Rise of Social Complexity in Southern Sri Lanka an archaeological analysis
The emergence of social complexity is one of the less researched fields of study in Sri Lanka. A special project aimed to research upon the origin of urbanism in southern Sri Lanka had been initiated between 1999 and 2004. 300k2 area of the lower basin of river Kirindi Oya, one of the five major rivers in the region was filed-walked using systematic stratified sampling technique. 127 hitherto unknown archaeological sites that represent 4 micro-environmental zones have been identified during the first field season. During the second field season 8 test excavations were conducted to obtain chrono-stratigraphic sequence of the sites. Explorations and excavations have yielded a substantially rich assemblage of material culture of the people who inhabited the area for nearly two millennia since 900 BCE. Four major aspects have been outlined as the impetus of the rising complexity of the region. They are (1). Longstanding practice of agro-pastoralism; (2). Use of iron as the principle medium of technology ; (3) Gradual increase of the inland trade and (4). Active participation to the long-distance trade. This paper provides an archaeology explanation to that developmental process.
In 1999, an archaeological project was launched by the Postgraduate Institute of Archaeology, Colombo, with the funding support of SIDA/SAREC and the scientific collaboration with the Department of Archaeology and Ancient History of the Uppsala University, to research upon the urban origins in Southern Sri Lanka. The prime objective of the project was to understand the development of cities and urban life in the region, which is elaborated by the ancient chronicles as a metropolitan center of the country, and the existing above ground monumental remains that reflect high esteem of art and architectural quality that would have been an integral part of a urban life way, which is still a terra incognita.
The term ‘southern Sri Lanka’ used here to denote the geographical region between river Walave in the west and the river Kumbukkan Oya in the east, including the coastal plain and the fertile hinterland consist of the alluvial tracts of the lower basins of other three rivers namely Malala Oya, Kirindi Oya and Manika Ganga, where the earliest historical settlements flourished in the early first millennium BCE (Somadeva 2006). This area was historically known as Rohan or Ruhuna, which was described as a separate administrative territory and one of the major cosmopolitan centers of the country since the late first millennium BCE.
The history of the archaeological research in this region was initiated in the latter part of the 19th century. Early pioneers were the P. Goldsmidt (1876) and E. Muller (1883) who traveled through difficult tropical jungle tracts in the area to copy ancient inscriptions and especially H.W. Parker who did several excavations at the ancient urban mound of Magama in Tissamaharama in the early decade of the 20th century (Parker 1910). Despite its early date, Parker’s work still exist as a point of departure of the archaeology of Ruhuna, which is now being widened its scope by the excavations conducted by the Sri Lanka-German joint project (Wheissar & Wijepala 1993; Somadeva 2006) and the Urban Origins in Southern Sri Lanka project conducted by the Postgraduate Institute of Archaeology.
The territory of historical Ruhuna was expanded several times in different periods. At least by the 10th century CE it was flourished up to the southern frontier of the central highlands from the southern coastline including the south and south-eastern dry zones and in the west, up to Kaluganga and to Mahavali Ganga in the east. Despite as the archaeological evidence has explicitly elaborates the expansion of the agro-pastoral settlements in the area were confined to the dry zone of Ruhuna.
This area consists number of characteristics that had been influenced the agro-pastoral settlements including a vast plain moist by five perennial rivers and their alluvial tracts, the mineral rich foot-hills such as Beralihela and Binkemhela, the lagoons and the coastal strip with scrub jungle vegetations where much favourable for pasture.
The geology of the area consists of metamorphic granite gneisses of Achaean origin (Fernanado 1988) and Miocene sedimentary lime stones. The coastal belt consists of Pleistocene sands. In several places, the gneiss bedrock is exposed at Veherakema (N 06° 17¢ 50¢¢, E 81° 16¢ 55¢¢) Kirinda and Pôtthevelagala (N 06° 15¢ 28¢¢, E 81° 20¢ 19¢¢) (Plate 3.9). Eroded remnants like Galapitagala, Beralihela and Binkemhela on the northern boundary of the study area and also the Chendugala isolated inselberg situated in the floodplain of the left bank of Kirindi Oya comprise elevated heights. These formations have been directly associated with the man-made landscape of the area. Some of those natural rock shelters were used by Buddhist monks, as habitation sites directly after the introduction of Buddhism c.300 BCE. Several inscriptions recorded from the cave roofs have been published elsewhere (Paranavitana 1970). The quartz intrusions of the northern extreme of the area associated with the lower Mänikgaňga basin would have been an impressive feature of the regional geology. From the beginning of phase III, the settlers of the area exploited these rocks for a variety of purposes especially for making sculptures and ornaments.
