While still studying at university, I saw a Taiwanese film on a whim during the 2014 edition of the Melbourne International Film Festival. Coming in blind, I didn’t know of the director or his body of work and was essentially just intrigued by some of the promotional images and the film’s description. That film happened to be Tsai Ming-Liang’s Stray Dogs (orig. Jiaoyou, 2013), a memory still treasured as one of my favourite cinema experiences. It was a divisive watch amongst the audience, there were many walkouts, perhaps the most I’ve seen during a screening, but I sat transfixed by its strange and uncompromising vision. It wasn’t my first flirtation with “slow cinema”: I had seen Edward Yang’s Yi Yi (2000) and Chantal Akerman’s Jeanne Diellman, 23, quai du Commerce, 1080 Bruxelles (1975), but it was certainly my first cinema experience of this type of filmmaking. Stray Dogs may be considered a less than ideal entry point to the work of Tsai: at this point, narrative is pared down to an abstract minimalism, continuity editing is disregarded, there may be multiple actresses playing the role of one character, and the duration of takes is pushed to the extreme with the newfound adoption of digital cameras. It was these bizarre idiosyncrasies and uncompromising decisions that simultaneously confounded and intrigued me. After my first experience with a Tsai film, I didn’t know what to think of it, all I knew was that I was deeply moved by an unusual directorial voice that I didn’t fully understand. So began my obsession with the work of this great director.
In the ensuing years, I made an effort to acquire and watch all the other Tsai films I could get my hands on, beginning first with his features and then extending to the various shorts and other productions he has created (a missed opportunity to see his VR project The Deserted (orig. Jia Zai Lanre Si, 2017) at MIFF 2018 is sorely felt). I discovered a filmography of startling consistency and originality, inspiring a general inability to articulate any rational criticism toward the films. Of course, there are preferences and favourites, but apart from a rare minor dud, such as the redundant The Skywalk is Gone (orig. Tian Ciao Bu Jian Le, 2002) we are talking about critical judgements of minor variation. In other words, my usual rating of a Tsai work generally only consists of determining whether it is “great” or in the upper echelon of “excellent”. That is, until I saw Goodbye Dragon Inn (orig. Bu San, 2003).

My initial experiences of Goodbye Dragon Inn left me feeling a little unworthy and sheepish with my Tsai Ming-Liang fan club membership. This is a film whose critical reputation has grown over the years in some circles as Tsai’s definitive contribution to cinema. Hell, even the man himself cheekily voted for it as one of his own submissions in Sight and Sound’s 2012 Greatest Films of All Time poll. A Tsai film is characteristically sparse, languid, with “nothing much happening.” The inexplicable magic of his filmmaking is to somehow make this moment-to-moment mundanity interesting, and ultimately, emotionally affecting. Yet my first watch of Goodbye Dragon Inn was only one of intermittent interest that contrasted with… well, boredom. Perhaps the greatest disappointment in my initial experience was that I simply wasn’t that moved by the film, a blasphemous statement in regards to a Tsai film.
Sometimes we are not in the right mindset to receive a film, we are too distracted, too tired, or desiring a different sort of cinematic expression than the one a film is offering. I considered my initial impression as a curiosity, a possible off night that hampered my enjoyment of the experience, or perhaps an honest reaction to a film that may not work. I treated it as a film to revisit in the future in order to re-evaluate my assessment.
A few years later, the second viewing proved even more disastrous. Not only did I have a similar response, I began to fall asleep and had to turn off the film midway. Now there were two cinematic sins committed in the House of Tsai. What is it about this particular film that didn’t work compared to all the others? The approach is relatively the same and even the runtime stands as one of his shortest at a concise 83 minutes. What is it about this particular film that makes it a more challenging watch?

Let’s talk about the subject matter: Goodbye Dragon Inn, as the name suggests, is about people in a rundown cinema watching King Hu’s Wuxia classic, Dragon Inn (orig. Lóng Mén Kè Zhàn, 1967). That is to say, the film literally consists of watching people watch a film. Minimal plot and action are hallmarks of Tsai’s cinema, but even his other pared down works have contrasts in events, time, and location. There is contrast of action between the characters: a Japanese man awkwardly uses the cinema as a location to cruise; Chen Shiang-chyi performs various maintenance tasks as a ticket lady; Miao Tien and Jun Shih appear as themselves, watching their own youthful performances in Dragon Inn. However, there is largely no variation of action within these characters: the Japanese man is always trying to cruise, the ticket lady is always performing her duties, the actors are always watching the film. It is easy to view the events as repetitive and largely uninteresting.
Another difference among Tsai’s works is the approach to intimacy within the character ensemble. A standard Tsai film will consist of a small cast of around three to four characters and create an intimate portrait by sustained screen-time of their everyday experience.
By contrast, Goodbye Dragon Inn has a loose and ephemeral approach: characters appear briefly and then leave; interactions between characters remain minimal or unreciprocated; even Tsai’s leading man, Lee Kang-Sheng, appears only in a minor role near the conclusion of the film. Chen Shiang-chyi, as the ticket lady, features as a protagonist of some sort, the only character to be present throughout the entire duration of the film, yet even she has no spoken dialogue or any interactions with the other participants at the cinema. The characters here can feel more removed and at a distance, their plights minor in the broader scheme of things, their presence cut short within the narrative, so it is not hard to be unmoved by the proceedings.

