Ludic and narrative sound in games
I believe that game sound design is not about designing sound, but about designing the game via sound. Every sound has a function. Every sound needs a purpose. As Joel Beckerman and Tyler Gray write in The Sonic Boom, “Sound for sound’s sake is often a missed opportunity.” But game audio is rarely discussed in these terms in the professional community. Like many others in this field, I used to focus on audio much more than on design during my years of learning. And here I propose a simple framework to shift the perspective and discuss sound in different terms than fidelity and acoustic features.
In 2015 Richard Stevens and Dave Raybould published an article exploring sound authenticity in Battlefield 4. The paper proposes a framework that divides the game audio into four functional spaces: ego-ludic, ego-narrative, socio-ludic, and socio-narrative. Here are the authors’ definitions:
- Narrative: Audio that draws us into a fictional world or narrative.
- Ludic: Audio that provides information to help the player achieve, or motivate the player towards achieving mastery.
- Social: Audio that is heard by all agents/entities in the game.
- Personal: Audio that is heard only by the player
Having a strong interest in design methodologies, I’ve been using this framework to analyze, categorize and prioritize sonic events in games. But methodologies have limitations, and facing one of them has changed how I draw the line between ludic and narrative in games.
You might have heard of the decades-long Ludology vs. Narratology debate. By looking at the image above one might think the framework follows the “ludic versus narrative” dichotomy, which I personally find somewhat counterproductive. Authors mention, however, that the lines between the spaces “are not ones of division but are purely for orientation”. Since “Personal” is intuitively seen as an opposite of “Social” it is easy to see “Ludic” as an opposite of “Narrative”, even though this is not the case here. To expand on this, I explicitly state that:
- most sounds in any game have both ludic and narrative functions;
- a ludic sound is not necessarily “anti-narrative” and vice versa.
These two statements are the foundation of the framework I propose in this post. Before proceeding, I need to clarify what I mean by “sound” and slightly pragmatize the definitions of ludic and narrative functions of a sound.
Definitions
By sound I mean a sonic event perceived by the player. I use it as a higher-level concept than an audio file or a sound event in the audio middleware terminology. Sound is subjective. The same sound event makes different sounds in different contexts, while different audio files can produce the same sound in similar contexts. Imaginary sound is a sound (Read Sonic Virtuality by Mark Grimshaw and Tom Garner if you are curious about this). The absence of an expected sound is also a sound because you always hear something. When you attentively listen and don’t hear what you expect, you still hear something else. I further elaborate on this definition in a separate post.
The ludic function is the sound’s ability to help the player deal with challenges and achieve their goals. Ludic sounds direct the players’ focus, inform about meaningful events and states of the gameplay systems, reduce cognitive and perceptual load, and improve the game’s usability and accessibility. The ludic value shows how well the sound aligns with the game- and UX-design direction.
The narrative function helps the player’s mind construct the story we want them to experience. Following the Theory of Narrative Thought, I see narrative as an attribute of human consciousness. Game narrative is subjective. It belongs to the player, while the narrative design is an intrinsic part of the game that shapes the narrative. In that sense, everything in the game belongs to a narrative (though I agree that not everything should be seen through the narrative lens) but not necessarily supports the narrative design. The narrative value reflects how well the sound aligns with the narrative design.
Both ludic and narrative functions can go to negative region. “Anti-ludic” sound misinforms the player, adds (or nudges the player towards) a non-fun challenge, masks more important ludic sounds, etc. “Anti-narrative” sound pushes the narrative in an unwanted direction or builds a competing narrative in the player’s mind.
Since sounds can be both ludic and narrative simultaneously, instead of contraposing these functions, I propose mapping them on a cartesian plane and analyzing the four quadrants.
The ludo-narrative plane
Synergistic sound supports both gameplay and narrative design, making the end product greater than the sum of its parts. Sounds in a well sound-designed game usually belong to this quadrant, even though most of them gravitate closer towards one of the axes. The enemy death screams from The Last of Us Part II make a perfect example of a synergistic sound. They both inform you that the threat is eliminated and make you feel in a certain way, supporting the message the game communicates. You can also think of a sound that plays when your shield breaks in a melee combat game. It makes you more cautious with your actions in the fight and contributes to your character's emergent story.
Misinforming sound supports narrative design but affects gameplay. The word “misinforming” is too narrow to describe every possible incoherence in this quadrant, but I struggled to find a better name. I welcome your suggestions!
Here is my favorite case of a misinforming sound. There is a hardcore multiplayer shooter, where skilled players carefully listen to the enemies’ footsteps to understand their location. One level is set during the thunderstorm. Some players wait when the thunder strikes to start moving, to “hide” the sound of their footsteps behind the rumble. But the sound of the thunder is purely aesthetic. It randomly plays on the client-side; every player hears it at different times, so the footsteps stay audible for the enemy.
Another example is a dialog that plays during an intense battle. If you focus on the voice-over, you perform worse in the action sequence. If you concentrate on the action, you miss some bits of the story. The player gets different information than what they need in a given moment, so, in a sense, it is still a case of misinformation.
