Navigation – Plan du site

A Chinese in the Nubian and Abyssinian Kingdoms (8th Century)

The visit of Du Huan to Molin-guo and Laobosa
Wolbert G.C. Smidt


La curiosité géographique de la dynastie T'ang permit à des extraits d'un récit de voyage du VIIIe s. rédigé par un officier militaire chinois, Du Huan, d'être conservés. Il est le premier Chinois dont la présence en Afrique de l'Est soit clairement attestée. Son récit mentionne plusieurs régions dont "Laobosa" (Abyssinie), évoquée pour la première fois dans une source chinoise ancienne. Les doutes concernant la localisation d'une autre région appelée "Molin" semblent pouvoir être levés aujourd'hui. Située probablement au nord de Laobosa, elle correspondrait à la plaine côtière désertique de l'Erythrée et du Soudan actuels. Ce texte livre de ce fait un des très rares témoignages contemporains sur le royaume axoumite tardif, accompagné de brèves notations sur les coutumes des populations de ces différentes régions.

Haut de page

Entrées d’index

Mots-clés :

récit de voyage

Géographique :

Haut de page

Texte intégral

(Du huan in chinese)

1Ancient contacts between China and the Aksumite kingdom, in today's Ethiopia and Eritrea, are mentioned from time to time by scholars1 and seem to be well established. But only very few publications go into more detail. Even if there were no historical or archaeological evidence, it would, however, at least be likely that Chinese and Aksumites did know of each other. The presence of Indian merchants, who were also in commercial contact with Chinese coastal towns, at the Adulis coast is attested in ancient texts and can also indirectly be drawn from archaeological findings of coins from ancient India in Eritrea and Ethiopia2. Red-Sea-merchants are also known to have been present in India, coming into direct contact with Chinese there3. Early Chinese sources of the Han dynasty (206 B.C. - 220 A.D.) mention products from the "South" like pearls, tortoise shells, incense and spices, part of which might have been imported from the Red Sea area - in exchange with silk4. Silk is one of the goods from China, which has probably reached the Red Sea already in early times5. The 1st century Periplus Maris Erythraei, which is one of the few ancient Greek sources on the Red Sea, also describes the Chinese coast and mentions silk as a trade product along the coasts of the Indian Ocean6.

2This is just to refer to some examples, which make it exciting to go further in the search for other cultural influences or exchanges which might reach not only as far as to the Agaeïs in the Mediterranean Sea in the North, but also as far as to the Indian and Chinese Ocean in the Far East. Traditionally scholars concentrated on influences on ancient Ethiopian history coming from Southern Arabia, ancient Egypt and Greece - leaving apart the most important influences from inner Africa which are still very much ignored7 -, but in the future we might still find more about aspects of Asian influences8.

3Contacts with China in ancient sources

4Aksum in late antiquity played a considerable international role. As a power controlling the African shores of the Southern Red Sea, it served as an intermediary between the Byzantine - Arabic trade routes and the Indian Ocean - and thus became an interesting potential ally to Byzantium. To give one example : several sources mention the wish of the East-Roman emperor Justinian to control the route to India via the Red Sea in the 6th century9. With the help of its ally Aksum, the Persian incursions into Southern Arabia were then, in fact, halted ; South Arabia was occupied by the Aksumites and trade routes were secured. As Procopius Caesarensis reports in his De bello persico, the East-Romans had always purchased silk from the Persians, their enemies ; now there was hope that silk trade would become possible by sea, which connected Aksum with India ­ a project which wasn't realised, however, as the ports on the way to India were controlled by the Persians. The influence of Indian scripts on the development of the Ge'ez script shows the cultural impact of the presence of Asians at the Aksumite court10.

5The Greek traveller Cosmas Indikopleustes, allegedly a merchant turned into a Byzantine monk, in his detailed Universal Christian Topography (545 A.D.) describes the trade routes of the Red Sea and the Indian Ocean, joining as remote areas as Adulis, the Ethiopian port, and China :

6Ceylon "being as it is in a central position is much frequented by ships from all parts of India and from Persia and Ethiopia (...) And from the remotest countries, I mean Tzinista [China] and other trading places, it receives silk, aloe, cloves, sandalwood and other products, and these in turn are passed on to markets this side, such as Malê, Kalliana (...) Persia and the Homerite [Himyarite] country (...) and Adulis11. "The sinologist Philip Snow notes that Ceylon was a place, where Chinese and Adulite merchants should have met occasionally12. There are Chinese travel accounts on Ceylon especially of the 5th century, mentioning the presence of a number of foreign merchants there. Cosmas himself notes that traders from Adulis were in fact in Ceylon. Contrary to Arab and Indian traders, the Chinese were not very active in sea commerce during most of the periods in their history. However, there were periods, during which they entered into active contact with Western and South-Western kingdoms. In fact, there are Chinese sources of the Han dynasty which seem to contain information on the African coast of the Red Sea.

7Several sources mention the kingdom "Huang-tchi" in the Western Sea (recorded between the 1st century B.C. and 1st century A.D.), which has been located in Africa. From the ancient descriptions of the travel by boat (analysed by Herrmann 1913), we know that it took over one year. Therefore it seems clear, that it was far from India, in a Western direction. It had been suggested that it corresponded to today's Eritrean coast, "Huang-tchi" being a deformation of "Ag'azi"13 . Recent archaeological findings reconfirm that the Eritrean highland areas close to the coast had already reached the stage of state formation very long time ago, but certainly in the period concerned. The product, which was sent by the Huang-tchi-ruler to the Chinese emperor, is African - a living rhinoceros. Rhinoceros horn is known from other sources of the same period as an African export product traded in the Indian Ocean14; also tortoise-shell and ivory are exported by Huang-tchi15.

8One more source of the 1st century A.D., the Houhan Shu, might also refer to the Eritrean coast : Gan Ying, a Chinese envoy to the East Roman regions via the sea in the year 97 A.D., never reached Egypt - but a country called "Dou-le". This has recently been identified as Adulis16, the ancient port of today's Eritrea. The envoy Gan Ying returned with an embassy from "Dou-le" to Luoyang, the Han capital. We don't, however, know much today on this period of Northeast-African history, called "Pre-Aksumite".

