THE PLACE NAMES OF EURO-AFRICA IN THE KANGNIDO

Nurlan Kenzheakhmet Nazarbayev University Astana,

he +RQLO NDQJQL \žNWDH NXNWR FKLWR ␧а⮶⨶↧ԓ഻ representations of the world, it has received much T 䜭ѻെ (Comprehensive Map of Integrated Lands attention, especially since there are controversial and Regions of Historical Countries and Capitals, challenges in interpreting its data. This article is one hereafter the Kangnido), the oldest Korean world map, of several by the author identifying the place names was compiled in 1402 (Fujii et al. 2007, pp. 56, 448- on the Central Asian and western parts of the map 54).1 The map [Fig. 1; Color Plate VII] is dominated and where possible determining the sources used by a huge in the center and an overly large ᵜ Korea in the east, but arguably, much of its interest Fig. 1. The Honkżji Kangnido, kept in the Honkżji Temple ݹሪ in the city of Shimabara, Nagasaki prefecture. Picture lies in the Central Asian and western territories it source:

The 14 (2016): 106–125 + Color Plate VII 106 Copyright © 2016 Nurlan Kenzheakhmet Copyright © 2016 The Silkroad Foundation by the cartographers (Kenzheakhmet 2011, 2015, LGHQWLÀHVRQO\SODFHQDPHVIRU:HVWHUQ5HJLRQRQ 2016). Many of the large issues about the conceptual them. world of the cartographers — whether or not they Sources of the Kangnido really understood the geography they were placing on the map — must be left for future discussion. For scholars, the paucity of authentic contemporary The discussion below will treat selectively a good sources poses the main obstacle to studying the many of the names which illustrate the complexities toponyms of the western region depicted on the RI LGHQWLÀFDWLRQ 7KH DSSHQGHG WDEXODWLRQ LQFOXGHV Kangnido. Among the sources are important many others as well. ones which have not previously been used for the reconstruction of the toponyms in the Kangnido, even The original map no longer exists; however, two though they were brought to light a long time ago. In th th important copies, drawn in the 15 to 16 centuries, particular, it is important to examine Chinese sources survive in Japan. Scholars consider that these two which may have served as the intermediaries beween FRSLHV³RQHSUHVHUYHG5\ƈNRNX8QLYHUVLW\·VŻPL\D those in Persian, Arabic and Turkic on the one hand, Library (hereafter referred to as the 5\ƈNRNX Kangnido) and what was inscribed on the Kangnido on the other. DQGWKHRWKHUDWWKH+RQNżML7HPSOHᵜݹሪ in the city of The western section of the Kangnido describes the Shimabara, Nagasaki prefecture (hereafter the +RQNżML general form of the Xiyu, which stretches from Africa Kangnido) — are cartogenealogical descendents of the and Europe in the west to Qumul in the east; from the RULJLQDOPDS7KH+RQNżML.DQJQLGRPHDVXUHVQHDUO\ Russian in the north to India, Sri Lanka and 220 by 280 cm and is painted on paper. In addition the Persian Gulf in the south; and includes Qirqiz (in to thousands of place-names, the map includes modern Khakasia) and Dasht-i-Qipchaq (in modern an afterword by the map’s authors, located at the Kazakhstan and Western Siberia). Chinese knew this bottom of the map, and a description of Yuan-dynasty vast region from three sources: administrative divisions at the top. The map was 1) The brisk trade relations that existed between the revised in Japan around 1560 (Miya 2006, p. 599; 2007, Arab World and other parts of the Old World beginning p. 14). According to Kenneth Robinson (2008, p. 62), well before the time of the map’s creation left their FKDQJHVLQWKHFLYLOSURYLQFLDODGPLQLVWUDWLRQSURÀOH mark on traditional Arabo-Persian geography, such as LQ&KRVžQG\QDVW\.RUHDGDWHVWKHFRPSOHWLRQRIWKH the anonymous ֺXGƈGDO¶ŒODP, and the works of Ibn +RQNżML .DQJQLGR WR VRPHWLPH EHWZHHQ  DQG .KXUUGœGKELK ,EQ 4XGœPDK DO%DOœGKXUū $ᒒmad 1549, or even as late as 1567. bin Yaᒒ\œ  DO

107 as Changbaliha 䮧ޛ䟼૸, Hudaba ᘭ 䚄ޛ, Yangyilihe 伺Ժ䟼ਸ, and Aimili ᝋᇫ䟼. Evidence in Chinese sources reveals the presence of scholars in WKH

108 Fig. 3. North Africa in the Honkżji Kangnido, with a number of place names captioned with their modern or ancient equivalents to provide some orientation.

part a pastiche of excerpts from other books, drawing heavily on the work of foreign envoys to the western regions. Fragments of his geographic accounts survive thanks to their inclusion in the geography section (dili ൠ⨶) of the Xin Tangshu (New book of the Tang [dynasty]). Besides, Jia Dan completed a map called Hainei Hua yi tu ⎧ ޗ㨟ཧെ (Map of Chinese and foreign lands within the seas). In the Jiu Tangshu (Old book of the Tang yutu entitled Huayi jianzhi 㨟ཧᔪ㖞 (Chinese and [dynasty]) appears the following statement: “He Non-Chinese administrative division). Luo Hongxian ordered an artisan to paint the Hainei Hua yi tu on a  S   VSHFLÀFDOO\ LQGLFDWHV WKDW LQ FUHDWLQJ KLV scroll. It was three zhang wide, and three zhang and atlas he consulted a Yuan-period map (now lost) by three chi high. Its scale was one cun to one hundred Zhu Siben ᵡᙍᵜ (1273–1333/?40), a fact which then li.”4 In fact, Jia Dan’s map and the geography section raises complex issues concerning the relationship in the Tangshu became a prototype for the Inner Asia between the navigation charts and the tradition of section of the Kangnido. the Islamic “academic” geography which was also available to the Yuan cartographers.   8QOLNH WKH 5\ƈNRNX .DQJQLGR WKH +RQNżML .DQJQLGR UHÁHFWV LQÁXHQFH IURP VHD FKDUWV LQ WKH Africa Guang yutu (Broad terrestrial map), a compilation The Kangnido depicts the general form of Africa, from of maps and geographic texts by Luo Hongxian 㖵 the Maghreb in the west to the east coast; from Egypt ⍚ݸ PDGHLQ7KLVLQÁXHQFHFDQEHVHHQLQWKH in the north to the Equator in the south [Figs. 2, 3]. image of southern Africa and the Indian Ocean and Contrary to what some modern authors have asserted, place names along the East African coast. While the there is little reason to believe that the Kangnido sea charts seem to have incorporated records from UHÁHFWV DQ\ UHDO NQRZOHGJH RI WKH VRXWKHUQ WLS RI actual voyages, to a considerable degree the littoral Africa, despite its roughly triangular shape. Chinese all the way from southeast Asia to southern Africa 5 mapmakers knew about North Africa — Egypt and is only vaguely recognizable. It is no surprise then the Maghreb — from Arabo-Persian maps and other WKDWRQWKH+RQNżML.DQJQLGRFLWLHVDQGRWKHUFRDVWDO foreign sources. Because of this, the Kangnido can help locations including Maqadashu (Mogadishu) in scholars reconstruct the Arabic worldview of Africa as Africa, the Malabar coast of India, Modena (Medina), they saw it around the 9th–10th centuries, a time during Baigeda (Baghdad), and Tianfangguo (Mecca) appear which the norms of classical Arab geography were as islands. The maps also contain information about being developed. legendary regions, such as the country of Wowo (r (EgyptڍœTZœT WKHFRXQWU\RIWKHKDLU\OLWWOHSHRSOHLQ 1. Mi: addition to other hard-to-identify places. In particular, The capital of Miᒲr during the tenth century was WKHFUHDWRUVRIWKH+RQNżML.DQJQLGRZHUHLQÁXHQFHG WKH FLW\ RI DO)RVWœW )XVWDW ³ZKLFK WKH .DQJQLGR in their world view of the world by two sea charts in simply calls Misi (MiᒲU  ,EQ.KXUUDGœGKELKS the Guang yutu: The “Map of Foreign Lands in the  DO0XTDGGDVū  S   &XULRXVO\ WKH Southeast Sea” (Dongnan haiyi tuᶡই⎧ཧെ) and the Arabic name Iskandiriya $OH[DQGUū$OH[DQGULD GRHV “Map of Foreign Lands in Southwest Sea” (Xinan haiyi QRW DSSHDU DQ\ZKHUH RQ WKH 5\ƈNRNX .DQJQLGR $ tu 㾯ই⎧ཧെ). In addition, the depictions of Hemaoli, QXPEHU RI XQLGHQWLÀHG SDJRGDVKDSHG EXLOGLQJV 'HOL*XOLDQGRWKHUSODFHVLQWKH+RQNżML.DQJQLGR represent the city, including one that clearly stands almost certainly came from an appendix in the Guang IRU WKH IDPRXV 0LQœU 7RZHU RU 3KDURV  Ꮱamd

