Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, LINGUISTICS ( (c) Oxford University Press USA, 2016. All Rights Reserved. Personal use only; commercial use is strictly prohibited. Please see applicable Privacy Policy and Legal Notice (for details see Privacy Policy).

date: 18 August 2017

Language and Linguistics in Pre-Modern China and East Asia

Summary and Keywords

Traditional Chinese linguistics grew out of two distinct interests in language: the philosophical reflection on things and their names, and the practical concern for literacy education and the correct understanding of classical works. The former is most typically found in the teachings of such pre-Qin masters as Confucius, Mozi, and Gongsun Long, who lived between the 6th and 3rd centuries bc, the latter in the enormous number of dictionaries, textbooks, and research works which, as a reflection of the fact that most Chinese morphemes are monosyllabic, are centered around the pronunciations, written forms, and meanings of these monosyllabic morphemes, or zi (“characters”) as they are called in Chinese. Apparently, it was the latter, philological, interest that motivated the bulk of the Chinese linguistic tradition, giving rise to such important works as Shuowen Jiezi and Qieyun, and culminating in the scholarship of the Qing Dynasty (1616–1911). But at the bottom, the philosophical concern never ceased to exist: The dominating idea that all things should have their rightful names just as they should occupy their rightful places in the universe, for example, was behind the compilation of Shuowen Jiezi and many other works. Further, the development of philology, or xiaoxue (“basic learning”), was strongly influenced by the study of philosophical thoughts, or daxue (“greater learning”), throughout its history.

The picture just presented, in which Chinese philosophy and philology are combined to form a seemingly autonomous tradition, is complicated, however, by the fact that the Indic linguistic tradition started to influence the Chinese in the 2nd century ad, causing remarkable changes in the analyzing techniques (especially regarding character pronunciation), findings, and course of development of language studies in China. Most crucially, scholars began to realize that syllables had internal structures and that the pronunciation of one character could be represented by two others that shared the same initial and final with it respectively. This technique, known as fanqie, laid the basis for the illustrious 7th-century rhyme dictionary Qieyun, the rhyme table Yunjing, and a great many works that followed. These works, besides providing reference for verse composition (and, consequently, for the imperial examinations held to select government officials), proved such an essential tool in the philological study of classical works, that many Qing scholars, at the very height of traditional Chinese linguistics, regarded character pronunciation as central to xiaoxue and indispensable for the understanding of ancient texts. While character pronunciation received overwhelming attention, the studies of character form and meaning continued to develop, though they were frequently influenced by and sometimes combined with the study of character pronunciation, as in the analysis of the relations between Old Chinese sound categories and the phonetic components of Chinese characters and in their application in the exegetical investigation of classical texts.

Chinese, with its linguistic tradition, had a profound impact in ancient East Asia. Not only did traditional studies of Japanese, Tangut, and other languages show significant Chinese influence, under which not the least achievement was the invention of the earliest writing systems for these languages, but many scholars from Japan and Korea actually took an active part in the study of Chinese as well, so that the Chinese linguistic tradition would itself be incomplete without the materials and findings these non-Chinese scholars have contributed. On the other hand, some of these scholars, most notably Motoori Norinaga and Fujitani Nariakira in Japan, were able to free themselves from the character-centered Chinese routine and develop rather original linguistic theories.

Keywords: history of linguistics, philology, Chinese linguistics, East Asian languages, Chinese, Chinese characters, Japanese, Korean, Vietnamese, Tangut

1. The Linguistic World of East Asia

East Asia, when taken as a cultural concept, refers to China, Japan, (South and North) Korea, and Vietnam, an area where the Chinese culture originated and prospered or had a profound influence. Native to this area are languages from a variety of families, notably the Sino-Tibetan, Austronesian, Austroasiatic, and Altaic families.1 But this genetic diversity has been obscured, to a significant extent, by some prominent regional commonalities: Many East Asian languages have a large number of Sinitic loanwords (expressing both learned and basic concepts), and many have developed shared typological features (e.g., tones). Additionally, various non-Chinese languages, including Japanese, Korean, and Vietnamese, have been or are still written in scripts based on or influenced by Chinese characters, though many such scripts have become obsolete due to language extinction or romanization.

What has helped to unite the linguistic world of East Asia, in addition to the extensive language contact over the course of long-term cultural exchange, is a general conception of the Chinese language and writing as a basis for higher learning and a conscious effort to pursue them in traditional East Asian society. Before deepened contact with the West and other changes brought that society into the modern era, Classical Chinese was widely taught and studied in East Asia, and with it, a systematic knowledge about language was widespread. This knowledge, as a whole, was built on the time-honored studies of Chinese by Chinese scholars (for which a brief timeline is given in Table 1 as a preliminary overview), but by origin, it was a combination of both Chinese and Indic elements, and during its transmission, this combination was further enriched by interpretations and innovations in and for other languages.

Table 1. A Brief Timeline of Chinese Dynasties and Representative Linguistic Works.9

Language and Linguistics in Pre-Modern China and East Asia

2. Language Studies in China Through the End of the Han Dynasty

2.1 Philosophy and Philology in Ancient China

Philosophy is fundamental to understanding the beginnings of Chinese linguistics, not simply because “language” was mentioned in the discussion of some philosophical schools, but more because no academic activity can be detached from its intellectual environment. From the earliest works of Chinese classics composed before the 6th century bc, including the Yi Jing (or I Ching, “Book of Changes”) and the Shi Jing (or Shih Ching, “Book of Odes”), it is already evident that the Chinese mind, although not unfamiliar with the supernatural, is predominantly concerned with human life in the real world and embraces it with a modesty and elegance that resists all excessiveness. At the same time, it sees all things human and nonhuman as operating on the same principles and endeavors to grasp them through such notions as yin, yang, and their interactions. These ideas, as Chinese philosophy flourished between the 6th and 3rd centuries bc, developed into a system of thoughts about the appropriate order of human society as well as the universe that a decent and moral life should follow.

It was within the context described that the Chinese started to reflect extensively on language. Indeed, for almost all the major thinkers between the 6th and 3rd centuries bc, language was an important topic, because the way we speak is directly associated with the way we think, and thus crucial to the appreciation and maintenance of order. Confucius famously said that the first thing he would do, were he to administer a government, was to “rectify names,” since incorrect names would result in incongruous speech and ultimately a series of social disorders (The Analects, 13.3). Laozi, on the other hand, highly stressed the limitation of language in revealing the eternal order, or Dao (or Tao, literally “way”), of the world, and his Dao De Jing (or Tao Te Ching) begins by claiming that “the Way that can be told is not the eternal Way” and that “the name that can be named is not the eternal name” (Dao De Jing, 1).

Between the opinions of Confucius and Laozi lie those of the other scholars. Mozi and his followers, for instance, were positive about the function of language, but more as a reflection of the actual world than as a direct means to maintain social order. They were the first to inquire into the relationship between ming (“name,” “word”) and shi (“actuality”), and to classify ming into da (“general word”), lei (“classifying word”), and si (“private word”) based on their belief that every ming should properly correspond to an actual referent as its shi (Mo Jing, 10.79). The most active in the mingshi discussion, however, were those who came to be known collectively as the “School of Names,” represented by Hui Shi and Gongsun Long. Hui Shi’s arguments, recorded as a series of paradoxes in Zhuangzi, highlighted the personal perspectives underlying ming, and were favored by the Taoists who held a negative attitude to the rigidified mingshi relationship. Gongsun Long, in contrast, emphasized the independence of ming as based on absolute universals. His arguments were therefore close to pure conceptual analysis and sometimes rather sophistic, as is shown in his famous contention that a white horse is not a horse, since “white-horseness” differs from “horseness” in various ways (Gongsunlongzi, “White Horse Discourse”).

The mingshi debates are routinely regarded as an important part in the beginning of Chinese linguistics, but may appear somewhat irrelevant to its later, philological, stages. However, as Fung (1948, pp. 9–10) has pointed out, even Gongsun Long, whose work seems to have little to do with politics, actually intended his arguments to “rectify the relationship between names and facts in order to transform the world” (Gongsunlongzi, “Ji Fu”), and this essential concern for correct words and order, which characterized the whole philosophical beginning of Chinese linguistics, brought about a long-standing normative dimension in the many philological works to come. Admittedly, the technical excursions in the mingshi debates were mostly rejected as irrelevant play of words by later scholars, but the basic attitude to language as a means to achieve order connected the philosophical beginning of Chinese linguistics with its later stages. In this sense, Xunzi, who is known (especially among modern Chinese linguists) for his statements on the conventional nature of names, represented a significant return from technicalities to the original intentions of the mingshi debates when he reemphasized the Confucian idea of “rectification of names” in the 3rd century bc (Ge, 2009, p. 207).

2.2 Establishment of Xiaoxue

As the philosophy about ming, shi, and order prevailed, practices in the interpretation of words also became widespread. Some of these interpretations, e.g., Confucius’ definition of “politics” as “uprightness” (The Analects, 12.17) based on the fact that the two words in the Old Chinese of his time were homophonous (as they still are in Modern Chinese), clearly carried a special implication, but others may have been less intentional than necessitated, such as the explanation of an obsolete word to those who did not know it. Both cases could be regarded as examples of achieving order through correct names, but in the latter case, a practical need stood out. Even at the time of the mingshi debates, simple commentaries and textbooks had appeared, for with the passage of time, the Chinese language and its script forms kept changing, making older texts less easy to read for later generations. During the Han Dynasty (206 bc–ad 220), when standardization of writing and destruction of many classical books at the order of the emperor in the previous Qin Dynasty (221 bc–206 bc) made it even more difficult for people to read ancient texts, especially the Confucian classics that had risen to political and academic prominence, a rapid boost of philological studies occurred. After the accidental discovery of certain Confucian works that had survived the Qin-Dynasty destruction, these studies further split between the “Old Text School,” who highly recognized the value of the pre-Qin documents, and the “New Text School,” who insisted on the authority of their post-Qin versions of Confusion classics. With their factual attitude and rigorous research, the Old Text School made unprecedented contributions, and it was due to their effort that a new branch of study—xiaoxue (“basic learning”; “philology”)2—was formally established.

