My Summer Scholars project explored interactive narrative through the lens of video games. I studied narrative theory, analyzed literary and story game case studies, and created an original branching narrative game.

First, I studied how the rise of interactive narrative has both deepened and subverted storytelling by disrupting established notions of audience/creator relationships, the nature of authorship, and the future of the format. I examined scholarship on the mechanics of interactive narratives and video games, and analyzed issues such as the Paradox of Choice, what goals interactive fiction should pursue, and how the genre can or should fit into designations like “literary.” I also explored how interactive fiction finds itself in conversation with philosophies of authorship written long before its heyday.

Second, the nonlinear novels and seven interactive story games selected for this project reveal a system of common hierarchies and structures critical to both analyzing and writing branching narrative games. Additional insights were gleaned from interviews with three professional narrative designers.

Third, I planned and wrote a branching narrative of my own. Through it, I experienced the unique creative challenges posed by nonlinear fiction, including the importance of planning, avoiding choice bloat, and balancing player choice with story structure. This practical dimension of the project adds a new perspective to the academic literature by looking forward to the future of storytelling overall rather than only looking backwards to specific existing works.

In the world of creative media, video games are young. Over the past several decades, they have evolved from simple tic-tac-toe challenges to the pixelated passion projects of early programmers to the vast and growing world of triple-A titans and creative indie developers we inhabit today. Compared to thousands of years of fiction and theater and over a century of cinema, video games are positively fetal. Yet their core attribute, the beating heart of video games and what draws us to them, has enabled a new mode of narrative engagement that is rewriting storytelling as we know it: interactivity.

This isn’t to say interactive stories have never existed in some form. Plays often indirectly involve the audience by breaking the fourth wall. Some even explicitly allow the audience to influence the plot itself, as in theatrical experiences like Sleep No More. Similarly, many kids have had the delightful experience of flipping to this or that page number in choose- your-own-adventure books to defeat this shark or befriend that dragon. Indeed, the earliest interactive story games were hypertext quests like Zork that functioned very much like a choose- your-own adventure book, enabling players to move between locations and interact with the environment by typing a finite set of commands into their consoles.

By now, however, digital media has enabled interactivity to balloon far beyond the possibilities of stage, paper, or even hypertext. Popular games now proudly boast extensive narrative components like rich worldbuilding, fascinating characters, and stories that fork and branch into elaborate trees based on the player’s actions. With digital media’s capacity to store near-endless amounts of data, it seems the only limit faced by a software-savvy writer is the extent of their imagination.

Of course, there are many who think that games have little to offer in the way of narrative, and they are often looked down upon by educators, scholarly discourse, and even public opinion. Despite the vast array of game genres and the millions of players who enjoy them, there remains a perception of games as a brainless pastime without much to add to “serious” creative discourse. But the world of video games is much broader than this narrow connotation allows. As this project endeavors to show, video games have a great deal to teach us about the workings of narrative, and the creative insights they offer are woefully understudied. The future of interactive narrative is bright and the possibilities endless.

It’s impossible to discuss the field of video game theory without acknowledging its tumultuous beginning as scholars struggled to figure out games’ relationship to other forms of media and academic disciplines. Do video games fall along the same narrative axis as traditional media, or do they comprise their own category distinct from other media, which is perhaps even mutually exclusive with narrative itself? Ludology, as the latter viewpoint called itself, focused on the centrality of gameplay, rules, mechanics, and player skills/strategies to differentiate video games from other media, and proponents pointed to the absence of seemingly ubiquitous narrative features like narrators/narratees, linearity, and fixed temporality. Narratology, on the other hand, sought to place video games within academia’s existing literary theories that encompasses movies, novels, and so forth.

Neither viewpoint ended up defining game discourse as it exists today, which has shaped itself into a field independent from both positions, though with elements of each. It would be shortsighted to completely disregard the challenge aspects of play that do not exist in traditional media, and it would be similarly hasty to point to these elements as reasons to exclude the possibility of storytelling. Strategy, reflexes, and certain skills are completely foreign to media like novels but highly relevant when discussing the video game genre. Similarly, game stories share principles that govern other types of storytelling media and narrative theory can often be applied to the story aspects of games as it would be to a movie.

That said, ludology, in my opinion, was certainly the weaker stance in this early debate. Marie-Laure Ryan, in her seminal work Avatars of Story, provides perhaps the most comprehensive takedown of ludology in chapter 8, “Computer Games as Narrative.” She dissects common ludological arguments, which largely boil down to simplistic assumptions about the nature of narrative. For example, the argument that games cannot be narratives because “they do not allow the rearrangement of events that marks the distinction between story and discourse” (Ryan, Avatars 185) is easily dismissed. While chronological scrambling might not be a common feature of games, Ryan points out that plenty of games use cinematic cut scenes to exposit the backstory of a character through flashback (Avatars 185). She adds that “I can think of some cases where flashbacks would not be detrimental to gameplay” (Avatars 185).

This essay was written over fifteen years ago, and Ryan’s prediction has since come to pass. As Dusk Falls, for example, is a branching narrative that spans multiple timelines. The first narrative takes place in present-day, with Zoe, a survivor of the hostage crisis the story centers on, grappling with PTSD. The second takes place within the crisis, where the player controls Vince, Zoe’s dad, whose actions determine the fate of the family and the extent of the trauma Zoe faced. A third timeline tracks the events leading up to the crisis, following the youngest and most innocent criminal, Jay, and his turbulent home life. All three timelines interact and intersect, and choices made in one define outcomes and crossroads in the others. Gameplay elements, like swiping, clicking, or circling the mouse, are mostly accessory to the story.

Another ludological argument asserted a similar fallacy by asserting that narrative must recount real or fictious events by way of one or more narrators to one or more narratees—meaning that any events that unfold in real time, rather than being recounted, are not narrative (Ryan, Avatars 184). This narrow definition would exclude not only games, but theater and cinema as well. Ryan correctly points out that “the trend today is to detach narrative from language and literature and to regard it instead as a cognitive template with transmedial and transdisciplinary applicability” (Avatars 184).

