In Maritime Southeast Asia, archaeological investigations of old historical period settlements really began in the 1970s. Before that, with the exception of the excavations conducted by Tom Harrison in Sarawak between the late 1940s and the mid-1950s,^[1] this type of research was sporadic. In the Malay Peninsula, the first to be mentioned is Ivor Hugh Norman Evans, who in 1931 excavated the Kuala Selinsing site (3rd century BCE – 10th century CE), in the Perak region.^[2] A few years later, Horace Geoffrey Quaritch Wales undertook excavations at Kota Tinggi and Johor Lama (Quaritch Wales 1940, 60–3) in Johor.^[3] In 1961, Alastair Lamb (1961) conducted excavations at Pengkalan Bujang (11th to mid-14th century CE) in Kedah.^[4] On the east coast, on Tioman Island, off Pahang state, remains of settlements dating to the 11th century onwards were excavated in the 1970s (Martin 1985; Kwan and Martin 1985). In Brunei, archaeological investigations of settlements dating to the historical period probably started at the beginning of the 1950s at Kota Batu with a follow-up in 1968 (Harrisson 1970), while excavations were undertaken the same year at Sungai Lumut, also in Brunei (Harrisson and Shariffuddin 1969). Archaeology in Singapore is discussed in detail elsewhere in this volume. In Indonesia, as early as the 1910s, excavations were conducted in Trowulan, which shelters the remains of the capital of the Majapahit kingdom, but they were mostly devoted to religious structures. In Sumatra, the first significant archaeological excavations on old settlements from the historical period date back to the 1970s, with research work undertaken by Edwards McKinnon in the Province of North Sumatra, mainly at Kota Cina (fig. 1). At approximately the same time, archaeological investigations started on the west coast of the same province, especially in the Barus area. The 1970s also saw the first archaeological research in South Sumatra province, mostly around Palembang, related to the history of Śrīvijaya.

For the last quarter of a century, research on old settlement sites dating to the historical period has been particularly active in North Sumatra Province, in the framework of French-Indonesian scientific cooperation, with financial support from the Consultative Commission for Archaeological Research Abroad under the French Ministry of Foreign Affairs. Intensive excavations and multidisciplinary studies were initially conducted on the west coast, precisely in the Barus region: Lubok Tua (1995–2000), a settlement site dating from the 9th century until the end of the 11th century CE;^[5] Bukit Hasang (2001–2005), a neighbouring site dating from the 12th century until the beginning of the 16th century CE.^[6] After 11 years of research in Barus, the team moved inland to Padang Lawas, a region known since the mid-19th century for its numerous Hindu-Buddhist remains. There, a multidisciplinary research project took place near Binanga, specifically at Si Pamutung (2006–2010), where a settlement dating from the 9th century until the 13th century CE was uncovered in association with a religious complex.^[7] These archaeological investigations undertaken between 1995 and 2010 brought a wealth of new data and hypotheses regarding the ancient history of the northern part of Sumatra for the period covering the 9th century until the 15th century CE. It was within this new context that, in 2010, the team decided to move toward the Straits of Melaka, specifically to the Kota Cina site. This major settlement site has indeed not been given the attention it deserves since pioneering research was conducted there in the 1970s.

The present village of Kota Cina is located at 03°43’ N and 98°39’ E, in the kelurahan of Paya Pasir, kecamatan of Medan Marelan, today part of the northern suburbs of Medan, a city of over two million inhabitants, and the capital of North Sumatra province. The site is also less than ten km away from the harbour of Belawan on the Straits of Melaka.

As analyses of the results are still underway, this article focuses on the aims of the project, on the context of the recent excavations, on methodological aspects regarding the study of the finds, with a few preliminary results, and offers some comparisons with sites recently excavated in North Sumatra.

