Contexts of the Song Lyric in Sung Times:
Communication Technology Social Change Morality

Stuart H. Sargent

The song lyric embodies in its very form the experiencing of emotion, its prosody creating a morphology of feeling unprecedented in Chinese poetry (see the chapter by Stephen Owen in this volume). This characteristic prosody reminds us of the genre's musical origins and of the fact that it was once preeminently performance literature. Yet from the beginning the song lyric also existed in handwritten texts. Moreover, it soon passed into print and, far from being ephemeral, was seen by posterity as the most important Sung contribution to Chinese poetry.

The following essay addresses the significance of the various means by which lyrics were preserved in writing. I shall give special attention along the way to the printing of song lyrics in Chiang-nan West Circuit and the ideological affiliations and changing geographical distribution of the song lyric composers there. Finally, I shall touch upon possible implications of the parallel flourishing of song lyrics and printing in China up into the thirteenth century, with regard to both the genre's evolution and its acceptance as an art form open to morally serious persons.

Performance and Text: Inked and Engraved Inscriptions

The song lyric, though it may occasionally share themes and imagery with poems (shih ), differs from poetry in that its characteristic phrasing

suggests very strongly its origins in oral performance, a musical performance in which pauses, turns, and variations in rhythm could mimic the experiencing of emotions. Yet this genre was stored in writing from the beginning and composed in writing centuries after its music was lost. To understand the genre, we cannot ignore the ways song lyrics functioned in written, and later printed, texts and the circumstances under which they became part of the print culture of the Sung dynasty.

It is because song lyrics were often written down that we know in remarkable detail the range of songs performed during the T'ang and Five Dynasties: many of the earliest examples were preserved in the dry air of cave 17 at Tun-huang. Included in the song lyrics recorded in blank spaces on the various Tun-huang documents is performance poetry of all types, from refined compositions to the popular song lyrics that we might have expected to have died with their oral culture.

We do not know who recorded these lyrics or why they did so. While the vast majority of the Chinese must have been illiterate until well into modern times, in absolute terms the number of merchants, government clerks, monks, and professional entertainers whose occupations required some degree of literacy was also quite high. Tun-huang, remote from the centers of agrarian Chinese civilization but an important stopping place on the trade routes through Central Asia, was a locus of activity for all these categories of people, and it must be their tastes that are reflected in the songs preserved there, whether the song texts were written down at Tun-huang or imported from other parts of China.

It is sometimes said that song lyrics found on the margins or backs of various documents were simply texts used for writing practice. But many of the words in the lyrics must have been of no immediate practical use for the novice accountant, monk, or clerk. I suggest that these notations represent rather the simple storage of favorite songs as an aid to memorization or a backup to recollection: once the song lyric was fixed in the mind or successfully performed, the written text had discharged its function.

Such notations contrast with cases in which chirographic song lyric texts were treated with more reverence, evidently preserved and treasured for their own sake. P. 3994, a four-page booklet datable to the first half of the ninth century, is an extant example of such a text. It contains six song lyrics: four of these are clearly elite compositions (three are attributed elsewhere to known literati), and two are rough and unrefined. Marsha Wagner observes:

The neat calligraphy of booklet P. 3994 suggests that this handbook was not the exercise of a schoolboy. It was perhaps copied out by an educated

man as a memento of a visit, perhaps to a Chiang-nan entertainment center. Or it may have been intended as an art object for trade. Whatever its purpose, it includes a wide stylistic range, from an abstract and evocative elite poem to a concrete and flat popular verse, with several intermediate gradations between these two extremes.^[1]

One of the "Palace Lyrics" by Lady Androecia (Hua-jui fu-jen) of the Later Shu court may help explain how P. 3994 came into being. It tells of the text of a new song lyric being bestowed on courtiers by their ruler in booklet (pen ) form, then copied avidly in (commercial?) scriptoria (shu-chia ).^[2]

We do not know whether such a booklet would have been read silently or used simply as an aid to a performer's memory. But in format and in its relation to performance, P. 3994 may be usefully compared with certain little printed pamphlets of sixteenth-century Germany: these contained one to four sets of lyrics (usually hymns, sometimes political propaganda), and on their title pages were designated the old tunes to which the new lyrics were to be sung.^[3] In both the Chinese and the German cases, the music to which the lyrics were to be sung must have been so familiar that the owner of such a booklet could have sung them as he or she read them, either mentally or aloud. (Of course, in terms of content, the Chinese booklets and the German pamphlets are very different; while the motivation for copying the former would appear to have been aesthetic, the latter were clearly printed as propaganda. But this distinction may be too crude. Those who sang the German songs may have found it musically pleasurable to do so; and those who obtained copies of the songs bestowed on the court by the king of Later Shu surely would have treasured them for social reasons as well as aesthetic ones.)

Xylographic or Wood-block printing would seem to be the next step in song lyric preservation, but two other kinds of publication that were common in Sung China—inked inscriptions and stone engravings—raise several interesting issues.

[1] Marsha L. Wagner, The Lotus Boat: The Origins of Chinese Tz'u Poetry in T'ang Popular Culture (New York: Columbia University Press, 1984), pp. 46–47.

[2] Ch'üan T'ang shih (Peking: Chung-hua shu-chü, 1960), 798.8979; hereafter cited as CTS . Dr. Shih I-tui notes that this reveals the value which the elite and popular cultures alike placed on such texts. See his Tz'u yü yin-yüeh kuan-hsi yen-chiu (Peking: Chung-kuo she-hui k'o-hsüeh ch'u-pan she, 1989), p. 39.

[3] See Kyle C. Sessions, "Song Pamphlets: Media Changeover in Sixteenth-Century Publicization," in Print and Culture in the Renaissance , ed. Gerald P. Tyson and Sylvia S. Wagenheim (Newark: University of Delaware Press, 1986), pp. 110–19.

The inked inscription written on a wall or other fixed surface resembles a manuscript text insofar as it must be copied manually to reach a wide audience. But it differs in that this form of publication was commonly done by the song lyricist himself as a personal statement. For example, Su Shih once went on a spring evening excursion that included a stop at a wineshop. When he awoke the following morning next to a bridge, he took out brush and ink and inscribed a song lyric to the tiao , or tune, "Hsi chiang yüeh" on a pier of the bridge.^[4] In such a case, we may ask, who was the intended audience? Su Shih's composition utilizes an allusive wit that would not have been immediately accessible to the casual tourist, I think—if there were any "casual tourists" in Huang prefecture, where he was in exile at the time (1080–85). It is arguable that Su Shih wrote his new work on the pier because he lacked paper and needed to write either in the process of composition or as a memorandum of what he had orally composed (a crucial distinction, but one inaccessible to us nine centuries after the fact). But in addition, he surely knew his song lyric would be circulated by others, as were so many of his jottings. Even if the inscription had been made for his own purposes, it spoke to a larger audience he would never see, and he knew from experience that it would reach that audience in spite of the apparent inefficiency and transiency of his medium.

The process by which a song lyric inscription would be circulated is seen in the story of the lovely song lyric Hsieh I (d. 1113) left on the wall of a rest station in Apricot Village in the same prefecture. It is said that the keeper of the station got so tired of people asking to borrow writing brushes to copy it that he smeared it over with mud!^[5] Again, the lyricist must have known that his inscription would find a larger audience. In fact, Hsieh I's works were quite popular in his day, and we might suspect him of writing the inscription as a conscious means of feeding that popularity.

There are also stories of women who came to tragic ends in times of national crisis and left song lyric inscriptions clearly intended as testimonies for others to read. Abducted by soldiers, the women somehow

[4] He evidently made a copy for himself at some point, adding a preface to detail the circumstances of the lyric's composition. See T'ang Kuei-chang, ed., Ch'üan Sung tz'u (Hong Kong: Chung-hua shu-chü, 1977), 1:284–85; hereafter cited as CST . For helpful notes, see Wang Shui-chao, Su Shih hsüan-chi (Shanghai: Shang-hai ku-chi ch'u-pan-she, 1984), pp. 285–86.

[5] Hsüeh Li-jo, Sung tz'u t'ung-lun (Hong Kong: Chung-liu ch'u-pan-she, 1974), p. 155; "Chiang shen tzu," no. 2 (CST , 2:650) appears to be the song lyric, although Hsüeh identifies it as "Chiang ch'eng tzu."

escaped supervision long enough to pen their song lyrics on a wall and (usually) commit suicide.