Ultimately, Ryan settles on an open, intermediary interpretation of games’ relationship to narrative that I tend to agree with: “some games have a narrative design and others do not” (Avatars 192). This connection is a continuum, not a dichotomy. It’s difficult to tease a narrative out of, say, Tetris, and even more difficult to assert that it was designed with one in mind. Other games, like Space Invaders, more readily open themselves to narrative interpretation (a spaceship fending off the encroach of hostile aliens). Still others are undeniably narrative- focused, with the gameplay and mechanical aspects secondary to the story they tell, a la As Dusk Falls. The simple fact that game writing awards exist should be an indication that games and narrative are not mutually exclusive. Absolutist statements do nothing to help describe or define video games as they truly are and only serve as an exercise in pedantry and to obscure both the genre itself and the narrative concepts it complicates.

Once one embraces the compatibility of games and narrative, the question arises: how do game narratives work? One of the most prominent and perplexing challenges narrative designers face is what I call the Paradox of Choice. Simply put, it is the conundrum created by the fact that more player freedom means less structure in which a story can exist. Ryan describes it thus: “if the player’s choices are too broad, there will be no guarantee of narrative coherence; if the choices are too narrow, the game will be boring” (Avatars 196).

The paradox forces the designers of interactive stories to be creative within this limitation. Open-world roleplaying games like Fallout, Skyrim, and others must consider the fact that players not only can go anywhere at any time but can also interact with the world in ways that could be potentially story-wrecking. For instance, if a player can kill any non-player character (NPC) in the world, they might well murder an undercover villain long before the final confrontation was designed to occur, thereby making that narrative thread moot. The stories of open-world games must therefore be highly adaptable, flexible, and reactive to whatever decisions the player makes.

The first broad solution to the problem of story-destroying freedom is restriction of player agency. For instance, designers might make certain characters unkillable, ensuring they’ll be present when a later event plays out. While this is perhaps the most straightforward way to ensure that important events can occur as planned, the obvious downside is that it shatters the fiction that the player inhabits a living, truly interactable world. In turn, this reduces the believability of the game and creates a frustrating distance between the player and the game. It gives the player a sense of being railroaded, or forced along a predetermined path regardless of what choices they might otherwise have made.

The same immersion problem is arguably worse in the case of game cutscenes. In these short, cinematic interjections, the player has no control over their character or the direction the story takes. Because writers typically insert cutscenes during narratively significant moments to ensure the story unfolds exactly as planned, the player is actively excluded from participating in the story. This is not to say cutscenes and interactive narratives are mutually exclusive—after all, player choices made earlier in the game might lead to different cutscenes at a given story beat— but the cutscenes themselves are, by definition, non-interactive. At best, they engage the player as a movie scene would and feel compelling and relevant to the game’s narrative. But at worst, they’re frustrating, restrictive distractions.

The other common path designers take to balance story with agency is focusing on player exploration and discovery over having a distinct plot. Rather than one specific narrative, storytellers might focus on creating a rich and compelling universe whose lore serves as its largest story component. Rather than participating in an explicit, unfolding story, the player is invited to uncover mysteries, secrets, and backstory at their leisure while they explore an extensively interactive world. By talking to NPCs and investigating significant locations, players can construct and uncover the world’s story-rich history without restriction of gameplay or agency. Additionally, engaging with these story elements is often entirely optional and ignoring them will not be detrimental to any player’s experience of the game. The main features are the mechanics and challenges of gameplay: combat, strategy, hand-eye coordination, and so forth.

At its best, the Paradox of Choice forces game designers to be creative and design solutions in which, as Ryan puts it, “gameplay and narrative remediate each other’s deficiency” (Avatars 198). Poetically, she describes video games as “an art of compromise” (Avatars 198) between these two forces.

Ultimately, all games will be limited in some sense by storytellers’ production capacity. Humans cannot design an infinite world with unlimited interactivity, whether said interactivity is in narrative form or embedded in the mechanics of gameplay. Some elements will have to be preprogrammed, hard-coded, and therefore not alterable by the player. And it’s likely impossible to have a game with both full narrative cohesion and complete player agency. This raises an interesting question, however: would the ideal game be one that’s infinitely interactive, able to sculpt a compelling narrative from whatever choice the player makes without limiting the player’s agency as a tradeoff? Or is that a poor ideal even in theory?

The “Holodeck” theory of game design bases itself on the Holodeck portrayed by Star Trek, in which a player can physically enter a simulated, three-dimensional, fictional world whose hyper-advanced AI will react coherently to anything they say or do, enabling it to sculpt a story out of any action. The Holodeck model was proposed by Janet Murray, one of the first theorists to explore technology’s influence on narratives in her controversial 1997 book Hamlet on the Holodeck: The Future of Narrative in Cyberspace. Murray set forth the idea that the Holodeck should be the ultimate aspiration of game design.

A perfect Holodeck experience would encompass the player completely, adapt to any choice and action the player makes, and ultimately put together a compelling narrative from those choices to create a satisfying game and story experience. While great bounds in AI advancement have taken place over the last year—and these are undoubtedly already being incorporated into the next generation of games—humanity is still far from developing something as immersive and comprehensively reactive as the Star Trek Holodeck. Should that be the goal?

For this project, I spoke to three narrative designers. Georgia Symons and Jason Bakker worked on Wayward Strand, while Brad Kane was the lead writer for As Dusk Falls. I posed the Holodeck question to all, and their answer was a resounding no—a sentiment with which I agree. Bakker pointed out that the design of Wayward Strand is pretty much directly antithetical to the Holodeck ideal in how it plays with and limits player agency. In the game, the player controls a thirteen-year-old girl, Casey Beaumaris, whose mother is the head nurse aboard the titular airship hospital. Casey has been conscripted to help in the short-staffed long-term care ward following the abrupt departure of a nurse, and her goal is simple: spend time with the other characters. The stories of the NPCs—most of whom are the elderly residents of the hospital—unfold whether Casey is present to witness them or not, and it is up to the player to decide which characters to talk to and what story paths to pursue (investigate the history of the airship? Learn about the recently deceased Hugo Falk? Solve a mystery about a goat?).