In the beginning, xiaoxue was mainly a reference to character form studies. In the Sui Dynasty (581–619), it began to include studies on character meaning and pronunciation as well, and a composition of three subfields, referred to as yinyun (“character pronunciation”), tizhi (“character form”),3 and xungu (“character meaning”), was finally recognized in the Song Dynasty (960–1279) (Hu, 2005, p. 2). These subfields are known in modern scholarship as yinyunxue, wenzixue, and xunguxue, a terminology established by Zhang Taiyan at the beginning of the 20th century. The threefold composition, centered around different aspects of Chinese characters (and in interesting contrast with the Western trivium), is deeply rooted in the fact that Ancient Chinese words (and Modern Chinese morphemes)—the essence of Chinese characters—are mostly monosyllabic and naturally show up as a basic unit for study.

In a sense, meaning is the ultimate goal of philology; therefore, many early interpretations of words already had a philological dimension in them. In the early Han Dynasty, a collection of such interpretations titled Erya (“Approaching the Proper Language”) gradually became popular. The book, as it has come down to us, contains nineteen chapters, arranged according to semantic categories like “kinship terms,” “music,” and “birds.” Experts disagree as to when and by whom it was compiled, but the book does include certain interpretations that have been found to have a pre-Qin precedent preserved in other sources. Another archaic feature of the book is that a thesauruslike format, which had occurred in the Pre-Qin classic Shizi, was employed in the first three chapters of Erya, in which groups of synonyms are listed with the last item of each group explaining all its preceding items, e.g.:

Language and Linguistics in Pre-Modern China and East Asia

For later scholars, Erya provided an encyclopedia of the ancients’ language and life, as well as an example to follow when similar books were made. But during the Han dynasty, it inspired a rather special work titled Fangyan (“Dialects,” literally “Regional Expressions”). This book, written, somewhat arguably, by Yang Xiong (53 bcad 18), was arranged in a similar way to that of Erya, with thirteen chapters each dealing roughly with a semantic or functional category. Moreover, it also employed the thesauruslike format quite frequently, but with particular notes on where each of the synonyms was mainly used, as in the following:

Language and Linguistics in Pre-Modern China and East Asia

Comparing this entry from Fangyan with the previous one from Erya, one can easily see the connections and differences between the two books. For those mainly interested in the understanding of classical texts, the regional information provided in Fangyan would not seem very important—indeed, its interest in living dialects found very little echo in the Chinese linguistic tradition. But from the modern linguistic point of view, the value of Fangyan cannot be overstated, for it laid out, for the first time, the dialectal situation in China from a systematically lexical approach. It should be mentioned, of course, that the author Yang Xiong already possessed certain dialectal records from previous folk song collectors of the government when he started his survey and that he did not travel around the country for his research, but mainly stayed in its capital Chang’an, interviewing speakers from different areas. In addition, Yang Xiong never had a modern conception of dialect and language based on genealogy. His fangyan meant simply “regional expressions,” and some of the expressions he recorded were actually from non-Chinese languages.

For the ancient Chinese, the greatest obstacle to the understanding of classical texts, besides changes of character meaning, was probably unfamiliar character forms. To facilitate the learning of characters, various textbooks had been composed before and during the Han Dynasty, but they invariably listed the characters with no explanation on their forms. It was only in the hands of the Old Text School that the study of character form really developed. The most representative of their work, without doubt, was the Shuowen Jiezi (“Explaining Simple and Analyzing Compound Characters”) by Xu Shen (ca. 58–ca. 147). In this book, over nine thousand characters were analyzed structurally on the basis of 540 “radicals”—i.e., basic “semantic components” with relatively fixed form and meaning. These formal components were also used as the criterion for the categorization of the characters, which meant that for the first time, characters were grouped not according to what words or morphemes they stood for (and thus what was signified), as in Erya or Fangyan, but according to how they were written. A similar concern for the written form also defined the presentation and explanation of the characters in Shuowen Jiezi: Normally, an old-fashioned seal script form would be given at the beginning of an entry, to familiarize the reader with the more ancient and “revealing” way the character had been written, and to facilitate the author’s account of the “original” meaning of the written form. These features could be seen clearly in the following examples, in which the characters 如‎ and 嫁‎, both having appeared in the aforementioned quotations from Erya and Fangyan, and both included now under the category of 女‎ (their shared radical) in Shuowen Jiezi, are introduced:

Language and Linguistics in Pre-Modern China and East Asia

Language and Linguistics in Pre-Modern China and East Asia

From these entries, the reader will know not only how 如‎ and 嫁‎ had been written before their time, but the rationale of their composition and, in connection to it, their original meaning. As Xu Shen saw it, every Chinese character belongs, according to its composition, to one of six general types: xiangxing (“pictograph”), zhishi (“ideograph”), huiyi (“semantic compound”), xingsheng (“phonosemantic compound”), jiajie (“phonetic loan”), and zhuanzhu (“semantic derivative”). Actually, the so-called “six types of characters” had been mentioned as early as in the pre-Qin documents, but Xu Shen was the first to give them a systematic definition. Though what he meant by zhuanzhu is not very clear, the whole system has been well accepted and is widely used even in modern scholarship. According to this system, 如‎ (“follow”), for example, is a semantic compound built from a “woman” and a “mouth” component, based on the idea that women typically follow what they are told, while 嫁‎ is a phonosemantic compound, because though the “woman” component still indicates its semantic relevance, the “home” component only points to its sound. An example of the pictograph would be the “woman” character itself, which is basically a human image with coiled arms (and originally kneeling legs), as can be seen in its seal script:

Language and Linguistics in Pre-Modern China and East Asia

Generally speaking, 女‎ is a simple character, while 如‎ and 嫁‎ are compound characters—exactly what wen (“simple character”) and zi (“compound character”) refer to in the title of Xu Shen’s book.

While Xu Shen’s original and comprehensive study of character form was a tremendous contribution to Chinese philology, it is worth noting, from the perspective of “understanding classical texts,” that the “original meanings” of characters in Xu Shen’s work are completely based on character form, and differ, in principle and often in practice, from the original meanings of the words written down by the characters. As Karlgren (1940, p. 13) has pointed out,

It should be observed that we must distinguish carefully between the primary sense of a word (word-stem of the language, quite irrespective of the script) and primary sense of a character (the Chinese ideograph). … An example is 裏‎li. The word-stem li means “in, inside”, and it is only in a very special, secondary, derivative concrete sense that it has come to mean “inside of a garment”. But the Chinese script inventor has seized upon this depictable special sense in order to find a suitable graph, and this has then by extension come to serve for the whole range of the stem li “inside”.

Actually, ancient Chinese scholars were not unaware of a “primary sense” of the word independent of its character form. For example, in the pre-Qin document Shizi, the word for “ghost” was explained with the word for “return” (i.e., “the returned after death”), not with reference to how they were written, but on the basis of their phonetic similarities and semantic relations. However, such interpretations, known as shengxun (“phonetic interpretation”), were often subject to the ideology of the time or the belief of the interpreter. They could be plausible, as in the abovementioned definition of “politics” as “uprightness” by Confucius, but more often, they were too far-fetched to be true, as in the definition of “man” as “(official) post” on the ground that the two words sounded similar in Old Chinese and that men should take an official post and establish achievements in the world. At the end of the Han Dynasty, a collection of such interpretations titled Shiming (“Interpretation of Names”) was compiled by Liu Xi (late 2nd century–early 3rd century). The book, in 27 chapters arranged according to semantic category, is significant as the first etymological work in Chinese linguistics and can be very insightful and enlightening in many places. But generally speaking, the interpretations it provided show a strong influence of the folk etymology of its time. Moreover, these interpretations were often misguided by a belief in the natural connection between sound and meaning.

3. Influence of Indic Linguistics and Developments of Xiaoxue

3.1 Overview

By the end of the Han Dynasty, linguistics in China had been developing without much external influence on two complementary dimensions. On the one hand, scholars were predominantly focused on the philological details in the study of classical works—for this reason, they were sometimes considered overmeticulous or even fussy by later scholars. But while doing all the detailed work, they harbored a deep philosophical concern for the correctness of language as a means to achieve social order. Xu Shen, for instance, clearly stated in his preface to Shuowen Jiezi that characters were “the basis of all learning and the foundation of government” and that once their secrets were known, “there would be no disorder.” Liu Xi’s phonetic interpretations were, of course, more directly motivated by the ruling doctrine of his time. In addition, the rulers of the country, being the most directly concerned about order, made various policies about language. The most famous of these was the standardization of characters in the Qin Dynasty, but during the Han Dynasty, official conferences were also convened by the emperor to discuss philology and had far-reaching influences on scholars across the country.

After the Han Dynasty, the philology–philosophy interaction in Chinese linguistics persisted, but at a deeper level, this academic world ceased to be autonomous, because Indic linguistics started to exert a substantial influence in China with the entry of Buddhism, and some of the techniques it provided, though incongruous with the interpretation of Chinese philosophy, gradually grew into an indispensable and even crucial part of Chinese linguistics and significantly determined the course of its development. As it turned out, the area where the Indic tradition was exclusively influential was the study of character pronunciation. This, without doubt, had to do with the fact that ancient Indian phoneticians were “above all their contemporaries and successors before the nineteenth century” (Robins, 1997, p. 175). Yet it has to be admitted that in the philology of the Han Dynasty, character pronunciation was a seriously understudied area as compared with character meaning and form: On the one hand, as has been revealed in the etymological work of Liu Xi, the Chinese lacked a proper method for phonetic analysis; on the other, because of the cross-temporal and cross-dialectal nature of Chinese characters, pronunciation was at least prima facie a less likely obstacle to the reading of classics, and its neglect by most scholars, except in the somewhat arbitrary “phonetic interpretations,” was nothing but understandable from a practical perspective.

As the Indian method of phonetic analysis was introduced into China, the situation started to change significantly. In the beginning, the study of character pronunciation was developed, not as a direct aid to the interpretation of the classics, but as a guide to verse composition. But it soon got the official recognition of the government in the Tang Dynasty (618–907) and joined the philology–philosophy interaction of the country. During the Song Dynasty, it was formally introduced into the study of ancient classics, and in the Qing Dynasty (1616–1911), it finally became the most essential tool in the philological analysis of classical texts. Naturally, the studies of character form and meaning kept advancing at the same time, but they were changed in a fundamental way by the newly developed knowledge of phonetics.

