2.2.1National Romanization

Because Chinese writing is slightly phonetic, it is easy for us today to forget how the widely-used Pinyin alphabet came into being. Paralleled with the Vernacular Literature Movement, there had been movements towards the unification of the National Language. The idea of writing Chinese with an alphabet had been taken up with much vigor, and more than one practical scheme of alphabetizing had been worked out. But there were many technical difficulties as well as social and political hurdles to be surmounted before any form of an alphabet could be used as a general means of writing.

In 1916, Chao’s first article “The Problem of the Chinese Language” appeared on The Chinese Students’ Monthly in May and June issues. It includes four parts: scientific study of Chinese philology, Chinese phonetics, the teaching of Chinese as it is, and proposed reforms. Hu Shih, his Cornell friend, contributed to the third part. Chao begins by indicating that “There exists a problem of the Chinese language. […] In taking up the problem, I should like to emphasize at the outset the distinction between scientific, or historic research on the one hand, and constructive reforms on the other” (Zhao & Huang 1998Zhao, Xinna & Peiyun Huang 1998 Zhao Yuanren nianpu 赵元任年谱 [Chronology of Yuen Ren Chao]. Beijing: Commercial Press.: 83, Chao 2002Chao, Yuen Ren 2002 Zhao yuanren yuyanxue lunwenji 赵元任语言学论文集 [Selected Essays in Linguistics] ed. by Zongji Wu & Xinna Zhao. Beijing: Commercial Press.: 668). This seminal paper sets the tone for his lifelong pursuit of language study, e.g. dialect and spoken Chinese. He points out that the pronunciation of the same character is vastly different in different dialects. He encourages linguists to investigate dialects because he claims that the interaction between spoken and written language necessitates a holistic study of language and provides the spoken language a status it deserves.

Chao’s stance as an ardent advocate of Chinese language reform was unique for his time. Like Hu Shih, he advocated the reform of Chinese language and character. He maintained that there was no contradiction between guided reform and the natural development of a language. Spearheading a National Romanization movement, he whole-heartedly swept away all theoretical oppositions to alphabetizing the Chinese script and shaped the Gwoyeu Romatzyh 国语罗马字 “Chinese Romanization scheme”. In 1923, he published two articles “Ten Objections to Romanizing Chinese” and “Principles for Constructing a Practical System of Romanization”, in which he pointed out: “It is a Romanization of Chinese, by the Chinese, and (primarily) for the Chinese” (Zhao & Huang 1998Zhao, Xinna & Peiyun Huang 1998 Zhao Yuanren nianpu 赵元任年谱 [Chronology of Yuen Ren Chao]. Beijing: Commercial Press.: 122).

On September 26th, 1925, the shù rén huì 数人会 “Society of a Few Men” was formed at Chao’s Tsing Hua home.1616.The name shù rén huì came from an anecdote in the preface to Qièyùn 切韵 (see footnote No.50 for more). At the famous rhyme discussion, a consensus was reached by the principle of “wǒ bèi shù rén, dìng zé dìng yi” 我辈数人，定则定矣 [We few men decide and it is decided]. Other members of the shù rén huì included important linguists Liu Fu (1891–1934), Qian Xuantong (1887–1939), Li Jinxi (1890–1978), Wang Yi (ca.1875–1960), Lin Yutang (1895–1976), and Zhou Bianming (1891–1984). During its 22 meetings, the draft of a National Romanization scheme was developed; and the Peking dialect was chosen as the standard pronunciation.1717.The dialect of Peking, on account of its having been the capital of China for more than one dynasty, has the greatest prestige. The Committee on Unification of the National Language in December 1925 approved this proposal.1818.The Committee on Unification of the National Language was organized under the auspices of the Ministry of Education, to unify the spoken language of the nation. The result was a standard Dictionary of National Pronunciation, with the pronunciation given in a new alphabet of 40 letters called National Phonetic Alphabet. The form of Mandarin thus defined, so far as the sound is concerned, is called guóyǔ 国语 “National Language”. When the government in 1928 adopted Chao’s proposed scheme of Gwoyeu Romatzyh, he was so enthusiastic that he recorded in his diary the announcement in this alphabet.

In 1959, Chao gave a series of 16 lectures on the topic of “The Problem of the Chinese Language” at the National Taiwan University. His lectures were compiled into a booklet published under the same title. For anyone interested in Chinese and its relevant problems, the publication is a must-read. Even today, many of his opinions in the booklet are not outdated, and contemporary scholars quote them.

2.2.2Chao’s surveys of dialects

China has many different dialects, which are unintelligible to each other if the speakers of the localities are far enough apart. But people of all provinces inherit the same classical literature, write in the same literary styles, and can communicate in writing with perfect ease without betraying their locality. There is no fundamental difference of vocabulary among the dialects. As to grammar or syntax, it is practically identical in all the dialects; in fact, even the literary idiom has about the same syntax as the colloquial. From these considerations, it follows that a thorough knowledge of any one of the dialects gives one the key to the whole language (Chao 1925Chao, Yuen Ren 1925 A Phonograph Course in the Chinese National Language. Shanghai: Commercial Press.: xii).

In September 1927, when Chao was still teaching at Tsing Hua, he conducted the first survey of the Wu dialects. At that time transportation was so difficult that Chao and his assistants had to take boats, ferries, and buses to reach the villages, since trains could only reach the main areas. The recording would take as long as 8 to 9 hours a day; and to be more efficient in the field, Chao prepared ahead for one or two months. Everywhere he went, he would learn the local dialect. For instance, after two weeks in Guangzhou, he could make a speech in Cantonese. His diligence paid off and the first survey Xiàndài wúyǔ de yánjiū 现代吴语的研究 [Studies in the Modern Wu Dialects] was published in 1928Chao, Yuen Ren 1928 Xiandai wuyu de yanjiu 现代吴语的研究 [Studies in the Modern Wu Dialects]. (= Monograph No.4 .) Peking: Tsing Hua College Research Institute.. As the first attempt using a modern approach, it laid a foundation for the systematic investigation of Chinese dialects.

The first fieldwork at the Academia Sinica was carried out in the provinces of Guangdong and Guangxi where the dialectal varieties were the richest. After nearly three months’ field work in 1928, he recorded a large amount of data in Yue dialect, but those data were not published, except for the textbook Cantonese Primer (1947Chao, Yuen Ren & Lien Sheng Yang 1947 Guoyu zidian 国语字典 [Concise Dictionary of Spoken Chinese]. Cambridge: Harvard University Press.) and the article “The Zhōngshān Dialect” (1949). Early equipments like a sliding pitch pipe were used in the survey of Wu dialects, and wax cylinders in the surveys of the Yue dialect. Despite the primitive conditions, the surveys were ably handled by Chao and his assistants.

In 1932, Chao’s dialect field work was interrupted by his appointment to succeed Mei Yiqi (1889–1962) as the Director of the Chinese Educational Mission in Washington, D.C., when Mei was called back to be the President of Tsing Hua University. While he was in the U.S., Chao ordered more sophisticated recording equipments for the Institute. The Chaos went back to Shanghai in the fall of 1933. In 1934, they moved to Nanjing where the Institute was relocated; and Chao soon resumed his dialect investigation. He and his assistants recorded speakers in An Hui province in 1934, and in Jiangxi and Hunan in 1935, and in Hubei province in 1936. Upon its publication in 1948, the report on the survey in Hubei attracted attention from linguists in and outside China, and it became a model for dialect surveys. Chao’s table of 3,567 characters has been widely referenced for surveys in the 1980s. After his remarkable work on six surveys in several provinces and areas, he gained a thorough understanding of the sound distinctions across these areas, and he concluded that “The geographical sound variance represents the historical evolution” (1959: 99).

