Identifying the Country of Meilugudun and the Significant Value of Zhou Qufei's Lingwai Daida

Ming Qing Yanjiu 21 (2017) 1–43

brill.com/mqyj

Identifying the Country of Meilugudun and the Significant Value of Zhou Qufei’s Lingwai daida

Victoria Almonte Tuscia University [email protected]

Abstract

During the last century, considerable interest arose regarding Chinese knowledge of western territories, with a long list of works being published on the topic. Joseph Needham’s Science and Civilisation in China (1959) states that Arab thinking had clearly influenced the Chinese conception of geography over the centuries. Zhang Xinglang analyses the relationship between the Chinese empire and countries overseas, focusing on Islamic countries and particularly those in the north of Africa. Feng Chengjun’s western territories toponyms and Gudai nanhai diming huishi have provided two pow- erful and even fundamental tools for the research presented here. The first gathers together a large collection of toponyms from various literary works; these are written in western language with their relative transcription or translation in Chinese. The second, the Gudai nanhai diming huishi, is divided in two volumes analysing many Chinese toponyms and their use in several geographical works. Li Qingxin’s Haishang Sichou zhilu, focuses on the development of the Maritime Silk Road and its economic-political consequences for China’s empire. Gabriele Foccardi’s research focuses instead on the motives for Chinese travellers and their expeditions, highlighting the historical and social differences between the different dynasties. Friedrich Hirth and William Rockhill provide a crucial literary resource with their translation of Zhao Rukuo’s work, Zhufanzhi (1966), as does J.V.G. Mills with his annotated translation of Ma Huan’s Yingya shenglan, a journey work of the fifteenth century. Yang Wuquan’s research into Zhou Qufei’s work, published in 1999, identifies several toponyms used by Zhou and compares several foreign geographical works. Zhou Qufei and Zhao Rukuo were both imperial officials during the Southern Song dynasty. They spent many years in the border territories of China: Zhou Qufei in Qinzhou, Guangxi province, and Zhao Rukuo in Quanzhou, Fujian province. Their works mention several toponyms never used before in Chinese texts: ‘Meilugudun’, or ‘Meilugu’ (as written by Zhao Rukuo), is one of these. The identification of this

© koninklijke brill nv, leiden, 2018 | doi 10.1163/24684791-12340012Downloaded from Brill.com09/30/2021 07:52:51AM via free access 2 Almonte toponym has not been determined until now. Therefore, the aim of this paper is to determine which kingdom was identified with the ‘Meilugudun’ toponym during the Song Dynasty. Two different questions are here discussed and resolved. First, can the land of Meilugudun be identified with the city of Merv in Turkmenistan? Second, do Zhou’s ‘Meilugudun’ and Zhao’s ‘Meilugu’ both stand for the same place? This paper can be divided into four sections. The first section focuses on Zhou Qufei, the second on Zhao Rukuo. The third analyses and compares previous scholars’ studies. The fourth proposes the new identification of the Meilugudun kingdom.

Keywords

Song dynasty – Chinese geographical texts – Zhou Qufei – Zhao Rukuo – Meilugudun

Introduction

This work belongs to the large field of studies of Chinese geographical works from the Song dynasty. Two Song works are analysed here: Lingwai daida (Notes from the Land Beyond the Passes1 or Various Replies from Beyond the Passes)2 written by Zhou Qufei in 1178, and Zhufanzhi (Description of the Foreign Lands or Description of Barbarous Peoples)3 written by Zhao Rukuo in 1225. Scholars can read and study these types of work only because of Ming and Qing dynasty copyists. Zhou Qufei’s Lingwai daida wasn’t printed until the twentieth century, but it was copied by Qing scholars and included in the en- cyclopedic work Siku quanshu 四庫全書, with an abstract dated 1781 (or the forty-sixth year of the Qianlong emperor). Lingwai daida is placed in historical section number 11 (Shibu shiyi 史部十一), geographical category number 8 (Dili shuba 地理類八), miscellaneous (Zaji 雜記). This paper takes that version as its point of departure.4 As you will read in the abstract below, Qing copyists

1 As has been translated by Hyunhee Park. Please see Park 2012: 46. 2 Henceforth Daida. 3 As translated by Hirth and Rockhill in 1911. See Hirth and Rockhill 1966: preface. 4 The author has analysed in detail the electronic version of Siku quanshu, distributed by Chinese University Press with Digital Heritage Publishing, Hong Kong, 1998. But the most reliable version of Lingwai daida is also the newest, that annotated by Yang Wuquan 杨武泉 and published in 1999. He not only analysed Zhou’s work, but also compared it with other geographical works, trying to identify place names. Please see Yang Wuquan 杨武泉, Lingwai daida jiaozhu 岭外代答校注, Zhonghua Shuju 中华书局, 1999. It is volume sixteen of the series Zhongwai jiaotongshi shuji congkan 中外交通史书籍丛刊.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 3 duplicated Zhou’s work from Yongle dadian 永樂大典. In that Ming dynasty version, Daida was divided into only two volumes; according to Qing scholars, Song dynasty bibliographical treatises were divided into ten volumes, as the original work seems to have been. Siku quanshu’s preface reads as follows:

嶺外代答十巻宋周去非撰。去非字直夫，永嘉人，隆興癸未進士。淳熙中官桂林通判。是書即作於桂林代歸之後自序謂。本范成大桂海虞衡志而益以耳目所見聞。録存二百九十四條。盖因有問嶺外事者倦於應酬。書此示之，故曰代答。原本分二十門。今有標題者凡十九門，一門存其子目而佚。其總綱所言則軍制戸籍之事也。[。。。] 書録解題及宋史藝文志並作十巻。永樂大典所載併為二巻。蓋非其舊今從原目仍析為十巻云。乾隆四十六年九月恭校上總纂官臣紀昀、臣陸錫熊、臣孫士毅。總校官臣陸費墀。

Lingwai daida written by Zhou Qufei during the Song dynasty. Qufei’s patronymic name was Zhifu and he was from Yongjia; during the 20th year of Longxing he passed the imperial examination. During Chunxi period, he was an official in Guilin, as an assistant prefectural magistrate. As you will read in the preface, Lingwai daida was written after he returned from Guilin. He was influenced by Fan Chengda’s work, Guihai yuheng zhi, and his memories were taken from what he saw and heard during his work. Lingwai daida is composed of 294 sections. Dealing with people and answering their questions about the Lingwai territories were wearing him out. As this book represents Qufei’s attempt to answer people’s questions, it was named daida. The original copy was divided into 20 chapters, but now there are only 19, as one was lost—it probably dealt with the military system. […] Shulu jieti,5 and other Song dynasty bibliographical treatises state it was divided into 10 volumes. However, Yongle dadian was divided into only two volumes. In order to respect the original composition, it is divided here into 10 volumes. The ninth month of the 46th year of Qianlong, edited by officials Ji Yun, Lu Xixiong, Sun Shiyi. Checked by official Lu Feichi.

The author has conducted below an analysis that takes the Qing version into consideration and focuses on Zhou’s geographical chapters.

5 Zhizhai Shulu jieti was a bibliographical work written by Chen Zhensun during the Song dynasty.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 4 Almonte

In the last century there arose considerable interest in Chinese knowledge of the western territories, with several works pertinent to the topic being published. Joseph Needham’s chief work, Science and Civilisation in China (1959), includes a volume entitled Mathematics and the Science of Heavens and the Earth. In this work he describes several geographical discoveries and expeditions accomplished by Chinese travellers during the previous centuries. He states that China’s geographical concept was clearly influenced by the thinking then prevalent in the Arab world.6 Zhang Xinglang analysed the relationship between China’s empire and overseas countries, focusing on Islamic countries and particularly those in the north of Africa. In his work Zhongxi jiaotong shiliao huibian, he analysed each Chinese dynasty and its respective commercial itineraries, with particular attention to the linguistic, geographical, historical, and political aspects of several foreign places, including Iran and lands in Central Asia. The high value of his work lies in the deep analysis of many toponyms developed during different centuries. His work is divided into four volumes, and the third volume focuses on ancient Chinese contacts with Arab countries. He first analysed different Chinese geographical works, then considered Arabic sources. For example, he collected every section in which the toponym Dashi (indicating Arab countries) appeared, highlighting parallels and differences; this was in such Chinese works as dynastic histories, Jia Dan’s writing, Du Huan’s work, Lingwai daida and Zhufanzhi. Then he closely studied Arab geographical works—such as those by Sulaiman, Al Masudi, Al Biruni, Idrisi, and Abulfeda—and how they described China’s empire.7 Feng Chengjun’s Xiyu diming and Chen Jiarong’s Gudai Nanhai diming huishi could be considered two influential and fundamental tools for present research. The first contains a large collection of toponyms drawn from various literary works; the toponyms are all written in western characters and are accompanied by a transcription or translation in Chinese. In this work it is remarkable to see the evolution of geographical lexical terms and the signs of po- tential influence from foreign languages. For example, in treating the toponym ‘Simhala’ (which originated in Sanskrit and indicates Sri Lanka), he considers many Chinese sources that used different toponyms, all of them linked to the Sanskrit name or (as Feng suggests) the Arabic transcription, ‘Silan’.8 Chen’s Gudai Nanhai diming huishi is divided into two volumes, where many Chinese

6 Needham 1959: 512. 7 Zhang Xinglang 2003, 3: 677–948. 8 See Feng Chengjun 1980: 85–86.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 5 toponyms and their use in several geographical works are analysed. Regarding ‘Mulanpi’ 木兰皮, it finds that Lingwai daida was the earliest work in which this toponym was used, stating that ‘Mulanpi’ was a Murabit transcription—in the eleventh century it indicated the al-Maghreb territories. Gudai Nanhai diming huishi then cites other Chinese works in which ‘Mulanpi’ appears.9 Li Qingxin’s Haishang sichou zhilu focused on the development of the Maritime Silk Road and its economic-political effects on China’s empire.10 Gabriele Foccardi’s studies focused instead on the motivations that pushed the Chinese travellers on their expeditions, highlighting the historical and social differences between the different dynasties.11 Hirth and Rockhill’s work, the unique English translation of Zhao Rukuo’s Zhufanzhi, must also be considered a crucial literary resource,12 as is Mills’ annotated translation of Ma Huan’s Yingya shenglan, a work chronicling the fifteenth century.13 In Yang Wuquan’s research on Zhou Qufei’s work, published in 1999, the author identified several toponyms used by Zhou and compared several foreign geographical works.14 Zhou Qufei and Zhao Rukuo were both imperial officials during the Southern Song dynasty. They spent many years in the border territories of China: Zhou Qufei in Qinzhou, Guangxi province, and Zhao Rukuo in Quanzhou, Fujian province. They mentioned several toponyms never used before in Chinese texts: ‘Meilugudun’ 眉路骨惇, or ‘Meilugu’ 眉路骨 (as written by Zhao Rukuo), is one of these. In particular, its identification has not been determined until now. Therefore, the objective of this paper is to determine which territory was identified with the Meilugudun toponym during the Song dynasty. Two different questions are here discussed and resolved: if ‘Meilugudun’, used by Zhou, and ‘Meilugu’, used by Zhao, represent the same country, can they be identified with the city of Merv in Turkmenistan? Is it possible Zhou was influenced by Arab geographical knowledge? This paper is divided into four sections. The first section focuses on Zhou Qufei, the second on Zhao Rukuo. The third analyses and compares previous scholars’ studies. The fourth proposes a new identification of Meilugudun and

9 Chen Jiarong, Xie Fang, Lu Junling 1986, 1: 192. 10 Li Qingxin 2006: 20. 11 Foccardi 1992: 32. 12 Hirth and Rockhill 1966: 1–39. 13 Mills 1970: 1–25. 14 Ibid.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 6 Almonte offers an explanation of why Meilugudun can be identified as the city of Merv in Turkmenistan during the Song dynasty.

