You can view a
version of this
article here.
The Future of a Luminescent Cloud
Recent Developments in a Pan-Asian Style

Hou Hsiao-Hsien, Tsai Ming-liang, and Hong Sang-soo are three of the most stylistically exciting directors to come out of Asia in recent years. It has been observed that all three have incorporated the rare combination of long take and static camera into their work, a distinctive style that has led to declarations of a pan-Asian Minimalism. James Udden explores the origins of such a style, its connection (or lack thereof) with tradition, and its significantly wide-ranging execution by three very different directors.

Historically there has been one almost invariable rule: the longer the takes are in a film, the more likely they will include the mobile camera. In fact, it is typical of a film primarily comprised of long takes to have camera movements in at least two thirds of its shots, if not more. Certain long take masters such as Jancso or Angelopoulos sometimes barely keep their camera still for an entire film.

The most notable exception to this rule apparently has occurred in Asia over the last two decades, first evidenced by the films made between 1984 and 1993 by the Taiwanese director, Hou Hsiao-hsien. Over this period Hou’s takes not only become longer on average, at the same time they become more static. In his first feature-length film as part of the Taiwanese New Cinema, boys from fengkuei (1984), the average shot length is “only” nineteen seconds per shot, and almost half of those shots contain camera movement. By 1993, with the Puppetmaster, the average duration of Hou’s shots reaches eighty-three seconds per shot, and yet now only 30% of these have any camera movements. Moreover, almost half of those are just a slight reframing at most; otherwise they are merely an overt camera movement for only a brief portion of an otherwise static long take.

From 1995 on, Hou inexplicably abandons this style starting with Good Men, Good Women. Still, other directors in East Asia evidently noticed what Hou had done up to 1993 and decided to take it up a notch. The first evidence was the Japanese director, Hirokazu Kore-eda, and his 1995 film, mabarosi, which averaged twenty-five seconds per shot and yet has camera movement in only six percent of them. Lee Kwang-mo’s springtime in my hometown (1999), from South Korea, averages nearly fifty seconds per shot, and yet only two percent of them have any camera movement whatsoever. The most consistent in pursuing this unusual long-take style have been the Malaysian-born Taiwanese director, Tsai Ming-liang, and the South Korean director, Hong Sang-soo. Both directors have even made long take films where literally every single shot is perfectly still. Hong’s the power of kangwon province (1998), does not contain a single camera movement despite an average shot length of just over thirty-three seconds per. Tsai’s what time is it there? (2001) averages more than a minute per shot and yet once again there is not even the slightest movement in any of them.

In no other time, or in any other part of the globe, have a set of directors so ardently pursued such a minimalist long-take style as what has occurred in East Asia. In effect, Hou, Tsai and Hong are at the forefront of what can be best described as a pan-Asian style. Yet certain questions arise from this. Is there any larger cultural meaning behind this, or is this merely a quirky aesthetic phenomenon? Furthermore, what are the limits of this sort of style, and how long can it sustain itself? Some tentative answers can be found in the most recent films available from the three most consistent practitioners of this style: Hou’s café lumiere (2003), Tsai’s the wayward cloud (2004), and Hong’s woman is the future of man (2004). On the surface these three films suggest a wavering from this pan-Asian style. Closer inspection, however, reveals an underlying dynamic and complexity that better captures contemporary Asian cinema than any single term can hope to do.

The Significance of the Pan-Asian Style

Why have Asian directors practiced this most unusual style, and why is this even significant? One might be tempted to chalk this up to culture, but we have to clarify what is meant by culture. If we mean traditional culture, then there is little to suggest why the long take and the static camera would be the result. Existing attempts to define the styles found in East Asian cinema suggest that a mobile camera has traditional roots, not a static one. [1] Within Taiwan often this is explained in terms of the practical limitations of the production situation found there. [2]