Chao’s earliest published work includes essays and translations that introduce Chinese readers to Western science and literature. His translation of Lewis Carroll’s Alice in Wonderland was an early experimentation on modern spoken Chinese. Published in 1922, it was seen as a landmark in Chinese literary history. Through his translation practice, he developed his own concepts; and he contended that fidelity should be observed as the main desideratum, but its dimensions were neither measurable nor completely independent (Chao 1976Chao, Yuen Ren 1976 Aspects of Chinese Sociolinguistics: Essays. Stanford: Stanford University Press.: 148–169).

Last but not least, his paper “The Non-uniqueness of Phonemic Solutions of Phonetic Systems” (1934) was a milestone in this first phase.1919.Chao said he was not satisfied with his Chinese title “Yīnwèi biāoyīnfǎ de duō néng xìng” 音位标音法的多能性 because it tended to confuse the theoretical work of building a system with the formal and practical work of marking the phonemes (Chao 1959Chao, Yuen Ren 1959 Yuyan wenti 语言问题 [The Problem of Language]. Taipei: School of Humanities at National Taiwan University.: 35). Writing in English, Chao claimed that the phonemic solutions of phonetic systems are neither unique nor definite. In the paper, he demonstrated remarkable ability to identify critical questions in linguistic theory.

In 1936, Chao’s decision to decline the offer from the University of Hawaii clearly reflected his enjoyment of his work, and it summed up his two major interests in this period. “For the present, my vocation is dialect survey, and my hobby is Romanization schemes for popular education, both of which have to be done right here in the country, so I shall have to stay put as long as I am able to carry on my work” (Zhao & Huang 1998Zhao, Xinna & Peiyun Huang 1998 Zhao Yuanren nianpu 赵元任年谱 [Chronology of Yuen Ren Chao]. Beijing: Commercial Press.: 212–213). At that time, he did not expect that the imminent war would cause him to change his mind and move to Hawaii in 1938.

2.3.1Chao’s language instruction in the American universities

In the spring of 1922, Chao offered the first Chinese class at Harvard since the death of Ge Kunhua (1836–1882).2020.Chinese language was not taught on a regular basis at Harvard until 1921. In 1937 Chinese and Japanese instruction, previously offered in the Department of Semitic Languages and History, found a more appropriated home in the newly created Department of Far Eastern Languages (also referred to at the time as the Division of Far Eastern languages). In 1941, the Department offered its first doctorate, jointly with the Department of History. Most doctorate degrees awarded in Far Eastern Languages between 1941 and 1972 were in fact joint degrees in History and Far Eastern Languages. He was appointed to a one-year renewable position as Instructor of Chinese. In the 1920s, Chinese was typically taught by asking students to translate chapters of Classical Chinese into English.

Departing from the conventional way of merely translating the text into English, Chao taught Classical Chinese as if it were a living language, and he encouraged students to read and answer questions aloud in class. He believed that the knowledge of the spoken language is the best introduction to an appreciation of the literature. His teaching order was pronunciation first, then syntax, vocabulary, and idioms. The Chinese tones are reputed among Western speakers to be the most difficult part of the language. It is so only because from a Western point of view, tones seem to be something added because people use pitch of the voice for expression rather than as constituent parts of words. Chao compared the three layers of sound: consonant, vowels, and tones, using terminology from physics. He emphasized the importance of getting the pronunciation correct in the first weeks (Chao 1925Chao, Yuen Ren 1925 A Phonograph Course in the Chinese National Language. Shanghai: Commercial Press.: xv).

The highlight of Chao’s teaching was the intensive ASTP program at Harvard.2121.Hu Shih recommended Chao to Carl J. Friedrich (1901–1984), the director of the Harvard School of Overseas Administration, saying that if he wanted to see success and effectiveness in this class, he should put it in Chao’s able hand, instead of the American instructors (Hu 1970Hu, Shih 1970 Hu Shih gei Yuen Ren Chao de xin 胡适给赵元任的信 [Seventy-eight Letters to a Friend]. Taipei: The Budding Publishing Co.: 6). Hu Shih was appointed in Jan, 1943 as research associate and consultant of the American Council of Learned Societies (ACLS) after his retirement from the Chinese Ambassador to the United States (Hu 2003Hu, Shih 2003 Hu Shih quanji [Collected Works of Hu Shih]. vol, LXIV. Anhui: Anhui jiaoyu chubanshe.: 521, 537). ACLS sponsored for Intensive Language Program (ILP) and cooperated closely with the United States Armed Forces Institute (USAFI). For the history of USAFI’s involvement in language teaching and dictionary compilation, cf. Hall (1991)Hall, Robert A. 1991 “165 Broadway: A crucial node in American structural linguistics . Historiographia Linguistca 18:1.153–166. . He was happily eager and busy to teach over two hundred students at one time.2222.Back in the late 20s, enrollment in Chinese language classes ranged from three to nine. In September 1959, when Chao’s class of Introductory Chinese at Berkeley had 25 students, he said it was “more than ever before” (Zhao & Huang 1998Zhao, Xinna & Peiyun Huang 1998 Zhao Yuanren nianpu 赵元任年谱 [Chronology of Yuen Ren Chao]. Beijing: Commercial Press.: 371). Many Chinese students from Harvard and MIT were hired as assistants in the drill sessions, and his audio-lingual method and direct methods made the program a great success. The textbook that grew out of this program, Mandarin Primer (1948), with accompanying character texts, is a highly concentrated body of materials summarizing the essential features of modern spoken Chinese. His daughter remembers:

It is not an easy textbook either for the students or for the teacher. But for those who are willing to put their full effort into it from the very beginning, it is a very efficient way to gain both the active and passive use of the language.(Pian 1995Pian, Rulan Chao 1995 “Autobiographical Sketches”. ACMR Reports 8:1.1–20.: 13)

Eventually, some Army students were launched into careers in Asian studies, and they became leading sinologists: for example, Frederick W. Mote (1922–2005) of Princeton and James Crump (1921–2002) of University of Michigan. It is no exaggeration to say that Chao eased the path of Westerners who desired to learn, in the Chinese language, the thoughts and ideals of the Chinese people.

Based on his experience in dialect teaching, Chao produced another textbook, Cantonese Primer (1947Chao, Yuen Ren & Lien Sheng Yang 1947 Guoyu zidian 国语字典 [Concise Dictionary of Spoken Chinese]. Cambridge: Harvard University Press.). When participating in the Chinese dictionary project at Harvard, together with Lien Sheng Yang (1914–1990), Chao composed The Guóyǔ zìdiǎn 国语字典 [Concise Dictionary of Spoken Chinese ] (1947Chao, Yuen Ren & Lien Sheng Yang 1947 Guoyu zidian 国语字典 [Concise Dictionary of Spoken Chinese]. Cambridge: Harvard University Press.), which remains the best dictionary of colloquial Chinese. Chao so much valued spoken language that he produced Readings in Sayable Chinese (1968Chao, Yuen Ren 1968 A Grammar of Spoken Chinese. Berkeley & Los Angeles: University of California Press.) specifically to benefit people in their speech.2323.Chao once criticized that Hu Shih and Lu Xun (1881–1936), the father of modern Chinese literature, wrote with a vernacular language that was not sayable.

Chao’s unmatched teaching has become legendary. Zhou Zhiping (b.1947), a former director of “Middlebury Chinese Summer School” and founder of “Princeton in Beijing”, reviewed Chao’s feat of finding a path for Chinese language teaching in the U.S. (2015aZhou, Zhiping 2015a “Meiguo zhongwen jiaoxue shi shang de Zhao Yuanren” (1) 美国中文教学史上的赵元任（上） [Yuen Ren Chao in the History of Chinese Language Instruction in the United States (1)]. Journal of International Chinese Teaching 国际汉语教学研究 5:1. 90–96., bZhou, Zhiping 2015b “Meiguo zhongwen jiaoxue shi shang de Zhao Yuanren” (2) 美国中文教学史上的赵元任（下） [Yuen Ren Chao in the History of Chinese Language Instruction in the United States (2)]. Journal of International Chinese Teaching 国际汉语教学研究 6:2. 83–92.). These programs of high repute adhere to Chao’s ASTP pedagogy — significant attention to accurate speech, an earlier introduction to Chinese characters, teaching reading through the spoken language, and teaching Classical Chinese through modern colloquial.