In this way, Wayward Strand is a game about radical reduction of agency, and this is arguably its central theme. Young Casey and the elderly residents have little control over the events unfolding around them, in contrast to most games, which position the player’s character as the centerpiece of the world and whose actions determine the fate of everyone and everything around them. In a very telling interaction that’s even featured in the trailer for Wayward Strand, Casey expresses her confusion to one of the hospital residents, Ida. “I hear people talk about things, but I’m not sure what they’re on about,” she says. To this, Ida replies, “That sounds a bit like the life of an older person. Nobody explains to us what’s going on. We just have to piece it together as best we can.” In effect, this can be understood as the thesis of the game. Because the player is so limited, Wayward Strand opens space for the NPCs to have more dynamic, engaging, and interpersonal storylines.

A Holodeck version of this same premise would completely miss the point of the game. By definition, a player character with the ability to do anything would disrupt the limitation that is so central to what makes Wayward Strand compelling to begin with. Bakker takes this a step further and argues that a similar necessary constraint applies to all stories, not just those focused on limiting agency as Wayward Strand does. What makes stories emotional and engaging, he says, is fundamentally different from the experience a Holodeck would provide.

Bakker likened the Holodeck to the experience of playing tabletop roleplaying games (TTRPGs) such as Dungeons & Dragons, which are arguably closer to the Holodeck than video games because nothing is pre-scripted and the game master (GM) must come up with reactions to players’ choices on the fly. In the Holodeck, the player would essentially be an inexperienced RPG player in a world that tries desperately to make them a hero. Even if the GM warps and reshuffles the story world around the player to make any action they take plot-relevant, the resulting narrative would still be unsatisfying. The player might act arbitrarily and inconsistently, or bumble about without motivation or direction—all poor traits in a main character. And even in TTRPGs, the story is still a collaboration between players and the GM, rather than the GM having ultimate power. Players cede power, and thereby part of their agency, to the GM. This agreement creates the understanding that the players will follow the GM’s story hooks to the plot points they’ve designed. Games also involve players working within the GM’s (limited) world to make the story more interesting, fulfilling, and exciting. All of this shows that even in a Holodeck-like scenario, audiences still crave structure.

There’s also the problem of players being unable to do the sorts of things required of heroes in the sorts of stories they wish to inhabit. Because the Holodeck’s infinite interactivity puts no limits on what the player can or can’t do, it must physically encompass the player to truly do its job. This is problematic for obvious reasons. Do you want to play a martial arts master in a high-octane heist? Well, you’d better learn martial arts, then. A soldier in a medieval battlefield? Time to learn sword fighting. A clever diplomat in a political drama? Better read up on rhetoric and public speaking. A hostage in a tense motel standoff (as in As Dusk Falls)? Try not to have a panic attack. While the narrative output of the theoretical Holodeck might be infinitely adaptable, the player is not, and a Holodeck scenario might ironically lead to the player feeling frustrated and disempowered because they personally are unable to do the sorts of things that they’d like to see from a main character in the story they’re hoping to experience. Essentially, they are reduced to playing themself, not a hero.

Marie-Laure Ryan takes a more conservative approach to the Holodeck theory in her essay “From Narrative Games to Playable Stories: Towards a Poetics of Interactive Narrative.” Though she still views the Holodeck as a theoretical design ideal worth considering, she writes, “The Holodeck, as a whole, may be a castle in the air, but this does not take anything away from the validity of its individual features as goals to pursue for researchers and developers of interactive narrative” (“From Narrative Games,” 47). From the Holodeck, she argues, we can take away four main features that game designers can more realistically implement: natural interface (47), which deals with the problem of players’ physical disempowerment discussed above; integration of user action within the story (50), which must address the Paradox of Choice; dynamic creation of the story (51), in which the Holodeck must compute the effects of player actions in real time and renew itself with each playthrough, as with TTRPGs; and the pleasures of interactive narrative (53), which she further divides into several different kinds of immersion that video games can foster.

Ryan’s interpretation of the Holodeck aligns much better with how game designers approach the major challenges of interactivity and the much more realistic goals many game companies have for the genre, especially in big-budget games and the emerging field of virtual/augmented reality, which provides a different form of interface than was even possible when Ryan published this essay in 2009. However, it is worth remembering that not all games strive for greater player agency, as in the case of Wayward Strand. While extension of player choice is fascinating and worthwhile to pursue, it can be just as interesting (or perhaps even more so) to break from this goal as it is to further it. Creativity thrives within limitation.

The infinite branching potential of the Holodeck also obviously complicates the concept of authorship, given that a hyper-advanced AI would be generating the story concurrently with any actions or choices the player makes (rather than any human writer). How this would change our definition of “author” might seem like a far-off theoretical question, but more subtly, and in my opinion, more intriguingly, the definition of authorship is already blurred even in entirely human-made games.

i. Common Definitions of Authorship

The question of “who’s the author?” is a deceptively simple-seeming one. Video games add an interesting wrinkle to this already tricky matter, which been posed to and grappled with by many other forms of media in the past. First, it’s helpful to identify general “common sense” views of authorial credit. The platonic ideal of an author is someone who conceives of, writes, edits, and publishes a novel without input from any other people throughout the process. A step away from this model, wherein the process includes other parties like editors, publishers, and beta readers, generally doesn’t sway the status of the original writer as the definitive author of the work. Sometimes works are written by multiple authors, in which case they all get the recognition, despite lack of clarity about who wrote which portion of the manuscript (or the use of a collective pseudonym). Anthologies and collections typically credit an editor and each individual contributor for the stories within it.

By contrast, stories written by credited or uncredited ghostwriters (as in many celebrity memoirs) are more difficult to pin down; still, authorship remains generally attributed to the name on the book cover. An even more thorny—and ethically questionable—complication is the possibility that stories might soon have AI contributions, in which case authorship might be attributed to the AI, to its programmers, to the company selling it, or to the millions of (uncompensated, unidentified, and nonconsenting) authors who wrote the works it was trained on. We simply don’t know yet.

Further complicating common notions of authorship are cinema and theater, simply because their production involves the efforts of so many people. Playwrights typically minimize stage direction to allow actors the freedom to make dramatic choices with their performances, and reading the script of a play is a far cry from seeing it performed. This suggests that an audience has not experienced the work itself by only consuming what the original writer put to paper—a play requires the participation of others to be complete. Surely the divergent ways different actors might interpret and perform the same characters, thereby altering an audience’s experience of the play, implies some kind of individual authorship through dramatic choice. Still, the authorship of a play is credited to the original writer, and individual performances are instances and derivatives of the original work. When one looks up the author of A Marriage Proposal, the search results show Anton Chekov.