Wei Chü-an (1268 chin-shih ) relates two such incidents, though the eventual fates of the song-lyricist subjects are not mentioned. One involves a very young woman whose father died in the Chin attack that toppled the Northern Sung dynasty and who was subsequently taken north as a prisoner. When the convoy reached Hsiung-chou (approximately a hundred kilometers west of modern Tientsin), she wrote the details of her plight on the wall of a way station there and added a short song lyric (forty-four syllables to the tune "Chien tzu mu-lan-hua").^[6] The other incident took place in 1277, in the last days of the Southern Sung. Passing troops kidnapped a woman. When they passed through Ch'ang-hsing, on the western side of Lake T'ai, she paused to inscribe a long and bitter lament in the form of "Ch'in-yüan ch'un" on the front of a wine storehouse, signing it "Liu of Yen-feng." Her work strikes me as a rather artless piece, but perhaps all the more pathetic for that very quality.^[7]

We may wonder whether anyone could have had the discipline or composure to write song lyrics under such difficult conditions, and suspect that some stories might have been made up to frame or explain song lyrics whose actual provenance was unknown. If the stories ring true, of course, the song lyrics themselves still might have been touched up or refined in the process of copying and recopying as they entered the written record. But Liu's song lyric was probably not, to judge by its style, refined in the course of its transmission. Her story also establishes that a woman of the late thirteenth century could indeed have known at least some song lyric meters quite well, for if this were not the case, Wei Chü-an's anecdote would have had little credibility. Liu must have been confident that others could make sense of her text, either because they, too, knew the music that determined its prosody, or because they were accustomed to reading song lyrics as purely written texts.^[8] Perhaps both are true.

[6] CST , 2:988; Mei-chien shih-hua (TSCC 2572), C.39.

[7] CST , 5:3420; Mei-chien shih-hua , C.39.

[8] Reading a lyric, it seems to me, would be somewhat analogous to reading texts in premodern Malay, whose orthography was based on Arabic script and did not represent many of the vowels. The reciter had to comprehend large chunks of text at once to remove the ambiguities in the orthography of the individual words. The fact that literature in the Malay world was largely an aural experience made this less problematic than it would have been in a print culture; the texts were full of the formulas and repetitions which met the expectations of the audience and also provided the large units of meaningneeded by the reciter. See Amin Sweeney, A Full Hearing: Orality and Literacy in the Malay World (Berkeley: University of California Press, 1987), pp. 85, 89–90. The absence of punctuation or line breaks in premodern Chinese orthography creates an ambiguity corresponding somewhat to the paucity of vowels in old Malay manuscripts. In the Chinese case, the individual morphemes are clearer, but the reader may hesitate in deciding the boundaries of words, phrases, and rhythmic units. Therefore, storage and transmission of lyrics in written form presupposes enough acquaintance on the part of the reader with the underlying music to make the meter intelligible.

The song lyric is functioning here in a stereotypical situation that probably reflects both real patterns of feminine resistance and paradigms of loyalty idealized by the survivors who did not martyr themselves. T'ao Tsung-i, who lived into the Ming dynasty, leaves among his jottings (concerning everything from Yüan dynasty bureaucracy to a castrated male dog's giving birth to puppies) a cluster of similar anecdotes about women who maintained their integrity after the fall of the Sung and invariably left behind poems or song lyrics as testimony. One such woman, captured in 1266 when Hang-chou fell, left a song lyric to the tiao "Man chiang hung" on a way station wall, lamenting the end of her life as a palace lady. (After resettling in the north, she became a Taoist nun.) Another, the wife of a man of Yüeh-chou, on the eastern shore of Lake Tung-t'ing, was taken east to Hang-chou. On the way she used her wits to avoid rape and her beauty to avoid execution; but when it became clear that her captor in the fallen capital was going to have his way with her, she asked for a few private moments to pray to her late husband, used the time allotted to write a song lyric to "Man t'ing fang" on a wall, and drowned herself in a pond. Her song lyric is a medium-length lament for the fallen dynasty. There are similar anecdotes involving other forms of literature.^[9]

These women captured T'ao's imagination. Their plights symbolized the helplessness of the whole society, and their refusal to yield to their captors was a model for male resistance. The symbolism is not perfect, to be sure: these women have the power to thwart the foe's invasion of their bodies by committing suicide; to prevent the invasion of one's homeland requires a collective sacrifice on the part of many individual defenders over an extended period of time. But resistance on either scale calls for resoluteness, and that is what the anecdotes are about. The song lyric, too, plays a crucial role in the stories: without these inscriptions to initiate the publication of the women's "fine reputations," their suicides would have been forgotten gestures.

[9] Ch'o-keng lu (TSCC 0218), 3.57–59. CST , 5:3420, has the lyric from the second anecdote mentioned above, with only one variant character, citing a Tung-yüan k'o-t'an .

Another kind of song lyric publication is the engraving of texts on stone. Here we are no longer dealing with the lyricist's own motives and his or her expectations as to who the audience might be, but with the motives of those who caused the engravings to be made. Perhaps, engraved in fine calligraphy and placed in scenic spots, such song lyrics served to enhance the landscape. But engravings might also have been executed for the purpose of making rubbings that could circulate in greater quantity and more efficiently than handwritten copies from a scriptorium.^[10]

For example, ten song lyrics by P'an Lang (d. 1009) on the various charms of Hang-chou and its West Lake were engraved on stone at a government office in Hsü-chou; later, a minor official employed in Hang-chou obtained the text and, thinking it appropriate to have the lyrics engraved at West Lake, did so in 1108.^[11] We would expect the text engraved at Hsü-chou, far from West Lake, to have been destined for distribution in the form of rubbings; the West Lake inscription might have been intended for either display or mass reproduction.

When lyrics are engraved for display, the question arises as to how intelligible they were to the nonspecialist public; it is likely that people from many walks of life would have seen the engraved text, like the inked inscription, whether they understood it or not. An intriguing variation in the rhyme schemes of the ten song lyrics by P'an Lang must have given pause to the casual reader. The first one rhymes abb , aacc , but in the other nine song lyrics the first line rhymes with no other line. Moreover, the second through seventh song lyrics establish an alternating pattern of three rhymes—xaa , bbcc —(nos. 2, 4, 6) and two rhymes—xaa , bbaa (nos. 3, 5, 7). The last three song lyrics revert to the three-rhyme scheme xaa , bbcc . Ideally, this cycle should have been performed, so the elements of repetition and surprise could have been highlighted by singer and musicians to give shape and pleasure to what would otherwise have been merely a challenging puzzle. But I doubt

[10] Almost all of Su Shih's letters were engraved on stone and circulated as rubbings. See Chu Ch'uan-yü, Sung-tai hsin-wen shih (Taipei: Shang-wu yin-shu-kuan, 1967), introduction.

[11] Jao Tsung-i, Tz'u-chi k'ao (Hong Kong: Hsiang-kang ta-hsüeh ch'u-pan-she, 1963), 2.34–35. The minor official was named Huang Ching; he became a chin-shih only in 1112; see Ch'ang Pi-te et al., eds., Sung-jen chuan-chi tzu-liao so-yin (Taipei: Ting-wen shu-chü, 1974–75), 4:2875. Individual lyrics from this cycle had attracted attention long before Huang Ching's time; see Chao Wan-li, ed., Chiao-chi Sung Chin Yüan jen tz'u (Peiping: Kuo-li chung-yang yen-chiu yüan, li-shih yü-yen yen-chiu-so, 1931), p. 2a. Only one lyric besides the decade on Hangchow is attributed to P'an Lang; see CST , 1:5–6, for the lot.

that many people in 1108 could sing them readily, for they are in an anomalous form of a tiao that is itself rare in the written record.^[12] It is likely, therefore, that those who caused P'an Lang's ten song lyrics to be inscribed at Hsü-chou and Hang-chou published them either for the specialist or for readers who would treasure them simply as mementos of West Lake (or of a song lyric performance heard there), regardless of their own ability to recreate them fully as musical or literary experiences.

