Witnessing the Illusion
As video games achieve greater realism, Eugenie Shinkle asks what the difference is between virtual game worlds and those depicted by landscape photographers.
by Eugenie Shinkle
Issue 51 Summer 2007
View Contents ▸
Videogames have come a long way in the last 25 years. The minimalist aesthetic of early videogames - clusters of pixels or slender wireframe outlines superimposed on a black void - has given way to lavishly detailed and ever more lifelike gameworlds. Ironically, however, it is Hollywood cinema, rather than the real world, that provides the benchmark for realism in today's videogames. These virtual environments may meet the standards of the average cinemagoer, but would they hold up to the more exacting standards of a landscape photographer? Using the task of image-making to gauge the realism of a virtual environment may seem counterintuitive, but comparing the two scenarios in terms of both the picture-taking experience and the resulting images ends up revealing some interesting things. Though a gamescape may successfully mimic the appearance of a real landscape, imagine the experience of a landscape photographer in a gamescape and it describes a profoundly different kind of picture-making.
Both the British and North American landscape traditions are based in the structural and aesthetic conventions of seventeenth-century Classical landscape painting. Landscape artists from Thomas Gainsborough to Ansel Adams have drawn on the formal definition of the landscape view set out in the work of painters like Claude Lorrain, Nicolas Poussin, and Salvator Rosa. Under this definition, a landscape consists of layers of receding strata leading the gaze into the picture from the foreground, towards a horizon line located about halfway up the picture plane. Framing side sections (called coulisses), masses of light and shade, and atmospheric perspective - the diminishment of detail and colour saturation with distance - provide additional depth cues.
Claude Lorraine Landscape with Peace and Justice Embracing, 1654
The notion of movement into the picture plane is key to Western notions of what a landscape is, and what it does. Structurally, a landscape is built around rules for perspectival composition that date back to the Renaissance. The landscape view shares its three key variables - viewing position, horizon line, and vanishing point - with rules for composing pictures in linear perspective. These rules were first set down by Leon Battista Alberti in 1435, and instructions for composing landscapes were often included in early treatises on perspectival composition and stage design. A landscape is designed as an invitation to the gaze; it promises unrestricted movement into its space. Historically, the ability to accomplish this movement with the eye - to visually traverse the distance between point of view and horizon - has acted as an important means of commanding and controlling space.
William Gilpin, View of the Gothic temple at Painshill, 1772
Though 'landscape' is often used as a generic term for a depiction of nature, the term has a more complex history. Historically, creating a landscape has acted as a kind of coping mechanism for the Western imagination: a way of bringing shape and order to otherwise shapeless terrain, and of claiming agency over a natural world that appears vast, intractable, and essentially hostile to humankind. Landscapes created in this tradition are much more than simple representations of nature. They are, first and foremost, constructed and framed spaces - enclosed, measurable, and spatially homogeneous. Arguably, it is the linked notions of enclosure and artificiality, rather than any necessary relationship to nature, that defines landscape as a category.
Screenshot from the game Witcher, published by Atari, 2007
The layout and appearance of a typical gamescape borrows heavily from the Western landscape tradition. Visually, gamescapes mimic the appearance of Classical landscape painting. They are typically composed of a clearly delineated foreground, with nearby landforms rendered in rich detail, and surface texture, brightness, detail, and contrast diminishing with distance. This is partly a matter of structural necessity (nearly all video games are built on top of an underlying perspectival framework), and partly a matter of convention. Controlling nature is not an issue for the designer of gamescapes, however. Instead, they are concerned primarily with creating a familiar and easily-navigable terrain to act as a stage for game events, and then directing the player's progress through this space. While landscape imagery provides the first, it is the less well known conventions of eighteenth century landscape garden design - a discipline also concerned with the control of nature and the movement of bodies through space - that provides the second.
The issue of control is one of the most conspicuous differences between a real landscape and a gamescape. In a real landscape, the photographer is free to go more or less wherever they want. If the view from one position isn't quite right, the photographer can usually move to a better vantage point. Provided there is solid ground underfoot, movement through the landscape is limited only by the weight of one's equipment and one's willingness to trespass. Most games, however, don't allow the player anywhere near this level of freedom. With few exceptions, videogames consist of a series of levels which are meant to be completed in a particular order. Individual levels are small segments of a larger gameworld, which means that the part of the gameworld that the player can actually interact with is relatively small and completely self-contained. The resulting restrictions on movement mean that the gamescape bears more resemblance to a landscape garden than it does to real space. Landscape gardens typically consisted of a series of carefully planned views linked by gravel paths. Although its space was open to the gaze, the visitor's ability to physically move through the garden was carefully regulated. As William Shenstone noted, 'When a building, or other object has been once viewed from its proper point, the foot should never travel to it by the same path, which the eye has traveled over before. Lose the object, and draw nigh, obliquely.' Departing from the intended route was certainly possible, but not encouraged, as doing so might reveal objects that were not meant to be seen: outbuildings, untended fields, or other eyesores which would shatter the carefully planned illusion of nature that the garden presented. Likewise in a game, most of the player's visible environment is designed purely to be seen, rather than actively explored. As well as obvious pathways, game designers use strategically placed objects and lighting effects to direct the player through space while retaining, as much as possible, the illusion of independent movement. Were our gamescape photographer unsatisfied with the view from their present position, however, they would likely find themselves unable to reach another. Attempts to depart from the intended route are often blocked by walls, cliff faces, impenetrable objects, or invisible boundaries.
