Galpaya Ancient Settlement Survey the results of the first field season conducted in a newly discovered site in Rathnapura district of Sabaragamuva province of Sri
The decline of the hunter-gatherer subsistence economy in Sri Lanka is placed somewhere in c. 1000 BCE. Excavations suggest that the agriculture was prevailed in c. 900 BCE in the north-central dry zone and the southern and southeastern dry zones. However the transition between these two cultures is still obscure. Very few excavations show that it was an abrupt change and the dynamics of it is still to be questioned. A recent field survey conducted by the Postgraduate Institute of Archaeology, Colombo at a site in Rathnapura district has revealed a set of promising result, perhaps to be able to address this problem. This article presents the preliminary discoveries of the surface survey conducted during the period between August and September 2006.
This article has structured into four sections. Section one presents a general introduction including the discovery of the site together with a short account of its environmental and historical context. The general description of the site appears in section two evaluates the degree of the disturbances occurred at site. The survey methodology, the analysis and the material culture are discussed in section three. The concluding section attempts to propose a tentative chronology to the site based on the artifacts and at the end of that section an effort is made to discuss the findings within a wider geographical area of the Indian Ocean region. A natural earthen mound consisting of a dense scattering of artifacts on its surface has been reported by the Department of Archaeology in the early 2004 (see also ASCAR 1962). An excellent preliminary report about the site` has been provided by the Department of Archaeology in the year 2005 including a brief account of the historical context of the location as well (Wijerathne 2005). Valamkatuyaya has been identified as an archaeological site, which bears a great potential to carry out a controlled survey aiming at an objective of resolving the problems associated with the intra-site level surveys as well as a location that provides sufficient material and spatial database to research upon the development of social complexity in ancient Sri Lanka. Besides, Valamkatuyaya is perhaps the only fairly extensive archaeological site that has been identified so far in the north-western frontier of Ruhuna (fig.1.1). The site lies in-between the Balangoda and the Rakvana escarpments and on the right bank of Kuda Oya which is one of the feeding channel of River Walave (fig.1.2). This suggests a geographic peculiarity of the location in comparison with other similar sites in the macro environmental context that were adaptive to dry harsh environment of the south and southeastern Sri Lanka.. The other archaeological sites scattered in the immediate surrounding area, such as Pallebadda, Pollamure, Diyahinna Vihara & Kirimakulgolla and Bambaragala bear early Brahmi inscription are early historic Buddhist monastery sites.
Fig. 1. The geographical setting of Galpaya within its surrounding Early Brahmi inscription sites
The closet to Valamkatuyaya is Bambaragala, which contains four cave inscriptions (fig.1.3) of the late first millennium BCE (Paranavitana 1970:59). This suggests that places like Valamkatuyaya might have been the locations encompassed in the regional settlement network of the area at least since the early historic period. Perhaps there will be another few hitherto unknown sites in the area that could bear similarities with that of Valamkatuyaya. For example, a location called Pussatota that is situated not far from Valamkatuyaya can be cited. It is a mound but the greater part of it had been destroyed by treasure seekers.
In a broader view one can argue that the archaeological sites like Valamkatuyaya, including the above-mentioned inscription bearing sites, may signify the settlement system of the northwestern periphery of the region, which is historically known as ancient Ruhuna.
Initially the Department of Archaeology has inspected the site and reported its archaeological significance. As a consequence the Department has invited the Postgraduate Institute of Archaeology (PGIAR) to carryout further scientific investigation at the site. The PGIAR has decided to launch a project in collaboration with the Department of Archaeology to explore the site within a broad research objective.
- 1 Objectives
A set of broader objectives had been elucidated at the outset of the project formulation. All of them are addressed to the central question of the rise of social complexity in late first millennium in Sri Lanka which is one of an unresolved archaeological problem in the country (e.g. Bandaranayake 1984). A detailed account of the project objectives and its rationale is to be provided in the final report. To achieve the research goals enlighten by the key issue of the rise of social complexity, the project proceeds through several phases seeking for some other scrutinized objectives which are specific to each phase and directly involve with the main objectives of the project.
Basically the present research attempts to answer two specific research questions pertaining to the rchaeology of Valamkatuyaya i.e.(1) Why people occupied the site and how they used and organized the site and; (2) To what degree that the archaeological record at the site manifests the organizational patterns of the human behavior of the past along the time and space.