109 $OOœK 0XVWDZIū DO4D]ZūQū DXWKRU RI WKH FHOHEUDWHG • Bu ᐳVWDQGVIRU)ƈDRU)XZZDKDWRZQLQ1RUWKHUQ geographical work entitled Nuzhat al-Qulub, says that, Egypt (Miller 1926-31, 1, p. 85; Miller 1986, p. 58). beside the city, but four leagues distant from it, is a • Further south along the Nile are several cities. strong castle, set on a hillock, which overlooks the sea: South of Gaosi 儈ᙍ (Qus) (al-Ya‘qubi 2011, p. 334) DQGE\UHDVRQRILWVKHLJKWLWLVNQRZQDVWKH0LQœU is Luodelini 㖵Ⲵ䟼ቬ, this Chinese name possibly a (Minaret or Tower) of Alexandria. This is one of the transcription of Radyny (= Meshed al-Radyny), the PRVWIDPRXVEXLOGLQJVRIWKHZRUOG DO4D]ZūQū QDPHRIWKHFLW\WKDWOLHVQHDU$VZDQ DO)LGœ 6 19, 2, p. 239; Ibn Ꮱauqal 1800, p. 33). 2QWKH+RQNżML  SS   ,GUūVū   S   FDOOHG LW Kangnido near the pagoda-shaped buildings are two $QIżU DO5DGLQL 0RUH OLNHO\ /XRGHOLQL WUDQVFULEHV ሻԆ ᐳ places: Xunta and Bu . The name Xunta is said 5ƈGL1ūO5XG1ūORU5ƈGKL1ūOWKH3HUVLDQQDPH WRUHÁHFW$UDELF6KDᒷœRQDQLVODQGLQ/DNH0DQ]DODK for the Nile (ҹXGƈG 1982, p. 68; Miller 1986, p. 53). east of Tinnis (Rapoport and Savage-Smith 2014, p. The Yongle dadian (The Great Encyclopedia of the 7 0LçNS  (PSHURU@  LGHQWLÀHV WKLV FLW\ WUDQVFULEHG South and east of Miᒲr, the map gives the names into Chinese as Lude Nile ⴗⲴ⌕ं, in the country of many places on both banks of the Nile River, of Misi’er ᇶᯟނ (Yao 1959, juan 3526, book 49). Liu such as Manyong, Bunasi, Abusaiyi, Liangyi, and Yingsheng argues that this name must be a Chinese $KDPLQJ ZKLFK DUH DPRQJ WKH PDQ\ LGHQWLÀHG E\ WUDQVODWLRQRIWKH3HUVLDQQDPH5ƈGL1ūOVWLOOXVHG ,EQ .KXUUDGœGKELK DQG RWKHU $UDELF JHRJUDSKHUV for the Nile today (Liu 2010, p. 93). ³0DQIDO)DL\ƈP,NKPūPZDDO'D\UDO%DKQDVœ • South of Luodelini, the map marks a place called DO)LGœ Mahadalai 哫૸㥵ֶ (an error for Mahadashu 哫૸㥵S,EQ.KXUUDGœGKELK DQG)XZZDK   S  DO4D]ZūQū , 2, p. 265; al- ᶏ  ZKLFK 6XJL\DPD  S  7DEOH  Ɋ   Ya‘qubi 2011, p. 332). LGHQWLÀHGDV0DTGDVKDZ RU0RJDGLVKR ,IVRWKHQ • The city of Abusaiyi 䱯ঌ䌭Ժ8 — Abusir, also the name appears twice on the map, once adjacent to known as Busir, in Arabic — appears in the world the Red Sea and once far to the southeast at a place RI HDUOLHU $UDELF JHRJUDSKHUV DO,GUūVū   labeled Mahehashu 哫ਸ૸਄, connected by the Nufa p. 306). Busir was an ancient city of Lower Egypt, ྤⲬ (Zufar). located at the modern Abu Sir Bana (Djedu). Busiris • Lastly, Kunzheba ഠ㘵ޛ may be a corruption of was also the name of a town in Middle Egypt, in the 4HQDRU4LQœDFLW\QRUWKRI/X[RU 0LçNS  neighbourhood of Memphis and the Great Pyramid. Its site is marked by the modern village of Abousir in 2. The Maghreb that district (Dictionary 1854, pp. 458-59). West of Tripoli routes led to the settled lands of what is now Tunisia, the region known to the Arabs • Near Miᒲr is a city that the Kangnido calls Manyong DV,IUūTū\D*HRJUDSKLFDOO\WKH0DJKUHEHQFRPSDVVHV 㴞⎼. Perhaps this refers to Memphis, which the present-day Morocco, Algeria, Tunisia, and Libya. Arabs call Manf or 0DQƈI, considering that the ,EQ .KXUUDGœGKELK JLYHV DQ DFFRXQW RI WKH KLJK second character, yong ⎼, may actually be a copyist road of the Maghreb, which extended east from error for a more likely character pu ⎖. The renowned Spain to Libya (Barqah, nowadays called al-Marj) 12thFHQWXU\ JHRJUDSKHU DO,GUūVū LQ KLV Nuzhatul ,EQ.KXUUDGœGKELKSS 7KHOLNHO\$UDE 0XVKWœTIū,NKWLUœTXOŒIœT names the town Menf (al- sources for the information on the Kangnido connect ,GUūVūSDOVR,EQ%Dᒷᒷnjᒷa 1958-2000, the region with areas to its south, where the medieval 1, p. 51). Sugiyama (2007, p. 58) points out that this Islamic maps show a huge lake (possibly representing name is written as Manfu 㴞⭛LQWKH+RQP\żMLDa Lake Chad) and other features which will be discussed Ming guo ditu and the Tenri Da Ming guo tu. IXUWKHU EHORZ $IWHU ÁRZLQJ RXW RI WKLV ODUJH ODNH • The town of Ahaming 䱯૸᰾, south of Fustat and WKH 1LOH VSOLWV LQWR WZR EUDQFKHV DQG WKHQ ÁRZV on the east bank of Nile, apparently stands for northward to lower Egypt. the modern-day city of Akhmim, which Ibn Near the north shore of this lake lies a city called .KXUUDGœGKELKDQGRWKHU$UDEJHRJUDSKHUVFDOOHG Xizhilitumasi ㍠ਚ䟼൏哫ᙍ, which is also shown $NKPLP,NKP\Q,NKPūP,NKPūPZDDO'D\U ,EQ surrounded by water. The name appears to be a .KXUUDGœGKELKSDORIWKHSHRSOH6LMLOPœVD@E\RQHRIWKHWUDGHUVRI 7DPDGXOW ZULWWHQE\,GUūVūDV7œPGDODW0LOOHU $ZGDJKRVWZKRZDVKLPVHOIRIWKHSHRSOHRI6LMLOPœVD 31, 1, p. 81). The ruins called Tamdoult Ou Akka, are in the sum of 42,000 dinars. I have never seen or heard visible about 13 km southwest of Akka on the south anything comparable to this story in the east. I told it in DO¶,UœTLQ)œUVDQGLQ.KXUœVDQDQGHYHU\ZKHUHLWZDV side of the Anti-Atlas (Levtzion and Hopkins 1981, regarded as a novelty. [Levtzion 1968, p. 225] p. 457). )UD 0DXUR D ÀIWHHQWK FHQWXU\ &DPDOGROHVH PRQN On the Mediterranean coast, the Kangnido features and mapmaker in Venice, referred to the city of a place called Hasatanina, that is, Constantine, the city 6LMLOPœVL\DK DV Siçilmensa in his well-known Mappa in Algeria that the Arabs conquered in the seventh mundi (Brooks 2009, pp. 294-95; Mauro 1966, p. 23, FHQWXU\DQGUHQDPHG4XVDQWLQD DO,GUūVūS B13). The Kangnido marks a place called Weitamao ⛪ 4RVDQWūQDLQ,EQ%Dᒷᒷnjᒷa 1829, p. 3; Quᒲtanᒷiniyya, ᆳটSUREDEO\WREHLGHQWLÀHGZLWKDPRXQWDLQHDVWRI according to Sugiyama 2007, p. 58). Moving east 6LMLOPœVL\DK$\GHPXU DO,GUūVū  along the coast, the map marks the capital of Ifriqiyah $IULFD  ,EQ .KXUUDGœGKELK·V 4D\UXZœQ which the In the west of the Maghreb, several of the names on Kangnido calls Kanpuban ⴻᲞ㡜, the site of modern- the map may be identiried with locations in what is GD\.DLURXDQLQ7XQLVLD DO,GUūVūS  today Morocco. Zaobian 㰫䚺, perhaps a corruption of Zaosa 㰫仟, • Ani 䱯ቬ and Aixue ᝋ㯋 are possibly Adna and UHIHUVWR6ƈVD SUHVHQWGD\6RXVVH