Mahler in Chinatown

It’s always worth tuning in to Donald Macleod’s Composer of the week on BBC Radio 3.

Even for a Mahler fanatic like me, last week’s programmes (based on the well-trodden theme of his years with Alma) were instructive.

I suppose this story belongs to the cliché of “international cultural exchange” (guoji wenhua jiaoliu 国际文化交流, which deserves an entry in my Catechism of Chinese cliché):

We all know (don’t we) about Mahler’s settings of Hans Bethge’s embroidered translations of Tang poems for Das lied von der erde (composed in 1909 but only performed after his death in 1911). In 1910, as he was fêted in New York, the Schirmers [some sources say the Roosevelts] took him and Alma on a visit to an opium den in Chinatown. Long before the stereotypes of Fu Manchu and Anna May Wong, this could have been an intriguing encounter.

Their visit to the “teeming” Lower East Side Jewish quarter must have been more conflicted (among myriad discussions of Mahler and Judaism, Norman Lebrecht, Why Mahler? is accessible). Alma’s portrayals are not always reliable, but here it’s worth citing her account (Gustav Mahler: memories and letters (pp.161–2)—more prurient than ethnographic:

We were invited by the music publisher, Schirmer, and his wife to dine with them one day and drive with them afterwards “down town”, into China town. The indispensable detective sat beside the chauffeur. We turned out of the busy streets into narrower ones which became by degrees quieter, narrower, darker and more uncanny. We got out, accompanied by the detective with a loaded revolver in his pocket, and went into an opium den. A creature with a sickeningly womanish face received us in an ante-room, where we had to put down a sum of money. He began at once to give us a long list of his successes with white ladies, and told us he acted female parts in the Chinese Theatre. A Chinese woman, of course, may not either act or look on in a theatre. He showed it in his face—it was the most degenerate man-woman face you could imagine. He showed us numerous photographs of American women he had—and he said the rest by gestures. Then he conducted us into several small but high rooms, empty in the middle but furnished with bunks along the sides, each of which contained a stretcher; and on each stretcher lay a doped Chinese with his head lolling into the room. Some of them raised their heads heavily as we approached, but at once let them sink again. It was a gruesomely horrible sight. They were simply dumped there to sleep off their intoxication. They might be robbed or murdered while they were in this state and know nothing about it. The whole scene resembled a baker’s shop with human loaves.

On now to a house of cards higher and higher, up into a room luxuriously furnished for strangers, cushions everywhere, and beside each cushion an opium pipe. And a Chinese, for payment, was ready to smoke a pipe on the spot while we watched him slowly succumb, rolling his eyes and twisting his limbs about. We were invited to smoke too but declined with horror. Next the theatre. Charming, but no play was being given. If it had been, no European would have been allowed among the audience. On again. Rats with long pigtails slunk nimbly and rapidly along the walls of the stinking street. Mahler said: “I can hardly believe that these are my brothers.”

On again. Small shops, small hotels, all silent. Finally, on the outskirts of this district we came on the habitat of a religious sect. There was a large hall at the far end of which sat a man with the face of a fanatic playing hymns on a harmonium in a pronouncedly whining style. The benches were occupied by a starving congregation. We were given the explanation. For listening to those hymns and joining in—a cup of coffee and a roll. What wretchedness in those faces! We pushed our way out, followed by hostile eyes, and for long afterwards we could still hear the flat notes of the hungry singers.

On again, and now the Jewish quarter. It was dark by this time. But here all was life and bustle, chaffering and shouting. The racial difference was staggering, but it was because the Jews worked day and night shifts to lose no time. The whole street was full from end to end of old clothes and rags. The air was heavy with the smell of food. I asked Mahler softly in his own words, “Are these our brothers?” He shook his head in despair.

With a sigh of relief we at last turned a corner and found ourselves in a well-lighted street among our own sort of people. Can it be that there are only class and not race distinctions?

Mahler’s music is so full of what would be known as folk and world music that his consternation is startling; can his success have made him so oblivious to his own background? And as ever, while trying to visualize the ethos of the time, we can only read this with later history in mind.