Both movies and plays also share the fact that massive crews are required to bring them to fruition—costume designers, managers, lighting experts, prop builders, and so forth. The scrolling credits at the end of a film often involve hundreds or thousands of names. Are these people, too, authors? Or does the sum of their work constitute an immaterial “collective” author of some kind? Interestingly, person who tends to receive the most credit for a movie’s creation is not the original scriptwriter(s), but the director (though sometimes they overlap). Video games are also group efforts, especially those produced by triple-A studios, where there may once again be hundreds of people in the credits. But when one considers interactivity, an additional complication appears.

Adapting the platonic ideal of authorship to an interactive context adds a whole new dimension to this debate. Even an interactive story that is written, coded, and published by a single person would not have as straightforward a claim to authorship as the writer of a linear narrative, for one main reason: the writer could never experience the entirety of their own story, even if they wrote every individual part of it.

Branching stories are a case in point. These interactive narratives are made of nodes that split off in different directions depending on the choices a player makes. These branches rapidly exponentiate. If a single node branches into two, and those two nodes branch into two more each, there are now four possible paths through the tree. The pattern will continue increasing at a rate of 2n nodes, and a full narrative map might have dozens of these splits. However, even this is oversimplified: branches will link back to each other, nodes will branch to more than two paths at a time, or the game will make use of variables and other tricks to determine what nodes the player is able to access based on their previous choices. This means that branching narratives are a confusing, interwoven tangle which, if each possible individual path through them was considered, might easily contain hundreds or thousands of unique stories. A single creator could not play through all of them. (If you think you’re up to the challenge, I invite you to try to chart every single route through the jumbled labyrinth of a branching narrative I wrote as part of this project. More on that later.)

Each person who interacts with a branching narrative will almost certainly have a different experience. Arguably, even taking on different strategies in combat-heavy games and dispatching of enemies in different ways might qualify as having a different story experience— after all, combat might well be part of the narrative. The creator of a branching narrative has assembled a collection of story parts (often in the form of nodes) but has not put them together— that’s up to the reader. The creator isn’t the determinant of the story the reader experiences.

ii. Death of the Narrative Designer?

One school of thought relevant to this question comes from Roland Barthes in his famous essay “The Death of the Author.” Barthes critiques the “capitalist ideology, which has afforded the greatest importance to the author’s ‘person’” (1), a figure who has come to overshadow the work they wrote, the readers who consume it, and even language itself. “The explanation of the work,” Barthes argues, “is always sought in the man who has produced it, as if, through the more or less transparent allegory of fiction, it was always finally the voice of one and the same person, the author, which delivered his ‘confidence’” (2). Barthes writes approvingly of the role surrealism has played in helping “secularize the image of the Author” (3). Secularization allows for broader analysis of a work and acknowledges the role of readers in its interpretation.

Without the supreme figure of the Author, a text transforms in ways that bear an uncanny resemblance to video games. For instance, Barthes claims that “the modern writer (scriptor) is born simultaneously with his text; he is in no way supplied with a being which precedes or transcends his writing [...]; there is no other time than that of the utterance, and every text is eternally written here and now” (3). In other words, the author is displaced by the text itself, which is conceived of as a presence that’s created in the process of being read and understood— rather like the creation of a video game story that actively unfolds as it’s played rather than arriving as a definitive pre-written, predetermined package that a passive reader simply consumes. Barthes argues that “to give an Author to a text is to impose upon that text a stop clause, to furnish it with a final signification, to close the writing” (5). One cannot “explain” the text by way of its creator—in fact, to Barthes, writing should even throw off and “evaporate” meaning altogether.

Barthes provides perhaps the most direct parallel to video games in his conclusion. He argues that the “multiplicity” of a text—the culture, history, etc. that a writer inevitably pulls from when they write—is united not in the author, but in the reader: “the reader is the very space in which are inscribed, without any being lost, all the citations a writing consists of; the unity of a text is not in its origin, it is in its destination” (6). The presence of a reader is when a text truly comes into existence. Certainly, it would be disingenuous to call the totality of a branching story map the true “story.” Rather, each path the reader/player takes through it constitutes the real story—hence, it exists only through the interaction of a reader, as Barthes suggests.

Barthes passed away in 1980, right around the time that the first video games were turning the corner into pop culture. Were he to see how far they’ve come, he’d likely applaud the how they’ve complicated the relationship between the creator and the reader/player. Yet the question now remains: what is that relationship, and how are we to define it?

iii. Authorship in Video Games

Unlike Barthes, I believe the “author” is still a worthwhile title and point of reflection if one is to fully understand a piece of interactive fiction. However, the role of any author on the creation end of the game production pipeline is necessarily reduced, not just because of large studios having teams of writers, but because of the nature of interactive narrative itself. Even in the “platonic ideal” of game authorship, the single writer behind the game cannot be considered the end-all be-all author, either. Because interactive narratives are choice- and engagement- based, any reader/player will largely be the master of their own plot, involved in the act of story creation. And what is an author if not a creator?

The realm of meaning, too, is almost entirely that of the reader/player. While the creator writes the potential plot points a reader/player might encounter, the way each piece of the narrative falls into place is determined by the path an audience member takes through them. A single node or passage of a story in isolation cannot create meaning; meaning is the sum of all parts. The question of what a story is “about” is essentially generated by each reader/player as they partake in one of many possible stories. One playthrough of a game might yield a story about bending beneath the crushing weight of social norms, while another playthrough of the same game might become about defying and casting off those same norms.

This isn’t to say that the reader/player entirely supplants the original creator. Their name still belongs at the top of the credits. But the title of authorship in interactive narratives must also be united in at least small part with each player who engages with the story. Every game contains a multitude of stories that any member can experience through their choices, leaving interactive stories in a strange gray area of authorship that simply does not exist in other forms of media.

This should prompt us as creators to reconsider the boundaries of what an author can be. Instead of an individual, the title of “author” can consist of a relationship between parties who experience a piece of media. Perhaps expanding this definition can also help theorists tackle other gray-area situations discussed above in the cases of theater, television, and cinema as well.