9T'ang dynasty contacts to Africa

10Much better documented, mainly through inscriptions, is the later period of the Aksumite kingdom, which was at the peak of its power in the 4th - 6th centuries A.D. In China the centre was weak in that time ; only under the T'ang dynasty (618 - 906 A.D.) did China win back its strength and some interest for foreign peoples reappeared. Therefore, the first dynasty which might again provide information on African ports after the Han is the T'ang dynasty. Snow provides us with a map showing localities at the East African coast where Chinese porcelain had been found. Unsurprisingly there are many findings at the Kenyan coast, but also some at the African Red Sea coast near major ports in the Northern Somalia-Djibouti-area, especially near Zayla', and Northern Sudan / Eritrea, near the Sudanese'Aydhâb, from the 14th century, and near the port Badi north of Adulis, from the T'ang Dynasty17. There is written evidence that 9th century T'ang dynasty merchants were present not only in India, but further West in Islamic areas18 - following their ancestors, who seem to have travelled to the Red Sea already in the 5th century19 -, which suggests, that some of them may have continued to ports at the African coast.

11A research for additional ancient Chinese texts, reflecting those contacts, could therefore be promising and might serve as a minor source for the understanding of the history of the Horn of Africa20. For example, we already know about one 9th century T'ang dynasty text mentioning "Bo-ba-li", the Berber coast (Somaliland) - about which later. Extracts in an ancient Chinese encyclopaedia of an 8th century text refer to the travel of a Chinese to Northeast Africa. The source itself had been lost. But the fragments are giving us an important impression of the geographical knowledge in this period of competition with their ambitious Arab neighbors.

12Du Huan : From Uzbekistan to Molin-guo and Laobosa

13In the case of the traveller Du Huan, it were the Arabic contacts to the African coast of the Red Sea which led to this early encounter between a black people and a Chinese. And the water-ways, which were used by Indians, secured his way back21. A military disaster in today's Uzbekistan was the initial catastrophe which led to the encounter between representatives of China and - as I will show - the Aksumite kingdom, two ancient "world powers"22.

14Du Huan from Wannian23, an officer of the T'ang dynasty army, was involved in a catastrophic defeat for the Chinese near Tashkent in the year 751 and was captured together with over 20,000 other Chinese soldiers by the "Dashi" (the Arabs) and then brought to "Yajuluo" (near Baghdad in Iraq), the centre of the Abbasid caliphate24. Most Chinese prisoners became gold- and silversmiths, painters, weavers, and paper fabricants. After long journeys through Arab countries, Du Huan returned by ship to Kanton (Guangzhou). There he wrote his Jingxingji ("Record of My Travels"), which was almost completely lost except a few extracts, which give information on central Asian and Arab countries among others.

15His relative Du You (735-812) in his encyclopaedia Tongdian mentions Du Huan's travels with the following words, in a chapter discussing the defense of the borders to the West :

16"The relative Huan was fighting in the West under General Gao Xianzhi ; in the 10th year of the government device Tianbao [751] he arrived at the Western Ocean ; in the first year of the government device Baoying [762] he returned to Kanton on board of a trade ship and wrote the Jingxingji."

17There are seven more notes on Du Huan's travels in Du You's encyclopaedia. On Molin, Du Huan himself is quoted as following :

18"We also went to Molin, Southwest of Yangsaluo25. One reaches this country after having crossed the great desert and having travelled 2,000 Li26. The people there are black, their customs rough. There is little rice and cereals and there is no grass and trees. The horses are fed with dried fish, the people eat XX [word not identified] and also Persian dates. Subtropical diseases [Malaria] are widespread. After crossing the inland there is a mountainous country27, there are a lot of confessions.

19The followers of the confession of the "Dashi" (the Arabs) have a means to denote the degrees of family relations, but it is degenerated and they don't bother about it. They don't eat the meat of pigs, dogs, donkeys and horses, they don't respect neither the king of the country, neither their parents, they don't believe in supernatural powers, they perform sacrifice to heaven and to no one else. According their customs every seventh day is a holiday, on which no trade and no cash transactions are done, whereas when they drink alcohol, they are behaving in a ridiculous and undisciplined way during the whole day.

20Within the East Roman confession28 the medical doctors know diarrhoea - or they recognise it already before the outbreak of the disease, or they open the brain and insects come out."

21Another passage in which Du Huan gives a brief geographical overview has been preserved in another work, the XinT'angshu of Ouyang Xiu. In chapter 221 (Dashi-zhuan) he notes :

22"Coming from Fulin29 (Baizhanting30) one reaches, after having crossed the desert in a Southwestern direction and having travelled for 2,000 Li, a country called Molin31, respectively Laobosa."

23Snow and Shen both think that Du Huan's mention of a third religion, the "Zimzim teaching" (in pinyin : Xunxun) in another fragment is also referring to Molin32. This is certainly due to the fact that the Molin-fragment mentions "many religions", but stops while describing the second. A third religion or confession, therefore, should be the "Zimzim" (or Zemzem). This is an Arabic term, in its strict sense meaning the Zoroastrians of Persia33, but in a wider sense pagans in general. The Zimzim "practise incest, and in this respect are worst of all the barbarians34". In fact some peoples of the African coast (such as the Beja, known as rebellious subjects of the Aksumite kingdom already from 4th century inscriptions), were neither Christians nor Muslims - the Beja were only christianised in the course of the 8th and 9th centuries. Other peoples in the area are, still today, following traditional religions.

24Further notes on Du Huan's travel report Jingxingji are found in the Taiping yulan, the Taiping huanyu ji, the Tongzhi and the Wenxuan tongkao, but without giving more information on Molin35.

25Identification of the countries visited Theories

26The discussion on which area exactly should be identified with Molin has brought a number of ancient African locations at the Red Sea coast and the Indian Ocean to our attention. A Chinese encyclopaedia of 1995 still notes that the location of Molin is not know36. After the publication of an important article in Chinese37, however, we have some convincing results, even if the exact etymology at least of Molin will stay subject to discussion. Over about 130 years a number of theories were developed, but all of them agreed that Molin should be located in Africa. We don't have many hints. It should be noted that the distance, which is given only roughly, was 2,000 Li (which corresponds to 1300 km approximately), and that the country Molin was lying in a Southwestern direction, seen from Jerusalem. If we follow this information, we probably have to locate the country visited somewhere in the Sudan or south of it ­ and as we know that Du Huan left Molin by sea, it should be a coastal country.

