T of C 
Qin as
Qin in
/ Song
Analysis History Ideo-
Personal email me search me
SQMP  ToC   /   Scroll     "Short Version"     18 Blasts: 1597 version     North and Central Asia 首頁
47. Long Version of Nomad Reed Pipe 1
- Huangzhong mode:3 1 3 5 6 1 2 3
大胡笳 2
Da Hujia
  See Illustrations for 18 Songs of a Nomad Flute
Also known as Hujia Shibapai (18 Blasts of the Nomad Reed Pipe), this melody relates the story of the abduction ca. 195 CE of Cai Wenji and her 12 years in captivity. Besides here, the same story is also told in her biography as well as with two unrelated melodies having related titles, Xiao Hujia (Short Version of Nomad Reed Pipe; #15 in Shen Qi Mi Pu Folio One) and a sung rendition of the aforementioned alternate title Hujia Shibapai.4 All of these titles can be found in early handbook lists.5

As for Hujia Shibapai, the two most significant melodies using this title are:

  1. The extended qin song Hujia Shibapai, discussed under its first surviving version, Luqi Xinsheng (1597); it was reprinted in 1611, but apparently then disappeared until there was a modern reconstruction.6
  2. The instrumental Hujia Shibapai commonly played in the Qing dynasty; its earliest publication seems to have been in 1689.7 It is not clear whether it actually survived into the modern repertoire or whether the modern version should be considered a reconstruction.8

To look at this another way, in surviving Ming dynasty qin handbooks the most common titles of melodies on this theme are Xiao Hujia (four occurrences to 1585) and the present Da Hujia (six occurrences to 1634). In contrast, after an interlude that included publication of the extended qin song, during the Qing dynasty Da Hujia and Xiao Hujia disappeared and, from at least 1689 to 1910, the new Hujia Shibapai flourished, with versions in over 30 handbooks.9

The pages just mentioned and the analyses in QSCB, Chapters 5b1 and 6b1-2, also discuss the titles and content of various other relevant artistic works in addition to the qin melodies.

As for the Song dynasty or earlier, all that actually survives are Hujia lyrics. These include at least three Hu Jia poems in 18 sections that relate this story. Two of these, one attributed to Cai Wenji herself,10 the other by Liu Shang (late 8th c.),11 are included in the Yuefu Shiji. The third is by Wang Anshi (1021-86).12

The story of Cai Wenji has also been the subject of several modern operas. For example, at least one Western-style modern opera (with Chinese melodic connections) uses in part the lyrics attributed to Cai Wenji herself as well as portions of the melody from 1722.13

Of course, the story was also related in various traditional Chinese operas.14

As for Hujia Shibapai in the Qing dynasty (and modern?) repertoire, mention has already been made that its tablature can be traced back only to 1689, the Chenjiantang Qinpu. This version introduced a Hujia Shibapai that seems still to be related to Da Hujia but was so different that perhaps it should be considered a new melody.15 The most two identifiable musical characteristics of Da Hujia in Shen Qi Mi Pu are the opening phrase, also found in later versions, and the theme which begins Section 3 and then begins nine later sections.16 This later version keeps the opening phrase but not the repeated theme. Versions related to this one occur in over 25 further handbooks to the 20th century.

It should perhaps be noted that whatever the melody or the title or the lyrics, the music is often still credited to the great Tang dynasty qin player Dong Tinglan. Qinshi Chubian, Chapter 5b1, seems generally to agree in its argument that Da Hujia and Xiao Hujia can be considered Tang dynasty melodies. However, Chapter 6b1-2 argues that the Qing dynasty Hujia Shibapai dates from the Song dynasty. One of its points is that during the Song dynasty the nationalistc sentiments of this story were particularly popular. However, this should also have been true at the beginning of the Qing dynasty, when China again had foreign rulers.17

Cai Wenji's abduction took place around the time of the death of her father Cai Yong (133-192). As for her ransom 12 years later by Cao Cao, it took place while he was still a Grand General supposedly loyal to the Han dynasty. The title given him in the preface below, Wei emperor Wu Di, is one postumously bestowed on him by his son, Cao Pi, who proclaimed himself Wei emperor Wen Di when his father died in 220.