Still, even if the origins of this style may be somewhat accidental in Taiwan, what has occurred over the last decade seems a more conscious appropriation by other Asian directors, especially given how literally some have practiced it. This appears to be a part of a larger, contemporary cultural phenomenon in East Asia. Despite the greater encroachment of Hollywood, McDonalds and Starbucks, there are many signs that as East Asia becomes more modernized, it also becomes more “Asianized.” No where is this more clear than in Asian popular culture, where trends and fads often defy longstanding geopolitical realities. Despite tensions across the Straits of Taiwan, for example, Taiwanese popular music has been the rage in mainland China, while Chinese serial dramas are watched by many Taiwanese. The Japanese may be generally despised throughout Asia, but this has done nothing to stem the popularity of Japanese pop music, clothing styles and animation throughout the region. Most recently there has been a Korean Wave throughout Asia as seen in the regional popularity of its movies, television programs and music. This is evidence of a pan-Asian identity in popular culture: East Asians are now more likely to consume and borrow from each other than from the West. What it may mean politically or ideologically in the long run remains to be seen. [3] This particular pan-Asian style — where the long take is coupled with a static camera – shows how this is occurring even outside of the mainstream. Yet it raises a few questions of its own.

The Limits of a Pan-Asian Style?

This brand of Asian filmmaking has been labeled “Asian Minimalism,” [4] which may be somewhat of an understatement. What makes this sort of style particularly audacious and risky is how much it reduces one’s choices as a filmmaker: not only is editing not often relied upon, one also does not have recourse to camera movement to get from point A to point B. One has to wonder if in some ways those who pursue this particular pan-Asian style often find themselves boxed into a corner – in some cases literally. So how long can this style really sustain itself?

For this reason, it is not surprising that only a handful have managed to maintain this style for a long period of time. After being accused of being too Hou-like with mabarosi, Kore-eda has largely abandoned the static long take ever since. The mainland Chinese director, Jia Zhangke, has openly acknowledged his debt to Hou, and it shows in many of his distanced and static long takes in platform (2001). In his next film, unknown pleasures (2002), by contrast, the camera is mostly hand-held and hardly ever keeps still. Yet if we look at the most recent films of the three masters of this pan-Asian, minimalist style, the numbers seem to indicate that even they are wavering.

Hou Hsiao-hsien never literally nor consciously pursued the static long take to begin with: as we have seen, he parted ways with this style as early as 1995. Yet his café lumiere posed a most interesting dilemma for Hou. This film was commissioned by Shochiku in Japan as a homage to Yasijuro Ozu on the hundredth anniversary of his birth. Hou has often been mistakenly compared to Ozu even though their styles are radically different. (For example, Ozu’s style is very much based on editing.) Yet Ozu over his career moved the camera less and less, especially in the later stages. [5] Thus, for Hou to do a proper homage to Ozu would seem to suggest that he should revert back to earlier form and revive his own static-camera aesthetic. Yet Hou does nothing of the sort. Although the average shot length of this film is shorter than his most recent films, coming in at “only” sixty-six seconds per shot, there is still camera movement in over three quarters of these shots.

Tsai Ming-liang, on the other hand, has been almost doggedly persistent in this trend. As already noted, what time is it there? has an average shot length almost identical to that of café lumiere — but without the slightest camera movement throughout the entire film. Yet even Tsai seems to have “peaked.” His next film, Goodbye, Dragon Inn (2003) has a slightly shorter average shot length at fifty-five seconds per shot, and yet now a little over ten percent of them move. In his most recent film, the wayward cloud, the average shot length has been reduced to just over a half a minute each, and yet it has camera movement in nearly a quarter of them. Thus, these three films combined suggest a sort of “regression” from a more pure minimalist style in 2001.

Hong Sang-soo’s most recent film suggests that he has abandoned the static camera just as Hou did a decade earlier. In his 2002 work, on the occasion of remembering the turning gate, the average length of each shot is now nearly a minute each, and still only under a quarter of those have any camera movement. By contrast, in his 2004 work, woman is the future of man, the average shot length has nearly doubled to just over a hundred seconds per shot, yet seemingly at the expense of the static camera: now almost eighty percent of these shots have camera movements, and over two thirds of those are overt in nature, not merely slight reframings. In other words, these numbers suggest that in fact this pan-Asian style may be dying out, and that these directors are slowly losing something that has distinguished them and East Asian cinema from the rest of the world.

The problem is that these numbers are just that: numbers. Such statistics do tell us something, but not everything. They are crude measurements at best, useful as starting points in trying to grasp what is occurring with these films; there is more to be discovered here than how much or how little the camera moves in the midst of long takes. In fact, these numbers may be an apt analogy to how this particular aspect of the pan-Asian style – namely the coupling of the long take with the static camera – may also be a starting point for these filmmakers and nothing more.

