1Atauro is a small island, covering an area of around 140 km2 with a population of 10,000 in 2022. Both volcanic and coral, it is located between the large island of Timor and the volcanic islands of Indonesia.
2When asked about the ancient sites that dot the territory of their village or hamlet, locals are quick to mention the fortified enclosures that once served as refuge from various enemies. The term used on the island to designate these strongholds is kota, a word of Sanskrit origin used widely throughout the Malay world to refer to a defensive site. These fortified enclosures were visited between 2014 and 2023, guided by informants who provided on the spot explanations or additional historical information. Selected sites with typical or remarkable layouts were surveyed with a GPS. In this article, we analyze these historical remains to understand the context that led to their construction, and to identify elements pertinent to a recent ethnohistory of the island of Atauro.
- 1 Ili Hngara Bitauk is the full name of this site, often referred to as Ili Hngara or simply Ili in t (...)
3On Atauro, fortified sites are reached by climbing one or more dry-stone walls, often over two meters high. These walls are built of coral limestone or volcanic tuff depending on the geological substratum. They may feature chicanes or narrow corridor entrances. These fortified enclosures are generally located on a high, defensive position such as the top of a cliff (Heuknan, Taklo...) or an easily accessible area on a steep slope, taking advantage of the incline below as a natural defense (Ili Lor, Ili Hngara Bitauk,1 Ili Ara, Kota A'i...) (figs. 1, 2, 3). Still other sites, such as Ili Tungas, a simple fortification on a slope, were chosen because of their remoteness or altitude from the coast. One site (Ili Hatu Tutun) is limited to the summit of gigantic limestone blocks, therefore naturally protected by its challenging access. Fortified sites range between those primarily protected by the natural topography of the site with or without walls added, and veritable strongholds enclosed by several rows of dry-stone walls.
Fig. 1 – Kota Ili Hngara Bitauk © Guillaud, IRD/Soares da Silva, SEAC
Fig. 2 – Kota Ruma No © Do Rêgo Soares, SEAC/Guillaud, IRD
Fig. 3 – Taklo © Do Rêgo Soares, SEAC/Guillaud, IRD
4It is not always easy, when visiting these sites today, to get a clear idea of what the ancient structures represented. This may be due to successive redevelopments, as at Ili Hngara Bitauk where rows of walls were built in several phases, or to the dismantling of walls to salvage stones as at Kota Hahan. Frequent earthquakes in Atauro may also have played a role as in the case of Ili Ralam where parts of the walls collapsed some forty years ago.
5The surface area of these complexes ranges from a few hundred square meters (Ili Hatu Tutun) to a few hectares (Abak), and even some fifteen hectares for the largest site (Taklo); however, the majority are in the half-hectare to one-hectare range.
6Fortified sites show various characteristics, although present environmental factors can affect interpretation. It is sometimes difficult to discern the internal organization of a site due to the vegetation covering it. Walking proves difficult on coral lapiaz surfaces, leading us to suspect that erosion may have carried away topsoil. However, some features can be discerned. Ili Hngara Bitauk has a grassy area near one of its entrances, described as a place for ceremonies or dances. Some sites feature piles of stones designated as tombs. Kota Aʼi culminates on a ridge with a view over the island and out to sea.
7On Atauro, in the course of various visits to the field, we have identified more than 20 of these fortified enclosures (fig. 4) — each almost systematically associated with a different nearby settlement —, but there are probably more which remain to be documented. This remarkable density is comparable to that noted in the far east of Timor by Lape (2006: 287) in the subdistrict of Tutuala, where 17 ʻstone structuresʼ have been recorded on an area equivalent to that of Atauro, and in the district of Lautem where McWilliam (2019: 250; 2020: 136–137) recorded some thirty similar sites, covering areas ranging from 0.5 to 3 hectares with cactus formations serving as protection. In Atauro, ancient rattan plants provided a defense system at Kota Wataran (Rattan Fort).
Fig. 4 – Location of documented fortified sites on the island of Atauro © Billault/Guillaud, IRD/SEAC
- 2 Raw material and know-how needed for pottery-making have been identified only in the Arlo area. Pot (...)
8On Atauro, water is a major constraint and a number of sites, especially in the north, are located far from any presently known source of drinking water. Giant clam shells (Tridacna gigas), probably used for storage, are found on many Atauro sites, and for a long time, water was carried from springs using sections of large bamboo.2
9Such fortifications are reported by numerous authors in most of the Lesser Sunda Islands. Their frequency has fueled an archaeological debate opposing primarily two hypotheses (see Schapper 2020) on the dating and reasons behind fortification construction: one stresses demographic growth and the ensuing problems of competition for resources (Earle 1991, O’Connor et al. 2012); the other highlights issues of climate change which would have led to major social changes (Field 2008, Field and Lape 2010). Other hypotheses combine both explanatory registers. The model proposed by Lape and Chao (2008), for example, dates the appearance of the earliest fortifications around the 12th–14th centuries, during the droughts caused by the El Niño phenomenon, and their perpetuation or densification over the following centuries in relation to social and historical processes (Lape and Krigbaum 2020: 49), especially slave trade.
