Video Games as Art?

Half-Life CoverSo I’m listening to one of my favorite soundtracks — from the game, Half-Life. Video games are becoming more and more like cinematic experiences. (In many cases, they are being converted into really bad cinematic experiences, such as the Doom movie or Alone in the Dark, but that’s not my point right now.) In addition to soundtracks, video games like Half-Life have plots, scenes, characters, and dialog. A lot of this is rudimentary — the dialog, for example, is pretty limited, and character development is sparse — but it adds a level of depth and complexity to games that only recently were as simple as Space Invaders.

Still, as Roger Ebert pointed out last year, it’s silly to think they rival movies as story-telling formats:

“[V]ideo games [are] inherently inferior to film and literature. There is a structural reason for that: Video games by their nature require player choices, which is the opposite of the strategy of serious film and literature, which requires authorial control.”

Ebert got a lot of hate-mail from gamers for this comment, but I think he’s essentially correct that games are inferior story-telling devices, at least given today’s technology. The more interesting question is whether the loss of “authorial control” that Ebert correctly ascribes as the fundamental difference between a game and a movie makes games “inherently inferior” as narrative devices.

Half-Life and Half-Life 2 illustrate both my points and Ebert’s.


Take dialog. Given current technology, there is one huge barrier to making games more like movies, and that is the fact that the character you play is invariably mute. In other words, while other characters in the game can and do talk to you, and they converse with each other, you can’t respond in any way. (Half-Life 2 pokes fun at this fact early on, when Alyx Vance, a Non-Player Character (NPC), says, “You’re kind of the quiet type, aren’t you?”). Various games have attempted to work around this problem — dialog trees, for example, are a feature in many adventure games. But dialog trees are poor substitutes for the ability to speak; they often either fail to present a line you’d like to utter, or offer you choices you wouldn’t have thought of. Furthermore, either they present you with the same choices repeatedly, which means that the optimal game strategy is simply to utter all of the available options and see what the responses are, or they artificially limit your choices, meaning that you are presented with a false dichotomy of what statements you can make. It’s no wonder gamers find dialog trees frustrating. Ironically, one of the better implementations of dialog in a game that I’ve come across was the old Infocom text adventures from a quarter-century ago. In several, you could talk to characters by typing commands such as “ASK SMITH ABOUT THE REVOLVER”.

A related problem is that the non-player characters you encounter cannot react spontaneously to events. This means that their speeches must be scripted, leading to an odd imbalance between the fully developed dialog they utter on some occasions, and one-liners they utter in response to your actions. The shortness of the improvised utterances makes them more flexible — like horoscopes, they can seem appropriate in a wide variety of circumstances. (E.g., Barney, the security guard, as you fiddle with something: “Man, I could sure go for a cold one right now.”) There are other problems with character speech, too, such as the characters continuing their set speeches oblivious to what you are doing, such as hitting them or leaving the area. Artificial intelligence is difficult enough to achieve for moving characters from point A to point B; artificially intelligent dialog is probably many years away.

Technology is not the only problem with games. Despite all of the accolades about their plots, if we held Half-Life and Half-Life 2 to the same standard as movies, they would be universally panned. Despite great scene-setting at the beginning, both end with barely any resolution of what went before. It’s as if Die Hard ended with a big explosion but without any explanation of who the terrorists were, or why they had taken over Nakatomi Plaza, or who John McLane was, or whether he makes it out of the building alive. I spent much of Half-Life 2 worried that, like a character from a Philip K. Dick story, I was being used against my will to expose the Resistance. In other words, I didn’t even know my own motivations. It’s a weird variation on the mind-body problem, where you feel like all body and no mind.

But let’s say those problems were all fixed; AI develops to the point where players can converse with NPCs, and game designers get more serious about their plot development. Is the lack of “authorial control” fatal to games as narrative?