Let’s start with Tsai Ming-liang and the wayward cloud. One question comes to mind: why is Tsai still the most persistent in keeping his long takes mostly static? The answer is because, compared to Hou Hsiao-hsien and Hong Sang-soo, Tsai Ming-liang is the most schematic of the three directors. Tsai’s films are minimalist in just about every sense of the term: not just minimal editing and minimal camera movement, but also minimal plot, minimal dialogue and even minimal character development. In fact, characters in his films are more like human cogs caught up in an absurdist universe of Tsai’s own creation. Almost all of these characters suffer from obsessive compulsive behavior always involving the central motifs in the film. Tsai’s uber-motif has always been water in some form, which seems to represent the indelible somatic and sexual reality of human beings. (For this reason, nearly every one of his films features scenes in bathrooms.) Yet often, this central motif interacts with one or two other motifs in the film as well. In the river (1997), for example, Kang-sheng seems unable to rid himself of something from the river after posing as a dead body for a film; later he is unable to shake a pain in his neck which gets worse and worse, leading to a chain of events that result in inadvertent incestuous encounter in a bathhouse. In what time is it there? Kang-sheng obsessively resets all clocks and wristwatches to Paris time, yet this leads him to a “twelve o’clock” encounter with a stranger in a bathroom. Meanwhile, his love interest in Paris, after drinking too much coffee, nearly has a lesbian affair with a woman she first meets in a bathroom and who subsequently offers her a glass of hot water.

To wit, there is just no escaping water in Tsai’s films, and whatever other motifs happen to be tagging along. While often humorous, these works seem like Kafkaesque worlds one can never escape. And the fact that Tsai persistently conveys this through minimal editing and camera movements makes all the more palpable this sense of being trapped. The pan-Asian style suits Tsai’s personal concerns exceptionally well.

the wayward cloud is essentially a continuation of all of this. Now water is inexplicably in short supply, but watermelons and water bottles are not. The characters are obsessed with both, and both objects are used for sexual activities, resulting in some of the most comical moments in the film. The absurdity of this cinematic universe is visually abetted by the use of extreme angles throughout. The opening minute and a half shot is a completely static long take of an intersection of two pedestrian underpasses, using a very wide lens, and thus creating a slightly distorted perspective. The few camera movements in the film also serve their purpose: they mostly occur in the absurdist musical numbers appearing from seemingly nowhere, much as had occurred in Tsai’s earlier film, the hole (1998). The only novel feature for Tsai is a possible new interest in Taiwan itself, which in his previous films has served merely as a convenient backdrop. Now various local cultural icons appear: dancers engaging in salacious gyrations with a statue of Chiang Kai-shek, the female protagonist working in the National Palace Museum, and the orgiastic denouement occurring next to a life-sized, cardboard cutout of flight attendants for China Airlines. Aside from this, this is the same Tsai Ming-liang as before.

A much more complicated picture arises with Hou Hsiao-hsien, the most complex of these three directors. One of the reasons he is no longer beholden to the static long take is because style is not his primary concern, but merely a means to other ends. [6] When he speaks of wanting to have a style that “belongs to the East,” what he ultimately seems to be referring to is indirectness. [7] His monumental film from 1989, city of sadness, is a milestone for Asian cinema despite the fact that few people can understand it after a single viewing. Yet Hou likes it less than his other films because he finds it, amazingly enough, “too direct.” [8] Given how Hou’s own criteria calls for a more poetic and indirect fashion of film narration, we can see why he used the static long take for so long. Both editing and camera movement have typically been used to literally “direct” a viewer’s attention – a static long take is one way to resist that. Nevertheless, over time he realized that indirectness could be done in multiple ways: through narrative structure, through the staging of his actors, and even through lighting. The static long take evidently became less and less useful for Hou as time went on.

café lumiere shows how indirect Hou can actually be. A single viewing of this film might leave the impression that there is not much to it; a subsequent viewing reveals underlying layers of both feeling and meaning beyond what one first suspected. For whatever reason, Hou has seemingly required a certain amount of distance from his subject matter to convey it successfully on film. In most cases, this distance is historical, and his films of the historical past are usually more compelling than those occurring in the present. Yet café lumiere could be considered the first Hou film of present-day subject matter that feels like his historical films, perhaps in this case due to a cultural/geographical distance, since this takes place entirely in Japan.