2.3.2A Grammar of Spoken Chinese

In both Cantonese Primer and Mandarin Primer, Chao included a chapter on grammar. Ever since then, he had been urged to write a full-length book discussing Chinese grammar. To get himself ready, he had been collecting materials, jotting down notes, and writing some articles on special topics in the grammar of spoken Chinese. He also read grammars written by others. In his autobiography, Chao recalled buying a copy of Mǎ shì wén tōng and finding it very exciting reading (Chao 1975Chao, Yuen Ren 1975 Yuen Ren Chao’s Autobiography: First 30 Years, 1892–1921, [Life with Chaos: The autobiography of a Chinese family], vol. II. Ithaca, N.Y.: Spoken Language Services, Inc.: 64). Later he read grammar books written by Wang Li, Gao Mingkai (1911–1965), Li Jinxi,2424.Li Jinxi was a friend of Chao, and also a member of the shù rén huì and Committee on Unification of the National Language. His Xīn zhù guóyǔ wénfǎ 新著国语文法 [New Grammar of the National Language] (1924) was the first grammar to systematically describe modern Chinese on English models. and Lü Shuxiang (1904–1998). In 1954, when he went to Europe on a Guggenheim Fellowship, Chao decided that he would write a book on grammar; and he would write it in a structuralist framework. He continued to collect raw materials and record live conversations during his sojourn in Taiwan in 1959Chao, Yuen Ren 1959 Yuyan wenti 语言问题 [The Problem of Language]. Taipei: School of Humanities at National Taiwan University.. Most of the actual writing was not begun until the Fall of 1960, when the project was funded by a grant under the National Defense Education Act (NDEA), which was established by Congress in 1958, following the Russian launch of the Sputnik. When A Grammar of Spoken Chinese (1968) got published, he was already 75 years old.

In his most cited book, Chao offered sharp insights into the nature of Chinese which can be summarized as follows: “The grammatical meaning of subject and predicate in a Chinese sentence is topic and comment, rather than actor and action” (1968: 69) and “The full sentence is made up of minor sentences and minor sentences are adequate and more primary than full sentences” (p. 83). Because these perceptions have been accepted by an increasing number of scholars, the monograph continues to remain a most influential work.

The President of Chinese Linguistic Society, Shen Jiaxuan (b.1946) wrote two articles in commemoration of the 120th birthday of Yuen Ren Chao. Shen (2012)Shen, Jiaxuan 2012 “ ‘Lingju’ he ‘liushuiju’ — wei Zhao Yuanren xiansheng danchen 120 zhounian er zuo” “零句”和“流水句” — 为赵元任先生诞辰120周年而作 [On Minor Sentences and Flowing Sentences in Chinese: In commemoration of the 120th birthday of Yuen Ren Chao]. Studies of the Chinese Language 中国语文 5.403–415. introduced in great length Chao’s minor sentences and Lü Shuxiang’s flowing sentences and challenged the universalness of structural recursion and a noun-verb distinction. Shen (2014)Shen, Jiaxuan 2014 “Hanyu de luoji zhegeyang, hanyu shi zheyang de — wei Zhao Yuanren xiansheng danchen 120 zhounian er zuo zhi er” 汉语的逻辑这个样，汉语是这样的 — 为赵元任先生诞辰120周年而作之二 [A Review of Yuen Ren Chao’s Two Articles on Chinese Grammar and Logic]. Language Teaching and Linguistic Studies 语言教学与研究 2.1–10. asserted that Chao insightfully indicated that Chinese logic is determined by the characteristics of Chinese grammar. These articles revived the concepts and theories raised by Chao half a century earlier.

More recently, Shen referred to Chao’s work in his new book Mingci he dongci 名词和动词 [Nouns and Verbs] (2015Shen, Jiaxuan 2015 Mingci he dongci 名词和动词 [Nouns and Verbs]. Beijing: Commercial Press.) where Shen stated that while nouns and verbs are two separate syntactic categories in Western languages, they are pragmatic categories in Chinese and they are not separate entities. In other words, Chinese nouns constitute a super-noun category with verbs as a sub-category. He said that his apparently “appalling” statements were indeed based on his total agreement with great scholars, like Yuen Ren Chao, Lü Shuxiang, Wang Li and Zhu Dexi (1920–1992) (Shen 2011Shen, Jiaxuan 2011 “Zhu Dexi xiansheng zui zhongyao de xueshu yichan” 朱德熙先生最重要的学术遗产 [The Most Valuable Legacy of Zhu Dexi’s Grammatical Theory]. Language Teaching and Linguistic Studies 语言教学与研究 4.7–19.).

Chao contributed greatly to the standardization and spread of Mandarin speech, dialect investigation, Chinese instruction, grammar, and general linguistic theories. His distinguished career won him the highest honor as he was elected the President of the Linguistic Society of America (LSA) in 1945. On Dec 31st, 1945, he presented his presidential address “The Logical Structure of Chinese Words” which was revised and published in Language in 1946. A special April-June issue of Language was dedicated to him in 1966. He also served as the President of the American Oriental Society (AOS) and gave a presidential address “What is Correct Chinese?” in 1961. The inaugural issue of Journal of Chinese Linguistics was dedicated to him in 1973. Earlier, in 1948, he had been elected member of American Academy of Arts and Sciences, section Phonology and Criticism. Chao is deservedly recognized as the most eminent linguist from China.

3.3.1Works on grammar

In grammar, Wang Li criticized the mistake of completely copying the Western model that was ill-suited for the analysis of Chinese. His first paper Zhōngguó wénfǎxué chūtán 中国文法学初探 “An exploratory study of Chinese grammar” (1936) emphasized the importance of a distinction between the Classical Chinese and modern Chinese grammars, and laid a solid foundation for a new study of grammar. In his eyes, this paper was a manifesto to launch his linguistic career. His next paper, Zhōngguó wénfǎ zhōng de xìcí 中国文法中的系词 “The copulative particles in Chinese grammar” (1937) proved that shì 是 “be” was not a copulative particle in Old Chinese, but evolved from the demonstrative shì to become a copula in modern Chinese. In Chinese, especially Old Chinese, copulative particle is not a necessary sentential component. Even verbs are not always necessary in Chinese sentences. This was the first paper on historical grammar, and it set a model for later grammarians. Developing his own system, his Zhōngguó xiàndài yǔfǎ 中国现代语法 [Modern Chinese Grammar] (1943–1944Wang, Li 1943, 1944 Zhongguo xiandai yufa 中国现代语法 [Modern Chinese Grammar]. Shanghai: Commercial Press.) was first intended as one book, but recommended by poet and scholar Wen Yiduo (1899–1946) to be published as two separate ones. Writer Zhu Ziqing (1898–1948), the then chair of the Department of Chinese at Tsing Hua University, wrote a long preface to Wang Li’s grammar book, titled Zhōngguóyǔ de tèzhēng zài nàlǐ 中国语的特征在那里 “The characteristics of Chinese”.2929.Zhu Ziqing was famous for his poetry and prose writing. People seldom knew that he studied linguistics in England. This preface not only commented on the extant grammars, praised Wang Li’s seminal work, but showed Zhu’s profound understanding of the unique characteristics of Chinese. Zhu gave a detailed summary of Wang Li’s pioneering work and pointed out his independence without blindly following Western models. (Zhu 1948Zhu, Ziqing 1948 Yuwen lingshi 语文零拾 [Selected Essays]. Shanghai: Mingshan shuju.: 54–65). When the second volume of Zhōngguó yǔfǎ lǐlùn 中国语法理论 [Theory of Chinese Grammar] (1944–1945Wang, Li 1944, 1945 Zhongguo yufa lilun 中国语法理论 [Theory of Chinese Grammar]. Shanghai: Commercial Press.) was published, the companion volumes highly established him in the study of grammar. He realized that the rule of grammar should only be drawn from the language facts and empirical data. In his works he analyzed, examined and concluded the characteristic features of Chinese grammar, based on the rich data from the eighteenth century vernacular novel Hóng lóu mèng 红楼梦 [Dream of the Red Chamber] (1791).