The variations in the annual rainfall are the most influential factors on the fauna and flora, which is unique to the area. The entire rainfall regime of the region is generated by the two monsoons. Heavy rains fall in the area during the northeast monsoon in the period between October and December. The area around Tissamahārama receives an average annual rainfall of 1159 mm. The southwestern monsoon does not bring much rainfall to the area and it is limited to an annual figure of 112.5 mm. During drought conditions, the required monthly effective rainfall figure of 100mm for cropping in the area is not obtained (Wickramatilleke 1963:31). During the period between May and September, the area experiences severe dry conditions of varying durations accompanied by warm dusty winds (Jameson 1941). As Wickramatilleke has pointed out (1963:30) the permanent settlements and the successful cultivation of crops (the only crop cultivated during the present day dry season in the area is rice (Oriza sativa indica)) is dependent entirely on sufficient supplies of water from large irrigation tanks scattered over the area. This situation has been slightly improved by the large-scale Lunugamvehera reservoir project directly associated with Kirindi Oya, which is now able to supply more water to the area.
Some variations are noted in the mean annual rainfall figures of the different rainfall stations situated within the area. This is obvious in the mean annual rainfall figures obtained between 1946 and 1950 from Uduwila and Weerawila stations, which are situated within the study area about 6.8km apart. This mosaic pattern has influenced the regime of effective rainfall of the area and together with other environmental factors such as soils and relief would have directly affected the agricultural activities of the area during the last 2000 years.
The soils of the research area belong to the Reddish-Brown earth (RBE) zone of the island. These soils are derived from the parent rocks of metamorphic origin. Several micro-soil regions can be observed within the present area. Most prominent types among them in the area are the solodized Solonetz soils and the alluvial soils. Solodized Solonetz soil represents the most part of the area especially in the western and eastern parts of the out side the floodplain. The some parts of the floodplain consist of alluvium.
The lowest part of the undulating terrain of the local landscape consists of other subtype of Reddish Brown Earth called low-humic gley soils. These are poorly drained, moderately fine to fine-textured soils suitable for rice cultivation with or without irrigation (Moorman and Panabokke 1961). Most of the area beyond the floodplains in both banks of Kirindi Oya consists of these soils and associated with the slash and burn cultivation which is still practiced today.
The floodplain area mainly consists of alluvium brought by the floods of Kirindi Oya. The entire area of rice cultivation is associated with this soil horizon. Furthermore, owing to the richness of the ferruginous mineral content, it is well suited for the production of pottery, bricks and other earthenware. The exploitation of the silty soils in the floodplain continued from the first occupation of the farming communities in the area until the present day.
According to the data gathered during the fieldwork, tropical dry evergreen thorn forest is most used by the present inhabitants of the area for timber and rarely for wild fruits. The dominant thorn forest tree variety is Manilkara hexandra, locally known as Palu. The timber of this species is very hard and still is used for beams, piles, ploughs, and sleepers etc. by the local people. Bees in particular are attracted to this tree for nesting (Worthington 1959:310). The coastal-strip of the area is dominated by scrub vegetation mostly consisting of Andara (Dichrostachys cinerea) and some varieties of succulent (Euphorbia nerifolia).
Apart from the above, cotton (Gossypium sp.) could also have been important. During the first phase of the fieldwork several dozen cotton plants were observed in different locations. Nowadays this variety is not cultivated on an industrial scale. It is reasonable to argue that in the historical period cotton would have been cultivated for the weaving industry. One of the cave inscriptions found at a site of Veherakema situated within the area dating to 250 BCE mentions a merchant’s guild (Paranavitana 1970:696a) which was then engaged in the craft of weaving (Siriweera 1978). Cotton was the main raw material for the production of thread for weaving at that time. Several inscriptions also dating to c. 250 BCE found in different sites in the island mentions craftsmen called pehekara. This is the Prakrit derivation of the Sanskrit word peshakara, which means ‘weaver’. Some chronicles written in about the fourteenth century also refer briefly to cotton chenas (slash and burn cultivation plots) (Devarakshita ed.1957: 13).