Hu Tzu (1082–1143) records two conflicting stories about the discovery of an old inscription that I believe must have been made for the production of rubbings. Known to us as "Fish Sport in the Spring Waters" ("Yü yu ch'un-shui"),^[13] the song lyric is perhaps pre-Sung in origin. According to one account, it was discovered in Yüeh-chou (modern Shao-hsing, Chekiang) during the Cheng-ho period (1111–18), on the back of an old stele. Such a "recycled" stone would not have served for display purposes, but it would have provided a ready-made surface for engraving a text for rubbings. According to the other source, the song lyric was found on a stone uncovered near K'ai-feng in the course of excavations undertaken for flood control. It had no name and, in the words of the former account, "no p'u ," no designation for music. In both accounts, the song lyric was presented to the emperor, Hui-tsung, who gave it its title and had it set to music.^[14] Now, if this song lyric, whose theme is the separation of a couple during spring, was in fact engraved without the name of its tiao , this could mean that the song was so popular originally that no such information was necessary. But in the twelfth century, music being extremely ephemeral in the absence of sophisticated notational techniques, Hui-tsung had to command his experts to provide new music to make it performable again: this suggests that a song lyric still could not be properly enjoyed without being performed musically, although it could be argued that Hui-tsung would have

[12] Of all the other lyricists represented in CST and born before 1108, only four use tiao by the same name, but in every case the last line of each stanza has two fewer syllables, there is a single rhyme, and the second line of the second stanza never rhymes. Yen Shu (991–1055) wrote two (CST , 1:95); Ch'ao Pu-chih (1053–1101), one (1:555); Wang Cho, one (2:1035); and Ts'ao Hsün (1098–1162), five, four of which are unique in rhyming the fourth syllable—like the first lyric in P'an Lang's series (2:1229 and 1231).

[13] CST , 5:3651.

[14] See Hu Tzu's T'iao-hsi yü-yin ts'ung-hua—hou-chi (1167), in Pi-chi hsiao-shuo ta-kuan (Taipei: Hsin-hsing shu-chü, 1983), vol. 35, 39.325. One account cited by Hu, in the Fu-chai man-lu , quotes the lyric; it is 89 syllables in length. The other, in Ku-chin tz'u-hua (cf. the Chiao-chi Sung Chin Yüan jen tz'u ed., pp. 18–19), quotes the lyric in 89 syllables but states that it has 94 syllables.

demanded the court performance of this old text in any case in order to endow its recovery with a significance auspicious for his reign.

Despite what I have just said, it should be noted that certain tiao could be performed without complex musical accompaniment. Some examples are found among the song lyrics that frequently served the proselytical purposes of Buddhism. In form, these songs were evidently thought to be derived from fishermen's songs; their melodies are variously identified as "Fisherman's Pride" ("Yü-chia ao"), "Oar Thrust" ("Po cho-tzu"), or "Fisherman's Song" ("Yü ko-tzu"). The early Sung Shih-shih yao-lan attests to contemporary "Ch'an masters in the south" using "Fisherman" and "Oar-Thrusting" song lyrics for preaching; the author of the Shih-shih yao-lan , a Hang-chou monk named Tao-ch'eng, traces this practice back to the Indian beginnings of Buddhism, citing a musician in Rajagrha who made songs to inspire believers.^[15]

Several anecdotes suggest that such songs required little more than the human voice for performance. In one case, a monk named Tsung-kao (also known as Ta-hui, 1089–1163) goes boating with a friend named Li P'eng. Beating time with an oar, Tsung-kao sings "The Fisherman" ("Yü-fu"), then challenges Li to write hymns of praise for the various masters in the lineage of his first teacher, which he will then sing to the same tune. Li P'eng produces ten such song lyrics "in jest" before their outing is finished.^[16] This episode tells us that these songs were sung with no musical accompaniment other than a percussion beat; we know also that Ou-yang Hsiu (1007–72) used the tiao "Yü-chia ao" to write a set of song lyrics on the twelve months for performance with a small drum (although there is some confusion as to which of two such sets is properly attributable to him).^[17] Perhaps the beat was often marked even more simply, by clapping the hands or striking two sticks together. Wang An-shih (1021–86) used to go walking in the hills with the sometime monk Yü-ch'an, who sang for him the "Yü-chia ao" he composed;^[18] presumably, there was no elaborate musical accompani-

[15] T.2127.305a. See also Wu Tseng's Neng-kai chai man-lu (1157) (TSCC ed.), 2.33; in Pi-chi hsiao-shuo ta-kuan (Yangchow: Chiang-su Kuang-ling ku-chi k'o-yin-she, 1984), vol. 4; it appears at 1.20a. Wu appears to copy Tao-ch'eng, but mentions the Indian precedent last, rather than first.

[16] Hsiao-ying, Kan-shan yün-wo chi-t'an , quoted by Kanda Kiichiro^[*] , Nihon ni okeru Chugoku^[*] bungaku , vol. 1, Nihon tenshi shiwa (Tokyo: Nigensha, 1965), p. 40.

[17] CST , 1:136–38.

[18] Wei Ch'ing-chih (13th cent.), Shih-jen yü-hsieh , 2d ed. (Shanghai: Shang-hai ku-chi ch'u-pan-she, 1978), 2:404.

ment for these al fresco performances. (We do not know the content of Yü-ch'an's songs, but the pair of "Yü-chia ao" lyrics in Wang's own collected works describe simple country scenes and quiet thoughts such as might have been experienced on such outings.)^[19] The fact that every line in this tiao rhymes, with no change in rhyme throughout, creates an insistent, recurrent element that suggests a strong percussive beat.^[20]

To sum up what we know so far: chirographic preservation of song lyrics served as an aid to memorization, but handsome texts could be treasured as souvenirs; inscriptions that were written on walls or engraved on stone were more likely to achieve successful "publication" of song lyrics, and suggest the existence of a fairly sizable public that could decipher or sing metrically complex texts.

Printing and the Song Lyric: The Social Dimension

Printed books containing song lyrics often tell us something about the motivations of their publishers, but do not necessarily tell us any more about how the printed song lyrics were used by the owners of the books than do the forms of publication discussed so far. Sample data from Chiang-nan West Circuit, where eighteen localities are known to have published books during the Sung,^[21] will acquaint us with the types of projects that resulted in song lyrics' being stored in print as well as the changing social context of their composition and preservation.

Twenty-five song lyricists from Chiang-nan West Circuit are represented in the Ch'üan Sung tz'u by twenty or more song lyrics. Of these twenty-five lyricists, twelve were published in one or more series put out by commercial publishers outside the circuit around the beginning of the thirteenth century. These are Yen Shu (991–1055), Ou-yang Hsiu, Yen Chi-tao, Huang T'ing-chien, Hsieh I, Wang T'ing-kuei (1080–1172), Yang Wu-chiu (1097–1171), Yüan Ch'ü-hua (1145 chin-shih ), Ching T'ang (1138–1200), Shih Hsiao-yu (1166 chin-shih ), Chao Shih-

[19] Lin-ch'uan hsien-sheng ch'üan-chi (Hong Kong: Chung-hua shu-chü, 1971), 37.400.

[20] As Ch'en Pang-yen pointed out in his paper for this conference, however, the nature of a given song lyric depends as much on its theme and the style of the poet as on the tiao selected. It should be noted that "Yü-chia ao" is the tiao most frequently used (14 times) in the song lyrics of Yen Shu (991–1055), who is known for his soft and romantic style. See Wu Hsiung-ho, T'ang Sung tz'u t'ung-lun , 2d ed. (Shanghai: Shang-hai ku-chi ch'u-pan she, 1978), p. 134.

[21] See Chang Hsiu-min, "Nan Sung (1127–1279) k'o-shu ti-yü k'ao," T'u-shu kuan 1961, no. 3: 53, for a chart of all 173 localities with known publication activity in the Southern Sung.

hsia (1175 chin-shih ), and Yang Yen-cheng (1145–?). I shall discuss these commercial editions in more detail later.

Within the circuit, perhaps the earliest printing of a native son's works that included song lyrics was the 1196 publication of Ou-yang Hsiu's collected works in Lu-ling, the administrative seat of Chi prefecture. Chi was not only home to the major kilns of Sung and Yüan times, but also one of the richest rice-growing areas in the country. Already fifth in production of chin-shih in the first century of the Northern Sung, the prefecture supported a large literati population.^[22] In fact, the entire Kan River basin from Chi prefecture north to Nan-ch'ang was as prosperous in terms of its agricultural yield as it was in terms of its examination graduates.

Ou-yang Hsiu was not born in the prefecture, but his family was registered in Lu-ling, and it is natural that a major compilation of the works of such an illustrious "native" should be undertaken and printed there. As early as 1122, a version of his works had been published by the Envoy Storehouse (kung-shih k'u ) in Lu-ling. Established to serve the needs of visiting capital envoys, such units had opportunities to siphon off funds for publication projects as well as for private favors.^[23] The 1196 edition of Ou-yang's works, more complete, was reprinted at least twice soon afterwards in other parts of the circuit.^[24] There is no evidence that this 1196 project was either a commercial or government-sponsored endeavor; it was probably a private effort by the local elite. One Lo Mi was responsible for editing the portion devoted to Ou-yang's song lyrics.^[25]

Chou Pi-ta (1126–1204), who was directing the Ou-yang Hsiu project, at about the same time also had a hand in printing the works of Wang T'ing-kuei from An-fu, about fifty-five kilometers to the north-

[22] John W. Chaffee, The Thorny Gates of Learning in Sung China: A Social History of Examinations (Cambridge: Cambridge University Press, 1985), pp. 111, 149.

