203

Shi Hui Notes about Directing Married to a Heavenly Immortal1 Recently the problem of how to shoot stage operas as movies has attracted the attention of quite some people. Does it become very difficult to give full play to the characteristic of film if one stays too loyal to the arts of the theater? Or do we necessarily bring damage to the arts of the theater if we go too far in “moviefication”? Is it best to shoot the opera without making any changes at all? Or is it right to go for “moviefication” without caring for the rest? Everybody has his own approach, and the opinions are not very unified. What I know in this respect is very limited, but let me mention some simple experiences and some suggestions based on my work in directing Married to a Heavenly Immortal. My earliest creative ideas I first have to acknowledge that I am an outsider when it comes to our national traditional theatrical arts. I am only an enthusiastic member of the audience, filled with a strong and passionate love. Whenever we came to some locality when our production team came to some place to shoot outside scenes, I would grab that opportunity to go and watch the traditional opera of that locality, and I would always greatly enjoy these performances. This applies even more to Peking Opera, which I would listen to from my earliest youth—I am truly addicted to it, I’m a true Opera fanatic! I’ve also studied it for a while and learned to sing a few lines. I’ve also performed on stage, and performed quite a range of characters, trying out every role type. I’ve even sung military plays. I didn’t hit anyone, and no one hit me. But according to what friends who watched my performances told me, my performance belonged to the “self-indulgent style.” Love does not equal understanding. Love does not mean not messing up when it becomes your turn to be responsible, and the people who criticize other people for messing up are not necessarily the people who will do things in style when they are in charge. Before I had undertaken the task of directing Married to a Heavenly Immortal, I had never watched Huangmei Opera. What I saw first of all was the adapted literary script (in its first draft). That had thrown off the shackles of the set conventions of the traditional theater and gave full expression to the nature and capacities of film; the fairytale2 features were very much emphasized. Because the script writer had also worked as a director, the script he had written was very rich in “camera-feeling.” At the same time it aroused in me some preliminary ideas for arranging this future film.

1

Shi Hui, “Tianxian pei daoyan shouji,” in Wei Shaochang, Ed., Shi Hui tanyi lu. Shanghai: Shanghai wenyi chubanshe, 1982, pp. 242-253. Originally published in Dianying yishu 1957 No. 5: 24-27. 2 The Chinese text systematically speaks about a “myth-movie” (shenhua pian), but “fairy-tale” probably conveys better the intention of the author.

204 I conceived of the idea of a new kind of cinematic style: a fairy-tale movie. The contents and incidents of the first draft of the literary script of Married to a Female Immortal allowed for the possibility. It also gave me a creative impulse—in order to shoot it completely in conformity with cinematic demands I wanted to compose a new score for which the original Huangmei Opera would only provide the materials; fully utilizing cinematic techniques I wanted to create an original fairy-tale film. “An immortal maiden descends to earth and marries a peasant who has sold himself as a slave”—this was such a daring and intelligent image! In my mind all kinds of pictures emerged: heavenly palaces floating in emptiness, immortal maidens riding clouds, scenes of the human world, a scholartree opening its mouth, a god of the soil serving as go-between, immortal maidens weaving silk, celestial soldiers and celestial officers from the Southern Gate of Heaven3 … This really fascinated me. Because we had done away with the set forms of the stage, the original music also would not be suitable, so the music for the whole play would have to be created anew, and dance movements also could be assigned to an extremely secondary position. Truly, this was a movie—a fairy-tale movie, preferably in full color! If I would shoot it without any change, sticking to the set forms,4 the music, and the dances of the stage performance, where would the creativity of us film workers be found? I was determined to undertake a daring experiment! After watching the stage performance Not much later the Anhui Province Huangmei Opera Company came to Shanghai, and the first play they performed was Married to a Heavenly Immortal. When I walked into the theater, I was carrying with me my original ideas about shooting this movie, and I had no intention at all to pay attention to the music and dances and characteristics of Huangmei Opera. I only wanted to absorb some materials that corresponded to my original conceptions. Strangely, however, the first thing to absorb me was the music of Huangmei Opera. Hearing that thick earthy breathing was really moving. What I want to talk about here is not some “novelty”, because I know that if it is only novelty it will not last long. What I liked was its clear national characteristics and heavy local coloring—its simplicity, its beauty, its appeal. This kicked my original ideas very far away. In the theater I stubbornly clung to some of those ideas, but the enthusiastic reaction of the audience made me waver again and again. In the end I realized that I had to abandon my original ideas, that I had to dispassionately watch the play as a member of the audience, and that once I had seen the play I had to dispassionately reconsider how to go forward. The more I saw the play, the more I liked it; the more I listened, the more I became entranced. The melodies were relatively easy to remember, and during the second performance I could already follow along and sing the first line of a couplet or continue the second line of a couplet. “Easy to remember” is, I am afraid, a characteristic of the music of our traditional opera. Even though “easy to remember” has some connection with “monotonous,” it is not the same as “monotonous.” Every genre of traditional opera
3 4