10Our aim here is to reconsider these fortified structures through the local representations and knowledge. Such an approach allows us to understand the place these structures occupy in the memory of inhabitants, as well as their patrimonial nature. This article deals not only with fortified sites, but also with the various conflicts that, in the minds of informants, are inseparable from these places and justified their construction. This shift of research perspective from archaeology to history opens up different avenues for investigation, such as the local context in which these structures were built and used, and the identity of the groups involved. The history of local conflicts and the monuments that evince them is important to understand the history of relations between groups, but also and above all to outline the history of the island’s settlement. The defensive sites discussed here are thus not only archaeological and historical sites, but also tangible objects that evoke social history and relations. Conflicts are still remembered, as they are at the root of important (re)compositions of local societies in different parts of the island.
11To investigate these fortified enclosures and conflicts, we have drawn on the knowledge received by informants from previous generations. Having started with the very probable hypothesis that oral tradition is a way of representing and transmitting history, we mobilized certain elements of Atauro’s ethnolinguistic groups’ and clans’ founding narratives, to evoke and illuminate the history of these places and of the island. A wealth of information was gathered in the course of over a hundred interviews of varying length, and initially delivered without chronological concern but with significant recurrences. The sections that follow are an attempt to organize this information.
12The compilation of oral information from different parts of the island highlights a period in local history described as ʻolder.ʼ At that time, outsiders would have arrived in large boats “with two or three sails,” to barter with the inhabitants. These exchanges are confirmed by the composition of the clan’s wealth partly made up of goods — such as swords, jewels, plates, pottery, textiles, gongs — brought in by foreigners which informants say were bartered for wax, sea cucumbers, dried fish, and various other sea products (source: Ili Timur, Doru). However, the deviousness of these foreigners is often emphasized (particularly in the village of Makili) because, aside from trading, they also committed murder, carried out massacres, and conducted kidnappings. When the origin of these traders-turned-enemies is known, this deviousness is associated, for example in Makili, with people from Hera and Buton, Makassar and Maumere; in Akrema, with people from Ternate; in Atekru and Doru, with people from Sulawesi.
13In addition, bartering was carried out with traders not only from distant archipelagos, but also nearby islands. Atauro’s nearest neighbor is Liran, which, according to oral tradition, is a fragment of Atauro detached by one of the arrows shot from the summit of Manukoko (source: Akrema). Liran is associated with several episodes in the history of Atauro. One narrative (collected in Vila-Maumeta) describes how a resident of Makili named Meharek went to Liran to exchange betel nuts and wax from Atauro for dried fish. Following a dispute, Meharek destroyed the property of his Liran allies, leading to a larger conflict. Meharek was chased back to Makili, where his pursuers were repelled (more on this later).
14Other information refers to traditional alliances between Atekru (West Atauro) and the large island of Wetar, and between Makadade and the islands of Alor and Wetar. These alliances were reactivated, for example, at times of crisis when islanders needed to seek refuge on neighboring islands (source: Arlo) as when part of the population of Arlo fled to Liran because the Portuguese tried to levy taxes. Yet inter-island relations were also marred by looting and theft perpetrated by people from Alor, Kisar, or Wetar (source: Makili).
15Scientific literature, most notably in linguistics, touches upon these traditional alliances. According to Hull (1998: 3–4 note 12), the Liran dialect is intelligible to Rasua speakers in northern Atauro. For Grimes and Edwards (forthcoming), the languages of Atauro are related to those of Wetar.
16The literature also provides information on slave trading activities and dates their extension or intensification in the region to the 17th century. Hägerdal (2012: 244, note 93) for example notes that the princes of Alor extended their trading activities as far as Wetar and Atauro around 1680, “to obtain slaves and wax for the Company” (VOC, the Dutch East India Company).
17Frequently mentioned by informants, the kingdom of Makassar, converted to Islam in 1605, was already an important merchant state by 1630. Over the “long sixteenth century” it extended its influence as a maritime power as far as Timor, Solor, and Bima (McWilliam 2020: 144; Gunn 1999: 77). The early 17th century also coincided with the intensification of human trafficking operations in the Lesser Sunda Islands region (McWilliam 2020: 146): with the Islamization of Java and the extension of Sharia law prohibiting slavery, Muslim kingdoms were seeking their slaves beyond Islamized areas, for example in Kisar, east of Atauro, from where slaves were destined for Banda’s nutmeg plantations.
18More generally, as Hägerdal (2012) reminds us, the early 17th century saw three exogenous powers — European, Eurasian, and Asian — competing throughout the region which was then experiencing high levels of instability and insecurity. Needham (1983) and Reid (1993) also point to an intensification of the slave trade in Eastern Indonesia in the late 18th–early 19th centuries.
19In Atauro, this memory of regional trade and conflict, with its related insecurity, dates back to a long period of time stretching between the 17th and 19th centuries.
20The presence of ships from the regional groupings of Sulawesi/Makassar, the Moluccas, or closer kingdoms such as Hera was ambivalent: foreigners were synonymous with danger, but were also, as we have seen, providers of goods. With neighboring islands, volatile alliances based on trade were always ready to tip over into conflict. Facing this uncertain context, the reaction of the locals was primarily to avoid coastal areas.
21Although it is difficult to identify which fortified sites might be associated with this period, informants clearly indicate that the populations often settled far inland, in naturally protected sites. The defensive site of Kota Aʼi had yet to be built when Akrema saw the arrival of the first boats from Ternate:
“The locals were on the coast for a celebration with palm wine. While they were drunk, a boat arrived, and the people on the boat massacred them. Only one elderly man, Bilewa, and an elderly woman survived, having remained in the mountains. The day after the massacre, the old woman came to fetch water and saw what happened; the two survivors put all the bones in the cave near the sea (Loniria) at Akrema. The boat, they were people from Ternate, Indonesians.ˮ (Martino da Silva, 4 July 2018, Akrema)
- 3 This system inspired local mythology: the character of Ilibalek (from the Hatu Dalas clan in Makili (...)
