multiple ideological and policy stances, sometimes even subscribing to contradictory policies. In this context, in the Sinhalese countryside that was more or less homogeneous in terms of class, religion, and ethnicity, caste appeared as a strong cleavage for parties to exploit at times of elections. The role of caste in elections has been well-documented in the Sri Lankan political party literature (Wriggins 1963; Woodward 1969; Kearney 1973; Jupp 1978; Jiggins 1979; Jayanntha 1992; Peiris 2014).

Under the two-party system, the relationship between the party and the caste of the voter became a complex but powerful one. Neither of the two main parties completely depended on the votes of one caste group, nor did they appear to represent the interests of one caste group. Instead, both parties started mobilising their voters through highly sophisticated patron–client networks spread across each electorate (Jiggins 1979; Jayanntha 1992; Peiris 2014), approaching multiple groups defined by the boundaries of caste identity (Peiris 2014). Caste received further impetus under the first-past-the-post electoral system as no political elite could win an electorate without successfully wooing the minority caste groups within the electorate. In this context, both parties maintained a second tier leadership from minority caste groups within each electorate to facilitate the process of bargaining with caste-based block votes in return for various patronage goods (Jiggins 1979; Jayanntha 1992; Peiris 2014).

Peiris (2014) illustrates how party organisers took care to include representatives from all caste and religious categories in their networks. Further, he demonstrates how in the field locations of Kelaniya and Weligama, both the leading parties (the UNP and the SLFP) had made sure to include party acti­vists from the Bathgama (labourer) and Wahumpura (jaggery) castes in addition to the Govigama majority, because they were the second and third largest caste groups in the area (Table 1).

This inclusion by the party organisation of grassroots leaders from each prominent caste group was not intended to reflect the interests of those caste communities, but rather to provide the assurance that they too were equal with the other caste groups (who were hierarchically superior) in an electorate, in terms of receiving patronage that the party organiser distributes. (Peiris 2017: 190)

As a result, caste interests were represented in national politics in the form of material benefit distribution. Hence, existing caste stratifications were accommodated, rather than diluted, in this system. Unlike during the early democratisation process, during this stage reaffirmation of caste not only served the political system but also the individuals within each caste group (or rather individuals of the numerically larger caste groups) (Tables 2 and 3). As a result, caste continued to assume considerable salience despite economic practices that symbolised caste stru­ctures continuing to weaken under emerging economic realities (de Zoysa 2013: 96).

With the introduction of the proportional representation system, party competition (even the intra-party variant) further intensified and attention to caste identities has grown proportionately due to the new rules of the game. Under this system, parties have to compete in geographically and population-wise much larger electoral districts. In addition, preferential voting allows three choices per voter within the selected party or coalition. This new electoral system promotes more intra-party competition than inter-party competition. As many candidates from the same party or the same coalition contest for the same electoral district, politicians have invented new cost-effective ways in reformulating the party–voter nexus (Peiris 2017). On the one hand the proportional representation system weakens the politician’s accountability to any geographically-bound community. On the other, voters too have been empowered by the availability of multiple choices, whereby they can maximise the utility of voting. Electoral engineering has become a complicated and sophisticated affair under the proportional representation system as it has paved the way for complex yet fragile political patronage network systems within electorates where politicians and voters both act as critical agents in the making of such networks.

Under the previous first-past-the-post electoral system, caste emerged as an important factor for both the politician and the voter during the time of election. However, not all caste groups received the same attention as the electoral system favoured the numerically bigger ones. The proportional representation system has further expanded the role of caste by granting agency for even numerically smaller caste groups dispersed across the electoral district to negotiate a better deal for their group during elections (de Zoysa 2013). Under the proportional representation system, politicians of downtrodden caste groups also can exploit existing caste rivalry within the district to amass sufficient votes to get elected to Parliament. Particularly, the lower caste groups and their social mobility are ­dependent on capitalising on their caste identity through means such as this. Hence, even marginal groups cannot really do away with their caste identities. They are encouraged to be part of the existing system to reap the benefits. Therefore, in the realm of politics, unlike Buddhism, what seems to prevail is a comparatively more pronounced practice of caste-based differentiation.