The Southern Gate of Heaven guards the connection between heaven and earth. The “set forms” refer primarily to the acting conventions of traditional theater such as Peking Opera.

205 has its own basic melody, and it is often the case that we only have to hum a melody and immediately can determine whether it is Peking Opera, Sichuan Opera, Yueju Opera,5 or Cantonese Opera. When you have watched a performance of whatever kind of opera, there will be a basic melody that will haunt you when you leave the theater, even if you cannot sing the words. This is the most important characteristic of our local theater; it is a characteristic that is quite different from western opera where every play has its own music. That future film should not have an original score of new melodies and new tunes, but would have Huangmei Opera music! As one scene after another was performed on the stage, I also step by step entered more deeply into the world of the play. I was deeply absorbed by the beautiful folk legend, and the scene of “Separating at the Scholartree” made even me shed tears of sympathy. I too had deeply fallen in love with this genre of local opera. I could not watch the play completely like other members of the audience. I wanted to understand from the performance the problems of Married to a Heavenly Immortal in terms of genre, script, music and dances. I wanted watch from different seats and angles. I wanted to watch from below stage and from backstage. This made me become even more acquainted with these problems, and it also definitely stimulated my creative ideas. I would very much like to stress the benefits I derived from watching over twenty performances. This transformed many of my ideas and fixed my basic direction and method in approaching this film. In the theater I paid special attention to the reaction of the audience: at what place they were happy and which place aroused their emotion. I derived great benefits from sitting amongst the audience and listening to their many comments. This made me suggest quite some proposals to the script-writer for rewriting the script. For instance, in the Magpie Bridge scene, the immortal maidens have sneaked out and watch the mortal world, and there is a long section of song and dance describing the fisherman and the woodcutter, the student and the farmer. The audience was deeply absorbed by this part of the play, and they showed their pleasure. This was the same at each performance, but in the first draft of the movie script this part of the play had been removed. In my opinion the reason the audience loved “the four characterizations” was first of all the beauty of its form, and secondly that is embodied the strong love of the people for those people who rely for their livelihood on their own labor. Because these two aspects were intertwined, the audience was very satisfied. Therefore, after an exchange of opinion with the author,6 we eventually restored this scene. Also, in the scene of “Meeting on the Road” there is an episode in which Seventh Sister blocks Dong Yong’s way on the main road; when Dong Yong sneaks off to a side road, she once again blocks his way on this side road; and when Dong Yong once again goes back to the main road, she also goes to the main road and blocks his way, leaving Dong Yong extremely unhappy. Following a conversation, she allows him to pass by her, but then she bumps into Dong Yong and says that he has bumped into her; when Dong Yong goes a second time, he manages to dodge her, with the result that Seventh Sister attempt to bump into him but fails to her great annoyance. The author and I originally had the same view of this episode: We felt that Seventh Sister blocking the road this way and
5

Also known as Shaoxing Opera. Performed by all female troupes, this genre enjoyed a remarkable popularity in Shanghai from the late thirties to the sixties. 6 Sang Hu, the script-writer.