For Der abschied, see also here.


A tribute to Laurence Picken

I was one of Laurence Picken’s more tangential disciples, but he remains among the great inspirations of my life.[1]

I know several of us have fond memories of turning up for lunch at his little house overflowing with books, a sherry followed by a carefully prepared meal, listening to him carefully explaining, non-stop for four or five hours (for all his encyclopaedic erudition, he knew nothing of small-talk) how the marker to the right of the column in some 11th-century Japanese zither manuscript had been misinterpreted—with liberal asides on plainchant, birdsong, and medieval Sogdian viniculture—to which I occasionally managed to interject “I say, fancy that…” And that’s how it went, every couple of months for about twenty-five years.

Apart from his immense scholarly arsenal, he was a true amateur, an enthusiast. He maintained a network of like-minded people, communicating extraordinary enthusiasm for a topic that, even by the high standards of obscurity of those topics that many of us here today pursue, was pretty arcane. His devotion to scholarship was nothing to do with conforming to institutional demands; as a bit of a Lone Ranger myself, I now realize where I got it from.

He corresponded indefatigably with scholars all over the world (not least Eastern Europe)—he had to wait far too long for the invention of email. Though I think his influence on Western scholarship on Chinese music has been disappointingly slight, his work on Tang music had echoes in that of Chinese scholars, including He Changlin and a group of scholars in Shanghai, from Ye Dong and Chen Yingshi to a newer generation. Senior scholars like Yang Yinliu, Huang Xiangpeng, Ren Erbei, and Yin Falu were themselves engaged in similar work through the 1950s, and would have relished a chance to exchange ideas with Laurence.

Having helped him out with his magnum opus Music from the Tang court for many years, I finally began going to China in 1986. My initial reason for going was to seek clues to Tang performance practice in living traditions there—how to play his transnotations in a convincing style. Except for his early and late visits, most of his life coincided with a period when few foreigners could gain meaningful access to living traditions in the PRC. And immediately I discovered a vast unknown treasury of living folk and ritual music, soon putting to one side what we might call historical musicology in favour of contemporary ethnography (see e.g. my Plucking the Winds, pp.169, 184–5). But what I really appreciate is that Laurence entirely understood, and was immensely generous and supportive of this churlish choice of mine.

A special edition of Early Music, edited by Richard Widdess, includes my succinct thoughts on the relation between “early music” and living traditions in China, with thoughts on notation and recreation:

  • “Source and stream: early music and living traditions in China”, Early Music August 1996: 375–88.

As I published a lengthy analysis of some of the pieces from the Hua family shawm band’s suites introduced in my 2007 book (and the accompanying DVD film, and an amazing CD), it reminded me that very few (Chinese or Western) have done any serious analysis of Chinese instrumental music—so Laurence’s project with Noel Nickson (however traditional in style) on the Tang repertoire remains a bold, comprehensive, and detailed body of work. My only reservation is that I’m not so keen on analyzing old scores when we can’t hear how they actually sounded; doing fieldwork in rural China, I’m happy if we can make an educated guess— within a living tradition—about how a score no longer in use, was performed 100 years ago, let alone 1,000!

A distant relative of the Cambridge early music movement (Dart, Munrow, Hogwood…), Laurence’s Tang music project was controversial, not least in Japan, where it challenged deeply-held assumptions about the sanctity of gagaku.[2] Most striking is his theory that in Japan the Tang scores were gradually retarded—ending up being played up to sixteen times more slowly, robbing the melodies of their melodic coherence. Generally this remains convincing, though our later experience of living genres in China like the temple music of Beijing, or nanguan in Fujian, might prompt us to refine it.

