At the Edge of the World of Islam

The article focuses on Ibn Baṭṭūṭa’s account of his journey through the Malay Archipelago in the 1340s, which is remarkable in so far as it captures the region at the early stages of Islamisation, when the first Muslim sultanates emerged in North Sumatra. It describes the Arab traveller’s encounters with both Muslim and Hindu-Buddhist inhabitants of the region, drawing the image of a frontier of the world of Islam. The article discusses the narrative techniques used in the travelogue and Ibn Baṭṭūṭa’s perceptions of the cultural border between Muslim and non-Muslim parts of the archipelago. As the author tries to demonstrate, this barrier is represented in the text as the one between the cosmopolitan but familiar Islamic world and the exoticised space beyond, and it appears to be more imagined than actually experienced, since the account seems to be largely shaped by the general narrative paradigm of the travelogue.


Introduction
Now a home for the world's largest Muslim population, the Malay Archipelago is likely to [1] have first come in contact with Islam as early as the seventh century. But the incorporation of the region into the Dār al-Islām started in the later medieval period when a number of Malay states converted to the new faith. This is when the Straits of Malacca was visited by the famous Moroccan traveller Ibn Baṭṭūṭa , who left what appears to be one of the few first-hand medieval Arabic accounts of the region. The traveller's description of his experiences in the archipelago can be found in the last chapters of his Riḥla. 1 Occupying only some ten pages, this description seems to be the longest piece of medieval Arabic text dealing with maritime Southeast Asia, and the only arguably original account after 1000. Furthermore, Ibn Baṭṭūṭa is the first Arab known to report on Islam in the Malay Archipelago and to witness its advent.
The account of Ibn Baṭṭūṭa is remarkable in so far as it captures the region at the early [2] stages of Islamisation when the first Muslim sultanates emerged in North Sumatra. It describes the traveller's encounters with both Muslim and Hindu-Buddhist inhabitants of the Malay world, drawing images of a borderland of the world of Islam. No less remarkable is that this borderland is observed and depicted by a person coming from the Islamic heartlands who was a native of the opposite frontier of the Dār al-Islām, i.e. the Maghreb. This is also where-at the court of the Marinid sultan Abū 'Inān (1348-1358)-the text of Ibn Baṭṭūṭa's travelogue was created by Ibn Juzayy al-Kalbī (1321-1357), a man of letters who had never been in the East. Either written down or constructed, the descriptions of the Malay Archipelago in the Riḥla tell us as much of the observer as they do of the observed, and of his audiences' worldview and expectations.
Southeast Asian chapters of the Riḥla contain certain inconsistencies in terms of itinerary [3] and chronology, which might make one question whether Ibn Baṭṭūṭa had ever been to Southeast Asia. As more and more borrowings from other sources are being revealed in other parts of the Riḥla, Ibn Baṭṭūṭa's credibility as an eyewitness remains under dispute. His brief trip to China, of which the voyage through the Straits of Malacca is a part, has been addressed by a number of scholars, their opinions ranging from distrusting particular accounts to disclaiming all the traveller's wanderings to the east of India (on the discussion see Yule 1916, 4:45-51;Yamamoto 1936, 102-3;Gibb 2008, 12-13;Elger 2010). Nevertheless, it appears to be generally assumed that Ibn Baṭṭūṭa reached at least the southern coast of China (Dunn 1986, 3, 252-53); hence, his accounts of the Malay Archipelago might be reliable. Gerald Tibbetts does not seem to question Ibn Baṭṭūṭa's veracity, as he attributes discrepancies in the traveller's accounts of Southeast Asia to displacement of passages by a later copyist or loss of pieces of the original text (Tibbetts 1979, 13).
However, in this essay I would like to focus not on what Ibn Baṭṭūṭa knew about the Malay [4] Archipelago, but on how he interpreted his knowledge or experiences and built them into his general picture of the world. Or, to be more precise, on how he and his co-author, Ibn Juzayy, shaped the narrative and presented their ideas of the region to the readers. Rather than attempting to spot the points of Ibn Baṭṭūṭa's itinerary on the map, it might be more rewarding to define invisible social and cultural boundaries he imposed on it. I will look into Ibn Baṭṭūṭa's perceptions of the cultural border between Muslim and non-Muslim parts of the archipelago, as far as those can be recognised by a modern reader. As I will try to demonstrate, this barrier is represented in the text as the one between the cosmopolitan but familiar Islamic world and an exoticised space beyond, and it appears to be more imagined than actually experienced.
Capturing the historical moment when the eastern edge of the world of Islam crossed [5] Sumatra, the account of Ibn Baṭṭūṭa introduces an interesting conception of the region represented in the opposition between the two countries, which the traveller calls Jāwa and Mul Jāwa. While the former is depicted routinely and realistically as another Muslim state on Ibn Baṭṭūṭa's way, the latter becomes a scene for amazement and cultural misunderstanding. Unique to the Riḥla, Ibn Baṭṭūṭa's conception of the Malay world seems to be a cultural rather than a geographical one. And it appears to have been the feeling of the frontier that informed this conception, along with the oppositions between the Islamic and "infidel" worlds, the Muslim Self and the non-Muslim Other.
The passages under discussion can be found between Ibn Baṭṭūṭa's descriptions of India [6] and China, as he allegedly passed through the Straits of Malacca on his way to China and back. The account of Samudra on the northern coast of Sumatra is followed by that of a country called Mul Jāwa and its port Qāqula, but on the way back to the west only Samudra is revisited. While the sultanate of Samudra is known from other local and foreign sources, the identification of the latter two place names remains problematic. Different locations in Java, Sumatra and the Malay Peninsula have been suggested, but none of them fully meet the case. What is clear is that both of the places belong to the Straits of Malacca area-the westernmost part of the Malay world and a hub of Indian Ocean commerce.

Ibn Baṭṭūṭa and the Indian Ocean World in the Fourteenth Century
The travels of Ibn Baṭṭūṭa took place in between 1325 and 1354, covering the entire medieval [7] Islamic world and beyond. At the time the political landscape was dominated by Mongol states: Yuan China, Ilkhanate and the Golden Horde; also Mamluk and Delhi Sultanates were on the rise. Still threatened from both east and west, the Middle East was recovering after the Crusades and Mongol invasions. With the fall of Baghdad in 1258 and the expulsion of crusaders from the ports of Levant at the end of the thirteenth century, the trade routes through Mesopotamia and the Persian Gulf went into decline. Communication between the East Mediterranean and the Indian Ocean concentrated in the Red Sea and Egypt, and the Mamluks held the key to this route. The Persian Gulf ports continued to operate, but never regained the role they had played in the Abbasid Era.
In the Maghreb, where Ibn Baṭṭūṭa was born, the fourteenth century was a heyday of [8] Marinids who reached the peak of their power by 1350. In 1325, Ibn Baṭṭūṭa left his home in Tangier in order to perform hajj and did not return until 1349. After a series of travels around North Africa and the Middle East, he headed north to the Golden Horde, and from there to Central Asia and India. Having reached Delhi in 1333, Ibn Baṭṭūṭa stayed at the court of Sultan Muhammad ibn Tughlaq (1325-1351) for around eight years, and spent some time in the Maldives after that. In 1345 he embarked on a journey from Bengal to China, passing through the Straits of Malacca on his way there and back. After his second visit to Sumatra, Ibn Baṭṭūṭa sailed directly to the Middle East and soon returned to Morocco, where he settled down in Fez at the court of Abū 'Inān and, in 1356, eventually put his Riḥla into writing.
Ibn Baṭṭūṭa's ambitious tour of the medieval world and his voyage to the East in particular [9] would not have been possible without a developed network of trade routes, on the one hand, and a social context that facilitated his mobility, on the other. The traveller joined caravans of merchants and pilgrims, royal corteges and diplomatic missions. Throughout the years of wanders he played many different roles, most of them within the repertoire of the class of 'ulamā', Islamic scholars. Ibn Baṭṭūṭa travelled as a pilgrim, student, scholar and a wandering ṣūfī, a jurist qualified enough to be a qāḍī (Sharia judge) and an adventurous career seeker. Occasionally a courtier, diplomat or administrator, or simply as an honourable learned guest, he benefited much from the hospitality and patronage of different rulers-especially those at the periphery of the Islamic world.
