1The island of Madagascar is situated in the southwestern Indian Ocean and is separated by about 400 km from the African coast. Despite its proximity to the African continent, it was the last large island in the world to be permanently settled. How and when Madagascar was settled remains poorly known. Traces of human occupation dating back to the first millennium ce have been identified along parts of the Malagasy coast (Vérin 1986, Dewar & Wright 1993). Historical, linguistic, genetic and archaeological evidence indicate that the ancestors of the Malagasy came from different parts of the Indian Ocean including Southeast Asia, East Africa, South Asia, and the Near East (Dewar & Wright 1993, Adelaar 2006).
2Although systematic archaeological research in Madagascar did not start until the 1960s, isolated excavations had already been carried out in various parts of the island in the late 19th and early 20th century. Excavations in northeastern Madagascar in the first half of the 20th century revealed the presence of a former prosperous civilization generally referred to by archeologists as the “Rasikajy civilization” (e.g. Vérin 1972, 1986; note that Deschamps refers to this civilization as Iharaniens). One particularity of this civilization is that it carved a wide variety of objects out of soapstone (also referred to in the English literature as “chlorite schist” and in the French literature as “chloritoschiste”). Gaudebout & Vernier (1941a) report that the word “Rasikajy” is used by the inhabitants of northeast Madagascar to designate a group of people that came from overseas and made objects from soapstone, whereas Wright & Dewar (2000) mention that the term is applied by the inhabitants of the northeast to the people of the distant past who mined chlorite schist and worked this soft stone into vessels.
3Most evidence for the Rasikajy civilization comes from excavations at a cemetery discovered in the late 19th century at the coastal town of Vohemar in northeastern Madagascar (Grandidier 1908). Although hundreds of tombs have been unearthed notably in the 1940s (e.g. Gaudebout & Vernier 1941a, Vernier & Millot 1971), the site of the former settlement that most likely existed nearby has never been found.
4Vérin (1986) proposed that the Rasikajy civilization was the result of biological and cultural intermingling between Islamized groups arriving at the island and local people already present in northeast Madagascar. The position of the dead in the tombs at the cemetery of Vohemar has been used to suggest that the civilization had muslim roots and the presence of imported pottery to indicate engagement in the Indian Ocean maritime trade (Dewar & Wright 1993).
5Differences of opinion exist regarding the timing of occupation of the settlement at Vohemar. Vernier & Millot (1971) consider that most imported pottery found in the tombs dates from between the 14th and 16th ce. In contrast, Vérin (1972, 1986), who did only have limited access to objects excavated at Vohemar, concludes that the finds date mostly from c. 1400 to 1750 ce.
6In this paper, we examine published data of excavations in northeast Madagascar in an attempt to constrain the origin of the Rasikajy civilization and the timing of their settlement. We first present an overview of past excavations at the cemetery of Vohemar and proceed with an analysis of the existing documentation of burial objects unearthed. We compare imported pottery found in the tombs with pottery found at other sites involved in the Indian Ocean maritime trade and with ceramic cargoes found in shipwrecks. This analysis provides constraints on the timing of the settlement. In addition, we examine the nature and characteristics of both imported and locally produced burial objects. From the burial objects, we deduce funeral rituals and we use these rituals as well as the nature and characteristics of locally produced soapstone objects in an attempt to trace the origin of the Rasikajy.
7In 1899, Guillaume Grandidier recognizes the presence of a large cemetery at Vohemar, but the few tombs that he examines contain only human bones and no objects (Grandidier 1908). Maurein carries out more successful excavations in 1906 and donates a collection of burial objects to the Musée d’histoire naturelle of Nîmes in France (Gaudebout & Vernier 1941a, van der Sleen 1960). In a note accompanying the donation, Maurein interprets the cemetery as an extensive Arab burial place with muslim tombs (van der Sleen 1960).