It appears, then, that the introduction of democracy and its interaction with patron–client traditions have encouraged the retention of caste identities among a majority of the Sinhalese peasantry. While it is true that caste may no longer exercise the capacity of defining the status of individuals to the same extent it did in the past, caste certainly remains an important factor that Sinhalese voters recognise as an intrinsic part of their identity. As a result of decades of caste-based electoral and patronage distribution practices, caste-based differentiation has continued to reproduce itself in the collective Sinhalese mindset. Sri Lanka’s democratisation process since the 1930s, then, has reproduced caste-based differentiation not only as external structures in electoral practices, but also as an internal cognitive structure of the voter, giving rise to a “social practice” and, thereby, a “habitus” within which they operate (Bourdieu 1989). In the current practice of demo­cracy, not only does the political system take note of caste identity, but individual caste groups themselves demand that such identity be electorally significant. The resultant politicisation of caste and its use as a political resource base has become one instrumental way in which caste-based differentiation has been perpetuated in Sinhalese society (Uyangoda 1998).

Some parallels to this phenomenon can be found in India. As Bayly (1999) notes, the classification of “backward castes” originally intended to empower (both electorally and thereby socially) traditionally marginalised groups, and the consequent “benefits” they were accorded (most prominently quotas in representative bodies, education, jobs, and so on), encouraged more and more groups to demand that they also be classified into such categories. This finally led to a situation where parties could garner massive popular support by promising an expansion of the scope of this word!


In this paper, we made an effort to interrogate how caste-based differentiation persists amongst the Sinhalese community ­despite the near absence of the economic structures that initially warranted caste stratifications. We examined the role of two supposedly egalitarian social structures—Buddhism and democracy—in encouraging caste-based differentiation amongst members of the Sinhalese community. We argue that both ­institutions in their social and institutional practices, as an ­unintended by-product, contribute towards reproducing caste-based differentiation among Sri Lankans, particularly the Sinhalese.

Despite the philosophy of Buddhism espousing an egalita­rian parallel society, the socialisation and institutionalisation of Buddhism in Sri Lanka over many centuries has produced a form of religion that one may call “Sinhalese Buddhism,” which has not only been structured by feudal caste hierarchies and practices, but also become an agent of reproducing caste-based differentiation in Sinhalese society. The internalisation of caste as demonstrated in the self-­organisation of the laity on occasions such as bana, funerals, and traditional processions as well as the organisational logic of the Sangha, has reproduced caste ostensibly in the realm of rituals. On the other hand, material practices such as caste-based hierarchical seating arrangements in some temples, division of labour in rituals, and sometimes the setting up of caste-designated temples also continue to structure the perceptions of devotees, resulting in what can be termed as “difference-producing mechanisms” in the guise of cultural and ritualistic practices. Thus, we argue that these practices of Sinhala Buddhism make silent but significant contributions to the reproduction of caste-based differentiation by transforming its means of sustenance from an economic to more of a cognitive basis.

Even though as a modern concept democracy promises equality, fraternity, and liberty, caste and caste-based differentiation have received new impetus especially in the practice of elections. We demonstrated how political parties, especially the two main parties, capitalise on caste divisions within Sinhalese electorates to expand their electoral bases. The patron-–client relationship, which has been the main electoral mobilising strategy since the introduction of universal adult franchise, has encouraged not only political parties but also voters to recognise caste as a powerful social cleavage. As such, one’s caste identity no longer only signifies recognition or stigma, but rather has become a means by which one could gain access to state resources through patronage
networks. In that sense, caste-based differentiation has become a more “overt” practice in politics, as compared to the realm of Buddhism. This leads to the conclusion that caste-based differentiation not only ­implies reinforcing hierarchy but also countering hierarchy as well. As in Buddhism, however, in politics too caste is reproduced not only as external structures in electoral practices, but also as an internal cognitive structure of the voter, giving rise to a “social practice” within which votes make their electoral rationale.

Despite their ideological commitment to egalitarianism, therefore, these two powerful religio-cultural and political ins­truments—either as cognitive or material structures—structure and are structured by caste-based differentiation, thereby contributing to the latter’s reproduction in the Sinhalese ­community. As is true in the case of any “habitus,” these intersecting subjective and objective conditions continue to dictate the choices, practices, and judgments of those who ­belong in that system, to the effect that caste-based differen­tiation is reproduced despite the absence of its initial economic rationale.