The story of Cai Wenji's abduction, in addition to its presence in all performing art genres, is also of note in fine art. The best introduction to this is in the book Eighteen Songs of a Nomad Flute, by Robert A Rorex and Wen Fong. It was published by Metropolitan Museum of Art, New York, following their acquisition of a 14th century scroll illustrating the cycle of 18 poems by Liu Shang. The fascinating illustrations bring the story very much to life. Each of the 18 section titles below is quoted from one of these 18 poems, in sequence, and so one can perform a fully-illustrated Da Hujia by combining the music with the scroll.18 In the last verse Wenji expresses her joy at being able to play the qin again.

Although most Chinese music has programmatic titles, in qin music the connection between the images and the melodies is rarely obvious. Da Hujia, however, has several very evocative passages, in particular one in Section 13 where Wenji must leave her children and return home.

Besides my own, there are recordings of the SQMP version by Guan Pinghu, Yao Bingyan, Chen Changlin and Gong Yi. There are also several recordings of Hujia Shibapai; these seem generally to be related to the 1722 version, which has a melody quite different from that of Da Hujia.

Original Preface19

The Emaciated Immortal, in accord with Qin History,20 says

the Official History of Han records:

Cai Yan, literary name Wenji, daughter of Cai Yong, had extensive learning and was a talented conversationalist; she was also very sensitive to music pitch. One night when her father was playing the qin a string broke. Hearing this she said, "That was the first string." Again hearing a string break she said, "That was the fourth string." This amazed her father. Later she was married to Wei Zhongdao of Hedong. When he died she returned home. During the great disorder at the end of the Han dynasty, Cai Yan was carried off by nomad cavalry, and in the frontier region she was a royal princess for 12 years. She bore two sons, and the chieftain greatly respected her. One spring while riding on a nomad cart she was emotionally moved by the reed pipes, and wrote a poem expressing her feelings. It said,

Nomad reed pipes are blowing;
Horses whinny on the frontier.
A solitary goose turns its head,
And its cry resounds, "Ying, ying."

Later Cao Cao, an old acquaintance of Cai Yong, ordered a great general to ransom Cai Wenji. She returned to Han, (but) her two sons remained with the nomads. Afterward, thinking longingly of Wenji, the nomads would roll up grass leaves for blowing as reed pipes, and play mournful sounds.

During the Tang dynasty Dong Tinglan, who was very good at the sounds of the Shen and Zhu schools, wrote down these sounds of the nomad reed pipe for the qin. This was his Long and Short versions of Nomad Reed Pipe.

18 sections, titled;
21 timings from my CD set

(00.00) 01. 紅顏隨虜 The pretty woman (must) follow the nomads (into captivity)
(00.54) 02. 萬里重陰 Darkened skies extend for 10,000 li
(01.18) 03. 空悲弱質 (At a desert encampment) helplessly resenting her weakness
(01.51) 04. 歸夢去來 Dreams of returning home come and go
(02.39) 05. 草坐水宿 Sitting on the grass and sleeping by the water
(03.35) 06. 正南看北斗 (So far north) it is to the south one looks to see the Big Dipper
(04.25) 07. 竟夕無雲 (Nomad music on) a cloudless night
(05.02) 08. 星河寥落 (As dawn approaches) the stars of the Milky Way thin out
(05.39) 09. 刺血寫書 Pricking blood (from her finger) to write a letter (home)
(07.03) 10. 怨胡天 Resenting the nomad skies (but finding love in the birth of a child)
(07.26) 11. 水凍草枯 Waters freeze over and the grass withers (marking the 12th year in captivity)
(08.10) 12. 遠使問姓名 A (Han) envoy from afar is asking after her name
(08.47) 13. 童稚牽衣 (As she prepares to leave) her (two) children pull at her clothing
(09.39) 14. 飄零隔生死 Drifting around separated (from her family, not knowing if they are) alive or dead
(10.01) 15. 心意相尤 (On the homeward journey) her heart and mind argue (at having to leave her husband)
(10.40) 16. 平沙四顧 The flat desert is everywhere one looks
(11.04) 17. 白雲起 White clouds rise (as they approach first Chinese garrison town)
(11.37) 18. 田園半蕪 The fields and gardens (of home) were half-neglected,
                      (but now she can play her qin again, expressing in music her sad story).
(12.16) --- 入無射泛 play harmonics of the wuyi mode
(12.04) --- 曲終 Piece ends

Return to the top

Footnotes (Shorthand references are explained on a separate page)

1. Translating 胡笳 Hu Jia (see under Xiao Hujia for 大 da and 小 xiao) 胡笳 Hujia   
"Hu Jia" (hujia) has been given a variety of translations, including "nomad flute", "nomad reed pipe", "barbarian reed pipe", and so forth. The following may explain some of the variety.