3.2 Phonetics as a Meeting Ground of the Chinese and the Indic Tradition

Just as the real study of character form must start with a structural analysis, the real study of character pronunciation must take the analysis of phonetic structure as a premise. However, with sound being transient, and the syllable being much more prominent than its component parts as the basic phonetic realization of the Chinese word/morpheme, the latter analysis proved considerably difficult for the Chinese. Although the fact that a few one-syllable contractions of two-syllable sequences had long existed in China (e.g., the contraction of 之‎ ([tjə], “it”) and 乎‎ ([ɡa], a particle) into 諸‎ ([tja]) in Old Chinese) suggests that the Chinese could have detected the internal structure of the syllable from their own observation, most scholars believe that enlightenments from Indic linguistics were indispensable to the official beginning of phonetic analysis in China. This idea also corresponds well with the dating of the invention of fanqie (literally “turn and match”) around the end of the Han Dynasty, which was not long after the introduction of Buddhism into China.

Fanqie is a technique for the first step of syllable analysis, according to which the pronunciation of one character is represented by two others which share the same initial and final (including the tone) with it respectively. For example, the pronunciation of 歌‎ (“sing”, pronounced [kɑ] with the level tone in Middle Chinese) could be represented by 古‎ (“ancient”, pronounced [ko] with the rising tone in Middle Chinese) and 俄‎ (“sudden”, pronounced [ŋɑ] with the level tone in Middle Chinese), as illustrated in Figure 1.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 1. Middle Chinese pronunciation of 歌‎ represented by 古‎ and 俄‎ according to the fanqie method.

The fanqie method was a revolution both in theory and in practice. Before its invention, scholars were used to indicating the pronunciation of a difficult character with a simpler one, but it was not always easy to find a simple character with exactly the pronunciation they needed. The fanqie method, besides exposing the internal structural of the syllable, made the choice of simple characters much easier. Starting from the Three Kingdoms Period (220–280), fanqie spellings flourished throughout the country. Some of the earliest examples can be found today in the Jingdian Shiwen (“Commentaries on Classical Texts”) by the Tang-Dynasty scholar Lu Deming (ca. 550–ca. 630) and several other works.

The fanqie method was not without disadvantages. For instance, a simple character with the desired initial or final (especially final) was not always easy to find. Thus, the fanqie spellings of some difficult characters had to involve equally or even more difficult characters. Moreover, as dialectal differences always existed, a fanqie spelling which sounded simple to one person could pose some difficulty to another. In addition, as the sounds of Chinese kept changing, the same initials and finals might become differently pronounced in different characters, turning the old fanqie spellings invalid. However, for the earliest researchers of fanqie, these were not yet the most serious problems. While fanqie spellings served their purpose as indicators of pronunciation for individual characters, the phonetic system underlying the various spellings (in which the same initial or final was often represented by different characters) was not directly clear. Faced with the numerous fanqie spellings and the pronunciations they stood for, a researcher would naturally want to classify them—classification being the first step in any study, including the study of character meaning and form as well. However, with character meaning and form, classification was relatively easy: In Erya, it was based on the semantic categories of the characters’ references; in Shuowen Jiezi, the semantic categories of the radicals were drawn upon. But since “initial” and “final” were such atomic notions that the Chinese mind still had to get used to, scholars encountered a problem in the classification of character pronunciations.

Among the first scholars to undertake the classification were Li Deng (3rd century), who compiled Shenglei (“Sound Categories”) during the Three Kingdoms Period, and Lü Jing (4th century), author of Yunji (“Collection of Rhymes”) in the Jin Dynasty (265–420). Unfortunately, their works have both been lost, but according to records, they both abandoned the semantic principle, which had long dominated dictionary-making before them, and chose to resort to the “five sounds” (i.e., gong, shang, jue, zhi, and yu) of traditional music theory—a framework derived from the traditional Chinese philosophy about the “five elements” (i.e., mu “wood”, huo “fire”, tu “earth”, jin “metal”, and shui “water”) of the world. It is no longer known how they actually classified characters under the “five sounds,” but judging from criticisms by later scholars, their classifications were not very successful.

At the same time, Indic linguistics continued to exert its influence, leading to the further discovery of the “four tones” in Middle Chinese—namely, ping (“level”), shang (“rising”), qu (“departing”), and ru (“entering”, i.e., checked, with a final plosive that is inaudibly released)4—during the Southern and Northern Dynasties (420–589). Regarding the exact manner in which Indic linguistics had helped to disclose the Chinese tones, the most influential theory was proposed by Chen (1934), who claimed that the four-tone idea derived directly from the three-tone system used in the recitation of Buddhist scriptures with the addition of the ru tone. This theory has been seriously challenged, but all agree that Zhou Yong (?– ca. 485), author of Sisheng Qieyun (“A Rhyme Book Based on the Four Tones”), and Shen Yue (441–513), author of Sisheng Pu (“The Four-Tone Compendium”), were the first Chinese scholars to describe the four tones systematically. Their works have long been lost in China, but it is believed that the “Tiao Sisheng Pu” (“Compendium Guide to the Four Tones”) chapter in the Bunkyō Hifuron (“Treatise on the Secret Treasury of the Literary Mirror”) by Kūkai (774–835), a Japanese monk who had studied in China during the Tang Dynasty, was copied from Shen Yue’s Sisheng Pu and should be the same as the original. The chapter clearly gives the names of the four tones, and contains a tone diagram in four parallel columns, as shown in Figure 2:

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 2. A tone diagram in the “Tiao Sisheng Pu” chapter of Bunkyō Hifuron.

Each of the columns here contains seven characters of the same initial and main vowel, except that the first three characters end in a nasal, the fourth ends in a plosive, and the last three have no consonant ending, while the first and last three characters have parallel tones. For example, the seven characters in the third column from right are pronounced, from top to bottom, kuɑŋ (level tone), kuɑŋ (rising tone), kuɑŋ (departing tone), kuɑk (entering tone), kuɑ (level tone), kuɑ (rising tone), and kuɑ (departing tone) in Middle Chinese. This arrangement shows that the author was not only familiar with tones, but had a clear conception about the internal structure of the syllable and the final.

With the development of fanqie and the discovery of tones, phonetic study became a popular pursuit among Chinese scholars, and affected the composition of literature to a significant extent. Throughout the Southern and Northern Dynasties, writers were extraordinarily conscious of the phonetic beauty of literary works, and various rules were proposed for verse composition. On the other hand, the widespread knowledge of the four tones finally enabled the successful compilation of a series of phonetic dictionaries, in which characters were no longer classified under the “five-sound” system, but on the basis of their tones and rhymes.5 These dictionaries, which came to be known collectively as “rhyme books,” were especially useful for poets, because they provided an easy reference to groups of characters that rhymed with each other. As required by the rules of rhyming, characters must have the same tone, the same main vowel, and the same ending (if there is one) to rhyme, but not necessarily the same “medial” (i.e., anything between the initial and the main vowel, if there is one). For example, a character pronounced [tuan] (level tone) would have the phonetic structure seen in Figure 3.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 3. Phonetic structure and rhyming part of [tuan] (level tone).

Thus, it could rhyme with [luan] (level tone), [lan] (level tone), and [lian] (level tone), but not [luen] (level, rising, or departing tone), [lua] (level, rising or departing tone), [luak] (entering tone), or [luan] (rising or departing tone). In a “rhyme book,” those characters with the same rhyme—i.e., the “main vowel (+final) + tone” part—will form a yun (“rhyming group”). Usually, each yun would be further divided into various xiaoyun (“small rhyming groups”), in which characters of exactly the same pronunciation are grouped together.

The most famous and authoritative of all the rhyme books is Qieyun, compiled by Lu Fayan (ca. 562–?) on the basis of his own research as well as the notes he had jotted down in a “phonetic salon” attended by eight scholars during the Sui Dynasty. This rhyme book, containing 12,158 characters according to Fengshi Wenjianji (“Feng’s Record of What He Sees and Hears”) of the Tang Dynasty, had been lost for a long time after the Song Dynasty, when its enlarged and adapted version, titled Guangyun (“The Expanded Rhyme Book”), was made and became its successful replacement. Although some incomplete copies of Qieyun were eventually recovered in Dunhuang at the end of the Qing Dynasty, scholars are still accustomed to using Guangyun as a reference to the phonological system underlying the original Qieyun.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 4. Two continuous pages from Guangyun.

Guangyun contains altogether 26,194 characters, grouped under 206 “rhyming groups,” which, according to their tones, are put into five volumes (with the ping-tone characters divided into two volumes because of their large number). Figure 4 shows two continuous pages from the second volume of the ping-tone characters in the book.

Reading from right to left and focusing on the highlighted parts, one encounters first “十一‎○唐‎”, which means that what follows are the eleventh “rhyming group” of the present volume, in which all the characters, represented by 唐‎ (pronounced [dɑŋ] with the level tone in Middle Chinese), rhyme with each other. At the end of the entry for 唐‎ is its fanqie, followed immediately by a note on the number of characters that share the pronunciation of 唐‎ (i.e., characters in the same xiaoyun, or “small rhyming group”, of 唐‎). This same pattern is repeated for the second “small rhyming group” starting with 郎‎, and the third “small rhyming group” starting with 當‎. Surely, the characters of all these “small rhyming groups” must rhyme with each other, because they all belong to the same “rhyming group”. The interesting thing is that if one looks at the fanqie of these three “small rhyming groups” in detail, one shall see that the final of唐‎ is represented by 郎‎, the final of郎‎ by 當‎, and the final of當‎ by 郎‎. This shows that although representative characters were used to name the various “rhyming groups” and “small rhyming groups,” there was no conception that the same final must be represented by one and the same character in practice.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 5. The 36 Chinese initials presented in the Song-Dynasty Qiyinlüe (with reconstructed sound values).

Generally, people were much more interested in tones and rhymes than in initials, because the former were important for verse composition, which was, starting from the Tang Dynasty, part of the national “imperial examinations” held to select government officials. However, the Buddhists, without the Confucianists’ worldly concern, but with enlightenments from their knowledge of Sanskrit, went off the beaten track. In the Tang Dynasty, a list of 30 Chinese initials was made, apparently by a monk called Shouwen. It is believed that this list was the forerunner of a better-known list of 36 initials from the Song Dynasty, as given in Figure 5 in a table from Qiyinlüe (“Compendium on the Seven Sounds”).