1 Gatherings of laymen under a hut on the compound of a temple to listen to sermons delivered by a Buddhist monk.

2 According to Mahavamsa, the island’s chief historical chronicle, prince ­Vijaya who is considered the father of the Sinhalese race, got down a Kshatriya princess from Madurai in South ­India to wed, and requested along with her 1,000 families of 18 service groups (Geiger 1912: 56–57). It is said that Vijaya refused to have his coronation without first marrying a Kshatriya princess, which would in due course ensure a royal bloodline (Geiger 1912: 56–57).

Along with Kshatriya came from India the trader caste of Vaishyas who eventually started exerting a considerable amount of influence in the island. Their influence owed largely to ­lucrative trade ties as well as their success in agriculture, a new avenue they had ventured into, inspired no doubt by the fact that the ­expansive flat lands in Sri Lanka’s dry zone and the weather conditions had inevitably made rice the staple meal. Naturally to be in control of these lands and thereby the staple grain was to be in control of significant influence. Hence, it could be that the initial strength of the Govigama caste was derived from being in control of most of the arable lands in a society where sustenance was mainly drawn from rice. “In the twelfth century, notably in the epigraphic records of Nissaņ­kamalla … there is information giving the impression that the Vaiśyas, refe­rred to therein as Govikula [Agricultural Caste], had emerged as an important social group” (Liyanagamage 2008: 176). Their ascendance in the social ladder resulted in them being “figured among the several factions which strove to capture the throne at the time” (Liyanagamage 2008: 176).

3 Such as the oft-quoted saying of the Buddha: “Not by birth does one become a Vasala (outcaste); Not by birth does one become a Brāhmaņa; by deed (alone) does one become a Vasala; by deed (alone) does one become a Brāhmaņa” (Suttanipāta: verse 136).

4 It should also be noted here that the nikaya stratifications within the Sri Lankan Buddhist order initially emerged not as a result of caste considerations, but rather of ideological disputes. The first of such divisions was between the Mahavihara and Abhayagirivihara, with the former “representing the more orthodox, conservative traditions and [the latter] the newer, more progressive tendencies” (Bandaranayake 1974: 18). In fact, the word nikaya is translated as “schools of religious opinion” (Bandaranayake 1974: 18; Gunawardena 1993: 7), which has no caste connotations whatsoever.


Bandaranayake, Senaka (1974): Sinhalese Monastic Architecture: The Viharas of Anuradhapura, Leiden: Brill.

Bayly, Susan (1999): Caste, Society and Politics in India, Cambridge: Cambridge University Press.

Bourdieu, Pierre (1989): “Social Space and Symbolic Power,” Sociological Theory, Vol 7, pp 14–25.

Chakravarti, Uma (2007): “Buddhism as a Discourse of Dissent? Class and Gender,” Religion in Context: Buddhism and Socio-political Change in Sri Lanka, J Uyangoda (ed), Colombo: Social Scientists’ Association, pp 7–23.

Dumont, Louis (1998): Homo Hierarchicus: The Caste System and Its Implications, Oxford: ­Oxford University Press.

De Jong, J W (1988): “Buddhism and the Equality of the Four Castes,” Acta Iranica, pp 423–31.

de Zoysa, Prasanna (2013): Caste Matters: Democracy, Caste and Politics in Sri Lanka after the Promulgation of the 1978 Constitution: A Case Study of Sinhala Society (unpublished MA ­thesis), Colombo: University of Colombo.

Geiger, Wilhelm (1912) trns: The Mahavamsa or the Great Chronicle of Ceylon, London: Oxford University Press.

Geiger, Wilhelm (1912): “Mahavamsa: The Great Chronicle of Ceylon” (nd):….

Gombrich, Richard and Gananath Obeysekera (1990): Buddhism Transformed: Religious Change in Sri Lanka, Delhi: Motilal Banarsidass.

Gunasinghe, Newton (2007): Changing Socio-Economic Relations in the Kandyan Countryside, Colombo: Social Scientists’ Association.

