Libie Nan 離別難
 T of C 
Qin as
Qin in
/ Song
Analysis History Ideo-
Personal email me search me
Japan Theme   Japan pu   Qin songs   Qin Ci   Qin Ci in Japan 聽 listen to my recording   網站目錄
Separation Sorrows
Subtitle: Seeing Jiang Yulu off to Jianxi2
"羽音 Yu yin"3
離別難 1
Libie Nan  
  Opening of the two Libie Nan from 1676 (complete pdf)4    
"Libie Nan" is the name of a cipai (ci lyrics pattern). The dictionary definition of "libie" says it refers to the act of separating. Likewise the subtitle "Seeing Jiang Yulu off to Jianxi" seems to indicate that the ci lyrics here should refer to the act of separation. This being the case, "Parting is Such Sorrow" might seem to be a more appropriate translation. In fact, though, many of the lyrics in this form seem to emphasize more the state of being apart. Thus, other possible translations in addition to the one used here include "Anguish of Separation", "Separation Hardships" and so forth.

In fact, the earliest known poems using this title are not in a ci form at all, and their titles might also better be rendered as either "Hardships of Separation" or "Anguish of Separation". The only story connected to those earlier poems tells of (or is told by) a woman who is already separated from her husband. There is no surviving old music for those earlier lyrics; and it should also be noted that those lyrics are poems of fixed line length, thus having no structural relationship to the later ci poems of this title. The relationship between the later ci poems and the earlier poems of the same title is thus uncertain.6

The two closely-related versions of the melody in the present handbook are the only known existing early musical settings of lyrics in this ci pattern. Both survive only in the Japanese handbooks attributed to Jiang Xingchou (1639-1695).7 In 1676 Jiang went to Japan where, as the monk Toko Etsu, he began the modern tradition of guqin playing in Japan. Because he brought with him qin tablature as well as qins, the Japanese handbook containing this music, as published in Qinqu Jicheng, is generally dated "1676". However, the source of the actual music in his handbooks is not always clear: the handbooks usually credit the lyricists but rarely state the specific sources of the music. It is generally assumed that Jiang brought the music from China, but he may in fact have created some of it himself either before or after he arrived in Japan.

These two settings of Libie Nan may be instructive about this. Differences between the music of the two versions are small.8 Both use the same ci lyrics by Zou Zhimo;9 and both Zou and the subject of his lyrics, Jiang Rui (1625-98),10 seem to have been younger contemporaries of Jiang Xingchou. Thus one might assume that the lyrics and the music were created around the same time as each other.

On the other hand, differences right at the beginning of these two settings may point to an earlier source for the music of the first version. Specifically, in the opening line of each version, as can be seen in the image at upper right, the lyrics of both are the same, but with the tablature paired to the lyrics you can see (if you understand qin tablature) that the second version (on the left half of the image) adds one extra punctuation mark, and three of the eight tablature clusters (the fifth to the seventh) are different from those of the earlier version (right half of the image). The result is that the tablature on the right fits better some earlier lyrics of the same title, specifically those by Liu Yong given below, while those of the second tablature exactly fit those by Zou Zhimo. This suggests the possibility that the tablature for the first version was taken from a pre-existing source that had applied them to lyrics that follow the standard pattern, perhaps even those by Liu himself, then later this problem was noticed and the tablature/melody was changed to fit the new lyrics.11

Further regarding the ci pattern called "Libie Nan", historically there were two distinctly different patterns. One has 87 characters;12 the other has 112, as here.13 Both forms are called "shuangdiao", meaning that they are both divided into two stanzas. The former is divided into 43+44 characters, the latter is 55+57.

The "representative piece" for the 112 character version of this title is said to be the above-mentioned ci poem by Liu Yong (987 - 1053).14 The two stanzas of his poem have the same character count (55+57) and almost the same phrasing as those of the piece in the present handbook, so the standard syllabic setting used for qin songs15 does make it possible to pair all lyrics in this form to the same melody. However, the phrasing difference mentioned above and below causes a problem. Perhaps further analysis of this will lead to further insights into the process of pairing qin tablature to ci lyrics, including also how the musical phrasing (tentative comment) may reflect the pattern's overall form (shuangdiao), pingze pattern and use of "yun" (rhyme). But without a transcription of the actual melody from 1676, or a full understanding of the requirements of the ci, any interpretation remains tentative. Hopefully further insights will emerge during the process of reconstruction (dapu).