Comparing this table with the Sanskrit consonant alphabet (Figure 6), one can easily see that although the Chinese initial system and the Sanskrit consonants differed in the actual sounds they included (e.g., Chinese has only one series of voiced obstruents, and Sanskrit has only one series of affricates), there is an obvious connection between the two in the way the sounds are laid out (the Chinese table should be read from right to left): Obstruents and nasals appear before glides and liquids and are grouped according to place of articulation; in each group, the voiceless unaspirated obstruent always comes first, followed by the voiceless aspirated, the voiced, and the nasal.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 6. The Sanskrit consonant alphabet (with reconstructed sound values).

As the study of tones, rhymes, and initials progressed, a new tool known as yuntu (“rhyme table”) was developed, again under the direct influence of Indic linguistics. Imitating the Siddham tradition of language education, in which consonants and vowels were combined one by one to give various syllables, these “rhyme tables” laid out the Chinese initials, rhymes, and tones like warps and wefts on the cloth and gave their combinations at each point where they crossed. With these tables, one could not only see how syllables were formed by their components, but was in a better position to understand the nature of the “small rhyming groups” in the rhyme books. The best-known “rhyme table”—and also the oldest we now have—is the Yunjing (“Mirror of Rhymes”) by an anonymous author, made in the Tang Dynasty or sometime later. Figure 7(a) is a page from this work.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 7. A page from Yunjing and the schematic presentation of part of it (in red square) with reconstructed sound values.

In this figure, the initials are given on the top, and the rhymes on the left. Since rhymes make no reference to the medial, a “rhyming group” may include several finals with different medials. The “rhyming group” under the 麻‎ rhyme in this table, for instance, has two kinds of finals, namely [a] and [ia], both of them level tone. Similarly, the馬‎ “rhyming group” has [a] and [ia] of the rising tone, and the禡‎ “rhyming group” [a] and [ia] of the departing tone.6 There are only three tones in this table, because the ru-tone rhymes always appear with their corresponding nasal-ending rhymes (e.g., the ru-tone [ak] and [iak] would appear together with [aŋ] and [iaŋ] of the other three tones).

Since each of the rhymes in this table includes two finals, they each need at least two horizontal lines to show how their finals are combined with the initials to produce the syllables of actual characters, which correspond to the “small rhyming groups” in the rhyme books, as well as syllables unused in the language, which are indicated by circles. The 麻‎ rhyme, for instance, is presented on two such lines under the initial [ȶ], so that the combinations [ȶa] and [ȶia] can be written in different slots (more clearly shown in Figure 7(b)). One can see that each rhyme is actually given four horizontal lines in this table, which seem to be more than necessary. However, besides accommodating the number of finals under the rhymes, these lines are roughly related to certain qualities of the finals. For instance, the first line is often used for those finals with a very open main vowel (like [ɑ]), and the third line is reserved for finals with the [i] medial (which is why [ȶia] should really be written on the third line, as indicated in Figure 7(b)). Another reason why these horizontal lines are necessary is that the vertical lines are actually not enough for the initials. Thus, as Figure 7(b) clearly shows, some vertical lines may contain two or even three initials, and when more than one initial is combined with the same final, some of the resulting syllables will have to be “squeezed” out of their original slot onto other horizontal lines, as are [zia], [sia], [dzia], and [tsia] in Figure 7(b).

After Yunjing, many other “rhyme tables” appeared, including the famous Qiyinlüe, Sisheng Dengzi (“A Standard Reference on the Four Tones”), and Qieyun Zhizhangtu (“A Manual of Qieyun in Chart Format”) from the Song Dynasty, and the Qieyun Zhinan (“A Guide to Qieyun”) from the Yuan Dynasty. They were not all made in the same way, but they all included some “tricks” like the ones in the preceding example from Yunjing. In these tricks, the horizontal lines, known as deng (“grade”) in Chinese, often played a crucial role. Therefore, the study of the “rhyme tables” came to assume the name dengyunxue (“study of graded rhymes”). Because of their special design as well as the change of language over time, the “rhyme tables” could become so tricky in some cases that menfa (“gateway”), a unique knowledge to guide the reading of the tables, was developed.

From the beginnings of fanqie to the establishment of dengyunxue, the techniques and enlightenments from Indic linguistics certainly changed the Chinese study of character pronunciation in a fundamental way. However, this does not mean that the study was turned into a field where, like what has happened in modern linguistics, non-Chinese methods were applied to Chinese facts and findings were happily made. Instead, the Chinese kept trying to adapt the non-Chinese methods so that the methods would at least appear Chinese and fit better into the Chinese epistemological system. Admittedly, for the classification task confronting Li Deng and Lü Jing, it seemed no longer necessary to appeal to the traditional “five-sound” system, but a bigger problem confronting all Chinese scholars was how the many phonetic techniques and findings should be accounted for in their philosophy of the world. Many Chinese scholars took painstaking efforts to solve the problem. The most famous and influential of them was probably Shao Yong (1011–1077) of the Song Dynasty. As a great master of traditional knowledge on yin, yang, the “five elements,” and the secret meanings of numbers, Shao Yong tried to incorporate the basic notions in phonetic studies into his grand picture of the world in his masterpiece Huangji Jingshi Shu (“Book of the Supreme Laws of the World”), assigning traditional philosophical values to initials, rhymes, tones, and their interactions. These studies were in many ways a dogmatic imposition of Chinese concepts and are often dismissed as unscientific and far-fetched by modern scholars, but they had far-reaching influences in the ancient world.

3.3 Contemporary Pronunciation Versus Ancient Pronunciation

When the foundations of character pronunciation studies were laid, the phonetic system in use was Middle Chinese, but as time went on, later scholars had to face the problem that their everyday speech was not always compatible with what they had learnt from the standard references. Here, the philology–philosophy interactions, strengthened by the government through their “official rhyme books” and administrative system, played a decisive role, as most scholars chose to keep to the “standard” pronunciation as much as possible in their research and composition, and in reading aloud from written texts. Naturally, the influence of the local speech was unavoidable, but to keep to the “standard,” one did not really have to follow the exact pronunciation of the ancients, because, despite their descriptions of sound values, the “rhyme books” and “tables” were mainly focused on which characters were sounded differently from which others, and in this sense, were much more phonological than phonetic. Therefore, as long as its phonological distinctions were maintained, the “standard” pronunciation would still stand, although its phonetic realization may have changed considerably. Under these circumstances, the Middle Chinese sound system continued to thrive and dominate philological studies. For instance, Qieyun, which was made an official reference in the early Tang Dynasty, generated an enormous number of offspring “rhyme books.” Most of them, including the famous Guangyun and Jiyun (“Amassed Rhymes”) of the Song Dynasty, kept the Middle Chinese system of the book largely unchanged.

However, even within the “standard” sound system of Middle Chinese, variation was inevitable, and could create some knotty problems. Modern scholars, for instance, disagree on whether Qieyun was based on one or more dialects and what the dialect(s) was/were. In ancient China, such problems were largely ignored, but during the late Qing Dynasty, the Cantonese scholar Chen Li (1810–1882), in his masterpiece Qieyun Kao (“Investigation of Qieyun”), made an outstanding contribution to the study of Qieyun by disclosing the actual sound system of the book through his ingenious xilian (“linking”) method. His findings proved, among others, that the system of initials in Qieyun was not completely consistent with the 36 initials popular since the Song Dynasty, although both the book and the table were assumed to reflect the same “standard.”

Unlike the supreme “standard” of Middle Chinese, contemporary pronunciation seldom caught the attention of most scholars unless certain phonological distinctions or fanqie spellings prescribed by the “standard” could no longer hold in the actual linguistic experience of most speakers, and an adaptation of the “rhyme books”—which could be rather considerable in certain cases—was desired. An interesting detail in these adaptations was that the old “rhyme books” to be adapted were often criticized for having deviated from the “standard” under the influence of the wuyin (“pronunciation of the Wu dialects”) in southeast China. Actually, for demographical reasons, the Chinese language changed more quickly in the north than in the south, and the Middle Chinese sound system was generally better preserved in the southern dialects of Chinese. But in the eye of most scholars, the fact that the language spoken in the traditional heart of civilization was more removed from the “standard” than the southern dialects simply did not make sense. Thus, their adaptation of the old “rhyme books” was often done under the name of restoration of the “standard” (Hirata, 2016).

Such was the mentality of the Yuan-Dynasty scholar Xiong Zhong (13th century) when he based his Gujin Yunhui Juyao (“Synopsis of A Collection of Ancient and Modern Rhymes”) on the contemporary pronunciation of his time. However, the book still retained the traditional Middle Chinese framework. In comparison, the Zhongyuan Yinyun (“Rhymes of the Central Plain”) by Zhou Deqing (1277–1365), also a Yuan-Dynasty scholar, was a real revolutionary work. As a reference book for the composition of qu, a popular form of poetry, Zhongyuan Yinyun gave a realistic presentation of a contemporary sound system (arguably that of the Yuan capital Dadu), classifying the rhymes into 19 categories and subsuming, under each category, rhymes of different tones. The fact that the traditional ru-tone characters were attached as subcategories to the other tone groups in the book suggests that the ru tone may have already disappeared in the north of China when it was written. Actually, shortly before Zhongyuan Yinyun, a unique book titled Menggu Ziyun (“Rhyme Book in Mogolian Script”) had already been made on the basis of contemporary pronunciation, using the Phags-pa script, an alphabetic writing system, to transcribe the “rhyming groups.” Following Zhongyuan Yinyun, a series of other untraditional “rhyme books” appeared, including the Yunlüe Yitong (“An Easy and Simplified Access to Rhymes”) of the Ming Dynasty (1368–1644), the Wufang Yuanyin (“Fundamental Sounds of the Five Directions”) of the Qing Dynasty, as well as a series of reference books for qu composition. At the same time, numerous “rhyme table” works were also made on the basis of contemporary pronunciation, including the Ming-Dynasty Qingjiao Zazhu (“Miscellaneous Writings Composed in the Green Outskirts”) and Jiaotai Yun (“Rhymes of Harmonious Match”), the Qing-Dynasty Lishi Yinjian (“Li’s Guide to Sounds”), and many others.

