publisher colophon


Han Songs and Ballads as Social Documents

The writings of a great civilization may be subsumed under the Great Tradition and the little tradition. To the Great Tradition belong the works of the classical philosophers, historians, poets, scholars, and scientists who form an educated élite of inspired talent. Thanks to the translations of Arthur Waley in England, Burton Watson in the United States, and other sinologists, their writings and those of anonymous classical authors have become familiar to Western readers.1 Inherent in the word ‘classical’ is the idea of serious literature from the ancient past written by members of the aristocracy, of court circles, and families whose social status and political position granted them an education.2 The Great Tradition in China instructed and entertained the generations of men whose destiny was to shape and govern the empire. Significant though its humanistic influence has been, it represents only the most visible stratum of civilization.

The less visible layers are perceived through the little traditions of art, crafts, dance, music, and song, which tell of the lives, thoughts, feelings, and aspirations of the nameless mass of people in rural and urban regions. They are the eloquent, enduring creations of anonymous crafts­men and makers: the magnificant tombs of ancient kings; vessels and ornaments of bronze and jade; paintings; sculpture; woven silk; and lacquerware. These material expressions of creative, but mostly unknown, people form much of the archaeological heritage of ancient China which has been revealed to us in the past few decades.3 Less tangible is the oral tradition of song, ‘a rich and disorderly pattern’, which has been fixed in texts and transmitted through the centuries.4 The folk-songs and ballads in this book belong to the oral tradition of the Han period, when the first great dynasty of China was founded and the empire was formed. They express the hopes and dreams of ordinary people, their routine lives, the tragedies which beset them, their brief moments of happiness, the values and beliefs they cling to. Such utterances from foot-soldiers, farm-hands, servant girls, entertainers, poor wives, orphans, and destitute husbands, many of whom were 2considered to be outside the traditional social classes, counterbalance the picture we have of Han China derived from the Great Tradition.

These songs and ballads are the legacy of the popular, oral tradition, works of art which have endured because of their simple, honest, and unfailingly human qualities. More importantly, perhaps, they constitute informal social documents which help to illuminate the darker, less familiar existence of the people.

Han Culture

In 202 BC Liu Pang, a commoner, founded the Han Dynasty. The Liu family succession was maintained until AD 220, except for a short inter­regnum between AD 9–23. During these four centuries, the imperial foundations of China were established and consolidated. Han China inherited a revolutionary sociopolitical structure from the short-lived Ch’in Dynasty which immediately preceded it.5 The Ch’in, informed by a Legalist and Governmental philosophy, conceived of a new social order in which the people were divided into functional classes and were accountable directly to the State. The Ch’in rulers dismantled the old aristocratic orders, imposing a new state-centred system. The economy came under central control; weights, measures, vehicle axle gauges, and coinage were standardized. The writing system was also made uniform. The people were mobilized for huge projects, such as the construction of the capital city and the building of the Great Wall, and were conscripted into the army. The early Han emperors endorsed the Ch’in institutional innovations, even extending some aspects of state management of the economy, such as state monopolies on salt, iron, liquor, and the marketing of grain.6

The Ch’in government had set in motion measures toward political and social cohesion and economic efficiency within its immediate boundaries. Yet because the dynasty was so short-lived (just 16 years), the practical implementation of its radical measures, the complete unification of the many regions of the empire, and the pacification of neighbouring countries or tribal zones were left to Han rulers.

The new social order envisaged by the first Ch’in emperor was posited upon a dynamic, omnipotent role for the emperor: he stood at the apex of society, aided by a meritocratic bureaucracy. The first Ch’in emperor, Cheng, was himself a man of extraordinary ability and ener­gy, and ably fulfilled his imperial role.7 Similarly, some Han rulers, such as the founder, Emperor Kao-tsu, Emperor Wu, and Emperor Kuang-wu, who restored the Han Dynasty, proved to be shrewd and competent administrators.8 Such a centralized political system which was dominated by one leader failed, however, when that leader was 3ineffectual, or was bored by affairs of state. Consequently, although on the surface the four centuries of the Han appear to enjoy cohesion and continuity, there were periods of severe political dislocation and social disorder.

Emperor Wu’s long reign of over half a century from the age of 16 to 70 years is a formative era in Han history. He introduced innovation and reform into many spheres of life. On the military front he extended his empire’s borders north-east into Korea, south into Yunnan and north Vietnam, east down to Chekiang and Fukien, and west to the Tarim Basin. These areas were colonized and assimilated into Han culture. The emperor’s armies also succeeded in expelling the northern enemies of China, the Hsiung-nu, as far away as the Gobi desert, but had to endure persistent and sporadic raids. Han envoys and generals even reached Ferghana to procure the famous horses of the region. In his economic policies, Emperor Wu made radical changes, such as the state appropriation of the lucrative salt and iron industries (previously under private ownership), and established a state mint. He also embarked on a policy of government marketing and trade, ostensibly to provide famine relief and to stabilize prices, but indirectly to ensure huge revenues for the treasury.

This dynamic ruler personally supervized the administration of his expanding empire and participated in the planning and construction of grandiose palaces and halls designed to reflect the glory of the Han. He built lavish mansions and parks. The splendour of his royal park was eulogized by one of his court poets, Ssu-ma Hsiang-ju, in ‘Rhapsody on Shang-lin Park’.9

Ch’angan, the capital city of the Former, or Western Han, was reconstructed and refurbished in 194 BC. Like many ancient Chinese cities it was formally designed as a rectangle, with its main gate facing south, the source of the cosmic Yang principle. At the beginning of the Christian era, the population of Ch’angan and the adjacent rural area was estimated to be about a quarter of a million, with approximately 80,000 inhabitants living within the city’s double walls.10 The rich lived in great luxury in storied, timbered mansions which were profusely carved and decorated in vivid colours. Silks and embroideries softened and brightened the interiors. They wore fine silks, furs, and jewels. At their banquets they enjoyed exotic menus served on dishes ornamented with gold or silver. Their carriages and horses were magnificently arrayed with lacquer, leather, silk, and jewels. Many of the wealthy residents attended court in an official capacity. Apart from their enjoyment of court entertainment, they often had their own orchestras and enter­tainers, with dancing, acrobatics, juggling, warrior dances with sword­play, cock-fighting, racing, and hunting.11

The numbers of people required for the service of the expanding class 4of nouveau-riche in Han China must have been considerable. Coachmen, stablelads, entertainers, cooks, stewards, builders, decorators, tailors, and scores of servants were resident in or near the mansion compounds within the city. Besides this, there were government offices to maintain throughout the city, which enlisted the services of many attendants and slaves.

In the countryside (termed yeh, or wilds) beyond the city walls lived the farmers and their hands. As has been the case throughout Chinese history, and in the present, the population was mainly rural. Although the Ch’in founder had inaugurated a system of land re-allocation so that the rural population could own land which previously had belonged to the aristocracy, the concept of an equitable system of land ownership became eclipsed during the Han, especially at the end of the Former Han (c. AD 9) and at the end of the Latter Han (c. AD 220). The recurring pattern was that due to natural disaster such as flooding, to heavy taxation, corvée duties when men were forced off the land each year for construction work or the army, and to the increase of family numbers owning the original parcel of allocated land, many displaced farmers and hands sold up and became tenant farmers or even serfs on large estates. The traditional contrast between city and country became more pronounced.

In this period, agriculture was the mainstay of the economy. It was a slow and painstaking process, occasionally based on animal-power, but more often on manpower. Very few technical contrivances aided the farmer, and heavy work like irrigation, sowing, harvesting, milling and so forth, were mostly manpower jobs.12 It was only towards the end of the Han era that some rudimentary mechanization was introduced, such as well-pulleys and milling rollers.

Many aspects of life in the city and country are reflected in the songs. The hymns hint at the pageantry of seasonal religious processions in the outskirts of the city. Political broadsides voice criticism of the govern­ment’s requisition of farm animals and the removal of men from the land at harvest-time, and the stockpiling of grain by fortune-hunting members of the imperial family. The war songs complain of con­scription, inadequate provisions, and the waste of human life in seemingly senseless construction projects far from home. Domestic ballads describe penury in the home, where infants are left to starve and the mother dies of neglect. Rustic fables express the grim and harsh reality of the countryside where death and destruction pose a constant threat. In the carpe diem ballads, especially, there appears a generalized sense of spiritual and social malaise; even at a banquet at which rich food and strong liquor are served, a singer sings a troubled piece about the meaning of existence.

In short, these popular ballads provide written evidence of the 5imbalance between certain sectors of Han society. While we may enjoy the luxuriant exuberance of the Han rhapsody and formal odes eulogizing the regime, the songs and ballads remind us of the people whose labour underwrote the glittering opulence of the city and whose own life was likely to be spent in poverty. The lack of merry, carefree songs and ballads testifies to the ordinary person’s miserable lot.

The Bureau of Music: Yüeh-fu

The repertoire of anonymous songs and ballads in the Han is inextricably linked to the establishment and development of the Han Bureau of Music, the Yüeh-ju. There is general agreement on the question of when the Han Bureau of Music was abolished: that occurred in the year 7 BC when Emperor Ai acceded to the throne and was convinced of the unethical, unorthodox trends in this government office.13 There is little agreement, however, on the questions of when and by whom this institution was established. The problem lies in the documentary evidence. Historical texts contemporary with the Han carry contradictory statements on the founder and the foundation date. The Latter Han historian Pan Ku (AD 32–92) states that the Bureau of Music was subordinate to another bureau, the Shao-fu or Lesser Treasury, which itself dated from the Ch’in and was responsible for the emperor’s personal needs—harem, household, robes, secretariat, missions, parks, arsenal, and so forth.14 The Former Han historian, Ssu-ma Ch’ien (?145–?90 BC) confirms the early date of the Bureau of Music when he states that the three early Han emperors who reigned between 195–141 BC ‘… neither added nor changed anything in the Bureau of Music; they simply practised ancient custom.’15 The existence of the Bureau in the early Han is even more definitely fixed by Pan Ku in his ‘Treatise on Rites and Music’ where he suggests it was active in the year 194 BC.16 Yet in the same treatise he implies a much later date, stating that Emperor Wu established the Bureau when he instituted new religious ceremonies between 114–113 BC.17 This text goes on to say that Emperor Wu appointed Li Yen-nien as his Master of Music, and that he often performed musical arrangements of poems and rhapsodies composed by Ssu-ma Hsiang-ju and others.18 In his biography of Li Yen-nien Pan Ku repeats his statement about the new cults, adding that Emperor Wu ‘… commissioned Ssu-ma Hsiang-ju and others to compose poems.’19 The problem here is one of chronology: Ssu-ma Hsiang-ju died c. 117 BC, about four years before Pan Ku’s founding date of between 114–113 BC of the Bureau of Music. Moreover, Li’s appointment as the court Master of Music took place at the time when his brother became a general, a promotion which is said to have occurred in 104 BC. A parallel text in the histories of both Pan Ku and Ssu-ma Ch’ien states that Li 6Yen-nien persuaded Emperor Wu to introduce music and dance into the service of thanksgiving for victory in Nan-Yüeh in 111 BC, which would indicate that the Bureau was in existence only quite late in the reign of Emperor Wu. Apart from the debate about the Bureau’s connection with Ssu-ma Hsiang-ju and Li Yen-nien as late as 114 BC, another group of texts suggests that the Bureau was formed in c. 121–120 BC, a date linked to the reported discovery of a ‘horse of heaven’.20

The earlier date of 120 BC for the establishment of the Bureau of Music by Emperor Wu is the one which is traditionally accepted. This date has its adherents among modern critics. J. R. Hightower, Burton Watson, and Hans H. Frankel accept that the Bureau was established by Emperor Wu around that year. Takeo Sawaguchi also tends toward this early date, arguing that Emperor Wu must have employed the musician Li Yen-nien at his court in 121 BC. Hellmut Wilhelm argues for a much earlier date, suggesting that the Bureau, ‘… like other institutions of Han times was a Ch’in establishment.’ P’eng Li-t’ien favours a pre-Ch’in date, suggesting that the Bureau originated in the late Warring States era. He cites as evidence the archaeological discovery of an inscribed official seal which reads: ‘The Seal of the Ch’i State Bureau of Music’.21 M. Loewe prefers a dating around 114 BC.22 G. S. Williams, after presenting the most important documentary evidence, declines to settle on a date.23

J.-P. Diény, in my view, offers the safest and soundest compromise in this conflict of textual evidence. He thoroughly reviews the texts, pointing out inconsistencies and exposing the flaws in various arguments. He concludes that the Bureau of Music existed from the beginning of the Han and was not created by Emperor Wu or during his reign. On the other hand, Diény points to the year 111 BC, when Li Yen-nien was appointed as Master of Music, as the commencement of a period when Emperor Wu possibly renamed an existing government office of music and, more importantly, completely reoriented the concept, functions, and scope of the office. So radical was this institutional reform, Diény argues, that it must have seemed at the time to be the creation of an entirely new institution.24 The mainspring of this reform was the introduction of popular elements into religious and secular ceremonies which reflected Emperor Wu’s taste for modern music from foreign countries and traditional folk music and dance.