Ross Dunn distinguishes between the traditional roles of a pilgrim, scholar and ṣūfī played [10] by Ibn Baṭṭūṭa in the Middle East and a new type of a "literate frontiersman" that was in high demand in the growing Muslim communities and at recently converted courts outside the Islamic heartland (Dunn 1993, 63). The expansion of Islam in the late medieval period created a cosmopolitan social order that enabled and encouraged migration of religious and intellectual specialists from the centre of the Islamic world to its periphery. The borderlands of Islam attracted them with great opportunities, as is demonstrated by Ibn Baṭṭūṭa's career in Delhi and those of other foreigners he encountered at the court, most of whom would have hardly achieved similar positions back in the Middle East. As a representative of this class of migratory 'ulamā', Ibn Baṭṭūṭa was a common social type, not remarkable but for the extent of his wanderings.
In the age of Ibn Baṭṭūṭa, the Indian Ocean shores were the frontiers of the world of Islam, [11] but they were also a special world themselves-the world that brought East Africa, Arabia, South Asia and the Malay Archipelago together through intellectual networks and maritime trade. Ibn Baṭṭūṭa's travels around the Indian Ocean and his accounts of the people he met highlight the mobility of Muslims around the region and the existence of a vast network of Islamic scholars (Pearson 2003, 75-76). Also, a certain cultural unity of the Indian Ocean world is visible through Ibn Baṭṭūṭa's descriptions of social practices and etiquette. As the traveller seems to have found many of them familiar wherever he went, it suggests the development of a uniform cosmopolitan order that would not have been possible without the continuous circulation of people around the region.
Another aspect related to this uniformity was the language. It is commonly assumed that [12] besides his native tongue, Ibn Baṭṭūṭa had only some limited knowledge of Persian-the court language of the Delhi Sultanate and a lingua franca of the Indian Ocean trade. However, he apparently could do without it, since the role of Arabic as a bearer of Islamic texts and ideas to East Africa as well as South and Southeast Asia enabled him to move around the region without difficulty. At the linguistically diverse frontiers of the Islamic world, religious conversion went hand in hand with the spread of Arabic as the language of faith and law, and later resulted in the formation of what Ronit Ricci defines as the "Arabic Cosmopolis" of South and Southeast Asia. 2 However, at the time when Ibn Baṭṭūṭa visited the Malay Archipelago, the region was only beginning to be absorbed by the expanding world of Islam.
Although Muslims appear to have been settling in Malay ports since the first centuries of [13] Islam, the circumstances for a large-scale conversion of local populations evolved only by the end of the thirteenth century. By that time the remnants of Srivijaya in East Sumatra almost lost their control of trade routes and vassal port-polities, giving way to the Javanese empire of Majapahit. The city-states of North Sumatra benefited from Srivijaya's decline, as well as from the increase in international commerce in the archipelago. In the thirteenth century the trade with the Middle East started to revive due to the insecurity of land routes and the establishment of Mamluk power on the western side of the Indian Ocean. Located at the entrance to the Straits of Malacca, North Sumatran ports flourished as transit points for spice and textile trade. The decay of Srivijaya enhanced their autonomy, and conversion to Islam became another step in this direction. However, the economic rise and political independence of Sumatran ports did not last long, as in 1350-soon after Ibn Baṭṭūṭa's visit-they faced the invasion of Majapahit.
The Malay sultanate of Samudra (also known as Pasai), where Ibn Baṭṭūṭa stopped on his [14] 2 Speaking of the period after the sixteenth century, Ricci describes the Arabic Cosmopolis as a "translocal Islamic sphere constituted and defined by language, literature and religion" (Ricci 2011, 4). way to China, converted to Islam around the 1290s and became one of the earliest Muslim states in Southeast Asia. In the fourteenth century it dominated the northern coast of Sumatra as a hub for the export of pepper, camphor and benzoin from the hinterland and a market for exchange of goods coming from east and west. Samudra is known from local chronicles and epigraphy as well as foreign written evidence, and is commonly identified with the kingdom of Samara visited by Marco Polo in 1292. As Polo reports it to have been pagan (Polo 1875, 2:274), the conversion is assumed to have taken place between 1292 and 1297, the latter being the date of death of Samudra's first Muslim ruler, Malik as-Saleh. His successor, Malik az-Zahir (d. 1326), expanded the boundaries of the sultanate with military campaigns and was known for his religious zeal. The court of this sultan and his successors was a home to theological discourses and Arabic and Persian literature translated into Malay. The sultans of Samudra, similar to other rulers of recently converted Islamic societies, [15] favoured foreigners of Middle Eastern origin and sought legitimacy and prestige by surrounding themselves with Islamic scholars and jurists, especially those claiming descent from the Prophet. As the Malay sultanates accepted the Shafi'i school of Islamic law, 'ulamā' belonging to it were particularly welcome. Starting from the thirteenth century, the flow of Arab migrants to Southeast Asia increased (Wade 2009, 234), both in light of political instability in the Middle East and the rising demand for literate Muslims in the Indian Ocean world. While in the previous centuries the majority of Arab visitors to the archipelago were traders or sailors, in the late medieval period visitors of a new kind appeared in the region. Ibn Baṭṭūṭa was one of them and the first who left a written account of his voyage.

Samudra and its Sultan
Ibn Baṭṭūṭa appears to be the first Arab writer to mention the sultanate of Samudra and [16] the first foreigner to report the conversion of Malay states to Islam. Most of what is known of the history of Samudra in 1300 to 1350 is derived from the early Malay chronicles of Hikayat raja-raja Pasai (from the second half of the fourteenth century) and Sejarah Melayu (fifteenth century). 3 Ibn Baṭṭūṭa's account presents interesting evidence complementing these texts, along with Chinese and European reports, archaeological finds and epigraphy. It contains, among other things, various ethnographic details related to court etiquette, wedding rituals, food, clothes and other cultural practices. These details appear to correspond to what is known from Malay chronicles and later ethnographic sources, and thus seem to confirm Ibn Baṭṭūṭa's credibility. Yet a problem arises regarding the sultan Ibn Baṭṭūṭa claims to have met, Malik az-Zahir, who is supposed to have died some twenty years before the traveller's visit. Different explanations have been offered for this chronological contradiction, but the issue remains unclear.
The narrative related to the sultanate of Samudra (called Sumuṭra in Ibn Baṭṭūṭa's text) [17] starts with the traveller's arrival on the island of Jāwa, which apparently indicates Sumatra or a part of it, and proceeds to a coherent description of the first four or five days of his stay there. Ibn Baṭṭūṭa lands at the port and is escorted by the sultan's court jurists to the capital, which is located inland. They head to the royal palace where the traveller and his companions are greeted by the sultan's deputy, treated with a meal and presented with sets of clothes. Afterwards, the guests are taken to a special pavilion where they stay for three days before the audience with the sultan occurs. Ibn Baṭṭūṭa mentions meeting a certain amīr Daulasa, whom he happened to have seen before at the court of Muhammad ibn Tughlaq in Delhi. This mention highlights Samudra's diplomatic relations with India, while the names of the court jurists who escorted the traveller to the capital indicate their Middle Eastern origins.
On the fourth day, Ibn Baṭṭūṭa meets the sultan at a mosque after the Friday prayer and [18] accompanies him to the audience hall where he observes the ceremony of greeting the ruler. The traveller's stay in Samudra is described in great detail, only two relatively isolated tales falling out of the narrative-those of the sultan and a rebellion of his nephew. About the rest of the fifteen days spent in Samudra, no details are provided. Afterwards Ibn Baṭṭūṭa asks permission to embark, "for it was the season, since the voyage to China is not organised at any time" (Beckingham 1994, 880)-it was likely the southwest summer monsoon that he was waiting for. On the way back from China the traveller had to stay in Sumatra longer-for two months, probably waiting for a favourable wind again. However, Ibn Baṭṭūṭa's account of his second visit to Samudra is brief and mostly deals with the wedding of the crown prince.
Ibn Baṭṭūṭa describes Samudra as a big city surrounded by wooden walls and towers and [19] located four miles away from the port. Due to the volatile natural conditions of the North Sumatran coast, the archaeological evidence of the period is scarce and the fourteenth-century location of neither the city nor its port has been identified yet. But it was a common practice for Malay riverine city-states to be located upstream for security reasons and separated from their trading ports on the sea coast (Airriess 2003, 92). Sejarah Melayu mentions a port called Jambu Air where those heading to Samudra used to land (Revunenkova 2008, 185, 200); in addition, Ma Huan, a companion of the Chinese admiral Zheng He who visited Sumatra in 1416, reports that "when a ship leaves Malacca for the west […] it first comes to a village on the sea-coast called Ta-lu-man; anchoring here and going south-east for about ten li (3 miles) one arrives at the said place. This country has no walled city. There is a large brook running out into the sea, with two tides every day" (Groeneveldt 1876, 85).