[23] See Li Chih-chung, "Sung-tai k'o-shu shu-lüeh," Wen-shih 14 (July 1982): 157. The title of the 1122 edition of Ou-yang's works is given there as Liu-i chü-shih chi , in fifty chüan with another fifty-chüan continuation. There is no mention of lyrics. Li also notes that an Envoy Storehouse at O-chou in Ching-hu North Circuit (near modern Wu-han) published the Hua-chien chi in 1187. On the establishment of the Envoy Storehouses at the beginning of the dynasty and the danger of storehouse funds' being misused, see Liu Ts'en (1087–1167), as quoted in Wang Ming-ch'ing, Hui-chu hou-lu (1194) (TSCC ed.), 1.208–9.

[24] Abe Ryuichi^[*] , "Tenri toshokan zo^[*] So^[*] Kin Gan hambon ko^[*] ," Biblia 75 (October 1980): 407. One of these later editions must be the 1198 version mentioned by Li Chih-chung, "Sung-tai k'o-shu," p. 159.

[25] Ronald C. Egan, The Literary Works of Ou-yang Hsiu (1007–72) (Cambridge: Cambridge University Press, 1984), pp. 2, 10.

west of Lu-ling but still in Chi prefecture. This edition apparently did not include Wang's song lyrics. (Most of his forty-two surviving lyrics are minor works celebrating plum blossoms, social outings, and the like.) The work was edited by a protégé and carried prefaces by such prominent locals as Hu Ch'üan (1102–80) and Yang Wan-li (1127–1206), both of whom, with Chou Pi-ta, were outspoken critics of the weak, despotic, and corrupt central government.

Wang T'ing-kuei was respected locally as a literary figure just as Ou-yang Hsiu was throughout the empire, and yet the publication of their literary works could be seen as a political act. If we look more closely at the prefaces to Wang T'ing-kuei's works and the identities of their authors, we find that Hu Ch'üan was one of a number of hostile opponents of Ch'in Kuei's policy of negotiation with the Chin state in 1138. It was because of a farewell poem Wang presented to Hu when the latter was on his way to exile—the result of petitioning for Ch'in Kuei's execution^[26] —that Wang T'ing-kuei himself was exiled to a remote post in Ching-hu North Circuit in 1148 (some sources say 1143 or 1149). In Ou-yang Hsiu's case, Lo Mi's decision to exclude song lyrics he felt unworthy of his image underscores the fact that the scholarly urge to preserve was at least matched by the urge to uphold certain moral ideals.^[27]

Lu-ling song lyricists Yang Yen-cheng (a cousin of Yang Wan-li) and Liu Kuo (1154–1206) were both associated with Hsin Ch'i-chi (1140–1207), and their styles show evidence of his influence. The works of Liu Kuo as edited by his younger brother included his song lyrics; they were printed in the mid-1230s, but the nature and the place of publication are obscure.

We do know that the song lyrics of Liu Hsien-lun (fl. late 12th cent.) were published in Lu-ling: a Chi prefecture edition is mentioned by Huang Sheng in the mid-thirteenth century.^[28] Among the thirty-one song lyrics ascribed to Liu in the Ch'üan Sung tz'u are song lyrics composed for birthdays and banquets, but there are also a few compositions voicing vaguely heroic frustrations, leading the modern scholar Hsüeh Li-jo to place Liu Hsien-lun, with Yang Yen-cheng, in a line of "indignant" song lyricists who shared the ethos of Hsin Ch'i-chi.^[29]

The publications by Chou Pi-ta that came out of Chi prefecture were presumably printed at private facilities. The expense need not have

[26] See Jao Tsung-i, Tz'u-chi k'ao , 4.130; and Teraji Jun's superb Nan So^[*] shoki seijishi kenkyu^[*] (Hiroshima: Keisuisha, 1988), pp. 172 and 471.

[27] See Egan, Ou-yang Hsiu , p. 194, for a summary of Lo Mi's decisions.

[28] Huang faults it for being incomplete; see Jao Tsung-i, Tz'u-chi k'ao , 4.170.

[29] Hsüeh Li-jo, Sung tz'u t'ung-lun , p. 235.

been great; xylography in China required far less capital investment and technological expertise than printing with movable type.^[30] Song lyrics were also printed at facilities one would normally suppose to have been publicly funded. We have already noted an 1122 Envoy Storehouse edition of Ou-yang Hsiu's works. Likewise, schools affiliated with various local administrative units could be mobilized to publish the works of an individual. It was in the school of his native Nan-ch'ang, the administrative seat of Hung prefecture in Chiang-nan West Circuit, that the works of Ching T'ang were printed in 1199, one year before his death.^[31]

Because xylography was relatively inexpensive, local government academies found publishing a lucrative way of raising funds.^[32] But the man who undertook the Ching T'ang project in Nan-ch'ang, Huang Ju-chia, may have had other motives: Huang considered himself a protégé of Ching, who at that time held high office in the Han T'o-chou administration. Huang's private publication in the same year of Lü Pen-chung's (1084–1145) works, possibly including song lyrics, is therefore somewhat puzzling at first glance.^[33] Lü (who was not a native of Chiang-nan West Circuit but had moved south to Hang-chou from Huai-nan West) took his intellectual direction from the major neo-Confucian thinker Yang Shih (1053–1135), which would seem to place him in the tradition of the "false learning" whose proponents were persecuted by the Han T'o-chou administration from 1195 through 1200. But Lü was also one of a number of scholars allied with Chao Ting (1085–1147), who, while he had influence at court from 1134 to 1138, had advocated a strong but cautious strategy against the Chin state.^[34] Ching T'ang held similar

[30] See Evelyn S. Rawski, "Economic and Social Foundations of Late Imperial Culture," in Popular Culture in Late Imperial China , ed. David Johnson, Andrew J. Nathan, and Evelyn S. Rawski (Berkeley and Los Angeles: University of California Press, 1985), pp. 17–22.

[31] Jao Tsung-i, Tz'u-chi k'ao , 4.165. The lyrics included were largely composed as responses to other people's lyrics or in connection with seasonal outings Ching took while serving in Szechuan. See CST , 3:1841.

[32] Ming-sun Poon, "Books and Printing in Sung China (960–1279)" (Ph. D. diss., University of Chicago, 1979), pp. 95–97.

[33] Li Chih-chung, "Sung-tai k'o-shu," p. 159, is my source for the date and publisher of this edition. The wai-chi in three chüan , which is mentioned by Li but not by Jao (Tz'u-chi k'ao , 3.95), would have been the place to look for lyrics, if they were included.

[34] Chao advocated concentration of military power, which was decentralized and somewhat uncontrollable at the time, under the emperor; he urged Kao-tsung to personally lead a punitive expedition against the puppet state of Ch'i, as long as it could be done without directly engaging Ch'i's Chin backers. See Teraji Jun, Nan So^[*] seijishi , pp. 111–36.

views in his generation, and Huang must have been confident in 1199 that his mentor and Lü Pen-chung shared like minds and, perhaps, moral fiber as members of administrations committed to aggressive but responsible foreign policies. At the time Huang undertook publication of their works, Han T'o-chou's reckless northern campaign was still several years in the future.

The son or protégé of a man had an understandable interest in preserving that man's works, but print technology preserves in surplus , ensuring that the works will be part of the acknowledged body of literature not only for the contemporary generation but for posterity as well. The power imparted by this multiplication could enhance ideological solidarity or regional pride, justifying investment of the resources of the local elite or their educational institutions in the publication of a native son's collected works, even if no blood relation or master-disciple relationship was involved. These motivations can be inferred in the publication of Wang T'ing-kuei, Ou-yang Hsiu, and possibly Liu Hsien-lun in Lu-ling, and of Ching T'ang in Nan-ch'ang.

A growing regionalism is suggested by changes in the career patterns and geographical distribution of song lyricists in Chiang-nan West Circuit. Table 1 is based on one by John Chaffee,^[35] with the addition of known song lyricists for the circuit who meet the criterion of a corpus of twenty or more extant song lyrics. It shows that the number of song lyricists rises dramatically along with the number of chin-shih in four of the five more populous core prefectures, but falls just as strikingly in Fu prefecture. One way to explain this anomaly would be to say that the disappearance of song lyricists from Lin-ch'uan was compensated for by the appearance of three Southern Sung song lyricists in Nan-feng, a mere hundred kilometers upriver in Chien-ch'ang military prefecture. The fact that the prose master Tseng Kung (1019–83), a native of Nan-feng, had founded a charitable estate in Lin-ch'uan suggests that the two towns could be considered part of the same subregion though they belonged to different prefectures. In this connection, the printing in Lin-ch'uan of the collected works of Ch'en Shih-tao (1053–1101), including his song lyrics, may be significant.^[36] Ch'en was not a native son; and I would speculate that this project may have been supported by

[35] Chaffee, Thorny Gates of Learning , Appendix 3, p. 197.

