“You are not Stanley”: Cyranoids in “The Stanley Parable.”

In 1897, Frenchman Edmond Rostand wrote a play entitled “Cyrano de Bergerac.” The play told a story which has since been retold in various other plays and films: an eloquent man (Cyrano) who is in love with a woman, but who finds himself too ugly to suit her (in particular, he is ashamed of his long nose), instead woos her by feeding lines of his own poetry through one of his more handsome but less eloquent friends – essentially using his friend as a puppet for his own affections. Confusion ensues when this leads to the damsel falling in love with the man’s friend, as opposed to the man himself.

This story has impacted more than literature alone. In the 20^th century, one psychologist, a Doctor Milgram, was well known for his controversial experiments on deference to authority (the infamous “Milgram experiments”); what Milgram was less known for was theory developed at the end of his life, which was directly related to the plot of “Cyrano.” Milgram hypothesized that people could not tell the difference between conversing with an ordinary person, and conversing with someone whose speech was transmitted to them by a third party. This latter kind of situation exactly mirrors the plot of the play; in reference to the play, Milgram termed a person whose speech originates somewhere other than their nervous system a ‘cyranoid’. For current purposes, I expand this definition to ‘an entity whose actions originate somewhere external to themselves’.  (You can read more on this in a very well-crafted 2014 Wired article on the subject of cyranoids.)

Milgram did not have the opportunity in his own life to conduct a large amount of research into the dynamics of cyranoids; however, since Milgram’s time, studies conducted since then (such as Corti & Gillespie 2014) have suggested that Milgram’s hypothesis was right: people seem unable to distinguish cyranoids from ordinary people.  This is the reason, for example, why unwitting participants in ABC’s hidden camera show, “Repeat After Me,” watch celebrities do and say whatever the show’s host (Wendi McLevon-Covey) tells them to do through a remote earpiece, without ever once doubting that the celebrities are acting and speaking genuinely.

Research in cyranoids is still wanting in terms of scientific data; however, there is another entirely distinct field in which cyranoids have substantial potential utility: the aesthetics of video games (in particular, I refer to narrative-driven video games).

The most general way in which cyranoids hold potential for game aesthetics is that, at first pass, it seems that most video games provide explicit cases of cyranoids. This is because of the avatar-player relationship: the player dictates the avatar’s actions, by definition of what it means for a character to be an avatar. In light of this, Milgram’s theory gives us a robust way of understanding just what sort of entity an avatar is: an avatar is an instance of a cyranoid. This also gives us a potential way into analyzing the relationship between avatars and NPCs: in light of how humans have been shown to behave around cyranoids, we might be able to point to cyranoid dynamics as the theoretical grounds for why NPCs treat avatars as “just another person” in their world, despite avatars being ontologically different from NPCs.

This general mode of analysis will take significant fleshing out, which is outside the scope of this paper. Presently, I wish to demonstrate the analytic utility of cyranoids by showing how they shed insight on the narrative dynamics on a well-known game that is explicitly concerned with how narrative functions in games: this is “The Stanley Parable” (Davey Wreden, 2011).

“The Stanley Parable” tells the deceptively simple story of a man, Stanley, who arrives at his office job one day to find all of his fellow employees missing. The game is voiced over by a narrator, who describes what Stanley is doing at each moment in the story, as well as what he is going to do next. The player, controlling Stanley from a first-person perspective, has the choice of either acquiescing to the narrator’s description of what Stanley is going to do next, or of doing something different. For example, there is a point in the game at which Stanley encounters a door on his left and a door on his right. At this juncture the narrator says that “Stanley went through the door on the right”; the player, then, can either acquiesce by directing Stanley through the left door, or contradict the narration by directing Stanley through the right door. When the player contradicts the narrator, he will initially try to recalibrate the narrative such that Stanley still arrives at the conclusion that the narrator had in mind; however, should the player continue disobeying, it will be impossible to return to the narrative’s story, and he will express his mounting frustration as the game evolves in all sorts of ways that he did not desire.

Ultimately, “Stanley” is a game that is explicitly concerned with problems of choice and the linearity of game narratives. Immense as these subjects are, there are many possible tacks to take in analyzing the game as an aesthetic and argumentative piece. What I now wish to show is that one particularly efficient and useful way into the game’s dynamics is to analyze its narrative as a conflict between three distinct conceptions of Stanley as a cyranoid: Stanley controlled by the narrator; Stanley controlled by the player; and Stanley controlled by the game designer.

One of the useful tools offered by Corti & Gillespie 2014 is a compact notation for describing cyranoids, of which I will avail myself in this analysis. The notation is as follows: a cyranoid, in which the actions some agent x are dictated by some third party y, is defined as {[x] y}. So, for example, the three competing conceptions of Stanley that I just picked out are codified as {[Stanley] Narrator}, {[Stanley] Player}, and {[Stanley] Game Designer}, respectively.