Previous research

John Anderson, an officer of the East India Company, was the first to mention the place-name Kota Cina, during a survey he conducted in the Strait of Melaka in 1823 (Anderson 1826, 294). Apart from a carved stone collected there by a Dutch civil servant 60 years later,^[8] Kota Cina fell into oblivion until the early 1970s when Edwards McKinnon and Luckman Sinar rediscovered the site (Milner, Edwards McKinnon and Luckman Sinar 1978; Edwards McKinnon 1984, 38, 40). Surveys followed by excavations undertaken by Edwards McKinnon himself from 1974 until 1977 in three areas (totalling a surface area of some 137 m²), then by archaeologists from the National Centre for Archaeological Research of Indonesia (approx. 60 m²) in 1977 and 1979 (Satyawati Suleiman 1976, 9, 25–6) uncovered several types of finds. It was also in 1977 that John Miksic conducted a geomorphological survey and stratigraphical study at Kota Cina, suggesting the presence of estuarine mud and the existence in former times of a stream dividing the site in two (Miksic 1979, xvii, 131–54, 281).

These initial investigations led to the recording and discovery of several remains (Edwards McKinnon 1984). Eight remains of brick structures were uncovered, sometimes in combination with stones. Edwards McKinnon suggested the existence of two shrines dedicated to Śiva, showing architectural influence from Sri Lanka and South India, a Buddhist vihāra (monastery), a pendopo (terrace with a superstructure), a stūpa, and three pedestals. He mentioned the uncovering of several inscribed artefacts, including one half of a soapstone object inscribed with Chinese characters, two gold leaves also inscribed with Chinese characters recovered at the base of a brick structure, as well as a number of small rings bearing the inscription ‘om’. Miksic observed stylised Old-Malay (or Pallava) characters on numerous earthenware sherds, interpreting them as pseudo-inscriptions with a decorative function (1979, 158, 192).

In the 1980s, the opening of a sand quarry at Paya Pasir, to the north of the site, led to the discovery of boat or ship timbers as well as large quantities of Chinese ceramics. A team of archaeologists managed to examine what was left of the finds and concluded that they belonged to at least two vessels built according to the Southeast Asian lashed-lug construction technique: a medium-sized vessel, 15–25 m in length, and a larger type. The Chinese ceramics were dated between the 11th and the 14th centuries CE (Manguin 1989, 205-209). Today, the area is submerged in the waters of Lake Siombak.

Besides remains uncovered during surveys and excavations, chance finds of stone sculptures made in various circumstances since WW2 need to be mentioned here. A dozen stone sculptures have been recorded. They include two seated Buddha images in a Cōḷa style (Satyawati Suleiman 1976, 25, no. 36–7 p. 51; Edwards McKinnon 1977, 33; Milner, E. Edwards McKinnon and Luckman Sinar 1978, pl. 1; Suleiman 1981, 17–18; Edwards McKinnon 1984, 18, 66–9). Edwards McKinnon has suggested that both statues are images of Amitābha dating to the beginning of the 12th century CE, and were made in Tamil Nadu (1984, 69–70; 1994, 17–20). In 1979, villagers discovered a headless four-armed standing statue and the lower part of a female deity. Edwards McKinnon identified the first as a Viṣṇu image and suggested a common origin with both Buddha statues. While remaining unsure about the identification of the second deity, he placed its production also in Tamil Nadu (1984, 66, 68, 72, 75; 1994, 20, 22). The other finds of stone sculpture include two yonis, a lingga, a square base, two pivot stones, and a small granite pillar base (Edwards McKinnon 1984, 46, 51, 54–5, 61–2). Moreover, at least four small bronze images were recovered at Kota Cina before the first excavations, including two Buddha images, a deity in a South Indian style (Satyawati Suleiman 1976, 25), and a small image which disappeared before it could be documented (Milner, Edwards McKinnon and Luckman Sinar 1978, pl. 3; Edwards McKinnon 1984, 47, 61, 92).