Compared with contemporary pronunciation, the Old Chinese pronunciation of the pre-Qin classical authors received much more attention from an early time on. During the Southern and Northern Dynasties, scholars were already aware that the rules of rhyming in the pre-Qin classical texts often differed from their own and formed the practice of altering character pronunciation where classical poems did not rhyme in their mouth. This practice, known as xieyin (“harmonizing sound”), was based on an incorrect conception about language change, but became very popular in the Song Dynasty. At the same time, with the accumulation of knowledge about the Middle Chinese sound system, some Song-Dynasty scholars began to examine the xieyin materials against the Middle Chinese “rhyming groups,” and came up with preliminary findings about the Old Chinese rhymes. These studies, epitomized in the works of Wu Yu (1100–1154), often took xieyin for granted and have been criticized for being careless and arbitrary in their treatment of rhyming relations. However, they initiated some of the most important methods in the study of Old Chinese pronunciation, including the use of poetical works and the phonosemantic compounds of the characters, as well as the sound system of Middle Chinese as a reference framework. These methods were given a more definite form in the works of the Ming-Dynasty scholar Chen Di (1541–1617), who abandoned xieyin for a more precise and systematic approach to rhyming in the pre-Qin classics, paving the way for the stupendous advancements in the study of Old Chinese pronunciation during the Qing Dynasty. When criticizing the xieyin theory, Chen made the famous statement that “the passing of time from the past to the present, the division of land into the south and the north, the replacement of old characters by new ones, and the change of pronunciation through time—these are all natural and inexorable processes.” This statement is thought to be the first clear formulation in Chinese linguistics of the idea of language change over time.

The categorization of Old Chinese rhymes occupied the center of philological scholarship throughout the Qing Dynasty, and boasts a long list of illustrious names: Gu Yanwu (1613–1682), author of Yinxue Wushu (“Five Books of Sound Studies”), Jiang Yong (1681–1762), author of Guyun Biaozhun (“Standards of Ancient Rhymes”), Dai Zhen (1724–1777), author Shengyun Kao (“Investigation of Rhymes”) and Shengyun Biao (“Table of Sound Categories”), Duan Yucai (1735–1815), author of Liushu Yinyun Biao (“Table of Rhyme Categories Underlying Character Structure”) and Shuowen Jiezi Zhu (“Commentary on Shuowen Jiezi”), Kong Guangsen (1753–1787), author of Shi Shenglei (“Sound Categories in Shi Jing”), Wang Niansun (1744–1832), author of Guyun Pu (“A Compendium of Ancient Rhymes”), Jiang Yougao (?–1851), author of Yinxue Shishu (“Ten Books of Sound Studies”), and so on. They applied the methods of poetic rhyming examination and phonosemantic compound analysis with unprecedented scrutiny, and, based on their findings and other evidence, classified Old Chinese rhymes into 10 or more categories. The most representative of these classifications, with the names, ordinal numbers and corresponding relations of the categories, are given in Table 2.

Table 2. Old Chinese Rhyme Categories According to Seven Qing-Dynasty Scholars (Source: Adapted from Yang, 2005, p. 119).

Language and Linguistics in Pre-Modern China and East Asia

This table is divided into two halves, separated by a horizontal double line in the middle, and the figures following the names of the scholars in the left column indicate the total number of rhyme categories each scholar proposed. Through the corresponding relations between the categories, one can see roughly how the ten categories of Gu Yanwu, the groundbreaker in the classification of Old Chinese rhymes, were further refined and divided in various ways by subsequent scholars.

Of the many divergences among Gu Yanwu and his followers, the most important, as shown in the table, was that while Dai Zhen (and some others) treated the Old Chinese sources of Middle Chinese ru-tone characters (whose pronunciations invariably ended in an inaudibly released plosive, i.e., -p, -t, or -k) as individual categories (i.e., categories 3, 6, 9, 12, 15, 18, 21, 23, and 25 in Dai Zhen’s classification), the other scholars usually included them under those categories with the same main vowels, but with no consonant ending in their Middle Chinese reflexes. For example, the characters 靄‎and 遏‎, pronounced [ɑi] and [ɑt] in Middle Chinese, belonged to different Old Chinese rhyme categories in Dai Zhen’s classification (i.e., Dai’s 20 and 21 in Table 2), but were treated as members of the same category by the other scholars. This divergence was caused by the enigmatic rhyming of characters of the 靄‎ category (which should not have had a [t]-ending in Old Chinese) with characters of the 遏‎ category (which probably ended in [t] in Old Chinese as it did in Middle Chinese) in Old Chinese poetry, which was confirmed by the phonetic connections revealed in the phonosemantically compounded character forms. For example, in Middle Chinese, the pronunciations [muk] (ru tone) and [bio] (falling tone) for the characters 木‎ and 附‎ did not rhyme in principle, but the two characters obviously rhymed in Old Chinese poetry; similarly, 寺‎ and 特‎ were pronounced [zi] (falling tone) and [dək] (ru tone) in Middle Chinese, but the ancient character-designers used the former as the phonetic component of the latter. These phenomena are still troubling modern scholars to a significant extent.7

The study of Old Chinese initials received less attention, but was nonetheless productive, mainly because of the work of Qian Daxin (1728–1804). Drawing on the abundant character variations and other evidence from ancient records, and comparing them with the 36 initials of (late) Middle Chinese, he correctly concluded that the Old Chinese labial and coronal initials were much less complicated than those in the Middle Chinese system.

In a sense, the study of character pronunciation before the Qing Dynasty, although fruitful, was not of the utmost importance to philologists, because it did not bear directly on the understanding of classical texts unless one takes into consideration the crude etymology in Shiming. But during the Qing Dynasty, the situation was completely reversed. With the sound system of Old Chinese gradually disclosed, character pronunciation study secured a central place in the whole of philology, because scholars began to realize that classical texts could not be thoroughly understood without a firm knowledge of the ancients’ pronunciation, which was often the basis for the correct understanding of the original meanings of the characters. Hence the famous claim by Duan Yucai that “what is the most important in the study of the classics is to find out the correct meaning, and the most proper way to do so is by means of phonetic analysis” (quoted from Wang, 1981, p. 129). In this sense, the development of pronunciation study in the Qing Dynasty could be understood from a broader perspective of the interaction between xiaoxue—the traditional Chinese philology—and daxue (“greater learning”)—the traditional Chinese study of philosophical thoughts—because the strong need for “correct understanding” and “firm knowledge” about classical texts, as Yü (1975, 1996) has pointed out, had a deep root in the “inner logic” of the intellectual history of China, and the progression of daxue turned out to be an important driving force for the development of Old Chinese pronunciation study in the Qing Dynasty.

In the investigation of Old Chinese pronunciation, the phonological orientation which had already shown up conspicuously in Middle Chinese studies became a defining quality, as scholars found it feasible to classify the rhymes, but difficult to determine the phonetic rationale underlying the classification. Duan Yucai, according to his contemporary Chen (1984, p. 14), famously said that if he could be enlightened on the phonetic nature of the distinction between the rhyme categories zhi (支‎), zhi (脂‎) and zhi (之‎), which he had proposed himself on the basis of firm evidence, he “would die content.” Of course, phonetics was never a weak spot in Indic linguistics; its underdevelopment in Chinese linguistics had a direct relevance to the use of characters, but the fact that traditional descriptions of sounds—not just speech sounds—based on the whole Chinese philosophical system had always existed should not be neglected. The problem confronting Duan Yucai represented a bottleneck for traditional Chinese linguistics in its culmination during the Qing Dynasty, and its solution had to wait till Bernhard Karlgren’s reconstruction using the Western method of phonetic description and historical comparison.

3.4 Developments in the Studies of Character Form and Meaning

In the study of character form, the tradition laid down by Xu Shen in Shuowen Jiezi continued to grow after the Han Dynasty. New dictionaries were made to enlarge the character inventory or to satisfy the practical need of the public. In the Jin Dynasty (265–420), a dictionary titled Zilin (“Forest of Characters”) was compiled by Lü Chen (4th century), which contained 12,824 characters, including many ancient and variant forms not found in Shuowen Jiezi. The book was very influential during the Southern and Northern Dynasties and the Tang Dynasty, but was lost after the Song Dynasty. In the Southern and Northern Dynasties, a more practical dictionary titled Yupian (“The Jade Book”) was compiled by Gu Yewang (519–581), in which all the characters assumed the popular style of regular script and were given detailed explanations. Afterwards, many more practical dictionaries appeared from the Tang to the Ming Dynasty, including the Ganlu Zishu (“Character Book for Success in Official Career”) and Wujing Wenzi (“Characters of the Five Classics”) of the Tang Dynasty, compiled for participants of imperial examinations, the Peixi (“Children’s Book,” literally “Spike of the Girdle”) of the Song Dynasty, used as a children’s textbook, the Longkan Shoujian (“Hand Mirror of the Dragon Niche”) of the Liao Dynasty (907–1125), which contained many folk character forms, and the Zihui (“Collection of Characters”) of the Ming Dynasty, which was very popular for its reader-friendly presentation. Of course, there were also more ambitious works, like the Pianhai (“Ocean of Texts”) of the Jin Dynasty (1115–1234), which contained 54,595 characters.

In the making of character-form dictionaries, Shuowen Jiezi was naturally used as a most fundamental reference. But with the development of new ideas and techniques, the layout of dictionaries changed considerably. The Tang-Dynasty Wujing Wenzi, the Liao-Dynasty Longkan Shoujian, and the Ming-Dynasty Zihui, for instance, contained only 160, 242, and 214 radical categories respectively, based on the authors’ combination of the 540 categories in Shuowen Jiezi. In Zihui, the categories were further ordered according to their number of strokes, which made the dictionary much easier to consult. However, even greater changes occurred under the strong influence of pronunciation studies. For example, the Tang-Dynasty Ganlu Zishu, while focusing on character form, was arranged by tones and “rhyming groups.” In Longkan Shoujian and Pianhai, although radical categories remained, these categories—and the characters under them—were arranged either completely by the four tones (in Longkan Shoujian) or by the combined use of the 36 initials, the four tones, and the number of strokes (in Pianhai).

