Imagined Imperial Tribute: Descriptions of Foreign Lands in the “Barbarians” Sections of Late-Ming Dynasty Daily-use Encyclopedias

Hsu Hui-lin (National Taiwan University)
Translated by Michael Day

Originally published as: Hsu Hui-lin 許暉林 (Xu Huilin). “Chaogong di xiangxiang: wan Ming riyong leishu <Zhuyi men> di yiyu lunshu” 朝貢的想像：晚明日用類書《諸夷門》的異域論述in Zhongguo wenzhe yanjiu tongxun 中國文哲研究通訊 (Newsletter of the Institute of Literature and Philosophy, Academic Sincia) 20.2 (2010): 169-192.

Introduction

The term “daily-use encyclopedia” (“riyong leishu”) was first introduced in the 1950s by Japanese scholar Sakai Tadao. It came to be widely used in the academic world to refer to popular reading material prevalent during the Ming and Qing dynasties that consisted of snippets of classified and compiled knowledge on various aspects of daily life.¹ Because such encyclopedias are rich in information on the daily lives of common people, scholars have used them to reconstruct the history of daily life in late imperial China.² However, if we define such encyclopedias as practical manuals providing everyday knowledge, we inevitably encounter interpretive difficulties when we investigate certain sections of such works. The “Barbarians” (“Zhuyi” 諸夷) sections are one example. In addition to concrete geographical information on the local folk customs of foreign lands, these sections are interspersed with numerous fantastic tales of foreign countries from the ancient Classic of Mountains and Seas (Shanhai jing 山海經).³ The nature of these sections obviously differs greatly from what is generally understood as practical knowledge.

The purpose of this essay is to attempt to understand the historical cultural significance of the “Barbarians” sections of late-Ming dynasty (1368-1644) popular encyclopedias in terms other than “everyday practicality.” I focus on the daily-use encyclopedias of this particular period because it was at this time — the late sixteenth and early seventeenth centuries — that these encyclopedias came to be produced and widely distributed in large numbers. This was also just the time that western world geographical knowledge was introduced to China, so we will be able to compare that knowledge with the treatment of foreign lands in the late-Ming daily-use encyclopedias.

At present, the academic discourse surrounding the “Barbarians” sections of late-Ming daily-use encyclopedias has focused on the ways in which they mingle concrete geographical information with fantastic mythical tales. In general, these encyclopedias are comprised of two registers of text and pictures, one on the bottom and one on the top of the page. In the “Barbarians” sections, the top register generally presents ancient mythical creatures described in the Classic of Mountains and Seas under the title “Strange Creatures of the Mountains and Seas”; the bottom register, under the title “Miscellaneous Records of Foreign Barbarian” describes foreign lands, including actual countries such as Japan, Goryeo [present day Korea], and Chenla [in present day Vietnam], in addition to imagined countries that exist only in the Classic of Mountains and Seas, like the Country of Dwarves, the Country of the One-Eyed, and the Country of Winged People.

Wang Cheng-hua believes that these legendary creatures and fantastic monsters from a wide range of places are expressions of the “curiosity” of the people of the Ming dynasty.⁴ Alternatively, Miura Kunio suggests that the mythical elements of the “Barbarians” sections represent a unique Ming worldview in which reality and imagination are combined.⁵ Wu Huey-fang, however, holding that the mixture of reality and imagination in the “Barbarians” sections of the encyclopedias was already apparent in the Yuan dynasty (1279-1368) daily-use encyclopedia Expansive Record of the Forest of Affairs (Shilin guangji 事林廣記), argues that it does not necessarily represent a mentality unique to the people of the Ming.⁶ To be sure, as Wu points out, Wang and Miura have not considered the “Barbarians” sections of late-Ming daily-use encyclopedias in the context of the longer history of encyclopedia production. However, Wu’s criticism itself requires further discussion. Firstly, Wu’s investigation of the phenomenon of “the mixture of reality and imagination in ‘Barbarians sections’” focuses mainly on daily-use encyclopedias, and does not discuss the earlier tradition of encyclopedias. Additionally, when criticizing the term “the uniqueness of the Ming people,” Wu does not proceed further to differentiate the mixture of reality and imagination in Expansive Record from that of the “Barbarians” sections of late-Ming daily-use encyclopedias. One purpose of this paper is to fill in the gaps in the approaches described above.

Prior to addressing the “Barbarians” sections of late-Ming daily-use encyclopedias, it is necessary to consider the relationship between the mythological and legendary countries in the Classic of Mountains and Seas, and “barbarians” as a category of information. “Barbarians” was first firmly established as a category of information in the biographies of barbarians included in the orthodox dynastic histories.⁷ In these biographies, we find no links between the Classic and accounts of foreign peoples. Indeed, the very first dynastic history, the first century BCE Records of the Grand Historian (Shiji 史記), expressed explicit reservations about the fanciful foreign peoples described in that text. For instance, the Grand Historian mentions that Zhang Qian and other envoys were sent to Daxia and, though successful in tracing the Yellow River to its source, failed to find the legendary 800-mile-high Kunlun Mountain, stating clearly “Therefore, what the Book of Documents states about the mountains and rivers of the nine ancient provinces of China seems to be nearer the truth, while I cannot accept the wonders recorded in the Basic Annals of Emperor Yu or the Classic of Mountains and Seas.”⁸ Later official historians concurred with this judgement.⁹

Indeed, prior historical works have diverged in their categorization of this work. The History of the Han classifies it as a work on geomancy; the History of the Sui Dynasty (Sui shu 隋書), the Old History of the Tang Dynasty (Jiu Tang shu 舊唐書), the New History of the Tang Dynasty (Xin Tang shu 新唐書), and General Catalogues of the Academy for the Veneration of Literature (Chongwen zongmu 崇文總目) all list it as a work of historical geography; the History of the Song Dynasty (Song shi 宋史) classifies it as a classic work on the five elements. Later, in the Ming dynasty, Hu Yinglin 胡應麟 (1551-1602) regarded it as “a progenitor of fantastic tales ancient and modern”¹⁰; and, in the Qing (1644-1911), the authoritative Complete Collection of the Four Treasuries (Siku quanshu 四庫全書) classified the text as a classic work of fantastic literature, opining that it is “in fact the oldest example of fiction.”¹¹ These differences in classification are, on the one hand, due to the varied, complex content of the Classic of Mountains and Seas, a work that is inherently difficult to categorize, and on the other, to changes in readers’ categorization standards over time. And we should note that, although the Classic was categorized as a work of historical geography in the Tang (618-907) and Song (960-1279) dynasties, it was still not considered to be an appropriate source for writings on foreign lands in the orthodox dynastic histories. It is not difficult to understand why: the notion of “barbarian” as it is used in the histories, derives from the tributary relationships long established between China and other peoples (“barbarians”). There was never any indication, however, that it was possible to interact with the strange creatures of the Classic of Mountains and Seas within the context of a tribute system.

The Classic of Mountains and Seas, although either ignored or repudiated by the compilers of the orthodox dynastic histories, did at last find a home in encyclopedic writings on “barbarians.” The strange creatures of the Classic were included, apparently for the first time, in the official Song-era encyclopedia Imperial Readings of the Taiping Era (Taiping yulan 太平御覽), compiled between 977 and 983. That is, the inclusion of the strange creatures from the Classic in the “Barbarians” sections of daily-use encyclopedias of the Ming was in fact a continuation of an earlier encyclopedic tradition, not a unique feature of these encyclopedias. Thus, the question is not simply why the creatures of the Classic appear in the late-Ming works, but why these works chose to follow the course set by Imperial Readings — despite the judgment of the orthodox histories — and include portions of the Classic in their sections on barbarians. The process of inclusion involved alterations in the form of the information adopted from the Classic. We will inquire into the social and cultural significance of these alterations in the late-Ming context.

This paper is divided into three main sections, which attempt to answer the questions raised above. Encyclopedias, by their nature, are compiled through classification and define categories of information. Therefore, in the first section of this paper, I will analyze how the changes in compilation methods influenced the formation of the “Barbarians” section in daily-use encyclopedias by comparing the compilation methods for the Expansive Record of the Forest of Affairs (Shilin guangji 事林廣記), a Yuan-dynasty daily-use encyclopedia, with those of the late-Ming daily-use encyclopedias. Secondly, while “Barbarians” is a traditional encyclopedia category, we still have to confront the question of how this category of information contributed to the encyclopedias’ stated goal of providing practical guidance for everyday life. By reconsidering the concept of “practicality,” I will probe the significance of the “Barbarians” sections of the encyclopedias in the sociocultural context of the late-Ming dynasty — a time when the tribute system was in fact breaking down and there was therefore little need for “daily-life” information about foreign lands. Finally, by examining late-Ming world maps, this paper will inquire into how the images of foreign lands in the encyclopedias influenced the way in which China received western geographic knowledge of the world. In sum, this essay will attempt to provide a fuller picture of cultural history than the two-dimensional, binary framework of concrete geographical information versus fanciful fabrications that past scholars have traditionally used when discussing the “Barbarians” sections of late-Ming daily-use encyclopedias.