The lack of grasslands along the coastal-strip and the existence of the sand dunes suggest little grazing potentials in the coastal area. Present-day herders take their cattle to the northern part of the research area or in the coastal area where limited areas of grasslands are available.
32 species of mammals have been found in the forested parts of the area (Jayawardhana 1993). Among them, six species have been identified as game animals. These are wild buffalo (Bubalus bubalis) , Sambur (Cervus unicolor), Spotted dear (Axis axis ceylonensis), Wild boar (Sus scrofa), Pangolin (Manis crassicaundara) and Monitor Lizard (Varanus bengalensis). The poachers of the area still hunt these animals rarely but hunting wild animals is banned by the Wildlife Conservation Department.
36 species of birds can be seen in the area. Most of them are seasonal and migratory. Among the endemic species there are two, which were important for the humans through out the history. These two are the Jungle fowl (Gallus lafayetti) and the Peafowl (Pavo cristatus). Peafowl has been hunted and its cooked flesh is much appreciated. Sometimes it is smoked and mixed with ghee when eaten.
The sea and also the artificial reservoirs of the area provide rich sources of edible fish. 9 saltwater fish species and 3 freshwater fish species have been identified. The lower part of the Kirindi Oya is dry for most of the year owing to the construction of the Lunugamvehera reservoir and the lower reaches of Kirindi Oya are no longer a rich habitat for the freshwater fish species. The presently available freshwater fish in the reservoirs are modern varieties.
The Micro zones
The micro-zone concept used here was developed on the basis of considering the natural settings of each zone of the area. This includes the distribution of soils and vegetation, as well as geomorphology, topography, natural drainage and the availability of a range of humanly exploitable resources. The results of the excavation of the eight selected sites in the area show that there was a great diversity in the resources used for subsistence between the micro-environmental zones.
Analysis of a SPOT 7 satellite image of the area with thirty-metre pixel resolution was used to differentiate between distinct landforms and other physical characteristics of the 7 landscape in the research area. The research area can be divided into four different microenvironments:
(1). a mountainous landscape in the northern part, with gravelly Regosols and thorny forest.
(2). flat terrain in the outer floodplain dominated by solodized Solonetz soils supporting scrub jungle vegetation
(3). a coastal strip with sandy Regosols and mixed vegetation.
(4). the floodplain with rich alluvial soils and a mixed vegetation.
127 sites were identified during the first field-season in summer (June-August) 1999 (Map 4.1). The majority of the archaeological sites identified within the core area are human habitations. Several sites fall into the category of non-occupational locations such as stone quarries, iron production sites and man made reservoirs. These sites are functionally diverse and belong to several phases between 900 BCE and 1400 BCE. The correlation of the C14 dates with the material and spatial data suggests seven major cultural phases in the LKB from the early agro-pastoral village to the latter quarter of the mature urban phase.
- Phase I – Early agro-pastoral village phase (900 – 500 BCE)
- Phase II – Early agro-pastoral – early urban transitive phase ( 500- 350 BCE)
- Phase III – Early urban phase (350 – 250 BCE)
- Phase IV – Mature urban phase [early quarter] (250 BCE –350 CE)
- Phase V – Mature urban phase [middle quarter] (350 –600 CE)
- Phase VI – Mature urban phase [latter quarter] (600 – 1000 CE)
- Phase VII – Urban decline (1000-1400 CE)
Eight sites were excavated (Map. 1) in 24 weeks from June to September 2000 and 2001. The aim of the excavations was to check the depositional environment of each site, to 8 understand its developmental stages in relation to the rise of the regional urbanism and to characterize the cultural developments in each of the micro-environmental zones of the research area. The selection of sites for the excavations was based both on an assessment of archaeological potential and on the micro-environmental zones.
There are several criterions that can be utilized to describe the origin and the development of social complexity in Ruhuna. Some of them are archaeological evidence and the rest are literary and epigraphical information. We hope, in its wider sense, the harmony of those two may provide a comprehensive framework for our purpose of reconstructing social complexity in southern Sri Lanka.