In the theoretical analysis of character form, centered around Shuowen Jiezi, the influence of pronunciation studies was even more profound, although the “six-type” theory about characters kept developing as a general framework. At the turn of the Five Dynasties (907–960) to the Song Dynasty, Xu Xuan (916–991) and his brother Xu Kai (920–974) each produced a finely revised edition of Shuowen Jiezi (with Xu Kai’s edition accompanied by his own commentary). Their focus was on correcting the mistakes in previous editions, but following the trend of the time, they also provided a fanqie spelling for each of the characters explained and even compiled another book based on materials from Shuowen Jiezi, using “rhyming groups” for categorization. Over a century later, the new book enlightened another one titled Shuowen Jiezi Wuyin Yunpu (“Chart of Shuowen Jiezi Based on Sounds and Rhymes”), which became so popular that for a time, the original Shuowen Jiezi of the radical-categorization format became unknown (He, 2013, p. 307).

In the Song Dynasty, attention to pronunciation further led to the development of the so-called youwenshuo (“right-part theory”), according to which the phonetic component, often occupying the right-hand part of the phonosemantically compounded characters, are not merely indicators of pronunciation, but have to do with meaning as well. This theory was never fully expounded and did not really work as a universal rule, but it was based on a perceptive observation of the phonetic relations between cognates reflected in the invention of their characters. In the Yuan and the Ming Dynasty, the idea of the interrelations between character form, sound, and meaning was again discussed by Dai Tong (1200–1285), in his Liushu Gu (“Origins of the Six Types of Characters), and Zhao Yiguang (1559–1625), in his gigantic Shuowen Changjian (“Extended Commentary on Shuowen Jiezi”). Of course, it was only in the Qing Dynasty, when systematic studies of Old Chinese pronunciation were developed, that reliable conclusions could be made about how character forms are connected through pronunciation. In this respect, Duan Yucai’s work was the most representative. In his Liushu Yinyun Biao, which has been mentioned in the preceding discussion of Old Chinese rhyme studies, he made it a principle that characters sharing the same phonetic component in form must belong to the same rhyme category in their Old Chinese pronunciation, and he made a list of 1,521 phonetic components classified under his 17 rhyme categories. According to Duan Yucai’s principle, the nine characters in Figure 8, for instance, should belong to the same rhyme category, because 之‎ is the phonetic component of 芝‎, 志‎, and 寺‎ (though this is no longer obvious in the regular forms of the characters), while 寺‎ is the phonetic component of 詩‎, 侍‎, and 時‎, and 時‎ is the phonetic component of 榯‎ and 鰣‎. Further, as之‎, 寺‎, and 時‎ are phonetic components under the same rhyme category, one can make the inference that all the characters with these phonetic components must also belong to this rhyme category. Quite remarkably, for many characters, the inference can be confirmed by the rhyming relations in classical poetry, though this is not always the case because most characters were not created at the same time as the poetical works.

Language and Linguistics in Pre-Modern China and East AsiaClick to view larger

Figure 8. Nine characters sharing phonetic components with each other consecutively (in seal and regular script) (adapted from Hu, 2002, p. 238).

Drawing on his finding about the relations between phonetic components and rhyme categories, Duan Yucai was able to make some very perceptive observations. For example, scholars before his time sometimes used sha (殺‎, to kill) to freely replace shi (弑‎, to kill (one’s superior)) when editing classical texts, probably because the two characters were close in meaning and sounded somewhat similar. However, Duan pointed out that since 殺‎ and 弑‎ had different phonetic components (杀‎ in 殺‎ and 式‎ in 弑‎), they belonged to different rhyme categories in Old Chinese, and since “phonetic loan could only happen between characters of the same rhyme category” (Duan, quoted from He, 2013, p. 535), 殺‎ and 弑‎, instead of being easily interchangeable, had to be quite distinct words to the ancients. This argument, grounded on purely linguistic evidence, was an insightful observation from the social perspective as well, considering how important social ranking was to the ancient Chinese.

Duan Yucai’s principle was widely shared by scholars of the Qing Dynasty, and proved crucial to many of their findings on Old Chinese pronunciation and character meaning. The essential relations it has established between character form and pronunciation are still recognized in modern research.

In the study of character meaning, the encyclopedic Erya of the Han Dynasty was followed by a large number of similar dictionaries, all with the morpheme ya (originally “proper (language),” but now “dictionary in the Erya tradition”) in the title. These include the Guangya (“Expanded Ya”) by Zhang Yi (3rd century) of the Three Kingdoms Period, the Piya (“Increased Ya”) by Lu Dian (1041–1102) of the Song Dynasty, the Erya Yi (“Wings of Erya”) by Luo Yuan (1136–1184), also of the Song Dynasty, the Tongya (“Comprehensive Ya”) by Fang Yizhi (1611–1671) of the Qing Dynasty, and many others. Of all the ya books, which collectively form the so-called yaxue (“study of ya”) tradition, the most important are Erya, Guangya, and Tongya. It has been said that in the study of historical Chinese lexicology, these three books are the must-haves for the pre-Qin period, the Han Dynasty and the Three Kingdoms Period, and for the Tang, Song, Yuan, and Ming Dynasties respectively (He, 2013, p. 413). On the other hand, there are also many other special ya books that are worthy of attention. For example, the Pianya (“Ya of Couplets”) by Zhu Mouhan (16th century) of the Ming Dynasty is a valuable collection of two-syllable words and expressions; the Biya (“Ya of Synonyms”) by Hong Liangji (1746–1809) of the Qing Dynasty contains many quotations from ancient works in which synonyms are given and explained.

Despite the encyclopedic knowledge provided by the ya books, of direct relevance to the understanding of character meaning in classical texts are the innumerable commentaries written by philologists over two millennia of time. Since the Han Dynasty, two canons of classical texts have existed in China: the works of the pre-Qin scholars (and some later works) and the Buddhist scriptures. Both canons have large numbers of commentaries. For the Buddhist canon, the most important commentaries (as far as Chinese philology is concerned) are two books of the same name, both written in the Tang Dynasty: the Yiqiejing Yinyi (“Sound and Meaning of All Sutras”) by Xuanying (7th century), in 25 volumes, and the Yiqiejing Yinyi by Huilin (736–820), in 100 volumes. For the Chinese canon, the most important commentators include Zheng Xuan (127–200) of the Han Dynasty, Lu Deming of the Tang Dynasty, Zhu Xi (1130–1200) of the Song Dynasty, and Duan Yucai, Wang Niansun, and many others of the Qing Dynasty. From a linguistic point a view, commentaries are not systematic studies of language and are not all about meaning, but deal with character form and pronunciation as well. However, they can hoard incredibly valuable information and insights. On the other hand, character form and pronunciation, especially the latter, have proved indispensable to the interpretation of classical works. Lu Deming’s commentary and the two Yiqiejing Yinyi, for example, greatly benefited from the widespread use of fanqie during the Tang Dynasty. In the Qing Dynasty, the systematic studies of Old Chinese pronunciation further led to revolutions in the investigation of etymological meanings and the identification of phonetic loans. Compared with the Han-Dynasty Shiming and the Song-Dynasty “right-part theory,” etymological work of the Qing Dynasty was much less arbitrary. For example, the Old Chinese first person pronouns 吾‎, 卬‎, 言‎, and 我‎ were recognized by Dai Zhen as cognates, not because they all sounded similar in his own tongue, nor because they had the same phonetic component in writing, but because they shared the same initial consonant and the same basic function in Old Chinese. For the initials in the cognate relations, Dai Zhen further made a distinction between tongwei and weitong, which were roughly equivalent to “with the same place of articulation” and “with the same manner of articulation” (He, 2013, p. 520). In the study of phonetic loans, the Qing-Dynasty scholars corrected many misunderstandings caused by the failure to see a phonetic loan for what it is (e.g., the use of 亡‎ (“die; kill”) to stand for 盟‎ (“ally”) on the basis of purely phonetic similarity). Their achievements are not only found in the many commentaries on classical texts, but are systematically presented in their studies of Shuowen Jiezi. Of the four best-known Qing-Dynasty scholars of Shuowen Jiezi, namely Duan Yucai, Gui Fu (1736–1805), Zhu Junsheng (1788–1858), and Wang Yun (1784–1845), Duan Yucai, and Zhu Junsheng are the most representative regarding the clarification of etymological meanings, extended meanings, and uses as phonetic loans of the characters.

Meanings—or uses—of function words caught the attention of scholars as early as in the Han Dynasty, when commentators made a distinction between ming (“content words”) and ci (“empty words,” “function words”), and started to discuss the various functions of ci. In these discussions, some germinal ideas in the line of grammatical studies appeared. After the introduction of Buddhism into China, Indian vyākaraṇa (“grammar”) works came to be known to certain Chinese (mainly Buddhist) scholars, who gave sporadic and sketchy descriptions of these books as well as the grammatical features of Sanskrit in their Buddhist writings. Modern scholars have found that it was also in these writings that the Chinese words for “grammar” and “syntax”—yufa and wenfa—made their first appearance (Sun, 2001, pp. 150–152, 171–177). However, the influence of the Indic grammatical tradition in China was generally very limited and never triggered a boost in the study of Chinese grammar. Even in the discussion of function words, it was not until the 14th century that the first book on this particular topic, the Yuzhu (“Function Words”) by Lu Yiwei (?–?) of the Yuan Dynasty, had appeared. This short volume, intended as a practical guide for students, listed 135 function words and expressions, classified them under 66 categories, and explained their uses through comparisons from multiple perspectives. Its rare interest in grammatical issues was continued in the Qing Dynasty by three other works: the Xuzi Shuo (“Explanations of Function Words”) by Yuan Renlin (16th to 17th century), the Zhuzi Bianlüe (“Brief Notes on the Analysis of Function Words”) by Liu Qi (16th to 17th century), and the Jingzhuan Shici (“Explanations of Words in the Classics and Commentaries”) by Wang Yinzhi (1766–1834). Like Yuzhu, Xuzi Shuo was also a practical aid for students. It introduced about 60 function words and laid special emphasis on the delicate variation of speaking manners in their use. In contrast, Zhuzi Bianlüe and Jingzhuan Shici were much more exegetical and included a large number of quotations from classical texts. In Zhuzi Bianlüe, over 470 function words were presented in the order of the four tones and the “rhyming groups,” but were actually classified into 30 functional categories. The classification, however, has been criticized for being inconsistent. In comparison with Zhuzi Bianlüe, Jingzhuan Shici is less inclusive, containing only 160 function words. But in a sense, the book is the epitome of the Qing-Dynasty philological approach applied to function words. Besides providing many more examples for each word explained, it employs the ancient initials as a framework for classification and presentation, not simply for convenience’ sake, but because the author intends to highlight the etymological relations between various function words as revealed through their Old Chinese pronunciation and use. For example, 由‎, 以‎, and 用‎, all meaning “with” or “by means of” and all sharing the yi initial (pronounced [j] in Middle Chinese and [l] in Old Chinese), appear in the same chapter and are discussed with cross-references to each other.