The various texts which mention the Bureau provide some vague information about its functions. Its alleged tasks were to collect popular songs from regions of the Han empire (especially Chao, Tai, Ch’in, and Ch’u) and from foreign countries; to adapt and orchestrate popular songs; to compose new hymns and put them to music; to put existing texts to new, modern tunes; to set to music compositions commissioned from famous poets, such as Ssu-ma Hsiang-ju; to participate in, and perform at, religious and secular functions.25 Virtually nothing is known 7of how the Bureau functioned, but its great popularity may be judged from a statistic given at the time it was abolished in 7 BC: no fewer than 829 personnel were employed in its different departments. Their number was reduced by about half and the functions of the Bureau were redistributed among other government offices.26 In the Latter Han the Bureau seems to have been revived, but it did not survive much beyond the Han.

The Genre of Yüeh-fu Songs and Ballads

In the post-Han literary tradition, certain hymns, songs, and poems acquired the name of the Han Bureau of Music, or Yüeh-fu, from which they were believed to have historically emerged. The first literary figure to use the generic term yüeh-fu is Shen Yüeh, late fifth century AD. In the ‘Autobiographical Postface,’ ch. 100 of his History of the Southern Sung, he enumerates 17 genres in which his older relative, Shen Liang, composed 189 pieces, the ninth genre being yüeh-fu. He failed to attach the term, however, to any text in the collection of Han and post-Han songs in the ‘Treatise on Music’ of his History. Therefore, although Shen may be said to be the earliest literary figure to use the term yüeh-fu for a composition, he was not the first to apply the term specifically to extant song-texts. The first to apply the term to such texts were Hsiao T’ung and Hsü Ling, compilers of early sixth century imperial anthologies. Anthology of Literature (Wen hsüan), compiled by Hsiao T’ung and others, and New Songs from a Jade Terrace (Yü-t’ai hsin-yung), edited by Hsü Ling, both contain valuable Han yüeh-fu material, so named and so dated. A somewhat earlier work of literary criticism, The Literary Mind and the Carving of Dragons (Wen-hsin tiao-lung) by Liu Hsieh (d. c. AD 523) established the authenticity of the genre of yüeh-fu songs and ballads by elevating them to a separate category after sao-elegies and shih-lyrics but before fu-rhapsodies.27 The terms yüeh-fu or ku (old) yüeh-fu, designating Han pieces from the perspective of post-Han writers and commentators, appear as a sub-section of shih poetry in Anthology of Literature, ch. 27–28, and independently in New Songs from a Jade Terrace, ch. 1. But already in these three post-Han sources the classification of pieces by generic titles began to be confused. For example, the Han Emperor Kao-tsu’s song, ‘A Great Wind Rises’, is classified as a miscellaneous song (tsa-ko) in the yüeh-fu section of Anthology of Literature, but as a plain yüeh-fu piece by Liu Hsieh. Similarly, the Han Emperor Wu’s song, ‘Autumn Wind’, is classified as a tz’u, or Ch’u elegiac piece in Anthology of Literature. Six centuries later, Kuo Mao-ch’ien classified Kao-tsu’s song under category 8, ‘Song-texts for Lute Pieces’, among ko-songs and tsao-lute songs, and Emperor Wu’s song under category 11, ‘Words for Miscellaneous Ditties’, among ko-songs.28 8

After the sixth century AD, the generic term yüeh-fu became fixed and, in the course of time, was attached to an increasing number and variety of songs to be found in Han and post-Han sources. The repertoire of yüeh-fu actually surviving from the Bureau cannot be ascertained because songs are not recorded under that generic title in extant Han documents. By the early twelfth century, though, they formed a vast repertoire of, so-called, yüeh-fu texts, notably anthologized by Kuo Mao-ch’ien.

The corpus of Han songs has come down in a piecemeal and rather battered fashion. The earliest source is Pan Ku’s ‘Treatise on Rites and Music’ and ‘Treatise on the Five Elements’. The former contains Han hymns, the latter, political lampoons. Another treatise by Pan Ku, the ‘Treatise on Literature’, tantalizingly lists compositions under twenty-eight headings but it does not give the text of the songs. So in cases where his titles correspond with those in the extant repertoire we cannot be sure that his tide represents the same surviving text. On the whole, however, his titles do not tally with those of the present repertoire. When Han historians and ritualists discussed music, it was mainly at the level of theory, or else in the form of scattered comments.29

The next, and most important, source is the ‘Treatise on Music’ in the History of the Southern Sung by the poet and musicologist Shen Yüeh (AD 441–512), compiled some two and a half centuries after the end of the Han. As was the case with Western ballads, ‘It was the literary men who made the ballad a prestigious enough article to be granted the status of [book form] in the course of time’.30 His ‘Treatise’ includes a valuable collection of ku-tz’u (old songs), some of which he lists under the rubric of hsiang-ho, ‘Concerted’ pieces, defined by him as old songs of the Han era.31

Chronologically, the next sources are the previously mentioned Anthology of Literature and New Songs from a Jade Terrace of the sixth century AD. The source par excellence of anonymous songs and ballads, however, is the Anthology of Yüeh-fu Poetry (Yüeh-fu shih chi) compiled by Kuo Mao-ch’ien. Very little may be gleaned about Kuo’s biography. He lived most of his life in the Northern Sung era. His grandfather, Kuo Pao, and his father, Kuo Yüan-chung, were both famous officials in Yün-chou (modern Shantung) in the first half of the eleventh century. Kuo Mao-ch’ien, was the eldest of five sons (and five daughters). His courtesy name was Te-ts’an. In about AD 1084 he took up a subordinate official post in the prefecture of Ho-nan. He became a specialist in music. Besides his great anthology, he compiled the Anthology of Miscellaneous Verse Forms (Tsa t’i shih chi), which is no longer extant, and some of his own poetry. It seems likely that he compiled his Anthology of Yüeh-fu Poetry by the end of the Northern Sung era, c. AD 1126, because it was already being mentioned by literary critics by the beginning of the Southern Sung, AD 1127.32

His anthology comprises nearly all extant yüeh-fu texts between the 9Han and T’ang to Five Dynasties. Kuo’s collection, allowing for some wrong attributions, is now considered to contain the main repertoire of traditional songs and ballads. They are hymns, ceremonial songs, marching songs, folk-songs, dance songs, songs for the lute, miscellaneous airs, ‘modern’ (i.e. Sui-T’ang) songs, and ‘new’ (i.e. T’ang) yüeh-fu. Kuo arranged this vast corpus of 5,290 pieces into one hundred chapters, subdivided into twelve categories, for the most part based on the musical setting of the song-texts.33 Unfortunately, the music has been lost. Kuo also arranged his material in a unique and innovative manner: he presents an original piece, such as an anonymous Han song with its variant text, and immediately after it he places influential imitations of it by known poets of the Han to the T’ang and Five Dynasties. This organizing principle is of great literary value, for it enables the reader to compare the original with its later imitations, and to contrast existing versions of the early text.

One’s first response to this vast corpus of songs and ballads is that it manifests generic diversity. Under one so-called genre of yüeh-fu are subsumed a variety of other genres or sub-genres, each with its own generic title. For example, ko (song), ko-tz’u (song-words), ko-ch’ü (song piece), hsing (suite [?ballad]), yin (lay), yin (lament), ch’ü-tz’u (melodic words), p’ien (piece [?folk-song]), tsa-ch’ü (miscellaneous piece), tsa ko-yao tz’u (words for miscellaneous ditties).34 By Kuo Mao-ch’ien’s time a yüeh-fu piece indicated almost any song prior to the Sung Dynasty.35

It is interesting to compare the continuing controversy among Western literary critics concerning the origins of the ballad, and what constitutes a ballad, with its Chinese counterpart. In both cases the genre defies an exclusive definition. As A. Bold succinctly defines the Western ballad, ‘The ballad can accept any number of epithets, and there are tuneless ballads as well as broadside ballads and popular ballads and literary ballads.’36 G. H. Gerould’s definition of the Western ballad neatly combines the two terms ballad and folk-song: ‘A ballad is a folk-song that tells a story with stress on the crucial situation, tells it by letting the action unfold itself in event and speech, and tells it objectively with little comment or intrusion of personal bias.’37 In generic terms we may define the anonymous Chinese ballad or song of the Han era as follows: it was composed by an anonymous amateur and, until fixed in a text, was in subsequent folk possession and became part of a popular repertoire; originally oral, it was later fixed in texts of varying degrees of literary quality and usually with a number of variant words or phrases; it was set to music which is now lost; it is relatively simple in diction, content, and point of view; it may either tell a story as a narrative ballad, or convey an emotion as a lyrical ballad; it may describe action and feeling with numerous musical devices, such as repetition and refrain; its title usually includes a term relating to song or melody. 10

Kuo Mao-ch’ien’s Twelve Categories of Yüeh-fu Songs and Ballads

The twelve categories which Kuo Mao-ch’ien used to order his yüeh-fu material did not originate with him. The specialization of types of music is already apparent in Pan Ku’s discussion of numerous sub-sections of the Bureau of Music at the time of its radical reorganization under Emperor Ai in 7 BC. Most of these sub-sections were categorized as religious (suburban sacrifice) or regional (Han-tan drumming, Huai-nan drumming, and the like). In the reign of the Han Emperor Ming (AD 58–75) four musical categories of song were distinguished: 1) The Grand Yü Music, 2) Music for Classical Odes and Praise Songs, 3) Music for Drumming Songs of the Yellow Gate, and 4) Music for Songs for Short Pan-pipes and Nao-bell. Another set of four classes of music for songs in the Han is attributed to Ts’ai Yung (AD 133–92), the statesman, classicist, poet, astronomer, and musician, who is also credited with authorship of the book, Song Melodies for the Lute (Ch’in ts’ao). These four were: 1) Sacred Mysteries of the Suburban and Temple (Sacrifices), 2) Imperial Banquets, 3) Court Entertainments with Archery (Contests), and 4) Songs for the Short Pan-pipes and Nao-bell.38 In the Northern Sung Dynasty, Kuo utilized much of this early system. The first category of the second set became his first category, ‘Words for Suburban and Temple Songs’. He combined the second and third categories of the same set to form his second category, ‘Words for Banquet and Archery Songs’. He also combined the last two categories of the first Han set, equivalent to the last category of the second set to form his own third category, ‘Words for Drumming and Blowing Pieces’. He probably did not derive this material at first hand, but through the T’ang critic Wu Ching (d. AD 749), who will be discussed shortly.