Fieldwork of phase 1 was based on two objectives. They are (1). Asses the impact caused by the post-occupational disturbances and the other site formation processes to make the present appearance of the site; and (2). Understand the organization patterns of the spatial behavior of the site reflect by the arrangement of the artifact patterns.
1.2. The area
The present site is laying between two major mountain ranges in the country i.e. the Rakvana- Bulutota massif and the Kaltota- Balangoda mountain ranges. The Kaltota Balangoda escarpment comprises several another low-height ridges namely Bambaragala, Heendodangahatenna, Ulgala and Hataramune. This low-height ranges stand as a rolling platform from the high-peak ridges called the ‘southern wall’ (Cooray 1984: 52), which divides the southern dry zone from the wet uplands.
The present site is lying in the transitional area between those two ecological zones (fig.1.4). The two mountain ranges in either side makes a valley type terrain forming a favorable environmental condition for agriculture especially rice and other drought resistant varieties of cereals like millet in and around the present site. Two varieties of millets i.e. finger millet and fox tail millet still grows in the area but farmers are discouraged to cultivate them due to the eclipse of different market oriented cash-crops.
The valley type terrain gradually becomes a vast hinterland in further south that was one of the fertile agricultural plains in the country even today. It was the major settlement area in south and southeastern Sri Lanka, which is historically known as Rohana or Ruhuna (Mv xxx).
The impact caused by this topographic setting to the regional environment is numerous. One of the notable indicators is the annual rainfall budget that (1250-1875mm) contrasts with the higher annual rainfall regime of the immediate wet zone (1875-2500mm) and the lower figures of the immediate dry zone (1250-1000mm). Especially rainfall variation makes the area as an intermediate zone in an ecological sense, between the wet uplands and the southern dry zone. We are not intended here to make a descriptive account on the environment in relation to the cultural development of the area because the environmental analysis is still in-progress.
Immediate east of the present site there are ruins of ancient runoff irrigation system, which is popularly known as the rain-fed tanks and it suggests the fact that the early farmers occupied the eastern sector of the area at least during the early centuries of the first millennium BCE. This phase of the cultural development is akin to phase I (900-650 BCE) of the settlement development of the Lower Kirindi Oya basin that has secured by a series of 14C dates (Sinclair & Somadeva et al 2003; Somadeva 2006). Perhaps this is one of the strong evidence to show the presence of a sedentary farming group in and around the area at least in the early centuries of the first millennium BCE. It is also a contrasting fact that the absence of the evidence of such a run-off irrigation system in the valley area where the present settlement mound and several other similar settlement mounds are situated (Map xxx). To make an explanation to this peculiarity needs more data on the environment and a well-stratified cultural sequence of the site.
The Mesolithic habitation of Bellan bandi palassa is situated about 5km of crow-fly distance in the northeast to the present site shows us one of the major sites of the earliest human occupation in the area (Deraniyagala 1958). The quartz implements in the natural rock shelters at Bambaragala suggest that the prehistoric communities had occupied some places around the present site, but the presence of the open-air sites like that of Bellan bandi pallassa are still unknown in the vicinity.
The evidence reported by Deraniyagala from Bellan bandi palassa show a close resemblance with some of the artifacts recovered from the surface of the Valamkatuyaya mound. Deraniyagala commented upon his findings and concluded a new development of the material culture he found. He states; “
….The presence of the semi-lunate microliths and the pebble hammer stones suggests that this was a Mesolithic culture that had evolved into the Neolithic (1958:246).
We assume that the developmental trend, which Deraniyagala had observed in his finding and the similar degree of technological advances shown by the findings of the present site had a set of complex eco-cultural motives. Such raison d’être is yet to be researched upon thoroughly after the proposed next year excavation at the site.
It is too early to infer some kind of interaction between the prehistoric Mesolithic groups and the early agricultural groups occurred in the area at a certain point in the past. However the abundance of the grindstones, querns and pestles together with finely prepared quartz implements and early BRW wares makes a reasonable arena to formulate such a hypothesis with a strong material basis. We are waiting to check the chronostratigraphy of the site in the year 2007 using couple of 14C dates and a fine analysis of the surrounding environment to provide a logical framework to this hypothesis.
