Procedural Diegesis

Treating the Game Engine as Co-Author

Essay - Procedural Diegesis

Let’s talk about narration and videogames. In this case, narration refers to a game’s story, as told by the writers and the game engine. When there is discord between narrators, the story suffers, and when there is harmony, the narrative is more persuasive. Let’s call this element of storytelling ‘procedural diegesis,’ knowing that it involves treating algorithmic and authorial processes as co-authors of a narrative. The procedural portion here highlights that we are interested in processes, systems of representation that unfold over time that are dictated by rules and/or conventions. By diegesis we mean to indicate the internal consistency of the narrative. Together, they represent a form of narrative criticism that cares very little for content but quite a lot about delivery. Like Ian Bogost’s procedural rhetoric, which has informed much of this article, this perspective enables one to critique representational processes, only this time we are looking for coherence between <em “>narrative processes. In that respect it is beneficial to think of each narrator (writer, physics engine, texture mapping, audio system, etc.) as a system.

Systems

In Niklas Luhmann’s systems theory, each ‘system’ (legal, political, or artistic, for example) makes its own distinctions as to what is information and what is noise. The premise here is that information is ‘a difference that makes a difference.’ What causes waves in the artistic community may appear as noise to the legal and political systems. And even when all three overlap, such as an artist portraying a prominent politician in the nude, each system interprets the information in its own terms (legalese, political posturing, art history). In games, the simple act of jumping can involve numerous systems, such as physics, lighting, graphics, and story:

The player leaps off the roof top and grabs the handle on the helicopter.

Each system understands this event according to its own variables and parameters (momentum, light source, moving image, climax). In respect to these overlapping systems, procedural diegesis refers to the correlation between differences made by the engine with that of the narrative. In its most basic form this might be as simple as characterizing ‘health points’ as ‘power’ to a life support suit. But as we will see, procedural diegesis is not about fidelity to ‘reality’ but is rather about consistency within its own narrative world.

In Luhmann’s theory systems are closed, capable only of internal distinctions and differentiation. Communication takes place through a mutual irritation between two systems. For instance, the player manipulates the N64 controller to move Mario around the screen. The console interprets the upward movement of the joystick as a series of integers that correlate to the redrawing of graphics. The belief that Mario is moving forward and the computational processes that redraw the graphics are two irreconcilable distinctions of the same irritation. What’s more, if we were to add a Rumble Pak to the controller we could enhance the communication between the two systems. In this case the Rumble Pak helps close the gamer-console system (that’s right, two systems interacting with one another can form a larger system). In other words, the additional feedback helps maintain the coherence between player and game by expanding the range of differences that make a difference. To return to the topic at hand, procedural diegesis, then, is the means of closing the system on the level of software as well as narrative.

In my own research I refer to this process as ‘closure,’ following McLuhan and gestalt psychology. Briefly stated, the suspension of disbelief is to literature what closure is to media in general. And what’s more, each medium requires its own unique form of closure. Filling in the action between comic book panels differs from the kind of sensory engagement that correlates the joystick with the in-game point of view. From this perspective, videogames have largely been subjected, from developers and critics alike, to the forms of closure present in other media (namely, literature and film).

Print

Consider the disconnect between reading a narrativized description of a gameplay session and the actual process of playing the game, in this case Dishonored:

Carnage. Drop assassinations with the knife out, an instant kill. A crossbow to loudly take out another foe, in two shots. Around five guards swarm in, wielding both sword and pistol, alert lines revealing their state of open aggression. The time power, upgraded to allow full-on freezetime, means they are all safely dispatched. Charging straight upstairs, again using Windblast on the [target] but at full force from the centre of the room, slamming him through a plate-glass window, out to the balcony, over the rail, onto the street many floors below. This time, everyone sees.