8Although it is quite likely that more tombs were searched after Maurein’s excavations in 1906, it is not until 1941 that a large-scale investigation of the cemetery at Vohemar is undertaken and 261 tombs are excavated by Gaudebout and Vernier (1941b). Their excavations cover groups of tombs at five sites labelled A (9 tombs), B (78 tombs), C (130 tombs), D (divided in three sub-sites, D1, D2 and D3 with a total of 38 tombs) and E (6 tombs). The tombs are generally marked by slabs of coralline material, which are placed vertically in the ground. The finest tombs are marked by a rectangle of four slabs that are either level with or slightly above the level of the ground (Gaudebout & Vernier 1941a). More than 800 objects or parts of objects are taken out of the tombs during excavations in the early 1940s (Vernier & Millot 1971). A first, succinct description of objects found in a few rich tombs and a map with location and orientation of the tombs are given in Gaudebout & Vernier (1941a). A more detailed documentation of the excavations in 1941 is not published until 30 years later when Vernier & Millot (1971) publish an illustrated catalogue of Malagasy objects at the Musée de l’Homme in Paris. This catalogue includes a limited number of objects unearthed during later excavations at the same site in 1942, 1948 and 1955. According to unpublished reports cited by Vérin (1986), another 312 tombs are investigated in 1942 by Gaudebout (grave sites G and H; see map of Gaudebout & Vernier 1941a), a further 42 tombs in 1948 by Poirier and an unknown number of tombs by Millot in 1955. However, except for the few objects documented in Vernier & Millot (1971), no results of the excavations in 1942, 1948 and 1955 have ever been published.
9Although more than 600 tombs have been investigated in the 1940s, only the results of the excavation work in 1941 covering 261 tombs are reasonably documented. Hence, the following description of burial objects unearthed from the tombs at Vohemar relies heavily on documentation provided by Gaudebout & Vernier (1941a) and Vernier & Millot (1971). In addition, a limited number of objects from the excavation by Maurein in 1906 are documented in van der Sleen (1960) and Vérin (1975, 1986). Unfortunately, a detailed recording of data (burial objects, skeletons), grave by grave, is lacking, prohibiting a dating of the different groups of tombs within the cemetery and preventing a sound sociological and economical study of the Vohemar settlement.
10Although Gaudebout & Vernier (1941a) send over 800 objects uncovered from the Vohemar cemetery to the Académie malgache in Antananarivo, the catalogue of Vernier & Millot (1971) shows that a considerable amount of these objects were subsequently transferred to the Musée de l’Homme in Paris. Vérin (1986) notes that only few of the objects excavated at Vohemar remained in Antananarivo. Recently, the collection at the Musée de l’Homme has been moved to the Musée du quai Branly and part of it can now be viewed on-line (http://www.quaibranly.fr/). The on-line catalogue also includes objects from excavations in 1948 and 1955 donated by Elie Vernier, imported Chinese bowls and plates donated by Anne Gaudebout in 1991 (probably a family member of P. Gaudebout, who was district administrator at Vohemar and carried out the excavations at the cemetery with Elie Vernier in the 1940s), and objects donated in 1995 by Lucie Vernier-Escande, the wife of Elie Vernier.
11The research by Vérin (1975, 1986) represents a monumental contribution to our knowledge of the trading ports in northern Madagascar. In the context of the cemetery at Vohemar, however, Vérin had difficulties accessing the collection of Malagasy objects at the Musée de l’Homme and writes: “… about 30 objects reluctantly shown to me in three afternoons spent at the Musée de l’Homme…” (Vérin 1986). Thus, Vérin’s assessment of the cemetery at Vohemar is mainly based on objects preserved at the Académie malgache in Antananarivo, the limited number of objects shown to him in Paris, and the Maurein collection in Nîmes. Although Vérin (1975) includes the study of Vernier & Millot (1971) in his reference list, he adds a note that he was not aware of it at the time of writing.
12Burial objects recovered from the tombs at Vohemar include soapstone objects such as tripod vessels, cups and pierced circular disks; iron objects such as knives, machetes and a saw; jewellery such as silver, bronze and golden rings, bracelets and necklaces; pottery such as dishes, bowls and jars; spoons made from shells, bronze mirrors and needles; glass ware such as flasks in different sizes; beads; decorated bone objects.