22In response to these risks, a system of clan watchers (asupaʼin in Hresuk) was set up on the heights to watch over the arrival of ships.3 This mistrust of the coastline continued until the early 20th century, when arrival of Makassar’s boats was still feared by the locals. Oral accounts and Portuguese maps from the 1960s confirm that many settlements remained inland until the arrival of the Indonesians, who forced the relocation of villages on the coast (towards the late 1970s).
23Atauro’s real conflicts, however, took place on the island itself. Oral tradition distinguishes between conflicts that pitted newcomers against groups previously settled on the island which were eliminated, and those that divided present-day society in the early days of its history. Let us begin with the wars of conquest against previous occupants of the island. A few clues help situate these events relatively within the history of settlement.
- 4 All these vanished humans or non-humans are fairly precisely located in what was the narrative orig (...)
24Atauro’s oral tradition mentions various beings, the Keti Kmolang, Takngan, Piknutun, and Aran Mameta4 (see also Rappoport, this issue), described as ʻdifferentʼ or ʻdistinctʼ from present-day humans, whether by their physical appearance, language, activities, or habits, or by all these traits at once. In recounts of the island’s skirmishes and conflicts, the first two groups in particular feature prominently and are associated with specific defensive sites.
25Island mythology describes the Keti Kmolang as an imperfect ʻfirst humanity,ʼ embodied by Pauketi, a woman rejected by the demiurge Ada Inan. Keti Kmolang are depicted as beings with small bodies and large heads — and sometimes pigs’ teeth or horns — and likened to cannibalistic, murderous devils (Guillaud 2019). The abductions and ambushes the Keti Kmolang allegedly carried out to procure human flesh eventually justified their extermination or assimilation by present-day human groups. Although mentioned all over the island, the Keti Kmolang are always associated with specific places such as caves or rock shelters, including more or less defensive areas (although no fortified sites).
26In Doru, the place known as Ili Rala Ina associates a low-walled area with a rock shelter where Keti Kmolang lived. The present-day inhabitants of Doru who claim to have felt forced to leave Ili Rala Ina because of these awkward neighbors, resettled at Ili Heru — a hill 275 m above sea level situated just behind the present-day village — where they again built fortifications. Both sites are located in an area with defensive features overlooking the coastline and surrounding areas.
- 5 This name originates from kako, meaning in the past, and natar, meaning ‘flat place’, in Raklungu: (...)
27Another site, Kakonatak5 near Dori (Makadade), has similar characteristics: a defensive position overlooking steep valleys surrounded by simple, now-collapsing walls. The site encloses a vast esplanade of compacted red earth, on which small obsidian and red limestone fragments as well as pottery sherds are now found. The site also includes a cave believed to have sheltered Keti Kmolang. This site is adjacent to a mythical place on the island, where the demiurge MoruaʼAran is believed to have molded mankind from wood, stone, earth, and leaves heated in water. Kakonatak is described not as a site of war, but of ceremonies and dances.
28Pilarilin, just above Dori, and Asame Leʼen (a rock shelter) in Berau, are also related to the Keti Kmolang. Pilarilin is a vast flat area accessible via a steep slope from the seaside. Its expanse is littered with various remains — including pottery sherds and shells, objects from outside the island (a bronze fragment), as well as iron smelting slag. These surface findings reflect multiple phases of occupation. Pilarilin features small rock shelters again associated with the Keti Kmolang. According to oral tradition, Pilarilin was the site of the final confrontation with the Keti Kmolang which resulted in the murder of Pauketi, the last woman of this group. More on this later.
- 6 The nature of this difference is still unclear: the Takngan were not cannibals but spoke a differen (...)
29The Takngan, another population described as human but ʻdifferent,ʼ6 were ousted from their territory by other groups. This population occupied the volcanic massif in the south-west of the island, an area known as Harak Opun, the ʻMasters of the West,ʼ or Tutun Opun, the ʻMasters of the Mountain,ʼ which refers to these former occupants. A war between the Takngan and three Makadade clans ended with the bloody massacre of the Takngan and the conquest of the entire south-western massif by the present-day Makadade clans. To this day, the Takngan are considered the rightful occupants of the area, which remains dangerous for anyone who has not made an alliance with them in the past (see Rappoport, Guillaud, this issue).
- 7 One of these megaliths is named Isu Talin, a reference to the wild, poisonous Entada rheedii liana, (...)
30Several sites are associated with the Takngan, the most important being the Ili Ralam fort. It lies in a volcanic zone and corresponds to an esplanade built on a coral limestone flat overlooking a slope. The site is enclosed by a now largely collapsed wall of coral blocks, and its ground surface shows pottery sherds. The epic account of the war between the Takngan and Makadade clans (Guillaud et al. 2023: 171–177) mentions that ladders were used to climb the walls of Ili Ralam, these ladders being pulled back once on the other side. Two small, elongated and partly polished monoliths made of volcanic rock can be seen in the immediate vicinity of this site.7 Still protected and feared, they are known as sharpening stones (adi), once used to take care of weapons before and after battle. A third, similar stone lies between coral rocks at the center of this site.