Shen Yüeh, the late fifth-century statesman, historian, and poet, is to be credited with the emphasis on anonymous popular songs of the Han, giving them the cachet of orthodoxy by including them in his ‘Treatise on Music’ in his History of the Southern Sung. He termed them hsiang-ho‚ ‘Concerted’, and wrote of them: ‘All the musical songs that survive today are Han period ditties from the streets and lanes.’39 We may translate ‘streets’ as the city, ‘lanes’ as the country. This group became Kuo’s fifth category, ‘Words for Concerted Songs’. This contains most of the songs and ballads generally regarded as the repertoire of anonymous Han yüeh-fu.

Shen Yüeh’s ‘Treatise on Music’ is a valuable literary document, for it constitutes the earliest source of Han yüeh-fu songs and ballads and of information about them. Its four volumes (ch. 19–22) comprise the following material: 1) a discussion of song and music from legendary times to Shen’s own day, AD 488; 2) an anthology of hymns and praise songs of the Wei, Chin, and Southern Sung Dynasties; 3) an anthology of ku-tz’u, old songs, anonymous, popular pieces deriving from the Han, 11and their imitations by famous poets of the immediate post-Han era; and 4) an anthology of song-texts for miscellaneous named dances and for nao-bell pieces.

Shen divided his material into four musical categories: 1) Hymns and Formal Court Pieces, 2) Hsiang-ho, ‘Concerted’ pieces, Ch’ing-Shang Pieces in Three Modes, Major Pieces, and Ch’u-mode, 3) Songs for Dances and Dance Melodies, and 4) Drumming and Blowing Pieces.40 Only with category 4) did he attach the label ‘Han’ to nao-bell songs, making the set of 18 Han Songs for the Nao-bell the earliest firmly dated pieces in his collection. (For the list of his categories, showing the Han songs I have translated, see p. 206.) Elsewhere he affixed the label ku-tz’u, ‘old songs’, to certain titles such as 16 Hsiang-ho pieces in ch. 21. Since he affixed post-Han dynastic labels to all other titles, i.e., Wei, Chin, and Southern Liu-Sung, it may be assumed that by ku-tz’u Shen meant songs of the Han era, or even earlier perhaps, in his estimation.

On the other hand, it would be rash to infer that these texts, designated as ku-tz’u and Han by Shen, were transmitted intact from the Han to the Southern Sung era. Shen Yüeh frequently states that the recension of the old texts was a preoccupation of several musicians and poets of the Wei court. From his discussion and song collection it is clear that the processes of restoration, reconstruction, reinterpretation, and imitation were initiated by poets and musicians such as Fu Hsüan (AD 217–278), and firmly established by Ts’ao Ts’ao (AD 155–220), Ts’ao P’ei (AD 187–226), Ts’ao Chih (AD 192–232), and Ts’ao Jui (AD 204–239), who were all born in the Latter Han era, some barely surviving the dynasty itself. Moreover, since Shen Yüeh adopted the format of placing some ku-tz’u tides after their poetic (not necessarily musical) imitations by the royal Ts’ao poets, it follows that those ku-tz’u titles must date at least from the Latter Han, and in some cases possibly earlier.

Shen Yüeh credits eight musicians of the Wei court with the valuable work of the recension and transmission of the texts of the anonymous popular songs of the Han, each specializing in different branches of music: one Mr. Sun, Ch’en Tso, Ho So, Chu Sheng, Sung Shih, Lieh Ho, and Hsün Hsü and his son Hsün P’an (‘Treatise on Music’, ch. 19.559, ch. 20.603, ch. 19.542). Hsün Hsü is especially commended for his work on old song titles and music, Hsün’s Poetry Notes. This is no longer extant, but is extensively cited in subsequent yüeh-fu literary criticism. Hsün also composed sets of hymns (two sets being preserved by Shen Yüeh in ch. 20). Shen also records that ‘“Ch’ing-Shang Song Poems in Three Modes: Level-mode, Clear-mode, and Zither-mode” are old texts which Hsün Hsü selected and realized’ (ch. 21.608).

The ‘Three Modes’ are interesting in so far as they differ from Kuo Mao-ch’ien’s later perception of ‘mode’ (tiao). In the first place, Kuo linked the Ch’u-mode, a quite separate modal category in Shen, to 12form four modes. Secondly, the appellation Ch’ing-Shang Kuo divorced completely from the modes, which he restricted to Hsiang-ho, ‘Concerted’ pieces. For Kuo, Ch’ing-Shang pieces meant certain post-Han popular songs. Kuo reallocated certain titles within Shen’s categories to different mode categories. For example, Kuo transferred several of Shen’s ‘Major Pieces’ titles to the ‘Zither-mode’ and one to the ‘Ch’u-mode’. One possible conclusion to be drawn from this is that the texts of yüeh-fu, once they were committed to writing from an earlier oral tradition, were more or less preserved unchanged, whereas the more fugitive form of music, not fully notated, altered through the centuries to suit the prevailing taste in music. Musicians wrote in different modes for the same song-text. Alternatively, during the seven centuries between Shen Yüeh’s times and Kuo’s, from AD 488 to about AD 1126, the original music of the Han songs was lost and the apprehension of mode per se became obscured, resulting in an antiquarian reconstruction of mode categories in the early twelfth century, or earlier, which sometimes proved to be quite faulty.

The next important classification of music is the nine categories of Wu Ching in the T’ang. He eliminated the hymn and praise song category, while up-grading popular songs to first place. In his Explanations of the Old Titles of Yüeh-fu he presented nine groups of song titles, indicating after each group to which musical or prosodic category the group belongs. His nine categories are: 1) Hsiang-ho, Concerted (pieces), 2) Sweeping Dance Songs, 3) White Hemp Songs, 4) Songs for the Nao-bell, 5) Horizontal Flute Pieces, 6) Ch’ing-Shang Pieces, 7) Miscellaneous Titles, 8) Pieces for the Lute, and 9) Old Titles and Miscellaneous Titles.41 Of these the first and fourth have been encountered before. Wu’s second and third categories became Kuo’s seventh, ‘Song-texts for Dance Pieces’. Wu’s fifth became Kuo’s fourth category, ‘Words for Horizontal Flute Pieces’. Wu’s sixth remained Kuo’s category number six, ‘Words for Ch’ing-Shang Pieces’. Wu’s seventh is probably the equivalent of Kuo’s ninth, ‘Song-texts for Miscellaneous Pieces’, while Wu’s eighth remains Kuo’s. It is Wu’s last category which seems the most deceptive. A valuable source of pre-Han, Han, and post-Han titles, it became Kuo’s eleventh category, ‘Words for Miscellaneous Ditties’. Thus far, ten of Kuo’s categories have been found to derive from pre-existing musical and prosodic categories of song. Kuo’s tenth and twelfth categories contain songs and poems dating from the Sui-T’ang to Five Dynasties era.42

The clue to Kuo’s borrowing from Wu’s scheme lies in his uncritical adoption of the first part of Wu’s last category, ‘Old Titles’, as his own eleventh. For in Kuo’s day some of these titles were believed to date from hoary antiquity and were alleged to be songs by the legendary Emperor Yao. It would have been more logical for Kuo to have opened his massive anthology with these ancient rimes, following on with the Han hymns 13and banquet pieces. In fact, a later anthologist of yüeh-fu, Tso K’o-ming (14th century), did precisely this when he arranged his yüeh-fu pieces in eight categories in Ancient Yüeh-fu. Tso began with ancient rimes, omitting hymns and formal ritual pieces; then he followed Kuo’s categories up to Kuo’s ninth; he omitted Kuo’s tenth and last categories. In this scheme Tso was clearly presenting his material in a more or less chronological manner, but he was also attempting to provide an ancient pedigree for the Han yüeh-fu, thus rendering them acceptable as orthodox literature. Tso’s scheme of the late Yüan was followed in the main by Feng Wei-no of the Ming, but thereafter the musical categories tended to be dropped and yüeh-fu merged with poetry in literary criticism. The yüeh-fu material was arranged chronologically with shih-lyric poetry as with Shen Te-ch’ien, and more recently with Ting Fu-pao and Lu Ch’in-li.43

While it appears that Kuo Mao-ch’ien merely adapted his scheme from existing schema from the Han to the T’ang, he may be judged original in the following ways. He introduced the organizing principle of arranging original songs and their later imitations chronologically within each of his twelve categories. He placed versions of the same title next to each other. He presented his categories more or less in sequence as each historically evolved. Also, he was the first to amalgamate the traditional categories of religion and court ceremonial with the more popular, secular musical idiom, although it is true that he was prompted by earlier exemplars of musical classification. Finally, he strove to preserve the remnants of musical information associated with many of his categories, even though by his day the musical accompaniment was largely lost. In this sense he was a true Sung scholar and an antiquarian of the first rank.

Sources of Kuo Mao-ch’iens Song-texts and Their Contexts

When Kuo Mao-ch’ien compiled his anthology of yüeh-fu songs and ballads in around AD 1126, he looked back on a period of over a millennium during which the genre had evolved and matured. In his own day a new song form, the tz’u lyric, had displaced the yüeh-fu and had become the fashionable medium for poets and song-writers to exercise their talents. Although the yüeh-fu had influenced the tz’u at its inception in the T’ang Dynasty, by Kuo’s time the yüeh-fu was regarded as an old-fashioned literary form which had enjoyed its renaissance in the T’ang. In a sense, the yüeh-fu was becoming defunct by the end of the eleventh century and was in need of conservation. Kuo admirably performed this task of preserving a vast repertoire of familiar and recherché titles. His was an age of literary compilation on a grand scale. Early in the Northern Sung, Li Fang and his colleagues were com­missioned by Emperor T’ai-tsung to compile a compendium of 14knowledge which resulted in The Four Books of the Northern Sung consisting of over 3,500 volumes. One unintended advantage of such compilations was that much fugitive material has been preserved.44

To complete his task of gathering virtually every song title prior to the Sung, Kuo referred to over 160 works, either at first or second hand.45 Even by Shen Yüeh’s day the works of certain authorities on yüeh-fu were no longer extant, such as Hsün Hsü (d. AD 289). Collections of songs and critical comments were included in treatises on music and literature in the official histories through the centuries, so that by the end of the Northern Sung a considerable amount of research material on old yüeh-fu had accumulated. Much of this material Kuo incorporated into the essays prefacing his twelve categories and some sub-categories. The main treatises from the histories, arranged chronologically according to the dates of their authors, are as follows:

Historical Records (Shih chi) ch. 24, ‘History of Music’(1st cent. BC)

History of the Han (Han shu) ch. 22, ‘Treatise on Rites and Music’ (1st cent. AD)

History of the Han (Han shu) ch. 30, ‘Treatise on Literature’ (1st cent. AD)

History of the Latter Han (Hou Han shu, hsü) ch. 1–3, ‘Treatise on Pitchpipes’, ch. 13–18, ‘Treatise on the Five Elements’ (3rd–4th cent.)

History of the Southern Sung (Sung shu) ch. 19–22, ‘Treatise on Music’ (5th cent.)

History of the Southern Ch’i (Nan-Ch’i shu) ch. 11, ‘Treatise on Music’ (5th–6th cent.)

History of the Wei (Wei shu) ch. 109, ‘Treatise on Music’(6th cent.)

History of the Chin (Chin shu) ch. 22–23, ‘Treatise on Music’(6th–7th cent.)

History of the Sui (Sui shu) ch. 13–15, ‘Treatise on Music’ (7th cent.)

Old History of the T’ang (Chiu T’ang shu) ch. 28–31, ‘Treatise on Music’ (9th–10th cent.)

New History of the T’ang (Hsin T’ang shu) ch. 11–22, ‘Treatise on Rites and Music’ (11th cent.)

Apart from this material from the histories Kuo made use of a number of works from the Han to the Sung by specialists on music and literature. One third of this material existed only in a fragmentary form (marked with an asterisk in the list below).

Record of Ritual (Li chi), ch. 11/19, ‘Record of Music’ (1st cent. BC)

* ‘Treatise on Rites and Music,’? Ts’ai Yung (2nd cent. AD)

Master Ch’iao’s Legal Precepts, Ch’iao Chou (3rd cent.)