1 Zhou Qufei

1.1 Zhou Qufei and His Life Zhou Qufei, also known as Zhifu 直夫, was born in Yongjia, Wenzhou province. There is little known about his life, but his birth was probably in 1133, 1134, or 1135, and his death in 1187, 1188, or 1189.15 He grew up in a familiar literary context: his grandfather Zhou Xingji 周行己 (courtesy name Gongshu 恭叔), was the main representative of the Northern Song School of Thought, called Yuanfeng jiu xiansheng 元丰九先生 (the nine scholars from Yuanfeng); his nephew Zhou Duanchao 周端朝 was one of six scholars at Imperial College during Ningzong’s 宁宗 reign. At the age of thirty, in 1163, Zhou received a Jinshi degree (Metropolitan Graduate) and became an imperial official. However, he did not have a successful career. He was a serious official and worked very hard during his service, but he died young and his prospects for advancement had been limited. The main sources about his life and work as a government official are his friends’ and colleagues’ writings. The Qing dynasty historian Wan Sitong16 described Zhou Qufei as one of the main disciples of Zhang Shi 张栻.17 At the Four Virtuous Men Temple (Si xian dian 四贤祠) in Guilin, there is an inscrip- tion where Zhou Qufei’s name appears as one of the thirty-three brightest disciples of the four ancient scholars (Zhou Dunyi, Li Yanping, Luo Qingshun, and Chen Baisha) that the temple is devoted to.18

15 Yang Wuquan 1999: preface part 1. For biographical information, see Yang Wuquan 1994: 80–84, Ling Yan 1998: 51–56, and Huang Quancai and others 1996: 76–82. 16 Wan states, “Zhang shi menren 张氏门人”. Wan Sitong (1638–1702) was a Qing dynasty historian and also wrote the sixteen-chapter Confucianism history Rulin zongpai and the Ming dynasty history Mingshigao 明史稿, which is divided into five hundred chapters. Wan Sitong 1935, 11: 7. 17 Zhang Shi (1133–1180) was the main scholar of Huxiang xuepai 湖湘学派 (Philosophical School of Hunan); he can be considered one of the three most important Neoconfucianist scholars, with Zhu Xi and Lu Zuqian. He wrote Nanxuan Anthology, Nanxuanji 南轩集. Yang Wuquan 1999: preface 6. 18 Zhejiang tongzhi, Wenzhoufu-Yongjiaxian 1991, 225: 3864.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 7

Lou Yao 楼钥,19 Yang Wanli 杨万里,20 and Fan Chengda 范成大21 talked about Zhou’s excellent service as an official.22

1.2 Lingwai daida: Historical Background, Structure, and Purpose Zhou Qufei wrote Lingwai daida in 1178, during the Southern Song dynasty (1127–1279). The Song had lost the northern half of its territory to the Jurchen Jin dynasty. The Jingkang zhichi or zhibian 靖康之耻 or 之变 (also known as Jingkang zhihuo 靖康之禍), with Jingkang being the name of the year, chron- icles the Song court’s retreat south of the Yangtze River and the establishing of its capital at Lin’an (now Hangzhou). Although the Song dynasty had lost control of the traditional “birthplace of Chinese civilization” along the Yellow River, the Song economy was still strong, as the Southern Song’s empire con- tained a large population and productive agricultural land. The empire also took in the ocean, as the Southern Song bolstered its naval strength; its ships defended the empire’s water and land borders and conducted missions to more than forty foreign countries. The Silk Road was now less important than the Maritime Silk Road. Stretching over the sea from southeast Asia to the far shores of the Indian Ocean, the water route tied China to Arab ports such as Aden. Along that route there travelled the products of various countries, and also geographical, linguistic, and historical information that was carried to remote areas. Of the Southern Song territories, Guangxi province was one of

19 Lou Yao 楼钥 (1137–1213) very likely was a close friend of Zhou Qufei. He was born in Yin county, in modern-day Ningbo of Zhejiang province. He passed the imperial examination in 1163 and began a career in office, one that included his trip in 1169 to the Jin dynasty territories as a member of a diplomatic mission. He was the author of Gongkui ji 攻媿集, a poetry collection 120 rolls in length. Zhang Jin 2004: 36–49. 20 Yang Wanli 杨万里 (1127–1206), was born in southern China’s Jishui county, in modern- day Jiangxi province. Yang was an important scholar, poet, and politician known for his strong resistance to the Jin dynasty. With Lu You 陆游, You Mao 尤袤, and Fan Chengda 范成大, he was one of the most important Southern Song scholars, known as Nan Song Si dajia 南宋四大家. See Hu Chuanzhi 2010: 62–69. In this work Yang’s 352 poems have been analysed. See also Shen Songjin 2006: 73–83. 21 Fan Chengda (1126–1193), also known as Zhineng 至能, was a Southern Song poet. Zhou Qufei mentioned him several times as Shihujushi 石湖居士. From 1173 he worked as an official in Jingjiangfu (modern-day Guilin), then in 1175 he moved to Chengdu in Sichuan province. He was the author of Guihai yuheng zhi 桂海虞衡志, a geographical treatise focusing primarily on the topography and commercial output of China’s southern provinces. See Miao Jing 1999: 3–9. See also Zhang Xuezhong and Yi Jun 2001: 121–25. 22 Fan Chengda said of Zhou Qufei “juan xiubasu (隽秀拔俗)”. In English this would be “he is a talented and outstanding writer”. Yang 1999: preface 6–7.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 8 Almonte the most active. A large number of imperial officials went to Guilin for work, which stimulated literary production. At the same time, contacts between foreign merchants, travellers, and locals became more and more frequent, and this proved fruitful for Chinese knowledge of the western territories. This is the background to Zhou’s Lingwai daida. He collected his notes in Guangxi province’s Qinzhou and Jingjiang district (Guilin territory), where he spent at least six years. During this period, he was also able to travel to Guilin, the main city of Guangxi. He gathered material and experiences regarding local customs and foreign traditions. Although he never travelled abroad, it is probable that he met Chinese merchants who travelled in the Song empire, or foreign merchants who lived in the Chinese territory. In several cases, Zhou states that a given piece of geographical information comes from travellers or interpreters. The “Geography” chapter, section 20 (“Xiangbisha 象鼻砂”, Xiangbi Desert), mentions a merchant/sailor (boshang 舶商) who provided information about itineraries leading to the east or west.23 Section 74, “Dialects”, refers to an interpreter who explained the difference between the Guangxi and Jiaozhi languages.24 Daida was a casual composition. Zhou didn’t have any intention of writing a geographical work; he was merely collecting a series of personal notes (suishi biji 随事笔记). But, as he states in Daida’s preface, he decided to answer once and for all the questions his friends kept asking about what he’d seen and heard in Guangxi province. He’d tried to answer their questions orally, but he soon realized that his material deserved a more accurate treatment. After he read Fan Chengda’s Guihai yuheng zhi 桂海虞衡志, he began writing Daida’s 294 sections. Zhou’s preface reads as follows:

盖甞(尝)随事筆記，得四百餘條。[。。。] 問以絶(绝)域事，驟莫知所對者，盖数数然。（然）至觸事而談，或能舉其一二。[。。。]晚得范石湖《桂海虞衡志》，又於藥裹得所抄名数，因次序之，凡二百九十四條。應酬倦矣，有復（僕）問，用以代答。[。。。].

23 Yang 1999: 37. The Chinese text is: 甞聞之舶商曰: “自廣州而東, 其海易行。自廣州而西，其海難行。自欽廉而西，則尤為難行。” In English: “I heard from a merchant traveler that from Guangzhou toward the East, navigation is very easy. From Guangzhou toward the West it is quite difficult. From Qinlian toward the West, it is more difficult.” 24 Yang 1999: 159–60. The Chinese text: 余又嘗令譯者以《禮部韻》按交阯語，字字有異，唯 “花” 字不須譯。又謂 “北” 為“朔”。In English: “An interpreter of Libuyun explained to me the difference from the Jiaozhi language. Every character is different, even the character ‘hua’ (flower) is not easy to translate. The character ‘bei’ (north) becomes ‘shuo’ (new moon).”

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 9

I collected notes about various items and topics, approximatively 400 or more. […] Someone asked me information regarding the very far territories, but orally I was not able to do a good job, I needed to organize these notes. Without organizing them, I can only explain a few things, not as much as I would like. […] Later I read Fan Shihu’s Guihai yuheng zhi, and from “medicine pocket”25 I took all my notes and counted them: there were 294. I got tired of dealing with people, so I decided to write this work in order to provide answers to all their questions. [This is precisely why he includes da 答 in the title.]

As the title implies, Zhou sought to answer questions he received about the territories beyond southeastern China—territories he called Ling (for ‘Lingnan territories’, today Guangdong and Guangxi provinces)—and foreign lands. As shown in the table below, Lingwai daida’s structure is very complex. It includes ten volumes and twenty-one chapters about various topics, orga- nized into 294 sections.26 Every volume starts with a new chapter and can include four chapters. Chapter length varies from a mere six sections (as in Chapters 2, 7, and 13) to all of forty-five (as in Chapter 16). Section length varies too: some have few characters, others a thousand or so.

Table 1 Lingwai daida structure

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

1 Dilimen 22 条: 地理門 1. Baiyue Gudi 百粤故地、 (Geography) 2. Bingbian 並邉、 3. Guangxisheng Bingzhou 廣西省併州、 4. Wuling 五嶺、 5. Huguangzhushan 湖廣諸山、 6. Guishan 桂山、 7. Guilin Yandong 桂林巖洞、 8. Lingyan 靈巖、 9. Luocongyan 羅叢巖、 10. Limushan 黎母山、

25 A figurative phrase indicating messiness. 26 Juan 卷 has been translated as ‘volume’, men 门as ‘chapter’, tiao 条 as ‘section’. Zhang Hui 1993: 29.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 10 Almonte

Table 1 Lingwai daida structure (cont.)

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

11. Guangxi Shuijing 廣西水經、 12. Zangkejiang 䍧牱江、 13. Lingqu 靈渠、 14. Guishui 癸水、 15. Longmen 龍門、 16. Xiangshan 象山、 17. Tianweiyao 天威遥、 18. Tianfenyao 天分遥、 19. Sanheliu 三合流、 20. Xiangbisha 象鼻砂、 21. Tianya haijiao 天涯海角、 22. Hu 潮.

Bianshuaimen 6 条: 邉帥門 (Border 1. Guangxi jīnglüè Anfushi 廣西經略安撫使、 commanders) 2. Qiongzhou jian Guangxilu Anfu Doujian 瓊州兼廣西路安撫都監、 3. Yongzhou jian Guangxilu Anfu Doujian 邕州兼廣西路安撫都監、 4. Yizhou jian Guangxilu Bingma Doujian 宜州兼廣西路兵馬都監、 5. Rongzhou jian Guangxilu Bingma Doujian 融州兼廣西路兵馬都監、 6. Qin lian Xidong Douxun Jianshi 欽亷溪峒都廵檢使.