From Expansive Record of the Forest of Affairs to the Comprehensive Compendium of Myriad Treasures

Although daily-use encyclopedias only truly came to be published in large quantities during the late-sixteenth and early-seventeenth centuries, the earliest extant daily-use encyclopedia is the Yuan dynasty version of Expansive Record of the Forest of Affairs, compiled by the Southern Song period scholar Chen Yuanjing 陳元靚.¹² Versions of the Expansive Record with additions, supplements, deletions, and revisions were constantly in circulation throughout the Southern Song, Yuan, and Ming dynasties. In this section, I hope to gain an understanding of the unique characteristics of the descriptions of foreign lands in the “Barbarians” sections of the late-Ming dynasty wanbao quanshu by comparing the compilation methods of the “Barbarians” sections of the different editions of the Expansive Record and various Ming dynasty daily-use encyclopedias, and by clarifying how the changes in compilation methods shaped the information on foreign lands in the encyclopedias from the Yuan to the late Ming. In the course of this comparison, I will examine the contrasts between different versions, the borrowing and copying of the text describing foreign lands, and even the unique layout and design of the encyclopedias. Only by taking the physical aspects of the manufacturing of these books into account can we properly understand the significance of late-Ming daily-use encyclopedias’ discussions of barbarians in Chinese cultural history.

As previously stated, the first inclusion of the strange creatures of the Classic of Mountains and Seas in the “barbarians” category can be traced to the late tenth-century Imperial Readings of the Taiping Era. In a section titled “Barbarians of the Four Directions”, it treats barbarians from the east, south, west, and north, in that order, drawing on materials from a range of sources, including biographies of barbarians in the orthodox dynastic histories, records of local customs and traditions, and anecdote collections (biji 筆記), as well as works such as the Classic of Mountains and Seas, Classic of Divine Marvels (Shen yi jing 神異經, 2^nd c. BCE), Record of Foreign Matters (Yiwu zhi 異物志, early 3^rd c.), Map of Foreign Countries (Waiguo tu 外國圖). Entries drawn from the latter are collected mainly in the final part of the “Southern Barbarians” section. It is apparent that the compilers gave priority to reliable information from the orthodox dynastic histories in arranging the entries. Early alternate editions of the Expansive Record also generally follow this approach.

The Expansive Record was published during the late Song dynasty, but no block-printed editions from the Song have survived. Three different lineages of Yuan dynasty versions are currently extant, those deriving from the 1325 woodblock edition reprinted in Japan in 1699 (hereafter referred to as the 1325 edition); from the New Edition, with Illustrations and Additions, of the Encyclopedic Expansive Record of the Forest of Affairs (Xinbian zuantu zengxin qunshu leiyao shilin guangji 新編纂圖增新群書類要事林廣記) of 1330-1333 (referred to below as the 1330s editions); and from the 1340 Encyclopedia Record of the Forest of Affairs, with Illustrations and Additions (Zuantu zengxin qunshu leiyao shilin guangji 纂圖增新群書類要士林廣記). The numerous Ming dynasty reprints are all new block-printed editions of one of the above lineages.¹³ As the contents of the “Barbarians” sections in the 1330s and 1340 editions are exactly the same, I will compare the 1325 editions to the 1330s editions.

In the 1325 Expansive Record edition, information on barbarians is presented in fascicle 8 under the title “Miscellaneous Records of Island Barbarians.” The content focuses mainly on countries that can be reached by sea from Chinese ports, including Champa, Panduranga, Chenla, and Srivijaya to China’s southwest, and extending as far as the Arabic empire and Sindhu; they are listed in order of increasing distance from China. ¹⁴ The entries for countries near southwest China list wind directions for voyages by boat and travel distances from Guangzhou. For instance, the entry for Champa reads, “The kingdom of Champa lies across the ocean to the southwest. Champa is a short distance from Guangzhou, taking eight days to reach with favorable wind.” The entry for Chenla states “Traveling by boat with the aid of the north wind, the journey takes ten days.” However, with regard to more distant lands not easily reached by boat from China, no information is provided. These entries make it apparent that the compilers based their writings on actual ocean voyages, and thus the lands closer to China are described in greater detail, while there is a relative lack of information on more distant lands.¹⁵ At the bottom of the “Miscellaneous Records” section, the compilers of the 1325 edition note that “This document is based on official Guangzhou ship records,” likely referring to the ocean voyage records of the Southern Song Guangzhou Seaport Office. Many parts of “Miscellaneous Records” were copied word for word from the Record of Foreign Matters (Zhufan zhi 諸蕃志)¹⁶ written by Song dynasty historian and official Zhao Rukuo 趙汝适 (1170-1228) during his tenure at the Quanzhou Seaport Office. Fascicle 9 of the 1325 edition, under the title “Strange Creatures of the Mountains, Seas, and Clouds,” presents various strange creatures from the Classic of Mountains and Seas; and then, fantastic humanoid creatures — immortals, people with giant ears, three-headed people, etc. —from the Classic. Thus the 1325 edition of the Expansive Record distinguishes actual countries from countries populated by legendary nonhuman or humanoid creatures.

Figure 1 : From “Miscellaneous Records of Island Barbarians,” New Edition of the Encyclopedic Expansive Record of the Forest of Affairs ) Xinbian qunshu leiyao Shilin guangji 新編群書類要事林廣記 (1699 Japanese reprint of 1325 block-printed edition)

Between the somewhat later 1330s edition and 1340s edition of the Expansive Record, compilation methods changed. In these two later editions, the corresponding section is entitled “Miscellaneous Records of Tributary Lands.” Departing from the 1325 edition, this section adds, after all the lands listed in the “Miscellaneous Records” section of the 1325 edition, ten-odd nonhuman kingdoms listed in works such as the Record of Divine Marvels, Record of Wide-ranging Knowledge (Bowu zhi 博物志, late 3^rd c.), Miscellaneous Tales from Youyang (Youyang zazu 酉陽雜組, 9^th c.), Records of Guangzhou (Guangzhou ji 廣州記, 4^th-5^th c.). The humanoid creatures from the Classic of Mountains and Seas featured in the 1325 edition are added at the end. That is, in the 1325 editions, actual foreign lands that can be reached by sea are separated from the humanoid (and non-human) creatures of the Classic of Mountains and Seas, Record of Divine Marvels, and Record of Wide-ranging Knowledge, but in the 1330s editions and the 1340 edition, they are grouped together. This difference in editions highlights the compilers’ differing images of foreign geography. In the 1325 edition, we see that human and nonhuman lands do not share either the same textual or geographic space. In the earlier editions, the nonhuman lands from the Classic are placed in a separate section, in a distinct mythological space. In fact, the original text of the Classic refers to these nonhuman lands as “guo” (“countries” 國), as in the Country of the Immortals, the Country of the Giant-Eared, the Country of the Three-Headed, and so on. While the descriptions of these countries in the 1325 edition of the Expansive Record are copied from the Classic, the titles of the entries refer to the “ren” (people人) instead of “guo,” emphasizing the beings rather than the lands they inhabit. This may be a subtle change, but it is maintained from beginning to end, suggesting that the compilers made the change in a deliberate, systematic way. The change emphasizes the absolute differences between varieties of creatures, not merely their cultural differences. In other words, these strange creatures of myth and legend do not comprise “countries.” This change is consistent with the compilers’ placement of nonhuman creatures in a section separate from that on real countries and peoples.

As previously stated, the 1330s edition groups both sections — that including fantastic humanoid creatures and that listing real countries — together in a single section. It would be overly simplistic, however, to understand this move as a mingling of reality and fantasy. To begin with, it is important to note that while the 1330s edition places the nonhuman creatures of the 1325 edition along with real human countries, the compilers of this edition also change the titles of the entries so that they again refer to “countries” instead of “people.” However, this does not indicate a return to the original text of the Classic of Mountains and Seas. In fact, this change is in accordance with the logic behind the compilation of the relevant section of the 1330s edition. This section has a multilayered structure: the first part presents actual countries found in official government records, the middle part reproduces the content of works of the recent past about legendary countries made up of both human and nonhuman creatures, and the final part presents nonhuman creatures from the ancient Classic of Mountains and Seas. These compilation methods reflect a traditional image of geography: lands nearer to China are seen as normal and populated by humans, while lands further from China are seen as abnormal and populated by nonhuman creatures. Of course, this accords with the traditional moralistic image of geography involving a Sinocentric world order. It must also be pointed out, however, that the three portions of this section are based on, from first to last in order, Song dynasty ocean voyage records, books of marvels from the third through the ninth centuries, and finally the very early Classic of Mountains and Seas. That is, the structure of the section proceeds from nearest to farthest in terms of geography, from human to nonhuman in terms of type of creature, and from most recent to most ancient in terms of textual source. Not only do these compilation practices transform the untamed humanoid creatures of the 1325 editions into geographical foreign countries, they also represent the concept of a single worldwide cultural order unifying non-Chinese cultures and nonhuman creatures. At the same time, they reorganize geographical knowledge from recent to ancient and situate it in a shared contemporary geographical and textual space.