31Another site where, according to oral accounts, the fate of the Takngan drew to a close is Ili Hatu Tutun (Fortress of the Stone Mountain). This naturally fortified, flat summit of large limestone blocks overlooking a rock shelter of the same name, was ideally suited for sheltered surveillance. The last Takngan supposedly perished at Ili Hatu Tutun—except for a pregnant woman who had her ear cut off in the fight.
32Looking for chronological clues regarding the Keti Kmolang and Takngan, the genesis of the world of Atauro as recounted today suggests a few hypotheses about the succession of a number of episodes. Today’s Atauro society refers first and foremost to the myth of the creation of the island by two brothers who used arrows to pull the island out of the sea. In another story, however, these same two brothers are confronted by the Keti Kmolang, who terrorized the island, driving its population away. The myth tells of the brothers’ victory over the Keti Kmolang in Pilarilin thanks to their knowledge of the forge, as heated metal was the only way to kill these extraordinary beings. Literally, this detail could signify the arrival of a population knowledgeable in forging and thereby gaining a foothold in Atauro. The myths also explain a historical fact: the supposed descendants of one of the two mythical brothers are responsible for the division of Atauro’s society into three ethnolinguistic groups, thus establishing the island’s present-day society.
- 8 In these stories, the Takngan are associated with the Piknutun, another group who occupied the foot (...)
- 9 “Kilikmau with her ear cut off.”
- 10 This negotiator was not yet evangelized, but his son became Christian. This episode dates from just (...)
33As for the war against the Takngan and the conquest of the southwest territory they occupied, the events appear to be much more recent: the names of the protagonists on both sides are still remembered. The three Makadade clans (represented by three warriors) exterminated the Takngan using ʻmagic stonesʼ that killed at a distance; this is reminiscent of rifles which the Takngan did not appear to possess.8 This narrative is best remembered for its conclusion, wherein Kilikmau Klingapa,9 a pregnant woman, survives the massacre and, pursued by the three clan chiefs, flees toward Adara where the inhabitants paid for her life. The man in the narrative from Adara who negotiates a bride price with the Makadade chiefs to save the woman’s life is the grandfather of the present informant. From several hints10, we can reasonably place the war against the Takngan in the first decades of the 20th century.
34While both episodes deal with the extermination of populations distinct from the present-day communities, they also refer to two very different moments in history. The Keti Kmolang represent a ʻtime beforeʼ in the island’s history, and their disappearance marks the establishment of the ʻmodernʼ society. Comparatively, the Takngan were driven out of their territory by the Adadi society which was already structured into clans, confirming its dating to a relatively recent period.
35Between these two episodes, the ʻmodernʼ society faced various internal conflicts, pitting the three ethnolinguistic groups against each other, as well as against clans within each of these groups.
36Oral tradition reveals several episodes and conflicts which suggest that no area on the island was spared a generalized situation of tension or insecurity due to internal conflicts. These conflicts, initiated at clan level, escalated to pit the island’s various ethnolinguistic components against each other, as well as against outside groups.
- 11 Clans were organized in pairs based on matrimonial alliances.
- 12 This episode seems to confirm the composite nature of this settlement, where clans with different s (...)
37Conflicts between clans over small territories led to the building of the first fortified sites. This phase of history, particularly well documented through oral accounts in the Makadade region, witnessed a conflict prior to the structuring of the Adadi society in its current form. Four clans (Tua Bere Doe, Kapitan, Ruma Lari, and Ruma Nokron) entrenched themselves in Ili Hngara Bitauk, to fight two other clans, Ruma Pas and Mangetu,11 sheltering in Kota Wataran (the Rattan Fortress). The alleged cause of this conflict was the refusal of the four Ili Hngara Bitauk clans to be ruled by a queen, as was the practice in the Ruma Pas clan.12 These two fortified sites would have been the first to be built on the island’s south-eastern massif.
38The fortified site of Ili Hngara Bitauk occupies the summit of a small hillock of raised coral, more or less surrounded on all sides by steep slopes. Kota Wataran, less than two kilometers to the north, overhangs a steep slope and, as its name suggests, was hidden in a rattan forest that was difficult to penetrate. Hidden passages through the thorny plant defenses would have provided access to the center of the site.
“Each clan (ruma lisan) had a chief. Since they were all chiefs, they went to war with each other. They fought for four years, without stopping. No one worked, all they did was fighting. There was nothing to eat, just the foods from the forest, and even that, after a while, there were none left.ˮ (Hermenegildo de Araujo, 14 Sept. 2017, Anartutu)
39The never-ending conflicts in the early days of the “modern” society were resolved through a pact among the seven clans in the Makadade area, commemorated by the laying of a stone at Ileti Paadu (ʻthe Forbidden Leti Tree,ʼ13 located behind the Makadade Catholic Church in Anartutu). This pact gave rise to the Adadi society in its present seven-clan configuration and led to the construction of the larger fortified area of Kota Hahan (ʻFortress of Voicesʼ).
40Similarly, in the Ili Timur area, accounts allude to an internal territorial war over territory between Manroni (entrenched in the Ili Ngura fort) and a clan from the same group, established at Atekru near the coast in the fortified Wunu Rala.