2 Waiguomen 10 条: 外國門 1. Annanguo 安南國、 (Shang 上) 2. Haiwai Liman 海外黎蠻、 (Foreign reigns 1) 3. Haiwai Zhufanguo 海外諸蕃國、 4. Zhangchengguo 占城國、 5. Zhenlaguo 真臘國、 6. Puganguo 蒲甘國、 7. Sanfoqiguo 三佛齊國、 8. Shepoguo 闍婆國、

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 11

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

9. Gulinguo 故臨國、 10. Zhunianguo 注輦國.

3 Waiguomen 14 条: 外國門 1. Daqinguo 大秦國、 (Xia 下) 2. Dashi Zhuguo 大食諸國、 (Foreign reigns 2) 3. Mulanpiguo 木蘭皮國、 4. Xitian Zhuguo 西天諸國、 5. Xitian Nannihualuoguo 西天南華囉罗國、 6. Dongnan Haishang Zhuzaguo 東南海上諸雜國、 7. Kunlun Cengqiguo 崑崙層期國、 8. Bosiguo 波斯國、 9. Danman 蜑蠻、 10. Sanfotuo 三佛䭾、 11. Yaoren 徭人、 12. Xinanyi 西南夷、 13. Tongdao Waiyi 通道外夷、 14. Hanghai Waiyi 航海外夷.

Bingmingmen 12 条: 兵民門a 1. Yan bian bing 沿邊兵、 (Soldiers) 2. Tuding Shubian 土丁戍邊、 3. Dongding Shubian 峒丁戍邊、 4. Tianzi jia 田子甲、 5. Dongding 峒丁、 6. Zhaiding 寨丁、 7. Tuding Baoding 土丁保丁、 8. Xiaoyong 效用、 9. Tuxuan 土宣、 10. Wumin 五民、 11. Duonong 惰農、 12. Sengdao 僧道.

4 Fengtumen 11 条: 風土門 1. Guangyou fengqi 廣右風氣、 (Local conditions) 2. Xuebao 雪雹、

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 12 Almonte

Table 1 Lingwai daida structure (cont.)

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

3. Zhangdi 瘴地、 4. Zhang 瘴、 5. Wushi 屋室、 6. Chaoju 巢居、 7. Tali 踏犁、 8. Zhuangtang 樁堂、 9. Songlao 送老、 10. Fangyan 方言、 11. Suzi 俗子字.

Fazhimen 6 条: 法制門 1. Zoubi 奏辟、 (Legal system and 2. Dingni 定擬、 institutions) 3. Shichang 試場、 4. Sheguan 攝官、 5. Nanhai Yifa 南海役法、 6. Changping 常平.

5 Caijimen 8 条: 財計門 1. Guangyou Caoji 廣右漕計、 (Wealth) 2. Guangxi Yanfa 廣西鹽法、 3. Jinglüe Simai ma 經略司買馬、 4. Yizhou maima 宜州買馬、 5. Magang 馬綱、 6. Yongzhou Hengshan Zhai Bo Yichang 邕州横山寨博易场、 7. Yongzhou yongping zhai bo yichang 邕州永平寨博易場、 8. Qinzhou Bo Yichang 欽州博易.

6 Qiyongmen 20 条: 器用門 1. Duanyan 端硯、 (Utensils) 2. Bi 筆、 3. Mo 墨、

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 13

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

4. Chaju 茶具、 5. Luobei 螺杯、 6. Yushan 羽扇、 7. Mandao 蠻刀、 8. Man jiazhou 蠻甲胄、 9. Man An 蠻鞍、 10. Man Nu 蠻弩、 11. Rong jian 融劒、 12. Li gong 黎弓、 13. Majian 藥箭、 14. Wuzhou tieqi 梧州鐵器、 15. Mulanzhou 木蘭舟、 16. Tengzhou 藤舟、 17. Kumuzhou 刳木舟、 18. Duo 舵 (in alcune versioni è yi 柂) 、 19. Manli 蠻笠、 20. Pilv 皮履.

Fuyongmen 10 条: 服用門 1. Tian 緂、 (Clothes and 2. Bu 布、 articles for daily 3. Yaobanbu 徭斑布、 use) 4. Shuichou 水绸 (in some editions is shuixi 水細) 、 5. Shuzi 綀子、 6. Annanjuan 安南絹、 7. Tan 氊、 8. Jibei 吉贝、 9. Chongsi 蟲絲、 10. Poshan poqun 婆衫婆裙.

Shiyongmen 7 条: 食用門 1. Jiu 酒、 (Food) 2. Cha 茶、 3. Shibinglang 食槟榔、 4. Laozha 老鮓、

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 14 Almonte

Table 1 Lingwai daida structure (cont.)

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

5. Yiwei 異味、 6. Zhaisu 齋素、 7. Shishui jieshui 實水潔水.

7 Xiangmen 7 条: 香門 1. Chenshuixiang 沉水香、 (Aromatics) 2. Pengcaixiang 蓬菜香、 3. Zhegubanxiang 鹧鴣斑香、 4. Jianxiang 箋香、 5. Zhongxiang 眾香、 6. Linglingxiang 零陵香、 7. Fanzhizi 蕃梔子.

Yueqimen 6 条: 樂器門 1. Pingnanyue 平南樂、 (Musical 2. Yao yueqi 瑶樂器、 instruments) 3. Yaogu 腰鼓、 4. Tonggu 銅鼓、 5. Guilinnuo 桂林儺、 6. Baijin guyue 白巾鼓樂.

Baohuomen 7 条: 寳貨門 1. Zhuchi 珠池、 (Precious goods) 2. Shezhu 蛇珠、 3. Pichenxi 辟塵犀、 4. Hupo 琥珀、 5. Chequ 硨磲、 6. Longxian 龍涎、 7. Dabei 大貝.

Jinshimen 13 条: 金石門 1. Shengjin 生金、 (Metals and 2. Danshashuiyin 丹沙水銀、 stones) 3. Lianshuiyin 煉水銀、

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 15

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

4. Yinzhu 銀朱、 5. Tong 铜、 6. Tonglǜ 铜绿、 7. Qianfen 鉛粉、 8. Zhongru 鍾乳、 9. Huashi 滑石、 10. Shiyan 石鷰、 11. Shixie shixia 石蠏石蝦、 12. Shimei 石梅、 13. Shibai 石柏.

8 Huamumen 45 条: 花木門 1. Gui 桂、 (Flowers and 2. Rong 榕、 trees) 3. Shamu 沙木、 4. Yanzhimu 燕脂木、 5. Sileimu 思櫑木、 6. Binglang 檳榔、 7. Guanglang 桄榔、 8. Yezimu 椰子木、 9. Zhu 竹、 10. Lizhi Yuanyan 荔枝圎眼、 11. Hongyan caoguo 紅鹽草果、 12. Bajiao huixiang 八角茴香、 13. Yuganzi 餘甘子、 14. Shili 石栗、 15. Shaoli 杓栗、 16. Jiaozi 蕉子、 17. Wulan 烏欖、 18. Youzi 柚子、 19. Baizi 百子、 20. Teng 藤、 21. Huateng 花藤、 22. Danpingjiao 膽瓶蕉、 23. Shuijiao 水蕉、 24. Hongjiaohua 紅蕉花、

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 16 Almonte

Table 1 Lingwai daida structure (cont.)

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

25. Nanshancha 南山茶、 26. Suxinhua 素馨花、 27. Molihua 茉莉花、 28. Shiliuhua 石榴花、 29. Shijunzihua 史君子花、 30. Tianse furong hua 添色芙蓉花、 31. Doukouhua 豆蔻花、 32. Paohua 泡花、 33. Mantuoluohua 曼陀羅花、 34. Junahua 拘那花、 35. Shuixihua 水西花、 36. Guomeihua 裹梅花、 37. Yuxiuhua 玉脩花、 38. Yuehe 月禾、 39. Dahao 大蒿、 40. Douguancao 都管草、 41. Jucao 蛆草、 42. Tonggucao 銅鼓草、 43. Shifa 石髮、 44. Biancai 匾菜、 45. Humancao 胡蔓草.

9 Qinshoumen 38 条: 禽獸門 1. Xiang 象、 (Animals and 2. Hu 虎、 birds) 3. Tianma 天馬、 4. Manma 蠻馬、 5. Guoxiama 果下馬、 6. Manquan 蠻犬、 7. Yuan 猨、 8. Bailu 白鹿、 9. Wei 蜼、 10. Renxiong 人熊、 11. Shanzhu 山猪、 12. Huayang 花羊、

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 17

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

13. Mianyang 綿羊、 14. Dali 大狸、 15. Fengli 風狸、 16. Yangshu 仰䑕、 17. Xiangshu 香鼠、 18. Shishu 石鼠、 19. Shexiang 麝香、 20. Lanfu 懶婦、 21. Shanta 山獺、 22. Shanfenghuang 山鳯凰、 23. Kongque 孔雀、 24. Yingwu 鸚鵡、 25. Wufeng 烏鳯、 26. Qinjiliao 秦吉了、 27. Feicui 翡翠、 28. Yan 鴈、 29. Lingwu 靈鶻、 30. Guzao 骨噪、 31. Zhen 鴆、 32. Chunchong 春虫、 33. Chunzi 鶉子、 34. Douji 鬭雞、 35. Changmingji 長鳴雞、 36. Chaoji 潮雞、 37. Zhenji 枕雞、 38. Fanmaoji 翻毛雞.

10 Chongyumen 12 条: 蟲魚門 1. Ranshe 蚺蛇、 (Insects and fish) 2. Liumugui 六目龜、 3. Bidaimao 鼊蝳瑁、 4. Shan 蟺、 5. Xunhuangyu 鲟鳇鱼、 6. Jiayu 嘉魚、 7. Heyu 河鱼、 8. Zhuyu xiayu 竹魚鰕魚、

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 18 Almonte

Table 1 Lingwai daida structure (cont.)

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

9. Guijiadie 鬼蛺蝶、 10. Heijiadie 黑蛺蝶、 11. Tianxia 天蝦、 12. Yu 猒.

Gujimen 9 条: 古跡門 1. Shaoshi 韶石、 (Historic sites) 2. Qincheng 秦城、 3. Lǜzhujing 緑珠井、 4. Gufuzhou 古富州、 5. Tongzhu 銅柱、 6. Zhiqisi 陟屺寺、 7. Jiaozhi 交阯、 8. Dan’er 儋耳、 9. Bingjing huoshan 冰井火山.

Mansumen 16 条: 蠻俗門 1. Mansu 蠻俗、 (Man people’s 2. Liaosu 獠俗、 customs) 3. Ruliao 入寮、 4. Guajian 掛劒、 5. Xiumian 繡靣、 6. Biyin 鼻飲、 7. Feituo 飛駞、 8. Tayao 踏揺、 9. Kuansai 款塞、 10. Muqi 木契、 11. Dabeng 打甏、 12. Diya 抵鵶(鸦)、 13. Shiqi 十妻、 14. Juanban 捲伴、 15. Doubaima 鬭白馬、 16. Yingmaoniang 迎茅娘.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 19

VOLUME CHAPTER SECTION tiao 条 juan 卷 men 门

Zhiyimen 15 条: 志異門 1. Tianshen 天神、 (Miscellany) 2. Shengfo 聖佛、 3. Ningjianyi 甯諫議、 4. Wupopo 武婆婆、 5. Zhuanzhi Dawang 轉智大王、 6. Xinsheng 新聖、 7. Jibu 雞卜、 8. Maobu 茅卜、 9. Nanfa 南法、 10. Jiagui 家鬼、 11. Tiaosheng 挑生、 12. Gudu 蛊毒、 13. Wangliang 罔兩、 14. Liuzhou wugong 栁州蜈蚣、 15. Guilin Houyao 桂林猴妖.