Unlike some scholars, I quite accept that the Tōgaku scores that Laurence collected do indeed represent Tang music. But I wish I could debate with him now. His tenet that we should read the scores “with no more information than that given in the manuscripts themselves”[3] may seem at odds with his following comment, “the attempt to determine what an ancient text meant at the time when it was written”. So I think he might concur with my response:

I agree absolutely that we mustn’t assume the way a piece is performed now is the way it was performed before; this was his way of explaining an alternative to the passive acceptance of modern-day gagaku performance practice in Japan. However, one cannot possibly “use only the information contained in the scores themselves”! Recreations of European medieval music (a tradition to which Laurence belonged) always try to extract as much information as possible from early instruments, treatises, anecdotal literature, iconography, society, and so on—and also, notably, from living traditions which have remained relatively stable, as performers of European medieval music do for folk singing and instrumental heterophony in Europe and North Africa. All such material is abundant for the Tang, and Laurence would have loved to make more use of it; one cannot possibly treat the score (a skeletal outline) as if it provides all the information necessary to performance (it doesn’t even do that for Bach or Mozart!), in some kind of cultural void. Of course, we need to select judiciously which cultures we use as our material. Music is never merely notes on a page!

Laurence retained committed to the qin after his initial studies with Zha Fuxi and Xu Yuanbai in wartime Chongqing, along with Robert van Gulik (imagine…). In the 1960s he provided notes for John Levy’s Lyrichord recordings of Daoist and Buddhist ritual in Taiwan and Hong Kong, a rare initiative for the time—Laurence would have been excited by later projects on the mainland. (I note, en passant, that one online catalogue, under Genre listing, gives “Non-music”!)

The interminably long titles of his articles were endearing—my prize goes to

The musical implications of Chinese song-texts with unequal lines, and the significance of nonsense syllables, with special reference to the art songs of the Song dynasty.

And his language was charming, with formulations like

In this context, sheng 聲 is to be understood as an acoustic phenomenon with extension in time—something organized so that (again in time) it may be complete or incomplete; in fact, a tune.

Apart from Laurence’s chamber music gatherings, another cherished memory is him playing Bach on the clavichord—above which a magnificently garish framed picture (gift from a friend in China) of the workers, peasants, and soldiers clutching the Little Red Book, celebrating the achievements of the Cultural Revolution.


And for what it’s worth (not, you realize, for what it’s not worth), here are my notes for the memorial service:

Music from the Tang court: Qinghai bo (Waves of Kokonor)

Rachel Harris (dizi flute)
Stephen Jones (sheng mouth-organ)
Sun Zhuo (zheng zither)
Simon Mills (changgo drum)
Richard Widdess (bo cymbals)

Laurence worked for several decades on recreating the music of the early-8th-century Tang Chinese court. His insights from deciphering scores exported from Tang China to Japan still deserve wider recognition.

We tried playing these transcriptions in the 1970s, with more enthusiasm than ideas about Tang performance practice, or indeed any Chinese performance practice—given that this was during the Cultural Revolution, when we had virtually no access to the practice of traditional music in China. I still have little idea of Tang practice, but trying to play such pieces under the influence of “ancient” genres still performed today for rituals in the north Chinese countryside—notably the shengguan wind ensemble of ritual specialists around Xi’an, Wutaishan, and Beijing—yields what I find rather attractive results.

Laurence changed the course of my life. I first went to China in search of clues from living music there about how to perform these scores, and he was most generous, as ever, in understanding my rapid conversion to the documenting of living traditions in China, postponing historical reconstruction—well, until now.

In returning to the piece Qinghai bo (Waves of Kokonor), we ornament the simple outline of the tune, in 12 bars of 8/4, as Laurence suggested; we model our version on shengguan music, and are also influenced by our playing of Shanghai teahouse music. Whereas Laurence convincingly showed that Japanese performance practice had retarded the melody substantially, we begin with a very slow ornamented version, and gradually strip the ornaments away as we speed up, as they still do in Shanghai. I have no evidence that this practice was used in the Tang—given that the piece seems to be in 8/4, the first, slow, version is most likely to be “original”, but the faster versions are closer to the way that Laurence would have heard it, so these successive versions are more like alternatives.

Today we use dizi flute, sheng mouth-organ, and zheng zither, all of which have early scores for this melody; accompanied by a small changgo drum, a rough approximation to the Tang jiegu, and a pair of small cymbals, as in north Chinese ritual music today.

