You Must Register


Finding the “I” in Player-Avatar Identity

In a search for writing about the player-avatar relationship I stumbled across the 2011 Philosophy of Computer Games conference. A lot of the papers are pretty good.What follows are brief and simplified summaries of some of them, along with my reactions.

Eric Olson, The Metaphysics of Avatars and Their Relations to Players

  • There are no concrete, active computer game characters distinct from players.
  • Identities in games are similar to pretending to be or acting as someone. Distinctions between being someone and pretending to be someone are familiar to us; at the very least, they aren’t specific to videogames.
  • The avatar on the screen represents two different things: the player and the character. It’s the same in a puppet show.

Peter Day, Darn it, “I” Keep Dying! The Metaphysical Self in Computer Games

[This is a good follow-up to Olson. As said above, Olson thinks that videogame characters are two different things, the player and character. But he doesn’t say how or why. Day tries to give us a clear picture of what’s going on.]

  • Day thinks that games have two languages: representation and input. The world of representations has the language of movies or books. Input is another language entirely. These dual languages create a dual self in the game.
  • Day gives us two different modes of “I”. He builds off Wittgenstein’s Tractatus. Here’s a verysimplified version of what Day proposes:
    • Meaning requires a self for whom things mean.
    • Games communicate meaning in two different forms, or languages.
    • There are two different selves, or “I”’s in a game, for which things mean.
    • One “I” in games is the “I” of representation. The other is the “I” of input. The former is what we get from the images, sounds, cut-scenes, stories, etc. The latter is what we get from understanding that “w” translates to “go forward”. These are two different modes of meaning making, creating two different senses of self.

Rune Klevjer, Telepresence and Diegesis in the Action-Adventure

[While Day and Olson seem to be compatible, Klevjer’s approach is different. He claims the main property of the player-avatar relation is prosthetic telepresence. He also thinks that avatars are different from puppets, computers roleplays for us, and games have no diegesis.]

  • The first aspect of player-avatar relations is prosthetic telepresence. Like a prosthesis, the avatar reacts to our input. It’s an extension of us. But the relation to the character is not just prosthetic.  We are telepresent in the game. Telepresence is not just a function of input; it results from the avatar belonging in the game world. A mouse cursor is a sort of prosthesis, but it is not a presence or avatar. For something to be telepresent, it has to be an object among objects; it must both affect and be affected by its environment. The avatar is both subject (of the game) an object (in the game).
  • Klevjer claims that avatar control in games is different from puppetry—during play an avatar has “no will, no intention”. It is “an empty shell”. It seems he thinks that during puppetry a puppet is imbued with will and intention by the puppeteer. The avatar, on the other hand, is lifeless until some sort of scripted event gives it that will or intention.
  • The computer often roleplays the character, not us. We don’t come up with what Nico would do or want in GTAIV. The computer fills that in for us.
  • There is no diegesis in the game world. Game worlds are real spaces and the experiences we have in them are real. There is no such thing as having an embodied presence in an imagined world; think of a cutscene v.s. gameplay.
  • We don’t need to “ponder the presumed lack of synchronization between how the telepresent player is experiencing his role in the game world, and what we are told that playable character is experiencing. The two exist on separate ontological planes.”
  • We sometimes feel a tie between our own and our character’s experiences—but during play there is an unbridgable ontological gap. The player and the character are entirely different entities.

Me, My Reaction 

Each writer here differs on the details in player-avatar relations. But there’s an important commonality between them: the player and the character are separate, irreconcilable identities. Each writer doesn’t think it’s useful to ask whether the player and the character overlap as selves. They (the player and character) don’t. They can’t. For Olson, without the player pretending to be the character, the character isn’t active—but the character is still separate from the player. For Day, two different identities are made from the two different languages: the character, and the player. For Klevjer, the character is roleplayed by the computer—but there is no diegesis that the player inhabits, so while play happens, the character doesn’t really exist.

I like what Day says. But I also think that the line between the two “I”’s a game creates is not as distinct as he claims. Can we really pull input apart from the representations? To answer this, it’s helpful to think of the game avatar as Klevjer’s “telepresent prosthesis”. It’s a good phrase, one that Klevjer doesn’t really follow through with it. As he says, the avatar is not just a prosthesis. Nor is the player just telepresent. The player-avatar is both simultaneously.  If there’s a distinct separation between the two, we have binaries and can analyze games according to one and not the other. This is silly. It harkens back to the old ludology/narratology “debate”. And if there’s an ontological gap between diegesis and play, then each has to be taken into account with a different ontology. Which leaves games in a funny place where we have to talk about representations in one way and “play” in another. As if they can’t actually overlap. I don’t like that. Because it’s not right.

The representations (environment in which we’re telepresent) and the input (the functions of the prosthesis) being present at the same time are what makes a game a specific game. So two games could have exactly the same input functions but different representations, or vice versa, and they would be totally different games. Pulling the prosthesis from the telepresence—taking the input from the representations, the player from the diegesis, the agency from the environment—betrays the whole purpose of the phrase “prosthetic telepresence”. With just one of those things you don’t capture a specific game. In Day’s terminology, to talk about one of the senses of “I” in game without having to talk about the other ignores what’s going on in the game.