With the exception of the revised editions of Expansive Records of the Forest of Affairs, the vast majority of Ming daily-use encyclopedias were published late in the dynasty, during the late-sixteenth and early-seventeenth centuries. The layout of these works is quite different from that of the Expansive Record. To reduce costs, in many cases Ming publishers (particularly in Jianyang, Fujian, an important commercial publishing site noted for its popular imprints) employed a layout that separated book pages into upper and lower registers, with each register presenting information on different topics. In the “Barbarians” sections of daily-use encyclopedias, not only was the traditional layout of vertical text on every page refashioned into pages divided into upper and lower registers, but illustrations were also inserted into the upper and lower registers for each entry. For instance, in one wanbao quanshu, the upper register of the “Barbarians” section is “Strange Creatures of the Mountains and Seas” and the bottom register is “Miscellaneous Records of Foreign Barbarians” (see Figure 2). This division appears at first glance to copy the separation between “Strange Creatures of the Mountains and Seas” and “Miscellaneous Records of the Island Barbarians” in the 1325 editions of the Expansive Record. However, it should be noted that, unlike the 1325 editions of the Expansive Record, in the wanbao quanshu, nonhuman creatures are not placed in the upper “Strange Creatures of the Mountains and Seas” section, but alongside human countries in the lower “Miscellaneous Records of Foreign Barbarians” section. For instance, the Country of Winged People of the Classic of Mountains and Seas is placed alongside the country of Pekalongan in the north of Java.¹⁷ There are countless such examples throughout late-Ming daily-use encyclopedias. This arrangement differs from both the 1325 and the 1330s editions of the Expansive Record of the Forest of Affairs; the former distinguished actual countries from humanoid and non-human creatures, and the latter placed humanoid creatures after accounts of (real) tributary lands in order to illustrate the concept of a stratified, Sinocentric, cultural order. In contrast, the wanbao quanshu mingle actual countries with nonhuman countries.

Figure 2: “Barbarians,” Complete Book of Myriad Treasures in Marvelous Brocade (Miaojin wanbao quanshu 妙錦萬寶全書 (1612)

This choice is best seen, however, not as a reversal of the practices of earlier daily-use encyclopedias, but rather as the result of compilers’ assimilation of contemporary intellectual discourse on foreign lands into the new encyclopedias. Miura Kunio points out that the “Barbarians” sections of the wanbao quanshu and other late-Ming daily-use encyclopedias seem to follow various versions of Record of Naked Creatures (Luochong lu 贏蟲錄; later retitled Yiyu zhi 異域志or Record of Different Regions ) by Zhou Zhizhong 周致中 (fl. 14^th c.) of the Yuan dynasty, with some changes to the order of the countries.¹⁸ Ge Zhaoguang observes that Record of Naked Creatures already, well before the late-Ming encyclopedias, mixes “real” countries reachable from China by sea with the nonhuman countries of the Classic of Mountains and Seas.¹⁹ The mixture of concrete geographical information and strange nonhuman creatures of the Classic of Mountains and Seas found in the “Barbarians” sections of late-Ming daily-use encyclopedias is not a unique feature of these works, but rather a continuation of a tradition in travel logs of the late Yuan and early Ming.

However influential this text may have been, the “Barbarians” sections of the late-Ming daily-use encyclopedias are not merely illustrated versions of Record of Naked Creatures. The unique page formatting requirements of the daily-use encyclopedias make the mixture of concrete geographical information with the fantastic accounts in the Classic of Mountains and Seas more conspicuous. Ideally, in works with two registers per page, the text on each register should come to an end on the same page. In actual practice, however, there are many cases in which the upper register has concluded while the lower has yet to come to an end (or vice-versa). Therefore, to even out the layout, content that ought to appear in the lower register is often shifted to the one above. For example, in Santai’s Infinitely Useful [Guide to] How to Do Everything Correctly for the Convenient Consultation of All the Four Classes under Heaven, Newly Cut (Xinke tianxia simin bianlan Santai wanyong zhengzong 新刻天下四民便覽三台萬用正宗, 1599), in order to accommodate layout requirements, not only is the original two-register layout changed to a layout with three or four equally sized registers, blurring the boundaries between the “mythical creatures” and (real) “barbarians” originally separated in the two-register layout, but the human countries that ought to appear in the lower “barbarians” register are also shifted to the upper register under the title “Strange Phenomena of the Classic of Mountains and Seas” (see Figure 3). Thus, the mixture of reality and fantasy, as Miura puts it, appears not only in the lower “Miscellaneous Records of Barbarians” section, but also at times crosses over to the top.

Figure 3: “Barbarians,” Santai’s Infinitely Useful [Guide to] How to Do Everything Correctly (1599)

The above comparison of compilation practices from the various editions of the Expansive Record of the Forest of Affairs to the daily-use encyclopedias of the late-sixteenth and seventeenth century shows that, from the late-Southern Song dynasty to the late-Ming dynasty, differing approaches were taken to compiling the “Barbarians” sections of daily-use encyclopedias: at first, concrete geographical information was strictly separated from the descriptions drawn from the Classic of Mountains and Seas; later, stratified compilation practices situated the nonhuman countries of the Classic at the outer periphery of geographical and textual space; and finally the categories mingled chaotically, blurring all boundaries.

A Form of Imagined Imperial Tribute

Ge Zhaoguang has asked: “Why, in spite of the fact that the Chinese had been acquiring an increasing amount of factual knowledge of foreign lands from the Yuan and Ming, were images of foreign lands consistently drawn from much earlier classical speculations and understandings?”²⁰ Amending this question to apply to the late-Ming daily-use encyclopedias, we can ask: Why do the encyclopedias not only mingle concrete geographical knowledge with the ancient — and fantastical — knowledge of the Classic of Mountains and Seas, but also conspicuously highlight this mingling?

Before answering this question, however, it is necessary to examine in greater detail the significance of the creation of this “fantastical” knowledge. To divide the countries described in the “Barbarians” sections of late-Ming daily-use encyclopedias into real and imaginary is simplistic. In fact, the information in the encyclopedias that is not drawn from factual accounts of foreign lands does not necessarily originate in mythical works or records of the strange like Classic of Mountains and Seas or In Search of the Supernatural (Sou shen ji 搜神記, mid-4th c.). In fact, many of these lands — Karasahr, the Uyghur Khaganate, and the Roman Empire — appear frequently in historical records; the only reason they appear fantastical is that they were no longer in existence in the Ming. These countries of the past, the imagined countries of Classic of Mountains and Seas, and the actual countries described in travel logs like Record of Foreign Matters are all mingled in the “Barbarians” sections of daily-use encyclopedias. For instance, the Complete Book of Myriad Treasures in Miraculous Brocade (Miaojin wanbao quanshu 妙錦萬寶全書) places the Country of the Impalers from the Classic of Mountains and Seas beside the “real” Tang-era western country of Karasahr.

The insertion of historical kingdoms into the “Barbarians” sections of late-Ming daily-use encyclopedias indicates that the “mingling” of information in these sections should not be seen merely as fact coexisting with fantasy. Rather, these sections adopt an informational framework that breaks down the boundary between real and fabricated, simultaneously presenting geographical knowledge from different time periods and adopting an expositional method in which countries interlink, interpenetrate, overlap, and interlock across textual and geographic space. The compilers were not overly concerned with whether these historical countries had disappeared by the late-Ming or whether they were anywhere to be found in contemporary geographic space. Looked at in this light, it is highly doubtful whether the compilers of daily-use encyclopedias saw the fantastic foreign creatures of Classic of Mountains and Seas as imaginary and the concrete geographic information as “real” — or whether the coexistence of concrete geographical information and information from the Classic of Mountains and Seas can be understood as a mixture of fact and fantasy. Perhaps the compilers of late-Ming daily-use encyclopedias did not regard the Classic as a “fantastical” or fabricated — as opposed to a factual — text, but simply as a component of ancient geographical knowledge. They present geographical knowledge in an entirely different way from the 1330s edition of the Expansive Record of the Forest of Affairs, using a distinctive method of compilation to organize ancient geographical knowledge.

A reader here might reasonably ask, “Why was it so important for daily-use encyclopedias to find a place for ancient geographical knowledge?” Of what practical “daily life” use was this knowledge? But, if, as Wang Cheng-hua points out, encyclopedias by their very nature collect and categorize different types of information, ²¹ the key questions become, rather: how did late-Ming daily-use encyclopedias situate this knowledge of foreign lands? What are the differences — and the significance of the differences — between their methods and those of earlier encyclopedias? The ways in which the “Barbarians” sections of late-Ming daily-use encyclopedias situate ancient geographical knowledge reveal the “everyday pragmatization” of ancient geographical knowledge in the late Ming.

