One of the most important sources for the Phags-pa script is the Yuan dynasty book Měnggǔ Zìyùn 蒙古字韻 (mong xol ts.hi 'win in the Phags-pa script) "Mongolian Letters arranged by Rhyme", which is a rhyming dictionary of over 9,000 Chinese ideographs arranged according to phonetic syllables written in the Phags-pa script.
This page presents my preliminary analysis of the text. An annotated transcription of the complete text may be made available at this site in the future.
It is not known when and by whom the original work was composed, but the only extant version was revised and edited by Zhu Zongwen 朱宗文 (whose Mongolian name was Bayan ᠪᠠᠶᠠᠨ) in 1308, based on a collation of several editions that were then current. In all likelihood this work was originally compiled by official order when the Phags-pa script was first promulgated, for use in teaching the new script to Chinese officials.
That Zhu Zongwen was only a revisor of the text and not the original author is made clear by the prefaces to the text written by Liu Geng and Zhu Zongwen himself, which are given below :
In the same way that [the Song dynasty annotator of Du Fu's poetry] Zhao Cigong was a loyal servant to The Poems of Du Fu, now Zhu Boyan is a loyal servant to The Mongolian Rhyming Dictionary for his work in expanding the text and rectifying its errors. However, using the National Script to write Chinese is a most difficult thing. Now that my friend Zhu has managed to write the National Language using the National Script, the depth of his scholarship far exceeds that of other people. This diagram (i.e. the diagram illustrating the pronunciation of Phags-pa letters at the beginning of the book) can most certainly be used as a guide [to the Phags-pa script] by later scholars. I once had two students who came to study calligraphy with me, both of whom were conversant with the study of Mongolian, and were quick to learn and talented. One was Ye Suke, the other one was Zhu Boyan.
Written by Liu Geng (courtesy name Langao) of Keshan on the day of the full moon of the third month of spring in the cyclical year wushen during the Zhida reign period [i.e. the year 1308].
The borders of our empire extend far and wide, and the languages spoken in different parts are not mutually intelligible. Even if you recognise a particular ideograph, but do not know how to pronounce it, then it is as if you are unable to speak. The Mongolian Rhyming Dictionary unites the written and the spoken forms of ideographs, and is truly the key to phonetics and a guide to rhyme. I once used the various Chinese rhyming dictionaries to try to verify the accuracy of this work, but due to the many textual corruptions and mistakes copied from one work to another, it was difficult to determine which [of various conflicting entries in different rhyming dictionaries] was correct and which was wrong. Only the Collected Rhymes Ancient and Modern classifies each syllable according to the four tones [of Classical Chinese]. Because of this I finally realised that the ideographs jiàn, jīng and jiān [all examples of the Ancient Chinese k initial] correspond to the Phags-pa letter GA. It can well be said that the thirty-six initials are covered with clarity and depth by the Collected Rhymes, and for this reason I used this [book] to carefully revise the various existing editions [of the Mongolian Rhyming Dictionary], and listed the errors that I found at the front of the book in the hope that future readers might correct my mistakes.
Written by Zhu Zongwen (courtesy name Yanzhang) of Xin'an on the day before the Festival of Pure Brightness in the cyclical year wushen during the Zhida reign period [i.e. the year 1308].
One Qing dynasty author, Luo Yizhi 羅以智, writing during the Daoguang period (1821-1850), describes a Yuan dynasty printed edition of Menggu Ziyun that he had seen, but the whereabouts of that copy is no longer known, and no other printed editions are known to have survived.
Several manuscript copies of Menggu Ziyun are also described in various Qing sources, but the only extant copy of the text that I know of is a manuscript copy held at the Britsh Library in London (callmark Or. 6972). This manuscript had been bought by the British Museum in 1909 from the widow of the famous Chinese Art Historian, Dr. S. W. Bushell (1844-1908), who may well have acquired the manuscript on his travels through North China and Inner Mongolia in 1872.
The description of the printed and manuscript editions described in Qing sources concur with the British Library manuscript, and they would all seem to be derived from Zhu Zongwen's revision of the text.
The British Library mansuscript has been published several times. An exact copy of the entire text (except for folio xia:5) is given in Basibazi yu Yuandai Hanyu 八思巴字與元代漢語 pages 95-127; whilst an annotated facsimile (including folio xia:5) edited by Professors Junast 照那斯圖 and Yang Naisi 楊耐思 has been published under the title Menggu Ziyun Jiaoben 蒙古字韻校本. The facsimile published by Professors Junast and Yang is a reprint of a facsimile originally published by the Kansai University 関西大學 in 1957, and has also recently been reprinted in the Xuxiu Siku Quanshu 续修四库全书 collection (Shanghai : Shanghai Guji Chubanshe 上海古籍出版社, -2002).