Notably, Ibn Baṭṭūṭa mentions wooden walls and later says they were built by the current [20] sultan after the revolt of his nephew, but these were probably destroyed by the time of Zheng He's voyage.
Ibn Baṭṭūṭa's report on Sultan Malik az-Zahir (al-Malik aẓ-Ẓāhir in the Riḥla, Maliku'l-Tahir [21] in Hikayat raja-raja Pasai) corresponds to what is known of Samudra's wars with the neighbouring Hindu-Buddhist rulers and theological disputes, which occurred at its court: He is Sultan Al-Malik Al-Ẓahir, one of the noblest and most generous of kings, [22] a Shāfī'ī in madhhab, and a lover of jurists, who come to his audiences for the recitation of the Qur'ān and for discussions. He often fights against and raids the infidels. He is unassuming and walks to the Friday prayer on foot. The people of his country are Shāfī'īs who are eager to fight infidels and readily go on campaign with them. They dominate the neighbouring infidels who pay jizya to have peace. (Beckingham 1994, 876-77) As was already mentioned, Sultan Muhammad Malik az-Zahir is known to have died in [23] 1326-that is, almost 20 years before Ibn Baṭṭūṭa's visit. The tombstone of this ruler, along with that of his predecessor Malik as-Saleh, can be seen in the vicinity of the modern city of Lhokseumawe (Tinggalan Sejarah 2014, 7). As some scholars point out, the name Malik az-Zahir was a hereditary honourific title borne by several rulers of Samudra, including Muhammad's son Ahmad (1326-c.1360), and it was most probably this sultan whom Ibn Baṭṭūṭa met (Hill 1963, 15). An interesting explanation for this confusion is offered by Naquib al-Attas, who suggests that Ahmad, having ascended the throne at an early age, was not officially recognised as a sultan and ruled under the name of Malik az-Zahir for a certain period of time after the death of his father and before his own marriage. Al-Attas refers to the Malay practice according to which a ruler's son could not inherit the throne unless he fathered a son, and argues that the wedding, which Ibn Baṭṭūṭa witnessed, was actually that of Ahmad (Attas 2011, 31-32). Nevertheless, a conflict in this speculation cannot go unnoticed: in Ibn Baṭṭūṭa's account Sultan Malik az-Zahir is present at his son's wedding and the two act as separate characters. If Sultan Ahmad was the bridegroom, who was the father?
The genealogy of Samudra rulers appears to be confusing even disregarding Ibn Baṭṭūṭa's [24] report, as the local chronicles, epigraphy and oral tradition offer conflicting versions of it. The names of Muhammad Malik az-Zahir and Ahmad, the father and the son, belong to the version of Sejarah Melayu, which is confirmed by an epitaph on a fifteenth-century tombstone. Meanwhile, according to Hikayat raja-raja Pasai, Ahmad was Malik az-Zahir's grandson, and according to the oral tradition, written down in 1931, his nephew (Cowan 1973, 256-59). While the name of Malik az-Zahir is present with minor spelling variations in all the mentioned sources (the sultan's personal name, Muhammad, is found in the epigraphy and oral tradition only), Ahmad is also called Ahmad Permadala Permala (in Hikayat raja-raja Pasai) and Raja Bakoy (oral tradition) (ibid.). Since Ahmad is known to have ascended the throne at a very young age, he must still have been young in 1345, but Ibn Baṭṭūṭa's description creates an impression of a mature ruler, the father of an adult son. As the beginning of Ahmad's rule is derived from the date of the death of Malik az-Zahir, it is not impossible that someone else, perhaps Ahmad's father and Malik az-Zahir's son or brother, might have ruled the sultanate in the 1340s when Ibn Baṭṭūṭa visited it.
Alternatively, the traveller might have never been in Sumatra, but someone else might have [25] met Muhammad Malik az-Zahir before 1326. I would speculate that Ibn Baṭṭūṭa could have met such a person at the Delhi court soon after his arrival there in 1333 (could it be the mentioned amīr Daulasa?), or, as suggested by Michael Laffan, the traveller's account might have been "drawn from descriptions of proud Sumatrans in Calicut" or any other South Indian port (Laffan 2009, 39). An encounter with a Malay community abroad could arguably have been enough for Ibn Baṭṭūṭa to get some idea of Malay customs, and pick up information on Samudra and its sultan.

The stone walled Qāqula
As opposed to Samudra, which does not appear in Arabic texts before Ibn Baṭṭūṭa, Qāqula 4 [26] is a place name deeply rooted in the Arab geographical tradition. References to this port can be found in Arabic accounts starting from the ninth century, but its location remains under dispute. Different identifications have been suggested, most of them placing Qāqula on the north-eastern coast of the Bay of Bengal. Nevertheless, this point of view disagrees with Ibn Baṭṭūṭa's itinerary, which seems to place Qāqula somewhere down the Straits of Malacca. The only explanation offered so far for this confusion is the displacement of paragraphs and corruption of the original text by the copyists.
Compared to Ibn Baṭṭūṭa's description of Samudra, passages dealing with Qāqula seem to be [27] much more inconsistent. They are not always properly related to each other and lack detailed descriptions of the traveller's personal experiences. His accounts of spices contain a number of botanical confusions, while those of local customs do not correspond to what is known of the Malay culture, finding more parallels in medieval Arabic and Western texts dealing with India. Also, the relation between Qāqula and another place name-Mul Jāwa-remains unclear from the narrative. As it appears from the text, after leaving Samudra, Ibn Baṭṭūṭa arrives in the country of Mul Jāwa, of which Qāqula is said to be a part. The description of the country is interrupted by a detailed excursus into spices, which is followed by two accounts that appear to be somewhat disconnected from each other-those of Ibn Baṭṭūṭa's arrival at the port of Qāqula and of his visit to the court of the sultan of Mul Jāwa. As there is little to suggest that the sultan of Mul Jāwa could reside in Qāqula, the problem arises of whether it is one or two different places described.
References to Qāqula can be found in classical and post-classical Arabic texts starting from [28] the ninth century. There is no agreement whether this place name refers to a port, an island or a country, and sometimes it is also applied to a sea or mountains. Qāqula is not only mentioned by Ibn Baṭṭūṭa but also by al-Ya'qūbī, Buzurg ibn Šahriyār, Abū Dulaf, al-Idrīsī, al-Qazwīnī and Ibn Sa'īd al-Maġribī, and it is often identified with Takola of classical writers and Ko-ku-lo appearing in Chinese texts. No local sources survived mentioning a similar place name, and scholars have to rely solely on foreign evidence. Arab geographers describe a citystate located on the seashore at a river estuary and refer to aloeswood as its major export commodity. Ibn Baṭṭūṭa's account corresponds to their descriptions of a flourishing port, and his mention of piracy and taxes points to Qāqula's location on a busy trade route. The traveller also mentions a wide stone wall, abundance of aloeswood and large number of elephants: We arrived at the port of Qāqula and found there an assemblage of junks prepared [29] for piracy and to fight any junk which might oppose them, for a tax is imposed on each junk. We disembarked and went to the city of Qāqula. It is a fine city with a wall of cut stone wide enough to take three elephants. The first thing I noticed outside the city was elephants with loads of Indian aloes wood which they burn in their houses; it is the price of firewood among us, or even cheaper. That, however, is when they sell it to each other. When they sell to (foreign) merchants a load costs a robe of cotton, cotton being more expensive than silk among them. (Beckingham 1994, 882) Ibn Baṭṭūṭa's voyage from Samudra to Qāqula is said to have taken twenty-one days, al- [30] though this distance might have referred to that between Samudra and Mul Jāwa. The traveller places Qāqula on the way from North Sumatra to China, which seems to indicate some point in the Straits of Malacca. However, the evidence from other sources is at odds with this. The location of Qāqula as it appears in other Arabic texts remains unresolved, but most of the locations suggested do not correspond to the itinerary of Ibn Baṭṭūṭa. Scholars attempted to identify Qāqula with different places in Southeast Asia, including those in Java, Sumatra and the coastal areas around Kelantan or Ligor. Gabriel Ferrand distinguishes two places of the same name-one on the eastern shore of Indochina and the other on the western coast of the Malay Peninsula (Ferrand 1913(Ferrand -1914. Most later scholars agree in favour of the eastern shore of the Bay of Bengal, but a more precise location is still uncertain. Tibbetts and Wheatley tend to look for Qāqula at the coast of Tenasserim-despite Ibn [31] Baṭṭūṭa's indication of it being southwards from Samudra (Tibbetts 1979, 97-98;Wheatley 1961, 228). As Tibbetts suggests, the accounts of Samudra and Qāqula could have been mis-placed in the text by a later editor or copyist, while in the original itinerary Ibn Baṭṭūṭa's visit to Qāqula preceded his stop in Sumatra (Tibbetts 1979, 97-98). Notably, all the speculations placing Qāqula so much to the north appear to be based on two accounts, one by Abū Dulaf and another by a Chinese source of the eighth century, which locate Qāqula (Ko-ku-lo) on the seashore to the northwest of Kalāh (Ko-lo). Abū Dulaf, in his description of the voyage from China, puts Qāqula to the west of Kalāh, and the lands of Pepper and Camphor between the two (Tibbetts 1979, 39-41). Chinese compiler Kia Tan provides similar directions: "On the northern coast (of the strait) is the country of Ko-lo and to the west of Ko-lo is the country of Ko-ku-lo" (Hirth and Rockhill 1911, 11). The position of Qāqula, therefore, appears to be directly correlated with that of Kalāh, another highly disputable port on the Malay Peninsula coast.