Archaeological evidence consists of artifacts obtained from stratified excavations in 8 different locations in the area and it provides a substantial foundation to reconstruct the degree of material engagement of the people through time. The artefacts reflect a marked diversity in their quality and quantity. This aspect could be assimilated with the series of available 14C dates to show how they were organized towards emerging social complexity.
Cultural Phase 1 ( 900-500 BCE) -The early farming village phase is ceramologically represented by the presence of the earliest Black and Red ware pottery together with two other wares i.e. Plain red ware and Black ware. Eight vessel types are represented in this phase.
Besides the earthenware, our knowledge of the material culture of this phase is meagre owing to paucity of evidence. Beads are however, abundant and the only other representative variety of artefact available. The beads of phase 1 are confined to three specific types i.e. barrel disc (clay), cylinder disc (clay) and short barrel (clay). A metal rod (kohl stick) made out of a copper alloy was reported as a surface find in the ancient settlement at Tambarava.
The characteristic cultural manifestations of this phase are the megalithic burials. Two such burial grounds have been recovered in the outer floodplain. The architecture of those burials clearly emphasizes an accumulation of social power. The harvest yielded from the extensive rice-fields developed in the outer floodplain might have stimulated the beginning of social differentiation.
The faunal remains unearthed from the excavation at Tambarava show that hunting was significant during phase 1. Three wild and one domesticated species were identified.
In phase 1, the association between the settlement and the small reservoirs is the characteristic landscape setting of the early farming villages. During the early part of this phase, it is likely that slash and burn cultivation prevailed. The present scrub jungle vegetation of the outer floodplain may be the outcome of the extensive clearance of the dry evergreen forest of the area at that time. The distribution of the ancient rice-fields in the outer floodplain suggest that most of the slash and burn cultivation plots had been converted into rice-fields in a later period of the same phase.
Six C14 dates obtained from two excavated sites furnish a chronological frame for this phase. Context 3, context 2 and context 1 of the excavation at Konvälana, context 15, context 13 and context 12 of the excavation at Ellegala are examples. No precise date has been produced by the excavation at Tambarava, but the artefacts and the date proposed by the relative chronological sequencing suggest that context 1 and context 2 of the settlement at Tambarava can be ascribed to phase 1.
Phase II – The most crucial period of the cultural development of the LKB was the early farming – early urban transition. It can be placed between 500- 350 BCE. The availability of the earthenware categories of this phase shows a marked difference from the previous phase. The use of BRW is reduced considerably and there is a remarkable increase of the use of PRW.
The number of pottery types has also increased during this phase compared to phase 1. 16 types appeared during phase 2. A piece of Red polished ware (RPW) is reported in context 11 of the excavation at Ellegala can be considered as a new ware type that appeared for the first time in the LKB.
The diversity of other artefacts also increased considerably. The appearance of the clay disks in context 23 of the excavation at Kirindagodana suggests that the use of tokens in economic transactions emerged during this period. The exploitation of marine resources for non-subsistence purposes such as conch-shells for bead making and bangle were also occurred. In context 17 of the Kirindagodana excavation, a fragment of shell-bead was reported. A bead made from Garnet from context 5 of the Polbindivala excavation, shows that mining activities were initiated during phase II. Three bead types occur in this phase with oblate disc (clay) type appearing as a new type in addition to the previous barrel disc and short barrel.
The appearance of the post-fire graffiti marks on potsherds is a marked characteristic that reflects the pre-urban and urban transition in the LKB. Several such graffiti marks have been discovered from the excavation at Kirindagodana.
Phase III (350- 250 BCE) – This phase is characterized by the initial urban manifestations of the LKB. The KAVA excavations in the ancient urban mound at Akurugoda have yielded several dates lying between 400-300 BCE for the basal cultural levels of the site (Wheissar et al 2001). This indicates that the development of the urban centre at Akurugoda began at the end of phase II.
A new pottery ware, Red painted ware ( RPW) appeared in phase III. The use of BRW further decreases during this phase. The most widespread was the Plain red ware ( PRW). 56 pottery types were recovered from the phase 3 levels of Kirindagodana.
The development of a site as a port at Kirinda in the southern shore was significant during this phase. The basal levels of phase III at the site have yielded several fishhooks that confirm the early association of the site with fishing activities.