2. The Site
Valamkatuyaya is a natural mound. Such are one of the common features of the undulating terrain of the Reddish Brown earth zone in Sri Lanka (Panabokka 1967; 1996). The approximate extent of the mound is 0.11km2 (length 585m x width 195m) and its artifact universe extends approximately ¾ of the total surface area. The entire surface of the mound bears a scattering of ancient potsherds. The scattering is moderately distributed and some places have high-density accumulations. Besides the earthen ware shreds, the surface artifact assemblage comprises prehistoric stone implements, grind stones, iron nails, beads of different varieties, clay disks, punch marked coins, roman coins and a considerable amount of Iron slag while forming an extremely complex artifact palimpsest .
The key attractor of the site might have been the man-made reservoir called Bisokotuvavava (Bisokotuva is the Sinhala equivalent for cistern sluice. Probably the term Bisokotuvavava is a modern connotation) situated in the western side. The distance between the mound and the reservoir does not exceed 10m shows a similarity between the layout of the present site and the ancient city mounds at Ridiyagama and Tissamaharama situate within a 40km radius. They all are adjacent to a large man-made reservoir that almost akin to that of Valamkatuyaya. The Bisokotuvava is now in a ruined state. Nearly 1km long earthen wall suggests that Bisokotuvavava was a main source of water used by the inhabitant of the nearby settlement during the time that the reservoir was functioned. However the artifacts collected from the mound, especially the prehistoric stone implements, have pointed out a fact that the reservoir is slightly later construction than the mound. Two inscriptions found in the vicinity of the mound are belonging to 9th century CE and 11th century CE are illegible and has no use of the identification of the historicity of the reservoir.
2.1. Site’s preservation
The factors of the site formation processes were thoroughly examined during the fieldwork of phase 1. The crucial among them is the destruction caused by the treasure hunters. At least half of the mound had been dug-out by them. The calculations show that approximately 5000m3 of earth had been excavated through illegal diggings during last 5 7 years. Most of the surface findings are a result of this conscious disturbance. Other alterations were mainly due to the natural processes and seem less influential to make up site into the present form. The earth mixed by the tree-fall and the activities by the burrowing animals like kangaroo rats and termites are discernible at some places.
3. Survey methodology
The northern sector of the mound was selected for the surface survey during phase 1 of the fieldwork. The land extent of xxxx meters has divided into 4 zones as A,B,C & D. Zone ‘C’ and ‘D’ was sampled using systematic sampling strategy and zone ‘A’ and ‘B’ were subjected to the random sampling using the ‘grab bag’ strategy.
Zone ‘C’ and zone ‘D’ again divided into 10x10m squares called fractions. Each fraction has sub-divided into 1x1m squares to increase the intensity of the surface sampling (see fig.xxx). The surface soil down to 4 to 6 cm was sieved using 3mm steel net following the 1×1 meter squares to collect the surface artifacts. Potsherds were weighed and the counts of the other artifacts have recorded according to their provenance. Zone ‘A’ & ‘B’ has field-walked following 10 meters transects. The provenance of the random samples was recorded as well. Sampling of phase 1 covers 22.62% of the extent of zone ‘C’ and 33.33% of the total extent of zone ‘D’.
The first level of analysis dealt with quantative data. Statistical data of the artifacts was subjected to the analysis to identify the probable activity areas of the site. To achieve this end, the statistical data of the artifact density has converted into a digital image. This image of density isopleths has explicitly reflected the concentrations and the discrimination patterns of artifacts on the surface (fig.xxx). The patterns outlined by the density isopleths might have been an outcome of various disturbances processes occurred during the recent times. It highlighted several high-peak amplitudes in a 3D representation suggesting the points of unusual accumulations are the locations of the intensive mixing of artifacts (e.g.xxxxx).
This image helps to distinguish between highly disturbed areas and the less impacted areas of the site surface. For example, according to figure xxxx the area covers by units 50 to 70 of transects 1 to 15 could be identified as less affected area of zone ‘C’ in comparison to the rest of the zone. Displacement of artifacts in an extensive scale could be inferred from units 1 to 40 of transects 1 to 10. The depression of two pits excavated by treasure hunters have been observed in the area between unit 70 and 80 of transects number 17 to 20. The high amplitude appeared in the 3D representation invariably associates with those disturbances that results a high accumulation of artifacts.