Here Shang Wei’s perspective on the practical nature of daily-use encyclopedias is worth considering. He believes that these encyclopedias are constituted not of faithful depictions of daily life in the Ming dynasty, but of discourses on the meaning of daily life. As the encyclopedias were continually reprinted, revised, and supplemented, this discourse shaped readers’ imagination of social life.²² Here, I have adapted Shang Wei’s perspective by going a step further, taking the view that the encyclopedias are practical in that they discuss daily pragmatic matters — that is, they profess to contain verifiable, actionable information that will achieve certain results and experiences that can be reproduced. For instance, the “Medical Learning” (“Yixue” 醫學) sections of Yuan and Ming daily-use encyclopedias often describe the immediate treatment effects of ancient medicinal formulas as “instantly effective upon administration,” “certain not to fail,” and “assured to be effective.” Moreover, when revised editions of these works were published, they always contained new and updated ancient formulas. While medicines said to be “effective” for particular conditions were numerous and constantly changing, the reliability of the formulas and the reproducibility of the treatment methods were constantly emphasized. That is, daily-use encyclopedias are practical not only in that compilers selected (or at least claimed to select) practical information for use in daily life, but also because they edited the entries to make it easy for every reader to understand information that was vague or hard to verify in the original source texts the encyclopedia compilers drew on. The “Jing Chen’s Chart” (“Jing Chen zhi tu” 競辰之圖) entry in the Expansive Record of the Forest of Affairs is a classic example. As Stephen West points out, “Jing Chen’s Chart” lays out a linear schedule that, using quantitative methods, provides a complete preparatory plan for imperial civil service examination candidates. This examination preparation schedule makes information objective, so that the acquisition of knowledge required for the imperial civil service examination no longer depends solely upon readers’ quick wits, but upon each individual’s replication of the specific steps necessary to prepare for the examinations.²³ To provide another example: to the readers of late-Ming daily-use encyclopedias, particularly those who were traveling merchants, information about travel routes was of vital importance, but also difficult to represent. The “Geography” (“Diyu” 地域) sections of late-Ming daily-use encyclopedias generally contain route guides in entries like “Travel Routes to the Two Capitals and Thirteen Provinces” and “Jade Mirror of Empire-wide Travel Routes”; these provided travelers with the names of places passed on the way from the capitals to the provinces, distances between places, and even information on pitfalls and dangers to be aware of while traveling. “Travel Route Songs,” versified summaries of this information in easy to memorize formulas, were appended to the ends of the route guides. Daily-use encyclopedias are practical not because their content can in fact be put into practice, but because they repeatedly emphasize and guarantee that particular goals (both abstract and concrete) can be attained through following a sequence of certain steps, using certain tools, within a certain time span. From this point of view, we should cease trying to explicate the relationship between “Barbarians” sections and actual daily life; rather, we should ask how the compilers of encyclopedias created and organized the contents of the “Barbarians” sections in accordance with this principle of daily life usefulness.

As the compilers excerpted the descriptions of foreign lands by scholars from the Yuan onwards in the “Barbarians” sections of the encyclopedias, they simultaneously edited the information to make it more practically useful. One of the most conspicuous manifestations of this transition to greater usefulness is the addition of the distances between China and the countries of the “Barbarians” sections; the late-Ming encyclopedias list the time required to complete such trips by specific transportation methods. They frequently state that one country or another was “X years and X months to Yingtian Prefecture²⁴ [on horseback].” To be sure, this was not the first time distances had been included, for both the Expansive Record of the Forest of Affairs and the Record of Different Regions often noted the distance between Chinese capitals, major cities, and “barbarian” countries. But, in these relatively early texts, distances appear to be indicated only for countries like Champa and Chenla that could be reached by sea from China. By contrast, the “Barbarians” sections of daily-use encyclopedias published in the late Ming began to indicate distances for an extremely large number of countries, including distant and legendary, human and nonhuman, countries. For instance, the Five Carts of Select Brocade for the Profitable Use and Convenient Consultation by All under Heaven (Tianxia simin liyong bianguan wuche bajin 天下四民利用便觀五車拔錦, 1597) states that the Country of Dogs, comprised of men with dogs’ heads and human bodies and Chinese women, is “two years to Yingtian Prefecture on horseback”; that the Country of Long-Haired People, who live in caves and have hair all over their faces and bodies, is “two years and ten months to Yingtian Prefecture on horseback”; that the Country of Monkey Men, who wage war to defend their homeland, is “three years to Yingtian Prefecture”; and that the Country of Galloping Spirits, where people grow hair from the knee down, have hooves, and are skilled runners who can travel one hundred miles by foot in the span of a day, is “two years by horse to Yingtian Prefecture.” Santai’s Infinitely Useful [Guide to] How to Do Everything Correctly contains similar descriptions. Assembled Illustrations of the Three Realms: Heaven, Earth, Man (Sancai tuhui 三才圖會)，compiled in 1607 by Wang Qi 王圻 (1530-1615) and published in 1609; the 1607 Gems of Learning from the Expansive Record of the Hanlin Academy Collected for the Use of All Four Classes of People, Newly Published with Additions and Revisions (Xinkan hanyuan guangji buding simin jieyong xuehai qunyu 新刊翰苑廣記補訂四民捷用學海群玉); and Orthodox [Guide to] a Myriad [Categories] of Useful Information for Scholars and Commoners Compiled at the Pavilion of Revered Culture [So That They] Need Not Seek the Advice of Others, Just Newly Cut (Dingqin Chongwen ge huizuan shimin wanyong zhengzong buqiuren 鼎鋟崇文閣彙纂士民萬用正宗不求人, 1607) use the same descriptions as well, but in a slightly different order. Other than the exclusion of the Country of Long-Haired People, the Marvelous Book of Myriad Treasures from the Forest of Literati for the Use of All, Newly Edited and Fully Supplemented (Xinban quanbu tianxia shiyong wenlin miaojin wanbao quanshu 新板全補天下便用文林妙錦萬寶全書, 1614) is exactly the same. It is worth noting that the Profound Collection of Marvelous Works from the Forest of Literati for the Consultation and Convenient Use of Scholars, Completely Supplemented and Newly Published (Xinkan quanbu shimin beilan bianyong wenlin huijin wanshu yuanhai 新刻全補士民備覽便用文林匯錦萬書淵海, 1610) contains no record of the Country of Dogs or the Country of Galloping Spirits, but does retain the Country of Monkey Men and the Country of Long-Haired People, and adds a description of the Country of Bird People, “where people have three claws,” which is “five months by horse to Yingtian Prefecture.” Daily-use encyclopedias published prior to 1610, however, all contain entries on the Country of Bird People, but do not include any information about its distance from China; this information was deliberately added to the Profound Collection of Marvelous Works from the Forest of Literati, suggesting that markers of distance from China (e.g., “X months by horse to Yingtian Prefecture”) were a means of granting special significance to the country described. Such descriptions reflect the idea that if these strange, distant, unknowable countries could be reached by particular transportation methods within a certain number of days, their relationship to China could be made concrete and experiential.

Figure 4: “Barbarians,” Profound Collection of Marvelous Works from the Forest of Literati for the Consultation and Convenient Use of Scholars, Completely Supplemented and Newly Published, 1610 woodblock edition

When we recall as well the previously noted tendency of late-Ming daily-use encyclopedias to include ancient geographical knowledge, it is clear that these descriptions of foreign lands formed a discourse that had both a historical dimension and a practical character: countries a short distance from China by sea, remote countries of legend, the humanoid creatures of the Classic of Mountains and Seas, and vanished countries of the past, although not noted in route guides, were made accessible to one another, by boat or by horse, through the jumbled layouts of the “Barbarians” sections of the encyclopedias. At the same time, the dress and customs of these countries were often described as similar to those of some other land. For instance, the inhabitants of the Country of People with Rear-Looking Eyes are said to be “as closely unified as the Tatars.” While such descriptions serve on the one hand to make these countries seem more real, on the other, they represent attempts to place these countries within a mutually influential cultural system. This discourse, unlike that of travel logs or route guides, is not provided for practical reference by travelers. Foreign travel is clearly not the point: phrases such as “X months by horse to Yingtian Prefecture” do not describe departing from China for these foreign lands, but arriving in China from the foreign countries described.

The direction of travel described here — from a foreign country to China — assumes a tributary relationship between “barbarians” and China. In fact, use of the phrase “X months by horse to Yingtian Prefecture” continues a tradition of writing about foreign lands originating in very early biographies of “barbarians”. From the History of the Han Dynasty onward, the orthodox dynastic histories begin using descriptions such as “X miles to the capital,” “X months and X days by sea,” and “its customs resemble those of country X in describing foreign lands.” During the Song dynasty, such descriptions of “barbarian” countries were not limited to historical works, but were also employed in large-scale official encyclopedias such as Outstanding Models from the Storehouse of Literature (Cefu yuangui 冊府元龜). When sea-based transportation proliferated after the Song dynasty, it became customary to refer to “barbarian” countries by indicating their travel distance from China. Thus, the “Biographies of Foreigners” section of the early-Ming History of the Yuan Dynasty (Yuan shi 元史) explains, “[Champa] is close to Qiongzhou, one day away by boat with favorable winds” and “[Countries such as Malabar] are approximately 3000 miles from Quanzhou. From these countries, Kish Island²⁵ can be reached in approximately fifteen days by sea with favorable wind.”²⁶ That is, descriptions such as “X months by horse to Yingtian Prefecture” in the “Barbarians” sections of late-Ming daily-use encyclopedias highlight one of the unique textual characteristics of these encyclopedias: the trend toward experiential, practical descriptions. At the same time, in light of the fact that such phrases replicate descriptions in the biographies of foreigners in the orthodox dynastic histories, they also represent a kind of discourse on the tribute system.