The total number of bead types used during this phase increased to thirteen. Ten new types occurred including circular (stone), long cylinder (clay), short truncated concave cone (clay), standard cylinder (glass), long truncated bi-cone (glass), short cylinder (clay), short truncated convex bi-cone (glass), standard cylinder with two concave ends (glass), oblate (clay) and long barrel (ivory, clay).
The distribution of the early Brahmi inscription within the core area has pointed out that the urban communities in the area adopted a full-fledged script during phase III. Buddhism began to flourish in the area not only as a new religion, but also as a new intellectual tradition. The emergence of the Painted Red Ware pottery and the decrease of the prevailed Black and Red ware tradition suggest an assimilation of a new cultural group into the area. A few Rouletted ware shreds reported from the urban mound, during the excavations of the KAVA archaeologists also show the presence of new pottery wares in phase III. The presence of the new cultural traits such as Buddhism and the Brahmi script has pointed out that the immigrant group was affiliated to the north Indian region. However the assimilation of immigrants and the host population seems a mutual one and cannot be inferred as a domination of a single side. The existence of the Rouletted ware in phase III explicitly shows that contacts with south India were also maintained during the early urban phase.
In a broader perspective, the cultural development of the early urban phase of the LKB can be seen as an interaction of two particular streams i.e. the local iron using agropastoral cultural tradition and the inspiration caused by the influx of communities via long-distance commercial linkages of the Indian Ocean and the Indian sub-continent.
Phase IV (250 BCE-350 CE)– This phase marks the urban period proper in the LKB. The characteristic feature of this cultural phase is the emergence of the active engagement in 12 foreign trade. The presence of two roman coins and glass beads in context 12 of the excavation at Kirinda port, glass beads in context 8 of the excavation at Kirindagodana and the glass beads reported from context 5 of the excavation at Ellegala indicated the emergence of foreign trade. The bead types used during phase 4 further increased up to nineteen. The new types available are short cylinders with two convex ends (glass), long convex cone (clay), standard barrel (clay) convex truncated bi-cone disc (clay) collared (stone), melon (glass), segmented (metal) and standard truncated convex bi-cone (bone).
A few new pottery ware types came into light during this period in addition to the distinctive foreign wares. Several pottery sherds identified as ‘mica-coated ware’ reported from context 9 and context 5 of the excavation at Ellegala. A deteriorated punch-marked coin and a Gajalakshmi coin from the surface of the excavated area of the mound at Kirindagodana suggest that those artefacts derived from the mature urban levels of the sites. (The upper levels of the stratification of Kirindagodana are from urban phase 1 (early quarter)).
The excavation at Kirinda port suggests that 30 pottery types were used during this phase. Nine of these are new in comparison to phase III. The new types emerged were the types 19, 22, 25, 36, 39, 56, 88, 89 and 99 (see ‘pottery’ in chapter 6).
The initial construction of some of the ancient architectural edifices situated within the core area is invariably associated with phase IV. Historical chronicles describe a number of Buddhist monasteries that were constructed by the political leaders of the area during the period between 250 BCE and the beginning of the present era.
The earliest forms of the urban tanks such as Tissavava, Yodhakandiyavava and Debaravava were also constructed during the early part of phase IV and subsequently enlarged in the latter part of the same phase.
In many discussions of the rise of the social complexity in a given context, population growth takes its position as an important parameter and also its prime mover (Steward 1949; Sanders and Price 1968; Carneiro 1970; Goody 1977). The stimulants given by population growth in a given society can vary. This topic is controversial and archaeologists face a number of difficulties and theoretical uncertainties in providing population estimates from the fragmented physical remains and the highly modified landscapes.
An attempt has however been made to evaluate the population growth of the LKB during the 700 year period from 600 BCE onwards.
The methods employed to reconstruct the ancient population figures in archaeology are inspired by both human geography and social anthropology. Among them the rank-size rule is widely used ( Vining 1955; Haggett et al 1977) and archaeological applications are promising (cf. McIntosh & McIntosh 1993).