The inability of interactive stories to name a single author also upends traditional notions of literature, which are usually focused on “great” historical works by “great” historical authors (Jane Austen, Franz Kafka, George Orwell, William Shakespeare, etc.). While these notions are mainly focused on novels and plays, cinema has now been around long enough as well that movies have been inducted into the “classics” category, such as Casablanca or The Godfather.

Video games, however, are new enough to the playing field that there’s little consensus on what constitutes a classic, and even less consensus on what it might mean for a game to be “literary.” While certain games have come to be considered iconic, this is usually tied to their place in the history of the genre or their recognizability rather than any “literary” quality they might possess. Many of these have spawned vast franchises, like Super Mario Brothers, Final Fantasy, or The Legend of Zelda. Others are widely known for being the first popular games of their kind in the early days, like Pac-Man or Tetris. But overall, games are not iconic because they embody qualities associated with “literature” in other mediums; instead, iconic status has more to do with memorability and prominence in pop culture. This, plus the fact that story games are evolving and improving, because they’re so technology-centric, and because they thrive along the cutting edge of tech, makes it difficult to pin down a historical work to pedestalize or deem foundational in the sense of a “literary classic.”

To me, this is a good thing. I am wary of the impulse to canonize any game as “literary,” a category I take issue with even in its present context. At its worst, the term has too often been used to differentiate “true” art from the “mass market,” lending itself to a kind of elitist academic snobbery that has done little except exclude most modern stories from an equivalent esteem and encourage disproportionate focus on outdated works over the thriving media landscape we have today. The existence of decades of analysis written about this or that “classic” only serves to foster an unproductive impression that these are the only stories worth analyzing or admiring, an exclusive class that modern stories cannot broach simply by nature of not having the existing groundwork of excessive scholarship about them.

A similarly troubling trend in this designation is the tendency to pit “literary” fiction against “genre” fiction, a false dichotomy that excludes fantasy, science fiction, romance, and other popular genres from literary status. Ignoring the fact that literary fiction is a genre, and that it sometimes cherry-picks works (like The Lord of the Rings) from these same categories, often this false distinction is used to look down on speculative fiction for being only “entertainment value” (and thereby implying that any stories written to entertain their audiences are also inferior). Literary gatekeeping essentially refuses “genre fiction” the academic esteem, attention, and respect given to “literature.” There’s also often the implication that the first or most famous work of a genre is also the best in that genre (such as The Lord of the Rings, Dune, Foundation, and others credited with a boom of a certain genre), when in fact subsequent works have done a much better job developing the ideas introduced then.

The more that is written about video games in academic circles, the closer we veer to instilling this problem in the game industry, as well. The “first is best” impulse has largely been avoided, given that the earliest games were extremely rough and, while important for kickstarting game development, have since been surpassed by obviously stronger works, often by the same companies who produced those early iterations. However, I sense an inclination for critics to turn away from the mainstream simply because it is mainstream and therefore overlook the contributions to game storytelling that big studio games might be able to offer. I worry that games that are confusing or poorly made might be upheld as literary classics simply because they don’t resemble mass market games like Overwatch or Fortnite.

One example of this tendency is the analysis of the titular game in Marco Carraciolo’s essay “Playing ‘Home’: Videogame Experiences Between Ludic and Narrative Interests.” Carraciolo analyzes how a game like Home constantly reminds the player of their choices and complicates temporality through a past-tense narrative voice despite the game unfolding in the present. These are interesting notions to contemplate. Yet the game Carraciolo upholds as exemplary of these techniques simply sounds like a bad game. He describes how “During most of the game, the player finds a number of leads that almost seem to coalesce into a coherent answer to all these questions, but the picture is never complete—there is always something that resists the player's attempts at making sense of the plot” (241). Similarly, “no matter what the player does or how carefully he or she explores the game world the plot still remains blurry, and the game keeps reminding the audience that they might have ‘missed something’ along the way” (Carraciolo 242). He describes the game’s ending: “no matter what the player chooses the story still won’t make any recognizable or satisfactory sense” (244); in fact, he admits that “the suspense and curiosity elicited by the game’s narrative cues are frustrated by a system with unexpectedly lets down the audience, asking them to choose their own story and leaving up in the air the previously established plot instabilities” (245).

The impression of Home one comes away with after reading this essay is not that it’s an introspective and clever look at the mechanics of game narratives, but simply a bad game that trips on the curtain and reveals those mechanics incidentally: all empty intrigue and no payoff. Its inability to tell a cohesive, compelling narrative is not an indicator that it is secretly good. While the author finds that it’s an insightful, important intuition, it’s taken an entire essay to justify this position while providing description that, to me, only serves to explain that Home is not a good game or a good story worth emulating or praising.

Carraciolo writes, in conclusion, that “there is no reason to drive a wedge between the canonical narratives of film and literary fiction and videogames: both can, if sufficiently thought- provoking, encourage recipients to connect their experiences to broader thematic concerns through interpretation, which is—of course—our main strategy for coping with the unusual, the incomprehensible, and the stubborn in artistic practices” (247). In other words, Carraciolo is suggesting that video games—namely, Home, as he has tried to prove in this essay—ought to be inducted into the pre-existing literary canon, which already includes the film and fiction genres.

It is here that I see the most worrisome impulse emerge. Inaugurating games into the present literary canon—and setting the precedent for what types of games can be canonized with the weak option of Home—would only further the problematic inclinations of the designation that I discussed earlier. Instead of trying to fit games into this prevailing paradigm, I would like to encourage game-interested academics to use this new genre to rethink what classifications like “literature” and “classic” mean to begin with and not take for granted that ambiguity, confusion, and experimentalism are necessarily indicators of depth. Games, perhaps more so than other media, are unabashedly entertainment, and they veer readily into speculative fiction with magical worlds to explore and great monsters to tame. None of this means they have nothing to say. I urge academics to engage with these games on their own terms.

Even more problematic than the canon wars is the public perception of video games, which currently considers them to have little artistic merit. Instead, they’re often characterized as brain-rotting, unproductive pastimes that encourage violence and sloth, meaning that games are often overlooked as the powerful form of narrative media that they are. Just as there’s little consensus about what it means for a game to be “literary,” there’s even less on whether a game can or should be literary to begin with. Some academics assume that analyzing them in any serious context an exercise in futility, characterizing the video game genre as one that cares more about gore splatter effects and explosions than narrative relevance. Viewing games in such a way only reveals that these academics have played few of them.