206 that way and even bumping into Dong Yong made little sense and that it came across as “far-fetched.” In the first draft of the movie script this episode had been omitted, and we had chosen for the solution to have Seventh Sister appeal to Dong Yong for his sympathy because she was “all alone in the world,” which would then develop into mutual love. But why did the audience love the original performance so much? Seventh Sister is a young girl, so it would seem to be highly irrational to have her bump on purpose into a man she doesn’t know, but why did the audience not experience it like this and felt that Seventh Sister was extremely likable? The actors also performed it in such a convincing way. I was puzzled by this problem for a long time. In the end I found an answer. This kind of treatment first of all was in accordance with the characters of the persons. Seventh Sister is neither the White Snake nor Zhu Yingtai. Apart from her daring, intelligence and diligence, she is also characterized by her naughtiness. “Diligent and naughty” is her unique character. Dong Yong cannot marry an elegant and sophisticated wife who is so weak she is blown over by the wind and loves to eat but is too lazy to work. He has to share his life with a woman who does not fear poverty and suffering and is willing to work. Only this would conform to the people’s imagination. The stage performance of the “three blockings and two bumpings” greatly strengthens the description of Seventh Sister’s characterization. At the same time it develops the characterization of Dong Yong as loyal and honest, good and innocent. The later developments are causally connected to this episode; this episode also constructs the unique atmosphere of Married to a Heavenly Immortal. Using cinematic methods we could give full expression to the fairy-tale atmosphere that was hinted at in this episode. The author agreed to my suggestion, and we restored this episode. Watching many performances, I reaped many benefits, but there is not need to go into further details. A crucial round-table conference This round-table conference was attended by all creative personnel, and the scriptwriter also attended at our invitation. At this conference, apart from conducting a detailed analysis, study, and discussion of the script, we also decided the crucial problem of “how to turn Married to a Heavenly Immortal into a movie.” At the beginning, the opinions were extremely divided. Some people argued that the best approach to the theatrical arts was to shoot the movie adopting one stable method, because for many performance conventions we could not create a set of performance methods suited to a movie in the very short period of time allotted for our task, so that is all would be a wasted effort. Basically they argued that we shoot the movie just the way the play was performed on stage, because that would result in a documentary record. But at the same time there were people what raised the following counter question: When Seventh Sister flies down from heaven, it is obvious that that is best filmed using cinematic techniques, so would you want her at all costs to run a circle on stage7 in front of the camera? If she runs a circle on stage, then that is easily understood by the theater audience, but in a movie that is quite different. A movie audience is not a theater audience. A theater audience will not demand that Seventh Sister flies up and down on
7

A convention to indicate that the performer has moved from one locality to the next in the story.

207 the stage, but a movie audience will definitely not be satisfied when Seventh Sister runs a circle in front of the camera. Starting our discussion from this point, with respect to the question whether a movie audience comes to watch a movie or comes to watch a staged play when it enters a movie theater, our general opinion was that is comes to watch a movie. Of course, you can say that this only applies to a “story-film” (feature film). If this movie is a documentary of a stage performance of an opera, the audience will come of course to watch the opera. This resembles the scenery, design and editing of a news feature—the audience comes for the facts. In this way what we had first to determine was whether the movie Married to a Heavenly Immortal would be a documentary of a stage performance (a so-called documentary movie), or whether we would shoot a creative movie (a feature film). Determining the task of the filming had a clear advantage. From the very beginning the leadership8 had assigned us the task of a “documentary,” and only later this was changed into the task of a “feature film.” This decision gave us a great inspiration: we could approach our creative work from a cinematic angle. But the leadership had attached one condition: We had to preserve the characteristics of Huangmei Opera, and this gave rise to contradictions, because between an opera and a movie there are numberless irresolvable contradictions. Even though the contents of a script may allow for the insertion of movie characteristics, still the operatic performance brings with it a whole set of problems, problems that could not be resolved in a short time, problems that would have to wait for experiments and discussion. Was the task of this movie a documentary recording the performance methods of Huangmei Opera? It was not. Was it the task of this movie to document each movement and each set form of a particular performer to serve as teaching materials for later generations? It was not. Moreover, the current performance conventions of Huangmei Opera are very different from the codified and unalterable conventions of Peking Opera and Kunqu. These conventions had a great elasticity, they did not have to suffer the limitations of set forms, and were a system on its own. Theatrical conventions such as qiyi,9 “the stiff,”10 circling the stage, the horsewhip,11 city towers,12 jiaotou,13 kutou,14 sijitou15 and fixed poses were all optional. These conventions did not constitute a codified, conventionalized performance method, where “everything goes wrong as soon as one link goes awry.” Lively and naturally, these conventions formed a system of their own, and especially the performances of comrade Yan Fengying highlighted this point. This implies that as far as Married to a Heavenly Immortal and cinematic form were concerned, the contradictions between the two were somewhat less than in the case of
8 9