Heji 合计 21 men 門 294 tiao 条 sections Total chapters a This chapter’s title was lost. According to Yang’s analysis, it could be titled Bingmin men, Soldiers. Yang Wuquan 1999: 130.

Daida’s structure is more detailed, with more extensive content than many other contemporary works, such as Guihai yuheng zhi by Fan Chengda. This has only thirteen chapters and about ten thousand characters.27 Lingwai daida is highly valued by scholars and can be considered an ency- clopedic work on Guangxi province during the Song dynasty.28 However, in

27 According to Zhang Hui’s studies, only a few Song works have the same structure of Daida. For a full discussion on this topic please see Zhang Hui 1993: 29. 28 For Daida’s contribution about Guangxi knowledge in the Song dynasty, see Ling Yan 1998: 55, Huang Quancai and others 1996: 101.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 20 Almonte this article we will focus on Zhou’s treatment of the country Meilugudun, in chapter 4 (“Waiguomen Xia”), section 40 (“Dashizhuguo”, Various Countries of the Arab Empire). As the following paragraphs show, the author first compares Zhou’s toponym and its description with Zhao Rukuo’s toponym and description; then, after consulting Arab sources, she determines the identification. Zhou’s work contains two chapters titled “Foreign Countries” (waiguo 外國”) and comprising twenty-four sections: ten in the first chapter (“Waiguo men shang”) and fourteen in the second (“Waiguomen xia”). He first describes Indochina, starting from Annan (Vietnam), Pugan (Burma), and Zhenla (Cambodia), and then focuses on the islands of Sumatra and Java (the kingdom of Sanfoqi and Shepo). Through the Indian Ocean, he reaches the kingdom of Gulin (south coast of India) and Zhunian (east coast of India). He describes Pakistan, Afghanistan, Central Asia, and the Arabian peninsula in the sections “Daqin” and “Dashi” (which contain the description of Meilugudun), and northern Africa and the Mediterranean Sea in the section “Mulanpi”. In the later sections Zhou returns east, describing the northern and western territories of the Indian peninsula, then moving toward the south-eastern sea and introducing several kingdoms. But section 45, “Kunlun Cengqi”, is thought to describe some territories on the east coast of Africa. The exact location of Bosi, the land described in section 46, has yet to be determined, but the place can be identified with some territories between Sumatra and southern Burma.29 Sections 47 through 50 focus on different tribes living near Guangxi or beyond Guanxi, in Vietnam territory. The last two sections (51 and 52) deal with different land and sea routes to China. As the section sequence implies, these two chapters are quite different in their content. The first includes ten countries that are very close to the Chinese empire, both geographically and culturally. Each section explains the diplomatic and commercial relationships between the countries, and then describes local customs, weather, and goods. The second chapter describes fourteen countries that are farther away and in less frequent contact with China’s empire; the material collected here is less substantial and less accurate. In the microstructure of the twenty-four sections, we can distinguish three different parts. First, Zhou explains how to reach the country, then he describes the local customs (with attention to their social, religious, political, and commercial aspects), and finally he introduces local goods and the most important products.30

29 See Almonte, unpublished doctoral dissertation, 2016: 188. 30 Shiji, Sima Qian’s work, also had a similar microstructure.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 21

1.3 Lingwai daida’s Meilugudun Section The “Dashi zhuguo” section is one of the biggest sections of the fourth chapter. It comes after “Daqinguo” and before “Mulanpiguo”, and contains six subsections, one for each Arab country: Malibaguo, Majiaguo, Baidaguo, Jiziniguo, Meilugudunguo, and Wusiliguo. In this paper only the description of Meilugudun is analysed.31 In Zhou’s work the passage concerning Meilugudun reads as follows:

有眉路骨惇國。居七重之城，自上古用黒光大石疊就，每城相去千步。有蕃塔三百餘，内一塔高八十丈，内有三百六十房。人皆纒頭搭項，寒即以色毛叚為衣，以肉麵為食，以金銀為錢。所謂鮫綃、薔薇水、梔子花、摩娑石、鵬（硼）砂，皆其所產也。

There is the Meilugudun country. It is in a seven-fold wall. From remote antiquity, they used tiers of shining big black stones, and each wall is distant from the other by a thousand paces. There are more than three hundred foreign pagodas [minarets], among which one is eighty-zhang high and has three hundred and sixty rooms. All the people wear turbans covering the neck. When it is cold they use also colored woolen materials. Their food consists of meat and bread. Gold and silver are used for coins. The substance called jiaoxiao, rose-water, gardenia flowers, bezoar stones [mosuo], and borax are all products of this country.32

Hirth and Rockhill translate jiaoxiao as ‘sea silk’, also known as byssus, which originates along the Mediterranean coast and particularly in Smyrna, on the western Anatolian peninsula. Byssus is a bundle of filaments secreted by many species of the bivalve mollusc, a secretion that allows the mollusc to stick to a solid surface. Jiao 鲛 means ‘shark’33 and xiao 绡 is a particular raw silk. In Hanyu dacidian it’s called “chuanshuo zhong jiaoren suo zhi de sijuan 、bosha 傳說中鮫人所織的絲絹、薄紗”,34 in English “the silk or muslin knitted by mermaids”.35 The Chinese scholar Yang Wuquan describes jiaoxiao as a kind of high-quality muslin or chiffon; in Chinese, this is gaoji bosha 高级薄纱. In

31 For the other toponyms, see Almonte, unpublished doctoral dissertation, pp. 120 and following. 32 Hirth and Rockhill 1966: 141–42. 33 Schafer translated this name as ‘kraken’ or ‘dragon’. For a more detailed discussion, see Schafer 1985: 220. 34 See Luo Zhufeng, Hanyu dacidian, 1994, volume 12: 1222. 35 Sometimes translated as ‘sea gauze’.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 22 Almonte particular, he quotes the work Shuyiji 述異記 (Description of Strange Events), in which the term jiaoxiao also occurs. In the southern sea (that is, near the coasts of Vietnam and Borneo) some mythological creatures knitted jiaoxiao, which had an approximate value of a hundred pieces of gold; clothes produced with this kind of cotton were very strong and waterproof.36 Qiangwei shui is translated by Hirth and Rockhill as ‘rose-water’ and is a typical precious product from the province of Fars, in modern-day Iran. Zhou does not describe it, Zhao, instead, does say it is from the Dashi countries (Arab countries) and that the flower from which it is made is not identical with the Chinese rose (qiangwei).37 Zhizihua has been translated as ‘gardenia flowers’ because there were different kinds of gardenia. Section 133 focuses on fanzhizi 蕃梔子, or ‘foreign gardenia’.38 Zhou wrote that it was from the Dashi countries (Arabic countries) but that people erroneously believed it came from Xizhu (indicating modern- day Xitianzhu, on the western coast of India), and in Buddhist works was called zhanpuhua 詹葡花, from the Sanskrit champak.39 Most likely Zhou specified this gardenia’s foreignness because there was a kind of Chinese native gardenia flower.40 Also, in the earlier work Youyang zazu, Duan Chengshi described zhizihua, traditionally from western regions (Xiyu) and called zhanbuhua 詹卜花.41 Mosuo stone is the bezoar stone, which Zhou Qufei also describes in the “Jiziniguo” subsection. The object is actually not a stone but a mass of undi- gested matter that has gathered in the gastrointestinal tract. Its usage was very common in Turkey and remains so in modern-day Iran and Afghanistan, where animals able to form bezoars, like sheep and deer, are fairly common.42

36 Translation is mine. Chinese text reads as follows: “南海出鲛绡纱，泉室（指渔人，即人鱼）潜织，一名龙纱。其贾百余金。以为服，入水不濡。 ” Yang Wuquan 1999: 106. Shuyiji was a collection of strange ancient events. It was completed during the Southern dynasties, from 420 to 589 AD. There are two books of this title and no one knows for sure who wrote them. 37 Hirth and Rockhill 1966: 203–04. For a full discussion on the topic (in particular the production of rose-water in Persian territories and its export around the world during the tenth through twelfth centuries), see Le Strange 1966: 293. 38 Yang 1999: 249. 39 Champak is translated as Michelia, belonging to the Magnolia family. See Schafer 1985: 167. 40 For a full discussion on this topic, see Schafer 1985: 199. 41 Duan Chengshi 段成式, Du Cong 杜聪, 2007: 125. 42 Laufer 1919: 525. See also Schafer 1985: 216. Schafer talked about a kind of bezoar extracted from the bellies of some serpents in Nam Viet.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 23

Borax (pengsha), or ‘tinkal’ in Sanskrit was imported to Italy by Marco Polo. Borax is a boron compound that occurs in evaporated deposits left by the re- peated evaporation of seasonal lakes. Today, the most important deposits are found in Turkey. But borax was first discovered in Tibet and was carried via the Silk Road to the Arabian peninsula in the eighth century. Arabs used it to pro- duce a particularly durable kind of glass. Borax traces have been found in Arab products from the eighth century and afterwards, and from the 10th century on this component also occurred in Chinese lacquers.43

2 Zhao Rukuo

2.1 Zhao Rukuo and His Work: Zhufanzhi Zhao Rukuo was a member of the Song imperial family. Like Zhou Qufei he lived between the twelfth and thirteenth centuries, during the Southern Song dynasty. Working as the Superintendent of Merchant Shipping (Shibosi) in Quanzhou, the great port in Fujian, he had more opportunities than Zhou Qufei to speak with the people who engaged directly in foreign trade. In 1225 he wrote Zhufanzhi (Description of the Foreign Lands). It contains two long chapters: one on countries (‘Zhiguo’, 志国) and one on products (‘Zhiwu’, 志物). The first chapter introduces the countries’ geography, people, and customs, and their relationship with China, while the second chapter discusses the various articles imported into China from foreign lands.44 Even if he cited other people’s works—such as dynastic histories, Du Huan’s Jixingji, and Zhou Qufei’s Lingwai daida—he was able to supply a large amount of information directly from Chinese and foreign traders. He was the first to give the names and facts about many countries from south-western Asia, Africa, and the Mediterranean.45 Hirth and Rockhill provided an annotated translation of Zhufanzhi that was published in St. Petersburg in 1911 and followed by an Amsterdam edition published in 1966. Their work gave full value to Zhao Rukuo’s account and to the rise and development of maritime intercourse between China and southern and south-western Asia. They pointed out which sections he based exclusively on oral information furnished by traders, and which ones he took from Zhou

43 Laufer 1919: 503. Also consult http://www.azom.com/article.aspx?ArticleID=2588 [accessed 31 August 2017] and see Needham 1959 3: 662 and following. 44 Park 2012: 50. 45 Hirth and Rockhill 1966: 36.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 24 Almonte

Qufei’s work. In any case, Zhufanzhi reveals the extent of the Chinese geographical knowledge about foreign places. The table below shows Zhufanzhi’s structure.