Despite what the above numbers tell us, for many who have seen this film and are familiar with Hou’s work, the camerawork in this case seems more static than his most recent films. A finer grained breakdown of the numbers accounts for this impression. While just over seventy-five percent of the shots do contain camera movement, less than twenty percent of them have camera movement for more than half of the duration of the shot. Twice as many shots feature overt camera movements, but which take up less than half of the shot’s duration. Just over twenty percent of the shots in the film have only the slightest reframing, meaning essentially no change in the shot’s composition. [9] Clearly then, these are not the more continuous camera movements found in goodbye south, goodbye (1997) and flowers of shanghai (1998); certainly they are not the baroque camera movements of Angelopoulos or Jancso.

Yet café lumiere is one of the most open-ended narratives Hou has made to date. There are three story lines, all joined in progress, all ending without any clear resolution. Among the usual quotidian details and everyday conversations, subtle hints are dropped as to what these underlying stories really are. As is typical of Hou, we are almost like outsiders who get mere snippets of other people’s lives, and we have to put together the missing pieces on our own.

When Yoko, the protagonist of the film, reveals that she is pregnant by her Taiwanese boyfriend, her back is completely to the camera, giving us no visual clues to her state of mind. In the next scene she is sweeping the ancestral tombs, yet the camera is so distanced that there is little indication as to how the rest of her family, including her father, have reacted. In the subsequent scene Yoko, her father and her stepmother eat noodles together, and yet all three have their backs to the camera, still giving us no visual clues as to any of their feelings. Only when the husband and wife are alone do we get any sense at all of how he is taking it: the stepmother complains that he always says nothing when something serious comes up, and yet the father still says nothing. Later scenes indicate that somehow they have accepted their daughter’s fate in the intervening ellipses. They visit her Tokyo apartment, and Yoko explains why she will not marry her Taiwanese boyfriend. Once again, aside from ceasing to eat, the father otherwise does not react. The film ends not long thereafter, answering none of the questions raised about her pregnancy.

Such open-endedness is echoed by the other two narrative strands, both also involving Yoko. One concerns her research into a Taiwanese composer from the Japanese era, Jiang Wengye. Once again, we see her snatching up random fragments of his past life, just as we are catching fragments of hers. One is left with an indelible sense of unbridgeable gaps between past and present: a bookstore owner knows nothing of Jiang’s past patronage of his shop; a coffee shop Jiang once frequented is now nothing more than an unremarkable office building. Through all of this Yoko is often accompanied in her quest by Hajime, a male friend whose relationship with her is never clearly defined. For some reason Hajime likes to record the various sounds of different trains in Tokyo, saying at one point that perhaps that this will be part of a research project as well. Yet their ambivalent relationship is summarized in one of the later shots in the film. From one train, the camera pans right to reveal Hajime on an adjacent train recording the sounds. As Hajime’s train trails off, the camera pans further right to show Yoko standing by the door, unaware that Hajime just passed. In one brilliantly executed shot, the camera movement suggests, with the greatest subtlety, that whatever unspoken feelings may or may not be there, some unknown fate will keep them apart. For all the seeming quietude, Hou’s world is one of flux and changeability. In this case the moving camera helps convey this.

Perhaps the most intriguing of these three directors, however, is Hong Sang-soo. Hong has always displayed an innovative way of handling narrative structure, as exemplified by the he said/she said versions of the same story in bride stripped bare by her bachelors (2000). He has also displayed a remarkable sophistication in dealing with the vagaries of male/female relationships, making these among the most adult films found anywhere in the world today. Even more striking is the often lingering question as to whether these are Korean themes, or universal ones: do men everywhere really want women to be sexually active virgins? What ultimately do women want in return? woman is the future of man brings all of this to a whole new level. Two old friends, Munho and Hunjoon, meet again after several years. Munho, a professor of Western Art, complains to Hunjoon, a western trained film director: “Koreans are too fond of sex. They have nothing better to do. There’s no real culture.” And yet the behavior of both men thereafter implies that they are no different.

One of the pleasures of any Hong film is the parallels drawn between characters, and this is where his stylistic strategies come into play. In the case of this film, camera movements are used for very calculated effect. This is most evident in the two long takes that occur in the restaurant when Hunjoon and Munho eat and talk. The first of these is just over six minutes in length, five and a half of which remains in a completely static framing of the two men eating by the window. The camera pans briefly three times: the first is a pan left from a counter showing a woman in a purple scarf about to leave the restaurant, ending with the two men seated by the window. Later Hunjoon, when alone at the table, asks another young woman working in the restaurant if she would act in his film. She refuses and the camera pans right as she walks to the same counter. The camera then pans left back to Hunjoon, only now that same woman in the purple scarf is outside standing across the street, evidently waiting for a ride. Hunjoon then exchanges furtive, suggestive glances with her.