41These wars illustrate the conflict-influenced establishment of society and the outlines of today’s territorial divisions. It was probably at this time that local society organized itself around three ethnolinguistic groups: Manroni, centered on the village of Ili Timur and theoretically covering the entire north of the island, Humangili, based in the Makili district, and Adadi, located in the Makadade district.
42Among the Adadi, now united, fratricidal conflicts were followed by a period of war against the Manroni: the seven Makadade based clans fought the seven Manroni clans for a border that fluctuated according to victories and defeats:
- 14 In other accounts, Er Leti (the source of the Leti tree), just below the Kota Hahan site, marked th (...)
“There was a war with Makadade. The border between Manroni and Makadade went from Beloi to Adara, and it passed through Loropui, near Abak.14 People could not cross that border, they would get killed.ˮ (Ruben da Cruz, 6 Dec. 2017, Bikeli)
43The wars were territorial, pitting the Adadi against the Humangili, and the latter against the Manroni. Involving larger groups, they probably motivated the reinforcement of the fortifications at Ili Hngara Bitauk, where today we can see a triple row of walls almost two meters wide in places. The two outer walls are pierced by chicane entrances to slow down attackers, while the inner wall features four narrow corridors easily defended by fighters overlooking them. Each of these entrances would have been entrusted to the guard of a different clan. This site is probably the most consistent with sites described elsewhere in East Timor. The imposing nature of the walls and the outer chicanes suggest the presence of firearms, some of them potentially heavy.
44Fighting the Adadi groups, the Manroni withdrew toward the Ili Lor site, which also features a double enclosure, the innermost of which is accessed by two narrow corridors and perhaps a third one (now collapsed), reminiscent of the fortifications at Ili Hngara Bitauk.
45Makili, for its part, boasts a vast fortified complex located above present-day Vila-Maumeta, known as Kota Heuknan. The entire population used to gather inside in times of danger, particularly during conflicts with Manroni groups over border issues (the limits being in the vicinity of Beloi).
- 15 Some of these conflicts remain unresolved to this day.
46Fortified sites increased in a context of growing insecurity. Each of the island’s localities appears to have been involved at one time or another in conflicts with other areas or even smaller localities. The conflicts often concerned borders,15 but not exclusively.
- 16 Tungas is a magical instrument used to hide the village.
- 17 The village of Maker would have enjoyed an intermediary status between different entities, offering (...)
47The first people arriving in Maker from the Makadade highlands settled on high ground near a spring, at Rareputi, where there is no defensive site but where traces of houses and ancient tombs can be found. However, the massacres perpetrated on the coast by the people from Makassar led to the construction of Kota Ili Tungas,16 a fortified enclosure at an altitude of 270 m asl. Based on genealogical information, this structure may have been built between 1875 and 1890.17 Kota Ili Tungas was abandoned when the Indonesians caused the population to migrate to the coast where a church was being established; nevertheless, a guardian remained on the site.
- 18 The inhabitants of Arlo allegedly failed to stop at Akrema, where they were supposed to ʻrequest pe (...)
- 19 Arlo’s warriors claimed to have called for a truce in order to practice meti, fishing on foot on th (...)
48The north of the island was not war-free. In Akrema, the vast fortified site of Kota Aʼi, high above the coast, is naturally defended by its steep seaward access, as well as by several rows of crumbling walls (seven, according to one informant) on the inland slope. People from Akrema used their magic powers to capsize a pirogue from Arlo, en route to the island of Liran,18 killed most of the survivors, and committed atrocities, triggering reprisals from the Arlo’s warriors who laid siege to Kota Aʼi. The war dragged on — this stronghold was deemed impregnable — but eventually the fort’s inhabitants were dislodged through a fake truce19 and, as a result of this ruse, were massacred by Arlo’s warriors.
49The Arlo region in particular seems to have crystallized antagonisms. Arlo seems at first to have been in conflict with Akrema and Bikeli, then in a later phase with the larger Makadade and Maker complexes. These latter conflicts ended in the mid–1940s. One finds in Arlo the defensive site of Ili Ara, ili and ara corresponding respectively to the stern and the bow of a boat, an image of a site made up of two parts separated by a wall. The site, at the top of a coral limestone massif, features a sinkhole that communicates with the Lepu-Kina cave in the depression below (where the present-day village is located).
- 20 This myth echoes the episode mentioned by Duarte (1984: 215): “Manrôni (...) decided to trade with (...)
50Not far away, Abak, at an altitude of 250 m asl, also features a defensive site enclosed by walls, now almost fully collapsed, near a water reservoir in a clay substrate. This site, located in the center of the island, is considered as a mythical place of the origins for the Manroni people. Humans are believed to have been born there from the scabbard (abak) of a saber or sword (opi),20 humanized by the two brothers who came down from Mount Manukoko.
- 21 Including wax, chickens, eggs, goats, men (for labor), and women.
- 22 Spillett (1999) indicates that Atauro was a dependency of the kingdom of Hera, a fact confirmed by (...)
51Informants often mention the last ʻgreat warʼ on the island of Atauro. At the beginning of the 20th century, war broke out when the Adadi refused to pay their tribute21 to the Portuguese. In an ongoing conflictual context, the Manroni seized the pretext of their enemy’s revolt to rally the Portuguese and their allies from Hera22 and Manatuto, tasked by the Portuguese with tax collection. Makili was also involved in this conflict on behalf of the Portuguese.