Table 2 Zhufanzhi’s structure

Volumes Chapters Sections

Juan shang 卷上 Zhiguo 志国 1. Jiaozhiguo 交趾国, (Volume no. 1) (Countries) 2. Zhanchengguo 占城国, 3. Bintonglongguo 宾瞳龙国, 4. Zhenlaguo 真腊国, 5. Dengliumeiguo 登留眉国, 6. Puganguo 蒲甘国, 7. Sanfoqiguo 三佛齐国, 8. Danmalingguo 单马令国, 9. Lingyasijiaguo 凌牙斯加国, 10. Foluoanguo 佛罗安国 11. Xintuoguo 新拖国 12. Jianbiguo 监篦国、 13. Lanwuliguo and Xilanguo 兰无里国、细兰 14. Shepoguo 阇婆国, 15. Sujidan 苏吉丹 16. Nanpaguo and Gulinguo 南毗国、故临国, 17. Huchalaguo 胡茶辣国 18. Maluohuaguo 麻啰华国, 19. Zhunianguo, Pengqieluoguo, and Nannihualuoguo 注辇国、鹏茄罗国、南尼华啰国 20. Daqinguo 大秦国, 21. Tianzhuguo 天竺国, 22. Dashiguo 大食国, 23. Majiaguo 麻嘉国, 24. Cengbaguo 层拔国, 25. Bipaluoguo 弼琶啰国, 26. Wubaguo 勿拔国, 27. Zhongliguo 中理国, 28. Wengmanguo 瓮蛮国,

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 25

Volumes Chapters Sections

29. Jishiguo 记施国, 30. Baidaguo 白达国, 31. Bisiluoguo 弼斯啰国, 32. Jiziniguo 吉兹尼国, 33. Wusiliguo 忽厮离国, 34. Lumeiguo 蘆眉国, 35. Mulanpiguo 木兰皮国, 36. Wusiliguo 勿斯里国, 37. Egentuoguo 遏根陀国, 38. Haishang zaguo 海上杂国: Antuomanguo 晏陀蠻国, Kunlun Cengqiguo 昆仑层期国, Shahuagongguo 沙华公国, Nvreguo 女人国, Bosiguo 波斯国, Chabishaguo 茶弼沙国, Sijialiyeguo 斯伽里野国, Mojialieguo 默伽猎国. 39. Boniguo 渤泥国, 40. Mayiguo 麻逸国, 41. Sanyuguo, Puliluguo 三屿国、蒲哩鲁国, 42. Liuqiuguo 流求国, 43. Pisheyeguo 毗舍耶国, 44. Xinluoguo 新罗国, 45. Weiguo 倭国

Juan Xia (卷下) Zhiwu 志物 1. naozi 脑子, Volume no. 2 (Products) 2. ruxiang 乳香, 3. moyao 没药, 4. xuejie 血碣, 5. jinyanxiang 金颜香, 6. duruxiang 笃褥香, 7. suhexiangyou 苏合香油, 8. anxixiang 安息香, 9. qizihua 桤子花, 10. qiangweishui 蔷薇水, 11. chenxiang 沉香, 12. jianxiang 笺香, 13. suzanxiang 速暂香, 14. huangshuxiang 黄熟香,

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 26 Almonte

Table 2 Zhufanzhi’s structure (cont.)

Volumes Chapters Sections

15. shengxiang 生香, 16. tanxiang 檀香, 17. dinxiang 丁香, 18. roudoukou 肉豆蔻、 19. jiangzhenxiang 降真香、 20. shexiangshui 麝香木、 21. boluomi 波罗密、 22. binglang 槟榔、 23. yezi 椰子、 24. moshizi 没石子、 25. wumenmu 乌樠木、 26. sumu 苏木 27. jibei 吉贝、 28. yexindian 椰心簟、 29. muxiang 木香、 30. baidoukou 白豆蔻、 31. hujiao 胡椒、 32. bichengqie 荜澄茄、 33. awei 阿魏、 34. luhui 芦荟、 35. shanhushu 珊瑚树、 36. liuli 琉璃、 37. maoerjing 猫儿睛、 38. zhenzhu 真珠、 39. chequ 砗磲、 40. xiangya 象牙、 41. xijiao 犀角、 42. wanaqi 腽肭脐、 43. cuimao 翠毛、 44. yingwu 鹦鹉、 45. longxian 龙涎、 46 daimao 蝳瑁、 47. huangla 黄蜡、 48. hainan 海南 (rouzhau 瓇州, changhua 昌化, jiyang 吉阳, wan’an 万安, Lí 黎, wuhuo 物货)

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 27

2.2 Lingwai daida and Zhufanzhi: Geographical Sections Comparing Zhou’s and Zhao’s works and their respective geographical sections, it is evident that Zhao mentioned fifty-eight countries, while Zhou mentioned forty. Only twenty-one countries occur in both works. This difference helped scholars identify how Zhou had influenced Zhao’s work, and to highlight the value of Zhou’s geographical sections. Table number 3 shows how much Zhao quotes verbatim (or nearly verbatim) from Lingwai daida. In the second column these 21 toponyms are listed, based on Lingwai daida’s sequence. The third column lists characters used by Zhao, and the fourth those used by Zhou. The last column contains notes regarding the extent to which Zhao copied Zhou’s sections.

Table 3 Zhou’s and Zhao’s geographical sections comparison

TOPONYMS Zhufanzhi Daida Notes

1. Zhancheng 725 270 Low correspondence 2. Zhenla 601 126 Low correspondence 3. Pugan 105 118 Partial correspondence 4. Sanfoqi 840 273 Low correspondence 5. Shepo 742 163 Partial correspondence 6. Gulin 139 180 Partial. In Zhao’s work there is no section titled “Gulin”, but it is a subsection of “Nanpiguo”. 7. Zhunian 850 259 Partial correspondence 8. Da Qin 499 257 Partial correspondence 9. Da Shi 1,320 21 Low correspondence. In Zhou’s work this section contains six subsections: “Malibaguo”, “Majiaguo”, “Baidaguo”, “Jiziniguo”, “Meilugudunguo”, “Wusiliguo”. 10. Majia 102 138 Partial correspondence 11. Baida 194 129 Partial correspondence 12. Jizini 193 121 Partial correspondence 13. Meilugudun 145 85 Partial correspondence. In Zhao’s work this section is titled “Lumeiguo” (蘆眉国); he also mentioned the ‘Meilugu’ 眉路骨 toponym.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 28 Almonte

Table 3 Zhou’s and Zhao’s geographical sections comparison (cont.)

TOPONYMS Zhufanzhi Daida Notes

14. Wusili 63 69 Nearly verbatim. Zhao used two very similar toponyms twice: ‘Wusiliguo’ 勿斯離国 for Mosul (as in Zhou’s section); and ‘Wusiliguo’ 勿斯里国, which probably meant territory in Egypt. Zhou did not mention this second toponym. 15. Mulanpi 244 210 Nearly verbatim. 16. (Xitian) 146 92 Partial correspondence. In Zhao’s Nannihualuoguo work this is a subsection of “Zhunian”. He used ‘Nannihualuoguo’. 17. Dongnan \ Haishang \ Dongnan Partial correspondence. Only the Haishang zaguo Haishang title is similar; Zhou and Zhao zhuzaguo 海上杂国 zhuzaguo describe different subsections. 东南海上诸杂国 18. Shahuagong 92 95 Nearly verbatim. 19. Nürenguo 187 108 20. Kunlun Cengqi 116 135 Nearly verbatim 21. Bosi 100 94 Nearly verbatim

As is visible in the table, in the Wusiliguo, Mulanpiguo, Shahuagong, Nvrenguo, Kunlun Cengqiguo, and Bosiguo descriptions, Zhao quotes Daida nearly verbatim, adding a phrase here or changing some words there. Eleven sections are based on descriptions taken from Zhou Qufei’s work, but with added information supplied by Chinese or foreign traders. They are “Puganguo”, “Shepoguo”, “Gulinguo”, “Zhunianguo”, “Daqinguo”, “Majiaguo”, “Baidaguo”, “Jiziniguo”, “Meilugudunguo”, “Nannihualuoguo”, and “Dongnan haishang zhuzaguo”. Among them only the “Shepo”, “Zhunian”, and “Daqin” sections are richer and more detailed than Zhou’s. The “Zhancheng”, “Zhenla”, “Sanfoqi”, and “Dashi” sections contain no overlap with Zhou’s sections, indicating they are based exclusively on oral information.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 29

2.3 Zhufanzhi’s Lumei Section Zhao’s section regarding the country of Lumei reads as follows:

蘆眉國,自麻囉拔西陸行三百餘程始到，亦名眉路骨國。其城屈曲七重，用黑光大石甃就，每城相去千步。有番塔三百除，內一塔高八十丈，容四馬並驅而上，內有三百六十房。人皆纏頭塌頂，以色毛段為衣。以肉麵為食，以金銀為錢。有四萬戶織錦為業。地產絞綃、金字越諾布、間金間絲織錦綺、摩娑石、無名異、薔薇水、梔子花、蘇合油、鵬砂、及上等碾花琉璃。人家好畜駞馬犬。46

If one travels by land toward the west for over three hundred cheng [stages] from Maluoba, one reaches Lumei, also called Meilugu country. The city wall is crooked, seven-fold, and built of large smooth flat black stones; each wall is a distant thousand paces. There are more than three hundred foreign towers [minarets], among which one is eighty-zhang high, in which four horses abreast can be driven up. It has three hundred and sixty rooms. All the people wear turbans covering the neck; their clothing is made of colored woolen materials. Their food consists of meat and bread. They use gold and silver coins. Forty thousand families are employed weaving silk brocades. The products of the country are byssus, gold spangled yuenuo cloth, brocades with alternating stripes of gold and silk, bezoar stones, wumingyi, rose-water, gardenia flowers, liquid storax, borax, and a superior quality of engraved opaque glassware. The people are fond of breeding camels, horses, and dogs.47

The section’s title is “Lumeiguo” (“蘆眉國”), not Zhou Qufei’s “Meilugudunguo”. Zhao’s section is very largely taken from the Lingwai daida.48 He added some words regarding local products, and an opening passage explaining how to reach Lumeiguo from Maluoba, travelling west for three hundred stages. When he talked about the towers (minarets) he wrote that these could be driven by four horses. He then provided information concerning the important production of silk brocades (forty thousand families were employed weaving them) and added some products that Zhou did not mention.49

46 Feng Chengjun 2011: 49. 47 Hirth and Rockhill 1966: 141. 48 Hirth and Rockhill 1966: 141. Feng Chengjun 2011: 49. 49 For example, yuenuobu (a kind of cloth), silk and gold brocades, wumingyi, suheyou (liquid storax), and opaque glassware.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 30 Almonte

The section is short on valuable original information, and nothing in the description of the place can help the reader pin down what countries Lumei and Meilugu actually were. Zhou describes his Meilugudunguo in his work’s “Dashi” section (“Dashi zhuguo”, “大食诸国”, “Arab Countries”), with Malibaguo, Majiaguo, Baidaguo, Jiziniguo, and Wusiliguo. In Zhufanzhi it follows the “Wusiliguo” section (probably Mosul in northern Iraq) and precedes “Mulanpiguo” (the western coast of Africa). The country might have been anywhere in a wide area stretching from the Arabian peninsula through Syria and Iraq to eastern Afghanistan. But there is not enough data to clearly identify it. Zhao mentions Maluoba as a starting point to reach Lumei, although in his work there is no section called “Maluoba”. After the country Huchala 胡茶辣国 (western India), there is Maluohuaguo 麻啰华国, which is most likely the Malwa territory (west-central India), as Feng Chengjun states in his notes to Zhufanzhi.50 It is likely that Zhao referred to this country when he said “if you travel three hundred stages51 west from Maluoba you reach Lumei”. Travelling north-west from Malwa territory, today one can reach Iran or Afghanistan. A second hypothesis holds that ‘Maluoba’ might be a different transcription of ‘Maliba’, indicating Merbat. In Lingwai daida there is a section titled “Malibaguo”, a place the Chinese scholar Yang Wuquan identifies with the port of Murbat, which is Mirbat in today’s Oman.52 The section “Majiaguo” (Mecca) also deserves mention; in it Zhou affirms that travelling west from Maliba country will take one to Majia in eighty cheng.53 From Mirbat, travelling west you can reach some places between the Arabian peninsula: Iraq, Syria, Turkey, and Egypt. Lumei (in Zhao’s work) and Meilugudun (in Zhou’s) may be the same place, but a firm identification is difficult, as we shall now discuss.