13Most of the pottery found at Vohemar is clearly of Chinese origin and remarkably diverse (Vernier & Millot 1971). Few earthenware pottery of unknown origin has been excavated including small bowls with handles or lugs, big bowls covered in red slip and decorated with arabesques and dishes decorated with unknown forms of writing (“terres cuites diverses” in Vernier & Millot 1971). Vérin (1986) interprets the earthenware pottery as imported Islamic pottery. Excavations in 1948 (Poirier) and 1955 (Millot) at Vohemar yielded receptacles made of local Malagasy pottery with identical shapes as similar Chinese or Islamic objects (Vérin 1986).
14The Chinese ceramics from Vohemar consists mainly of bowls, plates and dishes of different sizes and shapes. The ceramics include coloured and celadon stoneware, and monochrome and blue-and-white porcelain (Vernier & Millot 1971; Vérin 1975, 1986). Only a few polychrome plates have been described (Vernier & Millot 1971).
15Coloured stoneware comprises mostly white-, gray- and brown-glazed pottery such as bowls, plates, saucers and small jars. The celadon stoneware consists of a grayish to grayish-white body with a tendency to brown-red where left unglazed and a glaze varying in color through different shades of green. The monochrome porcelain includes white, grayish and “bluish-white” bowls and small plates, and double-gourd vessels with spotted marks. The blue-and-white porcelain ware is decorated with a large variety of motifs such as flowers, birds, dragons, lions and chi’lin, a mythic Chinese creature. Some of the blue-and-white plates and dishes have ribbed or fluted sides like some of the celadon ware.
16Over the last decades considerable research has been done on dating the export of Chinese ceramics through the investigation of shipwrecks (e.g., Sjostrand et al. 2006). Dated shipwrecks can give a reliable age for the export of a particular type of ceramics. However, as the number of recovered ceramic cargoes from shipwrecks increases, it is becoming increasingly clear that popular ceramic designs were made and exported for extended periods (Martin 2001)
17During the Tang dynasty (618-960) Arab and Persian trading vessels started to transport Chinese coloured stoneware and celadon to different regions of the Indian Ocean including the Philippine Archipelago, Indonesia and the Near East (Crick 2010). Export of these ceramics during the Song dynasty (960-1279) occurred more and more by Chinese trading vessels and during the Mongol Yuan dynasty (1279-1368), the trade export was further promoted and included celadons, white porcelain and the first blue-and-white porcelain (ibid.). After an initial period during the Ming dynasty (1368-1644) when sea trade was banned and only permitted under the tribute system, thus boosting lucrative smuggling operations, restrictions were lifted in 1567 and official trade was resumed (ibid.). In the early to mid-16th century, Portugese ships transported Chinese blue-and-white porcelain in their ships to Europe (e.g. Auret & Maggs 1982).
18Below, we describe a few examples of imported Chinese ceramic ware in more detail since they provide clues on their origin and on their approximate timing of production.
19A small, brown stoneware dish from Vohemar shows a raised twin-fish motif in the centre of the interior with the fish swimming in opposite directions (Vernier & Millot 1971:89). Celadon and stoneware dishes with identical motifs have also been found at the Santa Ana gravesite in the Philippines (Locsin & Locsin 1967:73, fig. 53), at kilns in the Longquan region in the province of Zhejiang in SE-China (Hughes-Stanton & Kerr 1982) and in cargoes from shipwrecks, such as the one dated at c. 1323 and found at Sinan, off the coast of South-Korea (National Museum of Korea 1977:28, 31, 99). At Santa Ana, raised twin-fish dishes have been found with coins of the Emperor Hui-Ts’ung during his reigning year of Ch’ung-ning, i.e. 1102-1106 ce (Locsin & Locsin 1967:122, fig. 96). Although the raised twin-fish dishes are generally considered to be typical Southern Song (1127-1279), their presence in the Sinan shipwreck illustrates that certain types of celadon wares continued to be exported during the Yuan dynasty (1279-1368).
Fig. 1. A brown stoneware dish swith a raised twin-fish motif.
Dish, Paris, Musée du quai Branly. Porcelain. 12.8 x 12.8 x 3.7 cm, 205 g. Inventory number: 71.1965.4.110
©2011. Musée du quai Branly/Scala, Florence.