The fieldwork conducted in the 1970s uncovered a large quantity of various types of finds. Excavations undertaken in one sector (110 m²) by Edwards McKinnon yielded some 477 kg of earthenware sherds, as well as some 260 kg of Chinese stoneware and porcelain sherds. Other finds included organic remains (timber, resins, various plants, and mostly faunal remains - 207 kg), Chinese copper coins,^[9] eight coins issued in Sri Lanka dating to the turn of the 13th century CE, a number of circular bronze discs, a bronze lamp with attached suspension chain, a bronze vessel, three bronze covers, a bronze ring set, two bronze fish hooks, a bronze pommel, a fragmentary bronze bell, two small gold pendants (or ornaments), two gold rings, five gold beads, some twenty fragments of thin gold leaves or foils, four lengths of fine-drawn gold wire, a gold cylinder, iron knife blades, an iron spearhead, numerous pieces of iron slags, a tuyère, a touchstone, more than 200 glassware shards, more than 200 glass beads, 38 stone beads, as well as several peg-like pointed artefacts carved from shale (Edwards McKinnon 1984).

No further archaeological excavations were conducted at Kota Cina until 2010 when archaeologists from the regional branch of the National Centre for Archaeological Research uncovered parts of a brick structure. At around the same time, the building of a small on-site museum with an adjacent pond yielded a large quantity of artefacts, later exhibited in the museum.

The early estimates regarding the surface area of the Kota Cina site varied widely, from more than 25 hectares to more than 100 hectares (Edwards McKinnon 1977, 20; 1984, 48; Miksic 1979, 214). After several dating adjustments since the 1970s, the most recent time bracket suggested by Edwards McKinnon (2009, 126, 133) regarding the ancient occupation of the site, mostly based on imported stoneware and porcelain, is "end of the eleventh century – end of the thirteenth century, or beginning of the fourteenth century”.^[10] While Edwards McKinnon (1984, 81, 362-363; 1996, 89) interpreted Kota Cina as a cosmopolitan settlement, highlighting the presence of Tamil and Chinese communities, Miksic at the same time favoured the idea of a settlement where the majority of the population was of local origin (Miksic and Yap 1992, 61). Edwards McKinnon proposed that due to its location in the Straits of Melaka and its access to natural resources of the hinterland, Kota Cina was a link in a network of trading places extending from the Red Sea to Quanzhou in South China, a network controlled by a Tamil trade guild, while relations were also probably established with Kedah (Manning, Edwards McKinnon, Treloar 1980, 103-104; Edwards McKinnon 1987, 87; E. McKinnon 1994, 23, 80; 1996, 97). Its appearance would result from a shift of the centre of gravity in the region from Barus on the west coast toward the east coast (Edwards McKinnon 1984, 81). Moreover, Miksic suggested that the site had established relations with East Java, from where it received spices harvested in Maluku (Miksic and Yap 1992, 74). According to Edwards McKinnon, besides its commercial functions, Kota Cina was also a Śaivite and Buddhist centre (Milner, Edwards McKinnon and Luckman Sinar 1978, 36; Edwards McKinnon 1984, 48, 58-9, 66, 68, 71, 79, 361).

A New Archaeological Approach to the Kota Cina Site

During several visits to the site since the 1990s, we became aware that the context had made it much more difficult to undertake excavations compared to the 1970s when Kota Cina was a small and sparsely populated village. Threats to the viability of the site for archaeological research were real, in the form of the multiplication of fish ponds as well as rapid housing and road intensification. In 2010, we considered the idea of a series of evaluation field projects, in order to prepare, in case of positive results, for the launch of an archaeological program along the same lines as the program conducted previously in Barus and Padang Lawas, while seeking to implement new approaches.

We will also revisit the question of the role of Kota Cina in the political and economic contexts of the Straits of Melaka, as well as its interactions with the Indian Ocean, the South China Sea and Java. The question of the political nature of Kota Cina has already been raised in the 1970s through several interesting hypotheses which need a reappraisal today, in the light of data and ideas on the history of the region accumulated during the last four decades. It was thus suggested that Kota Cina was, if not the capital, at least a component of the kingdom of Aru. The problem is that foreign sources seem silent regarding the northeast coast of Sumatra during almost the entire period of the Kota Cina settlement. Furthermore, if Kota Cina proves to have been abandoned at the end of the 13th century, this event would coincide with the earliest known mention of the place name Aru in Chinese sources. The history of the Aru polity itself has yet to be thoroughly investigated. As this place name occurs along the eastern coast of Sumatra between Langkat to the north and the mouth of the Barumun River to the south, two areas separated by some 250 km, its history has probably been marked by shifts in its centre of gravity. Lastly, there is the question of the abandonment of Kota Cina, which may coincide with the emergence of Singapore. Is this shift pure coincidence or the consequence of drastic changes occurring in the Straits of Melaka at the time? All these questions will, of course, benefit from the results of the research conducted recently in North Sumatra and in Singapore.