DWRZQQHDU7XQLV $TVDZKLFK$UDEJHRJUDSKHUVDOZD\VZULWHDV6ƈV DO,GUūVū   S  Rapoport and Savage- DO$GQk DQG DO6ƈV DO$TᒲD ,EQ .KXUUDGœGKELK Smith 2014, p. 451). Moni sounds like %ƈQDK or Bône. S$O%DOœGKXULS  Bone is modern Annaba, one of the most important FRPPHUFLDO FHQWHUV RI WKH $OJHULDQ FRDVW DO,GUūVū • Ashiye 䱯ཡ㙦, refers to ԵVūOœ$]ūOœ $FLOD³ 1836-40, 1, p. 246; Rapoport and Savage-Smith 2014, corrupted into Arzila, a charming little seaside resort p. 424). south of Tangier (Rapoport and Savage-Smith 2014, p. 447; Levtzion and Hopkins 1981, p. 444). The Arab geographers generally give meager accounts of other towns in the Maghreb. Groups ‡ 7KH .DQJQLGR LGHQWLÀHV &DVDEODQFD ZKLFK WKH of villages, so common elsewhere in Africa, did not Arabs (and Berber) called Anfa, as Anifa 䱯ቬ⌅ (al- exist. In addition, Portugal began conquering city ,GUūVūS0LOOHUS/HR after city in Northern Africa during the 15th century, Africanus 1896, 2, pp. 396, 552, 652). which meant that some places ceased to exist or, when • Malidasa 哫䟼㥵᫂UHIHUVWR0RVWœVDRU0LVLᒷœVD DO rebuilt, were renamed, which affected the makeup of ,GUūVūSDOVRVHH0LOOHUS all subsequent maps. 177). 3. South and Southeast Africa ‡ %DODZD\DQJ UHIHUV WR %DUJKZœᒷah, a Berber 7KH+RQNżML.DQJQLGRVHHPVWRKDYHPDGHXVHRID FRQIHGHUDWLRQHVWDEOLVKHGLQWKH7œPDVQœSURYLQFH QXPEHURI´PRGHUQµPDSV³WKDWLVPDSVUHÁHFWLQJ extending along the Atlantic coast of Morocco, contemporaneous knowledge of the world drawn EHWZHHQ6DOpDQG6DÀIURPWKHth to the 12th century from the accounts of explorers or navigators, even if the (Rapoport and Savage-Smith 2014, p. 422.). map also relied heavily on a classical understanding • Shanghema кਸ哫 is probably a mistake for of the world. In particular, the map’s geography of the Ahema 䱯ਸ哫ŒJKPœW7KHUHZHUHWZRWRZQVFDOOHG East African coast draws directly from the Guang yutu ŒJKPœWƇœUūTDDQGŒJKPœWƇœLODQRUŒJKPœW$\Oœ (The broad terrestrial map) of Luo Hongxuan (1504- LQ ZHVWHUQ 0RURFFR ŒJKPœW ƇœUūTD RU ŒJKPœW 1564). The overall sense of the African landmass is :XUūNDLVQHDUWKHPRGHUQYLOODJHRI2XULNLNP deceptively suggestive of our modern understanding. southeast of Marrakesh (Levtzion and Hopkins 1981 A close examination of detail though reveals a rather p. 443). In the Kangnido the name seems to follow the mixed picture of recognizable locations and hugely IRUPRIŒJKPœW$\Oœ DO0XTDGGDVūS distorted understandings of the actual geography. 111 Fig. 4. Detail of South and East Africa in the Honkżji Kang- nido, with some inserted captioning. Classical Arab geography, and through it WKH LQÁXHQFH RI WKH nd-century geography of Ptolemy, is abundantly evident here. Like Islamic maps, the Kangnido devotes a great deal of attention to the course of the Nile. Yet it is an odd Nile indeed, its lower reaches ÁRZLQJRXWRIWKHKXJHFHQWUDO$IULFDQODNH whereas its long upper course [Fig. 4], while deriving from the river’s source in what the map designates (more or less correctly) as the Zhebulu hama 䙉н冟૸哫 -DEœO DO4DPDU  literally the mountain of the moon, located not about midway down the continent but near its southern tip, debouches into the Red Sea, not the Mediterranean.9 Chinese maps of the Ming period, such as the Liangyi xuanlan tu ޙܰ⦴㿭െ (the World Observing Map, 1603) by Matteo Ricci and Tianxia jiubianfen yeren ji lucheng quantu ཙлҍ䗩࠶䟾Ӫ䘩䐟〻ޘമ (A Complete Map of the World, 1644) called it Yueshan ᴸኡ (mountain of the moon), the &KLQHVHQDPHIRUWKH-DEœODO4DPDU :DQJ et al. 1994, Pl. 59, p. 146). The body of water that is shown to occupy much of the interior of the continent surrounds an island called Huangsha 哳 ⋉, which literally means “Yellow desert.” It perhaps refers to ᐀aᒒœUū ULPœO LQ LQQHU $IULFD-D]ūUDWDO7LEU ,VODQGRI*ROG %LODG DO7LEU RU -D]ūUDW :DQNDUœ DO7LEU RQH RI the countries of Sudan situated south of the Maghreb (see Levtzion and Hopkins 1981, p. 169). The inscription designating the “Mountain of the We observe here the pictorial description of the Arab Moon”, Zhebulu hama 䙉н冟૸哫> -DEœODO4DPDU@LV legend that claims the existence of a large lake in the located on the Kangnido south of the large lake and FHQWUDO SRUWLRQ RI WKH $IULFDQ FRQWLQHQW $O,GUūVū between it and the symbols for the actual mountains. however, writing in 1145, makes of ‘Wangara’ a piece Almost directly east of that inscription, and east of the of land, an island surrounded by the ‘Nile’ to the east river is a place called Henayisigui ਸ㌽Ӗᙍ൝, which of Ghana. In his text he tells us that it is inundated the Da Ming hunyi tu calls Fanayisiwa ⌅䛓Ժᯟ૷, and E\WKH¶1LOH·ZDWHUVDQQXDOO\DQGWKDWDIWHUWKHÁRRG the Guang yutu calls Hanayisijin ૸㌽Ӗᙍ⍕ (Wang et al. people swarm over it to search for gold. In describing S/XRSS 7KHÀIWKFKDUDFWHU WKHODNH$O,GUūVūUHFRJQL]HGWKDWWKHLQIRUPDWLRQZDV is evidently incorrectly drawn, given they way in transmitted through al-Khwarizmi from Ptolemy and which the Da Ming hunyi tu inscribes the name above. wrote: The name Henayisiwa (also Fanayisiwa) transcribes the DOūVWLƈœ(Miller 1926-ڒThis lake is just beyond the equator, and touching it. In Arabic name for the Equator, Ha the lowest part of this lake in which the rivers collect, a 31, 1, p. 80; Takakashi 1963-75 (1963), pp. 87-90). mountain protrudes, splitting the main part of the lake into two, and extending from the lake to the northeast. Locations along the southeastern African coast are 2QHRIWKHEUDQFKHVRIWKH1LOÁRZVDORQJWKLVPRXQWDLQ all indicated in cartouches placed out in the ocean, on the western side. This is the Nil of Bilad al-Sudan, on thus making their relationship to the land somewhat which most of the towns are situated. The second branch of the Nil comes out of the lake on the eastern rift of the unclear. From south to north, starting at approximately PRXQWDLQ DQG ÁRZV WR WKH QRUWK WKURXJK WKH FRXQWU\ the latitude of the source of the Nile in the Mountains of the Nuba and the country of Egypt. [Levtzion and of the Moon, they include Maoxi habila ߂㾯૸ᗵࡼ Hopkins 1981, p. 115; Levtzion 2000, pp. 73-74] / Changxi habila ᰼㾯૸ᗵࡼ, Ala 䱯ࡼ, Wa’a’wa ૷䱯 112 )LJ  'HWDLO RI WKH 0HGLWHUUDQHDQ basin and portions of Europe in the Honkżji Kangnido, with selective inserted captioning. ૷, and Kexidabin 乶㍠ᢃ 䌄, all located south of the designation for the Equator. While the equivalent for Ala is uncertain (it might represent Sofala), Maoxihabila or Changxihabila occurs in the Da Ming hunyi tu and Guang yutu,10 perhaps a transliteration for Jazirat al-Qamar (Comoro Islands) (Miller 1986, pp. ;(ֺXGƈG 1982, p. 205 ;108-9 or perhaps -D]ūUDWDO4DQEDOƈ (Pemba Island) or Marsa Qanbala (Rapoport and Savage-Smith 2014, p. 434). On account of the position assigned this island, the name Changxi for the inhabitants of the Somali coast was given to the Mozambique channel. The similarity of sound between this name and that used by the Arabs of the