* Hsün’s Poetry Notes, Hsün Hsü (3rd cent.)

Record of Things Ancient and Modern, Ts’ui Pao (3rd–4th cent.)

Record of Neglected Data, Wang Tzu-nien (4th cent.)

* Arts Record of Correct Music…, Chang Yung (5th cent.) 15

* Arts Record of Banquet Music…, Wang Seng-ch’ien (5th cent.)

* Procedures for Controlling the Army, Liu Hsien (5th cent.)

Anthology of Literature, Hsiao T’ung et al. (early 6th cent.)

New Songs from a Jade Terrace, Hsü Ling (mid-sixth cent.)

* Record of Ancient and Modern Music, Chih-chiang (late 6th cent.)

* Record of Songs, anon. (pre-T’ang)

Compendium of Literature, Ou-yang Hsün (6th–7th cent.)

Commentary on Anthology of Literature, Li Shan (late 7th cent.)

Notes for Beginning Students, Hsü Chien (7th–8th cent.)

Explanations of the Old Titles of Yüeh-fu, Wu Ching (7th–8th cent.)

Comprehensive Survey of Classical Learning, ‘Music’ ch. 141–47, Tu Yu (8th–9th cent.)

Miscellaneous Records of Yüeh-fu, Tuan An-chieh (9th cent.)

Imperial Survey of the T’ai-p’ing Era, Li Fang (10th cent.)

Prize Blooms from the Garden of Literature, Li Fang (10th cent.)

From these works Kuo derived texts, background information, data on music, theoretical writings on music and literature, and variant texts and readings. To control and organize such a mass of data bears witness to his artistic and scholarly genius.

The Oral Nature of Han Songs and Ballads

Balladry has attracted a great deal of theorizing, sometimes informed by ideology. In eighteenth-century Europe, Johann Gottfried von Herder asserted that ballads originated with the folk and expressed the voice of the people.46 His theory nicely dovetailed with his desire for a national renaissance in Germany. In China, too, we find in pre-Han and Han texts political theorists, ritualists, and historians arguing that local songs produced by the folk represent the pure expression of the people.47 These writers insisted that the ruler should receive the songs collected from among the people, and in some cases portrayed the sage ruler as one who personally engaged in this activity. The political benefit was explained in this way: ‘Thus the king, without going to the window or leaving the door, knows fully the sufferings of his realm.’48 In their persuasions to a ruler the theorists would always argue from the stand­point of authority, the authority of the ancient sage rulers. Whether this theory was based on historical fact is subject to doubt, but clearly the theory became a potent political myth, for it promised that public opinion might be divined and even influenced. Several Han rulers, especially in the Latter Han, are reported to have sent officials to collect folk-songs, for example, Emperor Kuang-wu, and Emperors Ho, Shun, and Huan. The political theory of collecting folk-songs in the Han owes its origin to the parallel tradition that Confucius had compiled the Book of  16Poetry’s 305 songs from a collection of 3,000 Chou pieces.49 Unfortunately, we know as little about the true manner of the compilation of that classic anthology as we do about the provenance of Han ballads.

One historic episode in the life of the founder of the Han perhaps provides some clues about the oral origin of balladic art. In the winter of 195 BC, when Emperor Kao-tsu was on his way back to the capital, he stopped in the countryside at P’ei in the region of former Ch’u state, where he had begun his politico-military career. At a banquet in his honour, overwhelmed by the occasion with its memories of hardship and triumph, he composed a spontaneous song, ‘A Great Wind Rises’. Accompanying himself on a stringed instrument he chanted it. Then he formed a choir of some 120 local boys who joined in and repeated his song while the emperor danced, weeping with grief. After his death the choir was charged with performing the song at the memorial temple at P’ei.50 In this vignette we note several features that might pertain to other Han songs: the emperor was a commoner who had educated himself along the way to victory; he was a composer of simple song; he could play a musical instrument; he could dance, improvising the steps and tempo to suit his own composition; the context of his amateur composition was a banquet where wine (the other muse) flowed freely; the song, dance, and music were spontaneous expressions of profoundly felt emotion; some members of the community learned the piece and repeated it thereafter in the community as a remembered song-text with music and dance; the text was preserved in the official biography of the emperor. With an imperial patron like Emperor Kao-tsu, and later Emperor Wu, it is not surprising that the simple, beautiful art of balladry was nurtured and flourished, particularly in the Han.

When we search the anonymous Han ballads themselves for evidence of orality we may make a few observations. First the children’s songs, or t’ung-yao, are a sub-category of anonymous Han ballads which were deployed by historians as prophetic utterances, but were traditionally believed to have been produced in the ‘streets and lanes’, presumably by illiterates. Secondly, if we compare two formal versions of the same poem in the quite different styles of the ku-shih, or old poem, and the ballad, yüeh-fu, we notice at once that the version in the ku-shih style is concise, metrically regular, literary, and polished, while the other version in the yüeh-fu style is often prolix, metrically irregular, and without literary polish. This contrast is most clear in the ‘West Gate Ballad’ and ku-shih no. 15 of the ‘Nineteen Old Poems’. Clearly, the ku-shih version was a literary piece composed by a professional poet, the ballad, or yüeh-fu, was an oral piece composed by an amateur.51 Thirdly, the presence of textual versions among some ballads, that is ballads having two or more versions, one usually longer, suggests, but does not prove 17their oral origin. Variations on an original ballad continued apace from region to region and from time to time until the oral ballad died or was fixed in print. Conscientious archivists, such as Kuo Mao-ch’ien, happily preserved these alternative versions, even if he did err in the matter of their dating.52 It so happens that we possess two versions of ‘West Gate Ballad’. One is longer, and there are differences in diction and point of view. The long version treats the carpe diem theme and bases its point of view on the rejection of religion, philosophy, and drugs; the shorter version treats the same theme, but its point of view is based on the speaker’s wretched poverty. A song going by a different title, ‘A Yen Song, The Whenever Ballad’, appears to constitute even a further version on the same theme, and might in fact be the original among the three songs.

Apart from these oral aspects of Han yüeh-fu, there are oral features typical of popular songs and ballads, such as colloquialism, direct speech, earthy or folkloric subject matter, a rustic, unlettered point of view, nonsense words, and commonplace expressions.

The oral origin of anonymous Han ballads does not mean that they were invariably produced by illiterate folk. For instance, ‘The Ballad of the Prefect of Goosegate’, ‘The Lament of Liang-fu’, ‘The Ballad of Breaking Willow’, and the hymns are the products of educated composers. Their categorization as ballads rather than as odes or literary hymns must signify their subsequent folk possession rather than their folk composition. Moreover, several ballads contain hints of literariness in diction and content. We know, for example, that the pursuit of immortality through elixirs and cults was originally the province of the ruling classes. Later as the cult of immortality lost imperial patronage it filtered down to the people with its arcane lore and language. In other ballads, the presence of non-folk elements indicates that there was an interaction between popular and literary culture and a creative inter­relationship between city and country.

Many features which characterize the Western ballad are present in most anonymous Han ballads, and it seems useful to summarize them here: universality of point of view; stereotyped characterization, situation, and plot; detachment, impersonality, and objectivity in presentation; no localized action or setting; ahistoricity, or a lack of historical perspective; no analysis; no introspection; commonplace phrases; little plot; weak narrative line; abrupt entry into the balladic story and equally abrupt transitions; understatement; dialogue; nonsense words; alliteration; refrain; repetition; irregular metre; colloquialism; a homely style drawing on familiar material rather than from the literary tradition.18

The Art of Han Songs and Ballads

The Han song is storytelling set to music. Although the music can now only vaguely be hinted at, the songs contain striking features which allow them to be enjoyed as literature to be read rather than songs to be heard at a banquet or a party. Without the charm of music and the conviviality of the banquet the ballads and songs lend themselves to a different kind of interpretation and appreciation. Whereas oral art in performance is a fleeting experience, in which the note, once sung, lingers only in the memory, ballads and songs as literature may be savoured more deeply and examined in a more leisurely manner.

The singer of tales is essentially a narrator who introduces us to a crucial situation and shows us how certain characters respond to the crisis. Hans H. Frankel identifies three types of relationship between the narrator and the characters of early narrative yüeh-fu: 1) ‘objective third-person narration’, 2) ‘impersonation’, and 3) an interchange between objective narration and impersonation. From his analysis of Han and post-Han narrative ballads he finds that the third type is more common. What particularly distinguishes the genre is the ‘multiple and shifting role of a single performer as narrator and actor, impersonating one or more characters.’53 His classification system may be usefully applied to the larger repertoire of Han yüeh-fu I have translated.

Because the storyteller is a performer, the narration is enlivened by dramatic techniques. Sometimes we are rushed headlong into the situation when the characters are at a breaking-point: a sick wife lies on her deathbed and begs her husband to care for her infants; a man leaves home to escape bleak penury; a soldier comes home from a lifetime in the army to find his house in ruins and deserted; or a girl finds that the man she hoped to marry has gone off with another woman. The narrator sketches in some details of a character’s life prior to the action of the tale, for example, the opening lines ‘An orphan boy I was born’ and ‘At fifteen I joined the army’. Plot is not so much the unfolding of events by action as the revelation of events by the characters’ thoughts, words, and feelings. In ‘East Gate’ we follow a poverty-stricken husband’s emotional conflict as he decides whether to leave his wife and family. In ‘West Gate’ we listen to a man at a banquet as he questions the meaning of life. The endings of the ballads often leave a scene of despair and destruction, with infants wailing, an old man weeping, or an orphan longing to die. There is little character development, for this art form deals with one-dimensional types, as some titles suggest; ‘The Ballad of the Ailing Wife’, or ‘The Ballad of the Orphan Boy’.

The song narrative is characterized by abrupt transitions, a technique identified by F. B. Gummere as ‘leaping and lingering’.54 Action is introduced swiftly and then dallied with, while the listener or reader 19wonders what will happen next. In the fable about two white swans the sudden illness of the hen-swan is succinctly stated: ‘The wife suddenly falls ill’, and for three more lines her faltering flight is lingered over. In another fable, which tells of a butterfly snapped up by a nursing swallow (euphemistically described as ‘suddenly it meets a swallow’), the action is delayed as the swallow circles slowly round before popping the chewed morsels of butterfly into her chicks’ wide-open beaks. Perhaps because the Han songs and ballads are so early in the tradition of narrative art the transitions sometimes tend to be abrupt to the point of senselessness. In ‘East Gate’ the voice switches so awkwardly that it is difficult in trans­lation to assign dialogue to a specific speaker. In ‘Song of Melancholy’ the speaker remains the same throughout, but his actions are lurchingly conveyed: first he walks toward the sea, then we are told that he has not got out of bed. In ‘A Yen Song, A Ballad’ the narrative develops elliptically: a girl tells how her brothers are away from home amusing themselves, then she asks, ‘Old clothes who must mend? /New clothes who must sew?’ Later in the song the mistress’s husband is described lurking around the girl’s rooms. The girl addresses him with a plea and two proverbs to try and politely make him go away. In this case the abrupt transitions in the narrative are deliberately deployed by the singer who seeks to envelop the girl’s predicament in a set of codes, suggesting that his song deals with the social taboo of an unwanted pregnancy of an unmarried girl. The technique of ‘leaping and lingering’ is also skilfully used in ‘Orphan Boy’, where the swift change of setting, time, and action underscores the boy’s harried life. Other cases of ‘leaping’ narrative connections have less to do with technique consciously used than with problems of corrupt texts. ‘Vermilion Ibis’ contains strings of enigmatic phrases and words which are either nonsense like ‘hey nonino’ or else snatches of song of which the original meaning has been lost or deliberately censored. Clearly, in their original performance the music must have greatly enhanced the sense of the words. What seems meaningless to us was probably clarified by familiar musical refrains and interludes. The tone, tempo, and pitch of the human voice and musical instruments no doubt registered nuances of meaning which remains elusive today.

