The question has bothered me for a long time, and I guess it will continue to do so for years to come. I still remember the unease many of my friends felt after they watched this film by Zhang Yimou in 2002, and the feeling of betrayal among them could be understood in two registers.

(1) How can one accept that a director, who has always spoken openly his social and political criticism (and therefore had most of his internationally-acclaimed films banned by the authorities), kowtowed to the state apparatus by making what appears to be an unapologetic piece of propaganda? (2) How can one accept that a director, who has always valued the beauty and power of photographic reality, created the most incredible and phantasmagorical spectacle of computer-generated images (CGI's)? Perhaps we can work through these two levels of discomfort and see if we can come up with a new interpretation of the film.

This impression that Hero is a sellout probably comes from a straightforward reading of the film. According to such reading, four warriors, Broken Sword (Tony Leung), Flying Snow (Maggie Cheung), Sky (Donnie Yen), Moon (Zhang Ziyi) ask Nameless (Jet Li) to assassinate the King of Qin (r. 246-221 BCE; as the First Emperor, r. 221-210 BCE; Chen Daoming). Nameless hesitates at the last moment because he sees the King's potential to achieve and preserve the unity of the tianxia (empire).

Nonetheless, this “sellout” theory may become more difficult to explain as we consider the film's financial basis. The film's $30,000,000 budget was raised under the guarantee of three distributors, Beijing New Picture Film (China), EDKO Film (Hong Kong), and Miramax Films (US and UK). These distributors guaranteed, at the stage of preproduction, distribution in four major markets: China, Hong Kong, the US, and the UK. In this respect, Mark Harrison suggests that the film was perhaps one of the first instances of a piece of “national” propaganda that was backed up by a global financial market for their mutual benefit, since the continued coherence of the “Chinese nation” guanrantees the continued stability of the marketplace (See, “Zhang Yimou's Hero and the Globalisation of Propoganda,” in Millennium--Journal of International Studies, 34.2 [2006]: 569-72). According to such interpretation, the film's plot cannot be more metaphorical: global financiers, some of who might have different troubles with their imagined “China,” pay tribute to the “future emperor” in order to safeguard the integrity of the global market.

Too easy, perhaps?

Not unlike any “classical” narrative system, Hero is organised around an enigma, and in this case, an epistemological one. The narrative structure is modelled directly upon the cinematic classic Rashomon (Kurosawa Akira [1910-98], 1950), which has been known for its narrative non-linearity and unreliability. The spectators of Rashomon were therefore seduced by their curiosity of finding the truth despite the impossibility to find one. Nevertheless, the spectators of Hero today are confident enough to claim: “I can by-pass the structure of non-linearity, the questionable narrative, and even the narrative about disbelief, and reconstruct a perfectly straightforward and linear narrative that I fully believe.” The criticism against the film is therefore symptomatic of a fundamental epistemological turn: the spectators' affective responses to the cinematic image no longer give these spectators physical or intellectual knowledge about factive reality (or the narrative puzzle from which one could question this reality); rather, these responses are re-processed, re-configured, and re-assembled actively and effortlessly as pieces of information that affirm a pre-conceived notion of fictive reality. In other words, the film is merely one out of many “para-texts” (texts that put certain referential differences between them under erasure) that the spectators can browse through in order to reach a conclusion they have already expected to reach.

This spectatorial behaviour is mirrored in the film by the King of Qin, who interrogates Nameless about the “back stories” of the four warriors. The King's interrogation differs from the way it was done in Rashomon. In Rashomon, the spectators are interested to find out which testimony they should trust in order to find truth or reality. In Hero, the King of Qin has already had a preconceived notion of what reality is, and is only interested to entertain himself, and to re-confirm what he presumes as the truth. The basis of interrogation is also based on a presumed guilt: “I know that you are guilty, and I am only interested in knowing why, of all the well-known warriors, you are the chosen one.” Having been dissatisfied by Nameless's claim that he has defeated all the other warriors in order to “come so close” to the imperial seat, the King commands Nameless to tell an alternative version of the stories. As a result, this second version is best understood not as a presentation of the truth, but as a fantasy that the King desires to hear in order to reaffirm his preconceived notion of reality. Moreover, Nameless, being caught by the King in his palace and is awaiting his final execution, is merely a post-mortem abject who is kept alive by the King in order to tell stories (I borrow the term “post-mortem” from Thomas Elsaesser; see European Cinema: Face to Face with Hollywood [Amsterdam: Amsterdam UP, 2005], 125; see also, Juila Kristeva, Powers of Horror: An Essay on Abjection, trans. Leon S. Roudiez [New York: Columbia UP], 1982).