(Alec Meer. “Corvo Blimey: A Dishonored Preview.” Rock, Paper, Shotgun)

The catch here, and this can, and often does, favour the marketing of videogames, is that the closure that follows reading is often more ‘realistic’ (i.e. persuasive) than the in-game experience being described. A man “slamming through a plate-glass window, out to the balcony, over the rail, onto the street many floors below” can be read with impeccable detail, such as: ‘the glass shattered into hundreds of pieces, only to be further broken down under the victim’s soles as he backpedals in vain before cartwheeling over the balcony.’ Literacy teaches us to insert these details into narrative experiences. In videogames, the underlying physics engine, lighting, texture mapping, interactive materials, audio cues, and character animation are all engaged in this event—they are all making claims about the believability of the environment. Put differently, they are all involved in the rhetoric of the game, but too often they argue against one another. The underlying belief seems to be that realism in games is an end in itself and that the ‘literary’ narrative determines all other processes. This may hold true for static print and the directed gaze in film but it counters the potential of gaming, as the technical instantiation (the game engine) is portrayed as an inferior mimesis of the real world. Conversely, the more we can view these algorithmic processes as co-authors of events and less like limitations or constraints, the more ludic our stories will become.

In this respect, it’s no coincidence that ‘retro’ graphics have resurged as the mode d’emploi for developers interested in a different kind of narrative from the big developers (Pippin Barr’s Art Game, Anna Anthropy’s Dys4ia, and Superbrothers’ Sword & Sworcery come to mind ). Such titles embrace the limited capacity of their respective platforms. Playing upon audience familiarity and low-fidelity, these designers are seemingly cognizant of the fact that the player can be just as or more engaged when given less information. It’s as though for the human system, the differences that make a difference are more persuasive when made internally than when prompted externally. Thus, it’s all the more problematic to present tensions in the procedurality of an information-rich game.

Take Dishonored as an example. This game tells the story of a city perpetuated by whale oil and ravaged by plague. As a player you control a wrongfully accused bodyguard turned assassin. On your journey you pass among the scant few impoverished denizens of Dunwall, picking up artworks that are automatically converted to coin in your pocket. You can summon rats to kill patrolling guards, but the rodents will ignore any unconscious bodies you leave strewn about the way. The city is sprawling, and yet you set foot in every major quarter; its villains are numerous, and you meet every single one. Dishonored is not a bad game but there is tension between the narrators (in this case, the writing that establishes the narrative world and the algorithmic processes of gameplay). In other words, what the game says and what the game does are two distinct forms of narration. This is because the ‘story’ (the scripted events) is based largely on filmic techniques, whereas the gameplay is composed of ludic processes; both require different forms of closure in order to gain maximal rhetorical efficacy.

Cinema

In contrast to literature, closure in film is of a different sort. In movies the perspective is fixed and directed. What takes place off screen is assumed to be a continuation of on screen events, not only in terms of action but style and scale as well. In videogames, however, ‘off screen’ is largely a misnomer—the player has the capacity to observe his/her surroundings. Thus, unlike filmic approaches to game studies, I do not view videogames as an extension of a fundamentally cinematic mode but rather videogames are a categorically distinct medium that requires its own form of closure; this also necessitates a unique form of diegesis as well.

At this point it’s worth noting that there is a precedent for diegesis in games studies. In Gaming: Essays on Algorithmic Culture, Alexander Galloway differentiates between diegetic and nondiegetic modes in video games, explicitly taking his cues from literary and film theory (7). In one particular passage Galloway makes note of “enabling objects” that  are “integrated seamlessly into the world of the game using some sort of trick or disguise—What Eddo Stern calls ‘metaphorically patched artifacts’—as with the voice recorders in The Thing or the HEV suit charging stations that supplement health in Half-Life” (32). It follows that these moments “[reveal] the tension often present in games whereby diegetic objects are used as a mask to obfuscate nondiegetic (but necessary) play functions” (32). This would seem like the initial steps in a move towards developing what I’ve called procedural diegesis. And yet Galloway’s definition of diegesis in a videogame is “the game’s total world of narrative action” (7) and thus he includes cinematic interludes under the diegesis umbrella.

However, on a procedural level, in-game cut-scenes could not be more nondiegetic, for even though they perpetuate the narrative, they do so in a manner that fundamentally alters how the narrative progresses. Consider that not only do cut-scenes remove player agency, they typically involve camera angles outside of those used throughout the game (this is especially disruptive in shifts from first-person to a third-person or vice-versa) and they’re often rendered in greater detail than the actual gameplay. This violates the core principle of procedural diegesis on the basis that a fundamentally different set of processes and rules are being engaged to relate part of the narrative. And so throughout gameplay the player is shot dozens, even hundreds, of times, but in a cut-scene a single bullet can change the trajectory of the narrative. The differences that make a difference are inconsistent and this undermines the rhetorical efficacy of the narrative.

