The above video and below written essay comprise my final project for one of Dr. Laura U. Marks‘ arts and culture classes from this past spring. Both are reproduced here with little tweaking; both only begin to scratch the surface of this topic, and while I’d like to devote more time to a film by film analysis of the mass battle on film as its evolved over the decades, that’s probably something more suited to an actual graduate study. In the meantime, I hope that this whets the appetite for deeper discussion and raises questions about the digitally-aided mass battle — and its stunning rise and fall in recent Hollywood moviemaking.
Tom Gunning characterized the earliest, pre-narrative films as the “cinema of attraction.” This was filmmaking that was exhibitionist in nature; it was eager to show audiences both the real world reproduced on screen, and fancy tricks only available to the cinema (such as the visual effects employed by Georges Méliès). After 1906 and the rise of the narrative as the dominant framework for filmmaking, the cinema of attraction did not disappear but went underground “both into certain avant-garde practices and as a component of narrative films.”
What drove home the immersive effect of these early films on their spectators? Using the Lumiere Brothers’ “Train Arriving at a Station,” Akira Lippet answers that question by drawing out the spatial impact of the cinema of attraction as described by Gunning: “The image-become-life – and not the train – threatens to breach the space of the spectator and swallow him, her.” The cinema envelops its viewer and draws them into an interior psychology that belongs to the cinema surface alone. With the advent of the narrative, would this “swallowing” ability be retained in any form?
Over 100 years later, the cinema of attraction is still alive and indeed functioning as a component of narrative films. It survived over the decades, attracting audiences with each new innovation after another: the introduction of synchronized soundtracks, the creation of widescreen, 3-D, the rise of blockbuster practical effects, and more recently, the seemingly endless development of digital visual effects.
The rise of digital effects to become the cornerstone of the Hollywood blockbuster suggests that the swallowing power of the image surface has indeed survived and evolved with the cinema of attraction. What’s intriguing about the current period is the sense of reversal accompanying the application of artificial effects. When the cinema was born in 1895, it immediately found a foothold with films based in reality. The “actualities” were a raging success for simply showing audiences elements of real life recorded by the camera. Once Méliès began plying his trade, though, the cinema-as-magic-show model took off and became the most prevalent type of attraction until the narrative took over. The cinema of attraction’s means of enthralling transferred from that of total realism, to that of artificiality. The digital effects-dependent film of the late 20th/ early 21st century has seemed to take the opposite path. What began as an obviously artificial creation (as seen in The Abyss or Terminator 2) has become the most indispensable feature of the 21st century’s blockbusters, not least because of efforts to make such effects appear more realistic.
As Lev Manovich points out, though, perceiving what is real is not the same as perceiving that which is photoreal, and it is only to the photoreal that the current effects-driven cinema really aspires. The drive towards realism, of creating a seamless composite of the animated and the filmed with no visible seams, has itself been mirrored by recent efforts to craft blockbuster films that attempt to show the usual spectacles they’ve always shown, but without the glossy production values symptomatic of the Hollywood picture. As the modern war film was turned upside down by the unhinged documentary aesthetic of Saving Private Ryan, and the horror film adopted a video-diary style (pioneered by The Blair Witch Project) so the epic period film, which we will be focusing on, embraced incoherent, hand-held visual styles, cold palettes, and graphic violence. The past would be shown the way it was, warts and all.
In my video essay “Battle Surfaces,” I chronicle the visual progression of the epic battle sequence in cinema. For the purposes of my study, I’m focusing only on depictions of battles where the use of gunpowder and firearms is either completely absent, or very minimal. The blueprint of the standard epic-scale battle sequence is laid down in Alexander Nevsky (1936). Sergei Eisenstein’s combination of wide shots of crowds charging each other with close-ups of hacking and slashing, filling the frame with constant busyness, set the standard for how a battle would build peaks and valleys of movement and emotion. This is a pattern repeated again and again over the decades, particularly in Hollywood’s 1960s sword-and-sandal epics (Spartacus, El Cid, The Fall of the Roman Empire), and continued with updated brutality in Braveheart (1995) and Gladiator (2000). What all of these films have in common, is the use of real human extras to populate the wide shots. There may be hundreds (or in the case of Spartacus, thousands) of men on screen, but most films attempting to recreate a battle involving tens of thousands must rely on the power of implication, of the sense of offscreen space created by the moving frame. Such attempts at conveying the sense of battle were not helped by limp fighting styles (see the 60s’ films for examples) and clean coverage of the action. Every angle of every battle shot could be attributed to a human perspective of some kind; even the usual crane shot carries an implicit knowledge of limitation within itself, rarely rising high enough or moving forward in a way that doesn’t necessarily preclude human movement.
