Originally posted on the Chromed blog here.
Recently, aspiring game designer Eric Schwarz posted an article in which he argues that an unreliable narrator is inappropriate for interactive fiction. As March 32nd‘s protagonist is very much an unreliable narrator, and its story dependent on the choices of the player, I feel the need to defend the use of this literary device in video games. This post does not contain spoilers for March32nd.
A narrator-protagonist is an avatar through which the player experiences the world of the game. Like the player, the avatar lacks perfect knowledge of the game world. Given this situation, skilled narrative designers can match the options for discovery available to the player to those available to the protagonist, and in so doing allow the player to express his or natural curiosity in a way that doesn’t cause the character to take incongruous actions.
A protagonist whose knowledge is imperfect because their perception (or presentation) isn’t accurate allows designers to ratchet up the tension; the player is learning, but must be mindful that there are limitations to how accurately the protagonist provides them with information. These limitations don’t need to be a setup. They can be used to encourage the player to probe deeper with their own questions about the game world.
While protagonists with missing memories or dishonest intentions may frequently result in shallow “gotcha” reveals, an unreliable narrator doesn’t have to be setting the player up for a sort of reverse dramatic irony. Indeed, they can be used for good old fashioned dramatic irony. With an unreliable narrator, the player can make a choice knowing that the outcome will be different than what the character expects. This forces the player to give more consideration to their choices than if they were acting through a direct link into the world of the game.
Do these knowledge deficits create problems for interactive narratives? No. Players can still make choices, and the game will follow through on them. Schwarz suggests that an unreliable narrator takes away control of the narrative from the player in an attempt to forcibly retcon the player’s actions into alignment with the predetermined narrative. Some games do this, and as is noted, in a way that takes the player out of the game. I assert that the problem is in attempting to shoehorn the actions of the player back into the predefined narrative. When the actions of the player and the protagonist deviate this should be presented clearly at the time, in a way that makes sense within the context of the game. Done correctly, this sort of break can be very powerful.
Take for example, the scene in Final Fantasy VII when the player, acting through the protagonist Cloud, must swing his sword at an unarmed innocent. The player recoils at each press of the button, but must advance the blade, against both Cloud’s desires and their own. When the ultimate press is made, the deathblow does not come. Had the player killed the innocent, only to have Cloud suddenly remember that he did not take that action, it would have been jarring and cheap, but because the disconnect was built into the flow of the game, the scene was far more engaging that had it been a cut scene, or an impermanent act hidden by the unreliability of the narrator.
It is my intention, as well as that of the rest of the March 32nd team, to demonstrate all of this in practice. I look forward to feedback from the community once the episodes start coming out.