Researching a newly inscribed World Heritage Site
Clusters of massive stone jars in Laos have inspired considerable curiosity. Little is known about the people who fashioned them, while even the date they were created has not been conclusively resolved. Louise Shewan, Dougald O’Reilly, and Thonglith Luangkhoth explain what research is revealing about these mysterious megaliths.
The breathtaking, mountainous, and forested landscape of northern Laos conceals one of South-east Asia’s most mysterious and least understood archaeological cultures, known primarily for the massive stone jars they left behind. The megalithic jar sites of Laos comprise 1m- to 3m-tall carved stone jars scattered across the landscape, appearing alone or in clusters of up to several hundred. To date, it has been thought they are related to the funerary rituals of an elusive, powerful, and expansive group that existed during the Iron Age (c.500 BC-AD 500) – a dynamic period with evidence for increasing social and political complexity. The sites were brought to the attention of Western scholars by visitors and surveyors from as early as the late 1800s.
Significant research commenced with the pioneering expeditions by Madeleine Colani (1866-1943), a formidable French geologist and archaeologist from the École Française d’Extrême-Orient. Colani excavated at the now-famous site of Ban Ang (today known as Ban Hai Hin, and Site 1) and documented some 20 other sites in the region. Her work uncovered human skeletal remains and artefacts including ceramic vessels, stone and glass beads, spindle whorls, iron implements, bronze jewellery, ceramic ear-discs, and ground-stone objects. In some cases, human bone and glass beads were reportedly found inside the jars, but they were otherwise empty. Since Colani’s missions and impressive subsequent two-volume publication, archaeological excavation and research has been limited, due in part to the unexploded military ordnance contaminating the region (see www.uxolao.org/en), a tragic legacy of the Vietnam War.
Work resumed in the 1990s, conducted by Eiji Nitta and Thongsa Sayavongkhamdy. This resulted in the recovery of similar material culture to that reported by Colani, but it was also observed that chipped stone pavements surrounded the stone jars and, in some instances, buried limestone slabs mark the location of primary and secondary burials, while boulders marked the location of burial jars. A series of radiocarbon dates were also produced, which ranged widely from 7552 BC to AD 1214. Following this research, UNESCO-supported surveys in the 2000s – undertaken by a team led by Julie Van Den Bergh and Samlane Luangaphay – resulted in the creation of a detailed database of 58 GPS-located jar sites, complete with jar group numbers, and a list of an additional 26 unexplored sites. Since then, the number of recorded sites has steadily increased as a result of more recent survey and excavation. The total now stands at over 100 sites, with many more likely to exist, concealed by dense forest. In July 2019, UNESCO recognised the significance of these remarkable relics, when it added the Plain of Jars to its list of World Heritage Sites.
The jars of Laos are not the only such megalithic receptacles known in the wider region. Seemingly comparable examples are found in Assam, in Northeast India, where both the geography and environment are strikingly similar to those of upland Laos. Another intriguing example of megalithic jars occurs in Central Sulawesi, in Indonesia, where large stone vats called kalambas bear some resemblance to the jars of Laos.
Return to Ban Ang
In 2016, an international Lao–Australian team conducted excavation and survey at Ban Hai Hin (Site 1), creating a detailed inventory of the stone jars, burial-marker boulders, and sandstone discs. Each megalith was accurately geolocated, while their appearance and state of preservation were carefully registered to aid ongoing conservation measures. Site 1 is dominated by a limestone cave that Colani suggested functioned as a crematorium. The mouth of the cave faces over 316 stone jars, arranged in five groups. A small hill to the north is home to the largest and most-impressive jars, which are known as Group 1. The most-numerous set (Group 2) lies at a lower elevation, and was arranged in a crescent in front of the cave mouth. Three smaller groups of jars to the south make up the remainder of the site, bringing it to nearly 30ha in extent. Interspersed among the jars were more than 20 sandstone discs. Although these are often thought to be the jar lids, this interpretation is questionable, as there are many more jars than discs, and the discs have also been found to mark burial places. In addition, there were 308 boulders that are natural, but exotic – meaning they do not naturally occur at the site – all of which were captured with drone-acquired imagery. As noted, the boulders appear to be tombstones of sorts, heralding the position of buried ceramic funerary jars.
We opened three trenches around the Group 2 jars, which revealed the remains of at least 18 individuals, who were found in a variety of settings. Most individuals proved to be secondary burials, which could take the form of disarticulated bundles of bone or remains placed in ceramic vessels. For the first time, though, we also found what seems to be a primary extended burial, on the basis that the skeleton was completely articulated. More than 60% of the deceased were infants and children, almost half of whom died at the foetal stage or in early infancy. Extrapolating from this suggests that the area immediately around the Site 1 megaliths could, potentially, contain the remains of more than 8,000 individuals. Isotope analysis is now under way to investigate how these people lived and what sort of diet they enjoyed. These results will also be compared to baseline environmental isotopic data to explore where the interred individuals may have spent their childhoods.
Radiocarbon dating of charcoal samples produced dates spanning 8200 BC to AD 1200, with the majority suggesting that activity around the Group 2 jars occurred between the 9th and 13th centuries AD, making it considerably more recent than the Iron Age dates previously reported for the site. However, only one of these dates (AD 1163-1125) was produced by material retrieved from beneath a jar, so further evidence is needed to confirm their exact date of placement, as the burial activity is not necessarily contemporary with when the jars were installed. Currently, a programme is under way to secure further radiocarbon and optically stimulated luminescence (OSL) dates from sediments sealed beneath the jars, to try to pin down when they were positioned.
Another vexing issue is how such massive megaliths, some weighing in excess of 30 tonnes, were transported from their quarry across the rolling landscape. A soon-to-be-published geochronological study of jars from Site 1 has confirmed that they probably came from a quarry at Phukeng, some 8km away. This raises several, as-yet unanswered questions related to the method of megalith movement. Possibilities include the completed vessels being hauled from the quarry using a rolling pulley system or perhaps drawn by elephants or buffalo, but whatever the solution, or solutions, manoeuvring the megaliths was clearly a significant logistical and organisational undertaking.
Despite the fame of Ban Hai Hin, it is unusual among the jar sites in Xieng Khouang province, because most jars were placed not on the plain, but at higher altitudes perched on ridges and hill slopes. Prior to 2017, none of these had been extensively investigated. With this in mind, the research team focused its attention to the north-east of Site 1, on a remote, mountainous region. This is home to Site 52, which is situated at a height of about 1,300m near a Hmong village, in rugged and forested terrain. Some 415 jars, along with 219 discs and boulders, are laid out in six discrete groups among the trees and thick undergrowth. Several of the discs are decorated with carved zoomorphic figures.
Our aim was to collect data from Site 52 that could be compared to Site 1. To that end, eight areas were selected for excavation based on the presence of jars, discs, and boulders. The excavations revealed significantly less material culture than was encountered at Site 1 and – aside from a single tooth – no skeletal material. Despite the sparsity of artefacts, features such as the presence of sandstone chip pavements and exotic limestone blocks were closely comparable to elements associated with burials at Site 1. As the absence of human bone at Site 52 could be explained by it being eaten away by the acidic soil, it seems reasonable to suggest that the site was also mortuary in nature.