The rank-size analysis method is adopted here and shows three levels of population dispersion within the area of the LKB. The first is associated with the city center at Akurugoda and subordinate satellite urban settlements. Level two of the population dispersion can be seen in some sites along the bank of Kirindioya. Two possible reasons might be suggested to account for high population along the riverbank. Those are the attractiveness of the riverine trade and the alluvial rich levees situated along to the riverbanks. A comparatively low population was dispersed within the other parts of the area.
Other than the application of rank-size rule, another way of looking at the regional population growth of the area during a specific period in the past has been tested. It is the critical population reconstruction suggested by Allan (1949:14-15).
Population growth in the LKB was clearly associated with the intensification of agriculture and the development of the irrigation systems.
From 600 BCE to 300 BCE the irrigation system was mainly confined to small village reservoirs. The graph shows that there was a slight rate of change of the regional population initiating in the period around 300 BCE but of course such estimates are uncertain.
This change might be accounted for by a number of reasons. Agriculture was intensified during that period including the exploitation of new micro ecological resource zones (see chapter 4). Upgrading the efficiency of the existing irrigation system to meet new demands were important measures taken by the local community to intensify agriculture.
After circa 300 BCE, there was apparently a gradual increase of the population. During this period the acceleration of the technological advances in the irrigation system continued. Emergence of the cascade system connecting the small village reservoirs considerably improved water storage and circulation. Canal systems were required to connect the reservoirs and these enabled the distribution of water to more lands. It has been argued in chapter 4 that the most influential development that occurred in the regional settlements in the LKB basin was the emergence of the cascade system. It was crucial not only technologically but also socio-politically. It seems that by the time of cultural phase 1, the interconnection of the reservoir systems (cascades) were almost completed. The increased efficiency of water management system had a positive impact on the growth of the regional population. More forested lands were probably cleared for new agricultural fields. In a long-term perspective, these deforestation activities might have been resulted in negative impacts. The disappearance of the forest cover might have accelerated evaporation of the soil moisture. The content of the organic nutrients added by leafs become gradually reduced. Tilling and burning can loosen the soil and it could be a reason for the soil erosion. Clearance of forests have resulted a loss of habitats of animals. In sum, such were crucial for the sustainability of the human settlements during the subsequent periods.
The cascade system catered more to the wet rice cultivation than the slash and burn cultivation. The annual yield of the rice cultivation was more productive and relatively secure than the yield obtained from dry cereal cultivation. However, the slash and burn cultivation practice seems to have continued throughout different phases of the historical settlement development in the area. Other than rice (Oriza sativa indica), chronicles mention about finger millet (Eleusine coracana), cotton (Gossypium sp.), sesame (Sesamum indicum) and sugarcane (Saccharum officinarum) as other crops cultivated in the slash and burn cultivation plots.
The population of southern and southeastern Sri Lanka seems to have increased after c. 300 BCE (phase III) when the urban characteristics of the area first become apparent. As suggested in chapter 4, the proliferation of the sites into the floodplain and the northern hilly sector of the area suggest a dispersion of population throughout the LKB during this period.
The estimates of the population in the area show a cumulative growth up to 300 CE. If the level of resource availability, especially the carrying capacities of the lands available in the area for agriculture, is exceeded by the density of the regional population, then the dynamic of the ‘population pressure’ could emerge. But the present case in the LKB does not signal any over population threat affecting the area until the mature urbanism becomes apparent in the period between 250 BCE and 350 CE.
In a discussion of the population growth of the early period of the urbanism in the LKB, the influx of an alien group of people cannot be ignored. Several archaeological indicators confirm that a group of people who carried a new cultural tradition appeared in the area around 250 BCE. The appearance of the fully developed Brahmi script, Northern Black Polished ware (NBP) pottery, punch marked silver coins (PMC) and Buddhism as a religion, signifies the emergence of this new cultural tradition. (The impact generated by those newly emerged cultural traits on the society of the LKB is discussed below). Those traits were culturally dominant but cannot be regarded as the sole influential factor on the growing urbanism in the area. The amalgamation of two cultural traditions, i.e the iron using early farmers and herders and the newly emerged literate group can be regarded as an important factor, which accelerated the ongoing urban process in the LKB with a new impetus.
The migration of a new group of people into an area could effects society considerably but it does not immediately influence the growth of the regional population. However, some socio-cultural practices they brought such as new technology can be significant for growth of the local population on a long-term perspective.