The director-in-chief of the studio, and the studio’s Party committee. Breathing technique? 10 This refers to the stage technique of falling backwards fully stretched out, for instance when fainting or dying, in the manner of a rigid corpse. 11 The horsewhip is used to symbolize riding a horse. 12 The Peking Opera stage tends to be bare except for a small table and two chairs, which may be used to symbolize various objects, such as walls and towers. 13 A specific rhythm beaten out on the gong to underline violent passion, whether joy, rage, sadness or frustration. 14 A specific rhythm beaten out on the gong when one of the performers is weeping. 15 Four loud strokes on the large gong and the small gong, accompanying entrances, exits, and frozen poses.

208 other genres of traditional opera, and therefore there was no reason at all to have the immortal maiden run a circle on stage in front of the camera when she descends to earth—it was quite acceptable to use special effects to have her “fly down to the realm of men.” At the same time there was no need to maintain a performance mimicking action when the immortal maidens are weaving silk, and it was quite acceptable to use cinematic techniques to display their ingenuity and wisdom. At the conference everyone tended towards this view, and the comrades of the Huangmei Opera Company also were very supportive of this view, even though this still might create quite some contradictions during the actual filming, but this was something we would love to try out and experiment with. Eventually we decided that Huangmei Opera would serve as the basis, that we would to the greatest possible degree utilize the nature and possibilities of film, and that we would make a fairy-tale feature film with songs and dance. First of all we had to destroy the frame of the stage, jump out of this thing that shackled cinematic creativity, do away with everything that only belonged to the form of a performance on stage, change this beautiful lyrical fairy-tale into something even more moving, provide it with a strong emotive power, and use the special qualities of film to express through images everything the stage could not express. This work was not simple, it was filled with contradictions, our capacities were limited, and objectively there were limits to what we could do. But we wanted to experiment and daringly create a new modality of turning opera into a movie, whether we would succeed or fail. That we decided to work in this way should not lead to the conclusion that this is the only way to turn an opera into a movie. Our approach was determined on the basis of a concrete genre, of the concrete contents, of the concrete performance form, and the concrete performers. Pushing the Mill Together (Shuang tuimo, a Wuxi Opera) is a pure documentary movie,16 and equally very good, but that is because the performance characteristics of this play consist in people mimicking an action. On a completely empty stage one relies on the movements of performers to make the audience feel that they are busy pushing the mill and hauling water, and that’s why is was right to shoot this movie as a documentary. When some theatrical arts or the performances of some famous actors are intended as historical materials, as materials for study, or as teaching aids, we can achieve that task not only by a faithful registration, but also by the added means of “slow motion,” “diagrams” and “explanations on the side.” Married to a Heavenly Immortal is an item that has its complete story, and when performed on stage is also not extremely conventionalized. Its contents and incidents are so rich in fairy-tale atmosphere, and the music of Huangmei Opera is so rich in the breath of life and emotive power that we decided to treat it in the form of a “fairy-tale feature film with song and dance.” It’s my opinion that “turning an opera into a movie” provides an extremely broad array of performance methods depending on the differences in concrete circumstances, and that this necessarily can generate all kind of different movies. The crucial question is to base oneself on the concrete situation. Starting from the notion of a “fairy-tale feature film with song and dance” we concluded that the less dialogue would be the better, and that it would be even better if the dialogue would rhyme. In this way the dialogues would be rich in musicality and also be suited for dance. We did away with the “drum-beat” of the stage music, the most
16