Figure 5: “Barbarians,” Complete Book of Myriad Treasures in Marvelous Brocade (1612 woodblock edition)

In fact, descriptions such as “X months by horse to Yingtian Prefecture” are not the only expressions in the encyclopedias of the concept of the imperial tribute system. Often the “Barbarians” sections of these texts open with an illustration of foreigners entering the capital to offer tribute (see Figure 5), suggesting that imperial tribute was in fact the central theme of “Barbarians.” So, too, the “local products” from foreign countries listed in these sections indicate what was offered in tribute. Thus, in the 1607 Gems of Learning from the Expansive Record of the Hanlin Academy, the description of the Country of People with Miraculous Arms from the Classic of Mountains and Seas adds to the descriptions of the “local products” of this country, such as pearls and other rare items, in previous daily-use encyclopedias. Given the other indications that the tribute system was a theme in these works, the descriptions of local products was most likely not intended to tell Chinese merchants of opportunities for overseas commercial travel, but rather to inform readers of the kind of tribute that foreign peoples were likely to bring to China. Reflecting on the ways in which the “Barbarians” sections of the late-Ming daily-use encyclopedias reuse and reshape the information on the strange creatures of the Classic of Mountains and Seas presented in the earlier Expansive Record of the Forest of Affairs and the still earlier Imperial Readings from the Taiping Era, we can draw the following conclusions: The descriptions of foreign lands in the “Barbarians” sections of the late-Ming encyclopedias do much more than satisfy readers’ curiosity. They fuse practical everyday discourse with traditional historical depictions of barbarians, reconfiguring the relationship between China and foreign lands and emphasizing what might be called imaginary tribute brought to the Ming court by “barbarians” from across the sea.

The fixation of the “Barbarians” sections of these encyclopedias on imperial tribute is tied to the very pressing contemporary issue of the breakdown of tributary relationships. The tribute system, previously a rather loosely structured series of symbolically hierarchical relationships between China and its outlying polities, had been formalized in the early Ming dynasty through the institution of certain rituals of subordination and gift exchange. But, by late in the dynasty, this system was in decay. With the raids by southern pirates²⁷ in the mid-sixteenth century and later, and the invasions by the northern Jurchen people beginning in the late-sixteenth century, barbarians were no longer merely objects of curiosity. The Jurchen invasion not only put extreme pressure on frontier defense officials, it posed a genuine threat to the lives of ordinary people. For the people of the Fujian region, where daily-use encyclopedias were published in the greatest numbers, the pirate invasions were particularly threatening — truly a matter of life and death. When the imperial tribute system in fact did break down, and the central government could no longer ensure political and cultural order through tributary relationships, the “Barbarians” sections of daily-use encyclopedias were, in a sense, attempts by commoners to reestablish the tributary order through textual production. Commercial publishers, with their eyes on the market, supported this effort. In order to increase book sales, it was common for commercial publisher-booksellers to continually release revised and expanded editions of daily-use encyclopedias in hopes of attracting readers with the latest information. The frequent revisions and expansions also, of course, reflect the compilers’ constant collection of knowledge and information, in particular knowledge and information relevant to contemporary concerns. For the “Barbarians” sections, collection of new knowledge meant “collecting” tributary states and their products. By adding new vassal states and their local products with each revision, encyclopedia compilers were intimately involved in reimagining tributary relationships. To be sure, late-Ming daily-use encyclopedias did not construct knowledge of “barbarians” as the biographies of “barbarians” in the orthodox dynastic histories did, systematically categorizing them by north, south, east and west. However, encyclopedia compilers were engaged in the gathering, selection, and categorization of information, processes that imply an attempt to reconstruct some sort of order and structure. In this sense, curiosity and public interest in new knowledge were not the only reasons why late-Ming daily-use encyclopedias contained “Barbarians” sections or why new editions featuring new information about “barbarians” were continually released; the main reason is that, by gathering and categorizing knowledge, daily-use encyclopedia editors were attempting to construct a kind of order that no longer existed in China’s relationships with foreign lands.

By comparing the “Barbarians” sections of daily-use encyclopedias published in the late-Ming dynasty, the Qing dynasty, and later, we can further explain how the creation of this knowledge in the late Ming was the expression of a desire to maintain the tribute system. In the daily-use encyclopedias of the Qing, although phrases such as “X months by horse to Yingtian Prefecture” have not completely disappeared, they have become much less conspicuous. For instance, in the “Barbarians” section of the 1758 Comprehensive Compendium of Myriad Treasures, Newly Expanded and Published (Xinjuan zengbu wanbao quanshu 新鐫增補萬寶全書), six barbarian countries — namely Xifanguo, Bailiguo, Daluoguo, Shenlieguo, Tuma, and Sheluguo ²⁸— are described as “X months by horse to Yingtian Prefecture.” Such countries are proportionately no less prevalent than in the Ming dynasty daily-use encyclopedias, but it must be noted that this 1758 edition makes no mention of the countries of the Classic of Mountains and Seas that appear in such large numbers in late-Ming encyclopedias. We see a similar state of affairs in the 1895 Comprehensive Collection of Myriad Golden Treasures, Newly Expanded (Xinzeng xuanjin wanbao quanshu 新增懸金萬寶全書). Although the entry for the Country of the Impalers, one of the forty-seven countries listed (and one of the very few entries in this edition drawn from the Classic of Mountains and Seas), contains a description of local products (“winged horses, jiaoduan,²⁹ groundhogs, sandgrouses”), the phrase “X days by horse to Yingtian Prefecture” appears in only three entries (for Xifanguo, Bachijiguo and Buciguo), less than seven percent of all entries. This is much lower than the proportions of countries thus described in the late-Ming daily-use encyclopedias, where a minimum of 12.5% (in Profound Collection of Marvelous Works from the Forest of Literati) and a maximum of 25% (Comprehensive Compendium of Myriad Treasures in Miraculous Brocade, Five Carts of Select Brocade, Santai’s Infinitely Useful [Guide to] How to Do Everything Correctly) of entries are described in this way.

The above analysis shows how, in the process of constantly supplementing, revising, and reprinting, the compilers of late-Ming daily-use encyclopedias integrated the encyclopedias’ discourse on practicality with the descriptions of “barbarians” in the orthodox dynastic histories; linked the ancient geographic knowledge of the Classic of Mountains and Seas with the concrete geographical knowledge of the Yuan, Ming and later; and attempted to reimagine tributary relationships just at the time when the tribute system had broken down. But there is another aspect of the late-Ming cultural context that cannot be ignored: the introduction of western geographic knowledge in the early seventeenth century. By examining the world maps in popular circulation during the late Ming, the section below will inquire into how the imagined tributary relationships of the “Barbarians” sections of late-Ming daily-use encyclopedias influenced China’s reception of worldwide geographical knowledge from the West.

Daily-use Encyclopedias Encounter World Maps

On the basis of extant daily-use encyclopedias, it appears that late-Ming encyclopedias were published in large numbers in the 1590s. Slightly earlier, in 1583, Italian missionary Matteo Ricci (1552-1610) entered China through Macao in order to proselytize the Christian faith, bringing with him Western geographical knowledge of the world. Thus, it is almost impossible to discuss the “barbarian” tributary system imagined in daily-use encyclopedias without considering how the introduction of western geographical knowledge might have shaped that imaginary. In this section, I discuss how the notion of a Sinocentric tribute system absorbed the shock of western geographical knowledge structures. Other scholars have discussed in some detail the exchange of geographical knowledge between China and the West in the seventeenth century.³⁰ Here, rather, taking the historical reception of texts and their intertextuality as starting points, I would like to make some preliminary observations on the relationship between the imagined imperial tribute system of the late-Ming encyclopedias and Chinese maps, using two contemporary world maps as examples.

The earliest extant western world map showing the locations of the legendary ancient countries listed in the Classic of Mountains and Seas is Ricci’s “Complete Map of Mountains, Seas, and Lands,” published in 1602 in Feng Yingjing’s 馮應京 (1555-1606) Overview of Monthly Ordinances (Yueling guangyi 月令廣義).³¹ (This version is not to be confused with another of Ricci’s world maps, the large, six-panel stand-alone “Complete Map of All Countries on Earth,” also printed in 1602, in Beijing.) This map, like the series of world maps Ricci began to draw in 1584, conformed to the Sinocentric worldview in moving the first meridian so that China was slightly left of center on the map. It was a copy of the map of the same title that Ricci had produced in 1600 at the request of Wu Zhongming 吳中明, Director of Nanjing’s Ministry of Civil Appointments, which in turn is believed to be a copy of a map by Ricci carved on a stele in Nanchang, Jiangxi, in 1596. It was reprinted, in many different versions and under different titles, in texts and as stand-alone maps in the early seventeenth century. When Wang Qi and his son compiled Assembled Illustrations of the Three Realms in 1607, they made a further copy of this map, placing it on the front page of the geography section.³²

As Ricci’s original “Complete Map of Mountains, Seas, and Lands” has not survived, it is impossible to determine the changes the map went through as it was circulated and copied. What is clear, however, is that the contours of the 1596 and the 1602 maps are quite different. In the 1602 “Complete Map of Mountains, Seas, and Lands,” the latitude and longitude lines have been removed; and it depicts a Country of Dogs directly north of center. The Country of Dogs corresponds to the Feudal State of Dogs in the Classic of Mountains and Seas. However, Ricci’s 1596 map and the 1602 stand-alone “Complete Map of All Countries on Earth” neither make note of the Country of Dogs, nor contain a land mass with corresponding contours. Common sense allows us to conclude that Ricci is unlikely to have made such a striking addition to the 1600 “Complete Map of Mountains, Seas, and Lands” alone. Thus, we can reasonably assume that the Country of Dogs was added to this map as it circulated in such works as Overview of Monthly Ordinances and Assembled Illustrations of the Three Realms).