Fieldwork research at Kota Cina took place between 2011 and 2016, in the framework of a French-Indonesian archaeological project supported by the École française d’Extrême-Orient and the Consultative Commission for Archaeological Research Abroad under the French Ministry of Foreign Affairs.

A systematic pedestrian survey was undertaken as thoroughly as possible at the beginning of 2011 to obtain basic information regarding the present situation (fig. 2), to collect surface finds, and to draw the first topographical map of the site. A first overall picture of the site seen from the surface emerged at the end of this initial phase. Three short yearly diagnostic excavation seasons took place between 2011 and 2013 (fig. 3). Encouraging results led to three longer seasons from 2014 until 2016. All the fieldwork was conducted with the help of dozens of villagers, as well as students from Universitas Indonesia, Universitas Negeri Medan and Universitas Sumatera Utara.

Excavations in Kota Cina faced the usual problem with this type of site in North Sumatra: privately owned land plots, often belonging to several owners. At Kota Cina, this land issue has become more acute over the past several years due to the large number of land transactions, usually with the goal of converting fallow or cultivated plots into building lots (tanah kaplingan), making these areas therefore inaccessible for archaeological excavations. In such a context, our program turned rapidly into a rescue excavation. Moreover, for the last 50 years, the site has supplied Medan antique dealers, an unwelcome competition for archaeologists. Another problem is related to the nature of the site. Kota Cina presently borders nipah palm-mangrove swamps with numerous waterways. As the site is situated just above sea level with a tidal water table approximately 0.60 m below the ground surface. The use of water pumps proved necessary in many digs.

Previous analyses and hypotheses regarding the history of Kota Cina were mostly based on two trenches excavated in one area (110m²) in the 1970s. We therefore decided to conduct surveys and excavations all over the site, as systematically as possible, in order to get a more accurate overall picture of the site, leading to a better understanding of its history. The main aims were: to refine the chronology of the site occupation, to reconstruct the spatial evolution of this occupation, and to determine the origins, lifestyles and activities of the people living there. If the presence of Tamil, Chinese, and even Sinhalese communities is highly probable, did they live there together or successively? Furthermore, was the settlement organised along community lines, as described in local and foreign written sources referring to later urban centres in the region? What about populations of maritime Southeast Asia? It is difficult to imagine that they were completely absent. Here again, the study of the archaeological materials in order to identify their origins, site distribution, and dating, will be decisive in trying to answer these questions. Anticipating a wealth of organic and faunal remains, we decided to launch thorough paleoenvironmental and archaeozoological studies. Regarding the first, the idea is to try to reconstruct a picture of the landscape at the time of the ancient occupation of the site, in particular through the study of phytoliths. The second offers an opportunity to get a better understanding of the diet at the time. Both are probably a first for archaeological sites belonging to the historical period in Sumatra. Based on the remains uncovered so far, Hindu-Buddhist cults were certainly important. A number of hypotheses have already been suggested regarding the origins of the four stone statues uncovered on the site. It is clear that this corpus shows no similarities with the contemporaneous Padang Lawas statuary, both in terms of material and style.

Preliminary results

Since the 1970s, brick remains were known to exist on the site, but a systematic survey was never conducted. We chose a simple method, using several 1.5m-long iron rods to detect bricks underground. Wherever possible, these tests were conducted at regular intervals (usually one meter).