Consider a game like Bioshock, which unfolds entirely within the first-person perspective of the player. In this underwater dystopia, players are repeatedly aided by a character who utters the phrase ‘would you kindly.’ Later it is revealed that this phrase was supplanted in the mind of the main character as a post-hypnotic suggestion, enabling this apparent ally to bend the player to his will. And here we have a criticism of the average gamer that mindlessly follows orders, killing indiscriminately to appease the on-screen objective. The revelation of this manipulative utterance part way through the game is intended to be a dramatic moment: you’ve been manipulated all along and now that you know, you are free to forge your own path. And yet the structure of the game, the very procedural basis for progress that is hardcoded into the scripts that define objectives and enable progress from level to level, requires that you submit to the game’s own version of ‘would you kindly,’ which is ‘would you kindly play this way and perform these actions and engage in these processes.’ And so the designers’ glib commentary on the blind faith of the gamer cuts deeper than perhaps even they are aware, for such blind faith is the very basis for much of the industry, the very nature of the digital content churned out annually by large studios. This proves antithetical to the ludic nature of the game, for all subsequent play-throughs of Bioshock, despite this new information, unfold according to the same objectives. Thus, we have a narrative that calls into question the procedures of the game only to enforce their constancy. The only way this criticism really ‘lands’ is if the player has a degree of freedom over which objectives to pursue and how. And yet Bioshock is a definitively linear game and as such its argument is self-defeating. The processes the game has you engage with are at odds with the narrative itself.

Conclusion

 

Dishonored

Screenshot from Dishonored

If we are going to define digital games by their algorithmic or procedural nature (Galloway’s subtitle, after all, is “Essays on Algorithmic Culture”), then our terminology should reflect that fundamental distinction. Diegesis in reference to videogames, then, should refer to the narrative process and the rules that dictate that process. Non-diegetic games, like Bioshock, treat the game engine like a blank canvas, and thus content supersedes medium, message surpasses delivery. A game like Dishonored asks you to suspend your disbelief and accept that these NPC models are human, that this scenery is real, and that you are in control of the moral of the story.

Anthropy

Screenshot from Dys4ia

Anna Anthropy’s Dys4ia, on the other hand, attempts to replicate through play the experience of alienation and discomfort; the ludic elements, though nearly devoid of visual similitude to their real-life counterparts, replicate through their procedurality the experience of disconnection, rejection, and confusion. In the former, the story of Corvo the assassin could conceivably survive a translation to the silver screen. Dys4ia is inextricable from its play. As a critic and a gamer, I’m far more interested in the process of feeling something new, like gender dysphoria, than being told about an assassin that cuts throats at the press of a button. In fact, Anthropy sums this up rather nicely when she remarked that “[Dys4ia] was a story about frustration – in what other form do people complain as much about being frustrated? A video game lets you set up goals for the player and make her fail to achieve them. A reader can’t fail a book. It’s an entirely different level of empathy” (“Dys4ia tackles gender politics, sense of self, and personal growth… on Newgrounds“). Procedural diegesis draws attention to the simple fact that its not merely content that invokes empathy but the very algorithmic processes as well.

 

  • Pingback: Procedural Diegesis in Anna Anthropy’s dys4ia » chris dale dot game()

  • A. Duncan

    Great article, I completely agree that there needs to be greater harmony between ludic and narrative aspects of games, and you do a very good job at looking at how cut-scenes disrupt this. I wouldn’t be so quick to dismiss cut-scenes as a whole, however, as they can be used in a way which contributes both to narrative and gameplay. In Journey, for example, the cut-scenes serve to explain the world, but also hint to the player their eventual path, thus helping them prepare for future challenges in the gameplay (I’m talking here about the scenes with the white robed figures). Also, at the beginning of each level, the camera view expands, giving a wider view of the landscape in another kind of cut-scene. This also helps the player, however, as they can see, in the distance, objects which they will need to complete the level. Thus these cut-scenes both increase the narrative and visual enjoyment of the game, while simultaneosly contributing to its gameplay.

