Rhyme dictionary
A rime dictionary, rhyme dictionary, or rime book is a genre of dictionary that records pronunciations for Chinese characters by tone and rhyme, instead of by graphical means like their radicals. The most important rime dictionary tradition began with the , which codified correct pronunciations for reading the classics and writing poetry by combining the reading traditions of north and south China. This work became very popular during the Tang dynasty, and went through a series of revisions and expansions, of which the most famous is the Guangyun.
These dictionaries specify the pronunciations of characters using the method, giving a pair of characters indicating the onset and remainder of the syllable respectively.
The later rime tables gave a significantly more precise and systematic account of the sounds of these dictionaries by tabulating syllables by their onsets, rhyme groups, tones and other properties. The phonological system inferred from these books, often interpreted using the rime tables, is known as Middle Chinese, and has been the key datum for efforts to recover the sounds of early forms of Chinese. It incorporates most of the distinctions found in modern varieties of Chinese, as well as some that are no longer distinguished. It has also been used together with other evidence in the reconstructions of Old Chinese.
Some scholars use the French spelling rime, as used by the Swedish linguist Bernard Karlgren, for the categories described in these works, to distinguish them from the concept of poetic rhyme.
Pronunciation guides
Chinese scholars produced dictionaries to codify reading pronunciations for the correct recitation of the classics and the associated rhyme conventions of regulated verse. The earliest rime dictionary was the by Li Deng of the Three Kingdoms period, containing more than 11,000 characters grouped under the five notes of the ancient Chinese musical scale. The book did not survive, and is known only from descriptions in later works.Various schools of the Jin dynasty and Northern and Southern dynasties produced their own dictionaries, which differed on many points. The most prestigious standards were those of the northern capital Luoyang and the southern capital Jinling. In 601, Lu Fayan published his, an attempt to merge the distinctions in five earlier dictionaries. According to Lu Fayan's preface, the initial plan of the work was drawn up 20 years earlier in consultation with a group of scholars, three from southern China and five from the north. However the final compilation was by Lu alone, after he had retired from government service.
The Qieyun quickly became popular as the standard of cultivated pronunciation during the Tang dynasty. The dictionaries on which it was based fell out of use, and are no longer extant. Several revisions appeared, of which the most important were:
| Date | Compiler | Title |
| 601 | Lù Fǎyán 陸法言 | 切韻 |
| 677 | Zhǎngsūn Nèyán 長孫訥言 | 切韻 |
| 706 | Wáng Rénxū 王仁煦 | 刊謬補缺切韻 |
| 720 | Sūn Miǎn 孫愐 | 唐韻 |
| 751 | Sūn Miǎn 孫愐 | 唐韻 |
| 763–84 | Lǐ Zhōu 李舟 | 切韻 |
In 1008, during the Song dynasty, a group of scholars commissioned by the emperor produced an expanded revision called the Guangyun. The Jiyun was a greatly expanded revision of the Guangyun. Lu's initial work was primarily a guide to pronunciation, with very brief glosses, but later editions included expanded definitions, making them useful as dictionaries.
Until the mid-20th century, the oldest complete rime dictionaries known were the Guangyun and Jiyun, though extant copies of the latter were marred by numerous transcription errors. Thus all studies of the Qieyun tradition were actually based on the Guangyun. Fragments of earlier revisions of the Qieyun were found early in the century among the Dunhuang manuscripts, in Turfan and in Beijing.
When the Qieyun became the national standard in the Tang dynasty, several copyists were engaged in producing manuscripts to meet the great demand for revisions of the work. Particularly prized were copies of Wang Renxu's edition, made in the early 9th century, by Wu Cailuan, a woman famed for her calligraphy. One of these copies was acquired by Emperor Huizong, himself a keen calligrapher. It remained in the palace library until 1926, when part of the library followed the deposed emperor Puyi to Tianjin and then to Changchun, capital of the puppet state of Manchukuo. After the Japanese surrender in 1945, it passed to a book dealer in Changchun, and in 1947 two scholars discovered it in a book market in Liulichang, Beijing. Studies of this almost complete copy have been published by the Chinese linguists Dong Tonghe and Li Rong.