Read in this light, Hero is, first and foremost, a parody of the position Zhang Yimou took when he decided to direct the film: a storyteller who is deprived, politically, of his agency of storytelling. Not only that, Nameless is epistemologically confined not only by the King alone, but by the thousands of physical and CGI-generated soldiers, who demand his execution of the story, and eventually, of his biological body. In this sense, the storyteller, technically, does not “propagate” the ambition of the state; rather, the spectators from across the entire political spectrum expected his story to confirm their collective ambition, of which “state power” is merely a symptom. Put it simply, Hero is not so much a propaganda that speaks for the biopolitical power of the state than a film that is about this power.

The excessive use of CGI's in Hero, though aesthetically tasteful, responds to the spectators' demand (not unlike the King of Qin's) for a collection of images that corresponds to a database of images to which they have already had access in parallel media. The way these CGI's are related to the optical images, and the way they can be re-connected with the film's position in cinema's teleology can help us understand the ethical stance it takes. Hence, let us study one of the film's sequences closely.

In a fight between Nameless and Sky, Zhang Yimou borrows from the famous martial arts novel Shujian enchou lu (Book and Sword, Jin Yong [b. 1924], 1955) a mental combat between the Qianlong Emperor (r. 1735-96) and his fictional brother Chen Jialuo over a performance of qin (zither). Shot with a mixture of black-and-white and colour footage, and the employment of physical special effects (including the use of slow motion, wires, and rain machines), this part of the sequence uses optical cinematic techniques that create an image that confirms the spectators' preconceived notion of a mental combat (i.e. a mental combat as described by a martial arts novel).

At the climax, however, the fight is supposed to turn from being mental to physical. Ironically, the cinematic device now turns from optical to digital, with computer-generated raindrops that make visible the jianqi (the chi of sword), a concept originated from the martial-arts manhua (manga) in Hong Kong, popularised by the comic artist Ma Wing-shing (b. 1961) in his manhua series Fung Wan (Wind and Cloud, 1989-; figure 1), and first referred to cinematically in the film Swordsman (King Hu and Tsui Hark, 1990). How are these CGI's edited together?

Figure 1. Ma Wing-shing, Fung Wan (Wind and Cloud).

This sequence, I argue, does not present us a cause-and-effect chain (as in a classical Hollywood sequence); rather, it is organised around affective impacts, sometimes at the expense of maintaining the illusion of spatial and temporal continuity. In this sequence, the film first cuts to a close-up of a sword modelled upon the one used by Fung Wan's protagonist Bou Keng-wan (Cloud; figure 2), whose jianqii is made visible by the circular movement of raindrops following the sword's movement in slow motion. The film then cuts to an extreme close-up of Sky's eyes (figure 3). This is followed by a close-up of Nameless passing through Sky's wall of jianqi (figure 4). The film then shows a three-quarter shot of Sky moving his sword in a circular motion, whose jianqi is made visible by the circular movement of the raindrops (figure 5). The film then cuts back to the close-up of Nameless, passing through layers of jianqi, each being represented by a “wall” of raindrops (figure 6). The film then “jump-cuts” (not strictly, but technically) to a close-up of the sword passing through Sky's jianqi (figure 7); it then cuts to a long shot over Sky's shoulder, in which we see Nameless's body piercing through the walls of jianqi, and another jump-cut into a three-quarter shot (figures 8-9). The film then cuts to a medium close-up of Sky to show the impact of the action (figure 10). Now, the film cuts back to a repetition of figure 9 (figure 11). Without any pretence to create continuity, the film cuts to an extreme close-up of Nameless cutting Sky's sword into two halves (figure 12).

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

This sequence therefore concatenates a series of impacts, made visible and sensible by the CGI's and jump cuts, an organising principle not only of the martial arts film, but its immediate predecessor the martial arts novel (and its theatrical performances). Nonetheless, with its overt references to a sensorial database from its literary, manhua, and cinematic sources, these “impacts” are no longer there to induce affective responses; rather, they confirm the information that has long been deposited in the spectators' minds as cinematic and cultural memories. As a result, the sequence is not so much about mimesis; rather, it responds to the spectators' conceptual expectation: “Please re-configure our memories. Show us something that we cannot imitate, or better yet, something that Hollywood has yet to imitate.”

The interesting point here is that Zhang Yimou has planted within this sequence a subtle teleological reversal: a mental fight represented by a physical image; a physical fight represented by a digital (conceptual, non-physical) image. Is there an intersection in this seemingly oxymoronic relationship?

We need to bear in mind that in the terms of “martial arts novel philosophy,” a physical fight is merely a mimesis of the mental fight. In other words, the CGI sequence offers a mimesis of the mathematical relationships that are ontically and ontologically prior to its physical manifestation (similarly, a CGI is the product of the mathematical relationships that constitute it). Interestingly enough, the same can be said about the mental fight that is represented physically. The mental fight, manifested physically by the optical image, is in itself a mimesis of the mathematical relationships of the choreographed action.