The second volume of his Hànyǔ shǐgǎo 汉语史稿 [A Draft History of Chinese] (1957–1958) was regarded as seminal work on a diachronic study of Chinese grammar from Old Chinese to the present. Later it was published separately as Hànyǔ yǔfǎ shǐ 汉语语法史 [History of Chinese Grammar] (1989) after revisions and replenishment. Even today, this is still regarded as the most comprehensive, authoritative, and profound work in the field.

3.3.2Works on phonology

Upon returning to China from Paris in 1932, Wang Li taught general linguistics and Chinese phonology at Tsing Hua University and Chinese phonology at Yenching University. His Zhōngguó yīnyùn xué 中国音韵学 [Studies of Chinese Phonology] (1936) was the first comprehensive treatise on the subject written in Chinese. It revealed the veil of the traditional phonology that intimidated many beginners studying Chinese. He explained traditional phonology in modern terms, commented concisely on the achievements of Qing philologists with extensive citations and introduced the Old and Middle Chinese reconstruction of Bernhard Karlgren. Inspired by Zhang Binglin, Wang Li raised his own creative proposal of zhī/wēi 支/微 distinction in Old Chinese rhyme groups. His work transformed the traditional phonology into a modern science. A revised edition was published under a new title as Hànyǔ yīnyùn xué 汉语音韵学 [Studies of Chinese Phonology] (1956Wang, Li 1956 Hanyu yinyun xue 汉语音韵学 [Studies of Chinese Phonology]. Beijing: Zhong Hua Book Company.).

His three volumes of Hànyǔ shǐgǎo represent his crowning achievement in historical linguistics. Its first volume, Hanyu yuyin shi 汉语语音史 [History of Chinese Phonetics] (1985), deals with historical phonology. It divides the history into nine chronological periods from pre-Qin to modern times and remains a comprehensive work on the history dated back from Old to Middle Chinese, and from Middle Chinese to modern. His reconstruction of Old Chinese was very influential in China and established him as one of the three giant Old Chinese phonologists.3030.The other two phonologists renowned for reconstruction of Old Chinese are Li Fang-Kuei and Zhengzhang Shangfang (b.1933). An important paper for today’s reader to understand his reconstructive system and method on Old Chinese rhymes is his Xiānqín gǔyùn nǐcè wèntí 先秦古韵拟测问题 “Problems on the reconstruction of pre-Qin rhymes” (1964). Wang Li built his own theory of Old Chinese and applied his system to mark the rhymes of two classic poetic works. The two companions Shījīng yùndú 诗经韵读 [Rhyme Reader of Odes] and Chǔcí yùndú 楚辞韵读 [Rhyme Reader of Chu Songs] were both published in 1980.3131.The two books on rhyme readings were reprinted in Collected Works of Wang Li, vol. XII (2014Wang, Li 2014 Shijing yundu; Chuci yundu 诗经韵读; 楚辞韵读 [Rhyme Reader of Odes; Rhyme Reader of Chu Songs]. Collected Works of Wang Li , vol. XII. Beijing: Zhong Hua Book Company.). They offered an elaboration of the rhyming rules for both insiders and laymen to understand and grasp the Old Chinese pronunciation.

3.3.3Works on lexicology and dictionaries

In lexicology, Wang Li also had significant accomplishments and contributions. The third volume of his Hànyǔ shǐgǎo treats the development of the lexicon. It summarizes the works done by predecessors, contemporaries and the author himself, and facilitates the teaching and research in historical lexicology.

Based on his cogent criticism of the ancient lexicons and modern dictionary, Wang Li presented three principles that dictionaries should follow in his article Lǐxiǎng de zìdiǎn 理想的字典 “Ideal dictionary” (1945). His Liǎoyī xiǎo zìdiǎn chūgǎo 了一小字典初稿 “First draft of a small dictionary” (1946) was a pilot trial to apply these principles.3232. Liǎoyī is Wang Li’s courtesy name. For each character, there were pronunciation, parts of speech, meaning explanations, and examples. There was also an etymological index with an old pronunciation and quoted examples of earliest uses in Classics. Xīn xùngūxué 新训诂学 “New exegesis” (1947) was a breakaway from an old study and advocated a new study of glosses or explanations of words in Classics. He discussed the rules in the evolution of words and word meanings in various papers.3333.These articles can be found in Collected Works of Wang Li, vol. XIX (2015).

His Tóngyuán zìdiǎn 同源字典 [Etymological Dictionary] (1982Wang, Li 1982 Tongyuan zidian 同源字典 [Etymological Dictionary]. Beijing: Commercial Press.) was a milestone work on etymology, treating 439 groups of words of same origins, and involving over 3,000 words. Invited by Zhong Hua Book Company, Wang Li started to work on Gǔ hànyǔ zìdiǎn 古汉语字典 [Classical Chinese Dictionary] in 1984. One year later, he finished one fourth of the dictionary. Considering his age and state of health, his several students took over and completed it. It was published in 2000Wang, Li 2000 Wang Li yuyanxue lunwen ji 王力语言学论文集 [Selected Essays in Linguistics]. Beijing: Commercial Press. and it realized his dream of an ideal dictionary.

3.3.4Works on versification

Wang Li was interested in combining language study with classical poetic meter and form. After he completed three books on grammar, he wanted to work on the grammar of poetry which included prosody. He gave a course on “grammar of poetry” and lectured on the beauty of the rhythm and meter in classical poetry when he was teaching at the National Southwestern Associated University in Kunming in 1944. One year later he began the writing but covered more topics and went deeper than the lectures given. When it came out, Hànyǔ shīlǜxué 汉语诗律学 [Chinese Versification] (1958Wang, Li 1958 Hanyu shilüxue 汉语诗律学 [Chinese Versification]. Shanghai: Shanghai jiaoyu chubanshe.) was the first book on genre study from the linguistic perspective. He not only gave a systematic and comprehensive analysis of poetic meter and rhythm; but for the first time, he explained the grammatical features in poetry, cí 词 poetry and qǔ 曲 “songs of Yuan dynasty”, illustrating the differences in word orders and word choices from those in prose. It aroused a critical sensation because it broadened the study of poetic meter through a linguistic perspective.

When the modern poet Guo Xiaochuan (1919–1976) read his essay Lüè lùn yǔyán xíngshì měi 略论语言形式美 “On the formal beauty of language” (1962), he immediately wrote to Wang Li expressing empathy and resonance. Later Guo paid more attention to the rhythm and meter, which greatly enhanced the beauty of his poems when read aloud. During his stay in Paris, Wang Li translated the French poem Les Fleurs du mal by Charles Baudelaire (1821–1867) into classical style of Chinese poetry.3434.Talented as he was, Wang Li said in the prologue wèi xìn shīqíng jù bié cháng, píngshēng zì jiè nòng cí zhāng 为信诗情具别肠，平生自戒弄词章 [I believe that only people who have special sentiment could compose poems. Therefore I have been refraining myself from writing poems for my entire life]. He showed his talent in poetry writing at a young age but being so modest that he never considered himself a poet.

3.3.5Textbooks on Classical Chinese

Wang Li was a master at integrating teaching with research. As mentioned above, most of his books began as syllabi for his courses.3535.Wang Li often said jiǎng yì mén kè, xiě yì běn shū 讲一门课，写一本书 [When a course is given, a textbook is written]. (Wang & Zhang 2008Wang, Jiguo & Gu Zhang 2008 Guowen tongcai: Wang Li 国文通才：王力 [Biography of Wang Li]. Beijing: Peking University Press.: 146). He was so gifted that he could turn the boring grammatical lectures into vivid and interesting ones, attracting not only linguistic students but also students from Chinese literature. He wrote many textbooks and articles on teaching. Gǔdài hànyǔ 古代汉语 [Classical Chinese] (1962–1964Wang, Li 1962–1964 Gudai hanyu 古代汉语 [Classical Chinese]. vol. I-IV. Beijing: Zhong Hua Book Company.) is a textbook compiled under his leadership, combining not only the selected readings, but also high frequency words, descriptions of grammar, and essays on various aspects of traditional culture. It is the earliest work with this mosaic layout, and it continues to be an influential college textbook.