The mean value of the percentages of the pottery weights of each transects indicated in table xxxxx was subjected to a leaf-plot analysis. The resulted histogram (fig.xxx) shows a distributional trend occurs around 4% of the total weight (16 6564g) and the median (Md) of 10 values (4.76, 4.33, 4.59, 4.83, 4.22, 4.33, 4.58, 4.56, 4.01 &4.01) suggest 4.4% i.e. 7328g is the approximate probable pottery density of a 40m2 area of zone ‘C’. This figure proposes the probable area of least affected by the post-occupational disturbances and other natural taphanomical processes. Comparatively least affected area covers by transects 3, 4, 7, 9,11, 12, 14, 15, 19 & 20. The extent of this area is 400m2 . It is 50% of the total sampled area in zone ‘C’.
This identification derived from a simple statistical analysis is important to distinguish between highly altered surface segments and less disturbed zones. Further analysis such as the artifact class differentiation therefore could be focused upon least disturbed areas aiming to understand the micro level arrangement patterns of the artifacts. Perhaps such could help to infer some particular activity areas of the site.
The second level of analysis is in-progress based on the qualitative attributes of the artifacts. At this stage only a catalogue of the artifacts is presented in table xxxx below.
3.2. Material culture
Majority of the artifact assemblage collected from the site are earthenware shreds.
At this stage, fixation of a chronology for the site is a partial attempt and has a less scientific utility. Contrary perhaps such requires to portrait the archaeological potentials of the site for the purpose of planning further examinations and managing the site. Fixing a temporal line using an artifact palimpsest is theoretically polemical. The main reason for that is the difficulty of divorcing from the bias of ‘standard cultural sequence’ rooted 10 in our mental template that reflects culture specific, time-hierarchical artifact arrangements. For example we believe that in Sri Lanka, (as objectively substantiated by the research of last five decades) use of Iron preceded by the existence of Mesolithic stone implements and the emergence of pottery was contemporaneous to use of iron and the initiation of agriculture. But this generalization is time dependent and bears the insight of linear succession of the cultural progress. For instance, the three types of artifacts namely finely made quartz microliths, querns, pestles and grind-stones and BRW recovered from the surface of the present site do not tell us exactly whether their chronology is akin to our generalized time hierarchy (successive order from stone to pottery) or it has other mode of temporal expression that break this order. Nevertheless a tentative chronological scheme presented below (fig.xxx) is essentially relative and a momentary formulation that appreciates the above-mentioned theoretical uncertainties.
During phase 1 of the fieldwork it is only dealt with the surface materials. Most of them are derived through the post-occupational disturbances occurred at the site in different periods but they invariably represent the diversity of the material culture of the subsurface deposit.
The material manifestations of the five cultural phases outlined are as follows. Phase I – Hunter-gatherer/early agriculture transition or a ‘techno-culturally intermediate phase’ exemplified by the evidence of plant processing devises such as grind stones, querns and pestles together with finely made microliths and the presence of pottery)
Phase II- Early agricultural-pastoral village phase represented by BRW pottery, use of Iron and terracotta circular disk beads)
Phase III- Agricultural-pastoral village and early urban transitive phase represented by relatively sophisticated technology of the terracotta beads and the enhanced quality of the BRW pottery and the relatively extensive use of iron implements).
Phase IV- Urban phase exemplified by the presence of Brahmi letters, imported glass beads and Red polished ware pottery that suggest the links with long-distance trade, presence of punch-marked coins and the practice of an organized religion (Buddhism) exemplified by a miniature terracotta Buddha image)
Phase V- Urban continuum represented by late roman copper coins and the special pottery such as mica coated ware etc.)
It is highly recommended to confirm this phasing scheme through a series of absolute dates.
4.1. The wider context
The first field season has produced a greater quantity of material culture of the ancient inhabitants of the site. All of them are unstratified findings therefore temporally disoriented. In spite of that, we observed two possibilities of having a comparison between the similar materials found in different parts of the Indian Ocean region with that of the present site.
In one hand the assemblage of grind-stones, querns and pestles together with finely made thin quartz implements and pottery are comparable with that of unearthed in different location in the mainland India. Greater similarity could be observed from phase I and phase II materials excavated at Bagor in northwestern India. The Bagor excavators have commented upon their findings and states; ‘
…….Essentially the culture of phase II …….are best explained as due to cultural contact between the stone age population of Bagor and the agriculture- based village populations of Mewar and the neighboring Malwa (Lukas et al 1982: 14).