This movie was released in 1954. It was directed by Zuo Lin.

209 important reason being that the “drum-beat” had a very strong stage-feeling, and that if we would maintain it, one would very naturally be reminded of the stage. The “drumbeat” is the control center of the operatic music and has the function of a conductor: using their hearing it guides the cooperation of the orchestra and the performers on stage. But in a movie there is no need to see the orchestra, and therefore the method of conducting can be changed. In the process of singing the arias the rhythm of the “drum-beat” of one stress and three light beats may have two functions: One function is that it very objectively tells the performers “you have reached the middle beat, you have reached the stress,” while the other function is that is determines the speed of the tempi. The first is a kind of habit. When we asked our comrades the performers whether or not they could sing without the drum, they said they could do it, but the second function of determining the tempi depends on the drum. Apart form this there was still the other function of creating atmosphere. In this case the drum is used as one of the percussion instruments, and this has to be preserved because then the drum is treated as part of the orchestra. This decision was very daring, but it was very beneficial to the shape of our movie. We surmised that we also could not lose the characteristics of Huangmei Opera because of this, because we maintained the drum’s performance role as a musical instrument and did away with its role of the objective conductor that was devoid of any melody. Some experiences gained from the actual shooting Often things that make sense in theory do not make sense in practice. The decision that “Huangmei Opera would serve as the basis, that we would to the greatest possible degree utilize the nature and possibilities of film, and that we would make a fairy-tale feature film with songs and dance” was not an easy road at all. As soon as you started to give expression to it through shots, concrete problems arose. First of all, how to produce shots made us encounter many problems. I have watched Romeo and Juliet, starring Ulanova,17 quite some time, and this movie has greatly inspired me. I will never forget the shot in which Juliet runs down an endless stair when she rushes to the priest to ask him for the drug. This is a technique that combines ballet and cinema; at the same time it is an artful way to jump out of the stage. At the same time I also cannot forget how the camera movement is artfully combined with the dance. This provided me with and answer that said that if the director was not thoroughly acquainted with the complete music and dance of this movie, he would be unable to proceed in his creative work. Therefore, “the thorough knowledge of the music and the detailed awareness of the dance” became the important weapon of the director who was directing this kind of movie, and the shots would necessarily have to be produced from this basis. I once met a friend from the Peking Opera world, and he told me some of the circumstances of shooting “Creating Havoc in the Heavenly Palace” (Nao
17

Galina Ulanova (1910-1998) was the most famous ballerina in the Soviet Union during the 1940s and 1950s. The music for the ballet Romeo and Juliet was composed in 1935 by Sergei Prokoviev (1891-1953), and the ballet was first performed in 1940 by the Kirov Ballet in Leningrad (now St. Petersburg). A movie of the performance of the ballet by the Bolshoi Ballet, starring Ulanova, was released in 1954. The movie was introduced to Chinese audiences by Xiao Qian, “Yibu gesong qingchun, zhouzu fuxiu shili de wujupian—jieshao Luomi’ou yu Zhuliye,” Dazhong dianying 1956 No. 3: 24-25.