Most of the scholars locate Kalāh in the north of the peninsula or even in Burma, Kedah [32] being the southernmost suggested place; hence the position of Qāqula shifts further north. However, S. Fatimi's arguments in favour of Kalāh's identification with Klang, a port in the vicinity of modern Kuala Lumpur, should not be dismissed (Fatimi 1960). In light of his speculation, Qāqula could be sought more to the south and appear on Ibn Baṭṭūṭa's way through the Straits of Malacca. Brian Colless accepts Fatimi's identification of Kalāh, and Tibbetts leaves this possibility open, concluding that it might have been anywhere on the coast between Phuket and Klang. But both scholars still locate Qāqula to the north of Samudrain Tenasserim or Takuapa (Colless 1969, 31, 34-37, 41;Tibbetts 1979, 128). It is fair to say, though, that placing Kalāh in Klang does not seem to solve the problem of Ibn Baṭṭūṭa's itinerary in any case, since the distance between Samudra and any point north of Klang would be too short for a 21-day voyage down the wind.
Except for the conflicting accounts of two travellers, Abū Dulaf and Ibn Baṭṭūṭa, no other [33] Arabic text seems to give proper directions to locate Qāqula. But is it reasonable to identify a port known to the Chinese and Arabs in the eighth and tenth centuries with that visited by Ibn Baṭṭūṭa in 1345? In view of the volatile natural and political conditions in the region-silting of coastlines, shifts of trade routes and frequent relocations of power centres-one would hardly expect a flourishing port-polity to remain at the very same place for more than five hundred years. Considering post-classical Arab geographers' tendency to preserve outdated place names, it seems to be even less plausible. In this regard, Alastair Lamb's remarks on the shifting Indian Ocean entrepots appear to be of relevance to the matter: It is quite possible that the toponym Kalah became, in Arab literature, associated [34] less with one geographical spot than with a general region in which were entrepots of the Kalah type. If the entrepots tended to move up or down the west coast of the Malay Peninsula, so would the location of Kalah. The same argument, of course, could be applied to other names in the Arab geographical texts. (Lamb 1964, 2) Leonard Andaya accepts this point of view and concludes that Kalāh was one of the names [35] used by Arab traders to refer to any one of a number of ports located along the west Isthmian and peninsular coast and serving similar functions (Andaya 2008, 37, 57). Qāqula could possibly be another such name. Toponyms containing the element kuala ("estuary"), assonant to Kalāh and, to a lesser degree, to Qāqula, can be found all around the Malay Peninsula as well as in the central and northern lowlands of Sumatra (Airriess 2003, 86). Ibn Baṭṭūṭa might have visited any of them on the Malay or Sumatran coast of the Straits of Malacca and forgotten the proper name, as he stayed there for only three days. Also, Ibn Juzayy could have inserted a toponym familiar to Arab readers into the text, or even have interpolated the whole story if Ibn Baṭṭūṭa never went to a place called Qāqula.

Jāwa and Mul Jāwa
The Riḥla of Ibn Baṭṭūṭa is the only Arabic text where these two place names appear together.
[36] Both of them belong to the few toponyms that entered Arabic descriptions of Southeast Asia in the post-classical period and cannot be found in the geographies of the ninth and tenth centuries. The country of Jāwa is first mentioned in the thirteenth century by Yakut, who locates it at the border of China and states that it "resembles" India (Tibbetts 1979, 55). In the same century a detailed description of the island of Jāwa is provided by Ibn Sa'īd, who manages to incorporate almost all known Arabic place names related to the Malay Archipelago, old and new, into his peculiar conception of the region. He places Jāwa to the south of the Islands of Maharāja (the classical name for Srivijaya) and names a number of older toponyms-Lāmrī, Fanṣūr, Kalāh and the Mountains of Camphor-along with a new one, Malāyur, as the towns or regions located on the island (Tibbetts 1979, 58-59). Since at least three of them (Lāmrī, Fanṣūr and Malāyur) are associated with Sumatra and none with Java, Jāwa of Ibn Sa'īd is most likely to be Sumatra.
This also seems to be the case with Jāwa of Ibn Baṭṭūṭa, as he applies this name to the [37] island where Samudra is located as well as to the Muslim country ruled by Malik az-Zahir: We left these people and after twenty-five days we reached the island of al-Jāwa, [38] from which Jāwī incense takes its name. We saw it at a distance of half a day's sail. It is green and very well wooded […] These people buy and sell with little pieces of tin or unrefined Chinese gold. Most of aromatics there are in the part belonging to the infidels; they are less common in the part belonging to the Muslims. (Beckingham 1994, 876) Although this passage creates an impression that the island is divided into Muslim and [39] non-Muslim parts (a Muslim north and Buddhist south, or perhaps Islamised coastal areas as opposed to the hinterland), later Ibn Baṭṭūṭa calls Malik az-Zahir the sultan of Jāwa. While Jāwa the island is most probably Sumatra, Jāwa the country seems to refer only to the northern part of it, subject to the Muslim ruler of Samudra. The traveller's strong association of Jāwa with Islam can also be seen elsewhere in the text: when he reports on a Muslim community of Jāwa origin residing in a separate quarter in Barahnakār, 5 and when he finds a Jāwa junk with a Muslim crew in Guangzhou (Defrémery and Sanguinetti 1858, 4:225, 304).
Along with the place name Jāwa, Ibn Baṭṭūṭa uses the ethnonym al-jāwī, which seems to [40] indicate people coming from the western part of the Malay Archipelago in general. As a regional notion iqlīm al-Jāwa is also mentioned by thirteenth-century writer Ibn Mujawir (Laffan 2009, 47), and can be thus considered a third case of Ibn Baṭṭūṭa's usage of the place name, the three of them therefore being Jāwa the country (Samudra and its dependencies), Jāwa the island (Sumatra) and Jāwa the region (the western part of maritime Southeast Asia known to Arabs). Although later Jāwa became associated with Java only, applying this name to Sumatra is not surprising in light of the common confusion between the two islands typical of foreign sources. Classical Arab geographers used to call both Sumatra and Java Zābaj, 6 and European travellers, including Marco Polo, referred to Sumatra as Java Minor.
With the introduction of Jāwa into Arabic geographical texts older place names began to [41] disappear from them. Laffan assumes that Jāwa replaced Jāba of Ibn Ḫurdāḏbeh and argues that it was the Javanese spelling of the word that substituted the Malay version (Laffan 2009, 10-11). Tibbets also points to the absence of the place name Sribuza in Ibn Sa'īd's description of the island (Tibbetts 1979, 115), which might indicate the decline of Srivijaya and its disappearance from the map. During the thirteenth century, the economic role of Sumatran remnants of the empire decreased, the Javanese taking control of the region's spice trade and eventually subjugating Melayu-Jambi in 1286 (Hall 1981, 22, 26). Arab ships started to frequent Java, and their older bases in southeastern Sumatra were now controlled by the Javanese. It might be the case that the appearance of the place name Jāwa in the Arabic sources reflected this establishment of Javanese dominance in the western part of the archipelago.