Laurence didn’t allow purism to delay his exploratory renditions of these pieces: one of my enduring memories of him is his playing of the melodica, with a completely straight face—I’m sure he would have recognised that modern ritual specialists’ style on the sheng, with its addition of fifths and octaves to the melodic line, might make a more suitable model.

While this is far from a historically informed rendition, it marks an advance from our versions of the 1970s; Laurence would doubtless have many comments! The music at last sounds Chinese, if not necessarily Tang Chinese…


[1] Just a few partial references:
On film:
[2] Among much discussion, Richard Widdess provides context: “Historical ethnomusicology”, in Helen Myers (ed.), Ethnomusicology: an introduction (The New Grove handbooks in music), pp.219–37.
[3] Ibid. p.221.

Faqu tu 2, or tutu

At Cambridge, Paul Kratochvil was not alone in enjoying a bit of drôlerie.

Here’s another jeu d’esprit on the faqu 法曲, from a precious old letter that I just found from my teacher Denis Twitchett, sent (by post!) from Princeton in 1986. A tad more whacky than his magnum opus Financial administration under the T’ang dynasty, it deserves to take its place among his magisterial writings on Tang history:

Dear Steve,
Some further ruminations on that mysterious term faqu. I am reminded of the hypothesis (first adumbrated in my alas-as-yet-unfinished “Preliminary proleptical remarks proving beyond Reasonable Doubt that Li Bo [aka Patrick O’Leary] was the earliest Irish poet”) that faqu is a rough-and-ready transliteration of the greeting shouted at tax-collectors in the medieval Irish countryside, and also commonly exchanged by rival drivers of donkey carts involved traffic accidents. The etymology of its common form in Chinese is obscure; under the Liao dynasty a folk etymology suggested that it meant “May the Law twist [your private parts]!” It should not be confused with the alternative writing (found in non-Buddhist contexts) fa-k’iu 發具 (explained by Karlgren as “Get out your [ritual] implement!”). This is quite distinct from the forms fakefu 法可夫 or 伏軻䮛 (the latter writing mistranslated by Legge as “Kneel by the hubcap of the Prince Consort”), meaning, according to Admiral Ting (first Chinese to be trained in the Royal Navy), “Be on your way, Jack!”

“French pieces” were surely more normally written in Tang Chinese as faji 法伎 or more colloquially as fashi 法式 “French models”. The title of the well-known Song-dynasty work Yingzao fashi 營造法式 is now attached to a rather boring work on architecture which explains how to construct yourself a baroque dog-kennel without mod cons. This work, however, is clearly a forgery by a Ming author, horrified to the depths of his neo-Confucian sensibilities by the original contents of the book, originally a Song DIY handbook “Make yourself a French model”.

The French presence long predates the Tang, and surely explains Zhuangzi’s Frog in the Well (a fine metaphor for the petty self-interest of the French). This also explains the title of the Tang Li Wa zhuan 李蛙傳, “The story of Froggy Li [the French model]”, and how the word wa “frog” came to have its other meaning “lascivious” or wanton”.

A funky address

A very different kind of oeuvre of that époque—less verité—is Jacques Rivette’s 1974 film Céline et Julie vont en bateau.

One abiding memory is my favourite address, where much of the “action” (if that’s the mot juste) takes place:

7 bis, rue du Nadir-aux-Pommes

We often used this a forwarding address.

What an age—Rafaelson, Bertolucci, Godard, Chabrol… At the time [historical note—Ed.], the way to see all these films was to do something called “going to the cinema”. Late-night Marx brothers films at the Arts Cinema were a rite de passage too…

In my imagination I tend to reduce my student years to the Tang dynasty, violin, and Adnams, forgetting how important—and how modern—my wider education was—film, art, literature. But the gaps were remarkable too. Meanwhile, Li Manshan and his fellow Daoists were languishing in silence and hunger (my book, p.133):

“We just sat around at home, but we could never feel at peace”—always fearing a knock at the door.