Sometimes misinforming sound is good for the game. Many horrors scare us with creaks, scratches, and other intrusive, unsettling sounds. They have nothing to do with actual gameplay events and may disinform us about something happening next room or behind the corner. They are not just unhelpful, but sometimes disruptive from the gameplay progress perspective: we may take a wrong turn or become too scared to open some door. But all that is perfectly fine for the horror genre.
Depending on the context, misinforming sound can be acceptable in single-player story-driven games and undesirable in multiplayer competitive genres.
Destructive sound affects both gameplay and narrative design. It is rarely intentional. Most often, it is a result of a bug or bad implementation. Every game audio designer has heard a stuck endless loop that failed to stop when necessary. Such issues frequently happen during the development, but rarely end up in the released game because of the strong negative impact they have on the player experience. Performance-related dropouts or obviously missing sounds also belong to this quadrant. No one ever wants to have those sounds in the game.
Utilitarian sound supports gameplay but not the narrative design. Think of obviously unauthentic non-diegetic informative sounds like UI notifications. The quiet synthetic sound that plays every time you can climb on something in Death Stranding is one example of this. A repetitive voice alert informing you that your base is under attack in an RTS game also falls under this category. On top of that, you can think about sterile soundscapes in some competitive games, cleared from any non-informative sound that could distract the player with an unnecessary narrative.
Finally, some games release with an optional sound set designed to help visually impaired players. These sounds need to stand out from the soundscape, but the game is unplayable for part of the audience without them.
According to Stevens and Raybould, ego-ludic sound, which partially falls under this quadrant, is acceptable when the drivers for immersion are challenge-based. If you are curious to learn more about different types of immersion in games, you could start here.
The statement from above arguably holds for utilitarian sounds. We often deliberately break the narrative with informative sounds to make them more noticeable. But when we do, we look for some reasonable tradeoff. Bloodborne would be more comfortable to play if every time you kill an enemy you would hear fanfares, applause, and cheerful whistles. But these sounds would completely ruin the atmosphere, so it is hardly a reasonable tradeoff for this particular game.
Both misinforming and utilitarian sound are only welcome/acceptable when offering a reasonable tradeoff between ludic and narrative. We rarely welcome the sound that slightly supports the narrative at a huge ludic cost. So not all sounds in quadrants II or IV are equally good or bad for the game.
Here are few more examples of game sounds mapped on the ludo-narrative plane. The exact position of dots is subjective and debatable.
Note that since sounds are subjective, the same audio can move across the plane in a different context. The epic battle music that plays when you encounter an enemy in an open-world RPG has positive ludic and narrative values if you face a meaningful challenge. But think what happens if the same music starts when you meet a feeble level 1 goblin close to the endgame. The narrative value goes down because this battle doesn’t feel as epic as it sounds. If you choose to run past the enemy peacefully, the music will accompany your escape for a while with no good reason. The ludic value goes down because the music nudges you to engage in a boring fight to avoid degrading the narrative experience. Look at the visual representation of this shift:
Practical applications
Here are few ways to use this framework in practice. First, it can help us assess how well our sounds and sound systems work in the game. In other words, rate our work based not on how it sounds (which is also important!), but on what it does. The first step is to map the sound on the plane and think of the area or a trajectory it takes in various contexts. Then we separate the plane into several parallel “quality” segments, like this:
You can customize the segmentation to fit the priorities of your specific project. The top right and the bottom left parts of the plane are rather obvious and intuitively understandable. I’m most interested in two middle sections — Disruptive and Underperforming. Every game probably has at least a few sounds that fall into this area of “missed opportunities.” I wrote a separate post about disruptive audio last year, and even though I didn’t think in these terms back then, most of the arguments and examples from there still hold.
In many cases, sounds end up in this area for very trivial reasons like lack of shared goals or poor communication between the departments. Since disruptive and underperforming sounds are rarely obviously wrong, they are easy to overlook even when listening critically. This is where the framework becomes useful as a heuristic.
The ludo-narrative plane helps to quickly identify and evaluate these missed opportunities and suggest a possible fix for them. For instance, when we see that the audio becomes disruptive in some specific context, we can implement a small, context-specific fix. If a ludic sound has a strongly negative narrative value, we can think of changing either the sound or the contexts where it plays. If we see that sound consistently underperforms in every context, we should ask ourselves if we really need it in the game.
We also can use this framework when analyzing the reference games. In this case, we map the entire soundscapes the way we perceive them. The results will be highly subjective, but they help crystallize the direction, formalize the design goals, and set priorities for the game we are working on. Anything outside of the first quadrant is rarely useful for this purpose. Here is an example.
Finally, understanding the ludic and the narrative functions of a particular sound gets you a clear idea of who in the development team you can discuss it with. As I said I'm some previous posts, I firmly believe that good player experience is a result of tight interdisciplinary collaboration and knowing where to look for synergies is an essential quality for any game developer.
I hope this framework will inform some of your design decisions, but keep in mind that as any framework, it has limitations. If you want to improve, criticize it or propose a different use, reach out to me on LinkedIn or wherever else you can find me.