  1. "胡 Hu" (30073.000/12 "夷狄名 name of yidi", considered as northern tribal groups)
    This is a term that the Chinese applied to non-Han people or peoples originating north of the Han, ranging from the northeast to Central Asia. As discussed in the
    Wikipedia article "Wu Hu" (五胡), many peoples called hu were not nomadic, nor were they uncivilized. However, early Chinese literature generally seems to depict them in this way, hence the translations. Specifically (and perhaps in this case) the term often seems to apply to the 匈奴 Xiongnu (Wikipedia).
  2. "笳 Jia" (26511.0 "胡笳也,笛之一種 same as hujia, a type of flute"; image at right is from 30073.357)
    Although said to be a type of flute, the hujia is generally considered to be a reed type instrument, hence "reed pipe".

For issues in translating pai ("blasts", "songs", etc.), see another footnote.

2. Da Hujia references (see also QSCB analysis)
5960.738 大胡笳 da hujia says only "樂器名。參見十八拍、沈家聲條 name of a music instrument (sic), see 18 Sections and Shen Family Sounds."

See also Xiao Hujia, in particular the quotes under Xiao Hujia and Earliest Hujia titles

The poems on this theme attributed to Liu Shang and Cai Wenji, which are included in Yuefu Shiji, Folio 59, #3, are discussed further below. Qinshu Daquan (see QQJC, V, pp.261-268) also has these, plus one by Wang Anshi (1021-86; see below) and some additional material.

3. 黃鐘調;即無射 Wuyi (or Huangzhong) mode
For this tuning, slacken 1st, tighten 5th strings each a half step. For more details on this mode see Shenpin Wuyi Yi. For more on modes in general see Modality in Early Ming Qin Tablature.

4. 胡笳十八拍 Hujia Shibapai
For comment on the difficulty of translating this title, see above and under 1598. Note also that Hujia Shibapai is included in Zha Fuxi's index under Da Hujia (see the Appendix below).

5. History of Hujia titles
In the earliest known melody list, dated to the 7th century or earlier, Hujia seems to be only a mode name. Song dynasty lists include both Da Hujia and Xiao Hujia. As for "shibapai", I have not found much mention of this as a qin melody title prior to the Song dynasty. In fact, Seng Juyue includes a "Da Hujia Shibapai" only amongst its "less ancient" melodies.

On the other hand, Qinshu Cunmu suggests an early origin for Hujia Shibapai. Thus, although surviving melodies with this title do seem to be more recent than those with the titles Xiao Hujia and Da Hujia, the title Hujia Shibapai itself is not necessarily more recent. For more on this perhaps one should search in other media for occurrences of these titles.

6. Hujia Shibapai as a Qin Song
The earliest Hujia Shipapai set to lyrics, dated 1597, is introduced separately. The modern reconstruction must have achieved some popularity as when I presented my own reconstruction to some friends they recognized the melody from having heard it earlier.

(The Da Hujia set to lyrics in <1491 uses the same melody as in 1425; the lyrics seem to be new and properly paired, but are not naturally adaptable for singing.)