Unlike Jāwa, which was coming into use among late medieval Arab geographers, Mul Jāwa [42] cannot be found anywhere except in Ibn Baṭṭūṭa's account and that of a fourteenth-century Persian historian, Wassaf, who reports on a Mongol expedition to Java (Ferrand 1913(Ferrand -1914. Ibn Baṭṭūṭa thus appears to be the only Arab writer to mention this place name, and he describes it as follows: Mul Jāwa […] is the country of infidels. It extends for two months' travel. It has [43] aromatics, and good aloes of Qāqula and Qamāra, both places being in the country. In the country of Sultan Al-Ẓāhir, in al-Jāwa, there are only incense, camphor, some cloves and some Indian aloes. The greatest quantity of these occurs in Mul Jāwa. (Beckingham 1994, 880) It is notable that Mul Jāwa is opposed to Jāwa in two major aspects-religion and spice [44] trade. In his mention of Mul Jāwa as a place to which Malik az-Zahir's nephew fled, Ibn Baṭṭūṭa calls it a country of infidels, and later he repeats that the sultan of Mul Jāwa is an infidel. Muslim Jāwa thus seems to be clearly contrasted by "infidel" Mul Jāwa. At the same time, Muslim and non-Muslim lands are also compared in terms of the amount of spices available and the laws regulating ownership of trees, which appears to reflect the interests of the Muslim trading community. Also notably, the opposition of Muslim and non-Muslim lands in relation to the ownership of the trees is repeated in Ibn Baṭṭūṭa's accounts of India (Euben 2006, 76).
Due to the lack of data on Mul Jāwa in other sources, the identification of this place name [45] can be only derived from Ibn Baṭṭūṭa's account alone. The traveller's description of a vast infidel (apparently Hindu or Buddhist) country, extremely rich in spices and located east of Samudra, best corresponds to the Javanese empire of Majapahit. But the relation between the port of Qāqula and the country of Mul Jāwa adds much confusion. Qāqula, along with Qamāra (Cambodia), is said to be a part of Mul Jāwa or belong to it. As both places appear to be in mainland Southeast Asia, Mul Jāwa could be identified with the Malay Peninsula. However, "belonging" to a country can be understood in terms of allegiance, which does not necessary imply adjacent territories. Ibn Baṭṭūṭa's report on his arrival at Mul Jāwa is followed by the description of the port of Qāqula, and the latter by passages on the sultan of Mul Jāwa and 6 Some of the Arabic descriptions of Zābaj correspond to Sumatra, others to Java. Most of scholars agree that the toponym indicated not particular geographical area but the realm of Srivijaya in general (see Tibbetts 1979, 100-116).
his court. But as Tibbetts points out, Ibn Baṭṭūṭa never calls Qāqula the capital of the country, and he never mentions the sultan in the passages on Qāqula (Tibbetts 1979, 132). Therefore there seems to be no need to see the two places as one.  (Tibbetts 1979, 151), as does Colless, who assumes that Ibn Baṭṭūṭa visited the western part of the island (Colless 1975(Colless , 1969. However, west Javanese ports still seem to be somewhat off the shortest way to China, while Ibn Baṭṭūṭa's voyage to the Far East is often criticised for its brevity. Furthermore, the traveller's statement that they sailed along the shore of the country of Malik az-Zahir until they arrived at Mul Jāwa can be interpreted as indicating that they did not cross the Straits of Malacca or Sunda, and the next stop might still have been on the same island.
In this case the port visited by Ibn Baṭṭūṭa could be sought on the eastern coast of Sumatra, [47] possibly in the kingdom of Melayu-Jambi 7 at the estuary of Batanghari River, where a massive Buddhist temple compound of the eleventh through thirteenth centuries has been excavated. At the time of Ibn Baṭṭūṭa's visit Melayu was a subject to Majapahit and, at the same time, as an heir of Srivijaya, it might have still claimed some nominal control over the ports across the Straits of Malacca. 8 It might thus have been Majapahit, together with its dependencies, that was meant by a country of two months' journey length. The Mul Jāwa of Ibn Baṭṭūṭa, similar to the Zābaj of classical writers, might be identified with the Javanese empire as a political entity, but not with Java, Sumatra or the Malay Peninsula as geographical regions.

Court Etiquette and Ethnography
Ibn Baṭṭūṭa's accounts of both Samudra and Qāqula contain data on local customs, etiquette [48] and everyday practices, but the former includes much more of it. Passages dealing with the traveller's stay in Samudra contain, among other things, various ethnographic details related to greetings, food, clothes, material culture and social relations as well as the description of a marriage ceremony and insights into court etiquette. Some of the practices described by Ibn Baṭṭūṭa appear to be common for the courts of the medieval Islamic world in general; others are typical of the Malay states and references to them can be found in Malay historical chronicles.
Ibn Baṭṭūṭa's arrival at the sultan's palace is a glimpse into the courtly routines: [49] When we went towards the Sultan's house and were near it we found spears fixed [50] in the ground on either side of the road. This is a sign that people should dismount and no one riding should go any further. We dismounted and went into the audience chamber where we found the Sultan's deputy… He rose and greeted us. Their form of greeting is shaking hands. We sat with him and he wrote a slip of paper to the Sultan informing him about our coming, sealed it and handed it to one of the pages, who brought the answer written on the back. Then a page brought a 7 Probably Malāyur of Ibn Sa'īd. Compare also Marco Polo's account of a kingdom called Malaiur (Polo 1875, 2:261). 8 There is indirect evidence that in the late thirteenth century Melayu still claimed lands in the Malay Peninsula (Jack-Hinton 1964, 59).
buqsha, which is a clothes bag. The deputy took it in his hand, took me by the hand and took me into a little apartment they call a fardkhāneh […] It was his daytime rest room because it is the custom that the Sultan's deputy comes to the audience hall at dawn and does not leave it till nightfall. It is the same with the Wazīrs and great amirs. From the buqsha he took three aprons, one of pure silk and linen, three pieces of clothing which they call underwear, of the apron type, three pieces of different types which they call 'middle-wear', three woollen mantles, one of them white, and three turbans. I put on an apron in place of my trousers, according to their custom, and one of each kind of clothing. My companions took what was left. Then they brought food, which was mostly rice, then a kind of beer, and then betel, which is the signal for departure. We accepted it, we rose, and the deputy rose when we did. (Beckingham 1994, 877-78).
Dismounting as a sign of respect to a ruler was a common practice in the medieval East and [51] West (Ibragimov 1988, 116), as well as royal gifts of honourary robes. In the Malay world, ceremonial exchange of textiles and clothes also had important ritual and social functions: it bestowed spiritual protection and well-being on the receiver and cemented social relations and hierarchy (Hall 2001, 213-16). Frequent references to and detailed descriptions of textiles, their exchange and their ceremonial display can be found both in Hikayat raja-raja Pasai and Sejarah Melayu. In the sultanate of Malacca, the presentation of honourary robes came with ceremonies of appointment to high positions and the reception of diplomatic missions, the number and quality of pieces of clothing depending on the receiver's rank (Revunenkova 2008, 117, 216). As one can see from Ibn Baṭṭūṭa's account, his companions received different sets of clothes. The traveller also indicates the well-known Malay custom of wearing sarong (fūṭa) instead of trousers.
Another cultural element that played a significant ritual and etiquette role in the Malay [52] world was the practice of chewing betel. A mixture of betel leaves, areca nuts, lime and optional spices was served after feasts and official ceremonies as well as regular meals. As a sign of honour and hospitality, betel was offered to guests not only in Southeast Asia but all along the shores of the Indian Ocean. In the Middle Ages this practice spread beyond its cultural homeland up to South Arabia and Hijaz, as it is reported by al-Mas'ūdī (Maçoudi 1863, 2:84). By the time of his arrival at Sumatra, Ibn Baṭṭūṭa was long familiar with betel and apparently chewed it more than once in India and other places. His travelogue outlines the extent of the spread of betel chewing in the fourteenth century, as it contains multiple references to it not only in India, the Maldives and Southeast Asia but also in East Africa and the Mashriq.
Ibn Baṭṭūṭa reports multiple cases of being offered betel, and it can be assumed that he [53] adopted the practice during his long stay in India or at least joined in on social occasions. Throughout his accounts one also notices the similarity of betel etiquette all around the Indian Ocean world and its general association with communication and hospitality. A different function of betel specific to Malay culture is described in Ibn Baṭṭūṭa's account of the royal marriage ceremony: I was present on the day of the unveiling of the bride. I saw they had erected in [54] the middle of the audience hall a big tribune and spread it with pieces of silk. The bride came on foot from within the palace with her face visible. With her were about forty ladies, wives of the Sultan, his amirs and his Wazīrs, who held up her train. They were all unveiled. Everyone present, whether high or low, could look at them. This was not their practice except at weddings. The bride climbed onto the tribune. In front of her were musicians, men and women, playing and singing. Then the groom came on an elephant caparisoned, with a throne on its back with a canopy over it as over a palanquin. The said groom had a crown on his head. To his right and left were about a hundred sons of maliks and amirs, dressed in white, riding caparisoned horses, and with caps on their heads encrusted with precious stones. They were of the same age as the groom and were all beardless.