27First theories

28The first one who tried to locate Molin was Bretschneider in 1871, who identified it with the countries of the "moors", Mauritania or Libya. However, this doesn't fit into what we know about its approximate location. In 1885 the sinologist Friedrich Hirth, who referred to the above mentioned passage of the XinT'angshu, located Molin and Laobosa at the African Red Sea coast, in the area of the "Troglodytes" (the ancient Greek denomination of the inhabitants of the African Red Sea coast)38. Later in 1909, he underlined that Molin should be an East African state, but couldn't locate it more precisely39.

29Molin = Malindi

30In the early 15th century Cheng Ho (in pinyin : Zheng He), the famous Muslim eunuque admiral of the great Ming dynasty fleet, discovered the Eastern coast of Africa. Sailing via the Persian Gulf his ships first reached Aden at the mouth of the Red Sea, and then travelled further South to Somalia and the Kenyan coast. The fleet made several expeditions from 1417 to 1433, before the exploration was stopped probably for economic reasons. The identification of 8th century Molin with Malindi in Kenya40, which played such a prominent role in the Chinese discovery of Africa, is rather a result of Zheng He's fame, but doesn't correspond to Du Huan's texts. Malindi barely existed in the 8th century ; the multitude of confessions described by Du Huan and the allusion to disrespected kings doesn't seem to correspond to its past. But even more important is the argument, that Malindi could not in any way be reached after a journey from Jerusalem through a desert, followed by a mountainous country. A much more complicated journey through the land would have been necessary - and it was much longer distance than 2,000 Li.

31Molin = the god "Mahrem"

32Others locate it much more in the North, in the distance reported by Du Huan. It had been suggested that Molin should be another form of "Mahrem", the name of the Aksumite "sea god"41 But Mahrem, the most important god of Aksumite kings, was actually a god of war, not of the sea, and possibly linked with the moon42. No area was called after any god in that region. And, even more important, in the 8th century the ancient South Semitic gods had not been worshipped any more since centuries. Only one inscription could possibly suggest that the name of Mahrem might have survived for some time the rise of Christianism : in the mid-4th century trilingual inscription of king Ezana, which refers to three South-Semitic gods, new Christian monotheistic conceptions seem to have been linked with Mahrem, which appears as the ancestor of the king43. Only shortly afterwards Ezana adopted Christianism as a state religion. But after this isolated and only partial identification of Mahrem with characteristics of the Christian god, Mahrem doesn't appear any more. So, this theory still stays highly improbable.

33Molin = Malaô

34Surprisingly Molin has, so far, not been identified with the city of Malaô in the country of the "Barbaroi" on the Horn of Africa, corresponding to today's Berbera in Somaliland. The anonymous Periplus Maris Erythraei of the 1st century A.D. provides us with very detailed information for seafarers on the ports of the Red Sea, on the natural conditions, on trade and products, and, quite roughly, also on the political status of the regions bordering the Red Sea and the Indian Ocean. The names of places and peoples are given in Greek, but often are corrupt forms of local names. Locations of ports are described so exactly, that they are still identifiable today.

35Malaô was a port of the Berbera people, an ancient Somali group. Since the 13th century the name Berbera is recorded for the port, but the people's denomination is considerably more ancient. The original form of the Greek denomination Malaô is not known. Etymologically it might be connected in some way to the Somali (mentioned first in 15th century Ethiopian chronicles) - and the similarity to Molin is striking. It should, however, be noted that the multitude of confessions cannot be applied to the Berbera land. In the 8th century, they were already, at least partially, islamized and not subject of any kingdom. Quite important is another fact, still. There is another Chinese source of the 9th century which can be related to Berbera (Arabic Barbara)44. The area is called "Bo-ba-li", a denomination which was still used several centuries later in Chinese geographical texts45. It produced ivory and ambergris. The description of the people is worth to be quoted here, as the Somali nomads are clearly recognisable ; still today a mixture of raw blood and milk is consumed by several peoples of the Horn of Africa :

36They "do not eat the Five Grains, but only meat. They are given to sticking a needle into the veins of their cattle and drawing out the blood, which they mix with milk and consume raw. They wear no clothes, but merely use goatskins to cover the parts below their waists46."

37Molin, however, never appears in this context, which makes it improbable that it was located in the Malaô area. It should, however, be noted that this is the second ancient T'ang dynasty text providing information on an African area.

38Molin = Meroe

39Recently it had been suggested that Molin should correspond to "Meroe"47 If we can accept the hypothesis that the place-name Meroe still existed in that time it is a possible solution, even if the state of Meroe had collapsed long before (it had been raided in ca. 350 A.D. by the emergent Aksumite kingdom). But the name had not completely disappeared. The centre of the Christian Nubian State of Napata, a successor state of Meroe to the North in today's Egypt, is the city today called Merowi. It is possible that Du Huan went to a settlement with this name, and derived the country's name from the city. In fact, names of countries were often derived from settlements in ancient Northeast-Africa ; and such names often survived locally for long. Geographically it would perfectly correspond to the few data we have. It is not farer away than 2,000 Li approximately, it can be reached after a long journey through the desert, trade routes link it with a mountainous country (Abyssinia) and with coastal ports, from where merchants sailed to South Arabia and even further to India.

40Molin = Muqurra

41I myself would suggest an etymological connection with Muqurra, another ancient Christian Nubian kingdom south of Napata, which was centered in the city of Dongola (Dunqula). In the sources it also appears as al-Muqurra or as Mekuria. The unpronouncable "q" would have been left out (the "o" of Molin is long and has a falling and rising tone, which might point to a consonant between two vowels which had been dropped), the "r" replaced by "l", which would lead to a word similar to Molin. I believe, that this solution would be preferable to an etymological connection to Meroe. However, this would not change our idea about the area visited by Du Huan ; in contrary, it reconfirms it. The kingdoms of Napata and of Muqurra had merged before ca. 700 due to pressure by Arab armies (Egypt had been conquered in 640), the capital now becoming Dongola - and the name of Muqurra being used for the kingdom (which only in later sources was called the kingdom of Dongola

42after its capital). Trade routes linked it with ports like'Aydhâb and Badi north of the Aksumite port of Adulis and to the Christian Aksumite kingdom in the highlands, with which close connections are documented especially in the mid-8th century48. In competition with Islam the bishop Barnabas in 700 A.D. tried to christianise the Beja at the coast, while he was ministering the Christian merchants of the port of'Aydhâb. This also makes it possible, that Du Huan also understood the whole coastal areas as part of Molin (today corresponding to the Sudanese and north Eritrean coast). The expansion of Christianism was followed by a rise of a puritanical Islamic movement in the nearby Barka valley (today an Eritrean region bordering the Sudan) in 759 A.D. and finally reaching the highlands in 765. Some connection with Dongola had already been suggested by Shen, which I think can be reinforced by this etymological argument49.