52Considering Kota Hahan poorly protected and Ili Hngara Bitauk too small to receive them all, the Adadi clans retreated to the fortified site of Taklo, a vast flat above Berau. Taklo is vast, with walls overhanging steep slopes and three paths and two gates providing access. Despite their retreat to this defensive site, the war ended with the defeat of the Adadi by the other two groups, followed by the landing of the Portuguese in Berau and the subsequent surrender of the Adadi insurgents. De Magalhães (1916: 3–4) probably refers to this episode:
“In 1905, this [tax] collection gave rise to an indigenous intrigue, leading the energetic governor to assume it was a revolt and order the punishment of the alleged culprits, who were defeated by a column of irregulars under the command of Staff Sergeant Antonio Joaquim.”
- 23 They stopped during the Japanese presence in the archipelago between 1942 and 1945. The Japanese ne (...)
53The arrival of the Portuguese on the island at the beginning of the 20th century certainly marked an easing of conflicts. The first military post was established at Beloi in 1909, followed by another in the Makadade highlands in 1912 (De Magalhães 1916). Nevertheless, clashes and skirmishes continued in various localities. Weapons of war, machetes, and rifles were gradually stored in caves or limestone shelters, protected by their reputation for danger. The last magical warfare ceremonies were held as late as the 1940s.23 A Catholic mission was established in 1949 and set about converting the population, while Protestants had already begun evangelizing the island in the 1930s (Duarte 1984: 15), advocating an end to the violence.
54With the exception of some particularly extensive sites (such as Taklo, still occupied today), the fortified sites used as retreats in times of conflict were not permanent settlements. Ili Hngara Bitauk, for example, was the retreat of the Adadi, whose peacetime settlement was a few hundred meters below in Nusa Lau Turuk. Some informants have mentioned the seasonal nature of clashes, which tended to take place during the agricultural off-season between July and October (Ili Timur). Warriors adopted war names for the occasion, such as Leki Kare, Kutukia, and Kutu Reti, who became the famous Tohleki, Tohlepo and Retiau in the epic account of the war against the Takngan. Two points in particular, linked to the techniques of warfare, deserve analysis.
55The techniques of warfare most frequently mentioned by informants concern the weapons used and magical knowledge.
- 24 Bullets were made from coral, and rifles were fired with flints and gunpowder.
56Bows and arrows appear in some accounts, as during the conflict between the occupants of Kota Wataran and the warriors of Makili, the latter all pierced by the same arrow as they climbed in line up the steep path leading to the Rattan Fort. More often than not, however, it was firearms that left their mark on informants’ memories. For example, the three famous Makadade warriors used small black stones (bullets?)24 against the Takngan, an advantage which proved decisive:
“Each of the Ilik Telu [three warriors] threw a stone that killed everyone in front of their house. They threw stones for five days.
They asked, ʻDid our stones hit you?’
ʻYes, not only did they hit us, but they also killed many of us. Please stop, your stones are killing all our people!’” (Rodolfo de Araujo, 18 Oct. 2022, Biti)
57The names of the most famous warriors are still remembered, as in Arlo where Maisako and Madeira, reputed to be invincible, performed ballistic feats by firing bullets backwards over their shoulders without missing any of their targets.
58Whatever weapons were used, it was primarily the strength of the fighters’ magic that determined the outcome of the battle. Certain trees (Aseitan) were used for the magic needed to defeat the Keti Kmolang, while the forging techniques deemed of divine origin enabled their extermination. Arlo’s warriors crafted statues that enabled them to defeat their enemies by petrifying them as they crossed into Arlo’s territory. At all sites, fighters would ʻwarm upʼ their weapons in specific spots charged with power: at Kota Wataran, they would go first to Hamak Tutun, a place with a ritual platform topped by a forked wood (iteas in Raklungu), then run successively to Kanilai Tutun, Laudarak, Samalai, Hatasu, and Pretun Mera Tutun, before finally going off to battle, being sure of their success. The three Makadade warriors mentioned above (Tohleki, Tohlepo, and Retiau) also used magic, by blowing kapock trees to crush the last Takngan who had taken refuge on the top of Ili Hatu Tutun. The only survivor, Kilikmau with her ear cut off, tried a ruse: she walked backwards and escaped through the ashes of Ili Hatu Tutun, but her pursuers, using a magic Jew’s harp that only sounded when they were on the right track, caught up with her.
59Cunning and deception are valorized in epic accounts of these victories, often in the form of truces that are actually traps into which the enemy falls. These subterfuges explain how the last Keti Kmolang and Takngan chiefs were killed, how the warriors of Atekru were intoxicated and thrown into a ravine by those of Makadade, or how the inhabitants of Akrema were deceived by a false truce proposed by Arlo.
60Another practice, that of severing heads, is mentioned in most war stories. Severed heads, particularly those of chiefs, were used to appropriate the power of fighters:
“I am the grandson of Maisako, who was a warrior in the time of Karilihu, the ʻBlack Kingʼ (Aran Mametan) in Arlo; he was black, and he won the wars. Maisako fired all his bullets and the enemies killed him, cut off his head, and roasted it on stones, placing fresh corn underneath. The corn was then eaten to take the warrior’s strength.ˮ (Simaõ da Cruz, 19 Nov. 2016, Arlo)
- 25 This expedition was carried out on behalf of the Portuguese, and perhaps refers to the events repor (...)