F. B. Gummere also isolated a device in Western ballads which he termed ‘incremental repetition’.55 Although these Han examples of balladry and song do not reveal such a sophisticated use of this device as in the West, it is present. In ‘Two White Swans’, the narrative develops through the speech of a male swan explaining to its sick mate why he cannot remain with her. His speech takes the form of a repeated pattern, with slight changes to augment the narrative line: ‘I would carry you away in my beak,/ But my beak is sealed and will not open./ I would carry you away on my back,/ But my feathers and tail each day would be 20crushed.’ A more extensive use of the technique occurs in the final passage of ‘Mulberry on the Bank’, where the pretty mulberry-picker, Lo-fu, describes her dashing husband to a passing official who wants to seduce her: ‘At fifteen he was a county clerk,/ At twenty a palace official,/ At thirty a gentleman-in-waiting,/ At forty lord of his own city.’

The dramatic content of the narrative is increased by direct speech, sometimes in the form of dialogue between humans or between animals, sometimes an address by the narrator to the audience, or else a sustained monologue. The patterns of speech are vivid and realistic. In ‘We Fought South of the City Wall’ a dead soldier cries out to carrion crows: ‘In the wilderness we dead clearly lie unburied,/ So how can our rotting flesh flee from you?’ In ‘Orphan Boy’ a cruel brother orders the boy to ‘Get the meal ready.’ In ‘East Gate’ a desperate husband leaving home shouts at his wife who is clinging to his clothes, ‘Get off! Get away!/ I’m late in leaving./ I can’t go on living here when my hair’s nearly white.’ One of the uses of direct speech is the formula of question and answer: one character says: ‘Oh why don’t I do it?/ I just have no pole or oar.’ Sometimes the formula serves as a means of introducing a set descriptive passage, as in ‘They Met’ and ‘In Ch’angan’ where the question is put, ‘And your lord’s house?’ by someone whose carriage has been caught up with another person’s vehicle, and the answer constitutes the rest of the ballad describing the house. Occasionally the formula is a kind of riddle: ‘Old clothes who must mend?’ The formula is employed in a more refined way in ballads which bear a strong resemblance to the European chanson d’aventure. For example, in ‘Mulberry on the Bank’ the formula creates a battle of wits between the sexes. In ‘The Lunghsi Ballad’ the question, ‘What’s up there in the sky?’ introduces an answering passage which creates a subtly ironic counter­point between the ideal world and the real world.

Balladic art is informed by the regular use of ‘remembered phrases’, or commonplace phrases within an oral tradition. Han ballads and songs vary between maximal and minimal use of such phrases. Maximal use occurs in ballads like ‘Cocks Crow’, ‘They Met’, and ‘In Ch’angan’ where whole passages appear in all three ballads, differently positioned and worded slightly differently. On the whole, however, the ballads reveal relatively little use of commonplace phrases, in some cases just over 25 per cent.56 This may have to do with the brevity of the songs and ballads, most of them being no more than 24 lines long, and several being only quatrains. The subject of formulaic expression has generally been discussed on the basis of long epics in the European tradition, in which a composer-narrator’s creativity and ability to memorize were extended to their greatest.57 Implicit in the discussion of formulaic expression is a negative prejudice, exemplified by the judgement of Sir Walter Scott: ‘The least acquaintance with the subject will recall a great number of 21commonplace verses, which each ballad-maker has unceremoniously appropriated to himself, thereby greatly facilitating his own task, and at the same time degrading his art by his slovenly use of over-scutched phrases.’58 In other words, the use of formulaic expression is evidence of plagiarism, and plagiarism denotes a lack of individual creativity.

Examining the use of commonplace phrases in Han songs and ballads from a more positive standpoint three main artistic aspects may be identified: familiarity, literary quotation, and ironic wit. Familiarity is evident in the repetition of certain tropes, gestures, and names. Gates of yellow gold, going up the hall of the family home, leaving by the east or west gate of the city, tears falling on clothes, cart wheels turning in the guts, epithets such as ‘brilliant’ or binomes like ‘green, green’, and stars and planets indicating time are all examples of commonplace phrases used with such frequency that they become familiar to the repertoire. This kind of repetition is posited on the joy of recognition, similar to that aroused by nursery rhymes or a shaggy dog story. It is part of the naive simplicity of the pieces.

Literary quotation involves the use of phrases from the earlier written tradition which have become almost cliché, such as the Taoist allusion in the opening couplet of ‘Cocks Crow’, or the felicitation closure of ‘The White Head Lament’ borrowed from songs of ancestral sacrifice in the Book of Poetry, or binomes borrowed from the same classic. By the time the Han song-maker used such quotations he was probably no longer citing from the locus classicus in Chou literature but was borrowing from a more generalized ragbag of quotation and fancy phrasing. Although literary quotation is not so extensive in the yüeh-fu genre as it is in the shih lyric in the Han, when it is used as a conscious device it reveals great sophistication, such as the way the girl in ‘A Yen Song, A Ballad’ quotes allusively to deflect the curious stares of her mistress’s husband.

The third use of commonplace phrase displays the song-maker’s wit. For example, in ‘The Lunghsi Ballad’ and ‘Walking out of Hsia Gate’ virtually the same quatrain describing paradise appears. In ‘Hsia Gate’ the quatrain ends the piece, presenting a vision of paradise intrinsic to its theme of immortality. In ‘Lunghsi’ the quatrain opens the piece, positing a vision of an ideal world which the human world fails to attain, thus creating an ironic viewpoint. These and other uses of commonplace phrases and verses, far from demonstrating the makers’ lack of originality, seem to me to prove their conscious deployment of familiar material in a variety of ways in order to engage the listeners’ interest, to hold their attention, and to amuse them with cunning reversals of expected response.

As one might guess from ballads and songs where narrative and dramatic techniques are dominant, figurative language is sparingly used. This is not to say that lyricism and imagery are almost absent, but that 22they are subordinated to the demands of storytelling. The clearest illustration of this is in the makers’ use of nature imagery, which else­where in Chinese poetry forms an imaginative leitmotif. Nature sometimes provides the scenarios of the ballad, such as in the elixir pieces, where mountains belong to the lore of immortality. Nature may serve as an indicator of time or season. It may also create a parallel or contrast between the impassive world and human experience, as in ‘Dew on the Shallot’. A vignette based on natural imagery may establish an emotion in the piece, prior to the narrative proper, as in the opening lines of ‘Along the Embankment’, where luxuriant reeds suggest the beauty and youth of the female speaker, and, through their use in the making of mats, cushions, and beds, also denote the speaker’s humble position in relation to the man she is talking about. Nature symbolism features in a number of pieces, either lyrically as white mountain snow in ‘The White Head Lament’, or ironically as nest-seeking swallows in ‘A Yen Song, A Ballad’. Symbolism informs the basic structure of some didactic pieces, such as the songs on the poplar and date-tree. In sum, the song-maker does not linger over the imagery in the same way as the shih lyricist. Imagery is used to develop the narrative, to set a scene, to suggest a mood or emotion, or to pre-figure the ballad’s denouement.

The art of the Han songs and ballads rests firmly on realism. Magic, marvels, the extraordinary, do not generally appear in the narratives. There is little superstition, nothing of ghouls or the grotesque. The super­natural, of course, is present in the elixir pieces, but there the presentation of it is low-key and realistic within the confines of the theme. In general, the stories of Han songs and ballads are tied to everyday situations and the characters’ dilemmas arise from these situations. Their dilemmas stem from poverty, ill-treatment by a cruel relative, war, love, imprudent behaviour, and greed. In other words they are problems born of human failing and social ills. The predicaments of the characters are presented through concrete details rather than articulated as abstractions.

In contrast to Western ballads, the Han repertoire appears to exercise a self-imposed censorship on the topics of sex and violence, or perhaps that censorship was imposed by Han officials who simply failed to record lurid songs. The sexual violence of the English ballad, ‘Little Musgrave and Lady Barnard’ in lines like ‘He cut her paps from off her breast;/ Great pitty is was to see/ That some of this ladie’s heart’s blood/ Ran trickling down her knee’, is entirely absent from the Han pieces. Another curious omission is the lack of the heroic. Despite the tradition of chivalry in early Chinese writings, immortalized in the figure of the wandering knight, yu-hsia, there is no equivalent in the Han repertoire of the Robin Hood sequence.59 This lack underscores my earlier point about the avoidance of the extraordinary in most Han pieces, a mark of their 23realistic mode. In fact, the songs and ballads are anti-heroic. They reveal that the ordinary person does not wish to be a hero or to die for some abstract cause. The will to live, or rather just to survive in a hostile environment is the recurring refrain. The anti-war ballads voice this sentiment most strongly.

Realism is evident in the use of speech. The everyday idiom of ordinary people is reproduced in dialogue and exclamations. Realistic descriptions enhance mundane situations, such as the empty foodbin and bare clothes-hanger in ‘East Gate’. Of course, ‘realism’ is notoriously ambiguous in any literary discussion. It might well be argued that the Han singers’ penchant for misery is subjective and runs counter to realistic depiction. Yet in their new focus on other areas of experience than had been explored in the tradition, in their depiction of the unattractive face of common humanity, they sang of the lives they led, the world they knew and understood. Looking beyond the splendour of the imperial court eulogized by Han court poets in near-sycophantic rhapsodies and beyond the glitter of the capital city to the drabness of diurnal reality, the song-makers reminded their listeners of the forgotten and despised, the insulted and the injured, a class of ordinary folk whose otherwise unsung labour had built the empire.

Some Musical Aspects of Han Songs and Ballads

To read only the words of any well-known traditional or modern ballad or blues song prompts the question, why have they become so famous and popular? Once the ballad or song leaves the printed page and rejoins its musical setting the words come alive again. The definition of a ballad proposed by the musicologist B. H. Bronson, ‘Question: when is a ballad not a ballad? Answer: when it has no tune’, may be modified to suit the context of Han songs and ballads: they may be termed songs and ballads although they have lost their tunes.60 What little information survives about the Han pieces indicates that they were sung to musical accom­paniment. The earliest recorders of the Han repertoire, who were interested in music in relation to poetry, such as Shen Yüeh, classified what we now call popular Han songs and ballads, Han yüeh-fu, under musical headings, such as hsiang-ho, ‘Concerted’ piece and tsa-ch’ü, ‘Miscellaneous Melodic Piece’, instead of grouping them under a literary genre or sub-genre.61 Some songs recorded by Shen Yüeh are listed as ‘old songs’, ku-tz’u, indicating that music is their underlying system.62 He also provided information about the sort of musical instruments that were used with these popular songs. This information was expanded by Kuo Mao-ch’ien in his preface to the twelve major categories of his collection, so that there is some indication of what instruments were played in the performance of certain pieces.63 Yet since Shen Yüeh’s 24information dates from a period two centuries after the end of the Han and Kuo Mao-ch’ien’s from about a millennium after the Han, much of this information has to be treated with caution.

The songs and ballads themselves provide firm evidence of their musical background. Many have stanzaic divisions, or chieh. Some song­-texts are followed by an indication that the stanzas are divided into musical sections. These are a yen, or prefatory passage, and a tsü or finale passage. The yen and tsü each required different kinds of musical accompaniment. The yen section usually forms the longest part of the whole song. For example, in ‘Mulberry on the Bank’ the yen comprises 35 lines, the tsü 18 lines. In ‘Two White Swans’, long version, the yen is 16 lines, the tsü 10. On somewhat slender evidence the yen is believed to derive from a style of song native to Ch’u state in the late Chou era (a style Emperor Kao-tsu enjoyed), while the tsü is believed to have originated in Wu state in the late Chou. This musical designation yen may explain why some songs bear cumbrous titles like ‘A Yen Song, The Whenever Ballad’, or ‘A Yen Song, A Ballad’, or ‘A Yen Song, The Ballad of Lo-fu’ (another title for the piece known as ‘Mulberry on the Bank’). Such pieces may consist of a medley or suite of passages which originally constituted independents yen-style songs. There is only one extant anonymous yüeh-fu title bearing the musical designation tsü, Wu tsü hsing, but it is not clear if this is a Han text.64 Apart from the chieh, yen, or tsü some songs contain an envoi, luan, usually at the end of the whole piece, or in one rare case in the middle, and consisting of about two or four lines. In fact, this rare use of the envoi in the middle of the piece, ‘Ailing Wife’, may shed some light on the methods of composition of the Han song. In this instance, the envoi, which normally appears at the end, stands in the middle and divides the narrative into two parts; the second half takes on a new tone which is at odds with the first half, suggesting that two songs have been patched together to form a new piece.