43Laobosa = al-Habasha

44Finally, the geographical description and the second geographical term indicate which region is referred to by Du Huan. He mentions a great desert to be crossed ; and in fact we know that from ancient times the way through the Sudanese desert led to the edges of the Ethiopian highland, a trade route used since times immemorial. The only mountainous country which lies in this direction is highland Ethiopia, which was dominated by the Aksumite kingdom. The term "Laobosa" is easily identifiable : it can only be "al-Habasha", the ancient High Arabic term used for the Ethiopian highlands and its kingdom50. This appearance is already an interesting fact, as it is one more example for the use of this term51.

45The term al-Habasha is very ancient52. In the Ethiopian region it regularly appears in the mid-4th century inscriptions of the Aksumite king Ezana53, who controlled the Ethiopian highlands and the coast of Adulis, and had subdued a number of people even as far as Meroe. In classical Arabic this term is territorial (etymologically it is at the origin of the European term "Abyssinia"54). It corresponds perfectly to the Chinese "Laobosa", which equally was a territorial denomination55. Also in al-?abarî's Annals (9th/10th century) "al-Habasha" is a geographical term in contrast to the ethnonym "al-Habash". The Arab historiographer ibn Wadîh al-Ya'qûbî in his 9th century History describes "al-Habasha", lying north of the countries of the Berber (Somali) and of the Zanj (the "Blacks"). All these three terms appear later in Chinese sources ; it can generally be noted, that all early Chinese geographical terms related to Africa were Arabic loanwords.

46Du Huan's text mentions an erosion of state power, which corresponds to what we know from the late Aksumite kingdom, in which the central power was weakened, and especially on the coastal areas at least partially lost control after the Islam had started to dominate the coastal plains in the 7th century56. The way how Du Huan uses both terms, Molin and Laobosa, doesn't make it clear, if two neighbouring regions were meant or if it was only one region. It had been suggested that they were the same57. But Du Huan's text doesn't exclude that it is two countries58. The verb connecting the words Molin and Laobosa means "say" in ancient Chinese. It is possible to translate this passage as "countries called Molin and Laobosa" instead of "a country called Molin, respectively Laobosa" as we did above. For this case Shen suggests that Laobosa should lie further south.

47If we can believe that Molin was a derivation of Muqurra, and also taking into account the fact that it was only 2,000 Li far from Jerusalem (which is enough to reach the Sudan, but not highland Ethiopia), we should understand the terms as denoting two neighbouring countries, but closely interrelated (which was in fact the case between the Aksumite and Nubian kingdoms). It seems that the term Molin has been used to roughly denote the lowlands, including the coastal regions. Du Huan left Molin by boat ; for this he should have used a port, which was frequented by Arabic or possibly

48Indian ships. This suggests that it had been Adulis, the outlet to the sea of the Aksumite kingdom, or, as this port was in decline in that period, Badi.

49The year in which Du Huan visited and left Molin

50We know that Du Huan came back to China in 762 A.D., probably in June due to the monsoon. He should have left the African port of Molin at least in the year before, in 761 A.D., if we consider that he had first to make the journey via the Persian Gulf to India and then further. This date is most interesting, as it might even provide a reason explaining why Du Huan has visited Molin at all. As Molin should be identified with areas in today's Sudan and Eritrea, historical events of that time might be connected with his travel. In fact, the year 758 was of crucial importance for the Christian kingdom at Dongola. Dongola had broken off with Egypt in the 740s, which led to subsequent wars. As a result the king of Dongola had to send his son to the Abbasid Caliph in Baghdad in 758, as a hostage. This was followed by an exchange of embassies. This might well have been the occasion which brought Du Huan, who was based in Iraq, to these Southern kingdoms and provided him with a chance to travel back to China by sea. Shen speculates that Du Huan accompanied the Dongola prince back to his country59 ; we know that foreigners often acquired a position at Arabic courts, this might also have been the case with Du Huan60. In addition to that, we know that the Nubian kingdoms especially in the 8th century were in contact, both economically and politically, with the Aksumite kingdom. This would easily explain the mention of “Laobosa” (al-Habasha). The attractiveness of this hypothesis is certainly undeniable. But Du Huan's fragments don't give us any reason for his wide travels.

51Ethnographical information

52Du Huan's ethnographical sketches concern both the highlands (al-Habasha / Laobosa) and the desert lowlands near the sea. The highlands, which lie further "inland", are characterised by a great number of religions. His description of the region and its peoples is vivid, even if brief. The T'ang dynasty's understanding of Islam and Christianism is perfectly reproduced in Du Huan's text : they adore the heaven and no one else. The T'ang dynasty China knew both religions, Nestorian Christianism was well established61. The diversity of religions described by Du Huan corresponds to the late Aksumite kingdom and its coastal areas, Islam and Christianism mixing with ancient traditional religions in that time of continuous loss of power of the Aksumites and the rise of Islam at the coast. Still today traditional religions are practised in the Eritrean inland in some remote Kunama areas. These people, who lived most of the time outside direct control of any states, preserved an original local religion for very long.

53From Du Huan's descriptions we get a clear picture of a hot desert landscape at the coast : dry fish are fed to horses, there is a lack of cereals, there are no trees at all. This was in fact true for all coastal areas from the Sudan to the Eritrean coast. Further details can easily be related to the Aksumite kingdom : people take rest every seven days (practised in Islam and Christianism), they don't eat the meat of pigs (which is typical not only for Muslims, but also for Abyssinian Christians until today), alcohol is a traditional beverage, also for Muslims. Muslims are indirectly described as traders - as they stop their trade every seventh day - and they are acquainted with money. Certainly they traded also with Persia, as Persian dates are mentioned. In fact, Muslim traders also lived in the Abyssinian highlands in scattered villages amidst Christian areas and still do today. Genealogy still today is an important part of the identity of the people and do have a complex system of denoting degrees of blood relations - which corresponds to Du Huan's remark on the Muslims, who denote the different degrees of family relations. His remark, that they however don't care about it, seems to point to the decline of social stability of that period.