It is useful before turning to “Stanley” to consider how cyranoid dynamics function in other storytelling media, such as films and novels. It is clear enough that there exists no analogue in these media to the {[Avatar] Player} cyranoid that I pointed to as a typical feature of video games; we assume, unless explicitly shown otherwise (cf. Cyrano de Bergeron), that the actions of character in a story are not dictated by another character. However, there is another type of cyranoid that is virtually ubiquitous in these media, and which is so obvious that there is usually no point in noting it: the narrator of a given narrative determines the actions of characters as conceived within that narrative. So characters in these media are cyranoids of the form {[Character] Narrator}.

I will dwell for a moment on this last claim, because it is easy to doubt its veracity if one fails to parse ‘as conceived within that narrative’ correctly. Many narratives depend on the notion of an unreliable narrator, on whom the reader cannot depend for an accurate account of the events that are described within the narrative. In light of this, one might object that it cannot be the case that the narrator determines the actions of characters within a narrative in the way required of cyranoids. But what I am referring to by the ‘character’ in {[Character] Narrator} is the semantic complex resembling a particular entity within the narrative that is being told by a narrator. If there is some notion of objective truth in play, then it may well be that the narrative in question is untrue if the narrator is unreliable. However, this has no bearing on the status of the character as conceived by the narrative itself; it only has ramifications for the truth-value of the narrative and its constituents in relation to objective truth. So, saying that the narrator of a given narrative determines the actions of characters as conceived within that narrative is merely claiming that the actions of a character constructed within a narrator’s narrative are determined by that selfsame narrative, which is the product of the narrator. It is easy to read too far into this claim because, as I said, the claim is typically too trivial to bear mention.

The {[Character] Narrator} cyranoid becomes non-trivial when we turn to “The Stanley Parable” – in fact, part of what makes the game so interesting is that is underscores this concept and makes it nontrivial. The narrator of “Stanley” has a linear storyline formulated, which he is presumably aiming to recount to an audience – viz., the game begins by the narrator delivering the start of a story over a cutscene, which perfectly coheres with his descriptions; it makes sense that the narrator would expect the remainder of the story to progress in this way. The narrator, in other words, expects the {[Character] Narrator} dynamic to hold as in traditional narratives. This is the first of the three competing conceptions of Stanley: {[Stanley] Narrator}.

The problem with this conception of Stanley is that video games integrate choice and possibility in ways that traditional narratives don’t, as manifest in the relationship between player and avatar. The ability for the player to choose different ways to direct an avatar through the world frustrates any conception of narrative as a single, fixed path from a beginning to a conclusion. This peculiar dynamic of video games is what I described earlier as the {[Avatar] Player} cyranoid, and in this case is the second competing conception of Stanley: {[Stanley] Player}.

The conflict between the first and second conceptions of Stanley describes the friction of traditional storytelling in the medium of video games. We typically suppose in stories that the actions of characters are determined by some third party – a character with genuine free will, after all, would not function so much as a semantic component of a story, but rather as a genuine, living being. In this way, characters are always some type of cyranoid, narratologically speaking. But the question of who gets to determine the actions of characters is complicated by video games, and a tension emerges that is not present in other media. On the one hand, a typical narrative seems to require some kind of definite story arc, which would imply the {[Character] Narrator} framework of cyranoids; on the other hand, the person engaging a video game is explicitly able to control at least one character – the avatar – and this leads to a {[Avatar] Player} framework. When the character that we are analyzing is the avatar, we have a problem: two different entities, the player and narrator, are in a position to control the actions of the same character.

The conflict can be formalized using a model of game narrative that I call ‘narrative three-space’. The broad idea is that, because narrative-driven games have a main plot as well as a variety of sidequests and other things for the player to explore, we need three axes in order to represent the course of a video game narrative.  This can be graphically represented as follows.

In this model, the game’s narrative begins at the origin, (0, 0, 0). The z-axis describes progress in the main plot of the game, concluding at some maximal z-value, which typically represents the point at which the game’s credits roll. The (x, y) plane represents what I call the ‘exploratory domain’, which encompasses all actions the player can take that do not advance the game’s main plot. A point, which I label ‘lambda’, describes the current position of the avatar in the game’s narrative. The game’s narrative proceeds by the player relocating lambda by one of the vectors available to them at that position in the three-space, each of which represents a different available choice in the game’s narrative (so, for instance, a player may be able, at some point, to either battle a boss, pursue part of a sidequest, or start a different sidequest) Each point in the three-space represents a narrative event in the game. Given this narrative space, a playthrough of a given game is described by the path of lambda, as directed by the player, from the origin to the maximum z-value of the three-space.

Returning to “The Stanley Parable,” we can describe its narrative three-space in the following terms: the narrative dictated by the narrator is the ‘main plot’, represented by the z-axis. Every narrative that results from the player disobeying the narrator is a distinct path through the exploratory domain, described by vectors which have only x and y components (that is to say, they do not advance the main plot). Notice two things about this framework. First, the formalization makes clear that “Stanley” inverts the narrative structure of most video games: whereas the narratives of most video games are anchored by the main plot, which the player ultimately aims to see through to its conclusion, the majority of narrative in “Stanley” comes from avoiding the main plot. Second, this allows us to schematize the conflict between the {[Stanley] Narrator} concept and {[Stanley] Player} concept like so: the game highlights a conflict between lambda (i.e., Stanley) being guided by vectors with z-components, thereby advancing the main plot, and vectors with exclusively (x, y) components, advancing without regard to the main plot. This is a precise account of how “The Stanley Parable” pits traditional storytelling and video game storytelling against each other: linear storytelling is at odds with choice mechanics because the multidimensionality of an exploratory domain necessarily eschews any potential for a narrative that is perfectly linear. Put another way, the dynamics show that the kinds of strictly deterministic linearity we have been discussing and genuine agency on the part of the player are mutually exclusive.