Lyrics by Liu Yong
聽我的錄音 listen; 看五線譜 see transcription)
Transcription is tentative; recording is without voice 16

For this preliminary reconstruction of the tablature for the first version of the 1676 Libie Nan I substituted Liu Yong's lyrics for those by Zou Zhimo: as mentioned above, Liu's lyrics seem to fit the first version better. The content seems to be a recollection of a departed (deceased?) female friend (quite likely a courtesan). My preliminary translation is as follows (a more accurate translation is urgently needed).17

(The first 7 seconds of the recording have the closing harmonics)

Huā xiè shuǐ liú shū hū, jiē nián shào guāng yīn.
Flowers fade and water flows by in the blink of an eye, I sigh that the years have so few days and nights.

    Yǒu tiān rán, huì zhí lán xīn.
    By nature, she had a kind heart,

    Měi sháo róng, hé chì zhí qiān jīn.
    a beautiful appearance, worth countless treasures.

Biàn yīn shén, cuì ruò hóng shuāi, chán mián xiāng tǐ, dōu bù shèng rèn.
But for some reason, her complexion was weak, attention unfocused, unable to take responsibility.

    Suàn shén xiān, wǔ sè líng dān wú yàn, zhōng lù wěi píng zān.
    As for immortality, all the elixers had no effect, and on the road we cast aside bottle and hairpin (permanently separated).

Rén qiāo qiāo, yè chén chén.
Humankind is sadly silent, the night is dark and deep.

    Bì xiāng guī, yǒng qì yuān qīn.
    Closed the boudoir, forever abandoned the mandarin duck quilt.

    Xiǎng jiāo hún mèi pò fēi yuǎn, zòng hóng dū fāng shì yě nán xún.
    Thinking of the lovable spirit and charming soul not far away,
        sought by the
Hongdu alchemist but still hard to find.

Zuì kǔ shì, hǎo jǐng liáng tiān, zūn qián gē xiào, kōng xiǎng yí yīn.
Most bitter is this, good scenery nice day, in front of wine bottles songs and laughter, in vain thinking of the sounds she bequeathed.

    Wàng duàn chù, yǎo yǎo wū fēng shí'èr, qiān gǔ mù yún shēn.
    Look toward the break off point, dim and distant Sorcerer Mountain's Twelve (Peaks), a thousand old twilight clouds deep.

(See further comment on the requirements for this translation..)

Adapting the Liu Yong lyrics to the 1676 tablature
There being no tablature surviving specifically for Liu Yong's version, an attempt has been made here to see whether his lyrics can be sung to the 1676 tablature. Both 1676 settings are mostly syllabic, following the lyrics by Zou Zhimo. The punctuation of both is identical but omits some of the punctuation included above with Liu Yong's lyrics. Beyond this it may be significant to make the following points:

  1. The two 1676 versions use different parsing for the first line (11 characters); other than this the tablatures of both 1676 settings are nearly identical to each other. But since the parsing of the first line in Zou's poem is 4+3+4 instead of 6+5 (as in Liu's poem), and since in 1676 the tablature of the first version actually fits the 6+5 phrasing better than it does 4+3+4, while that of the second version is clearly designed for 4+3+4, perhaps this suggests that the tablature of the first 1676 version was taken from an earlier setting, but then someone realized the problem and did a re-setting to fit Zou's poem.

  2. Both the Zou and the Liu versions of the 112-character Libie Nan divide it into two verses, divided as 55+57 characters. Beyond that, based on the yun (rhyme), each verse seems to have five lines. The last two lines of one being virtually identical to those of the other, but the first three lines of each are very different. According to my preliminary transcription both sets of three lines of lyrics seem most naturally open to interpretion as two musical lines. If this is the case, then each of the two verses has four musical lines. In the first verse the first two lines of music divide the lyrics as 11+15, with the second musical line including two of the rhymes; in the second verse they divide them as 13+15, with the first musical line including two of the rhymes.

The ramifications of these two observations, if correct, are not clear.

None. Under the title it says,

Seeing Jiang Yulu off to Jian (送蔣馭鹿遊建 Sòng Jiǎng Yùlù Yóu Jiàn)
(Lyrics by) 鄒訏士 Zou Xushi.

Music18 and lyrics19 (XII/205)
Specifically 離別難,又譜 Libie Nan #2 (XII/207)
As can be seen in the image
above, this second setting omits the Japanese indications of pronunciation. Otherwise, as explained above, although the title, subtitle and lyrics of both versions are the same, the first setting seems to work better with the lyrics by Liu Yong.

For this reason I have done a tentative transcription of Libie Nan #1 substituting Liu's lyrics. For proper comparison a transcription of #2 using Zou's lyrics will also be necessary.