54Traditional medicine is described as strong among Christians ; as Christians were holders of the traditions of the Aksumite court, it is possible, that they also had deeper knowledge of it than the Muslim traders. Medicine in today's Ethiopia and Eritrea is in fact relying on very old traditions, often preserved within the low clergy of the Christian church, some practises preserving knowledge from ancient treatises which arrived in Aksum during the Byzantine Empire. It might be a bit surprising to hear that "they open the brain" ; but this

55should correspond to a very ancient medical practise still known today in rural areas of Eritrea and Ethiopia. Many diseases are healed by cutting the skin of the forehead, the eyebrow or the cheek. Then blood pours out and, together with it the bad spirit ("insects"), which caused the disease. Diarrhoea, also mentioned by Du Huan, is extremely widespread and traditional medicine is well known against it. Thus, Du Huan's notes correspond well to the coastal areas of today's Eritrea and give us a glimpse on the life in the late, politically weakened Aksumite kingdom.


57Indirect contacts between the Northeast African coast and China, mostly based on trade, seem to be documented since at least the 1st century A.D. (Han dynasty). This article focused on the first Chinese, whose presence in Africa is clearly documented. The geographical curiosity of the T'ang dynasty made it possible, that extracts of an 8th century travel report of a Chinese military officer, Du Huan, were documented and preserved. He visited Arabic countries and also Africa. The location of Molin-guo, an African country, seems to be clarified today. South of it lies Laobosa, the first mention of Abyssinia in an ancient Chinese source ; Molin should be located in a dry desert lowland in the Sudan and Eritrea62. This text is thus one of the very few ancient sources contemporary to the late Aksumite kingdom. Briefly, but with significant details, the peoples' customs are described, most of them still identifiable today.

58What makes the text fragments on Molin important, is, first, that it is the most ancient source which provides evidence of the presence of a Chinese in Africa63. Most other sources only reflect the fact that products from the African Red Sea coast reached China. Second, it is among the very few ancient sources contemporary to the Aksumite kingdom. Third, the Du Huan travel report fragments give an example on the ways used in that time to cross far distances. Probably rather unusual is the fact, that Du Huan reached the mountainous region south of Egypt by land ; how his way back was organised is not known in detail, but we know that he reached Kanton in Southern China by boat - using the sea routes to India, which we know from a few ancient sources. This shows that still in the period of decline, the sea trade from the African coast was not halted. Fourth, the geographical notions mentioned are reflecting information originating evidently in Arabic contacts, which is probably due to the fact that during the decline of the Aksumite kingdom the Red Sea trade was already in the hands of the Arabs. Finally, the brief descriptions of the multitude of religions give us an impression of the coexistence of religions in this period, on which only very few sources exist.

Haut de page


S. I. Berzina, 1984 : "Kushana coins in Axum", Information Bulletin, International Association for the Study of the Cultures of Central Asia, Issue 7, Moscow, 57-64.

H. Brakmann, 1994 : Die Einwurzelung der Kirche im spätantiken Reich von Aksum, Bonn.

E. Bretschneider, 1871 : On the Knowledge Possessed by the Ancient Chinese of the Arabs and Arabian Colonies, and other Western countries, mentioned in Chinese books, London.

F. Böckelmann, 1998 : Die Gelben, die Schwarzen, die Weißen, Frankfurt am Main.

L. Casson (ed.), 1989 : The Periplus Maris Erythraei, Text with introduction, translation, and commentary, Princeton, N.J.

E. Cerulli and G. S. P. Freeman-Grenville, 1960 : "Somali, History", The Encyclopaedia of Islam, New Edition, vol. IX, Leiden - London, 715-16.

Chinese Great Encyclopaedia, 1995 : "Du Huan", Zhong guo da bai ke quan shu ('Chinese Great Encyclopaedia'), Beijing, vol. 2, 1118.

V. Christides, 1994 : "New Light on Navigation and Naval Warfare in the Eastern Mediterranean, the Red Sea and the Indian Ocean (6th-14th centuries A.D.)", Nubica, Internationales Jahrbuch für koptische, meroitisch-nubische, äthiopische und verwandte Studien, III/1, 3-42.

L.M. Devic (ed. and tr.), 1883-1886 : Kitab Ajâ'ib al-Hind, Leiden.

A. J. Drewes, 1956 : "Nouvelles inscriptions de l'Éthiopie", Bibliotheca Orientalis 13, 179-183.

J. J. L. Duyvendak, 1949 : China's Discovery of Africa, London.

Gungwu Wang, 1966 : "'Public' and'Private' Overseas Trade in Chinese History", Sociétés et Compagnies de Commerce en Orient et dans l'Océan Indien, M. Mollat (ed.), Paris, 220-221.

E. Haberland, 1965 : Untersuchungen zum äthiopischen Königtum, Wiesbaden.

W. Hahn, 1999 : "Das Kreuz mit dem Abessinierland, Epigraphische Anmerkungen zu einer axumitischen Münzlegende", Mitteilungsblatt des Instituts für Numismatik der Universität Wien, 18, 5-8.

A. Herrmann, 1913 : "Ein alter Seeverkehr zwischen Abessinien und Süd-China bis zum Beginn unserer Zeitrechnung", Zeitschrift der Gesellschaft für Erdkunde, 10, Berlin, 553-561.

F. Hirth, 1885 : China and the Roman Orient, Leipzig - München.

F. Hirth, 1909 : "Early Chinese Notices of East African Territories", Journal of African and Oriental Studies, 30.

T. O. Höllmann, 1992 : "Wo der Himmel endet. Vom Umgang mit fremden Ländern, Völkern und Kulturen im chinesischen Kaiserreich", Focus Behaim Globus. Germanisches Nationalmuseum, Nürnberg 2. Dezember 1992 bis 28. Februar 1993, Teil 1: Aufsätze, Nürnberg, Ausstellungskataloge des Germanischen Nationalmuseums, Hrsg. v. G. Bott, 411-422.