    Another small issue I have with your argument is that you posit that the algorithmic processes and the narrative designers for the game form the authorial voice together: “Let’s call this element of storytelling ‘procedural diegesis,’ knowing that it involves treating algorithmic and authorial processes as co-authors of a narrative.” My one point of contention with this is that it leaves out the role of the player as an authorial voice in the game. The player, in having agency in their interaction with the environment (of course within the limits of the game) also determine the narrative. While their contribution can be more or less prevalent, it is always there, even in the most linear of games. This is an important point to me, and while it isn’t much of an issue in your article, I think it always needs to be kept in mind. A more complete discussion of this can be found in Aylish Woods’s excellent essay in Games and Culture, where she talks about “recursive space.” Also, an essay talking about this process (in a sense applying procedural diegesis in an extended way to a particular game) can be found in Game Studies, “The Algorithmic Experience: Portal as Art” by Burden and Gouglas.

    I hope my comment doesn’t seem overly critical, I agree with the vast majority of your argument and will likely be citing it in my upcoming paper!

    • Thanks for the feedback. Sorry my reply has been so delayed, I wanted to give some serious thought to your points. If I may expand on the above essay in response:

      1) Agreed. Cut-scenes can provide invaluable insight into the world and the opening shot that pans across the level can provide both ludic details (locations of enemies, patrol patterns, etc) and narrative details (why are these enemies here? to rob a bank). What I’m suggesting with procedural diegesis, however, is a more constrained approach in exchange for a more persuasive narrative. Let’s say that that opening pan shot showed me some robbers in the midst of a heist. As the camera pans I see a sniper on the roof overlooking the back door. The open shot ends at me, a cop, about to take on these crooks. However, I now know more about the environment then the player character–I can avoid the sniper at the back door and proceed with the mission or sneak up on the would-be assassin and take him out. Elsewhere the narrative will likely keep me ignorant of other details for the sake of propelling the narrative. Which is it? Am I the cop or am I someone controlling the cop who sometimes gets clues and sometimes has them withheld? I don’t think all (or even most) games should adopt a hardline approach in response to this disconnect but sometimes I want to experience a game with the same constraints as the character I’m playing.

      2) Agreed, once again. I really downplayed the role of the player here, perhaps too much. But in my defence I wanted to critique the design elements that were preventing me from believing that I am this character and that this world has rules that limit that individual. To fully embrace procedural diegesis is to, in a sense, procedurally recreate the constraints placed on the character through the (narrative) mechanics. So while the player participates in the authorship, they should do so through rules that facilitate your suspension of disbelief, not undermine it. To analogize, non-diegetic games are a bit like LARPing if you were allowed to break character at certain times for the sake of some larger narrative. Except with cutscenes you’re actually forced to break (with your) character. To push this even further, let’s say I’m LARPing in the park in my makeshift medieval garb when a professional crew shows up to ‘enhance the realism.’ Immediately their better gear makes me question the validity of my carefully wadded bedsheet-turned-tunic (normal in-game fidelity vs high-res cutscene). And since these fellows have a set routine for fighting I find myself watching the battle unfold rather engaging (removal of player agency through cutscene). By the end I’m having a really hard time believing I’m Steve the Squire out to prove myself to the Lords of Waterloo. We shouldn’t deny ourselves these other techniques but at the same time we risk losing the potential to persuasively communicate an experience if we remain committed to the idea of an overarching story as a higher ideal.

      Lastly, I really appreciate those article suggestions. I haven’t had time to look at them but that’s exactly what I was hoping for from our community–some good old knowledge sharing that a google search just can’t match! You were in no way overly critical, and so the above is merely a response to some very valid observations. If I may pass along some links of my own, I came across Clint Hocking’s ‘ludonarrative dissonance’ (http://clicknothing.typepad.com/click_nothing/2007/10/ludonarrative-d.html) article recently, which articulates some of the elements of procedural diegesis far better than I have and he wrote that 5 years ago! The term has since been picked up an expanded upon here as well (http://www.destructoid.com/ludonarrative-dissonance-the-roadblock-to-realism-235197.phtml) and it’s gained some wider recognition in games studies, so you might find it more useful than my somewhat awkward ‘procedural diegesis’. The terms differ in that I’m more interested in structure and the engine as narrator but many of the principles are the same. I hope that’s useful and thanks again for the insightful comments!