At his entry dinars and dirhams were strewn among the people. The Sultan sat on [55] a raised place from which he saw it all. His son dismounted, kissed his foot, and climbed the tribune to his bride. She rose and kissed his hand. He sat beside her and the ladies fanned her. They brought areca nuts and betel, which he took in his hand and put into her mouth. Then the groom took a betel leaf in his mouth and then put it in hers, and this was all done in the public eye. Then she did what he had done. Then she was veiled and the tribune with both of them on it was carried into the palace. The people ate and went away. Next day his father assembled the people and made his son heir apparent. The people swore allegiance to him and he gave them profuse gifts of robes and gold. (Beckingham 1994, 912-13) In what appears to be quite a detailed ethnographic description, Ibn Baṭṭūṭa depicts recog- [56] nisable elements of the Malay wedding ritual, among them the Hindu practice of enthroning the bride and the groom on a tribune. Betel also plays a part in Malay courtship and marriage ceremonies, as it is traditionally presented during the betrothal and has an explicit sexual symbolism (Reid 1985, 533). However, this particular ritual described by Ibn Baṭṭūṭa cannot be found in modern Aceh and, to my knowledge, is not mentioned in other sources, the closest practice to it apparently being that of a Javanese bride and groom throwing betel leaves at each other (Reid 1985, 533). Providing that another common element of Hindu-influenced Malay weddings is the bridal couple feeding each other turmeric rice, it could be assumed that Ibn Baṭṭūṭa's account features a variant of the ritual that did not survive to the present day.
The traveller's claim that recently converted Sumatran women covered their faces except [57] at weddings is somewhat confusing, but this might be attributed to the sultan's religious eagerness and limited to the ladies of higher rank at court. Otherwise, Ibn Baṭṭūṭa might be misrepresenting this in his desire to depict the sultanate as a stronghold of Islam. However, the majority of ethnographic details he provides on Samudra seem to be accurate. They correspond to what is known about the Malay culture from local chronicles and later ethnographic evidence, and thus either seem to confirm Ibn Baṭṭūṭa's credibility as an eyewitness, or at least his credibility as a reliable transmitter of some knowledge of the Malay world acquired from other sources.
As far as can be seen from the narrative, the traveller does not seem to have experienced [58] cultural shock in Samudra. His reception there did not differ much from those found in other parts of the Indian Ocean world: Ibn Baṭṭūṭa was greeted with presents, betel and honourary robes and introduced to the local community of 'ulamā'. Some of the local customs he finds interesting or unusual, since he considers them worth noting, but still appears to perceive them as acceptable variations of the norm. As opposed to this tolerance in the accounts of Samudra, in the few brief paragraphs dealing with the court of Mul Jāwa, Ibn Baṭṭūṭa is surprised by local mores at least twice: in regard to the sultan's habit to sit on the bare ground and after witnessing a ritual suicide. The traveller's interaction with the sultan is defined by a certain cultural miscommunication and a lack of mutual interest: the ruler laughs in reply to the guest's amazement at the local practices and expresses his surprise in turn.
The language barrier seems to have contributed to the cultural one. After greeting the [59] sultan with the traditional formula Al-salām 'alā man ittaba 'a al-hudā, 9 Ibn Baṭṭūṭa concludes that those present could understand nothing but the word salām. He mentions an interpreter who assisted their brief conversation with the sultan, and says he could not understand the speech of the man who killed himself in front of the sultan, so that a person from among the courtiers had to explain it. The account of a ritual suicide at the assembly is perhaps the most confusing part of Ibn Baṭṭūṭa's description of the Malay world. The traveller himself defines it as an 'ajība-a wonder: In this Sultan's assembly I saw a man with a knife like a billhook. He laid it on his [60] neck and spoke at length what I did not understand. Then he took the knife in both hands and cut his own throat. His head fell to the ground because the knife was so sharp and his grip of it so strong. I was astounded at what he had done. The Sultan said: 'Does anyone do this among you?' I said: 'I have never seen this anywhere.' He laughed and said: 'These are our slaves and they kill themselves for love of us.' He ordered the body to be carried away and burnt. The Sultan's deputies, the state officials, the troops and the common people went out to the cremation. He granted ample pensions to his children, wife and brothers, and they were highly honoured because of what he had done. Someone who had been present at that assembly told me that what the man had said had been an affirmation of his love for the Sultan, and a declaration that he was killing himself for love of him, as his father had killed himself for love of the Sultan's father, and as his grandfather had done for love of the Sultan's grandfather. (Beckingham 1994, 883-84) There seems to be no evidence of Malay or Javanese practices of suicide in the name of a [61] ruler, or that of royals sitting on the ground. However, most of the motifs that appear in Ibn Baṭṭūṭa's description of Mul Jāwa can be found in medieval Arabic and European accounts of India. Ralf Elger highlights the parallels between the two abovementioned motifs (suicide and sitting on the ground) in the Riḥla and Marco Polo's description of Ma'bar, and assumes that Ibn Baṭṭūṭa must have transferred them from one place to another (Elger 2010, 83). One more motif can also be added-that of the absence of horses, as is the case in both sources (Polo 1875, 2:324-25). This, however, does not necessary mean that Ibn Baṭṭūṭa borrowed from Marco Polo directly, since similar motifs related to India appear to have been circulating in both Western and Middle Eastern medieval literature. An account of religious suicides, very similar to that of Marco Polo and also related to Ma'bar, is provided by Odoric of Pordenone (c.1286-1331) (Svet 1968, 181). But what appears to be much closer to Ibn Baṭṭūṭa's version is the report of another European, Jordan of Severac (c.1280-1330): In this Greater India many people worship their idols the following way: if a man [62] falls ill or gets into trouble he vows to the idol and swears to keep the oath in case he recovers. After recovery these people… parade to the idol singing and dancing… and while worshipping it they have in their hands a two-handled knife, 9 "Greetings to whoever follows the true guidance," a formula used to address non-Muslims (Gibb 1929, 367). similar to those used in currying leather; and after having celebrated enough, one places the knife behind his neck, pulls it with force and cuts off his own head in front of that idol. (Svet 1968, 148-49) Both travellers mention a special knife used by the fanatics. Yule reports at second hand [63] that this crescent-shaped instrument called karavat could be still found in Bengal in his time (Polo 1875, 2:334). Suicides in performance of religious vows are known to have occurred in different parts of India, but the one described by Ibn Baṭṭūṭa is different as a deity is replaced by the king. Self-sacrificing for kings is also mentioned in medieval texts dealing with India, including that by Abū Zaid al-Sīrāfī, but is usually said to be performed by people throwing themselves into fire (Nainar 2011, 93-94). The two motifs might have converged in Ibn Baṭṭūṭa's account, or they converged before he borrowed it from an unknown source.
The same can be said on the confusion between Malay and Indian customs: rather than [64] blaming Ibn Baṭṭūṭa or Ibn Juzayy for replacing motifs in order to conceal borrowings, it must be noted that mixing up accounts related to India and Southeast Asia was common to Arab geographers already in the classical period. Classical writers and those who copied them do not tend to clearly distinguish the two regions, which is hardly surprising providing the extent of Indian influence in the west of the archipelago before the advent of Islam. While Ibn Baṭṭūṭa's accounts of Mul Jāwa most probably contain motifs derived from other texts, it is still not impossible that these rituals, along with other Buddhist or Shaivic cultural elements, were adopted and practiced in Southeast Asia at that date.
Arguing that Ibn Baṭṭūṭa's account is either authentic or fabricated seems to pose equal [65] difficulties, and, paraphrasing Hamilton Gibb, by exactly the same kind of reasoning it could be 'proven' that although Ibn Baṭṭūṭa undoubtedly was in the Malay Archipelago, he at the same time never went there (Gibb 2008, 12). Though it was possible to construct the description of the region without going there, nothing proves that Ibn Baṭṭūṭa actually did so. Unless strong evidence of fabrication is discovered, I would give him the benefit of doubt-to the same extent as to other medieval travellers who report very similar stories. As admitted by Ibn Khaldun, "it often happens that people are (incredulous) with regard to historical information, just as it also happens that they are tempted to exaggerate certain information, in order to be able to report something remarkable… Therefore, a person should look at his sources and rely upon himself… He should distinguish between the nature of the possible and the impossible. Everything within the sphere of the possible should be accepted" (Ibn Khaldun 1989, 146).