50 Feng Chengjun 2011: 23. 51 It is difficult to establish how long a cheng 程 was. 52 Yang Wuquan 1999: 101, note no. 2. He has considered Hirth and Rockhill’s, Feng Chengjun’s, Su Jiqing’s, and Zhang Xinglang’s theories. Hirth and Rockhill state that Zhou’s Maliba was Mirbat, on the Hadramaut coast of Arabia. Hirth and Rockhill 1966: 25 and 121. Zhang Xinglang affirms that the country referred to is on western India’s Malabar coast. Zhang Xinglang 2003 II: 820. Su Jiqing, in his section “Lijiata” (“哩伽塔”) talks about Turba, a Yemen port. Su Jiqing 2011: 211. 53 Translation is mine. Chinese text: “自麻離拔國西去，陸行八十餘程乃到”. Yang 1999: 99–100.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 31

3 Controversial Hypotheses from Previous Scholars

As Hirth and Rockhill argue, ‘Meilugudun’ may not be the name of any city or country, but a transcription of the Arabic word mulhidun, ‘infidels’, which means non-Muslims. Under this theory, Zhou’s Meilugudun (or Zhao’s Lumei) would be a territory of infidels, with the most likely suspect being Constantinople: as capital of the Byzantine Empire it would be the main in- fidel city for an Arab informant. Zhao’s toponym ‘Lumei’ might be the phonetic transcription of ‘Ruma’, the Arabic word for Rome or the Eastern Roman Empire.54 But ‘Meilugu’ or ‘Meilugudun’ is quite different from ‘Lumei’. Zhao gives a distance of three hundred cheng from Maluoba, but we don’t know how long a cheng was during the Song dynasty. It is also difficult to establish whether the person travelling went by foot, horse, or ship. If Maluoba was actually Mirbat in Oman, it seems conceivable that from there to Constantinople would be three hundred cheng.55 This would also explain the proximity of the country of Sijialiye (maybe Sicily in Italy) to the frontier of Lumei.56 Zhao’s passage reads as follows: “Sijialiyeguo, jin Lumeiguojie 斯加里野国，近蘆眉国界，。。。”.57 Zhou Qufei did not reference Sicily, Rome, or Italy. The second hypothesis is focused on correspondence between ‘Lumei’ and the Arabic word ‘Rūm Bilād al-Rūm’. In medieval Islamic geography, ‘Rūm Bilād al-Rūm’ referred to a wide territory beyond Islamic countries: Rome, the Byzantine Empire, and sometimes Christian Europe as a whole.58 It is likely that Zhou and Zhao referred to the Seljuk (or ‘Saljuk’, as transliterated by Le Strange)59 Sultanate of Rum. Seljuk Turkish rulers held power in western Anatolia from 1077 to 1257, after conquering Byzantine cities. The sultanate’s capital was Kuniyah, or Iconium.60 The rise of the Seljuk Turks entirely changed the face of Asia Minor, with the battle of Manzikart (or Malasjird) in 1071 followed by Rum’s establishment as one of the lands of Islam.61 With

54 Hirth and Rockhill 1966: 141–42. 55 Hargett argues that during Song dynasty a cheng was fifteen to twenty miles, that being the distance a traveller could cover in one day. Therefore, three hundred cheng was too big a distance when compared with the 3,200 miles between modern Mirbat and Istanbul. But we don’t have enough data to state how long was a cheng for Zhao Rukuo. See Hargett 2010:189–93. 56 Hirth and Rockhill 1966: 142. 57 Hirth and Rockhill 1966: 153–54. Feng Chengjun 2011: 58. 58 Durak 2010: 285–98. 59 Le Strange 1966: 141. 60 Kazhdan 1991 3: 1816. 61 Jaubert 1836 2: 310.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 32 Almonte

Rum lasting until the Mongol conquest in 1257, the timeline raises the possibil- ity that Zhou and Zhao referred to the Seljuk territories. Except for this point, there aren’t any analogies between ancient Iconium and the description from the two Chinese authors. Hirth and Rockhill underline the points of resemblance between Meilugudun and Damascus: the presence of number seven (sevenfold wall and the presence of seven city gates); the foreign tower (minaret or mosque) with 360 chambers may refer to the Djami mosque; the great importance to the silk brocade industry (mentioned only in Zhao’s section), can led also to the main center of Oriental trade, as did Damascus in Syria.62 We know that in Arabic the capital of a province may be called by the name of the province, even when it has a name of its own; thus, Damascus was commonly known as Ash-Sham (the capital of Syria).63 There isn’t any corroborating evidence to suggest that Zhou’s Meilugu could be Damascus in Syria: there are no linguistic analogies between the two toponyms ‘Damascus’ and ‘Ash-Sham’, on the one hand, and Meilugu on the other. According to Rawlinson and Le Strange’s hypothesis, and as Zhang Xinglang stated, Meilugudun could be the city of Balkh in northern Afghanistan, or the Zoroastrian temple of Naw Bahar. Yaqut’s (twelfth century) describes Naw Bahar as a great temple built in imitation of Mecca’s Kabah and with precious stones and a central building that comprised 360 chambers.64 This description certainly supports Zhang’s hypothesis, because there is a clear correspondence with Zhou’s section. In ancient times Balkh was capital of what the Greeks called Bactria, and it became the center of the Zoroastrian religion.65 Until the Mongol invasion (1220), Balkh was considered one of the most important cities of central Asia.66 But there is no exisiting linguistic or phonetic correspondence between ‘Meilugudun’ and the ancient toponym for Balkh, known also as Omm al-belad (Yaʿqūbī in the ninth century) or Balkh al bahiya (Moqaddasī in the tenth century). Lu Junling affirms that Zhao’s Lumei is Rome and that Zhou’s Meilugudun is the Zoroastrian temple in Balkh.67 He was the first scholar to give these

62 Hirth and Rockhill 1966: 141. 63 Le Strange 1966: 21. 64 Yāqūt 1955–1957 1: 713. 65 Zhang Xinglang 2003 2: 823 and following. Le Strange 1966: 420–22. Foltz 2010: 152. 66 Other Arab geographers have also described Balkh as one of the greatest cities of Khurasan. For a more detailed discussion, see Planhol and others, 1988 3, Fasc. 6: 587–96. 67 Yang Wuquan 1999: 105. And Ptak 1995: 56 and following.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 33 toponyms two separate identities. But, as with other theories discussed here, there is no firm evidence to settle things.68 Hirth and Rockhill state it was possible that Meilugudun and Lumei are names for a composite creation, one standing in for all territories remote from Chinese geographical knowledge.69

4 A New Identification: The Territory of Merv in Turkmenistan

In this paper, the author considers a new hypothesis: the Meilugudun mentioned by Zhou may indicate Merv, a land whose borders fell within those of modern-day Turkmenistan. In the twelfth century Merv provided an important oasis in Central Asia on the historical Silk Road, and Chinese and Arabic sources mention the country several times. In Hou Hanshu 后汉书, chapter 88 (“Xiyuzhuan 西域传”, Western Regions), section 78 (“Anxi 安息”) contains the toponym ‘Mulucheng’ 木鹿城, also known as ‘Little Anxi’ and indicating a place on the eastern border of the Anxi Empire (Parthian Empire), located two thousand li from Luoyang.70 Zhang Xinglang states ‘Mulucheng’ was the transliteration of ‘Maru’, or ‘Merv’. In Suishu 隋书, chapter 83 (“Liezhuan 列傳”) discusses the western regions in section 48, which also mentions Muguo 穆国, a possession of the Anxi Empire during antiquity. Du Huan in the eighth century used ‘Molu’ 末禄 (some editions also use ‘Zhuluguo’ 朱录国; it’s likely that he wrote the wrong character, ‘Zhulu’ 朱禄, instead of ‘Molu’ 末禄).71 Chinese scholar Song Xian says that in earlier times the country of Molu 末禄 was called Mulu 木鹿, with a different character, and indicates modern Mary province (Malei 马雷) in Turkmenistan. Today Mary province corresponds to the ancient cities of Merv, Marv, or Maru. In Xin Tangshu 新唐书, section “Dashi 大食”, we first read Mulu 木鹿 and then Molu 末禄, dealing with the eastern border of the Dashi territories.72 Zhang Xinglang states they are the same country, in Khorasan province. Since the ‘Mulu’ 木鹿and ‘Molu 末禄 toponyms are the same ones used in Du Huan’s work and in Hou Hanshu, Zhang Xinglang identifies them with Merv territory in modern Turkmenistan.

68 Lu Junling 2006: 34. 69 Hirth and Rockhill 1966: 142. 70 Zhang Xinglang 1966 2: 1027. 71 Song Xian 2011: 9. See also Zhang Yijun, Jing Xing Ji 200: 58. 72 Zhang Xinglang 1966 2: 696.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 34 Almonte

In Yuanshi 元史, chapter 63, Zhi number 15, Dili 6, appendix Xibeidi, we find all toponyms related to Bosi country (Persian territories). Among them is also Maliwu 麻里兀. Zhang states it was Merv city, belonging first to the Parthian Empire and then to the Khorasan territories.73 In Arab sources, from the ninth to thirteenth centuries, there are many references to Merv’s importance as well. Until the Mongol conquest (1220), Merv was one of the most famous centers of the Silk Road. Along with Nysabur, Herat, and Balkh, the city was one of the capitals of Khorasan, the eastern land of the Islamic territories.74 Merv was located on the Murghab River, also called the Merv River, whose waters fed four canals that radiated to various quarters of the city. In Ibn Khurradadhbih’s tenth-century work, Kitab al-masalik wa-al-mamalik,75 Merv is mentioned in the section describing the Baghdad-Khorasan route, the section describing the area toward the north-east, the section on the Merv-Chach route, and the one on Turks country. He noted the distance between Merv and Baghdad as 371 fars,76 and the Merv-Balkh route was 126 fars. Ibn Khurradadhbih said Merv was also known as Nil, and was composed of two districts on two different riv- ersides and with several estates. In his tenth-century work, Aḥsan al-taqasim fi maʿrifat al aqalim, Muqqadasi affirmed that Merv was one of the most beautiful cities of the past.77 Yaqut, in the thirteenth century, spent two years studying in Merv libraries, which were praised by every ancient scholar.78 Idrisi’s work, Nuzhat al Mustaq, compiled in 1154 in Sicily, mentions the toponym ‘Merw el Roud’.79 The place by that name is described as a flourishing and ancient city, placed in a large fertile plain. It had three mosques and an important keep on a mountain. According to Idrisi, the weather was temperate and very comfortable. Local products were soy, soy butter, and a high-quality cotton known as ‘Merv’s cotton’. Ghazni was among the city’s estates.80 According to Yakut’s and Muqqadasi’s accounts of Merv province, there were two Mervs: Murghab River flowed into Upper Merv (or Little Merv) and