3.3.6Works on language reform

Wang Li was generous and willing to put his linguistic expertise at the service of society at large. As an advisor to the State Commission on Language Reform in 1950s, he worked actively on the reform of characters, standardization and spread of Mandarin. He not only wrote articles urging the importance of simplification of characters, but he also participated in the design and implementation of Pinyin spelling scheme. He even compiled booklets to teach Mandarin to speakers of Wu dialects and Cantonese.

Wang Li trained many famous linguists, and he was generous to offer a hand to students in need. In 1985Wang, Li 1985 Hanyu yuyin shi 汉语语音史 [History of Chinese Phonetics]. Beijing: Zhongguo shehui kexue chubanshe., he donated the royalty from his published Collected Works and set up the “Wang Li Prize in Linguistics” at Peking University to inspire young scholars.

I now end this section with the quotes from two eulogies. In Memoriam written by Tsu-lin Mei (b.1933): “Wang Li devoted his entire life to the advancement of linguistics in China. Linguistics made him a great scholar, and he in turn made Chinese linguistics a modern discipline” (Mei 1986Mei, Tsu-lin 1986 “In Memoriam: Professor Wang Li”. Journal of Chinese Linguistics 14:2.333–336.: 336). In Memoriam written by Lü Shuxiang: “The reason why Wang Li could achieve so much in linguistics was due to his solid foundations in traditional Chinese philology and his training in scientific Western way. He could treat and employ the traditional heritage in a critical manner” (Lü 1986Lü Shuxiang 1986 “Daonian Wang Li jiaoshou” 悼念王力教授 [In Memoriam: Professor Wang Li]. People’s Daily 人民日报. May 15th.).3636. www​.ziliaoku​.org​/rmrb​/1986​-05​-15​-4

3.4.1The study of meaning

The Han dynasty saw the popularity of the Confucian canon and scripture reading. The study of characters, or xiǎo xué, was destined to be subordinate at its very beginning. Characters are the building blocks in numerous volumes of scriptures and Classics. Therefore, they became the object studied by early philologists dating back to the Former Han (also called Western Han, 202 B.C.–8 A.D.). The purpose of xiǎo xué differs from modern linguistics in its full commitment to scripture reading. There are three major areas of study in xiǎo xué: the shapes, sounds and semantic glosses of graphs. The distinction of the three kinds of treatises is not very clear-cut, because few deals only with the shape. And many treat both the sound and meaning of characters. According to Wang Li, the catalogues made by Sìkù quánshū zǒngmù tíyào 四库全书总目提要 [Critical Catalogue of the Qing Imperial Library] are based on the genre rather than the content (Wang 2015Wang, Li 2015 Long chong bing diao zhai wenji 龙虫并雕斋文集 [Popular Essays by Wang Li]. Collected Works of Wang Li , vol. XIX. Beijing: Zhong Hua Book Company.: 328). Though the terms might be borrowed for the sake of convenience, readers should bear in mind that the early study of sound and meaning were no equivalent to the contemporary linguistic definition of phonology and semantics. In addition to the purpose and the material being studied, i.e., the characters, there were critical differences in methodology.

China’s long tradition in philology has seen emergence of influential works, some of which have become classics in the field and have played a dominant role for a long time. Wang (2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 5) lists four monumental treatises on the study of meaning during the Han period. Ěryǎ 尔雅 [Examples of Refined Usage],3737.Separately, ěr means “near, close to”, and yǎ means “elegant and refined language”. Ěryǎ together means nearly standard language, i.e., using common language to explain dialects and special language. the earliest lexicographic work, contains brief definitions of some 4,300 words and phrases arranged by subject category. Like a thesaurus, a common word explains a group of synonyms. Although its authorship and dates remain unsettled,3838.The consensus is that the Ěryǎ was not made by one person or at one time and it was finalized at the end of Warring States (770–476 B.C.) through the work of generations of scholars. it has had a considerable impact on the interpretation of the Classics, and on the development of the language itself. Ěryǎ continues to interest scholars up to modern times as evidenced by the many commentaries and corrections.

One of the best known commentaries, Ěryǎ zhù 尔雅注 [Commentary to the Ěryǎ] was written by Guo Pu (276–324) in the Jin dynasty (265–420). He stated in his preface that his interest started at a young age, and it took him nearly twenty years to study Ěryǎ. He decided to annotate it because he believed that no previous annotators’ work was thorough and that they had many mistakes and missed points (Guo 1983Guo, Pu 1983 Erya Zhu 尔雅注 [Commentary to Erya]. In Siku Quanshu (Wenyuange edition). vol. 221.1–236. Taipei: Commercial Press.: 7).

The second treatise on the study of meaning was the earliest dialect geography, the Fāngyán 方言 [Dialects] by Yang Xiong (53 B.C.–18 A.D.),3939.Some say his year of birth was 58 B.C. a learned philosopher and poet. Besides the ready legacy from previous materials about languages in different areas, he travelled to record dialectal varieties. It took him 27 years to compile the work of fifteen volumes (the later version had thirteen). It includes more than 2,300 words collected from the spoken language of 14 geographical areas. Following the formula of Ěryǎ, it lists the words of similar meanings arranged by subject categories. The uniqueness is that Yang Xiong pointed out the fine distinctions between usages of different geographical regions,4040.An example goes like this. Dǎng 党, xiǎo 晓, zhé 哲, all mean “to know”. In Chǔ state people used “dǎng ”, or “xiǎo”, while in geographic borders between Qí and Sòng, people preferred to use “zhé” (Yang 1983Yang, Xiong 1983 Fangyan 方言 [Dialects]. commented by Guo Pu. In Siku Quanshu (Wenyuange edition). vol. 221.281–382. Taipei: Commercial Press.: 284). as well as the commonly used language that was based on the court and capital. The limitation of the work is that he used characters to record the pronunciation but that could not reflect the phonetic variance (Ji & Wang 1992Ji, Changhong & Peizeng Wang 1992 Zhongguo gudai yuyanxue jia pingzhuan 中国古代语言学家评传 [Biography of Linguists in Ancient China]. Jinan: Shandong jiaoyu chubanshe.). All in all, it offers a new perspective to look at the distinction between synonyms; and it creatively enriches the analysis of the meaning of words. In his preface to Fāngyán zhù 方言注 [Commentary to the Fāngyán], Guo Pu states that with this precious book in hand, people could know other places without actually being there; and they could name things without asking help (Guo 1983Guo, Pu 1983 Erya Zhu 尔雅注 [Commentary to Erya]. In Siku Quanshu (Wenyuange edition). vol. 221.1–236. Taipei: Commercial Press.: 284).

The first major Chinese lexicon on the study of shape was Shuō wén jiě zì 说文解字 [Explaining Single-component Graphs and Analyzing Compound Characters], authored by Xu Shen (ca.58–148?) of Han dynasty. It was the earliest comprehensive dictionary-like work to have survived.4141.Some said that the book was the completed in 100 A.D., but presented to the emperor only in 121 A.D. by the author’s son. Others said that the compilation took him 22 years from 100 to 121 A.D. By the Han period, xíngshēng zì 形声字 “a phonogram consisting of one ideographic and one phonetic element” achieved a predominant position. Both the semantic and phonological make-ups are essential to an understanding of how the Chinese writing system works. In total, Shuō wén jiě zì has 14 chapters and treats 9,353 characters arranged in 540 radicals (or classifiers) (Xu 1983Xu, Shen 1983 Shuo wen jie zi 说文解字 [Explaining Single-component Graphs and Analyzing Compound Characters]. In Siku Quanshu (Wenyuange edition). vol. 223.63–384. Taipei: Commercial Press.: 63). It was innovative to arrange in this method instead of the semantic category adopted by the first two treatises. The method was called xíngxùn 形训 “explaining the meaning in terms of the shape”. Because Chinese words are shape-based rather than phonetic-based, it is enormously helpful for understanding the history and etymology of Chinese characters. The analysis of the meaning was based on the shape so it belonged to the catalogue of Zì shū 字书 “books on characters” in Sìkù quánshū zǒngmù tíyào.