20Two white double-gourd pouring vessels spotted with dark markings have been found at Vohemar (Vérin & Millot 1971:128, fig. 137). Similar wares have been uncovered from the gravesite at Santa Ana (Philippines; Locsin & Locsin 1967:95, fig. 77; 97, fig. 79) and from a shipwreck at Pandanan (Philippines) carrying Chinese coins of c. 1430 (Diem 1997; 1998-2001). The spots have been obtained by applying drops of ferrous oxide on the white glaze, whereas the body of the spotted wares consists of a sugary white porcelaineous material identical to that of so-called Qingbai white wares and all early blue-and-white wares (Locsin & Locsin 1967). Qingbai white porcelain was produced at kilns in Jingdezhen and at many other southern kilns from the time of the Northern Song Dynasty (960-1125) until the 14th CE (Li et al. 2010).
Fig. 2. White double-gourd pouring vessel.
Pot with a handle, Paris, Musée du quai Branly. Porcelain. 12 x 8.5 x 7 cm,
154 g. Inventory number: 71.1949.12.16
©2011. Musée du quai Branly/Scala, Florence.
21Chinese potters at Jingdezhen (Jiangxi Province, China) produced the first blue-and-white ware during the Yuan dynasty (1279-1368) by decorating whitish kaoline clay with blue cobalt oxide under a clear glaze (e.g., Vainker 2005). Some of the white and blue-and-white porcelain bowls and plates from Vohemar carry Chinese inscriptions at their base that translate in “made under the great Ming dynasty” (Vernier & Millot 1971). The colour, shape and motifs of the blue-and-white bowls and plates from Vohemar suggest that most of the ware has been produced during the Yuan (1279-1368) and/or early Ming (1368-1425) dynasty. However, blue-and-white ware from later periods has also been excavated at Vohemar, notably a large blue-and-white bowl, which carries the inscription: “Great Ming Dynasty, Jiajing (‘Kia-tsing’) period”, i.e. 1521-1567 (Vernier & Millot 1971:111, fig. 117, 118).
Fig. 3. Blue-and-white plate considered to pertain to the Hongzhi (1487-1505) period (Monique Crick, pers. comm.).
Plate, Paris, Musée du quai Branly. Porcelain. 3.5 x 18 x 17.7 cm, 251 g.
Inventory number: 71.1949.12.129
©2011. Musée du quai Branly/Scala, Florence.
22Trade export of Longquan celadon ware throughout the Indian Ocean occurred over hundreds of years extending into the 14th and 15th century and exports of celadon ware overlapped with exports of white and blue-and-white porcelain (Crick 2010). It is therefore difficult to determine the duration of occupation of the settlement at Vohemar on the basis of the imported Chinese ceramics. Our own, preliminary investigations suggest that most of the imported ceramic ware at Vohemar, including the blue-and-white ware, predates the arrival of the first Europeans to Madagascar in 1500 ce. Monique Crick, Chinese export ceramic expert at the Foundation Baur in Geneva, classified the Chinese ware at Vohemar presented in the catalogue of Vernier & Millot (1971) and on the website of the Musée du quai Branly as pertaining to the 14th to 16th century, in agreement with the conclusions reached earlier by Vernier & Millot (1971). Celadon ware and spotted ware is considered to date from the early 14th century, whereas the majority of the Chinese blue-and-white ware dates from the 15th and early 16th century (Crick, pers. comm.). A few pieces (e.g., Vernier & Millot 1972, figs. 113. 115, 139, 140) were classified as somewhat later from the Jiajing period (1521-1567), and one bowl (Vernier & Millot 1972, fig. 110) was considered to date from the second half of the 16th century pertaining to the Wanli period (Crick, pers. comm.). The objects shown in the catalogue by Vernier & Millot (1971) do not provide evidence for the presence of 17th or 18th century Chinese ceramics at Vohemar (Crick 2010, pers. comm.).
23The presence of 16th century blue-and-white Chinese ware, however, indicates that the cemetery at Vohemar was still in use at the time that the first Portuguese ships reached the island of Madagascar. In 1500, a ship commanded by Diego Diaz was the first European ship to sail along the east coast of Madagascar. Other visits by Portuguese ships followed over the next decades.