This teleological reversal has therefore made visible the common mathematical root of the martial arts cinema, whether optical or digital. I will explain it via two historically distant, but perhaps archaeologically proximate routes.

First, an epistemological one: In Ethics, Spinoza has raised an epistemological doubt. For him, if the universe is constituted by one substance (with “God” as its origin), how can we explain, on a conceptual level, mathematical differences of the same substance (e.g. the difference between one man and two men would then be a numerical, not substantial difference)? If we push this doubt further, what is substance besides a set of numerical possibilities (Part 2, Proposition 44, Scholium, 152)? In response to this, Leibniz proposes that “God” is merely the Possibility from which mathematical relationships are generated (see Gottfried Wilhelm Leibniz, Discourse on Metaphysics and Other Essays, trans. Daniel Garber and Roger Ariew [Indianapolis: Hackett, 1991], Section 9, 9). Second, an ontological one: Where shall we find martial arts and cinema in a “Chinese encyclopaedia?,” i.e. can we take Foucault's wild experiment rather literally and see what it suggests (As suggested by Foucault, Les mots et les choses [1966; Paris: Gaillimard, 1999], 7)? In the Wulin jiushi (Old Affairs of Wulin, c. 1290, written by Zhou Mi, 1232-98), a Yuan-dynasty encyclopaedic work on folk cultures in Wulin (now Hanzhou, a commercial centre during the Southern Song dynasty), martial arts and proto-cinematic arts and crafts (e.g. the shadow play, the opera, the novel and its narration, painting, [ahem ... ] even hairdressing) are, curiously enough, being categorised within the same ontological category (Zhou Mi 周密, Zengbu Wulin jiushi 《增補武林舊事》 [Enlarged and Supplemented Edition of the Old Affairs in Wulin] [c. 1290; Taipei: Shangwu yinshu guan 商務印書館, 1982], chapter 3, 10b). Their affinity is their shared interest in making sensible mathematical relationships that would otherwise remain invisible and insensible to the spectators (to this list, therefore, we may add music).

In this light, cinema may not be ontologically rooted in the mimetic arts of painting and theatre; but as our old friends Plato and Aristotle have doubted, these mimetic arts have their ontological origin in music (and mathematics). It is also in this light that we can now say that the CGI can be put side by side with the optical image as both ontologically related to the quest for the mathematical origin of the order of things. These archaeological sites neither suggests that it is “business as usual,” nor do they suggest that with the advent of the CGI, there marks an ontological shift simply because of a technological change; rather, a new teleological picture can potentially be drawn by pushing further Bazinian ontology: What ontical-ontological “origin” does cinema try to put under erasure, a teleological “movement” that has so far been manifested by the human desire to defy death? (For further references of this ontological question, see Elsaesser, “Early Film History and Multi-Media: An Archaeology of Possible Futures,” in Wendy Hui Kyong Chun and Thomas Keenan, ed, New Media, Old Media: A History and Theory Reader [London: Routeldge, 2006], 13-26).

The teleological reversal in the sequence of Hero can be understood in this light, and can be made clear in ludological terms. In the fight sequence between Nameless and Sky, the “mental fight” represents a series of predictable outcomes (in their minds) within the limitation of the physical technology (hence, the use of physical and optical special effects). The “actualised” series of events, however, is determined by the random interventions between pure strategies, and the database of pure strategies is mirrored by an equally large sample space that is made available by digital technology.

In fact, with the aid of ludological interpretations, we can open up an ethical riddle planted within the film itself. Not unlike any other martial arts film, Hero can be understood in terms of a tension between predictability and unpredictability, co-operativeness and in-coperativeness, which lay out the paths of possibilities that the game can potentially follow. Nevertheless, in this game, there is one predictable element (mathematical constant): the King ultimately wins. The lure of the game for an artist, given the poiltical confines, is therefore not so much about how close one gets to the king, but how one maintains the maximum creative space in a perpetual state of checkmate.

What Zhang Yimou implies is that the state can exercise its biopolitical power by trapping the storyteller between two deaths. Nonetheless, precisely because of the storyteller's até, the storyteller, either as a martial artist or as a filmmaker, can still assert a profound relationship with a mathematical possibility that turns around a manifest state of political totality, albeit on an abstract and individual level. As Sima Qian argues, an assassin may not be a very ethical figure, but within the rather unethical political tension, he remains loyal to his “will” (Sima Qian 司馬遷, Shi ji 《史記》 [Records of the Grand Historian], vol. 8 [Beijing: Zhonghua shuju 中華書局, 2002], chapter 26, 2538). In one register, Nameless, as an assassin, seems to carry out the “will of the people”; but more important, what the “people” fail to recognise is precisely the individual “will” of the assassin himself, which surpasses the “will of the people.” By telling fantasised accounts of what “the people” want to hear, the artist is still free to maintain her/his relationship with a Possibility that stands prior to the constitution of the “people” as a totalising community.