7. Qing dynasty Hujia Shibapai: created ca. 1689 or dating from Song dynasty?
The Hujia Shibapai commonly played in the Qing dynasty (see comment on its survival) was first published in Chengjiantang Qinpu (1689), but is best known from the very similar version in Wuzhizhai Qinpu (1722). In Guqin Quji, Vol. I (pp. 135-151) there is a transcription of the 1722 version according to Wu Jinglue, for whom there is also a published recording. (It is also transcribed, but with somewhat different note values, in The Qin Music Repertoire of the Wu Family). In QSCB, Chapter 6b1-2, Xu Jian (who in a previous chapter had discussed Da Hujia and Xiao Hujia as Tang dynasty melodies but does not seem to mention anywhere the 1597 Hujia Shibapai) analyzes the 1689 version according to the 1722 version, concluding that this version dated from the Song dynasty. He gives little musical argument to support his opinion. The Wu Jinglue recording and transcription show that this version has a few melodic similarities with Da Hujia. The fact that Da Hujia survives first from Shen Qi Mi Pu (1425), a handbook noted for collecting earlier tablature, then later survives in only a few handbooks and with little variation, supports arguments for its antiquity: if it was highly respected but not part of the active repertoire, people would simply re-copy existing tablature. The five versions of Hujia Shibapai published between 1689 and 1722 might similarly suggest this was transmission of an earlier version except that, unlike Shen Qi Mi Pu, the 1689 handbook is not noted for having copied very early tablature; this then might more likely suggest it was a newly created piece.

8. Survival of Hujia Shibapai into the modern repertoire
Although the version recorded by Wu Jinglue (see above) has its earliest publication later than other Hujia, it is not clear whether one can say it actually survived into the modern repertoire or whether Wu or someone else revived it. There are also recordings of Hujia Shibapai by Guan Pinghu and Cai Deyun, but both mention only 1722 as their source, not a living tradition. In this regard it should be noted that there have been several reconstructions published of the 1425 Da Hujia (in addition to my own and those of Guan Pinghu and Yao Pingyan), and it may well be that one is more likely to hear this version today than the Qing dynasty one.

9. Tracing Da Hujia (see Appendix and compare Xiao Hujai)
For Da Hujia see Zha Guide, 8/77/119. The early versions related to the 1425 Da Hujia, where they have sub-titles, are all connected to the story as told in Liu Shang's poem. They occur in five handbooks from 1425 to 1596, then once again in 1634. The 1525 version is called simply Hujia; the 1634 one is called Hujia Shibapai. The later versions of Hujia Shibapai are also included in the chart.

10. Hujia poem attributed to Cai Wenji (see original text)
There is an early poem on this theme attributed to Cai Wenji (蔡琰 Cai Yan) herself, called Song of Grief and Resentment (Bei Fen Shi; see its original text; it is translated in Paul Rouzer, Articulated Ladies, 2001). It is also briefly quoted in the SQMP preface, as follows,

「胡笳動兮邊馬鳴, Nomad reed pipes are blowing; horses whinny on the frontier.
孤雁歸兮聲嚶嚶。」 A solitary goose turns its head, its cry resounding "ying, ying."

As for the earliest known publication of the Hujia Shibapai poem attributed to her today, it cannot be found earlier than the Song dynasty, two early sources being the 12th century Yuefu Shiji, Folio 59, #3 (pp. 860-865), or in the Afterword to the Songs of Chu by Zhu Xi of the Southern Song dynasty (see QSCB, Chapter 6b1-2). YFSJ says its poem is the original one, later imitated by Liu Shang. However, the Liu Shang poem, also a first person narrative, is known to have had some popularity during the late Tang, so it could well be that the poem attributed to Cai Wenji herself was the imitation. See Idema and Grant, p. 121ff. It is translated there as well as in Chang and Saussy, pp.22-30.

11. Hujia poem by Liu Shang (see original text)
This poem by 劉商 Liu Shang is in Yuefu Shiji, Folio 59, #3. There is a translation in Robert A. Rorex and Wen Fong, Eighteen Songs of a Nomad Flute, A Fourteenth-Centry Handscroll in the Metropolitan Museum of Art, New York, Metropolitan Museum of Art, New York, 1974. The scroll is discussed above. There is a copy of the scroll linked from another page, which also has the complete original text.

12. Hujia poem by Wang Anshi (see original text)
王安石 Wang Anshi (1021-86) is discussed under Qinshi Chubian, Chapter 6. His Hujia poem apparently has not been translated.

13. Modern musical settings of the Hujia story
There have been several orchestral arrangements based on or inspired by one of the early Hujia melodies; some of them have included lyrics. Usually the original melodies are simply arranged for another instrument or for orchestra. Two such examples are:

  1. "蔡文姬 Cai Wenji", an orchestration by 史志有 Shi Zhiyou of the music in sections 1, 12 and 18 of Wang Di's transcription of the 1597 Hujia Shibapai; in October 2009 I found a partial recording online but it has seend been moved or removed (try searching for "史志有" "蔡文姬")
  2. Sections from the same melody played solo on a 葫蘆絲 hulusi (Wiki); available on YouTube.