In a way, I guess these dual “I”’s might be considered units of the larger operation—but they seem to overlap in ways that makes me reluctant to read them as units.

I’ll have more on this later. But before that, I’ll be on an 8-hour bus ride from NYC to Pittsburgh tonight where I’ll try to write something about Morrowind.

 [Edit: Due to some technology-to-megabus integration issues I didn’t finish the Morrowind post. It’s now 2am so that’s not happening. But I have a lot of notes written. So something should be up tomorrow. It will also hopefully clarify some of today’s post. Now I am going to go do this:



The Project Gets Direction: Player-Character Relationships

The Prof and I have narrowed down the focus of the project. I’ll be taking a look at the relationship between player and the player’s avatar/character. More specifically (things will be getting more specific a lot), I’ll ask the questions “What perspective does a game put the player in? What is the point of view? In what ways does the game distinguish the player from the avatar?”

I hit on this in one of my earlier posts about Planescape: Torment. In other kinds of texts, like literature, perspective is pretty distinct.  Anna Karenina is third person. Love Is The Plan The Plan Is Death is second-person .The Left Hand of Darkness is first person. When perspective shifts in a work, we can tell. (Obviously there’s a little more depth to these points of views, i.e. the different modes of third person storytelling. The point is that we can still see these distinctions.)

 My intuition is that perspective works differently in some games. A game like PST seems to smudge second and third-person storytelling together.  I am not the Nameless One. But within the context of the story, I am the Nameless One. I read “You do this, you do that, you feel this”—but these second-person statements apply to the Nameless One. I act, but I’m not the one doing things. It’s blurry.

The natural response to this is that we are also the characters in stories we read. We project ourselves into the context of the characters’ world. I’m not Levin in Anna Karenina but I am projected into his world while reading his part of the story. This projection is part of what makes narrative compelling. It draws us into a work.  It prompts self-reflection, among other things.  Except I think a character like The Nameless One works differently. Because I don’t just identify with him; within the constraints of the game world I actively form what I’m identifying with. I direct and control him. I don’t just put myself in the context of his world. I act in it. It’s a different sort of relationship.

I’m reluctant to attribute too much to performance and agency in games. After all, as The Prof pointed out to me, we can act and perform in other media—it’s just harder. But there seems to be something in the player-avatar relationship worthy of pinning down.

Maybe games just exaggerate the element of self in a text. Maybe this is a function of player agency, or mechanics. Maybe games don’t do anything interesting regarding the player-character relationship. Or maybe there’s something totally new to discover while studying it. Whatever. There are some unanswered questions, and that’s what’s important.


So, I mentioned I’d be getting “more specific” a lot. This is where that happens. I want to avoid the pitfall of expecting to pin down what the player-character relationship is in all games. It’s overambitious (at least for me). So I’ll be looking at a few different games that are similar, but all seem to fall in different places on the spectrum of perspective. They will definitely all be roleplaying games. The Prof and I have a list started; each of these pieces of the list has its own different characteristic that might be good from a critical perspective. I hope to narrow it down to three games.

  1. Planescape: Torment
    1. This is going to be on the list. It’s too interesting not to be. Isometric camera.
  2. Fable
    1. This one is interesting because it offers a lot of player customization, with the ability to form a real sense of self as the avatar. But it also has a clear character background and fairly linear story. Third-person camera.
  3. The Elder Scrolls III: Morrowind
    1. This game sort of epitomizes the ability to “be” the avatar. Any quest or path can be totally disregarded. The game doesn’t overtly exclaim whether the player is good or evil. The world is huge and open-ended.  First-person(optional third-person) camera.
  4. The Temple of Elemental Evil
    1. Haven’t played this yet. I’m told it offers a static storyline to follow, which doesn’t really matter so much in the play experience. Isometric camera.
  5. Fallout
    1. Pretty good degree of character customization. Like Morrowind, facilitates thinking of the avatar as the self. (I think. Have only played a little in the past.) Isometric camera.

So there are three games here that are—at least superficially—really similar. Fallout, The Temple of Elemental Evil, and Planescape: Torment are all party-based isometric RPG’s. The draw with these three is that I don’t have to be distracted by different conventions, like camera view or player isolation. I can get a more microscopic view of what’s going on in these specific games that affects the player/avatar relationship, because it does seem to differ across all three. (For example, they are different in goal-direction and choice making).

The downside is that I wouldn’t be looking at how those different conventions (like camera view or player isolation) affect the player-character relationship. Camera is probably not incidental when we talk about how a player relates to her avatar, or what perspective a game is in. Same goes for games that have party members the player controls during the game. So I have to decide how far I want this analysis to stretch. How much do I want to account for? Any advice here would be good.

Again, my goal isn’t really to answer for the player-avatar relationship in all games. Just three (for now). And then maybe that will be useful when looking at other games.