[36] Knowledge of this edition comes by way of the thirteenth-century book collector Ch'en Chen-sun, who says it was published by a Liu Hsia-wei. See Chan Chih, ed., Huang T'ing-chien ho Chiang-hsi shih-p'ai chüan (Peking: Chung-hua shu-chü, 1978), 2:507–8. Liu entered the bureaucracy under the yin privilege but advanced to prominent national offices.

Tseng Kung's charitable estate because Ch'en had considered himself a student of the prose master in the 1070s.^[37]

But we are dealing with more than a geographical shift. The four Lin-ch'uan song lyricists, like Huang T'ing-chien of Hung prefecture and Hui-hung of Yün, left Chiang-nan West and became widely known figures with few ties to their home regions. Their interests were probably more closely linked to their status as members of the bureaucracy than to the local economy. The three Nan-feng song lyricists, on the other hand, were not prominent outside their home region. Chao Ch'ang-ch'ing, an imperial agnate who left behind a considerable number of song lyrics, does not seem to have had a significant political career. Like several other lyricists from the circuit, he had some connection with Chang Hsiao-hsiang (1132–70), a native of Li-yang in Huainan East Circuit who played a distinguished role in the early part of the third Sung-Chin war (1161), which toppled the vestiges of the Ch'in Kuei regime. Chang was a serious song lyricist who measured himself against Su Shih.^[38]

Chao Ch'ung-po, another imperial descendant in Nan-feng, similarly did not figure importantly in politics, though it must be noted that he often spoke out against various abuses during his career.^[39] Liu Hsün was a Sung loyalist who composed poems on the martyrs and patriots of the fallen dynasty and served as instructor in a Confucian School in Fuchien^[40] during the Yüan dynasty. His thirty surviving song lyrics sing of eremitism and nostalgia.^[41]

[37] The estate was still supporting Tseng agnates in 1333, when a shrine hall was erected in Lin-ch'uan to further focus their sense of history and unity. See the commemorative essay by Yü Chi (1272–1348) in his Tao-yüan hsüeh-ku lu (SPTK ed.), 35.304a–b. See also Robert Hymes, Statesmen and Gentlemen: The Elite of Fu-chou, Chiang-hsi , in Northern and Southern Sung (Cambridge: Cambridge University Press, 1986), pp. 106–9.

[38] Jao Tsung-i, Tz'u-chi k'ao , 3.119–20. Jao wonders if the editor of Chao Ch'ang-ch'ing's lyrics, Liu Tse, has any connection with the Liu bookshop and publishing business in Changsha; I wonder whether he has any connection with the Liu Hsiao-wei who published Ch'en Shih-tao's works in nearby Lin-ch'uan. Jao dates the lyric connected with Chang Hsiao-hsiang (CST , 3:1785) to 1264. On Chang, see Teraji Jun, Nan So^[*] seijishi , pp. 426–32, 470–72; and Wu Hsiung-ho, T'ang Sung tz'u , pp. 247–48.

[39] See, for example, Chou Mi's Ch'i-tung yeh-yü , 2.83–84 (TSCC ed.), relating Chao's support for an attack on corruption when he was an assistant minister in the Chief Office of Imperial Clan Affairs. Only twenty of his lyrics survive.

[40] Yves Hervouet, ed., A Sung Bibliography (Hong Kong: The Chinese University Press, 1978), p. 445 (where his name is romanized as Hsüan); Ch'ang Pi-te, Sung-jen chuan-chi tzu-liao so-yin , 5:3907. Hymes, Statesmen , p. 127, lists him as one of several Southern Sung "local advocates" in Fu prefecture who made appeals to the court for specific local relief—a pattern not seen in Northern Sung.

[41] CST , 5:3331.

Other studies have suggested that the national government was losing importance through the Southern Sung and into the Yüan both as a promising arena in which to compete for power and position, and as an effective arbiter of local affairs.^[42] During the effective reign of Ch'in Kuei, that is, 1142–55, most levels of government were monopolized by people who had personal ties to Ch'in or to the emperor; others were shifted from post to post so rapidly that their offices ceased to have any effective function. At the county level, fully one-third of the magistrate positions in the empire were vacant when Ch'in Kuei died, revealing both a remarkable failure to organize the 2,741 examination graduates of his regime^[43] into a working bureaucracy and a strong local sentiment against participation in or cooperation with the government. Even in the 122 counties of the Liang-che circuits and Chiang-nan East, where the urgent need to establish Southern Sung sovereignty insured that most county-level posts were filled, almost no appointees can be demonstrated to have belonged to Ch'in Kuei's faction, and many of those who would become influential in the dismantling of his administration after 1155 rose from those ranks.^[44]

This is not to say that the elite of Chiang-nan West Circuit cut themselves off from national politics; they did not. But the fact that the song lyricists cluster increasingly along the primary commercial routes through Nan-ch'ang, Feng-ch'eng, Lin-chiang military prefecture, Hsinkan, and Lu-ling suggests that their fortunes were tied more to the local economies than to imperial largesse. This change is apparent when one looks at the distribution of song lyricists within each prefecture, summarized in table 1.

In Hung prefecture, Huang T'ing-chien was from the northwest hinterland, but as soon as we enter the Southern Sung, we are confined to the Kan River. Yüan Ch'ü-hua's native Feng-hsin is upstream from Nan-ch'ang. Like Chao Ch'ang-ch'ing of Nan-feng, Yüan Ch'ü-hua is the author of a song lyric that can be connected with Chang Hsiao-hsiang: Chang, a noted calligrapher as well as a strategist, had so liked Yüan's song lyric on a terrace in Ch'ang-sha that he wrote it out in his own hand.^[45] This and similar song lyrics by Yüan are said to display

[42] See Robert P. Hymes, "Marriage, Descent Groups, and the Localist Strategy in Sung and Yüan Fu-chou," in Kinship Organization in Late Imperial China , 1000–1940 , ed. Patricia Buckley Ebrey and James L. Watson (Berkeley and Los Angeles: University of California Press, 1986), pp. 95–136; and Hymes, Statesmen , passim .

[43] Teraji Jun, Nan So^[*]seijishi , p. 506 n. 132.

[44] Ibid., pp. 323–91.

[45] Jao Tsung-i, Tz'u-chi k'ao , 4.137. Jao ascribes this to sometime between 1165 and 1174, when Chang was administrator in Changsha.

the heroic qualities of Hsin Ch'i-chi's tz'u .^[46] For Nan-ch'ang itself, we have Ching T'ang, exceptional as a prominent figure abroad as well as locally. Chao Shan-kua was an imperial agnate registered in Nan-ch'ang.^[47] Among his song lyrics are one following the rhymes of a lyric Hsin Ch'i-chi wrote in 1179 at O prefecture in Ching-hu North Circuit—Chao was administrator there in that year—and two following the rhymes of a lyric Hsin wrote in 1181 at a new home in Hsin-chou, 150 kilometers east of Nan-ch'ang in Chiang-nan East Circuit.^[48] Shih Hsiao-yu of Nan-ch'ang is known more for his song lyrics than anything else.^[49] His admiration for Chang Hsiao-hsiang was expressed in a song lyric written sometime between the time Chang was made a drafter in the Secretariat and his death in 1170.

Moving south up the river, we find that the two song lyricists from Lin-chiang military prefecture are both men of the Southern Sung and both from locales on the Kan River itself. Yang Wu-chiu, who refused to serve under the first Southern Sung emperor, exchanged many song lyrics with Hsiang Tzu-yin, a member of a prominent old K'ai-feng family who fled to Lin-chiang and invested large sums in a school and a charitable estate for Hsiang's family. (In 1138, Hsiang played an important role in the Chao Ting government as an advocate of using rewards and punishments to bring order to the Sung armies, but he also contributed to a split in his faction by siding with Ch'in Kuei and against his own friends in advocating a peace treaty with the Chin.)^[50] Chao Shih-hsia, an imperial agnate, lived upriver in Hsin-kan, eight kilometers south of Lin-chiang, although he was born before the fall of the Northern Sung and identifies himself in a postface (1187) to the Tung-ching meng-hua lu as a native of K'ai-feng.^[51] Much of his career was spent in Chiang-nan West Circuit (1172–74, 1179–86) and Ching-hu South (1167, 1188–89, 1197). His literary and political affiliations are unclear; his song lyrics are occasional in theme.