24A large dish with inscriptions forming part of the earthenware objects of unknown origin illustrated in Vernier & Millot (1971) has recently been identified by Mariam Rosser-Owen (2010, written comm.) as Manises pottery from Spain. The inscription and the motifs under the inscription band are almost identical in style with a dish (object number 302-1893) at the Victoria and Albert Museum and dated by Ray (2000) between 1500 and 1525. The inscription is a garbled version of “In principio erat verbum”. The inferred age of the earthenware Spanish dish falls within the age range of the Chinese ware.
Fig. 4. Dish with inscriptions identified as Manises pottery (province of Valencia, Spain; Mariam Rosser-Owen, pers. comm.)
Dish, Paris, Musée du quai Branly. Glazed earthenware. 5 x 26 x 26 cm, 645 g. Inventory number: 71.1918.104.22.168-2
© 2011. Musée du quai Branly/ Scala, Florence.
25A large number of objects carved from soapstone were found in the tombs at Vohemar (Vernier & Millot 1971, Vérin 1986). Mouren & Rouaix (1913) show that these and other soapstone objects found at other sites along the northeast coast were locally produced. They identify a number of sites in the Vohemar region where soapstone was quarried in the past and where remains of raw casts of soapstone objects were left behind. The same authors were the first to make a link between those who had worked the quarries and those who were buried at the cemetery at Vohemar. Gaudebout & Vernier (1941b) identify nineteen soapstone quarries in the Vohemar region. Subsequently, more soapstone quarries are identified further south near Antalaha in the Vohemar region (Vernier 1952, 1956) and further west in the Ambilobe region (Lods 1955). Soapstone quarries have also been identified nearly 750 km south of Vohemar in the region of Mananjary (Dalais 1919, unpubl., in Vérin 1986; Griffin 2009).
26Many of the excavated tombs at Vohemar included soapstone tripod vessels and lids. These objects show a simple, regular ornamentation consisting of circular grooves, indicating that a potter’s wheel (lathe) was used in the manufacturing process (Mouren & Rouaix 1913; Vérin 1986). No turned and polished soapstone objects were ever found at the quarries implying that only the rough cutting was done there and that the finishing of the extracted casts was done elsewhere (Vérin 1986).
27The tripod vessels show a remarkable resemblance to similar vessels found in ancient graves in China. In North China, tripod ceramic vessels (“ding tripods”) have been found in graves of the Dawenkou Culture dating back to the period between 3900 and 2400 BC (Fung 2000). During the Shang, Zhou and Han Dynasties bronze tripod vessels were used in funeral rituals and placed in tombs (de Groot 1892-1897, Eberhard 1969). Later during the Song and Yuan dynasties (960-1368), celadon tripod vessels modelled on burial ware from earlier dynasties were produced at the Longquan kilns in SE-China (Rotondo-McCord & Mowry 2000).
28Thus, Rasikajy artisans produced burial ware such as tripod vessels carved out of soapstone that was quarried locally. They used a potter’s wheel and modeled their tripod soapstone vessels on typical Chinese burial vessels.
29Besides tripod vessels, other soapstone objects found in the tombs at Vohemar include cups, pearls, pierced circular disks, flat-bottomed dishes, lamps, spindle whorls, a jar, and a turned bowl and pot (Vernier & Millot 1971; Vérin 1972, 1986).
Fig. 5a. Soapstone tripod vessel with lid.
© Institut de civilisations du Musée d’art et d’archéologie de l’université d’Antananarivo, Antananarivo, Madagascar.
Fig. 5b : Soapstone tripod vessel with lid. [Vessel, Paris, Musée du quai Branly.
Chlorite-schist. 33.3 x 29 x 29 cm, 6290 g. Inventory number: 71.1922.214.171.124-2.
©2011. Musée du quai Branly/Scala, Florence
Fig. 6. Pierced circular disk made from soapstone, possibly a bi-disk.
Paris, Musée du quai Branly. Chlorite-schist. 6.2 x 6.1 x 1.6 cm, 114 g. Inventory number: 71.1949.12.143
© 2011. Musée du quai Branly/Scala, Florence.