One of the first results of this survey has been to ascertain the basic plan of at least one brick structure excavated in 2013 when three sections of brick walls appeared at a depth of some 0.60 m. These walls are not oriented according to cardinal points and do not show any foundation trench. The inner and outer sides are made mostly of whole bricks, while the inner part of the wall is filled with broken bricks (fig. 5). No trace of mortar or any other binder has been identified. Two empty circular areas were very probably filled with wooden pillars supporting a timber-frame superstructure. It should be noted that two of the walls are not exactly parallel, raising the question of their belonging to the same structure. Finds recovered during the digging of these remains do not provide any clue regarding the function of this or these structures. The iron rods survey revealed the presence of two additional similarly oriented brick walls (fig. 6). If these four brick walls really belong to the same building, this compartmented structure would therefore have measured at least 14 m x 7 m.

Another interesting result of the systematic iron rod surveys came out some 260 m to the south, in the vicinity of a small Chinese shrine, where brick scatters were already noticed in the 1970s, and brick remains excavated in 2010. The survey led to the delimitation of a 40 m x 40 m area, very rich in bricks down to a depth of 1.20 m. This is very probably a complex consisting of several structures enclosed within a brick wall, possibly the main religious complex of the site. Unfortunately, despite several attempts, permission to excavate was not given by the landowners.

In the southern part of the site, levels of brick fragments were observed in one trench. Systematic surveys using iron rods conducted in 2015 and 2016 did not yield conclusive evidence of the presence of structures using permanent materials, except close to a recently built house. Very probably, most of what was left of the old structure had been destroyed during its recent construction or is now buried underneath, being therefore out of reach for excavation for decades to come.

At Kota Cina, accumulations of household waste resulting in shell middens of various thicknesses and sometimes showing successive layers are frequent in the northern part of the site (fig. 7). The maximum thickness reaches 60cm. Examples of successive layers have been observed in several trenches. These shell middens usually yield large quantities of finds. One of the aims of systematic soil core samplings, mostly along the North/South and East/West cardinal axes of the site (basically, at an interval of 20 m; some reaching a depth of four meters), was to detect these shell concentrations. Another aim was to quickly obtain stratigraphic information useful for the palaeoenvironmental study.

Compared to Barus and Padang Lawas, the presence of wood remains makes Kota Cina an exceptional archaeological site in northern Sumatra, as its humid environment helps preserve organic materials. The excavations have uncovered mainly two types of remains of wooden structures in the northern part of the site: truncated posts and wood scatters (fig. 8). Some pieces still bear sugar palm fibre cords (ijuk). More than 20 samples of wood have been selected to try to determine their species or genera. At least one of these wood remains is probably a boat timber which, by means of AMS radiocarbon dating, yielded a dating in the bracket mid-12th to mid-13th century CE^[11]. Traces of upright wooden structures were also identified through the uncovering of post holes.

Pedestrian surveys, systematic surveys using iron rods to detect brick concentrations, as well as more than 200 soil core samplings, some reaching a depth of four meters, were undertaken to complement data from the excavations. During the six seasons from 2011 until 2016, 69 trenches were excavated, the majority measuring 2 m x 2 m. The total excavated surface area is 482 m2 with an excavated volume of some 464 m3. The minimal surface area of the old settlement is estimated at some 25 hectares (ca. 300 m x 850 m). These excavations uncovered several types of structures and features briefly introduced here.

Another interesting feature is what we interpret as an “occupation surface”, consisting of one or several successive layers showing a number of artefacts in a flat position. One trench thus revealed ten successive layers of presumed occupation surfaces between depths of 22 cm and 75 cm below the present ground surface. As it is clear that these layers are not sealed, it is therefore impossible to state that each occupation surface is really posterior to the level situated immediately below. Nevertheless, they may reveal interesting spatial and chronological data.

Table 1. Main types of finds (2011-2016 excavations)

With more than 164,000 sherds for a weight exceeding 1.5 tons (Table 1), the earthenware corpus of Kota Cina is probably the largest archaeological collection of its kind gathered so far for a historical period site in Sumatra. The study of this collection, carried out with the help of a number of students from Medan, Jambi and Kendari universities, includes petrographic analyses, a comprehensive typology, as well as a statistical analysis of intra-site distribution of major types.