Figure 6: “Complete Map of Mountains, Seas, and Lands,” Overview of Monthly Ordinances (1602 woodblock edition)

In the copy of “Complete Map of Mountains, Seas, and Lands,” the Country of Dogs is not only directly north of center, but also very large compared to other countries—it is nearly as big as China. This modification to Ricci’s world map is extremely interesting. If Ricci moved the first meridian to conform somewhat to Sinocentric notions, placing the great Ming empire very close to the center in “Complete Map of Mountains, Seas, and Lands,” perhaps the modified version of Ricci’s map in Overview of Monthly Observance sent this message: China has given up its position at the exact center of the world, and the question of what to put in the empty space at the center of the map had become a pressing matter up for negotiation. The fact that the creator of this version of Ricci’s map filled the space at the center of the map with the State of Dogs from the Classic of Mountains and Seas suggests that Chinese of the late Ming drew on fantastic legends from ancient documents as ideological resources for resisting the structures of western geographical knowledge. We also see glimpses of this resistance in refutations of the geographical teachings of western missionaries by the anti-Catholic scholars of the late Ming and early Qing. Xu Dashou 許大受, Confucian scholar and contemporary of missionary Giulio Aleni (1582-1649), leveled this criticism: “There are dubious tales of great western nations whose people traveled great distances to reach our shores. If Chinese explorer Zhang Qian [fl. 2^nd c. BCE] went westward, tracing the Yellow River to its source, even reaching the Lunar Palace on the moon, how is it that no human being has ever set foot in such places [as the western nations drawn on Ricci’s map]? How is it that the Classic of Mountains and Seas, In Search of the Supernatural, Record of Vassal [States] (Xian bin lu 咸賓錄, 1591), Record of the Western Regions (Xiyu zhi 西域志, 646), and Expansive Record of the Taiping Era have not a word to say about such places?”³³ In this passage, we clearly see Xu Dashou citing the textual resources of the Classic of Mountains and Seas and In Search of the Supernatural to criticize western missionaries’ geographical knowledge of the world.³⁴ Thus, perhaps the copy of “Complete Map of Mountains, Seas, and Lands” places the Country of Dogs at the center of the map not only because of the conflict between western geographical knowledge of the world and the traditional Chinese view of the distinction between Chinese and “barbarians”, but also because the way in which foreign knowledge was replacing Chinese knowledge was a contentious subject in late-Ming world geographical discourse.

Figure 7: “Complete Map of the Nine Borders, Territorial Divisions, Human Traces, and Travel Routes” (1644 woodblock edition)

Below, by analyzing the intertextuality of late-Ming daily-use encyclopedias and world maps, I will explain how late-Ming Chinese drew on the resources of daily-use encyclopedias’ discussions of “barbarians” when faced with the challenge of western geographical knowledge. In addition to “Complete Map of Mountains, Seas, and Lands,” there is another map that includes countries listed in ancient texts like the Classic of Mountains and Seas in the format of a western world map: “Complete Map of the Nine Borders, Territorial Divisions, Human Traces, and Travel Routes” (“Tianxia jiubian fenye renji lucheng quantu” 天下九邊分野人跡路程全圖; hereafter “Complete Map of the Nine Borders”), printed in Nanjing by Cao Junyi 曹君義 in 1644 (see Figure 7). The product of a commercial publisher, this map is representative of the reception of western world maps by the general public. It is worth noting that it indicates the locations of the strange creatures of the four regions of the Classic of Mountains and Seas; in fact, it integrates the format of Ricci’s world map with the content of the “Barbarians” sections of late-Ming daily-use encyclopedias. It includes all the latitude and longitude lines, the stretch of ocean between Borneo and Australia as well as that between Borneo and the Philippines, the Arctic Ocean, and the Antarctic Ocean of Ricci’s world map, among other features. As Unno Kazutaka points out, however, the regions around China are a disjointed, miniaturized version of western world maps, and the map’s latitude and longitude lines are merely decorations. Unno makes special note of the “omission of part of the textual explanations surrounding the map.”³⁵

However, I believe that the adornments on the periphery of the map — the foreign countries on the edges and the textual explanations along its sides — are its most noteworthy features. Near the lower right periphery of the map, from north to south, we see in order notices of the Country of Women, the Country of Hairy People, the Country of the One-Eyed, and the Country of Dwarves, among other countries from the Classic of Mountains and Seas; these are all countries also listed in the “Barbarians” sections of late-Ming daily-use encyclopedias. In the panel to the left of the map, we find textual explanations of foreign countries such as, from top to bottom, Japan, Ryukyu, Xifanguo, Cochinchina, Champa, Java, Srivijaya, Malacca, Chenla, Brunei, Karakhoja, Kumul, the Almoravid empire, and Laos, as well as the Country of Red Barbarians, the Country of Erfutuo, the Country of Kezhi, Chola, the Country of Black People, Tibet, Chaoluguoe, and so on.³⁶ Not only do these countries appear in the “Barbarians” sections of daily-use encyclopedias, more importantly, the order in which they appear is nearly identical to that of the “Barbarians” entries of the 1614 woodblock edition Combined Edition of the Five Carts [of Knowledge] and the Comprehensive Collection of Myriad Treasures, Newly Cut (Xinke souluo wuche hebing wanbao quanshu 新刻搜羅五車合併萬寶全書, 1614). The text in the panel beneath the map is the “Travel Routes of the Two Capitals, Thirteen Provinces, Prefectures and Counties” invariably included in late-Ming daily-use encyclopedias. Of course, the route guide included in the map might have originated from a separate work not included in the encyclopedia, but since it is certain that Cao Junyi did make use of the “Barbarians” sections of daily-use encyclopedias when creating his map, it is very likely that the text on “barbarian” countries and travel routes at the edges of “Complete Map of the Nine Borders” is drawn entirely from contemporarily popular daily-use encyclopedias.

These country names and worldwide travel routes are not mere peripheral adornments, but a framework for viewers in both a concrete pictorial and an abstract sense. Given that daily-use encyclopedias were in wide circulation during the late-Ming dynasty, people could not have failed to notice, when consulting this world map — which appeared on the surface to be very different from traditional maps that highlighted the distinction between Chinese and “barbarians” — that the map’s periphery contained familiar travel information and content from the daily-use encyclopedias. Indeed, it is likely that viewers of the “Complete Map of the Nine Borders” saw this world map, with its latitude and longitude lines, as a graphic representation of the “Barbarians” sections of daily-use encyclopedias; and the travel route information as guides that the overseas “barbarians” of the Classic of Mountains and Seas could follow to reach China.

But here this information — explaining how long it would take to get to China from various foreign lands — signifies a vision different from the imagined imperial tribute missions in the “Barbarians” sections of late-Ming daily-use encyclopedias. In the context of Ricci’s world map, the suggestion in “Complete Map of the Nine Borders” that the overseas “barbarians” of the Classic of Mountains and Seas could reach China by following its travel route guides is significant in that it grants the nonhuman countries of the Classic the same informational authority as the western countries drawn from Ricci’s map. Thus, we must reexamine Huang Shijian’s view that the world maps of the Ming and Qing, including Cao Junyi’s “Complete Map of the Nine Borders,” incorporate Ricci’s map into a vision of Sinocentric unification that follows a firmly traditional worldview.³⁷ In fact, the above analysis of Cao’s map demonstrates that the framework by which it arranged and incorporated western geographical knowledge is not merely an abstract, general “vision of Sinocentric unification,” but also a framework for discussing the geography of China and foreign lands unique to late-Ming daily-use encyclopedias.

However, as late as the closing years of the Qing dynasty, world maps still had not fully supplanted the “Barbarians” sections of daily-use encyclopedias as the primary means of representing traditional Chinese knowledge of foreign lands. This has a great deal to do with the ways in which traditional Chinese knowledge of foreign lands was expressed. Late-Ming daily-use encyclopedias mainly communicate ancient geographical knowledge of foreign lands in narrative rather than cartographic form. Their narratives contain no quantifiable information on the positions of or distances between “barbarian” countries, their borders or their territories. These countries are described in relative rather than absolute terms. Between these countries, therefore, there are many unknowns and gray areas, and much room for extrapolation. When the people of the late Ming were forced to face the shock of western geographical knowledge, and it became necessary to reinterpret and resituate ancient geographical knowledge, the encyclopedias’ narratives filled in the gray areas, giving ancient knowledge a place and allowing it to peacefully coexist with concrete geographical information. For instance, a mythical country can be described as three years by horse to Yingtian Prefecture, and this will never conflict or clash with any other geographical description. However, such spaces cannot be found on western world maps drawn by precisely calculating the proportions of all land masses and seas. When the earth changes from a vast, unending plane with infinitely distant limits to a sphere where a long journey inevitably returns one to one’s starting point, the gray areas where the strange creatures of the ancient Classic of Mountains and Seas were said to dwell are eliminated. When the earth becomes a sphere, the space in which the “barbarians” of ancient treatises lived is flattened out. In this sense, rather than bringing knowledge of foreigners (or “barbarians”) to China, the introduction of western world geographical knowledge to China in the late-sixteenth century shrunk the space in which “barbarians” could exist. Through both narrative and pictorial means, however, late-Ming daily-use encyclopedias continued to present vivid images of numerous “barbarian” states traveling to the capital to give tribute across limitless space and limitless time.