This description is based on my own examination of the original manuscript held at the British Library.
The manuscript is bound in two fascicles, enclosed in a traditional style folding case with a faded yellow silk covering. The title 蒙古字韻全集 is inscribed on the front of the case, whilst a leather patch with the inscription "Dictionary of Rhymes. Chinese-Mongol. Brit. Mus. Oriental 6972." has been applied to the spine of the case. On the inside back cover of the first fascicle are written the words "Bought of Mrs. Bushell Apr. 6, 1909".
The dimensions of the fascicles are 24.7 x 17.3 cm. The original manuscript comprises 66 folios, and is written on very thin, orangish-brown paper. The individual folios of the manuscript have been mounted on sturdy white paper, and it is these backing papers that are bound into fascicles, not the manuscript itself. The dimensions of the manuscript folios are approximately 22.5 x 28.8 cm. (height and width of a complete folio). Each folio is folded down the middle, with the central fold facing the open edge and the two ends facing the spine, as is the standard practise. The volume designation (shang 上 or xia 下) and page number is written on the lower left edge of each folio.
Fascicle 1 comprises an unnumbered folio with the title and the volume designation (shang 上), followed by 33 folios numbered shang 1-33.
Fascicle 2 comprises an unnumbered folio with the title and the volume designation (xia 下), followed by 31 folios numbered shang 1-31. Folio xia:30b and xia:31a are blank except for the page number (下卅一) on the lower left edge of xia:31a. These two blank pages presumably represent one or more missing folios in the exemplar (a printed edition ?) from which this manuscript copy was copied. Although the blank page gives a page number of "31", this must be conjecture on the part of the copyist, and given the scope of the missing text (all of the syllables for -a and -e, as well as the first half of the "Taboo Ideographs" section), it is probable that actually two or three folios of the original text are missing.
The text, including the Chinese prefaces, reads vertically from left to right according to the rules for Mongolian textual layout.
The contents of the text are :
The manuscript is written in a single hand using black ink, although there are a few minor corrections to incorrect Phags-pa words in red ink (e.g. the letter SHA in the Phags-pa word corresponding to 十 shí "ten" on the last column of xia:12a is badly formed, and has been amended to the correct shape with red ink).
There are two marginal notes in black ink written on the backing paper above the manuscript, clarifying the reading of the Phags-pa word below :
Source : Basibazi yu Yuandai Hanyu 八思巴字與元代漢語 Plate 33.
There are no indications of when and by whom the manuscript was copied (as is sometimes provided by a copyist's colophon). Nor are there any impressions of the seals of previous owners, as are often found on manuscripts.
Fortunately we can approximately date the manuscript from the evidence of taboo ideographs :
That the ideographs used to write the personal names of the Kangxi, Yongzheng and Qianlong emperors are tabooed, but the ideographs used to write the personal name of the Jiaqing emperor are not tabooed, means that the manuscript was almost certainly copied sometime during the Qianlong period (1736-1795).
The main text comprises lists of Chinese ideographs with the same pronunciation (excluding tonal differences, which are not represented in the Phags-pa script), together with their pronunciation as spelled out in Phags-pa letters. The lists are ordered by rhyme, according to fifteen rhyme categories :
|1.||東 dōng||*-uŋ||-ung, -eeung|
|2.||庚 gēng||*-əŋ||-ing, -hing, -yung, -eeing, -wung, -ying|
|3.||陽 yáng||*-aŋ||-ang, -yang, -wang, -hang, -ong, -weeng|
|4.||支 zhī||*-ɿ/ʅ, *-i||-i, -hi, -eei, -ue, -eeue, -yue, -wi|
|5.||魚 yú||*-u||-u, -eeu|
|6.||佳 jiā||*-ai||-ay, -way, -yay, -hiy, -iy|
|7.||真 zhēn||*-ən||-in, -un, -eeun, -hin, -yin, -win|
|8.||寒 hán||*-an, *-ɔn||-an, -on, -wan, -yan|
|9.||先 xiān||*-æn||-en, -een, -ween, -eeon, -yen|
|10.||蕭 xiāo||*-au||-aw, -ew, -eew, -waw, -yaw, -weew|
|11.||尤 yóu||*-əu||-iw, -uw, -hiw, -yiw, -ow|
|12.||覃 tán||*-am, *-æm||-am, -em, -eem, -yam, -yem|
|13.||侵 qīn||*-im||-im, -him, -yim|
|14.||歌 gē||*-ɔ||-o, -wo|
|15.||麻 má||*-a, *-æ||-ee, -wa, -ya, -wee, -we, [-a, -e]|
These rhyme categories do not exhibit any evidence of historical phonetic features, such as the "entering" tone (i.e. final p/t/k), and appear to correspond to the vernacular Northern Chinese (Old Mandarin) spoken during the Yuan dynasty.