At the Edge of the World of Islam
Travel has always been seen as a quest for the Other and eventually a way to redefine and [66] rediscover the Self. Ibn Baṭṭūṭa's description of the Malay Archipelago seems to capture a rare moment when the traveller steps out of the world of Islam, crossing a critical cultural border and encountering the non-Muslim Other. But was the traveller's curiosity and interest in human beings that universal? Was he open to any encounter, or only to those of a certain kind? Marlène Barsoum argues that Ibn Baṭṭūṭa did not seek the exotic Other, and his striving for knowledge was limited to the Islamic world-a cosmopolitan but still homogenous society. She describes his cultural worldview as rooted cosmopolitanism, since the traveller had the luxury of covering vast territories without facing a truly foreign reality in terms of religion, language or social organisation (Barsoum 2006, 197-98). Roxanne Euben also observes Ibn Baṭṭūṭa's largely negative attitude towards Christians and Jews and his failure to differentiate between Hindus and Buddhists in India, resulting in his frequent references to undifferentiated "infidels" (Euben 2006, 77).
In most cases encounters with the religious Other do not seem to challenge the traveller's [67] cultural comfort as they occur within Islamic lands where non-Muslims were subjects to Muslim rule. Ibn Baṭṭūṭa's voyage to China thus presents an exceptional situation, as he traverses an ultimate border, leaving the world of Islam behind. The only other similar occasion appears to be his visit to Constantinople, where the traveller "experiences the city as a succession of multiple if unseen barriers: he is hemmed in by the unwelcome peal of church bells, nonplussed by pictures of 'creatures, both animate and inanimate,' and denied entry to the Hagia Sophia" (Euben 2006, 78). Neither did Ibn Baṭṭūṭa feel at home in China, as he describes his feelings as follows: China, for all its magnificence, did not please me. I was deeply depressed by the [68] prevalence of infidelity and when I left my lodging I saw many offensive things which distressed me so much that I stayed at home and went out only when it was necessary. When I saw Muslims it was as though I had met my family and relatives. (Beckingham 1994, 900) Otherwise, Ibn Baṭṭūṭa's cosmopolitanism and ability to embrace cultural complexities is [69] remarkable. But as soon as he crosses an invisible barrier somewhere on the way from Samudra to Qāqula, a certain discomfort and anxiety seem to evolve. In the country of Ṭawālisī, he tries to avoid attending a banquet "for they are infidels and it is not lawful to eat their food" (Beckingham 1994, 885). The traveller does not mention, however, how he dealt with this issue later on during his stay in China. Also, some traditional tales of exoticism come to the scene, contributing to the image of a barbaric and mysterious world. As pointed out by David Waines, the description of China is framed by two tales of the exotic and the monstrous, i.e. those of the female-ruled Ṭawālisī and of the ship's encounter with the Rukh, which precede and follow the China story and both occur in an unknown sea symbolically separating China from the rest of the world (Waines 2010, 191). The Malay Archipelago lies still before this water barrier, in between Muslim lands and outer space, and thus appears to play the ambiguous role of a frontier.
Apart from this role of the eastern edge of the world of Islam, mirroring Ibn Baṭṭūṭa's [70] homeland in the west, Southeast Asia does not seem to be of much interest or significance to the traveller. It was not his travel destination but a transit point on the way to his originally established goal-China. The fact that he originally intended to reach China by land indicates that Ibn Baṭṭūṭa was most likely not interested in the Malay Archipelago per se. Of the benefits reserved for 'ulamā', only Samudra could offer him some, and he enjoyed the hospitality of Malik az-Zahir for a while; but in Mul Jāwa he did not stay longer than needed for the ship to prepare for the next passage. Of the four searches of Ibn Baṭṭūṭa distinguished by Ian Nettonthose for the shrine, knowledge, recognition or power, and satisfaction of his wanderlust (Netton 2008)-non-Muslim Southeast Asia could fulfil the last one only, and the indifferent hospitality of Mul Jāwa's ruler would hardly have encouraged the traveller to stay. However, Ibn Baṭṭūṭa was still eager to satisfy his basic curiosity and add more routes to his ambitious travelogue.
The image of the Southeast Asian (or Malay) Other in the Riḥla can be viewed as derived [71] from two principal sources-Ibn Baṭṭūṭa's experiences of encounter and his (as well as Ibn Juzayy's) preconceptions which defined the interpretation and representation of those experiences in the text. As images of the Other are constructed in opposition to the Self, Ibn Baṭṭūṭa's own identity was one of the factors which informed his perceptions of foreign cultures. Throughout the Riḥla the traveller appears to identify as Arab, Maghribi, Muslim, Sunni, Maliki and white (and more implicitly as male, 'ālim and urbanite). André Miquel concludes that Ibn Baṭṭūṭa's Sunni identity prevailed over the Maliki one, as his attitude to all the four schools within the Sunni block seems to be positive (Miquel 1978, 75-77). The repertoire of identities appears to vary depending on where the traveller went: as aptly noted by Euben, the closer Ibn Baṭṭūṭa came to the borders of the world of Islam, ever broader allegiances were involved (Euben 2006, 76). Once he reached a frontier, regional and sectarian divisions of the Islamic heartland faded into insignificance, and oppositions between Muslims and non-Muslims, black and white came into play. Needless to say, the Sunni sultanate of Samudra, at the very edge of the umma, must have felt much safer and more pleasing to the traveller than neighbouring non-Muslim Malay states. This eastern edge of the world of Islam was also notably opposite to the Maghreb, where [72] Ibn Battuta started his riḥla. Similar to North Africa and al-Andalus, the Malay Archipelago contained a borderland with non-Islamic lands which was constantly negotiated in military campaigns. But unlike in the Arabs' struggle against the Reconquista at the time, here Muslims seemed to be winning, not losing. The way in which Southeast Asia mirrored Ibn Battuta's native Maghreb is graphically illustrated by reconstructions of al-Idrīsī's planisphere. In these, the Indian Ocean is depicted as being little larger than the Mediterranean Sea, while its single, eastern outlet is little wider than the Straits of Gibraltar. East appears to have mirrored west on the map as well as in the mind.
A source of preconceptions that shaped Ibn Baṭṭūṭa's description of Southeast Asia, appar-[73] ently through Ibn Juzayy, can also be found in the Arab geographical tradition related to the region. Along with India and China, Southeast Asia constituted the Islamic world's own Orient, rich in luxury trading commodities and inhabited by exotic peoples. Similarly to the later Orientalism of the West, the medieval Arab fascination with the East appears to have been initially informed by Greek representations of India as a marvellous wonderland and a home for various mythical creatures. With the extension of Arab geographical knowledge to the east, the Malay Archipelago was included into this Indian discourse, and the expansion of Islam to Hindustan later dislodged the most fantastic motifs to India's eastern borders and to the South China Sea.
Classical Arab geographers do not seem to distinguish Southeast Asia from India, viewing [74] it as part of a larger cultural zone of al-Hind. 10 The zone was considered to cover a vast area up to the borders of China, and its inhabitants were generally referred to as Indians. Similar climatic conditions and evident Indian influence in both maritime and mainland Southeast Asia perhaps contributed to this point of view. The majority of Arab writers place the "Indian islands" at the eastern margins of al-Hind, adjacent to China, and acknowledge their similarity with India proper. Scattered in the southern seas between India and China (and sometimes confused with the islands off the Eastern African coast), the Malay Archipelago of medieval Arabic texts leaves an impression of a vast and largely uncharted transit zone between the 10 According to Marshall Hodgson, Arabs followed the Iranian tradition of dividing the inhabited world into seven great realms-those of Arabs, Persians, Romans, Indians, Turks, Africans and the Chinese (the list could vary) (Hodgson 1977, 1:456).
Arab lands and China, between the known and unknown world (if one considers the eastern limits of the archipelago), and between civilisation and barbarity. It was not until the fourteenth century that Southeast Asia started to be recognised as a [75] separate cultural zone and the image of a liminal space and a wonderland gave way to a new conception of the region. This new move appears to be related to the expansion of Islam in the archipelago and the introduction of the ethnonym al-jāwī. In the earlier Arabic texts no special word for Southeast Asians can be found, as they are usually referred to as the people of a certain island or country. According to Laffan, the earliest known mention of a person identified as al-Jāwī can be found in a Yemeni chronicle and dates back to between the 1270s and early 1300s, while another one appears in 1355 on a tombstone in Pasai (Laffan 2009, 40). Meanwhile, Ibn Baṭṭūṭa applies this word as an adjective on different occasions in his Riḥla, giving the impression that it came into general use.