time to designate the big island of Pemba, Qanbalu, leaves little doubt that the Chinese name means the Zanj-i 4DQEœOD RU4DQEDOƈ ´=DQMRI4DQEDOƈµ.H[LGDELQ KDELWDWLRQZDV7DSUREDQř ᐅDEDUQœ ZKLFK3WROHP\ SHUKDSV UHIHUV WR 5D·V -DUGDIƈQ &DSH *XDUGDIXL ORFDWHGDW¼ƒQRUWKRIWKHHTXDWRUWKHH[DFWSODFHWR ancient Cape Aromata). $O-DUGDIƈQRU5D·V-DUGDIƈQ which the name referred remains uncertain. Just to the the name used by Arab navigators for the cape at the east of the mountainous island is a damaged section tip of the Horn of Africa, known today as Ra’s ‘Asir of the map, where, presumably, there might have and in European literature as Gardafui (Rapoport and been other names relating to the African coast. The Savage-Smith 2014, p. 445). map does include Mahehashu 哫ਸ૸਄ (Maqdashaw or Wa’a’wa ૷䱯૷ (:œTZœT) is a place visited by the Mogadishu), far to the southeast, connected by the Nufa ྤⲬ Arabs and located somewhere on the eastern coast  =XIDU LQWKH+RQNżML Kangnido and Guang yutu, of Africa. However, the Arab sources also refer to leaving questions as to whether the cartographer had an Asian Wowo ベベ (:œTZœT).11 $O,GUūVū EHOLHYHG any real idea of its location. In general, traditional Arab WKDW 6RIDOD ZDV FDOOHG WKH 6RIDOD RI WKH :œTZœT RQ geography preserved more and better information the basis that the indigenous speech resembled a WKDQLVUHÁHFWHGRQWKH.DQJQLGRUHJDUGLQJWKH(DVW 12 ZKLVWOLQJVRXQGRIDELUG DO,GUūVūS  African coast. Just to the north of the Equator on the map are Europe cartouches for Kuliu ᓛޝ (.XOZœ), Tibanu ởޛྤ One of the striking features of the Kangnido is its Dibanu ၓޛྤ in the Guangyu tu), and Hezu ்ং, the inclusion of Europe, even if its rendering of the last two placed in the ocean between the coast and Mediterranean world substantially distorts the way it DEDUQœ, the is depicted in the maps of classical Arab geographersמ a mountainous island. Tibanu refers to ֺXGƈG [Fig. 5]. While the map outlines what we can reasonably UDELF QDPH IRU WKH 7DSUREDQř RI 3WROHP\$ 1982, p. 57). The southernmost point of human interpret as coastlines (even if hugely deformed), the

113 Black Sea is entirely absent, and the Mediterranean by Muslims and known to Arab geographers by the not marked by any shading to distinguish water from name al-Andalus, from which evolved the name for land in the way that the map does for the oceans. the modern Spanish province of Andalucía. While Compounding the confusion in the eastern end of the some of the place names within al-Andalus in the 0HGLWHUUDQHDQZRUOGLVWKHFRQÁDWLRQRIWKH%DONDQV Kangnido are of Arabic origin, others originally were and Anatolia. The Kangnido’s Europe divides into 5RPDQ RU *RWKLF QDPHV WKDW ZHUH ODWHU PRGLÀHG ÀYH SDUWV QRUWKZHVWHUQ DQG FHQWUDO (XURSH WKH by Arabic speakers. The Kangnido marks a place it Iberian peninsula (Al-Andalus), the Italian peninsula, calls Zhebulifa 䙉н䟼⌅ at the southern end of the the Balkans, and the Crimean Peninsula. ,EHULDQ SHQLQVXOD 3URIHVVRU 6XJL\DPD LGHQWLÀHV WKH ,ƗULTמ-Northwestern and central Europe compound name Zhebulifa + Dapulu as Jabal al .1 otherwise known as (2007 p. 58: table 1, In the northwest, the mapmaker of the Kangnido Ɋ   +RZHYHU WKH QDPH Dapulu 䚄Პ冟 appears 䱯冟ቬ䱯 and Falixi ⌅ labels two countries as Aluni’a LQ WKH +RQNżML .DQJQLGR VHSDUDWHO\ IURP =KHEXOLID 䟼᱄. Aluni’a refers to 4DOƈQūD, the Arabic name for and probably refers to Tabira, an important town in Cologne (Köln) in Germany. Falixi transliterates%DUū], arf al-Gharb (Cabo de Sãoמ the Algarve (al-Gharb) or the Arabic name for Paris. In the Mingshi (History Vicente), the southern coastal region of modern- of the Ming [dynasty]), the characters Balaxi ᐤஷ㾯 day Portugal (Mónica Herrera-Casais 2008, p. 250). designate Paris, perhaps derived from the 13 There is in fact a better interpretation of the toponym Arabic name for the city. The Ming shi records: Zhebulifa. Indeed, the name Gibraltar derives from the ƗULT, which means “the mountמ-Balaxi is very far away from China. In the sixth year (1511) Arabic name Jabal al of the Zhengde reign, (Balaxi) sent the envoy Shadibai, to of ᐅƗULTµ¶$OūDO6KDUDIū Á² WKHth-century offer tribute. He advised that their country was located in Tunisian chartmaker and scholar, rendered the name the Southern Ocean. Initially, carrying his king’s orders, Herrera-Casais) ٭he had travelled four and a half years in an ocean-going Gibraltar into Arabic as Jabal al-Fat :Ibn Baᒷᒷnjᒷa writes ,٭ship.14 2008, p. 250). About Jabal al-Fat $FFRUGLQJWRWKH5\ƈNRNX.DQJQLGRVRXWKRI)DOL[L , WKHUHIRUH VHW VDLO IURP 6XEWD WR 6SDLQ DQG WKH ÀUVW Hill of victory). This) ٭Paris) lies a place named Ainu ᝋྤ, probably the place I saw was the Jabal al-Fat) historical province of Aunis in southwestern . is one of the greatest refuges of Islamism, and one To the southwest of Ainu is a place named Alazhila which forced sorrow down the necks of the idolaters. 䱯ࡼਚࡼ, probably 5DšœOD, the Arabic name for La From this place commenced Islamism, in the great YLFWRU\IRUKHUHODQGHG7œULN ᐅœULT ,EQ=LœGWKHVODYH Rochelle, the capital of Aunis. To the northeast of RI 0ƈVD ,EQ 1DVūU DW WKH WLPH RI KLV SDVVLQJ RYHU WR Ainu stood a place named Aminaxinan, represented Spain. From this circumstance it was named after him, by a pagoda-like building with a cross, so perhaps a DQGFDOOHG¶-DEDO7œULN ᐅœULT ,WLVDOVRFDOOHGWKH¶-DEDO FKXUFK6XJL\DPD SWDEOHɊ WKLQNV Fath (Hill of Victory), because his beginnings had their this may actually signify the papal city of Avignon, commencement here. [Ibn Baᒷᒷƈᒷa 1829, pp. 226-27; Ibn although it more likely is Venice (see below). Baᒷᒷƈᒷa 1958-2000, 4, p. 935] The Kangnido shows the Mediterranean coastline Thus, the name Zhebulifa most likely transliterates .٭from Dénia ('œQL\D) to Barcelona (%DUåOXQD) in Jabal al-Fat Catalonia, along the south of France, and extending 3URIHVVRU6XJL\DPDLGHQWLÀHVDQHDUE\FLW\QDPHG DVIDUDV*HQRD øDQXD LQ,WDO\7KHQDPHMalixilina Maliuli 哫ޝ䟼 UHFRUGHGRQWKH+RQNżMLEXWQRWRQWKH 哫䟼᱄䟼䛓 is either Marseille (according to Sugiyama) 5\ƈNRNX.DQJQLGR DVWKHFLW\RI0DUEHOOD S or, more likely, Barcelona (see below). To its north is WDEOHɊ $OWHUQDWLYHO\WKHQDPHPD\UHIHU Jiliruo ⊢䟼㤕ZKLFK6XJL\DPD SWDEOHɊ to DO0DUūUL or 0DZUƈU (Morón de la Frontera) in the  LGHQWLÀHVDV*LURQD+RZHYHUWKHÀUVWDQGODVW Seville province. In any event, we must be dealing characters in the name, ji ⊢ and ruo 㤕, could easily here with a town that was important in Arab Spain. have been mistakenly exchanged for the similar- The name Maliuli surely dervives from a rendering looking characters mo ⋂ and ku 㤖. If that is the case, in Arabic, and, like many of the cities on the Iberian then the actual characters, Moliku ⋂䟼㤖, may be a Peninsula in the Kangnido, it probably was a city on variant of the name Moniku ⋂ቬ㤖 that appears in the the border between Christian and Muslim territories. VDPHORFDWLRQLQERWKWKH+RQP\żMLDa Ming guo ditu Further north, the Kangnido plots a city it calls Falata and the Tenri Da Ming guo tu.15 Thus, the characters ⌅ࡼᆳ presumably the one the Arabs called Baladu- may refer to , the city-state between France Walid, Valladolid (al-Ya‘qubi 2011, p. 354). North of and Italy. it lies a city called Tangela శᠸࡼ, perhaps originally 2. Iberian peninsula (Al-Andalus) written as Tanfala శՀࡼ, which suggests the Arabic Spain, on the Iberian peninsula, was partly controlled name al-Tafaylla, in other words, the name for modern