Conclusion

This paper began by comparing various versions of daily-use encyclopedias, examining mainly the following three questions: How did Ming dynasty daily-use encyclopedias assimilate and transform the framework by which official Song dynasty encyclopedias categorized knowledge of barbarians? What was the social and cultural significance of this transformation? Secondly, from what perspective should we understand the “daily usefulness” of the “Barbarians” sections of late-Ming daily-use encyclopedias? And finally, what role did the images of foreign lands found in late-Ming daily-use encyclopedias play in the process of China’s reception of western geographical knowledge? By answering these three questions, this paper has sought to explain how the discussions of foreign lands in the “Barbarians” sections of the encyclopedias provide clues to comprehending the way the people of the late Ming imagined the world.

To begin with, the editors of Yuan-dynasty daily-use encyclopedias organized knowledge of foreign lands based on a two-dimensional distinction between fact and legend; slightly later editions adopted a multi-level organizational approach, taking concrete geographical knowledge as a starting point and gradually including the nonhuman countries of the Classic of Mountains and Seas, resulting in an imagination of foreign lands in which various beings and social strata mingled. These methods of compilation drew boundaries in order to impose a sense of order on discussions of foreign lands. Compared to these Yuan dynasty daily-use encyclopedias, the “Barbarians” sections of late-Ming daily-use encyclopedias seem jumbled and boundary-free. However, these later encyclopedias in fact emphasize that distant, mystical, unknowable foreign lands can be reached by certain transportation methods, an approach that makes their relationship to China concrete and (seemingly) experiential, overturning in a meticulous and innovative fashion the means by which past daily-use encyclopedias sought to understand foreign peoples. Through the processes of continual supplementation, revision, and reprinting, the compilers integrated the practical discussions of earlier daily-use encyclopedias with the writings on barbarians in the orthodox dynastic histories. They thereby broke down the firm cultural/textual boundaries of past daily-use encyclopedias and built a new relationship between ancient geographical knowledge like that of Classic of Mountains and Seas and the concrete geographical knowledge of the Yuan and Ming, in an attempt to reimagine tributary relationships at the very time the tribute system was breaking down. Such reimagining of the tribute system can also be understood as an alternate means of resisting the framework of western world geographical knowledge. The late-Ming appropriation and transformation of Ricci’s world map was born of the need, given the opportunities that the wide circulation of the encyclopedias provided for instruction of the public, to decide how to situate the new knowledge of foreign lands; equally important, however, was the need to reproduce, in both form and content, the imagination of foreign lands found in late-Ming daily-use encyclopedias. Thus, opposition to western geographical knowledge was expressed not only in daily-use encyclopedias, but also in world maps purportedly representing western geographical knowledge but shaped by its material expression in the encyclopedias. This essay has attempted to break free from the two-dimensional dichotomy between concrete geographical information and imaginary or fabricated information employed by past scholars in discussing the “Barbarians” sections of late-Ming daily-use encyclopedias; and to understand the discourse on foreign lands in these works as the result of both their effort to make knowledge as practical as possible and their attempts to reimagine the tributary system in the face of the breakdown of that system and the challenge posed by western geographical knowledge in the late Ming.

Finally, when we realize that, by constantly making revisions and issuing new editions, late-Ming daily-use encyclopedias were attempting to rebuild and maintain tributary relationships, it becomes indisputably necessary to reexamine the validity of the terms “daily-use encyclopedia” and “daily-use” in interpreting the late-Ming wanbao quanshu encyclopedias. By “daily-use,” Sakai Tadao is referring to concept of “practical everyday lifestyle guidance,” and needless to say, the “Barbarians” sections certainly cannot be understood as fulfilling this concept. Even if we redefine “daily-usefulness” as related to everyday practicality, based on the discussion above, it obviously cannot fully encompass the notion of cultural governance expressed in the “Barbarians” sections of late-Ming encyclopedias. From the Expansive Record of the Forest of Affairs editions to the wanbao quanshu, the history of the formation of “Barbarians” sections in fact reflects the process by which encyclopedia compilers redefined existing categories of information in reaction to crises in practical governance and knowledge systems. In this sense, the editors of the “Barbarians” sections of these encyclopedias not only undertook practical everyday discussions, they also displayed the sensitivity needed to gain a new understanding of a changing world and the timeliness required to fulfill this need by adjusting existing knowledge structures.