61Heads were also trophies that could be passed around, as at Ili Tungas (Maker) where warriors returning with their trophies from Dili25 aroused the envy of the people of Arlo in the early 20th century:
“The war with Arlo was the first, shortly after the fort was built. The people of Mingkoa Mingkaso [the man who built the fort] had cut off the heads of their victims and brought back a pirogue full of heads! The people of Arlo wanted them for their fort, and those here said:
ʻTry to come and take them.’
There was a war. The people from Arlo came, and they all died, killed by the people here.” (Cornelius Gomes, 10 Dec. 2017, Maker)
- 26 Mentioned above in connection with exchanges with Liran.
62Let us take a look at a story associated with fortified sites, that mentions an individual whose severed head, sword between teeth, inflicted a severe defeat on the enemy. In Heuknan (Makili), Meharek26 saw his harvest eaten up by pigs who eventually also devoured him, leaving only his head. His wife searched for him, and when calling him to the west, he answers from the east, and vice versa, thus signaling he has left the human world. She eventually found the head and brought it back wrapped in a cloth. The head asked her to sharpen a sword until a feather dropped on the blade split in two lengthwise. Meharek then requested his wife to place him in Fort Heuknan with his sword between the teeth. He routed the enemies, slicing their knuckles before rolling to the bottom of the fort in pursuit, and finally exterminating them all, except for one man who returned to Liran as witness of the defeat. The story goes on with the episode in which Meharek’s wife touched his head carelessly causing his death. An almost identical story, collected in Makadade, features an individual named Beresoe who, after being reduced to a severed head by pigs, decimated his Manroni enemies in the fort of Kota Wataran.
63The heads of Beresoe and Meharek were kept by their wives out of sight in the clan house, and thus assimilated to female prestige goods, as in Sumba Island (Hoskins 1989: 426).
64Hoskins (1989: 420) confirms that in then late 19th century Sumba, skulls were prestige items kept — along with other goods — in lineage houses but could also be purchased to adorn important houses. Cutting off an enemy’s head affected his descendants; afflicting them with fevers, crop failure, and disease until the death was avenged. Conversely, acquiring heads increased a group’s human and agricultural fertility (ibid.: 431).
- 27 Manukoko literally means ʻthe rooster crows.ʼ
65Mythology provides a final example of the power of severed heads. The head of the last Keti Kmolang woman, Pauketi, placed on a rock in a basket, was transformed the next day into two roosters shared between the two brothers of the story. While the elder went to the western massif, killed his rooster and ate it, the younger went to Manukoko where his rooster crowed in the morning,27 signaling to humans that the island had been liberated and that they were welcome to return from exile.
66Let us return to the debate between historians and archaeologists on the factors that triggered the construction of fortified sites. Logically enough, oral memory does not preserve any clear trace of ancient periods, such as the 12th to 14th centuries, when the first fortifications appeared in Timor. Other authors (notably Brockwell et al. 2020, about Vasino), relying on archaeological traces, make the same observation. In Atauro, according to oral accounts, internal and external conflicts that triggered the construction of the fortified sites appear relatively recent, the earliest being tentatively associated with events dating to the 17th century, while most of them occurred between the 19th and early 20th century. This does not prejudge the possibility of older uses of the same fortified sites, where surveys have so far proved impossible due to the absence of sediments.
67As the episodes unfold over the course of several centuries, the collective memory of the inhabitants paints a picture of deteriorating social relations, which are becoming the norm between groups, and whose various causes can be found in the literature.
68Schapper (2019) argues that in the 17th century, Dutch naval aggression — in particular the massacre of the Bandanese in 1621 — led to widespread fear among the indigenous populations. This event is considered as a potential trigger for the development of village fortifications in the southern Moluccas, where an endemic war culture took root. According to the same author (Schapper 2020: 243):
“The historical accounts we’ve seen elsewhere in the southern Moluccas indicate that warfare and raids were a cultural practice, not caused specifically by resource scarcity but by the normal way of life, with the smallest offenses or infractions leading to violence.”
69Discussing the dynamics of change leading to fort-building, McWilliam (2020: 135–136, 140) suggests:
“The new and combined impact of Portuguese colonialism and Islamic trading interests based in Sulawesi from the mid-16th century triggered a transformative change in Timorese social relations and residential patterns that gave rise to the emergence of fortified hilltop settlements. […] The four interrelated mechanisms involved in changing patterns of occupation are: (1) a boom in the sandalwood trade from the late 16th century onwards; (2) the introduction of maize as a staple food crop in Timor during the same period; (3) the new trade in modern weapons, particularly artillery and firearms; and (4) a significant increase in demand in the human slave trade.”
70These triggering factors could shed light on a number of phenomena observed at Atauro, while Atauro’s case could add new explanatory factors.
71The sandalwood trade is never mentioned in the local myths, but wax, another marketable commodity, seems to occupy an important symbolic place. Wax appears in the early episodes of the local mythology, in the form of the wax house built by the supreme being Ada Inan for the woman he finds on the Rea Nelir islet (which corresponds to the surroundings of today’s Mount Manukoko). According to McWilliam (1991: 54):
“During the 18th and 19th centuries, one of the region’s main export products was raw beeswax (nini) as well as sandalwood, for which Timor had long been famous. In addition to its use in candle-making, beeswax was sought after by traders to supply Java’s flourishing batik industry, particularly in the 19th century.”
72Wax would have been one of Atauro’s main exports at a time when sandalwood was in decline. The exportation of wax continued until the 1960s, with boats from Makassar visiting all the coastal areas to collect this abundant production.