F. Hommel, 1926 : Ethnologie und Geographie des alten Orients, München, Handbuch der Altertumswissenschaft, Abt. 3, Teil 1, Bd. 1.

H. Ibscher, 1940 : Manichäische Inschriften der Staatlichen Museen Berlin, Stuttgart.

A. K. Irvine, 1965 : "On the identity of Habashat in the South Arabian inscriptions", Journal of Semitic Studies, vol. X, Manchester, 178-196.

A. Kammerer, 1929 : La mer Rouge, l'Abyssinie et l'Arabie depuis l'antiquité, Le Caire.

Y. M. Kobischanow, 1969 : "The Sea Voyages of Ancient Ethiopians in the Indian Ocean", Proceedings of the Third International Conference of Ethiopian Studies, Addis Ababa, 19-23.

B. Laufer, 1919 : Sino-Iranica, Chicago.

J. Legge (tr.), 1886 : The travels of Fa-Hsien, Oxford.

A. Manzo, 1998 : "The dynamics of external contacts of Northern Ethiopia and Eritrea from proto-historical to Aksumite times, Late 2nd millenium BC - late 1st millenium AD", Äthiopien und seine Nachbarn, Ethiopia and its Neighbors, Orbis Aethiopicus, 3. wissenschaftliche Tagung des Orbis Aethiopicus, The 3rd Academic Conference of Orbis Aethiopicus, Gniew (...) 1997, Frankfurt, 35-52.

J. A. Mar, 1997 : Makéda ou la fabuleuse histoire de la Reine de Saba, Paris.

G. Mathew, 1956 : "Chinese Porcelain in East Africa and on the Coast of S. Arabia", Oriental Art, 2, 50-55.

A. Mordini, 1960 : "Gli aurei Kushana, del convento di Dabra-Dammo. Un indizio sui rapporti commerciali fra l'India e l'Etiopia nei primi secoli sell'era volgare", Atti del Convegno internazionale di studi etiopici, Roma, p. 249-54.

D. Morin, 1998 : "The Kingdom of Dankali, Ethiopia and Yaman (Baylûl Revisited)", Äthiopien und seine Nachbarn, Ethiopia and its Neighbors, Orbis Aethiopicus, 3. wissenschaftliche Tagung des Orbis Aethiopicus, The 3rd Academic Conference of Orbis Aethiopicus, Gniew (...) 1997, Frankfurt, 53-57.

S. Munro-Hay (awaiting publication) : "A Chinese Source for Aksumite History in the Sixth and Seventh Centuries AD", article prepared for the Pankhurst Festschrift.

S. Munro-Hay, 2002 (forthcoming) : "China", Encyclopaedia Aethiopica, S. Uhlig (ed.), vol. I., Wiesbaden.

T. Natsoulas, 2002 (forthcoming) : "Byzantine Empire", Encyclopaedia Aethiopica, Siegbert Uhlig (ed.), vol. I., Wiesbaden.

RIE, 1991 : E. Bernand, A. J. Drewes, R. Schneider, Recueil des inscriptions de l'Éthiopie des périodes pré-axoumite et axoumite, Paris.

Sergew Hable Sellassie, 1972 : Ancient and Medieval Ethiopian History to 1270, Addis Ababa 1.

S. E. Sidebotham, 1989 : "Ports of the Red Sea and the Arabia-India Trade", The Eastern Frontier of the Roman Empire: Proceedings of a Colloquium Held in Ankara in September 1988, D. H. French - C. S. Lightfoot (ed.), Oxford, BAR International Series 553 II, suppl.

Shen Fuwei, 1980 : "T'ang dai du huan de mo lin zhi xing" ('Du Huan's Journey to Molin in the T'ang Period'), Shijie lishi ('World History'), n° 6, Beijing [in Chinese].

P. Snow, 1988 : The Star Raft, China's Encounter with Africa, London.

Tse'n Chung Mien, 1982 : "Sui's North-Western Connections", The History of Sui and T'ang Dynasties, vol. I, Beijing, 48 [in Chinese].

J. Tubach, 1999/1994-1995 : "Die nestorianische Kirche in China", Nubica et Aethiopica.

International Jahrbuch für koptische, meroitischnubische, äthiopische und verwandte Studien, IV/V, 61-193.

S. Uhlig, 2001 : "Eine trilinguale 'Ezana-Inshrift", Aethiopica, Hamburg, n°4, 7-31.

R. Voigt, 2001 : "Die abessiniche Schrift (Teil 1)", Selam Eritrea, 15/16, Hamburg - Mainz, 25-29.

R. Voigt, 2002 (forthcoming) : "Abyssinia", Encyclopaedia Aethiopica, S. Uhlig (ed.), vol. 1, Wiesbaden.

E.O. Winstedt (ed.), 1909, The Chistian Topography of Indicopleustes, Cambridge.

Zhang Jun-Yan, 1983 : "Relation between China and the Arabs in Early Times", The Journal of Oman Studies, 1983, 91-109.

Zhang Xinglang (Chang Hsing-lang), 1930 : Zhong xi jiaotong shiliao huipian ('Collected Historical Materials on Sino-Western Contact), vol. 3, Gudai Zhongguo yu Feizhou zhi jiaotong ('Early contacts between China and Africa'), Beijing.

Haut de page


1 The first Ethiopian scholar to give an overview on contacts with China was Sergew Hable Sellassie, 1972. See also the work of the sinologist and geographer Herrmann, 1913, who first identified the far Western coastal kingdom of Huang-tchi (= "Ag'azi ?") with the Adulis coast (1st century B.C./1st century A.D.). This kingdom must have laid West from India and exported seemingly African products.

The idea of very ancient contacts with China also attracted the imagination of Ethiopian novel writers. See J. A. Mar, 1997 : this Ethiopian writer of partly German descent, is said to have written this novel, Makéda ou la fabuleuse histoire de la Reine de Saba, in the 1920s. He fabulates for example, that "a yellow prince from the country of miracles" came over the ocean and asked the queen of Saba for marriage, but was rejected - in spite of his miraculous wealth ... And her "green fleet" reached India, Malaysia, China, buying silk in great quantities.

2 On a 3rd century hoard of Indian coins in Däbrä Damo in the Aksumite kingdom, see A. Mordini, 1960.

On Kushana coins in Aksum, see S. I. Berzina, 1984.