In Chi prefecture, all song lyricists after Wang T'ing-kuei are from Lu-ling on the Kan River. We have already mentioned Yang Yen-cheng and Liu Kuo, both affiliated with Hsin Ch'i-chi, and Liu Hsien-lun, who was also vaguely in the same camp. After a brief hiatus, another

[46] Hsüeh Li-jo, Sung tz'u t'ung-lun , pp. 232–33.

[47] CST , 3:1980.

[48] See CST , 3:1983 and 1981; and Hsin Ch'i-chi, Chia-hsüan tz'u pien-nien chien-chu (Taipei: Chung-hua shu-chü, 1970), 1.55 and 1.76.

[49] CST , 3:2031.

[50] Teraji Jun, Nan So^[*]seijishi , pp. 121, 143, 156–57.

[51] Jao Tsung-i, Tz'u-chi k'ao , 4.162.

generation of Lu-ling song lyricists appears in the middle of the thirteenth century, taking us beyond the Sung. Liu Ch'en-weng (1232–97), P'eng Yüan-sun, Chao Wen (1239–1315), and Liu Chiang-sun (1257–?) are all Lu-ling song lyricists who survived into the Yüan. I have no information on the publication of their song lyrics as independent collections, but Liu Ch'en-weng, fearless critic of Chia Ssu-tao, and Chao Wen, protégé of Wen T'ien-hsiang, both possess the Lu-ling spirit of moral integrity. Liu Ch'en-weng and Liu Chiang-sun were both heads of academies (shu-yüan ).

Unlike the Northern Sung song lyricists from the peripheral areas of the circuit, the Kan basin lyricists are generally not themselves pivotal figures, either in literature or in administration. They appear generally to have avoided association with the more venal regimes of the age; instead, we often find them using song lyrics to establish or commemorate relationships with such figures as Hsin Ch'i-chi and Chang Hsiao-hsiang.

Printing and the Song Lyric: The Commercial Dimension

In Ch'ang-sha in Ching-hu South Circuit (modern Hunan), a remarkable project was undertaken by a bookstore in the first decade of the thirteenth century to publish the song lyrics of "a hundred poets" in series. Ninety-seven poets are known to have been covered, in 128 chüan . Nearly half of them have fewer than twenty song lyrics extant today; the major criticism leveled at this project was that the publisher became less and less selective in the quality and importance of the poets he included as he filled out the series.^[52]

Interestingly, only two song lyricists with twenty or more extant lyrics were native to the prefecture of which Ch'ang-sha was the administrative seat, and neither of them native to Ch'ang-sha itself: Wang Ining, who relieved the siege of T'ai-yüan in 1126 and held various posts outside the circuit into the 1140s,^[53] and Liao Hsing-chih (1137–89), whose career never took him outside the region.^[54] To be sure, a number of important song lyricists, many of whom were also important political figures, passed through Ch'ang-sha's prefecture: Huang T'ing-chien, in 1104; Li Kang (1083–1140), in 1126 and 1132; Chao Shih-hsia, in 1167

[52] See Wu Hsiung-ho, T'ang Sung tz'u , pp. 320–25, for the dating and contents of the series.

[53] CST , 2:1062.

[54] CST , 3:1834.

and 1197; Hsin Ch'i-chi in 1180; and Liu Sheng-chi in 1181. Nevertheless, in a region that was home to so few significant song lyricists, the bookstore must have pinned its hopes of turning a profit on a general audience of officials and merchants who would purchase song lyrics in print.

A second multivolume commercial series was the Ch'in-ch'ü wai-pien series published in Fu-chien. The scope and contents of the original are unknown; at present we are able to determine the names of only nine of the song lyricists published in the series. Starting with Ou-yang Hsiu and ending with Chen Te-hsiu (1178–1235), all are literary figures well known in their times, giving the impression that this publisher was more selective than the Ch'ang-sha bookseller.^[55]

Devoting one volume to each poet, these series were designed to appeal to the buyer who read song lyrics as literature or as the literary product of a certain admirable personality. Anthologies produced in the Southern Sung are similarly arranged by poet, indicating that they functioned in the same way. To be sure, the arrangement of a text is not an infallible indication of its intended use. The Hua-chien chi of 940 was originally compiled, according to its preface, to provide high-quality songs for performance ,^[56] and the title of the Tsun-ch'ien chi (In front of the winecups), an anthology of thirty-six Five Dynasties song lyricists compiled sometime in the second or third quarter of the eleventh century, signals a similar function—even though both anthologies are in fact arranged by poet, not by tiao or theme. Nevertheless, prefaces to some surviving printed editions of the Hua-chien chi , published in 1148, 1184, 1185, and around 1205, indicate that the anthology had come to serve the purposes of literary history: the preservation of the old and the skillful, or the recording of an age's best literary efforts at a time when other genres were supposedly in decline.^[57] Wu Ch'ang-shou is probably correct in his observation that most anthologies from Tseng Tsao's 1148 Yüeh-fu ya-tz'u on were compiled not for the sake of performance but for the sake of the texts themselves.^[58]

[55] See Wu Hsiung-ho, T'ang Sung tz'u , pp. 326–27. Wu identifies only eight titles in the series, neglecting to mention Chao Yen-tuan (1121–75); see Jao Tsung-i, Tz'u-chi k'ao , 4.136–37. Jao, on the other hand, does not include Chen Te-hsiu in his study, possibly because all but one of his lyrics (CST , 4:2423) are lost.

[56] See Kang-i Sun Chang, The Evolution of Chinese Tz'u Poetry: From Late T'ang to Northern Sung (Princeton: Princeton University Press, 1980), pp. 15–16.

[57] See the 1148 preface by Ch'ao Ch'ien-chih and the 1205 postface by Lu Yu, quoted in Wu Hsiung-ho, T'ang Sung tz'u , pp. 331–32.

[58] Quoted by Wu Hsiung-ho, T'ang Sung tz'u , pp. 328–29.

In Tseng Tsao's case, simple preservation of the materials collected by his family over the years was the expressed purpose.^[59] Besides the works of thirty-four mainstream song lyricists, he includes over a hundred acclaimed song lyrics of unknown authorship and suites of songs designed for concert performance with dance—a type of song lyric most Sung poets were neither trained nor required to produce, and therefore consigned to the margins of the canon.

The Effect of Print Culture on the Song Lyric

We have discussed motivations for preserving song lyrics in print: a strong interest in literary history seems to have been basic to these motivations, although this interest could have derived from aesthetic, antiquarian, or political and social concerns, depending on the individual case. It is important to keep these motivations in mind. But what, if any, influence did these motivations for preserving the genre in print have on the form or content of the song lyric itself? Attempting to answer this question involves us in broader issues crossing cultural and temporal boundaries.

Scholars (and others) have commonly divided humanity into the literate and the illiterate, with fundamental differences between the two in thought patterns, transmission of knowledge, and literary expression. The work of Ruth Finnegan and others has shown, however, that oral and written modes of communication take such a variety of forms and stand in so many relationships to each other in different societies that it is questionable whether this division is meaningful or even useful.^[60] Perhaps a transformation more fundamental than the move from orality to literacy in a given culture, then, is the leap to print. Print communication interacts with oral and written communication in complex and variable ways, but it stands in opposition to both chirographic and oral systems in that it permits the reproduction and storage of information in quantities far beyond the capacity of either.

We have seen that as the Sung progressed, more and more song lyrics passed into print in great volume. This was part of a wider cultural transformation. One of the first priorities in the use of the new technology of printing was the fixing and disseminating of vast seas of

[59] See his preface in Wu Hsiung-ho, T'ang Sung tz'u , pp. 338–39, or in the SPTK edition of Yüeh-fu ya-tz'u .

[60] Ruth Finnegan, Literacy and Orality: Studies in the Technology of Communication (Oxford: Basil Blackwell, 1988).

knowledge, surely because the standardization and preservation of information was one means of unifying a new empire, indoctrinating its bureaucrats, and defining the canons of powerful religions. The first Sung government printing project produced China's first printed code of laws, in 963. Eight years later, the first of six Sung printings of the Buddhist Tripitaka was undertaken. Classics, histories, medical treatises, and the like were also printed by the government. The state often rewarded officials for using the presses of local government units to print books (which the officials sometimes gave away as gifts or sold for private gain).