In one of the possible 19 endings to “The Stanley Parable”, the player “breaks the game” by making a choice that was supposed to be impossible within the context of the narrative: trapped in a room with a ringing phone, which Stanley supposedly must answer, the player instead directs Stanley to disconnect the phone. This leads the narrator to acknowledge the player directly, saying to her, “You are not Stanley!” This is the crux of the conflict between {[Stanley] Narrator} and {[Stanley] Player}: the narrator assumes he is describing a character, a complex semantic unit, the actions of which are entirely determined by the narrative; but in fact, as the narrator realizes in this outcome, he is interacting with an agent external to the world of the game, who imbues this character with the ability to choose whether or not to follow the narrator’s narrative.

The preceding dynamics alone would suffice for a game with insightful meta-commentary about video game aesthetics; however, “The Stanley Parable” goes one step further by also inviting the player to consider the fact that it is the product of game designers. In another possible ending, Stanley is taken to a pristine museum with vaulted ceilings, transported away from the narrator moments before Stanley’s death. A different voice poses a question to the player: “When every path you can walk has been created for you long in advance, death becomes meaningless, making life the same. Do you see now? Do you see that Stanley was already dead from the moment he hit start?” The exhibits in the museum are various pictures and stories from the real-world development of “The Stanley Parable”: concept art, early beta versions of various endings, snippets of narrator speech that were ultimately discarded, and so forth. After the interlude in the museum, the narrative returns to the moment immediately preceding Stanley’s death, and the voice implores the player to make “[her] only true choice”: quitting the game. This is our way into the third conception of Stanley as a cyranoid: {[Stanley] Game Designer}.

I have said that, as we can see by the narrative three-space, the player is able to make choices that lead Stanley to diverge from the narrator’s linear narrative; however, the choices available to the player are not unlimited. The totality of the game’s narrative, as represented by the three-space, is the product of the team that developed the game (I gloss the entire team as “Game Designer” in analysis); the Game Designer also codes the set of vectors available for the player to relocate lambda. As such, even though the player has choice within the narrative, the actions from which the player chooses are a function of the Designer – a third party. So in this way, it looks like we can actually define the player as a cyranoid in the form of {[Player] Game Designer}, so long as the player acquiesces to the act of playing the game. Since we have already determined that the player determines the actions of Stanley, the fact that the player’s possible actions are determined by the Game Designer lead to the conclusion that there is a third cyranoid as which we can define Stanley: {[Stanley] Game Designer}.

This also tracks with the narrative three-space model: this third cyranoid picks out the totality of the three-space, which the Game Designer created and over which the player seems to lack authority. This is the sense in which quitting the game is the only “real” choice available to the player: it is the only choice that is not fundamentally encoded by the game itself – and, as such, it is the only choice that is not the work of the Game Designer. This highlights a second tension that, unlike the first, is particular to video game storytelling: are the actions of the avatar determined more by the player or the game designer? “Stanley” suggests that the game designer ultimately has more authority, since the only true choice for the player is quitting the game; however, the answer is not clear-cut, which is part of makes the tension interesting to explore within a narrative context. After all, although the set of total choices available for the player to make through the avatar is determined by the game designer, the matter of which members of the set the player actually chooses to use in the playthrough does seem to be determined by the player herself. This means that there can exist a real push-and-pull between the degree to which a player defines the avatar and the degree to which the Game Designer defines the avatar. So a complete definition of a Stanley cyranoid would be most satisfying in a form along the lines of {[Stanley] Player/Game Designer}, where ‘Player/Game Designer’ represent a dynamic relationship in which both the Player and Game Designer exert partial influence over Stanley.

Cyranoid analysis allows us to see the crux of “The Stanley Parable” in a new light: on this reading, the game is an exploration of determining exactly what ‘Stanley’ is. It highlights how this seemingly simple question has drastically different answers depending both on what medium Stanley is represented in (traditional media versus video games) and on whether we emphasize which choices he actually makes (which are determined by the player) or all possible choice he could make (which are determined by the Game Designer). The game itself can thereby be seen as a study of precisely what constitutes the character, and how seemingly innocuous choices of how the character is represented can drastically alter the character’s existence. This demonstrates both the utility of cyranoids in video game aesthetics, and also what makes “The Stanley Parable” such an enticingly bizarre game to play.

Oh, and not that I concern myself with authorial intent, but I close with a moment of icing on the cake.  Doctor Milgram, to whom we owe the concept of cyranoids?

His first name was “Stanley.”