Although these are slightly fewer than the nineteen rhyme categories given in the Yuan dynasty rhyming dictionary, Zhōngyuán Yīnyùn 中原音韻, compiled by Zhou Deqing 周德清 in 1324, that is because some of the Zhongyuan Yinyun rhyme categories are amalgamated in Menggu Ziyun.
Within each rhyme category, syllables are ordered by their initial sound, according to the table of thirty-six initials given at the head of the book :
|Initial||Phags-pa Letter||Chinese Gloss||Nominal Phonetic Value||Transcription||Notes|
|32.||ꡣ||匣||ɣ||x||Notes : "ꡯ is the same"|
|33.||ꡖ||影||∅||-||Notes : "ꡗ is the same"|
|34.||ꡝ||喻||j||'||Notes : "ꡭ is the same"|
Source : Basibazi yu Yuandai Hanyu 八思巴字與元代漢語 page 97.
Unlike the rhyme categories, which clearly do correspond to Yuan dynasty phonetics, the 36 initials listed in this table are actually an idealised set of initials devised over a period of time from the Tang to the Song dynasty by Chinese phoneticists, and so represent an earlier stage in the history of Chinese phonetic evolution.
Considerable phonetic changes occured between the Song and Yuan dynasties, and these 36 initial sounds do not accurately reflect Northern Chinese as spoken during the Yuan dynasty (Luo Changpei 羅常培 posits 20 initials, whilst Wang Li 王力 posits 25 initials). In particular some of the 36 initials no longer represented valid phonemes in Yuan dynasty Chinese :
That the rhyme classes reflect Yuan dynasty Chinese, whereas the 36 initials reflect Song dynasty Chinese, can be explained by reference to Zhu Zongwen's preface to his edition of Mengguu Ziyun (see above), which suggests that Zhu Zongwen revised the original text of Menggu Ziyun by reference to the Chinese Rhyme Dictionary Gujin Yunhui Juyao 古今韻會舉要 [Collected Rhymes Ancient and Modern], written during the early Yuan dynasty by Xiong Zhong 熊忠. As is typical of many phonetic works of the period, Xiong Zhong's rhyming dictionary preserves historical phonetic distinctions that were no longer valid in the Old Mandarin dialect that was the official spoken form of Chinese during the Yuan dynasty :
It is my belief that Zhu Zongwen rearranged the original entries of Menggu Ziyun according to the phonetic distinctions found in Xiong Zhong's work, namely :
Zhu Zongwen seems to have recognised that there was no longer a valid distinction between Initials 9-11 and Initials 26-28, as these two sets of initials are represented by the same Phags-pa letters (JA, CHA and CA), and in the main text no Phags-pa syllables are given between Initials 25 [z] and 29 [ɕ] for any rhyme category. However, Zhu Zongwen artificially maintained the distinction between certain merged initials by using graphic variants of the same Phags-pa letter to represent them :
The variant forms of the letter FA are problematic. As there are three initials corresponding to Yuan dynasty [f] (Initials 17-19 : [f], [f'] and [v]), we would expect there to be three variant forms of the letter FA. However, only two distinct forms of the letter FA are used : the normal form of the letter FA ꡤ without a tail kink; and a variant form ꡰ with a tail kink. In some words the lower triangular component of the letter FA is missing, so that the letter looks like the letter HA, but this is clearly due to scribal corruption in the only extant mansuscript copy of the text. The two forms of the letter FA (with and without a tail kink) are distributed between the seventeen syllables with Initials 17-19 with apparent randomness, as shown in Table 3 :
|Initial||ꡤ or ꡯ||ꡰ or ꡜ|
There are two possible explanations for this :
Without the discovery of a more reliable version of the text, such as the Yuan printed edition mentioned by a 19th century Chinese source, it will probably be impossible to be sure which of these two hypotheses is correct. I originally favoured the first hypothesis, but now think that the second one is more credible.
The two forms of the letter SHA correspond to Initials 29 [ɕ] and 30 [ʑ]. The normal form of the letter SHA ꡚ is used for Initial 30, whereas a variant form of the letter SHA ꡮ with a sloping stroke is used for Initial 29.
The two forms of the letter HA correspond to Initials 31 [x] and 32 [ɣ]. The normal form of the letter HA ꡜ (with a tail kink) is used for Initial 31, whereas a variant form of the letter HA ꡯ (with no tail kink) is used for Initial 32.
However the variant form of the letter HA is only used to represent syllables with Initial 32 before the semi-vowel [j] or front vowels excluding [i] (i.e. syllables ending ya, ee, wee, we, ye, eeu, yu, eei and yi). The letter XA is used to represent Initial 32 before back vowels or [i] (i.e. syllables ending a, wa, o, wo, u and i). Thus the letter XA and the variant letter HA are mutually exclusive.