The unprecedented power of print could not be monopolized by the government, however. As early as the ninth century, print usurped one of the sacred functions of the court when privately printed family calendars went into circulation even before the government versions were approved.^[61] In the Northern Sung, the texts of court debates on foreign policy were leaked, printed, sold, and quickly found their way across borders and into the hands of the foreigners who were the subjects of those debates.^[62] The works of dissident officials became even more dangerous when they were picked up and disseminated by commercial publishers, Su Shih's works again being the prime example. If a text were proscribed, or had a limited market, the printing blocks could simply be stored away until conditions favored a renewed printing at some later date—a major advantage over movable type.

Most students who entered schools with hopes of embarking on official careers would encounter government-printed editions of the classics. Many of these students would never join the bureaucracy, and hundreds of private academy students would study with no intention of seeking government posts, but most of these literate men would participate in the economic life of the society at some point, whether as merchants or merely as managers of their patrimonies. They would have the means to purchase a variety of printed materials—including song lyrics.

These men would always be aware that a vast amount of knowledge and thought was safely fixed in print somewhere, and they knew it was often organized by chapter and verse for easy retrieval. The European experience tells us that printing was first harnessed to do what writing and memorization had always done badly and with great effort: to re-

[61] Tsien Tsuen-hsuin, Paper and Printing , part 1 of Joseph Needham, ed., Science and Civilisation in China , vol. 5, Chemistry and Chemical Technology (Cambridge: Cambridge University Press, 1985), p. 151.

[62] Poon, "Books and Printing," pp. 48–50, 55–63.

claim the knowledge of the ancients and support the oral, rhetorical tradition with encyclopedic compendia of formulas and tropes. Gradually, with knowledge stored in books, the function of writing changed from the retention of essential verities and the rehearsal of the rational to the exploration of the new, the unique, and the mysterious. Reports of voyages across the planet could be published and compared against the ancient texts; poets could safely scorn the commonplaces of rhetoric and pursue originality.^[63] In China, of course, print technology developed under vastly different linguistic, social, and material conditions; both the timing and the specific contours of the cultural and social results of the shift were bound to be different.^[64] But the Western evidence alerts us to the possibility that print might shape the literature that flourished in its environment.

It is striking how closely the rise and spread of print coincide with the development of the song lyric. To be sure, imported music, economic prosperity, relative domestic peace, and the large numbers of women supporting themselves as professional entertainers in the growing cities all helped to foster the new genre; the development of print culture cannot be given all the credit. Nevertheless, I would like to posit three theses for inquiry.

The first thesis is that when those members of a society who care about the preservation of knowledge are confident that it has indeed become securely preserved, their writings will place less emphasis on the shared tradition and more on individual feeling. Walter Ong has argued that this is what eventually brought about the "preoccupation with otherness, with what is different, remote, mysterious, inaccessible, exotic, even bizarre"—what we call Romanticism—after print had transformed Europe.^[65] The rise of the song lyric in the tenth and eleventh centuries certainly bespeaks an increasing interest on the part of the literate elite of China in exploring nuances of emotion through a new poetic form uniquely suited to evoking the process of feeling and inner reflection.

[63] See Walter J. Ong, Rhetoric, Romance, and Technology: Studies in the Interaction of Expression and Culture (Ithaca: Cornell University Press, 1971); and Elizabeth Eisenstein, The Printing Press as an Agent of Change: Communications and Cultural Transformations in Early-Modern Europe (Cambridge: Cambridge University Press, 1979).

[64] In China for example, the invention of inexpensive paper preceded the invention of printing by centuries. The availability of paper quickly encouraged scholarly breadth, although more complete organization and preservation of knowledge awaited the arrival of printing. See Shimizu Shigeru, "Kami no hatsumei to Gokan no gakufu^[*] , Toho^[*] gaku 79 (Jan. 1990): 1–13.

[65] Ong, Rhetoric, Romance, and Technology , p. 255.

It is clear that print did have an effect on the way Sung people read and wrote. The unprecedented flood of writing in Sung times, often in the form of collections of trivia, appears to reflect both the lure of publication in print and the new, casual function writing could assume when freed from weightier tasks.^[66] It is significant that "books which classify, itemize, or summarize so as to facilitate easy reference" were a major part of the book trade.^[67] The availability of such reference works, and indeed the abundance of books in general, clearly facilitated scholarship, but it was also said to devalue the individual text and encourage "lazy" perusal.^[68] In a manuscript culture, "lazy" perusal is a threat to society's store of knowledge and an insult to those who try to keep it intact. If print opened the gates of knowledge to less scholarly readers, this does not mean, however, that serious readers disappeared. Examination candidates expended great energy mastering canonical texts, and scholars sometimes devoted a good portion of their lives to a given classic. Yet print enabled these tasks to be supported by, or even to center on, the collocation of considerable amounts of knowledge—a task very different from the fundamental ones of preservation and transmission in a manuscript culture.

More study is needed on how print culture encourages new kinds of scholarship and aesthetic expression: the rise of the song lyric may not be comparable to Western Romanticism as a quest for the "exotic" and "bizarre," but perhaps the song lyric's adoption by the elite and scholarly class does signal a new attitude toward the function of literature in a new medium.

My second thesis is that the popularity of song lyrics in Sung times is an effect of printing, one closely related to the flourishing of prose writing at the same time: because both forms of literature are more difficult to memorize than pentasyllabic or heptasyllabic poems, they depend on print to enable them to compete with shih poetry in terms of dissemination.

It has been observed that in T'ang times the literature that circulated most widely was shih poetry because it was most easily memorized and orally disseminated.^[69] If song lyrics are a genre of performance literature, that does not necessarily mean they were easily memorized or disseminated: the musical talent their performance required limited their

[66] Poon, "Books and Printing," pp. 68–70.

[67] Ibid., p. 102, quoting Yüeh K'o (1183–1240), K'uei t'an lu (ca. 1214) (TSCC ed.), p. 78.

[68] Poon, "Books and Printing," pp. 79–80.

[69] Ibid., pp. 68–70.

reproduction as oral texts to specialists, though their audiences might have been extensive. Without print, the genre would have been more ephemeral and less likely to have developed in the direction of longer and more literary compositions.

This brings us to the third thesis: that print encouraged the development of increasingly refined, linguistically complex song lyrics in the thirteenth century, particularly as seen in the works of Chiang K'uei (ca. 1155–ca. 1221) and Wu Wen-ying (ca. 1200–ca. 1260), different as these two leading song lyricists are.

Through the late eleventh century, simple distinctions of level and deflected tones were generally sufficient to make a song lyric's sound structure fit the music, though there were places in some tiao where only more precisely designated tones would serve. Chou Pang-yen was the first to rigorously govern the tones of his syllables according to the music, but he still allowed a degree of variation. During his time, performance of the song lyric was usually accompanied by stringed instruments, which apparently permitted a looser fit between the music and the singer's voice. In the Southern Sung, wind instruments came to dominate, encouraging finer distinctions in the quality of each sung syllable. Perhaps it was the greater musical subtlety required of them that made it possible for Southern Sung experts to recognize Chou's achievements, if belatedly (Chou's contemporary song lyricists had scarcely seemed aware of his existence). Song lyricists such as Fang Ch'ien-li, Yang Tse-min, and Ch'en Yün-p'ing admired Chou so much that they wrote compositions following the rhymes of most of his song lyrics.^[70] That much was no great novelty, but they also followed the tonal patterns of his song lyrics in each syllable. Wu Wen-ying did the same in the over sixty tiao he took from Chou.

This obsession with perfecting the sound pattern of the words of a song is in one sense a sign of the lyric's separation from living musical practice, for it was a pattern abstracted from the words of a poet long dead that was directing the music, not the evolving music that was controlling the production of new texts. It was a remarkable event in the thirteenth century to find someone who could actually sing Chou Pang-yen's song lyrics. Even a new song by Wu Wen-ying himself, without a broad-based musical culture to sustain it, was likely to have been unsingable a mere decade or two after it was written.^[71]

[70] CST , 4:2488–504, 4:2999–3016, and 5:3113–34, respectively.

[71] See the preface by Chang Yen (1248–1320) to his "Hsi tzu chiang man," CST , 5:3475.

In the same age, however, the older musician–song lyricist Chiang K'uei was using a flute to determine the sound quality of each syllable of a song and writing his own music (sometimes before, sometimes after he wrote the words);^[72] the relationship between music and song lyric in such a case was tighter than before. But like composition by model rather than by musical sense, this tight relationship is also the product of an elite art. Great specialization is required to create, perform, and appreciate song lyrics under such conditions. Chiang K'uei's song lyrics neither came from nor spread among popular entertainers. This stands in striking contrast to the earlier case of Liu Yung, who adopted and revised so much non-elite music:^[73] the variability of Liu's tone patterns indicates that it was in performance, not in abstract planning, that word and note were harmonized.