30The pierced circular disks are particularly interesting as they might represent bi disks. Bi disks are round and flat objects with a small circular hole in the middle and were used as ritual artificacts in China for a long period starting in Neolithic times (Shu-P’ing 2000). In China, bi disks made of stone, jade or glass were placed on the body in graves of persons of high social status. Several pierced circular disks are shown in the catalogue of Vernier & Millot, one of them with un-deciphered inscriptions (Vernier & Millot 1971:47, fig. 36; 49, fig. 37).
31Soapstone cups were generally found at a depth of about 40 cm (Vérin 1986), much shallower than the bodies of the deceased persons. Vérin considers that these cups were used as incense burners and that they were originally placed on top of the tombs or on the surface.
32During the excavations by Gaudebout and Vernier at Vohemar in 1941, over hundred spoon-shaped shells were recovered from the tombs. They vary in length between 6 and 15 cm and are made from a seashell, the Turbo marmoratus (Vernier & Millot 1971), which is common in the Indian Ocean and along the coast of Madagascar. The shells have been strongly hollowed out and the nacre is generally polished. The spoon-shaped shells with a rounded bowl bear a strong resemblance to modern and ancient Chinese soup spoons. Similarly styled spoons have already been described from tombs of the Han Dynasty (Huang 2000).
Fig. 7. Shell spoon. [Spoon. Paris, Musée du quai Branly.
Shell. 5.5 x 10.5 x 5.7 cm, 35.9 g. Inventory number: 71.1949.12.2
© 2011. Musée du quai Branly/Scala, Florence.
33Iron slag heaps have been identified in many places in the vicinity of Vohemar (e.g. Vernier & Millot 1971). Although their age is not known, the presence of foundry tubes made from soapstone suggests that iron objects found in the tombs were produced locally.
34Metal jewellery found in the tombs includes rings, bracelets and different kinds of chains. Most of the jewellery is made of gold and silver. The discovery of a fragment of a soapstone mould used to make chain links near the Ankarana caves to the west of Vohemar (Vérin 1986) points to a local production of jewellery. It is possible that gold was mined locally. At present, gold is still produced in the Betsiaka District (Andavakoera valley) west of Vohemar (Rambeloson 1999). There are no known silver mines in Madagascar and probably silver jewellery was either produced overseas or made locally from imported silver coins. Vérin (1986) notes that Vernier wrote in a letter to Decary — who was administrator and ethnographer (Balard 2002) — on 8 November 1941: “One of the silver rings in package no. 2 has half-obliterated Chinese characters on the bezel.”
35Vérin (1986) notes that some pottery fragments found in the tombs at Vohemar are imitations of soapstone vessels such as receptacles, bowls and tripod vessels. In contrast to the soapstone vessels, these imitations have not been turned on a potters’ wheel (ibid.), and it is probable that the fragments represent local Malagasy pottery.
36The following description of the tombs and the position of the burial objects with respect to the body of the deceased are after Gaudebout & Vernier (1941) and Vernier & Millot (1971). At the time of the excavations in 1941, the location of most of the tombs could be identified by vertical slabs of sandstone or coralline matter that generally stand out up to 10 cm above the ground surface. The dimensions of the tombs are mostly about 180 cm by 70 cm, and the bodies of adults were generally found at a depth of between 150 and 200 cm, whereas children were buried less deep. The bodies in the tombs were placed on their right side, with the head towards the east and facing north. Remains of textile fragments indicate that the bodies of the deceased were wrapped in a shroud surrounded by a mat.
37Large dishes were placed vertically to the east of the skulls, whereas bowls were often found behind the head. Spoons made from shells were placed at the necks and bronze mirrors with their polished surface against the foreheads. Jewellery in the graves included silver bracelets, necklaces around the necks, and golden and silver rings on the fingers. Beads found in the graves and made of quartz, cornelian, soapstone, glass or metal have been interpreted as coming from necklaces or bracelets. Iron objects such as machetes and knives were also found alongside the bodies, whereas a soapstone vessel was often placed upside down at the feet of the body. A 14C date on bones from a skeleton excavated at Vohemar revealed an age of 760 ± 60 a (Vernier & Millot 1971), i.e. around 1200 ce.