The two forms of the letter YA correspond to Initials 33 [∅] and 34 [j]. The normal form of the letter YA ꡗ is used for Initial 33, whereas a variant form of the letter YA ꡭ with a trumpeted mouth is used for Initial 34.
Initials 33 and 34 are normally represented by the Phags-pa letters -A and 'A respectively, and the two forms of the letter YA are used in place of these two letters for some syllables. Presumably there is some rule for determining when to use the letter -A or when to use the letter YA for Initial 33, and when to use the letter 'A or when to use the variant form of the letter YA for Initial 34. However it is not immediately obvious what this rule is.
In summary, the differences between the traditional thirty-six initials and the usage of Phags-pa letters in Menggu Ziyun are :
This means that a total of thirty-five Phags-pa letters are used in Menggu Ziyun to represent the thirty-six traditional initials.
In the transcription used on this page, the various forms of the letters FA, SHA, HA and YA are differentiated by means of prime, double prime and triple prime marks after the letter.
As the forms of the letter FA used in Menggu Ziyun do not uniquely identify which historic initial they represent (e.g. the words 風 fēng "wind" and 方 fāng "square" both have an historic Initial 17 [f], but the former word is spelled with the form of the letter FA with a tail kink, whereas the latter word is spelled with the form of the letter FA without a tail kink), in the tables of Phags-pa syllables (Tables 4 through 10) the two forms of the Phags-pa letter FA have been artificially transcribed as f′, f″ and f‴, depending upon whether they represent Initials 17, 18 or 19 respectively. This means that thirty-six Phags-pa letters are designated by the transcription used on this page, even though there are only actually thirty-five distinct letter forms.
The correspondences between these thirty-six Phags-pa letters and the historic phonetic values that they represent are as follows :
It must be noted that the variant forms of the Phags-pa letters FA, SHA, HA and YA found in Zhu Zongwen's edition of Menggu Ziyun are not used in any other extant Yuan dynasty text to differentiate phonetic values of transliterated Chinese syllables (see my Description of the Phags-pa Script page for examples in support of this assertion), and would thus appear to be an innovation by Zhu Zongwen in order to artificially distinguish historical phonetic differences that were no longer found in standard Yuan dynasty Chinese. Possibly Zhu Zongwen's mother tongue was a dialect of Chinese in which these phonetic differences were still valid, which is why he wanted to distinguish them graphically in the written Phags-pa script.
Tables 4-10 detail all 813 Phags-pa syllables that are given in the extant manuscript (the total number of syllables should be slightly higher as the extant manuscript is missing a couple of folios). For each table the initial Phags-pa letters are listed down the left, and the finals are listed accross the top.
In the text, each Phags-pa syllable is followed by a list of Chinese ideographs which it represents, ordered by the four tones of Classical Chinese (Level, Rising, Falling, Entering). The number of ideographs for each syllable varies from one up to several dozen. In total the exatnt manuscript lists some 9,118 ideographs. In the tables below, only a single representative ideograph is shown for each syllable. In principle the selected ideograph is the first one in the list (hence normally in the Level tone), but where the ideograph is unusual (not in the basic CJK block of Unicode) or its Ancient Chinese reading anomalous, a different ideograph in the list has sometimes been chosen.
For each ideograph in the table, its reconstructed Ancient Chinese pronunciation is given in brackets. These are the phonetic reconstructions of the fǎnqiè 反切 pronunciations given in the Song dynasty Guǎng Yùn 廣韻 rhyming dictionary (completed in 1008, but based in part on the earlier Tangyun 唐韻, compiled by Sun Mian 孫愐 in 751, which itself was based on a Sui dynasty dictionary of pronunciation, Qieyun 切韻, compiled by Lu Fayan 陸法言 et al. in 601). The reconstructions are according to Hanzi Guyin Shouce 漢字古音手册. The tones have been omitted from the reconstructions given here, as tonal differences are not reflected in the Phags-pa spellings. Those Ancient Chinese pronunciations marked by an asterisk are not taken directly from Hanzi Guyin Shouce, but have been constructed from Guang Yun readings according to the principles followed in Hanzi Guyin Shouce, either because Hanzi Guyin Shouce does not include the particular ideograph, or because an alternative Guang Yun reading has been preferred.
The extant manuscript has the following shortcomings :
|'-||瞶 (= 聵)|
|z-||𧋍(⿰虫折 = 䱑)|
Phags-pa Home Page
Overview | Description | Script Styles
Forty-One Letters | Old Uighur Script | Baijiaxing Mengguwen
BabelStone Home Page