When we use a toaster, or a sweater, or a word processing software package, we have certain functional expectations. A toaster should caramelize bread evenly and consistently. A sweater should keep a body warm without fraying or stretching out from repeated use. A word processor should help automate the crafting of documents without requiring specialized expertise.
Some of our expectations of such objects are cosmetic. We like our toasters to match the décor in our kitchens, our sweaters to be woven with the colors and styles of the current season.
But the history of software as a tool for work has made most cosmetic demands for software relate to matters of usability: buttons and menus should be in convenient locations, actions should feel consistent and predictable, conventions set by previous iterations of a software package should be respected, even if lightly refined.
In the field of human-computer interaction (HCI), these values of software design are sometimes grouped under the term “transparency.” A good software tool, like a good toaster, is supposed to show us exactly how it should be used and then meet our expectations as users immediately and consistently.
Jay Bolter and Diane Gromala have suggested a different way to look at software, especially software that seeks to explore ideas rather than to serve as tools. Bolter and Gromala point out that the concept of transparency casts software as a window — a clear surface that seeks to disappear as it reveals a functional affordance.
This conception works well for tools but poorly for art. Instead, the two suggest another metaphor, a mirror. Unlike a window, a mirror’s job is to reflect back on its users, to give them a new perspective on themselves and their place in the world.
Video games are software, but they are not meant to serve the same function as spreadsheets. They are not tools that provide a specific and solitary end, but experiences that spark ideas and proffer sensations. Sure, video games have interfaces, like toasters have browning levers, like sweaters have cuffs, like word processors have font menus.
But too often we mistake the demands of these interfaces (and the in-game actions they facilitate) with the actions of tools. We gripe when a game doesn’t do what we expect, rather than asking what such an unexpected demand means in the context of the game.