Like earthenware, the stoneware and porcelain finds probably constitute the second-largest archaeological collection gathered so far for a terrestrial site in Sumatra. As was the case for Barus-Bukit Hasang and Si Pamutung, apart from a comprehensive catalogue, a statistical analysis of intra-site distribution will also be carried out to try to reconstruct the spatial evolution of the ancient occupation. At the present stage of the analysis carried out by Bing Zhao (CNRS, Paris), with the help of a number of students from Medan and Jambi universities, the majority of the wares date back to the 12th and 13th centuries. Sherds dating back to the 11th century and to the first half of the 14th century are also present, as well as a few sherds which could date back to the 10th century. This provisional result is in accordance with the most recent dating hypothesis suggested by Edwards McKinnon.

With more than 160 kg of faunal remains belonging to various genera and species, Kota Cina’s corpus is probably unparalleled in Indonesia, if not in Southeast Asia, for a site of this period. The progressive siltation and the presence of mangroves explain the exceptional conservation of these finds, currently analysed by Stéphane Frère (INRAP, Paris).

The size of the new corpus of Chinese coins, collected in excavations, is similar to the surface finds collected by Edwards McKinnon in the 1970s. While he observed that Chinese copper coins still littered the surface of the site at the time, today such surface finds are rather rare. This evolution is probably linked to the traffic of antiquities that has affected the site for half a century.

Of the two Sri Lanka coins collected during this program, one was found in excavation. By comparison, all eight coins reported by Edwards McKinnon in the 1970s were surface finds.

The excavations yielded some 1,000 shards of ancient glassware. Such a find is rather common in the northern half of Sumatra, and to date more than 40 sites in Southeast Asia are known to have yielded ancient glassware shards (Perret 2014). In the framework of our program, more than 80 shards from Kota Cina are currently being analysed by Laure Dussubieux (Field Museum, Chicago). A comparison with analyses carried out previously on some 180 shards from Barus and Si Pamutung should reveal interesting insights into the history of glassware circulation in Sumatra.

In addition to coins, Kota Cina’s excavations have yielded nearly 500 whole or fragmentary ancient metal objects, mainly iron and bronze. Although the vast majority of iron finds are covered with a thick gangue, which makes it impossible to know their exact shape, two main categories can be distinguished: blades and nails. Bronze fragments are usually very small, most of them belonging to containers. Several kilogrammes of iron slag and fragments of tuyeres were also found, indicating the existence of workshops on site.

Statistics on two types of finds offer the opportunity to situate Kota Cina among the other contemporary settlement sites of North Sumatra, where extensive research has recently been undertaken.

Abbreviations: KC, Kota Cina; BKH, Barus-Bukit Hasang (12th – early 16th c. CE);
PMTG, Padang Lawas-Si Pamutung (9th-13th c. CE)]>

Table 2. Comparison of densities among three sites of North Sumatra (earthenware)

Regarding earthenware (cf. table 2), the Kota Cina excavations yielded the highest densities in weight and quantity per cubic meter excavated. While the difference in the quantity of sherds is only 13 per cent compared to the sites of Bukit Hasang and Si Pamutung, the weight difference exceeds one third compared to Bukit Hasang and is multiplied by a factor of 2.7 compared to Si Pamutung. If the factor “duration of the occupation” is taken into account to obtain a density per cubic meter per century of occupation, Kota Cina (2.5 centuries) remains predominant with some 1,340g, while Bukit Hasang (4.5 centuries) records about 600g, and Si Pamutung (4.5 centuries) reveals a figure lower than 300g.

Abbreviations: KC, Kota Cina; BKH, Barus-Bukit Hasang (12th – 16th c. CE);
PMTG, Padang Lawas-Si Pamutung (9th-13th c. CE)

Table 3. Comparison of densities among three sites of North Sumatra (stoneware and porcelain)

As regards stoneware and porcelain (cf. table 3), with 112 sherds per cubic meter, Kota Cina ranks first, followed by Bukit Hasang and Si Pamutung. If the duration of the settlement is taken into account, Kota Cina shows the highest density per century (45), followed by Bukit Hasang (23), and lastly by Si Pamutung (9).