114 Tafalla in Navarre. To the east, the name Nadula tion of Qalibli (Gallipoli), geographic logic would probably indicates Andorra, now the microstate suggest instead that it could be Qalawriyah, the Arabic between Spain and France. name for Calabria in the Southern Italy, mentioned Continuing east to the Mediterranean coast, Talaruona in other Muslim sources (Rapoport and Savage-Smith Ԇᙍ䛓 Ԇ䟼䛓 Ԇࡼ㤕䛓, an error for Talakuna Ԇࡼ㤖䛓, refers to 2014, p. 421). The name Tasina / Talina may arakuna, modern Tarragona. The transcribe the toponym Taliya or Taliyan, the Arabicמ the Arabic name position of Maxilina ZKLFK 6XJL\DPD LGHQWLÀHG name for Italy, mentioned in other Muslim sources which ZLWK 0DUVHLOOH SHUPLWV DQ DOWHUQDWLYH LGHQWLÀFDWLRQ used this name for the language of Franks. However, with %DUåOXQD, the Arabic name for Barcelona. In the toponym may in fact be a corruption of $WUœQW the Kangnido, there are several place names with (=Turenum/Trani). For other locations on the map in an initial b or m, but their sound varies between the southern Italy, see my appended table. two consonants. This variation between b and m 4. The Mediterranean world and its islands. typically appears in the initial position and entails the Even though the map’s rendering of the replacement of m by b or b by m in the several place Mediterranean does not distinguish its waters from names, such as Maxilina, %DUåOXQD in Arabic. the adjacent land masses by the use of color, at least 3. The Italian peninsula some important geographic features found in or A large Italian peninsula cuts off the Iberian DORQJ LW DUH GHSLFWHG )RU H[DPSOH ZH ÀQG Sahalina peninsula on the northeast and turns to run into the ᫂૸䟼䛓 (Sicily), the name of the island written in a Mediterranean from the north. In the far north of cartouche. Lying at the heart of the Mediterranean, it the Italian peninsula is a pagoda-like object named is the sea’s largest island, a natural stopping place for Aminaxinan 䱯ᕼ䛓᱄ই, which probably is a corruption travellers, and close to both Rome and Tunisia on the of names like DO)LQū]L\D or al-Beneçia, the Arabic name 1RUWK$IULFDQFRDVWZKLFKSURIRXQGO\LQÁXHQFHGWKH IRU 9HQLFH ¶$Oū DO6KDUDIū Á ²  WKH VL[WHHQWK island’s history. The name Sahalina probably refers to century Tunisian chartmaker and scholar, renders 6LTDOūD, the Arabic name for Sicily (Ibn Ꮱauqal 1800 p. the Italian Venice into Arabic as )LQū]L\D ZDKL\D 53). Indeed, the older Chinese geography Zhufanzhi O%XQGXTū\D 9HQLFHWKDWLVDO%XQGXTū\D  +HUUHUD called the island Sijialiye ᯟ࣐䟼䟾 (Zhao 1996, p. Casais 2008, p. 252). 133), clearly derived from the Arabic name 6LTDOūD. In Arabic and Judaeo-Arabic sources, Palermo is usually Despite the map’s deformation of the shape of the LTLOOū\D (the city of Sicily), and often justי 0DGūQDW land masses and the relative placement of cities, we can ,LTLOOū\D (Sicily) (Rapoport and Savage-Smith 2014י identify with some certainty many of the geographical S   :KLOH WKH GDUN FRORU ÀOOLQJ WKH FRQWRXUV RI names in this region and at least suggest reasonable what in terms of its placement might be mistaken for possibilities for the identity of others even if the Sicily would seem to be make it impossible to read phonetic shifts may not always be readily explained DQ\LQVFULSWLRQVIURPWKHSKRWRJUDSKRIWKH+RQNżML by formal linguistic analysis. The name Lamo ࡼ⋂ , Kangnido, Sugiyama Masa’aki has made out on it the which appears in the center of the peninsula, clearly names Molide 㧛䟼Ⲵ and Falisang ⌅䟼ẁ, which he VWDQGV IRU 5ƈPD RU 5XPL\D WKH $UDELF QDPH IRU LGHQWLÀHV DV 0DUVDUD DQG 3DOHUPR  SS  ⴗⴹ Rome. The Zhufanzhi calls Rome Lumei . To the WDEOHɊɊ ,QIDFWLWVHHPVPRUHOLNHO\WKDW QRUWKRI5RPHWKHPDSOLVWVÀYHPRUHSODFHQDPHV Molide refers to Malta, and Falisang to Balzan, a town ࿣֐䛓 哫冟 ཙлᡦ ᆳ䟼 Nainina , Malu , Tianxiahu , Talitu of Malta. ֐⌅. The name Nainina ࿣֐䛓 seemsޛ ⿯, and Banifa to be an error, instead of Jinnina ࿇֐䛓, which suggests  7KH%DONDQV the Latin name for Genoa. Malu probably refers to 7KH FRQÀJXUDWLRQ RI WKH %DONDQV RQ WKH .DQJQLGR MalfPRGHUQGD\0HOÀLQWKH6RXWKHUQ,WDOLDQUHJLRQ is unusual. Some important cities of Balkans seem RI %DVLOLFDWD 2Q WKH 5\ƈNRNX .DQJQLGR 7LDQ[LDKX to be missing, the confusion compounded by the is written as Tiandinghu ཙбᡦ, perhaps Tarquinia fact that the landmasses of Asia Minor and the (Tarchuna), an ancient city of Viterbo. Talitu probably Balkans are subsumed into one. Many of its labels ƗUDQW, the Arabic name for Taranto in the far DZDLWLGHQWLÀFDWLRQ7KHUHDUHIRXUSODFHQDPHVHDVWמ denotes south of the peninsula. Banifa transliterates Benebent, of Aminaxinan (Venice): Saba ᫂ޛ, Chedala 䓺བྷࡼ, the Arabic name for Benevento. Shangnigu кቬਔ and Lataba ࡼᆳޛ. Saba probably u, modern Zatonڒaי wa orڒaי The southeastern tip of the Italian Peninsula, known refers the Arabic name as Apulia, is shown separated from the mainland by west of Zadar; Chedala is -œGDUD or -œGKDUD DO,GUūVū  a strait and bordered on the south by high mountains, the Arabic name for modern Zadar; Shangnigu suggesting that it is dangerous for ships. While the is Sabnaqi, the Arabic name for modern âLEHQLN name Halabayinu ૸ࡼޛԕྤmay represent a corrup- 6HEHQLFR  DQG ¿QDOO\, Lataba is Lastobon (modern-