Besides these formal elements of composition, there are other more obvious musical elements common to the songs and ballads of other cultures. These are the repetition of passages, either verbatim, as in ‘East Gate’, or with a slight modification, as in ‘Two White Swans’. Lines are repeated, especially in that most musical of songs, ‘South of the River’. Refrains occur throughout the song version of ‘Tung Flees!’ and in ‘A Crow Bore Eight or Nine Chicks’.

The metres generally consist of regular passages and extremely irregular passages, suggesting a relationship between text and music.65 A melodic line might match or be matched by the metre and at the same time might indicate through the use of irregularity of metre, changes of mood and tone. The metrical irregularity of the ‘Orphan Boy’ illustrates this bond between metre and emotion, and between metre and episodic structure, which suggests in turn a responsive, evocative musical back­ground. 25Moreover, the opacity of some song-texts argues for the clarifying presence of music. If, as I have proposed, some Han songs were composed by piecing together previously independent songs, or parts of songs, and by incorporating some echoes of their original music, those listening to the new composition would understand the context of the piece; they would remember the original tunes that went with the original songs and, if only bits were incorporated into the new song, would also remember the context of the original song. It is possible to surmise that just as Han songs and ballads were composed by patching together old and new lines to make a new text, so the music was composed by interweaving remnants of old musical scores into the new idiom. Certainly, the Han song and ballad repertoire, being mostly a popular, oral art form, was designed for lucid communication, especially since a song performance is fugitive and elusive. The listeners would have demanded comprehensibility. That Han music which is now lost must have helped to develop, explain, and emphasize the meaning of the words. But today, without their music, several passages in the song-texts remain opaque.

The Singers of Yüeh-fu in the Han

Hans H. Frankel recently posed the problem of balladic performance in this way: ‘An important question, to which there is no easy answer is this: Who were the singers of the yuèfǔ ballads?’66 If the question is focussed more specifically to Who were the singers of Han yüeh-fu? the answer becomes even less easy. In the first place, a Han yüeh-fu is not amenable to a strict definition. If we accept Kuo Mao-ch’ien’s definition, then the impromptu songs sung by early Han emperors, aristocrats, and courtiers, such as Emperor Kao-tsu’s ‘A Great Wind Rises’ and Emperor Wu’s ‘Autumn Wind’, would appear to be the earliest descriptions of Han yüeh-fu being sung.67 Yet these songs are not generally regarded as Han yüeh-fu proper, which tend to be anonymous, popular pieces. Secondly, the only recorded description of a trained singer and musician performing at the early Han court, namely Li Yen-nien, does not involve a yüeh-fu, but a piece usually classified as a pentasyllabic ko-shih lyric (except for its octosyllabic penultimate line).68 Thirdly, singers in Han China ranked too low in such government organs as the Bureau of Music to merit an official biography. The exceptions, such as Lady Li and Wei Tzu-fu who rose from the humble and anonymous level of entertainer to become a concubine and consort of Emperor Wu, prove the general rule.69

Although the question posed by Frankel cannot be answered in definite terms, some clues may be gleaned from various historical, literary, and archaeological sources. J.-P. Diény has pointed a way 26forward by his brilliant analysis of such Han sources. If we cannot know who sang yüeh-fu songs and ballads composed for and arranged by the Han Bureau of Music, we can at least deduce to some extent the taste of the audience for whom the Bureau of Music functioned. Diény presents a list of Han personages—emperor, prince, aristocrat, court lady, and courtier—who performed a song with music and dance, alone or accompanied by a favourite, usually at court. He terms such (court) performances ‘I’improvisation pathétique’, characterizing them as personal statements, spontaneously composed, and sincere, simply worded expressions of emotion.70 Clearly, since the Bureau of Music was rejuvenated by early Han emperors, especially by Emperor Wu, its immediate patrons and arbiters of taste were the Han ruler and his favourites at court. They would have imposed their taste in music, song, and dance on performances at banquets and parties, if not necessarily at the more solemn state ceremonies. Since the personal style of musical expression among early Han emperors, aristocrats, and courtiers was simple, direct, and sincere, their taste would have influenced the style of singing and the content of the song at court entertainment.71

It is an ironic fact that although as a group entertainers were classless, the dividing line between entertainers and members of the royal family was very narrow in terms of social manner and musical taste. This is especially true of Emperor Wu. His concubine, Lady Li, and his later consort, Wei Tzu-fu, attracted him by their artistic skills far more than the nobler breed of ladies with whom the royal harem was stocked. Li Yen-nien was promoted from kennel lad to Master of Music at court because Emperor Wu enjoyed his musical flair. But in general, enter­tainers were considered as household slaves, property to be disposed of at will. For example, when Emperor Wu first saw Wei Tzu-fu perform he was so attracted to her that he paid his sister, who owned her, a thousand catties of gold. And later in the Former Han, when Emperor Hsüan responded to a political crisis arising from bad harvests in 70 BC he ordered a reduction in the number of butchers and of musicians in the Bureau of Music (lumping them together), among other measures.72

Han entertainers acquired their training through their own family, for generations members of the musical world. Such families tended to come from certain parts of China, such as Han-tan, capital of the old kingdom of Chao in the north and neighbouring Yen, and from Ho-chien. Such places were famous for their glamorous singers with their popular style of performance. By the late Han poets were acknowledging the presence of these regional performers in the Han capital cities of Ch’angan and Loyang. In the course of the Han Dynasty, entertainment became less of a royal privilege than a right of the plutocracy. By the end of the period even eunuchs, such as Ma Jung, held their own musical extravaganzas in luxurious mansions.73 27

Diény has shown that the anonymity of singers as a group was further accentuated by their role in the musical ensemble during a performance. Drawing on archaeological evidence from Han funerary mural art, he presents a revealing picture of a Han singer at a performance: he or she is no more than one of an ensemble of instrumentalists, no one individual stands out from the group. The singer is usually close to a zither-player. Adding to the concept of a performing team is the fact that the singer and instrumentalists were no more important than the dancers. Furthermore, the same archaeological evidence shows that the role of song and dance was no greater than other performances by jugglers, acrobats, buffoons, wrestlers, and so forth.74 The performance was a series of variety acts, similar to the early European commedia dell’ arte, and not far removed in concept from modern Peking Opera. It may be that the style of variety performance popular at Han court enter­tainment, and entertainment in wealthy mansions, explains the extreme­ly heterogeneous nature of the repertoire I have translated. From this repertoire one might select songs which are maudlin, lyrically refined, humorous, dramatic, meditative, crude, and flippant, to create a programme suitable as a medley to entertain a party of celebrating people, whether in the Han era or two thousand years later in the West.

Criteria of Selection

My criteria for selecting songs and ballads from the large repertoire of the Han are as follows. I have chosen those titles which constitute the generally agreed corpus of anonymous Han pieces, and those titles especially which have been most imitated by later generations of poets.75 I have added a number of pieces in order to form a collection representa­tive of the traditional categories of Kuo and his predecessors in which Han songs appear. I have also included some titles so as to demonstrate the thematic diversity of the Han repertoire. The resulting total of my selection is 77 songs and ballads, of which seven are different versions of some texts. I have excluded titles from some of Kuo’s categories such as Ch’ing-Shang on the grounds that, in his view, Ch’ing-Shang pieces meant certain post-Han popular songs, and that some of his categories do not include Han material. I have also excluded some Han hymns and some Han Songs for the Nao-bell, besides some other anonymous popular Han material, in order to provide a representative selection and to keep this introductory book within reasonable limits of length and scope.

I have also decided to exclude some famous ballads, such as ‘A Peacock Southeast Flew’, because they are now thought to date from the Southern Dynasties, not from the Han as was traditionally believed. Absent too are some Han pieces attributed to certain authors, such as ‘The Imperial Guards Officer’ attributed to Hsin Yen-nien, because my selection 28represents the anonymous Han repertoire of popular songs and ballads. Conversely, I have included some pieces attributed to specific authors in the literary tradition because, on balance, they are to be judged as anonymous Han pieces. For example, ‘Watering Horses at a Long Wall Hole’, of which traditional critics and anthologists have been equally divided concerning its attribution—either to Ts’ai Yung or to an anonymous Han maker. Those familiar pieces I have excluded, such as those mentioned above, may be found in New Songs from a Jade Terrace.76

The texts in my selection are not based primarily on Kuo Mao-ch’ien’s Anthology. I have tried to cite the texts from their earliest sources. My finding is that 54 of the 77 texts appear in one Han source and six post-Han and pre-T’ang sources. This means that two-thirds of the texts are traceable back to pre-T’ang sources. Among them, the primary source is Shen Yüeh’s ‘Treatise on Music’, with 26 pieces. Next, numerically speaking, are Pan Ku’s treatises, with 12 pieces. Late Southern Dynasties texts overlap to some extent, but between them, that is, Anthology of Literature, New Songs from a Jade Terrace, and Compendium of Literature, 10 pieces are derived. T’ang sources produce few new texts, but a wealth of fragmentary material. Li Fang of the Northern Sung provides a useful number of fragments, corroborating earlier material. From Kuo’s Anthology of the late Northern Sung I have drawn 13 pieces not cited in earlier sources, plus three full texts of earlier fragments. Two late Ming sources provide five pieces either cited only in fragments or overlooked in earlier texts. These textual sources number 12 in all. The most important, in descending order, are: Shen Yüeh, Kuo, Pan Ku, Hsü Ling, and Hsiao T’ung and his collaborators. In view of the scattered nature of the material, which only acquired the status of respectable literature in the 5th century AD, the retrieval rate for pre-T’ang texts is very high. The full list of sources and pieces cited appears on pp. 209–10 below.

The principle on which I based my organization of this diverse material is thematic. It seems to me that the original classification system of Kuo Mao-ch’ien does not translate well into Western literature, especially as the musical notation has not survived. Moreover, pieces in a set like the ‘18 Han Songs for the Nao-bell’ do not cohere thematically or in terms of content, making them difficult to discuss without their original musical and social context. Although my thematic treatment is arbitrary and my arrangement one of personal choice, I feel that the pieces are well-suited to such a scheme. Naturally, since some songs contain a number of different themes, they might be placed in other thematic groups than mine. On the whole, however, the present arrangement reveals something of their rich diversity in theme, form, and content.


1 Arthur Waley, The Analects of Confucius (1938); The Book of Songs (1937); Burton Watson, The Complete Works of Chuang Tzu (1968); Records of the Grand Historian of China, 2 vols. (1961); Courtier and Commoner in Ancient China (1974); D. C. Lau, Mencius (1970); Lao Tzu, Tao Te Ching (1963); Angus C. Graham, Later Mohist Logic, Ethics and Science (1978); Wm. Theodore de Bary, Wing-tsit Chan, and Burton Watson, Sources of Chinese Tradition, 2 vols. (1960); Homer H. Dubs, The History of the Former Han Dynasty, 3 vols. (1938–55).

2 It would seem that many Chinese classical authors belonged to the lower ranks of the aristocracy, or, in the case of at least Mo Tzu, from the middle class. For a general study see Hsü Cho-yün, Ancient China in Transition: An Analysis of Social Mobility, 722–222 BC (1965).

3 Much of the material which appeared at exhibitions in the early 1970s is included in Historical Relics Unearthed in New China (1972).