I thank Professor Krishna Rao of the Department of Archaeology of Asmara University, Eritrea, for his kind information on this subject during my stay in Asmara in 2000.

3 Contacts between Yemenis and Chinese in India are reported already in the 5th century. In his travel report, Fa Xian observed in the year 414 the presence of a large number of Yemeni merchants in Ceylon. Cf. J. Legge (tr.), 1886, 104.
4 Cf. T. O. Höllmann, 1992, 411.
5 Silk started to be traded in Ptolemaic Egypt ca. 30 B.C., when Egyptian merchants entered into the trade with India (A. Herrmann, 1913, 55). Later, silk played an important role at the court of the Ethiopian kings.
6 It mentions the Chinese metropolis "Sinai" or "Thinai", which is a Sanskrit form of Ch'in, the dynasty which had unified China (L. Casson, 1989, 238). From here the silk ("Chinese cotton") is exported.
7 Already E. Haberland, 1965, recognised that the traditions of a Sacral Kingdom for instance were not influenced by Hamitic or Semitic traditions, but were of very ancient inner-African descent.
8 The origins of the art of silver jewellery in Northeast Africa for example might be related to India, even traditional music might have been influenced by Indian musicians. I thank Monsieur Michel Postel, directeur-fondateur of the Musée Asiatica at Biarritz (France), for his enlightening comments on that question.
9 See Y. M. Kobischanow, 1969, 21 ; T. Natsoulas, 2002.
10 See R. Voigt, 2001, 25-29.
11 P. Snow, 1988, 3 ; E. O. Winstedt, 1909, 322, where the regions mentioned appear as "Persis, Homiritia and Adulia".
12 P. Snow, 1988, 3.
13 Cf. A. Herrmann, 1913. Ag'azi is the name of a people dominating the area of today's Eritrea / Northern Ethiopia in the pre-Aksumite period, known from a few inscriptions. The pre-Aksumite kings - at least some of them - descended from the Ag'azi. Later, in the 6th century, also the Ethiopian ex-vassal king of South Arabia, Abreha, seems to have claimed descent from them. Cf. RIE, 1991, 8 (1-2), 5A (1-2), 10 (1-5) ; A. J. Drewes, 1956, 183.
14 The Periplus Maris Erythraei (op. cit., chapters 6 & 17) doesn't mention rhinoceros horn exported from India, even if this would also be possible, but in chapter 4 mentions it as an export product of Adulis.
15 A. Herrmann, 1913, 553 sq.

These are also major export products of Adulis according to the Periplus Maris Erythraei (op. cit., chapter 4).