Lyrics by 鄒祇謨 Zou Zhimo:

Yuè shuǐ Wú shān, dài piān zhōu, jiǔ yǐ qī chí.
    Zhǐ zhàngfū, nán běi dōng xī.
    Zhuó zhēng shān, hé bì zèng jiāng lí.
Yīng zhǐ yǒu, qí zì yíng náng, hǔ tiào lóng wò, luò zhǐ zhēng fēi.
    Guò Yántān, wèi fǎng Fùchūn jiù jī, chóng shàng Diàoyújī.

Yún rǎn rǎn, cǎo qī qī.
    Yì lín Qí, qiū shì xiū bēi.
    Tīng Wǔyí Màntíng xiān yuè, biàn cān luán Méi Fú yě xiāng qī.
Hé kuàng shì, dào chù féng yíng, tiān yá zhī jǐ, zhuó kuài xián bēi.
    Huái gù yuàn, chàng hǎo lì zhī wèi shú, quàn rǔ mò wàng guī.

Footnotes (Shorthand references are explained on a
separate page)

1. References for "Parting is Such Sorrow" (離別難 Libie Nan) (XII/205)
Also written as Li Bie Nan and translated as "Hard to Say Farewell"; other possible translations include "Sadness of Separation", "Separation Hardships" and, even more literally, "Separation is Difficult".

ZWDCD 43079.41 離別 libie says "分手也,近曰離,遠曰別": "Separate; if near then it is called 'li', if far then it is called 'bie';" "分手 fen shou" itself seems to emphasize the act of separation rather than the separation itself.

43079.43 離別難 says Libie Nan is 詞牌名 the name of a cipai, but it says nothing of the cipai structure of this title, such as the fact that it is a shuangdiao (i.e., has two verses). In addition, the only reference it gives is to structurally unrelated Tang dynasty Yuefu poems of this title.

First it quotes the introduction to this title as found in YFSJ 80/1131, which says (the understanding given here in English comes largely from Biographical Dictionary of Chinese Women, Volume II: Tang Through Ming; Palace Women, pp.307-8):

《樂府雜錄》曰:《離別難》,武后朝有一士人陷冤獄,籍其家,妻配入掖庭。善吹觱篥,乃撰此曲以寄情焉。 初名《大郎神》,蓋取良人第行也。既畏入知,遂三易其名曰《悲切子》,終號《怨回鶻》云。
Yuefu Zalu says: (as for) Libie Nan, in the palace of Empress (Wu Zetian) there was a gentleman who had been unjustly imprisoned and his family registered ("confiscated, including his wife"?). His wife thus became a servant in the imperial household. She was good at playing on a hornpipe (觱篥 bili: soft like an oboe? or raucus like a shawm?), and so she created this melody to convery her deep emotions. Its original name was "Spirit of Dalang (大郎神 Dalang Shen, 'Dalang' presumably being the husband's name)", drawing on the story of a good man 第行 (?). Fearing people would realize this she changed the name three times, (the other names being) "Mournful Ditty" (悲切子 Beiqie Zi)" the finally "Mourning with a Uighur [Pipe] (愁迴鶻 Chou Huihu)" ("怨迴鶻 Yuan Huihu"?), and so on.

It then has two poems with this title.

  1. Author not named but by implication it is the wife featured in the preceding story. (It is also in Complete Tang Poems, where she is referred to as 武后宮人 Palace Woman of Empress Wu.)
    (Translation copied from that by Lily Xiaohong Lee in the Biographical Dictionary of Chinese Women),

    此別難重陳,   It is hard to speak again of parting,
      花深復變人。   Flower petals fly away but they come back to haunt people.
    來時梅覆雪,   When you came, plum blossoms were covered with snow,
      去日柳含春。   When we parted, the willows hinted at spring.
    物候催行客,   The time came, you had to go.
      歸途淑氣新。   On your way home, the air would have been fresh.
    剡川今已遠。   Shanchuan is far, far away now,
      魂夢暗相親。   Secretly, I will follow you in my dreams.

  2. By 白居易 Bai Juyi (tentative translation):

    綠楊陌上送行人,   Green willows along the lane where I see off a traveler,
    馬去車回一望塵。   Horses go carts return and I immediately see dust.
    不覺別時紅淚盡,   Not conscious of a separation time when red tears were exhausted,
    歸來無淚可沾巾。   Upon return there are no tears that can wet my kerchief.

There is further comment below on making connections between old poems and later ci with the same title.

2. Subtitle
See preface.

3. Yu mode (羽音 yu yin)
Mode in the short songs of the handbooks preserved in Japan seems to be rather different from that in earlier Ming handbooks (see, e.g., criteria discussed under Shenpin Yu Yi as well as Modality in early Ming qin tablature.