Xu Shen was the first to make systematic use of the ideographic parts as radicals. Without his providing graphs in zhuànshū 篆书 “seal script”4242. Zhuànshū “seal script” got its name from its widespread use on seals. style and its accompanying analysis, it would be impossible to comprehend the actual meaning of characters, because the modern lìshū 隶书 “clerical script”4343. Lìshū was formerly known as “chancery script”, or “clerical script” from its association with various types of clerks employed by the government. style is much simplified and has evolved far away from its primitive pictorial roots. Unavoidably, his over-dependence on the shape caused him to take words too literally or to guess etymologies out of a fancied or conceived reading of the shape. However, this did not depreciate his influence. His system of 540 radicals has ever since served as a model for dictionary compilers for nearly 2,000 years. In the preface and the first paragraph of the prologue, Xu Shen introduced six methods whereby Chinese characters were coined and eight shapes of writing since the unification of the entire country under Emperor Qin Shi Huang (259–210 B.C.). Motivated to clear up the confusion caused by the disagreements between old and new school of Classics scholars, he adopted an effective approach in categorization and interpretation of graphs (Xu 1983Xu, Shen 1983 Shuo wen jie zi 说文解字 [Explaining Single-component Graphs and Analyzing Compound Characters]. In Siku Quanshu (Wenyuange edition). vol. 223.63–384. Taipei: Commercial Press.: 378). This preface was considered by some to be the first linguistic essay, by its own definition.

Shortly after Shuō wén jiě zì, Liu Xi (life dates unknown) of Eastern Han period, compiled the first etymological dictionary, the Shì míng 释名 [Explaining Names] (ca. 200 A.D.) to address objects that people named but didn’t know how the names came into being (Liu 1983Liu, Xi 1983 Shi ming 释名 [Explaining Names]. In Siku Quanshu (Wenyuange edition). vol. 221.383–424. Taipei: Commercial Press.: 385).4444.In the preface, Liu says that names of objects belong to different categories. For objects that people daily call but know nothing about their origins, he compiles the dictionary under 27 categories to discuss their origins. He used rhyming puns in both the literary and the spoken languages of his day to trace the origins of 1,500 words under 27 subject categories such as heaven, earth, body parts, kinship, food and drink, boats, and mourning rites, etc.4545.It gives the meaning of the graph, then the sound, followed by an explanation of the shape or an etymological analysis. An example is like this. Jiàn 涧 means “the water between hills”. Sounding like jiān 间 “between”, it means that “the creek is between the two hills” (Liu 1983Liu, Xi 1983 Shi ming 释名 [Explaining Names]. In Siku Quanshu (Wenyuange edition). vol. 221.383–424. Taipei: Commercial Press.: 389). He believed that the meaning of a character could be guessed based on its pronunciation. Though his shēngxùn 声训 “explaining the meaning in terms of the pronunciation” method was not so reliable, his attempt shed light on the phonological structure of Old Chinese before the third century A.D. (Wang 2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 7). But due to the asymmetry between the evolution of meaning and that of sound, Shì míng encountered difficulties when giving casual and far-fetching explanations.4646.Some graphs hold on to the same pronunciation though their meaning has changed, or other words whose pronunciation undergoes changes but their meaning remains the same. A far-fetching example is Zhì, shí yě, chóng shí zhī yě 痣，食也，虫食之也 [Hemorrhoids, which sound like eating, are caused by bug bites] (Liu 1983Liu, Xi 1983 Shi ming 释名 [Explaining Names]. In Siku Quanshu (Wenyuange edition). vol. 221.383–424. Taipei: Commercial Press.: 421).

In summary, the Han period saw the development of the earliest study of characters and their respective meanings. In addition to the advocacy of Confucian and Classics reading, the language per se led to the early obsession with glosses. Chinese has no inflections, nor is it phonetic-based. The earliest writing had a basically pictographic origin and each character represented a single syllable. Grammatical elements were particularly hard to represent in pictorial form (Norman 1988Norman, Jerry 1988 Chinese. Cambridge & New York: Cambridge University Press.: 59). The need to study the sound elements arose when Sanskrit, a phonetic-based language, was encountered during the sutra translation. So phonology came five centuries later and the first systematic grammar appeared at the end of the 19th century.

3.4.2The study of sound

The interest in sound arose during Northern and Southern dynasties (420–589) when the first wave of foreign influence struck China. Buddhism was introduced into China and Buddhists were inspired to understand Indian phonetics and create a phonological diagram to learn Sanskrit.4747.Buddhists regarded understanding Chan/Zen as ultimate enlightenment, and understanding phonetics as primary enlightenment. They saw the understanding of phonology as a necessary pathway to attain the higher enlightenment of Chan/Zen Buddhism. Also, poetry and prosody were of such a fashion that there was a pressing need for analyzing the rhyme and discovering the rhythm. Wang Li claimed that there was already a cutting of syllables into two segments in the 2nd century A.D., known as the shēngmǔ 声母 “initial” and the yùnmǔ 韵母 “final” in modern times. It was the poet Shen Yue (441–513) who first identified the four tones, i.e., píng 平 “even”, shǎng 上 “rising”, qù 去 “departing/going” and rù 入 “entering”, and to prescribe prosodic rules for poetry composition (2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 8).

3.4.2.1Earliest rhyme book Qièyùn

The Qièyùn [Standard Rhyme Dictionary]4848. Qiè means correct, normalized. Lu Fayan named his book Qièyùn to declare that it is a normalized and standard rhyme dictionary. by Lu Fayan (life dates unknown) appeared in 601 A.D. It had 195 rhyme rubrics and systematized the fǎnqiè 反切 “the alliterator-rhymer spelling system” formula which used the principles of shuāngshēng 双声 “alliteration” and diéyùn 叠韵 “rhyme” to specify the pronunciation of a syllable with two characters, the first showing the initial, and the second, the final.4949.An example of the fǎnqiè segmentation goes like this. The graph dong 东 “east” is cut as de 德 “virtue” and hong 红 “red” in which d is the initial (that alliterated with the glossed graph) and ong is the final (that rhymed with the glossed graph). So dong：de + hong = dong. Fǎnqiè segmentation was a revolution compared to the previous use of homophones or near-homophones. As the result of contact with learned Buddhist scholars from India, fǎnqiè remained the standard method until the twentieth century. It intrigued Qing philologists and modern historical linguists into reconstructing the pronunciation of Old Chinese by working back from the sounds of Middle Chinese.

Lu’s Qièyùn is a lost document. An augmented edition of the Qièyùn, the Guǎngyùn 广韵 [Extension of Qièyùn], appeared in the year 1008 where Qièyùn xù 切韵序 “Preface to Qièyùn” was extant. Lu Fayan at the beginning of the preface recalled the famous rhyme discussion at his home in Chang’an (modern Xi’an), the capital of Sui (581–618) and Tang (618–907) periods.5050.Eight scholars came to spend the night and they started a discussion when the dinner feast came to an end, among whom, were Yan Zhitui (531–591?), Xiao Gai (ca.535–610). One of the scholars, Wei Yanyuan (life dates unknown) said to Fayan: “We are debating heatedly about the differences between old and present sound and rhymes. There were so many disagreements. Why not take some notes? We few men decide and it is decided.” Fayan made an outline based on the discussion. But he did not start working on it until over ten years later when he was dismissed from his officialdom and returned to a private life of study and contemplation. He said that this seminal lexicon of five volumes was not his own creation but a record of those eight impressive minds at the discussion. The dominance of Qièyùn and its formula kept scholars away from studying the real language for the next 700 years.