When The Lord of the Rings stormed onto the screen in 2001, it opened with a mobile aerial shot of a battlefield where tens of thousands of creatures engaged in battle. The camera swooped like a bird over a landscape that could not exist, recording millions of discrete, intense actions that could not have all been carried out perfectly by human extras. The LOTR films went on to claim that image – the aerial, computer generated shot of thousands of computer programs – as their major legacy. A spate of similarly epic-sized films centered on huge battles soon followed. The epic battle was reborn.
The aerial CG shot offered two things to audiences. First, it provided the “wow” factor that the epic film, dormant since the end of the 60s,’ needed to draw crowds back to theatres. Filmmakers could now let the audience feel like gods, floating high above the action while enjoying the sheer scale of the image. This type of shot offered a higher-resolution, more dynamic, and more spatially-developed experience of the same basic camera angle developed by real time strategy (RTS) computer games, in which players assume god-like sight and managed armies and cities. The thrill of the entirety of a humongous spectacle could be savored, and the montage could stitch multiple takes of this panorama together to make something truly operatic and geographically relatable. And it all looked so photoreal, carefully blending thousands of digitally animated extras with the background, the film grain, and the colour palette. In cutting his battle scenes, Peter Jackson followed the example of Gladiator and Braveheart by covering the action with hand held cameras, cutting swift and often incomprehensible movements together before cutting back to the CG wide, stabilizing the scene in its vast steadiness.
But the CG aerial shot also offered audiences the tools for defusing its awe: it was not just its vastness alone, or its complexity of visual content, but the combination. It was too perfect. No real attempt at filming on such a scale could yield the right combination of well-timed choreography, frame size, and camera movement. And so almost as soon as it changed the nature of huge action scenes, it could already be defeated. The first few times it was used, it carried the potential of Lippet’s swallowing surface, both colliding with the viewer and sometimes enveloping them with the promise of a wider expanse beyond the frame; however subsequent over-use in post-LOTR films exposed the weaknesses of relying too much on the mass CG shot: it corralled such large elements (an army) into the frame and eliminated the centrifugal power of the shot. In addition, the behaviour of each soldier was bang-on. Not one extra was caught messing up. This is, of course, because the soldiers were animated figures operating under individual AI programs, and it is this algorithmic efficiency that calls attention to itself so readily.
My video essay intends to show the progression of the battle as filmmakers sought to make each one more convincing than the last. Limp choreography in the 60s’ gave way to fierce and graphic bloodshed in the 90s; with that problem solved, the question of conveying the scale of a real battle became the next obstacle to achieving total realism. And as we can see in shot after shot from the last decade, such realism is elusive when simulated by a computer.
Phenomenonologically, these aerial shots fail to genuinely engage the senses. The visual element is compelling, but the use of sound is often what gives away the artificiality; if we were to reach out and use every sense to feel the depth and scale of a wide aerial shot, we’d experience a distant perception of the battlefield’s din. These shots, though, are inevitably paired with a stirring score, which overpowers the ears; the sound mix is often turned up to keep excitement high, but to such a level that would be unrealistic if heard from so high. What these shots accomplish is much more metaphorical than literal; the sight of an army of good sweeping in to save the day says more about the emotional significance of the scene than it does of the sensual significance.