Laffan defines al-jāwī as "a pan-ethnic ascription used by Arabic-speaking outsiders cog-[76] nisant of the importance of Java in the 13 th and 14 th centuries, and (increasingly) an Islamic cultural one for insiders" (Laffan 2009, 43). The introduction of this ethnonym thus seems to mark the Islamisation of the archipelago, on the one hand, and the Arab rediscovery of Java and the region in general, on the other. Furthermore, it suggests not only an increasing presence of Arabs in Southeast Asia, but also that of Southeast Asian Muslim pilgrims and students in Yemen and Hijaz: The shift from fabulous Zâbaj, the source of monsters and spices, to Jâwa, a recog- [77] nizably Islamizing contact zone, would be confirmed only once Southeast Asians made their presence felt in the Middle East. (Laffan 2009, 40).
With the thirteenth-and fourteenth-century developments, the Malay Archipelago moves [78] closer to the Arabs as both the limits of the known world and the frontiers of the world of Islam shift to the east. Further Islamisation of the region and the establishment of the new conception of Jāwa in the Arab picture of the world continued for a long time after the Riḥla was written down. In the age of Ibn Baṭṭūṭa, these processes were still at an early stage. Only the western extremity of the archipelago entered the world of Islam while the rest still remained outside, but the campaigns of the Samudra sultans slowly pushed the barrier forward.
For Ibn Baṭṭūṭa, as the first Arab to report the advent of Islam in the region, Southeast [79] Asia was no longer a foreign land at the edge of the known world. It was now a constantly negotiated borderland, and this feeling of a frontier appears to have informed the opposition of Jāwa and Mul Jāwa in Ibn Baṭṭūṭa's account. In his conception of the region, the binary of Muslim/"infidel" plays a major role, bringing with it related patterns and cultural stereotypes. This opposition can also be generally interpreted as that between the Muslim Self and the non-Muslim Other, since the traveller appears to identify with the vast and cosmopolitan world of Islam and feel alienated in the space beyond.
The country of Jāwa-unspecified Muslim lands subject to the sultan of Samudra-is de-[80] scribed in a positive and realistic way as a stronghold of Islam: its wars with neighbouring kingdoms are unequivocally presented as jihād, and the sultan is praised for his religious knowledge and piety both in the Southeast Asian chapters and at the end of the Riḥla (Defrémery and Sanguinetti 1858, 4:332-33). In some cases Muslim categories seem to be superimposed, among them jizya collected not from the people of the Book, and Malay women veiling their faces. The "infidel" Mul Jāwa, alternatively, is depicted through the traditional topoi of classical Arabic accounts of Southeast Asia, such as descriptions of spices and bizarre customs of "Indian" people. The contrast between Ibn Baṭṭūṭa's representations of the two countries can also be seen in the very structure of the narrative and even its credibility: while the account of Samudra is detailed, fashioned to create an impression of plausibility and indeed finds many parallels in local sources, the passages dealing with Mul Jāwa and Qāqula are brief, inconsistent and more likely to contain borrowings.
Another point is Ibn Baṭṭūṭa's perceptions of local customs and etiquette. As mentioned [81] earlier, the traveller does not demonstrate any signs of cultural shock in Samudra, as he seems to have found many practices familiar and belonging to the cosmopolitan cultures of the Islamic and Indian Ocean worlds. In Mul Jāwa, on the contrary, he expresses amazement with local mores acting in the best traditions of the 'ajā'ib literature. Bearing in mind that, despite Samudra's recent conversion, medieval Malay states described by Ibn Baṭṭūṭa must still have been culturally very close, I would question whether the difference was that tangible. The religious and language barriers appear to have been much more significant for the traveller, and they apparently provoked the alienation of non-Muslim parts of the archipelago as a whole. Bringing up the classical discourse of exoticism, in this light, seems to be nothing but a narrative technique aimed to outline the invisible border which was, perhaps, more real for Ibn Baṭṭūṭa than for contemporary Malays.

Conclusion
Standing out among late medieval Arabic texts dealing with the Malay Archipelago as the [82] only original account, Ibn Baṭṭūṭa's description of the region appears to be, at the same time, an inherent part of the post-classical tradition. A number of chronological discrepancies, geographical confusions and other misrepresentations found in the text can hardly be explained by the corruption of the text by later editors and copyists alone. Along with the general discontinuity of the narrative structure they seem to indicate involvement of compiling methods and interpolation of material from different sources. And although there is no particular medieval text I could point to as Ibn Baṭṭūṭa's or Ibn Juzayy's source of borrowing, a number of indirect parallels with Arabic and European accounts make it highly possible that the traveller's description of the Malay Archipelago contains some amount of second-hand data. However, this does not shed light on whether Ibn Baṭṭūṭa had been to the Malay Archipelago or not. He might have gone all the way to China, visited only Samudra and turned back, or just never ventured to the east of India, but in any case other sources could have been used in his report.
Whether Ibn Baṭṭūṭa's account is original or not, the traveller's conception of maritime [83] Southeast Asia represents an interesting case of imposing invisible cultural borders on the map. The different ways Ibn Baṭṭūṭa describes parts of the Malay Archipelago that were already or had not yet been converted to Islam are revealed through the opposition of Jāwa and Mul Jāwa in his narrative. While Muslim Jāwa is described in a positive and realistic way, the "infidel" Mul Jāwa remains a scene for traditional tales of "the wonders of India" with its exotic spices and bizarre customs. In Jāwa Ibn Baṭṭūṭa seems to feel at home, recognising many elements of cosmopolitan Islamic culture, but in Mul Jāwa he is constantly amazed by the local mores. The religious and language barrier encountered by the traveller was perhaps much more tangible for him than for the local Malays, and it made him perceive the cultural differences as more intense than they were. Consequently, the recently converted Samudra was depicted as a normalised space and a stronghold of Islam, while the image of its non-Muslim neighbours accumulated various cultural stereotypes and preconceptions reserved for the exotic Other.
Ibn Baṭṭūṭa's conception of the Malay Archipelago can be viewed as a display of the intrinsic [84] spatial paradigm of the Riḥla, on the one hand, and as an inherent part of the Arab geographical tradition of describing the region, on the other. The second perspective can be referred to regarding his employment of the place names of Qāqula and Jāwa, the former belonging to the classical geographies and the latter representing the newly evolving conception of the region. With the introduction of Jāwa and the ethnonym al-jāwī into Arabic discourse, the western part of the archipelago gained recognition in the Middle East as a separate cultural zone. Ibn Baṭṭūṭa was among the first writers to apply these names; hence his accounts appear to mark the initial stages in the development of the modern Arab conception of the Malay world. As the new place name was associated with Muslim islanders, Ibn Baṭṭūṭa's report also marks the beginning of the expansion of Islam in the archipelago and the region's piecemeal incorporation into the umma.
The opposition between Jāwa and Mul Jāwa is unique to the Riḥla. This conception does not [85] appear either in earlier or in later Arabic texts, and perhaps reflects the temporary historical situation when the frontier of the Islamic world crossed Sumatra. Indeed, with the Islamisation of Java and the fall of Majapahit by the early sixteenth century, the Mul Jāwa of Ibn Baṭṭūṭa, as a powerful "infidel" state controlling the spice trade, ceased to exist, and the new Muslim Java became Jāwa. But as long as it was still there, Majapahit was an heir to all the topoi and the discourse of exoticism previously associated with Srivijaya. Ibn Baṭṭūṭa's description of the archipelago can thus be seen as a transitional phase in the development of a new Arab conception of the region, which marked the expansion of the Islamic world further to the east and the dislodging of the classical geographers' ideas of the "Indian islands" beyond its borders.
This perspective of Ibn Baṭṭūṭa's accounts of maritime Southeast Asia brings up a question [86] about the extent to which medieval geographical narratives could be shaped by cultural perceptions and images of the Other along with actual knowledge of physical geography. As I believe to have demonstrated in this essay, the representations of the region in the Riḥla appear to be based to a large degree on the contemporary political and religious situation and its interpretation by the traveller. The religious border produced a cultural barrier, and the latter affected Ibn Baṭṭūṭa's narrative. Misrepresenting Muslim North Sumatra in familiar terms, the traveller accepted it as the periphery of the world of Islam; and, othering non-Muslim parts of the archipelago, he alienated the space beyond. The feeling of a frontier could have been all the more critical as it mirrored the western edge of the umma which Ibn Baṭṭūṭa left behind. In this respect, his conception of Southeast Asia constitutes a part of the general spatial paradigm of the Riḥla and can be put into a broader context of geographic and ethnographic discourses in medieval Islamic culture.