Money money money

The term emoluments is suddenly enjoying a dubious revival with a clause in the US constitution that is among many currently battening down the hatches.

The term, while not constantly on the lips of the rap generation, evokes fond memories from my days studying Tang history.

True, this is scant consolation for the current Destruction of Civilization As We Know It.

Another namby-pamby term used in academia that always makes me giggle is honorarium. But since I very occasionally get one, I mustn’t bite the hand that feeds me.

Musos are more straight-talking. One day our Mozart recording sessions in St John’s Smith square were interrupted by deafening building work outside. Reluctant to send us all home, the conductor discussed with the record company whether they might offer the workmen some kind of bribe to knock it off. Meanwhile the orchestra, aware that we would still have to be paid even if the session had to be called off, wondered whether we might make them a better offer to get them to keep going.

This was around the time of a dispute between a certain conductor and the brass players about overtime. A trumpet player (legendary for many touring exploits besides) put their case with the classic remark,

It’s not the principle, it’s the money!

This actually goes back at least to Eisenhower in 1959.


Some may (wrongly) imagine folk cultures as a kind of “living fossil”, but in China, thankfully, few yet seek to recreate the performances of the past in arid concert halls. Or at least it’s still a small industry, such as attempts to recreate Tang music… And so far it’s been an exercise performed, with little or no concern for historical style, not by folk musicians but by urban educated pundits and conservatoire performers, trapped within their modern preconceptions. [1] Folk musicians, like symphony orchestras (at least until recently), are quite happy working within their evolving tradition, without agonizing over “preserving” some supposed “authenticity”.

Scholars of Daoist ritual aren’t necessarily seeking to “hear a centuries-old piece of music as it was heard when it was composed”. What they may do, though, is silently equate living performance with that of the Tang or Song dynasties. From my book p.369:

While Lagerwey’s fine accounts of Daoist ritual in Taiwan occasionally suggest clues to changing ritual practice in modern times, “our primary interest […] is less to give a complete description of actual practice […] than it is to analyze the deep structure of that practice” (Lagerwey 1987: 91)—an influential perspective that tends to lead to the noble yet arcane goal of studying texts as evidence for the ritual structures of medieval times.

And Daoist scholars do sometimes seek to recreate rituals from the memory of elderly Daoists, as in Shanghai—evidently a worthy “salvage” project, albeit without reference to the changing social context since their youth.

Context and style have changed far less than with WAM—but they have changed. Apart from more general social changes, early music studies influence me in noting all kinds of changes in the Li band’s performance practice since the 1930s (my book pp.358–60):

Recently they have reduced the personnel from seven to six, discarding one guanzi, and the dizi has hardly been needed since the 1990s,

Turning to rituals, since current practice is dominated by funerals, this might at first seem to be their tradition. But they clearly recall a tripartite system of funerary, earth, and temple rituals; even if the latter two are now virtually obsolete, it is clear from their manuals. Thanking the Earth, once their most frequently performed ritual, has been lost since 1954; people can now afford to commission it again, but don’t. Though some temples have been restored, those holding fairs are fewer, and ritual sequences have been simplified along the lines of funerals. Three-day funerals are less common; and when they are held, the old sequence has become simplified and homogenized.

As to ritual segments within funerals, some were already largely obsolete by the 1940s, while others could still be performed in the 1990s but weren’t. Some, like Communicating the Lanterns and Judgment and Alms, have been radically simplified into mere symbolic tokens since the 1990s. Some—such as Dispensing Food or those from the “outer five rituals” like Crossing the Bridges—have become virtually obsolete since the 1950s; Li Qing and his colleagues could perform Opening the Quarters and the Pardon, but his disciples have hardly needed to do so. Yet others were probably rare even by the 1930s (Presenting the Memorial, Roaming the Lotuses, Smashing the Hells) or already lost by then (Offering Viands). Though segments have been adapted under Li Manshan’s leadership, his elders were already doing so long before.