I would love to see games elevated to the respect for storytelling potential that novels, plays, and movies currently hold. We take for granted that other mediums are capable of telling thought-provoking, nuanced narratives that feature deep characters and rich thematic content—a supposition not commonly extended to video games. But games are just as able to tackle real- world issues, engage the reader in emotional and nuanced storylines, and imagine flawed characters navigating difficult worlds as other media—and they can do this while still being entertaining and fantastical. Games thus provide us with a unique opportunity to rethink literary canonization itself. Promoting exclusionary notions of “literature,” which would likely only uplift a select few games that are unlikely representative of the whole, is a poor way to do so.

The question of games as literature feeds into a larger point about the relevance of scholarly discourse to the game industry itself. Because it came up in both interviews, it’s worth mentioning that game designers themselves don’t feel particularly represented by the claims academics put forth when analyzing the games these selfsame designers create, nor by the marketing of games themselves on the part of the companies that sell them. The definition of a video game isn’t as clear-cut as the market would lead one to believe, and a lot of the distinctions between games and other forms of interactive media are theoretical rather than practical in nature. Ontology—the question of what something is and what it’s like—is always complex, even under the best of circumstances.

Brad Kane, for instance, argues that difference between a video game and other interactive media is really an academic one. As an example, HBO allows you to pause a movie and learn about what went on behind the scenes, but this is hardly a game despite being broadly “interactive.” Similarly, anything nonlinear is commonly viewed as a video game, but interactive theater also exists. Calling something a video game also carries social stigma that might repulse audiences who might otherwise greatly enjoy the stories they tell. Games are excluded from certain storytelling awards (such as the Emmys) on the basis that there cannot be tests of skill involved—even though, in games like As Dusk Falls (ADF), failure is as valid a story outcome as success and will lead to just as interesting of a narrative.

ADF’s “tests of skill” are quick-time (QT) events that determine the outcomes of physical conflicts. When the character they control gets in a fight, the player is prompted to swipe, click, or mash the mouse in certain amounts of time to win the fight. If they’re too slow and fail the QT action, the story accounts for their loss and branches accordingly (unlike games like Mass Effect where successfully winning a fight is crucial to progressing the story). ADF was designed around these gamified elements as much as it was around the story, resulting in a marriage of the two. However, Kane does not really think of ADF as a video game; he sees its peers as primarily TV shows, not games. If it were up to him, ADF would be classified as interactive television—even though from a studio standpoint, it makes more sense to be sold as a video game, and the academic outlook would likely categorize it as a game as well.

Wayward Strand also straddles definitions, in no small part due to its creators’ expertise and the ethos with which they approached the project. Georgia Symons, one of the game’s writers, never intended to do narrative design. Before her involvement in the project, she had no video game experience but a strong theater background, with some work on board games and escape rooms on the side. She noted that there’s recently been a shift in theater from audience- agnostic plays that do not acknowledge the existence of onlookers to bringing the audience back into the performances themselves. Before the Renaissance, theatergoers were much more involved in the enactment of a play than audiences are today; actors would speak straight to the spectators or respond to their reactions onstage, creating a low-level kind of interactivity. The most obvious incorporation of the audience in modern plays are directly interactive experiences like Sleep No More, but the encouragement of audience involvement is a larger trend that Symons has been privy to. She took this knowledge to Wayward Strand, the producers of which intentionally sought out a game industry outsider to bring fresh perspective to the project.

For Symons, the swing from theater to game writing came with a few major adjustments. She had to learn to let go of giving scenes a strong “narrative spine,” instead allowing the dialogue to go in multiple directions, with the options presented to the reader creating a broad tree of choices rather than sticking close to the trunk. Because of her theater background, she put much thought and care into creating outstanding moment-to-moment dialogue as well as cultivating the role of silence in scenes. Her background, then, was instrumental in making Wayward Strand what it is, and perhaps it also sculpts parts of the game more towards something like a play than like a traditional video game. (Interestingly, though perhaps coincidentally, the visual design of Wayward Strand also calls a stage to mind, with conversations taking place in rooms with an invisible “fourth wall” through which the player looks.)

When asked to define branching narrative, both writers responded similarly to Kane, pointing out that categorization questions can be distractions, and one must ask what their utility is in the actual creation of a game. Certainly, they help with marketing and identifying an audience, but from a creator standpoint, classification is less relevant and useful. Symons also pointed out that branching narrative and interactive narrative are not the same; branching is a subset of interactive narrative, yes, but there are a multitude of patterns and structures that fall under the same umbrella. This makes answering categorization questions a much more nuanced and interesting endeavor than just assessing whether a work is a video game or not.

Symons’ observations about the vast array of structures in interactive fiction is especially relevant when analyzing other story-based video games, as I’ll explore below.

Video games were my focus for this project, but they’re hardly the only form of media to experiment with narrative structure, which is why it’s worth appreciating other stories that push narrative boundaries. Because of their ink-and-paper limitations, none of the three novels I read for this project were “interactive” in the sense that video games can be. However, they challenge the limitations of conventional narrative in interesting ways that go beyond interactivity—most notably, by playing with temporality, perspective, and the reader’s relationship to the text.

Italo Calvino’s experimental novel If on a winter’s night a traveler is a mischievous, labyrinthine story. The plot, broadly speaking, follows a second person “Reader” who starts and then is foiled, in increasingly contrived ways, in finishing ten different stories. The titles of these stories ultimately add up to a single sentence/poem when combined at the novel’s end. Along the way, the male-gendered Reader meets Ludmilla, his love interest, and is swept up in a rather tongue-in-cheek conspiracy involving a devious translator, a struggling writer, a censorious autocracy, several fictional nations, and a shady publishing cabal.

Despite its nature as a printed text, If on a winter’s night nonetheless seems to invite the reader to interact with it in ways uncommon for novels. After all, the second person “you” seems to implicate the reader’s involvement with the way the story unfolds, even though “you” ends up being the Reader, a character who is separate from the actual reader. By setting this frame in the second person, Calvino implicitly claims that you would act the same way, and that the Reader’s actions are by extension your own.