An interesting opportunity arose recently when the newly reopened ACMI cinemas hosted a screening of the film to coincide with Nick Pinkerton’s new critical book on the subject. Goodbye Dragon Inn already had two strikes to its name by this point, but a chance to see a Tsai film projected onto the big screen was too good to miss. Even if it were another confirmation of disappointment, at least I would be able to bask in the cinematic glow of the recent 4K restoration.
Thankfully, such apprehensions turned out to be baseless. Let it be known that Goodbye Dragon Inn is in fact a great film. It may have taken me a few tries, but by golly, the film finally worked. While it can be argued that any film benefits from being seen in an actual cinema, if there is a film that immensely improves from the experience, it is certainly Goodbye Dragon Inn.
The principals of Tsai’s filmmaking are evident throughout his other work, but witnessing his virtuoso command of aesthetics projected onto the big screen allows a greater appreciation for just how effective they are. Tsai’s filmmaking primarily comprises of long takes, often using wide shots to frame his characters in relation to the environment. Close ups are used during significant moments, but are used sparingly to maintain their impact. Camera movement is minimal but does occur occasionally: a subtle pan to keep a character in frame, but predominately, he maintains a still image and a character’s emergence or disappearance within the composition usually acts as the sole movement in the scene. The framing and lighting by the cinematographer, Pen-Jung Liao, is impeccable. There’s an idiosyncrasy to Tsai’s camera placement, less emphasis on symmetry in favour of slightly off kilter composition by placing his camera in positions either elevated or beneath his characters. He also plays with the use of deep focus, as his characters sometimes occupy a very minimal part of the image or are obscured in some fashion.

In some ways, the manner in which Tsai creates his films encourages more active participation from the viewer. In the absence of conventional script and directorial cues, it’s up to the viewer to find engagement within this narrative sparseness. During this watch on the big screen, I found myself vigorously scanning the image of each scene, pouring over each detail of the mise-en-scène and the placement of characters inside the frame. The application of long takes feels purposeful and intuitive, the extended time allowing the audience space to examine the full density of his compositions.
These compositions are accompanied by equally dense sound design. In a film that eschews dialogue for the majority of its runtime, sound plays a significant role in conveying additional environmental detail. Throughout the film, we constantly hear the sound of rain and leaks in the derelict hallways, the ambient sound of Hu’s Dragon Inn playing in the background, the whir of the film projector in operation and the miscellaneous noises created by the audience members. The attention to detail and complexity of this ambient sound can appear to connote a type of “realism”, but Tsai also plays with sound as a formal device: for example, a particular surreal scene involves the sound mix becoming completely dead with the exception of the noises made from a lone lady eating peanuts.
One of the realisations during this watch of a film where “nothing really happens” was that the film, as a matter of fact, constantly has things happening: in particular, the parallel projection of Dragon Inn within the film, which is often unseen but is continuously heard. An interesting moment occurred during a long take of Chen Shiang-chyi staring at a pork bun absentmindedly in the projection room – just like the description reads, it is not the most exciting image to imagine, the duration of the take held to the point where my mind began to wander as well. But it was in the process of wandering in which I began to become strangely aware of the mechanical noises of the projector, the sounds of film within the film being played – sounds initially perceived as just background ambience. The interesting aspect of the film is how Tsai affords time for these moments of contemplation and unexpected revelations.

Perhaps the most interesting aspect that is evident when watching this at a cinema is the third parallel of action that occurs. In a bizarre way, the usual immersion breaking behaviours of your fellow audience members, be it coughs, the sounds of people eating, or people getting up to go to the bathroom, adds an extra layer of humour. The film, which is about people watching a film in the cinema, actively makes you self-aware of your own act of watching, the events on screen essentially being mirrored by your own participation. The lady sitting next to me who involuntary twitched her head at the sound of coughs and other audience noises became like an additional side character to the film.
The last aspect that was enhanced by seeing this in an actual cinema was how the film encourages the audience to contemplate the space they are in. One of the final images of the film is a shot of the empty cinema, the house lights on, and aside from Chen Shiang-chyi doing one final check up before leaving, is entirely absent of people. It is the last moment of this cinema in operation before it closes for good. It caused me to reflect on the memories at this particular cinema, recently reopened after being closed for two years, first due to renovations, then due to COVID-19. It made me remember the various old cinemas of my past that have since closed down. It is a bitter irony that the cinema in which I first saw Stray Dogs has now been demolished and converted into gentrified apartments. All things, inevitably, come to an end. But as the film concludes, Tsai leaves less on a note of despondency but one of a bittersweet touch through the use of an old Cantonese chanteuse ballad – itself a product of an era long passed. The formative places in our lives may not always exist, but at least we have the memories.
What a lovely piece. I love the contextualising at the start and you leave me, at the end, wishing I’d seen this movie… but also not wanting to watch it anywhere but a cinema.