16 Shen Fuwei, 1980, 50. The Chinese denomination as a "country" reminds of Cosmas who several centuries later does not only refer to Adulis as a port, but once also lets it appear as a country, as quoted above in footnote 14. This might correspond to some degree of local autonomy. It should in fact be expected, that the region, split up in very different climatic zones, before the rise of the Aksumites consisted of several independent areas. Epiphanius of Constantia lists people living at the African shore of the Red Sea in the 3rd century, among them the "Azomitorum cum (A)dulitibus", which according to L. Casson, 1989, 110 could in fact mean that the coastal area was in that time ruled quite autonomously from the Aksumites.
17 P. Snow, 1988, 8 ; G. Mathew, 1956, 50-55.
18 "Suleiman, a ninth-century Arab trader, refers to the presence of massive Chinese ships at Kulam-Malay in south-west India and at Siraf and Oman in the Gulf" (P. Snow, 1988, 8).
19 According to some recent authors Chinese sources confirm that Chinese junks sailed not only to the Persian Gulf, but also the Red Sea at least from the 5th century A.D. during the rule of the Liu-Sung dynasty already, cf. Jun-Yan Zhang, 1983 ; V. Christides, 1994, 34.
20 In a similar approach Stuart Munro-Hay is about to publish a text on a Chinese text on Arabic history, written in the 18th century on the basis of ancient Persian sources. Interestingly this text contains information on 5th century Aksum, which cannot be found in any other source. It seems that it is based on lost Arabic sources.
21 An early Arabic term shows that ships were known, probably already in those times, which were able to reach even China, markab sînî ("ship sailing to China"), cf. L. M. Devic, 1883-1886, 85.
22 As the Persian Mani put it in the 3rd century. According to him there were four "great kingdoms" in the world: First Persia, second Rome, third Ethiopia and finally Silis (China), cf. H. Ibscher, 1940, 188-190.
23 In Jingzhao, today the city of Xi'an in the province Shaanxi, cf. Chinese Great Encyclopaedia, 1995.
24 Ibid.
25 Jerusalem, as it has also been called in the ancient text Weilue, cf. Shen Fuwei, 1980, 47. Today, Jerusalem is transliterated as "Yelusaleng".
26 An ancient unit of measure, 2,000 Li corresponding to approximately 1,300 km.
27 In another version : "mountains and lakes one after the other".
28 Christianism.
29 In the T'ang period, the East Roman Empire was called Fulin. It is a derivation of the word "Roma", transported via the Soghdian "From (Frum)", derived from the Persian "Hrom" (cf. J. Tubach, 1999, 65). Du Huan in his Jingxingji notes : The empire of Fulin lies in the west of the country Shan. It is seperated from it by mountains and several thousand Li, and is also called Daqin” (cf. Shen Fuwei, 1980, 47). "Shan" is "Syria" (from Arab "Sham"), which had been lost by the Byzantine Empire to the Muslim armies not long before. The toponym Fulin is used for the whole East Roman territories, which had included Syria and Jerusalem until the expansion of Islam. Du Huan has started his journey from its Southern borders.
30 Byzantium.
31 Here different characters are used than in the work of Du You, but in both cases it is pronounced "Molin". Generally one can observe that Chinese prefer to transliterate foreign toponyms using characters which would not constitute a meaning in Chinese. Thus, "Molin" doesn't have any meaning in Chinese.
32 P. Snow, 1988, 3 sq. and Shen Fuwei, 1980, 48.
33 P. Snow, 1988, 4 even suggests that the Persians might have "imparted their Zoroastrian creed to some of the coastal people". However, we don't have any other hints to Zoroastrian beliefs at the coast, even if it cannot be rejected as completely impossible : some Afar groups at the Eritrean coast claim Persian origin (D. Morin, 1998, 53 n° 3 and 55), and imported Persian objects were found by archaeologists (A. Manzo, 1998, 42). After the 6th century, when the Aksumites had controlled the Red Sea and Southern Arabia, the Persian empire advanced into the area, took over Southern Arabia and even seem to have landed at the African coast before the expansion of Islam.
34 Du Huan, quoted by P. Snow, 1988, 4.
35 Chinese Great Encyclopaedia, 1995.
36 Chinese Great Encyclopaedia, 1995. It also ignores that the country of Fulin, which is as well described by Du Huan, is well identified (See footnote 30).
37 Shen Fuwei, 1980.
38 F. Hirth, 1885, 204-205.
39 Ibid., 46.
40 This hypothesis was established for many years, after Laufer had suggested it (B. Laufer, 1919, 339).
41 The Chinese scholar Shen simply notes : "This Christian kingdom was called "Molin" due to its former belief in sea gods." (Shen Fuwei, 1980) The problem, that Mahrem was certainly not venerated any more in that time, was solved with the following ­ erronous ­ argument : "The cult of the sea of the Ethiopian peoples was the reason why the country became a seafaring state. It is also due to this that in the 8th century, when Du Huan visited these areas personally, Ethiopia was still called Molin."
42 H. Brakmann, 1994, 37 sq.
43 In this inscription, King Ezana (' Z N) calls himself a son of M H R M, whom he identifies with the Greek god Ares. Later in the text, when he says that he has subdued the B G (the Beja in today's Sudan and Eritrea), he speaks of sacrifices to the three gods' S T R (traditionally connected with the heaven), B H R (connected with the land) and M H R M, see S. Uhlig, 2001, p. 26.
44 The Chinese scholar Tuam Ch'eng-Shih (in pinyin, Duan Chengshi, died 863 A.D.) is the first to use this name for the Somali coast in his work Yuyang za zu ("Assorted Dishes from Yuyang"), E. Cerulli, G.S.P. Freeman-Grenville, 715.
45 Later this country's name changed slightly into "Bi-ba-luo". It is known as such in the acts of the Kanton trade commissioner Chao Ju-Kuan (in pinyin : Zhao Rugua), Zhufanzhi, "Gazetteer of Foreigners", 1225 A.D., who seems to rely on a work of Chu Ch'ü-fei dated 1178. Chao describes their religion in words very similar to Du Huan's : "they serve heaven and do not serve the Buddha", a conception also used by Du Huan to describe the monotheistic character of religion. They compete "with each other for supremacy" ; even it is reported that before marriage a husband-to-be had to bring a severed man's sex (P. Snow, 1988, 14) - a custom still documented for the 20th century among the closely related Afar. The people produce camels and sheep, dragon's saliva, elephant and rhinoceros ivory, much putchuk, liquid storax gum, myrrh and tortoise shell. (Cf. E. Cerulli, G.S.P. Freeman-Grenville, 715 : the country's name is transliterated there as "Po-pa-li" respectively "Pi-pa-lo", in the pinyin transliteration it is "Bo-ba-li" etc.). For more details, see P. Snow, 1988, 11-13, 14 and 16.
46 Quoted after P. Snow, 1988, 13 and J. J. L. Duyvendak, 1949, 13.
47 S. Munro-Hay, 2002.
48 H. S. Sergew, 1972, 196 sq. ; A. Manzo, 1998, 40 and 45.
49 Shen Fuzei, 1980, 52. More on this hypothesis see underneith.
50 A. K. Irvine, 1965, 178-196.
51 Another example has been identified recently by W. Hahn, 1999, 7. He describes an ancient Aksumite coin of a king Noe around ca. 400 A.D. in which reference is made to the "Abyssinian land" (X[OR]AS ABACCIN[ON]).
52 The exact connection of this territorial term with the South-Arabian 5th century B.C. ethnic group HBST is still discussed among scholars. A. K. Irvine, 1965, 181 ; R. Voigt, 2002.
53 See A. K. Irvine, 1965, 182 ; cf. also the recent publication S. Uhlig, 2001, 16, on a trilingual mid-4th inscription, which also mentions the HBSHT among the peoples governed by Ezana, the king of the Aksumites (' K S M).
54 R. Voigt, 2002; cp. also the "Abasenoi" in ancient Greek sources, F. Hommel, 1926, 520.
55 There is another, later Chinese term, which had also been identified once with al-Habasha, Ha-pu-ni, a place described by the travellers of the Cheng Ho fleet in the 15th century (cf. J. J. L. Duyvendak, 1949). However, on a Chinese map it is close to Socotra and Aden and therefore should rather be identified with Ras Hafun on the Horn of Africa (cf. E. Cerulli - G. S. P. Freeman-Grenville, 1960, 715).
56 Cp. the sudden stop of coin minting in the mid-7th century, H. Brakmann, 1994, 108.
57 Shen Fuwei, 1980, 46.
58 I have to thank Xi Yiyang, Hamburg, for this information.
59 Shen Fuwei, 1980, 51.
60 Compare also : "The inclusion of a Chinese in the party may have been intended to awe them by exhibiting the vast range of peoples who were subject to Arab rule" (P. Snow, 1988, 4).
61 J. Tubach, 1999.
62 The idea that the coast of Molin described by Du Huan corresponded to the Eritrean coast had been accepted after it had been suggested by Shen Fuwei and Philipp Snow. Cf. Shen Fuwei, 1980 ; P. Snow, 1988, 4 ; F. Böckelmann, 1998, 264 ; S. Munro-Hay, 2002.
63 The Han dynasty sources on Huang-tchi, etc., might also refer to Africa, as we have seen, but Du Huan is the first to clearly refer to an encounter with a black people.
Haut de page

Table des illustrations

Fichier image/jpeg, 4,3k
Haut de page

Pour citer cet article

Référence électronique

Wolbert G.C. Smidt, « A Chinese in the Nubian and Abyssinian Kingdoms (8th Century) », Chroniques yéménites [En ligne], 9 | 2001, mis en ligne le 27 septembre 2002, consulté le 30 novembre 2015. URL :

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page
  • Logo Centre français d’Archéologie et de Sciences Sociales
  • Logo Institut de recherches et d'études sur le monde arabe et musulman
  • Logo Ministère des Affaires étrangères et européennes
  • Logo CNRS - Institut des sciences humaines et sociales