4. Language and Linguistics in the Non-Chinese World

In ancient East Asia, the character-centered study of Chinese was not only a traditional undertaking for Chinese scholars, but was widely pursued in the non-Chinese world as well. This was undoubtedly a result of the cultural and political influence of China, but as more and more Chinese morphemes were borrowed into the non-Chinese languages, linguistic knowledge of Chinese became indispensable to the investigation of these languages per se. At the same time, scripts based on or influenced by Chinese characters were designed and adopted by speakers of Tangut, Vietnamese, Korean, Japanese, and so on, and for some languages, systematic studies were established, some on the model of the Chinese xiaoxue, others with more innovations.

As early as in the 1st century bc, Chinese characters and classics had been introduced into Vietnam. At the turn of the 2nd to 3rd century ad (i.e., the end of the Chinese Han Dynasty), a large number of Chinese scholars, including Liu Xi, author of Shiming, fled to Vietnam to escape the social chaos in China and brought with them numerous Chinese books as well as the essence of the Han-Dynasty philology. Afterwards, Chinese studies became popular in Vietnam and flourished in the Chinese domination period of the 7th and 8th centuries (the Tang Dynasty in China) as well as the Lý and Trần Dynasties (1009–1400). Generally speaking, the Vietnamese took a much more practical than theoretical interest in the study of Chinese. Even in the works of such great scholars as Lê Quý Đôn (1726–1784), who stood as one of the most erudite Confucianists in the history of Vietnam and authored the encyclopedic Vân đài loại ngữ (“Classified Sayings from the Archives”), discussions on language and writing were more diffuse than systematic, though not without insights (Stankevič, 2000, p. 59).

In Korea, formal education of Chinese classics started in the 4th century ad, and with it, philological works like Yupian became known. In the 7th century, Korean students began to be sent to China in large numbers, bringing back many more philological works. The period when studies of Chinese were the most developed was the first two centuries of the Joseon Dynasty (1392–1897), which enjoyed such illustrious scholars as Sin Sukju (1417–1475) and Choe Sejin (1473–1542). The Koreans have preserved many valuable philological works from China. For example, the most complete version of the 100-volume Yiqiejing Yinyi by Huilin of the Tang-Dynasty is found in the Tripitaka Koreana of the 13th century. Moreover, Korean scholars composed numerous works of their own, including both dictionaries and textbooks, e.g., the Saseong Tonggo (“Comprehensive Survey of the Four Tones”) by Sin Sukju and the Saseong Tonghae (“Comprehensive Interpretation of the Four Tones”) by Choe Sejin, both “rhyme books” adapted from Chinese sources, and the Nogeoldae (“Old Cathayan”) and Bak Tongsa (“Pak the Interpreter”), both extremely popular textbooks teaching the colloquial Northern Chinese of the Yuan Dynasty. After the invention of Hangul (the indigenous Korean alphabet) in the mid-15th century, this phonemic script was used to transcribe both literary and colloquial Chinese pronunciations of the characters in Saseong Tonggo, Saseong Tonghae, and other works. At the same time, Hangul transcriptions of Korean pronunciations of the characters (ultimately also borrowed from China), known as eonhae (“explanatory notes in Hangul”), also abounded. These transcriptions are among the most precious parts of the Korean philological documents and provide valuable information for the study of historical Chinese phonology.

The study of Chinese in Japan started with the introduction of Chinese classics from Korea, which is traditionally traced back to the 3rd century ad. In the 7th century, regular contacts between Japan and China began, and large numbers of Japanese students were sent to China, while some Chinese scholars went to Japan. Since then, Chinese philology kept growing as a prominent branch of learning till the 19th century, and became a full-fledged tradition with Japanese characteristics. Compared with their Chinese teachers and colleagues, Japanese scholars often exhibited a broader vision, both in the materials they considered and in the methods they adopted. Regarding the materials, the Japanese were in possession of some important documents long lost in China. For instance, the famous Yunjing had been carefully studied in Japan for over 600 years when it was “refound” by Chinese scholars at the end of the Qing Dynasty. Similarly, the Sisheng Pu by Shen Yue is believed to be preserved as a chapter in Kūkai’s Bunkyō Hifuron. Moreover, as Chinese morphemes were borrowed into Japanese in different historical periods, different systems of pronunciation developed in Japanese for the Chinese characters,8 and added to the scholars’ “raw data.” For example, 京‎ is pronounced /keː/ in京师‎ (keishi, “capital”), but /kjoː/ in东京‎ (Tokyō, “Tokyo”), the former following the kan’on (“Han pronunciation”) system borrowed from the capital of China during its Tang Dynasty, the latter the go’on (“Wu pronunciation”) system which possibly originated earlier in the south of China and reached Japan via Korea. These systems, together with their Chinese sources as well as their Korean parallel, were all of concern to Japanese scholars in their study of Chinese character pronunciation. Methodologically, Japanese scholars were profoundly influenced by the Siddham analysis of Sanskrit pronunciation popular in Tang-Dynasty China and had a keen interest in the comparison between Chinese, Japanese, Korean, Sanskrit, and other languages. With the development of their syllabic scripts Katakana and Hiragana since the 9th century, they were able to transcribe character pronunciation in a similar way as Korean scholars later did with their Hangul. Of the many works on Chinese philology by Japanese scholars, some of the most important include the Bunkyō Hifuron and Tenrei Bansho Meigi (“A Dictionary of the Names and Meanings of Everything in Seal and Clerical Script”) by Kūkai, the Tōgū Setsuin (“The Togu Qieyun”) by Sugawara no Koreyoshi (812–880), the Shittan Zō (“The Siddham Repository”) by Annen (841–?), the Makō Inkyō (“The Polished Mirror of Rhymes”) by Bun’yū (1700–1763), the Kanji San’on Kō (“A Study of the Three Sound Systems of Chinese Characters”) by Motoori Norinaga (1730–1801), and the Kan-Go Onchō (“Characteristics of the Kan’on and Go’on Sound Systems”) by Ōta Zensai (1759–1829).

While studying and contributing to the philology of Chinese, speakers from the non-Chinese world started working on their own languages. Naturally, the first step in such work would have to be the invention of a writing system, and Chinese was certainly a ready source of inspiration. According to Zhou (1997), at least 15 languages in East Asia have developed individual scripts based on or with some relevance to Chinese characters. A summary of these scripts is given in Table 3.

Table 3. East Asian Scripts Developed on the Basis of or with Relevance to Chinese Characters.

Language and Linguistics in Pre-Modern China and East Asia

It should be pointed out that some scripts in the table were invented rather late (e.g., Lisu syllabary and the Laozhai Hmong characters were invented around the 1950s, and Nüshu script may have been invented in the Qing Dynasty), but are included because they certainly belong to the Chinese-character tradition. Also, most scripts did not come into being at once. For example, the Japanese Katakana and Hiragana derived gradually in the 9th century from the use of simplified Chinese characters or their components for phonetic transcription. From a linguistic point of view, the Korean Hangul deserves some special mention, because its design was cleverly based on certain phonetic observations (e.g., the use of ㄱ, a symbol depicting the tongue touching the velum, to represent /k/) and interestingly guided by the traditional Chinese philosophy about yin, yang, and the “five elements,” all of which were explained in detail in its “manual” Hunmin Jeongeum (“Proper Pronunciation for the Instruction of the People”), officially issued in 1446.

As the non-Chinese languages were written down, for some of them, systematic research work was further carried out. Relatively speaking, the studies of Japanese and Tangut—an extinct Tibeto-Burman language used in the Tangut Empire (1038–1227)—were the most developed, and it so happened that they stood at two extremes in their relationship with the Chinese philological framework.

The study of Tangut was almost completely modeled on Chinese philology. Not long after the invention of its script in the early 11th century—using strokes from Chinese characters to make radicals and combining the latter on ideographic and phonosemantic principles, Tangut began to have a “rhyme book”—the Wenhai Baoyun (“Treasured Rhyme of the Literary Sea”)—and a “rhyme table”—the Wuyin Qieyun (“Qieyun Based on the Five Sounds”). In both books, tones, rhymes, and initials were used to categorize the Tangut characters, and fanqie was adopted to indicate their pronunciation. In the 12th century, five comprehensive dictionaries were further compiled, centered around the sound, form, and meaning of Tangut characters as Chinese philological works were around those of Chinese characters: The Tongyin (or Yintong, both meaning “Identical Sounds”), in which 6,133 characters were categorized under nine initials, the Wenhai (“The Literary Ocean”), which was like a Tangut Shuowen Jiezi except that the characters were arranged in a similar way to that of Qieyun, the Shengli Yihai (“Ocean of Meanings Authorized by the Emperor”), an encyclopedia arranged by semantic categories, the Yitong Yilei (“Synonyms”), a Tangut thesaurus, and the Fanhan Heshi Zhangzhongzhu (“A Timely Tangut–Chinese Pearl in the Palm”), a Tangut–Chinese dictionary also arranged by semantic categories.