The tonal center for most of a melody in yu mode should be la (6), but not yet studied here.

4. Image: Opening of the two Libie Nan from 1676 (complete pdf)
These two settings, discussed further in the text below, are from Hewen Zhuyin Qinpu, published in Japan before 1676, but probably consisting of music brought there from China. The symbols by the Chinese characters in the first version tell Japanese how to pronounce the characters in Chinese. The second version omitted this, presumably because the lyrics are the same as in the first version.

There is no indication of why there are two quite similar versions. Did they develop independently, perhaps two different transcriptions of the same or different performances? There are some clear mistakes in the first version, and it would be very interesting to explore the possibility that the second version (which has its own problems) was an attempt to "correct" the first one. This is of particular interest in view of the discrepancy mentioned here and here in the pattern of the opening 11 characters.

6. Ci titles and their relationship or older forms of the same title
Ci poems were named after the titles of old song melodies that are now lost. In some cases there are existing old lyrics in that pattern, with the possibility that these were written by someone who knew (or even created) the original melody. Later, however, it is clear that only the poetic structure remained, not the original melody or perhaps even verses written by people who knew that melody. Furthermore, often the themes of poems following those ci structures have no connection to the original meanings of those titles.

Regarding the two Libie Nan structures discussed on this page, there are in fact very few surviving poems in either form. The ZWDCD entries (and presumably other sources) do not make it clear how they connect the earlier story and its associated lyrics to the later cipai of this name, other than through the general theme of separation. Such relationships seems to be a general issue for which there are few studies, at least in English.

7. Tracing 離別難 Libie Nan
Having two settings of the same song in one handbook is very rare (here is another), and this is the only such example from Japan. Zha Guide 35/--/507 lists Libie Nan only in the first handbook below, but it is also in several other handbooks copied in Japan. The ones I have found are:

  1. Hewen Zhuyin Qinpu (XII/168
    As discussed above, there are here two nearly identical versions
  2. Version 1 only; copied in this 1898 Toko Kinpu (XII/273)
  3. Version 1 only; original (?) Toko Kinpu (this handbook not in QQJC)

If there are any others in this pattern but with another title I have not yet found them. The fact that the other editions have only version 1 requires a close examination of the comments here. If that analysis is correct, could it be evidence that Hewen Zhuyin Qinpu actually includes at least some music that is later than that of these other handbooks?

8. Differences in the tablature/music
The close relationship between the two 1676 versions is emphasized by the fact that they are lined up identically in the book, as can be seen in this complete pdf. Differences in the music for the opening 11 notes are discussed here. Elsewhere differences in the second 1676 version may occur where a tablature sign in the first version is not clear. These are almost exclusively in the first section, most significant differences seeming to be in the passages accompanying the lyrics "著征衫、何必" and "應只有、奇字".

9. 鄒祇謨 Zou Zhimo
Here referred to by his style name, 鄒訏士 Zou Xushi (not 鄒訐士). For Zou see also #24 and #26.

10. Jiang Rui (1625-98)
蔣銳,字玉淵,號馭鹿,江蘇武進人 [original has 庶 for 鹿]; (1625-98). No family relationship with Jiang Xingchou.
9786.104 建溪 Jianxi is in Fujian.

11. Differences in the ci pattern
In the version on the right side of the image the slide up on 扁 bian suggests that this is the last note in the phrase (and that the punctuation should/could be there instead of after 山); by contrast the punctuation in the left side of the image puts it after 舟 zhou. This divides the first eleven characters on the right as 6+5 but those on the left as 4+3+4. This latter punctuation fits better the present lyrics while the phrasing on the right can more easily be made to fit the Liu Yong lyrics.

There is also some additional variety in the phrasing of the lyrics by Zou Zhimo given below but this is mostly in subdivisions - those marked there with , . In addition few of the characters in the tablature are different from those in the commonly published versions).