The story also plays with giving and taking agency; the Reader is active in that he pursues the novels that he is constantly interrupted in reading, and yet the circumstances of their disappearance are far beyond his control, leading to an interplay of intrigue and frustration. Of course, the literal reader of the novel has no agency, but they still feel the same thrill of the chase and annoyance at its foiling that the Reader does—the illusion of reader agency created by second person and identification of the Reader’s actions with any reader’s. So, although the reader cannot literally affect the story, If on a winter’s night still encourages a different relationship between the story, the reader, and the main character than most novels attempt.

Richard Powers’ The Overstory is more concerned with time than perspective, which is important given the novel’s central focus on trees and the great time scales in which they dwell. The modern world generally conceives of time as a linear motion, closely linked to human sensibilities and measured in days, months, and years that unfold in a straight path from past to future. Powers instead pushes his readers to reconceive time as a cycle—seasonal, regular, circular—that rotates on a scale much larger than a human life.

Unlike a traditional narrative, where the plot’s resolution is left unknown until the reader gets there, Powers frequently reveals long in advance what will become of the characters. The narrator is not bound by linear time; readers learn, for instance, that Ray will suffer a stroke dozens of chapters before it happens. Similarly, the most spiritual and nature-connected character, Olivia, has premonitions as well that seem to upset linearity; she communicates with beings of light that elevate her beyond immediate, trivial, human concerns and allow her to look at the greater picture that sprawls far beyond the lives of the humans inhabiting it. Despite being written in a medium that cannot branch or loop, The Overstory challenges its own linearity and stretches the boundaries of novel time. Everything comes around, so to speak—and these great natural cycles will outlast any temporary, linear human timescale.

Briar Rose by Robert Coover, a dark retelling of Sleeping Beauty, was probably the closest these novels came to having a true interactive component. Its sections are designed to be read in any random order, like a shuffled deck of playing cards. While one could technically do this with any novel, few are written to facilitate randomness and create a cohesive experience out of whatever order the reader chooses. The story of Briar Rose is less a story and more an atmosphere that the novel creates through presenting many versions of similar scenes featuring three major characters—Briar Rose, the wicked fairy who cursed her, and the prince attempting to rescue her—that coalesce into timeless loops of dreaming and searching that refuse to resolve. Never does the prince reach Briar, never does Briar awake, and the experience is an endlessly grim, cyclical, and surreal one. Even if a reader has read every section in the book, the components can still be endlessly reshuffled. Readers can make of this what they will—and whether the book is successful in what it attempts through these techniques—but Briar Rose avoids easily settling into any semblance of plot. It certainly underscores Brad Kane’s point that it’s incorrect to categorize any nonlinear narrative as a game, however tempting that might be.

All of this goes to show that one does not necessarily need a new storytelling medium to challenge existing conceptions of narrative. And while the nature of video games leaves them especially primed to try new things with their stories, it’s worth appreciating the genre-benders that have pushed limits in other mediums as well.

The category of “nonlinear narrative” is a broad label. When people use this term, they tend to mean one of two things. The first definition refers to stories that are internally nonlinear, which are told chronologically out of order with respect to the fictional timeline (e.g., the film Memento). The second definition, and the one this project is most interested in, is interactive stories, which are nonlinear in that there are multiple ways to experience the events of the story based on the audience’s interactions with it. Interactive stories can further be broken down into a few general kinds of interactivity:

INVESTIGATIVE. The audience is encouraged to uncover story elements by interacting with a staged environment; the narrative itself usually has either already happened, exists primarily as lore, or unfolds without the audience’s direct participation. The interactive element is solely or primarily environmental exploration; character interaction is minimal, unimportant, or peripheral. This is most commonly found in open-world RPGs (The Elder Scrolls, Skyrim), location-based hypertext stories, or interactive art instillations with a narrative component (Omega Mart, Otherworld). Interactive theatrical experience Sleep No More also mostly falls into this category.

IMPROVISATIONAL. The story is neither scripted nor staged; instead, it’s almost entirely spontaneous and results from direct interaction between the creator and audience. Prime examples of improvisational interactivity are tabletop roleplaying games, wherein an adlibbed story emerges from the collaboration between the Game Master and players, and improv theater, which is loosely structured around the actors’ playing games for which the audience provides prompts.

CONSTRUCTED. The audience is given a set of tools to create their own world in which they can orchestrate and imagine stories for themselves without one in specific necessarily being planned or intended at all by the developers. Story elements are low-level and not central to the interactive portion of the experience. If a player doesn’t build any story for themself at all and simply enjoys the gameplay, nothing is lost. Examples include card games such as Magic: The Gathering and video games like The Sims, DragonVale, and Minecraft.

BRANCHING. The story is pre-scripted to branch in any number of directions, enabling the audience to decide which of several options and paths they want to pursue. Aside from the player picking whichever of the available choices they please, no elements are spontaneous. This is the category into which story games and choose- your-own-adventure stories fall.

Branching narratives like the ones I studied for this project can be even further split into a few types of specific structures, which will be discussed later.

Choices generally affect the story on two different levels: character and event. Choices made during conversations, for example, tend to develop relationships with NPCs and shape the ways the NPCs view the player character (PC). These choices tend to have few immediate consequences and only affect the plot overtly in the long term. Whether you are rude, sympathetic, or neutral to a given character affects how they think about you and how they’ll interact with you in the future. If you acted nasty to an ally, they might refuse to work with you in the future, leading to plot challenges you wouldn’t have encountered with them on your side. In relationship-heavy games like Oxenfree, the player’s attitude towards PC Alex’s friends can easily alienate and turn them against her. The same is doubly true for games that are entirely based in character interaction. In Wayward Strand, the way in which you interact with characters (and who you choose to spend the most time with) affects what information they’re willing to share with you and how comfortable they are around you.

On the other hand, event-centric choices are concerned with the course of action you take in a scene, and their effects are generally seen right away. The very first choice The Stanley Parable presents to its player is whether to go through the left door, obeying the narrator, or the right door, defying him. In fact, The Stanley Parable directly confronts the Paradox of Choice in this moment by pointing out the contradiction posed by the illusion of decision in this moment (given that whichever way the player goes, they are following a path designed by the developers and are thus still within the structure of the game). The effects are seen immediately, most obviously in that the player is now witnessing a different part of the building than if they’d picked the other door.