4 Citing the felicitous phrase of Matthew Hodgart, ed., The Faber Book of Ballads (1965), p. 12.

5 Derk Bodde, China’s First Unifier, A Study of the Ch’in Dynasty as Seen in the Life of Li Ssu (1938), provides a detailed account of the Ch’in.

6 Michael Loewe, ‘The Grand Inquest—81 BC’‚ pp. 91–112 of Crisis and Conflict in Han China (1974), and Esson M. Gale, Discourses on Salt and Iron (1931).

7 Jean Lévi, Le grand empereur et ses automates (1985), presents a semi-fictional account of the first emperor, Ch’in Shih Huang-ti.

8 Hans H. Bielenstein, ‘The Restoration of the Han Dynasty’, 3 vols. (1954–67), provides a valuable study of the reign of Emperor Kuang-wu.

9 This rhapsody appears in Burton Watson, trans., Chinese Rhyme-prose (1971), pp. 38–51, ‘Sir Fantasy’. I have translated Shang-lin as Royal Forest in my texts. For an account of Emperor Wu’s expansionist policies see Yü Ying-shih, Trade and Expansion in Han China (1967).

10 Michael Loewe, ‘The Capital City of Ch’angan’ in Everyday Life in Early Imperial China (1968), pp. 128–36.

11 Id., ‘Life in the Cities’, pp. 137–51.

12 Id., ‛The Countryman and his Work’, pp. 163–79; Hsü Cho-yün, Han Agriculture, The Formation of Early Chinese Agrarian Economy (1980); Francesca Bray, Agriculture (1984).

13 Jean-Pierre Diény, Aux origines de la poésie classique en Chine (1968), p. 95; Loewe, Crisis, pp. 208–10; Anne Birrell, ‘Mythmaking and Yüeh-fu’ (1989), pp. 224–32.

14 Pan Ku, History of the Han (Han shu)‚ ch. 19, ‘The Table of Officials’, p. 731; Diény, op. cit., p. 86; Loewe, Crisis, pp. 200, 310, Table 6.

15 Ssu-ma Ch’ien, Historical Records (Shih chi)‚ ch. 24, ‘History of Music’, p. 1,177; the three emperors were Hui, Wen, and Ching. 176

16 Op. cit., ch. 22, p. 1,043: ‘In the second year of Emperor Hui the Filial [194 BC] he ordered the Director of the Bureau of Music, Hsia-hou K’uan, to arrange string and wind accompaniment for it [the set of hymns], and renamed it “Music to Set the World at Ease”.’ Diény, id., ‘La carrière du Yue-fou’‚ pp. 81–4, succinctly summarizes the statements concerning the founding of the Bureau in the historical texts.

17 Op. cit., p. 1,045: ‘At the time when Emperor Wu defined the ritual for suburban sacrifices, he sacrificed to T’ai-i at Kan-ch’üan … and to Empress Earth at Fen-yin …. Then he established the Bureau of Music.’

18 Ibid.

19 Op. cit., ch. 93.3725.

20 Loewe, Crisis, p. 196, n. 11, explains that even this date is changed in different sections of the History of the Han (Han shu), ch. 6 and 22. For fuller references to the ‘horse of heaven’, see p. 183, nn. 41–45 below.

21 James Robert Hightower, Topics in Chinese Literature (1950), p. 49; Burton Watson, Early Chinese Literature (1962), p. 289; Hans H. Frankel, ‘Yüeh-fu Poetry’(1974), p. 69; Sawaguchi Takeo, Gafu (1973), p. 35, argues that Li Yen-nien’s sister, Lady Li, is supposed to have originally entered Emperor Wu’s palace as an entertainer and then his harem in 121 BC. When Lady Li died in 120 BC, the Ch’i magician Shao­weng was commanded by the emperor to summon her soul. The magician lost favour, however, because his powers were unproven, and he was executed in 118 BC. Sawaguchi cites Pan Ku, History of the Han, ch. 97 and Historical Records, ch. 28 (see Watson, trans., Courtier, pp. 247–49, and Records, Vol. 2, pp. 41–2). Hellmut Wilhelm, ‘The Bureau of Music in Western Han’ (1978), p. 123. P’eng Li-t’ien, ‘A Discussion of Whether the Bureau of Music Arose in the Reign of the Han Emperor Wu’ (1937), p. 182.

22 Crisis, p. 196.

23 Gary Shelton Williams, ‘A Study of the Oral Nature of the Han Yüeh-fu (1973), p. 23.

24 Op. cit., p. 84.

25 Id., p. 25; Birrell, ‘Mythmaking,’ pp. 232–33.

26 Loewe, Crisis, p. 209, n. 59. For one set of figures for the Bureau’s personnel in 7 BC, see Pan Ku, id., ‘Treatise on Rites and Music’, pp. 1,073–74.

27 Shen Yüeh, History of the Southern Sung (Sung shu), ch. 100, p. 2,452. Although the traditional date for completion of his History is set at AD 488 (Shen, id., p. 2,466), Shen admitted in his presentation address to the throne that the ‘Treatises’ section of his History had yet to be finalized and added to his main text (id., p. 2,468), but the date for their final completion is not known, presumably some time between AD 488 and 512, the year of his death. Anthology of Literature (Wen hsüan), Hsiao T’ung (AD 501–31) et al., comp.; David R. Knechtges, trans., Wen xuan, or Selections of Refined Literature (1982), Vol. 1. p. 27, notes that yüeh-fu forms a sub-category of the shih lyric genre in the anthology (ch. 27–8); New Songs from a Jade Terrace (Yü-t’ai hsin-yung), Hsü Ling (AD 507–583), comp. c. AD 545; Anne Birrell, trans. (1982), the genre of yüeh-fu first appears there in Vol. 1, ‘Old Yüeh-fu Poems’. Vincent Yuchung Shih, trans., The Literary Mind and the Carving of Dragons (Wen-hsin tiao-lung) (1959, 1983), Chapter 7, ‘Musical Poetry (Yüeh-fu)’, pp. 77–87.

28 Wen hsüan, SPTK 28.37b, Wen-hsin tiao-lung, SPTK 2.5a; Wen hsüan, SPTK 45.26b; Kuo, ch. 58, p. 850, and ch. 84, p. 1,180; Birrell, id., pp. 233–35.

29 For example, Pan Ku, id., and Record of Ritual (Li chi), ch. 27, ‘Record of Music’, Séraphin Couvreur, trans., Li Ki, ou Mémoires sur les bienséances et les cérémonies‚ (1913) Vol. 1, pp. 45–114. See Walter Kaufman, Musical References in the Chinese Classics (1976). In Pan Ku’s ‘Treatise on Literature’, 2 song titles are listed under the tenth of the 28 headings, mostly regional, of 314 ko-shih sung poems: ‘Goosegate’ (ying-men) and ‘Lunghsi’, which may be the source of titles for the Han yüeh-fu pieces entitled 177‘The Ballad of the Prefect of Goosegate’ and ‘The Lungshi Ballad’ in Chapters One and Eleven above. (Han shu, c. 30, p. 1,754.)

30 Alan Bold, The Ballad (1979), p. 6; I have amended his phrase ‘hard covers’ to (book form). It is worth noting that in the main repository of yüeh-fu of the twelfth century, Anthology of Yüeh-fu Poetry, compiled by Kuo Mao-ch’ien in 100 chapters, the poets of Shen Yüeh’s era (the Liang) predominate among the total of poets from the Han to the end of the Southern Dynasties era—there are 73 poets and 546 yüeh-fu compositions. Moreover, of these Liang poets Shen Yüeh himself was the most prolific with 119 yüeh-fu pieces. Masuda Kiyohide, Research on the History of Yüeh-fu (1975), pp. 441–42, presents a breakdown of the 5,290 yüeh-fu pieces in Kuo.

31 These 16 ku-tz’u, plus one wrong attribution, and the ‘18 Han Songs for the Nao-bell’‚ 35 Han pieces in all, appear in his ‘Treatise on Music’, ch. 21–22 (see p. 206).

32 Masuda, op. cit., pp. 434–36, summarizes Kuo’s biographical data. Kuo Mao-ch’ien, comp., Yüeh-fu shih chi (Anthology of Yüeh-fu Poetry), 100 ch. (Chung-hua ed., 1979).

33 See pp. 207–08, Appendix II.

34 A precise definition of such terms as yin or hsing or tz’u has yet to be determined. For a discussion of the term hsing in many yüeh-fu pieces, see Shimizu Shigeru, ‘The Meaning of hsing’ (Ko no hongi, 1984), in which he suggests that the term in yüeh-fu pieces may have originated in the 3-tone music of ‘travelling bells’ (hsing chung) used by nobles in the Warring States era. It is interesting to note that no. 5 of Pan Ku’s 28 headings for ko-shih pieces reads: ‘Songs for Travel, Royal Tours of Inspection, and Excursions’ (Ch’u-hsing hsün-shou chi yu ko-shih), Han shu, ch. 30, p. 1,754.

35 Indicative of the confusion surrounding what constitutes a Han ballad is the controversy over the inclusion or non-inclusion of certain ku-shih, old poems, in the yüeh-fu repertoire. For example, ‘Uphill I Picked Sweet Herbs’, (known as the mi-wu poem), a shih, or ku-shih, and nos. 8 and 13 of the ‘Nineteen Old Poems’, also ku-shih, of which the last two poems appear in Kuo’s Anthology, ch. 74 and 61, ‘Song-texts for Miscellaneous Pieces’. Recognizing the value of Kuo’s classification system, but viewing the corpus from a literary standpoint, Hans H. Frankel makes distinctions between the oral and the literary traditions, the hymns and the ballads, the different regions, and different periods. He proposes a concise system of five yüeh-fu headings: 1) ritual hymns of the Han, 2) a special class of ritual hymns from the Southern Dynasties, 3) anonymous ballads of the Han, 4) anonymous ballads of the Southern Dynasties and Northern Dynasties, and 5) ballads in the yüeh-fu style by men of letters from the Han to the end of the T’ang; ‘Yüeh-fu Poetry’, pp. 71–2.

36 Op cit., p. 19.

37 Gordon Hall Gerould, The Ballad of Tradition (1932), p. 11.

38 For the four categories of Emperor Ming’s time see Chu Ch’ien-chih, ‘A Discussion of Yüeh-fu’‚ in A History of Chinese Music in Literature (1935), pp. 130–31. Wang Yün-hsi, A General Discussion of Yüeh-fu Poetry (1958), p. 65, points out that the last two categories represent popular music, in contrast to the serious, stately music of the first two. Shen Yüeh, op. cit., ch. 20, p. 565, gives Ts’ai Yung’s categories, without questioning the attribution. The terms Grand Yü Music, Yellow Gate, and nao-bell may be summarized briefly as follows. The Grand Yü Music (T’ai Yü Yüeh) was the name of a government office which was originally named T’ai Yüeh, Grand Music (its title was changed in AD 60). Bielenstein, The Bureaucracy of Han Times (1980), p. 164, n. 75, suggests that the title change was probably in response ‘to a prediction … [that] the Han dynasty would compose a music named Yü’. This office supervized the music for national festivals and important state banquets. The Yellow Gate (Huang-men) was an early Han office. In the reign of Emperor Wu and later it supervized court entertainment and major ceremonies, including various forms of musical entertainment. It seems that when the Bureau of Music was abolished in 7 BC, the 178Yellow Gate office assumed some functions. It enjoyed great prestige among the government bureaus in charge of music since it was directly associated with the emperor. The Yellow Gate’s Drumming and Blowing [Music] (Huang-men ku-ch’ui) was traditionally known as music accompanying banquets given by the Han emperor for his officials. The nao-bell was a hand-held, hollow, metal musical instrument, in a deep cup shape, emitting a clear musical sound when struck (it did not have a clapper). It was early on associated with military music.