3.4.2.2Living language study

It was not until the 13th to 17th centuries that the study of Chinese began to turn to the living language. The Yuan dynasty (1279–1368) witnessed a rapid rise of the opera-theater that had a major impact on the sound studies. Zhōngyuán yīnyùn 中原音韵 [Sounds and Rhymes of the Central Plains],5151.Central Plains refers to the Yellow River area if taken widely. Its narrow meaning is the area of Henan (south of the Yellow River). Zhou declared that the language of the Central Plains should be used as the norm and standard when he was consulted about the rhymes of Yuan Opera (Zhou 1960Zhou, Deqing 1960 Zhongyuan yinyun 中原音韵 [Sounds of Rhymes of the Central Plains]. Re-ed. Taipei: Guangwen Book Company.: 1–6). completed in 1324 by Zhou Deqing (1277–1365) and published in 1341, was the first rhyme book that depicted the live language in the northern area. Breaking the traditional practice, the author did not use the fǎnqiè formula, nor did he provide the initial consonant or meaning interpretations. It treats 5,866 characters, arranged into 19 rhyme groups according to the rhyming practice of the běiqǔ 北曲 “northern songs”. His rhyme table made great contributions to the theory of opera and qǔ “songs” study through the elaborate discussions about the system, prosody, composition, and singing techniques. His focus on the actual sound in real language led to the discovery that the pronunciation in the Yuan period had much similarity with modern Mandarin (Wang 2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 9).5252.Compared to those in the Tang or Song dynasties, the four tones of “level, rising, departing/going and entering” in the Yuan period underwent great changes. There are two subcategories in level and no more entering. Entering is distributed in the other three tones. Therefore, Zhōngyuán yīnyùn remains to be an important source to study the history of Mandarin.

Ming period (1368–1644) was the transition from Yuan to Qing in sound study. Another book intended to break the traditional Qièyùn formula was the Hóngwǔ zhèngyùn 洪武正韵 [The Correct Rhymes of the Hóngwǔ Reign] (1375). The authors’ ambition to correct old rhymes with the pronunciation in Central Plains was not realized due to different reasons (Deng 2002Deng, Wenbin 2002 Zhongguo gudai yuyanxue shi 中国古代语言学史 [History of Linguistics in Ancient China]. Chengdu: Bashu shushe.: 197). So it was criticized to be a rhyme table mixing old and new, northern and southern sound. But it is still a good source to study Mandarin and its formation.

Scholars of the late Ming period dug deeper into the initial, rhyme onset, place and manner of articulation, and tones. They began to notice the distinguishing effects of the mouth shape during the articulation. The concept of sìhū 四呼 “four vowel onset types” was developed gradually and used to categorize all the medials or onsets. It had a significant meaning and continues to be used in classifying modern Pinyin system.5353.Thanks are due to Prof. Richard VanNess Simmons for his presentation “The origin of the Chinese sìhū concept of syllable classification in the Ming dynasty” at the NAAHoLS, LSA’s 90th annual meeting in 2016.

To sum up, the introduction of Buddhism into China had a profound impact on the study of language. In addition to the above-mentioned influence on phonology and rhyme dictionaries, it also influenced the development of logical thinking. The logical thinking of the intellectuals in the Tang period when Buddhism reached its peak was more or less affected by the religion. It reflected in the structure of sentences in summarization and integration. Sentences and discourses were complete and coherent as well as concise. From the Tang period to mid-19th century, syntax didn’t undergo much change. Only after the Vernacular Literature Movement, it has experienced dramatic changes, deeply influenced by Western grammars (Wang 2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 461–463).

3.4.3A comprehensive study of Chinese in Qing era

This period was the heyday of philological studies in all respects. During the Qianlong and early Jiaqing reigns from 1796 to 1820, kǎojùxué 考据学 “evidential studies” flourished and the Qian-Jia evidential scholarship reached a new height. A great deal was achieved in systemizing, authenticating, and annotating an enormous corpus of classic texts (Ge 2015Ge, Liangyan 2015 The Scholar and the State: Fiction as political discourse in Late Imperial China. Seattle: University of Washington Press.: 31). In the study of Old Chinese sound, beginning with the great master Gu Yanwu (1613–1682), the Qing philologists made a sustained effort to classify the rhyme categories of Odes.

3.4.3.1Official dictionaries

Kāngxī zìdiǎn 康熙字典 [The Kāngxī Dictionary] (1716), the first comprehensive and closest to a modern dictionary, was composed under the auspice of the emperor.5454.It was the first time the name zìdiǎn 字典 “dictionary” was used. It took thirty compilers five years to finish working on the project. It treats 47,035 graphs and 1,995 variants and still serves as an indispensable tool for sinologists. The 214 radicals of the dictionary have been followed by all Chinese and foreign makers of Chinese dictionaries. Wang Li (2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 9) compared the dictionary to the four previous ones that adopted the same radical arrangement, and pointed out its greatest advantage was an exhaustive enumeration for the example uses of the characters as well as their first occurrence in Classics. In spite of its inadequacy in many respects, the system has gained an unearned increment through its sheer vogue.

3.4.3.2Evidential studies

The control of Qing rulers towards the culture and intellectual thoughts resulted in an escape from the reality and a renewed fascination in textual studies called Hàn xué 汉学 “Han learning”.

As a representative of the commentary works,5555.Other important commentaries and amendments are Shuō wén jiě zì yìzhèng 说文解字义证, authored by Gui Fu (1736–1805), Shuō wén jùdòu 说文句读 by Wang Yun (1784–1854), and Shuō wén tōngxùn dìngshēng 说文通训定声 by Zhu Junsheng (1788–1858). Shuō wén jiě zì zhù 说文解字注 [Commentary to the Shuō wén jiě zì] by Duan Yucai (1735–1815) has been highly acknowledged as “better than the original”. Duan spent thirty years annotating the Han scholar Xu Shen’s magnum opus Shuō wén jiě zì. His exhaustive quotation of ancient literature augmented his analysis of the meaning Xu Shen had attributed to the graphs. He traced their expanded and derived meaning as well. His efforts to establish the original pronunciation of the graphs led to a new discovery in the phonology of Old Chinese. Graphs which contain the same phonetic element should belong to the same rhyme category. Another great philologist of his time Wang Niansun (1744–1832), claimed that “Duan’s commentary work has been peerless in the past 1,700 years”.5656.Wang Niansun believed that the Classics could only be understood when the correct meaning and pronunciation of the graphs used in the ancient texts are known. He claims in the preface to Shuō wén jiě zì zhù, “The study of graphs, xiǎo xué, is done when the meaning and pronunciation are understood; the study of Classics is done when the study of graphs is understood.” (Duan 1981Duan, Yucai 1981 Shuo wen jie zi zhu 说文解字注 [Commentary to Shuo wen jie zi]. Shanghai: Shanghai guji chubanshe.: 1). Wang Li (2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 10) especially praised Duan’s scholarship shown in his critical and historical perspectives in commenting Shuō wén jiě zì.