In contrast to these is an opera by Bun-Ching Lam, which uses material (generally quite altered) from the Qing dynasty Hujia Shibapai.

18 Songs of a Nomad Flute: a new creation
This opera by Bun-Ching Lam (Wenji: Eighteen Songs of a Nomad Flute, 2002) has a libretto by Xu Wenying that uses selections of the poem attributed to
Cai Wenji herself. However, instead of drawing on qin music from 1597, Lam's opera adapts its from sections 1, 6-7, 13-14 and 16-18 of a transcription of a performance of the 1722 version of Hujia Shibapai played by 吳景略 Wu Jinglue as published in 古琴曲集 Guqin Quji, Vol.1 [Beijing, 1962], pp.135-151 (the original 1722 tablature is followed by the Cai Wenji lyrics, but they do not match the music by the traditional pairing method). Her rhythms are somewhat different from those in the performance by Wu on The Qin Repertoire of Wu Jing-lue, ROI RB-981014-2C, Hong Kong, 1998, so the transcription was presumably made from a different recording. The transcription, by 許健 Xu Jian, uses polyrhythmic meter to try to capture the nuances of Wu's free rhythms. Lam interprets these changing rhythms quite strictly. In her version some of these excerpts are played in a recognizable manner; others are quite altered, e.g., by playing them at about 1/4th the speed.

Lam's work is definitely new, drawing inspiration from qin materials.

14. Traditional Chinese operas on the Hujia theme
These include Wenji Goes to the Desert (文姬入塞 Wenji Ru Sai) and Wenji Returns to Han (文姬歸漢 Wenji Gui Han); both are discussed in LXS.

15. Origin of the Qing dynasty version
Xu Jian argues (see above) that this 1689 version actually originated in the Song dynasty; to the extent that it is sufficiently different from Da Hujia it is still a "new" melody, even if it dates from the Song dynasty.

16. Described as 從卷至盧 "From Roll to Reed"; see sections 3, 4, 5, 6, 7, 9, 10, 11, 12 and 15.

17. Dating the Hujia melodies
For Xu Jian's commentary on Hujia in Qinshi Chubian, see Chapter 6a2 as well as Chapters 5b1 and 6b1-2. The biography of 董庭蘭 Dong Tinglan (not yet translated) also has relevant information.

18. Hujia Scroll
I have shown this scroll at small gatherings, where someone can unroll the scroll as I play, and also as part of a stage performance where the images are projected using PowerPoint as I played the melody. I have not yet been had the opportunity to do this while someone recites Liu Shang's poem.

Some societies have a tradition of illustrated narrative storytelling. For example, in Rajasthan itinerant storytellers might as they go along point at a backdrop illustrating their story. Although a narrative series such as that in the Hujia scroll might seem ideally adapted for this sort of event, I am not aware of such a tradition in China.

19. For the original Chinese text see 大胡笳.

20. 琴史 Qin Shi
It is not clear if this is a book name or just the history of qin. Zhu Quan's sources are problematic. It is similar to but not directly from the version in Zhu Changwen's Qin History.

21. Section titles
Each section title is taken from the respective poem in the set of 18 poems by Liu Shang. The words added in brackets here are meant to help connect the section titles to the themes of Liu Shang's poems.

Return to the top

Chart Tracing Da Hujia
Xiao Hujia)
Comment; based mainly on Zha Fuxi's Guide, 8/77/119.