Structure
The Qieyun and its successors all had the same structure. The characters were first divided between the four tones. Because there were more characters of the 'level tone', they occupied two , while the other three tones filled one volume each. The last category or 'entering tone' consisted of words ending in stops -p, -t or -k, corresponding to words ending in nasals -m, -n and -ng in the other three tones. Today, these final stops are generally preserved in southern varieties of Chinese, but have disappeared in most northern ones, including the standard language.Each tone was divided into rhyme groups, traditionally named after the first character of the group, called the . Lu Fayan's edition had 193 rhyme groups, which were expanded to 195 by Zhangsun Nayan and then to 206 by Li Zhou. The following shows the beginning of the first rhyme group of the Guangyun, with first character 東 :
Each rhyme group was subdivided into homophone groups preceded by a small circle called a .
The entry for each character gave a brief explanation of its meaning.
At the end of the entry for the first character of a homophone group was a description of its pronunciation, given by a fanqie| formula, a pair of characters indicating the initial and final respectively. For example, the pronunciation of 東 was described using the characters 德 and 紅 indicating t + uwng =.
The formula was followed by the character 反 or the character 切 , followed by the number of homophonous characters. In the above sample, this formula is followed by the number 十七, indicating that there are 17 entries, including 東, with the same pronunciation.
The order of the rhyme groups within each volume does not seem to follow any rule, except that similar groups were placed together, and corresponding groups in different tones were usually placed in the same order. Where two rhyme groups were similar, there was a tendency to choose exemplary words with the same initial. The table of contents of the Guangyun marks adjacent rhyme groups as , meaning they could rhyme in regulated verse. In the above sample, under the entry for the rhyme group 刪 in the last part the table of contents is the notation "山同用", indicating that this group could rhyme with the following group 山.
The following are the rhyme groups of the Guangyun with their modern names, the finals they include, and the broad rhyme groups they were assigned to in the rime tables. A few entries are re-ordered to place corresponding rhyme groups of different tones in the same row, and darker lines separate the groups:
Phonological system
The rime dictionaries have been intensively studied as important sources on the phonology of medieval Chinese, and the system they reveal has been dubbed Middle Chinese. Since the itself was believed lost until the mid-20th century, most of this work was based on the.The books exhaustively list the syllables and give pronunciations, but do not describe the phonology of the language.
This was first attempted in the rime tables, the oldest of which date from the Song dynasty, but which may represent a tradition going back to the late Tang dynasty.
Though not quite a phonemic analysis, these tables analysed the syllables of the rime books using lists of initials, finals and other features of the syllable.
The initials are further analysed in terms of place and manner of articulation, suggesting inspiration from Indian linguistics, at that time the most advanced in the world.
However the rime tables were compiled some centuries after the Qieyun, and many of its distinctions would have been obscure.
Edwin Pulleyblank treats the rime tables as describing a Late Middle Chinese stage, in contrast to the Early Middle Chinese of the rime dictionaries.
Structural analysis
In his Qièyùn kǎo, the Cantonese scholar Chen Li set out to identify the initial and final categories underlying the fanqie spellings in the Guangyun.The system was clearly not minimal, employing 452 characters as initial spellers and around 1200 as final spellers.
However no character could be used as a speller for itself.
Thus, for example,
- 東 was spelled 德 + 紅.
- 德 was spelled 多 + 特.
- 多 was spelled 德 + 河.
More common segments tended to have the most variants.
Words with the same final would rhyme, but a rhyme group might include between one and four finals with different medial glides, as seen in the above table of rhyme groups.
The inventory of initials Chen obtained resembled the 36 initials of the rime tables, but with significant differences.
In particular the "light lip sounds" and "heavy lip sounds" of the rime tables were not distinguished in the fanqie, while each of the "proper tooth sounds" corresponded to two distinct fanqie initial categories.
Unaware of Chen's work, the Swedish linguist Bernard Karlgren repeated the analysis identifying the initials and finals in the 1910s.
The initials could be divided into two broad types: grave initials, which combine with all finals, and acute initials, with more restricted distribution.
Like Chen, Karlgren noted that in syllables with grave initials, the finals fell into two broad types, now usually referred to as types A and B.
He also noted that these types could be further subdivided into four classes of finals distinguished by the initials with which they could combine.
These classes partially correspond to the four rows or "divisions", traditionally numbered I–IV, of the later rime tables.
The observed combinations of initials and finals are as follows:
Some of the "mixed" finals are actually pairs of type B finals after grave initials, with two distinct homophone groups for each initial, but a single final after acute initials. These pairs, known as, are also marked in the rime tables by splitting them between rows 3 and 4, but their interpretation remains uncertain. There is also no consensus regarding which final of the pair should be identified with the single final occurring after acute initials.