- 28 This coming down from the mountain and discovery by both brothers of non-humans, plants or animals, (...)
73Following the construction of a wax house by the supreme being and the drying out of Atauro with the arrows of the two mythological brothers, the origin stories describe how the founding characters of present-day society — each beginning separately his descent from Mount Manukoko — ʻdiscoveredʼ various natural elements which they ʻmade human,ʼ thus creating the ancestors of today’s clans (Facal and Guillaud 2021; Facal 2023; Facal and Guillaud, this issue).28
- 29 Surveys conducted in several settlements on the west and north coasts of Atauro (Atekru, Adara, Akr (...)
74The emerging communities associated locals and newcomers, even if some groups remained on the margin.29 The various reconfigurations of power occurring during this stage of socio-genesis generated local conflicts, such as those observed in Makadade and Ili Timur. Is it possible to hypothesize that these reconfigurations were linked to the upheavals brought about by trade in commodities such as wax and slaves?
75Eventually the groups expanded across the island generating other conflicts, some of which appear to be linked to struggles over crucial resources. A detail related to the conflict with the Takngan is significant in this respect. The Makadade people are described entering into matrimonial alliances with the Takngan in order to obtain more land for cultivation, and to have then built dry-stone walls around their corn plots (source: Biti). The cause of the war is clearly associated with the Takngan’s rejection of this spatial enclosure — perhaps a sign of a different farming system, or even a different social system?
76This episode of territorial conquest is reminiscent of the combination of factors that led to Atoni expansion in West Timor, as described by James Fox (2003: 18):
“This combination of muskets, iron tools and corn, supplied mainly to Atoni groups, changed the face of West Timor. With a new, highly productive crop, the tools to plant it and the firearms to expand aggressively and open up new land on other people's territory, the Atoni population [...] spread rapidly across much of West Timor, assimilating other groups to Atoni ways of subsistence and culture.”
77Such a process, combining the arrival of settlers with technical changes, could also shed light on the most recent episodes in Atauro’s history. This process in itself is not original; what is original is the complete integration of the episodes of these arrivals into local mythology.
- 30 There are also contemporary interpretations, such as at Ili Hngara Bitauk where each gate is said t (...)
78Most fortified sites have been used on multiple occasions over the course of the island’s history, even in recent times. In Makadade, the comparison between a site like Kota Wataran, built with vegetal components before the establishment of Adadi society, and Ili Hngara Bitauk, used until the 20th century and equipped with three rows of walls, shows that they have been redeveloped and reinforced to adapt to changes in the weapons used: from bladed weapons and arrows originally, to rifles and probably cannons in Portuguese times. What is more, the builders not only chose naturally fortified sites, but seem also to have tested various defensive solutions (such as the enigmatic spiral of Kota Ruma No in Adara). Builders experimented with different configurations of the sites themselves, perhaps reflecting the evolution of the group’s social organization.30 What remains to be done, therefore, is to survey all these sites, and to conduct a more thorough study of them before the oral history is lost.
- 31 Some sites (Ili Ara in Arlo) feature piles of stones that could be tombs, but this is by no means s (...)
79In terms of their status, these fortified sites, linked to more or less ancient history, are associated with ancestors31 and as such join other types of sites received from ancestors, such as protected forests. Forts were once protected by the status of sacredness and dangerousness associated with everything that came from the ancestors, and although Protestant informants report having ʻcleansedʼ many localities of their sacred power (Silva, this volume), the fortified spaces still retain a certain association with danger. For example, it is sometimes not advisable to talk about the history of the place on site, or to go there at certain times (Ili Heru), while, in other cases the history can only be recounted on site (Ili Ralam). Similarly, McWilliam (2020: 137) confirms that “all sites are considered the abode of spirits, and therefore potentially dangerous to health and well-being.”
80The memory of these sites — of the groups who own them and on whose territories they are located — is also linked to current issues. Most of the time, no one particularly guards these fortified sites, but each remains within the lands of a specific clan whose approval is required to access it. This is even true of Ili Ralam, a Takngan stronghold before the group’s extermination. Even today, all clans still consider the Takngan group the owner of Ili Ralam and only clans who once formed alliances with the Takngan, and thus consider themselves custodians of their rights to the land, can take you there.
81The Ili Hngara Bitauk fort and the area surrounding it belong to one clan, Tua Bere Doe, and the forest that covers the site and its surroundings is protected; the forest trees may be cut only for collective uses (for the needs of the church or the village), such an action requiring the approval of the entire clan community. As McWilliam (2019: 254) points out, fortified sites are integrated into the clan’s territory, as are places of its migration. However, unlike the clan’s places of residence, or the graves marking its itinerary about which only concerned clans can speak, fortified sites have a more collective dimension: they are territorial ʻnodesʼ where allied clans came together during periods of war. Furthermore, during peacetime, important collective ceremonies were held there, as in the fort of Ili Hngara Bitauk (Makadade) where the Tolan rituals, now forbidden by the Christians, took place. In addition to their role as landmarks, the forts appear to have played a memorial role, as they readily symbolize the alliances and enmities that led to their construction. Today, the prospect of tourism on the island could target these ʻtourist objects.ʼ This would give rise to new, unprecedented challenges for the management of these complex sites with their ambiguous status, at once clannish but collective, desecrated but dangerous, and above all still charged with the energy of the many conflicts in Atauro’s long history.