The 'Die Hard' SceneOne of my favorite scenes from Half-Life illustrates the issue, the scene I like to call the “Die Hard” sequence. After a scientific experiment goes awry deep underground, you spend hours desperately trying to reach the surface to get help for your trapped colleagues. At last you reach the surface, but instead of rescuers, you find it crawling with marines dispatched with orders to kill anything that moves. In most games, the “solution” to any puzzle like this would simply be to kill all of the enemies so that you can take your time in finding an exit. But Half-Life was different; the number of kills you rack up is irrelevant. The “Die Hard” sequence was one of the first times I felt so totally outnumbered in a game that I simply ran for cover. An energetic music track starts up as you emerge from underground, and when I reached this point I quickly found my only hope of survival to be a mad dash past the marines, through a maintenance door and down an old air shaft — reminiscent of a certain scene in Die Hard.

I sometimes wonder if I would like that scene so much if I hadn’t found the maintenance door right away, and instead died repeatedly trying to fight the marines. In fact, I found the door and escaped on my first or second try, and just kept scrambling down as the marines pursued me. It makes me also wonder if I missed the excitement of other scenes because, by the time I found the “solution,” I had repeated the scene so many times it became frustrating (the insanely difficult prison shootout in Half-Life 2 comes to mind). Granted, different audience members will react differently even to a static movie scene, but in this example, the way the scene unfolds may be fundamentally different according to the player. The way I did it, it was more like a chase scene. But someone who repeats it several times before finding the exit may experience it as more like a puzzle, or a long, grueling battle.

If you want to watch a video that illustrates this sort of scene, take a look at this clip of the epic battle against the striders in the bank plaza from Half-Life 2. Some, but not all, of the scene is provided by the game: the music (which you can hardly hear), the sound effects, the design of the bank plaza, the line of dialog spoken by the NPC. But the choices of where to look (i.e., the camera angle), where to run, when to fire, and what to fire are completely up to the player. The player’s choices have further effects on the scene; what the player does affects what the striders do, and what happens to the NPC. While Valve has designed an overall scene — a complex battle in a courtyard against the striders — the particulars of that scene are largely unspecified.

In these scenes, it’s difficult to say that the game designer’s authorial vision is coming to fruition. It may, or it may not; there’s a large amount of randomness introduced by the player’s actions. As a result, the game designer’s intent can only be expressed in the most general terms (“I sure hope this plays out something like that scene in Die Hard“). It’s not like Stanley Kubrick filming the “All work and no play” scene in The Shining 120 times until he gets it right; it’s more like if Kubrick rented the Overlook Hotel, placed the manuscript on the table, and gave the camera to you. The result wouldn’t exactly be a “Kubrick film.”

So if Valve didn’t author the “Die Hard” scene or the bank plaza battle, does the player? Is it joint authorship? Or is the author dead?

Those are difficult questions, and I’m in the process of working through them for an article I’m writing. But for now, I would just note that these are different questions than the one Ebert starts with, whether video games are inferior to films and literature as story-telling devices. They are certainly different; and there may be an inherent loss of authorial control. But I think it is possible that a collective effort at story-creation, such as video games, can entertain and enlighten just as well as movies and literature, even if those goals are achieved through a different process.

You may also like...

4 Responses

  1. Zvi Rosen says:

    Interesting post (and hopefully enlightening to those who look down their nose at games). Two brief observations, a bit contradictory, that don’t quite do it all justice:

    Firstly, the lack of authorial control in a game is something of a myth, and Half-Life is a great example of this (I’ve not played 2, owing to time and system constraints). Going through Half-Life, especially for the first half of the game which is in underground passages, you are on a fairly linear path that the game’s designer’s have set you on. The trick is that it doesn’t feel like you are on a path, even if on reflection you notice you could only have gone forward that way. This also enabled the use of scripted sequences…scripts only work if you that the player will arrive at a certain place having already completed certain tasks. This holds true in the “Surface Tension” level discussed in the post as well…while finding that vent may have seemed fortuitous, the game design was meant to push you towards it. In fact, Surface Tension is commonly regarded as one of the greatest levels also because it managed to keep up the, err, tension in an open, aboveground, area, and ramp it up by using the added vulnerability of being in the open. This trend may be changing with more action-oriented “sandbox” games such as GTA, but most first-person shooters keep you on a rail, even if you’re not quite conscious of it (which was part of the joke of the Half-Life level “On a Rail”.