Screenshot from Halo, Bungie Studios, 2001
No matter how formal its design, a real landscape is subject to underlying geological, biological, and other physical processes. It has its own particular density and phenomenological presence - something that Lewis Baltz understood in his 1986 series San Quentin Point. Aiming his camera directly at the ground, Baltz disregarded the conventions of classical landscape and revealed the space in front of the camera's lens as something tactile and indexical - not a formal conceit, but a real place, bearing the marks of human use and misuse. The gamescape photographer would meet with something very different. With very few exceptions, gamescapes and the objects that fill them are designed and built entirely from scratch by wrapping texture-mapped 'skins' over wireframe matrices. Organic textures like rock, wood, and vegetation are particularly difficult to model convincingly, and surfaces that appear lifelike from a distance don't often hold up to close inspection: vegetation descends into meaningless visual noise, and supposedly rough textures like rock and wood reveal a wallpaper-like smoothness. Skinned objects and surfaces have no underlying object properties - they are smooth on the outside and hollow on the inside - and though they may occasionally be 'destructible', they are rarely malleable or deformable. Gameworld objects are iconic rather than indexical, and most of them register no trace of the photographer's presence.
Photograph from San Quentin Point by Lewis Baltz
Confined in a claustrophobic, insubstantial space, the gamescape photographer would also find themselves confronted with a perplexing array of spatial voids and discontinuities. Were a real world photographer to put their camera down in one place, they could, hypothetically, make their way from that position to features that appear on the horizon. In the real world, there is ground underfoot, and the interval between here and there is filled with solid, tangible things. This sort of continuous space is rare in video games, which have more in common with theatrical sets than they do with real space. Beyond the immediate foreground, objects and landscape features are mapped onto two-dimensional surfaces (called 'billboards') rather than rendered in three dimensions. Scenery to either side of the player's intended route may consist of little more than a series of two dimensional screens, stacked closely together to give the illusion of a more expansive space. The scale of distant features is often greatly exaggerated and the ground plane only intermittently present. Gamescapes are littered with spatial lacunae, heterogeneities, and inconsistencies - areas that aren't part of the experience and that can't be explored. The gamescape photographer could witness these illusions from a distance, but could never approach close enough to reveal them as such.
Theatre design for a woodland scene by Giacomo Torelli, 1643
Gamescapes, by and large, are not designed to be seen as still images. It is cinema, rather than the real world, that provides the benchmark for realism in videogames, and gamescapes have much in common with cinematic landscapes. Writing about the films of John Ford, Edward Buscombe remarks that 'Landscape in the cinema is never, or never for long, an object merely of contemplation. Narrative is all. In a film, landscape becomes scenery in another, theatrical sense, a backdrop against which the action is played.' Gamescapes perform a similar function. They are intended as theatres for gameplay, and most of the time that involves moving through the landscape (often at speed) rather than standing around contemplating it. Though gamescapes provide all of the structural attributes that should add up to a great, traditionally conceived landscape picture, they don't aspire to the condition of pictures. The gamescape has been framed in advance by the conventions of Western landscape, and by the game designer's intentions. Gamescapes are pragmatic, purposeful spaces; designed for a mobile eye - a glance rather than a gaze - they lend themselves to little more than stock re-presentations of Classical landscape. And although gamescape images may conform to the structural rules of Classical landscape, they lack the content that substantiates this form.
Screenshot from The Elder Scrolls IV: Oblivion, 2K Games, 2006
In fact, the kind of agency furnished by Classical landscape has little meaning in today's world. 'Landscape', as a term, no longer bears a necessary connection to nature, and the work of photographers like Ed Burtynsky, Jem Southam, Geoffrey James, John Davies, Richard Misrach, and others examines landscape as a process, a dynamic relationship between places and their inhabitants. Landscape representation is no longer concerned primarily with the command and control of space from a distance; these days, it is about the sort of political agency that is aligned with actions and gestures in the real world.
Photograph of Steamboat Springs, Washoe, Nevada, 1867 by Timothy O'Sullivan
Gamescapes may succeed in simulating the visual conventions of landscape but they don't come close to describing what it feels like to be in a landscape. As Merleau-Ponty observed, we inhabit space by means of the body; our experience of space is kinaesthetic and multisensory. The sense of being in a landscape, of feeling it, is instrumental in transforming static space into the more active configuration of place. It's also vital to the landscape photographer, for whom the act of photographing is, in part, an encounter with a place - and as such, it is a vital part of creating an image. Unlike a real landscape, which is a dynamic and evolving space, a gamescape is a static space. The player may move through the gameworld but they are intrinsically separate from it, and their actions typically have little lasting effect on their environment. Even the most realistically modelled terrain is no substitute for the sense of belonging that comes from inhabiting a space.
Screenshot from Myst Online: Uru Live, GameTap, 2007
Despite the introduction of new interfaces, like the Nintendo Wii, that address the player's body in more active ways, game design is still largely concerned with space as it is seen. There is still very little space for real bodies in the gamescape, which can be depicted or even imagined as a place, but cannot - at least not yet - be experienced as such by the player. And it's here that the difference between landscape photography and gamescape photography can be most clearly understood as the distance between represented and lived environments; between the limited and largely pictorial agency that is provided to the gamescape photographer, and the more complex forms of agency demanded of the landscape photographer in the real world.
Other articles by Eugenie Shinkle:
Other articles mentioning Lewis Baltz:
Other articles on photography from the 'Landscape' category ▸