210 tiangong).18 His original entry on stage [as Sun Wukong] took place while monkey soldiers, holding their banners, were entering stage to a rapid beat of the drum. He was hidden behind a big banner, covering his face with his gown. He released his gown only after the big banners had arranged themselves to his left and right, and he then briefly assumed a frozen posture. Then he flipped his long sleeves, displaying his intense satisfaction, and the image projected by the Great Sage was extremely impressive. But before the audience had had a chance to observe him very carefully, he immediately displayed his monkey nature, pulling his ears and rubbing his cheeks like a true naughty monkey. Afterward he pulled in his long sleeves, raised his gown with his feet, and returned to his seat. This was his design for the performance of Sun Wukong’s entry on stage. This made me think how one could capture such a design in one’s shots. If one would in the usual manner first use a distant view followed by an intermediate view, followed by a close view or a close-up, and finally by a full view of his resuming his seat, I would not be able to correctly give expression to the actor’s performance. But if the camera could follow the Great Sage through al his actions as he appeared when the banners parted and he covered his face with his gown, when he full satisfaction flipped his long sleeves, pulled at his ears and rubbed his cheeks, grabbed his sleeves and lifted his gown, that would probably be somewhat better. The expressive movement which the actor had created through long-term observation of human characteristics should be the most important starting point. On the basis of these two inspirations their more than twenty performances became the most important location for generating the shots. The stage performance of over three hours had to be shortened to the one hour and a half of a movie, so we had to change the original script. Eight lines became four, four lines became two; two arias became a single aria. We did away with all non-essential lyrics, and we added some film music that was descriptive of the surrounding atmosphere. The length of the transitions had to be adapted to the needs of the movie, so we had to change the original tunes. All this could not be separated for even a minute from the performance and dance of the actors. We knew that Huangmei Opera before Liberation had no accompaniment of string instruments—it did not even have the two-string fiddle! It only had percussion instruments, and the accompaniment of stringed instruments is a post-Liberation feature. Even the commonly used transition (musical example) is a new invention. Thanks to the unrelenting efforts in recent years of performers and composers, the music of Huangmei Opera has seen a continuous development. It may have its own characteristics, but it has not yet solidified. Comrade Yan Fengying can sing many folksongs, and many of the tunes for female voice have been created by her. … Comrade Wang Shaofang was originally a Peking Opera performer, and therefore he has borrowed some things from Peking Opera. Huangmei Opera originates from Huangmei District in Hubei Province, and only later it spread to Anhui, and therefore it also has a blood relation with Handiao and Huidiao. But Xipi, Erhuang, and Qinqiang also have each their own relationship with Huangmei Opera, and therefore we can say that the digestive
18

When the monkey king Sun Wukong has been ennobled as Great Sage Equal to Heaven but is not invited to banquet hosted by the Queen-Mother of the West, he steals her peaches of immortality, eats Laozi’s elixir of longevity, and returns to his original abode on Flower-Fruit Mountain. When the Jade Emperor dispatches his troops against Sun Wukong, they are soundly defeated. This Peking Opera was the representative play of Yang Xiaolou.

211 system of Huangmei Opera is extremely strong. In this way it shows a very large elasticity in the “level words,” “tune of the immortals,” “colorful tune,” “fired-up style,” “two-beat” and “eight-beat,” which are all manifestations of the basic tune. These can exist independently, but they can also be intermingled. This is a great difference with Peking Opera and Kunqu, and this makes it relatively lively. This was a great benefit for the movie. In dealing with the music we never departed from the needs of the movie camera, and this was a very beneficial condition. Everything was subservient to the needs of the camera, and the camera was subservient to the depiction of the characters and the original and relatively fundamental performance design of the actors. This was our fundamental guideline in the creative work of shooting Married to a Heavenly Immortal. In conclusion, I have to repeat once again that I am an outsider when it comes to traditional opera. I am only an amateur without any deep knowledge—at most I have only held half a weapon. I have mostly considered the problem from a cinematic standpoint, and in that way I have daringly experimented with this new genre. Unavoidably, I have created many defects and even failures. The contradictions between scenery and dance and the lack of unity of the movie as a whole testify to that. But the question is whether the road of this experiment can be pursued, so that these problems can be resolved one by one. Our most basic view has been that the concrete situation has to determine how the movie will be made. (1956)