In a less overly self-referential way, other games pull the same trick. As Dusk Falls is a great example. Decisions are split between huge branches, with significant impacts down the line, and dialogue trees without lasting story impact. For example, choosing to help Dale ransack the sheriff’s house in addition to completing the main plot point of stealing his money has few lasting impacts, while arguably the largest decision of the first “book”—whether to select Vince’s injured wife or six-year-old daughter for a hostage swap—has massive ramifications and can lead to character deaths if the player takes certain paths. The game does a good job balancing these two types of choice; one fleshes out character relationships, the other moves the plot forward in drastic ways. The outcomes are both immediate and delayed.

Of course, character and plot choices are not mutually exclusive. What a player decides to do in a particular scenario can affect both how that situation plays out and NPCs’ opinions of the player. But generally, choices are weighted more towards one or the other.

While game dialogue and relationship-focused choices tend to be organized in trees, the classic tree structure is only one of many ways to organize an interactive story’s overall plot, and few story games (especially longform story games) stick with a “true” tree format because of its exponential branching. Instead, games tend to use a combination of different structures depending on what level the game is currently operating on: moment, scene, or story.

Moment-to-moment interaction constitutes the smallest scale of a story, followed by scene-level conflict, followed by the overall plot itself. The structure of a single conversation might not necessarily mirror the structure used by the entire game itself. Dialogue might be a tree, but the plot overall might be linear.

On the broadest level, I’ve identified five common structures:

TREE. This is likely what most people think when they hear “branching narrative”: players traverse nodes that split into other nodes, which split into further nodes, and so forth based on their decisions. As mentioned above, this seems to be most used for in-game conversations.

The closest game to follow a true tree that I played was As Dusk Falls, which was impressive in its quantity and diversity of branches across both “books” in its story. To avoid exponential growth, however, it melded tree branching with the diamond structure discussed below; the game never allowed itself to branch more than three nodes deep at a time (except on crossroads choices) before returning to its main throughline.

DIAMOND. Players take branching paths through the same set of events that start and end at the same points (whether that be points in time or plot points) regardless of what choices the player made during the body of the diamond. The biggest benefit of diamond structure is that it provides a built-in defense against choice bloat, as it requires choices to resolve themselves as they reach the cusp of the next event.

Choice of the Deathless employed this structure most obviously; the player always faces the same challenges in the same order. It progresses through each of the cases they tackle in the same order: the exploitative contract, the conference, the goddess deposition, the demon negotiation, and the climax. However, the content of those challenges can change based on their stats and decisions. There is usually more than one simple outcome to each event even if they end at the same major plot point.

FOREST. Players follow a mostly linear narrative populated with encounters that branch into dialogue trees on the scene level. These encounters and their outcomes don’t affect the main narrative in major ways. However, it’s likely that the game is at its most interesting when the player pursues these side paths.

Wayward Strand is the most prominent example of this type of game. The story takes place over the same three days no matter what the player does, and the major events that unfold around Casey—Esther’s sepsis scare, Agnete’s visit, Mr. Pettigrew’s panic over the VIP—happen no matter what the player has her do. The events of the world are larger than her, and the main story follows the same path each time. But she can decide to stop and speak to whomever she pleases, and that yields different interactions and personal relationships. In other words, it’s always the same forest and the same path, but you can choose what trees you want to climb and how. (Credit to Georgia Symons for coming up with this metaphor for her game.)

QUEST. This structure is usually found in investigational narratives, but it can be a subset of branching narrative as well. While exploring a largely plotless world, players chance upon encounters that prompt them to complete tasks with story components. For example, they might negotiate a treaty between towns, track down and kill a local threat, or retrieve an item for an NPC. For the duration of the quest, a loose narrative unfolds.

This structure works well in open-world games because it populates the world with interesting characters while keeping the story-intensive elements optional, which doesn’t hinder the player’s explorations. They can choose whether to take on the quest or not, after all—though quests also usually come with a gameplay incentive like in-game money or gear that one receives when the quest is completed.

Big studio games like Mass Effect often utilize this structure, most likely because it’s much easier for their many writers to work on lots of small, compelling quests rather than a single massive story. Mass Effect’s quests and main story alike also use the success state structure described below.

SUCCESS STATE. To advance the plot (which itself might take on any of these other structures), the player needs successfully to complete a gameplay challenge. If they fail to complete the challenge, they must restart it and continue attempting until they pass. Note that this is different than a structure in which failing the challenge is also a valid story outcome; in success state structures, the player must win to continue. This is commonly used in combat games where the stakes are life or death, because if the main character is killed during conflict, there’s no way for the story to continue, so the player is sent back to the beginning of the fight to try again. When they win the conflict or successfully reach a target, the plot can proceed.

The downside of this structure is that, at its worst, it can feel like a movie that randomly pauses itself sometimes and forces you to win a game of basketball before you can continue. It doesn’t feel as genuinely interactive if the story must go on hold every so often in favor of a different kind of entertainment. Besides, it can be frustrating for people who are invested in the story but bad at the gameplay elements to have to struggle before they can go back to the aspect of the game they enjoyed.

Interestingly, The Price of Flesh also follows this route. Players can explore their surroundings while Celia is away, but if she returns while they’re outside their room, she’ll kill them. Hence, the success state is not simply surviving between her appearances, but rather, effectively balancing how long it takes to look for an escape with their estimation of how long Celia will be absent.

Obviously, not all branching stories fall cleanly into these five categories, and games often mix and match depending on the stories they want to tell and even their production limitations. A game without a team of writers, lots of time, and a huge studio budget to back it up will likely stick to a structure like the diamond pattern which, while more restrictive than the tree-based structures, has built-in rails to keep the story from getting out of hand. Programming capability also places some limits on who can build what types of games. After all, if one wishes to build a success state narrative, one needs to know how to code and implement interesting challenges as well as make them feel relevant and not tangential to the story parts of their game. There’s a reason most complex games need a studio behind them to reach their full potential— which has made the final part of my project, designing a branching narrative of my own, both especially interesting and especially challenging.

(See below for structure diagrams.)