Abbreviations: KC, Kota Cina; BKH, Barus-Bukit Hasang (12th – 16th c. CE);
PMTG, Padang Lawas-Si Pamutung (9th-13th c. CE)

Table 4. Relative representativity (density per cubic meter per site) in stoneware/porcelain and earthenware (quantity of sherds)

In quantity, the proportion per cubic m of stoneware and porcelain sherds against the total number of sherds is identical in the two coastal sites of Kota Cina and Barus-Bukit Hasang, with some 25 per cent. It would be interesting to compare these figures with other more or less contemporary settlement sites in Sumatra, in the Straits of Melaka, including Singapore, or in other regions of Southeast Asia.

Analyses of fieldwork data and finds are underway, and the publication of the collective volume on these new Kota Cina excavations is planned for the second half of 2022.

At this stage of the research, we wish to emphasise several observations and reflections, which, to us, seem to be interesting for the history of the Straits of Melaka at the time, including Singapore.

Almost 50 years after its rediscovery, with the exception of Singapore, Kota Cina remains surprisingly the only major coastal settlement site in the Straits of Melaka which has been extensively excavated so far which dates from the first half of the 2nd millennium CE. In Kedah, despite very fruitful research undertaken in the 1960s, the settlement sites of Kampong Sungai Mas and Pengkalan Bujang have not yet received all the attention they deserve. The same is true of contemporary settlement sites in the Batang Hari delta (Jambi Province). Discussing the place of Kota Cina in the economic history of the Straits will remain highly speculative as long as extensive excavations have not been carried out on these major contemporary settlement sites, without forgetting those of the northern tip of Sumatra.

Research into the history of Kota Cina has perhaps paid too much attention to its religious remains, a trend with roots in the colonial archaeology of the archipelago which emphasised the study of Hindu-Buddhist remains, whether monumental, images or inscriptions. Schnitgerʼs archaeological expeditions in Sumatra, from Palembang to Padang Lawas, in the 1930s, are emblematic in this respect. This focus on the field of religion, combined with the tenuous nature of other archaeological traces, led to the non-religious aspects of the history of the sites being ignored or downplayed.

The absence of contemporary mentions of a settlement situated on a major traffic route is puzzling in comparison with the relative abundance of written sources between the second half of the 1st millennium CE and the 19th century for the partly contemporary Barus, situated on a coast generally described as hostile to maritime communications. In fact, the large number of potential trade routes on the one hand, with all the possibilities of segmentation in Southeast Asia, and on the other hand, the social complexity of the communities active on these routes, can give rise to multiple hypotheses.

One of the factors that may explain the relative brevity of the ancient occupation of Kota Cina, and more generally of the North Sumatran coastal sites of the first half of the 2nd millennium CE studied so far, is the precarious supply of coastal trading places. Later examples show that control of harvesting areas, of routes towards the coast, and of intermediate markets raised issues of multiple conflicts that might disrupt the supply of coastal markets or divert the forest products. Barus chronicles, on the northwest coast, provide insight into more or less widespread conflicts occurring in the interior, conflicts that could result in a breakdown in the flow of commodities, and even its diversion toward other trading places, such as Singkel, some 75 km to the north. Major disruptions in hinterland trade routes may have led coastal settlements to engage in piracy in the Melaka Straits.

This relative volatility of the settlements, also fueled by internal political instability and threats coming from the sea, is probably one of the characteristics that differentiates the Straits of Melaka from territories with a mainly agricultural vocation. From this point of view, the Sultanate of Johor, about which nearly 30 moves of the royal residence are identified between 1511 and 1718, is an emblematic example. But this volatile character can also represent an opportunity offered to charismatic leaders to found new polities in the Straits, as told in Malay chronicles for example.

Bibliography