115 day Lastovo), the Slavic name of both an island and in the 10th century; its name in the Russian annals is a city. Even if the Black Sea itself seems to be missing, Korsun’, which has been proposed as the source for major locations connected with its western and the form of the name in the Arab sources (Konovalova northern shores are included on the map. Gusitanxina 2006, p. 177). On the east coast of the Black Sea lies a ƯQiyyah) is a phonetic rendering place named Hudadeli ᘭ䚄Ⲵ䟼, perhaps a corruptionڒanڒ僘ᙍశ᱄䛓 (Qus of Constantinople, present-day Istanbul. Along RI5XGLŒWūOWKH3HUVLDQQDPHRIWKHPRGHUQ9ROJD what might correspond to the sea’s western coast River. ⛿ᙍ䚄 is a toponym Wusida , probably a corruption Conclusion of Wulida ⛿䟼䚄, referring to 5ƈWœ river, an ancient name for the Danube; The toponym Xikena ᱄ݻ䛓 This study of the toponyms written on the Kangnido (the position of a large, red, gear-like marker on the LQ&KLQVHVLGHRJUDPVKDVUHVXOWHGLQWKHLGHQWLÀFDWLRQ river and west of the Gusitanxina) might correspond of most of them with toponyms recorded in the approximately to Šighnu or Šaghnu (,VWDJKQż — classical Arab and Persian geographies. In fact, it is Tomaschek 1886, p. 341), the Arabic name for modern remarkable that the names, when compared with Stagno (Ston), also known as Stagno Grande (Roman those given by Arabic geographers, show as close settlement Sta-g-n-um), in the Bay of Sabioncello, an approximation to original Arabic sound as the a historical town in Croatia (cf. Miller 1926-31, 1, p. is capable of expressing. Despite 73; 2, p. 127). The foundations of the Roman castrum the cartographic distortions in the western parts of can still be seen on Starigrad Hill. In the tenth century the Kangnido, it is possible to discern how the basic Ston was the seat of the diocese, probably located by outlines of landforms and the division of the world the church of Our Lady of Luzina. Shalehe ⋉′થ (an into several regions derive as well from Islamic error for Duolehe ཊ′થ refers to Adras, the Arabic cartographic traditions. The anonymous ֺXGƈG DO name for Durrës, historically Dyrrachium or Durazzo, ૽߹lam DQG ,EQ .KXUUDGœGKELK·V .LWœE DO0DVœOLN ZD the starting point for the Via Egnatia, the important ·OPDPœOLN are among the most detailed and earliest Roman (later Byzantine) road that led east across texts in that tradition and were based on itineraries of the southern Balkans. For other cities in the Balkans actual travel. That such information then was copied LGHQWLÀHGRQWKHPDSVHHP\DSSHQGHGWDEOH and transformed in subsequent Islamic “academic” geography is important to understand if we are to  &ULPHDQ3HQLQVXOD be able to say more (as remains to be done) about To the east along northern coast of Black Sea in the the direct Islamic sources used in the compilation of +RQNżML.DQJQLGRWKHUHLVGDPDJHREVFXULQJVRPHRI the Kangnido.16 Not the least of the issues here is the WKHQDPHV+RZHYHUWKH5\ƈNRNX.DQJQLGRLGHQWLÀHV question of the relationship between text and map, several important cities: Shatianpulu ⋉ཙᲞ冟, Qusun where the maps may be rather schematic and farther ᴢᆛ (Korsun), Puyu Პ։ (Fori, modern Foros), Chijina removed from any geographical reality than are the 䎔ਹ䛓 (Kikineiz, modern Opolzneve), Sudali 䙏䙩䟼 texts their cartographers consulted. ⌅ཡ ⋉ཙᲞ冟 (Sudaq) and Fashi . Shatianpulu is most Even given strong evidence about the Kangnido’s likely a copyist’s error for the very similar-looking reliance on Islamic sources for its depiction of the ⋉ཡᲞ冟 Shashipulu , which presumably was intended western regions, it is important to understand WRWUDQVOLWHUDWH6L]ƈEROLRU6L]XEOLWKH$UDELFQDPHIRU how, in captioning it using Chinese ideograms, the Sozopol, the oldest towns on the southern Bulgarian cartographers may have transformed their originals ⌅ Black Sea coast (Konovalova 2006, p. 153). Fashi and, at least to a certain degree, drawn on Chinese ཡ is most likely a copyist’s error for the very similar- cartographic traditions that preserved other evidence. ᙟཡ looking Qieshi . This suggests that Qieshi was The maps in the 16th-century Gang yutu seem clearly intended to transliterate Karši, the Turkic name for to have been such a source for the revised and fullest .HUFK6XJL\DPD SɊ LGHQWLÀHG6XGDOL version of the Kangnido. A full discussion of such as the Russian city of Suzdal’, which is so remote matters requires separate treatment, but it is to be from the other locations here as to be improbable. hoped that this article will contribute to such a study. Very likely, the character li 䟼in this name must have been brushed in error, where the cartographer must ABOUT THE AUTHOR have intended to write the similar looking character, Dr. Nurlan Kenzheakhmet is a visiting professor in the hei 唁. Sudahei would make a plausible transcription Department of History, Philosophy and Religion, School IRU6XGDT DOVRNQRZQDV6ROGDLD 4XVXQ .DUVƈQD of Humanities and Social Sciences, Nazarbayev University, Kherson or Korsun), formerly called Cherson Trachea Astana, Kazakhstan. He received his Ph.D. from Peking (outside present day Sevastopol), is to be distinguished University, has held a Japan Invited Research Fellowship in from another Kherson on the Dnieper River. The one Kyoto, and Humboldt and Henkel Foundation fellowships in the Crimea was taken by the Rus’ prince Vladimir in Germany. Within his main research specialty on the 116 historical geography of Central Asia, he has worked Harley and Woodward 1992 VSHFLÀFDOO\RQWKHJHRJUDSKLFDOOLWHUDWXUHDQGPDSVRIWKH J. B. Harley and David Woodward, eds. The History of Yuan and Ming dynasties and on Ming sources regarding Cartography. Vol. 2, Bk. 1. Cartography in the Traditional th th the Qazaq Khanate in the 15 –17 centuries. He has written Islamic and South Asian Societies. Chicago; London: Univ. of on the early Turkic monumental inscriptions from Inner Chicago Pr., 1992. Asia. E-mail: . Herrera-Casais 2008 REFERENCES 0yQLFD +HUUHUD&DVDLV ´7KH 1DXWLFDO $WODVHV RI ¶$Oū DO Aoyama 1938 6KDUDIūµSuhayl 8 (2008): 223-63. Aoyama Sadao 䶂ኡᇊ䳴. “Gendai no chizu ni tsuite” ݳԓ ֺXGƈG 1982 -ȃൠണȀǹǙǻ [On maps of the Yuan dynasty]. 7żKżJDNXKż ֺXGƈGDOࠣߥlam. “The Regions of the World”. A Persian Geog ᶡᯩᆖ๡ (Tokyo) 8 (1938): 1-49. UDSK\$+³$'Tr, and annot. by V[ladimir] Mi- DO%DOœGKXUL norsky. 2nd ed. Cambridge: Cambridge Univ. Pr., 1982 (repr. of 1970 ed.). $KPDGLEQ-DELUDO%DOœGKXULThe origins of the Islamic State: being a translation from the Arabic, accompanied with Ibn Ꮱauqal 1800 ƯIٿDQQRWDWLRQV JHRJUDSKLF DQG KLVWRULF QRWHV RI WKH .LWœE MuᒒDPPDG $Eƈ·O4œVLP ,EQ Ꮱauqal. .LWœE 0DVœOLN WDV IXWƈᒒDO%XOGœQRIDO,PDPDEXO$EEDV$KPDGLEQ-DELUDO Ibn ֺauqal. The Oriental Geography Ebn Haukal: an Arabian %DODGKXULE\3KLOLS.KƈULᏡitti. Vol. 1. New York: Columbia Traveller of the Tenth Century. Tr. by William Ouseley. University, 1916. London: Oriental Press, 1800. Ibn Baᒷᒷnjᒷa 1829 DO,GUūVū $EX ¶$EGDOOœK ,EQ %Dᒷᒷnjᒷa. 7KH 7UDYHOV RI ,EQ %DWƈWD. Tr. Muᒒammad Ibn-Muᒒammad al-Idrūsū. .LWœE QX]KDW DO and annot. by Samuel Lee. London: Oriental Translation PXåࡊWœT Iū¶KWLUœT DOœIœT *pRJUDSKLH '·eGULVL. Tr. and annot. Committee, 1829. by Pierre Amédée Jaubert, 2 vols. Paris: Imprimerie royale, Ibn Baᒷᒷnjᒷa 1958-2000 1836-40. _____. 7KH 7UDYHOV RI ,EQ %DWWXWD $'  Tr. and Juvaini 1958 annot. by H. A. R. Gibb. 5 vols. London: Hakluyt Society, ‘Ala-ad-Din ‘Ata-Malik Juvaini. [Tarikh-i Jahan-gusha] The 1958-2000. History of the World-Conqueror. Tr. John Andrew Boyle. 2 al-Būrnjnū 1910 vols. Manchester: Manchester University Press, 1958. Muᒒammad ibn Ahmad al-Būrnjnū. Alberuni’s India. An ac- Kauz 2013 count of the religion, philosophy, literature, geography, chronol- Ralph Kauz. “Some Notes on the Geographical and ogy, astronomy, customs, laws and astrology of India about A.D. Cartographical Impacts from Persia to China: The Persian 1030. Tr. and annot. by Edward C. Sachau. 