---

1. See Sakai Tadao 酒井忠夫, “Gen Min jidai no nichiyō ruisho to sono kyōikushiteki igi” 元明時代の日用類書とその教育史的意義, Nihon no kyōikushigaku 日本の教育史学1 (Oct. 1958): 67-94; Sakai Tadao 酒井忠夫, “Mindai no nichiyō ruisho to shomin kyōiku” 明代の日用類書と庶民教育, in Kinsei Chūgoku kyōikushi kenkyū 近世中国教育史研究, ed. Hayashi Tomoharu 林友春 (Tokyo: Kokudosha, 1958), 62-74.
2. The results of recent studies of daily-use encyclopedias are discussed in detail in the most recent work by Wu Huey-fang. See Wu Huey-fang 吳蕙芳, Ming Qing yilai minjian shenghuo zhishi de jiangou yu chuandi 明清以來民間生活知識的建構與傳遞 (Taipei: Taiwan Xuesheng Shuju, 2007), 11-54.
3. Translator’s note: The Classic of Mountains and Seas in its present form was completed in the second or third centuries BCE, although versions by multiple different authors existed as early as the fourth century BCE. It is now considered a work of geomythology.
4. See Wang Cheng-hua 王正華, “Shenghuo, zhishi yu wenhua shangpin: Wan Ming Fujian ban ‘riyong leishu’ yu qi shuhuamen” 生活、知識與文化商品：晚明福建版「日用類書」與其書畫門 [Lifestyles, information and cultural products: Fujian version daily-use encyclopedias of the late Ming and their illustrations], Jindai Shi Yanjiusuo jikan 近代史研究所集刊 41 (September 2003): 6. See Wang Cheng-hua’s article in this issue of Lingua Franca.
5. See Miura Kunio 三浦国雄, “[Wanbao quanshu] Zhuyi men shōron—Minjin no gaikokukan” 『萬寶全書』諸夷門小論—明人の外國觀 [(Complete record of myriad things) Short essay on barbarians — The view of the Ming people on foreign countries], Kangakukai shi 漢學會誌 44 (March 2005): 245.
6. Wu Huey-fang, Ming Qing yilai minjian shenghuo zhishi de jiangou yu chuandi, 34-35.
7. Translator’s note: The dynastic histories are a series of official histories of the successive dynasties of China. The first, the Records of the Grand Historian by Sima Qian司馬遷 (d. ca. 86 BCE), was completed around 94 BC.
8. Translator’s note: Cited from Burton Watson’s translation (Sima Qian, “Shi ji 123: The Account of Dayuan,” in Records of the Grand Historian: Han Dynasty II (Revised Edition), translated by Burton Watson (New York: Columbia University Press, 1993), 231-252. The original citation follows. “Shi ji: Dayuan zhuan” 〈大宛傳〉，《史記, chap. 123 in [Han] Sima Qian zhuan, [Liu Song] Pei Yin jijie 〔漢〕司馬遷撰，〔劉宋〕裴駰集解 [(Han) Author: Sima Qian, (Liu/Song) Pei Yin’s Collected annotations], 1962 Revised Edition (Beijing, Zhonghua Shudian, 1962), 3179.
9. See, for example, “Han shu: Zhang Qian Li Guang lizhuan” 〈張騫李廣利傳〉《漢書》, j. 61 in [Han] Ban Gu zhuan, [Tang] Yan Shigu zhu 〔漢〕班固撰，〔唐〕顏師古注, 1962 Revised Edition (Beijing, Zhonghua Shuju, 1962), 2705; and “Zhou shu: Yiyu chuanxu” 〈異域傳序〉《周書》, j. 49 in [Tang] Ling Hu Defen deng zhuan 〔唐〕令狐德棻等撰 , 1962 Revised Edition (Beijing: Zhonghua Shuju, 1962), 883.
10. “Shaoshi Shanfang bicong: Sibu zheng e xia” 〈四部正訛下〉，《少室山房筆叢》 , j. 16 in [Ming] Hu Yinglin 〔明〕胡應麟 , 1983 facsimile edition of Wen Yuange’s Siku quanshu (Taipei: The Commercial Press, 1983), 886: 3a.
11. “Siku quanshu zongmu tiyao” 《四庫全書總目提要》 in [Qing] Ji Yun zongzuan 〔清〕紀昀總纂 (Shijiazhuang: Hebei People’s Press, 2000), 3624.
12. On textual research on Ming and Qing daily-use encyclopedia editions, see Wu Huey-fang 吳蕙芳, Wanbao quanshu: Ming Qing shiqi de minjian shenghuo shilu 萬寶全書：明清時期的民間生活實錄 (Taipei: Department of History, National Chengchi University, 2001), 2: 355-385.
13. For studies of Shilin guangji edition-related issues, see Hu Daojing’s “Introduction” to the 1963 facsimile edition of the 1330s Shilin guangji as well as Morita Kenji 森田憲司, “Guanyu zai Riben de Shilin guangji zhu ben” 關於在日本的《事林廣記》諸本 in Guoji Song Shi Yantao Hui lunwen xuanji 國際宋史研討會論文選集, edited by Deng Guangming and Qi Xia (Baoding: Hebei University Press, 1992), 266-280. A revised version of the latter is included along with Hu Daojing’s “Introduction” in [Song] Chen Yuanjing: Shilin guangji—Fu lu 〔宋〕陳元靚：《事林廣記· 附錄》, 1999 version based on 1963 facsimile edition of Shilin guangji (Beijing: Zhonghua Shuju, 1999), 559-572.
14. Translator’s note: Champa and Chenla both refer to collections of polities along the coast of what today is central and south Vietnam; Panduranga was a major city in the far south; Srivijaya was the Buddhist empire (ca. late 7^th – early 11^th c.) based on the island of Sumatra (now in Indonesia). Sindhu refers to India.
15. However, it must be noted here that the contrast between “Daoyi zazhi” and “Shan hai yun yi” should not be understood as the difference between practical guidance for ocean voyages and mythical legends. The reason is that, taking these entries drawn from Zhufan zhi as guidance, it is difficult to imagine that a traveler could smoothly reach a foreign country from a Chinese port based merely on wind directions and the number of days. With regard to the practicality of daily-use encyclopedias, see the relatively detailed discussion below.
16. See [Song] Zhao Rukuo: Zhu fan zhi 〔宋〕趙汝适：《諸蕃志, 1985 Congshu jicheng chubian edition based on facsimile of the Xuejin taoyuan edition (Beijing: Zhonghua Shuju, 1985).
17. Translator’s note: Pekalongan is actually a city in northern Java; it was once the central seat of the kingdom of Kalingga and, later, Dvapa.
18. Miura Kunio, “[Wanbao quanshu] Zhuyi men shōron—Minjin no gaikokukan,” 232-233.
19. See Ge Zhaoguang 葛兆光, “Shanhai jing, Zhigong tu he lüxingji zhong de yiyu jiyi—Limasai laihua qianhou Zhongguoren guanyu yiyu de zhishi ziyuan ji qi bianhua” 《山海經》丶《職貢圖》和旅行記中的異域記憶—利瑪竇來華前後中國人關於異域的知識資源及其變化, in Ming Qing wenxue yu sixiang zhong zhi zhuti yishi yu shehui: Xueshu sixiang bian 明清文學與思想中之主體意識與社會：學術思想篇, ed. Zhong Caijun 鍾彩鈞 and Yang Jinlong 楊晉龍 (Taipei: Institute of Chinese Literature and Philosophy, 2004), 360.
20. See Ge Zhaoguang, “Shanhai jing, Zhigong tu he lüxingji zhong de yiyu jiyi—Limasai laihua qianhou Zhongguoren guanyu yiyu de zhishi ziyuan ji qi bianhua,” 350.
21. See Wang Cheng-hua, “Shenghuo, zhishi yu wenhua shangpin: Wan Ming Fujian ban ‘riyong leishu’ yu qi shuhuamen,” 6.
22. See Wei Shang, “The Making of the Everyday World: Jin Ping Mei and the Encyclopedias for Daily Use,” in Dynastic Crisis and Cultural Innovation: From the Late Ming to the Late Qing and Beyond, ed. David D.W. Wang and Wei Shang (Cambridge MA: Harvard University Asia Center, 2005), 63-92.
23. See Stephen H. West, “Tu 圖Good to Be True: Programmed Learning in Song” (paper presented at the conference ‘Discourses and Practices of Everyday Life in Imperial China,’ Department of East Asian Languages and Cultures, Columbia University, New York, October 25-27, 2002), 2, 16.
24. Translator’s note: Modern Shangqiu, Henan, formerly the southern capital of the Northern Song dynasty (960-1126).
25. Translator’s note: This term (阿不合大王城) may refer to another port on the Persian Gulf, but Lin Meicun believes it refers to Kish Island in present-day Iran. See Lin Meicun 林梅村, “Zhang Honglüe mu yu Dingxing jiaocan chutu Yuan dai gongting jiuqi: Jian lun fuliang ciju chuangshao Yuan qinghua zhi niandai” 张弘略墓与定兴窖藏出土元代宫廷酒器——兼论浮梁磁局创烧元青花之年代, Sina.cn, http://k.sina.com.cn/article_2145932192_7fe853a001900gof2.html?from=history (accessed December 27, 2019).
26. [Ming] Song Lian deng zhuan: Yuan shi 〔明〕宋濂等撰：《元史》, 1962 Revised Edition (Beijing: Zhonghua Shuju, 1962), chap, 210: 4660, 4669.
27. Translator’s note: The text uses a derogatory term referring specifically to Japanese pirates; however, by the Ming dynasty, not only Japanese but Korean, Portuguese and even Chinese pirates were among the crews.
28. Translator’s note: “Xifanguo” (“the Country of Western Barbarians”) appears to refer to areas in western China inhabited by ethnically Tibetan people. The other countries mentioned here seem to be purely fanciful, but unlike the names of the countries in Classic of Mountains and Seas, their names do not explicitly describe strange or exotic characteristics. The text provides an alternate reading for Sheluguo: “Geluguo.”
29. Translator’s note: A legendary beast resembling a deer with a horse’s tail and a single horn.
30. For the latest work in Japanese and Chinese on the influence of Ricci’s world map and East-West cultural exchange through maps, see Unno Kazutaka 海野一隆, Tōzai chizu bunka kōsh ōshi kenkyū 東西地圖文化交涉史研究 (Osaka: Seibundō, 2003) as well as Huang Shijian 黃時鑒 and Gong Yingyan 龔纓晏, Limasai shijie ditu yanjiu 利瑪竇世界地圖研究 (Shanghai: Shanghai Guji Chubanshe, 2004).
31. The “Sihai huayi zongtu” in the 1613 Tushu bian by Zhang Huang (1527-1608) also includes countries from Shanhai jing, including the Country of the Impalers (referred to as “Chuanxinguo” rather than “Chuanxiongguo”), the Country of Dwarves, the Country of People with Long Arms, and so on. There are notes to the right of the map, such as “This map of the Jambudvipa of the four seas of Buddhist scripture is provided for provisional reference,” making it clear that this map is based on a Buddhist worldview and not influenced by Ricci’s world map. See Zhang Huang 章潢, Tushu bian 圖書編, Vol. 969 of 1982 facsimile edition of Wen Yuange’s Siku quanshu (Taipei: The Commercial Press, 1982), chap. 29: 50a-51b.
32. On revised editions of Ricci’s world map, see Huang Shijian and Gong Yingyan, Limasai shijie ditu yanjiu, 3-47.
33. Xu Dashou 許大受, “Shengchao zuobi” 聖朝佐辟, j. 4 in Shouru Xu Changzhi ding: Shengchao poxie ji 收入徐昌治訂：《聖朝破邪集》, vol. 14 of Zhongguo zongjiao lishi wenxian jicheng 中國宗教歷史文獻集成 (Hefei: Huangshan Shushe, 2005), 3a.
34. Other scholars of the late Ming and early Qing saw the accounts in texts like Classic of Mountains and Seas and In Search of the Supernatural as criteria for judging whether western geographical knowledge should be accepted. Most notably, the late eighteenth-century editors of the authoritative Catalogue of the Four Treasuries [of the Imperial Library] (Siku tiyao 四庫提要), in evaluating Nan Huairen’s 南懷仁 (1623-1688) Illustrated Explanation of the Earth (Kunyu tushuo 坤輿圖說), took the view that Nan could credibly cite ancient Chinese texts like the Classic of Divine Marvels as accurate sources of geographical information. The Siku quanshu zongmu tiyao states that some of the information recorded in Illustrated Explanation of the Earth is identical to that of Classic of Divine Marvels, and thus “It is suspected that after they [the western missionaries] came to the East, they gained access to ancient Chinese writings and changed their views in accordance; thus these views may not be based entirely in fact. However, as numerous written works report, some of the rumors of merchant ships are clearly true. There may be some fabrications, but they are not completely false. There are numerous reports of strange happenings, and some are likely true.” See “Siku quanshu zongmu tiyao,” chap. 71 in Ji Yun zongzuan, 1921. This view is not unfounded. In autumn of the same year that Feng Yingjing’s Overview of Monthly Ordinances was published, Ricci created the famous “Complete Map of All Countries on Earth” in Beijing. On this map, to the northeast and southeast of the Mediterranean Sea, respectively, Ricci noted two countries from Classic of Mountains and Seas, the Country of the One-Eyed and the Country of Women. This obviously represents a compromise made for the sake of missionary work.
35. Unno Kazutaka 海野一隆, Ditu de wenhua shi 地圖的文化史, trans. Wang Miaofa 王妙發 (Beijing: New Star Press, 2005), 75.
36. Translator’s note: This second list of countries consists of a mixture of factual and fanciful lands drawn from various sources.
37. Huang Shijian 黃時鑒, “Cong ditu kan lishi shang Zhong Han Ri ‘shijie’ guannian de chayi—Yi Chaoxian de tianxia tu he Riben de Nanzhanbuzhou tu wei zhu” 從地圖看歷史上中韓日「世界」觀念的差異一以朝鮮的天下圖和日本的南瞻部洲圖為主, in Cong zhoubian kan Zhongguo 從周邊看中國, ed. National Institute for Advanced Humanistic Studies at Fudan University (Beijing: Zhonghua Shuju, 2009), 59.

Cite: Hsu Hui-lin, “Imagined Imperial Tribute: Descriptions of Foreign Lands in the ‘Barbarians’ Sections of Late-Ming Dynasty Daily-use Encyclopedias,” translated by Michael Day, Lingua Franca, Issue 6 (2020),
https://sharpweb.org/linguafranca/2020-hsu/.