39 Shen Yüeh, id., ch. 19, p.549 and ch. 21, p. 603, where he further defined hsiang-ho: ‘Hsiang-ho are old Han songs. Stringed instruments hsiang-ho [= play in concert] in turn, the one holding the rhythm-baton sings.’ Shen’s folk-song imitations appear in New Songs, Birrell, trans., pp. 136–43, 251, 275. In all he imitated 14 of the Han originals among his 119 yüeh-fu.

40 The term, Ch’ing-Shang, appeared in some ancient texts, sometimes signifying a government office. In the late Han and Wei (3rd century AD), it acquired the status of a sub-genre of popular song, signifying a style of clear (Ch’ing) singing in the Shang tone (Shang was one of the notes of the ancient pentatonic musical scale), which was associated with the season of autumn. After the Chinese court went into exile south of the Yangtse in AD 317, Ch’ing-Shang was influenced by anonymous love-songs of the southeast, known as ‘Wu Music’ (Wu sheng), and later by ‘Western Melodic Pieces’ (Hsi ch’ü) of the southwest, mainly the region of ancient Ch’u. The term Ch’ing-Shang became attached to the repertoire of popular Southern Dynasties love-songs, composed by anonymous folk-singers, and imitations of them by named court poets. This repertoire flourished in the 4th—6th centuries. It is important to distinguish between this usage of the term Ch’ing-Shang, and that understood by Shen Yüeh in his Treatise on Music’. His categorization of melodic pieces lists the term as a category after Hsiang-ho, or ‘Concerted’ pieces. Instead of taking Ch’ing-Shang as a category of songs and melodies dating from the Southern Dynasties, Shen includes Han pieces in it, relegating it to the same period as the Han ku-tz’u of the ‘Concerted’ category. Moreover, he assigns three modes to Ch’ing-Shang, whereas most medieval historians and later compilers separate the three modes and list them after ‘Concerted’ pieces. Although Shen was the earliest literary compiler of this song-text repertoire, it is impossible to determine whether his system of categorization is correct or misinformed.

41 Wu Ching, Yüeh-fu ku t’i yao-chieh, in Chin-tai pi-shu, ch. 34.

42 For a discussion of Wu Ching’s nine categories of yüeh-fu in comparison with Kuo’s twelve see Masuda, op.cit., pp. 511–31, 443–4.

43 Masuda, op. cit., pp. 445–6 briefly compares Tso’s and Kuo’s classification schemes. He points out, p. 14, that Tso traced the origin of yüeh-fu to the legendary Yellow Emperor. Feng Wei-no (d. AD 1572) compiled Notes on Ancient Poetry (Ku shih chi) in 156 chapters, a vast repository of traditional verse and songs. Shen Te-ch’ien (AD 1673–1769), Sources of Ancient Poetry (Ku shih yüan). Ting Fu-pao (1874–1952), comp., Complete Poetry of the Han, Three Kingdoms, Chin, and Northern and Southern Dynasties (1916). Lu Ch’in-li (1911–1973), comp., Poetry of the Ch’in, Han, Wei, Chin, and Northern and Southern Dynasties in 135 chapters (1982). The Four Sung Books edited by Li Fang et al. in the tenth century AD include the two valuable repositories of literature, Imperial Survey of the T’ai-p’ing Era (T’ai-p’ing yü-lan) and Prize Blooms from the Garden of Literature (Wen-yüan ying-hua), each in 1,000 chapters.

44 Masuda, op. cit., pp. 468–510, lists a number of fragments of old yüeh-fu preserved in Li Fang’s compilations and various pre-Sung collections.

45 See Nakatsuhama Wataru, Research on and Concordance to the Yüeh-fu shih chi, pp. 572–701.

46 Bold, id., p. 3. 179

47 Diény, op. cit.‚ pp. 5–8, discusses these texts from Discourses of the States (Kuo ) Record of Ritual, Historical Records, and History of the Han.

48 Diény, id., p. 6 (my translation from his French version), citing The Kung-yang Commentary (Kung-yang chuan)‚ SPPY 16.11 a, commentary, 15th year of Duke Hsüan.

49 Among the four sections of the Book of Poetry (Shih ching) the first, entitled Kuo feng, is believed to mean ‘Airs of the States’, or, since feng is a pun for wind, custom, or admonition, ‘Grievances of the States’. This section was traditionally believed to contain songs of the people. The poems in it are subdivided into regional categories, suggesting that they had indeed been collected from various parts of the Chou realm. Tradition also has it that the song titles classified under regional headings in Pan Ku, op. cit., ch. 30, pp. 1,753–55, ‘Treatise on Literature’ (I-wen chih)‚ are songs of the people collected from the different regions of the Han empire voicing their grievances. Also see Birrell, op. cit., pp. 232–33.

50 The episode appears in Ssu-ma Ch’ien’s official biography of Emperor Kao-tsu (Shih chi, ch. 8.389), Watson, trans., Records, Vol. 1, pp. 113–14. Diény, op. cit., Chapter 3, ‘Les conditions de l’essor de la póesie lyrique sous les Han’‚ pp. 45–6, includes this song in an interesting list of examples of spontaneous improvization in the early Han.

51 See Chapter Four, pp. 89–93.

52 Kuo’s method is to place these anonymous versions together. Sometimes he dates one version in the Han, the other in the Chin. But it is not always clear which version is the earlier, nor can a firm date be attached to them. In general, I have followed the hypothesis of J.-P. Diény that the long version is likely to be the earlier, the short the later; Diény, op.cit., p. 138, and Les dix-neuf poèmes anciens, 136–40.

53 ‘The Relation between Narrator and Characters in Yuèfu Ballads’ (1985), p. 107.

54 Gummere, op. cit., pp. 90–1. The device is noted by Gary S. Williams, op. cit., p. 96.

55 Id., pp. 90, 95, 117–34.

56 Williams, op.cit., p. 146, presents a figure of ‘… 85 percent half-lines occurring formulaically’ for the ballad ‘They Met’. The figures he presents in his statistical survey of a core group of 50 Han yüeh-fu are unreliable, for it is a smaller group than my own selection of 77 pieces, and it includes some pieces which are not anonymous, popular yüeh-fu, but shih or ku-shih. Moreover, his prime exhibit, ‘Two White Swans’, must be rejected because his claim for it that it ‘… does not appear to be a version of another song’ is untenable; in fact, there are two versions of the song, one appearing in Kuo’s Anthology, ch. 39, pp. 576–77, the other in New Songs, Vol. 1. Birrell, trans., p. 37.

57 See especially Milman Parry, ‘Studies in the Epic Technique of Oral Verse–making: 1, Homer and Homeric Style’ (1930), and Albert B. Lord, The Singer of Tales (1960).

58 Sir Walter Scott, Minstrelsy of the Scottish Border (1830, 1931), p. 505.

59 Bold, id., p. 50, citing Child’s category 81A. The earliest firm date for the poetic treatment of the yu-hsia theme in Chinese literature is Chang Heng’s (AD 78–139) descriptive passage in his rhapsody, ‘The Western Capital’, cited in James J. Y. Liu, The Chinese Knight-Errant (1967), p. 56.

60 Bronson, The Traditional Tunes of the Child Ballads, 4 vols. (1959–72), Vol. 1, Introduction, p. ix.

61 History of the Southern Sung (Sung shu)‚ ‘Treatise on Music’, ch. 21.

62 Id.

63 Kuo provides this information in various prefaces to his twelve categories and some sub-sections of them.

64 In Record of Things Ancient and Modern (Ku chin chu)‚ SPTK 2.2b, Ts’ui Pao (fl. AD 290–306) notes that a ‘Wu tsü piece is a native song of the people of Wu’. Lu Chi’s (AD 261–303) ‘Wu tsü Ballad’ contains the lines: ‘My audience, listen clearly!/ Listen 180to me sing a Wu tsü./ The Wu tsü first arose from Ch’ang Gate.’ (This was a gate in Wu’s capital city). Tso Ssu (? AD 250–?306) refers in his ‘Rhapsody on the Wu Capital’ to ‘the yen [songs] of Ching, dances of Ch’u,/ Wu yü [?glee songs] and Yüeh laments.’ (Ching was an old name for Ch’u. Wu, Ch’u, and Yüeh were states of ancient Chou.) These citations of Lu and Tso are from Anthology of Literature (Wen hsüan), SPTK 28.15b and 5.33b. The extant text of the anonymous Wu tsü hsing (A Wu Coda, A Ballad), appears in Kuo, ch. 64, p. 934, between Lu Chi’s piece and a sixth-century version.

65 In his essay ‘Assemblage and Segmentation in Yüeh-fu Song-texts’ (1947), Yü Kuan-ying suggests eight methods of composition, such as the combination of two songs to form one extended piece, the insertion of one or more existing songs into a composition, the insertion of a quatrain from another song’s closure to form a song’s opening, and the use of phrases from other pieces, such as the felicitation closure. Some of Yü’s eight methods overlap. I do not fully accept all of them, partly because it is impossible to date the songs in terms of earlier or later pieces, partly because some so-called ‘methods’ do not constitute a general practice of composition, but are unique techniques used by creative song-makers in one particular song. Yü, A Discussion of the Poetry of the Han, Wei, and Six Dynasties, pp. 26–38. The occurrence of pentasyllabic sections among passages of extreme irregularity in the Han yüeh-fu repertoire may provide important clues concerning the unsolved question of the rise of formal pentasyllabic verse (shih) which was a predominant verse form from the end of the Han to the T’ang.

66 ‘The Relation between Narrator and Characters in Yuèfǔ Ballads’, p. 126.

67 See n. 28, p. 176.

68 First recorded by Pan Ku, Han shu, ch. 97a, p. 3,951, trans. Watson, Courtier, p. 247. The octosyllabic line might also be metrically construed as two lines: 3. and 5-syllables.

69 For the official biography of Lady Li see Watson, id., pp. 247–51; for that of Wei Tzu-fu, see Watson, Records, Vol. 1, pp. 389–91. There were various names for entertainer: ch’ang-chia (singer), yüeh-chia (musician), or ou-che (popular song singer).

70 Aux origines, pp. 41–6. A fuller list of such court songs, or songs sung by eminent people in the Han, appears in Suzuki, Research, pp. 11–42, with texts and critical comments. Most of these texts were recorded with their social contexts in the Han histories. Sixteen of the most important examples of court members are listed by Suzuki, with a total of 25 songs: Hsiang Yü (1), Kao-tsu (2), Wu-ti (6), Chao-ti (2), Yu, Prince Yu of Chao (1), Tan, Prince Tz’u of Yen (1), Li Ling (1), Hsü, Prince Li of Kuang-ling (1), Ch’ü, Prince of Kuang-ch’üan (2), Lady Ch’i (1), Li Yen-nien (1), Ssu-ma Hsiang-ju (2), Tung-fang Shuo (1), Chao Fei-yen (1), The Wu-sun Princess (1), and Lady Hua-jung (1).

71 Diény, id., p.46. Royal patronage of music continued immediately after the Han with the Ts’ao family, rulers of the Wei Dynasty. Ts’ao Ts’ao is known to have been particularly fond of tan-ko‚ the unaccompanied singing of popular Han airs, and he, like his sons and grandson, was fascinated by the art of imitating Han yüeh-fu titles.

72 Diény, id., p. 94, quotes Emperor Hsüan’s edict; Watson, op. cit., p. 390, translates the anecdote about Wei Tzu-fu.

73 Diény, id., pp. 53, 52–4.

74 Id., pp. 67–74.

75 Shen Yüeh, op.cit., and Kuo, Anthology, form the main sources; in the former work anonymous Han songs or yüeh-fu are usually designated by the term ku-tz’u, old songs, in Kuo they are designated by the terms ku-tz’u, old words, or pen-tz’u, original words after their title.

76 Birrell, trans., Chapter 1. 181

Previous Chapter


Next Chapter

Chapter One: Hymns

Additional Information

MARC Record
Launched on MUSE
Open Access
Creative Commons
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.