2 vols. London: Gulf as Depicted in The Map of Integrated Regions and Kegan Paul, Trench, Trübner, 1910. Terrains and of Historical Countries and Capitals and Brooks 2009 The Map of Persia in the 1X]KDW DO4XOƈE of Hamdallah Michael E. Brooks. “Prester John: A Reexamination and 0XVWDZIūµ,Q(XUDVLDQ,QÁXHQFHVRQ, last accessed 4 October 2016. Nurlan Kenzheakhmet. “Karta Kannido 1402 g. – istochnik po istorii Eke Mongol ulus” [The Kangnido Map of 1402— Dictionary 1854 a source for the history of the Great ]. Dictionary of Greek and Roman Geography, Ed. William Smith. Zolotoordynskaia tsivilizatsiia (Kazan’) 4 (2011): 59-78. 2 vols. Boston: Little, Brown, 1854. Kenzheakhmet 2015 DO)LGœ _____. “Central Asian Place Names in the Kangnido.” Journal of Asian History 49/1-2 (2015): 141-60. ,VPœ¶ūO LEQ ¶$Oū $Eƈ DO)LGœގ al-ᏡDPDZū Géographie d’Aboulféda: texte arabe [.LWœE 7DTZūP DOEXOGœQ (Survey of Kenzheakhmet 2016 the countries)]. Ed. Joseph T. Reinaud and William McGu- _____. “Inner Asia in the Kangnido: Contexts for Early ckin de Slane. Paris: Imprimerie royale, 1840. Chinese Knowledge of Inner Asia.” In: International Confucian Forum – Exchanges and Mutual Learning among Fujii et al. 2007 Asian Civilizations: Beijing International Symposium, 2. Beijing, )XMLL -żML 6XJL\DPD 0DVD·DNL DQG .LQGD $NLKLUR HGV 2016: 229-62. ㎥ : ۿ H]X FKL]L JD NDWXUD VHNDL. བྷൠȃ㛆DLFKL QR VKż]żK' ണ࣭ൠണǡ䃎Ƞц⭼. [Portrait of the Earth: The World Talks ,EQ.KXUUDGœGKELK DERXWWKH3LFWRULDO0DS@.\żWR.\żWR'DLJDNX*DNXMXWVX $EX·O4œVLP¶8ED\G$OOœKE¶$EG$OOœK,EQ.KXUUDGœGKELK Shuppankai, 2007. .LWœE DO0DVœOLN ZD·OPDPœOLN [The Book of Itineraries

117 and Kingdoms]. Ed. Michael Jan de Goeje. Liber viarum et Mauro 1966 regnorum. Bibliotheca geographorum Arabicorum, vol. 6. Mauro, Fra. Il mappamondo di Fra Mauro. Tullia Gasparrini Leiden: Brill, 1889. Leporace and Roberto Almagià, eds. Rome: Instituto Konovalova 2006 SROLJUDÀFRGHOORVWDWR Irina G. Konovalova. $O,GUūVūRVWUDQDNKLQDURGDNK9RVWRFKQRL Miller 1926-1931 Evropy: tekst, perevod, kommentarii [Al-Idrisi on the countries Konrad Miller. Mappae Arabicae: Die Arabischen Welt- und and peoples of : text, translation and Länderkarten des 9.—13. Jahrhunderts in Arabischer Urschrift, commentarry]. Drevneishie istochniki po istorii Vostochnoi Lateinischer Transkription und Übertragung in Neuzeitliche Evropy. Moskva: Vostochnaia literatura, 2006. Kartenskizzen. 5 vols. Stuttgart, 1926-1931. Ledyard 1994 Miller 1986 Gari Ledyard. “Cartography in Korea”. Ch. 10 in: J. B. _____. Mappae Arabicae. Beihefte. Ed. Heinz Gaube. Beihefte Harley; David Woodward, eds. The History of Cartography. zum Tubinger Atlas des Vorderen Oriients. Reihe B Vol. 2, Bk. 2. Cartography in the Traditional East and Southeast (Geisteswissenschaften) Nr. 65. Wiesbaden: Reichert, 1986. Asian Societies. Chicago; London: Univ. of Chicago Pr., 1994: Ming shilu 1964-67 235-345. Ming shilu ᰾ሖ䤴 [The veritable records of the Ming]. Taipei: Leo Africanus 1896 Zhongyang yanjiuyuan lishi yuyan yanjiusuo, 1964-67. Leo Africanus. The History and Description of Africa. Tr. John Miya 2006 Pory; ed. Robert Brown 2 vols. London: Hakluyt Society, 1896. Miya Noriko ᇞ㌰ᆀ ´.RQ·LWVX .\żUL 5HNLGDL .RNXWR QR Zu” he no michi Nj␧а⮶⨶↤ԓഭ䜭ѻണnjȍȃ䚃 [An Levtzion 1968 approach to the ‘Map of Integrated Regions and Terrains Nehemia Levtzion. “Ibn-Hawqal, the Cheque, and and of Historical Countries and Capitals’]. In: Mongoru jidai Awdaghost.” The Journal of African History 9/2 (1968): 223- no shuppan bunka ɪɻȼɳᱲԓȃࠪ⡸᮷ॆ [The publishing 33. culture of the Mongol period]. Nagoya: Nagoya daigaku Levtzion 2000 shuppankai, 2006: 487-651. _____. “Arab Geographers, the Nile, and the History of Bilad Miya 2007 al-Sudan”. In: The Nile: Histories, Cultures, Myths. Haggai _____. Mongoru teikoku ga unda sekaizu ɪɻȼɳᑍഭǡ⭏Ȩ Erlich and Israel Gershoni, eds. Boulder; London: Lynne ǵц⭼ണ [The world map that the Mongol empire created]. Rienner Publishers, 2000: 71-76. Tokyo: Nihon keizai shimbun shuppansha, 2007. Levtzion and Hopkins 1981 Miåk 1916 _____, and J. F. P. Hopkins, tr. and ed. Corpus of Early Arabic Hans v. Miåk, ed. and tr. Afrika nach der arabischen Bearbeitung DPPDG LEQ 0ƈVœ DO٭Sources for West African History. Fontes Historiae Africanae. der des Claudius Ptolemaeus von Mu .Series Arabica iv. Cambridge, etc.: Cambridge Univ. Pr., ֽZœUL]Pū. Wien: Alfred Hölder, 1916 1981. DO0XTDGGDVū san٭Lightfoot and Miller 1996 Muᒒammad b. AᒒPDG6KDPVDO'ūQDO0XTDGGDVūA Dale R. Lightfoot and James A. Miller. “Sijilmassa: The rise DOWDTœVūPIūPDޏULIDWDOœTœOūP [The best divisions for knowl- and fall of a walled oasis in Medieval Morocco.” Annals of edge of the (regions)]. Ed. Michael J. de Goeje. Bib- the Association of American Geographers, 86/1 (1996): 78-101. liotheca geographorum arabicorum, 3, Leiden: Brill, 1976 (reprint of 1877 ed.). Liu 1975 Liu Shu ࢹ᫙ et al. eds. Jiu Tang shu 㠺ୀᴨ[The Old Book of DO0XTDGGDVū the Tang]. 16 vols. Beijing: Zhonghua shuju, 1975. _____. The Best Divisions for Knowledge of the Regions: a Trans- ODWLRQ RI $KVDQ DO7DTDVLP À 0Dޏrifat al-Aqalim. Tr. by Basil Liu 2010 Collins. Reading: Garnet, 2001 (1st ed. 1994). Liu Yingsheng. “A Lingua Franca along the Silk Road: Persian Language in China between the 14th and the 16th Needham et al. 1971 Centuries.” In: Aspects of the Maritime Silk Road: From the Joseph Needham, Wang Ling, and Lu Gwei-djen, eds. Persian Gulf to the East China Sea. Ralph Kauz, ed. Wiesbaden: Science and Civilisation in China, vol. IV: Physics and Physical Harrassowitz, 2010: 87-96. Technology, pt. 3: Civil Engineering and Nautics. Cambridge: Cambridge Univ. Pr., 1971. Luo 1969 Luo Hongxian 㖵⍚ݸ. Guang yutu ᔓ䕯െ [Enlarged Terres- Ogawa 1928-29 trial Atlas].Taibei: Xuehai, 1969 (reprint of 1579 ed.). Ogawa Takuji ሿᐍ⩒⋫. 6KLQDUHNLVKLFKLULNHQN\ƈ ᭟䛓↧ਢ ൠ⨶⹿ウ [Studies in Chinese historical geography]. 2 vols. Ma 1970 .\żWR.ǀbundǀ shobǀ, 1928-29. Ma Huan 傜⅒. Ying-yai Sheng-lan: “The Overall Survey of the Ocean’s Shores” [♋⏟㜌㿸].Tr. and ed. by Feng Ch’eng- DO4D]ZūQū Chün; intr., notes and appendices by J. V. G. Mills. Cam- ᏡDPGDOOœK 0XVWDZIū DO4D]ZūQū Geographical Part of the bridge: Hakluyt Society; Cambridge Univ. Pr.,1970. ׋X]KDWDO4XOƈEFRPSRVHGE\+DPG$OODK0XVWDZÀRI4D]ZLQ 118 iQ  . Ed. and tr. by Guy Le Strange. 2 vols. Leyden: Wada 1958 Brill, 1915-19. :DGD +LURQRUL ´0HLGDL QR WHSSż GHQUDL WR 2VXPDQ Rapoport and Savage-Smith 2014 teikoku: jingi fu to seiiki tochi jinbutsu ryaku” ᰾ԓȃ䡴 ⹢ՍᶕǽȲɁɦɻᑍഭ — ⾎ಘ䆌ǽ㾯ฏ൏ൠӪ⢙⮕ [Ming Yossef Rapoport and Emilie Savage-Smith, eds. An Eleventh- dynasty interaction with the — Sacred Century Egyptian Guide to the Universe: The Book of Curiosities. treasure and people of the Western Regions: A brief survey]. Leiden; Boston: Brill, 2014. Shigaku ਢᆨ 31: 1-4 (1958): 692-719. Rashūd al-Dūn 1998 Wade 2010 Rashūd al-Dūn ᐅabūb.