73 Zhang Xinglang 1966 2: 1164. Please see also Yuanshi 1976 5: 1574. 74 Le Strange 1966: 382. 75 Meynard 1865: 47 (Arabic text) and 162–69 (French translation). 76 Fars was an Arab length unit measure. One fars was 0.16 km, and 371 fars would be about 60 km. 77 Bosworth 2013, 6: 618–21. 78 Le Strange 1966: 401. 79 Jaubert 1836 1: 466. 80 Jaubert 1836 1: 467.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 35 towards Lower Merv (or Great Merv). The first was called Shahijan in Arabic, distinguished from Merw el Roud, Little Merv. Le Strange argues that ‘Shahijan’ was probably the Arab form of the old Persian ‘Shahgan’, meaning ‘kingly or belonging to the king’.81 The Great Merv consisted of an inner high-built citadel, surrounded by the city with four gates, beyond which were extensive suburbs stretching along the banks of the great canals.82 There were three mosques in Merv and a grand market. The domed house of Abu Muslim, the first Abbasid caliph, was also an important building, built of burnt brick. Merv’s market was a great place for all kinds of silk, cotton, mixed cotton and silk, sesame oils, aromatics, manna, and brass pots. Merv inhabitants were also highly skilled in working with precious metals, like gold and silver.83 Travellers could also find various other products from Khorasan province. Indeed, Merv was one of the biggest and richest cities of the eastern Arab territories and was situated on the major Khorasan road, where the crossing for Bukhara occurred.84 During the early Middle Ages, all provinces from Little Merv to Great Merv were under cultivation and studded with villages and towns. Chinese sources and Arab ones share some aspects. All travellers or geographers who deal with Merv cotton and precious metals also describe Merv’s mosques and river. Zhou talks about a sevenfold wall that surrounded the territory of Meilugudun. Merv’s walls, as archeological excavation shows us today, were built in different periods and with a variety of fortifications.85 Although we do not know exactly how many walls there were in Merv in ancient times, the presence of different fortifications has led the author to argue it was possible that there was no sevenfold wall and that Zhou fell prey to language problems or inaccurate sources. Merv territory, from Little Merv to Great Merv, was a marvel of towns and villages, each of which had its own mosque and market. Zhou says there are three hundred “foreign pagodas/towers” (most likely minarets), a figure that matches the number of mosques. It’s conceivable that in Merv province during the twelfth century there were three hundred mosques, or possibly Zhou used that number to indicate merely a large amount.

81 Le Strange 1966: 398. 82 De Goeje 1877: 330–31. 83 See London’s Global University website: http://www.ucl.ac.uk/merv [accessed 31 August 2017]. 84 Williams 2002: 4–15. 85 Hermann, “Medieval City of Merv”, https://maxvanberchem.org/fr/11-archeologie/46-the- medieval-city-of-merv?highlight=WyJtZXJ2IiwibWVydidzIl0= [accessed 13 September 2017]. Zavyalov 2007: 313–29. Williams 2007: 42–62.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 36 Almonte

According to Zhou’s account, one of these “foreign pagodas/towers” was eighty-zhang high and had 360 chambers. This description has led the author to argue that Zhou is confused between the Naw Bahar temple (the Zoroastrian temple situated near the city of Balkh, not far from Merv) and the domed house of Abu Muslim, who founded the Abbasid dynasty in the eighth century, located in Merv. This was a palace with a domed chamber in the center, surrounded by four iwans (typical Islamic architectural structure) opening to various courts, with many rooms for residents (women, servants, and so on).86 Zhou mentions jiaoxiao as one of the major products in Meilugudun. This term can indicate sea silk (also called byssus) or, more generally, a high-quality cotton. Although cotton was produced in numerous places throughout Central Asia, ancient Merv was celebrated for its production of silk and cotton, which was exported to other parts of Khorasan, not just to each settlement of the Silk Road.87 Zhou lists jiaoxiao first, possibly because it was one of the main products of this territory; we don’t know if it was a casual choice or not. But thanks to Arab references, we can conclude Merv’s precious cotton and silk were well known around the world. The exports provided a high income for all inhabitants and had been a very important economic component since the Sassanid dynasty, as archeological excavations reveal.88 Zhou most likely used the word jiaoxiao in an inappropriate way, because it was closer to the Chinese idea of high-value cotton and silk. Therefore, it was easier for the Chinese reader to understand. The use of gold and silver for coins, as mentioned by Zhou, brings to mind Merv inhabitants’ skill at handling precious metals. Several places in Central Asia during the Middle Ages doubtless used gold and silver coins. But, according to several references, Merv was also important because the oasis was found on one of the main arms of the Silk Road, the time-old route that reached from the Far East to Europe and Africa. Since the Achaemenian Empire in the sixth century, its importance gradually grew until the Mongol invasion. Even during the Caliphate of the eighth century, when the capital was settled in Baghdad, the province of Merv remained the most important point of the Silk Route in Khorasan.89 Using silver and gold to make coins shows the level of wealth and civilization of the people and demonstrates the importance and

86 Please see Williams 2002: 4–15. Corbishley and Cooke 2005: 17. 87 Le Strange 1966: 400–401. Corbishley and Cooke 2005: 4. Primary Arab sources: De Goeje 1877: 298–99, 310–12. Yāqūt 1: 534 and 827, 2: 610. 88 Hermann and others 2001 39: 9–52. Le Strange 1985: 397–401. Nowadays textile production is still an important source of income in Turkmenistan. 89 Le Strange 1966: 397–406.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 37 productiveness of Merv’s markets, which achieved levels never reached by other cities in Khorasan.90 We need to consider phonetic aspects too. Modern Chinese is different from the middle Chinese spoken during the Song dynasty. Pulleyblank highlights late-middle Chinese pronunciation (from the Tang dynasty): the toponym ‘Meilugudun’ likely sounded as [mi – luə – kut – tun].91 A second aspect is also very important: the difference between middle Chinese and southern dialects, and in particular the Yue dialect. This was spoken in Guangxi and Guangdong provinces, where Zhou lived and worked for a long time. According to Hanyu Dacidian, the pronunciation of ‘Meilugudun’ in the dialect of Guangxi province was [mei – lou – gwat – deon]. In both these cases it is possible to reveal a correspondence with the Arabic toponym used by Idrisi, Merw el Roud. Comparing the pronunciation of Arabic and Chinese toponyms (in Guangxi dialect), one discovers a kind of sound analogy, at least in the vocals: [mei – lou – gwat – deon] and ‘Merw el Roud’.92 This raises two questions. First, how is that possible? And is that enough for us to state that the Meilugudun described by Zhou was really the Merv territory in Khorasan described by several Arabic geographers? To answer the first question, it’s important to point out that Zhou lived a large part of his life in Guanxi province, working very closely with its inhabitants. Therefore, it is conceivable that his informers were local merchants, interpreters, or travellers. They used Guangxi dialect to speak, and to name foreign things and countries. Zhou’s transliteration of ‘Meilugudun’ was influenced by the Yue dialect’s pronunciation. In the same way it is likely that foreign and local travellers and merchants were influenced by the Arabic language due to commercial exchanges along the Silk Road. During the Middle Ages, important trade between Chinese and foreign merchants (Arab and Persian) led these people to obtain a great deal of information about foreign countries and cul- tures. The Silk Road was not just used to carry goods from one part of the world to another. Merchants and travellers used it to spread information and ideas as well, with new products being introduced to their culture, literature, and the arts. Merv certainly was very famous among Arab and Persian merchants, who spread its name and products over continents. Although Zhou probably never

90 De Goeje, Al Istakhri 1870–77 1: 260–61. 91 Pulleyblank 1984: 60 and ff. 92 This author has chosen Jaubert’s transliteration, but others are common. For example, Le Strange uses Marv ar Rud to indicate the same territory. Le Strange 1966: 398. The extra- bold characters indicate the sound analogy in the vocals.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 38 Almonte read Arab geographic works and never travelled to Khorasan, his geographical knowledge was influenced by Arab concepts thanks to this circulation of ideas. As for the second question, the similarity of the pronunciation of these two toponyms is not conclusive evidence—one must consider all the common aspects listed above. None of them gives us clear proof of the correspondence between Meilugudun and Merw el Roud. But together they can lead scholars to recognize an important resemblance. One last aspect should be considered: Zhou mentioned jiaoxiao, rose-water, gardenia flowers, bezoar stones (mosuo), and borax. There are no direct references stating these products were associated with Merv, but several Arab geographers have noted93 Merv’s importance as a territory, one placed on a crucial point of the Silk Road, near Bukhara and Samarkand. All kinds of goods were exchanged through the Merv markets, regardless of the goods’ provenance. Traders usually carried their goods in one big market along the Silk Road, mak- ing deals there and travelling back home with new loads. Therefore, it’s likely that Zhou’s informers had experience with those products in Merv markets. In a very international context, where people were engaged with the import or export of goods from the east to the Mediterranean and Europe (and vice versa), the place where some goods were produced did not always matter. Since previous hypotheses are not convincing and Meilugudun’s description in Lingwai daida deserves much more attention from scholars, the aim of this paper is to propose a new identification of the place named by Chinese toponym: in particular, with Merw el Roud, as quoted by Idrisi in 1154. Zhou was not the first Chinese author to mention the city of Merv, but he used a new transcription, a toponym never used before and probably influenced by the Arabic language. Therefore, we have the ‘Meilugudun’ toponym, far from ‘Muru’ or ‘Malu’, as written in Xin tangshu or in Yuanshi, but closer to ‘Merw el Roud’, as mentioned in Idrisi’s work, Nuzhat al Nustaq (1154). Idrisi in 1154 completed a description of the earth and a world map, compiled thanks to his many travels and research of previous Arab sources. His work often fails to provide accurate information about distant countries, instead quoting ancient Greek and Arab sources. Nevertheless, his work can be considered one of the greatest geographical works on Arab countries, due to the richness and high value of his descriptions and maps.94 Lingwai daida, compiled by Zhou in 1178, is particularly valuable for its data on Guangxi

93 See the notes above. 94 Ahmad 1987 2.1: 156–72.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 39 province, but his chapters regarding foreign countries are significant too, as they provide information never read before.95 Scholars such as Needham, Zhang Xinglang, and Tibbetts have said Arab geographers contributed to the growth of Chinese geographical knowledge during the Middle Ages.96 According to this research and thanks to an accurate comparison between Chinese and Arabic primary sources,97 it’s conceivable to argue Zhou was influenced by Arabic geographical concepts and toponyms (as shown above). Zhou’s descriptions often are incomplete and inaccurate, and this is due to several factors. First, his informers often were not travellers and did not have experience in the countries they described—they used only second-hand information. Even if they travelled to those countries, they didn’t necessarily know the local languages; thus, the information they provided could be in- exact and imprecise (as can be seen in the Meilugudun section and in every section that describes countries distant from China). Another aspect is worth pointing out: Zhou tries to fit his information (regarding foreign countries) into a Chinese frame of reference. In this way, readers who weren’t travellers or geographers could more easily understand his descriptions and grasp the greatness of foreign countries. Forcing his information to fit into a familiar frame does much to reduce the faithfulness and accuracy of his descriptions. Although these reflections are only hypotheses, they may help scholars gain more insight into Zhou’s work, which has been under-appreciated for too long. The uncertain identification and location of Meilugudun have attracted the attention of many Chinese and western scholars. The similarities between Arabic sources and Zhou’s section, listed above, should lead more scholars to recognize Lingwai daida’s great significance, notably as a cross-linguistic exchange between the Chinese and Islamic world. This remarkable book deserves much more attention from researchers.