12. Standard 87 character 離別難 平仄 Libie Nan pingze pattern
The standard pattern for the 87 character form, not used here, is a shuangdiao divided into 43 + 44. The 平仄 pingze pattern is as follows:

仄仄仄仄平平   平平仄仄平平
      仄平平仄仄   仄平平仄仄
      仄平平仄仄   仄平平
      平仄仄   平平仄   平平平仄仄平平

平仄仄   平平仄   平仄平   平平仄仄平平
      仄仄平平仄   仄仄平平仄
      仄平仄   仄平平   平仄仄   平平仄

Some websites, such as, give this poem by 薛昭蘊 Xue Zhaoyun as their representative example:

寶馬曉鞴雕鞍   羅帷乍別情難
      那堪春景媚   送君千萬里
      半妝珠翠落   露華寒
      紅蠟燭   青絲曲   偏能勾引淚闌干  

良夜促   香塵綠   魂欲迷   檀眉半斂愁低
      未別心先咽   欲語情難說
      出芳草   路東西   搖袖立   春風急

"Xue Zhaoyun 薛昭蘊 (fl. tenth century) once served in the Former Shu court as Attendant Gentleman (Shilang 侍郎). He has nineteen poems preserved in the Shu anthology." (See Yang Liu, Imagery of Female Daoists in Tang and Song Poetry, 2011 (online, p. 192).

13. Standard 112 character 離別難 平仄 Libie Nan pingze pattern
The standard pingze pattern for the 112 character form followed here (modeled after
Liu Yong's poem?) is as follows:

平仄仄平仄仄句 平平仄平平韻
    仄平平讀 仄仄平平韻
    仄平平讀 平仄仄平平韻
仄平仄讀 仄仄平平句 平平平仄句 平仄平平韻
    仄平平讀 仄仄平平平仄句 平仄仄平平韻

平仄仄句 仄平平韻
    仄平平讀 仄仄平平韻
    仄平平仄仄平仄句 仄平平平仄仄平平韻
仄仄仄讀 仄仄平平句 平平平仄句 平仄平平韻
    仄仄仄讀 仄仄平平仄仄句 平仄仄平平韻

In the actual lyrics "句 ju" is represented as ",", "讀 du" as "、", and "韻 yun" as "。".

The difference between Liu Yong's poem and the one by Zou Zhimo in the phrasing of line 1 is mentioned above. In addition, in the last line of each verse in Zou's poem it seems that the first "、" has been changed to a ",".

There is no information to suggest how or whether this pattern affects the music here.

14. 離別難 Libie Nan by 柳永 Liu Yong: the "代表作品 representative piece"
Liu Yong (987 - 1053) was from Fujian but he made a name for himself as a poet who habituated and wrote for the entertainment quarters of the Northern Song capital, Kaifeng, in today's Henan province.

Clearly this poem fits the 112 character Libie Nan structure, but beyond this I do not know the significance of the expression "representative piece". For example, are all pieces in the form representative, or is this one considered the standard by which others are to be judged? How long has this expression/concept been in use?

15. Syllabic setting
Some early Chinese sources seem to suggest that songs should have one note per syllable. However, in qin melodies with lyrics the common pairing practice was to have one character for each right hand stroke but no characters for left hand slides or, in some cases (such as here), left hand plucks.

16. Tentative transcription using Liu Yong's lyrics
Qin tablature usually tells all the notes quite clearly but gives only indirect indication of note values/rhythms. Therefore reconstructing ancient melodies begins with looking for structures, especially ones that suggest rhythms. Libie Nan is not untypical in that I have found that it can be interpreted in quite a structured manner: as can be seen from the transcription, the time signature is 2/2 throughout, with the piece divided into two sections of 33 measures each; the measures of both sections are arranged into four musical lines as follows: (4+4)+(4+4)+(4+4)+(2+3+4). In the last line of each section the extra measure is also played more slowly, emphasizing its finality. When listening without having seen the transcription this regular structure may be somewhat disguised by the fact that in places the rhythms are quite freely interpreted.

For the next stage of reconstruction I will need to examine the lyrics more carefully to see if they actually sound natural within such a structure; this may of course depend on who is singing. But even if the end result is or seems much more irregular, whatever connection it retains to this regular structure will make it seem more coherent and I believe (at least from a performer's standpoint) memorable.

17. Translation of Liu Yong's Libie Nan
Translating lyrics such as this is a special challenge. Ideally there should be a version that is brief enough that it can be followed by listening to the music, but this needs to be accompanied by copious notes and perhaps more detailed translation. See my Heat Sutra translation under Se Kong Jue for an example of this.

18. Music
Source unknown

19. Lyrics
There are some differences here with versions of these lyrics to be found published elsewhere. Specifically (line by line):

  1. "矣" was like "兵"; "征" was "徵"; "必" was "事"
  2. "盈" was "盛"
  3. "期" was "依"
  4. "" was ""

See also the comparison with the "representative version" by Liu Yong. My initial reading suggests that the first version from Jiang's handbook works better with Liu Yong's lyrics. If so, perhaps this might reveal something about how ci patterns were treated when applied to new music.

Return to the top, to Qin Poetry and Song or to the Guqin ToC.