As for the ritual manuals, we must take care to avoid some timeless ideal depiction. As the repertoire shrinks the manuals are not needed at all, but even before the 1950s many segments were performed without them. Li Qing probably didn’t know how to perform some of the rituals whose texts he copied in the 1980s. The lengthy chanted scriptures—around half of the total collection of manuals—were indeed placed on the table during performance, yet Li Qing and his colleagues could recite them so fluently that they barely needed to glance at them; now they are no longer expounded.

Along with the reduction in ritual repertoire, all three performance styles have been reduced—vocal liturgy, percussion items, and melodic instrumental music. The current repertoire of hymns is smaller than that in Li Qing’s score, so where there is a choice (as for Delivering the Scriptures and Transferring Offerings), that choice has become smaller. The “words of blessing” for Thanking the Earth are no longer performed, and fewer shuowen recited introits and mantras for offering paper are used. As the rituals that require them have been lost, instruments like the chaoban tablet, muyu woodblock, and qing bowl have fallen silent; and the dizi flute is no longer part of the melodic ensemble. The lengthy instrumental suites for Thanking the Earth and temple fairs are hardly performed, and the old variety of scales has been reduced. The repertoire of percussion items has also diminished.

But I don’t seek to lead the Li band towards reconstructing the practice of the 1930s, still less that of earlier ages.

Two small examples. Chatting with Li Manshan, I have mentioned how the “classic” instrumentation of the melodic ensemble that accompanies Daoist (and Buddhist) ritual around Beijing, and elsewhere in Shanxi, includes a ten-gong frame of yunluo—like Wutaishan further south, and Tianzhen (adjacent to Yanggao), where they still use a seven-gong frame, the lowest row missing. If Li Manshan felt so inclined, he could order a ten-gong frame, and “restore” it to the ensemble.

“But we don’t know how to play it!” he comments, reasonably.
“Even I could teach you!” I point out impertinently, adducing the common folk saying,
“A thousand days for the guanzi, a hundred days for the sheng; you can learn the yunluo by the fifth watch”.

But one reason I won’t press the idea is that, despite the Tianzhen yunluo, even his father didn’t recall a ten-gong frame. I may surmise that it must surely have been part of the band at some stage before the 20th century, but I don’t interfere. Li Manshan isn’t in the business of recreation, and neither am I. I describe, not prescribe—except when I transplant them to the alien context of the concert hall, when my subliminal influence, and their own perceptions of the demands of the situation, seem to prompt them to perform with somewhat more grandeur than in the casual current conditions of rural funerals.

Another instance: in my book I note that since 1953 there have been hardly any patrons commissioning the two-day Thanking the Earth ritual. Li Qing’s colleague Kang Ren (b.1925) described its sequence to us before his death in 2010; Li Manshan and Golden Noble were interested enough to take notes, but can’t mobilize their local patrons to invite them to do it. Most of its components could be recreated, if there were demand. But there isn’t. This is the kind of thing that Daoist scholars might commission specially as a worthwhile salvage project, but my gentle suggestions lead nowhere. Some other obsolete or rarely-performed funerary rituals (my book ch.13) could also be restored, just about. But local patrons wouldn’t welcome it—it’s inconceivable, until such time as they suddenly do request them.


[1] A rather different, if minor, case is recreations of obsolete rituals at the behest of local Bureaus of Culture. Such initiatives feel artificial, and scholars should take care both to point out the conditions under which they are made and to avoid silently equating them with some “authentic” folk practice. See e.g. Overmyer, Ethnography in China, pp.287–95.

Gems from Chinese history

Some quotes from The Cambridge history of China:

What is the use of a stele? —Sui emperor Wendi (“Just call me Wendi”) (vol.3, p.62). Cf. my book p.374!

The Cheng sisters were captured and decapitated. Mopping-up operations continued till the end of 43. —(vol.1, p.271). Messy business eh.

The “director of records for the empress” may have kept a record of the emperor’s cohabitations with the empress. This cannot have been a great burden. —(vol.1, p.503).

Hierarchies are an essential part of a well-ordered society, and they must be accepted voluntarily. —(vol.1, p.705). So there.