The song lyric was not yet a fossil, of course. We have noted stories of women abducted by hostile troops using their final moments to compose song lyrics. The tiao they employed must have been in common use, even though the details of performance would have been different from earlier decades. Both "Ch'in-yüan ch'un" and "Man t'ing fang," two of the tiao used by thirteenth-century women in the stories recounted above, are examples of patterns that remained relatively loosely defined by level/deflected tone distinctions alone.^[74]

But those elite Southern Sung song lyricists who, as we have seen, structured their tonal patterns after those of Chou Pang-yen were clearly estranged from the song lyric as oral performance. Chou himself, and Chiang K'uei even more, both consummate musical minds, would appear to have reversed that estrangement, but in fact their verbal patterns, so tightly bound to their music, have all the fixity of print . In performance they must have been exquisite, but they had been worked out beforehand through a sophisticated matching of word and music that reveals an established confidence in the authority and stability of the written or printed text.

If this seems paradoxical, we should remember that it was by writing (and rewriting) his text that Isocrates, the first Greek to write for (oral) publication by others, perfected the sounds and rhythms of an oral style.^[75] Similarly, we may say that the innovative and dense diction of

[72] On the above several points, see Wu Hsiung-ho, T'ang Sung tz'u , pp. 69, 70, 75, 139, 219, and 222; and Shih I-tui, Tz'u yü yin-yüeh , pp. 121–26.

[73] Shih I-tui, Tz'u yü yin-yüeh , p. 125.

[74] Wu Hsiung-ho, T'ang Sung tz'u , p. 69.

[75] Tony M. Lentz, Orality and Literacy in Hellenic Greece (Carbondale and Edwardsville: Southern Illinois University Press, 1989), pp. 122–35.

Wu Wen-ying was possible in the song lyric only after the genre had become as much a written as an oral one, if not more.

Although there are in some oral cultures literary compositions of great verbal richness and subtlety,^[76] the predictability and schematic patterning that generally support aural comprehension and oral composition become liabilities as members of a culture begin to compose in writing for a reading audience. We have noted that after the accumulation of knowledge in print had undermined the rationale for residual orality in Europe, the "commonplace" in writing became not a vital structuring device but a blemish. In the Malay world, study of the language of illiterate and literate storytellers (in our age of print) reveals that the latter give more weight to individual word choice and seek, almost unconsciously, variations in phrasing.^[77] Thus, I would suggest that there is a connection between Su Shih's innovative production of highly "literate" song lyrics and his awareness that his works were being circulated as printed texts.^[78] We should recognize, too, Wu Wen-ying's "poetics of density," which sharply limited the appreciation of his lyrics in performance,^[79] as the product of a mature print culture.

Print and Moral Distancing: A Cautious Proposal

When I began to research the contexts of the song lyric in Sung times, I wanted to understand two paradoxes: that the song lyric should have been a significant mode of expression for certain morally serious men; and that a genre of performance literature should circulate in print. I now think there is a relationship between these two phenomena. For reasons of space, I have had to reserve an extensive discussion of the writing of song lyrics by Buddhist monks for publication elsewhere. Let me simply state that some monks did write song lyrics in Sung times, as did generals and statesmen, despite the genre's association with love and the more personal emotions. Might the new print culture have fostered the consciousness that made it possible for such people to use a

[76] Finnegan, Literacy and Orality , pp. 69–77.

[77] Sweeney, Full Hearing , pp. 241–66.

[78] When Su Shih was tried in late 1079 on the charge of slandering the emperor, his accusers noted that his "satirical writings" had been printed and sold in the marketplace, and they produced four volumes of his printed works. See Murakami Tetsumi, "So Toba^[*] shokan no denrai to Tobashu^[*] shohon no keifu ni tsuite," Chugoku^[*] bungaku ho^[*] 27 (1977): 51–87.

[79] Grace S. Fong, Wu Wenying and the Art of Southern Song Ci Poetry (Princeton: Princeton University Press, 1987), esp. pp. 63–64.

genre that simulated the experiencing of emotion, but to use it without seeming to be trapped by emotion?

Such an attitude of detachment from one's morally suspect actions, "lodging" emotions and thoughts without being "detained" by them, was a stated philosophical ideal in the times;^[80] it might also have been a personality trait of some individuals. But detachment or distancing is sometimes said to be characteristic of a literate, analytical culture. Although the universality of such a rule is questionable, we must ask whether it wasn't the intensification of just such a culture through the flourishing of print in Sung China that enabled monks and morally serious men to believe they could "lodge" in the song lyric without being "detained" by it. The commonly presumed association of literacy and writing with abstract thought, logical argument, and distancing between performer and performance would seem to validate this theory.^[81]

It is certainly reasonable to postulate that the simultaneous rise of a new genre of literature and a new type of information technology points to some kind of cause-and-effect relationship between the literature and the technology, although it must always be stressed that this relationship is not necessarily unidirectional and that other factors—aesthetic, social, economic—play an important part in it. Since some kind of mechanism of moral distancing seems to have come into play at the same time as the rise of print culture to enable the song lyric to gain acceptance among statesmen and certain monks, print could be singled out as a factor in this distancing. I believe this is a useful line of inquiry that may throw new light on Chinese culture from Sung times on. But it raises at least two kinds of questions.

First, print was not the only new force in Sung society that can be associated with distancing. Money has a definite distancing effect, as reflected in the contrasting functions of money and food offerings in ritual contexts in China.^[82] It is well known that a money economy began to flourish in the Sung dynasty; would that not have had an effect on the culture at least as great as the effect of print? The song lyric was part of the new money economy, for the women entertainers among whom the genre developed in the growing commercial metropolises performed for money; and my discussion of the publication of lyrics has

[80] For some relevant texts, see Ronald Egan, "Ou-yang Hsiu and Su Shih on Calligraphy," Harvard Journal of Asiatic Studies 49 (1989): 402–12.

[81] To cite just two works in which these associations are articulated: see Sweeney, Full Hearing , p. 97; and Lentz, Hellenic Greece , pp. 3–6 and 177–81.

[82] See John McCreery, "Why Don't We See Some Real Money Here? Offerings in Chinese Religion," Journal of Chinese Religions 18 (1990): 1–24.

touched on many ways in which books, whether published commercially or by government schools, were printed for the same purpose. So an analysis of the relationship between exchange mechanisms and culture may be a fruitful path for future research.

Second, we find that it is very difficult to generalize about the influence of communication technology on consciousness. In the Chinese case, the common attribution to literacy of such virtues as analytical thinking, separation of performer and performance, and logic may be valid, since these are all features of Chinese thought in classical times, when literacy was becoming widespread among record keepers and rhetoricians who served the feudal courts and produced philosophical and historical texts. If there was linkage between print and the new kind of moral distancing that permitted the writing of song lyrics among the elite and the religious in Sung times, we must wrestle with the question of whether print culture is simply an intensification of literate culture or represents a leap to a new stage of culture. I have already noted that in terms of the efficiency of knowledge storage and retrieval, the fundamental dividing line in human cultures is not between oral and literate but between print on one side and the oral or chirographic on the other. We have also seen discernible effects of print on the kind of information that was stored and the manner of its use in the period under study. It would seem, then, that we must do more work to define the differences between the consciousness characteristic of print culture and that ascribable to mere literacy. The fact that much of the scholarship in this area does not distinguish between chirographic and print cultures when making comparisons with oral cultures complicates our work. The issues go far beyond the study of a single genre of poetry; they must be addressed diachronically and across several disciplines.

Indeed, the very assumption that aesthetic or moral distancing depends on either literacy or print should not go unquestioned; at the least, the terms used and the scope of their application must be carefully defined. For there are in fact many devices that produce aesthetic distancing in oral cultures (masks, for example, animal parables, evocation of ancestors as authority, etc.). Similarly, certain cultural environments have been known to foster a sophisticated awareness of language and linguistic questioning in peoples without any writing system.^[83] That is to say, literacy and/or print may not be necessary conditions for detachment from performance, even if they have apparently been sufficient conditions in some cultures. If the linkage I have suggested is

[83] Finnegan, Literacy and Orality , pp. 66 and 47–50.

valid for the song lyric in the print culture of Sung China, we still cannot easily extrapolate from it a general rule for the causes of "distancing."

A comparative approach has helped us understand some relationships between the song lyric and its social and material context; but this approach has also alerted us to the need to reexamine the simplistic dichotomies and supposed universals that underlie much of the existing scholarship. I believe that the hypotheses I have posited concerning the song lyric and the culture in which it developed can be accepted heuristically and will stimulate fresh approaches to Sung literature and culture. The wider theoretical implications this research on the song lyric seems to indicate should also be pursued, but only with the greatest care.