After a lengthy flashback, a second long take parallels this first one. Being exactly five minutes in length, more than four and a half of those minutes sustain the exact same framing of the two men eating by the window. This time it is Munho who is alone towards the end of the long take, and he asks the same restaurant worker if she would model for him in the nude. Once again, she refuses (using exactly the same words she used with Hunjoon), and a brief pan right follows her as she goes back to the counter. Yet when it pans back to Munho, the young woman in the purple scarf is once again standing across the street. This time Munho exchanges suggestive glances with her until her ride finally appears. Hong Sang-soo may be moving the camera in this ongoing, muted and adult game of hide and seek, but clearly it is part of an ongoing development that has already begun in films where he never moved the camera at all.

The latest films of Hou, Hong and Tsai all suggest that the pan-Asian style, as defined above, may be undergoing a significant transformation. Yet it should be clear that this is no cause for alarm. After all, the prolonged use of static long takes has been but a small part of much more interesting and complex development. The latest works of these three directors are still among the most subtle, complex and dynamic in world cinema today. These three are to be commended for having pushed an unusual stylistic strategy to its seeming limits. Yet they are also to be commended for not restricting themselves to that.

This is James Udden’s first contribution to SYNOPTIQUE.


1 See Linda Ehrlich and David Desser eds., Cinematic Landscapes: Observations on the Visual Arts and Cinema of China and Japan, (Austin: University of Texas Press, 1994). In Chinese the best source for this line of thinking is Lin Nian-tong, Chinese Film Aesthetics [Chung kuo tien ying mei hsueh], (Taipei: Yun Chen, 1991)

2 I have often gotten this explanation in informal conversations with members of the Taiwanese filmmaking community. Most recently, in late May of 2005, I spoke with Wu Nien-jen, who insists that these tendencies are the result of concrete environmental factors in Taiwan itself, most of all the conditions of production.

3 For a recent scholarly discussion of some of these issues see, Timothy J. Craig and Richard King, eds. Global goes Local : Popular Culture in Asia (Honolulu : Association for Asian Studies and University of Hawaii Press, 2002)

4 David Bordwell, Figures Traced in Light: On Cinematic Staging (Berkeley: University of California Press, 2005): 231-233.

5 Once again, a more complete discussion of both Ozu editing and static camera can be found in David Bordwell’s, Ozu and the Poetics of Cinema (Princeton NJ: Princeton UP, 1994).

6 Hou Hsiao-hsien, interview by Peggy Chiao in The New New Wave Of Taiwan Cinema: 90’s [Taiwan tianying 90 hsin hisn lang chao] (Taipei: Maitian, 2002): 99-101.

7 Hou Hsiao-hsien, interview by author, June 20, 2001, Sinomovie Company Office, Taipei, Taiwan

8 Hou Hsiao-hsien, interview by Emmanuel Burdeau, in Hou Hsiao-hsien, (Chinese language edition of 1999 work by Cahiers du cinema), (Taipei: Chinese Taipei Film Archive, 2000): 107.

9 As always, it must be noted that these are very fine grained distinctions. There is no precise dividing line between a shot with only slight reframing, one with more overt camera movement but only for part of the shot, and those for the majority of the shot. It is always a judgment call, and these figures have to be taken with a certain margin of error.

James Udden is an Assistant Professor of film Studies at Gettysburg College in Pennsylvania. He specializes in Asian cinema and has published in Asian Cinema, Modern Chinese Literature and Culture and soon in Post-Script. He is currently working on a book manuscript on the Taiwanese director, Hou Hsiao-hsien.

<<< :: Previous entry
American Samurai, Chinese Sheriffs, and Hollywood’s Tokyo

Next entry :: >>>
Fantasia Festival 2005 Report

[ Back to Top ]

ISSN 1715-7641
Copyright © 2004-2005 Synoptique and its respective authors. All rights reserved.
Reproductions of any portion of this website only with the
expressed permission of Synoptique and its respective authors.

 * * *