Wang Niansun, too, made numerous annotations and emendations on classical texts.5757.Wang Niansun’s son, Wang Yinzhi (1766–1834), also a philologist, wrote a book Jīngyì shù wén 经义述闻 “Account of what I have heard from my father concerning the interpretation of the Classics”. Due to the father and son’s outstanding achievements in philology, people like to mention them together as “Wang and son from Gāoyóu (their birthplace, a town in Jiangsu Province)”. His most prominent work was Guǎngyǎ shūzhèng 广雅疏证 [Commentary to Guǎngyǎ] which took him nearly ten years to complete, at the rate of commenting on three graphs a day. He rectified many earlier errors on graphs and supplemented missing places. Another thirty years were dedicated to rectifying this commentary, and 463 items were further revised. His meticulous and thorough knowledge about form, sound, and meaning of graphs in both ancient and contemporary periods enabled him to accomplish a work of notable magnitude. Duan Yucai admired him as the only one who could easily get the meaning by looking at the ancient pronunciation of graphs.5858.In the preface to Guǎngyǎ shūzhèng, Duan Yucai comments: “Huaizu (Wang Niansun’s courtesy name) can get the meaning of Classics through ancient pronunciation. He is the only scholar who excels at this in the whole nation!” (Wang 2004Wang, Niansun 2004 Guangya shuzheng 广雅疏证 [Commentary to Guangya]. Re-ed. Beijing: Zhonghua Book Company.: 1) The method is also called yīn shēng qiú yì 因声求义 “the pronunciation of a graph is seen as a vehicle of concepts and its meaning can be speculated through its pronunciation”. According to Wang Li (2015Wang, Li 2015 Long chong bing diao zhai wenji 龙虫并雕斋文集 [Popular Essays by Wang Li]. Collected Works of Wang Li , vol. XIX. Beijing: Zhong Hua Book Company.: 320), the merits of the two Qing scholars were their adherence to the pronunciation instead of the mere shape of characters. They quoted enough evidence and examples to prove, thus their conclusions were more credible than those far-fetched ones drawn only from the similarity in shape.

3.4.3.3Old Chinese pronunciation

Qing scholars excelled unprecedentedly in studying the pre-Qin pronunciation. At the beginning of Qing era, the master on Classics, Gu Yanwu opened a new chapter in the study of the rhymes of Odes. He compiled five books on the study of Old Chinese pronunciation and classified the old rhymes into ten rhyme groups. Later scholars built on his work and came up with more rhyme groups. Based on his lecture notes in the 1960s at Peking University, Wang Li wrote an entire book Qīngdài gǔyīnxué 清代古音学 [Qing Dynasty Studies of Ancient Pronunciation] (2013bWang, Li 2013b Qingdai guyinxue 清代古音学 [Qing Dynasty Studies of Ancient Pronunciation]. Collected Works of Wang Li, vol. VI. Beijing: Zhong Hua Book Company.). There are twelve chapters introducing in detail fifteen prominent figures in the study of Old Chinese pronunciation on rhyme groups and the initials. Wang Li (2013aWang, Li 2013a Hanyu shigao 汉语史稿 [A Draft History of Chinese]. Collected Works of Wang Li , vol.I. Beijing: Zhong Hua Book Company.: 11) thinks that the number of rhyme groups does not matter too much. It only shows the difference in scope but the belonging of certain characters to certain groups is an agreed conclusion.

To sum up, Qing period saw serious and flourishing works in dictionary compilation, Old Chinese phonological system and classic exegesis. A historical developmental perspective was established.5959.In the preface to Guǎngyǎ shūzhèng, Duan Yucai explains the essence of xiǎo xué as the study of graphs which includes the form, pronunciation and meaning, interdependent on one another. “There are form, pronunciation and meaning in xiǎo xué. The other two can be understood if one is known. But they are all changing. Thus there are forms old and new, pronunciations old and new, and meanings old and new, interdependent on one another. The other five can be understood if one is known” (Wang 2004Wang, Niansun 2004 Guangya shuzheng 广雅疏证 [Commentary to Guangya]. Re-ed. Beijing: Zhonghua Book Company.: 1). Qian-Jia scholars inherited the best of their predecessors, equipped with a solid Hàn xué background, and were honest and passionate in their intellectual pursuits. They advanced xiǎo xué to an unprecedented height to understand ancient thoughts and classic literature.

Once again, Qian-Jia scholars’ devotion to the Classics led to the negligence of the living language and blindness to alien perspectives. A series of significant questions were raised, but much of their work had to wait for Western linguists like Bernard Karlgren or Western-trained Chinese linguists before it could be interpreted in phonetic terms. Chao claims that by studying the sounds of modern dialects and comparing them with the systematization of traditional Chinese phonology, Karlgren had made a reconstruction in all detail (except the actual melodic values of the tones) of the ancient Chinese pronunciation of about 600 A.D., which, after some revisions, has gained wide acceptance among most Occidental and the majority of Chinese scholars (Chao 1961 [¹1948]Chao, Yuen Ren 1961 [1948] Mandarin Primer: An intensive course in spoken Chinese. Cambridge, Mass: Harvard University Press.: 4).

3.4.4The first indigenous Chinese grammar

Grammars did not have a place in the philological tradition. The focus of traditional training in China was on reading and reciting through which grammar was acquired subconsciously. “The development of grammar studies was initially tied to the activities of Western missionaries, who wrote numerous teaching grammars and language textbooks” (Gianninoto 2014Gianninoto, Mariarosaria 2014 “The Development of Chinese Grammars and the Classification of the Parts of Speech”. Language & History 57:2.137–148. : 137). Other than the earliest grammars written by Spanish missionaries in collaboration with native speakers, two treatises in Qing period were a necessary prelude: Zhùzì biàn lüè 助字辨略 [Compendium of Grammatical Particles] (1711) by Liu Qi (life dates unknown) and the Jīng zhuàn shì cí 经传释词 [Explanations of Words in the Classics and the Commentaries] (1798) by Wang Yinzhi (Lepschy 2013Lepschy, Giulio C. ed. 2013 [¹1994] History of Linguistics: The eastern tradition of linguistics. Vol. I. Repr., London & New York: Routledge. [1994]: 22). Adroit at using textual research, Wang Yinzhi gave elaborate explanation and categorization of xūzì 虚字 “empty words, or grammatical particles” in Classics (Wang 1984Wang, Yinzhi 1984 Jing zhuan shi ci 经传释词 [Explanation of Particles in the Classics and the Commentaries]. Changsha: Yuelu shushe.). Though the work followed the philological tradition, focusing mainly on the meaning interpretation, the abundant evidence and convincing conclusion ensured its authenticity among the literati of the 19th century. It was regarded as the main source of inspiration for the first indigenous grammar Mǎ shì wén tōng (Chappell & Peyraube 2014Chappell, Hilary & Alain Peyraube 2014 “The History of Chinese Grammars in Chinese and Western Scholarly Traditions”. Language & History 57:2.107–136. : 111).

The first comprehensive grammar Mǎ shì wén tōng was modeled entirely on grammars of Latin. Its author Ma Jianzhong was the first Chinese to earn a baccalaureate from a French university. After a professional career as a diplomat, Ma devoted the last decade of his life to writing the first grammar.6060.Some say the grammar was possibly co-written with his brother, Ma Xiangbo (1840–1939) who was a Chinese Jesuit priest, scholar and educator, the founder of the modern Fudan University. Before its publication, the Chinese had no idea of a noun-verb division. Ma points out that subject and predicate are the two most important components in a sentence. He also points out that in Chinese there are many sentences without subjects, and that adjectives can directly act as predicates (Ma 1935 [¹1898]Ma, Jianzhong 1935 [¹1898] Ma shi wen tong 马氏文通 [Basic Principles for Writing Clearly and Coherently by Mister Ma]. Shanghai: Commercial Press.) . However, his bias that “Ancients do better than contemporaries in writing” resulted in an array of 7,326 example sentences only from Classical Chinese, and it prevented a thorough understanding of the nature of modern Chinese.

This period is viewed as an infant phase of writing grammars under Indo-European frameworks without much regard to the characteristics of Chinese per se. The limitation of the grammarian is his adoption of a “band-aid approach” for some analyses which shows his over-dependence on the borrowed Western framework. Ma has been criticized for the distortion of the linguistic reality to allow for the aptness of the borrowed models. Yet the book had a significant historical value, and it laid a solid foundation for later scholars to make serious efforts on grammar (Guo 1998Guo, Xiliang 1998 “Zai tan Ma Jianzhong he Ma shi wen tong” 再谈马建忠和《马氏文通》[Ma Jianzhong and Ma shi wen tong Revisited]. Studies of the Chinese Language 中国语文 6.432–434.).