Gunawardena, Ralph (1993): Sivura Saha Nagula, Colombo: Social Scientists’ Association.

Jaffrelot, Christophe (2017): “Ambedkar, Buddhism and Democracy,” Wire, 13 December,

Jayanntha, Dilesh (1992): Electoral Allegiance in Sri Lanka, Cambridge: Cambridge University Press.

Jiggins, Janice (1979): Caste and Family in the Politics of the Sinhalese 1947 to 1976, London: Cambridge University Press.

Jupp, James (1978): Sri Lanka: Third World Democracy, London: Cass.

Kannangara, P D (1998): “The Caste Problem and the Study of Modern Period in Sri Lankan History,” Social Science Review, Studies in Social History of Sri Lanka, pp 135–48.

Kearney, R N (1973): The Politics of Ceylon (Sri Lanka), Ithaca and London: Cornell University Press.

Liyanagamage, Amaradasa (2008): Society, State and Religion in Premodern Sri Lanka, Colombo: Social Scientists’ Association.

Omvedt, Gail (2003): Buddhism in India: Challenging Brahmanism and Caste, New Delhi and Elsewhere: Sage Publications.

Perera, Punya (2018): Caste and Exclusion in Sinhala Buddhism, 7 March,

Peiris, Pradeep (2014): Catch All Parties and the Party-voter Nexus: A Study on the Nature of the Party-voter Relationship of the UNP and the SLFP (unpublished doctoral dissertation), Colo­mbo: University of Colombo.

— (2017): “Changing Dynamics of the Party–Voter Nexus,” Political Parties in Sri Lanka: Continuity and Change, A Shastri and J Uyangoda (eds), New Delhi: Oxford University Press, pp 77–99.

Queen, Christopher and Sallie King (1996): Engaged Buddhism: Buddhist Liberation Movements in Asia, New York: SUNY Press.

Ryan, Bryce (1953): Caste in Modern Ceylon: The Sinhalese System in Transition, New Jersey: Rutgers University.

Rogers, John (2004): “Caste as a Social Category and Identity in Colonial Lanka,” Indian Economic and Social History Review, Vol 41, No 1, pp 51–77.

Senaratne, Kalana (2017): “Buddhist Teachings, Ethnic Tensions & the Search for Solutions,” Transitional Justice in Sri Lanka, B Fonseka (ed), Colombo: Centre for Policy Alternatives, pp 100–19.

Silva, Kalinga (1997): Kulaya, Panthiya haa Wenas Wana Lanka Samajaya, Kandy: Community Education Center.

— (1999): “Caste, Ethnicity and Problems of National Identity in Sri Lanka,” Sociological Bulletin, Vol 48, Nos 1–2, pp 201–15.

Silva, Kalinga, P P Sivapragasam and Paramasothy Thanges (2009): Casteless or Caste-blind? ­Dynamics of Concealed Caste Discrimination, Social Exclusion and Protest in Sri Lanka, ­Colombo: Kumaran Book House.

Spencer, Jonathan (1990): A Sinhala Village in a Time of Trouble: Politics and Change in Rural Sri Lanka, Delhi: Oxford University Press.

Uyangoda, Jayadeva (1998): Studies in Caste and Social Justice in Sri Lanka: Caste in Sinhalese Society, Culture and Politics, Colombo: Social Scientists’ Association.

— (2012): “Local Democracy and Citizenship in the Social Margins,” Reframing Democracy: Perspectives on the Culture of Inclusion and Exclusion in Contemporary Sri Lanka, J Uyangoda and N De Mel (eds), pp 29–89.

Weber, Max (2001): “Buddhism: An Atheistic and Anti-Caste Religion?” Journal of Religious Culture, Vol 50, pp 1–10.

Woodward, Calvin (1969): The Growth of a Political System in Ceylon, Cambridge: Cambridge University Press.

— (1975): “Sri Lanka’s Electoral Experience: From Personal to Party Politics,” Public Affairs, Vol 47, No 4.

Wriggins, William (1960): Ceylon: Dilemmas of a New Nation, New Jersey: Princeton University Press.

Leave a Reply