In the study of Japanese, the various systems of Chinese pronunciation were admittedly an area shared with Chinese philology, and the so-called “fifty-sound table” of the Japanese syllabary, as well as the early description of Japanese pitch accents, also showed influences from the Chinese (and the Indic) tradition. However, even within the area of character pronunciation, Chinese philology was not of much help to the resolution of the spelling problems facing Fujiwara Teika (1162–1241) or the search for Old Japanese orthography by Keichū (1640–1701), both of which ultimately had to do with the sound changes of Japanese itself. And when it came to the large number of grammatical particles in their agglutinative language, Japanese scholars were completely on their own in their exploration. In the early poetic anthology Man’yōshū (“Collection of Ten Thousand Leaves”), two poems of the 8th century were already accompanied with “grammatical notes”, as they unusually dispensed with some necessary particles. In the Kokugaku (“National Learning”) movement of the late 17th and the 18th century, outstanding contributions to the study of Japanese grammar were made by Motoori Norinaga and Fujitani Nariakira (1738–1779). Both scholars discussed the Japanese word categories and grammatical particles in detail, Motoori in his Tenioha Himokagami (“Mirror of the Tenioha Particles”) and Mikuni Kotaba Katsuyōshō (“Conjugation Table of the Language of the Honorable Country”), and Fujitani in his Ayuishō (“On Ayui”) and Kazashishō (“On Kazashi”). In comparison, Fujitani was more original and unique. While Motoori occasionally drew upon Chinese terms to express his ideas, Fujitani’s terminology was mostly his own, and carefully based on metaphors from everyday life (Miller, 1975, pp. 1249–1251).

Further Reading

Auroux, S., et al. (Ed.) (2000). History of the language sciences, Vol. 1 (Chapters II, III, & IV). Berlin: Walter de Gruyter.Find this resource:

Brown, K. (Ed.) (2005). Encyclopedia of language and linguistics (2d ed.). Entries on “Chinese linguistic tradition”, “Japan: history of linguistics”, etc. New York: Elsevier.Find this resource:

Doi, T. (1976). The study of language in Japan: A historical survey. Tokyo: Sinozaki Shorin.Find this resource:

Gao, X.-F.高小方‎. (2005). Zhongguo yuyanwenzixue shiliaoxue (中國語言文字學史料學‎). Nanjing, China: Nanjing University Press.Find this resource:

Goddard, C. (2005). The languages of East and Southeast Asia: An introduction. New York: Oxford.Find this resource:

Halliday, M. A. K. (1981). The origin and early development of Chinese phonological theory. In R. E. Asher & E. J. A. Henderson (Eds.), Towards a history of phonetics (pp. 123–140). Edinburgh: Edinburgh University Press.Find this resource:

He, J.-Y.何九盈‎. (2013). Zhongguo gudai yuyanxueshi (中國古代語言學史‎) (4th edition). Beijing: The Commercial Press.Find this resource:

Hirata, S.平田昌司‎ (2016). Wenhua zhidu he hanyushi (文化制度和漢語史‎). Beijing: Beijing University Press.Find this resource:

Hu, Q.-G.胡奇光‎. (2005). Zhongguo xiaoxueshi (中國小學史‎). Shanghai: Shanghai People’s Publishing House.Find this resource:

Koerner, E. F. K., & Asher, R. E. (Eds.). (1995). Concise history of the language sciences: From the Sumerians to the cognitivists, Section III: Antiquity—The Far East. Oxford: Pergamon.Find this resource:

Lepschy, G. (Ed.). (1994). History of linguistics, Vol. 1. London: Longman.Find this resource:

Li, X.-K.李新魁‎, & Mai, Y.麥耘‎. (1993). Yunxue guji shuyao (韻學古籍述要‎). Xi’an, China: Shaanxi People’s Publishing House.Find this resource:

Miller, R. A. (1975). The Far East. In T. A. Sebeok et al. (Eds.), Current trends in linguistics, Vol. 13, Historiography of linguistics (pp. 1213–1264). The Hague: Mouton.Find this resource:

Norman, J. (1988). Chinese. Cambridge, U.K.: Cambridge University Press.Find this resource:

Pu, Z.-Z.濮之珍‎. (Ed.). (2004). Zhongguo lidai yuyanxuejia (中國歷代語言學家‎). Shanghai: Shanghai Cultural Publishing House.Find this resource:

Sybesma, R., et al. (Eds.). (2015). Encyclopedia of Chinese language and linguistics. Leiden: Brill. (Entries on “Chinese and Japanese”, “Chinese and Korean”, “Chinese and Vietnamese”, “Chinese linguistics in Japan”, “Chinese linguistics in Korea”, “Chinese loanwords in Vietnamese”, “děng”, “děngyùxué”, “Ěryǎ”, “fǎnqiè”, “rhyme table”, “Shuōwén Jiězì”, “xiéshēng”, “xùngǔxué”, etc.)Find this resource:

Wang, L.王力‎. (1981). Zhongguo yuyanxueshi (中國語言學史‎). Taiyuan, China: Shanxi People’s Publishing House.Find this resource:

Yu, Y.H.俞允海‎. (1999). Zhongguo yuyanwenzixue mingzhu tiyao (中國語言文字學名著提要‎). Beijing: China Youth Press.Find this resource:

Zhang, S.-L.張世祿‎. (1984). Zhongguo yinyunxueshi (中國音韻學史‎). Shanghai: Shanghai Shudian.Find this resource:

Zhao, Y.-T. (1559–1625). 趙蔭棠‎. Dengyun yuanliu (等韻源流‎). Beijing: The Commercial Press.Find this resource:


Chen Y.-Q. (1934). Sisheng san wen. Qinghua Xuebao, 9(2), 275–287.Find this resource:

Chen, L. (1984). Investigation of Qieyun (Qieyun Kao). Vol. 6. Beijing: Zhongguo Shudian.Find this resource:

Fung, Y.-L. (1948). A short history of Chinese philosophy. New York: Macmillan.Find this resource:

Ge, Z.-G. (2009). Zhongguo sixiangshi. Shanghai: Fudan University Press.Find this resource:

He, J.-Y. (2013). Zhongguo gudai yuyanxueshi. Beijing: The Commercial Press.Find this resource:

Hirata, S. (2016). Wenhua zhidu he hanyushi. Beijing: Beijing University Press.Find this resource:

Hu, A.-S. (2002). Yinyunxue tonglun. Beijing: Zhonghua Shuju.Find this resource:

Hu, Q.-G. (2005). Zhongguo xiaoxueshi. Shanghai: Shanghai People’s Publishing House.Find this resource:

Karlgren, B. (1940). Grammata Serica. Stockholm: Museum of Far Eastern Antiquities.Find this resource:

Li, F.-K. (1973). Languages and dialects. Journal of Chinese Linguistics, 1(1), 1–13.Find this resource:

Miller, R. A. (1975). The Far East. In T. A. Sebeok et al. (Eds.), Current trends in linguistics, Vol. 13, Historiography of linguistics (pp. 1213–1264). The Hague: Mouton.Find this resource:

Robins, R. (1997). A short history of linguistics. London: Longman.Find this resource:

Stankevič, N. (2000), La tradition linguistique vietnamienne et ses contacts avec la tradition chinoise. In S. Auroux et al. (Eds.), History of the language sciences, Vol. 1 (pp. 58–62). Berlin: Walter de Gruyter.Find this resource:

Sun, L.-M. (2001). Zhongguo gudai yufaxue tanjiu. Beijing: The Commercial Press.Find this resource:

Wang, L. (1981). Zhongguo yuyanxueshi. Taiyuan, China: Shanxi People’s Publishing House.Find this resource:

Yang, J.-Q. (2005). Hanyu yinyunxue jiangyi. Shanghai: Fudan University Press.Find this resource:

Yü, Y.-S. (1975). Some preliminary observations on the rise of Ch’ing Confucian intellectualism. Tsing Hua Journal of Chinese Studies (new series), 11(1 & 2), 105–146.Find this resource:

Yü, Y.-S. (1996). Zhang Xuecheng versus Dai Zhen: A study in intellectual challenge and response in eighteenth-century China. In P. J. Ivanhoe (Ed.), Chinese language, thought and culture: Nivison and his critics (pp. 121–154). Chicago: Open Court.Find this resource:

Zhou, Y.-G. (1997). Shijie wenzi fazhanshi. Shanghai: Shanghai Education Press.Find this resource:


(1.) The ancestry of some major East Asian languages, including Japanese and Korean, is still unclear. Furthermore, the Tai languages and the Hmong-Mien languages are treated by most Chinese scholars, who follow the classification of Li (1973), as subbranches of the Sino-Tibetan family, but in the West, they are normally regarded as separate families.

(9.) To avoid the confusion of the Jin (晋‎) Dynasty (265–420) and the Jin (金‎) Dynasty (1115–1234), dates are provided in the text wherever one of the two dynasties is mentioned. Otherwise, they are given only when a dynasty is mentioned for the first time.

(2.) Originally, the term xiaoxue referred to the ancient institution for primary education.

(3.) This subfield is sometimes further divided into pianpang (“character components”) and zishu (“character writing”), i.e., the internal structure and the written style of characters.

(4.) The exact phonetic values of these tones are not known for sure. Moreover, there is no consensus on when exactly the four tones came into existence. It may have been that they became full-fledged around the 3rd or 4th century, i.e., not too long before the Southern and Northern Dynasties. If that is the case, the development of the tones themselves may have also contributed to their discovery.

(5.) A “rhyme” in traditional Chinese linguistics is different from a “final” in that the former does not include the medial (if there is one). It is also different from the “rhyme” in modern linguistics, because it includes the tone of the syllable.

(6.) In each of these rhyming groups, there is actually a third final with a [u] medial (i.e., [ua] of the level, rising, and departing tones), which is not shown in this table.

(7.) A popular solution to the problem is to propose an original voiced-plosive ending for those characters whose pronunciation ended with a vowel in Middle Chinese, so that these characters could be phonetically paired up with their rhyming partners which had always had a voiceless-plosive ending (e.g., one could argue that the [i] in the Middle Chinese [ɑi] had originally been [d] in Old Chinese, and thus explain why the [t]-ending 遏‎ category could rhyme with the [d]-ending 靄‎ category). However, this would finally lead to the construction of an Old Chinese sound system with no open syllables, which is typologically rare and unacceptable to some scholars.

(8.) Besides the various systems of Chinese-based pronunciation, called ondoku (“phonetic reading”), Chinese characters in Japanese often had native Japanese pronunciations, called kundoku (“reading with reference to meaning”), when the characters are used to write native Japanese morphemes.