Add to these problems the issue of the drive towards realism affecting the types of stories being adapted after LOTR; the very approach towards even myths such as The Iliad, with the film adaptation (Troy, 2004) refusing to include supernatural elements and opting for a straight warfare-focused rendition, is telling of the urge to press the aerial shot into the service of realism. In that spirit, filmmakers kept expanding the height of the shot, turning little figures into little dots (Alexander did this to limited effect in the battle of Gaugemala sequence), perhaps in an effort to hide their artificiality. As the decade wore on, such large-scale shots either took the super wide approach or disappeared altogether (the fantasy-war film 300 is noticeably bare of such shots, for a film so focused on a massive army’s attack). By the time Ridley Scott made his third historical epic of the decade, Robin Hood (2010), he’d eschewed the wide-scale CG shots used on Kingdom of Heaven (2005) and, save for a few kilometers-wide panoramas, relied on practical footage to edit his climactic battle sequence; while the wide panorama used to be all about savouring the climax of two forces crashing together, its appearances in Robin Hood limit it to being merely a geographic tool. The battle scene came full circle; the director credited with reviving epics in 2000 effectively performed the funeral for the extravagant effects-driven epic ten years later. The general sentiment in this particular genre and other effects-heavy fields seems to be favouring a slow return to more practical effects work.
One other element may have contributed to the swift decline of the epic digital battle: easy public access to filmmaking documentaries (included on DVD releases of the epic films themselves) that expose the secrets of such sequences and demystify the image-making process. Once one sees how “easy” it is on a physical scale to create such sights, the “wow” factor may indeed decrease. A great part of the appeal of the old epics was the sheer pleasure of seeing so many real moving parts working together and knowing that someone had pulled off a great feat making it all work. The digital battle scene does not lack comparable expertise, but its creation within a world of code makes the achievement far less glamorous.
What does it all mean? The final movement of “Battle Surfaces” is called “Apatheosis,” which attempts to sum up the effect of so many digital battle scenes unleashed on the public over such a short span of time: apathetic climaxes. One money-shot looks like another, and when they are taken out of their original context, one can see how similar they tend to be. It’s intriguing to consider the possibility that such shots actually might serve a pacifistic agenda; though designed to awe and excite, the way in which such wide views reduce a body of humans to amorphous, algorithmic mob serves as a warning of sorts to the dehumanizing aspects of warfare, and of film as war propaganda.
But overall I want to emphasize the fact that these shots began life ten years ago as the newest and most exciting iteration of the cinema of attraction, and now find themselves virtually defunct, as far as Hollywood is concerned. The drive for realism has left them behind, and their ability to swallow the viewer has been exposed as shallow. As far as surfaces go, these ones have very little depth. Removed from their proper context, they have no purpose and no place to gather or extend energy. On the more modest scale of that first shot in Jackson’s Lord of the Rings, they contained some of the raw energy of the earliest cinema, and had a frame that was limited enough to point beyond itself. Once the filmmaking industry grabbed the concept, it grew, tired out audiences, and drew the outside into the frame. It gobbled up itself, not the viewer. And thus, it has to go.
Gunning, Tom. “The Cinema of Attraction: Early Film, Its Spectator, and the Avant-Garde.” Film and theory: An Anthology. Ed. Robert Stam and Toby Miller. Malden, Mass.: Blackwell Publishing, 2000.
Lippit, Akira Mizuta. Atomic Light (Shadow Optics). Minneapolis, MN, USA: University of Minnesota Press, 2005.
Manovich, Lev. The Language of New Media. Cambridge, Mass. : MIT Press, 2001.
 Gunning, Tom. “The Cinema of Attraction: Early Film, Its Spectator, and the Avant-Garde.” Film and theory: An Anthology. Ed. Robert Stam and Toby Miller. Malden, Mass.: Blackwell Publishing, 2000. p. 230
 Lippit, Akira Mizuta. Atomic Light (Shadow Optics). Minneapolis, MN, USA: University of Minnesota Press, 2005. P. 65
 Gunning, p. 231
 Manovich, Lev. The Language of New Media. Cambridge, Mass. : MIT Press, 2001. P. 200
 The type of battle scene that enjoyed such a lusty revival in the first decade of the 2000s was almost exclusively pre-Renaissance.
 Lippit, p. 73
 Another possibility: RTS games quickly caught up in graphics resolution and performance to mimic the best of the mass aerial shots, eliminating the need for players to vicariously experience the thrill in a theatre.
 Consider how the higher the perspective of the shot, less depth is perceived at ground level, crushing the armies flat against the distant ground.