ࡃ -DPL¶X·WWDZDULNK &RPSHQGLXP RI *HRII :DGH ´¶%DOD[L· DQG WKH 3œUVLV GXULQJ WKH 0LQJ Chronicles. Tr. and annot. by Wheeler M. Thackston. Dynasty: A Note.” In: Aspects of the Maritime Silk Road: [Cambridge, MA]: Harvard University, Deparment of Near From the Persian Gulf to the East China Sea. Ed. Ralph Kauz. Eastern Languages and Civilizations, 1998. Wiesbaden: Harrassowitz, 2010:173-79. Robinson 2008 Wang et al. 1994 Kenneth R. Robinson. “Mapping Japan in a Foreign Country Wang Qianjin ⊚ࡽ䘋, Hu Qisong 㜑੟ᶮ, and Liu Ruofang in the Fifteenth and Sixteenth Centuries”. In: Higashi ajia ࡈ㤕㣣. “Juanben caihui daming hunyi yitu yanjiu” 㔒ᵜ NDLLNL VKL NHQN\ƈ QL RNHUX VKLU\ż QR KDNNXWVX WR VDL NDLVKDNX ᖙ㔈བྷ᰾␧аമ⹄ウ [As regards the Da Ming Hun Yi Tu ᶡȪɀȪ⎧ฏਢ⹄ウȀ NRFKL]X JLVKL VKLU\ż EXQJDNX K\żJHQ (Amalgamated map of the Great Ming Dynasty) drawn ⨭ਢᯉ࣭᮷ᆖ㺘֯ڭǟǦȠਢᯉȃⲪᧈǽ޽䀓䟸 ਔൠണ࣭ - -. in colours on stiff silk]. In: Cao Wanru ᴩၹྲ et al., eds. Nagoya: Nagoya daigaku, 2008: 51-74. Zhongguo gudai ditu ji: Ming dai ѝഭਔԓൠമ䳶: ᰾ԓ [An Robinson 2010 Atlas of Ancient Maps in China—the Ming Dynasty]. _____. “Daoist Geographies in Three Korean World Maps.” Beijing: Wenwu chubanshe, 1994. Journal of Daoist Studies 2010/3: 91-116. Wheatley 2001 Sugiyama 2007 Paul Wheatley. The Places Where Men Pray Together: Cities in 6XJL\DPD0DVDNL´7ż]DLQRVHNDL]XJDNDWDUXMLQUXLVDLVKR Islamic Lands Seventh through the Tenth Centuries&KLFDJR  no daichihei” ᶡ㾯ȃц⭼ണǡ䃎ȠӪ于ᴰࡍȃབྷൠᒣ [The London: The University of Chicago Press, 2001. ÀUVWPDMRUKRUL]RQRIKXPDQNLQGUHSUHVHQWHGLQWKHPDSV Yao 1959 RIWKHZRUOGRI(DVWDQG:HVW@,Q)XMLL-żMLHWDO Yao Guangxiao ညᔓᆍ. Yongle dadian ≨倈བྷި. [Yongle en- Takakashi 1963-75 cyclopedia]. Beijing: Zhonghua shuju, 1959. Takakashi Tadashi 儈⁻↓ ´7ż]HQVHUX LVUœPX VHNDL]X³ al-Ya‘qubi 2011 VKXWRVKLWHNRQ·LWVXN\żULUHNLGDLNRNXWRQR]XQLWVXLWHᶡ╨ǰ ȠѝцȬɁɎόɨц⭼ണɔɔѫǽǬǻ␧а⮶⨶↧ԓ഻䜭ѻ Ahmad al-Ya‘qubi. Kitab al-Buldan [Book of countries]. Ed., ണȀǹǙǻ” [The Islamic worldview of eastern expansion]. tr. and annot. by Lidiia M. Semenova. Moskva: RAN, 2011. 5\ƈNRNXGDLJDNXURQVKƈ 喽䉧བྷᆨ䄆䳶, 374 (1963): 77-95; 400- Zhang 1962 401 (1973): 586-600; 9 (1975): 14-31. Zhang Tingyu ᕥᔧ⦹. Ming Shi ᰾ਢ [The History of Ming]. Tolmacheva 1969 Taibei: Guofang yanjiuyuan Ming shi biancuan weiyuanhui, 1962. Marina A Tolmacheva. “Vostochnoe poberezh’e Afriki v DUDEVNRL JHRJUDÀFKHVNRL OLWHUDWXUHµ >7KH HDVWHUQ OLWWRUDO Zhao 1996 of Africa in Arab geographic literature]. In: Strany i narody Zhao Rugua 䏉⊍䘲. Zhufan zhi jiaoshi 䄨㭳ᘇṑ䟺 Vostoka. Vyp. IX. Strany i narody Afriki. Moskva: Nauka, [Description of the foreign lands]. Ed. and annot. by Yang Glav. red. vostochnoi lit-ry, 1969: 268-97. Bowen ὺঊ᮷. Beijing: Zhonghua shuju, 1996. Tolmacheva 1987-88 _____. “The African Waq-Waq: Some Questions Regarding NOTES the Evidence.” Bulletin d’information. Union academique 1. In Chinese , the title would read Hunyi jiangli lidai internationale. Commission XXII. Fontes Historiae Africanae guodu zhitu. For a treatment of the Kangnido in the context 11/12 (1987-88): 9-15. of the history of Korean cartography, see Ledyard 1994, pp. Tolmacheva 2015 244-49, 265-67. On the different copies and later maps of the Kangnido, see Fujii et al. 2007, pp. 56; 448-54. Sugiyama _____. “The Indian Ocean in Arab Geography. Transmission Masa’aki (2007) discusses the Kangnido in Ibid., 54-83. Also of Knowledge between Formal and Informal Geographic see Robinson 2010, p. 92. Traditions.” Terra Brasilis, N. S., 6 (2015): 2-10. -DSDQHVHVFKRODUVKDYHPDGHVLJQLÀFDQWFRQWULEXWLRQVWR Tomaschek 1886 the study of Kangnido. The Japanese historical geographer Wilhelm Tomaschek. “Zur Kunde der Hämus-Halbinsel. 2JDZD 7DNXML   SS   ÀUVW EURXJKW WKH II. Die Handelswege im 12. Jahrhundert nach den 5\ƈNRNX .DQJQLGR WR VFKRODUO\ DWWHQWLRQ $IWHU 2JDZD Erkundigungen des Arabers Idrîsî.” Sitzungsberichte der see the careful study of the place names in the Kangnido by Philosophisch-Historischen Klasse der Kaiserlichen Akademie der Aoyama 1938. See also Wada 1958; Takakashi 1963-75; Miya Wissenschaften 113 (1886): 285–373. 2006; Miya 2007, p. 14. 119 3. Cf. Sugiyama 2007, p. 56, who regards the comparison is located near the ancient Egyptian city of Abydos. In the with the Catalan Atlas more grounded in history than the 5\ƈNRNX.DQJQLGRWKLVQDPHLVWUDQVFULEHGDV$ODVDL\L䱯 comparison with the Fra Mauro map. ࡼ䌭Ժ6XJL\DPD S LGHQWLÀHGLWZLWK$OH[DQGUū 4. Chinese text: 䅩ԔᐕӪ⮛þ⎧ޗ㨟ཧെÿа䔨ˈ  $OWKRXJK QRW RQ WKH 5\ƈNRNX .DQJQLGR WKH +RQNżML ᔓйиˈᗎйийቪˈ⦷ԕᣈᡀⲮ䟼(Liu 1975, 12: Kangnido shows the Chinese transcription Zhebulu hama 䙉 138:3787). н冟૸哫ZKLFK7DNDNDVKLLGHQWLÀHVZLWKWKH3HUVLDQL]HG Arabic name Djebel alqamar, ‘Mountains of the Moon’. See 5. On the problematic treatment of the Indian Ocean region Rapoport and Savage-Smith 2014, p. 494. and especially the Persian Gulf on this map and possible Is- lamic sources for it, see Kauz 2013. 10. About Da Ming hunyi tu and Guang yu tu, see Wang et al. 1994, pp. 51-55. 6. In the Ming shilu called Xigandaliye 㾯ᒢ䚄ଙ㙦 (Iskandi- riya). See Ming shilu 1964-67, Ming Taizong shilu, juan 56: 11. For a review of some of the issues requiring a new study [Yong-le: Year 4, Month 7, Day 29 (12 Aug 1406)] 㠺⑟九 about the location of :œTZœT, see Tolmacheva 1987-88. ⴞ䲣⾆㗙䚓ᆀ༛㢟ằ䚃᰾䚓ယ㿰 ᭯৺㾯ᒢ䚄ଙ㙦എഎ 12. See, e.g., the discussion in Tolmacheva 1969, but with ૸ⴤ俜ਸ唈ㅹֶᵍ䌌䡄ᒓᴹᐞ: “[Chen] Shi-liang, who LPSRUWDQW TXDOLÀFDWLRQV EDVHG RQ PRUH UHFHQW DQDO\VLV RI had been sent by his father Chen Zu-yi, a chieftain of Old evidence not from ”academic” geographical treatises but Port, and [Liang] Guan-zheng, who had been sent by his rather from actual travel itineraries (Tolmacheva 2015). uncle Liang Dao-ming, also a chieftain of Old Port, as well as the Muslim Ha-zhi Ma-he-mo from Xi-gan-da-li-ye, came  2WKHUVFKRODUVLGHQWLÀHG%DOD[LZLWK%UD]LORUZLWKWKH to Court. Paper money and silks, as appropriate, were con- HWKQRQ\P3œUVLV3DUVL6HH:DGH ᒤ䚓֯㠓ޝferred upon them.” 14. Chinese text: ᐤஷ㾯৫ѝ഻㎅䚐↓ᗧ 䋒䀰ަ഻൘ই⎧࿻ཹ⦻ભֶᵍ㡏㹼ഋᒤॺޕൠⲭ⋊ 7. Alternatively, the name XuntaLVVDLGWRUHÁHFW$UDELF IskandiriyaZKLFKZRXOGFRQÀUPWKHSURQXQFLDWLRQRIRXU (Zhang 1962, juan 325). name as Skinda.  0DSPDNHUV RI WKH +RQNżML .DQJQLGR ZULWH WKH 㤖 㤕 Ԇࡼ㤕䛓 Ԇࡼ㤖䛓 8. Alternatively, assuming a northern route, it could be the character ku as ruo . For example, city of $EVKœ\D, which appears in the world of earlier Ara- (Tarragona). 6HH6XJL\DPDS7DEOHɊ ELFJHRJUDSKHUV ,EQ.KXUUDGœGKELKSDO

120 121 122 123 124 Acknowledgement This article was written within the framework of a research project supported by the Humboldt Research Foundation.

125 Plate VII Kenzheakhmet, “.DQJQLGR,” p. 106. ᵜݹ LQWKHFLW\ 7KH+RQNżML .DQJQLGR NHSWLQWKH +RQNżML Temple ሪ RI6KLPDEDUD 1DJDVDNL SUHIHFWXUH

3LFWXUHVRXUFH KWWSVXSORDGZLNLPHGLDRUJZLNLSHGLDFRPPRQVG*HQHUDO0DS2I'LVWDQFHV$QG+LVWRULF&DSLWDOVMSJ!