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.