95 For a more detailed discussion, see Almonte, unpublished doctoral thesis, 2016: chapter three. 96 Needham and Wang 1959 III: 512. Zhang Xinglang 2003 II: 677–958 and III: 1279–1326. Ahmad 1987 2.1: 156–72. Tibbetts 1992 2.1: 108–36. Tibbetts 1992 2.1: 257. 97 For a more detailed discussion on this topic, see Almonte, unpublished doctoral dissertation, 2016: 179 and the following.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 40 Almonte

Bibliography

Ahmad, S. Maqbul, “Cartography of al-Sharif al Idrisi”, in Woodward and Harley, The History of Cartography, Islamic Tradition, 2.1, pp. 156–72, Chicago: University of Chicago Press, 1987. Almonte, Victoria, “L’ingerenza della lingua araba nella toponomastica cinese: il va- lore dell’opera Lingwai daida di Zhou Qufei”, unpublished doctoral dissertation, Sapienza, Rome University, April 2016. Bosworth, C. E. “Merv”, in Encyclopedia of Islam, New Series, 6 (2013): 618–21. Chen, Jiarong, Fang Xie, and Junling Lu (eds), Gudai Nanhai Diming Huishi 古代南海地名汇释 (Explanation of Antique Toponyms of South China Sea), Beijing: Zhonghua Shuju, 1986. Corbishley, Mike, Luise Cooke, Ancient Cities of Merv: Handbook for Teachers, London: University College of London, 2005. De Goeje, M. J. (ed.), Al Istakhri, Viae Regnorum, auctore Al Istakhri, Bibliotheca Geographorum Arabicorum, E. J.E., Leida, 1870–77, vol. 1. De Goeje, M. J. (ed.), Al-Muqaddasi, Descriptio imperii moslemici, auctore Schamso’d- dîn Abû Abdollâh Mohammed ibn Ahmed ibn Abî Bekr al-Bannâ al-Basschârî al- Mokaddasi, Lugduni Batavorum, Bibliotheca Geographorum Arabicorum, E. J. Brill, Leida, 1870–77. Duan Chengshi 段成式, Du Cong 杜聪, Youyang zazu 酉阳杂俎, Lishi Biji mingzhu congshu 历代笔记名著丛书, (Collezione di opere classiche di appunti sulle antiche dinastie) Qishu shushe 齐鲁书社, Jinan, 2007. Durak, Koray “Who Are the Romans? The Definition of Rūm Bilād al-Rūm (Land of the Romans) in Medieval Islamic Geographies”, in Journal of Intercultural Studies, Routledge, Taylor and Francis Group, 31, 3 (2010): 285–98. Fan, Chengda, Fan Shihu ji 范石湖集, (Collection of Fan Shihu’s Works), Shanghai Guji chubanshe, 1981. Fan, Chengda, Treatises of the Supervisor and Guardian of the Cinnamon Sea, translated by James M. Hargett, University of Washington Press, 2010. Feng, Chengjun (ed.), Rukuo Zhao (author), Zhufanzhi jiaozhu 诸蕃志校注 (Annotations on “Description of the Foreign Lands”), Quanzhou Wenku, Shanghai Cishu chubanshe, 2011. Feng, Chengjun and Junling Lu, Xiyu diming 西域地名 (Western Territories Toponyms), Beijing: Zhonghua Shuju, 1980. Foccardi, Gabriele, Viaggiatori del Regno di Mezzo, i viaggi marittimi dei Cinesi dal III secolo a.C. alla fine del XIX secolo d.C., Torino: Giulio Einaudi, 1992. Foltz, Richard, Religioni della via della Seta, New York: Palgrave Macmilliam, 2010. Hargett, James M., “Song Dynasty Local Gazetteers and their Place in the History of Difangzhi Writing”, Harvard Journal of Asiatic Studies, 56, 2 (1996): 405–42.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 41

Hermann, Georgina, “The Medieval City of Merv and the Buildings of the Merv Oasis”, https://maxvanberchem.org/fr/11-archeologie/46-the-medieval-city-of-merv?high light=WyJtZXJ2IiwibWVydidzIl0= [last accessed 13 September 2017]. Hermann, Georgina, and K. Kurbansakhatov, and others, “The International Merv Project Preliminary Report on the Ninth Year (2000)”, Iran, 39, (2001): 9–52. Hirth, Friedrich, and William Woodville Rockhill, Chau Ju-Kua: His Work on the Chinese and Arab Trade in the Twelfth and Thirteenth Centuries, Entitled Chu-fan-chi, Amsterdam: Oriental Press, 1966. Hu, Chuanzhi, “Lun Yang Wanli jiesong Jin shi shi 论杨万里接送金使诗”, (About Yang Wanli’s Poem on the Diplomatic Mission to Jin Dynasty), in Literary Heritage (Wenxue Yichan), n.p., (2010), 62–69. Huang, Quancai, and Sifei Lu and Dongfu Yang, “Zhou Qufei shili kao 周去非仕历考”, (“Considerations on Zhou Qufei’s Official Career”), in Guangxi Shifan Xueyuan xuebao (Zhexue Shehui Kexueban) 4 (1996): 76–82. Hucker, O. Charles, A Dictionary of Official Title in Imperial China, Taiwan Edition, 1987. Jaubert, P. Amedee, Geographie d’Edrisi, Imprimerie Imperiale, Parigi, 1836. Kazhdan, Alexander, “Rum”, Oxford Dictionary of Byzantium, New York: Oxford University Press, 1991. Laufer, Berthold, Sino–Iranica: Chinese Contributions to the History of Civilization in the Ancient Iran, Chicago, 1919. Le Strange, Guy, The Lands of the Eastern Caliphate, Frank Cass and Co. Ltd, 1966. Li, Qingxin, Haishang sichou zhilu 海上丝绸之路 (Maritime Silk Road), Beijing: Wuzhou chuanbo, 2006. Liu, Junling, “Lingwai daida yanjiu 《岭外代答》研究” (Research on Lingwai daida), unpublished master thesis, tutor prof. Miao Shumei, Henan University, 2009. Lou Zhufeng, Hanyu dacidian, Hanyu dacidian chubanshe, 1994. Meynard, Charles Barbier De, Les livres des routes et des provinces par Ibn Khurradadhbih, Imprimerie Imperiale, Parigi, 1865. Miao, Jing, “Fan Chengda sixiang chutan 范成大思想初探”, (Preliminary Research about Fan Chengda’s Thoughts), Journal of Liaocheng Teachers University (Liaocheng Shifan Daxue Yuanxuebao, Zhexue shehui kexuebao), 3 (1999): 3–9. Miao, Shumei, Songdai de guanyuan xuanren he guanli zhidu 宋代的官员选任和管理制度, Dingchaifa 定差法, (Imperial Officials’ Management and Election System during the Song Dynasty: Dingchai Law), Henan Daxue chubanshe, 1996: 196–206. Mills, John Vivian Gottlieb (ed.), Huan Ma (author), Ying-yai Sheng Lan: The Overall Survey of the Ocean’s Shores, Cambridge: University Press, 1970. Needham, Joseph and Ling Wang, Science and Civilisation in China, vol. 3: Mathematics and the Sciences of the Heavens and the Earth, Cambridge: Cambridge University Press, 1959.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access 42 Almonte

Netolitzky, Almut, Das Ling-wai tai-ta von Chou Chʼü-fei: e. Landeskunde Südchinas aus d. 12. Jh, Wiesbaden: Steiner, 1977. Park, Hyunhee, Mapping the Chinese and Islamic Worlds: Cross-Cultural Exchange in Pre-Modern Asia, New York: Cambridge University Press, 2012. Planhol, Xavier De, C. E. Bosworth, V. Fourniau, D. Balland, F. Grenet; “Balkẖ”, in Encyclopedia Iranica, 1988, vol. 3, Fasc. 6: 587–96. Ptak, Roderich, “Images of Maritime Asia in Two Yuan Texts: Daoyi zhilue and Yiyu zhi”, in Journal of Song-Yuan Studies 25 (1995): 47–75. Pulleyblank, Edwin G., Middle Chinese: A Study in Historical Phonology, Vancouver: University of British Columbia Press, 1984. Pulleyblank, Edwin G., Lexicon of Reconstructed Pronunciation in Early Middle Chinese, Late Middle Chinese, and Early Mandarin, Vancouver: UBC Press, 1991. Schafer, Edward, The Vermilion Bird: Tang Images of the South, Berkeley: University of California Press, 1985. Shen, Songjin, “Yang Wanli Chengzhai ti xinjie 杨万里《诚斋体》新解”, (Some New Considerations about Yang Wanli’s Name “Chengzhai ti”), in Literary Heritage (Wenxue Yichan), n.p., 2006: 73–83. Tibbetts, Gerald R., “The Balkhi School of Geographers”, in Woodward and Harley, History of Cartography: Islamic Tradition, vol. 2, book 1, chapter 5, pp. 108–36, Chicago: University of Chicago Press, 1992. Tibbetts, Gerald R., “The Role of Charts in Islamic Navigation in the Indian Ocean”, in David Woodward, J. Brian Harley (eds.), History of Cartography: Islamic Tradition, vol. 2, book 1, chapter 13, pp. 256–262, Chicago: University of Chicago Press, 1992. Vanoli, Alessandro, La migliore divisione per la conoscenza delle regioni, in I cammini dell’Occidente: Il Mediterraneo tra i secoli IX e X, Padova, Cleup, 2001. Wan, Sitong, Rulin Zongpai 儒林宗派 (Confucianism History), 11, Song, Siming Congshuben, 1935. Williams, T., “The City of Sultan Kala, Merv, Turkmenistan: Communities, Neighbourhoods and Urban Planning from the Eighth to the Thirteenth Century”, in Bennison, A. K. & A. Gascoigne (eds.), Cities in the Pre-Modern Islamic World: The Urban Impact of Religion, State and Society, pp. 42–62, London: Routledge, 2007. Williams, T, Ancient Merv: Queen of Cities, World Heritage 24, 4–15, UNESCO, 2002. Yan, Ling, “Du Lingwai daida zhaji 读《岭外代答》札记”, (Notes on Lingwai daida) in Journal of the Graduates, 19.4, Sun Yat-Sen University, 1998, pp. 51–56. Yang, Wuquan (ed.), Zhou Qufei (author), Lingwai daida jiaozhu 岭外代答校注 (Annotations on “Notes from the Land Beyond the Passes”), Beijing: Zhonghua Shuju, 1999. Yang, Wuquan, “Zhou Qufei yu Lingwai Daida 周去非与《岭外代答》”, (Zhou Qufei and his 《Lingwai daida》), in Zhongnan Minzu Xueyuan xuebao, 2 (1994): 80–84.

Ming QingDownloaded Yanjiu from Brill.com09/30/2021 21 (2017) 1–43 07:52:51AM via free access Identifying the Country of Meilugudun 43

Yaqut, Moʿjam al-boldān, 5 vols., Beirut, 1955–57. Yu, Beishan, Fan Chengda nianpu 范成大年谱, (Biography of Fan Chengda), Shanghai guji chubanshe, 1987. Yuanshi, Zhonghua shuju, Beijing, 1976. Zavyalov, V. A., “The Fortifications of the City of Gyaur Kala, Merv”, in Cribb, J. and G. Herrmann (eds.), After Alexander: Central Asia before Islam, pp. 313–29, Oxford: Oxford University Press, 2007. Zhang, Hui 张晖, Songdai biji yanjiu 宋代笔记研究, Huazhong Shifan daxue 华中师范大学, Beijing, 1993. Zhang, Xinglang, Zhongxi jiaotong shiliao huibian 中西交通史料汇编, Zhonghua shuju, Beijing, 2003. Zhang, Yijun, Du Huan, Jing Xing Ji, Zhonghua Shuju, Beijing, 2000.

Ming Qing Yanjiu 21 (2017) 1–43 Downloaded from Brill.com09/30/2021 07:52:51AM via free access