Among the other several dozens of sites, notably bear the characteristics of Ahar culture (Sankalia et al 1969) and the excavated site at Piklihal (Allchin 1960) had produced similar assemblages. Bagor phase I has three dates ranging from 4480 BCE to 3285 BCE and phase II from 2765 BCE to 2110 BCE. At Piklihal the phase called ‘lower Neolithic’ has a range of date from 2000-1250 BCE and ‘upper Neolithic’ ranges from 1250-650 BCE. The Ahar dates lies between 1900 BCE and 1200 BCE. The dates of Bellandandipalassa go back to 6500± 700 cal.BP. The two dates of 2050 and 2335-1825 cal. BP. dates of the same site was rejected by Deraniyagala (1992:701) based on a personal information about the contaminated nature of the sample. The Mesolithic openair habitation site excavated at Matota in the northwestern coast of the island has a range of date from 4310-3520 cal. BP. Unfortunately we do not have a full account of the artifacts recovered from the early levels at Matota to make a comparison.
If we take the lower date of Bagor and the upper date of Piklihal, the period covered show a new dynamic of transition of the Mesolithic cultures to a new direction. This was happened in different scales and different intensities and different techno-cultural modes as well. Our argument here is, Sri Lanka was a counterpart of this regional (or probably trans-continental) dynamism generated in the latter part of the first millennium BCE. In this sense the evidence derived from the sites like Valamkatuyaya shears significant cultural characteristics of this important regional cultural trajectory.
In the other hand the artifacts of the historic period recovered from the present site show an intensive link with different geographical regions in the peninsular India. Especially the abundance of late roman copper coins, which are a common occurrence in the archaeological sites in the greater part of the south and southeastern Sri Lanka including the present site, suggest probably the trade affiliations with the ports like Arikamedu (Somadeva 2006). Few glass beads recovered at the site perhaps can be ascribed to the genre of Indo-pacific glass beads. However, more empirical observations are required to a further confirmation. During the historical period, the site and several other places surrounded it seems played a key role of the urbanism in the south and southeastern part of the island (Somadeva 2006b) as well as linking Sri Lanka with the distance lands in the Indian Ocean region through trade and various other cultural discourses, which are yet to be researched upon.
(The authors are indebted to Professor Nimal de Silva, Director of the Postgraduate Institute of Archaeology for his genial support extended to the project from its outset. Thanks should also 13 goes to Dr. Senerath Dissanayake, Director General of the Department of Archaeology of Sri Lanka for his concern to give us the permission to carry out the preliminary survey at the site. The first field season of project was funded by the Postgraduate Institute of Archaeology of the University of Kelaniya, Colombo).
Bandaranayake, S. 1984. Problems in the classification of the historical societies of the Indian Ocean
region in a taxonomy of social types: a review of the Sri Lankan situation. A paper presented at the second
international conference on Indian Ocean Studies held in Perth, Western Australia, 5-12 December 1984.
Cooray, P. G. 1984. An introduction to the Geology of Sri Lanka (Ceylon). Colombo: National Museums
of Sri Lanka publication.
Deraniyagala, P.E.P. 1958. An open air habitation site of Homo Sapiens balangodensis. Spolia Zeylanica
Panabokke, C. R. 1967. Soil Science, The Soils of Ceylon and use of Fertilizers. Colombo: The Ceylon
Association for the Advancement of Science.
- Soils and Agro-Ecological Environments of Sri Lanka. Colombo: National Resource Energy
and Science Authority.
Lucas, J. R., V.N. Misra & K. A. R. Kennedy 1982. Bagor and Tilwara: Late Mesolithic Cultures of
Northwest India. Vol. 1. The human skeleton remains. Poona: Deccan Collage Postgraduate and Research
Deraniyagala, S. 1992. Prehitory of Sri Lanka. An Ecological Perspective. Colombo: Department of
Allchin, F. R. 1960. Andhra Pradesh Government Archaeological Series No. 1, Piklihal Excavations.
Hydrabad: The Government o Andhra Pradesh.
Sankalia, D. H., S. B. Deo, & Z. D. Ansari 1969. Excavations at Ahar (Tambavati). Poona: Deccan Collage
Postgraduate and Research Institute.
Somadeva, R. 2006a. Sri Lanka Andhra Sub system?: probably another way of thinking. Journal of Indian
Ocean Archaeology 2 (xxxxxxxx).