38The objects encountered in the tombs indicate that the settlers at Vohemar practiced specific funeral rituals, which are very similar to those practiced in the past in SE-China (Watson 1988). In China, incense burners were placed on top of the tombs or graves. Wares such as plates, bowls, spoons, jars and vessels were placed in the tombs for meals in the afterlife, and bronze mirrors were placed in front of the deceased person’s forehead for light in the afterlife (de Groot 1892-1897)
Fig. 8. Bronze mirror. [Mirror. Paris, Musée du quai Branly.
Bronze. 15 x 15 x 1 cm, 467 g. Inventory number: 71.1949.12.109
©2011. Musée du quai Branly/Scala, Florence.
39Imported Chinese ceramics (e.g. twin-fish dishes and double-gourd pouring vessels) have also been found at gravesites in the Philippines (e.g., Santa Ana — present-day Manila), and have been described in Locsin & Locsin (1967). The position of the ceramics in relation to the body of the deceased shows certain similarities with those at Vohemar. However, in contrast to the cemetery at Vohemar, the graves at Santa Ana are rarely delimited, randomly oriented and the position of the deceased does not seem to follow any ritual. At Santa Ana, the burial objects consist nearly exclusively of ceramics and only rarely of other objects such as knives, machetes, bracelets, gold jewellery or beads. No shell spoons, bronze mirrors or soapstone objects have ever been found at Santa Ana.
40Most of the tombs at Vohemar are roughly oriented east - west with the head of he deceased toward the east and facing north (Vernier & Millot 1971). The positioning of the bodies has been previously interpreted to be typical of muslim burials (Vernier & Millot 1971, Vérin 1986) and it can not be excluded that the Rasikajy at Vohemar had muslim roots. During the Mongol Yuan Dynasty and later during the Ming Dynasty Chinese muslims controlled much of the trade of goods leaving China (Crick 2010). The large Chinese fleets with hundreds of ships and thousands of crew that visited various regions of the Indian Ocean including the coasts of east Africa during the early 15th century were led by a muslim eunuch, Zheng He (Levathes 1986). Ma Huan who joined these tributary missions and wrote about them in detail (Ma Huan 1433) was also a muslim. Chinese muslim communities existed in Indonesia (e.g. Semarang, Java) in the 15th century (Suryadinata 2005).
41However, the east-west oriented tombs at Vohemar do not necessarily need to reflect muslim burial rituals. Similarly oriented graves are known from pre-Islamic sites in South Sulawesi (Indonesia) for the period ranging from approximately 1300 AD until 1600 AD (Bougas 1998, Bulbeck & Praseyto 2000). Burial objects in these graves include bronze and gold jewellery, iron weapons as well as Song, Yuan and Ming ceramics (Bougas 1998). Unfortunately, most of these gravesites have been looted (ibid.) and no tripod vessels, bronze mirrors or shell spoons have been documented.
42Although the majority of known soapstone objects originate from the tombs at Vohemar, a number of soapstone objects have been discovered at several other locations along the east coast of northern Madagascar. These objects include well-shafts found South of Vohemar at Bemanevika near the mouth of the Bemarivo river and at Angolovato near the mouth of the Mahanara river (Vérin 1986) and an unfinished basin measuring 143 by 94 cm left near a quarry at Amboaimohehy (ibid.). Other soapstone objects found in the Vohemar region include a block perforated with a circular hole and used as a foundry tube (Vernier & Millot 1971 49), and a knife sharpener with part of an iron blade still attached found in a grave at Antanandava (ibid.).
43About 750 km south of Vohemar, a soapstone sculpture of an animal is present in the village of Amobhitsara (e.g. Jully 1898, A. Grandidier & G. Grandidier 1908, Griffin 2009). The Malagasy refer to the statue as vatolambo (stone wild boar) or vatomasina (sacred stone). The sculpture measures 106 cm in height, is largely hollow and is described in detail by Molet & Vernier (1956), who suggest that it resembles the Sumatran elephant (Elephas maximus sumatranus). The presence of soapstone quarries in the immediate surroundings (e.g. Griffin 2009) suggests that the sculpture was made at or near Ambohitsara and that the Rasikajy migrated along the east coast of Madagascar. It is speculated here that the stone sculpture might have played a role in the funeral rituals of the Rasikajy. In ancient China, the path leading to tombs of important persons was often aligned with stone sculptures representing (mythical) animals or humans, a tradition that started in the Han Dynasty and lasted well into the 20th century (Paludan & Wilkinson 1998).