    On the flipside though, I think the criticism that games are not like movies only seems reasonable since movies are the most superficially similar pop-culture item. However, this criticism points out not only the greatest weakness of computer games, but also their greatest strength. Movies are a depictionary medium…art in its conventional form. The artist’s vision is presented to the viewer, whose interaction with the work is purely on a reactional level. Games, by contrast, are (shockingly enough) an interactive medium, in which the gamer does not merely react to the scenes presented, the gamer’s actions lead to them. While this limits the absolute power of the artist designing the work, it allows a depth of engagement movies don’t allow, and while this depth is hard to a acheive, especially in an ultimately immature medium, it has capacity to engage us in ways movies dare not, and many games have already done just that.

    A game has already been made which was little more than a series of vistas with little more than reaction from the player…it was called Myst, and was a dead end. The game paradigm has tremendous potential…let’s hope it’s used for more than brainless adolescent amusement (which of course most movies are also).

  2. Kornsac says:

    It’s unfortunate that we look to games to tell stories in the same manner as books and movies. Strict authorial control cannot offer choice, consequences, and deep involvement to an audience.

    Deus Ex is an example of a game that uses these strengths rather well. It is thought provoking in a way that a purely narrative story cannot be. It challenges the player’s notions of liberty and morality precisely because it offers choice. It allows the player to directly experience freedom and moral choice, so the theme is much more powerful than it would be if these things were discussed in terms of the actions of other characters.

    Game designers are able to choose how much freedom to give players, and this choice may be artistic as well as technical. A game can well guide players through a preconcieved story, but many of the smaller details are left to the player, rather than the author. It makes an experience unique to each person that plays the game.

    Games may not work as well for “good old-fashioned” storytelling, but for a designer who understands the use of player choice, a game can make a much more effective presentation of ideas than traditional non-interactive media.

  3. Bruce Boyden says:

    Zvi and Kornsac, great points.

    I agree that the amount of player choice in, e.g., Half-Life is subtly limited so that the player has the illusion of having a lot more freedom of movement than they in fact do. This allows the game designers to plot the overall journey and add in scripted sequences — they will always know where the player has been, where they are headed, and which direction they’ll be coming from. And I agree Half-Life does this very well (ironically, I think the illusion of freedom is preserved better in Half-Life than HL2). So the game designers clearly have a great deal of control over the way the game unfolds. But when it gets down to details — whether the scene I described becomes a “chase scene” or a long, hard slog — the game designers have very little control, beyond a few nudges. Partly I’m approaching this issue from a copyright perspective, and if you think of games as audiovisual works like movies, which courts traditionally have, one of the things that matters most is what appears on the screen. The level of detail Stanley Kubrick was seeking in the Shining is just not possible with video games, unless like in Myst player input is limited.

    One of the developments I expect to occur is that games will have multiple paths in the future. Open-ended games have been around at least since Elite, but they tend to have very little structure or overall narrative. It should be possible to build into one-player games, however, the type of structure that exists in MMORPGs, or for that matter, pen & paper RPGs — set pieces surrounded by large expanses of random monster generation. I’ve never played Deus Ex, but I’ve always wanted to for that very reason — and because I was a big fan of System Shock 2 and Thief, done by many of the same people. While that will be great for gameplay, I think it would reduce “authorial control” even further. That’s not a bad thing; it just means games are a different type of narrative exercise than films and literature.

  4. Daniel says:

    If you liked System Shock 2, you’ll love Deus Ex. Kornsac is right about the way the game challenges your views of freedom and allows you to choose different paths. At the same time, though, I felt like it was one of the more “cinematic” games I’ve ever played.

    The soundtrack for Deus Ex is also top-notch.