    (year; QQJC Vol/page)
Further information
(QQJC = 琴曲集成 Qinqu Jicheng; QF = 琴府 Qin Fu)
  1.  神奇秘譜
      (1425; I/149)
18T; Da Hujia (Folio 3; SQMP puts Xiao Hujia in Folio 1: "Most Ancient"); "Dong Tinglan"
Related versions below are #2,3,4,5,& 9
  2.  浙音釋字琴譜
      (<1491; I/255)
18TL; Da Hujia; = 1425 but adds lyrics (also has XHJ)
 "Dong Tinglan"
  3. 西麓堂琴統
      (1525; III/212)
18T; Hu Jia; differences, but basically follows 1425 (no Xiao Hujia);
"Dong Tinglan"
  4. 風宣玄品
      (1539; II/360)
18T; Da Hujia; same as 1425 (also has XHJ)
No preface
   . 琴書大全
      (1590; V/263-9)
No music, but extensive commentary and three poems,
18 sections each, by Wenji herself, Liu Shang, and Wang Anshi
  5. 文會堂琴譜
      (1596; VI/270)
18T; HJ18P, but basically same as above
No commentary  
  6. 綠綺新聲
      (1597; VII/33 [details])
HJ18P; Cai Wenji lyrics; fugu mode (but 7 strings)
Music completely different (attribution?); see comments above and elsewhere.
  7. 琴適
      (1611; VIII/46)
HJ18P (1st page missing); music and lyrics same as 1597;
Lyrics also placed at front of each section
  8. 理性元雅
      (1618; VIII/331)
HJ18P; lyrics; 9 string qin; no attribution
lyrics same as 1597, but melody again different
  9. 古音正宗
      (1634; IX/378)
Da Hujia; same as 1425
no commentary
10. 澄鑒堂琴譜
      (1689; XIV/342)
Hujia; hints at 1425 on 1st line, then seems almost completely different; no commentary;
This version became the standard one for the next two centuries
11. 琴譜析微
      (1692; XIII/132)
HJ18P; almost same as 1689
No attribution  
12. 響山堂琴譜
      (<1700?; XIV/122)
Hujia; 18 sections related to 1689
No commentary  
13. 蓼懷堂琴譜
      (1702; XIII/295)
18 Pai; almost same as 1689
No attribution  
14. 五知齋琴譜
      (1722; XIV/558)
HJ18P; the version most commonly played today, but almost same as 1689; attrib. Cai Yan
The complete Cai Yan lyrics (see source) are copied after the tablature 
15. 臥雲樓琴譜
      (1722; XV/125)
16. 存古堂琴譜
      (1726; XV/296)
17. 琴書千古
      (1738; XV/447)
18. 春草堂琴譜
      (1744; XVIII/280)
19. 蘭田館琴譜
      (1755; XVI/286)
HJ18P; compare 1689
(Sections sometimes divided differently)
20. 琴香堂琴譜
      (1760; XVII/177)
HJ18P; like 1689
21. 研露樓琴譜
      (1766; XVI/513)
HJ18P; 18T
22. 自遠堂琴譜
      (1802; XVII/480)
Hujia; like 1689
23. 裛露軒琴譜
      (>1802; XIX/110)
Hujia; Preface; short comment with each section
24. 響雪山房琴譜
      (>1802; XIX/418)
HJ18P; short comment with each section
25. 琴譜諧聲
      (1820; XX/172)
HJ18P; some sections have comments;
at end are comments about the tuning and melodies that use it
26. 琴學軔端
      (1828; XX/481)
XX/477 has Cai Wenji lyrics
27. 二香琴譜
      (1833; XXIII/174)
28. 悟雪山房琴譜
      (1836; XXII/402)
29. 琴譜正律
      (~1839; XXIII/62)
30. 稚雲琴譜
      (1849; XXIII/409)
18 Pai
31. 琴學尊聞
      (1864; XXIV/257)
HJ18P; "same as 1738"
31. 蕉庵琴譜
      (1868; XXVI/89)
same commentary as 1722 
32. 天聞閣琴譜
      (1876; XXV/609)
HJ18P; 16 Sections?
"= 1738"
33. 天籟閣琴譜
      (1876; XXI/214)
34. 響雪齋琴譜
      (1876; ???)
Not in QQJC
35. 希韶閣琴譜
      (1878; XXVI/383)
HJ18P; preface and afterword
Verse of Cai Wenji's poem placed after each section
36. 枯木禪琴譜
      (1893; XXVIII/113)
Hujia; preface
Cai Wenji's poem at end
37. 琴學初津
      (1894; XXVIII/390)
HJ18P; "仲呂變調宮音";
Cai Wenji poem at end
38. 琴學叢書
      (1910; XXX/331)
HJ18P; from 1722
Also in 琴府
39. 虞山吳氏琴譜
Da Hujia; from 1425

Return to the top, to the Shen Qi Mi Pu ToC, or to the Guqin ToC.