Fig. 9. Soapstone statue at Ambohitsara.
© Institut de civilisations du Musée d’art et d’archéologie de l’université d’Antananarivo, Antananarivo, Madagascar.
44Quarries from which soapstone was extracted have been discovered along the east coast as far south as Mananjary. However, objects made from soapstone and soapstone sherds are not restricted to the east coast, but have been found in other parts of Madagascar including southern Madagascar (Rakotoarisoa 1998) and northwest Madagascar (Radimilahy 1998) indicating trade with soapstone objects. An example is a pear-shaped weight (tronc de cone) with an Arabic inscription (in Vernier & Millot 1971) at Mahilaka, a trading port on the northwest coast since the end of the first millennium (Radimilahy 1998).
45There is also evidence for export of soapstone objects to the Comores (basin at Siwa; Barraux 1959) and Tanzania (tripod vessels in Kilwa; Chittick 1966a, 1966b).
46The locally produced soapstone objects, metal objects and shell spoons found in the tombs and the funeral rituals (e.g. bronze mirrors, plates, bowls, spoons and jars) suggest that the former Rasikajy civilization at Vohemar possibly had Chinese roots. The presence of soapstone objects in Comores, Somalia and Tanzania suggests that the Rasikajy participated in the Indian Ocean trade network.
47Comparisons with ceramics found in SE-China kilns and ceramic cargoes recovered from dated ship wrecks indicate a Yuan and early Ming dynasty origin (14th to early 16th ce) for most of the ceramics found at the gravesite and suggest that the Rasikajy had settled at Vohemar prior to the first visits by Europeans to the island, i.e. prior to 1500 ce. The presence of 16th century Chinese ceramics in the tombs indicates that Rasikajy were still being buried at Vohemar when the Portuguese made their first landings on the coast of Madagascar. The early 16th century Manises pottery at the Vohemar gravesite probably arrived in Madagascar with the Portuguese. Historical sources indicate that a Portuguese ship, commanded by Pedreanes, traded at Bemaro (Vohemar) in 1514 or 1515, where he bought a great deal of amber (Barros 1777). It is thus possible that Rasikajy and Portuguese traded directly at Vohemar. Megiser (1609) compiled descriptions on Madagascar from different sources, and noted that not only Islamic communities live in coastal towns alongside the indigenous population, but also white people presumably from China. The original German quote from Megiser (1609), which includes this description is given below:
Nu ist es auch zeit zu melden was für Leute hat in dieser grossen und mechtigen Insul Madagascar. Ist derwegen zuwissen, dass die geborne Inländer gar kolschwarz sind. Doch ist es nicht ohn es wohnen bei ihnen auch sonderlich in den Städten an den Meergränzen viel Mahometaner Mohren so auss Arabia dahin kommen welche weniger schwarz und etwas bräuner ausssehen. Es werden auch etlich gar weisse Leut unter inen gefunden die sollen auss der China dahin kommen sein.
48The timing of arrival of the Rasikajy in Madagascar is not well constrained. Vérin (1986) reports the presence of soapstone objects at Talaky in southern Madagascar and dates them to the 10th to 11th century. Soapstone production in the Vohemar region seems to have ceased in the 16th century. There are no clear indications for post-16th century objects in the graves at Vohemar (Crick, pers. comm.). Although an important necropolis existed at Vohemar, the site of the settlement that most likely existed nearby has not yet been found and the reasons for the demise of the Rasikajy civilization remain unknown.
49The presence of an overseas community with Chinese roots at Vohemar would not only have consequences for the maritime history of the Indian Ocean, but also for the history of Madagascar. The presence of such a community might have influenced customs and traditions of the local people.